income inequality in education essay

Support nonprofit news

The Hechinger Report

Covering Innovation & Inequality in Education

A decade of research on the rich-poor divide in education

Jill Barshay SENIOR REPORTER

(212)... More by Jill Barshay

Letters to the Editor

At The Hechinger Report, we publish thoughtful letters from readers that contribute to the ongoing discussion about the education topics we cover. Please read our guidelines for more information. We will not consider letters that do not contain a full name and valid email address. You may submit news tips or ideas here without a full name, but not letters.

By submitting your name, you grant us permission to publish it with your letter. We will never publish your email address. You must fill out all fields to submit a letter.

Thanks to Jill Barshay for the excellent column reminding us that there is much more to the rich/poor divide in our public schools than just the availability of digital devices and wi-fi. The real problem with equity in education is the lack of equity in school funding, which is an issue both of inequity in society and the ways in which public schools are funded (i.e., primarily local tax revenues).

Other barriers that kept the “school door blocked” for many low income students during this season of remote learning — and, presumably, next school year, as well — include: 1. Some with access to devices and wi-fi have had service disconnected at times due to unpaid (unpayable) bills. 2. Many have no private space in their homes from where to participate in synchronous learning/Zoom calls 3. With loss of family income and no child care, some have work or baby-sitting responsibilities that interfere with participation 4. Deadening effects of online learning cause many low-income students to disconnect and/or “drop out”. 5. In ability to access teacher supports and specialized instruction, esp. for English language learners and children with special needs. 6. Parent inability to assist students with computer routines, glitches, log-ins, etc

As districts address equity in the coming school year, we must also address the modes of learning that we consider both effective and valuable. If the top priority is engaging all students we need high engagement models based in trauma-informed practices, social and racial justice curricula, service learning, interdisciplinary project- and place-based learning, outdoor learning and other innovative ways to make education relevant to all students, regardless of their zip codes. Relax the standards. Cancel high stakes testing. Trust teachers to use their creativity to connect with every student and family. Otherwise, “remote” or “hybrid” learning, regardless of the availability of technology, will only be widening the gaps that structural racism has already created.

Why are we NOT reaching out to the teaching programs started by Marva Collins in Chicago and Ron Clark in Atlanta? Why are we NOT looking at a book called Schools That Work and viewing the achievements and strategies followed by successful programs. Let’s follow successful schools, successful environments in urban, rural, and suburban locations. As an eductor who started teaching in the Ocean-Hill Brownsville area of Brooklyn, N,Y. in 1971, there was a wildcat strike happening and this area was the where decentralization took place in N.Y.C. Rev. Al Sharpton’s church was down the block from I.S. 271. It took 2 years before a no nonsense, BLACK principal, took control over the choas and the movement of 125 teachers going and then coming to this “high poverty” intermediate school. There was stability of staff and the message was, you’re here to learn. I taught there for 7 incredible years and grew to understand what it was like being a minority teacher and human being. I then moved to Columbia, MD. where I lived in a planned community where diversity of color, homes, religions and belief in humanity living together as ONE took place. I taught in a white disadvantaged area for 2 years and observed the same behaviors students exhibited except there was no leadership at the top of this school. Now I teach in a suburban area for the last 31 years with limited diversity and succeeds because of innovative leadership, extraordinary teachers, and pretty high achieving students. Yes, I know every students must have access to technology as a MUST. Yes, I know urban education, rural education, and suburban education do education diffferently. Yes, I know poverty sucks, and I know distant learning may be around for a while. Change must come from the top. Let’s follow the successful educators, the successful programs, the dynamic elected officials who can shake up things so our students, our kids, our educational systems can be the change that can bring poverty to it’s knees.

I live on Long Island and know that whatever is written here about us is true. The Freeport Public School waste millions of taxpayers dollars throwing out teaching equipment, devices books that could be just given to the less fortunate schools next door-Queens, Brooklyn and the Bronx; where we see children suffering because of lack of proper learning tools. I am from the Caribbean where l taught for years. Oh l wish we were as privileged as these children. Maybe one day the disparity will end. Hopefully.

I enjoy reading this post. I am currently doing my thesis and the research question is: Do California K-12 public schools in lower-income communities offer the same level of academic curriculum as those in middle-income and wealthy communities? Do you have the reference page for those studies or even any peer reviewers where you got the information? I would like to review those studies and use them for my thesis. Thank you

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

Sign me up for the newsletter!

Submit a letter

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Publications
Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

Advanced Search
Journal List
Proc Natl Acad Sci U S A
v.117(32); 2020 Aug 11

A century of educational inequality in the United States

Michelle jackson.

a Department of Sociology, Stanford University, Stanford, CA, 94305;

Brian Holzman

b Houston Education Research Consortium, Rice University, Houston, TX, 77005

Author contributions: M.J. and B.H. designed research; M.J. and B.H. analyzed data; and M.J. wrote the paper.

Associated Data

The analysis code and auxiliary data required to produce the figures and tables in this paper can be accessed at https://osf.io/jxne5 . Code to produce estimates for each of the individual datasets (see Table 1 ) is also provided. Details on how to access these datasets are provided in SI Appendix (most datasets are available for download upon registration with the data provider, while others are accessible only with a restricted use license from the National Center for Education Statistics).

Significance

There has been widespread concern that the takeoff in income inequality in recent decades has had harmful social consequences. We provide evidence on this concern by assembling all available nationally representative datasets on college enrollment and completion. This approach, which allows us to examine the relationship between income inequality and collegiate inequalities over the full century, reveals that the long-standing worry about income inequality is warranted. Inequalities in college enrollment and completion were low for cohorts born in the late 1950s and 1960s, when income inequality was low, and high for cohorts born in the late 1980s, when income inequality peaked. This grand U-turn means that contemporary birth cohorts are experiencing levels of collegiate inequality not seen for generations.

The “income inequality hypothesis” holds that rising income inequality affects the distribution of a wide range of social and economic outcomes. Although it is often alleged that rising income inequality will increase the advantages of the well-off in the competition for college, some researchers have provided descriptive evidence at odds with the income inequality hypothesis. In this paper, we track long-term trends in family income inequalities in college enrollment and completion (“collegiate inequalities”) using all available nationally representative datasets for cohorts born between 1908 and 1995. We show that the trends in collegiate inequalities moved in lockstep with the trend in income inequality over the past century. There is one exception to this general finding: For cohorts at risk for serving in the Vietnam War, collegiate inequalities were high, while income inequality was low. During this period, inequality in college enrollment and completion was significantly higher for men than for women, suggesting a bona fide “Vietnam War” effect. Aside from this singular confounding event, a century of evidence establishes a strong association between income and collegiate inequality, providing support for the view that rising income inequality is fundamentally changing the distribution of life chances.

It has long been suspected that the takeoff in income inequality has made the good luck of an advantaged birth ever more consequential for accessing opportunities and getting ahead. The “income inequality” hypothesis proposes that intergenerational inequality—with respect to educational attainment, social mobility, and other socioeconomic outcomes—will increase as income inequality grows. Because this hypothesis shot to public attention with Krueger’s ( 1 ) discussion of the Great Gatsby curve, the proposition that high levels of income inequality have generated correspondingly high levels of intergenerational reproduction is now a staple of public and political discourse. Despite the prominence of this argument, the evidence in its favor is less overwhelming than might be assumed ( 2 ), and is largely limited to the empirical result that intergenerational income inheritance has increased in recent decades, at least in some analyses ( 3 , 4 ). Even this result has been contested and is far from widely accepted ( 5 ).

In this paper, we assess the plausibility of the income inequality hypothesis by examining changes over the past century in the income-based gaps in college enrollment and completion. This is a field in which descriptive evidence is key: Designs that would allow for convincing causal inference are in short supply, and where designs are available, the data are not. And yet most of the descriptive evidence in regard to the college level pertains only to recent decades, when both income inequality and collegiate inequalities have increased (refs. 6 – 8 ).

The trends through earlier decades of the century, within which the great U-turn in income inequality occurred, remain largely undocumented. To overcome this evidence deficit, we might be inclined to draw on evidence on other educational outcomes, such as test scores and years of schooling. Reardon’s analysis of family income test score gaps, for example, shows steadily rising gaps between cohorts born in the 1940s and those born in the present day (ref. 9 ; cf. ref. 10 ). But test scores are quite imperfectly correlated with educational attainment, and evidence from studies of inequalities in years of schooling would support different conclusions on trend. Hilger’s ( 11 ) analysis of long-term trends using Census data shows that there was a decline in the effects of parental income on child’s education between the 1940s and 1970s, while Mare ( 12 ) shows an increasing effect of family income on higher-level educational transitions for midcentury cohorts as compared to early-century cohorts. Taking these studies together, it is difficult to reach any firm conclusion about the income inequality hypothesis, as one might infer an increase, a decrease, or stability in collegiate inequalities during the midcentury, depending on which study is considered.

Extending the time series over the whole of the past century allows for a fuller assessment of the income inequality hypothesis, as the long-run historical series on income inequality exhibits a relatively complicated pattern, as opposed to the simple increase in the recent period. In much the same way as the magnitude of changes in income inequality could only be appreciated when considered in the long run, current levels of educational inequality must be evaluated and understood in full historical context ( 13 ). In a comprehensive extension of previous research on collegiate inequalities, we thus use all nationally representative data sources that we were able to locate and access. This strengthens the descriptive evidence that can be brought to bear upon the income inequality hypothesis.

In the following sections, we discuss the available data and the methods of analysis, and present our results on long-term trends in collegiate inequalities. We will focus on inequalities in completion of 4-year college, enrollment in 4-year college, and enrollment in any college (2- or 4-year). We will demonstrate an essential similarity in inequality trends across the range of collegiate outcomes. Although we will show that income inequality is strongly associated with inequalities at the college level, we will also highlight that it is not the only force at work.

College Enrollment and Completion in the Twentieth Century

The twentieth century was the first century in which education systems were widely diffused and, at least in principle, accessible to all social groups. The century witnessed substantial expansion at the college level: The college enrollment rate for 20- to 21-y-olds increased from around 15 % for the mid-1920s birth cohorts to almost 60 % for cohorts born toward the end of the century. * As Fig. 1 shows, rates of enrollment rose rapidly for cohorts born in the early century to midcentury, and flattened out and even declined for the midcentury birth cohorts, before resuming a steady increase for cohorts born in the later decades of the century.

An external file that holds a picture, illustration, etc.
Object name is pnas.1907258117fig01.jpg

Proportion of birth cohort enrolled in college ages 20 y to 21 y ( 14 ), and proportions completing 2- and 4-year college degrees, Current Population Survey March, Annual Social and Economic Supplement ( 15 ).

We see in Fig. 1 a stark reversal of the gender gap in college enrollment; for birth cohorts from the mid-1950s to mid-1990s, the proportion of women enrolled in college grew by around 30 percentage points, while the corresponding increase for men was just under 20 percentage points ( 16 , 17 ). The reversal occurred immediately after the rapid increase in enrollment rates observed for male birth cohorts at risk for service in the Vietnam War ( 16 ). A literature in economics has demonstrated that men born in the 1940s and 1950s were unusually likely to attend and graduate from college, although there is disagreement with respect to whether the observed increase in men’s college participation rates should be attributed to draft avoidance or to postservice GI Bill enrollments (ref. 18 ; cf. ref. 19 ).

Alongside trends in college enrollment, Fig. 1 presents rates of college completion by type of degree. While rates of completion of 2-year college are rather flat for cohorts born from the 1950s onward, rates of 4-year college completion have increased considerably. As the figure suggests, rates of 4-year college completion are highly correlated with rates of enrollment, but research shows that, over the past half-century, rates of college completion increased less sharply than rates of enrollment, because the college dropout rate increased ( 6 , 20 ).

Materials and Method

Although it is relatively straightforward to examine changes in rates of college enrollment and completion over time, it is rather less straightforward to examine income inequalities in collegiate outcomes across the span of the twentieth century, because data on parental income, college enrollment, and college completion are not routinely collected in government surveys. We must therefore piece together the trends in collegiate inequalities through the analysis of available sources of nationally representative data. We include results from the analysis of both cross-sectional surveys of adults and longitudinal surveys beginning with school-aged children, and, for a number of recent cohorts, we calculate estimates from tax data results in the public domain. Although this approach presents obvious challenges as regards comparability of data sources and measures, for much of the period that we cover, we have multiple estimates of collegiate inequalities for any given period of time. The datasets and their key characteristics are listed in Table 1 ; detailed descriptions of each dataset are included in SI Appendix .

Characteristics of the datasets included in the analysis

Dataset	Birth cohorts	Data collection	N
OCG 1973	1908–1952	Cross-sectional survey	25,163
NLS Young Men	1949–1951	Longitudinal survey	1,132
NLS Young Women	1951–1953	Longitudinal survey	752
PSID	1954–1989	Longitudinal survey	7,978
NLS72	1954	School cohort survey	9,637
HS&B	1962–1964	School cohort survey	18,805
NLSY79	1962–1964	Longitudinal survey	2,259
NELS	1974	School cohort survey	10,337
Add Health	1977–1982	Longitudinal survey	3,850
Chetty et al. (5)	1981–1993	Tax data	. 13 million
NLSY97	1980–1984	Longitudinal survey	5,254
ELS	1986	School cohort survey	9,990
HSLS	1995	School cohort survey	13,612

Add Health, National Longitudinal Study of Adolescent to Adult Health; ELS, Education Longitudinal Study; HSLS, High School Longitudinal Study.

The datasets cover cohorts born between 1908 and 1995, and it is only at the beginning and the end of the data series that our birth cohorts are represented by no more than one dataset. Although we aim to define cohorts according to year of birth, for some of the datasets we must construct quasi-cohorts based on age or grade, because year of birth was not recorded.

The biggest constraint that we face in analyzing income inequalities in collegiate attainment relates to gender. Data on the earlier birth cohorts come from the Occupational Changes in a Generation (OCG 1973) survey, which was administered in conjunction with the Current Population Survey ( 21 ). This survey was completed by men only, so we lack information on the educational attainment of women in the earliest birth cohorts. By presenting all results separately for men and women, patterns over time can be compared by gender.

The datasets were prepared to provide consistent measures of family income, college enrollment, and college completion. We produce simple binary variables that capture whether an individual completed a 4-year degree, whether an individual enrolled in (without necessarily completing) a 4-year degree program, and whether an individual enrolled in (without necessarily completing) a college program. Unfortunately, the tax data results pertain only to college enrollment per se, so we have fewer available data points for the analyses of 4-year completion and enrollment than for the analyses of enrollment in any college program. All samples are restricted to individuals who enrolled in high school, in order to maximize consistency across samples. In SI Appendix , we also include results for a smaller sample restricted to high school graduates ( SI Appendix , Fig. S6 ).

A more difficult variable to harmonize over time is family income. Although in some datasets family income is measured directly (e.g., annual net family income in dollars), in many of the available datasets family income is measured only as an ordinal variable. For these datasets, we employ the method used by Reardon ( 9 ) to calculate test score gaps from coarsened family income data; the method uses the proportions in each income category to assign an income rank to all of those in a given category, and income rank is then the explanatory variable in the analysis ( SI Appendix , SI Methods ).

We estimate logits predicting college enrollment and completion as a function of family income or income rank. Following Reardon ( 9 ), we fit squared and cubed terms to capture the nonlinear effects of income rank. Using the model, we estimate the enrollment and completion rates of those at the 90th percentile of family income and those at the 10th percentile. We choose the 90 vs. 10 comparison over other ways of defining inequality because it accords with past assessments and with the main source of trend in income inequality ( 9 ). † From these rates, we calculate log-odds ratios capturing, for example, the log-odds of completing a 4-year college degree for the 90 vs. 10 family income comparison.

We would be remiss if we did not note the difficulty in measuring family income reliably, particularly using one-shot measures, which are all that are available in almost all of the datasets that we analyze. Further worries might arise because some of the income measures are retrospective, or because the questions are asked of children, not parents. Although we would not minimize the danger of retrospection or of using children’s reports of family income, evidence suggests that child reports of parental socioeconomic characteristics are not substantially worse than parental reports of those characteristics ( 9 , 22 ). Furthermore, the types of errors that individuals make when reporting income appear to have changed very little over time ( 23 ), which is the key issue when mapping trend. To address concerns about the varying quality of the family income data, we multiply all log-odds ratios by 1 / r , where r is the estimated reliability of the family income measure (see SI Appendix , Table S5 for reliability estimates) ( 9 ).

We recognize that “researcher degrees of freedom” are of particular concern when presenting results from a large number of datasets ( 24 ). We provide additional results based on alternative specifications, in SI Appendix , and make our analysis code publicly available on Open Science Framework, https://osf.io/jxne5 .

The Great U-turn in Collegiate Inequality

We now examine collegiate inequalities for cohorts born between 1908 and 1995. Given data constraints, we are limited to examining inequalities over the whole period for men only, but we present results for women for a more limited range of birth cohorts.

In Fig. 2 we present, for the full male series, the estimated probabilities of completing 4-year college at the 90th and 10th percentiles of family income. ‡ We see in Fig. 2 that the increase in 4-year college degree attainment over the twentieth century was far from equally distributed across income groups. Men from the 90th percentile of family income were at the leading edge of the expansion; the figure shows a rapid increase in college completion rates through the 1940s birth cohorts, then a tailing off through the 1950s cohorts, followed by a further rapid increase for those cohorts born in the 1960s onward. In contrast, expansion at the bottom of the income distribution was more sluggish; 4-year college completion rates at the 10th percentile were less than 10 percentage points higher for cohorts born at the end of the century than for cohorts born at the beginning.

An external file that holds a picture, illustration, etc.
Object name is pnas.1907258117fig02.jpg

Probabilities of 4-year college completion at the 90th and 10th percentiles of family income, male birth cohorts, 1908–1986.

Fig. 2 shows that absolute differences in completion rates between income groups increased from the beginning to the end of the century. But this important result must be considered alongside changes over the century in the overall completion rate ( 12 ). Although the probability gap was small at the beginning of the century, the odds of college completion were around 7 times higher for the rich than for the poor, because the rich were able to secure a large proportion of the limited number of college slots. In relative terms, the poor born in the early century were more disadvantaged than their counterparts born in the 1960s, when 90 vs. 10 gaps in the probability of college completion were substantially larger. Although both probability gap and odds-ratio measures are informative, we focus from this point forward on odds-ratio measures of educational inequality, which are margin insensitive and thus feature relative—rather than absolute—advantage. But, in SI Appendix , we present probability plots for the three collegiate outcomes ( SI Appendix , Fig. S1 ), and include analyses based on probability gaps in SI Appendix , Table S3 . The key results hold for both types of analysis.

We plot, in Fig. 3 , the 90 vs. 10 log-odds ratios describing inequalities in collegiate outcomes for each of the datasets in our analyses, with trends estimated from generalized additive models (GAM). The GAMs are fitted to the plotted data points, with each point weighted by the inverse of the SE for the estimate. § In the earlier period covered by OCG, we fit the model to the estimates derived from analyses of single birth cohorts, but present point estimates representing groups of birth cohorts to show the consistency across these specifications. Confidence intervals are presented in SI Appendix , Fig. S2 ; figures showing 90 vs. 50 and 50 vs. 10 inequalities are included as SI Appendix , Figs. S3 and S4 .

An external file that holds a picture, illustration, etc.
Object name is pnas.1907258117fig03.jpg

The 90 vs. 10 log-odds ratios expressing inequality in 4-year completion, 4-year enrollment, and any college enrollment. ( Left ) Male birth cohorts, 1908–1995; ( Right ) female birth cohorts, 1951–1995.

We focus first on describing the trends for men, for whom we have results spanning the whole century. It is clear from Fig. 3 that the over-time trends are similar across the various collegiate outcomes and, further, that there is no simple secular trend for any of the outcomes under consideration. There are three key attributes of the trends that should be emphasized.

First, Fig. 3 shows that, toward the middle of the century, there was a great U-turn in collegiate inequality. Inequalities fell rapidly for cohorts born in the early to mid-1950s, then bottomed out until the mid-1960s, before ultimately rising steeply for cohorts born from the mid-1960s onward. The U-turn appears to be more pronounced for 4-year and “any college” enrollment than for completion of a 4-year degree, but it is present for all of the collegiate outcomes under consideration.

Had we measured collegiate inequalities in but a single dataset, we might be skeptical that our observed trend was on the mark and, in particular, that there was a rapid fall in inequality for the midcentury birth cohorts. But this trend is supported across all of the datasets from the period: OCG and National Longitudinal Study (NLS) Young Men show high inequality in the early 1950s; Panel Study of Income Dynamics (PSID), NLS72, and High School and Beyond (HS&B) pick up the lower inequality of the mid-1950s to the mid-1960s; and the subsequent uptick in inequality is captured in PSID, the school cohort surveys, and the National Longitudinal Studies of Youth (NLSY79&97). Indeed, Fig. 3 demonstrates that there is great consistency across a large number of different data sources. ¶ At the trough, inequality in 4-year college completion was reduced to a log-odds ratio of around 1.5, indicating that, even in this low-inequality period, the odds of those at the 90th income percentile completing a 4-year college degree were almost 4.5 times greater than the equivalent odds for those at the 10th percentile. Inspection of SI Appendix , Fig. S3 suggests that the U-turn observed in Fig. 3 is largely driven by changes in the top half of the income distribution: the U-turn is rather more pronounced for the 90 vs. 50 comparison than for the 50 vs. 10 comparison.

Second, if skepticism about a midcentury fall in collegiate inequality were to be sustained, suspicion would also have to fall upon all currently accepted results on over-time trends, which demonstrate a substantial increase in inequalities in college enrollment and completion between cohorts born in the midcentury and late century. If we were to impose a simple linear smooth on the century-long data series, this would indicate relatively modest increases in collegiate inequalities over the period taken as a whole (see dashed lines, Fig. 3 ). # Again, because the trends are mapped using multiple datasets, we are confident that the pattern of a U-turn in collegiate inequality is supported.

Third, any evidence of a U-turn must bring to mind the pattern of income inequality over the past century. As Piketty and Saez ( 27 ) described, toward the middle of the twentieth century, the share of income going to the top 10% rapidly declined, before rising again over the later decades of the century. The U-turn in collegiate inequality mimics this trend, although it is notable that, insofar as we see similarity in patterns of income inequality and collegiate inequalities, it is income inequality around year of birth that appears to matter most. But, despite the obvious similarities, there is at least one clear divergence in the pattern of collegiate inequality and income inequality: The U-turn in collegiate inequality comes very late. Income inequality begins to fall in the early 1940s, but inequalities in enrollment and completion begin to decline only for cohorts born in the mid-1950s. Men born in the mid-1940s onward were not just born into a period of low inequality, but they spent most of their formative years in a low-inequality society. Despite this, the evidence shows that collegiate inequality increased substantially for the cohorts born in the 1940s and early 1950s; the log-odds ratios describing inequality are increased by around a third over this short period.

Some of the same key features are visible in the results for women, shown in Fig. 3 , Right , although we only have access to data for women born after 1950. We see a basic similarity with the men’s analyses from the mid-1950s birth cohorts onward: Collegiate inequalities are relatively flat for the 1950s to 1960s birth cohorts, and increase for women born in the 1970s and onward. Just as with men, toward the end of the period we see flat and even declining inequalities in enrollment and completion. There are perhaps some subtle differences in the pattern by gender—the upturn in collegiate inequality begins, for example, several years later for women than for men—but we have little evidence here to support a conclusion of substantial difference in inequality for men and women over this period.

There is one notable difference between the men’s and women’s results, relating to the period when trends in male collegiate inequality substantially diverged from trends in income inequality. This exceptional period appears to be exceptional for men, but not for women. Although we cannot track collegiate inequalities for women across the whole midcentury period, the first data points in the female data series (NLS Young Women: 1951–1953 birth cohorts) are lower than the nearby estimates for men (NLS Young Men: 1949–1951 birth cohorts). ** This period of divergence between collegiate inequality and income inequality coincides with the period that we identified above as holding special consequences for men’s educational attainment: Men born in the 1940s and early 1950s were subject to the threat of military service in the Vietnam War.

There are no cohort studies of women that would allow us to compare male and female inequalities in college enrollment and completion throughout this period. We do, however, have access to data on men who fathered children who were at risk for service during the Vietnam War: The NLS Older Men survey can be used to track collegiate inequalities for the children of men who were aged 45 y to 59 y in 1966. The structure of this dataset is somewhat different from the datasets underlying our time series, but we nevertheless find confirmation, in Fig. 4 , that male and female inequalities diverged in the Vietnam years.

An external file that holds a picture, illustration, etc.
Object name is pnas.1907258117fig04.jpg

The 90 vs.10 log-odds ratios expressing inequality in 4-year college completion, 4-year enrollment, and any college enrollment, men and women born 1935–1943 and 1944–1952, NLS-Older Men data.

In the pre-Vietnam period, male and female collegiate inequalities were of similar magnitude. The log-odds ratio for 4-year enrollment, for example, was 2.3 for men (95% CI: 1.5, 3.1), as compared to 2.4 for women (1.7, 3.2). But, for the birth cohorts at risk for serving in Vietnam, the male log-odds ratio increased slightly, to 2.5 (1.8, 3.2), while inequality fell substantially for women, to 1.4 (0.8, 2.0) (see SI Appendix , Fig. S8 for a figure with CIs). These results provide support for the claim that men’s collegiate inequality was substantially and artificially raised relative to expected levels during this period because of the Vietnam War. Unfortunately, our data are not well-suited to evaluating why male and female collegiate inequality differed in the Vietnam period. But some evidence can be brought to bear on this question by comparing preservice and postservice inequalities in college participation for the men in OCG ( SI Appendix , Fig. S9 ). These data are more consistent with a draft-induced increase in male collegiate inequality than with a GI Bill-induced increase. ††

Bringing the results in Fig. 4 together with what is known about college enrollment and completion patterns during the Vietnam War period, it seems likely that the disproportionate increase in men’s college participation rates observed in Fig. 1 was achieved, at least in part, through a gender-specific change in the effect of family income on college enrollment and completion.

The Association between Income Inequality and Collegiate Inequality.

We now present a formal statistical test of the strength of the association between income inequality and collegiate inequality. We regress the log-odds for collegiate inequalities on income inequality, as measured through the share of wages going to the top 10% ( 27 ). ‡‡ In addition to the income inequality variable, for the full male series (1908–1995), we fit a “Vietnam effect,” with a dummy variable that isolates the cohorts at risk from the draft lotteries (i.e., 1944–1952 birth cohorts). We fit models to the full male series (1908–1995 birth cohorts), a compressed male series (1952–1995 birth cohorts), and the female series (1951–1995 birth cohorts). A full regression table with coefficients and standard errors is included as SI Appendix , Table S4 . §§ In Fig. 5 , we present estimates of the predicted increase in the log-odds ratios for an eight percentage point increase in the share of wages going to the top 10%; this increase is equivalent to the “takeoff” in income inequality that occurred between the midcentury and the 1990s. ¶¶

An external file that holds a picture, illustration, etc.
Object name is pnas.1907258117fig05.jpg

Predicted increase in collegiate inequality log-odds ratios associated with the top 10%’s share of wages increasing by 0.08 (equivalent to the takeoff in income inequality); 90 vs. 50 (dark gray), 50 vs. 10 (light gray), and 90 vs. 10 (total) comparisons.

The regression coefficients describing the associations between income inequality and 90 vs. 10 collegiate inequalities can be straightforwardly decomposed into two parts: an association between income inequality and the 90 vs. 50 log-odds ratio, and an association between income inequality and the 50 vs. 10 log-odds ratio. In Fig. 5 , the total height of each bar represents the predicted increase in the 90 vs. 10 log-odds ratio for an eight percentage point increase in income inequality, while the dark and light gray bars show the predicted increases in the 90 vs. 50 and 50 vs. 10 log-odds ratios, respectively.

Examining first the results for the 90 vs. 10 comparison, we see confirmation of a relatively strong association between income inequality and collegiate inequality over the full sweep of the twentieth century. For women, for example, the model predicts that an increase in income inequality equivalent to that observed in the takeoff period would increase the 90 vs. 10 log-odds ratio by around 1 for 4-year enrollment and completion, and by around 1.3 for enrollment in any college. Although there is variation in the strength of the association for the different outcome measures, the income inequality effects are large and positive in all of the analyses, indicating substantial support for the income inequality hypothesis.

Given that the takeoff in income inequality was largely characterized by the top of the income distribution moving away from the middle and bottom of the distribution, the income inequality hypothesis would predict larger effect sizes for the 90 vs. 50 comparison than for the 50 vs. 10 comparison. When we decompose the 90 vs. 10 results into 90 vs. 50 and 50 vs. 10 components, we see precisely this result. The income inequality effects for the 90 vs. 50 comparisons in all cases outweigh those for the 50 vs. 10 comparisons, particularly in the analyses of 4-year college enrollment and completion.

But the results also provide grounds for exercising caution when interpreting differences in effect sizes across the models, as the effect sizes in the full and compressed male series are more similar for the “any college” analyses than for the 4-year analyses, where the sample sizes are smaller. Even when analyzing all available datasets and exploiting the full range of variation in income inequality over the century, our statistical power is limited. This is even more clear when we extend the models summarized in Fig. 5 to include additional macro-level regressors that social scientists have previously used to predict inequalities at the college level. These additional variables include the economic returns to schooling, which are assumed to influence individual decisions about whether or not to invest in college education ( 33 ), and the high school graduation rate, which has been shown to influence educational expansion at the college level ( 34 ). As shown in SI Appendix , Table S1 , estimates from these models are more volatile, particularly for women.

The volatility arises because some of our analyses are, like past analyses, limited to more recent cohorts in which the takeoff assumes a monotonically increasing form. This makes it difficult to adjudicate between the large number of monotonically increasing potential causes. An important advantage of our full-century approach is that it reaches back to a time in which these competing causes did not always move together. In Fig. 6 , we present the results of a simulation exercise, in which we run 1,000 regressions for a range of different model specifications on the full and compressed male series, with each regression including a new variable containing random numbers drawn from a normal distribution ( μ = 0; σ = 1). We examine the stability of the income inequality effects with respect to inequality in college enrollment, for which we have the largest number of data points. We add to the basic model in Fig. 5 controls for time, either in the form of 1) a linear effect of year or 2) dummies for decades, and measures of the returns to schooling ( 33 , 35 , 36 ) and the high-school graduation rate ( 34 , 37 ).

An external file that holds a picture, illustration, etc.
Object name is pnas.1907258117fig06.jpg

Predicted income inequality effects (coefficients × 0.08) from 1,000 regressions of 90 vs. 10 inequality in “any college” enrollment on income inequality and random number variables, for various model specifications, for full and compressed series, men only. Models: 1, Inequality; 2, Inequality+year; 3, Inequality+controls; 4, Inequality+controls+year; and 5, Inequality+controls+decade.

As Fig. 6 shows, the income inequality effects estimated for the full male series are robust to the inclusion of other potential confounding variables. But Fig. 6 also highlights the extent to which a proper evaluation of the income inequality hypothesis requires researchers to exploit all of the available data. Although the bivariate analysis shows a similar effect of the income inequality variable in both the full and compressed series, the effects are a good deal more volatile in the more highly parameterized models in the compressed relative to the full series. *** The substantive implication of this analysis is clear: It is only with the full data series that we obtain relatively precise and reliable estimates of the association between inequality in collegiate outcomes and income inequality.

We have examined descriptive evidence on the association between inequality in collegiate attainment and income inequality over the past century. Although there has been much recent interest in the income inequality hypothesis, it has been difficult to make headway because commonly used datasets pertain only to recent decades, when income inequality was increasing. We have thus proceeded by reaching back to the very beginning of the twentieth century, assembling all of the available datasets, and harmonizing the variables in these datasets.

The results show that collegiate inequalities and income inequality are, in fact, rather strongly associated over the twentieth century. Just as with income inequality, we see evidence of a U-turn in 90 vs. 10 collegiate inequality, and evidence of a substantial takeoff in collegiate inequalities in recent decades. When we examine trends in 90 vs. 50 and 50 vs.10 inequalities, we find that the 90 vs. 50 trends mirror the 90 vs. 10 results. Taken together, our results offer solid descriptive support for the income inequality hypothesis.

Inequalities in collegiate attainment increased hand in hand with the expansion of college education in the United States. Rates of college enrollment and completion were higher at the end of the century than they had been at any time in the preceding hundred years, and yet, for these birth cohorts, we see substantial inequalities, as captured in both percentage point gap and odds ratio measures. In point of fact, the only time during the twentieth century for which we observe a reduction in educational inequality is during the period when expansion at the college level had paused. Although the counterfactual is obviously not observable, these results emphasize the importance of attending to the distribution of college opportunities in addition to overall levels of attainment. These distributional questions will take on even greater significance in the context of the economic and social crisis engendered by coronavirus disease 2019, a crisis that is likely to have enduring effects on both the distribution of income and access to the higher education sector.

Our analyses are not well suited to evaluating the mechanisms generating the association between income inequality and collegiate inequalities. However, given the pattern of collegiate inequality across the century, we suspect that a mechanical effect is likely to be responsible. If money matters, as we know it does, and growing income inequality delivers more money to the top, then, all else being equal, these additional dollars would in themselves produce growing inequality in college enrollment and completion. The mechanical effect is therefore a parsimonious account of the trend that we see here ( 8 ). That the over-time associations are substantially stronger for the 90 vs. 50 comparison as compared to the 50 vs. 10 comparison provides further suggestive evidence in this regard. Nevertheless, there is a period for which we undoubtedly hypothesize an increase in the relational effect of income: the Vietnam War. For the war to lead to increased collegiate inequality, the effect of income on educational attainment would have to increase, particularly given that income inequality was low and stable for these birth cohorts.

Whatever the mechanisms may be, the key descriptive result is that, over the course of the twentieth century, a grand U-turn in collegiate inequality occurred. Cohorts born in the middle of the century witnessed the lowest levels of inequality in college enrollment and completion seen over the past hundred years. Contemporary birth cohorts, in contrast, are experiencing levels of collegiate inequality not seen for generations.

Supplementary Material

Supplementary file, acknowledgments.

We thank David Cox, David Grusky, and Florencia Torche for their detailed comments on earlier versions of this paper, and also Raj Chetty, Maximilian Hell, Robb Willer, the Cornell Mobility Conference, the Stanford Inequality Workshop, the Stanford Sociology Colloquium Series, and University of California, Los Angeles’s California Center for Population Research seminar for useful suggestions. Additionally, we thank Stanford’s Center for Poverty and Inequality, Russell Sage Foundation and Stanford’s United Parcel Service (UPS) Fund for research funding, Stanford’s Institute for Research in the Social Sciences for secure data room access, and the American Institutes for Research for data access. We are grateful to the editor and reviewers for their helpful and productive suggestions.

The authors declare no competing interest.

This article is a PNAS Direct Submission. E.G. is a guest editor invited by the Editorial Board.

Data deposition: Code for data analysis is archived on Open Science Framework ( https://osf.io/jxne5 ).

*Throughout this paper, we use the term “college” as a shorthand for “2- or 4-year college.”

† We also include results based on comparing income quartiles in SI Appendix , Fig. S5 .

‡ The probabilities are estimated from the logit model, and we fit a GAM to establish trend. See SI Appendix , SI Methods for more details.

§ We determine the appropriate number of degrees of freedom for the trend lines by fitting a series of GAMs and comparing model fit (using the Akaike Information Criterion). For the analysis of college enrollment for male birth cohorts, we use the stepwise model builder in R’s gam package to find the best-fitting model ( 25 , 26 ). As we have fewer point estimates in the other analyses, the stepwise approach is less reliable, and we therefore choose smoothing parameters that provide a reasonable (and conservative) summary of the trend.

¶ It is also clear that some datasets are outliers from the trend. It is not surprising to see variation across samples, and we highlight this variation only because it illustrates a potential danger of using but one or two datasets to establish a trend. The estimates for National Education Longitudinal Study (NELS) (1974), for example, are substantially higher than the surrounding estimates based on one-shot income measures, and there is a surprising degree of cross-cohort volatility in the PSID estimates.

# The linear trend is strongest for 4-year completion, and weakest for enrollment in 4-year college. For all collegiate outcomes, the GAM offers a significant improvement in fit over the simple linear model.

**It would be possible to track male and female educational inequality with respect to parental education or socioeconomic index scores (SEI) ( 28 ), but the sample sizes are, unfortunately, too small for a detailed analysis of gender differences in educational attainment by birth cohort. This approach is also unattractive given that parental education, parental income, and SEI were only weakly correlated in this period ( 29 ).

†† Note that, while previous research has suggested that high-socioeconomic status (SES) individuals might have taken advantage of the GI Bill to a greater extent than low-SES individuals ( 30 ), SI Appendix , Fig. S9 provides little evidence that collegiate inequality was substantially affected. See SI Appendix for further discussion of this point.

‡‡ We choose the wages measure because, for the bottom of the income distribution, wages are a more important component of income than the types of income included in the alternative measures (e.g., capital gains). We measure wage inequality in year of birth. Surprisingly, given the prominence of the income inequality hypothesis, there is not yet adequate guidance in the literature as to the age at which income inequality most influences outcomes, although in the “money matters” literature there has been particular emphasis on the prenatal period, the postnatal period, and early childhood as the lifecourse moments when money matters most ( 31 , 32 ).

§§ In the 4-year analyses, we weight the data by the inverse of the standard errors underlying the estimates. In the analysis of any college enrollment, we do not weight the data, as this data series includes the tax data estimates. Given the size of the samples underlying these estimates, weighting would allow the relationship that pertains in the tax data for cohorts born in the 1980s and 1990s to have a disproportionate influence on the estimated century-long relationship between income inequality and inequality in college enrollment.

¶¶ The estimates in Fig. 5 are obtained by multiplying the income inequality coefficients in SI Appendix , Table S4 by 0.08.

***See SI Appendix , Fig. S10 for similar figures for 4-year enrollment and completion.

This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907258117/-/DCSupplemental .

Data Availability.

Featured Topics

Featured series.

A series of random questions answered by Harvard experts.

Explore the Gazette

Read the latest.

Your kid can’t name three branches of government? He’s not alone.

‘We have the most motivated people, the best athletes. How far can we take this?’

Houghton Library acquired a 1948 edition of "The Green Book," a travel guide for Blacks during segregation times..

Harvard Library acquires copy of ‘Green Book’

The costs of inequality: Education’s the one key that rules them all

When there’s inequity in learning, it’s usually baked into life, Harvard analysts say

Corydon Ireland

Harvard Correspondent

Third in a series on what Harvard scholars are doing to identify and understand inequality, in seeking solutions to one of America’s most vexing problems.

Before Deval Patrick ’78, J.D. ’82, was the popular and successful two-term governor of Massachusetts, before he was managing director of high-flying Bain Capital, and long before he was Harvard’s most recent Commencement speaker , he was a poor black schoolchild in the battered housing projects of Chicago’s South Side.

The odds of his escaping a poverty-ridden lifestyle, despite innate intelligence and drive, were long. So how did he help mold his own narrative and triumph over baked-in societal inequality ? Through education.

“Education has been the path to better opportunity for generations of American strivers, no less for me,” Patrick said in an email when asked how getting a solid education, in his case at Milton Academy and at Harvard, changed his life.

“What great teachers gave me was not just the skills to take advantage of new opportunities, but the ability to imagine what those opportunities could be. For a kid from the South Side of Chicago, that’s huge.”

If inequality starts anywhere, many scholars agree, it’s with faulty education. Conversely, a strong education can act as the bejeweled key that opens gates through every other aspect of inequality , whether political, economic , racial, judicial, gender- or health-based.

Simply put, a top-flight education usually changes lives for the better. And yet, in the world’s most prosperous major nation, it remains an elusive goal for millions of children and teenagers.

Plateau on educational gains

The revolutionary concept of free, nonsectarian public schools spread across America in the 19th century. By 1970, America had the world’s leading educational system, and until 1990 the gap between minority and white students, while clear, was narrowing.

But educational gains in this country have plateaued since then, and the gap between white and minority students has proven stubbornly difficult to close, says Ronald Ferguson, adjunct lecturer in public policy at Harvard Kennedy School (HKS) and faculty director of Harvard’s Achievement Gap Initiative. That gap extends along class lines as well.

“What great teachers gave me was not just the skills to take advantage of new opportunities, but the ability to imagine what those opportunities could be. For a kid from the South Side of Chicago, that’s huge.” — Deval Patrick

In recent years, scholars such as Ferguson, who is an economist, have puzzled over the ongoing achievement gap and what to do about it, even as other nations’ school systems at first matched and then surpassed their U.S. peers. Among the 34 market-based, democracy-leaning countries in the Organization for Economic Cooperation and Development (OECD), the United States ranks around 20th annually, earning average or below-average grades in reading, science, and mathematics.

By eighth grade, Harvard economist Roland G. Fryer Jr. noted last year, only 44 percent of American students are proficient in reading and math. The proficiency of African-American students, many of them in underperforming schools, is even lower.

“The position of U.S. black students is truly alarming,” wrote Fryer, the Henry Lee Professor of Economics, who used the OECD rankings as a metaphor for minority standing educationally. “If they were to be considered a country, they would rank just below Mexico in last place.”

Harvard Graduate School of Education (HGSE) Dean James E. Ryan, a former public interest lawyer, says geography has immense power in determining educational opportunity in America. As a scholar, he has studied how policies and the law affect learning, and how conditions are often vastly unequal.

His book “Five Miles Away, A World Apart” (2010) is a case study of the disparity of opportunity in two Richmond, Va., schools, one grimly urban and the other richly suburban. Geography, he says, mirrors achievement levels.

A ZIP code as predictor of success

“Right now, there exists an almost ironclad link between a child’s ZIP code and her chances of success,” said Ryan. “Our education system, traditionally thought of as the chief mechanism to address the opportunity gap, instead too often reflects and entrenches existing societal inequities.”

Urban schools demonstrate the problem. In New York City, for example, only 8 percent of black males graduating from high school in 2014 were prepared for college-level work, according to the CUNY Institute for Education Policy, with Latinos close behind at 11 percent. The preparedness rates for Asians and whites — 48 and 40 percent, respectively — were unimpressive too, but nonetheless were firmly on the other side of the achievement gap.

In some impoverished urban pockets, the racial gap is even larger. In Washington, D.C., 8 percent of black eighth-graders are proficient in math, while 80 percent of their white counterparts are.

Fryer said that in kindergarten black children are already 8 months behind their white peers in learning. By third grade, the gap is bigger, and by eighth grade is larger still.

According to a recent report by the Education Commission of the States, black and Hispanic students in kindergarten through 12th grade perform on a par with the white students who languish in the lowest quartile of achievement.

There was once great faith and hope in America’s school systems. The rise of quality public education a century ago “was probably the best public policy decision Americans have ever made because it simultaneously raised the whole growth rate of the country for most of the 20th century, and it leveled the playing field,” said Robert Putnam, the Peter and Isabel Malkin Professor of Public Policy at HKS, who has written several best-selling books touching on inequality, including “Bowling Alone: The Collapse and Revival of the American Community” and “Our Kids: The American Dream in Crisis.”

Historically, upward mobility in America was characterized by each generation becoming better educated than the previous one, said Harvard economist Lawrence Katz. But that trend, a central tenet of the nation’s success mythology, has slackened, particularly for minorities.

“Thirty years ago, the typical American had two more years of schooling than their parents. Today, we have the most educated group of Americans, but they only have about .4 more years of schooling, so that’s one part of mobility not keeping up in the way we’ve invested in education in the past,” Katz said.

As globalization has transformed and sometimes undercut the American economy, “education is not keeping up,” he said. “There’s continuing growth of demand for more abstract, higher-end skills” that schools aren’t delivering, “and then that feeds into a weakening of institutions like unions and minimum-wage protections.”

“The position of U.S. black students is truly alarming.” — Roland G. Fryer Jr.

Fryer is among a diffuse cohort of Harvard faculty and researchers using academic tools to understand the achievement gap and the many reasons behind problematic schools. His venue is the Education Innovation Laboratory , where he is faculty director.

“We use big data and causal methods,” he said of his approach to the issue.

Fryer, who is African-American, grew up poor in a segregated Florida neighborhood. He argues that outright discrimination has lost its power as a primary driver behind inequality, and uses economics as “a rational forum” for discussing social issues.

Better schools to close the gap

Fryer set out in 2004 to use an economist’s data and statistical tools to answer why black students often do poorly in school compared with whites. His years of research have convinced him that good schools would close the education gap faster and better than addressing any other social factor, including curtailing poverty and violence, and he believes that the quality of kindergarten through grade 12 matters above all.

Supporting his belief is research that says the number of schools achieving excellent student outcomes is a large enough sample to prove that much better performance is possible. Despite the poor performance by many U.S. states, some have shown that strong results are possible on a broad scale. For instance, if Massachusetts were a nation, it would rate among the best-performing countries.

At HGSE, where Ferguson is faculty co-chair as well as director of the Achievement Gap Initiative, many factors are probed. In the past 10 years, Ferguson, who is African-American, has studied every identifiable element contributing to unequal educational outcomes. But lately he is looking hardest at improving children’s earliest years, from infancy to age 3.

In addition to an organization he founded called the Tripod Project , which measures student feedback on learning, he launched the Boston Basics project in August, with support from the Black Philanthropy Fund, Boston’s mayor, and others. The first phase of the outreach campaign, a booklet, videos, and spot ads, starts with advice to parents of children age 3 or younger.

“Maximize love, manage stress” is its mantra and its foundational imperative, followed by concepts such as “talk, sing, and point.” (“Talking,” said Ferguson, “is teaching.”) In early childhood, “The difference in life experiences begins at home.”

At age 1, children score similarly

Fryer and Ferguson agree that the achievement gap starts early. At age 1, white, Asian, black, and Hispanic children score virtually the same in what Ferguson called “skill patterns” that measure cognitive ability among toddlers, including examining objects, exploring purposefully, and “expressive jabbering.” But by age 2, gaps are apparent, with black and Hispanic children scoring lower in expressive vocabulary, listening comprehension, and other indicators of acuity. That suggests educational achievement involves more than just schooling, which typically starts at age 5.

Key factors in the gap, researchers say, include poverty rates (which are three times higher for blacks than for whites), diminished teacher and school quality, unsettled neighborhoods, ineffective parenting, personal trauma, and peer group influence, which only strengthens as children grow older.

“Peer beliefs and values,” said Ferguson, get “trapped in culture” and are compounded by the outsized influence of peers and the “pluralistic ignorance” they spawn. Fryer’s research, for instance, says that the reported stigma of “acting white” among many black students is true. The better they do in school, the fewer friends they have — while for whites who are perceived as smarter, there’s an opposite social effect.

The researchers say that family upbringing matters, in all its crisscrossing influences and complexities, and that often undercuts minority children, who can come from poor or troubled homes. “Unequal outcomes,” he said, “are from, to a large degree, inequality in life experiences.”

Trauma also subverts achievement, whether through family turbulence, street violence, bullying, sexual abuse, or intermittent homelessness. Such factors can lead to behaviors in school that reflect a pervasive form of childhood post-traumatic stress disorder.

[gz_sidebar align=”left”]

Possible solutions to educational inequality:

Access to early learning
Improved K-12 schools
More family mealtimes
Reinforced learning at home
Data-driven instruction
Longer school days, years
Respect for school rules
Small-group tutoring
High expectations of students
Safer neighborhoods

[/gz_sidebar]

At Harvard Law School, both the Trauma and Learning Policy Initiative and the Education Law Clinic marshal legal aid resources for parents and children struggling with trauma-induced school expulsions and discipline issues.

At Harvard Business School, Karim R. Lakhani, an associate professor who is a crowdfunding expert and a champion of open-source software, has studied how unequal racial and economic access to technology has worked to widen the achievement gap.

At Harvard’s Project Zero, a nonprofit called the Family Dinner Project is scraping away at the achievement gap from the ground level by pushing for families to gather around the meal table, which traditionally was a lively and comforting artifact of nuclear families, stable wages, close-knit extended families, and culturally shared values.

Lynn Barendsen, the project’s executive director, believes that shared mealtimes improve reading skills, spur better grades and larger vocabularies, and fuel complex conversations. Interactive mealtimes provide a learning experience of their own, she said, along with structure, emotional support, a sense of safety, and family bonding. Even a modest jump in shared mealtimes could boost a child’s academic performance, she said.

“We’re not saying families have to be perfect,” she said, acknowledging dinnertime impediments like full schedules, rudimentary cooking skills, the lure of technology, and the demands of single parenting. “The perfect is the enemy of the good.”

Whether poring over Fryer’s big data or Barendsen’s family dinner project, there is one commonality for Harvard researchers dealing with inequality in education: the issue’s vast complexity. The achievement gap is a creature of interlocking factors that are hard to unpack constructively.

Going wide, starting early

With help from faculty co-chair and Jesse Climenko Professor of Law Charles J. Ogletree, the Achievement Gap Initiative is analyzing the factors that make educational inequality such a complex puzzle: home and family life, school environments, teacher quality, neighborhood conditions, peer interaction, and the fate of “all those wholesome things,” said Ferguson. The latter include working hard in school, showing respect, having nice friends, and following the rules, traits that can be “elements of a 21st-century movement for equality.”

In the end, best practices to create strong schools will matter most, said Fryer.

He called high-quality education “the new civil rights battleground” in a landmark 2010 working paper for the Handbook of Labor Economics called “Racial Inequality in the 21st Century: The Declining Significance of Discrimination.”

Fryer tapped 10 large data sets on children 8 months to 17 years old. He studied charter schools, scouring for standards that worked. He champions longer school days and school years, data-driven instruction, small-group tutoring, high expectations, and a school culture that prizes human capital — all just “a few simple investments,” he wrote in the working paper. “The challenge for the future is to take these examples to scale” across the country.

How long would closing the gap take with a national commitment to do so? A best-practices experiment that Fryer conducted at low-achieving high schools in Houston closed the gap in math skills within three years, and narrowed the reading achievement gap by a third.

“You don’t need Superman for this,” he said, referring to a film about Geoffrey Canada and his Harlem Children’s Zone, just high-quality schools for everyone, to restore 19th-century educator Horace Mann’s vision of public education as society’s “balance-wheel.”

Last spring, Fryer, still only 38, won the John Bates Clark medal, the most prestigious award in economics after the Nobel Prize. He was a MacArthur Fellow in 2011, became a tenured Harvard professor in 2007, was named to the prestigious Society of Fellows at age 25. He had a classically haphazard childhood, but used school to learn, grow, and prosper. Gradually, he developed a passion for social science that could help him answer what was going wrong in black lives because of educational inequality.

With his background and talent, Fryer has a dramatically unique perspective on inequality and achievement, and he has something else: a seemingly counterintuitive sense that these conditions will improve, once bad schools learn to get better. Discussing the likelihood of closing the achievement gap if Americans have the political and organizational will to do so, Fryer said, “I see nothing but optimism.”

Correction: An earlier version of this story inaccurately portrayed details of Dr. Fryer’s background.

Illustration by Kathleen M.G. Howlett. Harvard staff writer Christina Pazzanese contributed to this report.

Next Tuesday: Inequality in health care

Share this article

Good genes are nice, but joy is better

Harvard study, almost 80 years old, has proved that embracing community helps us live longer, and be happier

Examining new weight-loss drugs, pediatric bariatric patients

Researcher says study found variation in practices, discusses safety concerns overall for younger users

Shingles may increase risk of cognitive decline

Availability of vaccine offers opportunity to reduce burden of shingles and possible dementia

The impact of education costs on income inequality

Research Article
Open access
Published: 03 April 2024
Volume 71 , pages 553–574, ( 2024 )

Cite this article

You have full access to this open access article

Fa-Hsiang Chang 1

884 Accesses

Explore all metrics

Reducing education costs is a crucial policy element with broad support in the U.S. However, is lowering the cost of learning a panacea for eliminating income inequality? In this paper, I theoretically examine the relationship between income inequality and the cost of education by building a three-stage overlapping generation model with two sectors and two education systems. In contrast to conventional studies treating education as a unified concept or in hierarchical order, I consider two types of education, each targeting the training of workers for different roles in production. Workers who decide to spend time learning and improving creativity skills can work to produce intermediate goods used in current production and illuminate future production technology. Coders who produce industrial robots are one example of workers who receive this type of education. The other type of education only improves workers’ efficiency and helps them become experts in positions. For instance, office clerks who receive computer training can become more productive in dealing with their daily tasks. I pick a reasonable set of parameters, and the simulation result implies that reducing the cost of practical training may end up enlarging income inequality. The key is whether the effect resulting from the wage gap between jobs ( wage effect ) dominates the effect of changing the share of workers in jobs ( composition effect ). Reducing the cost of learning creativity encourages marginal experts to learn creativity and marginal basic workers to receive training, leading to declining wage gaps and income inequality.

Employment Market Effects of Basic Income

Optimal linear income taxes and education subsidies under skill-biased technical change

Academic or vocational education? A comparison of the long-term wage development of academic and vocational tertiary degree holders

Explore related subjects.

Artificial Intelligence

Avoid common mistakes on your manuscript.

1 Introduction

Is lowering the cost of learning a panacea for eliminating income inequality? The answers are highly debatable in economics and vary across countries and time. Back to Schultz ( 1963 ), investing in education is seen as a means to reduce income inequality by enhancing human capital. There are a bunch of empirical findings that support this positive relationship. See Gregorio and Lee ( 2002 ), Sylwester ( 2002 ) and Coady and Dizioli ( 2018 ) for a broad range of countries; Abdullah et al. ( 2015 ) for Africa, for example. However, it is essential to note that the relationship between education and income inequality is complex. According to Knight and Sabot ( 1983 ), education expansion increases the proportion of educated workers ( composition effect ). However, it may lead to a decline in wage premiums ( wage effect ), resulting in an ambiguous effect on inequality. Moreover, in some cases, education may primarily serve as a signal rather than directly reducing inequality (Spence 1973 ). Some empirical findings support that education may have little or even a detrimental effect on income inequality (see Battistón et al. 2014 for most countries in Latin America; Rodríguez-Pose and Tselios 2009 for regions in the European Union; etc.). Ram ( 1989 ) reviews previous theoretical and empirical papers and reaches the same conclusion.

However, the current studies either treat education as a homogeneous concept or categorize it in a hierarchical manner (e.g., Rodríguez-Pose and Tselios 2009 ; Abdullah et al. 2015 ) without considering its categorization based on the labor market objectives. Different jobs require specific and distinct skill sets in today’s technologically advanced world with rapidly evolving knowledge. Some jobs demand creative problem-solving abilities, while others focus on repetitive tasks. As a result, different types of education and training are required and established. This paper theoretically examines the relationship between the cost of training and income inequality, considering the two types of training based on the type of workers they nurture.

Following Becker and Tomes ( 1979 )’s string of thoughts, the income inequality in the economy is composed of inter-generational and within-cohort income differences. In this paper, I establish a three-period overlapping generation model to discuss the inter-generational income differences. In this paper, I do not focus on on-the-job training. I assume that individuals make education decisions when young, work during middle age, and retire in old age. To account for the within-cohort differences, I assume individuals are heterogeneous in endowed learning ability and choose education and occupation endogenously.

Before delving into the detailed model setup, it is worthwhile to elucidate the concept of training and the associated costs in this paper. In this context, the cost of receiving training incurs not only financial expenses but also consumes considerable time and effort, especially in an economy characterized by rapid knowledge iteration. With the higher skill requirements in the labor market, the costs of training increase when the teaching effectiveness remains constant. The reason is that acquiring the necessary knowledge and qualifications can become more challenging and time-consuming.

In this paper, I classify two types of training based on workers’ roles in production. The economy features two representative firms producing two types of final goods, requiring workers in three distinct types of occupations. Footnote 1 The first type of occupation demands workers capable of performing simple, repetitive tasks without extensive training or professional knowledge, referred to as basic workers. For instance, a janitor falls into this category. The other types of occupations require a certain level of extensive experience or training for more efficient work than basic workers. An example is the first-line supervisor, who is considered an expert due to the extensive knowledge required for the role. Candidates often undergo training and promotion from within similar fields. The type of education that increases workers’ production efficiencies is called practical training in this paper. A janitor, for example, can choose to receive practical training to learn structural cleaning techniques and be promoted to a first-line supervisor role, directly overseeing and coordinating the work activities of cleaning personnel. Basic workers and experts serve as inputs in the production of one type of final goods referred to as “non-automatable” goods in my model, such as service goods. The last group of occupations requires workers with solid professional foundations to explore cutting-edge technology or knowledge. Those are workers who receive the second type of education to enhance their creativity so they can produce intermediate goods used in current production and illuminate future production technology. For instance, coders who produce industrial robots are one example of workers who receive this type of education. Hence, the second type of final goods is called “automatable” goods because production requires intermediate goods (e.g., robots) produced by creative workers (e.g., coders) and physical capital (e.g., machines). Unlike “non-automatable” goods, “automatable” goods do not require labor in production directly. For example, cars can be considered “automatable” goods due to the widespread use of robots and machines in their production.

In the U.S., reducing education costs is a key element of policy with broad support. However, my paper argues that the impact of policy on income inequality depends on the targeted type of education. The paper calibrates the U.S. data and simulates the consequences of reducing two types of education costs on income inequality. The results demonstrate that reducing the cost of learning creativity reduces income inequality while reducing the cost of practical training may end up enlarging income inequality. I focus on within-group income inequality, as the simulated income differences between middle-aged and older workers with similar learning abilities do not change remarkably with reduced education costs. The reduction in the cost of learning creativity attracts more experts to become creative workers and also attracts basic workers to receive practical training. Moreover, the wage gap between basic and creative workers shrinks, leading to a decline in within-cohort and overall income inequality. However, reducing practical training costs makes becoming experts an attractive option to both marginal basic workers and marginal creative workers. So, assuming prices do not change when the training costs decline, the share of experts increases, and both the numbers of basic and creative workers shrink. However, the widening wage gap between creative workers and basic workers attracts workers to become creative workers. Thus, only the share of basic workers declines, and the numbers of experts and creative workers rise. Therefore, the effect of the declining cost of practical training on income inequality depends on the predominance of the composition effect and wage effect.

This study presents a novel explanation of the relationship between education and income inequality without asserting the superiority or universality of different educations. Existing literature examines the complex relationship from two perspectives. One perspective focuses on the different returns on education. According to Mincer ( 1974 ), encouraging individuals to pursue higher education may exacerbate income inequality, as the returns to higher education remain relatively higher than that on compulsory education. Other studies highlight the influence of learning ability and parental investment on educational choices. For instance, Yang and Qiu ( 2016 ) analyzes the impact of innate ability, compulsory education (grades 1–9), and non-compulsory education (grades 10–12 and higher education) on income inequality in China, emphasizing the significance of family investment in early education as a determinant of income inequality. Similarly, Prettner and Schaefer ( 2021 ) emphasizes the importance of inheritance flows on education choices and income inequality. This paper does not explicitly model a joint family utility function but assumes resource transfer from middle-aged to young individuals. The second perspective focuses on the signaling theory. For example, Hendel et al. ( 2005 ) explores the signaling role of education combined with households’ credit constraints, suggesting that affordable borrowing or lower tuition can encourage high-ability individuals to leave the unskilled pool, thereby increasing the skill premium. However, these studies generally consider education in a hierarchical order and do not discuss the types of training that lead to different job choices at the same stage of life. This paper fills this gap in the existing literature. It addresses the importance of skill selection in light of labor market polarization, as highlighted by studies such as Autor and Dorn ( 2013 ), Deming ( 2017 ), and Deming and Kahn ( 2018 ). For instance, Autor and Dorn ( 2013 ) emphasizes the reallocation of low-skill workers to service occupations due to the reduced cost of automating routine and codifiable job tasks, leading to earnings growth at the extremes of the income distribution. By incorporating different types of education based on the skills they develop, this paper contributes to understanding the relationship between education and income inequality.

The paper is organized as follows. Section 2 introduces the model setup and some preliminary theoretical results. Section 3 introduces calibration strategies and presents simulation results. Section 4 provides the concluding remarks.

2 The model

I develop a three-stage overlapping generation model featuring endogenous education and occupation choices in the economy with two final goods. All markets are perfectly competitive, and there are no uncertainties. The model incorporates three key features. Firstly, the economy includes two representative firms producing different goods, and each requires different types of workers in the production technology. Secondly, I introduce two education types that are aligned with the skills they impart. Individuals can opt for (1) no education, (2) practical training, or (3) training in creativity skills. Individuals with perfect foresight choose the types of education they pursue during their youth. Different educational backgrounds lead to distinct roles in production, resulting in different labor incomes when they are middle-aged - the only working period in my model. Thirdly, individuals are heterogeneous in my model. Individuals differ in their endowed learning ability $z \sim F(\cdot )$ . This endowed learning ability does not play a role if individuals choose not to receive education. Thus, the model allows me to discuss the behaviors of workers who receive different types of education across cohorts and how the change in the cost of each type of education impacts them. The model details are introduced in the following sections.

