Projected Trust Fund Build-up: Social Security Issues
by Alicia H. Munnell
Senior Vice President and Director of Research
Federal Reserve Bank of Bosto
This portion of the discussion has a much more limited focus than the morning session - that is, we are supposed to concentrate on the implications of the projected trust fund buildup for the Social Security system rather than for fiscal policy and the economy generally. The major question is whether the future of the Social Security system hinges on how we resolve the controversy about building up reserves. The answer is that Social Security will be just fine no matter which course of action we take; the possibility of saving through the Social Security trust funds is an opportunity, not a necessity. First, consider the history of the projected buildup.
I. A Little History
Folklore not withstanding, the surpluses originated with the 1977 amendments. As you remember, in order to strengthen the financing of the program, this legislation extensively revised both the benefit and the revenue provisions. Most of the additional revenue in the 1977 legislation came from increasing a planned 2011 rate hike and making it effective in 1990. This meant that the rate hike, which was originally scheduled to coincide with rising benefit costs as the baby boom generation retired, would now take effect while costs were declining as a percent of payrolls. The combination of increased revenues and lower costs was projected to produce surpluses in the program through the year 2010. If you go back and look at the 1978 Trustee's Report, you will find that by the year 2005 trust fund assets were scheduled to equal 2.5 times annual outlays. Because the 1977 legislation established actuarial balance for only 50 years, the trust fund reserves were then depleted in the early 2020s.
The world, as we all know, did not turn out as expected and further remedial legislation was needed. The 1983 amendments were designed both to relieve a short-term financing crisis and to remove the long-term deficit left by the 1977 amendments. In the process, the legislation greatly increased the size and extended the duration of the surpluses. This was accomplished primarily by increasing revenues somewhat through the taxation of benefits and reducing costs significantly through the extension of the retirement age. As a result of the 1983 legislation, the intermediate cost estimates now indicate that fund assets will equal 4 times annual outlays in 2005, rather than 2.5 times as projected under the 1977 legislation. Assets will ultimately rise to a peak of 5 times annual outlays in 2018, and then decline in order to cover annual deficits in the years between 2018 and 2051.
The main point of this history is that the projected buildup was a by-product of moving a scheduled tax rate from 2011 to 1990 in the 1977 legislation and then decreasing significantly long-run costs in the 1983 legislation. No one involved in the process ever considered the advantages or disadvantages of trying to increase our national saving rate through the Social Security program. If they had, I trust they would never have devised the plan of building up reserves and then drawing them back down. This pattern implies that we are facing a temporary blip in the cost of Social Security as the baby boom generation passes through - a rat being digested by a snake. In other words, the implication is that the number of beneficiaries per hundred workers will go from its current level of 30 up to 54 and then back down to around 30. This is not the case; rather, a permanent decline in the birth rate will cause a permanent increase in the ratio of beneficiaries to workers. Hence, if the decision were made to pre-fund some Social Security benefits, it makes far more sense to build up the fund and maintain it as a permanent source of partial funding rather than drawing it down.
In short, the history of these surpluses indicates that we really have a clean slate and can decide to do anything we like: we can use Social Security reserves and use them to fund other government spending, or we can return to pay-as-you-go. Regardless of the option selected Social Security will be fine.
II. Do We Want To Increase National Saving?
Despite the apparent ground swell of support for more saving, it is legitimate to question whether we as a nation are saving too little. Honest economists will admit that you do not get the answer to this question by comparing the U.S. saving rate to that of Japan. Nor do you get the right answer by comparing our current saving rate to that experienced before 1980. Nor do you get an answer by talking about how taxes drive a wedge between the return on saving to society and that actually received by the individual investor. It is very difficult to prove theoretically that our current rate of saving is less than optimum. As a result, the argument for more saving usually comes down to a value judgment that we are not putting aside "enough" for our children and their children. Specifically, people argue that if the baby boom generation does not save in advance for its retirement, it will place an unsupportable burden on the following generation.
The other side of the argument, of course, is that because of its large size, the baby boom generation has had a difficult time. These people attended overcrowded schools, faced severe competition in finding jobs, and experienced slow advancement once on the job. The mere size of this group has contributed to the virtual freezing of real wage growth for a decade and has made it difficult to find affordable housing. The answer to the question whether this generation should be the one to contribute towards its own retirement in addition to financing benefits for current retirees may not be as obvious as it first appears.
III. How Will More Saving Affect Social Security?
On the most superficial level, whether or not the buildup in the trust funds actually produces increased national saving will have no effect on Social Security. Under either scenario, replacement rates for a worker with a history of average earnings will be roughly 41 percent and the combined employee-employer OASDI tax rate (under the IIB assumptions) could probably stay around its current level of 12.4 percent over the entire projection period.
So what is all the hoopla about? Why does it matter whether or not the buildup of reserves actually produces increased national saving? It matters because how burdensome Social Security feels in the future will depend on how rich people are. If the trust fund assets are used to increase national saving and investment, then people will have more money with which to pay retirement benefits for the baby boomers and subsequent generations. If people in the future have lots of income to start with, they will still have lots more than we do now, even after they pay the higher costs of retirement benefits.
