Racial and ethnic inequality in education has a long and persistent history in the United States. Beginning in 1954, however, when the Supreme Court ruled in Brown v. Board of Education that racial segregation of public schools was unconstitutional, some progress has been made in improving racial educational disparities. But that progress has been slow, uneven, and incomplete.
One key set of measures of racial educational equality are racial achievement gaps—differences in the average standardized test scores of white and black or white and Hispanic students. Achievement gaps are one way of monitoring the equality of educational outcomes.
The series of figures below describe recent trends and patterns in racial achievement gaps.
Every few years, a sample of 9-, 13-, and 17-year-olds from around the United States are given tests in math and reading as part of the National Assessment of Educational Progress (NAEP). NAEP, sometimes called "The Nation’s Report Card," is designed to provide the public and policymakers with an objective assessment of the math and reading skills of American children. Because NAEP has used the same tests since the 1970s, we can use it to compare the reading and math skills of children today with those of their parents’ generation. We can also use NAEP to examine trends in the white-black and white-Hispanic achievement gaps. These trends are illustrated in the figure below.
White-black and white-Hispanic achievement gaps have, in general, narrowed substantially since the 1970s in all grades and in both math and reading. The gaps narrowed sharply in the 1970s and the first half of the 1980s, but then progress stalled. In fact, some of the achievement gaps grew larger in the late 1980s and the 1990s. Since the 1990s, however, achievement gaps in every grade and subject have been declining. As of 2012, the white-black and white-Hispanic achievement gaps were 30-40% smaller than they were in the 1970s. Nonetheless, the gaps are still very large, ranging from 0.5 to 0.9 standard deviations.
Achievement gaps have been closing because Black and Hispanic students’ scores have improved very rapidly over the last 30 years. Indeed, among Black and Hispanic students, the average 9-year-old student today scores almost as well on the NAEP math tests as the average 13-year-old did in 1978; the average 13-year old today scores almost as well as the average 17-year-old in 1978. In other words, black and Hispanic students today are roughly three years ahead of their parents’ generation in math skills. In reading, they are roughly two to three years ahead of their parents. White students’ scores have also improved, but not by as much. These trends are illustrated in the figure below.
The white-black and white-Hispanic achievement gaps vary considerably among states. This is evident in the figures below, which show state-level achievement gaps for the years 1990-2013. These gaps are estimated from a version of the NAEP tests (called Main NAEP") that has been given to samples of students in each state every two years since 2003 and in some states from as early as 1990.
In some states, particularly those in the upper Midwest, like Wisconsin, Michigan, Illinois, and Minnesota, the white-black achievement gap has generally been larger than a standard deviation over the last decade, regardless of grade or subject. Some other states, like Connecticut and Nebraska, also have white-black gaps this large, as does the District of Columbia, where the gap is well over 1.5 standard deviations. In states with small black populations, like West Virginia, Hawaii, Idaho, Wyoming, Montana, Vermont, and New Hampshire, for example, the gaps are consistently smaller, typically only half as large as in the states with the largest gaps.
The same is true of the white-Hispanic achievement gaps. In some states, most notably the New England states of Connecticut, Massachusetts, and Rhode Island, but also in California, Colorado, Minnesota, and in the District of Columbia, the white-Hispanic gap is quite large, on the order of 0.90 to 1.00 standard deviations (or 1.5 standard deviations in the District of Columbia). In West Virginia and Vermont, however, the gaps only 0.30 or smaller, only one-third the size as in the states with the largest gaps.
In some cases, the gaps are large because white students in these states score particularly high on the NAEP tests; in other cases, the gaps are large because black or Hispanic students score poorly. For example, the large white-Hispanic gap in California is largely due to the low average scores of California Hispanic students (who have among the lowest average scores in the country in math or reading), not the high performance of white students (who perform at roughly the average among white students nationally). Conversely, the large white-black gap in Minnesota is not due to black Minnesota students’ particularly low scores (they are near or slightly below the national average), but is due to the fact that white students in Minnesota have very high scores.
Although the white-black and white-Hispanic achievement gaps have been narrowing nationally over the last decade or more, the rate at which these gaps are changing varies among the states. The Main NAEP data provide repeated measures of the size of these achievement gaps in each of the states between 1990 and 2013.
On average, the within-state white-black achievement gaps have been narrowing at a rate of roughly 0.05 standard deviations per decade since 2003. The corresponding rate for white-Hispanic gaps is roughly 0.10 standard deviations per decade. Although the gaps are, on average, closing, they are doing so very slowly, compared to their current size.
Nonetheless, there are some states where the gaps are closing much more rapidly. In seven states (District of Columbia, New York, West Virginia, Louisiana, Arkansas, New Jersey, and Michigan) the white-black gaps are closing at a rate of at least 0.20 standard deviations per decade, four times the rate in the average state. The white-Hispanic gap has narrowed at this rate in 10 states, many of them Southern and Midwestern states with small but growing Hispanic populations (District of Columbia, Delaware, Georgia, Tennessee, Nebraska, Iowa, Illinois, Indiana, Arizona, and Michigan).
Likewise, many states that have seen no significant change from 2003-2013 in the white-black achievement gap (21 states) or the white-Hispanic gap (28 states). In only three states (Maine, Vermont, and Colorado) has the white-black gap increased significantly in the last decade. In only one has the white-Hispanic gap increased (West Virginia).
The figure below illustrates the trend in each state’s achievement gaps from 1990-2013.
One potential explanation for racial achievement gaps is that they are largely due to socioeconomic disparities between white, black, and Hispanic families. Black and Hispanic children’s parents typically have lower incomes and lower levels of educational attainment than white children’s parents. Because higher-income and more-educated families typically can provide more educational opportunities for their children, family socioeconomic resources are strongly related to educational outcomes. If racial socioeconomic disparities are the primary explanation for racial achievement gaps, we would expect achievement gaps to be largest in places where racial socioeconomic disparities are largest, and we would expect them to be zero in places where there is no racial socioeconomic inequality.
The figure below suggest this explanation is at least partly true. Achievement gaps are strongly correlated with racial gaps in income, poverty rates, unemployment rates, and educational attainment. When these four factors are combined into a single index of racial socioeconomic disparities, the correlation between state achievement gaps and state racial socioeconomic disparities is high: for white-black gaps the correlation is 0.61-0.68; for white-Hispanic gaps it is 0.83-0.86. A large part of the variation among states’ racial achievement gaps is attributable to variation in states’ racial socioeconomic disparities.
Nonetheless, even in states where the racial socioeconomic disparities are near zero (typically states with small black or Hispanic populations), achievement gaps are still present. This suggests that socioeconomic disparities are not the sole cause of racial achievement gaps.
Despite the fact a state’s racial socioeconomic disparity is a very good predictor of its racial achievement gap (as is clear in the figure above), some states with similar levels of socioeconomic disparities have substantially different achievement gaps. For example, New Jersey and Wisconsin have very similar (and very high) levels of white-black socioeconomic disparities, but the white-black math achievement gap in Wisconsin is considerably larger (roughly 0.25 standard deviations larger) than in New Jersey. This suggests that socioeconomic disparities are not the sole cause of racial achievement gaps. Other factors—including potentially the availability and quality of early childhood education, the quality of public schools, patterns of residential and school segregation, and state educational and social policies—may play important roles in reducing or exacerbating racial achievement gaps.