Reassessing the Labor Supply Curve
Prasch, Robert E., Journal of Economic Issues
It is no exaggeration to state that the supply and demand model of price determination is the most widely accepted theory in contemporary economics. In the professional and textbook literature, this model is used to determine the price of every kind of commodity from theater tickets to corn, oil, votes and wages. However, problems can arise when a model as abstract as the theory of supply and demand is too readily applied to situations in which it may be inappropriate. Indeed, Institutional economists have long maintained that the application of the supply and demand theory to the labor market is often an important instance of such an error [Galbraith 1997; Lester 1947, 1941; Power 1999; Prasch 2000, 1999, 1998b].
This paper will argue that labor is qualitatively different from other commodities in at least one crucial attribute. Specifically, when the price of a typical commodity falls, the conventional theory of supply maintains that suppliers will substitute out of the business of providing that particular item, and devote their productive capacity to the provision of some other item. In the standard models, this substitution is thought to occur with zero transactions costs. Now this idea is not necessarily wrong. For example, it is reasonable to suppose that a low price for minivans will induce automobile manufacturers to supply more trucks or sedans. But is it plausible to apply this line of reasoning to the aggregate labor market? Or the market for unskilled labor?
The problem is that most people depend upon the sale of their labor for the bulk of their income and, indeed, for their livelihood. In light of this fact, can we plausibly conclude that a low wage will induce more than a minority of people to substitute into leisure and forgo any effort to earn a living? After all, not many people live off the bounty of the land, depend upon their savings, or count on the beneficence of family and friends to meet their day-to-day needs. While this is the implication behind the conventional depiction of the labor supply curve, it is not a plausible representation of the situation faced by the overwhelming majority of people in the American labor market, to say nothing of conditions in Third World labor markets.
This paper will argue that with a more plausible specification of the relationship between income and the value of leisure, we can develop a more accurate model of the labor market without necessarily losing the primary merit of the received model--its analytic and pedagogical simplicity. Moreover, with an improved understanding of labor market dynamics we can see that progressive era labor market legislation, designed to mitigate the effects of bargaining power inequality in low wage labor markets, was in fact based upon sensible economic foundations [Prasch 2000, 1999, 1998b]. Rather than representing a "trade-off" between "equity and efficiency," minimum wage and maximum hours legislation may actually enhance, rather than diminish, the performance of the labor market.
The Neoclassical Labor Supply Schedule
The upward sloping factor supply curve holds a prominent place in textbook discussions of the labor market and the distribution of income. As one among several productive "factors" the conventional, or textbook, labor supply schedule shows that the quantity of labor supplied is a direct function of the inflation-adjusted, or real, wage. Moreover, this schedule is constructed from an aggregation of the labor-leisure choices made by all actual and potential workforce participants in the community [cf. Taylor 1998, 326-329; Katz and Rosen 1991, 133-150].
Despite this apparent consensus at the textbook level, there is some dissatisfaction with the conclusions that are typically drawn from the conventional model. An important problem is posed by our actual experience with labor markets. Over the years, numerous empirical studies have indicated that the labor supply schedule is either vertical or sloping downward to the right [Bluestone and Rose 1998, 28-30; Dessing, 1999; Gordon 1996, chap. …