1994-96 Advisory Council Report
The Individual Accounts Plan
Edward M. Gramlich and Marc M. Twinney
The Council is putting forward two starkly different visions for Social Security, summarized respectively in the MB plan and the PSA plan. We are often asked why we can't support either one -- why add yet a third plan? The reason is that we have serious reservations about both of the other approaches. The IA plan that we support avoids the problems we see with these other approaches.
This IA plan is the essence of simplicity. One part is to bring the main OASDI system into long term actuarial balance by benefit cuts rather than by tax increases. Some will object to the size of the benefit cuts, but these are the absolute minimum necessary given the long term financial difficulties facing Social Security. The other component of the plan is to add non-budget defined contribution accounts to Social Security. These in effect make up for the benefit cuts and provide, on average, the same benefits as under present law, for all income and age cohorts in the population.
As compared to the MB plan, there are six important advantages with this IA plan:
The IA plan provides new saving. We can debate about whether Social Security in the 21st century should be entirely a defined benefit system, as it would be in the MB plan, or have some new defined contribution elements, as in the other two plans. But somehow or other there must be some new saving soon to finance the nation's retirement system in the 21st century. The IA plan provides that new saving and the MB plan does not.
The lack of new saving provided by the MB plan means that there is no new national income in the 21st century to support whatever benefits are to be paid. It also means that as the Social Security Trust Fund accumulates a huge amount of common stock and correspondingly sheds a huge amount of government bonds, US private citizens will be holding that much less common stock and that many more bonds. In the end the MB plan represents a giant asset switch with the private sector, with the nation as a whole no better off and future living standards no higher. There is no economic or social point to such an asset switch.
The IA plan gives people an explicit stake in Social Security. The Council has heard over and over how people, especially younger people, do not trust Social Security, do not think any money will be there for them. Small scale individual accounts will not totally change this view, but they should change it in a positive direction. With the IA plan all workers will have and manage their own accounts, and these pre-funded accounts will be protected from many of the political, economic, and demographic uncertainties facing the OASDI system.
The IA plan reduces risk. The IA plan is sometimes criticized as adding too much risk to Social Security. The criticism is totally misguided. The stock market assets backing the individual accounts are index funds, just as in the MB plan. If the overall return on stock does not hold up, returns from the individual accounts will not be as large as forecast, and benefits in the MB plan will not be as large as forecast either. There is no added risk in the IA plan, and indeed even slightly reduced risk because of the lower amount of equity holdings.
The IA plan does not threaten the public-private separation. Under the IA plan all equity investments are under the discretion of individuals through the funds they elect, just as with other defined contribution accounts in the United States today (including the federal employees' thrift plan).
Under the massive central government stock purchases of the MB plan, by contrast, there could be any number of new difficulties. If particular firms are felt to be in violation of some ethical norm, some foreign policy principle, some labor standard, or even some political standard, there could be calls to dump the OASDI stock holdings of these firms. That change might sound benign, but it would disrupt the hands-off equity index fund principle of the MB plan and lead to a chance of political interference in the operations of private business.
The IA plan does not threaten fiscal integrity. Since no government budget funds are being used to buy stock, there should be no interference with the federal budget process. Under the MB plan, on the other hand, suppose the OASDI Trust Fund were to wiggle out of its long run actuarial difficulties by buying stock and selling bonds. What is to assure that the Medicare Trust Fund, or the Highway Trust Fund, might not follow the same approach? What is to stop the whole federal government from buying stock every time budget tightening was needed, as opposed to cutting spending or raising taxes? Indeed, if it is a good idea for the federal government to own a trillion dollars of common stock through Social Security, why does not the government now just borrow a trillion dollars and buy this much stock? It has proven very difficult for the federal government to practice responsible and prudent deficit reduction policies over the last two decades, and the MB plan introduces a huge new political obstacle to these efforts.
The IA plan does not reduce the revenues of the Medicare Trust Fund. Both of the other plans, for different reasons, divert revenues that now go to the Medicare Trust Fund. The IA plan does not do this, and is thus the only plan that preserves even the present precarious state of Medicare financing. Marc Twinney's statement gives the impact of this change on the overall money's worth calculations. That impact is modest, but if it were adjusted for, the IA plan would compare more favorably with the other plans in money's worth terms.
Turning now to the PSA plan, the IA plan improves on it in five ways:
The IA plan entails no transition costs. The IA plan is designed to phase in naturally. The proceeds from the individual accounts gradually build up as the central OASDI benefits phase down, so that all classes of workers receive about the same benefits and the system changes naturally from the present system to the new system.
