The effect of means-tested income support for the elderly on pre-retirement saving: evidence from the SSI program in the U.S.

https://doi.org/10.1016/S0047-2727(98)00004-8Get rights and content

Abstract

We attempt to draw inferences about behavioral responses to means-tested income support for the elderly by examining the effects on saving of the Supplemental Security Income (SSI) program for the aged in the U.S. Part of this program provides payments to the poor elderly, operating as a means-tested public retirement program. We exploit state-level variation in SSI benefits to estimate the effects of SSI on saving, using data from the 1984 Survey of Income Program Participation. We find evidence that high SSI benefits reduce saving among households with heads who are approaching the SSI eligibility age and are likely participants in the program.

Introduction

The elderly are one of the exceptional groups in American society with access to a significant cash safety net, a means-tested program called Supplemental Security Income (SSI). Although the potential disincentives posed by means-tested income-support programs for other populations (such as AFDC) have received considerable attention from both researchers and politicians, the potential disincentives posed by such programs for the elderly have largely escaped attention. In particular, the SSI program may create disincentives for saving as individuals approach the age of eligibility, because of both asset and income tests. Given that public policy often attempts to ensure that elderly individuals accumulate sufficient wealth to support their retirement – for example, Social Security and ERISA – it would be ironic if we also have policies in place that discourage saving among the more disadvantaged elderly population that is potentially eligible for SSI. However, there is no empirical evidence on whether SSI actually leads to discernible disincentives to save.1

In this paper, we examine the effects of SSI for the aged on saving.2 The federal government sets eligibility criteria and benefit levels for the federal component of the program, specifying maximum benefit levels for couples and individuals that are reduced by income from other sources, including Social Security benefits and Disability Insurance. (The first $20 of non-means-tested transfer income, the first $65 of earned income, plus one half of remaining earnings, are disregarded in reducing SSI benefits.) Thus, other sources of income influence both eligibility for and potential payments under SSI. Financial resources also affect eligibility. For example, as of 1985 individuals with over $1600 in countable assets, and couples with over $2400 in countable assets, were ineligible.3 In September 1984 (corresponding roughly to the time period covered by our data), there were 1.55 million persons receiving SSI payments who were eligible because of age (1995 Green Book).4

In addition to the federal component of the program, states may supplement federal SSI benefits. For example, in January 1985 the maximum federal benefit was $325 for an individual, and $488 for a couple. The highest state benefit was in California, which resulted in a maximum combined benefit of $504 for an individual, and $936 for a couple.5 In December 1985 the average federal benefit paid was $146 for individuals, and $232 for couples, and the average state supplements were $97 and $257, respectively (Kahn, 1987), with 39% of SSI recipients receiving state supplements.

We exploit the state-level variation in SSI benefits to estimate the effects of SSI on saving. We use data mainly on male householders from the 1984 panel of the Survey of Income Program Participation (SIPP), covering individuals in the 1983–1986 period, to estimate the effects of state SSI supplements via a difference-in-difference approach that controls for variation in saving across states and across different types of individuals. We find evidence that high SSI benefits reduce saving among households with heads approaching the age of SSI eligibility, and who are likely to end up participating in the program.6 This evidence is robust to many (but not all) of the changes in the sample, specification, or definition of variables that we consider. Overall, then, we view the evidence as providing a relatively consistent picture of dissaving effects induced by the SSI program.

Section snippets

The data

We use a sample drawn from the SIPP, which when weighted serves as a nationally-representative sample of households. The SIPP attempts to gather detailed and reliable data on income and welfare program use that are impractical to collect in the larger Current Population Surveys. Households are interviewed every 4 months (each 4 month interval is referred to as a “wave”) for 2–3 years. Most questions are asked retrospectively about the previous 4 months. However, questions about wealth holdings

SSI participation

The descriptive statistics in columns (2) and (3) of Table 1 provide demographic and other information for men aged 65 and over, classified by whether or not they are actual SSI participants. Participants are much more likely to have less than a high school education, and virtually guaranteed not to be college graduates. They are also more likely to be black, and never married, divorced, widowed, or separated. Finally, not surprisingly, SSI participants are more likely to ever have been

Conclusion

The goal of this paper is to draw inferences about the effect of SSI for the aged on pre-retirement saving. SSI for the aged effectively operates as a means-tested retirement income program, in that eligibility depends on financial resources in the form of assets and income. As a consequence, more generous SSI benefits may induce less saving or more dissaving at ages near retirement. We use state-level variation in generosity of supplemental SSI payments to identify the effects of SSI, studying

Acknowledgements

Neumark was a Visiting Scholar at the Federal Reserve Bank of Cleveland (FRBC), and Powers was an Economist at the FRBC, during the period when this research was initiated. Neumark gratefully acknowledges support from NIA grant K01-AG00589. We are grateful to seminar participants at the Cleveland Fed and Michigan State University, and to Alan Gustman, Paul Menchik, James Poterba, Mark Schweitzer, and anonymous referees for helpful comments, and to Kristin Roberts for research assistance.

