Skip Navigation

This study was prepared by Roberton Williams and Roald Euller of the Human Resources and Community Development Division and Carmela Pena of the Human Resources Cost Estimates Unit of the Budget Analysis Division, under the supervision of Nancy M. Gordon, Martin Levine, and Charles Seagrave. Questions regarding this analysis may be addressed to Roberton Williams or, for budget issues, Carmela Pena.
 
 

SUMMARY

INTRODUCTION

OPTIONS TO CURTAIL COLAS

Curtailing COLAs in All Cash Transfer Programs
Curtailing COLAs Only for Non-Means-Tested Cash Transfer Programs
Further Limiting Effects on Low-Income People

ANALYTIC APPROACH AND LIMITATIONS

Methodology
Limitations of the Analysis

EFFECTS OF CURTAILING COLAS

Budgetary Effects
Effects on Program Recipients
Comparison of Effects
Tradeoffs Between Budgetary Savings and Effects on Beneficiaries

APPENDIX: EFFECTS OF OPTIONS ON POVERTY GAPS AND RATES
 
 

Reducing or temporarily forgoing cost-of-living adjustments (COLAs) in federal cash transfer programs could substantially reduce government expenditures by providing lower benefits than would occur under current law. While most of the savings would result from benefit reductions for people who are not now classified as poor--that is, those with incomes above the official poverty thresholds--limiting COLAs would also reduce the incomes of significant numbers of low-income people and would move some of them into poverty. COLA reduction options could be structured to mitigate the impact on low-income people, however, with varying effects on the net budgetary savings.

This paper examines the potential savings and the impact on low-income people of four approaches to curtailing COLAs in cash transfer programs. The first approach would reduce or eliminate COLAs for one or more years for all current recipients under all cash transfer programs. The second alternative would exempt the means-tested programs from COLA reductions. The final two alternatives would provide additional protection for low-income recipients.

While many methods could be used to protect the incomes of the poor and the near-poor, the two examined here would provide COLAs for some Social Security and Railroad Retirement benefits. The "Poverty COLA" approach would increase program payments only if annual benefits based on a single earnings record were below the poverty threshold, while the "COLA Cap" alternative would grant COLAs on the first $5,000 of annual benefits based on a single earnings record.¹ Although both options are designed to protect low-income beneficiaries, they would also provide benefit increases to some recipients with total incomes well above the poverty line. Further, these options would be difficult to administer, since they would require that the Social Security Administration provide COLAs to some but not all beneficiaries, or for only part of most recipients' benefits.

Within each approach, four specific options are considered. Benefits could be frozen at current levels for one or three years, or currently legislated COLAs could be reduced by three percentage points, again for one or three years. As one would expect, so long as inflation exceeds 3 percent each year, the benefit freezes would result in larger budgetary savings and greater effects on beneficiaries than the COLA reductions. Also, longer periods of curtailing COLAs would have greater impacts than shorter periods.

The Summary Table shows the potential tradeoffs between budgetary savings and effects on low-income program participants for a one-year benefit freeze. For example, a one-year benefit freeze for all non-means-tested cash transfer programs would reduce net federal outlays by $43 billion over the next five years but would also reduce total income to the poor by about $400 million per year and could cause 420,000 additional people to fall into poverty. The Poverty COLA option would generate 25 percent smaller budgetary savings but would have only about one-third the adverse effect on the poor and near-poor.² The COLA Cap approach would protect even more low-income people, but, because it would provide COLAs to many more people with higher incomes, would yield much smaller savings.

The results of this analysis should be viewed with caution for a number of reasons. First, budgetary savings and estimated impacts on poverty statistics are not directly comparable because they are based on different data sources. Second, effects on beneficiaries reflect the population as it was in 1983, not as it will be in the future when COLA changes might be implemented. Third, the Bureau of the Census definition of poverty is used; since the value of in-kind benefits such as food stamps or housing assistance is excluded from income, official poverty statistics may overstate need. Fourth, severe data limitations mean that "the beneficiary impact analysis can only be indicative of the actual effects. Other reasons for caution are presented in the text.

This document is available in its entirety in PDF.

1. Under both the Poverty COLA and COLA Cap options, the benefit measure used to determine eligibility for and size of the COLA would be total benefits--both primary and dependents'--based on a single earnings record; that is, the Social Security/Railroad Retirement record showing the earnings of one worker.

Under both the Poverty COLA and the COLA Cap, some families would be given COLAs on more than the amounts of benefits described above. For example, if both members of a married couple were beneficiaries as a result of their own earnings, under the COLA Cap, the couple could get COLAs on up to $10,000 of annual benefits. On the other hand, a married couple receiving benefits as a worker and dependent spouse (that is, based on the earnings of only one worker) could get a COLA on no more than $5,000 of annual benefits. This approach would be necessary because the Social Security Administration cannot determine whether primary beneficiaries are in the same family.

2. Actual increases in poverty gaps and rates would be somewhat smaller than shown here for the Poverty COLA and the COLA Cap. See text for details.