| 摘要 | I explained to your staff director that I am not an expert on Social Security. But since he persisted in inviting me despite that, I will offer some thoughts on the subject.
As I see it, our present situation is that we have a problem and an opportunity. The problem is that by common estimates the reserves in the Social Security trust fund plus the inflow of payroll tax receipts dedicated to the fund will, at some future date, be insufficient to pay the benefits provided by existing law. The opportunity is that by common estimates the Federal budget–the unified budget including the trust accounts–will be in substantial surplus for the next several years, if present tax rates and expenditure programs are continued. I emphasize the words. “by common estimates”. I recognize the uncertainty of all these estimates, but I am unable to make such estimates by myself, and I only proceed on the assumption that the estimates are correct.
So, we have two questions. What should we do about the deficit in the Social Security accounts, which will become evident some time in the future if nothing is done? What should we do about the surplus in the unified budget that is now present?
I shall start with the Social Security question.
One course would be to cut the benefits that are provided in existing law. The government has no legal obligation to pay the existing benefits. No one can sue you if you decide to change the law. But I believe that there is a moral obligation not to disappoint the expectations on which generations of workers have counted, and I think that most people would agree with that.
There are some changes of benefits that could be defended as not denying previous expectations. People on the average live longer than they used to live, and will receive benefits for a longer period than was contemplated even a few years ago. Some adjustment was made for that in 1983 by an increase in the retirement age. A further increase could probably be justified. Also, when social security benefits were first indexed to the cost of living, in 1972, we did not realize by how much the official measurement might overstate the increase in the cost of living. Some adjustment for that might be justified .
Whether adjustments of this kind would be sufficient to bring the Social Security accounts into long-run balance I don’t know. I shall be assuming that they are not sufficient, and a significant deficit remains.
Before I proceed further I want to make clear that my adherence to the present benefit schedule is entirely based on the fact that it has become expected. If we could start over, I would prefer a different system, that would be smaller and cheaper, providing a low-level equal retirement benefit for everyone. In fact, it might be a good idea to start developing a system that would apply to everyone born after the year 2000, who could not claim that his or her expectations had been denied. But that is a subject for another day.
So, suppose we start with the proposition that even after the benefit scale has been adjusted the reserves and future receipts of the system will be unequal to the planned benefits. What will be the source of the additional funds that will be needed?
The additional funds will have to come out of the income of the rest of the society–the part that is not drawing the benefits. There are two questions here. One is how to make the transfer of income from the rest of the society to the beneficiaries as painless as possible. The other is how to make the commitment of the rest of the society to support the beneficiaries as strong as possible.
The surplus now in prospect can provide an answer to both questions. The best way to facilitate the support of the next generation for the social security beneficiaries of that time is to make the national income of the next generation higher–that is, to raise the rate of economic growth over the next generation. And the most reliable thing the government can do to raise the income of the next generation out of which benefits will be paid is to save the budget surpluses now in prospect. By saving them I mean refraining from cutting taxes or enlarging expenditure programs but using the surpluses to reduce the outstanding Federal debt to the public. If we do that the savers who would otherwise be holding Federal debt will have funds available to invest in private productive assets, and that will raise the income of future generations.
I am not a fanatic about the surpluses. There may be tax cuts or expenditure increases of small size and large benefits that should be made. Contrary to what Senator Dirksen said about 50 years ago, a billion here and a billion there no longer adds up to money. But I am talking about saving most of the surplus.
Even if we have done what we can to raise the income of the next generation, how can we make sure that the active earners of the next generation will make the payments needed for the benefits of retirees? The literal answer is that we cannot. We cannot now commit the voters and Congress of the year 2040 to pay those benefits. But we can do some things that will add assurance. If we do not cut taxes now or embark on additional expenditure expenditures now it will be less necessary in the future to raise taxes or cut expenditures, which is always difficult.
Beyond that, we of this generation could make a contribution to the Social Security reserves. That would be an act of only symbolic significance, without immediate fiscal or economic consequences. The Secretary of the Treasury could write a check for any amount, say $100 billion, and deposit it in the Social Security trust fund, where it would be used to buy Treasury bonds. That would not change the unified budget surplus, it would not change the debt held by the public, it would not change national savings and would not change the rate of economic growth. But the symbolism might be important, because the enlarged reserve of the trust fund might seem to make more concrete the obligation of the next generation to pay the promised benefits.