References (14)

  • Attanasio, O.P., 1993. A cohort analysis of saving behavior by U.S. households. NBER Working Paper No....
  • Gruber, J., Yelowitz, A., 1997. Public health insurance and private savings. NBER Working Paper No....
  • Jianakoplos, N.A., Menchik, P.L., Irvine, F.O., 1989. Using panel data to assess the bias in cross-sectional inferences...
  • A.L. Kahn

    Program and demographic characteristics of supplemental security income recipients, December 1985

    Social Security Bulletin

    (1987)
  • K. McGarry

    Factors determining participation of the elderly in SSI

    Journal of Human Resources

    (1996)
  • P.L. Menchik

    Economic status as a determinant of mortality among black and white older men: Does poverty kill?

    Population Studies

    (1993)
  • Mitchell, O.S., Zeldes, S., 1996. Social security privatization: A structure for analysis. NBER Working Paper No....
There are more references available in the full text version of this article.

Cited by (44)

  • Optimal consumption, investment and housing with means-tested public pension in retirement

    2017, Insurance: Mathematics and Economics
    Citation Excerpt :

    Nevertheless, studies in which the Age Pension is based on a means-tested pension function are limited. Researchers have mainly focused on the economic effects of different means-tested policies and the impact on savings (Hubbard et al., 1995; Hurst and Ziliak, 2004; Neumark and Powers, 1998 to name a few), or how annuitization is affected by the means test (Bütler et al., 2016). In Australia, which is a very special case of means testing, there are two main approaches for modelling Age Pension: expected utility maximization via dynamic stochastic programming and overlapping generation (OLG) models based on dynamic general equilibrium.

  • The effect of expanding Medicaid eligibility on Supplemental Security Income program participation

    2017, Journal of Public Economics
    Citation Excerpt :

    The SSI program's stringent financial eligibility criteria create disincentives for prospective and current beneficiaries to work and accumulate assets (Daly and Burkhauser, 2003). Empirical evidence demonstrates that SSI induces some moral hazard for at least a subset of beneficiaries (Neumark and Powers, 1998, 2000; Powers and Neumark, 2005; Kaushal, 2010). A large body of work on SSDI provides additional support for the idea that disability benefit programs can have work and asset disincentive effects (Gruber and Kubik, 1997; Black et al., 2002; Autor and Duggan, 2003; Chen and van der Klauuw, 2008; Maestas et al., 2013; French and Song, 2014; Moore, 2015; Shu, 2015).

  • Trade-offs in means tested pension design

    2014, Journal of Economic Dynamics and Control
    Citation Excerpt :

    The increase in the taper rate therefore induces individuals to save more or work extra hours. This result is consistent with empirical evidence documented in previous empirical studies, such as Neumark and Powers (1998), Neumark and Powers (2000), Disney and Smith (2002) and Friedberg (2000). This result is also consistent with Sefton et al. (2008), who analyze a calibrated multi-period overlapping generations model of the U.K. However, when the government raises the taper rate from 0.0 to 0.4, there is an increase in capital stock and labor supply.

  • An inter-temporal general equilibrium analysis of the Australian age pension means test

    2011, Journal of Macroeconomics
    Citation Excerpt :

    Simulation results by Sefton et al. (2008) demonstrate that the reduction in the taper rate of the means tested benefits from 100% to 40% implemented in the UK in 2003 would increase labour supply and saving of low-income households but have the opposite effects on high-income households, resulting in higher overall employment and lower aggregate saving. Similarly, Neumark and Power (1998) find empirical evidence that high Supplementary Security Income benefits for the elderly in the US, which are means tested, reduce pre-retirement saving of those who are likely participants in the program. Our simulation results demonstrate that the hypothetical pension policy changes considered have a significant behavioural impact on middle-income households that are income tested in the pre-reform case with the existing means test.

  • Introduction to senior hunger special issue

    2022, Applied Economic Perspectives and Policy
View all citing articles on Scopus
View full text