Sources of change in federal transfer payments to persons: an update.

Holloway, Thomas M. ; Reeb, Jane S.

Sources of Change in Federal Transfer Payments to Persons: An Update

TRANSFER payments to persons by the Federal Government are income payments, generally in monetary form, for which no current services are rendered. From 1970 to 1985, transfer payments, which are components of both Federal expenditures and personal income, increased by about $300 billion--an annual rate of growth of 13 percent. Because their growth rate was higher than that of Federal expenditures (11 percent) and personal income (10 percent), the transfer payments share increased from 30 percent of Federal expenditures in 1970 to 37 percent in 1985 and from 7 percent of personal income to 11 percent during the same period. Over this period, however, these shares did not increase continuously; they fluctuated throughout the period, with a peak in 1983 and a decline thereafter (table 1). The recent decline was largely the result of the recovery from the 1981-82 recession.

This article discusses the sources of change in transfer payments. It uses the same analytical framework and categorizations of transfer payments as an article in the October 1982 SURVEY OF CURRENT BUSINESS on the same subject. The first section describes the framework. The second section provides an overview of the sources of change in total transfer payments. The section presents revised estimates of the sources of change from the earlier article and updated estimates for 1982-85. The third section focuses on the sources of change in the major categories of transfer payments. Emphasis is placed on developments in the 1980's.

Analytical framework

Changes in transfer payments can be attributed to several sources: (1) Automatic cyclical effects, (2) automatic inflation effects, and (3) legislation and other factors. Automatic cyclical effects reflect automatic responses of transfer payment programs to fluctuations in economic activity--fluctuations indicated by changes in the unemployment rate. In cyclically sensitive programs, increases in the unemployment rate increase payments; decreases in the unemployment rate decrease payments.

Automatic inflation effects reflect the automatic responsiveness of transfer payment programs to changes in prices. Inflatio-sensitive programs can be categoraized as indexed or non-indexed. Indexed programs, which include most transfer payment programs, are linked by legislation to changes in a specific price index. Non-indexed inflation-sensitive programs automatically respond to inflation through responsiveness to nominal wages (e.g., unemployment insurance) or price changes of goods and services covered by the program (e.g., medicare).

Legislative changes and other factors reflect discretionary policy actions, the effects of demographic changes, noncyclical growth in real wages, and other factors not attributable to automatic cyclical and automatic inflation effects. The legislative changes and other factors source is derived as a residual by subtracting automatic cyclical effects and automatic inflation effects from changes in total transfer payments. Because it is a residual, the causes of its fluctuations vary from quarter-to-quarter.

Of changes in transfer payments and estimates of automatic cyclical effects (cycle-induced changes), automatic inflation effects (inflation-induced changes), and changes due to legislation and other factors. The table also shows transfer payments in dollar levels and as percentages of personal income.

Cycle-induced changes generally increased transfer payments during recessions and decreased them during recoveries (second panel in chart 2). Chart 3 shows changes in the unemployment rate gap--an indicator of changes in economic conditions. The close relationship between changes in the unemployment rate gap and cycle-induced changes in transfer payments in chart 2 apparent. Sharp cycle-induced increases accompanied the 1974-75, 1980, and 1981-82 recessions; cycle-induced declines accompanied the subsequent recoveries. Because automatic cyclical effects tend to be offsetting over time, the sustained--and often large--increases accompanying the 1980 and 1981-82 recessions were offset by sustained--and often large--declines during 1983-85.

Inflation-induced changes increased transfer payments in all quarters (third panel in chart 2). The sharp upward movements starting in 1975 mainly reflected cost-of-living adjustments (COLA's) to indexed programs--especially to Social Security. The relationship between the inflation are shown in chart 3 and inflation-induced changes in chart 2 is not obvious. The reason is that the COLA's reflect an adjustment in a single quarter based on several earlier quarters of inflation. The lags often exceed 6 months. The indexing provisions of Social Security illustrate the lag relationship. Under current law, the Social Security COLA occurs in January on the basis of the change in the Consumer Price Index (CPI) from the third quarter of 2 years prior to the third quarter of the prior year. Consequently, changes in the inflation rate may not be reflected in inflation-induced changes for a considerable length of time.

The decelebration of inflation during the 1980's that was evident in chart 3 was reflected, with a lag, by smaller inflation-induced changes in chart 2. Nevertheless, inflation-induced changes consistently contributed to increases in transfer payments because prices generally continued to rise, albeit at a slower rate. Consequently, automatic inflation effects, unlike automatic cyclical effects, tend to be cumulative. Current benefit levels reflect not only the most recent inflation adjustments, but the inflation adjustments of the past as well.

