The Effects of Irregular Real Wage Growth and Varying Rates of Price Inflation on Social Security Lifetime Wealth Transfers Across Birth Cohorts

Abstract

The paper estimates the effect of irregular real wage growth and varying rates of price inflation on the net value of the Social Security retirement program (Old Age and Survivors Insurance, or OASI) for the 33 birth cohorts born between 1917 and 1949. The net value of Social Security employed is the “lifetime net benefit rate”, which is the expected present value of lifetime benefits less the expected present value of lifetime taxes expressed as a percent of the expected present value of lifetime earnings. In order to isolate the effect of real wage growth and price inflation on the lifetime net benefit rate (LNBR), OASI taxes and retirement benefits are simulated for a representative person who pays a 10.6 percent OASI tax rate at all ages and has a normal retirement age and mortality probabilities projected for the 1949 birth cohort, but who experiences real earnings levels and rates of price inflation that are cohort-specific. The paper finds that irregular real wage growth and irregular rates of price inflation cause the LNBR for the 33 birth cohorts to vary between -4.31 percent and -3.27 percent. Variation caused by irregular real wage growth is 40 percent larger than total variation, and variation caused by price inflation is about equal to total variation, which is possible because the correlation between the two components is -0.7. The implication is that if the joint distribution of price inflation and real wage growth in the future is like that in the recent past, then somehow eliminating only the irregular real wage growth component would not be helpful, and somehow eliminating only the price inflation component would be counterproductive. Reducing the sensitivity of the LNBR to the macro economy would require reducing both sources of variation in this case. The effect variation in real wage growth has on the LNBR would have insurance value if cohorts whose lifetime earnings are low relative to trend tend to have higher LNBRs, and cohorts whose lifetime earnings are high relative to trend tend to have low LNBRs. The paper rejects this possibility. No correlation is found between the effect of real wage growth variation on LNBRs and its effect on the present values of lifetime earnings. Because wage indexing in the current Social Security benefit formula appears to have no insurance value, an alternative benefit formula based on a notional account balance is explored. This alternative formula has the virtue that the LNBR does not vary with real wage growth while benefits are as generous and progressive as under the current benefit formula. The paper also considers reforms to reduce the sensitivity of real benefits to price inflation when real wage growth is held constant. Surprisingly, straightforward reforms to correct the anomalies in the calculation of initial benefits do not significantly reduce the estimated variability of real benefit levels across birth cohorts when real wage growth is held constant. It is demonstrated that this finding is due to the fact that the COLA depends on lagged price changes. Given that the COLA must be based on lagged price changes, it is concluded that improving the COLA would require making annual payments making up for inaccuracies in the COLA.