What’s behind huge academic gap at U.S. public schools?

"The socioeconomic profile of a district is a powerful predictor of the average test score performance of students in that district," says Sean Reardon. "Nonetheless, poverty is not destiny: There are districts with similarly low-income student populations where academic performance is higher than others." (Credit: Ian Koski/Flickr)

Almost every school district enrolling large numbers of low-income students has an average academic performance significantly below the national grade-level average, according to new research based on data set recently created from more than 200 million test scores.

The research also revealed that nearly all US school districts with substantial minority populations have large achievement gaps between their white and black and white and Hispanic students.

“Poverty is not destiny.”

The data provide the most detailed account yet of academic disparities nationwide. They comprise reading and math test results of some 40 million 3rd to 8th grade students during 2009-13 in every public school district in the country, along with information about socioeconomic status, school district characteristics, and racial and economic segregation.

“We don’t administer a single standardized exam to all US students, so a clear picture of the differences in academic performance across schools and districts has been elusive up until now,” says Sean Reardon, professor of education at Stanford University, who devised the statistical methods that make it possible to compare the mandatory tests administered in different states. “It’s now much easier to identify school districts and communities where performance is high, compare them with demographically similar ones that do less well and try to determine what’s behind the differences.”

Patterns in education inequality

Reardon and colleagues were able to identify some key patterns:

  • One-sixth of all students attend public school in school districts where average test scores are more than a grade level below the national average; one-sixth are in districts where test scores are more than a grade level above the national average.
  • The most and least socioeconomically advantaged districts have average performance levels more than four grade levels apart.
  • Average test scores of black students are, on average, roughly two grade levels lower than those of white students in the same district; the Hispanic-white difference is roughly one-and-a-half grade levels.
  • Achievement gaps are larger in districts where black and Hispanic students attend higher poverty schools than their white peers; where parents on average have high levels of educational attainment; and where large racial/ethnic gaps exist in parents’ educational attainment.
  • The size of the gaps has little or no association with average class size, a district’s per capita student spending, or charter school enrollment.

The researchers stress the findings don’t prove cause and effect, but they do point to promising areas for further study.

[Most urban schools in U.S. get a failing grade]

“The socioeconomic profile of a district is a powerful predictor of the average test score performance of students in that district,” Reardon says.

“Nonetheless, poverty is not destiny: There are districts with similarly low-income student populations where academic performance is higher than others. We can—and should—learn from such places to guide community and school improvement efforts in other communities.”

In a paper posted online accompanying the data, Reardon specifically examines the relationship between segregation and academic achievement, using 16 different measures of segregation to identify which aspects of racial segregation are most strongly associated with academic achievement. “The racial difference in the proportion of students’ schoolmates who are poor is the key dimension of segregation driving [the association],” he writes.

The findings suggest that racial segregation is inextricably linked to unequal allocation of resources among schools; and that policies that don’t address this will fail to remedy racial inequality, Reardon says. “In sum, racial integration remains essential for reducing racial disparities in school poverty rates.”

[Fewer poor students ditched schools on probation]

In another paper, researchers focus on how geography correlates with disparities by race and ethnicity. Large white-black achievement gaps in such major school districts as Atlanta, Georgia; Auburn City, Alabama; Oakland, California; Tuscaloosa, Alabama; Charleston, South Carolina; and Washington, DC. They also find significant black-white gaps in a number of smaller school districts that are home to major universities: Berkeley, California; Chapel Hill, North Carolina; Charlottesville, Virginia; Evanston, Illinois; and University City, Missouri. The list of places with the largest white-Hispanic gaps includes Atlanta, Berkeley, Chapel Hill, and Washington, DC.

The testing data have a small margin of error and shouldn’t be used to rank school districts whose performance differs only slightly, Reardohn says.

Further, the achievement patterns identified in the research don’t indicate which school districts are more effective than others, he says. “Test scores are shaped by many factors: home environments, neighborhoods, childcare and preschool experiences, and after-school experiences, as well as by school experiences.”

Other researchers from Stanford and from the University of California, Berkeley, and Harvard University contributed to the research.

The two studies and the data can be downloaded for free from the Stanford Education Data Archive.  Reardon’s paper will be published in an upcoming issue of the  Russell Sage Foundation Journal of the Social Sciences. The US Department of Education’s Institute of Education Sciences, the Spencer Foundation, and the William T. Grant Foundation supported the work.

Source: Stanford University