Past attempts to estimate the magnitude of compensating wage differentials have been hindered by the biasing effects of omitted variables and measurement error. We argue that a wage change formulation, estimated with panel data that contain worker reports of their own job characteristics, reduces both of these biases. Our empirical results, based on a large panel of workers in Sweden, confirm these conjectures by giving many more reasonable coefficient estimates for a wage change equation than for a wage level formulation.

This paper summarizes results from a validation study in which administrative records from a large manufacturing company are used to validate survey responses of a sample of workers from that company. Relatively little evidence of bias is found in reports of prior-year earnings and unemployment, the more salient fringe benefits, and union status. Quite diverse amounts of error variance are found in the survey reports, ranging from very small for reports of job tenure, annual earnings, and annual unemployment to very large for work hours and especially for a measure of hourly earnings constructed by dividing annual earnings by annual work hours. Furthermore, it is found that the conventional assumptions regarding measurement error in earnings models are not justified. In particular, there appears to be an important correlation between the measurement error in the reports of earnings and the level of job tenure, producing a bias in the estimated payoff to job tenure of roughly 30%. This result suggests a need for a greater understanding of the correlates of survey response errors.

M OST theoretical and empirical work on the determinants of individual earnings has placed special emphasis on an individual's educational attainment and on his or her years of work experience. Although interpreting the effects of education on earnings is relatively straightforward,1 the proper interpretation of the effects of experience on earnings is much less clear. It is certainly true that earnings tend to rise with years of experience, but the nature of the underlying mechanism that generates that increase is still largely an unresolved issue. An understanding of the way in which experience increases earnings is especially critical for the analysis of wage differentials by race and sex. Previous empirical research has shown that black men and both black and white women have flatter experience-earnings profiles than white males and that differences in the returns to experience account for a large portion of observed wage differences.2 The most widely accepted interpretation of the relationship between experience and earnings is that of the human capital model, which considers years of work experience as a proxy for unobservable investment in on-the-job training.3 According to the human capital model, wage differentials among individuals over the life-cycle are largely the result of differential patterns of investment in human capital, primarily in the form of investments in on-the-job training. Most human capital training models have been developed for the case of training that increases worker productivity in more than one firm (general training) as opposed to specific training that increases productivity in only one firm. It is frequently argued that because women expect to have a less regular pattern of labor force participation, they have a shorter work horizon than otherwise similar men and, thus, they have clear economic incentives to invest in less on-the-job training.4 As a result, women will, in general, have accumulated less human capital than men with the same number of years of experience and, consequently, their returns to experience would be expected to be lower. For black males, lower human capital investment is attributed to discrimination and/or their presumed poorer quality of schooling. Discrimination reduces the value of any potential investment, while poor schooling is thought to increase the costs of acquiring training.5 An alternative view of the earnings-experience relationship draws on models of labor market segmentation.6 These models differ from the human capital model primarily in their focus on the characteristics of jobs and job markets, rather than the characteristics of individuals. Earnings are thought to be largely determined by the labor market in which an individual works rather than the skills (or human capital) he or she possesses. Training itself is viewed as being largely technologically determined by the design of jobs, so that a specified amount of training is intrinsic in any given job. An individual acquires training by first gaining access to a job that provides training; that is, jobs and job markets intercede between an individual and investment in on-the-job training. Segmented market theorists usually argue that because hiring decisions involve a considerable amount of subjective input there is ample opportunity to practice discrimination. They cite entry level discrimination as a major institutional barrier between the primary and secondary sector, Received for publication January 17, 1978. Revision accepted for publication November 1, 1978. * Institute for Social Research and University of Michigan, and University of Delaware, respectively. 1 A recent review of this literature is given in Blaug (1976). 2 For example, see Mincer and Polachek (1974), Blinder (1973). 3The basic references are Becker (1964), Ben-Porath (1967), and Rosen (1972). 4 Mincer and Polachek (1974); Johnson and Stafford (1974). 5 The effect of discrimination on investment varies in different versions of the human capital model. It has no effect in a Ben-Porath type model, but reduces optimal investment in Rosen's model. 6 This model was popularized by Doeringer and Piore (1971).

