ISR|SRC|Michigan Retirement Research Center|UofM
Mailing List | News | Contact Us
Follow onTwitter
Main photo Promoting research on retirement and Social Security policy

Volume 15, Issue 3 - September 2015

Investigators present findings at Retirement Research Consortium’s annual meeting

The event also featured work by NBER and Boston College researchers. Panels focused on retirement incentives, asset accumulation, health, health reform and retirement, employment and retirement, immigration, and saving behavior.

Summaries of MRRC researchers’ presentations

David Neumark (University of California at Irvine), “Does Eliminating the Earnings Test Increase Old-Age Poverty of Women?”

The Senior Citizens’ Freedom to Work Act of 2000 repealed the Social Security retirement earnings test (RET), which set a threshold for wages over which SSA benefits would be reduced for beneficiaries who had reached full retirement age (FRA). Neumark and co-author Theodore Figinski (U.S. Department of the Treasury) use the 2000 reform to test whether the RET elimination had adverse consequences for the incomes of older women whose own or spouse’s behavior was affected by the RET elimination. Neumark and Figinski focus on women because they generally outlive men and are, therefore, more likely to ultimately have a greater reliance on SSA benefits. The researchers used the RAND Health and Retirement Study, a processed version of the raw HRS data files that helps facilitate research analysis.

Neumark and Figinski’s findings:

They confirmed past findings that RET elimination led to earlier claiming of benefits for both men and women.

Social Security benefits at the individual and family level were generally lower as a result of RET elimination. This was most strongly true for husbands.

RET elimination “is associated with higher incomes and hence lower incidence of poverty initially—when women are at or just above age 70—but lower income and higher poverty as women get older”—about five years later in the authors’ sample.

“For older women, poverty, and low income more generally, may have been increased by RET elimination, with the effect arising at later ages when these women are more likely to have been widowed, and when there may no longer be higher income from increased labor supply at earlier ages.”

Eric French (University College of London and Institute for Fiscal Studies), “The Effect of the Affordable Care Act on the Labor Supply, Savings, and Social Security of Older Americans”

French and co-authors Hans-Martin von Gaudecker (University of Bonn) and John Bailey Jones (SUNY-Albany) are looking at the impact of the ACA’s expansion of Medicaid and private nongroup insurance provisions on labor supply and saving of Americans ages 50 and older.

The authors have extended the structural labor supply and retirement model in French and Jones (Econometrica, 2011). The life-cycle model of labor supply looks at medical expense uncertainty, health insurance, saving decisions, choice to work, and how many hours to work, along with detailed modeling of medical spending and insurance. The authors used data from the Medical Expenditure Panel Survey (MEPS) to estimate their medical payments model. “We find the model parameters that yield the closest match between data simulated from the model and data assets and labor supply taken from the [HRS]. We find that the model fits the data well with reasonable parameters.”

French explained that the authors have taken the time to build their model with a graphics processing unit (GPU) rather than a standard CPU. (More on GPU models in a future newsletter.) While the initial model building is more time consuming, data processing is exponentially faster using a GPU. French stressed that the data runs they have completed so far are preliminary and incomplete.

David Love (Williams College), “The Comprehensive Wealth of Immigrants and Natives”

Love’s paper, co-written with Lucie Schmidt (Williams College), extends research that shows immigrants have less annuitized wealth from pensions and Social Security benefits by asking how they might compensate for this: By saving more private wealth or by spending down their private wealth more slowly.

The authors use the 1998-2012 HRS to calculate comprehensive wealth (total financial and nonfinancial assets, plus annuitized assets) for immigrants and natives. By looking at median profiles of annualized comprehensive wealth over the retirement period and tracking trajectories of annualized wealth, Love and Schmidt can see whether households are drawing down resources faster or more slowly than a life-cycle model would predict. “We attempt to make sense of these patterns with a life-cycle framework that incorporates some likely suspects for explaining the observed trajectories: uncertain longevity, an explicit bequest motive, precautionary saving in retirement, and housing.”

Some of the paper’s conclusions include:

Immigrants have lower wealth overall than natives, but appear to decumulate resources in retirement at a slower rate

A life-cycle framework points to the combination of home ownership and an explicit bequest motive to explain slower decumulation.

More recent immigrants have low levels of total resources and are likely to have difficulty maintaining adequate levels of spending in retirement.

Emma Aguila (University of Southern California), “Social Security Contributions and Return Migration Among Older Mexican Immigrants”

Aguila serves as an advisor to the Mexican Health and Aging Study (MHAS), which is modeled after the HRS, and is one of the first panel surveys on the 50 and older population in the developing world. MHAS is “well-suited for examining older Mexican migrants as it over-sampled regions with strong U.S. migration patterns.” Aguila used the MHAS’ 2003 and 2012 waves for this study. The sample includes Mexican-born males 50 years or older who reported ever being to the U.S. Women are excluded due to small sample size.

