Labor Supply and Tax Rates: Comment

In a recent paper in this Review (1983) James Gwartney and Richard Stroup (G-S) take issue with what they assert is the "widespread view" (p. 447) that because changes in tax rates have a theoretically indeterminate effect on individual labor supply, the impact of changing tax rates is also indeterminate at the aggregate level. Gwartney and Stroup suggest, for example, that if the initial output of public goods is optimal and the government cuts taxes and expenditures by equal amounts, the forgone public goods would be valued as highly as the private goods that individuals purchase with their increase in disposable income. Hence, aggregate real income would not rise, but remain at its pre-tax-cut level. Gwartney and Stroup state that under these circumstances, a tax cut's impact on aggregate labor supply would not be indeterminate, but labor inducing, because in the aggregate there is no income effect, only a positive substitution effect.1 Gwartney and Stroup illustrate this point by examining the aggregate labor supply effects of an increase in transfer payments financed by an exactly offsetting increase in income taxes. They argue that if the transfer payments are income conditioned, this policy would decrease net wage rates for transfer recipients, as well as for taxpayers, thereby producing work-reducing substitution effects for all members of society. On the other hand, the aggregate change in disposable income in the economy would be zero since, by design, the expansion in transfers is exactly offset by the increase in taxes. They conclude that, unless recipients and taxpayers respond differently to a given change in disposable income, " the income effect of the tax-transfer program will leave both the aggregate supply of labor and consumption of leisure unchanged. [But] Of course, the substitution effect will still be present. Unambiguously, it will induce individuals to work less, although how much less is strictly an empirical issue" (p. 450). In this comment, we demonstrate that the G-S tax-transfer illustration and the policy implications drawn from it are highly misleading. We first show that the G-S conclusions about the relationship between tax rates and labor supply are based upon strong assumptions about the degree of homogeneity in the population and that without these assumptions, their conclusions are not guaranteed to hold. We then use microsimulation techniques to challenge the impression given by the G-S paper that more redistribution must necessarily imply a reduction in the labor supply and output of the economy.