After Sean Reardon, Ph.D., Professor of Poverty and Inequality in Education at Stanford University, spoke at Evanston Township High School on May 4, several members of the District 65 School Board have argued that Dr. Reardon’s research demonstrates that school districts such as Berkeley and Atlanta have twice the income gap as District 65, but the achievement gap in District 65, Berkeley and Atlanta is about the same.
They argue that since black students in District 65 come from families with higher incomes than black students in Berkeley and Atlanta, they should be performing at higher achievement levels than black students in Berkeley and Atlanta.
The School Board members rely on a chart entitled, “White-Black Achievement Gap by SES Difference” that Dr. Reardon used in his presentation on May 4 in support of their position. That chart is reproduced at below as (“Figure 1”).
While Figure 1 reflects that District 65 has one of the largest achievement gaps in the nation, Dr. Reardon’s study shows that black students in District 65 have, on average, scored significantly better than black students in Berkeley and Atlanta, and better than 78% of the school districts in the nation.
In addition, in calculating the socioeconomic status (SES) rating for black students in District 65 that is shown in Figure 1, Dr. Reardon used a median income of $55,535 for black families who had students attending the District. District 65’s free-lunch data, though, suggests that the median income of black families with students attending the District may be two-thirds that amount.
Dr. Reardon said his nationwide study demonstrates there is “a very strong relationship” between socioeconomic status (SES) and average academic achievement.
“We have an enormous amount of educational inequality in the United States driven largely by differences in the kind of opportunities that children have in their early childhood and in their early elementary schools,” said Dr. Reardon on May 4. “In most places, those opportunity gaps are evident in third grade – or earlier when we have earlier data to look at. In most places, school systems are unable to undo that.”
He adds, though, that disparities in SES are not the only explanation for black/white achievement gaps, and gave as some examples discrimination and unequal expectations or treatment in school.
This article is not intended to suggest that District 65 is making sufficient progress in raising the achievement of black students, but it provides additional information concerning the achievement profile of black students at District 65 and the nature of the achievement gap at District 65. In addition, the article provides additional information concerning the income levels of black families in the District, which bears directly on the extent of the opportunity gap between white and black students in the District.
Figure 1, titled White-Black Achievement Gap by SES Difference, plots all school districts in the nation with at least 100 black students/grade in 2009-2013. The vertical line shows the magnitude of each school district’s black/white achievement gap for the period 2009-2013. The horizontal line shows the magnitude of the difference in SES between black and white students.
The chart identifies eight school districts that have the highest achievement gaps in the nation: Evanston (District 65), Berkley, Washington D.C., Atlanta, Chapel Hill, Oakland, Madison, and Minneapolis.
The achievement gaps in these districts, expressed in standard deviation units, range from a low of 1.247 at Minnesota to a high of 1.602 at Berkley. The gap in District 65 is 1.491. A standard deviation is “very roughly” equivalent to about three grade levels in achievement. See Sean F. Reardon, Demetra Kalogrides, Andrew Ho, Ben Shear, Kenneth Shores, Erin Fahle. (2016). Stanford Education Data Archive (Gap C, Version 1.1, ) (“SEDA”).
A New York Times article, “Money, Race, and Success: How Your School District Compares” (April 29, 2016), provides an estimate of the grade level performance of black students and white students in each of the eight districts, using data available in Dr. Reardon’s study. A zero represents average grade level. A negative number indicates the number of years (or fraction or years) below grade level. A positive number indicates the number of years (or fraction of years) above grade level.
The data shows that while District 65 has a large achievement gap, both black and white students in Evanston are scoring significantly better than black and white students in Berkley, Atlanta, and the five other districts. The data is illustrated in Figure 2.
“In Evanston, black students score a little bit below the national average of all students in third grade and a little below the national average at eighth grade. Not much of a difference,” said Dr. Reardon on May 4.
He added that black students in District 65 scored higher on average than black students in 78% of the school districts in the nation.
White students score at the top of the nation, said Dr. Reardon. “The average white third-grader in Evanston is scoring at the seventh-grade level nationally. The average eighth-grader is scoring at about the 12th-grade level.
