July 15, 2005
JS-2535
Remarks by Assistant Secretary for Economic Policy Mark Warshawsky on Social
Security Reform and the Implications for National Savings
The Urgency for Social Security Reform and the Implications of Effective and Fair
Reform for National Saving
Remarks to the University of Chicago Graduate School of Business Alumni Finance
Roundtable
Thank you for the kind introduction. It is both an honor and a pleasure to speak with you today on the important topic of Social Security reform. As you are aware, Social Security reform is a major priority of this Administration and we are working hard with members of Congress to get it done.
Today I will explain why it is so important that responsible Social Security reform occur now, and why one element of a successful reform plan must be personal retirement accounts that give individuals more control over their financial futures. I will also say a few words about the implications for national savings of making Social Security permanently solvent in a fair manner.
The Size of Social Security's Financial Shortfall
For many years the Social Security Trustee Report has featured a 75-year measure of Social Security's financial shortfall, which is now estimated at $4.3 trillion or 1.92 percent of taxable payroll. But that measure substantially understates the true size of the funding gap because the short projection period excludes from the calculation a larger share of program benefits than it does of program taxes. The Social Security Actuary's estimate of the infinite horizon imbalance, which for the last 3 years has been reported in the Trustees' Report, is $11.1 trillion or 3.5 percent of taxable payroll.
The Administration believes that reform should make Social Security permanently solvent. The permanent solvency goal was recently unanimously endorsed by the Senate in an amendment to the Senate Budget Resolution. That amendment states that "the American people including seniors, workers, women, minorities, and disabled persons should work together at the earliest opportunity to enact legislation to achieve a solvent and permanently sustainable Social Security system."
Intergenerational Equity: Why Social Security Must be Reformed Now
It is clear that the Social Security system is not financially viable and must be fixed. How to close the permanent financing gap raises difficult questions over how the net benefits of Social Security should be shared across generations. In this context, it is important to recognize that the large unfunded obligations in the system are primarily the consequence of the past generosity to generations that are now either dead or retired. Of course, those early generations are beyond reform's reach, so the entitlement reforms needed to close the financing gap must fall entirely on later generations.
Viewing Social Security from the perspective of how it affects generations and individuals explains why it is imperative that Social Security be reformed now. Delaying reform only reduces the options for fairly distributing the benefits of Social Security across generations. Most people agree that it would not be fair to alter Social Security's promises to retirees and near retirees. The longer reform is delayed, the fewer generations that are left to participate in a reformed entitlement system so as to close Social Security's funding gap, and the more severe those reforms will be.
To make this point more concretely, consider a policy of closing Social Security's permanent financing gap by immediately increasing the payroll tax rate by 3.5 percentage points. If the tax increase were instead delayed until 2041 when the trust fund is depleted, the requisite tax increase would be 6.3 percentage points. Clearly, I do not advocate any of these policies, in particular because tax rate increases result in lower incentives to work. My point is that there is no doubt that fairness to future generations requires that action be taken now.
I would also point out that purely pay-as-you go financing of Social Security would be grossly unfair to future generations. For example, one way to make Social Security solvent would be to leave benefits unchanged and to raise payroll taxes year by year beginning when the Trust Fund is exhausted. According to current projections, the payroll tax rate under that policy would steadily rise beginning in 2041 and reach 19 percent at the end of the 75-year projection period. No reasonable person would consider it as fair to leave taxes and benefits unchanged for 35 years while asking future generations to pay such high tax rates to get the same benefit replacement rates as people receive now. The same conclusion would apply to a policy of cutting benefits on a pay-as-you-go basis. The implication is that a fair reform would share the net benefits of Social Security with near-term birth cohorts more equitably than would occur with pay-as-you-go financing, which, in turn, implies that a fair reform would result in an accumulation of either Trust Fund or PRA balances. Trust Fund and PRA balances of course constitute pre-funding.
The Implications for National Saving of Effective and Fair Reform
A fair and permanent fix for Social Security would probably cause national saving to increase. This is a result of the way reform would likely distribute Social Security's net benefits across generations, a choice that should be made wholly on fairness grounds. It is important to remember that fairness is the goal and increased national saving is the implication, not the other way around.
Why would a fair Social Security reform cause national saving to increase? To answer that question, I first note that Social Security is essentially a combination of a forced savings program, a wealth transfer program within generations, and a wealth transfer program between generations. This is apparent when you think about how Social Security affects you personally: First, it either adds or subtracts from your lifetime wealth, and second it takes income from you in your working years and gives it back along with some additional accumulation in your retirement years. The first part is a lifetime wealth transfer, that is a rate of return on contributions either in excess of or less than the fair investment return, and the second part is forced savings.
