Effect of School Poverty on Racial Gaps in Test Scores: The Case of the Minnesota Basic Standards Tests, The

Journal of Negro Education, The, Winter 2004 by Myers, Samuel L Jr, Kim, Hyeoneui, Mandala, Cheryl

Using data from 1996, 1998, and 1999 Minnesota comprehensive statewide testing of eighth graders, we examine whether Black students perform worse than White students because Blacks are more likely to attend high-poverty schools. We find the impacts of school poverty on Black students' test scores are miniscule. The results decompose aggregate racial gaps in test scores between portions due to unequal endowments-including differences in school poverty rates-and unequal impacts that these endowments have on test scores. Ultimately, school poverty or other characteristics of students, their schools or the programs in which they participate explain little of the test score gap. We conclude that much of the Black-White test score gap can be attributed to racial differences in treatment.

The purpose of this paper is to examine whether school poverty explains racial differences in test scores. The urgency of exploring this issue is reinforced by the passage of the No Child Left Behind Act of 2001 legislation (PL 197-110). One practical application of the act is to create greater accountability among schools throughout the nation. Persistent achievement gaps-based on race or other indicators-will now signal schools that are failing. Students in failing schools will have the option of transferring to other, better performing schools. Failing schools may lose students and possibly public support. Succeeding schools may fear the influx of hoards of students from failing schools. Thus, school districts across the nation will want to know with greater specificity what can be done to reduce the racial gaps in student performance so that they can avoid the costly label "failing school" and avert the likely crisis of students fleeing the school district.

An example of the widely held belief that the poor performance of African Americans in inner-city public school districts is due to school poverty can be found in Minnesota policy debates. In the spring of 1997, Minnesotans were bombarded with press accounts documenting the poor achievement of many students on the Minnesota Basic Standards Test administered to all eighth graders in the state's public schools in 1996 and again in 1997. Reports of the 1996 results revealed wide racial and economic disparities in performance and significant disparities across jurisdictions. These accounts, and the public commentary surrounding them, often attributed racial disparities in test scores to racial differences in income or poverty of students attending particular schools. These are powerful attributions principally because public school funds in Minnesota are allocated in part by formulae that take into account the poverty of school districts. Poorer districts, by state law, receive larger supplemental allocations of public dollars (Hoxby, 2001).

Although the conventional wisdom is that racial differences in student performance can be explained by school poverty, rejecting this wisdom invites an alternative and often more disturbing explanation-there is a race effect, or an unexplained portion of the racial gap in test scores that cannot be attributed to racial differences in characteristics of students, schools, neighborhoods, or home environment. We measure the race effects using recent test score data from Minnesota and conclude that school poverty impacts are small and often statistically insignificant. Race effects dominate the results.

Background

It is well known that African Americans consistently score less well than Whites on standardized tests (Chubb & Loveless, 2002; Coleman et al., 1966; Grassland, 1971; Fischer et al., 1996; Jencks & Phillips, 1998; Lubienski, 2002; Maruyama, 2003; Orfield, Eaton, & The Harvard Project on School Discrimination, 1996; Phillips, Brooks-Gunn, Duncan, Klebanov, & Crane, 1998). Various explanations for these discrepancies include differences in family background (Brooks-Gunn, Klebanov, & Duncan, 1996; Ceci, 1990; Currie & Thomas, 1995; Johnson, 1993; Venezky, Kaestle, & Sum, 1987), economic resources (Brooks-Gunn et al., 1996; Crane, 1996; Fischer et al., 1996; Johnson, 1993), school resources (BaIi & Alvarez, 2003; Ferguson, 1991; Kozol, 1991), neighborhood poverty (Menacker, 1990; Turley, 2003; Yancey & Saporito, 1994), teacher qualifications (Ehrenberg & Brewer, 1995; Ferguson, 1991, 2003; Wayne & Youngs, 2003), and academic background (Johnson, 1993). Others have attributed the gap in test scores to differences in innate cognitive ability (Herrnstein & Murray, 1994; Lynn, 2002), although there is great dispute over the size of the effects of cognitive ability on achievement (see Johnson, 1993), as well as the causes of those effects. Some of the current literature on achievement tests shows persistent racial gaps even when controls are made for variables such as family background and socioeconomic status (Fischer et al., 1996; Kao & Thompson, 2003; Eubienski, 2002).

There remains a significant debate, evident even in previous issues of this Journal (Armor, 1997; Trent, 1997), as to whether the race effect remains after one controls for school poverty. The "race effect" is understood to mean an independent influence of the race of students on test performance that remains after one controls for covariates such as school poverty. The presence of a race effect that remains after controlling for theoretically relevant determinants of student achievement can be interpreted to mean that there is racial discrimination.