US Social Security

by John Q on December 4, 2004

I’ve read lots of pieces on proposals to reform the US Social Security system, both positive and critical. Unfortunately, most of them include claims that are at best half-true and most of the rest assume a high level of knowledge of the issues. Over the fold, I’ve added a lengthy piece trying to explain the issues. Although I’m actively involved in debate on some of them, I’ve done my best to give a neutral presentation, at least until the final assessment of the proposals currently being discussed by the Administration and Congressional Republicans. This is primarily a matter of political judgement and can be summed up fairly quickly.

The Republican proposals involve accounting transfers amounting to trillions of dollars between different government accounts and newly created individual accounts. These transfers will almost certainly be packaged up with substantive changes to the Social Security system. Whether you support them depends on which you think is more likely:

* The transfers will be used to facilitate tough but necessary increases in contributions relative to benefits, eliminating the funding deficit. In doing this, the President and Congress will demonstrate their commitment to promoting the long term interests of the American people, even at the expense of short-term political pain

* The transfers will provide an ideal opportunity for all manner of pork-barrelling, from handouts to existing retirees to cosy deals for Wall Street investment banks, with accounting tricks being used to provide cover for a claim that the system has been restored to solvency

You may be able to guess which of these I think more likely, but you’ll have to read (or scroll) to the end to find out.

The problems of financing public and private pension schemes in the face of a growing proportion of retired workers have raised problems all around the world. Nowhere, it seems, has the debate been more complex and confusing, than in the United States. In view of the role of the US dollar as a reserve currency, and the current budgetary problems of the US government, the difficulties of the US Social Security Fund are a matter of global significance.

The beginning of the problem is the fact that the taxes currently being levied aren’t going be enough to fulfil the promises that have been made to beneficiaries. There are various ways of measuring the shortfall, but the most relevant is that, if the problem were to be fixed with an immediate injection of cash, the amount required is around $5 trillion or about 50 per cent of current GDP The fund won’t be exhausted until 2042 on current calculations, but something will need to be done well before then.

There are at least four distinct, but interrelated issues.

First, there is a proposal to recognise part or all of the unfunded Social Security liability explicitly, by having the US government borrow money and transfer it to the Social Security fund.

Second, there are proposals for partial privatization, that is, allowing individuals to allocate part of their contributions to a personal account, which they could invest at their discretion, and reducing benefits to those individuals accordingly.

Third, there are proposals that part or all of the Social Security fund should be invested in stocks rather than, as at present, in government bonds.

Finally, there are measures to meet the shortfall either by changing the rules of the Social Security system or by finding savings elsewhere in the budget.

To see what’s going on, it may be helpful to look first at the source of the shortfall. Social Security is a defined benefit scheme, operated on a ‘pay-as-you-go’ basis. The benefit is calculated on the basis of an individuals highest 35 years of earnings, and is not directly related to contributions. Payments to current retirees are made out of the contributions of current workers.

This was a hugely beneficial deal for the first cohort of people to retire under social security. Having made contributions for only a few years, they received an earnings-related pension for the rest of their lives. Compared to a fully-funded scheme, this means that Social Security built up a large deficit in its early years. However, because both incomes and the working-age population were growing fast, this was not an immediate problem.

In the future, with a reduction in the working population relative to the retired, the process will eventually be reversed. Either some group of participants will receive less than they put in, in present value terms, or the government will have to make up the deficit out of its general revenue.

The first proposal discussed above, to refinance the social security fund with new government debt may be seen as a simple recognition of the well-known funding deficit. In isolation, it looks like a desirable move, bringing a real, but unrecognised future obligation on to the balance sheet. Since it does not involve any new obligations, it is reasonable to argue, as Republican proponents of the proposal has done, that it should not count towards measures of the budget deficit, or be treated as an increase in the net debt of the public sector. Before accepting this conclusion, however, it is necessary to consider the interaction between this proposal and the other options being put forward at present.

