Educational Evaluation and Policy Analysis

This article provides a summary of the results of the first five years of the Milwaukee Parental Choice Program, which was the first program in the United States to allow students to attend private schools with public vouchers. I begin with a brief discussion of the theoretical and research issues. Following a description of the initial program and subsequent changes, I outline who participated in the program—including characteristics of students and families and schools. I then describe the results in terms of the effects on families, student outcomes, and schools. I conclude with a discussion of the implications for this type of program and more open-ended voucher programs. For those holding extreme positions on this controversial issue, there will be both ammunition and frustration, for the results contain both positive and negative elements. The quality of both the public and private schools and therefore student outcomes vary within and between schools, and that variance is more extreme than in middle-class or wealthy communities. Some schools are excellent, and families fight to get in them and stay in them. Some are so bad that they fail and, if they are private, cease to exist—often in mid-year. The general results of the voucher program follow that pattern: Some results are clearly positive, some can be interpreted either way, and others are negative.

This article considers social and ethnic inequality in access to resources for mathematics learning in eighth grade: favorable school disciplinary climate, advanced course offerings, teacher subject-matter preparation, and emphasis on reasoning during classroom discourse. Data are from 41 states and territories1 participating in the 1992 Trial State Assessment (TSA) of the National Assessment of Educational Progress (NAEP). Socially advantaged students typically had greater access to these resources than did socially disadvantaged students. Access also depended on student ethnicity. However, the degree of social and ethnic inequality in access varied significantly across states. New methods for assessing and displaying state-to-state variation in social and ethnic inequality are illustrated. We argue that “report cards” displaying state differences in student proficiency are, by themselves, misleading; state differences in access to key educational resources provide an important supplement.

This study relates various types of educational spending to mean mathematics achievement and its social distribution among students. A nationally representative database of 7,217 12th graders was developed by linking the 1992 National Assessment of Educational Progress in mathematics to the U.S. Common Core of Data and the School District Data Book. Through applying hierarchical linear modeling to the database, the study found that spending on instruction and capital expenditures, while not related to mean achievement levels, were related to differences in achievement between SES groups; lower spending levels are associated with greater achievement gaps within schools.

In reviewing educational trends in Japan during the post-World War II era, the authors found that current national data sets do not contain questions about family background or rank of school attended that are detailed enough to permit a significant analysis of the impact of family background on educational attainment. Given Japan’s stature in the current community of nations, the continued government aversion to collecting family income or educational background data, along with precise data on rank of school attended and/or types of cram school, appears negligent. The authors discuss both cultural and political reasons that may create a reluctance to collect such data.

Research shows that employers are dissatisfied with their ability to hire good workers out of high school (Barton, 1990; Cappelli & Rogovsky, 1993). This article considers whether employers could benefit from using high school grades to identify workers whom they will value more in the long run. Using the High School and Beyond data on the sophomore cohort, this article examines the effects of high school grades on long-term productivity as measured by earnings. It finds that high school grades do have a strong and significant effect on earnings 9 years after high school for both men and women, those with and without bachelor’s degrees, and controlling for race/ethnicity, SES, region of the country, and whether the school is public or private. Using a fixed-effect model, it also demonstrates that these findings are robust even after controlling for school-level differences. The article further confirms other researchers’ findings of no or negative short-term effects of high school grades on earnings. It argues that this connection between grades and long-run productivity suggests that employers could use high school graduates’ grades to identify workers they will value.

This rejoinder is a response to the three Ballou critiques. It argues that Ballou offered some interesting and creative ideas to check the quality of the data set, but gave no conclusive evidence to demonstrate that the findings are invalid. Further analyses of the data set are discussed.