No abstract

No abstract

The authors use multinomial logit and multiple discriminant analyses to predict the probabilities that an individual will attain each of several occupational categories based on that individual's characteristics and qualifications. By estimating the parameters of this model from a sample of men and then applying them to a sample of women, the authors simulate the occupational distribution that these women would have attained had they been treated as if they were men. Even after making adjustments for taste differences between men and women, the authors find that their hypothetical results vary substantially from women's actual occupational distribution. They conclude that a significant portion of occupational segregation by sex can be attributed to discrimination. THE two most important manifestationsof sex discrimination in the labor market are wage differentials within occupations and differences in the occupational distributions of men and women. To understand fully the nature and impact of this discrimination, we must understand both underlying processes: that of wage determination and that of occupational attainment. Many researchers have used wage determination models to measure the extent of sex discrimination, frequently including some attempt to account for the variance in earnings across occupations.-We choose instead to focus directly on the process of occupational attainment.2 We develop a model that uses microdata to predict the probabilities that an individual will end up in each of several occupational categories. 'This commonly takes the form either of including occupational variables as regressors or of looking only at intra-occupational wage differentials. See, for example, Alan S. Blinder, "Wage Discrimination: Reduced Form and Structural . For an interesting way of directly comparing earnings distributions both within and between occupations, see Edward N. Wolff, "Occupational Earnings Behavior and the Inequality of Earnings by Sex and Race in the United States," Review of Income and Wealth, Series 22, No. 2 (June 1976), pp. 151 -66.2We use the term "occupational attainment" to mean only the net effect of all the forces and processes that determine the occupation of a particular individual. This should not be confused with the sociological use of the term nor with the sociological models of status attainment. Readers interested in a model

No abstract

N THEIR RECENT Growth and Change article, Pro-I fessors Lu and Tweeten analyze a set of survey data collected by the state of Oklahoma in an attempt to determine the impact of busing on student achievement.' They estimate a linear regression equation with student composite achievement test scores as the dependent variable. Their explanatory variables include a socioeconomic index and variables measuring the amount of time a student spends riding a bus, watching television, and working after school. Lu and Tweeten find that the amount of time spent on a bus has a significantly negative impact on achievement test scores for the two younger groups studied (4th and 8th grade students) but no significant effect on the test scores of the oldest group (1 1 th grade students). This article presents a reanalysis of their data with what is considered to be a more complete model and concludes that time spent on a bus has no significant effect on test scores for any of the three age groups.2 The different conclusion for the two lower grades is a result of including additional student background characteristics, most importantly the size of the community in which each student has spent most of his or her life.Since none of the students in the sample were being bused for the purpose of school desegregation, Lu and Tweeten argue that their results can be interpreted as measuring "the influence of busing per se without the statistically confounding effect of currently emotional i~s u e s . "~ While this article takes no issue with that claim, it does question the significance and correctness of their conclusions on the grounds that their regression model is incomplete and that certain of their technical methods are inappropriate. The Lu and Tweeten Modelgiven by(1) C = fl0 + O18 + where C is composite test score, B is hours of bus riding per day, S is a socioeconomic index, W is hours worked after school per week, and T is hours of television The author is an assistant professor of agricultural The regression model estimated by Lu and Tweeten is + p3W + O4T + u economics at the University of California-Davis.watched per day. The residual term u therefore includes the effects of all omitted relevant variable^.^ However, to avoid biasing the coefficient and standard error estimates for the included variables, the model should include ds many variables as are thought to affect test performance for which a reasonable measure is available, especially since the Oklahoma survey contains a large amount of such information.'The survey design and data sample are described in the article by Lu and Tweeten. Briefly, the data include survey information collected from a statewide sample of students, parents, and school administrators. Students in 4th, 8th, and 11 th grades were sampled, and a set of test scores is available for each student. The problem of incomplete information is dealt with here, as it apparently was by Lu and Tweeten, by including in the analysis only those students for whom all relevant information is available.6An Expand...