This article investigates error properties of survey reports of labor market variables. We use the Panel Study of Income Dynamics (PSID) Validation Study, a two-wave panel survey of workers employed by a large firm that shared its detailed payroll records. Individuals' reports of annual earnings are fairly accurate. Errors are negatively related to true earnings, reducing bias due to measurement error when earnings are used as an Independent variable. Biases are moderately larger for changes in earnings. Earnings per hour are less reliably reported than annual earnings. Biases in estimating earnings functions are relatively small, but those in labor supply functions may be important.

This study proposes using correlations between neighboring children in their later socioeconomic status to bound the proportion of inequality in socioeconomic outcomes that can be attributed to disparities in neighborhood background. We apply this approach to educational attainment data from the Panel Study of Income Dynamics, which has sampled neighboring children and followed them into adulthood. We find that, once the effects of a few readily observed family background characteristics are accounted for, the correlation between neighboring children in their eventual educational attainment is only about 0.1. Given that even this figure is inflated by neighbors' similarity in unmeasured aspects of family background, the results suggest a limited role for neighborhood factors in accounting for inequality in educational attainment.

How Welfare Policies Affect Child and Adolescent Achievement by Elizabeth Clark-Kauffman, Greg J. Duncan and Pamela Morris. Published in volume 93, issue 2, pages 299-303 of American Economic Review, May 2003

The inability of formal skills such as schooling and on-the-job training to account for most of the valiation in adult labor-market success has spurred investigations of other factors. Cognitive ability, as measured by IQ-type tests such as the Armed Forces Qualifying Test has received the bulk of this attention. The consensus from research by economists and sociologists is that cognitive ability plays a significant although far from overwhelming role as a determinant of adult success (e.g., Christopher Jencks, 1979; John Cawley et al., 1997). Cognitive skills do not appear to account for more than about one-fifth of the association between completed schooling and earnings (Samuel Bowles et al., 2000). Similar magnitudes emerge in studies examining the strength of parents' cognitive ability as a predictor of the intergenerational success of their children (Meredith Phillips et al., 1998). The search for key noncognitive factors important in predicting both intraand intergenerational success has been wide-ranging. Jencks (1979) demonstrates that, net of background, formal schooling, and cognitive skills, personal traits such as industriousness, perseverance, and leadership have noteworthy associations with subsequent earnings and occupational status. Measures of social-psychological constructs such as locus of control (Julian Rotter, 1966) and self-esteem (Morris Rosenberg, 1965) have been shown to be predictive of subsequent labor-market success in some cases (e.g., Rachel Dunifon and Duncan, 1998; Arthur Goldsmith et al., 2000). Our focus in this paper is on the role of a different kind of personal characteristic: organization and efficiency, as operationalized in our data set by the housework-hours-adjusted score from five annual interviewer assessments of the cleanliness of a respondent's dwelling at the time of the interview. We argue that keeping a and organized home reflects an overall ability and desire to maintain a sense of order in a wide range of life activities. People whose homes appear clean both value order and demonstrate the ability to impose a degree of order at home. It is likely that people who are able to maintain such homes carry over the ability and desire to be organized to other aspects of their lives, such as work and parenting. The ability to maintain a degree of organization may be a skill that would command a reward in the labor market. Additionally, children raised in more organized households may be more successful in school and work. The results presented here indicate that, net of socioeconomic-status background, cognitive ability, completed schooling, housework time, and a host of other factors, the cleanliness rating of one's home is predictive of: (i) one's own earnings 25 years later; (ii) children's subsequent completed schooling; and (iii) children's earnings measured 25 years later. These results suggest the utility of continuing the search for noncognitive factors in intraand intergenerational models of schooling and labor-market success.