2.1 Firms and technologies

In this economy, two representative firms produce two types of final goods. The first type, referred to as “automatable” goods (e.g., cars), involves tasks in production that do not require manual labor. The firm producing “automatable” goods depends on physical capital and intermediate goods produced by creative workers as inputs. For instance, coders write programs to develop software, and machines are operated on assembly lines to produce the goods. The production of “automatable” goods ( $Y_t$ ) at time t follows the Cobb-Douglas production function:

where $D_Y$ represents total factor productivity in “automatable” goods production, $K_t$ denotes the physical capital stock at time t , and $A_t$ denotes the stock of intermediate goods produced by creative workers at time t . Here, $\alpha$ is the output elasticity of capital, with $0< \alpha < 1$ . In each period t , the production of intermediate goods ( $A_t$ ) by creative workers follows a linear fashion:

where $e^A e^{z^i}$ represents the productivity of a creative worker i after receiving training in creativity. $\bar{z}_{t-1}$ denotes the threshold value of endowed learning ability that divides experts and creative workers for those born at time $t-1$ and work in period t . This threshold value changes over time in the short run but remains a fixed number in the steady state.

The second type of goods is “non-automatable” goods, such as service goods, where production is dependent only on labor directly. “Non-automatable” goods are produced by workers who do not receive training in creativity skills. I assume the “non-automatable” goods production function is linear, where basic workers and experts are perfect substitutes, and the only difference between basic workers and experts is their productivity. For instance, certified massage therapists can undergo practical training, obtain a license, and transform into licensed massage therapists. Their treatments improve as they acquire knowledge in areas such as anatomy and physiology, ultimately passing the bodywork licensing exam administered by states. In the production function of “non-automatable” goods, I exclude physical capital, as it typically plays a minor role in production. For example, a massage therapist relies more on massage skills than machines to treat customers. Thus, I assume the production function of the “non-automatable” goods is

where $D_S$ denotes the total factor productivity in the production of “non-automatable” goods, and $G_t$ is the number of basic workers without training. The endowed learning ability thresholds, $\underline{z}_{t-1}$ and $\bar{z}_{t-1}$ , represent the minimum and maximum values for individuals working at time t . Detailed explanations and closed-form expressions for these thresholds are provided in Sect. 2.2.1 .

2.1.1 Firm’s problem

“Automatable” Goods . I assume the “automatable” good is the numeraire. Given all the prices, the representative firm that produces “automatable” goods solves the following maximization problem to determine the capital demand and intermediate goods demand,

where $m_t$ is the price of the intermediate goods, and $r_t$ is the rental rate of physical capital. The demands for capital and intermediate goods can be expressed in the following first-order conditions,

“Non-automatable” Goods . Given all prices, the representative firm producing the “non-automatable” good maximizes profit by determining the demand for basic workers and experts:

where $q_t$ represents the price of the “non-automatable” goods, and $w_t^G$ is the wage of a basic worker. Given the linear production function, the existence of equilibrium requires the cost of hiring one basic worker and one expert to be equal, clearing the two labor markets,

2.2 Individuals

The economy is populated by overlapping generations of agents, each living for three periods. Within each generation, a continuum of individuals with different learning abilities is born. In this paper, I assume the endowed learning ability of individual i is drawn from a uniform distribution $z^i \sim U(0,1)$ for simplicity. For each period, the economy includes individuals with the same endowed learning ability but born in different generations. In other words, in period t , the economy includes individuals with the same $z^i$ but born in generations $t-2, t-1, t$ . For simplicity, I use subscripts y , m , and o to denote whether the individual is young, middle-aged, or old, respectively, and use t to indicate the current time of the economy.

Individuals in this model value the consumption of two goods throughout their lifetime with a logarithmic utility function. For each individual i born in generation t , the utility is defined as:

where $0<\kappa <1$ , and c , s denote the consumption of “automatable” goods and “non-automatable” goods separately. $c^i_{y,t}, s^i_{y,t}$ show the consumption of “automatable” goods and “non-automatable” goods for individual i born in generation t during their young age at time t . Similarly, $c^i_{m,t+1}, s^i_{m,t+1}$ represent the consumption of “automatable” goods and “non-automatable” goods for the same individual during their middle age at time $t+1$ . Finally, $c^i_{o,t+2}, s^i_{o,t+2}$ signify the consumption of “automatable” goods and “non-automatable” goods for individual i during their old age at time $t+2$ .

Individuals own the factors of production in this economy. They supply physical capital and labor to the firms through factor markets. The physical capital market allows individuals to transfer resources to the next period. Since I assume individuals are heterogeneous, they can engage in borrowing and lending through the trading of one-period capital. In the paper, I assume “automatable” goods can be consumed or invested in each period, but “non-automatable” goods can only be consumed. The conversion rate between “automatable” goods and physical capital is one.

In this model, Individuals endogenously make educational choices when they are young. I assume the young have no initial endowment on the physical capital, so they have to borrow from the capital market in that period to fulfill the needs of consumption and education. Let $f^i_t$ denote the fixed cost of learning, $b^i_{y,t+1}$ as the amount of acquired physical capital stock for the agent i in period t when young. The subscript $t+1$ signifies that the return will be repaid in period $t+1$ . Here, $b^i_{y,t+1}<0$ because the young are borrowing capital. Then, the budget constraint for the individual i born in generation t when young is

Agents are assumed to work full-time and can only work when they are in middle age. Based on their educational choices when young, individuals work in different occupations when they are in middle age. Those who receive no education when young become basic workers in middle age to produce the“non-automatable” goods with productivity normalized to one. For the individual i who receives practical training when young becomes an expert in middle age to produce the “non-automatable” goods with productivity $e^H e^{z^i}$ , and individual i receives training in creativity skill when young becomes a creative worker to produce the intermediate goods in “automatable” goods production with productivity $e^A e^{z^i}$ . Let $I^i_{t+1}$ denotes the labor income for individual i born in generation t in the middle age, where

Furthermore, in the economy at time $t+1$ , middle-aged workers are required to repay $r_{t+1}$ units of interest for each unit of capital borrowed at youth. An individual i born in generation t also engages in the investment of physical capital since it serves as their sole source of income in old age. The physical capital depreciates at the rate of $\delta$ ( $0<\delta <1$ ). Let $b^i_{m,t+2}$ represent the physical capital stock that will yield the capital return for the individual when they reach old age. Therefore, this middle-aged individual will invest $b^i_{m,t+2}-(1-\delta )b^i_{y,t+1}$ units of physical capital. Thus, the budget constraint for the middle-aged individual i born in generation t is

When the individual i who was born at generation t becomes old, the only source of income comes from the physical stock of capital. The budget constraint for this individual when old is

2.2.1 Individual’s problem

For each individual i born in generation t , the agent solves the following maximization problem:

The intertemporal budget constraint for this individual can be summarized as follows:

where $\tilde{r} = r-\delta$ . From the intertemporal budget constraint, the lifetime disposable income matters for allocations across the life cycle. The lifetime disposable income is the subtraction of the present value of income and educational costs. Since $f^A > f^H$ , the wage of creative workers should be greater than the wage of experts, which requires $e^A m_{t+1}> e^H w_{t+1}^G$ , to ensure the existence of equilibrium. Let me denote $z_{1,t} = \frac{(1+\tilde{r}_{t+1})(f^A-f^H)}{e^Am_{t+1}-e^H w^G_t}$ , $z_{2,t} = \frac{w^G_t + (1+\tilde{r}_{t+1})f^A}{e^Am_{t+1}}$ , $z_{3,t} = \frac{w^G_t +(1+\tilde{r}_{t+1})f^H}{e^H w^G_t}$ .

Proposition 1

(Occupational & Educational Choices) If $z_{1,t}>z_{2,t}$ , agents with $z^i \in \left( \ln (z_{1,t}), 1 \right]$ invests in creativity and become creative workers; agents with $z^i \in \left[ \ln (z_{3,t}), \ln (z_{1,t})\right]$ are experts with training; and agents with $z^i\in \left[ 0, \ln (z_{3,t}) \right)$ receive no education and become basic workers. If $z_{1,t} \le z_{2,t}$ , agents with $z^i \in \left( \ln (z_{2,t}), 1 \right]$ receive training in creativity and become creative worker; and agents with $z^i \in \left[ 0, \ln (z_{2,t}) \right]$ receive no education and become basic workers.

See Appendix A.1. $\square$

Thus, if $z_{1,t}>z_{2,t}$ , $\bar{z}_t = \ln (z_{1,t})$ and $\underline{z}_t = \ln (z_{3,t})$ (case 1). Otherwise, $\bar{z}_t =\underline{z}_t = \ln (z_{2,t})$ (case 2). This proposition posts constraints in parameter values, which I use in the calibration section. For the individual i born in generation t , the demands of two types of goods in each period satisfy,

The supply of capital for the individual when young is

The supply of capital for the middle-aged individual is

2.3 Markets clear conditions

This economy has two labor markets, two goods markets, and one capital market. I denote $G_{t}$ as the demand of basic workers, $H_{t}$ as the demand of experts by the firm produces “non-automatable” goods at time t , and $E_{t}$ as the demand of creative workers by the firm produces “automatable” goods at time t . $C_{y,t}, C_{m,t}, C_{o,t}, S_{y,t}, S_{m,t}, S_{o,t}$ as total consumption of “automatable” goods and “non-automatable” by the young, middle-aged and old individuals at time t respectively. I denote $B_{m,t}, B_{y,t}$ as the supply of the aggregate capital from young and middle-aged workers in period t , respectively. The market equilibrium conditions can be shown as follows.

Labor Markets.

Goods Markets.

Capital Market.

2.4 Equilibrium

A sequential markets equilibrium consists of allocations $\{ c^{i}_{m,1},s^{i}_{m,1}, c^{i}_{o,1}, s^{i}_{o,1}, c^{i}_{o,2}, s^{i}_{o,2}$ , $b^{i}_{y,1}, b^{i}_{m,1}, b^{i}_{m,2},\{c^{i}_{y,t},s^{i}_{y,t}, c^{i}_{m,t+1},s^{i}_{m,t+1}, c^{i}_{o,t+2}, s^{i}_{o,t+2}, f^{i}_t, b^{i}_{y,t+1}, b^{i}_{m,t+2}\}_{i}$ , $K_t, A_t, Y_t, S_t, G_t, H_t, E_t\} _{t=1}^{\infty}$ and prices $\left\{ w^G_t,m_t,q_t,r_t\right\} _{t=1}^\infty$ , such that:

Given $\{w^G_t,m_t,q_t,r_t\}_{t=1}^\infty$ , each generation determines optimal consumption bundles and capital investments across the life cycle, educational and occupational choices as mentioned in Sect. 2.2.1 .

Given $\{w^G_t,m_t,q_t,r_t\}_{t=1}^\infty$ , two representative firms determine labor demand and capital demand, as mentioned in Sect. 2.1.1 .

Goods markets, labor markets, and capital markets are clear in each period as mentioned in Sect. 2.3 .

In this paper, I focus on the analysis of the steady-state equilibrium. The steady-state equilibrium conditions are summarized in Appendix A.2.

3 Simulation

3.1 calibration strategy.

To conduct numerical analyses, I begin by selecting a model parameterization as detailed below. The model’s steady-state equilibrium is calibrated to the U.S. average statistics from 2002 to 2020. In this model, workers live for three periods, equating each to 25 years to align with real-life data.

There are three types of occupations in the model, and I characterize detailed occupations in Standard Occupational Classification (SOC) into three categories manually based on their job requirements. Basic workers perform simple, repetitive tasks without extensive training, including roles like hand makers, attendants, and helpers. Experts, requiring practical training, manage and adapt to job responsibilities through extensive experience or professional knowledge, often bringing some levels of innovation to their roles; this category includes occupations such as managers, administrators, and superintendents. Creative workers engage in roles that demand a solid professional base and creativity to explore frontier technologies and knowledge, such as directors, artists, and scientists. The appendix A.4 details these categories and lists specific job titles within each.

I obtain employment and mean hourly wage data for each occupation from the U.S. Occupational Employment and Wage Statistics (OEWS) program. By calculating the average employment share and hourly wage for each occupational category, it is found that 18% of workers are basic workers, 70% are experts, and 12% are creative workers ( $G = 0.18, H = 0.70$ , $E = 0.12$ ), with basic workers earning an average hourly wage of $13.29 ( $w^G = 13.29$ ). Additionally, I compute the wage premium of experts over basic workers and the wage premium of creative workers over experts, finding the latter exceeds one, indicating $e^A m> e^H w^G\ \text {and}\ f^A> f^H$ . These wage premiums, however, are not used for calibration but to check the model’s fit.

Additionally, I derive total labor income from the wages and salaries reported in the National Income and Product Accounts (NIPA) by the U.S. Bureau of Economic Analysis (BEA). Footnote 2 In the model, I assume the middle-aged population supplies the total time endowment for work, and the overall population in each generation is normalized to 1 in the economy. Consequently, the income in the model is interpreted as per-hour income. Thus, I divide the total wages and salaries by the total hours worked each year. The total hours worked data is obtained from the Bureau of Economic Analysis (BEA). I find the total labor income is $35.34 ( $TotLI = 35.34)$ . Utilizing occupational employment shares, I then calculate the learning ability threshold values, $\underline{z} = G = 0.18$ , and $\bar{z} = 1-E = 0.88$ .

I categorize parameters into two groups: fixed and free. Fixed parameters are derived directly from data or are widely accepted in the literature. Free parameters are determined iteratively by substituting data values into theoretical moments. Table 1 lists all parameters in the model and calibrated values.

Fixed Parameter. I set the discount factor to $\beta = 0.99^{25} = 0.778$ , following precedents in literature such as Hurd ( 1990 ) and Ludwig and Vogel ( 2010 ). Considering the annual capital depreciation rate of 0.05, as cited by Kulish et al. ( 2010 ) and equating one period in the model to 25 years, I adjust the depreciation rate to $\delta = 1-(1-0.05)^{25} = 0.723$ . Furthermore, I normalize the total factor productivity for “non-automatable” goods D S to one.

Free Parameter. Due to the indistinguishability of individuals’ outcomes from the costs of two types of education ( $f^A$ and $f^H$ ) and the productivity enhancer difference between them ( $e^H$ and $e^A$ ), I equate $e^H$ to $e^A$ . Using the obtained data and equilibrium conditions, I can calibrate values of all parameters as listed in Table 1 . I first assume $e^H$ is known, and the calibration process unfolds through three nested loops to determine the capital income ( rK ), the total education spending ( $TotEdu= f^HH+f^AE$ ), and the net interest rate ( $\tilde{r}$ ). The methodology and rationale for solving these three variables are discussed in the subsequent paragraphs.

To calibrate $\alpha$ , which signifies the capital income share in total “automatable” goods production, I compute $\alpha = \frac{rK}{Y} = \frac{rK}{rK+mA}$ , with $mA = TotLI-qS$ . Here, TotLI represents the total labor income, and qS shows the income from “non-automatable” goods, equating to the labor income of both basic workers and experts. For a given level of $e^H$ , I can compute $qS = w^GG+e^Hw^G(e^{\bar{z}}- e^{\underline{z}})$ by using available data. Therefore, by determining rK , the calibrated value of $\alpha$ can be obtained.

Then, to calibrate the set $\{\kappa , f^H, f^A, D_Y\}$ , the focus narrows to solving for total education spending ( $TotEdu= f^HH+f^AE$ ), and net interest rate ( $\tilde{r}$ ). $\kappa$ represents the proportion of consumption expenditure on “non-automatable” goods relative to total consumption, calculated as $\kappa = \frac{qS}{qS+TotC}$ . Determining $\kappa$ requires the total consumption on “automatable” goods TotC . For a given level of $\{e^H, rK, \tilde{r}, TotEdu\}$ , I obtain $K = \frac{rK}{r} = \frac{rK}{\tilde{r}+\delta }$ . From the “automatable” goods resource constraint, the total number of consumption on the “automatable” goods is $TotC = Y - \delta K - TotEdu$ , with $Y = mA + rK$ reflecting the competitive market condition.

Thus, I get the calibrated value of $\kappa = \frac{qS}{qS+TotC}$ . Moreover, I compute the cost of practical training $f^H = \frac{(e^H e^{\underline{z}}-1)w^G}{1+\tilde{r}}$ from the basic worker labor market clear condition (Eqn. (A6)), and the cost of receiving training in creativity skill $f^A = \frac{TotEdu - f^H H}{E}$ from the definition of total spending on education in the economy. Lastly, the total factor productivity in “automatable” goods production, $D_Y$ , is determined as $D_Y = \frac{Y}{K^\alpha A^{1-\alpha }}$ , where $A = e^H(e-e^{\bar{z}})$ .

The following process illustrates how I solve these three key variables through a structured three-layer loop process for a given level of $e^H$ :

Outer Loop: This step focuses on determining capital income ( rK ) by satisfying the capital market equilibrium condition (Eqn. (A10));

Middle Loop: The objective here is to find the total education spending ( TotEdu ) using the labor market clear condition for creative workers (Eqn. (A8));

Inner Loop: This loop involves calculating the net interest rate ( $\tilde{r}$ ) to satisfy the “automatable” goods market clear condition (Eqn. (A9)).

I calibrate the value of $e^H = 1.620$ to ensure the existence of an equilibrium where $e^Am > e^Hw^G$ , $f^A > f^H$ , and the capital income share $\alpha$ are all consistent with the data. Footnote 3

Model Fit. I compute the effective hourly wage from the model for different types of workers and compare it with those from OEWS hourly wage information, which are the untargeted data moments in the calibration. All wages are deflated using 2012 PCE. Table 2 presents the wage premium of the other two types of workers to experts from model and data. From the table, the model generates similar relative wage premiums of basic and creative workers to experts as in the data. Additionally, the model’s calibrated ratio of total educational spending to GDP is approximately 7.11%, aligning closely with the OECD’s reported educational spending to GDP ratio in the U.S. of about 6.1%. Footnote 4 Thus, I conclude that the model fits the data sufficiently.

3.2 Simulated income inequality

In this paper, I focus on the effect of changing educational costs on long-term income inequality. The total income consists of labor and capital income, and I rule out young workers in the income inequality analysis since they have zero income. In the model, only middle-aged workers can work and gain labor income, and the total income of the old is capital income only. This paper investigates two types of education, illustrating how identical cost reductions may have distinct effects on income inequality measurements. I explore the consequences of the income Gini index and the income decile ratio in subsequent sections.

Scenario 1: Decrease $f^H$ when $f^A$ is fixed

In the first scenario, suppose the government decides to only reduce the cost of receiving expert training range from $[f^H_{ss}-0.2, f^H_{ss}]$ , where $f^H_{ss}$ is the calibrated steady-state cost of practical training. The simulated pattern of the steady-state Gini index for income is shown in Fig. 1 . The solid line in Fig. 1 a represents the total income inequality in the economy with both middle-aged and old workers. The line in Fig. 1 b shows the income inequality among middle-aged workers or among old workers. From a theoretical perspective, the Gini coefficient of income distribution is expected to be equivalent between middle-aged and elderly workers in this model, given that the income of middle-aged workers is several times greater than that of elderly workers. Thus, two within inequality curves are the same (See Appendix A.3 for proof).

Figure 1 shows that reducing the cost of practical training leads to a decrease in both within-group inequality and overall income inequality. Total inequality surpasses within-group inequality due to the existing income dispersion between middle-aged and elderly workers. However, the decline in the cost of practical training does not seem to impact between-group inequality.

Change in income Gini index if $f^H$ declines

Simulated steady state wages

Simulated steady state occupational distribution

To analyze the shifts in within-group inequality associated with changes in $f^H$ , I examine wage and occupational distribution patterns, illustrated in Figs. 2 and 3 . A relative decline in practical training costs draws individuals with endowed learning abilities across both thresholds towards experts, reducing the shares of basic and creative workers, assuming fixed prices. However, an increasing wage gap between creative and basic workers incentivizes a shift towards creative workers. Consequently, the proportion of basic workers decreases, while the numbers of experts and creative workers rise. Thus, despite the growing wage disparity, the composition effect leads to reduced income Gini index with further decreases in the cost of practical training.

The findings suggest that reducing the cost of practical training could lead to higher income inequality when measured by the income decile ratio. This rise is primarily attributed to the widening wage gap between creative and basic workers, which amplifies within-group income inequality among middle-aged workers. Figure 4 captures the variations in overall (Fig. 4 a) and within-group (Fig. 4 b) income decile ratios as a result of decreasing costs for practical training.

Change in income declice ratio if $f^H$ declines

Scenario 2: Decrease $f^A$ when $f^H$ is fixed

In this scenario, I simulate the long-term consequences when the government decides to reduce the cost of creativity education from $[f^A_{ss}-0.2,f^A_{ss}]$ , where $f^A_{ss}$ is the calibrated steady-state cost of creativity training. Figure 5 depicts the resulting change in the steady-state income Gini index. It demonstrates that both the overall income Gini across middle-aged and elderly workers and the within-group inequality decrease as the cost of investing in creativity is reduced. Similar to the first scenario, the reduction in educational costs on creativity training does not appear to affect the between-group income Gini.

Change in income Gini index if $f^A$ declines

Similar to the analysis in the first scenario, I present the wage and occupational distribution patterns in Figs. 6 and 7 . According to Figs. 6 and 7 , although the wages of basic workers and the price of intermediate goods both increase, the differences in wage gaps decline when the cost of investing in creativity declines. Moreover, the decline in the cost of investing in creativity results in fewer basic workers but more experts and creative workers. Thus, the within-age group inequality decreases because both wage gaps decrease and shares of low-wage earners (basic workers) decline. The finding also implies that altering the income inequality measure to the income decile ratio could also lead to a decline in income inequality. Figure 8 captures the variations in the overall income decile ratio as a result of decreasing costs for creativity skill training.

Change in income declie ratio if $f^A$ declines

4 Conclusion

This paper shows how lowering costs for different types of education can affect income inequality in various ways. Reducing costs for practical training, which focuses on repetitive tasks, may actually make income inequality worse. I construct a three-period overlapping generation model without considering family utility function but incorporating two types of education for workers in different roles. Individuals are in the process of receiving training in their youth, actively producing different types of goods in middle age, and retiring in old age.

The change in income inequality among middle-aged workers mainly determines the variations in total income inequality when the cost of education declines. When practical training costs decline, more middle-aged workers aim to become experts, positioning themselves in the middle-income range. This shift reduces the number of basic and creative workers, driving their wages up and widening the wage gap. This wage gap increase makes creative jobs more appealing, leading to a decrease in basic workers and an increase in experts and creative workers. This scenario suggests that changes in workforce composition (composition effect) are more influential than changes in wage differences (wage effect) in lowering the income Gini index as the cost of practical training decreases. However, if the widening pay gap (wage effect) takes precedence, income inequality could increase. This becomes apparent when measuring inequality differently, such as with the income decile ratio.

A reduction in the cost of creative training not only attracts experts to become creative workers but also attracts more workers to receive practical training. Hence, both the shares of higher-income and middle-income groups increase, and the wage gap between workers at both ends of the income distribution narrows. Therefore, both wage and composition effects explain the declining income inequality.

In this paper, I assume the type of education cost is a sunk resource and is determined exogenously. Suppose the policy goal is to reduce income inequality. In that case, the government should focus on reducing the cost of creative training or enhancing teaching effectiveness on creative training. For instance, the government can develop an institution to train more educators to provide creative training or encourage teaching innovation to enhance teaching effectiveness.

There are several limitations of the paper. Firstly, I assume that two types of education training exist: practical training and creative training. In the real world, some of the existing education training would involve both of them. As a result, it is common for individuals to obtain both creative training and practical training at the same time. However, there are still some educational programs that would focus on either one of them. For example, the coding program will focus on logical training, which can enhance workers’ ability to produce “automated” goods. Secondly, I assume individuals have perfect information for themselves and can foresee the future. Hence, they can decide on the type of education they want without any uncertainties. In the real world, individuals could have incentives to receive both creative training and practical training in order to diversify their human capital investments.

Data availability

All used data are public.

Refer to Appendix A.4 for additional details on these occupational types.

The data are collected from Table 2.1. Personal Income and Its Disposition, with wages and salaries deflated by PCE in 2012.

The calibrated value of $e^H$ ranges from 1.429 to 1.620 to guarantee an equilibrium where $e^Am > e^Hw^G$ , and $f^A > f^H$ . When $e^H=1.429$ , the capital income share is almost 0. I pick $e^H=1.60$ since the calibrated $\alpha$ is closer to 0.3.

This total educational spending includes expenses from primary through tertiary education.

Abdullah A, Doucouliagos H, Manning E (2015) Does education reduce income inequality? A meta-regression analysis. J Econ Surv 29(2):301–316

Article Google Scholar

Autor DH, Dorn D (2013) The growth of low-skill service jobs and the polarization of the US labor market. Am Econ Rev 103(5):1553–1597

Battistón D, García-Domench C, Gasparini L (2014) Could an increase in education raise income inequality? Evidence for Latin America. Latin Am J Econ 51(1):1–39

Becker GS, Tomes N (1979) An equilibrium theory of the distribution of income and intergenerational mobility. J Polit Econ 87(6):1153–1189

Coady D, Dizioli A (2018) Income inequality and education revisited: persistence, endogeneity and heterogeneity. Appl Econ 50(25):2747–2761

Deming D, Kahn LB (2018) Skill requirements across firms and labor markets: evidence from job postings for professionals. J Law Econ 36(S1):S337–S369

Google Scholar

Deming DJ (2017) The growing importance of social skills in the labor market. Q J Econ 132(4):1593–1640

Gregorio JD, Lee J-W (2002) Education and income inequality: new evidence from cross-country data. Rev Income Wealth 48(3):395–416

Hendel I, Shapiro J, Willen P (2005) Educational opportunity and income inequality. J Public Econ 89(5–6):841–870

Hurd MD (1990) Research on the elderly: economic status, retirement, and consumption and saving. J Econ Lit 28(2):565–637

Knight JB, Sabot RH (1983) Educational expansion and the Kuznets effect. Am Econ Rev 73(5):1132–1136

Kulish M, Kent C, Smith K (2010) Aging, retirement, and savings: a general equilibrium analysis. BE J Macroecon 10(1)

Ludwig A, Vogel E (2010) Mortality, fertility, education and capital accumulation in a simple olg economy. J Popul Econ 23:703–735

Mincer JA (1974) Schooling and earnings. In: Schooling, experience, and earnings, pp 41–63. NBER

Prettner K, Schaefer A (2021) The u-shape of income inequality over the 20th century: the role of education. Scand J Econ 123(2):645–675

Ram R (1989) Can educational expansion reduce income inequality in less-developed countries? Econ Educ Rev 8(2):185–195

Rodríguez-Pose A, Tselios V (2009) Education and income inequality in the regions of the European union. J Reg Sci 49(3):411–437

Schultz TW (1963) The economic value of education. Columbia University Press, New York

Spence M (1973) Job market signaling. Q J Econ 87(3):355–374

Sylwester K (2002) Can education expenditures reduce income inequality? Econ Educ Rev 21(1):43–52

Yang J, Qiu M (2016) The impact of education on income inequality and intergenerational mobility. China Econ Rev 37:110–125

Download references

Acknowledgements

I gratefully acknowledge funding from Wenzhou Kean University, IRSP (Grant IRSPG202107). I thank Yang Xuan for the helpful discussions on earlier versions of this paper and her research assistant in calibration.

This research is funded by Wenzhou Kean University, IRSP (Grant IRSPG202107).

Author information

Authors and affiliations.

Department of Economics, Wenzhou Kean University, 88 Daxue Road, Wenzhou, 325060, Zhejiang, China

Fa-Hsiang Chang

You can also search for this author in PubMed Google Scholar

Contributions

The work is purely done by the author, and all faults are mine.

Corresponding author

Correspondence to Fa-Hsiang Chang .

Ethics declarations

Conflict of interest.

The author has no relevant financial or non-financial interests to disclose.

Ethics approval

Not applicable.

Consent to participate

Consent for publication, code availability.

The codes can be applied upon request.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file 1 (pdf 450 KB)

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Chang, FH. The impact of education costs on income inequality. Int Rev Econ 71 , 553–574 (2024). https://doi.org/10.1007/s12232-024-00452-z

Download citation

Received : 01 September 2023

Accepted : 22 February 2024

Published : 03 April 2024

Issue Date : September 2024

DOI : https://doi.org/10.1007/s12232-024-00452-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Occupational choice

JEL Classification

Find a journal
Publish with us
Track your research

Education and inequality in 2021: how to change the system

Research Associate at the University of Geneva's department of Education and Psychology; Campus and Secondary Principal at the International School of Geneva's La Grande Boissière, Université de Genève

Disclosure statement

Conrad Hughes does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

AUF (Agence Universitaire de la Francophonie) provides funding as a member of The Conversation FR.

View all partners

Students sitting on their chairs and desks taking a test outside.

Since its earliest traces, at least 5,000 years ago , formal education – meaning an education centred on literacy and numeracy – has always been highly selective. Ancient Egyptian priest schools and schools for scribes in Sumeria were only open to the children of the clergy or future monarchs.

Later on, the wealthy would use private tutors, such as the Sophists of Athens (500 - 400 BCE). Ancient Greek schools, such as Plato’s Academy and Aristotle’s Lyceum , were restricted to a small elite group. Formal education was reserved for male children who were wealthy, able, and privileged.

Through time, even after learning societies began to flourish, it was still an education for some and not for everybody.

In the 1800s Black people were denied access to quality education in the United States. In European colonies, education was used to strip people of their cultural heritage and relegate them to a future of menial labour.

Education has always been less accessible to women than men. Even today, over 130 million girls are still out of school. Although the difference between girls and boys is lessening, the disparity disadvantaging girls persists . From a socioeconomic perspective, in many countries, private schools continue to grow alongside compulsory state schools, offering a different style of education, sometimes at a very high price.

Today, progress to attain the dream of universal access to education is slow. UNESCO’s Education for All and the UN’s Sustainable Development Goal 4 , which aims to “ensure inclusive and equitable quality education and promote lifelong learning opportunities for all”, are still far from materialising: roughly 260 million children are still not in school . The COVID-19 pandemic has made the situation worse: remote learning is inaccessible to roughly 500 million students . Estimates are that over 200 million children will still be out of school by 2030 .

In my study “Education and Elitism” , the overarching question that runs through the book is about the future of education worldwide: What are the prospects for the future? Are we facing an even more enclaved, pauperised majority while a tiny minority become more powerful and wealthy?

Certain paths could open up. On the one hand, places in selective institutions could become even more difficult to access while private education strips ahead of national standards. On the other hand, changes might make education more inclusive: this would include scholarships, cheaper private education, more robust state systems and deep assessment reform.

Prospects for the future

Scholarship programmes: These allow the brightest and poorest access to transformative learning ecosystems . However, this contributes to a brain drain and does not develop the local educational sector , particularly in Africa.

Cheaper private education: A movement of accessible private schools is growing . This allows more children to access some of the value-added features of such systems – more curriculum flexibility, smaller class sizes, more individual student tracking. However, there are reports that this is widening social divides , as the public system isn’t improving fast enough to keep up.

More robust state systems: UNESCO estimates that it would cost a total of US$340 billion each year to achieve universal pre-primary, primary, and secondary education in low- and lower-middle-income countries by 2030. The average annual per-student spending for quality primary education in a low-income country is predicted to be US$197 in 2030. This creates an estimated annual gap of US$39 billion between 2015 and 2030. Financing this gap calls for action from private sector donors, philanthropists, and international financial institutions.

Online learning: The COVID-19 lockdown has brought inequalities to the surface. However, the rise of online learning worldwide has been phenomenal. This opens up the potential to widen access to learning socioeconomically and, if delivered by skilled facilitators, academically . There is a problem, though: online instruction lacks the emotional quantum that face-to-face learning creates. Because of this, motivation levels and persistence tend to be low in online learning environments . And importantly, in many countries, many students still don’t have access to the internet.

A way forward: reforming the system

Perhaps the most substantive movement to reduce inequalities would not be to accelerate access to a broken system but to reform the system itself .

It is time to look further than narrow academic metrics as the only way of describing young people’s competences. The whole educational system across high schools, in every country, needs to change dramatically. Assessment models should recognise and nurture more varied and multiple competences, in particular, attitudes, skills and types of knowledge beyond those concentrated in constructs that are favoured by socioeconomic background, such as literacy and numeracy .

Read more: Education needs a refocus so that all learners reach their full potential

Until universities and employers look beyond traditional metrics, it will be difficult to break a circuit that favours, for the large part, middle class, socially and ethnically privileged candidates.

To truly break away from a millennia of elitist, selective systems , the approach needs to move from pure academics to a credit system that captures many more stories of learning. This new credit system should be known as a passport, meaning students have stamped it with the various competences such as lifelong learning and self-agency that they have developed throughout their learning (in an out of school), allowing them to be recognised on numerous different fronts.

A coalition of schools from every continent is working on this project, now seeking universities to sit around the table in order to bring this work to its conclusion. This would mean co-designing an elegant, life worthy transcript to allow more access to more children based on more expansive criteria.

Peacebuilding
Education inequality
Women and girls
Out of school children

Director of STEM

Community member - Training Delivery and Development Committee (Volunteer part-time)

Chief Executive Officer

Finance Business Partner

Head of Evidence to Action

Report | Children

Education inequalities at the school starting gate : Gaps, trends, and strategies to address them

Report • By Emma García and Elaine Weiss • September 27, 2017

Download PDF

Press release

Share this page:

This report was produced in collaboration with the Broader, Bolder Approach to Education .

What this study finds: Extensive research has conclusively demonstrated that children’s social class is one of the most significant predictors—if not the single most significant predictor—of their educational success. Moreover, it is increasingly apparent that performance gaps by social class take root in the earliest years of children’s lives and fail to narrow in the years that follow. That is, children who start behind stay behind—they are rarely able to make up the lost ground.

Using data from two academic cohorts, the kindergarten classes of 1998 and 2010, this study examines the relationship between children’s socioeconomic status (SES) and their cognitive and noncognitive skills when starting school. We find that large performance gaps exist between children in the lowest and highest socioeconomic-status (SES) quintiles and that these gaps have persisted from the 1998 cohort to the 2010 cohort. The positive news is that the gaps have not grown, even as economic inequalities between these two groups of students have grown. The negative news is that the gaps have not narrowed, despite the fact that low-SES parents have substantially increased their engagement in their children’s early education.

Why it matters: These performance gaps reflect extensive unmet needs and thus untapped talents among low-SES children. The development of strong cognitive and noncognitive skills is essential for success in school and beyond. Low educational achievement leads to lowered economic prospects later in life, perpetuating a lack of social mobility across generations. It is also a loss to society when children’s talents are allowed to go fallow for lack of sufficient supports. The undeniable relationship between economic inequalities and education inequalities represents a societal failure that betrays the ideal of the “American dream.”

What can be done about it: Greater investments in pre-K programs can narrow the gaps between students at the start of school. And to ensure that these early gains are maintained, districts can provide continued comprehensive academic, health, nutrition, and emotional support for children through their academic years, including meaningful engagement of parents and communities. Such strategies have been successfully implemented in districts around the country, as described in this report, and can serve to mitigate the impact of economic inequalities on children’s educational achievement and improve their future life and work prospects.

For further discussion of policy solutions, see the companion to this report, Reducing and Averting Achievement Gaps: Key Findings from the Report ‘Education Inequalities at the School Starting Gate’ and Comprehensive Strategies to Mitigate Early Skills Gaps .

Executive summary

High and rising inequality is one of the United States’ most pressing economic and societal issues. Since the early 1980s, the total share of income claimed by the bottom 90 percent of Americans has steadily decreased, with the majority of income gains going to the top 1 percent. These trends would not be such a major concern if our education system compensated for these inequities by helping level the playing field and enabling children to rise above their birth circumstances.

But that is hardly the case. Rather, the fraction of children who earn more than their parents (absolute mobility) has fallen from approximately 90 percent for children born in 1940 to 50 percent for children born in the 1980s. And the tight links between economic inequalities and achievement gaps cast doubt on asserted equality of opportunity that promotes social mobility and puts the “American Dream” within viable reach.

Extensive research has conclusively demonstrated that children’s social class is one of the most significant predictors—if not the single most significant predictor—of their educational success. Moreover, it is increasingly apparent that performance gaps by social class take root in the earliest years of children’s lives and fail to narrow in the years that follow.

Much is known about the determinants and mechanisms that drive early skills gaps among children of different backgrounds, but our failure to narrow social-class-based skills gaps from one generation of students to the next calls for further analysis to determine the degree of influence these factors have and how interventions employed in recent years to address these factors have or have not worked and why. Moreover, shifting economic and demographic landscapes emphasize the need for more robust policy strategies to address the gaps. This three-part study thus combines a statistical analysis of early skills gaps among a recent cohort of children and changes in them over time with a qualitative study of multifaceted, school-district-level strategies to narrow them.

What we do: Questions, data and methodology

In this paper, we:

Use data from the National Center for Education Statistics (NCES): the Early Childhood Longitudinal Study of the Kindergarten Classes of 1998–1999 and 2010–2011 to measure gaps in skills by social class. To measure gaps by social class, we use the socioeconomic status (SES) metric (primarily), a composite of information on parents’ educational attainment and job status as well as household income. We compare the average performance of children in the top fifth of the socioeconomic status distribution (high-SES) with the average performance of children in the bottom fifth (low-SES). Skills measured include reading and mathematics, as well as self-control and approaches to learning as reported by both teachers and parents.
Examine SES-based gaps at kindergarten entry among the most recently surveyed cohort (the kindergarten class of 2010–2011). We study how gaps manifest in both cognitive and so-called noncognitive skills, as both skill types are important components of children’s development.
Compare these SES gaps with those of an earlier cohort (1998–1999), with a focus on changes in the skills gaps between children in the high- and low-SES quintiles. We also analyze how sensitive gaps are to the inclusion of key determinants of student performance, such as family composition, children’s own characteristics, pre-K participation, and parental and educational practices at home.
Review a set of 12 case studies of communities that have employed comprehensive educational strategies and wraparound supports to provide more children (especially low-income children) with strong early academic foundations, and to sustain and build on early gains throughout their K–12 school years.
Based on examples from these diverse communities, we discuss implications: strategies that districts can employ and district and state policy changes to make those strategies easier to adopt and more sustainable. The report ends with conclusions and recommendations for further research, practice, and policy.

What we find

Our quantitative research produces a broad set of findings:

Very large SES-based gaps in academic performance exist and have persisted across the two most recent cohorts of students when they start kindergarten. The estimated gaps between children in the highest and lowest fifths of the SES distribution are over a standard deviation (sd) in both reading and math in 2010 (unadjusted performance gaps are 1.2 and 1.3 sd respectively). Gaps in noncognitive skills such as self-control and approaches to learning are roughly between one-third and one-half as large (unadjusted performance gaps are about 0.4 sd in self-control, and slightly over 0.5 sd in approaches to learning in 2010).
SES-based gaps across both types of skills among the 2010 kindergartners are virtually unchanged compared with the prior academic generation of students (the class of 1998). The only unadjusted cognitive skills gap between children in the high-SES and low-SES fifths that changed significantly over this period was the gap in reading skills, which increased by about a tenth of a standard deviation. Gaps in approaches to learning as reported by teachers and in self-control as reported by parents shrank between 1998 and 2010 by roughly the same amount (0.1 sd). Gaps in mathematics, in approaches to learning as reported by parents, and in self-control as reported by teachers did not change significantly.
This means that though part of the SES gap is attributable to differences in these characteristics and in family investments between children in the high and low parts of the SES distribution, a substantial share of SES-related factors is not captured by these controls, but is important to explaining how and why gaps develop, and thus how to narrow them.
Moreover, the capacity for these other factors to narrow gaps has decreased over time—as a whole, they accounted for a smaller share of the gaps in 2010 than they had in 1998. This suggests that, while such activities as parental time spent with children and center-based pre-K programs cushion the negative consequences of growing up in a low-SES household, they can do only so much, and that the consequences of poverty are increasingly hard to compensate for. This resistance of gaps to these controls is thus a matter of serious concern for researchers and policymakers alike.
These children’s likelihood of attending center-based pre-K did not change significantly across generations (about 44 percent for both cohorts: 44.3 percent in 2010 vs. 43.7 percent in 1998). However, in 2010 their parents reported having a somewhat larger number of books at home for the children, and there was also an increase in both indices of activities (literacy/reading activities and other educational and engagement activities).
In addition to doing more for their children, low-SES parents have greater expectations for their children’s educational attainment—a much smaller share saw them going no further than high school graduation, while a much greater share anticipated their children attaining bachelor’s and even advanced degrees in 2010.
They were slightly more likely to live with two parents (the share not living with two parents decreased from 11.1 percent in 1998 to 9.6 percent) and to have attended center-based pre-K (the share in center-based pre-K increased from 65.8 in 1998 to 69.9 percent in 2010).
The share of high-SES homes reporting having more than 200 children’s books slightly increased in 2010, as did parents’ expectations for their children’s educational attainment.
Although research uses various indicators to measure individuals’ social class, from composite measures such as the socioeconomic status index we use to single indicators such as mother’s education or income, some sensitivity of the results to the indicator used is found. In our analyses, we find that all are equally reliable social-class proxies for the estimation of early achievement gaps, though absolute gaps and trends in them vary slightly depending on the indicator used.

Our qualitative review of community interventions also provides valuable information:

A growing number of school districts across the country have embraced systems of comprehensive enrichment and supports for many or even all their students, based on the understanding that nurturing healthy child development requires leveraging the entire community. These districts took different approaches to enacting those comprehensive strategies, based on each community’s particular mix of needs and assets, ideological leaning, available sources of funding, and other factors. But all begin very early in children’s lives and align enriching school strategies with a targeted range of supports for children and their families.
Moreover, school districts embracing what we refer to as “whole-child” approaches to education are seeing better outcomes for students, from improved readiness for kindergarten to higher test scores and graduation rates and narrower achievement gaps. They thus can provide guidance to other districts and to policymakers regarding how to implement such approaches, what to expect in terms of benefits, and which policies at the local and state levels can advance those approaches.

Conclusions

While the persistence of large skills gaps at kindergarten entry is troubling, the fact that, by and large, they did not grow in a generation—despite steadily increasing income inequality compounded by the worst economic crisis in many decades—is a good thing. But we must still be very concerned about these gaps. We would have liked to see evidence that parents’ increased dedication to and investments in their children’s early development, and increased investments in pre-K programs and other early education and economic supports, closed these gaps. However, the data suggest that these efforts simply contained them, and that these positive trends were insufficient to narrow the skills gaps at kindergarten entry. This failure to narrow gaps points to a lack of appropriate policy response at all levels of government, the neglect of decades of research across multiple disciplines on child development, and the resulting waste of critical opportunities to nurture an entire generation of children.

The policy recommendations of this report strengthen the idea that we need much greater investments in pre-K programs and continued comprehensive support for children through their academic years, including meaningful engagement of parents and communities, if we are to substantially improve the odds for disadvantaged children, in light of their extensive unmet needs and untapped talents.

Introduction: Facts about income inequality and its growth over time

One of today’s most pressing economic issues is the worrisome level of income inequality. Since 1979, the total share of income claimed by the bottom 90 percent of Americans has steadily decreased (Bivens 2016). In 1979, that 90 percent received about 67 percent of cash, market-based income (i.e., pretax income). By 2015, their share had decreased to about 52 percent of pretax income. The majority of income gains during this period went to the top 1 percent (EPI 2013; Mishel and Schieder 2016; Saez 2016). Polls reflect widespread concern about income and wage inequalities and associated trends and the desire for policies to address these inequalities ( New York Times 2015).

Rising inequality might not be such a major concern if our education, economic, and social protection systems acted as compensatory mechanisms, helping individuals, and especially children, rise above their birth circumstances and improve their mobility. But that is hardly the case. Rather, the fraction of children who earn more than their parents (a measure of what social scientists refer to as absolute mobility) has fallen from approximately 90 percent for children born in 1940 to 50 percent for children born in the 1980s (Chetty et al. 2016). Children of certain ethnic and racial minorities who are disproportionately likely to live in concentrated poverty are also more likely to do so over prolonged periods of time (Sharkey 2013). And the close connections between education inequalities and economic inequalities cast doubt on assertions that America provides “equality of opportunities” that promotes social mobility (Mishel 2015).

The influence of income inequality affects multiple aspects of society’s functioning, from health outcomes and even life expectancy to democratic ideals (Putnam 2015; Schanzenbach et al. 2016; Stringhini et al. 2017). In the education arena, children’s socioeconomic status (SES), of which income is a key component, is considered one of the most significant predictors—if not the most significant predictor—of educational success. A number of studies show the strong relationship between social class (of which socioeconomic status is a frequent measure) and test scores, educational attainment, and college attendance and completion (see Duncan, Morris, and Rodrigues 2011; García 2015; García and Weiss 2015; Lee and Burkam 2002; Mishel et al. 2012; Putnam 2015; among others).

As a result of these trends and associations, achievement gaps by social class have grown substantially since the 1960s, especially between children at the highest end of the income distribution and all of the others (Reardon 2011). Some researchers have identified a large increase in parental investment in education among high-SES parents as one driver of the divergence in education outcomes (Duncan and Murnane 2011), among other contributing factors, such as time parents spend with their children and time parents devote to education-enhancing activities (Morsy and Rothstein 2015; Van Voorhis et al. 2013): Spending on education-enhancing activities by parents in the top income fifth nearly tripled between the 1970s and the 2000s (from $3,500 in 1972 to $8,900 in 2006), while such spending by parents in the bottom income fifth remained low and changed much less (from $800 in 1972 to $1,300 in 2006) (Duncan and Murnane 2011). 1 More time can mean more frequent interactions during playtime, more time spent reading to children, and other parenting practices that contribute to children’s learning and development (Barbarin et al. 2010). In general, more leisure and educational time with children can promote their development and school readiness (Brooks-Gunn and Markman 2005; Hart and Risley 1995; Phillips 2011; Rothstein 2004; Van Voorhis et al. 2013; Waldfogel 2006). Given the evidence that parental engagement and spending directly and continuously translate into improvements in children’s achievement and preparation, the presence of the various achievement gaps are not surprising.

Education researchers and policymakers have long been attentive to issues related to equity—by race/ethnicity, SES, gender, and other characteristics. At least since the 1966 publication of the “Coleman Report” by sociologist James S. Coleman and coauthors, researchers and policymakers have understood the critical impacts of race, poverty, and segregation on educational attainment (Coleman et al. 1966). And educational inequities remain a major problem today. Rigorous research demonstrates that inequalities in both opportunity and outcomes along the lines of race and social class begin early and often persist throughout students’ K–12 years and beyond, and that they are much larger in the United States than in comparable countries (Bradbury et al. 2015; Putnam 2015). Some of the research carefully describes the specific contexts and challenges that minority and lower-social-class students face and how these challenges create early education gaps. Other studies illustrate the consequences of these gaps for children’s later learning and development (Duncan et al. 2007; Duncan and Magnuson 2011). 2 And though this body of research is smaller, a few studies have looked at trends in inequalities across cohorts (Carnoy and García 2017; Magnuson and Duncan 2016; Reardon 2011; Reardon and Portilla 2016), with mixed or inconclusive findings regarding the changes in the gaps. 3 In addition, these latter studies, however, do not address causes that could drive changes in the gaps over time. As such, there is a need both for a better understanding of these causes and for strategies to counter them.

In this paper, we describe recent skills gaps and trends in them by social class, as measured by socioeconomic status; analyze some of the major factors driving the gaps; and explore a set of diverse school district-level initiatives that are helping to narrow gaps. The paper is structured in three sections.

First, we examine social-class-based gaps at kindergarten entry among the most recently surveyed kindergarten cohort (the kindergarten class of 2010–2011). We study how gaps manifest in both cognitive and so-called noncognitive skills, as both skill types are important components of children’s development.
Next we compare these gaps with those of an earlier kindergarten cohort. We look at changes from 1998 to 2010 in the skills gaps between children in the top and bottom social-class quintiles (primarily using SES as the proxy for social class). We also analyze how sensitive gaps are to the inclusion of several key determinants of student performance, such as children’s own characteristics, family composition, and parental and education practices at home.
Then we review a set of case studies of school districts that have employed comprehensive educational strategies to provide more children (especially low-income children) with strong early academic and life foundations, and to sustain and build on early gains throughout the K–12 school years.
Finally, we look at the implications of our findings, and, based on the case study examples from diverse communities, we discuss strategies that districts can employ along with district and state policy changes that will make those strategies easier to adopt and more sustainable.

For the first two analyses, we use two nationally representative studies from the National Center for Education Statistics (NCES): the Early Childhood Longitudinal Study of the Kindergarten Classes of 1998–1999 and 2010–2011. These data provide information about children’s skills and about the children themselves, such as their race/ethnicity, socioeconomic status, language spoken at home, etc. The data also provide information on the children’s experiences in their early years, such as how actively their parents engaged them in enriching activities, whether they attended prekindergarten care, and the number of books the child has (see Appendix A). This information allows us to test the associations between children’s characteristics and their educational outcomes at school entry. For the second analysis, we draw on 12 case studies of community and school districts employing comprehensive educational strategies (Weiss 2016a–h). We explore the qualitative information provided on investments these districts have made in early childhood education, on both within-school and broader K–12 supports for children, and on evidence that these investments are delivering both improved academic achievement and broader gains for children. Based on this evidence, the report ends with conclusions and recommendations for further research, practice, and policy. Appendices A and B provide detailed discussions of the data and methodology used in this paper.

How large are recent performance gaps at kindergarten entry?

This section documents inequalities among the most recently tracked cohort of students as they entered kindergarten in 2010. It provides us with the most recently available view of the various aspects of gaps at the school starting gate, all of which are critically important for understanding the implications of those gaps. The findings below draw on the Early Childhood Longitudinal Study of the Kindergarten Class of 2010–2011, and we use data from the fall measurement in the kindergarten year. (This section partly builds on our previous work; see García 2015 and García and Weiss 2015. See Appendices A and B for details on the variables and methodology used.)

Our decision to examine performance in both cognitive and noncognitive skills reflects growing acceptance that children’s development is a complex process in which both skill types build on and interact with each other, and on evidence of the roles that both types of skills play in the education process and adulthood outcomes (see García 2015; García and Weiss 2016; Levin 2012a, 2012b). Traits and skills such as critical thinking, creativity, problem-solving, persistence, and self-control are vitally important to children’s full development, and are nurtured through life and school experiences. These skills, sometimes referred to as noncognitive or social and emotional skills, tend to develop—or lag—in tandem with cognitive skills. Noncognitive or social and emotional skills are thus linked to academic achievement, and also to outcomes in adult life, such as productivity and collegiality at work, good health, and civic participation.

For these analyses, we use a measure of socioeconomic status that has three components: the educational attainment of parents or guardians, parents’ occupational prestige (determined by a score), and household income (see more details about the SES construct in Tourangeau et al. 2013, 7-56 to 7-60). We divide children of the 2010–2011 kindergarten class into five groups based on SES quintile. To measure the gaps in performance by socioeconomic status, we compare the average performance of children in the top fifth of the SES distribution with the average performance of children in the bottom fifth. This provides an estimate of the relative advantage of a child in the top fifth of the SES distribution (referred to in this report as “high-SES”) with respect to a child in the bottom fifth (“low-SES”).

Children are not equally prepared for school when they enter kindergarten, and our analyses show that students’ social class strongly determines their relative position in the performance distribution. Most socioeconomically disadvantaged children lag substantially in both reading and math skills, and these skills levels rise along with socioeconomic status (sometimes referred to as socioeconomic gradients). Children in the highest socioeconomic group score significantly higher in reading and math than children in the lowest socioeconomic group. As Table 1 shows, the relative unadjusted gaps in reading and math, i.e., the advantages of high-SES children relative to low-SES children in 2010 are 1.17 and 1.25 sd, respectively (Table 1 also shows that, after controlling for clustered data, the gaps are 0.94 and 0.91 sd, respectively). 4 Reading and math skills advantages of children in the middle of the SES distribution relative to the lowest SES group are roughly half as large as the advantages of high-SES children to the lowest SES group. 5

Children in the lowest socioeconomic quintile also lag substantially in noncognitive skills, based on assessments by both parents and teachers, although these gaps are smaller than those in reading and math. Socioeconomic-based gaps in self-control and approaches to learning are approximately one-third to one-half as large as gaps in reading and math. 6 In 2010, children in the high-SES quintile scored 0.38 sd and 0.51 sd higher in self-control and approaches to learning as reported by teachers (0.36 sd and 0.56 sd after clustering; see Table 1) than children at the low-SES quintile (see Figure A ). Using parents’ assessments of the same skills, the gaps are 0.39 sd and 0.56 sd, respectively (0.33 sd and 0.46 sd after clustering; see Table 1).

Our analyses also document stark socioeconomic disparities in inputs, child and family characteristics, and other factors that can affect school readiness ( Table 2 ). Here too we find a correlation between socioeconomic status and other factors that impede educational development. Low-SES students are more likely than their high-SES peers to be immigrants and less likely to speak English at home, to live with two parents, to have participated in center-based pre-K care activities in the previous year, and to have engaged in early literacy practices at home. Among children in the low-SES group, half (50.4 percent) are Hispanic, 23.1 percent are white, 19.6 percent are black, and 2.5 percent are Asian. 7

Though these gaps in both cognitive and noncognitive skills are troubling and call for policy recommendations, better policy solutions can be designed if we understand how these gaps have changed over time and what factors have played a role in those changes. Education outcomes are the product of a combination of multiple factors, which can reinforce or mitigate relative advantages or disadvantages in a dynamic fashion. We examine these issues in the rest of the paper.

How do the performance gaps in the 2010–2011 kindergarten class compare with the gaps in the prior generation?

The analyses presented in this section compare the inequities in inputs and the performance gaps between high-SES and low-SES students who began kindergarten in 2010 with the gaps among high-SES and low-SES schoolchildren in the prior academic generation, the 1998 cohort. We also analyze factors that have had major influences on the changes in performance of kindergartners, and briefly discuss the research and policy implications of our findings.

How have the characteristics of the children in the lowest and highest SES groups changed in a generation?

We first analyze children’s characteristics by SES quintiles in the two cohorts. This enables us to identify differences in the characteristics of low-SES kindergartners in 2010 versus in 1998. These changes may help explain why the performance gaps we are studying grow or shrink (for example, if children in the low-SES quintile in 2010 were more likely than their 1998 peers to have access to public programs such as pre-K, they might be more prepared for kindergarten, and thus the relative advantage of high-SES children might shrink). 8

Table 2 shows the student and family characteristics of the kindergarten classes of 1998–1999 and of 2010–2011, by SES quintile. The table also includes pre-K care arrangements and two indices of developmental activities parents undertake with their children—indices of “literacy/reading activities” and “other activities”). 9 The table also summarizes parents’ expectations regarding their children’s educational attainment. To some extent, expectations are based on hope, but they can also respond to behavioral patterns children are exhibiting that hint at their future success. Expectations can also influence outcomes by representing how motivated parents are for their children’s education. The ECLS-K survey does not ask parents how their expectations (and changes in their expectations) affect their provision of educational activities or support, but their answers to the expectations question can be used as a reasonable proxy of the degree to which parents are aware of their children’s education and willing to support it. 10

The most significant changes in children’s characteristics by SES quintile are for children in the bottom of the distribution. In 2010, a greater share of children in this group are Hispanic (50.4 percent, an increase of 10.6 percentage points relative to the 1998 share of 39.8 percent), live in homes where the main language is not English (40.3 percent, an increase of 9.1 percentage points from 31.2 percent in 1998), and are immigrants (49.8 percent, an increase of 19.5 percentage points from 30.3 percent in 1998). In 2010, a greater share of children do not live with two parents (54.9 percent, an increase of 9.3 percentage points from 45.6 percent in 1998), and live in poverty (84.6 percent, an increase of 13.3 percentage points from 71.3 percent in 1998). These substantially greater disadvantages for children at the bottom of the SES scale could all be reflections of both the much weaker national economic context in 2010 versus 1998 and the growing inequality described above.

These children’s likelihood of attending center-based pre-K did not change significantly across generations (about 44 percent for both cohorts), but they were more likely to be looked after by parents or relatives (with the share increasing from 46.4 percent in 1998 to 50.9 percent in 2010). These children’s parents also reported having a somewhat larger number of books at home for the children, and there were increases in their indices of educational and engagement activities (two composite measures, with the literacy/reading index measuring how frequently parents read books to their child, tell stories, sing songs, and talk about nature and how frequently the child reads picture books and reads outside of school, and the “other” index measuring how frequently parents and children play games or do puzzles, play a sport or exercise together, and build something or play with construction toys; and how often parents help children do arts and crafts and involve children in household chores). These parents’ expectations about their children’s educational attainment also changed significantly: the share who expected their children to attain no more than a high school diploma decreased by more than half (from 24.1 percent in 1998 to 11.4 percent in 2010), and the share of parents who expected their children to attain at least a bachelor’s degree increased, markedly for those expecting their children to obtain an advanced degree (a master’s degree, Ph.D., or M.D.).

Among children in the high-SES quintile, the group in 2010 includes a lower share of white children (falling from 78.8 percent in 1998 to 71.3 percent) and a larger share of Asian children (increasing from 4.7 percent in 1998 to 8.7 percent). Children in the high-SES group became slightly more likely to live with their two parents (the share of children who lived with one parent decreased from 11.1 percent in 1998 to 9.6 percent), and to have attended center-based pre-K (65.8 percent in 1998 and 69.9 percent in 2010). We only see a small increase in the reported number of books at home. 11 The share of homes reporting having more than 200 books—the maximum—increased slightly in 2010, across all SES quintiles except for the middle quintile). As was true of low-SES parents, those in the highest quintile raised their expectations for their children’s educational attainment from 1998 to 2010. Compared with the 1998 cohort, a larger proportion of high-SES children in the 2010 cohort were expected by their parents to attain an advanced degree (master’s degree or higher), while a lower share expected their children to attain a bachelor’s degree only.

How did the performance gaps between the children in the lowest and highest SES groups change in a generation?

Changes over time in the input factors by socioeconomic status (child and family characteristics, early-education practices, and parents’ expectations) explored above have been found by researchers to have major impacts on the outcomes (test scores on reading and math, and measures of noncognitive skills) explored in this section. 12 In other words, we would expect that changes in the unadjusted skills gaps (gap measures that do not include controls for child and family characteristics, early-education practices, and parents’ expectations) would partially reflect the compositional differences between the class of 2010–2011 and the class of 1998–1999. For example, we would anticipate that if the more recent generation’s low-SES parents read to their children more frequently, helped them do more arts and crafts, or had higher expectations for them, these factors would correlate with narrowing skills gaps. Also, we would expect that the adjusted skills gaps (gap measures that are net of the influence of child and family characteristics, early-education practices, and parents’ expectations, and thus reflect the SES gaps) would be different for the two cohorts if the correlations between inputs and outcomes had changed over time or if the share of children’s outcomes the adjustments account for had changed over time.

To understand these factors’ potential influence on gaps, we examine both unadjusted and adjusted gaps in the tables in this section. We also examine gaps by some of the components of the SES index, such as household income or mother’s educational attainment, and by other variables that are sometimes used as proxies of the child’s socioeconomic background, such as number of books in the home. If the gaps by SES components and proxies somewhat differ, this tells us that researchers’ choices about how to divide children into groups and compare them matter—both for their findings and for their policy recommendations.

Table 3 shows the unadjusted and adjusted gaps between the standardized scores in reading and math of kindergarten children in the top SES quintile relative to the bottom SES quintile in 1998 and the change in that gap by 2010. 13 Table 4 performs the same analysis for gaps in measured noncognitive skills. The tables show two somewhat perplexing patterns. On the one hand, the cognitive and noncognitive skills gaps between high-SES and low-SES children are large and statistically significant in both cohorts. But while significant social-class-based performance gaps persist from one kindergarten generation to the next, there is not the same consistency in how the high-SES to low-SES gaps change. For some cognitive and noncognitive skills, the performance gaps grow, while for others the gaps shrink, or remain the same from one generation to the next (which may complicate the process of understanding why performance gaps have changed over time).

Beginning with our unadjusted model (data column one), the only substantial increase in the gap between high- and low-SES children from 1998 to 2010 was in reading skills, which increased by one-tenth of a standard deviation. There were no significant changes in gaps in math skills, which, as the literature indicates, are less sensitive than reading skills to parents’ activities at home (see Rothstein 2004, 2010). Similarly, gaps in approaches to learning as reported by parents and in self-control as reported by teachers did not change significantly, and gaps in approaches to learning as reported by teachers and in self-control as reported by parents shrank by roughly the same amount as the reading gap (about a tenth of a standard deviation—0.12 and 0.08 sd, respectively). Figure A provides a graphic illustration of the unadjusted gaps in cognitive and noncognitive skills of high- and low-SES children across the two cohorts.

The additional models estimated for each outcome and shown in Tables 3 and 4 offer other key findings. In Model 1, we used the full samples for the two cohorts but did not include any controls that capture characteristics of children or their parents or the early education practices in which families engage. Model 2 partitions the data into schools and classes, or clusters, so that the subjects in the clusters are more similar to one another than to those in other groups. Under this adjustment, the gaps shrink substantially, by between 15 and 25 percent across the skills, and the regression fit improves significantly (see increased adjusted R-squared, i.e., this model explains more of the total variation in the outcomes than the first model). This clustering takes into account school segregation, that is, that children are not randomly distributed but tend to concentrate in schools or classrooms with children of the same race, social class, etc. Clustered estimates provide a comparison of the skills gaps of peer students—those in the same schools and classrooms—rather than a comparison across schools. García (2015) and Magnuson and Duncan (2016) offer these estimates too.

How do child and family characteristics, activities, and expectations affect SES-based performance and performance gaps?