There is no necessary connection between the size of this contribution to the social security reserves and the size of the unified budget surplus. Nevertheless, the idea might be easier for people to understand and accept if the policy were seen as contributing a surplus to the reserves. The surplus in the social security accounts is automatically contributed to the reserves. By present calculations in about five years the surplus in the unified budget will begin to exceed the surplus in the social security accounts, That is, the rest of the budget, excluding the social security accounts, will begin to run surpluses. One possible policy would be to contribute these surpluses, in addition to those in the social security accounts, to the social security reserves. That might continue until the reserves were sufficient, along with future payroll tax revenues, to pay the promised benefits for some long period into the future.
Up to this point I have not mentioned the magic word “Privatization”. Privatization has two meanings. The first is that the funds of the system should be invested, in whole or in part, in private assets, instead of in Treasury securities, as they now are. The second meaning is that the funds should be owned, in whole or in part, privately by individual covered workers or beneficiaries.
These two meanings are not necessarily connected. One can imagine a system in which each covered worker receives each year a Treasury non-negotiable bond with his name on it, equal to some fraction of his contribution for that year, and redeemable for cash when he begins to draw his benefits. This bond would be his private property, and the system could be called privatization. It would have one advantage. It would reduce the apparent budget surplus and so reduce the temptation for the government to spend the surplus. Otherwise this nominal form of privatization would make no difference.
The great interest in privatization is connected with the first meaning, to invest some of the funds in private assets, rather than in Federal securities. The attraction of this idea is the observation that on the average private investments yield more return than Treasury securities yield. Investment of the social security funds in private assets would increase the earnings of the system and so enable the system to pay the planned benefits. This spread, the excess of the yield of private assets over the yield of government securities, would seem capable of bringing the system into balance without–and this is the critical point–without having to take anything away from anyone.
I believe that on this critical point the argument for privatization is wrong unless privatization adds to total saving. If it does not add to total saving it will not add to total national income, and if it does not add to total national income the increased earnings of the social security accounts will be balanced by less earnings of the rest of the society. Privatization would then be an indirect way of abstracting money from the rest of the society and transferring it to Social Security beneficiaries.
I have this schematic view of the way the transfer would occur.. There are two kinds of assets in the country–private assets and Treasury securities. The private assets yield more return than the Treasury securities. There are two portfolios in the country–the Social Security accounts and the sum of private portfolios. At present the Social Security accounts are entirely invested in Treasury securities and they earn a lower rate of return than the private portfolios. Suppose we now decide to invest some of the Social Security accounts in private assets, and therefore invest less in government securities. Then more government securities have to be held in private portfolios. But the private people whose portfolios those are cannot be forced to hold more Treasury securities. They have to be induced to do so. And what will induce them is a rise in the yield on the Treasury securities, narrowing the spread between the yield on Treasury securities and the yield on private assets. So taxpayers will have to pay more interest on the Federal debt, and that is the form in which they will transfer income to the Social Security accounts. It seems to me that either private investors will earn less on their total assets or taxpayers will pay more in interest on the publicly-held debt. The increase in the income of the Social Security accounts will not be costless for the rest of the society.
The picture might be different if privatization resulted in an increase in total saving. Then one could see an increase in the total national income and an increase in the total earnings of assets, so that the Social Security accounts could earn more without anyone earning less. But this seems to me an entirely unlikely development. This is probably clearest if the privatization takes the form of changing the assets held in a collective pool in the Social Security accounts. Nothing has happened to the surplus or deficit in the Federal unified budget. And nothing has happened to make private people want to save more. The rate of return on Treasury securities will be higher than it was and the rate of return on private assets will be lower, because the Social Security accounts will now be in the market buying some of those assets.
But the situation is not really different if some of the Social Security funds are returned to covered workers and they are allowed to invest them as they choose. The amount that would be returned to them would have been saved in the Social Security accounts anyway. They will have no different incentive to save any other income than they already have.
I understand that his is a complicated subject and that there are people whose opinion I respect who disagree with me about the consequences of privatization for total saving. But this is where I come out. Setting up private retirement accounts out of money that would otherwise have gone into the Social Security accounts is a way of getting that money out of the apparent surplus and so reducing the temptation of the government to spend it. But that would happen whatever the private retirement accounts were invested in.
In conclusion:
1. Except for possible adjustments relating to the retirement age and the indexing formula we should meet the benefit schedules that are in current law for all workers now covered.
2. We should preserve most of the forthcoming surplus in the unified budget–that is, not use it for tax cuts or expenditure increases but use it to retire some of the publicly-held debt.
3. We should contribute the publicly-held debt we are retiring to the Social Security accounts as a symbol of our intention to provide the benefits that presently-covered workers expect under current law.
Herbert Stein is a senior fellow at AEI. |
