The Cost of Reducing Economic

The publication of Charles Murray’s (1984) book Losing Ground has rekindled interest in the role ofgovernment in promoting economic equality using redistributive tax and expenditure programs. Murray concludes that spending on poverty programs may well have increased the number of poor people and that eliminating many of these programs might be a better antipoverty strategy. These conclusions reflect the view that the adverse incentive effects of social programs, such as the discouragement of work effort and saving by the poor, more than offset the positive contributions the programs make by providing income transfers. Many scholars, however, have criticized Murray, arguing that much evidence suggests the disincentive effects of social programs are not large enough to offset the income-enhancing effects of the policies (see, for example, Danziger and Gottschalk 1985). We tend to agree that Murray has overstated his case, but we also believe thatthe debate has obscured an important aspectof the issue, namely, the importance of distinguishing between the overall effects of social programs and the effects of small changes in the scale of these programs. Much of the recent debate has addressed the queslion: if we eliminate poverty programs, will poverty rise or fall? An equally, if not more, important question is: if we reduce spending on poverty programs, will poverty rise or fall? These questions may not have the same answers. The essential distinction is between the overall, or average, effects and the effects of a small change, or marginal effects. As economists have longbeen aware, marginal and average effects can differ greatly,