POVERTY AND ECONOMIC GROWTH: HAS TRICKLE DOWN PETERD OUT?

The importance of the "trickle-down" effect of economic growth is a key issue in policy debate regarding tax reform, government expendi- tures, and efficiency-equity tradeoffs. For this reason, a recent study by Thornton, Agnello, and Link (1978) examining the relationship between poverty and economic growth from 1947 to 1974 deserves careful review. The authors (hereafter TAL) show that the trickle-down effect of economic growth on poverty, defined using a fixed threshold, is signifi- cantly less in the post 1963 period than in the 1947-1963 period. They conclude that trickle down has "petered out" and predict it will vanish in the future (pp. 386, 394). Moreover, they argue that there has been no statistically significant trickle-down effect in the post 1963 period for any demographic group when a definition of poverty using an increasing threshold is used. TAL draw what they regard as an obvious policy implication: "Since primary reliance on future economic growth to reduce poverty will be largely unsuccessful, expanded programs directed specifically at poor families will be required if poverty is to be elimi- nated" (p. 394). The purpose of this note is to show that TAL's results are sensitive to the particular specification of their model and that their specification may be inappropriate. In particular, the finding that trickle down has petered out result directly from their use of a questionable dependent variable. Reestimation of the model (with their data) using an alternative dependent variable indicates that there exists no significant evidence that the relationship between poverty and economic growth has changed in the post 1963 period. In addition to challenging the validity of TAL's empirical results, this note questions the interpretation and policy im- plications drawn from their analysis.