“Evanston has one of the largest achievement gaps in the country,” said Dr. Reardon. “But it’s a large achievement gap, not because the black students are failing, but because the white and affluent students are being enriched by so much opportunity.
“It’s clear that the gaps are already there at third grade. They’re not getting appreciably smaller over third to eighth grade, but something is happening that is creating really different opportunities prior to that time.”
When there are large achievement gaps in school districts like Evanston, “one reason is there are socio economic disparities,” Dr. Reardon said, referring to differences in household income levels and educational levels of parents. “There’s differences in the resources that families can provide their kids.”
In Figure 1, Evanston is an outlier among the eight school districts identified as having the largest achievement gaps, when viewed in the context of SES differences. The chart shows that Evanston has an SES rating between 2 and 3, while the other seven school districts each have a rating of 4 or more.
In determining an SES rating, Dr. Reardon said he took into account six factors: median family income, percent of adults with a bachelor’s or higher degree, poverty rate, unemployment rate, SNAP eligibility rate, and the percent of families headed by a single parent.
Dr. Reardon told the RoundTable he used estimates for each of the six factors provided in the 2005-2009 American Community Survey (ACS) tabulated by the U.S. Department of Education, National Center for Education Statistics, Education Demographic and Geographic Estimates Program (EDGE). The estimates are averages for the five year period.
He told the RoundTable he used the 2005-2009 ACS because subsequent ACS surveys tabulated by EDGE did not include estimates for each of the six factors. He added, “We’ve found that the demographics of communities didn’t change much over time, so the 2005-09 estimates are generally a good proxy for those in 2009-13.”
The achievement gap data is for the period 2009-13. In theory, the SES data would be for the same period.
Dr. Reardon’s team developed a measure for each school district’s SES which was standardized to have a mean of zero and a standard deviation of 1.
Thus, school districts with an SES of zero are at the national average. Those above a zero have an SES above the national average. Those with an SES below zero, have an SEs below the national average.
For Evanston, Dr. Reardon used a median income of $55,535 for black families and $142,617 for white families in determining SES ratings. After taking the median incomes and other factors into account, the SES rating was -0.56 for black students and +1.90 for white students. The disparity between white and black students is 2.46 standard deviation units.
For Berkley, he used a median income of $35,954 for black families, and $115,216 for white families. After taking the median incomes and other factors into account, the SES rating is -2.85 for black students and +1.57 for white students. The disparity is 4.42.
Because the difference in the SES rating between black and white students in District 65 seemed low, the RoundTable asked District 65 to provide the percentage of black students who qualified for the federal free-lunch program at various points in time. Paul Goren, Superintendent of District 65, told the RoundTable that in the school year 2009-2010, 56% of the District’s black students qualified for free lunch, meaning that they are from households with incomes less than 130% of the federal poverty rate.
In 2009-2010, the income guidelines to qualify for free lunch for a family of three was $23,803; for a family of four, $28,665; for a family five, $33,527, and for a family of six, $38,389.
Assuming that 90% of the black families who qualified for free lunch were from families of five or less, the median income for black families would be below $33,527, significantly less than the $55,535 used for Evanston in Dr. Reardon’s analysis. It would also be lower than the $35,954 used for Berkley.
In 2013, 68.5% of black students attending District 65 qualified for free lunch. This was a 21% jump from 2009. In that year, the income guidelines to qualify for free lunch for a family of three was $25,389; for a family of four, $30,615; for a family five, $35,841; and for a family of six, $41,067.
Assuming that 73% of the black families who qualified for free lunch were from families of five or less, the median income for black families would be less than $35,841, again significantly below the $55,535 used by Dr. Reardon.
The accompanying Figure 3 shows the percentage of black students who attended District 65 in 2009 and in 2013 – 2017 who qualified for free lunch and who qualified for free- and reduced-fee lunch. About 70% of the District’s black students have qualified for free lunch in the 2013 – 2017 time frame.
