Retirement Research Consortium

Households that delay claiming Social Security are, in effect, making additional purchases of the Social Security annuity. Theoretical calculations show the delayed claiming is optimal, even for high mortality households. Yet most claim well before the theoretically optimal age. This paper investigates whether subjective mortality beliefs contribute to the prevalence of early claiming. The value of delay depends not only on life expectancy, but also on the degree of uncertainty surrounding the age of death. Using data from the Health and Retirement Study, we show that women approaching retirement understate their probabilities of surviving to age 75 by an average of 10 percentage points, whereas men's forecasts are, on average, correct. But both men and women exhibit greater confidence of their ability to forecast their age of death, relative to the predictions of life tables. But these subjective mortality beliefs have little effect on the value of Social Security or the optimal claim age, and cannot explain the prevalence of early claiming. We also find that self-assessed survival probabilities do not predict survival after controlling for health and socio-economic status, indicating a potential for medical underwriting to reduce adverse selection in the annuity market.

Tax-deferred savings accounts such as IRAs are the major policy tool for increasing savings in the U.S. Unfortunately, the evidence on the impacts of these accounts on saving is mixed, largely because of the lack of good wealth data on a large population of individuals. In this paper, we use tax data from Denmark that provide accurate measures of wealth for all households in a long panel. We study the impacts of a tax reform in 1999 that significantly altered the tax-advantage of pension contributions on the quantity of pension contributions. In this paper, we demonstrate that this reform had large and statistically significant effects on pension contributions. These results motivate future research to evaluate the aggregate impact of the reform on total savings and wealth accumulation.

We investigate expectation formation, using the 1994–2008 Health and Retirement Study (HRS), merged to a number of variables on local and national macroeconomic conditions. We find that individuals update their expectations about the future economy in response to these macroeconomic conditions. Respondents revised their expected probability of a major depression upward substantially during 2008, the onset of the current economic downturn and on average during the entire survey period when their state's unemployment rate was higher and when average monthly stock prices fell. Increasing stock prices are also associated with a decreased expectation of future stock price increases, which is consistent with beliefs of mean reversion in stock prices. Furthermore, expectations of double digit inflation increase when job gains increase, perhaps reflecting beliefs of a tradeoff between unemployment and inflation. In many cases, we find that college graduates' expectations are more responsive than those of less educated individuals. We also find that macroeconomic conditions are correlated with subjective probabilities of one's future labor supply. House price increases are associated with a decreased expected probability of working past age 62 and higher unemployment is associated with an increased expected probability of working past age 65. The unemployment rate effect is larger for less-educated individuals, while the house price effect is larger for more-educated individuals. This may be because college graduates are less likely to be affected by business cycle fluctuations, and more likely to be homeowners. Together, these results provide strong evidence that expectations are shaped by fluctuations in the macroeconomy.

This paper estimates the short-run impact of the Pension Protection Act of 2006 (PPA2006) on holdings of employer stock in defined contribution pension plans. PPA2006 allowed participants in plans with employer stock to diversify their holdings. However, stand-alone ESOPs, i.e., those that do not allow employee elective deferrals or after-tax contributions, were exempt from this provision. Using detailed Form 5500 financial data for stand-alone ESOPs and those that allow employee elective deferrals or after-tax contributions, so-called KSOPs, from 2003-5 (before) and 2007-9 (after) the PPA and a quasi-experimental empirical framework, two primary empirical findings emerge. First, the share of plan assets in company stock fell 7 percentage points for KSOPs, because of the diversification provisions in PPA2006, a substantial decline. There was no change in holdings for stand-alone ESOPs. Second, most of the decline occurred in plans that had between 25–50% of plan assets in employer stock. Nonetheless, in 2009 still two-thirds KSOPs had more than 10% of assets in company stock, the statutory limit for defined benefit pension plans.

This paper uses the substantial increase in prescription drug insurance coverage from the adoption of Part D to generate new estimates the impact of coverage on drug utilization and cost-related non-adherence. The analysis uses detailed panel data on the elderly before and after the implementation of Part D drawn from the 2005 and 2007 Prescription Drug Study (PDS), administered as a supplement to the Health and Retirement Study (HRS), a large nationally representative survey of Americans aged 50 and older. Fixed effect estimates suggest that gaining coverage results in a 15% increase in utilization. These results are consistent with the lower end of estimates in the literature. Gaining coverage also is associated with a 20–50% reduction in the incidence of cost-related nonadherence. However, even among the uninsured, only a relatively small proportion of drugs (12%) are associated with episodes of cost-related non-adherence. So, these large reductions apply to a small slice of all drugs.

