This paper reexamines an old question: what is work and why do we do it? A straightforward ’economic ’ answer is that work is any time activity for which we receive a positive wage, but this begs the question which activities command a positive wage. Now the answer depends on market level considerations such

The paper considers an industry in which individual output follows a stochastic growth process with a cumulative effect. It analyzes the roles of labor mobility and capital market conditions in the determination of wage contracts. Positive costs of mobility are shown to be necessary for the provision of wage and employment insurance when workers have no access to the capital market. When insurance is provided, wages grow less than average productivity. If the capital market is perfect, wage insurance will be provided even in the absence of costs of mobility. In this case, wages grow faster than average productivity.

The consequences of introducing risk-averse workers with a known, finite retirement age into a simple job-matching framework are explored. The job-matching model includes two jobs. In the "primary" job, workers and firms learn about the worker's productivity by observing his output over time. In the "secondary" job, productivity of any worker is known with certainty at the outset. Workers start off in the primary job but, based on their observed performance, may choose to switch to the secondary job. The model is developed in two stages. First, the finite retirement age is introduced while the usual assumption of risk-neutral workers is retained. In this model it is shown that all workers who achieve at least a certain cumulative output record, denoted r*, by a certain age, denoted T, will thereafter remain in the primary job until retirement. Those who do not achieve r* by age T switch to the secondary job at or before age T and remain there. Second, risk aversion is introduced along with finite retirement age. Here the turnover aspects of the equilibrium are similar to the risk-neutral case, but firms provide workers with wage insurance. This takes the form of wages that are downward rigid. As a result, it may be optimal ex ante for workers to delegate the job choice decision to the employer, and, ex post, not all separations will be voluntary for the work r.

Between-family differences in expenditures and output reflect the effect of simultaneous increases in children's ability on the willingness of parents to transfer resources to them. Within-family differences also reflect the attitudes of parents toward disparity among children. In this paper we characterize the conditions on parents' preferences that determine whether between-family differences exceed within-family differences. For additive utility, within-family differences in expenditures always exceed between-family differences. This may also be true for the max-min utility function if an increase in ability reduces the marginal utility of income. Within-family differences in output (utility or income) can also exceed between-family differences. In this case, the implication for income distribution is that equality is enhanced by a higher correlation of ability between brothers.

Analysis of longitudinal earning data indicates higher earnings growth for scientists of the same experience but more recent vintage. Theoretical justification for such a relation is suggested, and the implied biases in cross-section data are noted. Because of a basic identification problem, an alternative interpretation, time-experience interaction, is also considered. The general conclusion is that earning growth is not uniform or neutral. There is no simple mechanical method by which lifetime profiles can be inferred from single cross-section data.

The joint determination of work and investment in human capital over the life cycle is analyzed. At low rates of impatience investment is decreasing throughout life, as in simpler models which assume hours of work to be fixed. The demand for leisure over the life cycle is "U shaped." Wages rise to a single peak which occurs after the peak in hours of work. Distinctly different patterns arise when the rate of impatience is high. Such individuals may prefer an increasing hours of work profile, and schooling need not be concentrated at the beginning of life. Conditions are provided to determine a critical level of time preference which is sufficient to induce a "normal" life-cycle pattern for investment and work.