Transition Costs: Facts and Fantasies

October 13, 1999

Global Pensions, a U.K.-based magazine focusing on international investment and retirement benefits, interviewed Bill Shipman, a principal of State Street Global Advisors and co-chair of Cato's Project on Social Security Privatization. Shipman outlines the problems facing Social Security today and the opportunities reform presents for the future.

KG: You argue that the State-funded pension system is generating inadequate benefits and that its unfunded liability could already be as high as US$8 trillion. What are you advocating?

BS: We're facing a huge problem if we don't change the system. The critical point that I have expressed everywhere is the difference in benefits that one could reasonably expect to have at retirement under the present social security Pay-As-You-Go (PAYG) system, versus a market-based savings and investment system.

The difference is dramatic. The existing system has low benefits relative to the taxes paid in relation to the market-based system. And it will be unable to pay out these benefits by 2015 if the revenues received are as projected.

Even if the system were sustainable, the benefits received are low. So the two main issues are sustainability and rate of return.

Does the embryonic legislation to reform Social Security concur with the models you've been advocating?

Each piece of legislation is different. But all, or almost all, of these bills, make reference to some form of market-based financing. So in this sense, they are sensitive to things that I have been talking about for a number of years.

Given that debt finance may be necessary during any transition period, how difficult will it be to secure government guarantees for this?

Although it doesn't matter how the transition is financed, my point is that if we introduce an alternative system incurring no extra cost, government spending will be no greater. The cost, if we do not, will be all the taxes we must pay over the next 75 years and all the additional taxes needed to maintain retirement benefits at current legislated levels. These additional taxes are the unfunded liability, which is in the region of $3 trillion to $8 trillion. You cannot get away from that. But if you move to an alternative system, and the unfunded liability is less, then the cost of this is less than staying with the existing system.

In your projections, you cite one case where a worker born in 1930 would have fared worse under a hypothetical market-based system than under the prevailing Social Security system. How could such a scenario happen -- and could the same circumstances recur?

There is one case, where a worker who invested in bonds would have been worse off than under the Social Security system. Although it would be folly to invest in a bond-only fund for 46 years, this is still a relevant question. My view is that the government should top it up so that you get to a level that would be applicable under the Social Security system.

Although the probability that this would happen is very low, it does depend on the age of the worker, the savings rate and market returns. For a worker bom in 1976, for example, the maximum return under a market-based system in my projections is 371% greater than under the PAYG. On the day that he retires, his stock portfolio would have to fall by 78% in one day - which has never happened - to dip below the social security equivalent.

But even this scenario makes no real sense as I have assumed that all the gains are invested in S&P 500 stock - which fails to give enough diversification.

Your projections assume pre-retirement nominal investment returns ranging from 7% to 10%. How would a market crash skew your assumptions?

The effects of any crash depends on age and timing. For a person born in 1976, let's say, earning $68,400 now, I'm assuming that the individual will get a nominal return of 7.8% annually from a balanced portfolio of 60% stocks and 40% bonds. The S&P index has been running at about 11% since 1926, so 7.8% is quite low in comparison to that historical norm. Now when such a person retires, his stock portion of the balance fund would actually have to fall by 115% to lapse back to the level of benefits paid by Social Security. This is impossible, as no sum of money can fall by more than 100%. And even if it fell by 100%, his pool of retirement money would still produce an annuity greater than the Social Security equivalent.

The numbers seem so extraordinary because you get the effect of 46 years' worth of compounding under the marketbased system. Only a dramatic drop in the market-based system can reduce it significantly.

Given the certainty still attached to the idea of State funding, don't you worry that the alternative- a substantial cut in benefits or a substantial rise in taxes to fully fund benefits, to a figure of around 17% of payroll that you cite in a recent report - may yet prove the more attractive prospect?

I think it will be quite the opposite, but I may be wrong. That's why I am suggesting that nobody should be forced out of the social security system. If the two systems ran side by side, I believe it would lead to a dramatic shift from the PAYG-based system to the market-based system.

I want people to have choice - and if the market-based system is superior, you shouldn't have to require people to go into it. Free choice should prevail.

In terms of a reformed system, you advocate the design of market-based portfolios of multiple asset classes. Do you envisage these various portfolios being designed by employees or fund managers?

To start off, they would be designed by a board of trustees using the same model as the employer-sponsored defined benefit plan. This would give three balanced funds, with the middle model typical of the employer-based DB plan, split between stocks, bonds and cash. The model to the right would feature more stocks, whereas the model to the left would include less, relative to the middle portfolio.

So you choose the model, the investment manager follows the guidelines, and all the mone y goes into one portfolio. And where all this is beyond somebody, they don't make a choice but simply default into the middle portfolio.

Now as time goes on, your level of assets builds and other people in the market will want to manage them. They would have to follow the investment guidelines, but might offer something above and beyond this. Perhaps a voice-response unit so you could query the portfolio. This way you could choose to opt out into say, Fidelity's balanced fund, or stay as you are. This would be your choice, but neither the employer nor the employee would be responsible for delivering asset performance.

Leaving the choice to the employee brings the possibility of moral hazard. You can imagine the scenario: An individual believes he or she is the greatest thing on the investment earth and chooses to invest in commodity contracts. These do well, and the returns received are far in excess of the Social Security equivalent.

But what happens if the commodities perform badly?At a political level it is the rest of society who bails him out. This causes a moral hazard, with a situation where he or she can only win and society can only lose - and this trade-off makes little sense. With a diversified portfolio, this possibility is remote.

How many providers would you envisage in terms of your optimum model and what might administrative fees cost?

In the short-term, the market-based system would go out to bid. I don't know how many providers would step up, but I imagine that the trustees would choose the providers. In the long run, other providers would come in. I would open it up to all those that meet stringent guidelines, which could probably number well into the thousands.

I don't think that there should be price controls. If you look, the costs are already low now because of the competition among institutions.

What's the best-case scenario? How would people fare under a marketbased system where returns are higher than your top assumption of 10%?

The upside is there even if things don't go well. simply because the market-based system is better than the PAYG system. But if things do go better than I argue in the projections, individuals will have more retirement income than I predicted. Just as importantly, the government costs incurred in maintaining the 42% replenishment rate will go down as people reach this target earlier. This gives a win-win situation for everybody.

On a more general note, do you think governments across the world are grasping the nettle of pensions reform that you feel is so necessary?

Clearly these ideas are being accepted and acted upon all over the world. You only have to look at countries like Australia, Singapore, Chile, Argentina, Peru, Mexico, Venezuela and Poland. It is taking place now all over the world and will be one of the most significant social issues that we deal with and resolve over the next quarter century.

How does the present unfunded liability in the US compare with other major industrial nations?

Well, the $3-$8 trillion is the actuarial figure for the US Social Security system. When you look at the other OECD nations, the unfunded liabilities can range from anything between 90% to 250% of GDP, the figure for Italy. In fact the figure is as high as 400% for Brazil. With the exception of the UK, which is in very good shape, the unfunded liabilities are greater than the outstanding explicit debt of these countries, usually multiples of this in fact. So this is not just a retirement finance issue.

The US case is fairly typical of the rest of the world, although the numbers may be different. This is a global problem, and a true global opportunity. What I have said in some of my speeches is that we will look back at the period as a true global financial renaissance. Taxes will be lower, but savings will be higher and the burden of debt will finally begin to fall. Greater transparency will prevail, transforming a rather opaque system in the US that the average person finds difficult to understand. People want to say: "Here's what I put in, here's what I'll get out". So the move to transparency will be all to the good.

2001 Index | 2000 Index | 1999 Index | 1998 Index