We next examine the contribution of the certain variables of interest to SES-based performance gaps. We approach this in two ways. First, we examine the changes in the gaps (Tables 3 and 4, Models 3 and 4) and the overall reduction in the gaps that results from controlling for children and their family characteristics, early literacy practices, and parental expectations of educational achievement ( Table 5 ). Second, we assess the influence of select early educational practices on performance and how that influence has changed over time by looking at the associations between these inputs and performance ( Table 6 ).

Models 3 and 4 in Tables 3 and 4 use the samples that result from removing observations without full information for the controls of interest. 14 Adding controls is important because performance gaps based on socioeconomic status may be explained by differences in variables other than the child’s socioeconomic status. In other words, we aim to determine which part of the gap is attributable to children’s SES, net of other factors that matter for performance. Thus, in the third data column (Model 3), we add controls for individual and family characteristics (gender, race/ethnicity, whether English is the primary language spoken at home, disability, age, whether children live with two parents) and early educational and play activities (center-based pre-K care, indices for literacy/reading activities and other activities, and total number of books the child has). Model 3 also includes the interactions between the early education variables with time. 15 In the fourth data column (Model 4), we control for the same factors as in Model 3 but add controls for parental expectations of children’s educational attainment (whether they expect their children’s highest level of education attained will be high school diploma or less, some college or vocational studies, bachelor’s degree, or advanced degree) and their interaction with time. 16 We describe these results in the next section.

Including covariates changes the estimates of SES-based skills gaps in various ways. First, the gaps between the top- and bottom-SES quintiles shrink, showing that SES-based gaps are partially explained by the variation in the controls (which is not visible in the tables). 17 Second, controls do not significantly change the SES-based gaps over time, in general; i.e., the coefficients associated with changes in the gaps between high- and low-SES children remain almost the same, or change very minimally, depending on the skill measured. The statistical significance of the SES-based skills gaps in 1998 is not affected by the inclusion of the controls (see rows “Gap in 1998–1999” in tables), but the statistical significance of the changes in the gaps between 1998 and 2010 (see rows “Change in gap by 2010–2011” in tables) is somewhat affected by the inclusion of the controls (note that the sizes of the coefficients measuring gaps in 1998 change after the inclusion of the controls, but that the sizes of the coefficients measuring changes in them between 1998 and 2010 do not change significantly). In reading, the change in the gap between 1998 and 2010 diminishes and becomes statistically insignificant in the last model (the relative gap increases by 0.08 sd but this change is not statistically significant), meaning that adding parental expectations of education accounts for some of the increase in the gap detected in Models 1 to 3. The only SES-based skills gap that shows a statistically significant increase from 1998 to 2010 once parental expectations are controlled for is the gap associated with parents’ assessment of approaches to learning, which increases by 0.11 sd. Gaps between high- and low-SES children in cognitive and noncognitive skills after adjustments are made are shown in Figure B .

As mentioned above, the fact that the skills gaps decrease after controls are taken into consideration affirms that SES-based gaps are due in part to variation in the controls among high- versus low-SES children. This trend can be seen in Table 5, which, as noted above, shows the overall reduction in gaps that results from controlling for child and family characteristics, early literacy practices, and parental expectations of educational achievement. With respect to cognitive skills, the 1998 gaps shrink by 46 percent and 53 percent, respectively, after the inclusion of the covariates. About half of the gaps are thus due to other factors that are associated both with SES status and with the outcomes themselves. The reduction in the 1998 gaps for noncognitive skills varies from 28 percent (approaches to learning as reported by teachers) to 74 percent (approaches to learning as reported by parents). (For self-control as reported by teachers, the reduction is 51 percent versus 35 percent when reported by parents.)

While the gaps hold after the inclusion of controls across outcomes, gaps in 2010 are less sensitive to the inclusion of the covariates than they were in 1998. This trend can also be seen in Table 5. 18 Declining values from 1998 to 2010 indicate that factors such as early literacy activities and other controls are not, as a group, explaining SES-based gaps as much as they had a decade prior. This change could be due to the failure of the index to fully capture parents’ efforts to nurture their children’s development and/or the index becoming somewhat out-of-date. In any event, the resistance of gaps to these controls should worry researchers and policymakers. The waning influence of these controls makes it harder to understand what drives SES gaps. It also suggests that the gaps may be growing more intractable or, at least are less easily narrowed via the enactment of known policy interventions.

Finally, we examine the association of performance outcomes (not performance gaps) with selected early educational practices, including having attended center-based pre-K, literacy/reading activities and other activities, and total number of children’s books in the home (Table 6). 19 We are mainly interested in two potential patterns: whether these factors are associated with outcomes (and, if so, how intense the associations are), and whether the relationships have changed over time.

In keeping with established research, having attended center-based pre-K is positively associated with children’s early reading and math skills. For 1998, the estimated coefficients are 0.11 sd for reading skills and 0.10 sd for math skills, substantial associations that do not change significantly over time. In other words, attending pre-K in 1998 improved kindergartners’ reading skills by 0.11 sd and improved kindergartners’ math skills by 0.10 sd relative to not attending pre-K. However, while center-based pre-K continues to reduce self-control as reported by teachers in 2010, the effect is less negative in 2010 (the 0.06 improvement from 1998 to 2010 shown in the bottom panel of the table shows us that the effect in 2010 was -0.07 [-0.13 plus 0.06], compared with -0.13 sd in 1998). We find no independent effect of center-based prekindergarten schooling (i.e., no effect in addition to SES, in addition to other individual and family characteristics, or in addition to other SES-mediated factors), on approaches to learning or on self-control as reported by parents. 20

The number of books children have at home likewise supports their skills at the beginning of kindergarten. Indeed, this factor is positively associated with all outcomes but self-control reported by parents. The coefficients are very small, of about 0.01 to 0.02 sd (associated with changes in outcomes for each 10 additional/fewer books the child has, as expressed by the continuous scale with which number of books in the home is measured, which is divided by 10 for the analyses (as mentioned in Appendix A), and these relationships do not change over the time period.

The two types of parenting activities that are summarized by the indices “reading/literacy activities” and “other activities” show interesting correlations with performance and patterns over time. On the one hand, the “reading/literacy activities” index (a composite of how frequently parents read books to their child, tell stories, sing songs, and talk about nature, and how frequently the child reads picture books and reads outside of school) is strongly and positively associated with all outcomes other than children’s self-control as reported by the teacher. The associations with cognitive skills, especially with reading, are strong and statistically significant—0.17 sd for reading performance and 0.07 sd for math—and these associations did not change significantly between 1998 and 2010. For noncognitive skills, the relationships are strong for those assessed by parents, though they shrink by about half over time: self-control is 0.14 sd in 1998 and decreases by 0.08 sd by 2010; approaches to learning is 0.32 sd in 1998 and decreases by 0.17 sd by 2010). The relationship is much weaker, though still statistically significant, for teachers’ assessed approaches to learning (it is 0.03 sd in 1998 and does not change significantly by 2010).

On the other hand, the index that measures other enrichment activities that parents do with their children (a composite of how frequently parents and children play games, do sports, build things, work on puzzles, do arts and crafts, and do chores) shows significant correlations with all of the skills, but they may be either positively correlated or negatively correlated, depending on the skill. For cognitive skills, the associations are statistically significant and negative, though stronger and somewhat more meaningful or more intense with reading achievement (-0.12 sd in 1998) than with math achievement (-0.04 sd). 21 These associations did not intensify nor weaken over time. For noncognitive skills the associations are highly positive and statistically significant, and very strong for parents’ assessment of approaches to learning (0.29 sd in 1998). As explained by García (2015), these correlations between “other activities” and noncognitive skills as assessed by parents could be bidirectional: engaging children in enrichment activities might enhance their noncognitive skills, but, at the same time, parents who are more inclined to participate in their children’s early play and educational time are probably more likely to perceive or judge that their engagement has an impact on their children’s skills. But the fact that both the frequency with which parents engage in most of these activities and the importance of this index for parent-assessed skills increased noticeably from 1998 to 2010 (by 0.22 sd for self-control and 0.27 sd for approaches to learning) suggests that parents are growing more informed and involved in their children’s early education over time. It also indicates that parents are increasingly acting on this knowledge and that this involvement will continue to grow, albeit potentially with decreasing marginal returns to time and resources invested. The association between “other activities” and teachers’ assessments of children’s noncognitive skills is also positive but weaker than that of parents’ assessments (about 0.03 sd for approaches to learning and 0.05 sd for self-control), and remained unchanged during the time period studied.

Finally, we find a strong association between parental expectations for their children’s educational attainment and all measured skills. In other words, net of socioeconomic status, the higher the expectations, the higher cognitive skills children have, and the higher the assessments by parents and teachers of children’s noncognitive skills. The parental expectations portion of the table measures children’s performance relative to children whose parents’ expectations are the lowest (high school diploma or less). While the expectation that a child will pursue some vocational education or complete college has a statistically positive influence on all skills measures except for reading, the expectation that their children will complete a bachelor’s degree or more education has a stronger influence, including on reading skills: between 0.11 to 0.16 sd higher in reading and between 0.17 to 0.22 sd higher in math in 1998. High expectations for children’s educational attainment also have a statistically positive effect on noncognitive skills. When the expectation is for an advanced degree (master’s or higher), coefficients vary from 0.12 sd in self-control by teachers to 0.38 sd in approaches to learning by parents in 1998. In addition, most of these associations—particularly the cognitive gradients—grow in 2010. Relative to children whose parents have low expectations, children whose parents have the highest expectations for their children’s attainment (graduate studies) perform much better in reading and math than in 1998 (relative gaps grow by 0.19 and 0.12 sd respectively). A similarly stronger association is noted for noncognitive skills assessed by teachers (though not for parents’ assessments of their children’s skills).

Sensitivity analyses: Do performance gaps vary based on which proxy for social class (socioeconomic status) is used?

Part of the challenge to making conclusive statements about trends in education gaps by social class is the existence of multiple valid proxies for measuring children’s social class or socioeconomic status. 22 Although researchers treat these proxies as equivalent, and even interchangeable, the lack of a comparison of results obtained using various indicators limits our capacity to extract major conclusions on social-class trends and their drivers, and hence hinders the plausibility and effectiveness of the policy recommendations that build on any specific indicator’s findings (net of other methodological and instrumental differences that may exist across studies).

We thus conduct analyses using several of the main proxies employed to measure socioeconomic status. The purpose of these analyses is twofold. The first purpose is to test the sensitivity of the estimated relative gaps, and of trends in them, to changes in the measurement of this key predictor of education performance. (In other words, if all the indicators are reliable proxies of SES, gaps and trends obtained using the various metrics should be similar.) The second purpose is to increase the comparability of the results of studies addressing trends in education inequalities that use various metrics of social class. This is an important issue; in addition to helping reconcile diverse results found in the literature, these analyses may reveal why patterns differ, and have significant policy implications.

As such, instead of the SES composite measure we use to estimate SES-based gaps in this report, we use three alternative indicators to run our analyses: mother’s educational attainment, household income, and number of books the child has in the home. Unlike the SES composite measure, two of these measures offer the advantage of being directly comparable over time. Both mother’s educational attainment and number of books the child has are objective categories. As a limitation, and mainly associated with the information that is available in the raw data, none of these categories can be transformed into a percentile-variable without major transformations. (The adjustments to ensure comparability over time are explained in Appendix A. See Reardon and Portilla 2016 for an analysis with a transformation of the income variable that offers a proper percentile comparison, based on the methodology developed by Reardon 2011.) Still, they are variables associated with social class and can be ordered in groups or categories that identify high- and low-social-class statuses. Thus, with the necessary caution when interpreting and using the findings, we offer this comparison of results as a sensitivity analysis.

We create five categories with these indicators, maintaining the structure of comparing “high-SES” (top quintile) with “low-SES” (bottom quintile) as in Tables 1–5 (note that we are using “SES” interchangeably with “social class” here). For simplicity, Tables 7–9 show only the results from two models: one without covariates (Model 1, baseline estimates) and one with all covariates (Model 4, fully adjusted estimates). We focus on the findings for the baseline relative gaps in 1998 and 2010 first ( Figures C–E ). The overall patterns found in the results suggest that all social-class gaps are statistically significant and sizable. However, the exact sizes of the gaps vary depending on the social-class indicator used and the outcome being assessed. Also, the changes in the gaps over time vary depending on the indicator used to capture children’s social class.

In addition to these general findings, we note some more detailed ones. For 1998, gaps by mother’s educational attainment (Figure C; Table 7) are the largest across all indicators (except for the gap in self-control as assessed by teachers, which is slightly smaller than the gaps as measured using household income and number of books the child has), while gaps by number of books (Figure E; Table 9) are the smallest across all indicators (except for the gap in approaches to learning as assessed by parents, which is slightly larger than the gap for household income). Again, according to the 1998 data, the coefficients of gaps by mother’s educational attainment are generally larger—and in three cases much larger—than those obtained using number of books in the home as the indicator of social class. For example, the relative gap is 1.29 sd in reading and 1.46 sd in math when mother’s education is the SES proxy, compared with gaps of 0.74 sd and 0.97 sd when number of books in the home is the SES proxy.

It is also important to note that gaps by mother’s educational attainment (Figure C; Table 7) and income (Figure D; Table 8)—two of the five components of the SES construct—are very close to the ones obtained by our SES composite measure (as shown in Figure A). All in all, results seem internally consistent as well as generally consistent with prior results on this topic (Reardon and Portilla 2016).

In terms of changes in the performance gaps over time (unadjusted), the findings vary depending on which indicators of social class are used, with mother’s education and household income being the indicators associated with the largest changes in the gaps. Changes in the performance gaps in cognitive skills between 1998 and 2010 by our composite SES measure and books are similar: an increase in the reading gap between children in the top and bottom quintiles of about a tenth of a standard deviation (0.10 sd with the composite SES measure [Figure A] and 0.08 sd if SES is proxied with books), and no significant change in mathematics (there are some differences in the noncognitive outcomes).

However, by mother’s educational attainment, there are no changes in relative reading and approaches to learning gaps reported by parents over time, and a significant reduction in the gaps in the remaining outcomes. Meanwhile, income-based gaps for the two cognitive skills—reading and math—decreased by -0.13 and -0.23 sd respectively, and for approaches to learning as reported by teachers by -0.13 sd. No significant changes occurred for the remaining noncognitive skills.

In sum, this sensitivity analysis demonstrates that all of the indicators are reliable proxies of SES for the estimation of early achievement gaps, though absolute gaps may vary slightly depending on the indicator used. However, the proxies are not equally reliable when we assess trends in the gaps by SES or their drivers. As such, aside from differences in the definitions and procedures used to construct each SES proxy, the proxies should not be treated as fully equivalent. The decomposition conducted here helps clarify the different weights that various components of SES may have in driving changes in gaps by social class. For example, variation in income across groups over time is associated with decreased performance gaps in the cognitive skills between 1998 and 2010, and variation in educational attainment quintiles or categories over time is associated with decreased performance gaps across cohorts in most noncognitive skills. But variation in books in the home over time and among groups is associated with increased gaps in reading and in parents’ assessed approaches to learning. Such findings also point to very different policy solutions: if mothers’ education is the main driver, enhancing that will improve children’s prospects. On the other hand, findings that indicate that income inequality is the larger culprit would point to the need for policies that reduce such inequalities. Future research should consider and look more closely into these questions.

What can we learn from these analyses?

The multiple factors and relationships examined in this section can now be examined from a policy perspective. If the aim is to increase equity, to improve children’s development across the board, and to improve our understanding of children’s development, there are two major policy recommendations:

Directly support less-resourced families so that they have greater access to educational and economic resources (for the latter, see García and Weiss 2017). All the early educational and play activities measured, which include center-based pre-K care and literacy/reading and other activities, as well as the number of books a child has, are positively associated with children’s readiness, and in part account for social-class gaps, but are much less accessible to children of lower socioeconomic status. Virtually all of the associations between these factors and outcomes were strong and positive (with a handful of exceptions), and some even grew over time. A related research recommendation of particular interest would be to examine whether the intensity of these activities or practices has any threshold level of effectiveness (after which point they no longer affect children’s development). 23 Also, it would be helpful to understand why parents’ expectations of their children’s educational attainment increased so much and how this has affected children’s development. For example, do parents have a better understanding of the relationship between educational attainment and prospects for success in life and the workforce? Are children performing better because their parents expect more, or because parents who expect more are also delivering more in the form of enriching activities?
Design and implement strategies that compensate at the community level for children’s lack of access to key foundational resources (economic and educational). These strategies can be considered indirect supports for less-resourced families that reduce inequities and complement the direct supports described above. Examples of communities that have enacted such comprehensive support initiatives provide a good starting point to explore how and why they emerge; the types of supports they provide (from preschool programs and home visits with parents to enriching summer programs, school-based health clinics, and more); the challenges of scaling them up and sustaining them; the benefits they deliver for students, and particularly for disadvantaged students; and their implications for policy at the local, state, and even federal levels. The next section of this report thus presents an analysis based on qualitative data from promising initiatives in a dozen school districts across the country (Weiss 2016a–h).

What are pioneering school districts doing to combat these inequities and resulting gaps?

This section of the report draws on a set of case studies published by the Broader, Bolder Approach to Education (BBA), a national campaign that advances evidence-based strategies to mitigate the impacts of poverty-related disadvantages on teaching and learning. 24 The case studies feature school districts that have employed comprehensive educational strategies to ensure that more children, especially low-income children, have strong early academic and life foundations, and that resulting early gains are sustained and built on through children’s K–12 years. (These strategies are often referred to as “whole-child” approaches to education, in reflection of their holistic nature.) We explore the premise that school districts that take a whole-child approach to education and a whole-community approach to delivering it are likely to enjoy larger gains in academic achievement and to narrow their race- and income-based achievement gaps. In doing so, we are building on evidence suggesting that consistent, strong supports for children and their families—both in and out of school—can avoid the “fade-out” seen among graduates of many pre-K programs and even enhance those programs’ early benefits.

This section is thus divided into four parts: (1) an introduction to the case study districts, followed by discussions of (2) how these districts invest in early childhood care and education, (3) how the districts’ investments in K–12 strategies sustain and boost the early childhood investments, and (4) how academic gains and narrowing achievement gaps indicate that the investments are paying off. Table 10 provides basic information on the 12 school districts/communities studied; Appendix E at the end of this report provides more information on key characteristics of these districts. 25

Introduction to the case studies: Why these districts enacted whole-child strategies

Large and growing disparities in the economic well-being of children in America and extensive evidence linking those disparities to widely diverging educational outcomes have prompted action among a growing number of communities and school districts. Heeding the evidence that out-of-school factors play even larger roles than school-based factors in school performance, these districts are seeking ways to mitigate the poverty-related impediments to effective teaching and learning.

These districts have benefited from a substantial body of research on strategies with promise to address core challenges that students and schools face—strategies that have been shown to shrink achievement gaps by narrowing major disparities in opportunity (Carter and Welner 2013). The first, and perhaps best-documented, of these strategies is high-quality early child care and education, especially when it engages parents early and in meaningful ways. High-quality early childhood education programs not only narrow achievement gaps at kindergarten entry but also deliver long-term benefits to children, their families, and society as a whole (Chaudry et al. 2017; Rolnick and Grunewald 2003).

Programs that support students’ physical and mental health and improve their nutrition are also known to reduce chronic absence and keep students focused and learning, and thus improve their academic performance (CDC 2016). Well-designed after-school and summer-enrichment programs likewise boost achievement, both directly and indirectly by enhancing students’ engagement in and attachment to school (Peterson 2013).

Whole-child approaches integrate these and other strategies into a comprehensive set of aligned interventions, leveraging the whole community’s resources to meet the broad range of student needs. While the impact of such comprehensive approaches has not been studied as extensively as the individual components, considerable theoretical and emerging empirical research point to the strong potential of such strategies to boost achievement and narrow gaps (Child Trends 2014; Oakes, Maier, and Daniel 2017; Weiss 2016i).

This section of the report seeks to add to that knowledge base by sharing qualitative information on how such comprehensive approaches have emerged and grown, what they look like when they are successfully implemented, and what types of outcomes and benefits result and how outcomes vary across diverse communities.

How are whole-child initiatives launched?

Each of the districts studied has distinct circumstances, and thus distinct reasons for coming to the conclusion, as a community, that it needed to take a comprehensive approach to education. At the same time, demographic trends that are affecting virtually every state—and many, if not most, school districts across the country—have played major roles in that decision in every case. 26 Indeed, community and school leaders in all of these districts cited students’ poverty (and, in some districts, demographic shifts) as posing challenges that required looking beyond the school walls to address.

How these factors triggered the initiative’s launch varied, but poverty was at the core in each community’s decision. For example, in 2008, community leaders identified East Durham as one of Durham, North Carolina’s, most distressed areas, based on a community risk assessment conducted by Duke University’s Children’s Environmental Health Initiative. The 120-block area’s 11,000 residents had a 40 percent poverty rate and a homeownership rate of just 19 percent, along with high rates of crime and unemployment, putting its 3,000 children and youth at high risk of academic failure (Weiss 2016e).

Across the country, in Vancouver, Washington, the share of children eligible for subsidized school meals rose from 39 percent to over 50 percent in less than a decade, such that, by 2015, in some central-city schools, more than four in five students qualified for subsidized school meals in 2015 (Weiss 2016b). In another distressed community, in north Minneapolis, median family income was just $18,000 in 2011, and fully one-fourth of the 5,500 Northside students were homeless or “highly mobile” (in such unstable housing that they were at risk of homelessness) (Weiss 2016d). In Pea Ridge, Arkansas, schools “had difficulty finding resources that met the needs of kids,” says superintendent Rick Neal. “We knew that we were not identifying all the needs that were there. I think that’s the way a lot of districts are” (Weiss 2016f). And in the early 1990s, the Tangelo Park neighborhood in Orlando, Florida—an isolated enclave of 3,000 residents, almost all low-income and African American—caught the attention of hotelier and philanthropist Harris Rosen, who was looking for a neighborhood in which to invest (Alvarez 2015).

Each of these districts took different approaches to enacting those comprehensive strategies, based on the community’s specific mix of needs and assets, ideological leaning, available sources of funding, and other factors. One of the most politically progressive of the districts studied, Montgomery County Public Schools (MCPS) in Maryland, paved the way for a whole-child approach in the early 1970s when it enacted housing policy that uses mixed-income residential developments to create communities with families of different income levels. In the 1990s, the county developed Linkages to Learning, a “community schools”–type approach targeted to engaging and partnering with low-income and immigrant parents and families and connecting them with a broad range of community resources (MCPS 2016). (Community schools are known for building partnerships with community agencies and private service providers to meet student and family needs.) Austin Independent School District (AISD), also in a politically progressive jurisdiction, began its whole-child efforts through parent- and community-organizing in schools. It has since invested in social and emotional learning and in a community schools strategy (CASEL 2017).

At the other end of the spectrum are whole-child approaches in Joplin, Missouri, and Pea Ridge, Arkansas, districts located in more politically conservative southern states. These districts operate under the umbrella of Bright Futures USA (a spinoff national nonprofit that began with Joplin’s Bright Futures initiative). The Bright Futures districts take a more individualistic angle, asserting that every member of the community has “time, talent, or treasure” to offer that can help children overcome disadvantage and ensure more equal opportunity (Weiss 2016a).

Two other districts have modeled their efforts on the Harlem Children’s Zone (HCZ). The Northside Achievement Zone in Minneapolis is funded through a grant from the federal Promise Neighborhoods initiative, enacted by the Obama Administration to help more communities dramatically improve the academic success for low-income children by adopting HCZ-like strategies. The East Durham Children’s Initiative in North Carolina is entirely privately funded so far (Weiss 2016e).

In both Kalamazoo, Michigan, and Orlando, Florida, pledges of “Promise” college scholarships have evolved into broader whole-child efforts (Alvarez 2015; Miller-Adams 2015).

Districts also take different approaches based on density. New York City—home to dozens of full-service community schools supported by the Children’s Aid Society and rapidly expanding to more—and Boston—home to the City Connects initiative—leverage a broad range of their respective cities’ arts and cultural offerings, along with health and nutrition and other social services (Weiss 2016g, 2016h). Cultural offerings to supplement other well-rounded services are also part of the full-service community schools district initiative in Vancouver, Washington. In contrast, Partners for Education, which serves the isolated region surrounding Berea College in Kentucky, was the first rural organization to receive a Promise Neighborhood grant and, thus, is a pioneer in exploring how well the model works outside the urban context (Berea College 2013).

What do whole-child initiatives do?

The sections below describe commonalities across these different approaches in terms of investments in children’s earliest years (before school starts), building on these investments throughout children’s K–12 years (both in and out of school), and the gains students and schools enjoy as a result of those investments. 27

How the case study districts invest in early childhood care and education

In keeping with their whole-child approaches to education policy and practice, every one of the 12 districts highlighted as a BBA case study has made investments in early childhood care and education, many of them substantial. These districts’ efforts begin long before children enter school and go beyond pre-K offerings to equip parents in the effort to ensure their children’s readiness for school.

One-on-one engagement with new parents

Investing in babies by engaging parents can include providing new parents with key information about child development and how to keep children healthy and safe. In Joplin, Missouri, Bright Futures Joplin partners with two of the area’s hospitals to deliver new baby “kits” with child development and early literacy information and is trying to raise funds to sustain the project long term and to expand it to reach every new parent (Weiss 2016a). In Vancouver, Washington, 6,000 “literacy packets” are delivered annually to families with children up to age five, providing child-development activities and lessons that families can complete at home (Weiss 2016b).

The districts leverage partnerships to connect parents with a range of school and community resources that support children from birth through kindergarten entry. In Eastern Kentucky, the whole-child program called Partners for Education works with Community Early Childhood Councils to host events such as Week of the Young Child, the Dolly Parton Imagination Library, and Kindergarten Transition Programs (Weiss 2016c). In Montgomery County, Maryland, “Judy Centers”—early child care and family education centers—leverage partnerships with social service agencies and local community nonprofits to increase parents’ access to mental health, nutrition, and other key services (Maryland State Department of Education 2017).

Educating and engaging parents early helps prepare children for school both academically and more broadly for healthy development. Those are the twin goals of the Minneapolis Northside Achievement Zone (NAZ), where currently only one in four preschoolers in the zone is ready for kindergarten based on standardized tests. To improve those odds, the zone has a team of “NAZ Navigators” who work with families to set and track progress toward goals in early childhood and to link this area of family support to goals in academics, housing, career and finance, and behavioral health (Weiss 2016d).

Parenting classes

Parents are children’s first and most important teachers. Like the one-on-one strategies described above, classes for parents provide information on child development, early literacy, health, and constructive disciplinary practices, and offer more specific guidance tailored to specific parents’ needs. Almost every district studied provides new-parent classes. The 1-2-3 Grow and Learn program is a weekly 90-minute literacy-rich program for young children and their parents offered at 12 elementary schools in high-poverty Vancouver neighborhoods. It lays the foundations for school readiness through social and education experiences. In addition, the district’s Family and Community Resource Centers offer parent workshops, groups, and courses to help parents support their children’s learning, while empowerment and skill-enhancement programs—such as job preparation, housing assistance, and parent leadership advisory groups—strengthen parents’ basic skills. Family Academy classes in the North Minneapolis Northside Achievement Zone include “College Bound Babies” (for parents of children up to three years old), which teaches early literacy, numeracy, and positive discipline skills, and “Foundations,” which empowers parents to feel confident talking with their children’s teachers and advocating for their children and their children’s schools.

In many cases, districts employ a combination of one-on-one and group supports, along the lines of Early Head Start. 28 The East Durham Children’s Initiative, a private program modeled loosely after the Harlem Children’s Zone, includes Durham Connects, a home visiting program that supports zone families with children up to age 3 and is followed by weekly or biweekly in-home parent education and support provided by two nonprofit social service providers, Healthy Families Durham and Jumpstart (Weiss 2016e). In Montgomery County, Maryland, family social workers collaborate with classroom teachers to help them develop Family Partnership Agreements, which are based on the strengths, needs, and personal goals of each family. A social worker–led team follows up by phone and with visits. In two of the district’s highest-poverty schools, these supports are complemented by early child care and family education centers (Judy Centers), which provide comprehensive early childhood education and support to children from birth to age five and their families (Marietta 2010).

Big investments in prekindergarten programs

Almost every state in the country now invests at least minimally in pre-K programs for disadvantaged children, and a growing share of states make these programs widely available. 29 Most of the districts we studied, however, have gone far beyond state programs through one or more strategies and funding mechanisms.

A few of these districts benefit from high-quality state pre-K programs that serve a large share of children, freeing the districts to invest in other aspects of early childhood enrichment. The Partners for Education initiative based in Berea, Kentucky, leverages the state pre-K program, which serves all three- and four-year olds who are either low-income or have other risk factors. This enables Partners for Education to use Promise Neighborhood grant funds to place early childhood specialists in pre-K classrooms throughout the four-county region (the region is a Promise Neighborhood region, which means that federal funds are available for a variety of education- and health-related investments). The specialists also provide coaching, professional development, and support for Head Start classrooms, as well as in-home tutoring over the summer.

In East Durham, North Carolina, strong state early education programs are supplemented by partner-led low-cost half-day preschool and a summer kindergarten readiness program, and home visits by parent advocates provide a range of supports, such as connections to state pre-K. In Kalamazoo, Michigan, the Pre-Kindergarten Early Education Program (PEEP) offers half- or full-day pre-K classes in elementary schools for four-year-olds at or below 250 percent of the federal poverty level, per state law, but it adds transportation and meals for those children. PEEP also works with other programs such as Head Start to provide families who are ineligible for PEEP with other options for low- or no-cost quality early education (KPS 2017).

Other districts with less comprehensive state support use federal resources to expand local options. For example, Vancouver draws on both state and federally funded early learning programs to provide pre-K in seven schools, along with district-supported programs for children in Title I schools. As of fall 2015, Vancouver’s new early learning center serves up to 100 additional children or more, with hot meals and playground space from an adjacent elementary school. Montgomery County also enhances state and federal programs with district-level investments: it provides the same literacy-rich curriculum in its Head Start classrooms as in district pre-K classrooms. And Montgomery County uses a blend of federal Title I and Head Start dollars to offer full-day Head Start in 18 of the poorest schools, serving 460 children (Marietta 2010). The Northside Achievement Zone in north Minneapolis uses federal Race to the Top Early Learning Fund money for scholarships for three- and four-year-olds to attend high-quality pre-K, serving 127 children in 2012–2013 and 156 in 2013–2014.

Local programs can also fill in where state programs are weak. Austin, Texas, uses local funds to provide enriching, hands-on full-day programs for the four-year-olds who would otherwise participate in lower-quality half-day state programs. Austin also provides a half-day program for three-year-olds who aren’t served by the state. Families who qualify for both state pre-K and Head Start also receive nutrition, health, and other services (AISD 2017).

Pea Ridge is another community using local resources to supplant state resources. A lack of available seats for children who are eligible for the state’s high-quality Arkansas Better Chance (ABC) pre-K program prompted Pea Ridge to seek a grant to open its own program, which serves 40 children: 20 at-risk children, who receive tuition scholarships, and 20 others whose parents can pay tuition (Weiss 2016f). Missouri’s pre-K program also has too few slots, so Bright Futures Joplin is building a new early childhood learning center that will be funded jointly by the district and the state.

Strengthening the transition to kindergarten

Featured districts also build on pre-K gains and help narrow school-readiness gaps with such programs as full-day kindergarten. Montgomery County Public Schools first started full-day kindergarten in “red zone schools,” those deemed to be most affected by high rates of student poverty, in 2000. Full-day kindergarten has since expanded to every school in the district (Marietta 2010). And Vancouver offers Kindergarten Jump Start, a school readiness program, at all 21 elementary schools, and full-day kindergarten; both programs seek to enhance the transition from pre-K into formal schooling.

Other investments in young children and their families

In addition to the above range of supports for infants, toddlers, and preschoolers and their parents, several of the districts studied by BBA have made additional investments in young children and their families. The Community Storywalk in Clay County, Kentucky, and the Born Learning Trail in Joplin, Missouri, provide opportunities for parents and paid caregivers to learn with their children in a hands-on way through outdoor and physical activities. In Eastern Kentucky, Partners for Education’s Promise Neighborhood grant supports work by national nonprofit Save the Children to improve the health and education outcomes of the region’s children through a literacy program that provides kids ages 5–12 with books and tools to develop strong reading skills. The Promise Neighborhood grant also allows Partners for Education to offer the Children’s Healthy Choices program, which provides healthy snacks and 30 minutes of daily physical activity for children in districts across Eastern Kentucky.

Joplin’s Little Blue Bookshelf program gives age-appropriate books to those children whose families cannot afford them, making the goal of 1,000 hours of reading by kindergarten a viable reality for every child. And the city’s Lend & Learn Libraries provide stimulating toys and socialization time for young children and their parents.

How the school districts invest in K–12 strategies to sustain and boost their early childhood investments

The whole-child approaches these communities embrace for children from birth to five years old continue as those children transition to kindergarten and through elementary, middle, and high school. This represents a sharp difference from most other districts, which focus heavily on narrow academic factors and assessments and thus neglect characteristics emphasized in pre-K, such as building strong teacher–student relationships and attending to the full range of children’s assets and needs. As these examples illustrate, students continue to benefit from a more comprehensive approach to education and there is an array of strategies school districts can use to deliver that comprehensive approach.

Enriching K–12 curricula and activities to sustain pre-K’s whole-child emphasis

A broad set of investments and activities can help sustain pre-K’s whole-child approach, including enhancing classroom experiences, aligning classroom lessons with out-of-school activities that expand children’s worldviews, and using targeted strategies to improve students’ readiness for college, careers, and civic engagement.

Schools that ensure hands-on learning both in and out of the classroom make the most of this opportunity. Joplin and Pea Ridge students and their teachers enjoy service learning projects that are a core component of the Bright Futures strategy. These range from kindergartners organizing coat drives and canned food drives for their neighbors to high school students designing and implementing water research projects and reporting on the health and safety of Joplin’s water supply to the city’s water management agency. In East Durham, partnerships with community agencies and nonprofits enable clubs, field trips to museums, and other enrichment activities.

After-school and summer programs help students build on what they learned during the school year, broaden students’ worldviews and skills, and reduce summer learning loss. In most of the districts studied, schools partner with organizations such as the YMCA, Boys and Girls Clubs, Boy Scouts, and Girl Scouts to provide out-of-school enrichment programs that range from organized sports and help with homework to math and book clubs, theater, and robotics. In addition to boosting student engagement, some focus in particular on academic and college preparatory help, and many also provide snacks or even full meals. Summer camps in Boston and East Durham and book deliveries and clubs in Pea Ridge and Eastern Kentucky—where online options help bridge long distances in rural areas—keep students reading, engaged, and on track for fall classes.

In several districts, the focus on nurturing not only students’ academic skills but also their social and emotional skills strengthens the transition to kindergarten and development throughout the K–12 years. Vancouver’s schools teach and model social and emotional learning in classrooms as part of the district’s work to improve school climate and track student data on engagement and mental health. Under City Connects—the whole-child collaboration among Boston College, Boston Public Schools, and community agencies—school coordinators meet at the start of the year with teachers to discuss the particular strengths and needs of each student and develop plans to support teachers with academic and enrichment activities and meet student needs with small-group sessions on healthy eating and dealing with bullies, referrals to mental health providers, and a range of other supports (Weiss 2016g).

Two districts have made social and emotional learning a particularly high priority. Austin is one of eight districts working with the Collaborative for Academic, Social, and Emotional Learning (CASEL) to comprehensively embed social and emotional learning in teacher training, teacher standards, curricula, and metrics for assessing student and school progress (CASEL 2017). In Montgomery County, former superintendent Joshua Starr drew on the Common Core’s emphasis on problem-solving and critical thinking to lead the design of a new curriculum and classroom practices that nurture social and emotional skills. These are complemented by enhanced support for teachers to nurture social and emotional learning in daily classroom practice, by standards-based report cards that track key social and emotional skills, and by constructive disciplinary policies that reengage students and build their soft skills instead of punishing them for infractions. 30

Several of the districts focus in particular on helping students—many of whom will be the first in their families to go to college—prepare for and make that leap. Strategies include middle-to-high-school transition programs in Joplin and Vancouver and clubs and specialized courses that advance students’ social and organizational skills in Vancouver and Montgomery County. In East Durham, three initiatives (Communities in Schools Durham, Student U, and Citizens in Schools) support youth who are preparing for graduation. They offer site-based mentoring from current undergraduates. Middle and high school students in the North Minneapolis Northside Achievement Zone receive similar assistance. And Vancouver’s GRADS Teen Parent program helps teen parents stay in school, graduate, and be more effective parents. De-tracking, an intentional decision to not separate students who are achieving at different levels into different classrooms or types of courses, which is the norm in Austin and in some Montgomery County high schools, helps ensure that college preparatory classes serve students of all income levels rather than just wealthier, nonminority students. 31

College readiness is also a high priority for many Bright Futures districts. In Joplin, programs such as Operation College Bound enhance students’ understanding of and access to postsecondary education, complementing initiatives that help students navigate transitions to higher education and other sensitive periods of their academic lives. And in Pea Ridge, specialized high schools such as the Manufacturing and Business Academy and Pea Ridge Academy provide targeted support for students who want to go straight to jobs and careers or need special academic supports.

Mentoring and tutoring to get and keep students engaged

In the case study districts, the whole-child approach includes understanding the critical importance of one-on-one relationships with caring adults who support children’s academic and broader needs. Strategies can be as simple as the car and bus “buddies” who greet children in Pea Ridge each morning as they arrive at school, or as intensive as the volunteer “lunch buddies” who meet regularly with Joplin and Pea Ridge students to eat with them, talk about their days, and offer guidance. Northside Achievement Zone in North Minneapolis partners with Big Brothers Big Sisters to connect students with mentors, and over 500 volunteer mentors in Vancouver, Washington, support students in Family and Community Resource Centers.

These relationships are key to efforts in large urban districts and remote rural ones. The Children’s Aid Society has partnered with the New York City Department of Education to integrate a strong school curriculum with out-of-school enrichment programming, as well as provide child and family support services designed to remove barriers to students’ learning (Weiss 2016h). Children’s Aid community schools offer both tutoring and mentoring among their after-school options, as do Boston’s City Connects schools. In Eastern Kentucky, to bridge the long distances between one school and community and another, mentors use Skype to connect with eighth- and ninth-graders in Promise Neighborhood area schools.

Supports for student health and family wellness as a tool for sustaining early gains

Several of the districts studied have established health clinics in some or all of their schools, including Montgomery County, Vancouver, and New York City. In some other districts, such as Austin, school coordinators can arrange for mobile clinics to come to schools. These clinics provide basic preventive care through immunizations and check-ups, along with prescriptions and other care for sick children, physical and mental health screenings, follow-up counseling, mental health care, and even crisis intervention when needed.

Nutrition is another critical factor that affects physical and mental health and thus learning. In East Durham, Back Pack Buddies and summer lunch programs prevent hunger and keep kids nourished. Food and clothing pantries plus social media outreach in Pea Ridge and Joplin enable counselors and teachers to meet targeted immediate needs so students can focus and learn. Montgomery County has expanded its breakfast-in-the-classroom program to serve all students in a growing share of schools (MCPS 2017).

Many of these districts look beyond meeting students’ basic health and nutrition needs to advancing their and their families’ wellness and strengthening their ties to the community. Vancouver’s GoReady! back-to-school festivals provide backpacks, school supplies, shoes and socks, immunizations and dental screenings, and even haircuts, plus resources from community partners . In Eastern Kentucky, physical and mental health supports provided through state-supported Family Resource and Youth Service Centers are complemented by school–community collaborative activities through a run/walk club, a summer fitness program, a Jump Start program, and gardening and food preservation activities. And the East Durham Children’s Initiative runs a Healthy Living Initiative that refers families to nutrition counseling programs, Zumba classes, cooking demonstrations, and walking groups; it also distributes children’s bicycles and partners with local farmers markets to provide families with fresh produce.

Though research has long affirmed the importance of parental engagement, many schools struggle to meaningfully engage parents. The case study districts show how it can be done. In the rural regions around Berea, Kentucky, where physical distance makes engagement difficult, Partners for Education’s Families and Schools Together project convenes parents, school staff, and local agency professionals to help parents build social networks. In the North Minneapolis Northside Achievement Zone (NAZ), a high-poverty heavily minority area, regular one-on-one meetings between parents and “connectors”—specialized social workers who grew up in the area, are familiar with its challenges, and are a core component of the NAZ strategy—provide opportunities to conduct family needs assessments and provide referrals to relevant services. These regular meetings lead to deeper parental engagement in their children’s schools.

And full-service community schools such as those in Vancouver and New York City specialize in parent outreach and engagement. Community schools in these districts draw on parental input to shape school policies and practices and provide parents with an opportunity to meet one another. For example, a “parents’ coffee room” in a New York City school with a large Dominican population evolved from simply providing a space for parents to hang out after student drop-off to a center for parent-led workshops, parent–student collaborative plays, and more.

Other targeted supports provide added help for the most vulnerable students and their families. In Vancouver, for example, student advocates conduct home visits to parents of kindergartners and first-graders who are at risk of chronic absenteeism. In these visits, the advocates emphasize the importance of attendance and brainstorm with parents ways to reduce specific barriers to attendance. Complementary in-school efforts reward strong attendance. High-risk Montgomery County Public Schools students benefit from an unusual, but very effective, system of targeted support. Specifically, the districts’ funding system redistributes money from wealthier schools to higher-poverty schools, enabling the latter to provide smaller classrooms, more individualized attention, and more specialists in English language learning, special education, and other areas (Elmore, Thomas, and Clayton 2006).

How academic gains, including smaller achievement gaps, indicate that the investments are paying off

Providing children from birth through 12th grade and their families with targeted supports both within and outside of school has enabled these communities to make progress toward a range of goals. First, compared with students in peer districts, these districts’ students tend to have better outcomes on traditional measures of academic achievement such as test scores and graduation rates. Second and just as, if not more, important, these districts have improved students’ kindergarten readiness, engagement, and health and well-being, and helped the students be better prepared for college, careers, and civic engagement. This is true in large part due to these districts’ intentional bucking of a growing trend of diverging practices in which students in high-poverty schools are subject to narrow academic drilling while students in wealthy schools benefit from a broader set of activities and learning experiences beyond a narrow focus on preparing for standardized tests. These districts ensure enrichment for all students, regardless of socioeconomic status. Finally, in contrast with the national trend in recent decades of rapidly growing achievement gaps between wealthy and poor students, these districts are also narrowing race- and income-based achievement gaps: while all students are gaining ground, those who started off behind tend to see the largest gains.

Most of the data presented in this section do not come from experimental studies; with a few exceptions (which are noted in the case studies), they rely on nonexperimental comparisons with a similar nontreatment group, such as other low-income children in the district or other high-poverty districts in the state. However, they are gathered from official district, state, or federal resources in all cases, except for the minority of cases in which such data are not publicly available. Perhaps most importantly, in contrast with many other programs that have reported substantially improved outcomes for very vulnerable groups of students, these programs do not cherry-pick students to get these results. Rather, these initiatives serve all students in the enrollment area for a school, a cluster of schools, or, in many cases, an entire district; as described above, they are serving some of the nation’s most vulnerable students and their families. 32 Moreover, many of these efforts are, for lack of a better term, “turnarounds.” That is, students in an existing system that is considered to be failing are offered a new approach in the same school building, making the large gains reported particularly striking given the notable lack of similar progress from much-larger-scale, more publicized attempts at employing other turnaround strategies. 33

Establishing more expansive goals and implementing ways to track progress toward those goals also offers timely guidance, given that the Every Student Succeeds Act (ESSA) asks states, districts, and schools to do just that. These districts have not only set broader goals, they are demonstrating real progress toward achieving these goals. Because of their success, many now serve as role models for other districts or entire regions, and a few are beginning to influence state policy as well.

Higher rates of kindergarten readiness predict school success

Some of the kindergarten readiness efforts described above have translated into improved readiness to learn and, thus, greater odds of success in kindergarten and throughout the K–12 years. In Eastern Kentucky, East Durham, and Minneapolis, children who participated in early learning programs significantly increased their rates of kindergarten readiness across a range of metrics and developmental domains. A study of Montgomery County Public Schools found much larger gains in reading for children in the full-day Head Start program than for children in the half-day program, with full-day students more than doubling their reading scores over the year and especially pronounced gains for the most vulnerable students: Hispanics and English language learners (Marietta 2010).

Rising test scores and narrowing gaps in core academic subjects are an important sign of sustained early gains

While only one of many indicators, rising test scores and narrowing gaps in core academic subjects are an important sign that schools in case study districts have sustained and enhanced early gains. Despite serving a higher percentage of low-income, black, Hispanic, and English language learner students than the district average, Austin’s Alliance Schools—schools in which community organizers have worked to empower parents in conjunction with teacher advocacy efforts—saw substantial gains in scores on the Texas Assessment of Academic Skills, the state’s main standardized test, in the three years after parent-organizing efforts began. Increases varied from four points to 15–19 points, with the latter increases occurring in schools with the highest levels of parental engagement (Henderson 2010). Subsequent rollout of social and emotional learning in district schools (some of which were also Alliance schools) produced gains in the share of students deemed proficient on the State of Texas Assessment of Academic Readiness (STAAR, the next-generation state assessments) in the years following that rollout, with students in the first set of schools with social and emotional learning programs scoring higher on state math and reading exams than those in later school cohorts. The small group of Minneapolis Northside Achievement Zone students who were tested increased their proficiency on the Minnesota Comprehensive Assessments (MCA) exam, with the share scoring as proficient rising from 14 percent in the 2012–2013 academic year to 22 percent in 2013–2014. 34 Students who had enrolled in the Northside Achievement Zone in 2013 had larger gains than those who enrolled in 2014, and, overall the largest proficiency gains were among first- and second-graders, with the smallest gains in middle schools.

Despite serving a much poorer and socially and economically isolated student body than in state schools overall, the Eastern Kentucky schools served by Partners for Education have seen substantially higher increases in test scores: from 2012 to 2015, math test scores in the Promise Neighborhood region rose 7.0 percentage points compared with 4.4 percentage points across the state, and reading scores rose 7.3 percentage points, compared with 5.8 percentage points statewide.

An independent study of middle school students who participated in the after-school programs run by Children’s Aid Society community schools in New York City had bigger gains in math and reading test scores than peers who did not participate. They also had higher relative increases in school attendance and in teacher-reported “motivation to learn.” And while the Children’s Aid Society did not make early childhood education investments a core component of its strategy, its Zero-to-Five program, which connects the federal Early Head Start and Head Start programs, produced relative test score gains among participants. Specifically, a study found that participants outperformed their peers 97 percent of the time on third-, fourth-, and fifth-grade standardized tests in math and reading, demonstrating a significant long-term positive effect (Caspe and Lorenzo Kennedy 2014).

Increases (or lack of decreases) in reading scores over the summer months (between the end of the school year and the start of the following year) can be an especially important indicator of sustainable academic achievement, since low-income students tend to lose substantial ground when they are out of school for the summer. Students who attended the North Minneapolis Northside Achievement Zone’s extended learning summer programs increased their reading test scores between the end of one school year and the beginning of the next, a period when scores normally decrease. And an evaluation of students who attended the East Durham Children’s Initiative’s summer camp in the summer of 2014 found that they lost no ground in literacy over those months.

Case study districts with more mature initiatives and those offering higher or more intensive doses of whole-child interventions are producing particularly large academic gains. Students enrolled in City Connects elementary schools in Boston score significantly higher on tests of both academic and noncognitive skills in elementary and secondary school, with the highest-risk students, such as English language learners, showing especially large gains. Scores of City Connects elementary school students on the Stanford Achievement Test version 9 increased between one-fourth and one-half a standard deviation greater than scores of their non–City Connects peers. And graduates of City Connects secondary schools are more likely to attend one of Boston’s three most selective public high schools.

Better student attendance and engagement are also predictors of academic gains

Chronic absenteeism depresses achievement, particularly among low-income students. A 2009 study found that New York City Children’s Aid Society’s community schools had “far higher” attendance than peer schools, and that schools with health centers tended to have higher attendance than those without health centers (Clark et al. 2009). Students attending City Connects high schools in Boston have significantly lower rates of chronic absenteeism than their peers (Boston College Center for Optimized Student Support 2012). In Joplin, Missouri, attendance rates among high school students increased 3.7 percentage points, rising from 91.3 percent in 2008 to 95.0 percent in 2012; black and Hispanic students closed gaps with their white peers over that period. At the same time, reportable disciplinary incidents—which keep students out of school and are found to drive at-risk students to disengage—dropped by over 1,000, from 3,648 in 2008 to 2,376 in 2012. 35

Every infant and toddler in East Durham whose family participated in the Healthy Families Durham home visiting program is up to date on immunizations; this helps at-risk children avoid missing school due to illness. In Pea Ridge, collaboration with one of the city’s doctors enabled the district to provide physical exams for high school students who would otherwise go without them. This not only improved their health but enabled them to participate in the kinds of extracurricular sports activities that boost student engagement. And City Connects’ practice of helping families draw on Medicaid coverage and of referring eligible students to insurance-eligible providers increases students’ access to both physical and mental health care. Given extensive evidence linking reduced absenteeism and improved physical and mental health to academic gains, these initiatives’ records of boosting both attendance and health represent another pathway to student success. 36

Increases in advanced coursework and completion of associated exams suggest improved college and career readiness

Because most of the initiatives studied have been in place for less than 10 years, and a few for five or fewer, there is less evidence of their impact on high school graduation and college enrollment. Nonetheless, the degree to which low-income and minority students in these districts perform better and have seen greater gains on these key indicators than their peers in comparable districts or across the state highlights the promise of comprehensive education approaches and, in some instances, their capacity to sustain and even boost children’s early gains.

Parent-organizing in Austin helped establish a program to get more low-income and minority middle school students into rigorous science and math programs, enabling them to successfully compete for slots in the prestigious LBJ High School Science Academy. From the 2007–2008 to the 2014–2015 academic year, the number of Kalamazoo Public School students taking Advanced Placement (AP) courses more than doubled, with low-income and African American students experiencing the largest absolute gains in participation and Hispanic students experiencing the largest percentage gains. Black and low-income students roughly quadrupled their participation in such courses; 263 black students and 193 low-income students took AP classes during the 2014–2015 academic year, up from 63 and 53 respectively in 2007–2008 (Miller-Adams 2015). Over the same period, the number of Hispanic students taking AP courses increased by a magnitude of 10—from just 8 to 78. And in Vancouver, which also made socioeconomic diversity of students in advanced courses a priority, enrollment in AP courses rose by 67 percent overall from 2007–2008 to 2013 –2014, and nearly three times as fast, by almost 200 percent, among low-income students.

Higher graduation rates and increasing college attendance of disadvantaged students are another measure of success of comprehensive strategies

In the early 2000s, the graduation rate at Austin’s Reagan High School fell below 50 percent and enrollment dropped to just 600 students. By 2015, with the benefit of a community schools strategy, the school was serving more than 1,200 students and had a graduation rate of 85 percent.

In the first six years of Bright Futures, Joplin’s graduation rate rose from 73 to 87 percent; from 2012 to 2015 it rose 13 percentage points, versus just 5 percentage points across the state as a whole. At the same time, the cohort dropout rate fell from 6.4 percent to 2.8 percent, with the dropout rate for black students falling slightly more. And in Kalamazoo, incentives to finish high school have proven to be powerful tools for disadvantaged students when combined with mentoring, tutoring, and after-school options. The district’s graduation rate rose from 64 percent in 2009 to 69 percent in 2014, with “five-year cohort graduation rates consistently higher than four-year rates, suggesting that some students may be opting to stay in school an extra year (or even just for the summer) to complete the credits necessary to get a high school diploma” (Miller-Adams 2015, 67). Moreover, African American girls in Kalamazoo graduate at higher rates than their peers across the state, and 85 percent of those graduates go to college.

Initiatives that have had time to mature have made particularly large gains. Montgomery County’s Linkages to Learning initiative began in 1993 and it substantially expanded its pre-K program around a decade later; a county policy responsible for improved racial integration has been in place even longer, since the early 1970s. Hispanic, low-income, and African American students in Montgomery County Public Schools are much more likely than their counterparts across the state to graduate from high school—80.0 vs. 77.5 percent, 81.0 vs. 77.8 percent, and 86.4 versus 80.5 percent, respectively. And from 2011 to 2014, a period when the share of students in poverty and the share of minority students rose in the district, overall graduation rates rose 2.9 percentage points, from 86.8 to 89.7 percent. There were much larger gains for Hispanic and black students, whose graduation rates rose (respectively) by 4.7 percentage points (from 75.3 to 80.0 percent) and 5.1 percentage points (from 81.3 to 86.4 percent), thus narrowing their gaps with their white peers by 3.4 and 3.8 percentage points, respectively (MCPS 2015). Participation in Boston’s City Connects program, which began in 2001, cuts a student’s odds of dropping out of high school nearly in half: 8.0 percent versus 15.2 percent for comparison students (Boston College Center for Optimized Student Support 2014). In Vancouver, the four-year graduation rate rose from 64 percent in 2010 to almost 80 percent in 2013, and the five-year rate rose from 69 percent in 2010 to over 80 percent in 2013. Vancouver’s Hispanic students had five-year graduation rate gains of over 15 percentage points.

Strong parent and community engagement is another sign of progress

The comprehensive, whole-child, whole-community approaches in the featured school districts have built strong school–community partnerships. Two indicators of the strength of the partnerships are the levels of parent and community engagement. In Joplin, 194 more adults are now serving as mentors and tutors than five years ago. And the American Association of School Administrators, National School Public Relations Association, and Blackboard Connected selected Vancouver Public Schools Superintendent Steve Webb and Chief of Staff Tom Hagley for their 2011 Leadership through Communication Award for their successful efforts to increase family engagement in high-poverty VPS schools.

Parental engagement boosts student achievement both directly and through other improvements to families’ situations. As they work actively with their “connectors,” Northside Achievement Zone parents in North Minneapolis become more likely to make academics a priority, to engage with their children’s schools, and to be focused on sending their children to college. The support also helps more families connect with stable housing, substantially reducing the number of times that some vulnerable families move. In 2014–2015, up to 300 Austin families benefited from help with legal, employment, health, and housing issues at the family resource center, which also provides classes for parents, including English language learning classes. And Montgomery County Public Schools social workers who specialize in early childhood education make an average of 200 home visits, 1,000 phone contacts, and 300 direct contacts with parents at school or conferences each month. These lead to roughly 1,000 monthly referrals to community services—many of them emergency interventions dealing with food, clothing, and housing—that help families meet their children’s basic needs and, thus, support their children’s education (Marietta 2010).

In some cases, engagement enhances school leadership. Through access to supports such as social services and adult education, parents of students in New York’s Children’s Aid Society community schools got more involved in their children’s schools, took more responsibility for their children’s schoolwork, reported feeling more welcome within the schools, and were observed to be a greater presence in the community schools than in comparison schools. And over 2,000 Kentucky parents have undergone training at the Berea Commonwealth Institute for Parent Leadership since its creation in 1997. Many of these parents have gone on to join school boards, serve on school councils, and engage in day-to-day educational advocacy.

Expansion of these initiatives shows that other districts, and even state policymakers, consider them successful

After City Connects succeeded in improving student achievement in over a dozen of Boston’s highest-poverty schools, the initiative caught the attention of state policymakers, who recruited City Connects to help turn around schools in Springfield, home to another large high-poverty urban district in Massachusetts. Aided by federal School Improvement Grant funds, City Connects has operated in Springfield since 2010, expanding from six to 13 schools in its first four years there. In New York City, the Children’s Aid Society played a central role in Mayor Bill de Blasio’s 2016 decision to employ a community schools strategy to turn around 100 of the city’s most struggling schools. And in both Vancouver and Austin, district leaders have led advocacy efforts to bring community schools to other communities in the region and to support the introduction of state-level legislation to enhance the work.

Bright Futures began in Joplin, Missouri, in 2009 but is now a national organization. Bright Futures USA has 50 affiliates in eight states, many of which—such as Pea Ridge—are just two or three years old. The newest affiliate, in Fairbanks, Alaska, has just been made official. In Virginia, Dave Sovine, superintendent of a second-year affiliate, Frederick County Public Schools, is reaching out to several of his counterparts across the region to create the first regional Bright Futures initiative (Gizriel 2016). If established, this would allow for the kind of cross-district collaboration identified by Bright Futures founder C.J. Huff as critical to breaking down the silos created by arbitrary boundaries that reflect political preferences rather than children’s daily realities. 37

As this report demonstrates, very large social-class-based gaps in academic performance exist and have persisted across the two most recently studied cohorts of students starting kindergarten. The estimated gap between children in the top fifth and the bottom fifth of the SES distribution is over a standard deviation in both reading and math in 2010 (unadjusted performance gaps are 1.17 and 1.25 sd respectively). Gaps in noncognitive skills such as self-control and approaches to learning—which are critical not only as foundations for academic achievement but also more broadly for children’s healthy development—are about half as large (about 0.4 sd in self-control, and slightly over 0.5 sd in approaches to learning in 2010).

Another important finding from our study is that gaps were not, on average, sensitive to the set of changes that may have occurred between 1998 and 2010: gaps across both types of skills are virtually unchanged compared with the prior generation of students—those who entered school in 1998. The only cognitive gap that changed substantially was in reading skills, which increased by about a tenth of a standard deviation. The gaps by SES in mathematics, in approaches to learning as reported by parents, and in self-control as reported by teachers did not change significantly. And relative gaps in approaches to learning as reported by teachers and in self-control as reported by parents shrank between 1998 and 2010, by about a tenth of a standard deviation. 38

We also find that, while taking into account children’s personal and family characteristics, parental activities, and other factors reduces the gaps somewhat, it does not come close to eliminating them. This means that there is a substantial set of SES-related factors that are not captured by the traditional covariates used in this study but that are important to understanding how and why gaps develop. Moreover, the capacity for these other factors—child and family characteristics, early education investments, and expectations—to narrow gaps has decreased over time. This suggests that, while such activities as parental time spent with children and center-based pre-K programs cushion the negative consequences of growing up in a low-social-class context, they can do only so much, and that the overall toxicity of lacking resources and supports is increasingly hard to compensate for. The resistance of gaps to these controls should thus be a matter of real concern for researchers and policymakers.

These troubling trends point to critical implications for policy and for our society: clearly, we are failing to provide the foundational experiences and opportunities that all children need to succeed in school and thrive in life. The failure to narrow gaps between 1998 and 2010 suggests, too, that investments in pre-K programs and other early education and economic supports were insufficient to counter rising rates of poverty and its increasing concentration in neighborhoods where black and Hispanic children tend to live and learn.

But there is also good news. The case study review in the previous section of this report explores district-level strategies to address these gaps, strategies that are being implemented in diverse communities across the country. The most effective ones begin very early in children’s lives and are sustained throughout their K–12 years and beyond. The communities studied all employ comprehensive educational approaches that align enriching school strategies with a range of supports for children and their families. Their implementation is often guided by holistic data and, to the extent possible, this report provides a summary, as well, of student outcomes, using both traditional academic measures and a broad range of other measures.

These findings also point to further research questions that need to be addressed, including why gaps changed or did not change, for whom they changed (or did not change), and what is the absolute change in children’s skills over time. 39

Parents are doing what they need to do, and a growing number of communities are, too, but as a society, we are still falling far short

Over the period studied, parents across all social class groups became more involved in their young children’s early education and development, with increases in involvement being especially pronounced among low-SES parents. Parents were more likely in 2010 than in 1998 to read regularly to their children; to sing to them; to play games with them; and to enroll them in center-based pre-K programs. Parents in 2010 also had significantly higher expectations for their children’s educational attainment, and mothers themselves were more highly educated—both factors that are associated with higher achievement for those children. In other words, parents’ actions show that they are doing more of what the brain science indicates they need to do, which either suggests that information about children’s needs during those years is more widely disseminated than it was for the prior cohort we studied, or that parenting styles have changed in a way that benefits the development in the early years.

And, as the case studies indicate, the number of communities that have embraced systems of comprehensive enrichment and supports (“Broader, Bolder Approaches to Education”) is growing. As these communities have shown, such comprehensive education policies are feasible; embedded in these policies is an understanding that children’s development involves nurturing a variety of competencies throughout the stages of development, that there are many individuals participating in these processes, and that coordinated efforts by various stakeholders are needed to put these processes to work. Key principles that span across the case studies include very early interventions and supports, parental engagement and education, pre-K, kindergarten transitions, whole-child approaches to curricula, and wraparound supports that are sustained through the K–12 years. Given the significant need for more such strategies, it is important to understand the factors that drove their enactment in a diverse set of communities, and to continue to monitor both the challenges these communities (and others like them) encounter and the outcomes/benefits of the initiatives.

However, despite the abundance of child development information available to researchers and parents—about the serious impacts of child poverty, about what works to counter those effects, about the importance of the first years of life for children, and about the value of education—our data indicate insufficient policy response at all levels of government. Pre-K programs have expanded incrementally and unevenly, with both access and quality still wildly disparate across states and overall availability severely insufficient. There is a dearth of home visiting programs and of quality child care (Bivens et al. 2016). Child poverty has increased (see Proctor, Semega, and Kollar 2016 for recent trends in child poverty rates). And the schools these children enter face increasing economic and racial segregation but with even fewer resources than they had in 1998 to deal with them (Adamson and Darling-Hammond 2012; Baker and Corcoran 2012; Carnoy and García 2017). And while a growing number of districts have embraced Broader, Bolder approaches, that number is failing to keep up with high and growing need.

