Comparing Proposals for Social Security Reform
Wednesday, September 01, 1999
by Liqun Liu and Andrew J. Rettenmaier
Table of Contents
Lessons for Reform
We have completed an extensive analysis of four major reform proposals. The analysis has been sufficiently broad to allow us to make some generalizations about an entire class of proposals to reform Social Security by moving from pay-as-you-go finance to a partially or fully funded system.
"The solution requires most of the projected budget surplus -- investing those funds in income-earning assets."
Benefits Can Be Secured without New Taxes. In principle, it is possible to secure the benefits promised by the current system without any increase in tax rates or in government borrowing. However, accomplishing this goal requires moving to a funded system (either in part or in whole) under which each generation saves for its own retirement benefits. This solution, in turn, requires that we dedicate most of the projected budget surpluses - investing those funds in income-earning assets.
"The tax burden needed to guarantee benefits is less than the current payroll tax."
Under three plans considered here (Gramm-Domenici, Archer-Shaw and the NCPA-PERC plan), retirees receive at least as much as or more than Social Security benefits promised by law by the year 2075. [See Figure XII.] Yet in all three cases, the tax burden needed to guarantee those benefits to all workers is less than the current payroll tax, and in the case of Gramm-Domenici and NCPA-PERC, it is significantly less. [See Figure XIII.]
Figure XIII also reveals two other important features designed to secure future benefits. First, note that:
- Whereas the bipartisan plan ultimately reduces benefits by 22 percent, the Gramm-Domenici plan increases them by 20 percent.
- Yet the burden for taxpayers is very similar under the two plans.
This comparison shows that investing in income-earning assets and taking advantage of compounding is far more powerful than reducing benefits.
Note also that the longer funds stay in the PRA accounts, the more they gain through compounding. The NCPA-PERC plan is able to accomplish much more than the other plans over the same time period because it skews contributions toward younger workers who must leave their funds invested for more years before they qualify for retirement. For example:
- Both the Gramm-Domenici plan and the NCPA-PERC plan commit about the same number of dollars to PRA accounts for the first 40 years.
- Yet in the year 2075, the burden for taxpayers under the NCPA-PERC plan is little more than one-third of the burden under the Gramm-Domenici plan.40
"Skewing account deposits toward young workers has a huge effect after only one generation moves through the workforce."
There Are No Long-Term Transition Costs. We define transition costs as any increase in government outlays occasioned by a reform proposal minus any increase in government revenues occasioned by the reform. As Figure XIV shows, under all four reform plans, these costs are positive and large in the early years. In the later years, as PRA accumulations stimulate a larger economy, the transition costs become negative and more than offset the initial positive costs.
We conclude that moving from a pay-as-you-go system to a funded system imposes no long-term costs. Put a different way, the government - and the people - profit from the transition. Figure XIV reaffirms the conclusions of the previous section, although from a different perspective.
- The Gramm-Domenici plan is less costly (society "profits" more) than the bipartisan plan even though Gramm-Domenici ultimately promises a 20 percent increase in benefits, while the bipartisan plan promises a 20 percent cut.
- At the same time, the Gramm-Domenici plan is twice as costly (society "profits" 50 percent less) as the NCPA-PERC plan.
- These results are another reflection of the fact that compounding is a far more powerful tool than cutting benefits and that skewing PRA deposits toward young workers has a huge effect after just one generation moves through the workforce.
Benefits Can Be Fully Funded. Social Security has a large unfunded liability because, from the beginning, it has operated on a pay-as-you-go basis. Therefore, any reform plan that retains pay-as-you-go financing - even if only on a partial basis - risks carrying the current problems into the future. For that reason, it is interesting to ask how easy or difficult it is to move to a fully funded system.
Figure XV shows the results of the four plans as they are currently structured:
- The NCPA-PERC plan moves to full funding for each group of new retirees by mid-century, and the Gramm-Domenici plan achieves full funding by the end of the century.
- The bipartisan plan achieves slightly more than two-thirds funding by mid-century, but retains that feature indefinitely.
- The Archer-Shaw plan achieves funding for about half of promised benefits for new retirees by mid-century and also retains that structure indefinitely.
- Both the bipartisan and the Archer-Shaw plans could easily achieve full funding in the next century by using extra revenues generated by a larger capital stock to fund larger deposits to PRA accounts.