Skip Navigation

Mr. Chairman and Members of the Committee:

There are a number of ways to analyze governmental programs, especially those that span many years and multiple generations. We need to be careful about the question we are attempting to answer and which analyses to apply. Further, we need to be careful not to mix the analyses and their respective concepts.

In the case of Social Security and Medicare, we pay the benefits of our parents and grandparents through taxes on current workers, that is, on us--both payroll and income taxes--and both programs will take much more from our children in order to fund them when we retire.

This year, Social Security and Medicare will account for 6.5 percent of gross domestic product (GDP). By 2030, those programs will grow to 11.0 percent of GDP. Moreover, the number of beneficiaries will grow much faster than the number of workers paying taxes to support those programs. The ratio of covered workers to beneficiaries will drop from about 3.4 this year to about 2.3 by 2030.

How we analyze these spending commitments and demographic changes is vitally important.

One approach, the one you've been dealing with thus far today, is to use actuarial techniques to project costs and income, and to focus on the revenues specifically dedicated to the program. That approach can help us ascertain whether a program, abstracted from the rest of the budget and the program's effect on the economy, is stable on its face--usually over long periods of time.

One measure of actuarial long-run viability is "solvency." That is, are expected revenues and expenditures roughly equal over proscribed but long periods? Another measure is the comparison of the present value of total expected revenues and total expected obligations--whether the program is "funded" in some sense. Within each of those measures are variations on the concepts that can give fairly disparate results.

In the end, actuarial analysis is limited, however, to a relatively narrow analysis of one program at a time, without consideration of the rest of the federal budget or the economy.

Actuarial analysis can reveal whether a program as designed appears to be stable over time. It cannot tell you, however, if the program is sustainable over time--whether the federal budget or the U.S. economy will support the level of transfer of resources from the working population to the retired population.

A second approach to the analysis involves the programs' interaction with the rest of the budget. In the case of Medicare, the interactions are direct because it is "on-budget" along with other non-Social Security programs. Even Social Security, although technically off-budget, can have striking on-budget effects. But whether a program is on- or off-budget, it is the combined effects of all taxes and spending that determine the federal government's impact on the economy--on whether public debt is increasing or decreasing, for example.

Mr. Chairman, a quick example of the difference between actuarial and budgetary accounting might be helpful.

If you chose to transfer general revenues--say, from the on-budget surplus to the Medicare Part A trust fund--the two kinds of analyses would yield very different conclusions. Under an actuarial approach, the trust fund balance, and therefore its projected solvency and unfunded liability, would all be improved--if the transfer was large enough, the trust fund could be made "solvent" forever.

The trust fund would look better because there would be more official government debt credited to it. That debt and any interest on it, however, would have to be redeemed in the future by raising taxes, cutting spending elsewhere in the budget, or borrowing from the public--effects much the same as those occurring if there had been no transfer at all.

Another obvious example is the construction of the Part B trust fund. It is actuarially sound, or "adequately financed," but only because it has an unlimited draw on the general funds of the Treasury. Again, the trust fund appears sound, but the growth in Part B spending has direct and potentially dramatic effects on the rest of the budget and the economy.

Last, Mr. Chairman, these programs are susceptible to economic analysis--the interaction of the programs and the economy.

Let me give one important example. This chart, which the Committee has seen before, represents CBO's current best projection of the amount of resources we baby boomers will consume after we retire. We will consume in just these three programs almost as much of the economic output in 2030 as does the entire federal government today. That result is driven by the fact that we will almost double the number of retirees while the number of workers barely increases.

This measure depends on only two factors: the size of the economy and the amount of resources obligated for retirees. It has nothing to do with the existence of a trust fund or any balances within it. It does not matter if the program is solvent or has incurred unfunded liabilities. And the only way to alter this future is to alter one of the two factors--change the size of the economy or the amount of benefits.

By way of summary, let's compare the actuarial, the budgetary, and the economic effects of the time when dedicated revenues to either of these trust funds no longer cover expenditures for that year. In yesterday's reports, that year for both programs happens to be 2016.

In both cases, the actuaries estimate that there will still be positive and growing trust fund balances in 2016. Therefore, the actuarial analysis would suggest ample resources to meet obligations.

The budgetary analysis would denote the transfer of general revenues to the trust funds as an intragovernmental interest payment. Those general revenues could not be used for other spending or debt reduction; indeed, the transfers would have to be funded by a tax increase, reductions in other spending, or the Treasury's issuance of debt.

Similarly, the economic analysis would pose the question, where's the cash? The Treasury would have to have the cash to honor the checks sent to retirees and medical providers. It could get the cash in only three ways: cut other spending, raise taxes, or borrow from the public. This analysis also suggests that it doesn't matter whether there are balances in a trust fund or, indeed, whether there is a trust fund at all--the cash would have to be generated to cover any shortfall in revenues.

Mr. Chairman, until now, we've been discussing how to finance the promises made to retirees, but a clearer picture emerges if we think of these long-tailed programs in terms of consumption--how the elderly spend the money. After all, facilitating consumption is the purpose of the transfers.

When I retire, I will use Social Security funds to buy groceries, clothes, transportation, and so forth--most of which will be produced about the time I use it. In other words, I will be competing with my children and grandchildren for the goods they are producing. What I eat, what I wear, what I drive, they cannot. That is why measures such as program spending as a percentage of GDP may be more relevant and "real" than trust fund or actuarial balances.
 

---