In sum, it is actually positive, and somewhat impressive, that gaps by and large did not grow in the face of steadily increasing income inequality, compounded by the worst economic crisis in many decades (EPI 2012, 2013; Saez 2016). But it is disappointing and troubling that new policy investments made in the previous decade were insufficient to make even a dent in these stubborn gaps. We cannot ensure real opportunities for all our children unless we tackle the severe inequities underlying our findings. And while momentum to enact comprehensive and sustained strategies to close gaps is growing, such strategies are not being implemented nearly as quickly as children need them to be.

Next policy steps

These data on large, stubborn gaps across both traditional cognitive and noncognitive skills should guide the design of education policies at the federal, state, and local levels; the combined resources and support of government at all three levels are needed if we are to tackle these inequalities effectively. 40

Policymakers can begin by learning from the small-scale, district-level strategies presented in the review of case studies above (see the section “What are pioneering school districts doing to combat these inequities and resulting gaps?” above). Looking at these case studies, policymakers can ask: What are the key strategies these communities employed, what main components characterize these strategies, and how did these communities effectively implement the strategies? What challenges did these communities face, what was needed to overcome the challenges, and how can we shape policies that better support other communities’ abilities to respond to such challenges and, to the extent possible, avert them? The latter set of questions is particularly pertinent to issues of scalability, financing, and sustainability, all of which have posed significant challenges for the districts studied and others like them. Policymakers can further ask: What other sources or examples might we learn from? Obvious ones include other districts that employ “community schools” strategies (as Vancouver, New York City, and Austin do) and Promise Neighborhood initiatives beyond Berea/Eastern Kentucky and the Northside Achievement Zone. Bright Futures affiliates now exist in 50 districts across eight states—and the program continues to grow—offering another set of communities to look to.

Also, new opportunities under the Every Student Succeeds Act (ESSA)—from funding to expand and align early childhood education programs to broader and more supports-based educator- and school-accountability systems—provide another avenue for exploration and educational improvement. This is already the focus of states and districts across the country—as well as of education policy nonprofits and associations—and is a focus that has the potential to inspire viable larger-scale models (Cook-Harvey et al. 2016).

We must take action, in particular, in those areas of policy related to early education in which we have seen little or no progress over the past decade. These include child care: comprehensive supports that engage parents as partners in their children’s education must start early and be of high quality to prevent the emergence of gaps and provide time to close any gaps that emerge (Bivens et al. 2016, among others). Quality preschool, among the most-agreed-upon strategies to avert and narrow early gaps, continues to be much talked about but far too little invested in and far too infrequently and shoddily implemented. The advantages of preschool have been known for decades, and significant progress has been made in preschool enrollment over that time; however, preschool enrollment stagnated soon after 2000 (Barnett et al. 2017; U.S. ED 2015) and there continue to be significant inequities in access (see Table 2; García 2015) and, just as important, in quality (NIEER 2016). And the gains made through these early, whole-child-oriented supports must be sustained through children’s K–12 years, with attention to issues of funding levels and equity, racial and socioeconomic integration, and enriching opportunities in the hours after school and in the summer months.

Altogether, this report adds to the strong evidentiary base that identifies strategies to reduce the education consequences of economic inequality. It also sheds light on the need to conduct further research on the channels that drive or cushion changes in readiness. A close follow-up of these trends in the near future and of the measures adopted to really tackle inequities will not only determine what type of society we will be, but will also say a lot about what type of society we actually are. This study, affirming a growing number of other studies on these issues, points to an “American Dream” that is alive in public pronouncements but dormant and pale in reality.

About the authors

Emma García is an education economist at the Economic Policy Institute, where she specializes in the economics of education and education policy. Her areas of research include analysis of the production of education, returns to education, program evaluation, international comparative education, human development, and cost-effectiveness and cost-benefit analysis in education. Prior to joining EPI, García conducted research for the Center for Benefit-Cost Studies of Education and other research centers at Teachers College, Columbia University, and did consulting work for the National Institute for Early Education Research, MDRC, and the Inter-American Development Bank. García has a Ph.D. in economics and education from Teachers College, Columbia University.

Elaine Weiss served as the national coordinator for the Broader, Bolder Approach to Education (BBA) from 2011 to 2017, in which capacity she worked with four co-chairs, a high-level task force, and multiple coalition partners to promote a comprehensive, evidence-based set of policies to allow all children to thrive. Weiss came to BBA from the Pew Charitable Trusts, where she served as project manager for Pew’s Partnership for America’s Economic Success campaign. Weiss was previously a member of the Centers for Disease Control and Prevention’s task force on child abuse and served as volunteer counsel for clients at the Washington Legal Clinic for the Homeless. She holds a Ph.D. in public policy from the George Washington University and a J.D. from Harvard Law School.

Acknowledgments

An earlier version of this paper was prepared for “Strong Foundations: The Economic Futures of Kids and Communities,” the Federal Reserve System Community Development Research Conference, Washington, D.C., March 23–24, 2017. We appreciate the feedback we received from our discussant Richard Todd and from the audience. The authors gratefully acknowledge Rob Grunewald and Milagros Nores for their insightful comments and advice on earlier drafts of the paper. Special gratitude is expressed to Sean Reardon, for his advice and thorough guidance on the sensitivity analyses affecting the measurement of the cognitive skills and their implications for our study, and for sharing useful materials to help test our results. We thank Ben Zipperer and Yilin Pan for their advice on issues associated with multiple imputation of missing data. We are also grateful to Lora Engdahl and Krista Faries for editing this report, and to Margaret Poydock for her work preparing the tables and figures and formatting the report. Finally, we appreciate the assistance of communications staff at the Economic Policy Institute who helped to disseminate the study, especially Dan Crawford, Kayla Blado, and Elizabeth Rose.

Address correspondence to: Economic Policy Institute, 1225 Eye St. NW, Suite 600, Washington, D.C., 20005. Email: [email protected] ; [email protected] .

Figures and tables

Unadjusted cognitive and noncognitive skills gaps between high-ses and low-ses children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010.

	Gap between high-SES (fifth) and low-SES (first) quintiles in 1998	Change in gap from 1998 to 2010
Reading	1.07	0.10
Math	1.26
Self-control (by teachers)	0.39
Approaches to learning (by teachers)	0.63	-0.12
Self-control (by parents)	0.47	-0.08
Approaches to learning (by parents)	0.54

The data below can be saved or copied directly into Excel.

The data underlying the figure.

Notes: SES refers to socioeconomic status. The gaps are the baseline unadjusted standard deviation scores for high-SES children relative to low-SES children. The gap in 2010 equals the gap in 1998 plus the change in the gap from 1998 to 2010. For example, the gap in approaches to learning as reported by teachers in 2010 is 0.51 sd (0.63 – 0.12). For statistical significance of these numbers, see Tables 3 and 4, Model 1.

Source: EPI analysis of ECLS-K, kindergarten classes of 1998–1999 and 2010–2011 (National Center for Education Statistics)

Copy the code below to embed this chart on your website.

Fully adjusted cognitive and noncognitive skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010

	Gap between high-SES (fifth) and low-SES (first) quintiles in 1998	Change in gap from 1998 to 2010
Reading	0.60
Math	0.61
Self-control (by teachers)	0.18
Approaches to learning (by teachers)	0.44
Self-control (by parents)	0.29
Approaches to learning (by parents)	0.13	0.11

Note: SES refers to socioeconomic status. The gaps are standard deviation scores for high-SES children relative to low-SES children after adjusting for all family and child characteristics, pre-K schooling, and enrichment activities with parents, and parental expectations for children’s educational attainment. The gap in 2010 equals the gap in 1998 plus the change in the gap from 1998 to 2010. For statistical significance of these numbers, see Tables 3 and 4, Model 4.

Unadjusted cognitive and noncognitive skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, using mother's educational attainment as a proxy for socioeconomic status

	Gap between top and bottom quintiles in 1998	Change in gap from 1998 to 2010
Reading	1.29
Math	1.46	-0.15
Self-control (by teachers)	0.32	-0.10
Approaches to learning (by teachers)	0.64	-0.24
Self-control (by parents)	0.47	-0.14
Approaches to learning (by parents)	0.66

Notes: The gaps are the baseline unadjusted standard deviation scores for high-SES children relative to low-SES children where high-SES children have mothers in the top quintile of the education distribution and low-SES children have mothers in bottom quintile of the education distribution. The gap in 2010 equals the gap in 1998 plus the change in the gap from 1998 to 2010. For statistical significance of these numbers, see Table 7, Model 1.

Unadjusted cognitive and noncognitive skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, using household income as a proxy for socioeconomic status

	Gap between top and bottom quintiles in 1998	Change in gap from 1998 to 2010
Reading	1.09	-0.13
Math	1.31	-0.23
Self-control (by teachers)	0.42
Approaches to learning (by teachers)	0.60	-0.13
Self-control (by parents)	0.44
Approaches to learning (by parents)	0.44

Notes: The gaps are the baseline unadjusted standard deviation scores for high-SES children relative to low-SES children where high-SES children are in households with incomes in the top quintile of the income distribution and low-SES children are in households with incomes in bottom quintile of the income distribution. The gap in 2010 equals the gap in 1998 plus the change in the gap from 1998 to 2010. For statistical significance of these numbers, see Table 8, Model 1.

Unadjusted cognitive and noncognitive skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, using number of books the child has in the home as a proxy for socioeconomic status

	Gap between top and bottom quintiles in 1998	Change in gap from 1998 to 2010
Reading	0.74	0.08
Math	0.97
Self-control (by teachers)	0.32
Approaches to learning (by teachers)	0.46
Self-control (by parents)	0.28
Approaches to learning (by parents)	0.58	0.09

Notes: The gaps are the baseline unadjusted standard deviation scores for high-SES children relative to low-SES children where high-SES children have a number of books in the home in the top quintile of the books-in-the-home distribution and low-SES children have a number of books in the home in the bottom quintile of the books-in-the-home distribution. The gap in 2010 equals the gap in 1998 plus the change in the gap from 1998 to 2010. For statistical significance of these numbers, see Table 9, Model 1.

Reading and math achievement gaps, and principal noncognitive skills gaps between high-SES and low-SES children at the beginning of kindergarten in 2010–2011, under unadjusted and clustered models

	Reading		Mathematics		Self-control (by teachers)		Approaches to learning (by teachers)		Self-control (by parents)		Approaches to learning (by parents)
	1 (unadjusted)	2 (clustered)	1 (unadjusted)	2 (clustered)	1 (unadjusted)	2 (clustered)	1 (unadjusted)	2 (clustered)	1 (unadjusted)	2 (clustered)	1 (unadjusted)	2 (clustered)
Gap in 2010–2011	1.169***	0.944***	1.250***	0.911***	0.386***	0.363***	0.513***	0.562***	0.391***	0.326***	0.563***	0.460***
	(0.024)	(0.036)	(0.024)	(0.034)	(0.029)	(0.041)	(0.027)	(0.041)	(0.028)	(0.041)	(0.028)	(0.044)
Controls
Demographics	No	No	No	No	No	No	No	No	No	No	No	No
Education and engagement	No	No	No	No	No	No	No	No	No	No	No	No
Parental expectations	No	No	No	No	No	No	No	No	No	No	No	No
School fixed effects	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Observations	14,090	14,090	14,040	14,040	12,180	12,180	13,280	13,280	12,890	12,890	12,900	12,900
Adjusted R2	0.165	0.281	0.190	0.276	0.021	0.114	0.034	0.105	0.018	0.028	0.037	0.118

Note: Using the full sample. For statistical significance, *** denotes p < 0.01, ** denotes p < 0.05, and * denotes p < 0.1. The number of observations is rounded to the nearest multiple of 10. Sizes may differ from those inferred from Tables 3–6, and from those in García 2015, due to differences in the sample sizes or to rounding.

Source: EPI analysis of ECLS-K, kindergarten class of 2010–2011 (National Center for Education Statistics)

Child and family characteristics, main developmental activities, and parental expectations for children, kindergarten classes of 1998–1999 and 2010–2011, by socioeconomic status (SES)

1998–1999		Low-SES (quintile 1)	Low-middle SES (quintile 2)	Middle SES (quintile 3)	High-middle SES (quintile 4)	High-SES (quintile 5)	All quintiles
Child and family characteristics and main developmental activities
Race/ethnicity	White	26.40%	53.70%	61.20%	68.10%	78.80%	57.70%
	Black	26.20%	17.80%	15.50%	12.00%	6.40%	15.60%
	Hispanic	39.80%	21.20%	15.80%	12.70%	6.80%	19.20%
	Hispanic English language learner (ELL)	28.40%	9.50%	4.80%	3.10%	1.40%	9.40%
	Hispanic English speaker	11.50%	11.70%	10.90%	9.60%	5.40%	9.80%
	Asian	2.30%	1.70%	2.30%	2.70%	4.70%	2.70%
	Other	5.30%	5.60%	5.30%	4.40%	3.40%	4.80%
Poverty status	Lives in poverty	71.30%	22.30%	10.60%	4.20%	1.10%	21.80%
Language	Child’s language at home is not English	31.20%	12.00%	7.00%	6.10%	5.30%	12.30%
Family composition	Not living with two parents	45.60%	30.50%	23.80%	15.80%	11.10%	25.10%
	Number of family members	4.84	4.55	4.42	4.36	4.40	4.51
	First- or second-generation immigrant	30.30%	15.10%	12.80%	13.10%	15.40%	17.30%
Pre-K care arrangements	Pre-K care	64.20%	70.90%	76.50%	81.00%	87.80%	76.20%
	Pre-K care, center-based	43.70%	45.00%	50.20%	55.40%	65.80%	52.20%
	Parental care	30.50%	22.60%	17.20%	15.40%	9.90%	18.90%
	Care by relative	15.90%	18.30%	16.20%	11.80%	6.60%	13.70%
	Care by nonrelative	5.30%	8.20%	10.90%	11.60%	13.70%	10.00%
	Care by multiple sources	4.60%	5.90%	5.50%	5.80%	3.90%	5.20%
Activities indices	Literacy/reading	-0.221	-0.059	-0.010	0.070	0.193	-0.003
	Other educational and engagement activities	-0.114	-0.011	0.014	0.042	0.071	0.002
Number of books	Average number	32.4	58.1	74.3	87.9	107.3	72.5
Number of books, grouped by least to most	0–25	61.70%	31.60%	20.20%	11.30%	5.00%	25.50%
	26–50	23.10%	34.80%	30.80%	30.60%	21.40%	28.20%
	51–100	11.30%	23.40%	32.90%	36.00%	41.00%	29.10%
	101–199	1.80%	4.00%	5.70%	6.60%	9.50%	5.60%
	More than 200	2.10%	6.20%	10.30%	15.50%	23.00%	11.50%
Parents’ expectations for their children’s educational attainment
Highest education level expected	High school or less	24.10%	15.20%	7.70%	3.70%	1.20%	10.20%
	Two or more years of college, vocational school	16.40%	21.80%	21.40%	11.60%	3.80%	14.90%
	Bachelor’s degree	33.20%	38.70%	46.70%	58.80%	57.20%	47.10%
	Master’s degree	9.20%	9.40%	10.30%	13.60%	22.80%	13.10%
	Ph.D. or M.D.	17.10%	15.00%	13.90%	12.30%	15.00%	14.60%

2010–2011		Low-SES (quintile 1)	Low-middle SES (quintile 2)	Middle SES (quintile 3)	High-middle SES (quintile 4)	High-SES (quintile 5)	All quintiles
Child and family characteristics, and main developmental activities
Race/ethnicity	White	23.10%	45.50%	56.80%	69.00%	71.30%	52.90%
	Black	19.60%	17.00%	13.40%	9.40%	5.80%	13.20%
	Hispanic	50.40%	28.30%	19.70%	12.20%	8.60%	24.10%
	Hispanic English language learner (ELL)	36.10%	11.90%	5.20%	2.10%	0.90%	11.40%
	Hispanic English speaker	14.30%	16.30%	14.40%	10.10%	7.70%	12.60%
	Asian	2.50%	2.80%	3.20%	4.40%	8.70%	4.20%
	Others	4.40%	6.40%	7.00%	4.90%	5.60%	5.70%
Poverty status	Lives in poverty	84.60%	35.70%	10.90%	3.10%	0.60%	25.50%
Language	Child’s language at home is not English	40.30%	15.60%	8.00%	5.00%	7.00%	15.30%
Family composition	Not living with two parents	54.90%	41.70%	34.10%	19.30%	9.60%	31.80%
	Number of family members	4.81	4.62	4.53	4.44	4.46	4.57
	First- or second-generation immigrant	49.80%	25.70%	18.90%	17.20%	21.60%	26.10%
Pre-K care arrangements	Pre-K care	66.60%	75.60%	81.60%	85.00%	88.30%	79.30%
	Pre-K care, center-based	44.30%	47.00%	53.10%	61.60%	69.90%	55.10%
	Parental care	34.90%	25.40%	19.10%	15.40%	12.00%	21.40%
	Care by relative	16.00%	19.70%	17.40%	12.70%	8.60%	14.90%
	Care by nonrelative	3.30%	5.50%	7.40%	7.30%	6.90%	6.10%
	Care by multiple sources	1.50%	2.40%	3.10%	2.90%	2.70%	2.50%
Activities indices	Literacy/reading	-0.231	-0.038	0.033	0.094	0.171	0.008
	Other educational and engagement activities	-0.049	0.022	0.029	0.026	0.001	0.006
Number of books	Average number	35.2	57.6	74.1	90.8	106.3	73.1
Number of books, grouped by least to most	0–25	59.30%	33.60%	19.40%	11.50%	5.00%	25.50%
	26–50	24.70%	31.70%	32.50%	26.90%	22.40%	27.70%
	51–100	11.20%	24.80%	32.30%	39.00%	41.70%	30.00%
	101–199	1.70%	3.10%	5.50%	6.50%	7.70%	4.90%
	More than 200	3.10%	6.80%	10.30%	16.20%	23.20%	12.00%
Parents’ expectations for their children’s educational attainment
Highest education level expected	High school or less	11.40%	6.20%	5.00%	2.40%	1.00%	5.20%
	Two or more years of college, vocational school	16.70%	25.00%	17.20%	9.80%	3.20%	14.40%
	Bachelor’s degree	34.80%	39.10%	47.00%	57.10%	53.10%	46.30%
	Master’s degree	10.70%	12.30%	14.60%	16.80%	26.60%	16.20%
	Ph.D. or M.D.	26.40%	17.30%	16.20%	13.90%	16.10%	17.90%

Note: SES refers to socioeconomic status.

Reading and math skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, under unadjusted to fully adjusted models

	Reading models				Mathematics models
	1 (unadjusted)	2	3	4 (fully adjusted)	1 (unadjusted)	2	3	4 (fully adjusted)
Gap in 1998	1.071***	0.846***	0.641***	0.596***	1.258***	0.932***	0.668***	0.610***
	(0.024)	(0.032)	(0.031)	(0.031)	(0.022)	(0.033)	(0.030)	(0.031)
Change in gap by 2010	0.098***	0.122***	0.096*	0.080	-0.008	0.025	0.053	0.051
	(0.033)	(0.046)	(0.051)	(0.052)	(0.032)	(0.045)	(0.047)	(0.048)
Controls
Demographics	No	No	Yes	Yes	No	No	Yes	Yes
Education and engagement	No	No	Yes	Yes	No	No	Yes	Yes
Parental expectations	No	No	No	Yes	No	No	No	Yes
School fixed effects	No	Yes	Yes	Yes	No	Yes	Yes	Yes
Observations	30,950	30,950	26,050	26,050	31,850	31,850	26,890	26,890
Adjusted R2	0.152	0.243	0.289	0.293	0.189	0.265	0.331	0.336

Notes: Models 1 and 2 use the full sample; Models 3 and 4 use the complete cases sample. Robust standard errors are in parentheses. For statistical significance, *** denotes p < 0.01, ** denotes p < 0.05, and * denotes p < 0.1. The number of observations is rounded to the nearest multiple of 10. SES refers to socioeconomic status.

Noncognitive skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, under unadjusted to fully adjusted models

	Self-control (reported by teachers) models				Approaches to learning (reported by teachers) models
	1 (unadjusted)	2	3	4 (fully adjusted)	1 (unadjusted)	2	3	4 (fully adjusted)
Gap in 1998	0.394***	0.304***	0.217***	0.182***	0.630***	0.630***	0.493***	0.435***
	(0.025)	(0.037)	(0.037)	(0.038)	(0.024)	(0.035)	(0.036)	(0.037)
Change in gap by 2010	-0.009	0.065	0.078	0.085	-0.117***	-0.066	-0.042	-0.043
	(0.037)	(0.054)	(0.060)	(0.061)	(0.035)	(0.053)	(0.057)	(0.057)
Controls
Demographics	No	No	Yes	Yes	No	No	Yes	Yes
Education and engagement	No	No	Yes	Yes	No	No	Yes	Yes
Parental expectations	No	No	No	Yes	No	No	No	Yes
School fixed effects	No	Yes	Yes	Yes	No	Yes	Yes	Yes
Observations	29,500	29,500	25,080	25,080	31,260	31,260	26,460	26,460
Adjusted R2	0.019	0.117	0.173	0.175	0.040	0.117	0.199	0.204

	Self-control (reported by parents) models				Approaches to learning (reported by parents) models
	1 (unadjusted)	2	3	4 (fully adjusted)	1 (unadjusted)	2	3	4 (fully adjusted)
Gap in 1998	0.467***	0.424***	0.357***	0.291***	0.539***	0.479***	0.215***	0.132***
	(0.025)	(0.036)	(0.039)	(0.040)	(0.025)	(0.032)	(0.033)	(0.033)
Change in gap by 2010	-0.076**	-0.084	-0.032	0.001	0.024	-0.024	0.096*	0.112**
	(0.037)	(0.054)	(0.060)	(0.061)	(0.036)	(0.053)	(0.055)	(0.056)
Controls
Demographics	No	No	Yes	Yes	No	No	Yes	Yes
Education and engagement	No	No	Yes	Yes	No	No	Yes	Yes
Parental expectations	No	No	No	Yes	No	No	No	Yes
School fixed effects	No	Yes	Yes	Yes	No	Yes	Yes	Yes
Observations	30,400	30,400	27,220	27,220	30,420	30,420	27,240	27,240
Adjusted R2	0.022	0.037	0.075	0.079	0.035	0.057	0.218	0.228

Reductions in skills gaps between high-SES and low-SES children after accounting for missingness and covariates, 1998 and 2010

	Year	Reduction	Change in reduction from 1998 to 2010 (in percentage points)
	Year	Reduction
Reading	1998	45.5%
	2010	42.9%	-2.6
Math	1998	52.6%
	2010	48.6%	-4.1
Self-control (reported by teachers)	1998	50.8%
	2010	32.6%	-18.1
Approaches to learning (reported by teachers)	1998	28.3%
	2010	20.3%	-8
Self-control (reported by parents)	1998	35.3%
	2010	34.3%	-1.1
Approaches to learning (reported by parents)	1998	73.5%
	2010	56.0%	-17.5

Note: SES refers to socioeconomic status. Declining values from 1998 to 2010 indicate that factors such as early literacy activities and other controls were not as effective at shrinking SES-based gaps in 2010 as they were in 1998.

Summary of association between cognitive and noncognitive skills at kindergarten entry and selected early educational practices, fully adjusted differences (Model 4)

	Reading	Math	Self-control (reported by teachers)	Approaches to learning (reported by teachers)	Self-control (reported by parents)	Approaches to learning (reported by parents)
Correlations between selected practices and skills measured at kindergarten entry in 1998
Center-based pre-K	0.106***	0.097***	-0.125***	-0.001	-0.006	0.018
	(0.016)	(0.015)	(0.018)	(0.018)	(0.019)	(0.016)
Number of books	0.012***	0.016***	0.004**	0.008***	0.002	0.006***
	(0.002)	(0.002)	(0.002)	(0.002)	(0.002)	(0.002)
Reading/literacy	0.166***	0.068***	0.010	0.030*	0.143***	0.315***
	(0.016)	(0.015)	(0.018)	(0.016)	(0.018)	(0.017)
Other activities	-0.115***	-0.036***	0.047***	0.033**	0.046***	0.292***
	(0.015)	(0.014)	(0.017)	(0.016)	(0.017)	(0.016)
Correlations between parents’ expectations about their children’s highest level of educational attainment and skills measured at kindergarten entry in 1998
Two or more years of college/vocational school	0.029	0.066**	0.072*	0.115***	0.180***	0.136***
	(0.025)	(0.026)	(0.042)	(0.037)	(0.038)	(0.033)
Bachelor’s degree	0.114***	0.172***	0.141***	0.211***	0.272***	0.228***
	(0.023)	(0.023)	(0.036)	(0.032)	(0.036)	(0.030)
Master’s degree or more	0.160***	0.220***	0.120***	0.219***	0.254***	0.377***
	(0.026)	(0.025)	(0.039)	(0.034)	(0.036)	(0.033)
Changes from 1998 to 2010 in the correlations between selected practices and skills measured at kindergarten entry
Center-based pre-K	-0.005	-0.036	0.060*	-0.010	-0.020	0.010
	(0.025)	(0.025)	(0.032)	(0.031)	(0.031)	(0.026)
Number of books	0.002	-0.001	0.001	0.002	-0.002	0.004
	(0.003)	(0.002)	(0.003)	(0.003)	(0.003)	(0.002)
Reading/literacy	0.018	0.008	0.015	0.014	-0.079***	-0.173***
	(0.025)	(0.024)	(0.031)	(0.028)	(0.030)	(0.027)
Other activities	-0.008	-0.016	0.031	0.020	0.218***	0.265***
	(0.025)	(0.024)	(0.029)	(0.028)	(0.029)	(0.025)
Changes from 1998 to 2010 in the correlations between parents’ expectations about their children’s highest level of educational attainment and skills measured at kindergarten entry
Two or more years of college/vocational school	0.121**	0.106*	0.201**	0.204***	-0.030	0.151**
	(0.055)	(0.059)	(0.081)	(0.072)	(0.084)	(0.066)
Bachelor’s degree	0.139***	0.103**	0.136*	0.174***	-0.084	0.100
	(0.048)	(0.051)	(0.070)	(0.063)	(0.078)	(0.061)
Master’s degree or more	0.186***	0.117**	0.140*	0.189***	-0.041	0.076
	(0.052)	(0.054)	(0.074)	(0.066)	(0.081)	(0.063)
Observations	26,050	26,890	25,080	26,460	27,220	27,240
Adj.R2	0.293	0.336	0.175	0.204	0.079	0.228

Notes: The robust standard errors are in parentheses. For statistical significance, *** denotes p < 0.01, ** denotes p < 0.05, and * denotes p < 0.1. The number of observations is rounded to the nearest multiple of 10.

Cognitive and noncognitive skills gaps between high-SES and low-SES children using mother's educational attainment as a proxy for socioeconomic status (SES), under unadjusted and fully adjusted models

	Reading		Math		Self-control (reported by teachers)		Approaches to learning (reported by teachers)		Self-control (reported by parents)		Approaches to learning (reported by parents)
	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)
Gap in 1998	1.294***	0.696***	1.457***	0.681***	0.317***	0.076	0.638***	0.409***	0.471***	0.254***	0.655***	0.221***
	(0.038)	(0.058)	(0.036)	(0.050)	(0.039)	(0.048)	(0.038)	(0.042)	(0.039)	(0.049)	(0.039)	(0.045)
Change in gap by 2010	-0.020	-0.075	-0.154***	-0.119*	-0.099*	0.046	-0.237***	-0.141*	-0.136**	-0.093	-0.084	-0.004
	(0.051)	(0.082)	(0.049)	(0.070)	(0.055)	(0.081)	(0.053)	(0.074)	(0.053)	(0.080)	(0.053)	(0.070)
Controls
Demographics	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Education and engagement	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Parental expectations	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
School fixed effects	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Observations	26,660	23,880	27,570	24,710	25,790	23,170	27,200	24,380	27,280	25,040	27,290	25,050
Adjusted R2	0.134	0.282	0.166	0.328	0.009	0.172	0.029	0.199	0.017	0.079	0.032	0.223

Notes: Model 1 uses the full sample; Model 4 uses the complete cases sample. Robust standard errors are in parentheses. For statistical significance, *** denotes p < 0.01, ** denotes p < 0.05, and * denotes p < 0.1. The number of observations is rounded to the nearest multiple of 10.

Cognitive and noncognitive skills gaps between high-SES and low-SES children using household income as a proxy for socioeconomic status (SES), under unadjusted and fully adjusted models

	Reading		Math		Self-control (reported by teachers)		Approaches to learning (reported by teachers)		Self-control (reported by parents)		Approaches to learning (reported by parents)
	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)
Gap in 1998	1.090***	0.384***	1.308***	0.443***	0.419***	0.119**	0.603***	0.325***	0.443***	0.272***	0.436***	0.073
	(0.042)	(0.058)	(0.041)	(0.060)	(0.045)	(0.050)	(0.044)	(0.049)	(0.045)	(0.051)	(0.044)	(0.052)
Change in gap by 2010	-0.127**	-0.006	-0.230***	-0.060	0.049	0.228***	-0.128**	0.008	0.044	0.106	0.032	0.051
	(0.060)	(0.084)	(0.059)	(0.082)	(0.066)	(0.081)	(0.064)	(0.079)	(0.065)	(0.084)	(0.064)	(0.080)
Controls
Demographics	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Education and engagement	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Parental expectations	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
School fixed effects	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Observations	28,650	26,050	29,560	26,890	27,550	25,080	29,110	26,460	28,170	27,220	28,190	27,240
Adjusted R2	0.103	0.276	0.143	0.321	0.023	0.174	0.036	0.199	0.019	0.079	0.019	0.226

Cognitive and noncognitive skills gaps between high-SES and low-SES children using number of books child has in the home as a proxy for socioeconomic status, under unadjusted and fully adjusted models

	Reading		Math		Self-control (reported by teachers)		Approaches to learning (reported by teachers)		Self-control (reported by parents)		Approaches to learning (reported by parents)
	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)	1 (unadjusted)	4 (fully adjusted)
Gap in 1998	0.736***	0.347***	0.966***	0.424***	0.324***	0.105***	0.455***	0.241***	0.283***	0.117***	0.583***	0.136***
	(0.028)	(0.034)	(0.027)	(0.031)	(0.029)	(0.035)	(0.028)	(0.033)	(0.029)	(0.037)	(0.028)	(0.033)
Change in gap by 2010	0.083**	-0.540***	-0.019	-0.818***	-0.068	-0.126	-0.058	-0.244	-0.044	-0.248	0.085**	-0.026
	(0.039)	(0.184)	(0.038)	(0.188)	(0.042)	(0.225)	(0.041)	(0.184)	(0.041)	(0.216)	(0.039)	(0.178)
Controls
Demographics	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Education and engagement	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Parental expectations	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
School fixed effects	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No	Yes
Observations	29,060	26,050	29,920	26,890	27,730	25,080	29,350	26,460	30,200	27,220	30,220	27,240
Adjusted R2	0.080	0.270	0.120	0.314	0.012	0.172	0.024	0.194	0.009	0.075	0.047	0.226

'Whole-child' case study initiatives, by service area

Part of school district	Entire school district	Across multiple school districts
Austin, Texas	Joplin, Missouri	Eastern Kentucky*
Boston, Massachusetts	Kalamazoo, Michigan
Durham, North Carolina (East Durham)	Montgomery County, Maryland*
Minneapolis, Minnesota (North Minneapolis)	Pea Ridge, Arkansas
New York, New York	Vancouver, Washington**
Orange County, Florida (Tangelo Park)

*Indicates that while the initiative covers the entire county or region, a portion of the county or region receives more intensive services. **Indicates that the initiative will cover the entire school district under plans to expand.

Source: Case studies published on the Broader, Bolder Approach to Education website (www.boldapproach.org/case-studies)

1. Values are in 2008 dollars.

2. Early investments in education strongly predict adolescent and adult development (Cunha and Heckman 2007; Heckman 2008; Heckman and Kautz 2012). For instance, students with higher levels of behavioral skills learn more in school than peers whose attitudinal skills are less developed (Jennings and DiPrete 2010). In general, as Heckman asserted, “skills beget skills,” meaning that creating basic, foundational knowledge makes it easier to acquire skills in the future (Heckman 2008). Conversely, children who fail to acquire this early foundational knowledge may experience some permanent loss of opportunities to achieve to their full potential. Indeed, scholars have documented a correlation between lack of kindergarten readiness and not reading well at third grade, which is a key point at which failing to read well greatly reduces a child’s odds of completing high school (Fiester 2010; Hernandez 2011).

3. Research by Reardon (2011) had found systematic increases in income gaps among generations. Recent studies by Bassok and Latham (2016) and Reardon and Portilla (2016), however, show narrower achievement gaps at kindergarten entry between a recent cohort and the previous one, and thus a possible discontinuation or interruption of that trend. (Bassok et al. [2016] use an SES construct to compare relative teacher assessments of cognitive and behavioral skills among low-SES children versus all children, adjusted by various other characteristics; Reardon and Portilla [2016] look at relative performance of children in the 90th and 10th income percentiles, and use age-adjusted, standardized, outcome scores.) Research by Carnoy and García (2017) shows persistent social-class gaps, but no solid evidence regarding trends: their findings for students in the fourth and eighth grades, in math and reading, show that achievement gaps neither shrink nor grow consistently (they are a function of the social-class indicator, the grade level, or the subject).

4. Clustering takes into account the fact that children are not randomly distributed, but tend to be concentrated in schools or classrooms with children of the same race, social class, etc. These estimates offer an estimate of gaps within schools. See Appendix B for more details.

5. Results available upon request. See García 2015 for results for all SES-quintiles (the baseline or unadjusted gaps in that report correspond with Model 2 in this paper).

6. The Early Childhood Longitudinal Study asks both parents and teachers to rate children’s abilities across a range of these skills. The specific skills measured may vary between the home and classroom setting. Teachers likely evaluate their students’ skills levels relative to those of other children they teach. Parents, on the other hand, may be basing their expectations on family, community, culture, or other factors.

7. See García 2015 for a discussion of which factors in children’s early lives and their individual and family characteristics (in addition to social class) drive the gaps among children of the 2010 kindergarten class.

8. Note that the SES quintiles are constructed using each year’s distribution, and that changes in the overall and relative distribution may affect the characteristics of children in the different quintiles each year (i.e., there may be some groups who are relatively overrepresented in one or another quintile if changes in the SES components changed over time).

9. The detailed frequency with which parents develop or practice some activities with their children at home and others is available upon request.

10. Literature on expectations and on parental behaviors in the home find that they positively correlate with children’s cognitive development and outcomes (Simpkins, Davis-Kean, and Eccles 2005; Wentzel, Russell, and Baker 2016). This literature acknowledges the multiple pathways through which expectations and behaviors influence educational outcomes, as well as the importance of race, social class, and other factors as moderators of such associations (Davis-Kean 2005; Redd et al. 2004; Wentzel, Russell, and Baker 2016; Yamamoto and Holloway 2010).

11. This may be affected by the fact that the highest number of reported books in 1998 was “more than 200,” while in 2010 parents could choose from more categories, up to “more than 1,000.” We had to use 200 as our cap in order to compare data for the two kindergarten classes.

12. Evidence also points to many other factors that affect children’s school readiness, and these, too, likely changed over this time period. For example, access to prenatal care, health screenings, and nutritional programs could all have affected children’s development differently across these two cohorts, but we do not have access to these data and thus cannot control for them in our study. For links between school readiness, children’s health, and poverty, see AAP COCP 2016; Currie 2009; U.S. HHS and U.S. ED 2016.

13. Models include all quintiles in their specification. Tables that offer a comparison for all quintiles relative to the first quintile are available upon request. We focus the discussion on the gap between the top and bottom.

14. As a result, sample sizes become smaller (see Appendix Table C1). Assuming “missingness” (observations without full information) is completely at random, the findings are representative of the original sample and of the populations they represent. Analytic samples once missingness is accounted for are called the complete case samples. We tested to see whether the unadjusted gaps estimated above with the full sample remained the same when using the complete case samples. For Model 1, we found an average difference of 0.01 sd in the estimates of 1998 SES gaps, and an average difference of 0.02 sd in the estimates of the change in the gaps. For Model 2, the differences were 0.01 sd for the gaps’ estimates and 0.04 for changes in the gaps’ estimates. In terms of statistical significance, there are no significant changes in the estimates associated with the 1998 gaps, but there are two changes in the statistical significance of the estimates associated with the changes in the gaps by 2010 – 2011, and one change in the magnitude of the coefficient. The first change in the statistical significance of the estimates associated with the changes in the gaps by 2010 – 2011 is the change in the gap in approaches to learning as reported by parents, which is statistically significant when using the restricted sample (0.07 sd, at the 10 percent significance level, Model 1); and the second is the change in the gap in math which also becomes statistically significant when using the restricted sample (0.09, at the 10 percent significance level, Model 2). Finally, the one change in the magnitude of the coefficient, in this model, is the estimate of the change in the gap in reading, which increases when using the restricted sample (from 0.12 sd to 0.18 sd). Results are available upon request.

15. These interactions between inputs and time test for whether the influence of inputs in 2010 is smaller than, the same as, or larger than the influence of inputs in 1998. Also, although only the fully specified results are shown, as noted in Appendix B, these sets of controls are entered parsimoniously in order to determine how sensitive gaps and changes in gaps over time are to the inclusion of family characteristics only, to the added inclusion of family investments, and, finally, to the inclusion of parental expectations (for the inclusion of parental expectations, we incorporated interactions of the covariates with time parsimoniously as well). For all outcomes, and focusing on the models without interactions between covariates and time, we find that all gaps in 1998 continuously shrink as we add more controls. For example, in reading, adding family characteristics reduces the gap in 1998 by 11 percent, adding investments further reduces it by 15 percent, and adding expectations further reduces it by 9 percent. In math, these changes equal to 16 percent, 13 percent, and 10 percent. For changes in the gap by 2010–2011, for both reading and math, adding family characteristics and investments shrink the changes in the gaps, but adding expectations slightly increases the estimated coefficients (which are statistically significant for reading, but not for math in these models. For self-control (as reported by teachers) and approaches to learning (by parents), which are the only two noncognitive skills for which the change in the gap is statistically significant, adding family characteristics reduces the change in the “gap [by 2010–2011” coefficient], but adding investments increases it, and adding expectations further increases the changes in the gaps by 2010–2011. These results are not shown in the appendices, but are available upon request.

16. The interactions between parental expectations of children’s educational attainment and the time variable test for whether the influence of expectations in 2010 is smaller, the same, or larger, than the influence of expectations in 1998.

17. The change in the skills gaps by SES in 2010 due to the inclusion of the controls is not directly visible in the tables in this report. To see this, see the comparison of estimates of models MS1–MS3 in García 2015. The change in the skills gaps by SES in 1998 is directly observable in Tables 3 and 4 and is discussed below.

18. The numbers in the “Reduction” column in Table 5 (showing the shares of the SES-based skills gaps that are accounted for by controls) are always higher for 1998 than for 2010.

19. Please note that until this point in the report we have been concerned with SES gaps and not with performance directly (though SES gaps are the result of the influence of SES on performance, which leads to differential performance of children by SES and hence to a performance gap). The paragraphs above emphasize how controls mediate or explain some of the skills gaps by SES, so, in a way, controls inform our analysis of gaps because they reveal how changes in gaps may have been affected by changes in various factors’ capacity to influence performance. Now the focus is on exploring the independent effect of the covariates of interest on performance. In this report, because we address whether the education and selected practices affect outcomes, the main effect is measured for the 1998 cohort, and we measure how it changed between 1998 and 2010. The detailed discussion for the correlation between covariates and outcomes in 2010 is provided in Table 3 in García 2015.

20. This variable indicates whether the child was cared for in a center-based setting during the year prior to the kindergarten year, compared with other options (as explained in García 2015, these alternatives include no nonparental care arrangements; being looked after by a relative, a nonrelative, at home or outside; or a combination of options. Any finding associated with this variable may be interpreted as the association between attending prekindergarten programs, compared with other options, but must be interpreted with caution. In other words, the child may have attended a high-quality prekindergarten program, which could have been either private or public, or a low-quality one, which would have different impacts. He or she might have been placed in (noneducational) child care, either private or public, of high or low quality, for few or many hours per day, with very different implications for his or her development (Barnett 2008; Barnett 2011; Magnuson et al. 2004; Magnuson, Ruhm, and Waldfogel 2007; Nores and Barnett 2010). For the extensive literature explaining the benefits of pre-K schooling, see Camilli et al. 2010, and for a meta-analysis of results, see Duncan and Magnuson 2013. Thus, more detailed information on the characteristics of the nonparental care arrangements (type, quality, and quantity) would help researchers further disentangle the importance of this variable. This additional information would provide a much clearer picture of the effects of early childhood education on the different educational outcomes.

21. Because these associations seemed counterintuitive, we tested whether they were sensitive to the composition of the index. We removed one component of the index at a time and created five alternative measures of other enrichment activities that parents do with their children. The results indicate that the negative association between the index and reading is not sensitive to the components of the index (the coefficients for the main effect, i.e., for the effect in 1998 range between -0.14 and -0.09, are all statistically significant). For math, the associations lose some precision, but retain the negative sign (negative association) in four out of the five cases (minimum coefficient is -0.06). As a caveat, these components do not reflect whether the activities are undertaken by the child or guided by the adult, the time devoted to them, or how much they involve the use of vocabulary or math concepts. The associations could indicate that time spent on nonacademic activities detracts from parents’ time to spend on activities that are intended to boost their reading and math skills, among other possible explanations. These results are available upon request.

22. Note that in this section, “social class” and “socioeconomic status” (SES) are treated as equivalent terms; in the rest of the report, we refer to SES as a construct that is one measure of social class. See Appendices C and D for discussions of two other sensitivity analyses, one based on imputation of missing values for the main analysis in this paper, and the other on the utilization of various metrics of the cognitive variables. Overall, our findings were not sensitive to various multiple imputation tests. In terms of the utilization of different metrics for the cognitive variables, some sensitivity of the point estimates was detected.

23. With certain activities that are already so provided to high-SES children, there may be little room for doing more for them. For example, there are only 24 hours per day to read to your child, so there is a cap on reading from a cap on time. But perhaps there is still room to improve the influence of reading, if, for example, the way reading is done changes.

24. Eight of the 12 districts explored in this paper are the subjects of published case studies. Case studies for the other four are in progress and will be published later this year. When citing information from the published case studies, we cite the specific published study. For the four that are not yet published, we refer to the original sources being used to develop the case studies.

25. Missing or incomplete cells in the table indicate that data were not available on that aspect of student demographics or other characteristics. As per the source note, most data came either from the districts’ websites or from NCES.

26. In the country as a whole, poverty rates, which had been rising prior to 2007, sped up rapidly during the recession and in its aftermath (through 2011–2012), and minority students (mainly Hispanic and Asian) grew as a share of the U.S. public school student body. Between 2000 and 2013, even with a decline in the proportion of black students, the share of the student body that is minority (of black or Hispanic origin) increased from 30.0 percent to 40.5 percent, and the proportion of low-income students (those eligible for free or reduced-price lunch) also increased, up from 38.3 percent of all public school students in 2000 to 52.0 percent in 2013 (Carnoy and García 2017). The Southern Education Foundation revealed a troubling tipping point in 2013: for the first time since such data have been collected, over half of all public school students (51 percent) qualified for free or reduced-priced meals (i.e., over half of students were living in households at or below 185 percent of the federal poverty line). Across the South, shares were much higher, with the highest percentage, 71 percent—or nearly three in four students—in Mississippi (Southern Education Foundation 2015).

27. A full cross-cutting analysis of why and how these districts have employed whole-child/comprehensive educational approaches will be published as part of a book that draws on these case studies.

28. The federal Early Head Start (EHS) program includes both a home visiting and a center-based component, with many of the low-income infants and toddlers served benefiting from a combination of the two. Studies of EHS find improved cognitive, behavioral, and emotional skills for children as well as enhanced parenting behaviors.

29. According to one important source for data on access to and quality of state pre-K programs, the State of Preschool yearbook produced annually by the National Institute for Early Education Research (NIEER) at Rutgers University, as of 2015, 42 states and the District of Columbia were funding 57 programs. Moreover, programs continued to recover from cuts made during the Great Recession; enrollment, quality, and per-pupil spending were all up, on average, compared with the year before, albeit with the important caveat that two major states—Texas and Florida—lost ground, and that “[f]or the nation as a whole,…access to a high-quality preschool program remained highly unequal, and this situation is unlikely to change in the foreseeable future unless many more states follow the leaders” (NIEER 2016).

30. Elaine Weiss interview with Joshua Starr, June 2017.

31. Murnane and Levy 1996; Elaine Weiss interview with Joshua Starr, June 2017.

32. In recent years, a growing number of reports have emerged that some charter schools—which are technically public schools and often tout their successes in serving disadvantaged students—keep out students unlikely to succeed through complex application processes, fees, parent participation contracts, and other mechanisms, and then further winnow the student body of such students by pushing them out when they struggle academically or behaviorally. For more on this topic, see Burris 2017, PBS NewsHour 2015, and Simon 2013.

33. See AIR 2011 and Sparks 2017. The federal school improvement models, in order of severity (from lightest to most stringent) are termed “transformation,” “turnaround,” “restart,” and “closure” (AIR 2011, 3).

34. While the cut score on any given assessment/test needed for a student to be considered “proficient” is an arbitrary one, and, in Minnesota and many other states, changes from year to year and from one assessment to another, these gains are a helpful indicator of program effectiveness, as they are comparable over the time period described.

35. Joplin statistics are from internal data produced for the superintendent at that time that are no longer available.

36. Attendance Works , a national campaign to reduce chronic absence, points to a range of studies that document and explain the connections between chronic absenteeism, student physical and mental health, and student achievement. Areas of research include elementary school absenteeism, middle and high school absenteeism, health issues, and state and local data on how these problems play out, among others.

37. Elaine Weiss interview with C.J. Huff, June 2016.

38. See Appendix D for a discussion of results using other metrics for reading and math achievement. Results are not meaningfully different across metrics, though the point estimates differ slightly.

39. This last feature will be explored in a companion paper to this one, as soon as the necessary information is released by NCES. (As Tourangeau et al. [2013] note, the assessment scores for the 2010–2011 cohort are not directly comparable with those for the 1998–1999 cohort. We are waiting on the availability of this data to conduct a companion study that allows us to learn whether starting levels of knowledge rose over these years, and what the relative gains were for different demographic groups.)

40. We acknowledge that there are multiple noneducation public policy and economic policy areas to be called upon to address the problems studied in this report, namely, all the ones that ensure other factors that correlate with low-SES are attended, and, obviously, the ones that lead to fewer low-SES children. These other policies could help ensure that more children grow up in contexts with sufficient resources and healthy surroundings, or would leave fewer children without built-in supports at home that need to be compensated for afterwards. We made these points in two early studies, and in the policy brief companion to this study (García 2015; García and Weiss 2015; García and Weiss 2017). A similar comprehensive approach in terms of policy recommendations was used by Putnam (2015).

AAP Council on Community Pediatrics (AAP COCP). 2016. “Poverty and Child Health in the United States.” Pediatrics vol. 137, no. 4. pii:e20160339.

Adamson, Frank, and Linda Darling-Hammond. 2012. “Funding Disparities and the Inequitable Distribution of Teachers: Evaluating Sources and Solutions.” Education Policy Analysis Archives vol. 20 (November), 37.

Alvarez, Lizette. 2015. “ One Man’s Millions Turn a Community in Florida Around .” New York Times , May 25.

American Institutes for Research (AIR). 2011. School Turnaround: A Pocket Guide .

Austin Independent School District (AISD). 2017. “ Pre-K 4 ” (section on the AISD website).

Baker, Bruce D., and Sean P. Corcoran. 2012. The Stealth Inequities of School Funding . The Center for American Progress.

Barbarin, O.A., J. Downer, E. Odom, and D. Head. 2010. “Home–School Differences in Beliefs, Support, and Control during Public Pre-Kindergarten and Their Link to Children’s Kindergarten Readiness.” Early Childhood Research Quarterly vol. 25, no. 3, 358–72.

Barnett, W. Steven. 2008. Preschool Education and Its Lasting Effects: Research and Policy Implications . Great Lakes Center for Education Research and Practice.

Barnett, W. Steven. 2011. “Effectiveness of Early Educational Intervention.” Science vol. 333, no. 6045, 975–78. doi:10.1126/science.1204534.

Barnett, W. Steven, Elizabeth Votruba-Drzal, Eric Dearing, and Megan E. Carolan. 2017. “Publicly Supported Early Care and Education Programs.” In The Wiley Handbook of Early Childhood Development Programs, Practices, and Policies , Elizabeth Votruba-Drzal and Eric Dearing, eds. Malden, Mass., and Oxford: John Wiley.

Bassok, Daphna, Jenna E. Finch, RaeHyuck Lee, Sean F. Reardon, and Jane Waldfogel. 2016. “Socioeconomic Gaps in Early Childhood Experiences: 1998 to 2010.” AERA Open vol. 2, no. 3.

Bassok, Daphna, and Scott Latham. 2016. “ Kids Today: Changes in School-Readiness in an Early Childhood Era .” EdPolicyWorks Working Paper Series no. 35.

Berea College. 2013. “ U.S. Secretary of Education Visits First Rural Promise Neighborhood ” (news release). November 12.

Bradbury, Bruce, Miles Corak, Jane Waldfogel, and Elizabeth Washbrook. 2015. Too Many Children Left Behind: The U.S. Achievement Gap in Comparative Perspective. New York: Russell Sage Foundation.

Brooks-Gunn, Jeanne, and Lisa Markman. 2005. “The Contribution of Parenting to Ethnic and Racial Gaps in School Readiness.” Future of Children vol. 15, no. 1, 139–68.

Bivens, Josh. 2016. Progressive Redistribution without Guilt. Using Policy to Shift Economic Power and Make U.S. Incomes Grow Fairer and Faster . Economic Policy Institute.

Bivens, Josh, Emma García, Elise Gould, Elaine Weiss, and Valerie Wilson. 2016. It’s Time for an Ambitious National Investment in America’s Children: Investments in Early Childhood Care and Education Would Have Enormous Benefits for Children, Families, Society, and the Economy . Economic Policy Institute.

Boston College Center for Optimized Student Support. 2012. The Impact of City Connects: Progress Report 2012 .

Boston College Center for Optimized Student Support. 2014. The Impact of City Connects: Progress Report 2014 .

Burris, Carol. 2017. “ What the Public Isn’t Told about High Performing Charter Schools in Arizona .” Washington Post Answer Sheet blog, March 30.

Camilli, Gregory, Sadako Vargas, Sharon Ryan, and W. Steven Barnett. 2010. “Meta-Analysis of the Effects of Early Education Interventions on Cognitive and Social Development.” Teachers College Record vol. 112, no. 3, 579–620.

Carnoy, Martin, and Emma García. 2017. Five Key Trends in U.S. Student Performance. Progress by Blacks and Hispanics, the Takeoff of Asians, the Stall of Non-English Speakers, the Persistence of Socioeconomic Gaps, and the Damaging Effect of Highly Segregated Schools . Economic Policy Institute.

Carter, Prudence L., and Kevin G. Welner, eds. 2013. Closing the Opportunity Gap: What America Must Do to Give Every Child an Even Chance . New York: Oxford Univ. Press.

Caspe, Margaret, and Joy Lorenzo Kennedy. 2014. Sustained Success: The Long-Term Benefits of High Quality Early Childhood Education. New York: Children’s Aid Society.

Chaudry, Ajay, Taryn Morrissey, Christina Weiland, and Hirokazu Yoshikawa. 2017. Cradle to Kindergarten: A New Plan to Combat Inequality . New York: Russell Sage Foundation.

Chetty, Raj, David Grusky, Maximilian Hell, Nathaniel Hendren, Robert Manduca, and Jimmy Narang. 2016. “ The Fading American Dream: Trends in Absolute Income Mobility since 1940 .” NBER Working Paper no. 22910.

Child Trends. 2014. Making the Grade: Assessing the Evidence for Integrated Student Supports .

Clark, H., et al. 2009. Study Comparing Children’s Aid Society Community Schools to Other New York City Public Schools (All Schools and Peer Schools ). ActKnowledge.

Coleman, J.S., E. Campbell, C. Hobson, J. McPartland, A. Mood, F. Weinfeld, and R. York. 1966. Equality of Educational Opportunity . Washington, D.C.: U.S. Office of Education.

Collaborative for Academic, Social, and Emotional Learning (CASEL). 2017. “ Partner Districts: Austin ” (webpage). Accessed August 31, 2017.

Cook-Harvey, C.M., L. Darling-Hammond, L. Lam, C. Mercer, and M. Roc. 2016. Equity and ESSA: Leveraging Educational Opportunity Through the Every Student Succeeds Act . Palo Alto, Calif.: Learning Policy Institute.

Cunha, Flavio, and James J. Heckman. 2007. “The Technology of Skill Formation.” American Economic Review vol. 97, no. 2, 31–47.

Currie, Janet. 2009. “Healthy, Wealthy, and Wise: Socioeconomic Status, Poor Health in Childhood, and Human Capital Development.” Journal of Economic Literature vol. 47, no. 1, 87–122.

Davis-Kean, Pamela E. 2005. “The Influence of Parent Education and Family Income on Child Achievement: The Indirect Role of Parental Expectations and the Home Environment.” Journal of Family Psychology vol. 19, no. 2 (June 2005), 294–304. doi:10.1037/0893-3200.19.2.294.

Duncan, Greg J., Chantelle J. Dowsett, Amy Claessens, Katherine A. Magnuson, Aletha C. Huston, Pamela Klebanov, Linda S. Pagani, Leon Feinstein, Mimi Engel, and Jeanne Brooks-Gunn. 2007. “School Readiness and Later Achievement.” Developmental Psychology vol. 43, no. 6, 1428–46.

Duncan, Greg J., and Katherine A. Magnuson. 2011. “The Nature and Impact of Early Achievement Skills, Attention Skills, and Behavior Problems.” In Whither Opportunity?: Rising Inequality, Schools, and Children’s Life Chances , Greg J. Duncan and Richard Murnane, eds., 47–69. New York: Russell Sage Foundation.

Duncan, Greg J., and Katherine Magnuson. 2013. “Investing in Preschool Programs.” Journal of Economic Perspectives vol. 27, no. 2, 109–32.

Duncan, Greg J., Pamela A. Morris, and Chris Rodrigues. 2011. “Does Money Really Matter? Estimating Impacts of Family Income on Young Children’s Achievement with Data from Random-Assignment Experiments.” Developmental Psychology vol. 47, no. 5, 1263–79. doi:10.1037/a0023875.

Duncan, Greg J., and Richard Murnane. 2011. “Introduction: The American Dream, Then and Now.” In Whither Opportunity?: Rising Inequality, Schools, and Children’s Life Chances , Greg J. Duncan and Richard Murnane, eds. New York: Russell Sage Foundation.

Economic Policy Institute (EPI). 2012. “ The Great Recession .” State of Working America feature.

Economic Policy Institute (EPI). 2013. “ Inequality.is ” (interactive website).

Elmore, Richard, David Thomas, and Tonika Cheek Clayton. 2006. Differentiated Treatment in Montgomery County Public Schools . Public Education Leadership Project at Harvard University.

Fiester, Leila. 2010. Early Warning! Why Reading by the End of Third Grade Matters. KIDS COUNT Special Report . Annie E. Casey Foundation.

García, Emma. 2015. Inequalities at the Starting Gate: Cognitive and Noncognitive Skills Gaps between 2010–2011 Kindergarten Classmates . Economic Policy Institute.

García, Emma, and Elaine Weiss. 2015. Early Education Gaps by Social Class and Race Start U.S. Children Out on Unequal Footing. A Summary of the Major Findings in Inequalities at the Starting Gate . Economic Policy Institute.

García, Emma, and Elaine Weiss. 2016. Making Whole-Child Education the Norm: How Research and Policy Initiatives Can Make Social and Emotional Skills a Focal Point of Children’s Education . Economic Policy Institute.

García, Emma, and Elaine Weiss. 2017. Key Findings from the Report “Education Inequalities at the School Starting Gate” . Economic Policy Institute.

Hart, Betty, and Todd R. Risley. 1995. Meaningful Differences in the Everyday Experience of Young American Children . Baltimore, Md.: Brookes.

Heckman, James J. 2008. “Schools, Skills, and Synapses.” Economic Inquiry vol. 46, no. 3, 289–324.

Heckman, James J., and Tim Kautz. 2012. “Hard Evidence on Soft Skills.” Labour Economics vol. 19, no. 4, 451–64.

Henderson, Anne T. 2010. Community Organizing to Build Partnerships in Schools: The Alliance Schools Movement in Austin . Annenberg Institute for School Reform.

Hernandez, Donald J. 2011. Double Jeopardy: How Third-Grade Reading Skills and Poverty Influence High School Graduation . Annie E. Casey Foundation.

Gizriel, Sarah. 2016. “ Bright Futures Looking to Expand to Schools across Shenandoah Valley .” localDVM.com , December 9.

Jennings, J.L., and T.A. DiPrete. 2010. “Teacher Effects on Social and Behavioral Skills in Early Elementary School.” Sociology of Education vol. 83, no. 2, 135.

Kalamazoo Public Schools (KPS). 2017. “ PEEP Information and Applications ” (webpage). Accessed August 31, 2017.

Lee, Valerie E., and David T. Burkam. 2002. Inequality at the Starting Gate . Washington, D.C.: Economic Policy Institute.

Levin, Henry M. 2012a. “More Than Just Test Scores.” Prospects vol. 42, no. 3, 269–84.

Levin, Henry M. 2012b. “The Utility and Need for Incorporating Noncognitive Skills into Large-scale Educational Assessments.” In The Role of International Large-Scale Assessments: Perspectives from Technology, Economy, and Educational Research , Matthias von Davier et al., eds. Springer.

Magnuson, Katherine, and Greg J. Duncan. 2016. “Can Early Childhood Interventions Decrease Inequality of Economic Opportunity?” RSF: The Russell Sage Foundation Journal of the Social Sciences vol. 2, no. 2, 123–41.

Magnuson, Katherine A., M.K. Meyers, C.J. Ruhm, and Jane Waldfogel. 2004. “Inequality in Preschool Education and School Readiness.” American Educational Research Journal vol. 41, no. 1, 115–57.

Magnuson, Katherine A., Christopher Ruhm, and Jane Waldfogel. 2007. “Does Prekindergarten Improve School Preparation and Performance?” Economics of Education Review vol. 26, no. 1, 33–51.

Marietta, Geoff. 2010. Lessons for PreK-3rd from Montgomery County Public Schools: An FCD Case Study . Foundation for Child Development.

Maryland State Department of Education. 2017. “ Judy Centers ” (webpage). Accessed August 31, 2017.

Miller-Adams, Michelle. 2015. Promise Nation: Transforming Communities through Place-Based Scholarships . Kalamazoo, Mich.: W.E. Upjohn Institute for Employment Research.

Mishel, Lawrence. 2015. “ The Opportunity Dodge .” American Prospect , April 9.

Mishel, Lawrence, Josh Bivens, Elise Gould, and Heidi Shierholz. 2012. The State of Working America, 12th Edition , An Economic Policy Institute Book. Ithaca, N.Y.: Cornell Univ. Press.

Mishel, Lawrence, and Jessica Schieder. 2016. Stock Market Headwinds Meant Less Generous Year for Some CEOs . Economic Policy Institute.

Montgomery County Public Schools (MCPS). 2015. Graduation Rate Rises, Gap Narrows for MCPS Class of 2014 (public announcement). January 27.

Montgomery County Public Schools (MCPS). 2016. Linkages to Learning (brochure).

Montgomery County Public Schools (MCPS). 2017. “ Maryland Meals for Achievement ” (webpage). Accessed August 31, 2017.

Morsy, Leila, and Richard Rothstein. 2015. Five Social Disadvantages That Depress Student Performance: Why Schools Alone Can’t Close Achievement Gaps . Economic Policy Institute.

Murnane, Richard J., and Frank Levy. 1996. Teaching the New Basic Skills: Principles for Educating Children to Thrive in a Changing Economy . New York: The Free Press.

Murnane, Richard J., John B. Willett, Kristen L. Bub, and Kathleen McCartney. 2006. “Understanding Trends in the Black-White Achievement Gaps during the First Years of School.” Brookings-Wharton Papers on Urban Affairs.

Najarian, M., K. Tourangeau, C. Nord, K. Wallner-Allen, and J. Leggitt. Forthcoming. Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011), First-Grade and Second-Grade Psychometric Report . Washington, D.C.: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

National Center for Education Statistics (NCES) (U.S. Department of Education). Early Childhood Longitudinal Study, Kindergarten Class of 1998–99 (ECLS-K 1998–1999) .

National Center for Education Statistics (NCES) (U.S. Department of Education). Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K 2010–2011) .

National Institute for Early Education Research (NIEER). 2016. The State of Preschool 2015: State Preschool Yearbook .

New York Times /CBS News. 2015. “ Americans’ Views on Income Inequality and Workers’ Rights ” (poll results). June 3.

Nores, Milagros, and W. Steven Barnett. 2010. “Benefits of Early Childhood Interventions across the World: (Under) Investing in the Very Young.” Economics of Education Review vol. 29, no. 2, 271–82.

