Income, Consumption, and Wage Taxation in a Life-Cycle Model: Separating Efficiency from Redistribution
提出一种简化方法，分离稳态福利变化中的效率与再分配成分，并应用于代际交叠模型，发现从所得税转向消费税或工资税的效率收益可能并不显著。
The desirability of switching from an income tax base to a wage or consumption tax base hinges critically on the potential efficiency gains from doing so. The overlapping-generations life-cycle model developed by Lawrence Summers (1981) cannot easily be used to answer this question. Because individuals of different ages alive at the time of the tax substitution differ in their shares of income, consumption, wages, and savings propensities, steady-state welfare changes include redistribution of taxes and factor incomes. The effects of redistribution on the calculation can be controlled for by tracing the transition. In another paper, Summers (1980) used myopic expectations (individuals assume that current prices will prevail) and solved the economy's equilibrium for 100 years. Gains and losses were discounted and summed. Alan Auerbach, Laurence Kotlikoff, and Jonathan Skinner (1983), hereafter referred to as AKS, employed a perfect-foresight model. They used lumpsum payments to keep the utility of those alive at the time of the tax change fixed and allowed equal equivalent variations to all future generations.1 A perfect-foresight model requires simultaneous solution of many years of equilibria (150 years in the AKS model). Although these techniques eliminate distributional elements, such models are so costly to construct, modify, and use that the study of relative efficiency of alternative tax regimes is now greatly hampered by computational complexity. This study has two major contributions. First, it introduces a much simpler method of separating the efficiency and redistributional components of steady-state gains. Rather than tracing the transition, it uses a system of transfers which ensures that changes in welfare in the steady state do not arise from redistribution of income to or from transitional generations. While this approach does not allow calculation of the transition, its application to the AKS model suggests that very little information about the magnitude of efficiency effects is lost by forgoing the tracing of the transition. For estimating the relative efficiency of income, consumption, and wage taxation, or the extent to which assumptions drive the results, this approach greatly simplifies the researcher's task.2 This simplification becomes increasingly valuable as more detail is introduced into models, which could involve thousands of equations using an approach that traces the transition. Secondly, my calculations suggest that switching tax bases may not be worthwhile. This result differs from the results of Summers (1981) and Auerbach and Kotlikoff (1987), which strongly suggest that the consumption tax is always superior on efficiency grounds. Using a 0.25 intertemporal substitution elasticity and a Cobb-Douglas production function, the steady-state wel*Congressional Research Service, Library of Congress, Washington, DC 20540. I thank Christophe Chamley, Don Fullerton, Yolanda Henderson, Don Kiefer, Larry Kotlikoff, Jonathan Skinner, Tom Woodward, Dennis Zimmerman, and three anonymous referees for helpful comments and discussions. The views in this paper do not necessarily represent those of the Congressional Research Service or the Library of Congress. 1The AKS model is confined to one good. Charles Ballard (1984, 1989), Lawrence Goulder (1985), and Ballard and Goulder (1987) develop multisector models. 2The suggestion of a method for separating efficiency from redistribution is not intended to imply that there is no value to tracing the transition. Tracing the transition can provide information about intergenerational redistribution in an uncompensated change and insights into the realism of the model.