Consumer Expenditure Responses to Income Redistribution Programs
利用1960-61年消费者支出调查数据，估计22类消费品的收入弹性，预测两种负所得税计划下家庭支出变化，发现约三分之一新增可支配收入用于基本必需品，且不同融资方式对消费刺激效果差异显著。
REFORMING welfare programs targeted on low income households will change the disposable income of families and individuals affected by the reform. Financing the new program with an increase in personal taxes redistributes income, on net, from high and middle income households to low income ones. This empirical study investigates expenditure responses elicited by these changes in disposable, income. Regression equations for a complete set of 22 individual expenditure items are estimated with the 1960-61 Bureau of Labor Statistics Survey of Consumer Expenditures (SCE) data, described in U.S. Department of Labor (1971). Income coefficients from these equations are used to predict household expenditure responses to microsimulated disposable income changes in the SCE data from two separate surtax-financed negative income tax (NIT) programs. The regression analysis tracks nonlinear in income Engel curves for each of the commodities. The two simulated NIT programs redistribute $3.04 billion and $33.4 billion in disposable income from surtaxpayers to NIT recipients. About one third of this new disposable income is spent on basic necessities (food, clothing, and shelter) by program beneficiaries. A deficitfinanced NIT program provides relatively strong stimulation to only clothing markets among the necessity items, but a surtax-financed program injects a relatively large stimulus into both food and shelter markets. In general, markets for nondurables are relatively stimulated while durable good markets either grow less than normal or contract under a modest guarantee surtaxfinanced NIT program. From a macroeconomic perspective, a deficit-financed NIT program increases total consumer expenditure by between 75% and 100lo of program net costs, while a surtax-financed program generally stimulates the economy by raising aggregate consumption by between 10% and 50% of the net program cost. The paper is organized as follows. Section II presents the expenditure regression models and reports regression statistics, income elasticities, and selected income spending propensities from the full set of expenditure equations estimated under both current and permanent income models of consumer behavior. Section III discusses the NIT simulations, and section IV reports and analyzes the expenditure response predictions to the simulated NIT programs. Conclusions and suggestions for future research appear in section V.