Nores, Milagros, and W. Steven Barnett. 2014. Access to High Quality Early Care and Education: Readiness and Opportunity Gaps in America . New Brunswick, N.J.: Center on Enhancing Early Learning Outcomes.

Nores, Milagros, and Emma García. 2014. “Language, Immigration and Hispanics. Understanding Achievement Gaps in the Early Years.” Paper presented at the Association for Public Policy Analysis and Management Fall Research Conference, November 6–8, Albuquerque, N.M.

Oakes, Jeannie, Anna Maier, and Julia Daniel. 2017. Community Schools: An Evidence-Based Strategy for Equitable School Improvement , Learning Policy Institute, June 5.

PBS NewsHour . 2015. “ In Reforming New Orleans, Have Charter Schools Left Some Students Out? ” (news segment).

Peterson, T.K., ed. 2013. Expanding Minds and Opportunities: Leveraging the Power of Afterschool and Summer Learning for Student Success . Washington, D.C.: Collaborative Communications Group.

Phillips, Meredith. 2011. “Parenting, Time Use, and Disparities in Academic Outcomes.” In Whither Opportunity?: Rising Inequality, Schools, and Children’s Life Chances , Greg J. Duncan and Richard Murnane, eds. New York: Russell Sage Foundation.

Proctor, Bernadette D., Jessica L. Semega, and Melissa A. Kollar. 2016. Income and Poverty in the United States: 2015 . U.S. Census Bureau, Current Population Reports, P60-256(RV).

Putnam, Robert. 2015. Our Kids: The American Dream in Crisis . New York: Simon and Schuster.

Ready, Douglas D. 2010. “Socioeconomic Disadvantage, School Attendance, and Early Cognitive Development.” Sociology of Education vol. 83, no. 4, 271–86.

Reardon, Sean F. 2007. “ Thirteen Ways of Looking at the Black-White Test Score Gap .” Working paper.

Reardon, Sean F. 2011. “The Widening Academic Achievement Gap between the Rich and the Poor: New Evidence and Possible Explanations.” In Whither Opportunity?: Rising Inequality, Schools, and Children’s Life Chances , Greg J. Duncan and Richard Murnane, eds. New York: Russell Sage Foundation.

Reardon, Sean F., and Ximena A. Portilla. 2016. “Recent Trends in Income, Racial, and Ethnic School Readiness Gaps at Kindergarten Entry.” AERA Open vol. 2, no. 3, 1–18. doi: 10.1177/2332858416657343.

Redd, Z., L. Guzman, L. Lippman, L. Scott, and G. Matthews. 2004. Parental Expectations for Children’s Educational Attainment: A Review of the Literature . Prepared by Child Trends for the National Center for Education Statistics.

Rolnick, Art, and Rob Grunewald. 2003. “Early Childhood Development: Economic Development with a High Public Return.” The Region vol. 17, no. 4, 6–12.

Rothstein, Richard. 2004. Class and Schools: Using Social, Economic, and Educational Reform to Close the Achievement Gap . Washington, D.C.: Economic Policy Institute; New York: Columbia University Teachers College.

Rothstein, Richard. 2010. “Family Environment in the Production of Schooling.” In International Encyclopedia of Education , Dominic J. Brewer, Patrick J. McEwan, eds. Oxford: Elsevier. doi: 10.1016/B978-0-08-044894-7.01233-1.

Saez, Emmanuel. 2016. Striking It Richer: The Evolution of Top Incomes in the United States (Updated with 2015 Preliminary Estimates) .

Schanzenbach, Diane, Megan Mumford, Ryan Nunn, and Lauren Bauer. 2016. Money Lightens the Load . The Hamilton Project, Brookings Institute.

Selzer, Michael H., Ken A. Frank, and Anthony S. Bryk. 1994. “The Metric Matters: The Sensitivity of Conclusions about Growth in Student Achievement to Choice of Metric.” Educational Evaluation and Policy Analysis vol. 16, 41–49.

Sharkey, Patrick. 2013. Stuck in Place: Urban Neighborhoods and the End of Progress toward Racial Equality . Chicago, Ill.: Univ. of Chicago Press.

Simon, Stephanie. 2013. “ Class Struggle: How Charter Schools Get Students They Want .” Reuters , February 15.

Simpkins, Sandra D., Pamela E. Davis-Kean, and Jacquelynne S. Eccles. 2005. “Parents’ Socializing Behavior and Children’s Participation in Math, Science, and Computer Out-of-School Activities.” Applied Developmental Science vol. 9, no. 1, 14–30. doi:10.1207/s1532480xads0901_3.

Southern Education Foundation. 2015. A New Majority: Low-Income Students Now a Majority in the Nation’s Public Schools . January.

Sparks, Sarah D. 2017. “Billions in School Improvement Spending but Not Much Student Improvement.” EdWeek , January 19.

StataCorp. 2015. Stata: Release 14 [statistical software]. College Station, Texas: StataCorp LP.

Stringhini, Silvia, et al. 2017. “Socioeconomic Status and the 25×25 Risk Factors as Determinants of Premature Mortality: A Multicohort Study and Meta-Analysis of 1.7 Million Men and Women.” The Lancet . Published online January 31, 2017. doi:10.1016/S0140-6736(16)32380-7.

Tourangeau, K., C. Nord, T. Lê, A.G. Sorongon, and M. Najarian. 2009. Early Childhood Longitudinal Study, Kindergarten Class of 1998–99 (ECLS-K): Combined User’s Manual for the ECLS-K Eighth-Grade and K–8 Full Sample Data Files and Electronic Codebooks (NCES 2009-004) . U.S. Department of Education. Washington, D.C.: National Center for Education Statistics.

Tourangeau, K., C. Nord, T. Lê, A.G. Sorongon, M.C. Hagedorn, P. Daly, and M. Najarian. 2013. Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011), User’s Manual for the ECLS-K:2011 Kindergarten Data File and Electronic Codebook (NCES 2013-061). U.S. Department of Education. Washington, D.C.: National Center for Education Statistics.

Tourangeau, K., C. Nord, T. Lê, K. Wallner-Allen, M.C. Hagedorn, J. Leggitt, and M. Najarian. 2015. Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011), User’s Manual for the ECLS-K:2011 Kindergarten–First Grade Data File and Electronic Codebook, Public Version (NCES 2015-078) . U.S. Department of Education. Washington, D.C.: National Center for Education Statistics.

Tourangeau, K., C. Nord, T. Lê, K. Wallner-Allen, N. Vaden-Kiernan, L. Blaker, and M. Najarian. 2017. Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011) User’s Manual for the ECLS-K:2011 Kindergarten–Second Grade Data File and Electronic Codebook, Public Version (NCES 2017-285) . U.S. Department of Education. Washington, D.C.: National Center for Education Statistics.

U.S. Department of Education (U.S. ED). 2015. A Matter of Equity: Preschool in America .

U.S. Department of Health and Human Services (U.S. HHS) and U.S. Department of Education (U.S. ED). 2016. Policy Statement to Support the Alignment of Health and Early Learning Systems .

Van Voorhis, F.L., M.F. Maier, J.L. Epstein, C.M. Lloyd, and T. Leung. 2013. The Impact of Family Involvement on the Education of Children Ages 3 to 8: A Focus on Literacy and Math Achievement Outcomes and Social-Emotional Skills . MDRC.

Waldfogel, Jane. 2006. “What Do Children Need?” Public Policy Research vol. 13, no. 1, 26–34.

Weiss, Elaine. 2016a. Bright Futures in Joplin, Missouri . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016b. Vancouver Public Schools (Vancouver, WA) . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016c. Partners for Education at Berea College, Berea, Kentucky . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016d. Northside Achievement Zone (North Minneapolis, MN) . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016e. East Durham Children’s Initiative (East Durham, NC). A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016f. Bright Futures (Pea Ridge, AR) . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016g. City Connects (Boston, MA) . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016h. The Children’s Aid Society Community Schools (New York, NY) . A Broader, Bolder Approach to Education.

Weiss, Elaine. 2016i. A Broader, Bolder Education Policy Framework . A Broader, Bolder Approach to Education.

Wentzel, Kathryn R., Shannon Russell, and Sandra Baker. 2016. “Emotional Support and Expectations from Parents, Teachers, and Peers Predict Adolescent Competence at School.” Journal of Educational Psychology vol. 108, no. 2, 242–255.

Yamamoto, Yoko, and Susan D. Holloway. 2010. “Parental Expectations and Children’s Academic Performance in Sociocultural Context.” Educational Psychology Review vol. 22, no. 3, 189–214. doi:10.1007/s10648-010-9121-z.

Appendix A. Data

Introduction.

Our research benefits from the existence of two companion studies conducted by the National Center for Education Statistics (NCES), the Early Childhood Longitudinal Study of the Kindergarten Class of 1998–1999 and the Early Childhood Longitudinal Study of the Kindergarten Class of 2010–2011 (hereafter, ECLS-K 1998–1999 and ECLS-K 2010–2011). The data from these studies come with multiple advantages and a few disadvantages.

The studies follow two nationally representative samples of children starting in their kindergarten year and continuing through their elementary school years (eighth grade for 1998–1999 cohort and fifth grade for the 2010–2011 cohort). The tracking of students over time is one of the most valuable features of the data. The studies include assessments of the children’s cognitive performance and knowledge as well as skills that belong in the category of noncognitive, or social and emotional, skills. The studies also include information on teachers and schools (provided by teachers and administrators) and interviews with parents.

Another valuable feature of the data is the availability of two ECLS-K studies (ECLS-K 1998–1999 and ECLS-K 2010–2011), which allows for cross-comparisons “of two nationally representative kindergarten classes experiencing different policy, educational, and demographic environments” (Tourangeau et al. 2013). The two studies are 12 years apart, or a full school cycle apart: when the 2010–2011 kindergarten class was starting school, the 1998–1999 class was starting the grade leading to their graduation. A comparison of the studies thus offers insightful information about the consequences of changes in the system that may have occurred during an entire cohort’s school life. For the 2010 study, the sample included 18,174 children in 968 schools. i The 1998 study sample included 21,409 children in 903 schools. ii

This existence of data from two cohorts is also a limitation to the current study, as explained by Tourangeau et al. (2013), who note that the assessment scores for the 2010–2011 class are not directly comparable with those developed for the class of 1998–1999. Although the IRT (Item Response Theory) procedures used in the analysis of data were similar across the two studies, each study incorporated different items, which means that the resulting scales are different. Tourangeau et al. (2013) state that “a subsequent release of the ECLS-K: 2010–2011 data will include IRT scores that are comparable with the ECLS-K 1998 cohort.” Up to the point of publication of the current study, this information had not yet been released, and we use standardized scores, instead of raw scores, for the outcomes examined. We can assess changes in the relative position in a distribution (i.e., how far apart high- and low-SES children are in 1998 and how far apart high- and low-SES children are in 2010), but not overall changes in their performance (i.e., it is not possible to ascertain whether performance has improved overall, or if gaps are smaller or larger due to an improvement in performance of children at the low end (specifically the lowest fifth) of the distribution or due to a decrease in the performance of children at the high end (highest fifth) of the distribution, etc.). A full comparison remains to be produced, upon data availability.

We use data for the first wave of each study, corresponding with fall kindergarten (or school entry).

For the analyses, we use the by-year standardized scores corresponding to the fall semester. (The 1998 IRT scale scores for reading and mathematics achievement and assessments of noncognitive skills are standardized using the 1998 distribution and its mean and sd; for 2010, we use the mean and sd of the 2010 distribution.)

Cognitive skills

Cognitive skills are assessed with instruments that measure each child’s:

Reading skills: print familiarity, letter recognition, beginning and ending sounds, rhyming words, word recognition, vocabulary knowledge, and reading comprehension
Math skills: conceptual knowledge, procedural knowledge, and problem-solving; number sense, properties, and operations; measurement; geometry and spatial sense; data analysis, statistics, and probability; and patterns, algebra, and functions

Principal noncognitive skills

We use the term “principal” to identify a set of noncognitive skills that are measured by both the ECLS-K 1998–1999 and 2010–2011 surveys, and that have been relatively extensively used in research.

Teachers are asked to assess each child’s:

Self-control: ability to control behavior by respecting the property rights of others, controlling temper, accepting peer ideas for group activities, and responding appropriately to pressure from peers
Approaches to learning: organizational skills (keeps belongings organized); curiosity (is eager to learn new things); independence (works independently); adaptability (easily adapts to changes in routine); persistence in completing tasks; focus (ability to pay attention); and ability to follow classroom rules

Parents are asked to assess their child’s:

Self-control: ability to control behavior by refraining from fighting, arguing, throwing tantrums, and getting angry
Approaches to learning: persistence (keeps working at something until finished); curiosity (shows interest in a variety of things); focus (concentrates on a task and ignores distractions); helpfulness (helps with chores); intellectual curiosity (is eager to learn new things); and creativity (in work and play)

For the analyses, we use the following set of covariates. The definitions, and the coding used for the covariates, by year, are shown in Appendix Table A1 .

Appendix B. Methodology

Gaps by socioeconomic status.

The expressions below show the specifications used to estimate the socioeconomic status–based (SES-based) performance gaps. For any achievement outcome A , we estimate four models:

Model 1 shows the unadjusted (descriptive) differences for children belonging to different racial/ethnic groups or SES quintiles (the reference group is children in the lowest SES quintile, “low SES”).
Model 2 adjusts for school clustering of students in different schools (i.e., gaps of students in the same schools). The purpose of this clustering is to account for school segregation (i.e., concentration of children of the same race, socioeconomic status, etc., in schools, which causes the raw average performance of students to differ from the adjusted-by-clustering average). It offers a comparison of the gaps shown by peer students in the same schools and classrooms (García 2015; Magnuson and Duncan 2016 offer these estimates as well).

These estimates build on all the available observations (i.e., only those children who have missing values in the outcome variables are eliminated from the analysis).

Because of lack of response in some of the covariates used as predictors of performance, we construct a common sample with observations with no missing information in any of the variables of interest (see information about missing data for each variable in Appendix Table C1 ). We estimate two more models: iii

Model 3 shows gaps adjusted for child and family characteristics, prekindergarten care arrangements, number of books the child has, and early literacy practices at home iv
Finally, Model 4 shows the fully adjusted differences (adjusted for child and family characteristics, prekindergarten care arrangements, early literacy practices at home, number of books the child has, and parental expectations)

The equation below shows the equation we estimate for Models 1 through 4.

$A_{i, s}^{c,nc}= \delta_{o}+\delta_{1}SES2_{i,s} +\delta_{2}SES3_{i,s}+\delta_{3}SES4_{i,s}+\delta_{4}SES5_{i,s} +\delta_{5}Year2010_{i,s}+\delta_{6}Year2010xSES2_{i,s}+\delta_{7}Year2010xSES3_{i,s}+\delta_{7}Year2010xSES4_{i,s}+\delta_{8}Year2010xSES5_{i,s}+Controls+\alpha_{s}+\epsilon_{i,s}$

Appendix C. Sensitivity analysis (I): Multiple imputation

Following standard approaches in this field, we use multiple imputation to impute missing values in both the independent and dependent variables, for the analysis of skills gaps and changes in them from 1998 to 2010 by socioeconomic status (main analysis). See share of missing data by variable in Appendix Table C1 . We use the mi commands in Stata 14, using chained equations, which jointly model all functional terms. The number of iterations was set up equal to 20. Imputation is performed by year.

Our functional form of the imputation model is specified using SES, gender, race, disability, age, type of family, number of books, educational activities, and parental expectations, as well as the original cognitive and noncognitive variables, as variables to be imputed. We use various specifications, combining different sets of auxiliary variables, mi impute methods, and other parameters, to capture any sensitivity of the results to the characteristics of the model. For example, income, family size, and ELL status are set as auxiliary variables and used in several of the imputation models. Another imputation option that was altered across models is the use of weights, as we ran out of imputation models using weights and not using them.

In the imputation model, in order to impute categorical variables’ missingness, we use the option augment, to prevent the large number of categorical variables to be imputed from causing problems of perfect prediction (StataCorp. 2015). The rest of the variables are first imputed as continuous variables. In a second exercise, we also impute SES and educational expectations as ordinal variables (also using the option augment).

In order to calculate the standardized dependent variables, we use the variables derived from the imputation variables (also known as passive imputation). This “fills in only the underlying imputation variables and computes the respective functional terms from the imputed variables” (StataCorp. 2015). In one case, we imputed the dependent variables directly as continuous variables (though we anticipated that the distribution of the scores imputed this way would not necessarily have a mean of 0 and a standard deviation of 1).

Using the imputed data, we estimate Models 1 through 4 following the specifications explained above (from no regressors to fully specified models).

The main findings of our analysis are not sensitive to missing data imputation. The estimates of the gaps in 1998 and the changes in the gaps from 1998 to 2010 are consistent across models in terms of statistical significance. There are some minor changes in the sizes of the estimated coefficients, especially those associated with the changes in the gaps (though all are statistically not different from 0, as discussed in the report using the results from the analysis with the complete cases). There are also some minor changes in the standard errors, though they are small enough to widen the coefficients’ statistical bandwidth to not include the 0.

Appendix D. Sensitivity analysis (II): The different scores available in ECLS-K and the sensitivity of the results to changing them

Children’s reading and mathematics skills are measured using several different metrics in ECLS-K. Among these, the best-known or more commonly used metrics in research are the IRT-based theta scores and the IRT-based scale scores (IRT stands for Item Response Theory). NCES provides data users with definitions of these metrics and recommendations on how to appropriately choose among the different metrics. NCES explains that both theta and IRT-based scale scores are valid indicators of ability. This makes them suitable for research purposes, even though each is expressed in its own unit of measurement. NCES recommends that analysts “consider the nature of their research questions, the type of statistical analysis to be conducted, the population of interest, and the audience” when choosing the appropriate score for analysis (see Tourangeau et al. 2013).

Although nothing would indicate that this could be the case, our work noted that results of analyses such as the one developed in this study are in some ways sensitive to the metrics used as dependent variables. v Thus, the purpose of this appendix is to illustrate the differences in the results associated with different analytic decisions in terms of the metrics used. As we will see, in essence, point estimates depend on the metric used, but the results do not change in a meaningful way and conclusions and implications remain unchanged. That is, although caution is required when interpreting the results obtained using different combinations of metrics, procedures (including standardization), and data waves, it is important to state that the main conclusions of this study— that social-class gaps in cognitive and noncognitive skills are large and have persisted over time — hold . So do the policy recommendations derived from those findings: sufficient, integrated, and sustained over-time efforts to tackle early gaps in a more effective manner.

The scores: Which one to use and definitions

NCES makes the following recommendations for researchers who are choosing among scales (see Tourangeau et al. 2013): vi

When choosing scores to use in analysis, researchers should consider the nature of their research questions, the type of statistical analysis to be conducted, the population of interest, and the audience. […] The IRT-based scale scores […] are overall measures of achievement. They are appropriate for both cross-sectional and longitudinal analyses. They are useful in examining differences in overall achievement among subgroups of children in a given data collection round or in different rounds, as well as in analysis looking at correlations between achievement and child, family, and school characteristics. […] Results expressed in terms of scale score points, scale score gains, or an average scale score may be more easily interpretable by a wider audience than results based on the theta scores. The IRT-based theta scores are overall measures of ability. They are appropriate for both cross-sectional and longitudinal analyses. They are useful in examining differences in overall achievement among subgroups of children in a given data collection round or across rounds, as well as in analysis looking at correlations between achievement and child, family, and school characteristics. […] The theta scores may be more desirable than the scale scores for use in a multivariate analysis because generally their distribution tends to be more normal than the distribution of the scale scores. However, for a broader audience of readers unfamiliar with IRT modeling techniques, the metric of the theta scores (from -6 to 6) may be less readily interpretable. […]

The two scores are defined as follows (see Tourangeau et al. 2013, section “3.1 Direct Cognitive Assessment: Reading, Mathematics, Science”):

The IRT-based scale score is an estimate of the number of items a child would have answered correctly in each data collection round if he or she had been administered all of the questions for that domain that were included in the kindergarten and first-grade assessments. To calculate the IRT-based overall scale score for each domain, a child’s theta is used to predict a probability for each assessment item that the child would have gotten that item correct. Then, the probabilities for all the items fielded as part of the domain in every round are summed to create the overall scale score. Because the computed scale scores are sums of probabilities, the scores are not integers. The IRT-based theta score is an estimate of a child’s ability in a particular domain (e.g., reading, mathematics, science, or SERS) based on his or her performance on the items he or she was actually administered. […] The theta scores are reported on a metric ranging from -6 to 6, with lower scores indicating lower ability and higher scores indicating higher ability. Theta scores tend to be normally distributed because they represent a child’s latent ability and are not dependent on the difficulty of the items included within a specific test.

Reardon (2007) describes the calculation of the theta scores in the following manner: vii

For each test [math and reading], a three-parameter IRT model was used to estimate each student’s latent ability…at each wave…. The IRT model assumes that each student’s probability of answering a given test item correctly is a function of the student’s ability and the characteristics [discrimination, difficulty, and guessability] of the item…. Given the pattern of students’ responses to the items on the test that they are given, the IRT model provides estimates of both the person-specific latent abilities at each wave… and the item parameters. (Reardon 2007, 10) viii

He also notes that “[b]ecause the ECLS-K tests contain many more ‘difficult’ items than ‘easy’ items, the relationship between theta and scale scores is not linear (a unit difference in theta corresponds to a larger difference in scale scores at theta=1 than at theta=-1, for example). The scale scores are difficult to interpret as an interval-scale metric (or are an interval-scaled metric only with respect to the specific set of items on the ECLS-K tests),” while he shows that the “theta scores are interval-scale metrics, in a behaviorally-meaningful sense” (Reardon 2007, 11, 13). ix

The analyses

For the analyses, both the scale and the theta scores need to be standardized by year (the original variables are not directly comparable because they rely on different instruments, as explained by NCES, and the resulting standardized variables have mean 0 and standard deviation 1). This is a common practice in the education field, as it allows researchers to use data that come from different studies and would not have a common scale otherwise. We need to take into consideration that the underlying units of measurement for each variable are different, but after standardization, the metrics are common, expressed in standard deviations and represent the population’s distribution of abilities.

The distributions of the scale and theta scores are shown in Appendix Figures D1 and D2 . In each figure, the plots reflect a more normally distributed pattern for the theta scores (right panel) than for the scale scores (left panel). The companion table, Appendix Table D1 , shows the range of variation for the four outcomes (mean and standard deviations are 0 and 1 as per construction).

We next offer a comparison of the results obtained when using the scale scores versus using the theta scores ( Appendix Table D2 ). We highlight the following main similarities and differences between the results obtained using the scale scores and the results using the theta scores.

Gaps are all equally statistically significant and persistent.
For example, looking at the unadjusted estimates in reading, the gap in 1998 between high- and low-SES children is 1.071 sd if using the scale scores and 1.233 sd if using the theta scores. In math, the gap between high- and low-SES children in 1998 is 1.258 sd if using the scale scores and 1.330 sd if using the theta scores.
Looking at the adjusted estimates in reading, the 1998 gap between high- and low-SES children is 0.596 sd if using the scale scores and 0.684 sd if using the theta scores. In math, the gap between high- and low-SES children is 0.610 sd if using the scale scores and 0.632 sd if using the theta scores.
For example, looking at the unadjusted estimates in reading, the change in the gap between 1998 and 2010 for high- and low-SES children is 0.098 sd if using the scale scores and -0.052 sd (not statistically significant) if using the theta scores. In math, the change in the gap between high- and low-SES children is -0.008 sd (not statistically significant) if using the scale scores and -0.078 sd if using the theta scores.

In Appendix Table D3 , we compare the results obtained using the different scales and the different proxies of socioeconomic status (our composite SES index, mother’s education, number of books, and household income).

Gaps are larger, as mentioned above, when we use the theta scores than when we use the scale scores.
Among the four social-class proxies, the largest gaps are associated with mother’s education, and the smallest gaps are associated with number of books. All are statistically significant.
Looking at the unadjusted gaps, we note that trends are the same (and similar in size) if income is used as the proxy. For mother’s education, the change in the gap between 1998 and 2010 is -0.020 sd in reading (not statistically significant) and -0.154 sd in math if using the scale scores and -0.135 sd in reading and -0.218 sd in math if using the theta scores.
With respect to the adjusted gaps, changes in the gaps are larger when using the theta scores both for household income and mother’s education as indicators of social class. Using the theta scores, the gaps in reading and math shrank over time, while using the scale scores, the only significant reduction was in math when mother’s education was the social class proxy.

Other considerations

There are two other significant pieces of information affecting the cognitive scores in more recent documentation released by NCES. In 2015, NCES announced in its ECLS-K User’s Manual that a

change in methodology required a re-calibration and re-reporting of the kindergarten reading scores since the release of the base-year file. Therefore, the kindergarten reading theta scores included in the K-1 data file are calculated differently than the previously released kindergarten theta scores and replace the kindergarten reading theta scores included in the base-year data file. The modeling approach stayed the same for mathematics and science, so the recalculation of kindergarten mathematics and science theta scores was not needed. (Tourangeau et al. 2015)

Following up on this, the most recent (2017) data user’s manual explains that

The method used to compute the theta scores allows for the calculation of theta for a given round that will not change based on later administrations of the assessments (which is not true for the scale scores, as described in the next section). Therefore, for any given child, the kindergarten, first-grade, and second-grade theta scores provided in subsequent data files will be the same as theta scores released in earlier data files , with one exception: the reading thetas provided in the base-year data file . After the kindergarten-year data collection, the methodology used to calibrate and compute reading scores changed; therefore, the reading thetas reported in the base-year file are not the same as the kindergarten reading thetas provided in the files with later-round data [emphasis added]. Any analysis involving kindergarten reading theta scores and reading theta scores from later rounds, for example an analysis looking at growth in reading knowledge and skills between the spring of kindergarten and the spring of first grade, should use the kindergarten reading theta scores from a data file released after the base year. The reading theta scores released in the kindergarten-year data file are appropriate for analyses involving only the kindergarten round data; analyses conducted with only data released in the base-year file are not incorrect, since those analyses do not compare kindergarten scores to scores in later rounds that were computed differently. However, now that the recomputed kindergarten theta scores are available in the kindergarten through first-grade and kindergarten through second-grade data files, it is recommended that researchers conduct any new analyses with the recomputed kindergarten reading theta scores. For more information on the methods used to calculate theta scores, see the ECLS-K: 2011 First-Grade and Second-Grade Psychometric Report (Najarian et al. forthcoming). (Tourangeau et al. 2017)

Therefore, because of these changes in NCES methodology and reporting, and in light of the comparisons in this appendix, one could expect additional slight changes in the estimates using the IRT-theta scores for reading for kindergarten if using rounds of data posterior to the first round (and probably if using the IRT-scale scores as well, as these values are derived from the theta scores), relative to the first data file of ECLS-K: 2010-2011 released by NCES in 2013. We would not necessarily expect, though, any changes when using the standardized transformation of those scores, because NCES’s documentation does not mention changes to the distribution of the scores, only to their values. We will explore these issues further upon the release of the scores that are comparable across the two ECLS-K studies without any transformation.

Appendix E. Descriptions of 12 community-level whole-child education initiatives

Initiatives that serve part of a school district, austin, texas.

The needs of children in Austin Independent School District (AISD) schools with the highest concentrations of poor, immigrant, and non-English-speaking families are supported through a combination of parent-organizing (schools with parent-organizing programs, led by the nonprofit Austin Interfaith, form a network of “Alliance Schools”), intensive embedding of social and emotional learning (SEL) in all aspects of school policy and practice, and the transformation of schools into “community schools” (i.e., schools that are hubs for the provision of academic, health, and social services).

Organizing partners: Austin Interfaith (a nonprofit of congregations, public schools, and unions that is part of the national Industrial Areas Foundation [IAF]); the Collaborative for Academic, Social and Emotional Learning (CASEL); the American Federation of Teachers (AFT); and the National Education Association (NEA).
Schools and students reached: The IAF/Alliance Schools network extended at its zenith into one-fourth of AISD elementary schools and one-half of AISD high-poverty elementary schools. CASEL worked in five high schools, and in the seven middle schools and 43 elementary schools that feed into these high schools, to embed social and emotional learning in school policies and practices. A middle school and a high school have been transformed into community schools and serve as the models for planned districtwide expansion of the “community schools” strategy into all AISD schools.
General makeup of the student body: In the district overall, 60 percent of students qualify for subsidized meals, i.e., are eligible for free or reduced-price lunch (FRPL); 28 percent are English language learners (ELL); and 10 percent are special education students. In schools targeted for whole-child supports, relative to the general student body, students are poorer, more heavily minority and immigrant, and more likely to be living in single-parent households.
Key features: Parent-organizing with teachers in Alliance Schools enables parents to partner with teachers to advocate for comprehensive supports for their children. Also, social and emotional learning (SEL) is embedded in all aspects of school efforts in the high schools and the feeder elementary and middle schools that worked with CASEL. Finally, health and other wraparound supports in high-needs middle and high schools, along with other community schools features, are expanding to additional district schools.
Core funding: The district received a CASEL grant to embed social and emotional learning in school policies and practices, and also received in-kind support from the NoVo Foundation in the form of technical assistance. The United Way of Greater Austin provides funds for wraparound support, and AFT and NEA fund community schools work and expansion.

Boston, Massachusetts

The City Connects program provides targeted academic, social, emotional, and health supports to every child in 20 of the city’s schools with the highest shares of low-income, black, Hispanic, and immigrant students.

Organizing partners: Boston College Center for Optimized Student Support, Boston Public Schools (BPS), and community agencies.
Schools and students reached: The 20 BPS schools in the program serve more than 8,000 of the city’s most disadvantaged students (out of 125 BPS schools and 56,000 students).
General makeup of the student body: The 20 urban schools serve neighborhoods that are poor and racially and ethnically diverse, with a heavy concentration of Hispanic English-language learners. Over 80 percent of the students in these schools are FRPL-eligible and roughly half do not speak English at home.
Key features: School site coordinators in each school connect students with a tailored set of services and enrichment opportunities provided by a variety of public and private agencies. Universal state health care supports all students’ physical and mental health needs, and the city’s Universal Pre-Kindergarten (UPK) program now offers quality pre-K for all four-year-olds in Boston.
Core funding: In addition to school district budget revenue, federal Race to the Top funds allocated to City Connects help defray costs. Several private foundations support various aspects of City Connects’ work.

Durham, North Carolina

The East Durham Children’s Initiative (EDCI) concentrates services and supports for the children and their families living in a 120-block, heavily distressed area of concentrated poverty and high crime within the city.

Organizing partners: Community leaders launched EDCI and engaged the Duke University Center for Child and Family Health to grow capacity. EDCI is now a fully staffed nonprofit that runs the initiative.
Schools and students reached: The 120-block area targeted by EDCI serves students in two neighborhood elementary schools, one middle school, one high school, and two charter schools.
General makeup of the student body: The 120-block area is urban and poor with a predominantly black but very diverse student body. In Durham schools overall, 66 percent of students are FRPL-eligible, nearly half are black, almost one-third are Hispanic, and 18 percent are white.
Key features: EDCI is a place-based initiative modeled on the Harlem Children’s Zone, providing a pipeline of high-quality cradle-to-college-or-career services. These include early childhood supports (that complement state pre-K programs), health and mental health services, and after-school and summer enrichment activities.
Core funding: EDCI has an annual fund receiving contributions from individuals, corporations, fundraising events, and private foundations; it neither seeks nor receives public funding.

Minneapolis, Minnesota

The Northside Achievement Zone (NAZ) is a Promise Neighborhood, a designation awarded by the U.S. Department of Education Promise Neighborhoods program to some of the most distressed neighborhoods in the nation. Through the program, children and families who live in the 13-by-18 block NAZ receive individualized supports.

Organizing partners: NAZ, the Promise Neighborhood grantee organization, is guided by a 20-member board of directors consisting of local leaders.
Schools and students reached: The 13-by-18 block zone in North Minneapolis serves 5,500 students in 10 public, charter, and parochial K–12 schools, including one high school.
General makeup of the student body: In this racially concentrated area of poverty, almost all residents are African American, and median family income is $18,000. One-third of children are homeless or “highly mobile” (not technically homeless but without stable housing).
Key features: “Connectors” are in essence case managers who help families develop achievement plans, and “Navigators” connect families with community resources to move toward goals. The zone offers access to high-quality pre-K and parenting supports, as well as mentoring, enrichment, college preparatory support, and after-school and summer programs.
Core funding: NAZ is anchored by a federal Promise Neighborhood grant. NAZ also receives private grants and is able to leverage federal Race to the Top Early Learning Challenge funds to support pre-K scholarship slots.

New York, New York

Through a collaboration between The Children’s Aid Society and the New York City Department of Education, 16 community schools in some of the most disadvantaged neighborhoods in three of the city’s five boroughs provide wraparound health, nutrition, mental health, and other services to students along with enriching in-and-out-of-school experiences, amplified by extensive parental and community engagement.

Organizing partners: The Children’s Aid Society, the New York City Department of Education, the New York State Education Department, and other local and state agencies.
Schools and students reached: Sixteen community schools in three boroughs serve some of the poorest immigrant and minority students in a school system of roughly one million students.
General makeup of the student body: Students in Children’s Aid Society community schools are disadvantaged relative to the system overall, which serves a heavily low-income and minority student body: more than three quarters of New York City public school students are FRPL-eligible, 13 percent are English language learners, and nearly one in five receive special education services. These schools also have high concentrations of students of color: 27 percent are African American and 41 percent are Hispanic.
Key features: Close coordination with local and state education, health, and other agencies along with community partnerships at each school enables wraparound health, mental health, and after-school and summer enrichment, as well as deep parental and community engagement.
Core funding: A range of public dollars, including federal Elementary and Secondary Education Act (ESEA) Title I funds and funds from the federal 21st Century Community Learning Centers program, together with state and local funding for after-school and other programs, is supplemented by funds from individuals and foundations.

Orange County, Florida

The Tangelo Park Project (TPP) provides cradle-to-college support for all children residing in Orlando’s high-poverty, heavily African American Tangelo Park neighborhood.

Organizing partners: The Tangelo Park Program board, along with Harris Rosen (the hotelier who envisioned and funds the program), work in close collaboration with the Tangelo Park Civic Association and the University of Central Florida.
Schools and students reached: The program serves all children in the Tangelo Park neighborhood.
General makeup of the student body: Virtually all residents in the low-income neighborhood are African American or Afro-Caribbean.
Key features: Universal college scholarships—called “Promise” scholarships because they are guaranteed by an established fund—are supported by quality neighborhood-based early childhood education, health, counseling, and after-school and summer programs.
Core funding: Harris Rosen funds early child care providers and universal college scholarships. Rosen also supports other services, such as a lifeguard at the YMCA, as needed.

Initiatives that serve all of a school district

Joplin, missouri.

Joplin’s Bright Futures initiative (which has spawned dozens of other Bright Futures affiliate districts under a Bright Futures USA umbrella since it launched in 2010) has a rapid response component that addresses children’s basic needs (within 24 hours of a need being reported), while strong school–community partnerships help meet students’ longer-term needs. Bright Futures also provides meaningful service learning opportunities in every school.

Organizing partners: The Joplin School District’s superintendent and top leadership, in collaboration with parents and community, faith, business, and social service leaders.
Schools and students reached: Bright Futures serves all of the district’s 7,874 students in all 17 schools.
General makeup of the student body: Joplin is a heavily white community. As of 2015, nearly two-thirds (61 percent) of Joplin students are FRPL-eligible and 16 percent are classified as needing special education; just 3 percent are English language learners.
Key features: The Bright Futures USA framework has three components. First, a rapid response system is designed to meet any student’s basic health, nutrition, or physical need within 24 hours of such a need being reported; this system is supported by combined resources from social service agencies, businesses, faith organizations, and individual community members. Second, school- and community-level councils build community leadership and partnerships with schools to meet longer-term needs and sustain systems. Third, service learning opportunities are embedded in all schools to help develop children as citizens. Teachers lead the service learning and receive training to do so. In addition to these three components, Joplin also provides pre-K for at-risk students, as well as tutoring, mentoring, and after-school and college preparatory programs based on student need.
Core funding: Federally funded Americorps VISTA volunteers provide in-kind support; funds from the state departments of Elementary and Secondary Education and of Economic Development support Bright Futures work and conferences; and the regional Economic Security Corporation and a range of private funders supplement these federal and state funding sources.

Kalamazoo, Michigan

The “Kalamazoo Promise,” a guarantee by a group of anonymous local philanthropists to provide full college scholarships in perpetuity for graduates of the district’s public high schools brought Kalamazoo Public Schools (KPS), the city, and the community together to develop a set of comprehensive supports that enable more students to use the scholarships.

Organizing partners: Kalamazoo Promise and Kalamazoo Public Schools, the local school district, in collaboration with Communities in Schools Kalamazoo (CIS) and other nonprofit entities.
Schools and students reached: All KPS students (12,216 in 25 schools) who graduate from Kalamazoo public high schools are eligible for Promise scholarships. CIS works in all schools but to varying degrees and with varying levels of financial support.
General makeup of the student body: In this combination urban–suburban district, a large majority of students (over 70 percent) are FRPL-eligible, 12 percent receive special education services, and 7 percent are English language learners. The share of African American students grew from less than one-third in 1987 to over half 30 years later; over this period the share of Hispanic students increased as well.
Key features: The anchor for comprehensive supports is universal “Promise” college scholarships, which have spurred community leadership to provide quality pre-K programs and wraparound health, mental health, and other supports, and to launch a districtwide effort to create a college-going culture and resources to support that culture.
Core funding: Anonymous donors have committed to funding Promise scholarships in perpetuity. CIS is supported by a combination of Title I funding, which helps support school coordinators; 21st Century Learning grants for after-school activities; and private individual and philanthropic donations.

Montgomery County, Maryland

All students in Montgomery County Public Schools (MCPS) benefit from zoning laws that advance integration and strong union–district collaboration on an enriching, equity-oriented curriculum. These efforts are bolstered by extra funding and wraparound supports for high-needs schools and communities.

Organizing partners: MCPS, Montgomery County Education Association (the local teachers union), Montgomery County Council, and Linkages to Learning (a joint initiative of MCPS and the county council that provides an integrated focus on health, social services, community development, and engagement to support student learning, strong families, and healthy communities.)
Schools and students reached: All 160,000 students in more than 200 schools are served via some services. Higher-poverty schools and their communities receive additional funds and supports that are broader and more intensive. For example, Linkages to Learning serves more than 5,400 individuals—students and their family members—per year at 29 schools. Over 3,700 of them receive comprehensive behavioral health or social wraparound services to mitigate the effects of poverty and reduce nonacademic barriers to learning.
General makeup of student body: The MCPS school district as a whole is racially and socioeconomically diverse: 30 percent of students are Hispanic, 29 percent are white, 22 percent are African American, 14 percent are Asian, and 35 percent are FRPL-eligible (more than 40 percent of students have been FRPL-eligible at some point). On the poorer, Eastern side of the county, where more intensive whole-child supports are provided, the 10 highest-poverty schools have student bodies that are at least 80 percent FRPL-eligible.
Key features: Mixed-use housing policies that enable racial and socioeconomic integration advance school-level integration that boosts low-income students’ learning, which the district enhances through various forms of support, including high-quality early childhood education, parent and community outreach, reallocation of funds to high-needs schools and students, nutrition and health services, and an emphasis on social and emotional learning.
Core funding: MCPS is heavily locally funded, with almost no federal Title I dollars. The district’s whole-child approach draws on a combination of school district and county revenues, along with federal funding for Head Start programs, state pre-K dollars, and assorted other grants.

Pea Ridge, Arkansas

The Pea Ridge School District, a small suburban–rural district outside Fayetteville, Arkansas, is among the newer affiliates of Bright Futures USA, a national umbrella group that grew out of Bright Futures Joplin. As a Bright Futures affiliate, Pea Ridge is making good progress toward identifying and meeting students’ basic needs, engaging the community to meet longer-term needs, and making service learning a core component of school policy and practice.

Organizing partners: Pea Ridge School District and Bright Futures USA.
Schools and students reached: Eight hundred and fifty students are served in one primary school, one elementary school, one middle school, and one high school, as well as an alternative high school and a new career-tech charter high school.
General makeup of the student body: The suburban–rural district is mostly white, with a small but growing Hispanic population, and predominantly middle-income with pockets of both higher-income families and families in poverty.
Key features: The first component of the three-part Bright Futures USA framework is a rapid response system to meet every student’s basic health, nutrition, and physical needs within 24 hours through a combination of social service agency, business, faith, and individual community contributions. Other components include school- and community-level councils, which build community leadership and partnerships with schools to meet longer-term needs and sustain systems, and service learning embedded in all schools that is enhanced by supportive training for teachers. Pea Ridge also provides pre-K for at-risk students, as well as tutoring, mentoring, and after-school and college preparatory programs for students who need them.
Core funding: State funds support meals and other needs for high-poverty schools, and Pea Ridge has secured a three-year private grant to support access to pre-K for low-income students.

Vancouver, Washington

Family and Community Resource Centers (FCRCs) currently serve 16 of the highest-needs Vancouver Public Schools (VPS) district schools, with mobile and lighter-touch support in other schools and plans to expand districtwide by 2020.

Organizing partners: School district leaders coordinate the program with the support of six central-office staff (three of whom just support FCRCs). Technical and other assistance is provided by the Coalition for Community Schools.
Schools and students reached: FCRCs serve 23,500 students in 16 VPS schools: 11 elementary schools, two middle schools, two high schools, and the Fruit Valley Learning Center (a combination elementary school and community center that also offers child care and Head Start programs). Plans are being made to expand FCRCs to all 35 VPS schools by 2020.
General makeup of the student body: As of 2015, more than half of students were FRPL-eligible, with FRPL-eligibility rates in some central-city schools exceeding 80 percent. More than one in five students speak a language other than English at home and 12.5 percent of students are special education students; in FCRC schools, the shares of non–English speakers and special education students are even higher.
Key services: VPS supports a range of early childhood education programs, including quality pre-K; middle and high school in-school enrichment; after-school and summer programs (provided by VPS partners); and help for parents and families through workshops, assistance, and referrals to a range of community resources.
Core funding: District and Title I funds, which support basic FCRC needs, are supplemented by cash and in-kind donations from faith-based, social service, business, and association partners.

Initiative that serves multiple school districts

Eastern (appalachian) kentucky.

A federal Promise Neighborhood grant helps Berea College’s Partners for Education provide intensive supports for students and their families in four counties in the Eastern (Appalachian) region of Kentucky and provide lighter-touch supports in an additional 23 surrounding counties. (Berea College, which was established in 1855 by abolitionist education advocates, is unique among U.S. higher-education institutions. It admits only economically disadvantaged, academically promising students, most of whom are the first in their families to obtain postsecondary education, and it charges no tuition, so every student admitted can afford to enroll and graduates debt-free.)

Organizing partners: Berea College launched Partners for Education (PfE), which is now a fully staffed nonprofit that runs the initiative.
Schools and students reached: PfE serves 35,000 students in 22 schools in Clay, Jackson, Knox, and Owsley counties; tens of thousands more are served less intensively in an additional 23 counties in the region.
General makeup of the student body: The Appalachian region is rural, very poor, and heavily white. The regional poverty rate is around 27 percent (in 2015), and reaches as high as 40 percent in some counties. About 80 percent of students are FRPL-eligible and 97 percent are white.
Key features: Family engagement specialists meet directly with families and help coordinate services provided by a range of community partners. Other specialists provide basic academic, college preparatory, and health and other wraparound services to students.
Core funding: Federal Promise Neighborhood, Full Service Community Schools, and Investing in Innovation grants are the most prominent sources of funding, but the initiative receives a range of other cash and in-kind supports.

Appendix tables and figures

Covariates from these models : ecls-k 1998--1999 and 2010--2011.

Source: ECLS-K, kindergarten classes of 1998–1999 and 2010–2011 (National Center for Education Statistics)

Missing data

	1998	2010
Variable	Percent missing	Percent missing
Race/ethnicity
White	0.2	0.5
Black	0.2	0.5
Hispanic	0.2	0.5
Hispanic English language learner (ELL)	6.6	11.8
Hispanic English speaker	6.6	11.8
Asian	0.2	0.5
Others	0.2	0.5
Socioeconomic status	5.9	11.9
Family composition: Not living with two parents	15.5	26.3
Mother’s education	7.5	42.8
Pre-K care, center-based	16.8	17.4
“Literacy/reading activities” index	15.6	26.4
“Other activities” index	15.6	26.5
Parents’ expectations for children’s educational attainment	16.1	26.5
Number of books	16.3	26.7
Outcomes
Reading	17.7	13.8
Math	13.0	14.2
Self-control (by teachers)	13.8	25.4
Approaches to learning (by teachers)	10.4	18.7
Self-control (by parents)	15.8	27.3
Approaches to learning (by parents)	15.8	27.3

Note: For detailed information about the construction of these variables, see Appendix Table A1.

Distribution of standardized scale and theta scores in mathematics, by year

Scale scores, 1998 (left) and 2010 (right).

Theta scores, 1998 (left) and 2010 (right)

Distribution of standardized scale and theta scores in reading, by year

Descriptive statistics of standardized scale and theta scores, by year (not weighted)

	1998				2010
	N	(Mean, sd)	Min	Max	N	(Mean, sd)	Min	Max
Scale score–reading	17,620	(0,1)	-1.39	10.13	15,670	(0,1)	-2.4	4.06
Theta score–reading	17,620	(0,1)	-2.72	4.30	15,670	(0,1)	-3.47	5.01
Scale score–math	18,640	(0,1)	-1.69	9.86	15,600	(0,1)	-2.22	4.23
Theta score–math	18,640	(0,1)	-3.13	4.48	15,600	(0,1)	-5.78	6.28

Note: N is rounded to the nearest multiple of 10.

Reading and math skills gaps between high-SES and low-SES children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, using scale and theta scores as dependent variables

	Model 1 (unadjusted)				Model 4 (fully adjusted)
	Full sample				Restricted sample
	Scale scores		Theta scores		Scale scores		Theta scores
	Reading	Math	Reading	Math	Reading	Math	Reading	Math
Gap in 1998	1.071***	1.258***	1.233***	1.330***	0.596***	0.610***	0.684***	0.632***
	(0.024)	(0.022)	(0.024)	(0.022)	(0.031)	(0.031)	(0.032)	(0.031)
Change in gap by 2010	0.098***	-0.008	-0.052	-0.078**	0.080	0.051	-0.016	-0.002
	(0.033)	(0.032)	(0.033)	(0.032)	(0.052)	(0.048)	(0.054)	(0.050)
N	30,950	31,850	30,950	31,850	26,050	26,890	26,050	26,890
Adj.R2	0.152	0.189	0.170	0.197	0.293	0.336	0.336	0.353

Notes: Standard errors are in the parentheses. N is rounded to the nearest multiple of 10. Asterisks denote statistical significance: *** p < 0.01, ** p < 0.05, * p < 0.1.

Source: ECLS-K, kindergarten classes of 1998-1999 and 2010–2011 (National Center for Education Statistics)

Reading and math skills gaps between high-social class and low-social class children at the beginning of kindergarten in 1998 and change in gaps by the beginning of kindergarten in 2010, using scale and theta scores as dependent variables

		Model 1 (unadjusted)				Model 4 (fully adjusted)
		Full sample				Restricted sample
		Scale scores		Theta scores		Scale scores		Theta scores
		Reading	Math	Reading	Math	Reading	Math	Reading	Math
By SES	Gap in 1998	1.071***	1.258***	1.233***	1.330***	0.596***	0.610***	0.684***	0.632***
		(0.024)	(0.022)	(0.024)	(0.022)	(0.031)	(0.031)	(0.032)	(0.031)
	Change in gap by 2010	0.098***	-0.008	-0.052	-0.078**	0.080	0.051	-0.016	-0.002
		(0.033)	(0.032)	(0.033)	(0.032)	(0.052)	(0.048)	(0.054)	(0.050)
By mother’s education	Gap in 1998	1.294***	1.457***	1.412***	1.502***	0.696***	0.681***	0.739***	0.685***
		(0.038)	(0.036)	(0.038)	(0.035)	(0.058)	(0.050)	(0.048)	(0.044)
	Change in gap by 2010	-0.020	-0.154***	-0.135***	-0.218***	-0.075	-0.119*	-0.135*	-0.182***
		(0.051)	(0.049)	(0.051)	(0.048)	(0.082)	(0.070)	(0.075)	(0.067)
By number of books	Gap in 1998	0.736***	0.966***	0.847***	1.032***	0.347***	0.424***	0.388***	0.438***
		(0.028)	(0.027)	(0.028)	(0.026)	(0.034)	(0.031)	(0.033)	(0.031)
	Change in gap by 2010	0.083**	-0.019	-0.015	-0.088**	-0.540***	-0.818***	-0.594***	-0.829***
		(0.039)	(0.038)	(0.039)	(0.038)	(0.184)	(0.188)	(0.181)	(0.174)
By household income	Gap in 1998	1.090***	1.308***	1.214***	1.320***	0.384***	0.443***	0.429***	0.439***
		(0.042)	(0.041)	(0.042)	(0.041)	(0.058)	(0.060)	(0.049)	(0.050)
	Change in gap by 2010	-0.127**	-0.230***	-0.247***	-0.292***	-0.006	-0.060	-0.058	-0.099
		(0.060)	(0.059)	(0.060)	(0.059)	(0.084)	(0.082)	(0.076)	(0.072)

Notes: Standard errors are in parentheses. Asterisks denote statistical significance: *** p < 0.01, ** p < 0.05, * p < 0.1.

Endnotes to the appendices

i. The sample design used to select the individuals in the study was a three-stage process that involved using primary sampling units and schools with probabilities proportional to the number of children and the selection of a fixed number of children per school. In the last stage, children enrolled in kindergarten or ungraded schools were selected within each sampled school. A clustered design was used to limit the number of geographic areas and to minimize the number of schools and the costs of the study (Tourangeau et al. 2013, 4-1).

ii. The dataset in the first year followed a stratified design structure (Ready 2010, 274), in which the primary sampling units were geographic areas consisting of counties or groups of counties. About 1,000 schools — 903 for 1998 and 968 for 2010—were selected, and about 24 children per school were surveyed. Assessment of the children was performed by trained evaluators, while parents were surveyed over the telephone. Teachers and school administrators completed the questionnaires in their schools.

iii. As a sensitivity check, we estimate Models 1 and 2 using Models 1’s and Model 2’s specifications but using the restricted sample (these results are not shown here, but are available upon request).

iv. As a sensitivity check, we estimate Model 3 parsimoniously, by including family characteristics only, and then adding family investments (prekindergarten care arrangements, early literacy practices at home, and number of books the child has), and then adding parental expectations (with and without interactions with time); results of the sensitivity check are not shown, but are available upon request).

v. We refer to the fact that we are using the same data and that the scale and theta scores are based on the same instruments and are not independent from each other. Advice on this possibility is found in Reardon (2007), who cites work by Murnane et al. (2006) and Selzer, Frank, and Bryk (1994) that also warn about this option.

vi. From NCES: “IRT uses the pattern of right and wrong responses to the items actually administered in an assessment and the difficulty, discriminating ability, and guess-ability of each item to estimate each child’s ability on the same continuous scale. IRT has several advantages over raw number-right scoring. By using the overall pattern of right and wrong responses and the characteristics of each item to estimate ability, IRT can adjust for the possibility of a low-ability child guessing several difficult items correctly. If answers on several easy items are wrong, the probability of a correct answer on a difficult item would be quite low. Omitted items are also less likely to cause distortion of scores, as long as enough items have been answered to establish a consistent pattern of right and wrong answers. Unlike raw number-right scoring, which treats omitted items as if they had been answered incorrectly, IRT procedures use the pattern of responses to estimate the probability of a child providing a correct response for each assessment question” (Tourangeau et al. 2017, 3-2).

vii. The quoted text is abridged to remove variables and formulas specific to Reardon’s study and not central here.

viii. Also, “the estimated scale score is the estimated number of questions the student would have gotten correct if he or she had been asked all of the items on the test. The estimated scale score is obtained by summing the predicted probabilities of a correct response over all items, given the student’s estimated theta score and the estimated item parameters” (Reardon 2007, 11).

ix. They are equally spaced units along the scale without a predefined zero point.

See more work by Emma García and Elaine Weiss

Chronicle Conversations
Article archives
Issue archives
Join our mailing list

Primary school math students in the MatiTec program in Santa Fe, Mexico City, 20 March 2012. Talento Tec. Wikimedia Commons

Recognizing and Overcoming Inequity in Education

About the author, sylvia schmelkes.

Sylvia Schmelkes is Provost of the Universidad Iberoamericana in Mexico City.

22 January 2020 Introduction

I nequity is perhaps the most serious problem in education worldwide. It has multiple causes, and its consequences include differences in access to schooling, retention and, more importantly, learning. Globally, these differences correlate with the level of development of various countries and regions. In individual States, access to school is tied to, among other things, students' overall well-being, their social origins and cultural backgrounds, the language their families speak, whether or not they work outside of the home and, in some countries, their sex. Although the world has made progress in both absolute and relative numbers of enrolled students, the differences between the richest and the poorest, as well as those living in rural and urban areas, have not diminished. 1

These correlations do not occur naturally. They are the result of the lack of policies that consider equity in education as a principal vehicle for achieving more just societies. The pandemic has exacerbated these differences mainly due to the fact that technology, which is the means of access to distance schooling, presents one more layer of inequality, among many others.

The dimension of educational inequity

Around the world, 258 million, or 17 per cent of the world’s children, adolescents and youth, are out of school. The proportion is much larger in developing countries: 31 per cent in sub-Saharan Africa and 21 per cent in Central Asia, vs. 3 per cent in Europe and North America. 2 Learning, which is the purpose of schooling, fares even worse. For example, it would take 15-year-old Brazilian students 75 years, at their current rate of improvement, to reach wealthier countries’ average scores in math, and more than 260 years in reading. 3 Within countries, learning results, as measured through standardized tests, are almost always much lower for those living in poverty. In Mexico, for example, 80 per cent of indigenous children at the end of primary school don’t achieve basic levels in reading and math, scoring far below the average for primary school students. 4

The causes of educational inequity

There are many explanations for educational inequity. In my view, the most important ones are the following:

Equity and equality are not the same thing. Equality means providing the same resources to everyone. Equity signifies giving more to those most in need. Countries with greater inequity in education results are also those in which governments distribute resources according to the political pressure they experience in providing education. Such pressures come from families in which the parents attended school, that reside in urban areas, belong to cultural majorities and who have a clear appreciation of the benefits of education. Much less pressure comes from rural areas and indigenous populations, or from impoverished urban areas. In these countries, fewer resources, including infrastructure, equipment, teachers, supervision and funding, are allocated to the disadvantaged, the poor and cultural minorities.
Teachers are key agents for learning. Their training is crucial. When insufficient priority is given to either initial or in-service teacher training, or to both, one can expect learning deficits. Teachers in poorer areas tend to have less training and to receive less in-service support.
Most countries are very diverse. When a curriculum is overloaded and is the same for everyone, some students, generally those from rural areas, cultural minorities or living in poverty find little meaning in what is taught. When the language of instruction is different from their native tongue, students learn much less and drop out of school earlier.
Disadvantaged students frequently encounter unfriendly or overtly offensive attitudes from both teachers and classmates. Such attitudes are derived from prejudices, stereotypes, outright racism and sexism. Students in hostile environments are affected in their disposition to learn, and many drop out early.

The Universidad Iberoamericana, main campus in Sante Fe, Mexico City, Mexico. 6 April 2013. Joaogabriel, CC BY-SA 3.0

It doesn’t have to be like this

When left to inertial decision-making, education systems seem to be doomed to reproduce social and economic inequity. The commitment of both governments and societies to equity in education is both necessary and possible. There are several examples of more equitable educational systems in the world, and there are many subnational examples of successful policies fostering equity in education.

Why is equity in education important?

Education is a basic human right. More than that, it is an enabling right in the sense that, when respected, allows for the fulfillment of other human rights. Education has proven to affect general well-being, productivity, social capital, responsible citizenship and sustainable behaviour. Its equitable distribution allows for the creation of permeable societies and equity. The 2030 Agenda for Sustainable Development includes Sustainable Development Goal 4, which aims to ensure “inclusive and equitable quality education and promote lifelong learning opportunities for all”. One hundred eighty-four countries are committed to achieving this goal over the next decade. 5 The process of walking this road together has begun and requires impetus to continue, especially now that we must face the devastating consequences of a long-lasting pandemic. Further progress is crucial for humanity.

Notes 1 United Nations Educational, Scientific and Cultural Organization , Inclusive Education. All Means All , Global Education Monitoring Report 2020 (Paris, 2020), p.8. Available at https://en.unesco.org/gem-report/report/2020/inclusion . 2 Ibid., p. 4, 7. 3 World Bank Group, World Development Report 2018: Learning to Realize Education's Promise (Washington, DC, 2018), p. 3. Available at https://www.worldbank.org/en/publication/wdr2018 . 4 Instituto Nacional para la Evaluación de la Educación, "La educación obligatoria en México", Informe 2018 (Ciudad de México, 2018), p. 72. Available online at https://www.inee.edu.mx/wp-content/uploads/2018/12/P1I243.pdf . 5 United Nations Educational, Scientific and Cultural Organization , “Incheon Declaration and Framework for Action for the implementation of Sustainable Development Goal 4” (2015), p. 23. Available at https://iite.unesco.org/publications/education-2030-incheon-declaration-framework-action-towards-inclusive-equitable-quality-education-lifelong-learning/ The UN Chronicle  is not an official record. It is privileged to host senior United Nations officials as well as distinguished contributors from outside the United Nations system whose views are not necessarily those of the United Nations. Similarly, the boundaries and names shown, and the designations used, in maps or articles do not necessarily imply endorsement or acceptance by the United Nations.

The "Memories" exhibition displayed in the Visitor’s Lobby of the United Nations Secretariat Building, August 2023. United Nations Office on Counter-Terrorism Office

Voices for Peace: The Crucial Role of Victims of Terrorism as Peace Advocates and Educators

In the face of unimaginable pain and trauma, victims and survivors of terrorism emerge as strong advocates for community resilience, solidarity and peaceful coexistence.

Sailors for Sustainability: Sailing the Globe to Document Proven Solutions for Sustainable Living

Most of the solutions we have described are tangible examples of sustainability in action. Yet our sailing journey also made us realize that the most important ingredient for a sustainable future is sustainability from within. By that we mean adopting a different way of perceiving the Earth and our role in it.

What if We Could Put an End to Loss of Precious Lives on the Roads?

Road safety is neither confined to public health nor is it restricted to urban planning. It is a core 2030 Agenda matter. Reaching the objective of preventing at least 50 per cent of road traffic deaths and injuries by 2030 would be a significant contribution to every SDG and SDG transition.

Documents and publications

Yearbook of the United Nations
Basic Facts About the United Nations
Journal of the United Nations
Meetings Coverage and Press Releases
United Nations Official Document System (ODS)
Africa Renewal

Libraries and Archives

Dag Hammarskjöld Library
UN Audiovisual Library
UN Archives and Records Management
Audiovisual Library of International Law
UN iLibrary

News and media

UN News Centre
UN Chronicle on Twitter
UN Chronicle on Facebook

The UN at Work

17 Goals to Transform Our World
Official observances
United Nations Academic Impact (UNAI)
Protecting Human Rights
Maintaining International Peace and Security
The Office of the Secretary-General’s Envoy on Youth
United Nations Careers

Contributor Notes

This paper presents new results on the relationship between income inequality and education expansion—that is, increasing average years of schooling and reducing inequality of schooling. When dynamic panel estimation techniques are used to address issues of persistence and endogeneity, we find a large, positive, statistically significant and stable relationship between inequality of schooling and income inequality, especially in emerging and developing economies and among older age cohorts. The relationship between income inequality and average years of schooling is positive, consistent with constant or increasing returns to additional years of schooling. While this positive relationship is small and not always statistically significant, we find a statistically significant negative relationship with years of schooling of younger cohorts. Statistical tests indicate that our dynamic estimators are consistent and that our identifying instruments are valid. Policy simulations suggest that education expansion will continue to be inequality reducing. This role will diminish as countries develop, but it could be enhanced through a stronger focus on reducing inequality in the quality of education.

I. Introduction

The persistence of high and, in many countries, rising income inequality over recent decades is a growing concern for policy makers worldwide, and has received increasing attention both from economists and in public debate ( OECD, 2008 ; Clements and others, 2015; Dabla-Norris and others, 2015 ) . Rising inequality has been attributed to a range of factors, including the globalization and liberalization of factor and product markets; skill-biased technological change; increases in labor force participation by low-skilled workers; declining top marginal income tax rates; increasing bargaining power of high earners; and the growing share of high-income couples and single-parent households ( OECD, 2008 ; Alvaredo and others, 2013 ; Hoeller, Joumard, and Koske, 2014). However, many of these factors have also had beneficial effects on growth and poverty reduction both nationally and globally (Chen and Ravallion, 2010; Milanovic, 2012).

The focus of this paper is on the relationship between education expansion and income inequality . Expansion of education is often seen as an important policy instrument for combating rising income inequality over the medium term. Not only is education expansion viewed as being important for promoting economic growth ( Barro, 2013 ; Hanushek, 2013 ), but it can also help to break the intergenerational transmission of poverty and reduce inequality of opportunity ( Corak, 2013 ), which in turn reduces future income inequality. Reducing income inequality through education expansion would also reduce the need for fiscal redistribution through distortionary fiscal policies such as progressive income taxes or means-tested transfers. So, from this perspective, education expansion has a “win-win” potential to simultaneously achieve both efficiency and equity objectives.

