Social Security Annuities and Transfers

The division of social security (OASI) benefits into an annuity portion and a transfer portion has been well documented. I have discussed this issue extensively in previous work (1987b, 1988, 1990, and forthcoming), as did Burkhauser and Warlick (1981) previously. My methodology is quite similar to theirs. The annuity portion is defined as the benefit level the worker would receive on the basis of his(her) contributions into the social security system (OASI) if the system were actuarially fair. The calculation is based on the worker’s estimated earnings history and actual social security tax rates. The transfer portion is the difference between the actual social security benefit received and the actuarially fair annuity equivalent. As we shall see below, it has been uniformly positive for workers who have retired on or before 1983.

Burkhauser and Warlick examined the relative proportions of annuity versus transfer benefits by income class and age group. However, they did not conduct an extensive examination of the overall distributional implications of who social security transfer portion. Nor did they consider the tax implications of treating social security transfers as taxable income. These are the principal subjects of the current paper. With regard to the distributional implications of the social security system, I will examine three sets of issue. First, I will consider what the relative magnitudes have been of the annuity and transfer portions of social security income. Since I have data for three years, a related issue is whether the relative proportions have changed over time. Second, I will consider how the social security transfer portion has affected the distribution of income among elderly households. Has the transfer component been neutral or has it tended to redistribute income toward lower income elderly households? Third, the same issue can be addressed with regard to household wealth, in which social security benefit flows are transformed (capitalized) into wealth equivalents.

From a policy point of view, the more interesting issue is how do the total taxes of the elderly change with the removal of the exclusion of social security transfer income — that is, when social security transfer income is treated as taxable income. There are three questions of interest. First, how does the change in tax treatment affect the post-tax distribution of income. Second, which groups of elderly are most affected by the change in tax treatment. Third, what is the total challge in the magnitude of tax revenues.

As a final point of policy interest, I will also consider whether the extra revenues generated by the new tax treatment of social security income can serve as a "social security capital fund" to reduce the growing wealth gap among age groups in the U.S. As will become apparent in the analysis, the social security system has been quite generous to today’s elderly, providing them with benefits far in excess of their contributions into the system. Moreover, young families have fared rather poorly over the last several decades in regard to their income and wealth accumulation. I will propose a policy vehicle below, called a "social security capital fund", which can serve as an additional source of capital for today’s young workers. The source of the funding can potentially come from the extra tax revenues from elderly households. It is thus also of interest to analyze whether the additional tax revenues are large or small relative to the wealth holdings of young households and whether such a fund can make a significant difference in the well-being of younger families.

Working Paper No. 58(1.88 MB)