The RoundTable provided the 2013 data to Dr. Reardon and asked if he had any explanation for the inconsistency between the median income figure he used for black families and District 65’s free-lunch data. He told the RoundTable that the percentage of black students who qualified for free/reduced price lunch in District 65 is “much higher than we’d expect from a median income of about $55,000 (for black families). There is, of course, some error in the ACS data since it is based on samples, but I doubt that is the primary reason for the discrepancy. So I don’t have a good answer for you. The two pieces of data about socioeconomic characteristics of black families in Evanston do not seem commensurate with one another. But I can’t say which is wrong or why.”
In a subsequent email, Dr. Reardon suggested that the RoundTable contact Doug Gerverdt, the program manager at NCES for EDGE. Mr. Gerverdt offered six possible areas to consider. Five did not appear to apply or explain the discrepancy. See sidebar. The sixth point made by Mr. Gerverdt is one raised by Dr. Reardon, “ACS income estimates, like all survey estimates, have a margin of error. It’s the best guess based on information reported from sampled households, but it may not be the same value you’d get if all households reported the information.”
In order to qualify for free-lunch status, a family must submit a signed application demonstrating an income less than 130% of the poverty rate, or be qualified under another federal program which has the same or lower income eligibility requirements.
SES and Opportunity
While there is uncertainty about why there is an apparent inconsistency between the estimate of median income for black families used in Dr. Reardon’s study and the maximum median income indicated by District 65’s free-lunch data, there is a clear pattern that shows for the last five years, approximately 70% of black students enrolled in District 65 qualified for free lunch, and an additional 6-8% qualified for reduced-fee lunch.
In contrast, about 6% of white students qualified for free- or reduced-fee lunch.
“There’s lots of reasons that provide opportunities for white students in Evanston,” said Dr. Reardon on May 4. “The same is not true for African American and low-income students. They’re not failing. They are well above what you would predict, but they are still well below the level of scores of white and affluent students.
“To reduce educational inequality we need to think about strategies that involve whole communities, not just the school system,” said Dr. Reardon. “We need to think about early childhood opportunities, equality of neighborhood environments, supports for families with children to make sure they can be the best parents, the first teachers, for their children. We need to think then about how the school system can sort of integrate with those other community organizations so there’s a wrap-around focus on equity.”
In response to a question from the RoundTable, Doug Geverdt, program director at National Center for Education Statistics for EDGE, listed six potential data issues to consider in exploring why there is a discrepancy between the estimate of the median income for black families used in Dr. Reardon’s analysis and District 65’s free-lunch data. Two potential issues would only arise only if the median income for black families that Dr. Reardon used in his analysis was based on income estimates for all black families in District 65, rather than just families in District 65 who had children enrolled in the District. Dr. Reardon told the RoundTable that the ACC/EDGE data he used was for black families that had children attending District 65.
Another question posed by Mr. Geverdt is whether District 65 exercised provisions under the Community Eligibility Program that result in school-wide eligibility of all students for free lunch. Raphael Obafemi, CFO at District 65, said the program is not applicable to District 65.
Mr. Geverdt referred to a federal law under which students are directly certified for free lunch if their families qualify for payments under the Supplemental Nutrition Assistance Program (SNAP) or under Temporary Assistance for Needy Families (TANF). The income level to qualify for SNAP is the same as the income level to qualify for free lunch; the income level to qualify for TANF is lower than the income level to qualify for free lunch, so students directly enrolled through these programs would appear to meet the income eligibility requirements for the free-lunch program. There are also provisions that enable children to be directly certified for free lunch if Medicaid data demonstrates they meet the income criteria for free lunch. Students qualifying through these programs are from households with incomes less than 130% of the federal poverty rate.
Dr. Geverdt also raised a potential issue that the ACC/EDGE data being used in Dr. Reardon’s analysis was for the period 2005-2009, while the free-lunch data that the RoundTable provided to Mr. Geverdt pertained to 2013 and subsequent years. After receiving Mr. Geverdt’s coments on this issue, the RoundTable obtained free-lunch data from District 65 for the year 2009-2010, which District staff said was the earliest year for which data was available.