Involuntary retirement, U.S. Social Security Program participation and the great recession.

Author:Seligman, Jason S.

    This article considers involuntary retirement and Social Security participation patterns ahead of and since the Great Recession. Involuntary retirement covers two interacting phenomena: economic and health related dislocations. Both phenomena inform SSA program participation. In looking at involuntary retirement, I find that three out of ten retirees reported involuntary retirement over the period of 1992-2011, of which roughly half were primarily related to health and half were primarily related to other dislocations--most notably job loss and business closure. Following the onset of the Great Recession, there have been great increases in reports of involuntary retirement over the 20082011 period. Over this period involuntary retirements grew roughly 78% faster than voluntary retirements. (1) Non-health related involuntary retirements grew fastest of all--in line with a large negative economic shock. Incomes have fallen for all retiree groups. Among those hardest hit, were folks who retired receiving no benefits (primarily under age 62) citing health as a reason; post retirement income for this group fell 38% while their ranks grew 270% faster than average. (2)

    The U.S. retirement and disability programs have both seen increases in participation and reductions in revenues as payrolls declined, exacerbating financial pressures on the DI system and accelerating the rate at which combined trust fund assets are declining towards zero. The programs are likely to be reformed in the next few years as the DI trust fund is exhausted. This article is intended to help readers understand the relative importance of these programs for various types of retirees as reform designs are contemplated.

    Findings here are in line with programmatic reform which marginally reduces the importance of tagging and instead increases the benefit gradient to more continuously encourage workforce participation among those whose health and economic prospects allow it. As outlined in Akerlof (1978), "tagging" welfare benefits (offering benefits to only those in discretely defined categories, such as those defined to be disabled) can be efficient; lowering inframarginal tax costs and/or affording increases in benefit generosity. However this strategy can leave many who are proximate to definitional thresholds with much lower standards of living than persons more or less like them challenging notions of horizontal equity and fairness.

    Social Security is a hybrid program, comprised of both earned and welfare benefit components. (3) In particular, ahead of the Early Retirement Age (ERA), Old Age Insurance (OAI) is generally unavailable. Ahead of the Full Retirement Age (FRA), more generous benefits are granted to those tagged as Disabled. However, receipts for DI are inadequate to sustain full benefits past 2016. In particular both DI applications and awards have increased substantially since the end of the 20th century--in recessionary and growth periods alike. (4)

    Theoretically, the distinction between health and economic causes would appear to inform program selection, health oriented dislocations should target DI application and other elders exits should target the U.S. retirement program--the Old Age and Survivors Insurance (OASI) program. In fact both programs appear to contribute to the welfare of those affected by both types of involuntary retirement. Empirically, this paper finds that both voluntary and involuntary retirees primarily participate in OASI at ages 62 and beyond. (5)

    Involuntary retirees can receive DI benefits if they are in poor health, however the program's definition of disability is restrictive such that not all applicants reporting poor health qualify. For those who cannot collect DI benefits an alternative is to collect OASI benefits early--much as the voluntary do. This piece examines the extent to which this strategy is used by those in poor health.

    The large and increasing proportion of involuntary retirement reports has important policy implications for fundamental questions that we as a society need to revisit regarding retirement. What are reasonable Full and Early retirement ages? How are these related to the incidence of disability and application to the DI program? What U.S. programs do individuals select, conditional on circumstances, and how do they perceive their retirement? What are their health and income experiences in retirement? This article offers as its main contribution how involuntary retirement interacts with OASI and DI benefits--improving understanding of the way in which individuals engage U.S. Social Security Programs. Three basic methods of investigation are employed:

  2. estimation of differential retirement hazards,

  3. analysis of program uptake, retirement satisfaction and income by characteristic/type and,

  4. estimation of impacts of the Great Recession on each of the above.

    Hazards consider differences in voluntary and involuntary retirement behavior. Uptake distinguishes between reported retirement type and Social Security program type. Taking account of retirement satisfaction helps to control for outcomes which may appear concerning but are possibly more holistically satisfactory--critical to any understanding of what "involuntary retirement" is. Accounting for the impact of the Great Recession generally helps place cohort-specific late career macroeconomic shocks in the context of work experiences and the programs.


    Social Security's OASI program serves a population of retired workers and their eligible dependents. The "Full" Retirement Age (FRA) is but one age at which OASI benefits can be received. The earliest age that a discounted benefit can be based on one's own work record is age 62. At age 65 Medicare benefits are generally available. The DI program serves those unable to work (participate in a meaningful Substantial Gainful Activity) ahead of the FRA. Successful application to the DI program generally comes with accelerated Medicare coverage. Thus, at ages ahead of the FRA the DI program offers more attractive benefits--as covered in Duggan, Singleton, and Song (2007). The DI program is important to consider as a potential substitute for OASI or other retirement programs in the context of involuntary retirement.

    Theoretically, OASI and DI program participation decisions may occur exclusive or inclusive to one another, with inclusive decisions motivated via the benefit differentials and the opportunity for dual application past age 61. Historically the economic assumption of rational actors operating under full information has motivated adoption of an inclusive {OASI; DI} multinomial decision model in theoretic work, for example Diamond and Sheshinski (1995). However, empirically, the interaction between these programs is not well understood and anecdotally the likelihood of dual application is somewhat a function of the field office an applicant happens upon. Some offices may be more likely to explain dual application options than others. Indeed a literature on the heterogeneity of disability determination outcomes can be traced through Black, Daniel, and Sander's (2002) work on community differentials while Maestas, Mullen and Strand (2013) have more recently documented examiner-specific effects. For those who apply online, the SSA's Benefit Eligibility Screening Tool (BEST) is a likely reference for applicants weighing options. It provides another possible dimension for heterogeneity in empirically observed behaviors. For example, applicants using BEST who answer "yes" when asked if they are disabled, and are insured are given the opportunity to apply for either OASI or DI benefits, but are not offered a dual application option. However, some response patterns require the applicant to visit a field office. In short, both the degree to which program participation decisions are inclusive and the ease with which a potential DI applicant can estimate a probability of acceptance conditional on health are debatable, due in part to non-standard variance across the information and application processes.

    Because SSA's Disability Insurance program can be financially attractive when compared to traditional retirement ahead of the FRA one may suspect that persons would prefer to describe themselves as disabled when they exit the labor force ahead of the FRA. This may be true in particular when their early exit causes a discrete drop in consumption, including consumption of health insurance. Lachance (2008) offers a model whereby disability (modeled as an exogenous shock), frustrates efforts to smooth lifecycle consumption in the traditional model offered by Samuelson (1958). The Lachance model can be driven by shocks that increase costs (disutility) of work or by shocks that decrease benefits from employment. Further, the model can be generalized from disability (health shocks) to job dislocations in which firm, industry, or union specific returns are structurally disrupted. In such situations future employment might reasonably be expected to yield lower wage and/or benefits, and perhaps higher costs. This is due to depreciated human capital, which when coupled with the loss of social-professional networks, impact the expected utility and productivity of search.

    In the Lachance model an induced disutility shock drives negative outcomes (for example: disability motivates labor force exit and a permanent drop in consumption). Of course empirically, not all permanent drops in consumption can be linked to disability, and not all are entirely involuntary. Those near threshold savings values for exit may feel they make voluntary choices motivated by health while others facing similar health shocks express different feelings.


    Against this backdrop there is increasing awareness that Disability Insurance awards appear to grow fastest during recessions, as...

To continue reading