It is important to point out that if the IIB assumptions prove correct, the difference between how well off people would be with and without the additional saving is fairly small. If I remember the Aaron-Bosworth-Burtless numbers correctly, our grandchildren will have 235 percent of what we have today without the additional saving, 242 percent with it. The key point may be, however, that without the additional saving, we may not end up with the kind of productivity growth embodied in the IIB assumptions. At lower levels of productivity growth, the incremental contribution of the additional saving becomes more important.
What happens if trust fund surpluses do not produce any new saving? That is, if they are simply offset by deficits in the rest of the federal budget. In this case, the surpluses will have contributed nothing to overall saving and capital accumulation and taxpayers will be no richer than they would have been otherwise. The full burden of supporting the beneficiaries will fall on the taxpayers in the second half of the period - just as if the system had been financed on a pay-as-you-go basis all along. The only effect of accumulating Social Security surpluses would have been to alter the composition of federal revenues over time. General government expenditures during the first half of the period would be financed by the relatively regressive payroll tax rather than the more progressive income tax, and future benefit payments would be financed by general revenues.
IV. How Do We Get More Income In 2030?
The conventional argument, and the one just alluded to, is that whether or not government saving actually occurs will depend on how the assets in the Social Security trust funds are used. If the reserves are used to finance current consumption - for example, to pay for current outlays in the rest of the budget - no real saving will occur. On the other hand, if the government alters its spending and taxing patterns to produce surpluses at the federal level - not just in the Social Security trust funds - the nation will enjoy higher saving and investment.
That argument is basically correct. The only difficulty is that it characterizes all government spending as consumption. Clearly, the building of roads, bridges and other types of physical infrastructure is just as much an investment as the construction of any factory in the private sector. Equally important, however, is investment in human beings, because future output will depend upon having a healthy and educated work force. In other words, money spent on nutrition programs for pregnant women, on health care for poor children, and on Head Start programs will contribute just as much as physical investment to ensuring that we produce lots of income in the future.
The implication is that if we were doing a careful and serious job of assessing whether trust fund surpluses were adding to national saving, we would have to do more than look at whether the non-Social Security portion of the budget were in deficit or surplus. If the Social Security trust fund surpluses were used to finance new health or educational programs, they would be adding to future income just as surely as if they had been invested in General Motors stock.
V. And What About Investment Policy for the Trust Funds?
And what about investing in General Motors stock, or student loans, or more realistically, long-term government securities, FNMAs, GNMAs, state and local bonds? Personally, I think this is a topic we could safely postpone for quite a long time and, in fact, may never have to address.
In my view, the assets in the trust funds cannot really be considered available for alternative investments until an adequate contingency reserve has been accumulated. As you know, the financial health of a pay-as-you-go system is very sensitive to the performance of the economy. Hence, it is important to have an adequate reserve in case an economic downturn causes a drop in revenues or rapid inflation causes benefits to increase faster than payroll tax receipts. My calculations indicate that a reserve equal to 150 percent of annual outlays would be required to weather the kind of back-to-back recessions that we experienced during the 1970s. (I know the 150 percent figure is correct since Bob Ball thinks it's too low and Bob Myers thinks it's too high.) The most recent projections by the Social Security Administration indicate that this level of contingency fund will be reached in 1995. Until then, it would not make sense to invest this contingency fund in anything other than the current highly liquid government debt.
After an adequate contingency reserve is established and the money continues rolling in, we can have extended discussions about exciting alternative investments. There really is no problem - realistically plenty of financial instruments should be available. The only issue is whether higher returns should be earned in the trust funds or in the private sector.
Let me conclude. Circumstances have created a situation where the United States has the unique opportunity to augment national saving by accumulating assets in the Social Security trust funds. This is probably the most desirable course of action. Saving in advance would make it easier to support the additional burden created by the retirement of the baby boom generation. It would also increase political support for Social Security: when critics argue that private pensions are superior because they are funded, supporters of Social Security can point to the billions of dollars in the trust funds.
What many fear may happen, however, is that reserves amassed in the Social Security trust funds will be used to finance current consumption - for example, by paying for current outlays in the non-Social Security portion of the budget. Although the Social Security program would look just the same under this scenario, no saving would occur and the full cost of baby-boom and subsequent benefits would be paid by future generations. It also would have the undesirable distributional consequences of financing current general government activities by the more regressive payroll tax. Nevertheless, this approach is very close to pay-as-you-go financing and therefore, while undesirable, perfectly manageable.
The alternative is to resort to explicit pay-as-you-go financing. If OASDI reserves are not accumulated during the period from now to 2018, then OASDI taxes will have to be raised in 2018 to finance annual deficits on a current cost basis. It is important to note, however, that the payroll tax rate increase required would only be between 1 and 2 percent each for the employee and the employer. These are not small amounts, but they are manageable.
In short, we are not faced with another Social Security crisis, but rather with an opportunity to increase saving and investment. Regardless of whether or not we choose to seize this opportunity, Social Security will function perfectly well.