Just Who Loses?: Discrimination in the United States, Volume 2

10. Just Who Loses?

10 Just Who Loses? T he presentation of empirical results has reached its end. Over the course of this journey we noted the central distinction between inequality and discrimination; outlined the limitations of common alternative explanations of inequality; described and justified the specific research decisions implemented to produce these analyses; and presented the results of investigations of the association between discrimination and education, labor supply, socioeconomic status, occupational autonomy, occupational segregation, earnings, poverty, and mortality, mostly over six decades in U.S. history. Now I briefly summarize some of the key conclusions of the investigation. The work presented here stands on the foundation conveyed in Theorizing Discrimination in an Era of Contested Prejudice, the first volume in this effort, which itself stands on the work of several scholars in multiple fields. In that volume the sociotheoretic, epistemological, and substantive limitations of the dominant perspective on and approach to discrimination are conveyed, and the basis for the approach followed here is developed and described. The findings here could not have been produced absent a rejection of the dominant perspective and the erection of a foundation for research more attuned to the way in which discrimination really exists and operates as a social phenomenon. On the basis of that foundation I obtained findings that are at variance with the stylized understanding of the role of discrimination in the United States. One caution is in order before summarizing the results. Some of the findings may be surprising. Such findings should motivate additional 296 / Chapter 10 research, no matter how solid the theoretical basis of those findings may appear to be. This is the nature of research. In 2012 physicists obtained results consistent with the clearly theorized, long-sought, heretofore unseen Higgs boson (Higgs 1964; CMS Collaboration 2012). Despite the strong consistency between their results and theoretical characteristics of the particle , the researchers soberly note that more research is needed to confirm what they seem to have found. Despite the vast distance between the fields in question, I echo their caution in this very different enterprise. A Broad Assessment of the Findings The most important conclusion from our results answers one interpretation of the question that titles this volume: when it comes to discrimination, all groups lose. Certainly, the contour of discrimination effects is not homogenous. Sometimes some groups appear to gain an apparent increment of advantage in connection with discrimination. Sometimes discrimination is associated with advantages for targets. Yet in every sphere, the vast majority of statistically significant effects reflect decrements connected to discrimination, and those decrements are levied on targets and nontargets as well. The lay, the elite, and the legal understanding of discrimination is that it is a zero-sum phenomenon. According to this view, whatever targets lose, nontargets, or some subset of them, gain. While the legal system may require some assumption of rough balance of losses and gains, the social system need not comply with that assumption. The analytic approach sets aside that assumption, and doing so reveals that targets and nontargets lose. In education we found that discrimination is associated with the special education placement of blacks and whites and boys and girls. The association remained even after other factors were controlled. Matters were complex, however, because discrimination sometimes is associated with higher probability of placement in special education, whereas at other times discrimination is associated with lower probability of placement in special education. The former pattern is consistent with warehousing of students under discrimination, while the latter pattern is consistent with a more passive neglect of students under discrimination. For measured achievement, however, matters were not complex. Race discrimination was associated with decrements in measured achievement for whites and blacks, and sex discrimination was associated with decrements in measured achievement for boys and girls. The race discrimination effect appeared to operate through structural factors such as school characteristics, because inclusion of such factors completely explained the race discrimination effect. However, sex discrimination effects were not explained by the inclusion of controls. The effect of discrimination on measured achievement is substantial. Some of the discrimination effects are larger than the estimated effect on Just Who Loses? / 297 achievement of having a nonintact family, while others are larger than the effect of having a parent who graduated from high school compared to having a parent who dropped out of high school. Because family structure and parental education are demonstrably important to students’ measured cognitive achievement, showing that...