The second proposal, reallocation of some contributions to individual accounts, with a corresponding reduction in future benefits would, in principle, make no difference to the deficit. However, there would be substantial problems in matching the shift in contributions to the reduction in benefits. Benefits for future retirees are a complex function of past earnings, on which contributions have already been paid, and future earnings, which would attract smaller contributions to the fund. Supposing contributions were reduced by 20 per cent tomorrow, it would be very difficult to work out the appropriate reduction in benefits for someone who retired in, say, ten years time.

The third proposal is to invest some of the social security fund in stocks rather than bonds. This would happen more or less automatically with individual accounts, assuming they were treated like existing 401(k) funds. However, it would also be possible to diversify the investments of the official Social Security fund, as was proposed during the Clinton administration.

Historically, stocks have yielded higher returns than bonds in the long run, a phenomenon known as the equity premium. Higher returns would make it easier to make up the current shortfall.

The problem is that the equity premium may be presumed to reflect, in some sense, the greater riskiness of stocks. Assuming investors are averse to this risk, we might expect to see any allocation of individual accounts to stocks being offset by other portfolio shifts out of stocks and into bonds. The net result will be a wash. A possible exception to this argument arises with workers who have no personal financial assets apart from their Social Security entitlements. Such workers might wish to diversify into stocks but be constrained from doing so at present.

The Clinton proposal is more controversial. Some economists have argued that the same risk premium should apply to public as well as private capital markets. Others (including me) have argued that the observed market equity premium is much too high to be a socially optimal estimate of the cost of risk and must be due to capital market failures of various kinds. Under these assumptions, an increase in public holdings of equity, for example through Social Security diversification, will yield a net benefit. On this analysis, the higher average returns of equity investment could be used to meet at least some of the funding deficit, while fluctuations in returns could be smoothed using the government’s taxation and borrowing powers.

With the controversial exception of diversification, all of the proposals discussed so far amount to a reallocation of existing contributions and commitments, with no change in the aggregate balance. The crucial problem is that of dealing with the funding deficit, whether or not this is brought on-budget. In the late 1990s, it seemed likely that the problem would be addressed by improvements in the general budget balance. The aim at that time was to place required contributions to social security in a ‘lockbox’ and ensure that the budget, exclusive of the lockbox, was in balance or surplus. The accumulated surpluses projected at the time would have been easily sufficient to address the Social Security shortfall with no changes in benefits.

On current indications, however, there is little likelihood of a surplus in the budget, excluding Social Security any time soon. It is therefore necessary to consider either an increase in contributions, a reduction in benefits or a tightening of eligibility. The most likely option is an effective reduction in benefits, by indexing them to consumer prices rather than, as at present to earnings. This would mean that retirees would get a fixed real income, based on their lifetime earnings, rather than sharing in the benefits of general wage growth.

The issues are logically independent, but that doesn’t mean they are politically separate. We can imagine a couple of opposing scenarios. On the one hand, the proponents of policies such as explicitly recognising debt and the introduction of private accounts could seek a broad and well-informed debate leading to a comprehensive analysis of all the policy options, including options to address the funding shortfall. In an ideal political world, this would undoubtedly be the outcome.

On the other hand, it might be suggested that policies involving accounting transfers of trillions of dollars between government accounts and from governments to private individuals would provide an ideal opportunity for all manner of pork-barrelling, from handouts to existing retirees to cosy deals for Wall Street investment banks. It might also be suggested that the difficulty of matching reductions in contributions with reductions in benefits could be addressed by ensuring that nearly everyone was promised that they would be better off. Finally, it might be suggested that a combination of creative accounting and rosy scenarios could be used to justify an announcement that the problems of Social Security had been solved when in reality the accumulated shortfall was worse than ever.

Recent observations of the US Congress, the Bush Administration and the accounting treatment of pensions in the private sector make it fairly clear that the second of these scenarios is considerably more plausible than the first. Current holders of US Treasury bonds will rightly be alarmed if attempts to address the funding deficit are bundled into a complex refinancing package involving a substantial increase in official government debt.



