Privatizing U.S. Social Security: Some Possible Effects on Intergenerational Equity and the Economy

In the early 1980s, when Chile first privatized its social security system, few observers could have predicted that its action would precipitate the worldwide interest in privatization that prevails today. In large part, the current interest grows out of a belief that privatizing social security can enhance national economic growth and lead to a higher standard of living for workers and retirees. Indeed, many analysts have attributed Chile’s current economic success to social security privatization. Whether that is actually the case will no doubt be a subject for study by economic historians for years to come. In the meantime, a question that many researchers and others are asking is “Will privatization yield the expected benefits in the United States?” Among the researchers who have studied this question are Seidman (1986) and Feldstein (1995), who use partial equilibrium frameworks, and Arrau (1990), Arrau and Schmidt-Hebel (1993), Raffelhuschen (1993), and Huang, Imrohoroglu, and Sargent (1997), who use general equilibrium frameworks. This article uses simulations in a stylized model—the Auerbach-Kotlikoff Dynamic Life Cycle model—to explore some possible economic outcomes of privatizing Social Security in the United States: first, the potential impact on the macro economy; second, how different generations may be affected (in particular, whether privatization will enhance or undermine intragenerational equity); and, third, some efficiency gains that might arise from privatization. The parameters in this model (hereinafter referred to as the A-K model) closely parallel those of U.S. economic conditions and fiscal policies. In addition, the model has more detailed equilibrium frameworks than those in the other models cited above. Previously used in a number of simulation studies (see Kotlikoff, 1996 and 1998, and Kotlikoff et al., 1997 and 1998), the A-K model has recently been enhanced to include intragenerational heterogeneity, kinked budget constraints, and a more realistic formulation of income taxation. All the variations of the A-K model comprise three basic elements: (1) workers are required to contribute to private retirement accounts, (2) social security benefits are awarded to those who are workers and retirees during the transition in accordance with the benefits they have accrued before the transition, and (3) these transitional benefits are paid for by means of a wage tax, an income tax, a consumption tax, or a combination of income taxes and deficit financing. Simulations in the A-K model indicate that the long-run effects on the macro economy differ substantially from the short-run effects, depending on how the transition is financed. Because consumption-tax financing produces much more rapid economic gains than either wageor income-tax financing (see Kotlikoff et al., 1997 and 1998), I discuss only the consumption-tax simulation in this article.