Skip Navigation

Mr. Chairman, I am pleased to appear today to discuss the treatment of indexed programs in the federal budget. I would like to discuss the case for including indexed programs in the effort to reduce the growth in federal spending this year, and some options Congress might consider if it should decide to reexamine indexing.

Congress is currently engaged in an historically significant effort to limit sharply the rapid growth in federal spending. Between fiscal years 1960 and 1980, federal spending has grown at an annual rate of 3.6 percent in real terms, or from 18 to 22 percent of GNP. This growth has been concentrated in benefits to individuals, which rose from 4.1 percent of GNP in 1960 to 8.9 percent in 1980. The rest of the federal budget has been shrinking--in relative terms--from 14.0 percent of GNP in 1960 to 13.1 percent in 1980.

The bulk of the expenditure on individuals is found in entitlement programs such as Social Security, Medicare, and Medicaid, in which benefit levels are either explicitly or implicitly linked to rises in the price level. The amount of outlays on formally indexed entitlement programs is projected at $195 billion in fiscal 1981, or 30 percent of total federal expenditures. This means that without a change in existing legislation, each 1 percent change in the Consumer Price Index automatically triggers nearly $2 billion in extra benefits.

The reason for indexing these programs was to preserve the real level of benefits against erosion from inflation. Without such protection, nominal benefits lose their purchasing power as prices rise. But in periods of rapid price increase, widespread indexation leads to high fixed costs and makes the management of fiscal policy much more difficult. The rapid rise in the level of federal spending that is the automatic response of indexation makes it harder to bring inflation under control; it even lends momentum to inflation because the beneficiaries have no incentive to adjust their behavior to take account of rising prices. Moreover, since the largest single program of this group, Social Security, is funded by payroll taxes, any increase in funding tends to exert further pressure on the price level by raising business costs. There is also a growing concern as to the fairness of fully indexing federal benefits while the wages of the working population lag behind. All of this is aggravated by the evidence that the Consumer Price Index has been overstating the actual rise in the cost of living, primarily because it has overstated housing costs.

Thus, serious efforts to cut the growth in federal spending will be handicapped if they do not include a reexamination of indexing. If defense spending is to grow, if taxes are to be cut, if interest payments are to be made, and if the budget is to be balanced, the remaining portion of the budget will require extremely drastic cuts unless something is done to moderate the growth of indexed programs.

The Administration has elected not to address the indexing problem in its proposed cuts for fiscal 1981 and 1982--with the exception of converting semiannual indexation formulas to annual ones. Its assumption is that the growth in these programs will diminish in the near future as inflation falls. But the increase in indexed benefits in fiscal 1982 will be determined by price increases that have already occurred. Moreover, there is the risk that inflation may not subside as fast as the Administration expects. If its forecasts are wrong, the cost of maintaining current indexing practices will be a high one.

Addressing current indexation practices this year instead of later has appeal for three reasons:

• Modification of indexation practices would permit spending cuts to be spread more evenly across the budget, avoiding the charge that some groups are being asked to make large sacrifices while other groups remain unaffected.
  
  
• The real wages of the working population have declined in the past two years, while beneficiaries of indexed federal programs have been receiving double-digit increases in their nominal benefits. The resulting transfer of income from one group to another has become one of the more conspicuous by-products of the practice of indexation.
  
  
• The CPI has in the past several years risen much more rapidly than any other measure of final product prices. A tendency to exaggeration and volatility is the product of a flawed treatment of homeownership costs and other shortcomings of the CPI as a cost-of-living measure. Continued use of the CPI in its present form constitutes a liability in the management of federal expenditure.

It might actually be easier to explain and gain acceptance of an indexing change this year than at some future time when inflation was lower, real wages were rising and the pressure for across-the-board cuts had diminished. There are several options for changing indexing practices that the Congress might consider:

1. Increases in benefit levels might be capped, either through an automatic procedure--such as limiting the increase in benefits to, say, 85 percent of the rise in the CPI, or through a more discretionary process such as is employed by the President and the Congress for the determination of increases in the pay of federal workers.
   
   
2. The CPI could be modified to improve its validity as a measure of the cost of living. Since it takes time to develop a new measure, an existing one might be chosen which contains fewer distortions than the current one--such as the experimental CPI-X1 which modifies the treatment of homeownership, or another consumption price measure such as the PCE chain-weighted index.
   
   
3. Benefits could be indexed to the increase in either a wage measure or a price measure--whichever rose the least.

There are arguments both pro and con for all of these choices. Simple capping would be an easy way of assuring immediate budgetary savings, but some may be concerned at its arbitrariness. A discretionary capping would permit a more flexible approach that would be able to respond to extraordinary events such as price increases in oil and other energy supplies. Substitution of another index in place of the current CPI is attractive because it deals with one of the fundamental causes of the current situation. But simple substitution would not save money if the current CPI were to begin distorting in the other direction--as it is capable of doing. Although future price changes might be more suitably measured by a different index, it is costly to remain locked into a very high level of benefits that is the cumulative result of past distortion in the CPI. The "switching" proposal--that is, to a wage or a price index, whichever rises least--has much to recommend it as a safety device. When real wages decline the financing of federal expenditures is made more difficult. A wider sharing of the burden of an economic setback lessens the burden on any one group. Still, there will be a redistribution of the burden as real wages regain their former level unless beneficiaries are allowed to catch up as well.

CBO has made some preliminary estimates of the cost savings that would result from three of these indexing options if applied to the Social Security program. These estimates were published in February in Reducing the Federal Budget; Strategies and Examples. The estimated savings for fiscal 1982 in limiting increases to 85 percent of the Consumer Price Index are $2.8 billion, in shifting to the PCE chain index $1.9 billion, and for the switching proposal without a catch-up $3.8 billion. The cumulative five-year savings for these proposals are estimated at $43.9 billion, $10.2 billion, and $25.2 billion respectively. A breakdown of the estimated savings under each of the three proposals is provided in a table attached to my written statement. If these indexing changes were adopted for other indexed entitlement programs, the savings would be correspondingly enlarged.