Partial Benefits in the Social Security Disability Insurance Program

• Date: 01 June 2015
• DOI: http://doi.org/10.1111/jori.12028
• Published date: 01 June 2015
• Author: Na Yin

PARTIAL BENEFITS IN THE SOCIAL SECURITY DISABILITY

INSURANCE PROGRAM

Na Yin

ABSTRACT

The current U.S. Social Security Disability Insurance program is an all-or-

nothing system that has been criticized for creating strong work

disincentives. In an empirically grounded and calibrated life-cycle model,

I simulate behavioral responses to a partial disability benefit system, a policy

alternative to the current program, which allows individuals to claim partial

disability and combine earnings with disability benefits. Simulation results

show financial savings for the program as well as welfare improvements for

individuals with disabilities.

INTRODUCTION

The Social Security Disability Insurance (DI) program is the primary public long-term

disability insurance program for disabled workers in the United States. It is intended

to insure against substantial loss of earnings capacity due to severe health problems.

According to the Social Security Administration (SSA), $128.9 billion was paid to 10.6

million beneficiaries (disabled workers and qualified dependents) in 2011. The

program has grown significantly over time and is likely to grow faster as baby

boomers reach ages with increased disability rates and because increasing life

expectancy raises the duration of benefit receipt among those entering the DI

program. As SSA trustees have pointed out in their annual reports (2012, p. 4), the

projected growth of the DI program will endanger the solvency of not only the DI

Na Yin is an Assistant Professor of Economics and Public Affairs at the School of Public Affairs,

Baruch College, City University of New York. The author is also a faculty associate at CUNY

Institute for Demographic Research. Na Yin can be contacted via email: na.yin@baruch.cuny.

edu. Yin would like to extend deep gratitude to Hugo Benı

´tez-Silva for great advice and

guidance throughout this project. Acknowledgment also goes to Selcuk Eren, Sanders

Korenman, Cordelia Reimers, Steve Schmidt, Kalman Rupp, David Stapleton, and two

anonymous referees for their helpful comments and suggestions. Yin would like to

acknowledge the financial support from the Center for Retirement Research at Boston College

and hospitality provide by RAND at an early stage of the project. Any remaining errors are my

own.

© 2014 The Journal of Risk and Insurance. 82, No. 2, 463–504 (2015).

DOI: 10.1111/j.1539-6975.2013.12028.x

463

program but also the Old Age (OA) program if it necessitates a reallocation of Social

Security payroll tax revenues between the two programs. Since the inception of DI, a

series of policy changes has been implemented to improve work incentives and

program solvency, but the effects of those reforms have been limited.

1

In this article, I examine the effect of a partial disability benefit program, a policy

alternative that fundamentally changes the DI program and that could potentially

better foster work among the disabled. This policy alternative, which offers partial

benefits to disability applicants who can still work, would be an important

complement to the “$1 for $2 Benefit Offset” program that SSA is currently

evaluating in a national demonstration project.

2

The Benefit Offset policy gradually

reduces the full benefit for DI beneficiaries who return to work. In effect, the Benefit

Offset policy offers partial benefits ex post (to current beneficiaries who return to

work), while the partial benefits discussed in the present article are offered ex ante (to

applicants who still work). The combination of the Benefit Offset policy, an ex post

partial benefit system, with an ex ante partial benefit system, would result in a

“symmetric” DI system that could curb opportunistic behavior and promote

continuous attachment to the labor force among the disabled.

Public contributory disability programs with partial benefit options are not at all

unusual in OECD countries.

3,4

For example, Sweden implemented a 1SEK for 2SEK

Offset policy in January 2010. As the Swedish Minister of Social Security, Husmark

Pehrsson (2009), pointed out, a policy of active social inclusion must “contain two

important elements—first, early intervention during the sickness spell, and second, a

policy to remove barriers to reenter the labor market after a period of sickness or

unemployment. Only if we combine these two strategies will it be possible to

maintain a generous welfare system and a high labor force participation rate

simultaneously.” The promise of a partial benefit system is the simultaneous

promotion of support to disabled individuals and program solvency through the

encouragement of labor force participation whenever possible.

