Search Results

This paper estimates the marginal returns to college for individuals induced to enroll in college by different marginal policy changes. The recent instrumental variables literature seeks to estimate this parameter, but in general it does so only under strong assumptions that are tested and found wanting. We show how to utilize economic theory and local instrumental variables estimators to estimate the effect of marginal policy changes. Our empirical analysis shows that returns are higher for individuals more likely to attend college. We contrast the returns to well-defined marginal policy changes with IV estimates of the return to schooling. Some marginal policy changes inducing students into college produce very low returns.

Published abstract reprinted by permission of the copyright owner. Show less

Three hypotheses are given to explain why a married woman's work hours might be related to her husband's education, even controlling for his wage rate. Data for a single cohort of women from the [National Longitudinal Survey of Youth] NLSY 1979 suggest that women's work hours are positively related to spousal education at the time of marriage but also fall more rapidly over time after marriage among those with the most educated husbands. Cross-sectional data from the [Current Population Survey] CPS for 1980-2010 indicate that the latter effect appears to have increased since 2000. Both men's and women's preferences for a traditional division of labour within the household are found to be negatively related to the husband's education among newlyweds but to rise faster over the course of a marriage when the husband is highly educated. Overall, the results provide evidence consistent with both marital sorting on the basis of attitudes to female work and changes in tastes that are influenced by marital quality. Little support is found for the argument that spousal education measures non-market productivity.

Published abstract reprinted by permission of the copyright owner. Show less

The large increase in computer use has raised the question whether people have to be taught computer skills before entering the labour market. Using data from the 1997 Skills Survey of the Employed British Workforce, we argue that neither the increase in computer use nor the fact that particularly higher skilled workers use a computer provides evidence that computer skills are valuable. We compare computer skills with writing and math skills and test whether wages vary with computer skills, given the specific use that is made of computers. The regression results show that while the ability to write documents and to carry out mathematical analyses yields significant labour-market returns, the ability to effectively use a computer has no substantial impact on wages. These estimates suggest that writing and math can be regarded as basic skills, but that the higher wages of computer users are unrelated to computer skills.

Published abstract reproduced by permission of the copyright owner. Show less

A quantitative model of two-sided search with ex-ante heterogeneity in both worker and entrepreneurial skills is proposed. It is possible to characterize both the competitive equilibrium and the optimal solution numerically. The competitive equilibrium is shown to be suboptimal. Less-skilled workers and firms are too selective, not matching with their comparable counterparts. High-types, on the other hand, are not selective enough. The model shows promise as a tool for evaluating the effects of labor policies (and other changes in the economy) on the composition of unemployment and on unemployment duration, as well as on wage distributions. The effect of introducing a simple unemployment insurance scheme is then twofold. First, it increases unemployment by allowing a greater proportion of low types not to match, which decreases output. Second, it decreases mismatch, which has a positive effect on output. It is possible to have a positive effect of unemployment insurance on productivity and find the optimal level of unemployment insurance. Finally, it is shown that assuming risk-neutral workers in this model is not innocuous.

Published abstract reprinted by permission of the copyright owner. Show less

Census data for 1990/91 indicate that Australian and Canadian immigrants have higher levels of English fluency, education, and income (relative to natives) than do US immigrants. This skill deficit for US immigrants arises primarily because the US receives a much larger share of immigrants from Latin America than do the other two countries. After excluding Latin American immigrants, the observable skills of immigrants are similar in the three countries. These patterns suggest that the comparatively low overall skill level of US immigrants may have more to do with geographic and historical ties to Mexico than with the fact that skill-based admissions are less important in the US than in Australia and Canada.

Published abstract reprinted by permission of the copyright owner. Show less

This paper integrates two strands of literature on overskilling and disability using the 2004 British Workplace Employment Relations Survey (WERS). It finds that the disabled are significantly more likely to be mismatched in the labour market, to suffer from a pay penalty and to have lower job satisfaction, the effects being stronger for the work-limited disabled. Giving workers more discretion over how they perform their work would significantly reduce these negative effects.

Published abstract reprinted by permission of the copyright owner. Show less

This paper is concerned with the determinants of English language proficiency (speaking, reading and writing) among immigrants. It presents a model of immigrant destination language acquisition based on economic incentives, exposure to the destination language, and efficiency in second language acquisition. A unique data set, the Longitudinal Survey of Immigrants to Australia, is used to test the model. This survey had three waves, at about six months, 18 months and three and half years after immigration. The analyses are performed by wave, type of language skill and gender using probit analysis. Bivariate probit analysis is used across waves. The hypotheses are supported by the data. The bivariate probit analysis indicates a positive correlation in the unexplained component that declines with time between waves, indicating a regression to the mean in English language proficiency.

Published abstract reprinted by permission of the copyright owner.

A related paper by the authors is indexed at TD/TNC 60.660. Show less

We consider the effect of legal access to alcohol, which is known to increase drinking behavior, on academic performance. We first estimate the effect using an [regression discontinuity] RD design but argue that this approach is not well-suited to the research question in our setting. Our preferred approach instead exploits the longitudinal nature of the data, essentially identifying the effect by comparing a student's academic performance before and after turning 21. We find that students' grades fall below their expected levels upon being able to drink legally, but by less than previously documented. We also show that there are effects on women and that the effects are persistent. The main results are robust to the inclusion of individual fixed effects, individual trends, and individual quadratics, in addition to other controls, that account for the expected evolution of performance as students make progress towards their degrees.

Published abstract reprinted by permission of the copyright owner. Show less

Using American panel data from the National Educational Longitudinal Study of 1988 (NELS:88) this paper investigates the effect of working during grade 12 on attainment. We exploit the longitudinal nature of the NELS by employing, for the first time in the related literature, a semiparametric propensity score matching approach combined with difference-in-differences. This identification strategy allows us to address in a flexible way selection on both observables and unobservables associated with part-time work decisions. Once such factors are controlled for, insignificant effects on reading and math scores are found. We show that these results are robust to a matching approach combined with difference-in-difference-in-differences which allows differential time trends in attainment according to the working status in grade 12.

Published abstract reprinted by permission of the copyright owner. Show less

Our paper investigates the relative effects of wage subsidies and further vocational training on the subsequent employment prospects of previously unemployed program participants. First, we outline a theoretical approach based on a firm's hiring decision. For the relative effectiveness of both labor market programs the assumption concerning the formation of human capital is crucial and leads to competing hypotheses for the medium and long term. On the assumption that wage subsidies have no effect on human capital they improve individuals' employment prospects less than training programs. Contrariwise, on the assumption that the formation of human capital on subsidized jobs equals that by formal training subsidization has the same employment effect as a training program. Second, we test the two hypotheses empirically, using a large administrative data set from Germany and statistical matching techniques. Our treatment groups consist of unemployed persons taking up subsidized employment or entering a further vocational training program, respectively, during March 2003. To exclude unemployment after program end we estimate the effect of keeping a subsidized job versus participating in training and taking up a job immediately afterwards. The results strongly support the latter of our competing hypotheses: previously subsidized individuals and trained individuals who found a job immediately afterwards have the same employment rates. This leads to the conclusion that firms value training on a subsidized job as much as formal training programs.

Published abstract reprinted by permission of the copyright owner. Show less