The paper extends the existing empirical literature in a number of dimensions . First, it expands the econometric analysis to address key estimation challenges not addressed in the existing literature, more specifically the issues of the endogeneity of the education and income inequality relationship and the persistence of income inequality over time. Second, it uses a new database on income inequality that expands the period of analysis while recognizing the need to use comparable measures of income inequality. Thirdly, it allows for heterogeneity in the relationship between education expansion and income inequality across advanced and developing economies to capture possible differences in returns to education. Finally, it also allows for heterogeneity in the relationship between education expansion and income inequality across working-age groups since there is evidence that education and experience are complementary inputs in human capital formation so that returns to education, and thus income inequality, can be expected to increase with working age.

The structure of the paper is as follows . In Section 2 we briefly discuss the conceptual framework underpinning the analysis of the impact of education expansion on income inequality and outline our empirical strategy for estimating this relationship. Section 3 discusses the data used in the analysis. Section 4 presents results based on these data and estimation methods, and compares them with the existing literature. Based on these results, Section 5 uses simulation analysis to discuss the implications of past and future changes in education outcomes for income inequality. Section 6 concludes.

II. Economic Theory and Empirical Estimation

A. Economic Theory

The standard theoretical framework for analyzing the relationship between education expansion and income inequality is the traditional human capital model . This model implies that the distribution of income (or earnings) is determined by both the level and distribution of education (or schooling) across the population. Using this model, earnings (Y) of an individual with S years of schooling can be approximated as 1 :

where Y 0 is the earnings of individuals with zero formal education, r is the rate of return to an additional year of schooling, and u captures other factors that influence earnings independent of education. The dispersion of earnings across individuals in a population can then be written as follows, with bar superscript denoting mean values:

Therefore, an increase in education inequality, Var(S), keeping the average level of schooling and other factors constant, unambiguously results in higher income inequality—i.e., the first two terms are unambiguously positive. However, the impact on income inequality of increasing the average level of schooling, S ¯ , keeping other factors constant, will depend on the relationship between r and S, i.e. Cov(r,S) —i.e., on the combined effect of the third and fourth terms. If the return to an extra year of schooling is constant across levels of schooling, so that Cov(r, S)=0, then an increase in the average level of schooling will unambiguously result in higher income inequality. Similarly, if the return to an extra year of schooling is higher at higher levels of schooling (Colclough and others, 2010; Castelló-Climent and Doménech, 2014 ), so that Cov(r, S)>0, then an increase in the average level of schooling will also unambiguously result in higher income inequality. However, if returns are lower at higher levels of education, as suggested by much of the empirical literature (Psacharopoulos and Patrinos, 2004), so that Cov(r, S)<0, then this will attenuate the increase in income inequality and, if sufficiently negative, may actually result in an increase in average schooling leading to a net decrease in income inequality. 2

B. Empirical Estimation

To test the empirical relationship between income inequality and the average level of education and education inequality, we use the following country-panel specification :

where subscripts refer to country i and year t respectively, I is a measure of income inequality, E is average years of education, σ is a measure of education inequality, X denotes other variables that impact income inequality independently of education outcomes, α captures unobserved time-invariant country-fixed effects, and ε captures other unobserved determinants that can vary across countries and time periods.

The data sources for the key income inequality and education variables used in the analysis are as follows (see Appendix 1 for details on the other explanatory variables included in the regression):

Income Inequality (I): For our analysis we use the Gini coefficient for disposable income inequality as our dependent variable since this is the inequality index that is most widely available and used in the related literature. We use an updated Gini coefficient database based on that assembled by Bastagli, Coady and Gupta (2012), which emphasized the need for comparability of country inequality estimates across time. Gini estimates are taken at five-year intervals.

Average Education (E): The average level of education of a country is taken as the average years of school attainment for the population aged 25 and over from Barro and Lee (2013) . These data are collected from census and survey information in five-year intervals, as compiled by UNESCO, Eurostat, and other sources. Average education is constructed based on the distribution of education attainment in the population over age 25, by five-year age groups and, for most cases, in six attainment categories: no formal education, incomplete primary, complete primary, incomplete secondary, complete secondary, and complete tertiary.

Education Inequality (σ): The inequality of education in a country is taken as the Gini coefficient of years of education for a given five-year interval based on educational attainment data from Barro and Lee (2013) . Appendix 2 discusses the construction of this Gini in more detail.

Our point of departure is the papers by De Gregorio and Lee (2002), Castelló-Climent and Doménech (2014) , and Dabla-Norris and others (2015) —henceforth referred to as DGL, CCD, and DNO, respectively . DGL used country-panel data for around 70 advanced and developing economies covering the five-year periods from 1965 to 1990. They estimated the relationship between education outcomes and income inequality (as captured by the Gini coefficient for disposable income) using the technique of seemingly unrelated regressions (SURE). Exploiting only the cross-country variation, they assumed that coefficient estimates were common across panels and thus ignored country-fixed effects. If these assumptions are valid then SURE provides more efficient estimates than OLS. As with previous attempts to estimate the relationship between education and income inequality 3 , they found that income inequality increases with education inequality and decreases with the average level of education. However, the positive coefficient on inequality of education was statistically insignificant when controls for a country’s GDP per capita and the level of public social spending were added.

Similar results were found by CCD, who extended the database to 2010 and used a fixed-effects estimation model to control for unobserved factors (such as historical factors, institutions, or culture) specific to a country that affect the level of income inequality but do not change over time . They also used the Gini measure of inequality from the SWIID database (unlike DGL who used the WIID database). Since there are valid concerns about the use of these Gini coefficients, in the current paper we use the Gini database constructed by Bastagli, Coady and Gupta (2012), which emphasizes the importance of comparability of inequality measures over time. 4 CCD also controlled for other factors such as technological change (i.e, a time trend and the share of high-technology exports in total exports) and globalization (i.e., proxies for trade and financial openness). Reflecting their primary focus on “the race between education and technological change” they also include variables for the ratio of the average years of tertiary to primary education in the population 25 years and older as well as the Gini coefficient for education outcomes. When controlling for fixed effects, the authors find a positive relationship between inequality of education outcomes and income inequality, although this is not always significant. They also find a negative relationship between the relative supply of skills, which we interpret as their proxy for the level of education, and income inequality, although again this is not significant in all specifications.

DNO also used a fixed-effects estimation model with average years of education and education inequality as dependent variables along with other dependent variables similar to those used in CCD . Their analysis does not find any statistically significant relationship between the income Gini and education inequality and education levels. They find a negative but insignificant relationship between education inequality and income inequality and a positive (negative) but insignificant relationship for education level in advanced (emerging) countries. They do however find a significantly positive (negative) relationship between education level and the share of income accruing to the top (middle) income decile.

This paper extends the estimation strategy to address two remaining econometric issues, namely, persistence and endogeneity :

Persistence of Income Inequality : Income inequality tends to change only slowly over time with very little within-country variation over the sample period, suggesting that there may be some, possibly unobserved, slowly-changing factors that explain this persistence. For example, this state dependence could reflect factors that prevent intergenerational mobility so that it is harder for a person born poor to achieve social mobility than for a person born in the middle class ( Corak, 2013 ). If these unobserved factors are correlated with education outcomes, then the estimated OLS and fixed-effects coefficients can be biased.

Endogeneity of Education Outcomes : Any observed relationship between education outcomes and income inequality may reflect reverse causation, i.e., current income inequality also affects current educational attainment and its dispersion. Therefore, any unobserved factors that affect income inequality and also education outcomes can bias the estimated relationship between education outcomes and income inequality.

To address these two issues, we use dynamic panel estimation techniques . To control for persistence, it is common to include past income inequality levels as an additional independent variable. However, by construction, this implies that the exogeneity assumption in the fixed-effects estimator is violated so that fixed-effects estimates are then biased (Nickell, 1981). To address this problem, Arellano and Bond (1991) suggest using a first-differenced GMM (Diff-GMM) estimator that also deals with the endogeneity problem by first differencing the data and then deploying suitably lagged values of the independent and dependent variables as instruments. 5 However, Blundell and Bond (1998) show that the Diff-GMM estimator suffers from the weak instrument problem when the number of time periods is small and that this bias is exacerbated when the time series are persistent. Building on Arellano and Bover (1995), the system GMM estimator (Sys-GMM) developed by Blundell and Bond (1998) addresses this weak instrument problem by exploiting level restrictions which remain informative even in the presence of persistence. Thus, where the number of time periods is small and in the presence of persistence, Sys-GMM estimator can produce dramatic efficiency gains over the basic Diff-GMM estimator. 6 For this reason, our preferred model is the Sys-GMM estimator.

Recent research has shown that Sys-GMM may equally suffer from the weak instrument problem, particularly when the time series is large and when substantial unobserved heterogeneity exists (Hayakawa, 2006; Bun and Windmeijer, 2010) . We therefore complement our Sys-GMM estimates by using the approach adopted by Barro and Lee (2010), which instruments a cohort’s schooling with the educational attainment of its parents (Ins-GMM). Finally, to further test the robustness of our estimates, we employ the long-difference instrumental variables estimator proposed by Hahn, Hausman and Kuersteiner (2006), which has been shown to be much less biased and more efficient than conventional implementations of the Sys-GMM when the number of time periods is small and in the presence of persistence.

III. Data Description

As indicated, the measure of income inequality used in the analysis is the Gini coefficient for disposable income, taken from the database constructed by Bastagli, Coady and Gupta (2012) . More specifically, we use Gini coefficients for five-year intervals from 1980 to 2010. Based on this, Figure 1 shows the profile of regional average income inequality over the last four decades. Whereas Gini coefficients are available from 1980 for advanced economies, for many other regions they are only available from 1990 onwards.

Disposable Income Inequality by Region, 1980–2010

Citation: IMF Working Papers 2017, 126; 10.5089/9781475595741.001.A001

Download Figure
Download figure as PowerPoint slide

Average income inequality is lowest in advanced countries (ADV) but has increased steadily over the last four decades . In Emerging European countries (EE), income inequality exhibits a step increase between 1990 and 1995, a period of substantial structural change associated with transition to market economies, but has stayed relatively stable (with a slight decrease) since then. Average income inequality in the Asian and Pacific (AP) and in the Middle East and North Africa (MENA) regions are higher than in ADV and EE and have also increased steadily over recent decades. More striking is the relatively large level of income inequality in countries in sub-Saharan Africa (SSA) and especially in Latin America and the Caribbean (LAC). 7 In SSA, while income inequality has been declining slowly it still remains among the highest in the world. Similarly, although average income inequality in LAC has declined sharply since its peak in 2000 it still has the most unequal income distribution in the world.

Figure 2 plots the inequality of education against average years of schooling for different country groupings . As expected, the data suggest an inverted-U relationship with education inequality increasing over lower levels of schooling before declining. However, the precise relationship seems to differ across country groups. Noticeable is the relatively sharp increase in education inequality in MENA and SSA countries as education expands from initial low levels. As a result, at around an average of 4 years of schooling, education inequality was substantially higher in MENA and SSA than in AP or LAC. Therefore, MENA and SSA appear to have experienced a much more unequal early expansion of education than in other regions, although there is some evidence that education inequality is beginning to decrease sharply as access to education expands further. 8

Education Levels and Education Inequality

IV. Empirical Results

To facilitate comparison with the existing literature, we start by estimating the above equation using SURE and fixed-effects estimators . Table 1 (column 1) shows the correlations between both education levels and education inequality and income inequality based on an OLS estimator. Whereas income inequality is lower at higher education levels (consistent with decreasing returns to education), income inequality decreases with education inequality, although this relationship is insignificant. Together, variation in these variables explains 22 percent of the total variation in income inequality. Most of the variation in income inequality is explained by cross-country variation as opposed to changes over time. For instance, when time dummies are added (column 2) the explained variation increases to just 30 percent. However, when country-dummies are also added the explained variation increases to 87 percent. While both coefficients are negative and significant when time dummies are included, both are insignificant when country dummies are added highlighting the importance of controlling for country-fixed effects.

Education and Income Inequality: OLS, SURE and Fixed Effects Estimates

OLS Simple (1) OLS Time (2) OLS Time and Country (3) OLS All (4) SURE (5) FEALL (6) Gini Education −2.909

(0.77) −18.094*

(0.09) 0.000

(1.00) −5.847

(0.55) −1.498

(0.88) 3.300

(0.68) Schooling −1.716***

(0.00) −3.101***

(0.00) −0.395

(0.58) −0.797

(0.14) −0.980*

(0.09) −0.795*

(0.09) GDP 9.322**

(0.03) 11.858***

(0.01) 7.696**

(0.03) GDP −0.571**

(0.02) −0.647***

(0.01) −0.398**

(0.02) Openness −7.479**

(0.03) −2.945

(0.15) 3.411

(0.25) Social Protection −0.499***

(0.00) −0.491***

(0.00) −0.040

(0.65) Population over 65 0.256

(0.31) 0.409*

(0.07) 0.419**

(0.03) Population less than 15 0.349**

(0.02) 0.523***

(0.00) 0.185

(0.15) Inflation 0.001***

(0.00) −0.017

(0.52) 0.000***

(0.00) Urban 0.069**

(0.04) 0.003

(0.95) −0.145**

(0.02) Capital Account Openness 0.045**

(0.03) 0.016

(0.35) 0.005

(0.73) N 873 873 873 418 32 418 Adjusted R 0.22 0.29 0.85 0.78 0.82 0.17

Credit Growth				0.031** (0.02)	0.021** (0.03)	0.018*** (0.00)

When other determinants of income inequality are added to the specification (column 4) , the coefficient signs on both education level and education inequality remain negative but the coefficients become smaller in absolute terms (compared to column 2) and insignificantly different from zero. The explained variation decreases to 79 percent reflecting the absence of country dummies. Estimating this relationship using SURE results in a smaller (in absolute terms) and still insignificantly negative coefficient on education inequality, but the negative coefficient on education level becomes larger in absolute terms and significant. Applying fixed effects results in the negative coefficient on education levels decreasing in absolute terms and remaining significant. Although the coefficient on education inequality now becomes positive it remains insignificant.

Table 2 presents the results when we attempt to control for endogeneity and persistence . With the first-difference estimator, the coefficient on education levels remains negative, decreases in absolute terms compared to the fixed effects estimate but now becomes insignificant. Consistent with theory, the coefficient on education inequality remains positive although still insignificant. Under both difference GMM (Diff-GMM) and system GMM (Sys-GMM) the positive coefficient on education inequality increases around six fold and becomes significant at the 10 and 5 percent levels respectively, while the coefficient on education levels becomes positive but remains insignificant. 9 When we instrument education levels and education inequality using parents’ schooling levels (Ins-GMM) the coefficient on education inequality remains positive, large and significant, while the positive coefficient on education level now becomes significant. Therefore, the coefficients on education inequality and levels remain quite stable across all the GMM estimators. The Arellano Bond tests for the serial correlation of the disturbances indicate that our GMM estimators are consistent. Moreover, both the Sargan and Hansen tests for over-identifying restrictions do not reject the hypothesis that our instruments are valid in our GMM estimations. Under the long-distance estimator, the positive coefficient on education inequality becomes smaller and insignificant while the coefficient on education levels remains positive but is smaller and insignificant. However, the insignificance found when running the long distance estimator could be explained by the reduction in the number of observations that we have to incur to implement this estimation.

Education and Income Inequality: Difference, GMM, Long Distance Estimates

First Difference (1) Diff-GMM (2) Sys-GMM (3) Ins-GMM (4) Long Difference (5) Sys-GMM (6) Gini Education 6.213

(0.42) 36.816*

(0.05) 34.445**

(0.02) 30.466**

(0.03) 12.101

(0.18) 31.038**

(0.03) Schooling −0.390

(0.36) 1.087

(0.32) 1.310

(0.13) 1.453**

(0.05) 0.194

(0.73) 0.965

(0.24) Gini Education Advanced −15.740**

(0.02) Schooling Advanced 0.123

(0.52) GDP 11.566**

(0.03) 13.420*

(0.06) 11.582***

(0.00) 11.464***

(0.00) 6.986*

(0.06) 14.077***

(0.00) GDP −0.599**

(0.03) −0.714*

(0.06) −0.694***

(0.00) −0.703***

(0.00) −0.463**

(0.03) −0.830***

(0.00) Openness 4.481**

(0.04) 4.755*

(0.07) 0.820

(0.78) 0.325

(0.91) 2.953

(0.19) −0.860

(0.67) Social Protection −0.068

(0.50) −0.120

(0.53) −0.236

(0.16) −0.162

(0.27) −0.115

(0.26) −0.239

(0.12) Population over 65 0.299

(0.23) 0.382

(0.12) 0.157

(0.56) 0.014

(0.95) 0.373*

(0.07) 0.004

(0.98) Population less than 15 0.132

(0.44) 0.010

(0.97) 0.028

(0.81) 0.103

(0.32) 0.120

(0.31) 0.033

(0.82) Inflation 0.001***

(0.00) 0.000***

(0.00) 0.000*

(0.06) 0.000*

(0.05) 0.000

(0.64) 0.001**

(0.01) Urban −0.124

(0.14) −0.195*

(0.09) −0.019

(0.51) −0.025

(0.41) −0.094

(0.18) 0.001

(0.97) Capital Account Openness 0.014

(0.28) 0.014

(0.43) 0.025*

(0.08) 0.026*

(0.08) 0.010

(0.48) 0.033**

(0.01) Credit Growth 0.017**

(0.01) 0.014

(0.20) −0.001

(0.90) 0.004

(0.56) 0.005

(0.56) 0.008

(0.33) N 322 309 402 402 164 402 R 0.10 0.37 # of Instruments 47 51 53 73 AR(1) Test p-val. 0.046 0.002 0.002 0.006 AR(2) Test p-val. 0.133 0.163 0.150 0.145 Hansen J Test p-val. 0.136 0.188 0.160 0.850 Sargan Test p-val. 0.182 0.517 0.454 0.060

L.gini_net		0.416** (0.01)	0.644*** (0.00)	0.593*** (0.00)	0.588*** (0.00)	0.503*** (0.00)

When we allow these coefficients to differ between advanced and other countries, we find that the positive coefficient on education inequality is halved in advanced while the coefficient on education level remains positive and insignificant in both sets of countries . This result suggests that the impact of education inequality on income inequality is not homogeneous across different levels of development and shows that reducing education inequality is an even more important policy for developing countries while becoming less important as countries develop.

The literature on the returns to education levels points to a “fanning out” of these returns with age (a proxy for years of experience) consistent with the returns to education accruing later in an individual’s working career . Therefore, inequality in years of education is more likely to translate into higher income inequality among older workers than among younger workers ( Table 3 ). To test this, we add education inequality for older workers as an additional explanatory variable. Consistent with the literature on returns to education, we find that higher inequality in education levels among older workers (keeping inequality of education for all workers constant) is associated with a substantially larger level of income inequality. In addition, the acceleration of the skill-biased technological change means that higher education levels for the young could reduce the income gap between older and younger generations and thus income inequality. We test this by including average education for the young as an additional explanatory variable and find that higher education levels for the young are indeed associated with lower income inequality.

Education and Income Inequality: Cohort Effects

OLS (1) FE (2) First Difference (3) Diff-GMM (4) Sys-GMM (5) Ins-GMM (6) Gini Education 2.102

(0.55) −0.990

(0.83) 1.256

(0.80) 7.043

(0.47) 5.225

(0.37) 1.719

(0.67) Schooling −0.317

(0.50) −0.268

(0.58) −0.792

(0.16) 1.076

(0.27) 1.986***

(0.00) 1.958***

(0.01) Gini Education Old −6.720

(0.79) 21.803

(0.37) 10.180

(0.68) 72.386

(0.23) 63.903*

(0.08) 69.535*

(0.09) Schooling Young −0.188

(0.86) −0.901

(0.11) −0.689

(0.23) −1.342

(0.27) −2.830**

(0.01) −2.250**

(0.04) GDP 10.337**

(0.01) 8.340**

(0.03) 12.837**

(0.03) 12.232

(0.14) 12.327***

(0.00) 11.999***

(0.00) GDP −0.631***

(0.01) −0.406**

(0.03) −0.665**

(0.03) −0.627

(0.15) −0.731***

(0.00) −0.734***

(0.00) Openness −7.426**

(0.04) 4.342

(0.11) 4.782**

(0.03) 4.917*

(0.10) 1.318

(0.59) 1.390

(0.59) Social Protection −0.488***

(0.00) −0.003

(0.97) −0.073

(0.53) −0.413*

(0.09) −0.167

(0.21) −0.155

(0.28) Population over 65 0.298

(0.33) 0.353**

(0.05) 0.355

(0.16) 0.746***

(0.01) 0.012

(0.96) −0.020

(0.94) Population less than 15 0.358**

(0.03) 0.054

(0.65) 0.119

(0.49) 0.012

(0.96) 0.171

(0.22) 0.252**

(0.02) Inflation 0.001***

(0.00) 0.000*

(0.08) 0.001***

(0.00) 0.001***

(0.00) 0.001*

(0.07) 0.001

(0.12) Urban 0.059*

(0.08) −0.101

(0.14) −0.099

(0.21) −0.222**

(0.03) −0.017

(0.58) −0.025

(0.49) Capital Account Openness 0.047**

(0.02) 0.014

(0.31) 0.014

(0.29) 0.010

(0.52) 0.030**

(0.01) 0.032**

(0.01) N 395 395 301 283 374 374 # of Instruments 45 50 52 AR(1) Test p-val. 0.374 0.007 0.008 AR(2) Test p-val. 0.149 0.189 0.162 Hansen J Test p-val. 0.375 0.627 0.533 Sargan Test p-val. 0.357 0.389 0.473

Credit Growth	0.034** (0.01)	0.020*** (0.00)	0.017** (0.02)	0.013 (0.11)	0.008 (0.36)	0.011 (0.25)

V. Policy Simulations

In this section, we use the results from our preferred Sys-GMM estimation to simulate the impacts of education inequality and level on income inequality to get a sense of their quantitative importance in determining income inequality . We start by analyzing how much of the change in income inequality over the last fifteen years from 1990 to 2005 can be attributed to changes in education outcomes. We then analyze how changing education outcomes over the subsequent two decades might impact income inequality.

A. Education and Past Income Inequality

Trends in income inequality between 1990 and 2005 varied substantially across regions . On average in the sample, inequality decreased in both SSA and MENA countries while other regions experienced increases with the Gini increasing by over 5 points in both AP and EE ( Table 4 ). Across all regions, decreases in the inequality of education reduced income inequality, ranging from a 1.5 point decrease in ADV and EE to a 4.8 point decrease in MENA. However, increases in the level of education increased income inequality across all regions by 1.5 to 2.2 points. While the net effect of changing education outcomes was to slightly increase income inequality in ADV, the effect was negative in all other regions. These results highlight the important role education investments can play in mitigating increases in income inequality, in particular the key role played by education expansion strategies that emphasize more equal access to education

Implications of Changing Education Outcomes for Income Inequality, 1990–2005

Regions Inequality in 1990 Inequality in 2005 Change in Income inequality From education inequality (absolute change) From years of schooling (absolute change) Total change in Income inequality from education Number of countries Advanced Economies 28.1 30.4 2.3 −0.9 1.5 0.5 32 Emerging Europe 25.3 33.0 7.7 −2.9 1.5 −1.4 11 Latin America & the Caribbean 46.4 48.0 1.5 −2.8 1.9 −0.9 21 Asia & Pacific 33.5 38.6 5.1 −3.8 1.9 −1.8 15 Middle East & North Africa 38.7 37.2 −1.6 −4.8 2.2 −2.7 7 Sub-Saharan Africa 49.4 43.6 −5.8 −3.6 1.7 −1.9 17

B. Education and Future Income Inequality

Table 5 presents projections of the impact of future changes in education levels and education inequality on income inequality from 2005 to 2025 . The average level of education is projected forward based on known education levels for current working cohorts while future working cohorts are assumed to have the same education outcomes of the youngest current working cohort (i.e., those aged 20–25 years in 2005). These changes in education levels are then used to project the impact on education inequality. The results indicate that changes in education outcomes will continue to have a dampening impact on income inequality over this period in most regions, with the exception of ADV, where it will increase income inequality by 0.1 points. Similar to the period 1990–2005, continued decreases in education inequality will have a dampening impact on income inequality while increasing education levels will increase income inequality, with both impacts being smaller than for the earlier shorter period. However, the results for ADV suggest that the income inequality reducing impact of education can be expected to decrease as the level of education increases, and the inequality of education decreases, in emerging and developing economies and converges to levels observed in advanced economies. This highlights the importance of also focusing on reducing the inequality of education quality (e.g., as captured by cognitive skills) to enhance the income inequality reducing impact of education expansion.

Implications of Changing Education Outcomes for Income Inequality, 2005–2025

Regions Inequality in 2005 From education inequality (% improvement) From years of schooling (% improvement) Total Number of countries Advanced Economies 30.4 −0.7 1.2 0.1 35 Emerging Europe 33.0 −1.4 1.1 −0.1 12 Latin America & the Caribbean 48.0 −2.9 1.4 −1.3 25 Asia & Pacific 38.6 −3.0 1.3 −1.5 23 Middle East & North Africa 37.2 −4.1 1.6 −2.3 18 Sub-Saharan Africa 43.6 −3.2 1.1 −2.0 29

VI. Conclusions

This paper presents new results on the relationship between education expansion and income inequality . It extends the existing literature in a number of dimensions. First, it addresses key econometric issues ignored in the existing literature related to the need to allow for the persistence of income inequality and the endogeneity of education and inequality outcomes, both of which require the use of dynamic panel analysis. Second, the analysis tests for heterogeneity in these relationships across country income groups as well as across different age cohorts. Finally, the paper uses a new database on income inequality that addresses concerns about the quality of the income inequality data currently widely used in the literature, and also extends the period of the analysis.

The analysis demonstrates clearly the importance of controlling for persistence, endogeneity and heterogeneity . When dynamic panel estimation techniques are applied, the positive relationship between education inequality and income inequality becomes substantially larger, statistically significant and stable across the various estimators. This is consistent with our theoretical insights based on the human capital model and confirms that education expansion reduces income inequality through decreasing the inequality of education. However, the relationship between income inequality and schooling levels is found to be positive but small and not always statistically significant. Statistical tests indicate that our dynamic estimators are consistent and that that our identifying instruments are valid.

Our policy simulations confirm that the net impact of education expansion over the last fifteen years has been to reduce income inequality, especially in emerging and developing economies . Although the magnitude of the net impact on income inequality varies across emerging and developing economies, it is always inequality reducing. This reduction reflects decreasing education inequality, which is only partly offset by the inequality-increasing effects of rising education levels. In advanced economies, education expansion is associated with a net increase in income inequality. This reflects the relatively smaller impact of decreasing education inequality at the lower levels of education inequality observed in advanced economies being offset by the income inequality-increasing impact of rising levels of education (consistent with constant or increasing returns to additional years of education).

Projections forward over two decades suggest that education expansion will continue to have an inequality-reducing impact in emerging and developing economies . Even though the inequality-reducing impact of falling education inequality is offset by the inequality-increasing impact of rising education levels, the net impact is still inequality reducing. The inequality-increasing impact of education expansion in advanced economies suggests that the inequality-reducing role of education expansion in emerging and developing economies will diminish as these countries develop. Therefore, other policies will also be needed to address rising income inequality. Among these, it is likely that focusing on reducing the inequality of education quality (e.g., improving cognitive skills) can help to enhance the role of education expansion as a force for reducing income inequality.

Appendix 1. Explanatory Variables and their Sources

The explanatory variables used in this paper were selected based on factors identified in the literature as important determinants of income inequality. In this appendix, we briefly discuss the variables and their sources.

Level of income . We test the Kuznets inverted-U hypothesis. According to this hypothesis, there is a positive correlation between income inequality and per capita income at low levels of income, which eventually becomes negative after a country reaches a specific level of development. In order to test this hypothesis, we use the log of ppp per capita income and its square. We use the variable “GDP per capita, ppp” in the World Bank’s World Development Indicators (WDI).

Social public spending . Spending on social programs is expected to be progressively targeted and thus to reduce inequality. Government social expenditure is obtained from IMF, Government Finance Statistics Yearbook. It includes pension and other welfare benefits. For most countries data is available from 1990, but advanced economies have available data from 1970.

Trade openness . According to the Heckscher-Ohlin and the Stolper-Samuelson theoretical framework, greater trade openness should increase the relative demand and prices for unskilled labor and lead to a more equal distribution of wages in low-skilled-labor abundant countries and a more unequal distribution of wages in high-skilled labor abundant countries. We use the commonly used variable (exports + imports) / GDP to capture a country’s degree of trade openness.

Capital account openness . By relaxing credit supply constraints, capital openness is expected to reduce income inequality if institutions are strong ( LaGarda et al, 2016 ). In this paper, we use the variable “cap100” of the IMF’s Annual Report on Exchange Restrictions to reflect the degree of openness of the capital account. Higher values of the indices represent greater openness; lower values of the indices represent greater restrictiveness.

Credit to the private sector . Similar to capital account openness, greater availability of credit should minimize the amount of credit-constrained people in a country and reduce income inequality. In our paper, we use the variable “credit to the private sector/GDP” from the IMF’s Government Finance Statistics Yearbook.

Dependency ratio . Theoretical economic models posit that, all other things being equal, an ageing population and a rising dependency ratio tend to increase income inequality (Von Weizsäcker, 1995). We use population data from the World Bank’s WDI. Specifically, we use the proportion of the elderly (above 65 years old) and the children (below 16 years old) in the population.

Inflation . Several empirical and theoretical studies have analyzed the relationship between inflation and income inequality. The data on these studies have consistently shown a positive correlation between these two variables. In this paper, we use average annual CPI inflation as measured in the IMF’s Government Finance Statistics Yearbook.

Urban population . Reflecting the gap in income between rural and urban populations, different urbanization levels can lead to differences in income inequality. We use the percentage of the population living in urban areas in the World Bank’s WDI.

Summary Statistics

Mean SD Obs GDP per capita PPP (constant 2000 US$) 5617.89 8872.32 1588 Social spending (%GDP) 6.30 6.10 813 Trade openness ((Exports+Imports)/GDP) 0.22 0.16 775 Capital account openness (Cap100) 58.53 29.38 1134 Credit to the private sector (%GDP) 37.75 38.02 1400 Elderly population (% total population) 6.05 4.06 1985 Children (% total population) 35.47 10.05 1985 Inflation (% average) 37.35 412.93 1267 Urban (% population) 48.57 25.40 2150

Appendix 2. Constructing a Measure of Education Inequality

There is no measure of education inequality in Barro and Lee database, which is the source for our education data. Thus, we estimate each country’s education inequality using the standard method to calculate income inequality from a Lorenz curve.

In order to implement this method, we need a disaggregated measure of schooling attainment. The education attainment data provided in the Barro and Lee database is presented as the fraction of the population with no education, and with incomplete and complete primary, secondary and tertiary schooling. Thus, the first step in the estimation of the education Gini for education inequality is to assign an average education level to each of these categories of education achievement.

While assigning a specific average to these categories might be considered somehow arbitrary, the estimation results did not change much when we assigned slightly different numbers to these categories. Specifically, we assigned a value of 1 for the category no education, a value of 4 for incomplete and a 7 for complete primary education. We assigned a value of 10 for incomplete and a 13 for complete secondary education. Finally, we assigned a value of 16 for incomplete and 19 for complete tertiary education.

Finally, with these average numbers for each education category, we can calculate total number of years of education for different segments of each country’s population. Using these total years of education, we simply apply the concept of the Lorenz curve using the population of each segment as weights.

Acemoglu , Daron , 1997 , “ Matching, Heterogeneity, and the Evolution of Income Distribution ,” Journal of Economic Growth , Vol. 2 ( March ), pp. 61 – 92 .

Search Google Scholar
Export Citation

Adelman , Irma and Cynthia T. Morris , 1973 , Economic Growth and Social Equity in Developing Countries , Stanford University Press , Stanford, California .

Ahluwalia , Montek S. , 1976 , “ Income Distribution and Development: Some Stylized Facts ,” American Economic Review , Papers and Proceedings , Vo. 66 ( May ), pp. 128 – 35 .

Alvaredo , F. , A. Atkinson , T. Piketty , and E. Saez , 2013 , “ The Top 1 Percent in International and Historical Perspective ,” Journal of Economic Perspectives , Vol. 27 , No. 3 , pp. 3 – 20 .

Anand , Sudhir and Ravi Kanbur , 1993 , “ The Kuznets Process and the Inequality-Development Relationship ,” Journal of Development Economics , Vol. 40 ( February ), pp. 25 – 52 .

Atkinson , Anthony B. , 1997 , “ Bringing Income Distribution in from the Cold ,” Economic Journal , Vol. 107 ( March ), pp. 297 – 321 .

Barro , Robert and Jong-Wha Lee , 1994 , “ Sources of Economic Growth ,” Carnegie - Rochester Series on Public Policy , Vol. 40 ( June ), pp. 1 – 46 .

Barro , Robert and Jong-Wha Lee , 1996 , “ International Measures of Schooling Years and Schooling Quality ,” American Economic Review , Papers and Proceedings , Vol. 86 ( May ), pp. 218 – 23 .

Barro , Robert , 1999 , “ Inequality, Growth, and Investment ,” mimeo , Harvard University , Cambridge, Massachusetts .

Barro , Robert , 2013 , “ Education and Economic Growth”, Annals of Economics and Finance , Vol. 14 ( 2 ), pp. 301 – 28 .

Barro , Robert , and Jong-Wha Lee , 2015 , Education Matters: Global Schooling Gains from the 19th to the 21st Century , Chapter 2 ( London : Oxford University Press ).

Becker , Gary S. and Barry R. Chiswick , 1966 , “ Education and the Distribution of Earnings ,” American Economic Review , Papers and Proceedings , Vol. 56 ( May ), pp. 358 – 69 .

Becker , Gary S. and Nigel Tomes , 1986 , “ Human Capital and the Rise and Fall of Families ,” Journal of Labor Economics , Vol. 4 ( July ), S1–S39.

Benabou , Roland , 1994 , “ Education, Income Distribution, and Growth: The Local Connection ,” NBER Working Paper 4798 , National Bureau of Economic Research , Cambridge, Massachusetts .

Benabou , Roland , 1993 , “ Inequality and Growth ,” NBER Macroeconomics Annual , pp. 11 – 92 .

Bulir , Ales , 1998 , “ Income Inequality: Does Inflation Matter? ” IMF Working Paper WP/98/7 ( Washington : International Monetary Fund ).

Castelló-Climent , Amparo and Rafael Doménech , 2014 , Working Paper 12/28, BBVA Research .

Chenery , Hollis B. and Moshe Syrquin , 1975 , Patterns of Development , 1950–1970 , Oxford University Press for the World Bank , London .

Chiswick , Barry R. , 1971 , “ Earnings Inequality and Economic Development ,” Quarterly Journal of Economics , Vol. 85 , pp. 21 – 39 .

Corak , Miles , 2013 , “ Income Inequality, Equality of Opportunity, and Intergenerational Mobility ,” Journal of Economic Perspectives , Vol. 27 (No. 3 , Summer ), pp. 79 – 102 .

Dabla-Norris , Eras , Kalpana Kochhar , Nujin Suphaphiphat , Frantisek Ricka , and Evridiki Tsounta , 2015 , “ Causes and Consequences of Income Inequality: A Global Perspective ”, IMF Staff Discussion Paper SDN/15/13 ( Washington : International Monetary Fund ).

De Gregorio , Jose , 1995 , “ Comment ,” in Rudi Dornbusch and Sebastian Edwards (eds), Stabilization, Economic Reform and Growth ( Chicago : University of Chicago Press ).

De Gregorio , Jose and Se-Jik Kim , 2000 , “ Credit Markets with Differences in Abilities: Education, Distribution and Growth ,” International Economic Review , Vol. 41 ( August ), pp. 579 – 607 .

De Gregorio , Jose and Jong-Wha Lee , 1999 , “ Education and Income Distribution: New Evidence from Cross-Country Data ,” Development Discussion Papers No. 714 ( Cambridge, Massachusetts : Harvard Institute for International Development ), www.hiid.harvard.edu_pub_ddps.html ).

Deininger , Klaus W. and Lyn Squire , 1996 , “ A New Data Set Measuring Income Inequality ,” World Bank Economic Review , Vol. 10 ( September ), pp. 565 – 91 .

Deininger , Klaus W. and Lyn Squire , 1998 , “ New Ways of Looking at Old Issues: Inequality and Growth ,” Journal of Development Economics , Vol. 57 ( December ), pp. 259 – 87 .

Durlauf , Steven N. , 1996 , “ A Theory of Persistent Income Inequality ,” Journal of Economic Growth , Vol. 1 ( March ), pp. 75 – 93 .

Gallop , J. , Steven Radelet , and Andrew Warner , 1998 , “ Economic Growth and Income of the Poor,” mimeo ( Cambridge, Massachusetts : Harvard Institute for International Development ).

Galor , Oded and Daniel Tsiddon , 1997 , “ The Distribution of Human Capital and Economic Growth ,” Journal of Economic Growth , Vol. 2 ( March ), pp. 93 – 124 .

Galor , Oded and Joseph Zeira , 1993 , “ Income Distribution and Macroeconomics ,” Review of Economic Studies , Vol. 60 ( January ), pp. 35 – 53 .

Gottschalk , Peter and Timothy M. Smeeding , 1997 , “ Cross-National Comparisons of Earnings and Income Inequality ,” Journal of Economic Literature , Vol. 35 ( June ), pp. 633 – 87 .

Hanushek , E. , and L. Wobmann , 2010 , “ Education and Economic Growth”, International Encyclopedia of Education , Vol. 2 , pp. 245 – 52 .

Hanushek , E. , 2013 , “ Economic Growth in Developing Countries: The Role of Human Capital ”, Economics of Education Review , 37 , 204 – 212 .

Jenkins , Stephen , 2015 . “ World Income Inequality Databases: An Assessment of WIID and SWIID ,” Journal of Economic Inequality , Vol. 13 ( December ), pp. 629 – 71 .

Knight , John B. and Richard H. Sabot , 1983 , “ Educational Expansion and the Kuznets Effect ,” American Economic Review , Vol. 73 ( December ), pp. 1132 – 6 .

Kuznets , Simon , 1955 , “ Economic Growth and Income Inequality ,” American Economic Review , Vol. 45 ( March ), pp. 1 – 28 .

LaGarda , Guillermo and Kevin Gallagher , and Jennifer Linares , 2016 , “ Capital Openness and Income Inequality: A Cross-Country Study and The Role of CCT in Latin America ,” Munich Personal RePEc Archive papers (June) .

Li , Hongyi , Lyn Squire , and Heng-fu Zou , 1998 , “ Explaining International and Intertemporal Variations in Income Inequality ,” Economic Journal , Vol. 108 ( January ), pp. 26 – 43 .

Londoño , Juan Luis , 1996 , “ Poverty, Inequality, and Human Capital Development in Latin America, 1950–2025 ,” World Bank Latin American and Caribbean Studies 2 ( Washington : World Bank ).

Marin , A. and George Psacharopoulos , 1976 , “ Schooling and Income Distribution ,” Review of Economics and Statistics , Vol. 58 , pp. 332 – 8 .

Montenegro , Claudio and Harry Patrinos , 2014 , “ Comparable estimates of returns to schooling around the world ”, Policy Research Working Paper WPS 7020 . Washington, DC : World Bank Group .

Organisation for Economic Co-operation and Development (OECD) , 2008 , Growing Unequal? Income Distribution and Poverty in OECD Countries . Paris : OECD Publishing .

Psacharopoulos , George , 1977 , “ Unequal Access to Education and Income Distribution: An International Comparison ,” De Economist , Vol. 125 , pp. 383 – 92 .

Psacharopoulos , George and Maureen Woodhall , 1985 , Education for Development: An Analysis of Investment Choices , Oxford University Press , New York .

Ram , Rati , 1984 , “ Population Increase, Economic Growth, Educational Inequality, and Income Distribution: Some Recent Evidence ,” Journal of Development Economics , Vol. 14 ( April ), pp. 419 – 28 .

Ram , Rati , 1989 , “ Can Educational Expansion Reduce Income Inequality in Less-Developed Countries? ” Economics of Education Review , Vol. 8 ( 2 ), pp. 185 – 9 .

Sarel , Michael , 1997 , “ How Macroeconomic Factors Affect Income Distribution: The Cross-Country Evidence ,” IMF Working Paper WP_97_152 ( Washington : International Monetary Fund ).

Winegarden , C. R. , 1979 , “ Schooling and Income Distribution: Evidence from International Data ,” Economica , Vol. 46 , pp. 83 – 7 .

Schooling is taken here as a proxy for human capital. More generally, it can also be viewed as a proxy for other forms of human capital accumulation such as on-the-job training. For a discussion of the importance of schooling quality, see Barro (2013) and Hanushek and Wobmann (2010) . For a review of country-level estimates based on household survey data, see Montenegro and Patrinos (2014) .

Note that the total impact of education expansion on income inequality will depend on the relationship between average schooling and its dispersion. For instance, at the early stages of development, both the average level and dispersion of schooling are typically low. Expansion of education will initially tend to increase income inequality as a few more individuals gain higher education and earnings, but eventually lower the inequality of education and earnings as education becomes more widespread. The increase in the supply of high-education individuals will also tend to decrease the skilled wage premium and thus also income inequality. The net effect of education expansion on income inequality will therefore depend on the relative sign and magnitudes of these “composition” and “compression” effects, and is more likely to be positive at low levels of development and education attainment ( Knight and Sabot, 1983 ).

For reviews of past studies, see Psacharopoulos and Woodhall (1985 , pp264-70) and Ram (1989) .

The use of country fixed-effects estimators will control for time-invariant differences in income inequality measures across countries, e.g. due to the use of different household surveys based on incomes, expenditures or consumption. See Jenkins (2015) for a detailed and critical discussion of the issues that arise with the use of the WIID and SWIID income inequality databases.

This approach is typically seen as superior to that suggested by Anderson and Hsiao (1982) that includes the dependent variable lagged two periods as an independent variable in the differenced equation, which results in biased coefficients when the number of time periods is small.

Note also that the implicit assumption in Sys-GMM is that independent variables are predetermined (or weakly exogenous), depending only on past values of income inequality. For example, when a family decides on the choice of education in year t, it takes into account income developments up to this year and does not anticipate future income developments. This assumption can be tested with a Hausman test. The assumption of weak exogeneity also implies a lack of autocorrelation in the error terms. Testing for lack of second-order correlation in the difference equation is therefore equivalent to testing the validity of the weak exogeneity assumption. For this reason, in addition to the Hausman test, we also do the AR(2) test suggested by Arellano and Bond (1991).

Note that income inequality measures in both LAC and SSA are typically based on consumption or expenditure inequality, which tends to be lower than the income inequality measures typically used in ADV and EE. Therefore, the gap between ADV and other country groups may actually be higher.

Whereas in ADV and EE much of the expansion in education levels between 1950 and 2010 came from a decreasing share of the working population with primary education, in other regions, but especially in MENA and SSA, it has been driven by a decreasing share of the working population with no formal education.

We conducted a Wald test to assess the importance of keeping both education inequality and level in the regressions. We rejected the null of the test at 1 percent significance so decided to keep education level in the regressions.

Same Series

Taking Down the Wall: Transition and Inequality
Inequality and Growth: A Heterogeneous Approach
Growing (Un)equal: Fiscal Policy and Income Inequality in China and BRIC+
National Income and Its Distribution
Inequality, Gender Gaps and Economic Growth: Comparative Evidence for Sub-Saharan Africa
Rent Seeking and Endogenous Income Inequality
Inequality of Opportunity, Inequality of Income and Economic Growth
Measuring Income Inequality and Implications for Economic Transmission Channels
Sharing the Growth Dividend: Analysis of Inequality in Asia
Heterogeneity and Persistence in Returns to Wealth

Other Publishers

Asian development bank.

Trends and Drivers of Income Inequality in the Philippines, Thailand, and Viet Nam: A Decomposition Analysis
ADB Economics Working Paper Series No. 529: Human Capital Spending, Inequality, and Growth in Middle-Income Asia

Inter-American Development Bank

Inequality, Democracy, and Persistence: Is There a Political Kuznets Curve?
The 1990s in Latin America: Another Decade of Persistent Inequality
On Compulsory Voting and Income Inequality in a Cross-Section of Countries
Foreign Aid, Income Inequality, and Poverty
International Remittances and Income Inequality: An Empirical Investigation
Education, Family Background and Racial Earnings Inequality in Brazil
Can Financial Market Policies Reduce Income Inequality?
Migration and Education Inequality in Rural Mexico
Income Inequality and Economic Growth: Evidence from the American Data
Capital Openness and Income Inequality: Smooth Sailing or Troubled Waters?

International Fund for Agricultural Development

Asia's rural-urban disparity in the context of growing inequality.

International Labour Organization

World of Work Report 2008: Income inequalities in the age of financial globalization
Global Wage Report 2014/15: Wages and income inequality

The World Bank

Informal Sector Heterogeneity and Income Inequality: Evidence from the Democratic Republic of Congo
Persistent Misallocation and the Returns to Education in Mexico
Top Incomes and the Measurement of Inequality in Egypt
Egypt: Inequality of Opportunity in Education
Global Inequality in a More Educated World
Global income inequality: what it is and why it matters
Mexico: two decades of the evolution of education and inequality
Measuring education inequality: Gini coefficients of education
Life satisfaction and income inequality
Democracy and income inequality: an empirical analysis
Share on facebook Share on linkedin Share on twitter

View raw image
Download Powerpoint Slide

[185.39.149.46]
185.39.149.46

Character limit 500 /500

Growing Income Inequality Threatens American Education

America has always taken pride in being the land of opportunity, a country in which hard work and sacrifice result in a better life for one’s children. Economic growth has made that dream a reality for generations of Americans, including many people who started out poor. The quarter century following World War II was a golden era for the U.S. economy, as high- and low-income families shared the benefits of substantial economic growth. But storm clouds began to gather in the 1970s. In particular, computer-driven technological changes favoring highly educated workers, plus demographic shifts such as the rise of single-parent families, have produced sharply growing income gaps among families.

In the past, America’s public schools have responded well to the challenges of a changing world. Indeed, America’s world leadership in education has fueled much of its prosperity and made the 20th century the “American Century” (Goldin & Katz, 2008). But technological changes, globalization, and rising income inequality have placed great strains on the decentralized American approach to public education. We are constantly reminded that the math, science, and language skills of our children and young adults lag far behind those of children in other countries. In international rankings, our college graduation rate has fallen from first to 12th.

In this article—the first of two appearing in consecutive months—we describe the origins and nature of growing income inequality and some of its consequences for American children. We document the increased family income inequality that’s occurred over the past 40 years. An increase in income disparity has been more than matched by an expanding gap between the money that low- and high-income parents spend on enrichment activities for their children.

Most distressingly, increasing gaps in academic achievement and educational attainments have accompanied the growth in income inequality. Differences in the reading and math achievement levels of low- and high-income children are much larger than several decades ago, as are differences in college graduation rates.

What accounts for these widening gaps? Drawing from the first part of our recent book, Restoring Opportunity: The Crisis of Inequality and the Challenge for American Education (Harvard Education Press and the Russell Sage Foundation, 2014), we explain that the evidence supports pathways operating through both families and schools. In addition to growing differences in the resources spent by poor and rich families on their children, declining real incomes for low-income families have affected maternal stress, mental health, and parenting. storing the educational opportunities that children from low-income families need if they are to lead productive and fulfilling lives.

Rising residential segregation by income has led to increasing concentrations of low- and high-income children attending separate schools. Peer problems, geographic mobility, and challenges in attracting and retaining good teachers have made it difficult to provide consistently high-quality learning experiences in schools serving a large proportion of low-income students.

Next month’s article draws from the second part of Restoring Opportunity to describe ideas based on proven policy approaches that will enable the country to make progress on the enormous task of re-storing the educational opportunities that children from low-income families need if they are to lead productive and fulfilling lives.

Widening Gaps

Based on U.S. Census Bureau data, the left-hand bar in each set of bars in Figure 1 shows the average income in a particular year (in 2012 dollars) for children at the 20th percentile of the nation’s family income distribution. This means that, in a given year, 20% of children lived in families with incomes below that level, while 80% had incomes above it. In 1970, the dividing line was drawn at $37,664.

The middle bar in each set shows the average family income in a given year at the 80th percentile of the distribution, which was about $100,000 (in 2012 dollars) in 1970. The right-hand bar in each set shows the average income for very high-income families — those with incomes exceeding those of 95% of U.S. families (a little more than about $150,000 in 1970).

In contrast to the two decades before 1970, when the incomes of these three groups grew at virtually identical rates, economic growth over the next four decades failed to lift all boats. In 2010, family income at the 20th percentile was more than 25% lower than the inflation-adjusted corresponding family income in 1970. In contrast, the real incomes of families at the 80th percentile grew by 23% to $125,000 over these four decades, while the incomes of the richest 5% of families rose even more. Census Bureau data also show that the decline of the incomes of families at the lower end of the spectrum is reflected in the nation’s child poverty rate: Over 16 million U.S. children—more than 20%—were living in poor families in 2012, up sharply from the 15% child poverty rate in 1970.

During this same time period, the gap between the average reading and mathematics skills of students from low- and high-income families increased substantially. As illustrated in Figure 2, among children who were adolescents in the late 1960s, test scores of low-income children lagged behind those of their better-off peers by four-fifths of a standard deviation — which represents about 80 points on the scale used to measure SAT scores. Forty years later, this gap was 50% larger, amounting to nearly 125 SAT-type points (Reardon, 2011). We were surprised to discover how much the income-based gap grew during this period in view of the decline in the racial gap in test scores in the decades following the 1954 U.S. Supreme Court decision in Brown vs. Board of Education .

Given the importance of academic preparation in determining educational success, it should come as no surprise that growth in the income gap in children’s reading and mathematics achievement has contributed to growth in the corresponding gap in the rate of college completion (Figure 3, which is based on Bailey & Dynarski, 2011). Among children growing up in relatively affluent families, the four-year college graduation rate of those who were teenagers in the mid-1990s was 18 percentage points higher than the rate for those who were teenagers in the late 1970s. In contrast, among children from low-income families, the graduation rate was only 4 percentage points higher for the later cohort than for the earlier one. Analysts differ in their assessments of the relative importance of college costs and academic preparation in explaining the increasing gulf between the college graduation rates of affluent and low-income children in our country. However, both cost burdens and academic performance are rooted, at least in part, in the growth in family income inequality.

Inequality Affects Skills Attainment

American society relies on its families to nurture its children and its schools to level the playing field for children born into different circumstances. More than any other institution, schools are charged with making equality of opportunity a reality. During a period of rising inequality, can schools play this critical role effectively? Or has growing income inequality affected families, neighborhoods, and schools in a manner that undercuts the effectiveness of schools serving disadvantaged populations?

Very young children tend to be completely dependent on their families to provide what they need for healthy development (Duncan & Magnuson, 2011). Children growing up in families with greater financial resources score higher on many dimensions of school readiness upon entering kindergarten. An obvious advantage of a higher family income is that it provides more resources to buy books, computers, high-quality childcare, summer camps, private schooling, and other enrichments. In the early 1970s, high-income families spent just under $3,000 more per year (in 2012 dollars) on child enrichment than low-income families (Figure 4; Duncan & Murnane, 2011). By 2006, this gap had nearly tripled, to $8,000. Spending differences are largest for enrichment activities such as music lessons, travel, and summer camps (Kaushal et al., 2011). Differential access to such activities may explain the gaps in background knowledge between children from high-income families and those from low-income families that are so predictive of reading skills in the middle and high school years (Snow, 2002).

Parents also spend different amounts and quality of time interacting with their children. High-income parents spend more time than low-income parents in literacy activities with their children. Most disparate is time spent in “novel” places — other than at home, school, or in the care of another parent or a childcare provider. Between birth and age six, children from high-income families spend an average of 1,300 more hours in novel contexts than children from low-income families (Phillips, 2011). These experiences, financed by the higher incomes of more affluent families, also contribute to the background knowledge that is so critical for comprehending science and social studies texts in middle school.

It is difficult to untangle the precise effects of a multitude of family-related factors — income and expenditures, family structure, time, and language use — on the disparities in children’s school readiness and success that have emerged over the past several decades. But the evidence linking income to children’s school achievement suggests that the sharp increase in the income gap between high- and low-income families since the 1970s and the concomitant increase in the income-based gap in children’s school success are hardly coincidental.

In particular, two experimental studies in the 1970s examined the overall effects on children of income supplements that boosted family income by as much as 50% (Maynard, 1977; Maynard & Murnane, 1979). At two of the three sites, researchers found that children in families randomly assigned to receive an income supplement did significantly better with respect to early academic achievement and school attendance than children in families that received no income supplement.

Still more evidence on policy-relevant effects of income increases comes from a study that takes advantage of the increasing generosity of the U.S. Earned Income Tax Credit (EITC) between 1993 and 1997 to compare children’s test scores before and after the credit was expanded (Dahl & Lochner, 2012). The authors found increases in low-income children’s achievement in middle childhood that coincided with the EITC expansion.

The strongest research evidence appears to indicate that money matters in a variety of ways for children’s long-term success in school. While some children always have enjoyed greater benefits and advantages than others, the income gap has widened dramatically over the past four decades and, as these research studies suggest, this has been a significant factor in widening the gap in children’s school success as well.

Researchers have long known that children attending schools with mostly low-income classmates have lower academic achievement and graduation rates than those attending schools with more affluent student populations. Less well understood are the ways in which student body composition shapes school functioning and children’s developmental trajectories and long-run outcomes.

In recent decades, it has been largely through an increase in income-based segregation of neighborhoods and schools that growing inequality of family income has affected the educational attainments of the nation’s children. Residential segregation by income has increased substantially in recent decades, as high-income families buy homes in neighborhoods where less-affluent families cannot afford to live, and poor families are increasingly surrounded by neighbors who are poor as well (Reardon & Bischoff, 2011). This reduces interactions between rich and poor in settings ranging from schools and child care centers to libraries and grocery stores. Without the financial and human resources and political clout of the wealthy, institutions in poorer neighborhoods, including schools, may decline in quality.

Perhaps most important, increasing residential segregation by income has led to increasing school segregation by income. From 1972 to 1988, schools became more economically segregated, and teenagers from affluent families were less and less likely to have classmates from low-income families (Altonji & Mansfield, 2011). As a result, a child from a poor family is two to four times as likely as a child from an affluent family to have classmates in either elementary or high school with behavioral problems and low skills. This sorting matters because the weak cognitive skills and behavioral issues of many low-income children have a negative effect on their classmates’ learning.

Student mobility resulting from these residential changes poses another threat to achievement. Urban families living in poverty move frequently, and, as a result of school sorting by socioeconomic status, children from poor families are especially likely to attend schools with relatively high numbers of new students arriving during the school year. Recent research has shown that children attending elementary schools with considerable student mobility make less progress in mathematics than do children in schools with less student turnover. Moreover, these negative effects apply to students who themselves are residentially stable as well as to those who are not and are likely to stem from disruption of instruction caused by the entry of new students into a class (Raudenbush, Jean, & Art, 2011).

Poor teacher quality, too, contributes to the weak performance of students in high-poverty schools. A substantial body of research has shown that schools serving high concentrations of poor, nonwhite, and low-achieving students find it difficult to attract and retain skilled teachers. In addition to preferring schools with relatively low proportions of low-achieving students, teachers favor schools in neighborhoods with higher-income residents and less violent crime (Boyd et al., 2011). In high-poverty schools, teacher commitment, parental involvement, and student achievement all tend to be lower.

Yet another challenge facing many of the nation’s schools concerns the school placements of new immigrants, many of whom speak little English. Today’s immigrants are more likely than their predecessors in the early 1970s to come from high-poverty countries. Black and Hispanic immigrants to New York City are much more likely to be poor than are white immigrants from Eastern Europe, and they are more likely to attend elementary and middle schools with native-born black and Hispanic students who are poor (Schwartz & Stiefel, 2011). Thus, while immigrants are not segregated from the native-born in New York City schools, their residential patterns contribute to segregation of schools by socioeconomic status and race.

Helping Low-income Children

By widening the gap in educational opportunities between children from low- and higher-income families, increasing income inequality jeopardizes the upward socioeconomic mobility that has long held our pluralistic democracy together. Improving educational outcomes for children growing up in low-income families is therefore critical to the nation’s future and requires a combination of policies that support low-income families and measures to improve the quality of schools that low-income children attend.

The United States has implemented a range of policies to raise the buying power of low-income families, including the Child Tax Credit, the Earned Income Tax Credit, cash assistance programs, and the Supplemental Nutrition Assistance Program (formerly Food Stamps). Recent studies show that the increases in family incomes produced by these programs result in improved educational outcomes for young children and health in adulthood (Hoynes, Schanzenbach, & Almond, 2013). Unfortunately, these programs are under attack as Congress seeks ways to reduce the federal budget deficit.

Improving the quality of schools attended by low-income children poses even more important and difficult challenges. As a nation, we have failed to appreciate the extent to which technological innovations have brought changes in the skills needed to succeed in today’s economy. Moreover, the rising economic and social inequality produced by technology and globalization has weakened neighborhoods and families in ways that make effective school reform all the more difficult. For a variety of historical reasons, our nation has not learned how to provide the consistent supports that schools—especially those serving large numbers of low-income children—must have to succeed.

Discussions of school reforms often center on simplistic silver bullets: more money, more accountability, more choice, new organizational structures. None of these reforms has turned the tide because none focuses directly on improving what matters most in education: the quality and consistency of the instruction and experiences offered to students. In our companion article, which will appear next month, we detail the building blocks that we consider essential for an “American solution” to the serious problems facing our nation’s schools.

Altonji, J.G. & Mansfield, R. (2011). The role of family, school, and community characteristics in inequality in education and labor market outcomes. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 339-358). New York, NY: Russell Sage Foundation & Spencer Foundation.
Bailey, M.J. & Dynarski, S.M. (2011). Inequality in postsecondary education. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 117-132). New York, NY: Russell Sage Foundation & Spencer Foundation.
Boyd, D., Lankford, H., Loeb, S., Ronfeldt, M., & Wyckoff, J. (2011). The effect of school neighborhoods on teachers’ career decisions. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 377-396). New York, NY: Russell Sage Foundation & Spencer Foundation.
Dahl, G.B. & Lochner, L. (2012). The impact of family income on child achievement: Evidence from the earned income tax credit. American Economic Review , 102 (5), 1927-1956.
Duncan, G.J. & Magnuson, K. (2011). The nature and impact of early achievement skills, attention skills, and behavior problems. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 47-70). New York, NY: Russell Sage Foundation & Spencer Foundation.
Duncan, G.J. & Murnane, R. J. (2011). Introduction: The American dream, then and now. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 3-26). New York, NY: Russell Sage Foundation & Spencer Foundation.
Goldin, C.D. & Katz, L.F. (2008). The race between education and technology . Cambridge, MA.: Harvard University Press.
Hoynes, H.W., Schanzenbach, D.W., & Almond, D. (2012). Long run impacts of childhood access to the safety net. Unpublished NBER Working Paper No. 18535.
Kaushal, N., Magnuson, K., & Waldfogel, J. (2011). How is family income related to investments in children’s learning? In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 187-206). New York, NY: Russell Sage Foundation & Spencer Foundation.
Maynard, R.A. (1977). The effects of the rural income maintenance experiment on the school performance of children. American Economic Review, 67 (1), 370-375.
Maynard, R.A. & Murnane, R.J. (1979). The effects of a negative income tax on school performance: Results of an experiment. Journal of Human Resources, 14 (4), 463-476.
Phillips, M. (2011). Parenting, time use, and disparities in academic outcomes. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 207-228). New York, NY: Russell Sage Foundation & Spencer Foundation.
Raudenbush, S.W., Jean, M., & Art, E. (2011). Year-by-year and cumulative impacts of attending a high-mobility elementary school on children’s mathematics achievement in Chicago, 1995-2005. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 359-376). New York, NY: Russell Sage Foundation & Spencer Foundation.
Reardon, S.F. (2011). The widening academic achievement gap between the rich and the poor: New evidence and possible explanations. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 91-116). New York, NY: Russell Sage Foundation & Spencer Foundation.
Reardon, S.F. & Bischoff, K. (2011). Income inequality and income segregation. American Journal of Sociology, 116 (4), 1092-1153.
Schwartz, A.E. & Stiefel, L. (2011). Immigrants and inequality in public schools. In G.J. Duncan & R.J. Murnane (Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 419-442). New York, NY: Russell Sage Foundation & Spencer Foundation.
Snow, C. (2002). Reading for understanding: Toward a research and development program in reading comprehension. Santa Monica, CA: Rand Corporation.

All articles published in Phi Delta Kappan are protected by copyright. For permission to use or reproduce Kappan articles, please e-mail [email protected] .

Sign Up for The Savvy Principal

Edweek top school jobs.

Image shows a multi-tailed arrow hitting the bullseye of a target.

