Preston J. Miller - Former Vice President and Monetary Adviser
Published December 1, 2002 | December 2002 issue
The long-term outlook for the federal budget is miserable. The Congressional Budget Office (CBO) projects that by 2025 the federal budget will have embarked on an unsustainable path: Spending will steadily rise relative to both revenues and the size of the economy, resulting in steadily rising deficits and debt burdens. Should we be concerned now about a problem that only exists in forecasts? If yes, how can sustainability best be achieved?
People could easily conclude that long-run budget projections should be taken with a grain of salt. Even short-run budget projections are known to be wildly erratic. Consider this: In January 2001, the CBO projected that the budget would be in surplus in fiscal 2002 (year ended September) by $313 billion and would have an accumulated surplus over the 10-fiscal-year span, 2002 to 2011, of $5.6 trillion. The pressing question of the day concerned what private assets the government should buy with its surplus money once all its debt was retired.
However, due to unexpectedly weak economic conditions and new legislative actions, the budget outlook abruptly shifted. By August 2002, the CBO revised down the budget balance in fiscal 2002 to a deficit of $157 billion and the 10-year accumulated surplus to $0.3 trillion. So, people might conclude, if even short-run budget projections can vary so much over such a short period, why should we have any confidence in long-run projections?
The answer, in general, is we shouldn't, but in this case we still should be concerned. That is because in this case the budget is the "canary" and not the "mine shaft." We may not know much about the canary's health over the next 50 years, but we can be pretty sure that there will be gas leaking into the mine. The gas, of course, reflects the changing age composition of U.S. society. In this case, the long-run budget outlook mainly reflects long-run demographic projections.
While virtually all forecasts will prove to be wrong to some degree, it is reasonable to hold long-run demographic projections with more confidence than long-run budget projections. The demographic projections depend on things such as birth rates, life expectancy and immigration levels. Long-run budget projections depend on all these same things and much more, like the evolving state of the economy, the distribution of income, and the size and shape of spending programs and tax policies. Thus, a lot more can go wrong with budget projections than with demographic projections.
The demographic projections come from the Social Security Administration. They show the ratio steadily rising of persons aged 65 or older, who generally are retirees drawing benefits, to persons aged 20 to 64, who generally are workers paying taxes. In particular, these projections show the ratio going from roughly 20 percent in 2000 to almost double that by the mid-2030s. The demographics imply that should productivity grow within a plausible range, the relatively small number of workers would not be able to produce enough goods and services to adequately satisfy both themselves and the growing numbers of retirees. In other words, the economic pie would be too small.
Under a wide range of economic assumptions, this problem manifests itself in budget terms as a steadily rising share of the economy going to Social Security, Medicare and Medicaid. Thus, the budget solution requires too large a payroll tax increase to fund existing retiree programs or too large a cut in retiree programs to keep payroll taxes unchanged. However, the basic problem of the economic pie being too small remains no matter the budgetary outcome. Moreover, any proposal for dealing with the problem will not succeed unless it results in a bigger economic pie.
How large is the problem? Suppose the population and economic pie grow as the Social Security Administration and CBO, respectively, now project. Then, if benefits remain at their legislated levels, combined employer and employee payroll taxes in 2030 would have to increase by over 80 percent compared with their current levels. At the other extreme, if payroll taxes were fixed at their current rates, then average retiree benefits would have to be cut in 2030 by roughly 45 percent from what they would be otherwise.
Because of the source of projected budget deficits, it does seem likely that government policies will have to be changed to put them on a sustainable course. Although many things can be done to make the policies sustainable, we want the changes to also treat the underlying problem of the too-small economic pie. Policy changes then should elicit reinforcing private responses.
A useful way to begin thinking about the problem is to imagine it is suddenly announced that retiree benefits will be cut across the board in a few decades. (To make the argument concrete, the focus will be on Social Security, but the reasoning applies as well to health programs.) How would young people react? Would private responses correct the underlying problem, or would there still be room for government actions? Posed in these terms, the question is not about how to share limited resources between generations, as the question is usually framed. Rather, it is about how young individuals, faced with a decline in future wealth, would want to reallocate their effort and resources over their lifetimes. This latter way of thinking about the problem recognizes that all old people were young once.
The young first would have to decide for themselves whether to believe the announcement. If they did not, there would be no reason to adjust their behavior. However, if they did, they would worry that they need more of their own resources to finance their retirements. A reduction in future benefits would affect their behavior much like an increase in a current nondistorting tax. The young then would want to adjust their economic behavior on many margins, such as working more hours over their lifetimes, saving more out of income and holding less-liquid assets with higher long-term payoffs. While individuals would be acting in their own best interests, in aggregate the actions would lead to a bigger economic pie by generating more labor input and a larger, more productive capital stock.
Private actions would alleviate the problem of the economic pie being too small, but important distributional problems would remain. Not all people would have believed the announcement and, hence, would not have adequately provided for their retirement. Moreover, not all people would have earned enough income to enable them to adequately provide for retirement. Thus, transfers to poor retirees from workers or wealthier retirees still would be needed.
The important point is that Social Security is not just a pension planit also is a social insurance program for the elderly. Private responses, in general, can deal with a reduction in private retirement benefits, but they cannot replace a reduction in social insurance.
Considerations about the budget outlook, the underlying demographic problem, private responses and the roles of Social Security suggest a solution that has received little, if any, attention: The government could fix payroll tax rates at their current amounts and require that in any year no more benefits are paid out than what are received in payroll revenues.
Then, people could decide for themselves whether or not to believe the Social Security demographic projections. Those who believed them and could act would take actions that, in aggregate, would tend to increase the economic pie. Those who did not believe them, or could not act, would find, were the projections borne out, that they had too few resources when retirement came.
And if the projections were borne out, Congress would find that it had too few resources to maintain full benefits for all retirees. This would force it to target the benefits to the retirees most in need, thus maintaining the social insurance role of Social Security.
The proposed policy is one way to return the federal budget to a sustainable path, stimulate private initiatives to alleviate the problem of the too-small economic pie and maintain the social insurance role of Social Security. However, the proposal is neither problem-free nor complete. Because it offers social insurance, it creates a moral hazard problem: Low-income workers might want to save even less than they might otherwise as they rely on future government benefits in their retirement years. Nevertheless, this problem seems smaller under the proposed program than it is under the current rules.
Under the proposal, the social insurance would be restricted to those deemed most in need, if revenues were inadequate. No such restriction is present in current rules. The proposal is incomplete because it does not describe how Congress should allocate resources when the resources provided by workers at the time are less than needed for full benefits. There are many ways to reduce benefits for those less in need, but the principle remains that the actions government takes should elicit reinforcing private responses.
No matter how we as a society decide to deal with the aging of America, it is important that the actions taken by government to solve the budget problem reinforce actions taken by the private sector to enlarge the economic pie.