Knowledge that Transforms

We use a unique data set to investigate whether a matching model can describe turnover in an accounting firm. The main focus of the article is to determine whether the probability of separation from employment varies in the way described by Jovanovic. The evidence suggests that as tenure increases both terminations and quits follow the predicted pattern.

The theory of compensating differentials predicts a negative relationship between wages and good working conditions, while the theory of segmentation predicts a positive one. Combining the hedonic wage model and the wages‐employment collective bargaining model, we show the relevance of a further factor: a union power effect. Then we test the validity of this effect with French cross‐section data. Empirical results confirm the predictions of the model, that is, the coexistence of a negative relationship between wages and good working conditions for the whole sample (market effect) and a positive relationship in highly unionized sectors (union power effect).

This article explores human capital acquisition decisions when job placement helps determine competition for a worker. With asymmetric information, workers may invest in firm‐specific capital without long‐term contracts. Specific investment increases promotion chances (and hence wage competition), shifting competition back to a time when firms are symmetrically uninformed. If general human capital is the efficient (output‐maximizing) investment, then an equivalent firm‐specific investment maximizes expected career wages. This is a general result for sellers in second‐price auctions: sellers (of labor) invest to maximize the expected second‐highest bidder valuation (wage), not the winner's expected valuation.

A worker's decision whether to apply for public transfers may depend not only on his expected level of forgone labor earnings but also on his degree of uncertainty about such earnings. This article provides theory and evidence about the effects of earnings and eligibility uncertainty on participation decisions. The application rate to the Social Security Disability Insurance program is estimated to be about 15% higher than it would be in the absence of earnings risk. As an application to tax policy, optimal marginal wage tax rates may be higher than indicated in previous analyses involving wage uncertainty.

This article builds a simple model of worker sorting and matchspecific productivity to explain interindustry wage differentials. High‐skilled workers sort themselves into the industries offering more jobs that are better matched to them, and those industries pay higher wages (on average). In job transition following an exogenous job separation, the likelihood of industry switching is higher among marginal workers—low‐(high‐) skilled workers in high‐(low‐) wage industry. Empirical findings from a sample of exogenous job separations created from the Displaced Worker Surveys are generally consistent with the implications of the model.

Technological innovation has reduced the effect of diabetes. Diabetic behavioral modification in the face and expectation of medical improvement should lead to improved labor market outcomes. This article uses three cross‐sectional data sets, from 1976, 1989, and 1992, to document improvements in diabetic labor market performance. Women diabetics have significantly increased their labor force participation while male diabetics have slightly reduced their participation relative to nondiabetics.

This study quantifies the roles of national, region‐specific, and industry‐specific shocks in aggregate employment fluctuations in U.S. regions and industries. Variation among the growth rates of major regions and industries is decomposed into unobserved national, region‐, and industry‐specific components. The results reject the view that any heterogeneity in regional fluctuations is attributable to differences in industry composition. After controlling for industry mix effects, roughly 40% of the variance of the cyclical innovation in any region's growth rate is particular to that region. In addition, region‐specific shocks appear to propagate across regions over time.

Using Italian Social Security records for male workers from a sample of firms in Turin from 1981 to 1983, we show that conditional on the worker's own wage the average wage in the establishment for similar workers is negatively related to quits. We also find that this variable predicts future wage growth. This is consistent with an economic model in which workers compare the longrun value of employment opportunities when making quit decisions.

We attempt to further characterize the search strategies of the employer. In the article, we discuss how characteristics of the employer or conditions that the employer faces affect the optimal search strategies and the probability of filling a vacancy in each period. Semiparametric and parametric methods are used to estimate hazard rates of filling vacancies. The results suggest that for the given sample of vacancies, the general form of the hazard function is nonmonotonic. Additionally, the results suggest that those employers who have advance notice of the vacancy may search longer than those employers who do not.

This article reexamines the negative seniority‐earnings relationship for academic economists. The empirical results show that the anomalous negative seniority effect found in earlier academic market studies holds in the absence of direct measures of research productivity. The negative effect, however, eventually disappears as more comprehensive measures of publishing, citations, and other productivity measures are included in the wage equation to control for the quantity and quality of faculty productivity. Faculty with greater seniority appear to be rewarded relatively less simply because many have been relatively less productive than their colleagues with less seniority at similar stages in their careers.