SSI and SSDI, the two work disability programs administered by the Social Security Administration (SSA), have been marked by concerns about target efficiency since their inception. This study uses SSA administrative data linked with National Health Interview Survey data (NHIS) to examine health status, labor force participation at time of NHIS interview, and linked mortality data to examine mortality during the period following NHIS interview. The self-reported health status data present two strong and consistent patterns: denied applicants report being in considerably worse health than non-applicants, and beneficiaries appear to be sicker yet. In logit models among disability beneficiaries, women are significantly less likely to report excellent/very good health, but race has no significant effect. While being female decreased the probability of good health, it has no significant effect on the probability of reporting no work limitation at time of interview among beneficiaries. Although race was not significant in the model of self-reported health, both Hispanics and non-Hispanic blacks are significantly more likely to report no work limitation at time of interview. This study has important limitations. NHIS respondents who link to the SSA administrative data may not be representative of all individuals with disability application histories. In addition, individuals must live long enough after disability determination to be drawn into an NHIS sample, and these results reflect the experience of that subsample of disability applicants who do not die during the determination process or soon thereafter.

Social Security Disability Insurance (SSDI) applications and receipts vary greatly by state. This paper investigates the extent to which this geographic variation in SSDI applications reflects differences in health, demographics, and employment characteristics, state policies,, and politics. We find that demographic, health, and employment characteristics of the state have the greatest effect on state-level variations in SSDI application rates, explaining over 70 percent of the variation. State policy concerning mandated employer-sponsored disability insurance (also known as temporary disability insurance or TDI) has a small negative effect on overall SSDI applications. This finding supports the principle underlying many recent SSDI reform plans: temporary disability insurance coverage could save the SSDI program considerable funds in the long-run. Further, when we look to explain variation within a state, we find that state changes in health insurance regulation are negatively correlated with the SSDI application rate. This could be an indication that the Affordable Care Act (ACA) may have spillovers to the SSDI program.

As policy makers consider changes to the Social Security and Medicare programs, it is important to know how the economic status of the older population was impacted by the financial crisis and the Great Recession. Because consumption is the best measure of material well-being we estimate the effect of the recession on spending in the older population based on data from the Health and Retirement Study and the Consumption and Activities Mail Survey. Our method is to compare spending change between 2007 and 2009 with spending change over three previous "normal" transitions: 2001–2003, 2003–2005 and 2005–2007. We find that spending declined from 2007 to 2009 by 3.6 to 7.0 percentage points more than would be expected based on the changes from the earlier two-year time periods where the range is due to differing statistical methods. We compared changes in spending by stock owners with changes in spending by nonowners both in the years 2001–2007 and in 2007–2009. We found that owners reduced spending by an additional 3.8 to 6.2 percentage points compared with nonowners over the 2007 to 2009 period. These results are consistent with an effect of stock market losses on spending, but there are other possible contributing factors such as losses of housing equity. Those who became unemployed between 2007 and 2009 reduced spending by about 19%, indicating a lack of complete insurance against unemployment. We conclude that the shocks generated by the Great Recession substantially affected the economic position of the older population as indicated by the reduction in spending.

Data from the Health and Retirement Study is used to investigate how the "Great Recession" affected the wealth and retirement of the near retirement population, ages 53 to 58 in 2006. Due to falling house prices, their retirement wealth fell by 2.8 percent between 2006 and 2010. In contrast, in past years members of older cohorts had added 5 percent to their assets over the comparable age span. Cushioned by Social Security, the retirement wealth of poorer households was less affected by the recession. Labor market effects of the Great Recession were more modest. The changes in full-time work and retirement are similar to the changes observed for members of older cohorts at comparable ages.

Both unemployment insurance (UI) extensions and the availability of disability benefits have disincentive effects on job search. But UI extensions can reduce the efficiency cost of disability benefits if UI recipients delay disability application until exhausting their unemployment benefits. This paper investigates whether UI eligibility, extension, and exhaustion affect the timing of disability application and the composition of the applicant pool. Jobless individuals are significantly less likely to apply to Social Security Disability Insurance (SSDI) during UI extensions, and significantly more likely to apply when UI is ultimately exhausted. Healthier potential applicants appear more likely to delay, as state allowance rates increase after a new UI extension. Though simulations find that extensions do not decrease SSDI costs, the benefits of UI extensions still may be understated—permanent disability benefits are diverted to shorter-run unemployment benefits and, potentially, new jobs, while easing the burden on the nearly-insolvent SSDI Trust Fund.