1

For details on work incentives, see the “Annual Statistical Report on the Social Security

Disability Insurance Program,” available at www.ssa.gov.

2

SSA has recently completed pilot demonstration projects on the Benefit Offset policy in four

states (Connecticut, Utah, Vermont, and Wisconsin) and found little evidence of change in

participants’ employment behavior. However, limitations of the pilot project design and

implementation made inference about the effects of the Benefit Offset policy difficult. A

national demonstration project, the Benefit Offset National Demonstration or BOND, is now

under way. See www.bondssa.org and www.ssa.gov/disabilityresearch/demos.htm for

details of the BOND project. Weathers and Hemmeter (2011) describe the BOND pilot projects

and present preliminary results.

3

See Appendix A for a discussion of disability benefit regulations and recent developments in

OECD countries.

4

Partial benefits are offered in some other disability programs in the United States, such as

workers’ compensation and Department of Veterans Affairs (formerly Veterans Administra-

tion) systems. See Bhattacharya et al. (2010) on workers’ compensation rating systems.

464 THE JOURNAL OF RISK AND INSURANCE

Introducing Partial Benefits Into the DI Program

The current DI program is an all-or-nothing system that implements a full disability

definition: eligibility requires an individual to demonstrate an inability to engage in

any substantial gainful activity for a long period of time. This definition of disability

has remained unchanged for nearly six decades. Yet the world has changed in many

important respects. Employability of many disabled individuals has been improved

by advances in the medical capacity to treat health conditions and rehabilitative and

accommodating technology, changes in the composition of medical conditions of

disability applicants, and the nature of work.

5

These changes have renewed interest in

the work disincentives associated with the DI program.

Moreover, disability (or the extent of work limitation) by nature is more a continuum

than a dichotomy, as some people have severe health limitations that prevent

employment (the fully disabled), while others have specific health problems that

interfere with work but do not necessarily make work impossible (the partially

disabled). The binary Social Security disability definition, coupled with program-

matic restrictions on work among recipients, is likely to result in high welfare costs for

those who retain some work capacity but can hardly make ends meet by relying solely

on their labor earnings. These individuals may either continue to work despite health

limitations or drop out the labor force and apply for disability benefits. Benı

´tez-Silva

et al. (2004, Table I, p. 651) use longitudinal household survey data to estimate the

disability rate by the final DI award decision. They find that 21.5 percent of DI

applicants who are finally awarded benefits report no disability or work limitation

and 58.8 percent DI applicants who are finally rejected report severe work limitations.

The structure of a dichotomous system also poses challenges for disability examiners,

especially in marginal cases. For example, a severely limited individual might be

deemed eligible if her/his disability is close enough to the strictest disability

definition. Yet a similar case may be rejected as not satisfying the disability severity

requirement.

6

Introducing partial disability eligibility with partial benefits could better match

individuals’ severity of work limitations and possibly improve program efficiency as

well as individuals’ welfare. The Social Security Advisory Board reports have twice

recommended (2003, 2006) enhancements of the current all-or-nothing disability

eligibility through a program to recognize varying degrees of disability. Under a

5

For example, over the last 20 to 30 years, the DI caseload has shifted from one characterized by

individuals with circulatory, neoplasms, and infectious diseases to one dominated by

individuals with mental and musculoskeletal impairments. (SSA’s Annual Statistical Report

on the Social Security Disability Insurance Program, various years.)

6

Several studies (Nagi, 1969; Smith and Lilienfeld, 1971; Benı

´tez-Silva, Buchinsky, and

Rust, 2004) have estimated the classification errors in the DI program: award errors, admitting

an applicant when s/he is not sufficiently disabled, and rejection errors, denying an application

when the applicant is fully disabled. Around 20 percent of DI awardees are estimated to be not

disabled enough for the benefits, while the rejection errors vary widely from 20 to as high as 60

percent.

PARTIAL BENEFITS IN THE SOCIAL SECURITY DISABILITY INSURANCE PROGRAM 465