Vaughn Hopkins 12.04.04 at 11:27 pm

In my opinion much of the reason for wanting to privatize SS by the Bush administration is to avoid ever having to pay back the billions of dollars borrowed from the SS “trust fund” to avoid having to raise taxes on the wealthy over the years. How ever you look at it the low to middle class earners have been paying a disproportionate share of the federal budget since FICA tax rates were raised beyond what was needed to pay current SS outflow. Once the FICA tax rates cease to be more than is needed, that share will no longer be paid by our FICA taxes, and an honest administration would raise high bracket income tax rates to compensate, and would pay back the borrowed money long before any consideration of raising the FICA tax rate or cutting benefits.


Giles 12.05.04 at 12:02 am

On the other hand the fact that Bush is now on his last term means he’s more beholden to history than to the electorate – which increases the chance that tough but unpopular descisons can be made. The second factor is that the Baby Boom generation, which is coming up for retirement are pretty implacably anti republican – so there’s no good political reason to pork barrel them.


Vaughn Hopkins 12.05.04 at 12:35 am

Bush has never been beholden to the electorate. He is beholden to his contributers and the businesses who picked him as President. That is why he will try his best to deliver this plum of a benefit to the stock brokerages, and secondarily to the holders of big blocks of stock who will benefit by the price rises the added buying will cause.


jam 12.05.04 at 12:53 am

The problem with these discussions is they all assume that the demographic changes over the next few decades will affect only Social Security. To realize that this is nonsense requires no more than three minutes thought, but, in A. E. Housman’s wonderful dismissal, “thought is irksome and three minutes a long time.”

There are at least two other effects of the demographic changes which will have indirect effects on Social Security. (1) Real prices for goods and services preferentially demanded by the aged will increase. (2) Real wages for work in that part of the service sector which can’t be outsourced will increase. The first is due to there being relatively more aged; the second due to there being relatively fewer workers. However, the first will cause at least some approaching retirement to postpone it (since it will be more costly than planned) and therefore require lesser Social Security outlays than currently anticipated; the second will mean greater income to Social Security than currently expected. Both will at least delay the 2042 date when the Trust Fund surplus is expected to be dissipated under current assumptions.

I do not pretend these are the only possible effects of demographic change which have indirect effects on Social Security solvency. They are simply two I’m aware of. I’m sure there are more. Any reasonable analysis of the effect of demographics on Social Security solvency needs to begin with a model of the effect of demographics on the economy as a whole and then specialize down to the effect on Social Security. To do otherwise is to play politics with the issue.


Giles 12.05.04 at 1:14 am

A third endogeity problem is that
3) The equity premium is partly the result of demographic factors – a boom generation creates excess demand for bonds when young (to finaince mortgages) and stock when older (to finance retirement).


Jim Harrison 12.05.04 at 1:16 am

Nobody seems to take notice that an influential segment of the administration is made up of ideologue who always hated social security and are determined to destroy it . Policy wonk discussions that assume the good will of the Republicans are highly unrealistic. I

By the way, assuming that the stock market will continue to grow at high enough rates to outperform social security is probably also unrealistic since the valuation increases the optimists rely on took place during a period of exponential population growth. The demographic transition has probably changed the rules.


remi 12.05.04 at 1:40 am

I think equity premium will collapse when retired baby boomers withdraw from their or reallocates to less risky assets.


fyreflye 12.05.04 at 2:54 am

The notion that Bush will feel beholden to History ignores the fact that he belongs to a religious cult that expects History to come to an end soon with The Rapture.


Sebastian Holsclaw 12.05.04 at 7:56 am

I’m not happy about age-related wealth distribution. Why should a poor or middle-class worker have their money diverted to older middle and upper-class workers? If we want to have system which guarantees a minimum retirement, fine. But there is no need to support the middle-middle class and up. If you set the retirement level in the lower-middle class zone, most people above that income level during their working years won’t want to cut back to that level afterwards. Then, we would have a system which wasn’t ridiculously expensive and which wasn’t tranfering wealth from middle class workers to any of the richer retirees.


joel turnipseed 12.05.04 at 8:30 am

I’m not an economist — but my grandfather is: and his reactions to this surprise & enlighten me. He is a long-time liberal (he was an early ambulance-chaser, or Forensic Economist), with a large private investment account and very nice pensions from both the Navy (Lt. Commander, no less) and his university. What’s shocking is, when I bring up either means-testing or reduction in benefit, he rails at the very injustice of the suggestion.

