Labor Supply and Tax Rates: Reply
回应了关于税率变化与劳动力供给关系的评论，强调传统劳动-休闲分析因忽略政府支出对福利的影响而失效，并讨论了公共品和转移支付的不同作用。
The comments of Firouz Gahvari and of Cecil Bohanon and T. Norman Van Cott provide useful extensions of our earlier analysis. Nonetheless, our central point remains intact: the traditional labor-leisure analysis is invalid because it ignores the effects of changes in government spending on individual welfare. Gahvari points out that, in the case of public goods, the linkage between changes in tax rates and labor supply is more complex than we implied. Individuals, unable in the large number case to transform leisure into public goods, will be affected differently when the government provides a public good rather than an transfer. In Gahvari's world, where government goods are irrelevant to all private decisions (complete separability in the utility functions and no ability to purchase public goods privately), the quantity of government goods can be safely ignored in the analysis of private decisions. The decision proceeds as it would if tax revenues were totally wasted, even though the citizens' total utility is assumed constant when tax revenues change, with changes in government goods exactly offsetting the utility impacts of the change in private goods. However, we think it is misleading to label the ambiguously signed element beyond the substitution as an effect. How can there be an income effect when total utility remains constant? As the quotes cited by Gahvari from our initial paper (p. 447) illustrate, this is not the envisioned by the traditional work-leisure analysis, which refers to a change in of (utility) and reflects the notion that a tax cut will encourage individuals to work less (consume more leisure) by giving them a higher standard of living through more after-tax pay.' Implicitly, this view ignores the negative impact on living standards associated with the reduction in the supply of government provided goods (or transfers). Gahvari recognizes that when government goods replace private goods such as public education, medical services, food stamps, or cash (and ignoring any cross elasticities), our original analysis stands and there is only the substitution effect. Bohanon and Van Cott make another refinement, pointing out some secondary effects of government's tax-transfer activity. While changes in tax rates and in government-provided goods will influence the individual's budget constraint, or ability to trade off among goods, a complete analysis must also account for the fact that any such shift will move the individual into a new region of his indifference surface. His willingness to trade off among the goods may well change. The individual's view of the substitutability or complementarity among government-provided goods, private goods, and leisure becomes relevant. If the government good is strongly enough complementary to leisure and/or substitutable for private goods, the standard substitution could indeed be overcome. Returning to our original paper, we reiterate its central point. The income effect component of traditional work-leisure analysis for an individual ignores the individual's utility derived (foregone) from increased (decreased) government spending accompanying changes in revenues. To treat that indi-