Primarily what matters for national saving is how Social Security transfers wealth across generations. As you perhaps remember from Economics 101, taking wealth from one 50-year-old and giving it to another 50-year-old has on average little or no effect on national consumption, and hence has little or no effect on national saving. Also, any undesired forced savings would rationally be undone by anyone with the means, which is anyone with the ability to borrow money or to redirect their discretionary savings. But intergenerational wealth transfers do affect national saving: Taking from a 40-year-old to give to a 60-year-old can be expected to reduce national saving because the consumption of a rational forward-looking 60-year old is more responsive to a financial windfall than is the consumption of a rational forward-looking 40-year-old. This is a generally accepted implication of Franco Modigliani and Richard Brumberg's life-cycle theory of consumption. That theory states that a financial windfall causes a person's consumption (including the services of consumer durables and not the initial purchase) to increase in a smooth manner over their planning horizon. Because young people have longer planning horizons than older people, their propensities to consume a windfall are smaller than those of older people.
As I mentioned earlier, Social Security has made huge wealth transfers from current and future workers to generations now either retired or dead. Our theory tells us that those wealth transfers caused national saving and national wealth to decline relative to the hypothetical of no Social Security. And indeed one of the motivating factors behind the creation of a pay-as-you-go Social Security system in the Great Depression years was to depress savings. But those wealth transfers cannot be undone. All Social Security reform can do is divvy up the bill racked up by early birth cohorts, and the way that bill is divvied up across generations affects national saving. The more of the bill that is paid off early, the less that will be paid off later and the better off will be future generations. And the mechanism by which paying off the bill early increases the wellbeing of future generations is through additional national saving.
With this as background, we can see that Social Security reform's effect on national saving depends on how reform allocates Social Security's net benefits across generations relative to what people currently expect. Expectations matter of course because current consumption depends on peoples' projections of future disposable income flows. I hazard to guess that most people expect that Social Security reform will be delayed until the absolute last moment. If that is true, then producing a permanent reform now, as I expect will occur, would result in more of Social Security's net benefits being allocated to future generations than is currently expected. Relative to current expectations, such a reform transfers wealth from currently living generations to later generations and would cause national saving to increase.
Personal Retirement Accounts Are Necessary to Ensure That Reform's Intentions Are Realized
The Administration believes strongly in personal retirement accounts (PRAs). PRAs provide individual control and ownership, and expand opportunities for individuals to partake in the benefits of participating in private capital markets. Individual control and ownership means that people would be free to pass the value of accounts to their heirs through bequests.
But perhaps most importantly, PRAs are a necessary part of reform if the intended allocation of Social Security's net benefits across generations is to be realized. Imagine, for example, that a reform is implemented that is judged permanently solvent but which relies on very large Trust Fund accumulations and no PRAs. In order for the reform to work as intended, those Trust Fund accumulations must represent real saving. But such "saving" would almost certainly be undone by political pressures to increase government spending and hence produce larger deficits outside of Social Security. The only way to truly save for our retirement and give our children and grandchildren a fair deal is with personal accounts. Personal accounts serve as private and therefore effective "lock boxes". When pre-funding is done using a personal account, there is no pressure to increase government spending, because this pre-funding belongs to individuals and does not appear on the government balance sheet as budget surpluses. And I would note that individuals would not be able to withdraw funds from their PRAs until retirement.
As many of you are aware, increased short- and medium-term public debt issuance would be necessary to help fund PRAs. You might ask whether this debt would neutralize the direct effect PRAs would have on national saving. The answer is no. The incremental public debt would be very different than debt that we are familiar with because it is countered by an approximately equal reduction in the government's obligations to pay future defined benefits. That is, a comprehensive aggregation of government liabilities would be about unchanged relative to the case of no PRAs. Hence, unlike public debt issued to fund increased spending, public debt issued to help fund PRAs would not reduce national saving and would not significantly affect interest rates.
CONCLUSION
To conclude, let me say that I am encouraged that Social Security reform is finally being earnestly debated, and that all parties are motivated to make Social Security fair and permanently solvent. Today, my small contribution to this debate consists of four points:
1. Social Security as currently designed cannot be sustained. We know with absolute certainty that Social Security will ultimately be reformed. The only question is when and how.
2. Social Security reform is urgent. The longer reform is delayed, the more unfair reform will be to future generations, and the more difficult it will be for individuals to plan their financial futures.
3. Making Social Security permanently solvent in a fair manner requires that retirement incomes be pre-funded to a large extent, and the pre-funding must be done in PRAs rather than the Social Security Trust Fund. Any attempt to pre-fund retirement incomes in the Trust Fund would be undone by excessive government spending outside of Social Security.
4. An effective and fair Social Security reform would very likely cause national saving to increase. Increased national saving is an implication of fairly distributing Social Security's net benefits across generations.