In the PSA plan, by contrast, higher taxes and large-scale borrowing are necessary to take the nation through the transition phase. Even if financed by a 70-year payroll tax, the financing mechanism now proposed by the PSA advocates, there would still need to be borrowing of up to 2 percent of GDP, $150 billion at present rates, in the next ten years, to finance the transition. Yes, this borrowing would be paid off beginning in forty years, but that is still a huge amount of borrowing for a very long time. Both the higher borrowing and the higher taxes are politically unpopular and economically unwise.
The IA plan provides for adequate basic Social Security benefits underneath the individual accounts. The IA plan phases down the benefit schedule, largely at the top of the income scale. But even for high wage workers, in the end more than sixty percent of the benefits come from the central OASDI system, as opposed to just twenty percent in the PSA plan.
To illustrate some of the risks in the PSA plan, when this plan hits its long run equilibrium, the central flat benefit paid by the OASDI Trust Fund would be at about the poverty line (it would be less than that now, but rise over time in real terms because of the wage indexing). That is simply not enough social protection for our taste. Most workers will invest their PSAs well and have a decent retirement income, but some may not, and they may be forced down to this low base level standard of living. Most workers will retire when the stock market is performing normally and will not be at the mercy of a sharp decline in stock values, but some may not and may be forced down to a lower standard of living. Most disabled workers may become disabled near the end of their career and have accumulated PSAs to draw on, but some may not and be forced down to a lower standard of living.
The IA plan imposes prudent restrictions on the management of the individual accounts. Switching to a greater public reliance on defined contribution pension accounts does introduce some new risks into the retirement saving system. Analysts may well underestimate the financial prudence of average wage earners, there may not be wide-scale abuses in privately-held individual accounts, or pervasive bad investments. At the same time, with more than 100 million working Americans of all wage levels holding individual accounts, there are bound to be some abuses, some instances of unwise investment, some people who arrive at retirement age with minimal amounts in their individual accounts.
The IA plan deals with these risks by constricting individuals' portfolio selection to government-managed index funds. For the PSA plan, by contrast, it is very hard to see how the private PSA accounts can be effectively regulated. Many of these PSA accounts will be small, and will be held in a disparate set of financial institutions. As with IRAs, there are also political risks that Congress will permit the privately-held PSAs to be used for non-retirement needs, hence defeating the purpose of stimulating retirement saving, and letting another set of people arrive at their retirement age with only the low flat benefit.
The IA plan keeps fund administrative costs low. Since all the money for the individual accounts will be collected by the government and allocated to peoples' individual accounts in a way that these individuals have designated, the money does not have to be managed account-by-account. Allocations to separate funds can be aggregated and large checks can be sent to the money managers.
For the PSA plan, by contrast, these individual accounts will have to be managed account-by-account, providing for excessive record-keeping costs for very small accounts. While some large funds may take advantage of economies of scale in managing assets, others may not. There could also be large private advertising expenses.
The IA plan requires annuitization of the accounts. Increasingly workers will be living very long retirement lifetimes, often more than half their working careers. It will be very hard for these workers, on retirement, to determine how much money they will need to provide for their very old age, in the face of inflation and many other uncertainties. Hence the IA plan simply converts fund accumulations to an annuity on retirement, and maintains the real value of this annuity under inflation, just like present Social Security benefits.
As it is presently written, the PSA plan permits workers attaining age 62 full access to their accounts that have been accumulated over an entire working career. The government is in effect saying to people that it does not trust them to save for the future when they are younger than 62, so it requires them to hold PSAs. But once these people become 62 they suddenly become wise and responsible, and the government no longer requires them to preserve their assets beyond that date. This segmentation seems strange. Some restrictions on the potential overspending of the newly-retired seem to us sensible; some annuitization important.
We have tried to resolve these serious problems with both alternative plans by proposing a plan that avoids their difficulties. On one side, it increases national saving and provides a way to get retirement funds into equities without threatening the public-private balance or the integrity of the federal budget process. On the other side, it entails no transition costs at all, it preserves the important social protections of the present day Social Security program, it provides for prudent and efficient management of funds, and it protects people against the financial risk of living a very long life by requiring annuitization of their IA funds.
Social Security would be preserved, replacement rates for all income groups of workers would be preserved, no transition financing would be necessary, and the asset-management restrictions would be prudent. There would be a sound balance of defined benefit and defined contribution accounts, of stocks and bonds. The IA plan seems clearly the most sensible way of compromising among the many retirement saving objectives now facing the nation.