Changes in transfer payments attributable to legislation and other factors were very volatile throughout the period (fourth panel in chart 2). In some quarters, changes due to this source could be identified as discretionary policy actions--such as legislated Social Security benefit increases in the early 1970's, the one-time payment under the Tax Reduction Act of 1975, and legislated temporary unemployment benefits associated with the 1981-82 recession. However, in most quarters, changes due to this source were not associated with any one program and reflected a complex mix of many small changes.

Sources of change in major transfer

payment programs

Federal transfer payments may be placed in five categories: (1) Social Security (excluding medicare); (2) other retirement and disability programs, which consist of Federal civilian retirement, military retirement, railroad retirement, workers' compensation, and black lung benefits; (3) medicare: (4) programs with a needs test, which consist of supplemental security income (SSI), food stamps, veterans pension and disability benefits, and the earned income credit; and (5) other programs, which include unemployment benefits, veterans readjustment, veterans life insurance, and military medical insurance. The remainder of this section highlights the sources of change in each of the five categories. For each category, table 3 shows the dollar levels, the dollar levels as percentages of personal income, and the sources of change. Chart 4 shows each category as a percentage of personal income. The table and chart show clear differences in the trends and sources of change among the categories.

Social Security.--Social Security (excluding medicare) was the dominant category throughout the period. In 1985, benefits amounted to $183 billion, and accounted for 50.0 percent of transfer payments and 5.6 percent of personal income. Over the period, Social Security as a percentage of personal income had an upward trend, with two periods of decline or leveling (1977-79 and 1983-85). Although the automatic cyclical effects accounted for some of the change in the category, the automatic inflation effects were much more important. Since indexing of Socil Security began in 1975, most of the change could be attributed to inflation-induced changes. Prioor to that time, the legislation and other factors source--partly reflecting legislated increases--accounted for most of the change. The combination of the deceleration of inflation during the 1980's and the omission of the Social Security COLA in 1983 contributed to much smaller inflation-induced changes in recent years than during 1980-82.

Other retirement and disability programs.--Benefits paid under the five programs in this category amounted to $50 billion in 1985, and accounted for 13.6 percent of transfer payments and 1.5 percent of personal income. As a percentage of personal income, this category increased during the early 1970's, but did not change much since then. None of the programs are cyclically sensitive, but all of them are indexed. As in the case of Social Security, the combination of the deceleration of inflation and shifts in the effective dates of COLA's contributed to inflation-induced changes that were much smaller in recent years than those during 1980-82.

Medicare.--Medicare benefits amounted to $70 billion in 1985, and accounted for 19.1 percent of transfer payments and 2.1 percent of personal income. Over the period, medicare as a percentage of personal income had a strong upward trend with no periods of decline. While the rate of increase in this percentage slowed since 1983, the percentages of all other categories declined. Medicare is not cyclically sensitive, but it does automatically respond to inflation through its sensitivity to health care costs. Inflation-induced changes consistently accounted for one-half or more of the annual changes in medicare benefits since the mid-1970's. Inflation-induced changes in the past few years did not decrease as much from their peak year levels as in the Social Security and other retirement and disability categories. In those categories, inflation-induced changes were noticeably smaller in recent years than their peak year values. Part of the explanation lies in the difference between overall price changes ad medical care price changes. Based on the all items measure of the CPI used to index most of the programs in the first two categories and shown in chart 3, the annual inflation rate was 9.1 percent from 1975-81 and only 4.0 percent from 1981-85. The comparable estimates using the medical care component of the CPI were 9.7 percent from 1975-81 and 8.2 percent from 1981-85. Thus, inflation-induced changes in medicare remained relatively large because increases in medical care prices did not decelerate as much as increases in overall prices.

Programs with a needs test.--Benefits paid under the programs in this category amounted to $35 billion in 1985, and accounted for 9.4 percent of transfer payments and 1.0 percent of personal income. As a percentage of personal income, this category increased noticeably in 1974 with the start of the SSI program. Since then, the percentage had a slight downward trend. The automatic cyclical effects reflect the cyclical sensitivity of food stamps. The automatic inflation effects reflect the inflation sensitivity of food stamps, SSI, and part of veterans pensions. In most years, these automatic effects were not very large.

Other transfer programs.--Benefits paid under programs in this category amounted to $28 billion in 1985, and accounted for 7.8 percent of transfer payments and 0.9 percent of personal income. The category is dominated by unemployment benefits. As a percentage of personal income, the category increased during recessions and declined during recoveries. For example, there was a sharp decline from 1983-84 during the economi recovery. The decline was an important source of the decline in total transfer payments as a percentage of personal income during the same period. Although automatic inflation effects account for some of the change in the category, the automatic cyclical effects are much more important. Because of the importance of the cyclical effects, even the total dollar amount of the category declined in many years.