Sign Up & Sign In

Inequality undermines the value of education for the poor

Subscribe to the center for economic security and opportunity newsletter, melissa s. kearney and melissa s. kearney nonresident senior fellow - economic studies , center for economic security and opportunity , the hamilton project phillip levine phillip levine nonresident senior fellow - economic studies , center for economic security and opportunity.

March 16, 2016

High school dropout rates are higher in cities and states with greater income inequality . This does not just reflect the different demographics across places. As we document in our forthcoming contribution to the Brookings Papers on Economic Activity, children from lower socio-economic backgrounds are more likely to drop out if they live in a more unequal city or state. The question is: why? Perhaps children from lower socio-economic backgrounds perceive a lower return to staying enrolled in school. They might be correct.

Unequal places, unequal returns to schooling

Brad Hershbein’s recent blog (“ A college degree is worth less if you are raised poor ”) shows that gains to post-secondary education are lower for those from poorer backgrounds. Our own work points to an additional factor: inequality. Places with greater “lower-tail inequality” (the ratio of income at the 50th percentile of the income distribution to the 10th percentile) show the lowest wage gains to education for those from low-SES backgrounds.

Using data from the 1979 National Longitudinal Survey of Youth, we examine outcomes for children from three socio-economic categories, based on their mother’s level of education (no high school diploma, high school graduate, any college). Specifically, we measure the percentage wage increase associated with each additional year of school. We also compare results in states with low, high and medium levels of lower-tail income inequality. On average, an extra year of school is associated with a 10 percent higher wage. This is consistent with the broader research literature on the causal impact of education on earnings. But there is striking variation between states with different levels of income inequality:

In the more equal states, the wage gains associated with education vary only slightly by SES background. But there are big class gaps in the mid-range and high inequality states. In more unequal states, children from low-SES households see much lower rewards, in terms of wages, from each additional year of education. (An interactive map showing inequality rankings, along with dropout rates is available here .)

How might inequality impact returns to education?

This pattern has a number of possible explanations. Perhaps in more unequal states, schools attended by low-SES children are particularly weak, whereas in more equal states school quality is less varied. In unequal states, poor children might live in very isolated, segregated neighborhoods. Or perhaps there are simply fewer decently-paying jobs for high school graduates in less equal places, which low-SES young people struggle to secure. A better understanding of the complex relationships between inequality, dropout rates, and returns to education may help to address the broader challenge of reducing class gaps, and promoting upward mobility.

Explaining Achievement Gaps: The Role of Socioeconomic Factors

Racial achievement gaps in schools are well documented and remain a significant cause of concern in education. Troubling too is that the role of socioeconomic disparities in mediating these gaps remains unresolved.

To better understand the relationship between race and socioeconomic status (SES) in producing achievement gaps, SUNY Albany's Paul L. Morgan and Eric Hengyu Hu examine two waves of data from the federal Early Childhood Longitudinal Study. Results show that a broad set of family SES factors explains a substantial portion of racial achievement gaps: between 34 and 64 percent of the Black-White gap and between 51 and 77 percent of the Hispanic-White gap, depending on the subject and grade level.

While SES accounts for much of the racial achievement disparities, closing these gaps requires a comprehensive approach, including improving school quality and supporting family stability. As essential steps toward equity, the authors recommend investments in early childhood education and income supplements, such as expanding child tax credits. Download Explaining Achievement Gaps: The Role of Socioeconomic Factors or read the full report below.

By Michael J. Petrilli

In 2004, superstar economists Roland Fryer and Steven Levitt published a seminal paper , Understanding the Black-White Test Score Gap in the First Two Years of School. Using then-brand-new data from the federal Early Childhood Longitudinal Study, Kindergarten Class of 1998–1999 (ECLS-K), they found:

In stark contrast to earlier studies, the Black-White test score gap among incoming kindergartners disappears when we control for a small number of covariates. Real gains by Black children in recent cohorts appear to play an important role in explaining the differences between our findings and earlier research. The availability of better covariates also contributes. Over the first two years of school, however, Blacks lose substantial ground relative to other races. There is suggestive evidence that differences in school quality may be an important part of the explanation.

To say the findings were “mixed” dramatically underplays how good the good news was and how bad the bad news was.

The good news was twofold. First, as the authors wrote, Black kindergarteners at the time were making strong gains over previous cohorts. Indeed, child poverty dropped dramatically in the 1990s, particularly for Black children, and this was showing up in stronger readiness for school.

It was also good news—great, actually—that Fryer and Levitt could completely erase the racial achievement gap when controlling for “a small number of covariates.” These included some traditional measures of socioeconomic status (SES), such as family income and parental education levels, but also health-related factors, such as the child’s birthweight and births to teenage moms.

These findings are hugely consequential for America’s longstanding debates around racial inequality. They directly rebut the hateful arguments of white supremacists who posit that achievement gaps are a sign of Black Americans’ genetic inferiority. And they throw cold water on the claims by some on the far left that bigotry and racism in schools are at the heart of all racial disparities in student achievement in the U.S.

Instead, the explanation for racial achievement gaps is much more straightforward, though still tragic: The vast racial disparities in socioeconomic conditions and prenatal and early-life health experiences explain the achievement gaps we see between racial and ethnic groups, at least at school entry. That suggests, per the Fryer and Levitt analysis, that universal, race-neutral interventions designed to improve the academic, social, economic, and health conditions of the poor would lift all boats and would also narrow racial gaps. (Not that those interventions are easy or always obvious.)

But the bad news was really bad, too. Namely, once children entered school, Black students started losing ground, likely because the schools they attended were lower quality than the ones attended by White students, even after controlling for SES. Changing that fact has, of course, been a major focus of education reform.

That was twenty years ago, and those of us at the Thomas B. Fordham Institute were curious to see if anything had changed. We knew that racial achievement gaps had continued to narrow until the early to mid-2010s. And we knew that the federal government had released a newer ECLS dataset, the ECLS-K: 2011. We wondered: Had the relationship between socioeconomic achievement gaps and racial/ethnic achievement gaps shifted? Was the Black-White gap still growing during elementary school? And how did all of this look for the White-Hispanic gap (also explored by Fryer and Levitt) and for subjects beyond just reading and math?

To find out, we turned to SUNY Albany's Paul Morgan. Paul is one of the nation’s leading scholars on disparities in education and health care. He’s made a career out of shaking up conventional wisdom—for example, finding that Black students are actually less likely to be identified for many disability conditions (like specific learning disabilities) in analyses controlling for academic achievement. He understood the complex relationships between the variables we were interested in, plus had a great deal of experience with the ECLS data.

He worked with Eric Hengyu Hu, an education policy and postdoctoral researcher experienced in analyzing the two ECLS datasets. They got to work, diving into the data from the older and newer ECLS-K datasets. What they found was largely consistent with Fryer and Levitt’s study, although they were able to add some new understandings, as well.

Key Findings

At the heart of Hu and Morgan’s study is a set of “SES-Plus” variables.

Table F-1. Family SES measures included in the study

Note: These measures are included in or derived from the federal Early Childhood Longitudinal Study datasets. For more information on how these data are used, see .
	Mother’s education background
	Father’s education background
	Mother’s occupation prestige
	Father’s occupation prestige
	Household income
	Whom child lives with
	Cognitive stimulation
	Emergent literacy activities
	Parent-child activities
	Family rules for TV
	Parental warmth

Since we were most interested in understanding the relationship between socioeconomic status and racial achievement gaps, Hu and Morgan did not look at health-related covariates, such as child’s weight at birth, or the age of the mother at first child’s birth, which Fryer and Levitt had included. As a result, the racial achievement gap did not “disappear,” as it had for Fryer and Levitt. But it did decrease significantly, just by controlling for the “SES-plus” factors.

Here’s what they found. (See the main body of the study for more details.)

Finding 1: Taken together, family SES+ factors explain between 34 and 64 percent of the Black-White achievement gap (depending on subject and grade level) and between 51 and 77 percent of the Hispanic-White achievement gap.

Figure F-1. Family SES+ explains more of the Black-White achievement gap in first grade reading than in other subjects and grade levels.

Figure F-2. Family SES+ explains more of the Hispanic-White achievement gap than the Black-White achievement gap.

Finding 2: Household income and mother’s education are the SES+ factors that best explain Black-White and Hispanic-White achievement gaps, respectively.

Figure F-3. Among individual SES+ factors related to science achievement gaps, household income best explains the Black-White gap and mother’s education best explains the Hispanic-White gap.

Figure F-3. Among individual SES+ factors, household income best helps explain the Black-White gap in reading achievement and mother’s education best helps explain the Hispanic-White gap.

Finding 3: Family SES+ indicators, and the extent to which they explain racial/ethnic achievement gaps, are stable over time (1998-99 and 2010-11).

Table F-2. Various indicators of family SES+ are moderately correlated with each other across the two kindergarten cohorts.

Note: For all bolded correlation coefficients, their statistical p-values were smaller than 0.05 level.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) Mother’s education background
(2) Father’s education background
(3) Mother’s occupation prestige
(4) Father’s occupation prestige
(5) Household income
(6) Household structure
(7) Cognitive stimulation						0.01
(8) Emergent literacy activities
(9) Parent-child activities
(10) Family rules for TV						0.01
(11) Parental warmth			0.00	0.01

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) Mother’s education level
(2) Father’s education level
(3) Mother’s occupation prestige
(4) Father’s occupation prestige
(5) Household income
(6) Household structure
(7) Cognitive stimulation
(8) Emergent literacy activities
(9) Parent-child activities
(10) Family rules for TV
(11) Parental warmth		0.01	0.01	0.01	0.01

Finding 4: The inclusion of family SES+ helps explain racial and ethnic excellence gaps.

Figure F-4. Family SES+ factors explain between 36 and 60 percent of the Black-White excellence gaps.

Figure F-5. Family SES+ factors explain between 52 and 69 percent of the Hispanic-White excellence gaps.

Making sense of the findings

These findings are generally consistent with Fryer and Levitt’s study from two decades ago. Socioeconomic factors can explain a large proportion of racial achievement gaps. But the current study adds a great amount of detail and nuance to our understanding of the relationships at play, while raising new questions:

1. How can we explain the different patterns for the Black-White achievement gap for reading, on the one hand, and math and science, on the other? Why is first grade reading such an outlier, given that it’s the only subject and grade combination where we see SES explaining a majority of the Black-White gap (about two-thirds)—especially when we combine that pattern with the finding that the Black-White reading gap continues to grow as students make their way through school? Here’s one hypothesis: As scholars, including E.D. Hirsch, Jr., have long argued, initial reading skills are more closely correlated to family SES than are math and science ones. This is likely because parents play a larger role, especially in a child’s first five years, in transmitting language abilities than they do for math and science. That can occur via behaviors, such as reading to their children, but also through their own use of verbal language. The advantages of high SES—and disadvantages of low SES—thus show up more for students’ initial reading skills than for their math and science ones. As students get older and benefit from classroom instruction, their relative advantages and disadvantages start to matter less. As Paul Morgan explains, “Children from higher SES families, who are disproportionately White and Asian, have a head start in terms of acquisition of early reading skills, so their better reading abilities show up early on the reading achievement measures. Over time, those from lower SES families acquire these early reading skills, including while attending school, and so the SES/racial gap narrows and begins to approximate those in the other subjects.” That’s good news from an equity perspective, but let’s not forget that the Black-White achievement gap (including in reading) continues to grow as students age through elementary school. Consistent with Fryer and Levitt’s paper, that likely means that we still haven’t closed the “school quality gap” between Black students and their White peers.

2. Why does SES explain so much more of the Hispanic-White gap than the Black-White gap? One explanation might be that Hispanic children being raised in Spanish-speaking families have latent potential that is obscured by their lack of English skills (which become stronger as the grade level increases). It may also be helpful to ponder what might be included in the “not SES” category. As explained earlier, possibilities include health-related factors, such as low-birth weight and being the child of a teenage mom—factors related to poverty that affect Black children more than their Hispanic peers. [1] It might also include various forms and effects of racism and bias, which might affect Black children at higher rates. For lower-income Black children, who are more likely to experience deep, persistent poverty than other groups, the combination of “adverse childhood experiences” might exacerbate inequalities. And for middle class Black children, bias, stereotype threat, and related factors might be especially at play. This might also be why the Black-White achievement gap grows over the course of elementary school, while the Hispanic-White gap shrinks. As Eric Hengyu Hu pointed out, “research by von Hipper et al. (2018) using both old and new ECLS data found school years tend to equalize early-grade Hispanic-White gaps but not Black-White gaps.” That might be because of the greater challenges Black students face outside of school, but it is likely also because of their inequitable access to effective schools.

3. What’s the role of household structure in the Black-White and Hispanic-White gaps? Hu and Morgan find that “family structure explains between 1 and 22 percent of the gaps, but is more important for explaining the Black-White achievement gap (10 to 22 percent of the gap explained) than the Hispanic-White achievement gap (1 to 4 percent of the gap explained).” That makes sense, given that Hispanic students are far more likely than their Black peers to live in two-parent families (74 percent versus 40 percent, respectively)—a rate which is much closer to that for White children (84 percent).

But these findings likely understate the role of family structure, especially for Black children, given the relationship between the number of parents in the household and household income. As shown in Table F-2, there’s a correlation of 0.32 between these two variables for the latest ECLS cohort, which is quite strong. On top of the many non-material benefits of growing up with two loving parents, it’s clearly the case that two incomes are usually better than one when it comes to boosting families out of poverty. And increasing the proportion of two-parent, two-income families in the Black community would thus help to narrow the Black-White achievement gap, as well.

None of this lends itself to simple takeaways, but the authors’ recommendations in the report—especially their suggestion to invest in early childhood education and to supplement families’ incomes, perhaps via an expanded child tax credit—deserve serious consideration.

As has been clear since the Coleman Report, when it comes to the interplay between race, poverty, and schooling, the honest read is that it’s complicated. What’s undeniable, though, is that much hard work remains, especially when it comes to providing effective schools to marginalized students, especially those who are Black. Let’s keep at it.

Introduction

Significant racial and ethnic achievement gaps exist between students in the U.S. by elementary school. [2] However, the underlying causes for these achievement gaps differ. [3] Thus, a better understanding of why racial/ethnic achievement gaps occur can help inform policies that promote educational and societal opportunities for all students.

One factor for racial/ethnic achievement gaps is between-group differences in socioeconomic status (SES), particularly exposure to poverty. For example, Black and Hispanic students perform, on average, at significantly lower levels academically than Asian and White students, which is primarily because Black and Hispanic students are more likely to grow up in less-resourced homes and neighborhoods. [4] According to this explanation, racial/ethnic achievement gaps result from socioeconomic factors; therefore, addressing these gaps would emphasize race-neutral policies and practices that lessen the negative effects of economic adversity.

Moreover, other factors contributing to racial and ethnic achievement gaps include bias, cultural insensitivity, stereotypes, and individual and systemic racism. Here, socioeconomic factors are simply one part of the story. [5] For example, why else would upper-middle-class Black students tend to perform worse than upper-middle-class White and Asian students? Or why do achievement gaps among fourth graders persist even when accounting for exposure to economic adversity? [6] This all suggests the need not for race-neutral but race-conscious policies (e.g., ensuring that Black or Hispanic students are taught by Black or Hispanic teachers and introducing ethnic studies curricula during K–12 schooling, using affirmative action in higher education) to address racial and ethnic achievement gaps.

Our understanding of the extent to which SES explains racial and ethnic achievement gaps during elementary school is limited in several important aspects. That is, available research mainly analyzed cross-sectional data rather than longitudinal data, used imprecise measures of SES (e.g., receipt of free or reduced-price lunch status), examined achievement gaps in certain academic subjects while excluding others (e.g., reading but not mathematics and science), and did not assess how SES may have changed as an explanatory factor across different cohorts of U.S. elementary students. [7]

Our study examines the extent to which socioeconomic factors explain gaps in reading, mathematics, and science achievement among racial and ethnic groups of U.S. elementary students. We use four macro- and eleven micro-level measures of family background to identify factors that best explain these achievement gaps. Our analyses include descriptive statistics and regression models. The results provide nonexperimental evidence of factors that might be the focus of experimentally assessed policies and practices attempting to lessen racial and ethnic achievement gaps in U.S. elementary schools. They also help determine the extent to which SES explains these gaps. Furthermore, we expect that the findings of this study will provide insights into the ongoing discussion of whether race-neutral or race-conscious policies are more effective in addressing these gaps.

We examine the following research questions:

To what extent does a broad set of family SES indicators explain initially observed racial and ethnic achievement gaps?
To what extent do specific family SES indicators explain racial and ethnic achievement gaps?
To what extent do the family SES indicators correlate, and how have they changed over time?
To what extent does family SES help explain racial and ethnic disparities among high achievers?

A Broader View of Socioeconomic Status

Researchers often used receipt of free or reduced-price lunch or household income to represent a family’s SES. [8] However, SES is most certainly a much broader factor, encompassing social patterns and aspects of family life that may relate to, but are not solely dependent on, household income. In this study, we used federal data on two cohorts of kindergarten students. Accordingly, our report included eleven indicators of a student’s family life (which can be aggregated into four key factors) for a more detailed view of the relationship between SES and student racial or ethnic background and academic achievement, including but not limited to the family’s household income (Table 1).

Table 1. Family SES measures included in the study

Note: These measures are included in or derived from the federal Early Childhood Longitudinal Study datasets. For more information on how these data are used, refer to the section.
	Mother’s education background
	Father’s education background
	Mother’s occupation prestige
	Father’s occupation prestige
	Household income
	Whom child lives with
	Cognitive stimulation
	Emergent literacy activities
	Parent-child activities
	Family rules for TV
	Parental warmth

Initial Racial and Ethnic Gaps in Academic Achievement

Before delving into the role of SES in academic achievement, it’s essential to first understand the existing racial and ethnic gaps in general. Figure 1 presents data from the federally administered Early Child Longitudinal Study (2010-11 kindergarten cohort). The figure demonstrates the racial and ethnic gaps in assessment scores using the largest student group (White students) as the reference group. On average, Black and Hispanic students score substantially lower than White students in all subjects, whereas Asian students score slightly higher than White students. Figure 2 illustrates the disparities between different ethnic groups in terms of math and reading scores. Regarding math scores, the Black-White gap tended to grow throughout elementary school, whereas the Hispanic-White gap narrowed slightly.

Figure 1. Racial/ethnic gaps in student achievement in fifth grade are substantial.

Figure 2. The Black-White achievement gap grows across elementary grades.

Disentangling Race and Class

SES factors, including parental education, income, and occupation, strongly predict children’s academic achievement, [9] with higher SES consistently associated with greater academic achievement. [10] Being from a higher SES family undoubtedly provides students with many advantages, such as greater access to higher-quality educational resources, enriched learning environments, and increased parental time and involvement in their education. [11] Prior studies suggested that parental education plays an outsized role in shaping children’s academic trajectories. This could be because parental education is often associated with a stronger emphasis on the value of education, which may lead to more positive learning environments for children. [12] There’s also the possibility that adults who have the skills—cognitive and otherwise—to persist in their own educational attainment are likely to bequeath similar skills to their children.

A challenge often encountered during the analysis of social patterns is the presence of many factors that may correlate with a given outcome, as well as explanatory factors that may correlate with each other and other factors. Analyzing differences in academic achievement by race, ethnicity, or family SES background highlights this problem. For example, Figure 3 depicts how household income varies for students in the 2010-11 ECLS-K kindergarten cohort. About half of White students (49 percent) and Asian students (49 percent) are being raised in families who are in the top three income categories. In contrast, less than one in five Black students (17 percent) or Hispanic students (17 percent) come from such families.

Instead, Black and Hispanic students are much more likely to live in poverty than their White and Asian peers. Most Black students (58 percent) and Hispanic students (56 percent) come from families in the bottom three income categories. Only one in five White students (19 percent) and one-fourth of Asian students (24 percent) come from families with the lowest income levels.

Figure 3. Household income varies across racial and ethnic groups from the kindergarten cohort of 2010-11.

As stated above, economic factors such as household income are only one aspect of SES, as there are other SES factors associated with race and ethnicity. For example, household structure is a factor that measures whether a child lives with a single parent, two parents, or other guardians. Figure 4 illustrates the significant variation in household structure among different racial and ethnic student groups. That is, 93 percent of Asian students and 86 percent of White students live in two-parent households. On the other hand, just 48 percent of Black students do. Moreover, Hispanic students, who are nearly as likely as their Black peers to live at low-income levels (Figure 3), have a significantly higher probability of living in two-parent families than those peers (79 percent versus 48 percent, respectively).

Figure 4. Household structure varies across racial and ethnic groups, as per the kindergarten cohort of 2010-11.

The correlations among family SES variables are often strong, but not always. Table 2 lists the correlation coefficients between family SES factors and additional home environment measures (collectively, we refer to the SES factors and the home environment factors as “SES+”). These coefficients have a possible range from +1 (perfect direct correlation) to -1 (perfect inverse correlation). All the (bolded) statistically significant correlation coefficients in Table 2 are positive except for the relationship between household structure and parental warmth, which indicates a very weak negative correlation (-0.06). Apart from the parental warmth factor, all relationships between SES+ factors are positive, but they range from practically and statistically insignificant positive correlations (e.g., the 0.01 coefficient for household structure and family rules for TV) to strong positive correlations (e.g., the 0.65 correlation for mother’s educational background and father’s educational background).

Table 2. Various indicators of family SES+ are positively associated with each other.

Note: The statistical p-values were smaller than 0.05 for each correlation coefficient in bold.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) Mother’s educational background
(2) Father’s educational background
(3) Mother’s occupational prestige
(4) Father’s occupational prestige
(5) Household income
(6) Household structure
(7) Cognitive stimulation						0.01
(8) Emergent literacy activities
(9) Parent-child activities
(10) Family rules for TV						0.01
(11) Parental warmth			0.00	0.01

These analyses suggest that students’ race and ethnicity, family SES+, and academic achievement are interrelated in multiple ways. Throughout the rest of this study, we will examine to what extent accounting for family SES+ helps explain initially observed racial and ethnic achievement gaps.

Data and Methods

This report uses federal data from the public-use version of the Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 [13] (ECLS-K:1998-99), and the Early Childhood Longitudinal Study, Kindergarten Class of 2010-11 [14] (ECLS-K:2010-11). The former is a longitudinal study that tracks the same cohort of children from kindergarten through the eighth grade. The latter follows a different cohort of children from kindergarten through the fifth grade. Both datasets have extensive information on student-level academic achievement, sociodemographic characteristics, and home environments for children who entered kindergarten during the fall of 1998 and spring of 1999, as well as the fall of 2010 and spring of 2011. The total number of students included in ECLS-K:1998-99 was 21,409, while that in ECLS-K:2010-11 was 18,174.

Table 3 displays the frequency at which data were gathered for students participating in each study.

Table 3: Data gathering intervals for ECLS-K:1998-99 and ECLS-K:2010-11

The ECLS-K datasets include individually assessed reading, mathematics, and science achievement measures in each grade reported on a consistent scale. [15] We used scores from first, third, and fifth grade reading, mathematics, and science assessments as continuous measures of academic achievement. See the Appendix for a description of the content assessed on each of the three subject-specific tests.

The ECLS-K datasets include the parent-reported race and ethnicity of individual students. Possible responses included the following: White, non-Hispanic; Black/African American, non-Hispanic; Hispanic, race specified; Hispanic, no race specified; Asian, non-Hispanic; Native Hawaiian or Other Pacific Islander, non-Hispanic; American Indian or Alaska Native, non-Hispanic; and more than one race, non-Hispanic. We combined the responses of Hispanic, race specified, and Hispanic, no race specified, into one Hispanic group. We also merged the categories of American Indian or Alaska Native, non-Hispanic, and Native Hawaiian or Other Pacific Islander into one American Indian, Native American, Pacific Islander (AINAPI) group.

We constructed family SES measures from parent surveys in the fall or spring of the student’s kindergarten year. The highest education level for mothers and fathers included five categories. These categories are as follows: (a) not high school graduate (reference group); (b) high school graduate or equivalent (e.g., GED); (c) some college; (d) bachelor’s degree; and (e) master’s degree or higher. The occupational prestige scores for mothers and fathers were measured based on occupations coded using the “Manual for Coding Industries and Occupations,” which was created for the National Household Education Surveys Program and uses an aggregated version of occupation codes. [16] Household income was derived from parental reports and divided into 18 categories ranging from $5,000 or less to $200,001 or more. [17] The household structure variable included four categories that refer to who raises the student: (a) two parents (b) single mother; (c) single father; and (d) other guardians (e.g., grandparents). [18]

Additionally, we considered five broad measures of “home opportunity factors.” These factors are used to assess parent engagement and aspects of the home environment, including (a) cognitively stimulating activities (e.g., playing games or doing arts and crafts); (b) emergent literary activities (e.g., reading to your child; number of books the child owns); (c) parent-child activities (e.g., visits to the zoo, bookstore, or library); (d) parental warmth (e.g., expressions of love and affection); and (e) family TV rules (e.g., how much time the child is allowed to watch TV and when). Additional description of these measures can be found in the Appendix .

We also created a composite variable called “SES+” by combining the family SES variables, including household income and mother’s occupational status, with the home environment variables, such as parental warmth and emergent literacy activities.

This report includes an analysis of student race and ethnicity, family SES and home environment measures, and three measures of academic achievement. Here, we analyzed reading, mathematics, and science scores separately from the spring of first, third, and fifth grade across the two cohorts. The fall or spring kindergarten measurements are the primary predictors. The findings are derived from correlation and regression analyses. [19] We used sampling weights to ensure that the results were nationally representative. Additional details of the analyses can be found in the Appendix .

A vital aspect of this analysis involves examining the reduction of the racial and ethnic achievement gaps after including family SES factors in the regression models. Each regression was run twice: once without family SES+ factors and again with SES+ factors included in the model. We refer to these in shorthand as “reduction rates,” which is synonymous with “percentage of achievement gap explained by SES+” and represents the coefficient for the race/ethnicity variable in the second model divided by the coefficient for the race variable in the first model.

For instance, for the 2010-11 kindergarten cohort, the Black-White reading gap in first grade is -0.45 SD and statistically significant (p < .001). Including the mother’s educational background in the regression reduces the estimated Black-White reading gap in first grade to -0.29 SD, which continues to be considerably significant (p < .001). We calculated this reduction as a percentage decrease as follows: (0.45 - 0.29) / 0.45 * 100 = 36 percent reduction. Thus, statistically adjusting for the mother’s education background reduces the estimated Black-White reading gap in first grade by 36 percent for the kindergarten cohort of 2010-11.

This section examines the degree to which family SES+, including all eleven indicators, explains the racial and ethnic achievement gaps in analyses across grades, subjects, and ECLS-K cohorts. We focus on gaps among the three largest racial and ethnic student groups, including the Black-White and Hispanic-White gaps, for which we compare reduction rates across ECLS cohorts and grades. (Again, these rates refer to the difference in the effect size of racial and ethnic categories before and after including family SES+ factors in the regression models.)

Figure 5 shows that the inclusion of SES+ factors explains nearly two-thirds of the first grade Black-White reading achievement gap but less than half of those gaps in fifth grade reading and other subjects, regardless of grade level.

For the Black-White reading achievement gap, the reduction rate is notably decreased from first to fifth grade (64 percent to 48 percent). This suggests that SES+ is somewhat less influential in later grades for reading. This could be either because of a lengthening time interval between the measurement of the two factors or because family SES+ became increasingly less predictive of reading achievement as Black students age.

For the Black-White mathematics and science achievement gaps, the role of SES+ remains stable across grades. (For analysis of the role of SES+ in explaining Black-White achievement gaps in the earlier ECLS-K cohort, see the Appendix , Figure A1 .)

Figure 5. Family SES+ explains more of the Black-White achievement gap in reading than in other subjects.

Figure 6 shows that the Hispanic-White achievement gap is considerably better explained by SES+ factors than the Black-White achievement gap. All the analyses show that SES+ factors explain more than half of the achievement gap, and in some analyses, SES+ factors explain about three-fourths of the gaps, namely, in first grade reading (74 percent) and fifth grade reading (77 percent).

Figure 6. Family SES+ explains more of the Hispanic-White achievement gap than the Black-White achievement gap.

SES+ factors better explain the Hispanic-White achievement gaps in fifth grade than in first grade, regardless of subject. For mathematics and science achievement gaps, the difference across grades is more substantial than in reading. SES+ factors explain 59 percent and 51 percent of the math and science gaps, respectively, in first grade, but they explain 67 percent and 66 percent of those gaps, respectively, in fifth grade. (For analysis of the role of SES+ in explaining Hispanic-White achievement gaps in the earlier ECLS-K cohort, see the Appendix , Figure A2 .)

To summarize, including the SES+ explanatory factors in an analysis of racial and ethnic achievement gaps reduces the estimated size of the gaps, but they remain. This is particularly evident in the Black-White achievement gap, where SES+ factors generally explain less than half of the gap in mathematics, science, and to some extent, reading.

Finding 2: Household income and mother’s education are the SES+ factors that best help explain the Black-White and Hispanic-White achievement gaps, respectively.

Next, we break down the measures of SES+, evaluating the extent to which they individually explain racial and ethnic achievement gaps. These analyses are designed to determine which of the study’s different family SES+ indicators best explains the observed racial and ethnic achievement gaps. Each measure of SES+ was incorporated separately into our regression model. Continuing our focus on the three largest student racial/ethnic groups, we examined the achievement of first-grade students in the more recent ECLS-K cohort. (Similar findings were observed for other grade levels and the earlier ECLS cohort.)

Overall, household income and mother’s education are the two SES+ factors that best help explain the achievement gaps. Figure 7, Figure 8, and Figure 9 depict the extent to which each SES+ factor accounts for the racial and ethnic achievement gap in first grade for each subject, respectively. Household income is the primary SES+ factor for explaining the Black-White achievement gap in all analyses , explaining between 30 percent to 56 percent of the Black-White gap, depending on the subject. It also important for explaining the Hispanic-White achievement gap, explaining between 29 and 45 percent of the gap, depending on the subject. Family opportunity factors, such as emerging literacy activities and family rules for television, explain very little of the racial and ethnic achievement gaps in all analyses.

Mother’s education is the most critical SES+ factor for explaining the Hispanic-White achievement gap in all analyses , accounting for 37 percent to 55 percent, depending on the subject. Moreover, this factor significantly explained the Black-White achievement gap, with values ranging between 20 and 36 percent, depending on the subject.

Father’s education is also an essential family SES+ factor for explaining these gaps . It is just as important as the mother’s education for explaining the Black-White achievement gap, explaining 20 to 36 percent of the gap, depending on the subject. It is also of similar significance to household income for explaining the Hispanic-White achievement gap, explaining 29 to 43 percent of the gap, depending on the subject.

Compared to the other four SES+ factors, parent occupational prestige and household structure are less influential in explaining racial and ethnic achievement gaps . Mother’s occupational prestige explains between 13 and 25 percent of the gaps, depending on race and subject. In contrast, father’s occupational prestige explains 9 to16 percent of the gaps, depending on race and subject. Family structure explains between 1 and 22 percent of the gaps . However, it explains the Black-White achievement gap (10 to 22 percent) better than the Hispanic-White achievement gap (1 to 4 percent).

Figure 7. Among individual SES+ factors, household income best explains the Black-White gap in reading achievement and mother’s education best explains the Hispanic-White gap.

Figure 8. Among individual SES+ factors, household income best explains the Black-White gap in math achievement and mother’s education best explains the Hispanic-White gap.

Figure 9. Among individual SES+ factors, household income best explains the Black-White gap in science achievement and mother’s education best explains the Hispanic-White gap.

Overall, the differences in the results between the two ECLS-K cohorts, which are 12 years apart, are relatively minor, implying that the relations between the variables of interest did not change substantially over this period. These similarities are shown in the analysis above, and they also apply to individual elements of SES+ and the relationships between the SES+ factors.

Table 4 displays the similarities in each component of SES+ across students in the ECLS-K:1998-99 and ECLS-K:2010-11 cohorts. For example, the share of students whose parents had “some college” education is identical across cohorts. However, there were some slight differences. Parental education levels increased between cohorts, with the share of mothers with bachelor’s degrees rising from 16 percent to 20 percent and the share of parents with graduate degrees rising even more (e.g., 5 percent to 10 percent for mothers). The percentage of students living in two-parent households increased from 74 percent to 79 percent between the two cohorts.

Table 4. Family SES+ components did not change substantially across the two cohorts.

Notes: Sampling weight was applied to generate nationally representative results. Percentages or means with standard deviation (SD) are reported. ECLS-K:2010-11 = Early Childhood Longitudinal Study: Kindergarten Class of 2010-11; ECLS-K:1998-99 = Early Childhood Longitudinal Study: Kindergarten Class of 1998-99. Family opportunity factors are standardized measures incorporating information from a variety of questions. As a result of standardization, all family opportunity factors have a mean of zero and a standard deviation of one and, therefore, are excluded from this table.
Multidimensional measures of family SES+	ECLS-K:1998-99		ECLS-K:2010-11
	Mother	Father	Mother	Father
	%/Mean (SD)

Not high school graduate	15%	14%	14%	14%
High school graduate	37%	38%	29%	33%
Some college	27%	21%	27%	21%
Bachelor’s degree	16%	17%	20%	20%
Master’s degree or higher	5%	10%	10%	12%
	43.10 (11.01)	42.68 (10.74)	44.45 (11.87)	43.14 (10.96)

$5,000 or less	3%		3%
$5,001 to $10,000	4%		4%
$10,001 to $15,000	7%		6%
$15,001 to $20,000	7%		7%
$20,001 to $25,000	7%		8%
$25,001 to $30,000	9%		5%
$30,001 to $35,000	7%		5%
$35,001 to $40,000	7%		5%
$40,001 to $45,000	11%		3%
$45,001 to $50,000			4%
$50,001 to $55,000	18%		3%
$55,001 to $60,000			3%
$60,001 to $65,000			3%
$65,001 to $70,000			3%
$70,001 to $75,000			4%
$75,001 to $100,000	10%		13%
$100,001 to $200,000	8%		17%
$200,001 or more	3%		4%

Two parents	74%		79%
Single mom	21%		17%
Single dad	2%		1%
Other guardians	3%		2%

Table 5 depicts the correlation coefficients demonstrating the relationships between SES+ factors for both cohorts. As discussed above, these coefficients can range from +1 (perfect positive relationship) to -1 (perfect negative relationship). The coefficients in Table 5 range from weak positive correlations (e.g., for household structure and parent educational background) to strong positive correlations (e.g., for mother’s educational background and father’s educational background).

These relations are quite similar across the cohorts, with all correlations between SES+ factors in a similar range from one cohort to the other. For example, the relationship between household structure and mother’s education is 0.16 in the later cohort and 0.19 in the earlier cohort (column 1). The reduction rates, such as those shown in the analysis above, are also stable across cohorts ( see Appendix , Figures A1 and A2 ).

Table 5. Various indicators of family SES+ are moderately correlated with each other across the two kindergarten cohorts.

Note: The statistical p-values were smaller than 0.05 for each correlation coefficient in bold.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) Mother’s education background
(2) Father’s education background
(3) Mother’s occupation prestige
(4) Father’s occupation prestige
(5) Household income
(6) Household structure
(7) Cognitive stimulation						0.01
(8) Emergent literacy activities
(9) Parent-child activities
(10) Family rules for TV						0.01
(11) Parental warmth			0.00	0.01

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) Mother’s education level
(2) Father’s education level
(3) Mother’s occupation prestige
(4) Father’s occupation prestige
(5) Household income
(6) Household structure
(7) Cognitive stimulation
(8) Emergent literacy activities
(9) Parent-child activities
(10) Family rules for TV
(11) Parental warmth		0.01	0.01	0.01	0.01

Finally, we examine whether the racial/ethnic achievement gaps are moderated by SES+ factors differently based on student performance levels. Differences in the proportions of student groups within the highest education levels are often termed “excellence gaps.” [20] This analysis uses the top quartile as the cutoff point to determine whether each student is an advanced achiever in reading, mathematics, or science in the first and fifth grades. (The analysis is limited to the kindergarten 2010-11 cohort.)

Figure 10 illustrates that family SES+ factors explain 60 percent of the reading excellence gap in the first grade and half in the fifth grade. SES+ factors account for 38 to 45 percent of the math excellence gaps and 36 to 45 percent of the science excellence gaps for Black and White students, depending on the grade.

Figure 10. Family SES+ factors explain 36 to 60 percent of the Black-White excellence gaps.

Figure 11 shows that the inclusion of SES+ factors explains a larger share of Hispanic-White excellence gaps than Black-White excellence gaps across the board. More than half of the Hispanic-White excellence gap is explained by SES+ in every subject and grade. Furthermore, more than two-thirds of these gaps are explained by SES+ in reading.

Figure 11. Family SES+ factors explain between 52 and 69 percent of the Hispanic-White excellence gaps.

Our findings suggest that students’ SES and home factors help to explain initially observed racial and ethnic achievement gaps. In many cases, the analyzed SES+ factors explain more than half of racial and ethnic achievement gaps. At the same time it is evident that SES, no matter how broadly construed, does not fully explain the racial gaps. SES+ can be less predictive over time, and it was found to be a less explanatory factor for racial than ethnic achievement gaps.

Educational policy solutions should reflect this complexity, as well as the comprehensive nature of the problem. Any number of well-executed policies would likely narrow achievement gaps of all kinds. Below, we present a few ideas, none of which are novel but all of which might help. Most are not cost-neutral. Moreover, we emphasize race-neutral policies in light of our findings, which reveal that family SES+ helps to substantially or fully explain racial and ethnic disparities in achievement. Race-conscious policies might also be helpful in further reducing these achievement gaps.

Whatever the approach, there is no denying the urgency of making the U.S. educational system more equitable.

The following are the proposed solutions:

1. Support programs to help parents earn their high school diplomas or higher education credentials: Because parental education, especially that of mothers, strongly correlates with children’s academic success, policymakers should consider increasing access to adult education and lifelong learning opportunities. This could include funding for adult education classes, online learning platforms, and community college courses. [21]

2. Focus on early childhood education: Because achievement gaps are already evident by elementary school, including as early as kindergarten, investing in high-quality early childhood education programs, especially in underprivileged communities, may be beneficial in mitigating the effects of socioeconomic disparities. [22]

3. Provide economic support and financial aid for low-income families: Income support programs that provide financial assistance should be implemented or enhanced to ensure that low-income families have the necessary resources to support their children’s education. [23]

4. Address racial and ethnic disparities: Policies that directly address the racial and ethnic achievement gaps should be developed and implemented, including the adoption of curricula that reflect diverse cultures and programs that specifically support underrepresented students. There is some evidence to indicate that student-teacher racial and ethnic matching may be of benefit, although whether such matching will address racial and ethnic disparities in achievement during elementary school is still unclear. [24]

The time to act is now. By enacting comprehensive and inclusive policies, we can narrow achievement gaps and create a more just educational landscape for the next generation.

This appendix provides further information related to the methodology and supplementary documentation of findings.

Additional Notes on Methodology

Description of Assessment Measures

Reading Achievement. The reading assessment included questions measuring basic skills (print familiarity, letter recognition, beginning and ending sounds, rhyming words, word recognition), vocabulary knowledge, and reading comprehension. Reading comprehension questions asked the child to identify information specifically stated in the text (e.g., definitions, facts, supporting details), make complex inferences within and across texts, and consider the text objectively and judge its appropriateness and quality.

Mathematics Achievement. The mathematics assessment was designed to measure skills in conceptual knowledge, procedural knowledge, and problem-solving. The assessment consisted of questions on number sense, properties, and operations; measurement; geometry and spatial sense; data analysis, statistics, and probability; and patterns, algebra, and functions.

Science Achievement . The science assessment included questions about physical sciences, life sciences, Earth and space sciences, and scientific inquiry. Meanwhile, for ECLS-K:1998-99 dataset, in the spring of first grade, student’s general knowledge was measured, which consisted of items that assessed knowledge in the natural sciences and social studies on a single scale. The social studies subdomain included questions that measured children’s knowledge in a wide range of disciplines, such as history, government, culture, geography, economics, and law. The science subdomain included questions from the fields of earth, space, physical, and life sciences.

Description of Home Opportunity Factors

Cognitive stimulation was a standardized sum of nine questions that assessed the frequency that parents engaged in activities with their children during a typical week. The activities included storytelling, singing, arts and crafts, playing games or puzzles, engaging in science projects or discussing nature, playing with construction toys, performing household chores, exercising or playing sports, and practicing reading, writing, or numeracy skills.

Emergent literacy was a standardized composite score of five items that evaluated literacy activities. Three items assessed the frequency of parental engagement in book reading and picture book reading with their children, as well as the children’s reading activities outside of school. Two items reported the number of books the children owned and the amount of time parents spent reading to their children. We combined the standardized scores of the first three items with those of the final two items to create the standardized composite score.

Parent-child activities was a standardized composite score of six items that assessed the frequency of parent-child engagement in activities over the prior month, including visits to libraries, bookstores, art galleries, concerts, zoos, and sports events. Twelve additional questions evaluated whether children participated in extracurricular activities, such as academic programs (e.g., tutoring or math lab), lessons in dance, music, drama, art, or crafts, organized athletic or club programs, volunteer work, and other forms of instruction (e.g., non-English language classes or religious instruction).

Family TV rules was a standardized composite of three binary questions indicating whether the family had established rules regarding: allowable TV programs, the maximum number of hours children could watch TV, and what time of day children could watch TV.

Parental warmth was a four-item scale that asked parents to self-assess their relationship with their children, specifically assessing their expressions of love, affection, quality time spent together, and child-parent closeness. These items were originally scaled from one to four, indicating “completely true” to “not at all true.” We reverse-coded the responses so that higher scores indicated greater warmth.

Description of Statistical Analyses

We conducted ordinary least squares (OLS) regression using the continuous version of outcomes, where we regressed student’s reading, mathematics, and science achievement from first, third, and fifth grade on student’s race or ethnicity. Then we added different family SES indicators separately, then together, through a serious of models for each grade level and subject. Each model incrementally adds variables to parse out their unique contributions. Model 1 begins with race or ethnicity as the main predictor. Models 2 to 12 added each SES indicator of mother/father’s education background, mother/father’s occupational prestige, household income, and household structure, as well as five indicators of home opportunity factors, separately. Model 13 added all SES+ indictors together. The following equation represents the fully adjusted Model 13:

Then we used 25 percent as the cut-off point to identify students who were high or low achievers. We considered as high achievers those students scoring above the 75th percentile of the academic achievement distribution in a specific grade. We considered as low achievers those students scoring below the 25th percentile of the academic achievement distribution. In this way, our outcomes became dummy variables, and we conducted both OLS and logistic regression models. The modeling strategy is the same as above.

Additional Documentation of Findings

Figure A1. Similar reduction rates of Black-White achievement gaps appear across cohorts.

Figure A1. Similar reduction rates of Black-White achievement gap across cohorts.

Figure A2. Similar reduction rates of Hispanic-White achievement gaps appear across cohorts.

Figure A2. Similar reduction rates of Hispanic-White achievement gap across cohorts.

[1] “Low birth-weight babies by race and ethnicity in United States.” Kids Count Data Center. Accessed: 12 August 2024. https://datacenter.aecf.org/data/tables/9817-low-birth-weight-babies-by…

[2] Barshay, Jill. “Proof Points: Tracing Black-White Achievement Gaps since the Brown Decision.” The Hechinger Report, May 13, 2024. https://hechingerreport.org/proof-points-black-white-achievement-gaps-since-brown/ ; “Racial and Ethnic Achievement Gaps.” The Educational Opportunity Monitoring Project: Racial and Ethnic Achievement Gaps. Accessed July 29, 2024. https://cepa.stanford.edu/educational-opportunity-monitoring-project/achievement-gaps/race/ .

[3] Roland G. Fryer, Jr. and Steven D. Levitt, “Understanding the Black-White Test Score Gap in the First Two Years of School,” The Review of Economics and Statistics 86, no. 2 (May 2004): 447-464, https://doi.org/10.1162/003465304323031049 ; Megan Kuhfeld, Elizabeth Gershoff, and Katherine Paschall, “The Development of Racial/Ethnic and Socioeconomic Achievement Gaps During the School Years,” Journal of Applied Developmental Psychology 57 (July 2018): 62-73, https://doi.org/10.1016/j.appdev.2018.07.001 ; Paul L. Morgan, George Farkas, Marianne M. Hillemeier, and Steve Maczuga, “Science Achievement Gaps Begin Very Early, Persist, and Are Largely Explained by Modifiable Factors,” Educational Researcher 45, no. 1 (January 2016): 18-35, https://doi.org/10.3102/0013189X16633182 ; David M. Quinn, “Kindergarten Black–White Test Score Gaps: Re-examining the Roles of Socioeconomic Status and School Quality with New Data,” Sociology of Education 88, no. 2 (April 2015): 120-139, https://doi.org/10.1177/0038040715573027 ; David M. Quinn and North Cooc, “Science Achievement Gaps by Gender and Race/Ethnicity in Elementary and Middle School: Trends and Predictors,” Educational Researcher 44, no. 6 (Aug/Sept 2015): 336-346, https://doi.org/10.3102/0013189X15598539 ; Sean F. Reardon, Joseph P. Robinson-Cimpian, and Ericka S. Weathers, “Patterns and Trends in Racial/Ethnic and Socioeconomic Academic Achievement Gaps,” in Handbook of Research in Education Finance and Policy , 2nd ed., eds. Helen F. Ladd and Margaret E. Goertz (New York: Routledge, 2015); Sean F. Reardon and Claudia Galindo, “The Hispanic-White Achievement Gap in Math and Reading in the Elementary Grades,” American Educational Research Journal 46, no. 3 (Sept 2009): 853-891, https://doi.org/10.3102/0002831209333184

[4] Although Hispanic students may be of any race, throughout this report we simplify the student groupings by referring to Hispanic students of any race as “Hispanic” and non-Hispanic students of other races by their racial category.

[5] Fryer and Levitt, 2004; Reardon et al., 2015; Reardon and Galindo, 2009; F. Chris Curran, “Income-Based Disparities in Early Elementary School Science Achievement,” The Elementary School Journal 118, no. 2 (Oct 2017): 207-231, https://doi.org/10.1086/694218 ; Daphne A. Henry, Laura Betancur Cortés, and Elizabeth Votruba-Drzal, “Black-White Achievement Gaps Differ by Family Socioeconomic Status from Early Childhood through Early Adolescence,” Journal of Educational Psychology 112, no. 8 (Nov 2020): 1471-1489, https://doi.org/10.1037/edu0000439 ; Jung-Sook Lee and Natasha K. Bowen, “Parent Involvement, Cultural Capital, and the Achievement Gap Among Elementary School Children,” American Educational Research Journal 43, no. 2 (Summer 2006): 193-218, https://doi.org/10.3102/00028312043002193

[6] Burchinal, Margaret, Kathleen McCartney, Laurence Steinberg, Robert Crosnoe, Sarah L. Friedman, Vonnie McLoyd, Robert Pianta, and NICHD Early Child Care Research Network. "Examining the Black–White achievement gap among low‐income children using the NICHD study of early child care and youth development." Child development 82, no. 5 (2011): 1404-1420. https://doi.org/10.1111/j.1467-8624.2011.01620.x ;

[7] See above note 2.

[8] See discussion of using free or reduced school lunch measure as SES here: Domina, Thurston, Nikolas Pharris-Ciurej, Andrew M. Penner, Emily K. Penner, Quentin Brummet, Sonya R. Porter, and Tanya Sanabria. "Is free and reduced-price lunch a valid measure of educational disadvantage?." Educational Researcher 47, no. 9 (2018): 539-555. https://doi.org/10.3102/0013189X18797609 .

[9] Reardon et al., 2015; Henry et al., 2020; Pamela E. Davis-Kean, “The Influence of Parent Education and Family Income on Child Achievement: The Indirect Role of Parental Expectations and the Home Environment,” Journal of Family Psychology 19, no. 2 (June 2005): 294-304, https://doi.org/10.1037/0893-3200.19.2.294 ; Amy J. Orr, “Black-White Differences in Achievement: The Importance of Wealth,” Sociology of Education 76, no. 4 (Oct 2003): 281-304, https://doi.org/10.2307/1519867

[10] Quinn, 2015; Curran, 2017; Nikki L. Aikens and Oscar A. Barbarin, “Socioeconomic Differences in Reading Trajectories: The Contribution of Family, Neighborhood, and School Contexts,” Journal of Educational Psychology 100, no. 2 (May 2008): 235-251, https://doi.org/10.1037/0022-0663.100.2.235 ; Selcuk R. Sirin, “Socioeconomic Status and Academic Achievement: A Meta-Analytic Review of Research,” Review of Educational Research 75, no. 3 (Fall 2005): 417-453, https://doi.org/10.3102/00346543075003417

[11] See the factsheet provided by American Psychological Association (APA) with additional resources: https://www.apa.org/pi/ses/resources/publications/education

[12] Davis-Kean, 2005; Annette Lareau, “Invisible Inequality: Social Class and Childrearing in Black Families and White Families,” American Sociological Review 67, no. 5 (Oct 2002): 747-776, https://doi.org/10.2307/3088916

[13] ECLS-K: 1998-99; https://nces.ed.gov/ecls/kindergarten.asp

[14] ECLS-K: 2010-11; https://nces.ed.gov/ecls/kindergarten2011.asp

[15] Trained field personnel individually assessed reading, mathematics, and science achievement in each grade using untimed and item response theory (IRT) scaled measures. The assessment process consisted of two stages. The first stage included items of varying difficulty levels that determined the student's initial performance level. This was followed by one of three second-stage assessments that included additional low-, middle-, or high-difficulty items. We used scores from these measures of academic achievement as continuous variables.

[16] Centers for Disease Control and Prevention (CDC). Census 2010 Occupation and Industry Coding Instructions . National Institute for Occupational Safety and Health (NIOSH). August 11, 2011. https://www.cdc.gov/niosh/ topics/coding/pdfs/ Census2010CodingInstruction. pdf . https://www.cdc.gov/niosh/topics/coding/pdfs/Census2010CodingInstruction.pdf . There are 22 occupation codes in this coding scheme. If an occupation could not be coded using this manual, the Standard Occupational Classification Manual—1980 was used to identify the appropriate code. Then based on these occupation categories, they were recoded to reflect the average of the 1989 General Social Survey (GSS) prestige scores. Although the GSS prestige scores are from 1989, they are still being used by the current GSS survey and matched to 1980 census codes.

[17] For ECLS-K:2011, the household income included 18 categories: 1) $5,000 or less, 2) $5,001 to $10,000, 3) $10,001 to $15,000, 4) $15,001 to $20,000, 5) $20,001 to $25,000, 6) $25,001 to $30,000 … 16) $75,001 to $100,000, 17) $100,001 to $200,000, and 18) $200,001 or more. For ECLS-K, the household income included 13 categories: 1) $5,000 or less, 2) $5,001 to $10,000, 3) $10,001 to $15,000, 4) $15,001 to $20,000, 5) $20,001 to $25,000, 6) $25,001 to $30,000, 7) $30,001 to $35,000, 8) $35,001 to $40,000, 9) $40,001 to $50,000, 10) $50,001 to $75,000, 11) $75,001 to $100,000, 12) $100,001 to $200,000, and 13) $200,001 or more. We treated them as a continuous variable in our analyses.

[18] Very little research on racial and/or socioeconomic gaps considers the role of family structure, so inclusion of this variable fills a hole in the literature that tends to be focused on race, class, and gender differences. For more, see https://www.city-journal.org/article/measure-what-matters

[19] Ordinary least squares (OLS) regression models include adjusted robust standard errors with clustering at the school (kindergarten) level. Multiple imputation is used to address missing data.

[20] Jonathan A. Plucker and Scott J. Peters, Excellence Gaps in Education: Expanding Opportunities for Talented Students (Cambridge, MA: Harvard Education Press, 2016).

[21] U.S. Department of Education provides different kinds of resources under Division of Adult Education & Literacy: https://aefla.ed.gov/ .

[22] Schoch, Annie D., Cassie S. Gerson, Tamara Halle, and Meg Bredeson. "Children's Learning and Development Benefits from High-Quality Early Care and Education: A Summary of the Evidence. Research Highlight. OPRE Report 2023-226." Office of Planning, Research and Evaluation (2023). https://www.acf.hhs.gov/opre/report/childrens-learning-and-development-benefits-high-quality-early-care-and-education

[23] Gennetian, Lisa A., and Katherine Magnuson. "Three Reasons Why Providing Cash to Families With Children Is a Sound Policy Investment." Journal of Human Resources 55, no. 2 (2018): 387-427. Sherman, Arloc, and Tazra Mitchell. "Economic security programs help low-income children succeed over long term, many studies find." Washington: Center on Budget and Policy Priorities .” https://www. cbpp. org/sites/default/files/atoms/files/7-17-17pov. pdf (2017).

[24] Paul L. Morgan and Eric Hengyu Hu, “Fixed Effect Estimates of Student-Teacher Racial or Ethnic Matching in U.S. Elementary Schools,” Early Childhood Research Quarterly 63 (Nov 2022): 98-112, https://doi.org/10.1016/j.ecresq.2022.11.003

Acknowledgments

This report was made possible through the generous support of the Achelis and Bodman Foundation and the Thomas B. Fordham Foundation. We are especially grateful to Paul L. Morgan and Eric Hengyu Hu for conducting the analysis and authoring the report. In addition, we extend our appreciation to Andrew Conway, professor at New Mexico State University, and Jing Liu, assistant professor at the University of Maryland, College Park, for their timely and helpful feedback on a draft of the report. We also extend our gratitude to Super Copy Editors for copyediting and Dave Williams for designing the figures and tables. At Fordham, we would like to thank Chester E. Finn, Jr., Michael J. Petrilli, Amber M. Northern and Adam Tyner for reviewing drafts; Heena Kuwayama for assisting with editing; Elainah Elkins for handling funder communications; Victoria McDougald for her role in dissemination; and Stephanie Distler for developing the report’s cover art and coordinating production.

Eric Hengyu Hu, Ph.D., is a Postdoctoral Associate at the Institute of Social and Health Equity, University at Albany, SUNY. Dr. Hu’s research focuses on sociodemographic disparities in the academic, cognitive, and social-behavioral development of early elementary students. He can be reached through email at [email protected] and Twitter/X @Eric_hhy.

Paul L. Morgan, Ph.D., is the Empire Innovation Professor, Social and Health Equity Endowed Professor in the Department of Health Policy, Management and Behavior, School of Public Health, and Inaugural Director of the Institute for Social and Health Equity, University at Albany, SUNY. He can be reached through email at [email protected] and Twitter/X @PaulMorganPhD.

Related Resources

Excellence Gaps by Race and Socioeconomic Status

The Black-White Achievement Gap: Why Closing It Is the Greatest Civil Rights Issue of Our Time

Finally, some straight talk on the achievement gap

Better Schools Won’t Fix America

Like many rich Americans, I used to think educational investment could heal the country’s ills—but I was wrong. Fighting inequality must come first.

Man in graduation cap standing on trend line

Long ago, I was captivated by a seductively intuitive idea, one many of my wealthy friends still subscribe to: that both poverty and rising inequality are largely consequences of America’s failing education system. Fix that, I believed, and we could cure much of what ails America.

This belief system, which I have come to think of as “educationism,” is grounded in a familiar story about cause and effect: Once upon a time, America created a public-education system that was the envy of the modern world. No nation produced more or better-educated high-school and college graduates, and thus the great American middle class was built. But then, sometime around the 1970s, America lost its way. We allowed our schools to crumble, and our test scores and graduation rates to fall. School systems that once churned out well-paid factory workers failed to keep pace with the rising educational demands of the new knowledge economy. As America’s public-school systems foundered, so did the earning power of the American middle class. And as inequality increased, so did political polarization, cynicism, and anger, threatening to undermine American democracy itself.

Taken with this story line, I embraced education as both a philanthropic cause and a civic mission. I co-founded the League of Education Voters, a nonprofit dedicated to improving public education. I joined Bill Gates, Alice Walton, and Paul Allen in giving more than $1 million each to an effort to pass a ballot measure that established Washington State’s first charter schools. All told, I have devoted countless hours and millions of dollars to the simple idea that if we improved our schools—if we modernized our curricula and our teaching methods, substantially increased school funding, rooted out bad teachers, and opened enough charter schools—American children, especially those in low-income and working-class communities, would start learning again. Graduation rates and wages would increase, poverty and inequality would decrease, and public commitment to democracy would be restored.

But after decades of organizing and giving, I have come to the uncomfortable conclusion that I was wrong. And I hate being wrong.

Read: Education reform and the failure to fix inequality in America

What I’ve realized, decades late, is that educationism is tragically misguided. American workers are struggling in large part because they are underpaid—and they are underpaid because 40 years of trickle-down policies have rigged the economy in favor of wealthy people like me. Americans are more highly educated than ever before, but despite that, and despite nearly record-low unemployment, most American workers—at all levels of educational attainment—have seen little if any wage growth since 2000.

To be clear: We should do everything we can to improve our public schools. But our education system can’t compensate for the ways our economic system is failing Americans. Even the most thoughtful and well-intentioned school-reform program can’t improve educational outcomes if it ignores the single greatest driver of student achievement: household income.

For all the genuine flaws of the American education system, the nation still has many high-achieving public-school districts. Nearly all of them are united by a thriving community of economically secure middle-class families with sufficient political power to demand great schools, the time and resources to participate in those schools, and the tax money to amply fund them. In short, great public schools are the product of a thriving middle class, not the other way around. Pay people enough to afford dignified middle-class lives, and high-quality public schools will follow. But allow economic inequality to grow, and educational inequality will inevitably grow with it.

By distracting us from these truths, educationism is part of the problem.

W henever I talk with my wealthy friends about the dangers of rising economic inequality, those who don’t stare down at their shoes invariably push back with something about the woeful state of our public schools. This belief is so entrenched among the philanthropic elite that of America’s 50 largest family foundations—a clique that manages $144 billion in tax-exempt charitable assets—40 declare education as a key issue. Only one mentions anything about the plight of working people, economic inequality, or wages. And because the richest Americans are so politically powerful, the consequences of their beliefs go far beyond philanthropy.

A major theme in the educationist narrative involves the “ skills gap ”—the notion that decades of wage stagnation are largely a consequence of workers not having the education and skills to fill new high-wage jobs. If we improve our public schools, the thinking goes, and we increase the percentage of students attaining higher levels of education, particularly in the STEM subjects — science, technology, engineering, and math—the skills gap will shrink, wages will rise, and income inequality will fall.

Annie Lowrey: Wages are low and workers are scarce. Wait, what?

The real story is more complicated, and more troubling. Yes, there is a mismatch between the skills of the present and the jobs of the future. In a fast-changing, technologically advanced economy, how could there not be? But this mismatch doesn’t begin to explain the widening inequality of the past 40 years.

In 1970, when the golden age of the American middle class was nearing its peak and inequality was at its nadir, only about half of Americans ages 25 and older had a high-school diploma or the equivalent. Today, 90 percent do. Meanwhile, the proportion of Americans attaining a college degree has more than tripled since 1970. But while the American people have never been more highly educated, only the wealthiest have seen large gains in real wages. From 1979 to 2017, as the average real annual wages of the top 1 percent of Americans rose 156 percent (and the top .01 percent’s wages rose by a stunning 343 percent), the purchasing power of the average American’s paycheck did not increase.

Some educationists might argue that the recent gains in educational attainment simply haven’t been enough to keep up with the changing economy—but here, yet again, the truth appears more complicated. While 34 percent of Americans ages 25 and older have a bachelor’s degree or higher, only 26 percent of jobs currently require one. The job categories that are growing fastest, moreover, don’t generally require a college diploma, let alone a STEM degree. According to federal estimates, four of the five occupational categories projected to add the most jobs to the economy over the next five years are among the lowest-paying jobs: “food preparation and serving” ($19,130 in average annual earnings), “personal care and service” ($21,260), “sales and related” ($25,360), and “health-care support” ($26,440). And while the number of jobs that require a postsecondary education is expected to increase slightly faster than the number that don’t, the latter group is expected to dominate the job market for decades to come. In October 2018 there were 1 million more job openings than job seekers in the U.S. Even if all of these unfilled jobs were in STEM professions at the top of the pay scale, they would be little help to most of the 141 million American workers in the bottom nine income deciles.

It’s worth noting that workers with a college degree enjoy a significant wage premium over those without. (Among people over age 25, those with a bachelor’s degree had median annual earnings of $53,882 in 2017, compared with $32,320 for those with only a high-school education.) But even with that advantage, adjusted for inflation, average hourly wages for recent college graduates have barely budged since 2000 , while the bottom 60 percent of college graduates earn less than that group did in 2000. A college diploma is no longer a guaranteed passport into the middle class.

Meanwhile, nearly all the benefits of economic growth have been captured by large corporations and their shareholders. After-tax corporate profits have doubled from about 5 percent of GDP in 1970 to about 10 percent, even as wages as a share of GDP have fallen by roughly 8 percent. And the wealthiest 1 percent’s share of pre-tax income has more than doubled, from 9 percent in 1973 to 21 percent today. Taken together, these two trends amount to a shift of more than $2 trillion a year from the middle class to corporations and the super-rich.

The state of the labor market provides further evidence that low-wage workers’ declining fortunes aren’t explained by supply and demand. With the unemployment rate near a 50-year floor, low-wage industries such as accommodations, food service, and retail are struggling to cope with a shortage of job applicants—leading The Wall Street Journal to lament that “low-skilled jobs are becoming increasingly difficult for employers to fill.” If wages were actually set the way our Econ 101 textbooks suggested, workers would be profiting from this dynamic. Yet outside the cities and states that have recently imposed a substantially higher local minimum wage, low-wage workers have seen their real incomes barely budge.

