In lieu of an abstract, here is a brief excerpt of the content:

Chapter 1 Inequality and Black-White Achievement Trends in the NAEP KATHERINE MAGNUSON, DAN T. ROSENBAUM, AND JANE WALDFOGEL H ow do recent changes in economic inequality and related social dimensions of inequality relate to trends in black-white test score gaps? In this chapter, we provide new evidence on the question, analyzing the links between inequality and black-white achievement trends for nine-year-olds using the National Assessment of Educational Progress Long-Term Trend data (NAEP-LTT) from 1975 to 2004. The NAEP-LTT series is the best single source of data on trends in black and white test scores, but contains only limited data on family-level factors. We therefore augment it by merging in information about the average characteristics of black and white families with children, by state and year, from the March Current Population Survey (CPS). These average characteristics cannot tell us about an individual child’s family but can tell us about the average family contexts that children of each race group experienced in a particular state and year. We also merge information about state-level annual income inequality among all families with children from the March CPS. This measure will not capture an individual child’s position in the income distribution but will provide information about overall income inequality in that state and year. To briefly preview the results, we find that controlling for the child’s characteristics, as reported in the NAEP, and the average characteristics of families from the same race group in that child’s state, from the March CPS, does help explain a portion of the black-white test score gaps in math and reading, with parental education appearing particularly important. We also find that when income inequality in a state is higher, children’s math and reading test scores are lower. However, income inequality seems to adversely affect a broad range of children, suggesting that increases in 33 inequality have suppressed performance overall but may not have widened black-white achievement gaps. Background As noted in the introduction to this volume, different socioeconomic standing among white and black children has been a common explanation for test score gaps (Magnuson and Duncan 2006). Parents’ education, in particular , has been singled out as an important determinant of children’s achievement and a possible source of convergence (Haveman and Wolfe 1995). Although the educational attainment of both white and black parents has been increasing, the black gains have been relatively larger, suggesting that they might explain the comparatively larger test score gains among black students relative to their white counterparts. Michael Cook and William Evans (2000), for example, found that the convergence of parental education between white and black students accounted for 25 percent of the convergence in thirteen-year-olds’ NAEP-LTT blackwhite test score gaps between 1970 and 1988. Likewise, both David Grissmer, Ann Flanagan, and Stephanie Williamson (1998) and Mark Berends and Roberto Peñaloza (chapter 2, this volume) have pointed to changes in parental education as an important source of gap convergence through the early 1990s. The role of parental education in explaining patterns of test score gaps in more recent years, however, has not been considered (see also chapter 2 this volume). National trends in maternal education suggest that the greatest convergence probably occurred before the early 1990s. The gap in years of education between white and black mothers in the March CPS narrowed a full year, from 1.7 years (11.3 versus 9.4 years) in 1967 to just 0.7 years (13.4 versus 12.7 years) in 1994. Since 1994, however, the gap has remained constant or increased as both white mothers’ and black mothers’ education has increased at similar rates, to 13.9 and 13.0 years, respectively, in 2005.1 It is important to note that tracking national trends in parental education is complicated by measurement challenges. An important source of error in the March CPS and other studies is the practice of asking about years of education completed but not about degrees attained. In particular , respondents who did not complete high school but obtained a General Educational Development (GED) certificate may report that they have completed twelve years of schooling, the same level reported by respondents with a high school degree, even though studies have shown that a GED is not equivalent to a high school degree (Heckman and LaFontaine 2006). Given that the proportion of GED recipients has risen over time, CPS trends overstate growth in true educational...

Share