Which is to say: as someone who’s already paying his, and in a decade or so will be paying his and my parents’ SS, it strikes me as almost politically impossible to get these kinds of programs on sane ground (this is true in Europe, true, no?).


victor falk 12.05.04 at 9:02 am

I think Sebastian makes an excellent point; in Sweden this is a known as the “general welfare principle”.
The idea is that if everybody both gives some and gets some, then everybody will support the (Social-Democratic Workers’ Party’s) welfare state.
Hence, even a Swedish millionaire gets benefits, such as children allowance, to the tune of thousands of euros a year…
It seems to work pretty well; the overwhelming majority of people in Sweden are totally or partially dependent on the public sector for their income, and the social-democrats have been the government party since 1932, with only two brief spells in opposition, 1973-79 and 1991-94…


Brett Bellmore 12.05.04 at 2:17 pm

I think it’s rather amusing to project government policy and demographic trends out two, three, or even four decades, without taking into account technological and medical progress.

On the technological front, we have the prospect of some sort of self-reproducing technology causing the productivity curve to go briefly “vertical”, by decoupling industrial production from human labor. This could be in the form of widespread “lights out” manufacturing, or the more spectacular promise of molecular nanotechnology. Either way, the ratio of laborers to retirees becomes irrelevant, while the problems of social security are swamped by the question of what to do with all those healthy workers for whom no work can be found. Well, maybe they can be reeducated for more intellectual jobs; It’s beginning to look like education will be as capable of being automated as assembling widgits.

On the medical front, we are finally begining to understand the underlying mechanisms of aging, and they look like they can be fixed. Much beyond a decade out, we can no longer take for granted the inevitability of people aging and becoming too decrepit to work. And once old doesn’t equal “unable to work for a living”, what justification remains for intergenerational income transfers? The elderly can work for their living, just like everybody else. Assuming everybody else isn’t unemployed because of countertop factories capable of manufacturing anything they’re given the design for, for pennies a pound.

As I say, in an age of incredibly fast technological progress, it’s amazing to see how far out economists will try to push their models.


fantazia 12.05.04 at 2:27 pm

Chickens, eggs, and not counting them before they are hatched, Brett. Government policy based on the Singularity solving everything is all very well, but where’s the harm in making the (admittedly flighty) assumption that we won’t be transcending into Godhood in the next two decades when planning our affairs?


Boffo 12.05.04 at 3:41 pm

Re: Bush in his second term

While the truism is that presidents in their second term act with an eye on their historical legacy, as Giles says, intention does mean outcome. In other words, Bush may be more beholden to history than to the electorate, but who is beholden to him?

Instead, presidents lack the leverage over Congress (not to mention other political actors) they enjoyed in their first term. This was quite clear, for example, during Reagan’s second term even before Iran-Contra further sapped his strength and the Dems regained the Senate. So what does this mean for Bush? I anticipate a scenario much like when Reagan announced that welfare reform would be a priority for his second term — the agenda was taken over by Congress. So, even more than what Bush would do with such a proposal, our focus ought to be what Hastert, Frist, House Ways and Means, and Senate Finance would do.

As we’ve seen only weeks after the election with the intelligence bill, the balance of power seems already to be shifting.


Boffo 12.05.04 at 3:44 pm

Oops, correction to previous post (must proofread more carefully):

intention does NOT mean outcome


abb1 12.05.04 at 4:46 pm

Why should a poor or middle-class worker have their money diverted to older middle and upper-class workers?

How to kill a social program:
1. make the middle and upper-class ineligible
2. kill the rest.


Nicholas Weininger 12.05.04 at 7:26 pm

How common, internationally, is the practice of indexing state old-age pensions to earnings growth rather than consumer price inflation?


Jason McCullough 12.05.04 at 10:35 pm

“Why should a poor or middle-class worker have their money diverted to older middle and upper-class workers?”

