Knowledge that Transforms

We examine the (sequential) introduction of early retirement provisions to Canada's two public pension plans. These reforms provide a unique opportunity to assess the effect of public pension plan parameters on labor supply behavior, free of the biases that potentially affect the simple time‐series or cross‐section inference presented in many previous studies. We find that the reforms led to an increase in pension receipt but had little immediate effect on labor market behavior. This is due to the fact that men who initially took advantage of the early retirement provisions would otherwise have had limited labor market participation.

This article considers whether two commonly used sources of information on employer tenure, the Panel Study of Income Dynamics and the Current Population Survey, yield systematically different trends in employer tenure. Little evidence of a discrepancy between the data sets in the 1980s or 1990s is found when comparable samples, variable definitions, and time frames are used. Neither data set shows a significant trend in the share of workers with 1 year or less of tenure, while both data sets show an increase in the fraction of men with less than 10 years of tenure starting in the late 1980s.

I study workers' incentives to invest in general human capital (education) in the presence of search‐induced unemployment. Workers queue for jobs, and firms prefer to hire the most productive applicants because of rent sharing. As a result, an unemployed worker's ranking relative to other job seekers will influence his job‐finding rate. This creates a “rat race,” where workers invest in education partly in order to achieve a better ranking. In equilibrium, identical workers may have incentives to diversify in terms of education, and the investments in education may exceed the socially optimal level.

This article provides evidence on changes in short‐term job instability and insecurity using the Survey of Income and Program Participation. Monthly measures from this data set are contrasted with annual measures from the Survey of Income and Program Participation and the Panel Study of Income Dynamics. Neither data set shows an increase in job turnover during the 1980s and 1990s. We also examine indicators of increased insecurity. These include the probability that a job ends involuntarily, is followed by a spell of nonemployment, or that the subsequent job has lower wages. These indicators of insecurity also show no upward trend.

Estimated negative substitution effects on work hours question the empirical validity of the classical labor supply model. Estimates are reconciled by allowing a dual choice of hours and effort for piecerate workers. In such a model, these negative substitution effects result from substituting on‐ and off‐the‐job leisure. We test our model using controlled experimentation on human subjects. These experiments, while not naturally occurring environments, represent real economic choices and can generate data unavailable elsewhere (e.g., effort data). The results support our model, and they have implications both for labor management and for empirical research focusing only on the hours choice.

I examine the extent to which workers who lose jobs obtain work in alternative employment arrangements, including temporary work and independent contracting, and obtain voluntary or involuntary part‐time work. I find that job losers are significantly more likely than nonlosers to be in both temporary jobs (including on‐call work and contract work) and involuntarily part‐time jobs. I also find evidence that temporary and involuntary part‐time jobs are part of a transitional process subsequent to job loss leading to regular full‐time employment.

This article attempts to determine whether wage records reported by employers to state unemployment insurance (UI) agencies provide a valid alternative to more costly retrospective sample surveys of individuals as the basis for measuring the impacts of employment and training programs for low‐income persons. We analyze UI data and survey data for a sample of low‐income adults and youths from 12 sites in the National Job Training Partnership Act (JTPA) study. Our comparison indicates that impact estimates based on UI data and survey data were usually comparable. However, average surveyreported earnings were higher than average UI‐reported earnings.

This article defines and analyzes job security in the context of implicit contracts designed to overcome incentive problems in the employment relationship. Contracts of this nature generate predictions concerning the relationship between job security parameters—such as worker seniority and sectoral economic conditions—and the probability of separations. To test these predictions, I estimate binomial and multinomial models of job separations using Panel Study of Income Dynamics (PSID) data for the years 1976‐93. The results are consistent with a decline over time in the incentives to maintain existing employment relationships for male workers and for skilled white‐collar women.

We study the effect of new technologies (NT) on wages and employment using a unique panel that matches data on individuals and on their firms. As in the United States, we show that computer users are better paid than nonusers (15%–20% more). But these workers were already better compensated before the introduction of the NTs. Total returns to computer use amount to 2%. Measurement errors do not affect our estimates. Furthermore, computer users are protected from job losses as long as bad business conditions do not last too long. This result holds even after controlling for possible selection biases.

We estimate interindustry wage differentials using new French longitudinal data that allow a tracking of workers and their firms over time. We find that, when measured on a cross‐sectional basis, they primarily reflect the interindustry variations in unmeasured labor quality. However, interindustry wage differentials are only a minor component of interfirm wage differentials. The average differential in wages paid to the same workers by different firms is about 20%–30%. In a given industry, wage policies are more favorable to workers in large, capital‐intensive firms.