All of which suggests that income inequality has exploded not because of our country’s educational failings but despite its educational progress. Make no mistake: Education is an unalloyed good. We should advocate for more of it, so long as it’s of high quality. But the longer we pretend that education is the answer to economic inequality, the harder it will be to escape our new Gilded Age.

Read: The 9.9 percent is the new American aristocracy

H owever justifiable their focus on curricula and innovation and institutional reform, people who see education as a cure-all have largely ignored the metric most predictive of a child’s educational success: household income.

The scientific literature on this subject is robust, and the consensus overwhelming. The lower your parents’ income, the lower your likely level of educational attainment. Period. But instead of focusing on ways to increase household income, educationists in both political parties talk about extending ladders of opportunity to poor children, most recently in the form of charter schools. For many children, though—especially those raised in the racially segregated poverty endemic to much of the United States—the opportunity to attend a good public school isn’t nearly enough to overcome the effects of limited family income.

As Lawrence Mishel, an economist at the liberal-leaning Economic Policy Institute, notes, poverty creates obstacles that would trip up even the most naturally gifted student. He points to the plight of “children who frequently change schools due to poor housing; have little help with homework; have few role models of success; have more exposure to lead and asbestos; have untreated vision, ear, dental, or other health problems; … and live in a chaotic and frequently unsafe environment.”

Indeed, multiple studies have found that only about 20 percent of student outcomes can be attributed to schooling, whereas about 60 percent are explained by family circumstances—most significantly, income. Now consider that, nationwide, just over half of today’s public-school students qualify for free or reduced-price school lunches , up from 38 percent in 2000. Surely if American students are lagging in the literacy, numeracy, and problem-solving skills our modern economy demands, household income deserves most of the blame—not teachers or their unions.

If we really want to give every American child an honest and equal opportunity to succeed, we must do much more than extend a ladder of opportunity—we must also narrow the distance between the ladder’s rungs. We must invest not only in our children, but in their families and their communities. We must provide high-quality public education, sure, but also high-quality housing, health care, child care, and all the other prerequisites of a secure middle-class life. And most important, if we want to build the sort of prosperous middle-class communities in which great public schools have always thrived, we must pay all our workers, not just software engineers and financiers, a dignified middle-class wage.

Today, after wealthy elites gobble up our outsize share of national income, the median American family is left with $76,000 a year. Had hourly compensation grown with productivity since 1973—as it did over the preceding quarter century, according to the Economic Policy Institute—that family would now be earning more than $105,000 a year. Just imagine, education reforms aside, how much larger and stronger and better educated our middle class would be if the median American family enjoyed a $29,000-a-year raise.

In fact, the most direct way to address rising economic inequality is to simply pay ordinary workers more, by increasing the minimum wage and the salary threshold for overtime exemption; by restoring bargaining power for labor; and by instating higher taxes—much higher taxes—on rich people like me and on our estates.

Educationism appeals to the wealthy and powerful because it tells us what we want to hear: that we can help restore shared prosperity without sharing our wealth or power. As Anand Giridharadas explains in his book Winners Take All: The Elite Charade of Changing the World , narratives like this one let the wealthy feel good about ourselves. By distracting from the true causes of economic inequality, they also defend America’s grossly unequal status quo.

We have confused a symptom—educational inequality—with the underlying disease: economic inequality. Schooling may boost the prospects of individual workers, but it doesn’t change the core problem, which is that the bottom 90 percent is divvying up a shrinking share of the national wealth. Fixing that problem will require wealthy people to not merely give more, but take less.

This article appears in the July 2019 print edition with the headline “Education Isn’t Enough.”

About the Author

Causes and Consequences of Income Inequality – An Overview

Rising income inequality is one of the greatest challenges facing advanced economies today. Income inequality is multifaceted and is not the inevitable outcome of irresistible structural forces such as globalisation or technological development. Instead, this review shows that inequality has largely been driven by a multitude of political choices. The embrace of neoliberalism since the 1980s has provided the key catalyst for political and policy changes in the realms of union regulation, executive pay, the welfare state and tax progressivity, which have been the key drivers of inequality. These preventable causes have led to demonstrable harmful outcomes that are not explicable solely by material deprivation. This review also shows that inequality has been linked on the economic front with reduced growth, investment and innovation, and on the social front with reduced health and social mobility, and greater violent crime.

1 Introduction

Income inequality has recently come to be viewed as one of the greatest challenges facing the world today. In recent years, the topic has dominated the agenda of the World Economic Forum (WEF), where the world’s top political and business leaders attend. Their global risks report, drawn from over 700 experts in attendance, pronounced inequality to be the greatest threat to the world economy in 2017 ( Elliott 2017 ). Likewise, the past decade has seen leading global figures such as former American President Barack Obama, Pope Francis, Chinese President Xi Jinping, and the former head of the International Monetary Fund (IMF), Christine Lagarde, all undertake speeches on the gravity of income inequality and the need to address its rise. This is because, as this research note shows, income inequality engenders harmful consequences that are not explicable solely by material deprivation.

The general dynamics of income inequality include a tendency to rise slowly and fluctuate over time. For instance, Japan had one of the highest rates in the world prior to the Second World War and the United States (US) one of the lowest, which has since completely reversed for both. The United Kingdom (UK) was also the second most equitable large European country in the 1970s but is now the most inequitable ( Dorling 2018 : 27–28).

High rates of inequality are rarely sustained for long periods because they tend to lead to or become punctuated by man-made disasters that lead to a levelling out. Scheidel (2017) posits that there in fact exists a violent ‘Four Horseman of Leveling’ (mass mobilisation warfare, transformation revolutions, state collapse, and lethal pandemics) for inequality, which have at times dramatically reduced inequalities because they can lead to the alteration of existing power structures or wipe out the wealth of elites and redistribute their resources. For instance, the pronounced shocks of the two world wars led to the ‘Great Compression’ of income throughout the West in the post-war years. There is already some evidence that the current global pandemic caused by the novel Coronavirus, has led to greater aversion to income inequality ( Asaria, Costa-Font, and Cowell 2021 ; Wiwad et al. 2021 ).

Thus, greater aversion to inequality has been able to reduce inequality in the past, this is because, as this review also shows, income inequality does not result exclusively from efficient market forces but arises out of a set of rules that is shaped by those with political power. Inequality’s rise is not inevitable, nor beyond the control of governments and policymakers, as they can affect distributional outcomes and inequality through public policy.

It is the purpose of this review to outline the causes and consequences of income inequality. The paper begins with an analysis of the key structural and institutional determinants of inequality, followed by an examination into the harmful outcomes of inequality. It then concludes with a discussion of what policymakers can do to arrest the rise of inequality.

2 Causes of Income Inequality

Broadly speaking, explanations for the increase in income inequality have largely been classified as either structural or institutional. Historically, economists emphasised structural causes of increasing income inequality, with globalisation and technological change at the forefront. However, in recent years opinion has shifted to emphasise more institutional political factors to do with the adoption of neoliberal reforms such as privatisation, deregulation and tax and welfare reductions since the early 1980s. They were first embraced and most heavily championed by the UK and US, spreading globally later, and which provide the crucial catalysts of rising income inequality ( Atkinson 2015 ; Brown 2017 ; Piketty 2020 ; Stiglitz 2013 ). I discuss each of these key factors in turn.

2.1 Globalisation

One of the earliest, and most prominent explanations for the rise of income inequality emphasised the role of globalisation ( Borjas, Freeman, and Katz 1992 ; Revenga 1992 ). Globalisation has led to the offshoring of many goods and services that used to be produced or completed domestically in the West, which has created downward pressures on the wages of lower skilled workers. According to the ‘market forces hypothesis,’ increasing inequality is a response to the rising demand for skills at the top, in which the spread of globalisation and technological progress have been facilitated through reduced barriers to trade and movement.

Proponents of globalisation as the leading cause of inequality have argued that globalisation has constrained domestic state choices and left governments collectively powerless to address inequality. Detractors admit that globalisation has indeed had deep structural effects on Western economies but its impact on the degree of agency available to domestic governments has been mediated by individual policy choices ( Thomas 2016 : 346). A key problem with attributing the cause of inequality to globalisation, is that the extent of the inequality increase has varied considerably across countries, even though they have all been exposed to the same effects of globalisation. The US also has the highest inequality amongst rich countries, but it is less reliant on international trade than most other developed countries ( Brown 2017 : 56). Moreover, a recent meta-analysis by Heimberger (2020) found that globalisation has a “small-to-moderate” inequality-increasing effect, with financial globalisation displaying the largest impact.

2.2 Technology

A related explanation for inequality draws attention to the impact of technology specifically. The advent of the digital age has placed a higher premium on the skills needed for non-routine work and reduced the value placed on lower skilled routine work, as it has enabled machines to replace jobs that could be routinised. This skill-biased technological change (SBTC) has led to major changes in the organisation of work, as many full-time permanent jobs with benefits have given way to part-time flexible work without benefits, that are often centred around the completion of short ‘gigs’ such as a car journey or food delivery. For instance, the Organisation for Economic Co-operation and Development (OECD) estimated in 2015 that since the 1990s, roughly 60% of all job creation has been in the form of non-standard work due to technological changes and that those employed in such jobs are more likely to be poor ( Brown 2017 : 60).

Relatedly, a prevailing doctrine in economics is ‘marginal productivity theory,’ which holds that people with greater productivity levels will earn higher incomes. This is due to the belief that a person’s productivity is equated to their societal contribution ( Stiglitz 2013 : 37). Since technology is a leading determinant in the productivity of different skills and SBTC has led to increased productivity, it has also become a justification for inequality. However, it is very difficult to separate any one person’s contribution to society from that of others, as even the most successful businessperson owes their success to the rule of law, good infrastructure, and a state educated workforce ( Stiglitz 2013 : 97–98).

Further criticisms of the SBTC explanation, are that there was still substantial SBTC when inequality first fell dramatically and then stabilised in the period from 1930 to 1980, and it has failed to explain the perpetuation of both the gender and racial wage gap, “or the dramatic rise in education-related wage gaps for younger versus older workers” ( Brown 2017 : 67). Although it is difficult to decouple globalisation and technology, as they each have compounding tendencies, it is most likely that globalisation and technology are important explanatory factors for inequality, but predominantly facilitate and underlie the following more determinant institutional factors that happen to be already present, such as reduced tax progressivity, rising executive pay, and union decline. It is to these factors that I now turn.

2.3 Tax Policy

Taxes overwhelmingly comprise the primary source of revenue that governments can use for redistribution, which is fundamental to alleviating income inequality. Redistribution is defended on economic grounds because the marginal utility of money declines as income rises, meaning that the benefit derived from extra income is much higher for the poor than the rich. However, since the late 1970s, a major rethinking surrounding redistributive policy occurred. This precipitated ‘trickle-down economics’ theory achieving prominence amongst American and British policymakers, whereby the benefits from tax cuts on the wealthy would trickle-down to everyone. Subsequently, expert opinion has determined that tax cuts do not actually spur economic growth ( CBPP 2017 ).

Personal income tax progressivity has declined sharply in the West, as the average top income tax rate for OECD members fell from 62% in 1981 to 35% in 2015 ( IMF 2017 : 11). However, the decline has been most pronounced in the UK and the US, which had top rates of around 90% in the 1960s and 1970s. Corporate tax rates have also plummeted by roughly one half across the OECD since 1980 ( Shaxson 2015 : 4). Recent International Monetary Fund (IMF) research found that between 1985 and 1995, redistribution through the tax system had offset 60% of the increase in market inequality but has since failed to respond to the continuing increase in inequality ( IMF 2017 ). Moreover, in a sample of 18 OECD countries encompassing 50 years, Hope and Limberg (2020) found that tax reforms even significantly increased pre-tax income inequality, while having no significant effect on economic growth.

This decline in tax progressivity has been a leading cause of rising income inequality, which has been compounded by the growing problem of tax avoidance. A complex global web of shell corporations has been constructed by international brokers in offshore tax havens that is able to keep wealth hidden from tax collectors. The total hidden amount in tax havens is estimated to be $7.6 trillion US dollars and rising, or roughly 8% of total global household wealth ( Zucman 2015 : 36). Recent research has revealed that tax havens are overwhelmingly used by the immensely rich ( Alstadsæter, Johannesen, and Zucman 2019 ), thus taxing this wealth would substantially reduce income inequality and increase revenue available for redistribution. The massive reduction in income tax progressivity in the Anglo world, after it had been amongst its leaders in the post-war years, also “probably explains much of the increase in the very highest earned incomes” since 1980 ( Piketty 2014 : 495–496).

2.4 Executive Pay

The enormous rising pay of executives since the 1980s, has also fuelled income inequality and more specifically the gap between executives and their employees. For example, the gap between Chief Executive Officers (CEO) and their workers at the 500 leading US companies in 2016, was 335 times, which is nearly 10 times larger than in 1980. It is a similar story in the UK, with a pay ratio of 131 for large British firms, which has also risen markedly since 1980 ( Dorling 2017 ).

Piketty (2014 : 335) posits that the dramatic reduction in top income tax has had an amplifying effect on top executives pay since it provides them with much greater incentive to seek larger remuneration, as far less is then taken in tax. It is difficult to objectively measure an individual’s contribution to a company and with the onset of trickle-down economics and accompanying business-friendly climate since the 1980s, top executives have found it relatively easy to convince boards of their monetary worth ( Gabaix and Landier 2008 ).

The rise in executive pay in both the UK and US, is far larger than the rest of the OECD. This may partially be explained by the English-speaking ‘superstar’ theory, whereby the global market demand for top CEOs is much higher for native English speakers due to English being the prime language of the global economy ( Deaton 2013 : 210). Saez and Veall (2005) provide support for the theory in a study of the top 1% of earners from the Canadian province of Quebec, which showed that English speakers were able to increase their income share over twice as much as their French-speaking counterparts from 1980 to 2000. This upsurge of income at the top of the labour market has been accompanied by stagnation or diminishing returns for the middle and lower parts of the labour market, which has been affected by the dramatic decline of union influence throughout the West.

2.5 Union Decline

Trade unions have typically been viewed as an important force for moderating income inequality. They “contribute to wage compression by restricting wage decline among low-wage earners” and restrain wage surges among high-wage earners ( Checchi and Visser 2009 : 249). The mere presence of unions can also drive up the wages of non-union employees in similar industries, as employers tend to give in to wage demands to keep unions out. Union density has also been proven to be strongly associated with higher redistribution both directly and indirectly, through its influence on left party governments ( Haddow 2013 : 403).

There had broadly existed a ‘social contract’ between labour and business, whereby collective bargaining establishes a wage structure in many industries. However, this contract was abandoned by corporate America in the mid-1970s when large-scale corporate donations influenced policymakers to oppose pro-union reform of labour law, leading to political defeats for unions ( Hacker and Pierson 2010 : 58–59). The crackdown of strikes culminating in the momentous Air Traffic Controllers’ strike (1981) in the US and coal miner’s strike (1984–85) in the UK, caused labour to become de-politicised, which was self-reinforcing, because as their political power dispersed, policymakers had fewer incentives to protect or strengthen union regulations ( Rosenfeld and Western 2011 ). Consequently, US union density has plummeted from around a third of the workforce in 1960, down to 11.9% last decade, with the steepest decline occurring in the 1980s ( Stiglitz 2013 : 81).

Although the decline in union density is not as steep cross-nationally, the pattern is still similar. Baccaro and Howell (2011 : 529) found that on average the unionisation rate decreased by 0.39% a year since 1974 for the 15 OECD members they surveyed. Increasingly, the decline in the fortunes of labour is being linked with the increase in inequality and the sharpest increases in income inequality have occurred in the two countries with the largest falls in union density – the UK and US. Recent studies have found that the weakening of organised unions accounts for between a third and a fifth of the total rise in income inequality in the US ( Rosenfeld and Western 2011 ), and nearly one half of the increase in both the Gini rate and the top 10%’s income share amongst OECD members ( Jaumotte and Buitron 2015 ).

To illustrate the changing relationship between inequality and unionisation, Figure 1 displays a local polynomial smoother scatter plot of union density by income inequality, for 23 OECD countries, 1980–2018. They are negatively correlated, as countries with higher union density have much lower levels of income inequality. Figure 2 further plots the time trends of both. Income inequality (as measured via the Gini coefficient) has climbed over 0.02 percentage points on average in these countries since 1980, which is roughly a one-tenth rise. Whereas union density has fallen on average from 44 to 35 percentage points, which is over one-fifth.

Figure 1:
Gini coefficient by union density, OECD 1980–2018. Data on Gini coefficients from SWIID (Solt 2020); data on union density from ICTWSS Database (Visser 2019).

Gini coefficient by union density, OECD 1980–2018. Data on Gini coefficients from SWIID ( Solt 2020 ); data on union density from ICTWSS Database ( Visser 2019 ).

Figure 2:
Gini coefficient by union density, 1980–2018. Data on Gini coefficients from SWIID (Solt 2020); data on union density from ICTWSS Database (Visser 2019).

Gini coefficient by union density, 1980–2018. Data on Gini coefficients from SWIID ( Solt 2020 ); data on union density from ICTWSS Database ( Visser 2019 ).

In sum, income inequality is multifaceted and is not the inevitable outcome of irresistible structural forces such as globalisation or technological development. Instead, it has largely been driven by a multitude of political choices. Tridico (2018) finds that the increases in inequality from 1990 to 2013 in 26 OECD countries, was largely owing to increased financialisation, deepening labour flexibility, the weakening of trade unions and welfare state retrenchment. While Huber, Huo, and Stephens (2019) recently reveals that top income shares are unrelated to economic growth and knowledge-intensive production but is closely related to political and policy changes surrounding union density, government partisanship, top income tax rates, and educational investment. Lastly, Hager’s (2020) recent meta-analysis concludes that the “empirical record consistently shows that government policy plays a pivotal role” in shaping income inequality.

These preventable causes that have given rise to inequality have created socio-economic challenges, due to the demonstrably negative outcomes that inequality engenders. What follows is a detailed analysis of the significant mechanisms that income inequality induces, which lead to harmful outcomes.

3 Consequences of Income Inequality

Escalating income inequality has been linked with numerous negative outcomes. On the economic front, negative results transpire beyond the obvious poverty and material deprivation that is often associated with low incomes. Income inequality has also been shown to reduce growth, innovation, and investment. On the social front, Wilkinson and Pickett’s ground-breaking The Spirit Level ( 2009 ), found that societies that are more unequal have worse social outcomes on average than more egalitarian societies. They summarised an extensive body of research from the previous 30 years to create an Index of Health and Social Problems, which revealed a host of different health and social problems (measuring life expectancy, infant mortality, obesity, trust, imprisonment, homicide, drug abuse, mental health, social mobility, childhood education, and teenage pregnancy) as being positively correlated with the level of income inequality across rich nations and across states within the US. Figure 3 displays the cross-national findings via a sample of 21 OECD countries.

Figure 3:
Index of health and social problems by Gini coefficient. Data on health and social problems index from The Equality Trust (2018); data on Gini coefficients from OECD (2020).

Index of health and social problems by Gini coefficient. Data on health and social problems index from The Equality Trust (2018) ; data on Gini coefficients from OECD (2020) .

3.1 Economic

Income inequality is predominantly an economic subject. Therefore, it is understandable that it can engender pervasive economic outcomes. Foremost economically speaking, it has been linked with reduced growth, investment and innovation. Leading international organisations such as the IMF, World Bank and OECD, pushed for neoliberal reforms beginning in the 1980s, although they have recently started to substantially temper their views due to their own research into inequality. A 2016 study by IMF economists, noted that neoliberal policies have delivered benefits through the expansion of global trade and transfers of technology, but the resulting increases in inequality “itself undercut growth, the very thing that the neo-liberal agenda is intent on boosting” ( Ostry, Loungani, and Furceri 2016 : 41). Cingano’s (2014) OECD cross-national study, found that once a country’s income inequality reaches a certain level it reduces growth. The growth rate in these countries would have been one-fifth higher had income inequality not increased, while the greater equality of the other countries included in the study helped to increase their growth rates.

Consumer spending is good for economic growth but rising income inequality shifts more money to the top of the income distribution, where higher income individuals have a much smaller propensity to consume than lower-income individuals. The wealthy save roughly 15–25% of their income, whereas low income individuals spend their entire income on consumer goods and services ( Stiglitz 2013 : 106). Therefore, greater inequality reduces demand in an economy and is a major contributor to the ‘secular stagnation’ (persistent insufficient demand relative to aggregate private savings) that the largest Western economies have been experiencing since the financial crisis. Inequality also increases the level of debt, as lower-income individuals borrow more to maintain their standard of living, especially in a climate of low interest rates. Combined with deregulation, greater debt increases instability and “was a major contributor to, if not the underlying cause of, the 2008 financial crash” ( Brown 2017 : 35–36).

Another key economic effect of income inequality is that it leads to reduced welfare spending and public investment. Since a greater share of the income distribution is earned by the very wealthy, governments have less income available to fund education, public amenities, and other services that the poor rely heavily on. This creates social separation, whereby the wealthy opt out in publicly funding services because their private equivalents are of better quality. This causes a cycle of increasing income inequality that is likely to eventually lead to a situation of “private affluence and public squalor” ( Marmot 2015 : 39).

Lastly, it has been proven that economic instability is a by-product of increasing inequality, which harms innovation. Both countries and American states with the highest inequality have been found to be the least innovative in terms of the amount of Intellectual Property (IP) patents they produce ( Dorling 2018 : 129–130). Although income inequality is predominantly an economic subject, its effects are so pervasive that it has also been linked to a host of negative health and societal outcomes.

Wilkinson and Pickett found key associations between income inequality for both physical and mental health. For example, they discovered that on average the life expectancy gap is more than four years between the least and most equitable richest nations (Japan and the US). Since their revelations, overall life expectancy has been reported to be declining in the US ( Case and Deaton 2020 ). It has held or declined every year since 2014, which has led to a cumulative drop of 1.13 years ( Andrasfay and Goldman 2021 ). Marmot (2015) has provided evidence that there exists a social gradient whereby differences in affluence translate into increasing health inequalities, which can be shown even down to the neighbourhood level, as more affluent areas have higher life expectancy on average than deprived areas, and a clear gradient appears where life expectancy increases in line with affluence.

Moreover, Marmot’s famous Whitehall studies, which are large-scale longitudinal studies of Whitehall employees of UK central government, found an inverse-relationship between salary grade and ill-health, whereby low-grade workers were four times as likely as high-grade workers to suffer from ill-health ( 2015 : 11). Health steadily improves with rank and the correlation is little affected by lifestyle controls such as tobacco and alcohol usage. However, the leading factor that seems to make the most difference in ill-health is job stress and a person’s sense of control over their work, including the variety of work and the use and development of skills ( Schrecker and Bambra 2015 : 54–55).

‘Psychosocial stresses,’ like those appearing in the Whitehall studies, have been found to be more common and frequent amongst low-income individuals, beyond just the workplace ( Jensen and van Kersbergen 2017 : 24). Wilkinson and Pickett (2019) posit that greater income inequality engenders low self-esteem, chronic stress and depression, stemming from status anxiety. This occurs because more importance is placed on where people fit in a hierarchy with greater inequality. For evidence, they outline a clear relationship of a much higher percentage of the population suffering from mental illness in more unequal countries. Meticulous research has shown that huge inequalities in income result in the poor having feelings of shame across a range of environments. Furthermore, Dickerson and Kemeny’s (2004) meta-analysis of 208 studies found that stress-hormone (cortisol) levels were raised particularly “when people felt that others were making negative judgements about them” ( Rowlingson 2011 : 24).

These effects on both mental and physical health can be best illustrated via the ‘absolute income’ and ‘relative income’ hypotheses ( Daly, Boyce, and Wood 2015 ). The relative income hypothesis posits that when an individual’s income is held constant, the relative income of others can affect a person’s health depending on how they view themselves in comparison to those above them ( Wilkinson 1996 ). This pattern also holds when income inequality increases at the societal level, because if such changes lead to increases in chronic stress, it can increase ill-health nationally. Whereas the absolute income hypothesis predicts that health gains from an extra unit of income diminish as an individual’s income rises ( Kawachi, Adler, and Dow 2010 ). A mean preserving transfer from a richer to poorer individual raises the health of the poorer individual more than it lowers the health of the richer person. This occurs because there is an optimum threshold of income required to maintain good health. Thus, when holding total income constant, a more equal distribution of income should improve overall population health. This pattern also applies at the country-wide level, as the “effect of income on health appears substantial as countries move from about $15,000 to 25,000 US dollars per capita,” but appears non-existent beyond that point ( Leigh, Jencks, and Smeeding 2009 : 386–387).

Income inequality also impacts happiness and wellbeing, as the happiest nations are routinely the ones with low inequality, such as Denmark and Norway. Happiness has been proven to be affected by the law of diminishing returns in economics. It states that higher income incrementally improves happiness but only up to a certain point, as any individual income earned beyond roughly $70,000 US dollars, does not bring about greater happiness ( Deaton 2013 : 53). The negative physical and mental health outcomes that income inequality provoke, also impact key societal areas such as crime, social mobility and education.

Rates of violent crime are lower in more equal countries ( Hsieh and Pugh 1993 ; Whitworth 2012 ). This is largely because more equal countries have less poverty, which leads to less people being desperate about their situation, as lower-income individuals have been shown to commit more crime. Relatedly, according to strain theory, more unequal societies place higher social value in achieving economic success, while providing lower means to achieve it ( Merton 1938 ). This generates strain, which may lead more individuals to pursue crime as a means of attaining financial success. At the opposite end of the income spectrum, the wealthy in more equal countries are also less likely to exploit others and commit fraud or exhibit other anti-social behaviour, partly because they feel less of a need to cut corners to get ahead, or to make money ( Dorling 2017 : 152–153). Homicides also tend to rise with inequality. Daly (2016) reveals that inequality predicts homicide rates better than any other variable and accounts for around half of the variance in murder rates between countries and American states. Roughly 90% of American homicides are committed by men, and since the majority of homicides occur over status, inequality raises the stakes of disputes over status amongst men.

Studies have also shown that there is a marked negative relationship between income inequality and social mobility. Utilising Intergenerational Earnings Elasticity data from Blanden, Gregg, and Machin (2005) , Wilkinson and Pickett (2009) first outline this relationship cross-nationally for eight OECD countries. Corak (2013) famously expanded on this with his ‘Great Gatsby Curve’ for 22 countries using the same measure. I update and expand on these studies in Figure 4 to include all 36 OECD members, utilising the WEF’s inaugural 2020 Social Mobility Index. It clearly shows that social mobility is much lower on average in more unequal countries across the entire OECD.

Figure 4:
Index of social mobility by Gini coefficient. Data on social mobility index from World Economic Forum (2020); data on Gini coefficients from SWIID (Solt 2020).

Index of social mobility by Gini coefficient. Data on social mobility index from World Economic Forum (2020) ; data on Gini coefficients from SWIID ( Solt 2020 ).

A primary driver for the negative relationship between inequality and social mobility, derives from the availability of resources during early childhood. Life chances have been shown to be determined in early childhood to a disproportionately large extent ( Jensen and van Kersbergen 2017 : 29). Children in more equitable regions such as Scandinavia, have better access to resources, as they go to similar schools, receive similar educational opportunities, and have access to a wider range of career options. Whereas in the UK and US, a greater number of jobs at the top are closed off to those at the bottom and affluent parents are far more likely to send their children to private schools and fund other ‘child enrichment’ goods and services ( Dorling 2017 : 26). Therefore, as income inequality rises, there is a greater disparity in the resources that rich and poor parents can invest in their children’s education, which has been shown to substantially affect “cognitive development and school achievement” ( Brown 2017 : 33–34).

4 Conclusions

The causes and consequences of income inequality are multifaceted. Income inequality is not the inevitable outcome of irresistible structural forces such as globalisation or technological development. Instead, it has largely been driven by a multitude of institutional political choices. These preventable causes that have given rise to inequality have created socio-economic challenges, due to the demonstrably negative outcomes that inequality engenders.

The neoliberal political consensus poses challenges for policymakers to arrest the rise of income inequality. However, there are many proven solutions that policymakers can enact if the appropriate will can be summoned. Restoring higher levels of labour protections would aid in reversing the declining trend of labour wage share. Similarly, government promotion and support for new corporate governance models that give trade unions and workers a seat at the table in ownership decisions through board memberships, would somewhat redress the increasing power imbalance between capital and labour that is generating more inequality. Greater regulation aimed at limiting the now dominant shareholder principle of maximising value through share buy-backs and instead offering greater incentives to pursue maximisation of stakeholder value, long-term financial stability and investment, can reduce inequality. Most importantly, tax policy can be harnessed to redress income inequality. Such policies include restoring higher marginal income and corporate tax rates, setting higher corporate tax rates for firms with higher ratios of CEO-to-worker pay, and establishing luxury taxes on spiralling compensation packages. Finally, a move away from austerity, which has gripped the West since the financial crisis, and a move towards much greater government investment and welfare state spending, would also lift growth and low-wages.

Alstadsæter, A., N. Johannesen, and G. Zucman. 2019. “Tax Evasion and Inequality.” American Economic Review 109 (6): 2073–103. 10.3386/w23772 Search in Google Scholar

Andrasfay, T., and N. Goldman. 2021. “Reductions in 2020 US Life Expectancy Due to COVID-19 and the Disproportionate Impact on the Black and Latino Populations.” Proceedings of the National Academy of Sciences 118 (5), https://doi.org/10.1073/pnas.2014746118 . Search in Google Scholar

Asaria, M., J. Costa-Font, and F. A. Cowell. 2021. “How Does Exposure to Covid-19 Influence Health and Income Inequality Aversion.” IZA Discussion Paper. no. 14103. Also available at https://ssrn.com/abstract=3785067 . 10.2139/ssrn.3907733 Search in Google Scholar

Atkinson, A. B. 2015. Inequality: What Can Be Done? London: Harvard University Press. 10.4159/9780674287013 Search in Google Scholar

Baccaro, L., and C. Howell. 2011. “A Common Neoliberal Trajectory: The Transformation of Industrial Relations in Advanced Capitalism.” Politics & Society 39 (4): 521–63, https://doi.org/10.1177/0032329211420082 . Search in Google Scholar

Blanden, J., P. Gregg, and S. Machin. 2005. Intergenerational Mobility in Europe and North America . London: Centre for Economic Performance. 10.1017/CBO9780511492549.007 Search in Google Scholar

Borjas, G. J., R. B. Freeman, and L. F. Katz. 1992. “On the Labor Market Effects of Immigration and Trade.” In Immigration and the Workforce , edited by G. J. Borjas, and R. B. Freeman, 213–44. Chicago: University of Chicago Press. 10.3386/w3761 Search in Google Scholar

Brown, R. 2017. The Inequality Crisis: The Facts and What We Can Do About It . Bristol: Polity Press. 10.2307/j.ctt22p7kb5 Search in Google Scholar

Case, A., and A. Deaton. 2020. Deaths of Despair and the Future of Capitalism . Princeton: Princeton University Press. 10.1515/9780691217062 Search in Google Scholar

Center on Budget and Policy Priorities (CBPP) . 2017. Tax Cuts for the Rich Aren’t an Economic Panacea – and Could Hurt Growth. Also available at https://www.cbpp.org/research/federal-tax/tax-cuts-for-the-rich-arent-an-economic-panacea-and-could-hurt-growth . Search in Google Scholar

Checchi, D., and J. Visser. 2009. “Inequality and the Labor Market: Unions.” In The Oxford Handbook of Economic Inequality , edited by B. Nolan, W. Salverda, and T. M. Smeeding, 230–56. Oxford: Oxford University Press. Search in Google Scholar

Cingano, F. 2014. Trends in Income Inequality and its Impact on Economic Growth . OECD Social, Employment and Migration Working Papers, No. 163. Paris: OECD Publishing. Search in Google Scholar

Corak, M. 2013. “Income Inequality, Equality of Opportunity, and Intergenerational Mobility.” Journal of Economic Perspectives 27 (3): 79–102, https://doi.org/10.1257/jep.27.3.79 . Search in Google Scholar

Daly, M. 2016. Killing the Competition: Economic Inequality and Homicide . Oxford: Routledge. 10.4324/9780203787748 Search in Google Scholar

Daly, M., C. Boyce, and A. Wood. 2015. “A Social Rank Explanation of How Money Influences Health.” Health Psychology 34 (3): 222–30, https://doi.org/10.1037/hea0000098 . Search in Google Scholar

Deaton, A. 2013. The Great Escape: Health, Wealth, and the Origins of Inequality . Princeton: Princeton University Press. 10.1515/9781400847969 Search in Google Scholar

Dickerson, S. S., and M. Kemeny. 2004. “Acute Stressors and Cortisol Responses: A Theoretical Integration and Synthesis of Laboratory Research.” Psychological Bulletin 130 (3): 355–91, https://doi.org/10.1037/0033-2909.130.3.355 . Search in Google Scholar

Dorling, D. 2017. The Equality Effect: Improving Life for Everyone . Oxford: New Internationalist Publications Ltd. Search in Google Scholar

Dorling, D. 2018. Do We Need Economic Inequality? Cambridge: Polity Press. Search in Google Scholar

Elliott, L. 2017. “Rising Inequality Threatens World Economy, Says WEF.” The Guardian. Also available at https://www.theguardian.com/business/2017/jan/11/inequality-world-economy-wef-brexit-donald-trump-world-economic-forum-risk-report . Search in Google Scholar

Gabaix, X., and A. Landier. 2008. “Why Has CEO Pay Increased So Much?” Quarterly Journal of Economics 123 (1): 49–100, https://doi.org/10.1162/qjec.2008.123.1.49 . Search in Google Scholar

Hacker, J. S., and P. Pierson. 2010. Winner-Take-All Politics: How Washington Made the Rich Richer – And Turned Its Back on the Middle Class . New York: Simon & Schuster. Search in Google Scholar

Haddow, R. 2013. “Labour Market Income Transfers and Redistribution.” In Inequality and the Fading of Redistributive Politics , edited by K. Banting, and J. Myles, 381–412. Vancouver: UBC Press. Search in Google Scholar

Hager, S. 2020. “Varieties of Top Incomes?” Socio-Economic Review 18 (4): 1175–98. 10.1093/ser/mwy036 Search in Google Scholar

Heimberger, P. 2020. “Does Economic Globalisation Affect Income Inequality? A Meta‐analysis.” The World Economy 43 (11): 2960–82, https://doi.org/10.1111/twec.13007 . Search in Google Scholar

Hope, D., and J. Limberg. 2020. The Economic Consequences of Major Tax Cuts for the Rich . London: London School of Economics and Political Science. Also available at http://eprints.lse.ac.uk/107919/ . Search in Google Scholar

Hsieh, C.-C., and M. D. Pugh. 1993. “Poverty, Inequality and Violent Crime: a Meta-Analysis of Recent Aggregate Data Studies.” Criminal Justice Review 18 (2): 182–202, https://doi.org/10.1177/073401689301800203 . Search in Google Scholar

Huber, E., J. Huo, and J. D. Stephens. 2019. “Power, Policy, and Top Income Shares.” Socio-Economic Review 17 (2): 231–53, https://doi.org/10.1093/ser/mwx027 . Search in Google Scholar

International Monetary Fund (IMF). 2017. Fiscal Monitor: Tackling Inequality . Washington: IMF. Search in Google Scholar

Jaumotte, F., and C. O. Buitron. 2015. “Power from the People.” Finance & Development 52 (1): 29–31. Search in Google Scholar

Jensen, C., and K. Van Kersbergen. 2017. The Politics of Inequality . London: Palgrave. 10.1057/978-1-137-42702-1 Search in Google Scholar

Kawachi, I., N. E. Adler, and W. H. Dow. 2010. “Money, Schooling, and Health: Mechanisms and Causal Evidence.” Annals of the New York Academy of Sciences 1186 (1): 56–68, https://doi.org/10.1111/j.1749-6632.2009.05340.x . Search in Google Scholar

Leigh, A., C. Jencks, and T. Smeeding. 2009. “Health and Economic Inequality.” In The Oxford Book of Economic Equality , edited by W. Salverda, B. Nolan, and T. Smeeding, 384–405. Oxford: Oxford University Press. 10.1093/oxfordhb/9780199606061.013.0016 Search in Google Scholar

Marmot, M. 2015. The Health Gap: The Challenge of an Unequal World . London: Bloomsbury. 10.1016/S0140-6736(15)00150-6 Search in Google Scholar

Merton, R. 1938. “Social Structure and Anomie.” American Sociological Review 3 (5): 672–82, https://doi.org/10.2307/2084686 . Search in Google Scholar

Organisation for Economic Co-operation and Development (OECD) . 2020. “Income Inequality” (Indicator) . Paris: OECD Publishing. Also available at https://data.oecd.org/inequality/income-inequality.htm . Search in Google Scholar

Ostry, J. D., P. Loungani, and D. Furceri. 2016. “ Neoliberalism: Oversold? ” Finance and Development 532: 38–41. Also available at https://www.imf.org/external/pubs/ft/fandd/2016/06/ostry.htm . Search in Google Scholar

Piketty, T. 2014. Capital in the Twenty-First Century . Cambridge: Harvard University Press. 10.4159/9780674369542 Search in Google Scholar

Piketty, T. 2020. Capital and Ideology . Cambridge: Harvard University Press. 10.4159/9780674245075 Search in Google Scholar

Revenga, A. 1992. “Exporting Jobs? The Impact of Import Competition on Employment and Wages in U.S. Manufacturing.” Quarterly Journal of Economics 107 (1): 255–84, https://doi.org/10.2307/2118329 . Search in Google Scholar

Rosenfeld, J., and B. Western. 2011. “Unions, Norms, and the Rise in U.S. Wage Inequality.” American Sociological Review 78 (4): 513–37. 10.1177/0003122411414817 Search in Google Scholar

Rowlingson, K. 2011. Does Income Inequality Cause Health and Social Problems? York: Joseph Rowntree Foundation. Search in Google Scholar

Saez, E., and M. Veall. 2005. “The Evolution of High Incomes in Northern America: Lessons from Canadian Evidence.” American Economic Review 95 (3): 831–49, https://doi.org/10.1257/0002828054201404 . Search in Google Scholar

Scheidel, W. 2017. The Great Leveller: Violence and the History of Inequality from the Stone Age to the Twenty-First Century . Princeton: Princeton University Press. 10.1515/9781400884605 Search in Google Scholar

Schrecker, T., and C. Bambra. 2015. How Politics Makes Us Sick: Neoliberal Epidemics . New York: Palgrave Macmillan. 10.1057/9781137463074 Search in Google Scholar

Shaxson, N. 2015. Ten Reasons to Defend the Corporation Tax . London: Tax Justice Network. Also available at http://www.taxjustice.net/wp-content/uploads/2013/04/Ten_Reasons_Full_Report.pdf . Search in Google Scholar

Solt, F. 2020. “Measuring Income Inequality across Countries and over Time: The Standardized World Income Inequality Database.” Social Science Quarterly 101 (3): 1183–99. Version 9.0, https://doi.org/10.1111/ssqu.12795 . Search in Google Scholar

Stiglitz, J. 2013. The Price of Inequality . London: Penguin Books. 10.1111/npqu.11358 Search in Google Scholar

The Equality Trust . 2018. “The Spirit Level Data.” London. Also available at https://www.equalitytrust.org.uk/civicrm/contribute/transact?reset=1&id=5 . Search in Google Scholar

Thomas, A. 2016. Republic of Equals: Predistribution and Property-Owning Democracy . Oxford: Oxford University Press. 10.1093/acprof:oso/9780190602116.001.0001 Search in Google Scholar

Tridico, P. 2018. “The Determinants of Income Inequality in OECD Countries.” Cambridge Journal of Economics 42 (4): 1009–42, https://doi.org/10.1093/cje/bex069 . Search in Google Scholar

Visser, J. 2019. ICTWSS Database . Version 6.1. Amsterdam: Amsterdam Institute for Advanced Labour Studies (AIAS), University of Amsterdam. Search in Google Scholar

Whitworth, A. 2012. “Inequality and Crime across England: A Multilevel Modelling Approach.” Social Policy and Society 11 (1): 27–40, https://doi.org/10.1017/s1474746411000388 . Search in Google Scholar

Wilkinson, R. 1996. Unhealthy Societies: The Afflictions of Inequality . London: Routledge. Search in Google Scholar

Wilkinson, R., and K. Pickett. 2009. The Spirit Level: Why Equality is Better for Everyone . London: Penguin Books. Search in Google Scholar

Wilkinson, R., and K. Pickett. 2019. The Inner Level: How More Equal Societies Reduce Stress, Restore Sanity and Improve Everyone’s Well-Being . London: Penguin Books. Search in Google Scholar

Wiwad, D., B. Mercier, P. K. Piff, A. Shariff, and L. B. Aknin. 2021. “Recognizing the Impact of COVID-19 on the Poor Alters Attitudes towards Poverty and Inequality.” Journal of Experimental Social Psychology , https://doi.org/10.1016/j.jesp.2020.104083 . Search in Google Scholar

World Economic Forum. 2020. The Global Social Mobility Report 2020 . Geneva: World Economic Forum. Also available at https://www3.weforum.org/docs/Global_Social_Mobility_Report.pdf . Search in Google Scholar

Zucman, G. 2015. The Hidden Wealth of Nations: The Scourge of Tax Havens . Chicago: University of Chicago Press. 10.7208/chicago/9780226245560.001.0001 Search in Google Scholar

X / Twitter

Supplementary Materials

Please login or register with De Gruyter to order this product.

Journal and Issue

Articles in the same issue.

Access to Education: The Impact Of Inequality On Education

By GGI Insights | August 29, 2024

Addressing inequality requires investment in infrastructure, quality teachers, financial support, and promoting diversity and inclusion. An equitable education system provides all students with the resources to succeed and break the cycle of poverty. Inequality affects us in many ways, one of which is education. But it’s often overlooked just how significant inequality is in access and quality of education. From socioeconomic and institutional barriers to disparities in funding, inequality has far-reaching impacts on how we learn.

Inequality leaves an undeniably deep mark in areas of public education. Low-income families often need help to afford private schools, and neighborhoods with money troubles mean a need for more resources for their public schools. This can create a disheartening cycle that never stops turning: the disadvantaged aren’t provided with equal education opportunities compared to wealthier peers, making it much more difficult for them to excel academically or professionally.

The repercussions of income inequality also impact the quality of education kids receive. High schools in financially strained districts are undersupplied in terms of textbooks and tech--which, unfortunately, damages the lifelong learning opportunities of students. Underqualified or inexperienced teachers also worsen this issue and further reduce the quality of education.

'Ignite Potential' merch – wear your dedication to Quality Education. Each item symbolizes support for enlightening minds globally. Shop, support, make an impact!

What is Inequality?

Inequality is an unequal distribution of resources, opportunities, and privileges in a society. This unevenness can manifest in various forms, such as economic inequity (unbalanced wealth or income), social discrimination (such as race, gender, or ethnicity), and political unfairness (disproportional access to power and representation).

Such disparities dramatically impact individuals and broader society - they can create divides between social classes and lead to unrest in efforts to alter the systems that facilitate these inequalities. Moreover, those disadvantaged by inequality often need more resources or opportunities to reach their potential. In this respect, we should all strive for a more equitable world.

Inequality and Its Impact on Education

Individuals from socioeconomically deprived backgrounds often lack access to the same educational resources as their more privileged peers. This means that minority students of different economic standings may not enjoy parity regarding access to high-quality instruction or educational materials. Additionally, due to the absence of equitable wealth and income allotment in recent years, students of lesser financial success may need help to cover the expenses incurred by university courses or college tuition.

In a consequence of these endemic inequities, movement up the social ladder becomes a difficult task, while academic accomplishments tend to suffocate. Furthermore, the disparity in education amplifies inequality in other facets, such as gender and racial divides.

When individuals are denied equity in schooling, they may feel pressured to fall into destitution or face injustice in other forms. Inequality in education has lingering ramifications on individuals’ lives and can extend further to societies as a collective.

How Does Inequality Affect Access to Quality Education?

Inequality continues to plague access to top-notch education in countless ways. One of the telltale indications is the divergence in resources between schools.

Generally, those located in affluent neighborhoods have a greater abundance of funds allocated for educational materials, resources, and teachers than those in low-income areas. As a result, white students with higher socioeconomic status are exposed to better facilities, lower dropout rates, more educational options, higher salaries, and advanced instruction than black students.

Disparities based on race and gender also play an influential role in gaining access to quality education.

This can range from receiving unfair treatment within classrooms to the appropriation of inadequate funds for school system supplies or resources tailored for particular demographics. Everyone deserves equity in getting the education they need and deserve; anything less comes short of what our society should strive for.

It's undisputed that social and economic inequities can disproportionately negatively impact a given student's access to quality education. For example, those living in poverty typically lack the essential resources to facilitate learning—laptop computers, textbooks, transportation costs, etc.—placing them at a severe disadvantage. Furthermore, this issue is heightened by the persistent stress due to the hardships of their home life, which can impede their performance in school.

Schools must stand against these glaring discrepancies and guarantee that each student is granted equitable access to a high-caliber public education. To start, they should allocate more financial aid to underprivileged black schools just like white schools, so they boast the same infrastructure as their wealthier counterparts. In addition, classrooms should remain free from discrimination, and customs need to be established that allow struggling students to feel accepted and heard. Ultimately, if schools can develop strategies based on the realities of poverty and apply them fairly, then each pupil has a fair chance of educational success.

Ways in which Inequality is Impacting Educational Quality

In today's world, inequality has a significant impact on educational quality. Inequalities in access to resources, finances, and opportunities lead to vast disparities between those students with privilege and those without. Schools often need more funding in low-income areas to provide their students with an adequate education.

This can lead to overcrowding, unsanitary conditions, scarce learning materials, outdated technology, and more.

As a result, students who do not have access to other forms of education inequality may receive a different quality than those in wealthier school districts.

Inequalities such as racism, sexism, and classism continue to shape how students experience school differently. Historically marginalized groups in the United States do not always receive the same attention or resources from their schools as majority groups do.

This creates a cycle in which marginalized students are disadvantaged when accessing higher education opportunities or applying for better jobs requiring higher education.

Educational inequality is prevalent throughout the nation, with racism and sexism playing a major role in the discriminatory hiring practices of many schools. Minority and female teachers often find themselves overshadowed by their white and male counterparts, being overlooked for leadership positions or denied jobs altogether due to their gender or race.

Secondary school administrators must reflect upon their biases and strive to create an inclusive environment where all members of society are given the same opportunities. To truly develop a diverse workforce, active steps must be taken to ensure that people from all backgrounds are adequately represented in the staff roster.

By consciously creating a diverse and equitable hiring process, schools can foster an environment that celebrates diversity and encourages students and faculty to reach their fullest potential. To reduce inequalities and provide a higher quality education for all students, it is essential to focus on policies and initiatives that address the root causes of disparities in educational quality.

This includes investing in resources, technology, and infrastructure in low-income areas and working to create an environment of acceptance, inclusion, and understanding. It is also important to ensure that hiring practices are fair and equitable and to provide teachers with the training and specialist support they need to build strong relationships with their students.

Expanding access to extracurricular activities, career guidance services, and higher education opportunities can help create a level playing field for all students in their cognitive skills.

Causes of Inequality in Education

Inequality in education is one of the most significant issues of our time. It affects educational outcomes for students from lower socioeconomic backgrounds and those from minority communities. This inequality results in disparities in opportunity, resources, and outcomes for certain groups of students compared to others.

Exploring the causes of these inequalities helps us to identify potential avenues for change. Many contributing factors can be traced to systemic and institutional racism that has historically denied access and opportunity to young people based on their race or ethnicity. These inequities are often perpetuated by policies that do not prioritize racial equity or by biased teachers, most times, who unknowingly favor certain groups over others

Other causes include:

Poverty, often forces families with low incomes to make tough decisions between necessities like food and housing versus paying for quality education and extracurricular activities.
Unequal distribution of resources amongst school districts.
Need for adequate support systems geared towards low-income students.

The consequences of inequality in education are long-lasting, making it essential to address them now. For example, significant disparities exist between student populations from different backgrounds regarding college enrollment rates, college graduation rates, test scores, academic achievement, course selection, access to high-level classes and qualified teachers, and availability of adequate learning materials. In addressing these deep-rooted challenges, charities for education play an indispensable role, often stepping in to provide resources and advocacy where public systems are lacking.

The Influence of Racial, Gender, and Economic Discrimination

When considering how different forms of discrimination have impacted social identity development, it is crucial to consider both direct and indirect influences. For example, racial and gender disparities may affect societal interactions, such as unequal access to resources or preferential treatment among specific demographics.

Economic inequity can often determine access to secondary education or quality of life, thus influencing how people view themselves compared to others.

In addition, unconscious biases can lead individuals to come up with inaccurate stereotypes based on a person’s race or gender, which further impacts how they view themselves and who they form relationships with. All of these issues work together to shape how people perceive their identities in relation to others.

The Impact of High Student-to-teacher Ratios

High student-to-teacher ratios can significantly impact a student's educational experience. If a school has a high student-to-teacher ratio, the students receive less attention from the teacher and may need help to keep up with the rest of the larger class. This can lead to frustration, confusion, and alienation from their peers from early childhood.

Additionally, large class sizes mean less opportunity for individualized instruction tailored to each student's unique needs.

As a result, some students may need to catch up or feel overwhelmed by the pace of instruction. The lack of individual interaction between teacher and learner can also make it hard for teachers to get to know their students and form meaningful relationships with them, limiting the amount of guidance they can provide both in and outside the classroom. Overall, high student-to-teacher ratios can lead to an unsatisfactory learning environment that places significant strain on both students and teachers alike.

The Impact of Inequality on Educational Outcomes

Inequality in access to quality education contributes significantly to poor academic performance. In addition, socioeconomic disparities have an enormous impact on educational opportunities; students from low-income backgrounds often need help to attend high-performing schools or receive tutoring and additional resources, making it more challenging to achieve satisfactory results.

In addition, inequality in terms of race, gender, and other factors can influence the educational environment by creating a climate of exclusion and marginalization, which can lead to lower engagement and fewer positive learning outcomes for certain groups of students.

Inequality in access to technology can further impede academic development, as many courses now rely heavily on internet-based resources and materials that may not be available at home. All these factors can make it more difficult for certain African Americans to succeed academically. The push for education through technology is critical in such scenarios, helping to mitigate disparities by providing alternative, accessible platforms for learning.

Inequality can exist in various forms, including economic, social, educational, and political. For example, economic inequality can lead to disparities in access to resources such as quality education and employment opportunities. Similarly, social inequalities can lead to stratified education attainment based on factors such as race or class. In addition, educational inequalities can manifest in unequal access to information and resources needed for literacy development. Finally, political inequality can lead to a lack of representation or advocacy for those experiencing lower literacy attainment levels. It is important to identify these underlying factors when examining how inequality leads to higher illiteracy rates as they provide insight into potential courses of action that could be taken to reduce it.

Inequality has a profound impact on access to and quality of education. A significant consequence of inequality is the socioeconomic gap, which can significantly differ between high-income and low-income families in terms of educational resources. High-income families often provide their children with better educational opportunities than those from low-income backgrounds, such as private schools, advanced tutoring sessions, and specialized educational services.

These differences can create further gaps by creating an unequal playing field for students from different economic backgrounds. In addition to the educational access gap created by inequality, there is also an impact on the quality of education received. Research has shown that students from lower-income backgrounds often receive a lower-quality education due to limited resources and other factors, such as fewer qualified educators or inadequate school facilities.

This may lead to less access to higher-quality learning experiences and fewer opportunities for success in academic pursuits. This inequality in education has been observed in countries worldwide, making it one of the leading causes of inequality within societies today.

Popular Insights:

Shop with purpose at impact mart your purchase empowers positive change. thanks for being the difference.

Prospecting Strategies: Maximize Lead Generation and Sales Performance

What do retinoids do: skin benefits, uses, and potential side effects, sales prospecting: effective techniques for boosting lead generation, retinoids: benefits, uses, and side effects for skin care, face masks for acne: top solutions for clearer skin, acne scars: effective treatments and prevention tips for clearer skin, aloe vera for acne: effective uses and benefits for clear skin, acne scar removal: effective treatments and proven home remedies, rice mask: benefits, application tips, and recipes for glowing skin, acne scarring: effective treatments and prevention techniques.

IMAGES

Education and Income Inequality
Education and Income Inequality Relationship
Income Inequality Essay
Income Inequality in US Essay Example
Income Inequality Essay Example
Income Inequality Essay

VIDEO

Income Inequality- Economic Mobility
Income Inequality, why can’t this problem be solved?? EP1
Income inequality #education #economy #legaladvice #legalknowlege #news
491: Income Inequality and Poverty
China
Why Income Inequality Leads to Social Unrest #income #money

COMMENTS

PDF Inequality Matters
the top 1 percent of earners earn 22 percent of all income in the United States (Piketty & Saez 2013) is evidence of a distributional income inequality. A host of material, social, historical, and political conditions has increased the likelihood of disproportional access to opportunities, producing this distributional inequality.
A decade of research on education inequality in America
In 2014, I documented a 40 percent jump in the number of school-aged children living in poverty between 2000 and 2012 from one out of every seven children to one out of every five students. In the most recent report, for the 2016-17 school year, the poverty rate declined from 21 percent in 2010 to 18 percent in 2017.
A century of educational inequality in the United States
Inequalities in college enrollment and completion were low for cohorts born in the late 1950s and 1960s, when income inequality was low, and high for cohorts born in the late 1980s, when income inequality peaked. This grand U-turn means that contemporary birth cohorts are experiencing levels of collegiate inequality not seen for generations.
How COVID taught America about inequity in education
Community colleges, for example, have "traditionally been a gateway for low-income students" into the professional classes, said Long, whose research focuses on issues of affordability and access. "COVID has just made all of those issues 10 times worse," she said. "That's where enrollment has fallen the most.".
PDF Income-based Inequality in Educational Outcomes: National Bureau of
Educational Inequality There are large income-related gaps in educational attainments in the United States. For example, among students who turned 24 in 2006, those from the top quartile of family income had completed nearly four more years of schooling, on average, than those from the bottom quartile (Bailey & Dynarski, 2011).
THE IMPACT OF EDUCATION ON INCOME INEQUALITY
This paper presents new evidence on the relationship between education and income inequality by drawing evidence from 145 countries between 1996 to 2016.
The costs of inequality: Education's the one key that rules them all
Third in a series on what Harvard scholars are doing to identify and understand inequality, in seeking solutions to one of America's most vexing problems.. Before Deval Patrick '78, J.D. '82, was the popular and successful two-term governor of Massachusetts, before he was managing director of high-flying Bain Capital, and long before he was Harvard's most recent Commencement speaker ...
Income Inequality in College Enrollment and Degree Attainment During
Short of a massive undertaking of updating Chetty et al.'s (2014) analysis of administrative data linking parental income to adult children's college outcomes, it will be difficult if not impossible to get firmer, more definitive answers regarding income inequalities in higher education due to the Great Recession or COVID pandemic.
Income Inequality and Education Revisited: Persistence ...
Across all regions, decreases in the inequality of education reduced income inequality, ranging from a 1.5 point decrease in ADV and EE to a 4.8 point decrease in MENA. However, increases in the level of education increased income inequality across all regions by 1.5 to 2.2 points.
Full article: Introduction to "Racial Inequality and Education
Beyond income and wealth, inequality affects many aspects of life in American society. From access to transportation and health care to Internet services, employment opportunities, and education, inequality is shaping the character and quality of life for most Americans.
American Higher Education and Income Inequality
Abstract. This paper demonstrates that increasing income inequality can contribute to the trends we see in American higher education, particularly in the selective, private nonprofit and public sectors. Given these institutions' selective admissions and commitment to socioeconomic diversity, the paper demonstrates how increasing income inequality leads to higher tuition, costs, and financial ...
The impact of education costs on income inequality
In the U.S., reducing education costs is a key element of policy with broad support. However, my paper argues that the impact of policy on income inequality depends on the targeted type of education. The paper calibrates the U.S. data and simulates the consequences of reducing two types of education costs on income inequality.
Education and inequality in 2021: how to change the system
The average annual per-student spending for quality primary education in a low-income country is predicted to be US$197 in 2030. This creates an estimated annual gap of US$39 billion between 2015 ...
Education inequalities at the school starting gate: Gaps, trends, and
Executive summary. High and rising inequality is one of the United States' most pressing economic and societal issues. Since the early 1980s, the total share of income claimed by the bottom 90 percent of Americans has steadily decreased, with the majority of income gains going to the top 1 percent.
Recognizing and Overcoming Inequity in Education
Equity and equality are not the same thing. Equality means providing the same resources to everyone. Equity signifies giving more to those most in need. Countries with greater inequity in ...
A century of educational inequality in the United States
Income inequality begins to fall in the early 1940s, but inequalities in enrollment and completion begin to decline only for cohorts born in the mid-1950s. Men born in the mid-1940s onward were not just born into a period of low inequality, but they spent most of their formative years in a low-inequality society.
Income Inequality and Education Revisited in: IMF Working Papers Volume
This paper presents new results on the relationship between income inequality and education expansion—that is, increasing average years of schooling and reducing inequality of schooling. ... Education Levels and Education Inequality. Citation: IMF Working Papers 2017, 126; 10.5089/9781475595741.001.A001. Note: Lines are based on non ...
Growing Income Inequality Threatens American Education
In 2010, family income at the 20th percentile was more than 25% lower than the inflation-adjusted corresponding family income in 1970. In contrast, the real incomes of families at the 80th ...
Inequality undermines the value of education for the poor
Places with greater "lower-tail inequality" (the ratio of income at the 50th percentile of the income distribution to the 10th percentile) show the lowest wage gains to education for those ...
Explaining Achievement Gaps: The Role of Socioeconomic Factors
These included some traditional measures of socioeconomic status (SES), such as family income and parental education levels, but also health-related factors, such as the child's birthweight and births to teenage moms. These findings are hugely consequential for America's longstanding debates around racial inequality.
Education, inequality and social justice: A critical analysis applying
Economic capital may be generated through inherited wealth, family income or engagement in the economy for financial return. Social capital is accrued through social networks, the family and wider community interactions. ... Measuring inequality using Sen's concepts of capabilities and functionings will illuminate pathways for addressing some ...
Better Public Schools Won't Fix Income Inequality
It's worth noting that workers with a college degree enjoy a significant wage premium over those without. (Among people over age 25, those with a bachelor's degree had median annual earnings ...
Causes and Consequences of Income Inequality
Therefore, as income inequality rises, there is a greater disparity in the resources that rich and poor parents can invest in their children's education, which has been shown to substantially affect "cognitive development and school achievement" (Brown 2017: 33-34).
Access to Education: The Impact Of Inequality On Education
In addition to the educational access gap created by inequality, there is also an impact on the quality of education received. Research has shown that students from lower-income backgrounds often receive a lower-quality education due to limited resources and other factors, such as fewer qualified educators or inadequate school facilities.

A decade of research on the rich-poor divide in education

Share this:

Related articles

Jill Barshay SENIOR REPORTER

Letters to the Editor

Submit a letter

A century of educational inequality in the United States

Brian Holzman

Associated Data

Significance

College Enrollment and Completion in the Twentieth Century

Materials and Method

The Great U-turn in Collegiate Inequality

The Association between Income Inequality and Collegiate Inequality.

Supplementary Material

Data Availability.

Featured Topics

Explore the Gazette

Your kid can’t name three branches of government? He’s not alone.

‘We have the most motivated people, the best athletes. How far can we take this?’

Harvard Library acquires copy of ‘Green Book’

The costs of inequality: Education’s the one key that rules them all

Plateau on educational gains

A ZIP code as predictor of success

Better schools to close the gap

At age 1, children score similarly

Possible solutions to educational inequality:

Going wide, starting early

Share this article

Good genes are nice, but joy is better

Examining new weight-loss drugs, pediatric bariatric patients

Shingles may increase risk of cognitive decline

The impact of education costs on income inequality

Cite this article

Similar content being viewed by others

Employment Market Effects of Basic Income

Optimal linear income taxes and education subsidies under skill-biased technical change

Academic or vocational education? A comparison of the long-term wage development of academic and vocational tertiary degree holders

1 Introduction

2 The model

2.1 Firms and technologies

2.1.1 Firm’s problem

2.2 Individuals

2.2.1 Individual’s problem

Proposition 1

2.3 Markets clear conditions

2.4 Equilibrium

3 Simulation

3.2 Simulated income inequality

4 Conclusion

Data availability

Acknowledgements

Author information

Contributions

Corresponding author

Ethics declarations

Ethics approval

Consent to participate

Additional information

Supplementary Information

Supplementary file 1 (pdf 450 KB)

About this article

Share this article

JEL Classification

Education and inequality in 2021: how to change the system

Disclosure statement

Prospects for the future

A way forward: reforming the system

Director of STEM

Community member - Training Delivery and Development Committee (Volunteer part-time)

Chief Executive Officer

Finance Business Partner

Head of Evidence to Action

Education inequalities at the school starting gate : Gaps, trends, and strategies to address them

Executive summary

What we do: Questions, data and methodology

What we find

Conclusions

Introduction: Facts about income inequality and its growth over time

How large are recent performance gaps at kindergarten entry?