Why should a worker be able to buy capital, and then used the increased output due to that capital’s to lay claim to younger worker’s production? The strange things “free market” people get up to these days.


pgl 12.05.04 at 11:55 pm

Very good throughout – and I especially liked the part where the higher expected return from stocks is offset by the extra risk COMBINED with the insightful (oft said by economists and never recognized by policymakers in the US) that individuals will shift their private accounts away from stock to bonds aka the “wash”. You also note the different between the social risk premium v. private side (aka the Clinton idea). Besides an old paper by Kent Smetters on this score – what other literature addresses this important point?


Sebastian Holsclaw 12.06.04 at 12:12 am

Jason, you have a definitional problem if you define Social Security as part of the free market.

Abb1, there are all sorts of programs where a huge section of the populace is ineligible. Take welfare in general, for instance. Even among Republicans, there isn’t a large constituency to eliminate it–though there was an appropriate push to encourage you not to live your whole life on the dole.


John Quiggin 12.06.04 at 1:22 am

“You also note the different between the social risk premium v. private side (aka the Clinton idea). Besides an old paper by Kent Smetters on this score – what other literature addresses this important point?”

Blows own trumpet

Grant, S. and Quiggin, J. (2002), ‘The risk premium for equity: implications for the proposed diversification of the social security fund’, American Economic Review, 92(5), 1104–15.


Jason McCullough 12.06.04 at 10:40 am

“Jason, you have a definitional problem if you define Social Security as part of the free market.”

I don’t think I’m the one having definitional issues. Apparently buying a weighted cross-section of investment goods through a heavily regulated monosopy stock “market” by paying a significant toll both on entrance and staying in the market is a-ok, but doing the exact same thing at one-tenth the cost by buying government bonds is communist or something.


james 12.06.04 at 4:51 pm


The government (social security) buying government bonds is not the same as say GE buying government bonds. An IOU from yourself to yourself is very different than an IOU from yourself to someone else. Especially if you are the one we makes the laws concerning repayment. All the goverenment is doing is using ss taxes as revenue.


Jason McCullough 12.06.04 at 6:31 pm

James, I’m not seeing it.

Buying stock for retirement investment is laying claim to the production of tomorrow’s workers; so is a government generational transfer program.


Tracy 12.07.04 at 8:53 am

Although I think finance is a topic worthy of study in itself, and often it is appropriate to think about things at the level of finance, I wonder if the discussion of the aging population problem spends too much time at the level of finance and not enough time at the level of things you can kick.

When all the baby boomers have retired and us workers are supporting some ridiculous ratio of retired people, it’ll be food and housing and health-care and clothing and the like we’ll be consuming, not stocks and bonds. While we could do some storing of some physical things ahead of time, on the whole food and health-care need to be produced shortly before they’re consumed – I wouldn’t want to live mostly off dried and canned food and I can’t have a heart bypass now since my heart is perfectly healthy, even if we could tell for certain I will need one in 40 years time. And while houses and the like can be built ahead of time, they do need pretty constant maintenance over time which takes labour. So, barring the invention of Star Trek’s replicator, what we consume in the year 2024 will have a lot to do with what we produce in the year 2024 (and the few years before that). Increasing savings would presumably mean a higher capital stock and to the extent the savings fund increased successful R&D, higher knowledge and thus higher productivity. But, since the baby boomers are set to start retiring at about the same time around the developed world, surely the price of financial assets will just fall as as a larger proportion of the population is trying to sell than buy them? And whether the savings are held by government or private individuals won’t make a difference to that?

I know the more sophisticated arguers for privatised accounts base them on private investors having more incentive to invest in income-producing things than the government and thus increasing the production each year in the future. But as far as I can tell, either workers or retired people will be enjoying a fall in their share of earnings, regardless of whether we save now or pay increased taxes in the future. Which implies that policies should be focussed around ensuring that there isn’t a fall in absolute incomes. In the case of NZ, breaking the link between wages and pensions is the obvious solution to the pensions problem. Don’t know about the US.

Comments on this entry are closed.