MHAS asks if respondents have ever contributed to Social Security in the U.S., whether they have received U.S. Social Security benefits, and whether they expect to receive such benefits in the future. Aguila notes that because some respondents might not be aware of their contributions to Social Security, the study is likely to underestimate the proportion of all return migrants who have done so.

Aguila found:

Sixteen percent of Mexican males in Mexico in both 2003 and 2012 reported having returned from the U.S. at some point.

Of these men, 40 percent (2003) and 32 percent (2012) reported having contributed to the U.S. Social Security system while here.

Compared to those who did not contribute, those who did were more likely to be U.S. citizens or legal permanent residents, report higher levels of education (college or more), and to have spent more years in the U.S.

“While illuminating, the broader macroeconomic implications of these findings hinge on the characteristics of those who contributed…[contributors] are only eligible to receive benefits in the U.S. if they obtained U.S. legal status before retirement and contributed to the U.S. Social Security system for at least 40 quarters (10 years).”

Aguila’s findings raise questions about retirement security for such immigrants. As she points out, “Technically, a [now] legal immigrant is eligible to collect benefits from wages made as a [previously] undocumented immigrant, but few are thought to provide the required documentation.”

Marco Angrisani (University of Southern California Center for Economic and Social Research) “Nonmonetary Job Characteristics and Employment Transitions at Older Ages”

Because of the emphasis on encouraging longer work as the population ages, policymakers might find information on how nonmonetary job characteristics influence retirement decisions useful. Angrisani and University of Southern California researchers Arie Kapteyn and Erik Meijer ran parallel studies using objective Occupational Information Network (O*NET) measures of job attributes and HRS information on self-assessed nonmonetary job characteristics and employment transitions and retirement plans. Data came from the years 2002-2012 and was restricted to individuals ages 51-79, excluding self-employed and disabled.

Some of the trio’s findings:

There are strong, statistically significant relationships between labor force transitions and job characteristics.”

“Objective measures are more powerful determinants of retirement, while self-reported ones are more important drivers of the decision to move from full-time to part-time.”

An increase of one standard deviation in the level of objective (O*NET) physical demand decreases the probability of remaining in full-time employment by two percentage points.

Perceived and objective social skills requirements and objective cognitive requirements are associated with greater distance from planned retirement age, as well as with higher likelihood of working past age 65.

Angrisani emphasized that the findings, while suggestive that nonmonetary job characteristics are important to retirement decisions, are preliminary in uncovering causal relationships. “Unobservable individual characteristics responsible for sorting into specific occupations may also shape retirement decisions.”

The researchers plan to look at biomarkers, retrospective information on early life environment, and attitudinal questions about individual traits to refine their study and infer causal links between nonmonetary characteristics and retirement timing.

Helen Levy (University of Michigan), “The Effect of Health Reform on Retirement”

Levy won the meeting award for most succinct summary of findings: “Health reform didn’t increase retirement in 2014.” More specifically, “We find no evidence of an increase in retirement among individuals ages 55 to 64 in 2014 compared to 2013. We also do not find any differential trend in retirement in states that have chosen to expand Medicaid under the Affordable Care Act compared with those that have not.” Levy, along with University of Michigan co-authors Sayeh Nikpay and Thomas Buchmueller, used data from the basic monthly Current Population Survey (CPS) for their analysis. Supplemental analyses came from the 2012 Health and Retirement Study.

Levy suggested that it’s probably too soon to look for the ACA’s effect on retirement, which most observers expect to reduce labor supply among workers nearing retirement. (Congressional Budget Office, 2014) “Several factors may have led prospective retirees to exercise caution in relying on ACA coverage in 2014. First, there were well-publicized obstacles to enrollment in health insurance exchanges in the first open enrollment period in fall 2013/winter 2014. Second, prospective retirees may have been prudently waiting to see whether the ACA reforms survived significant legal challenges that were not resolved until a U.S. Supreme Court ruling (King v. Burwell) in June 2015.” As the reforms become more established, the availability of subsidized coverage not tied to employment “may result in increases in early retirement over the next decade.”

Brooke Helppie McFall (University of Michigan), “Changing Work Demands and Compositional Changes in Occupations: Effects on Expected

Helppie McFall presented initial findings for her project with fellow University of Michigan faculty Amanda Sonnega and Robert J. Willis. The trio’s project asks “What changes do we see in occupations held by HRS respondents over time? What do we know about each of these occupations and the older workers in them, and how might the characteristics, demands, and skills required by these occupations affect retirement timing?”

The group used HRS public use data for 11 core surveys fielded from 1992 to 2012, which include observations for more than 37,000 individuals. They aggregated the 900-plus three-digit detailed occupations used in HRS restricted data into 17 occupational categories. They also used self-reports about occupational characteristics such as amount of physical effort and stress, and whether the respondent’s job has gotten more difficult. They also used the measure of whether or not the employer would allow, if the respondent desired, reduced work hours. In addition to HRS data, the trio used O*NET data on occupation-specific information on the activities and abilities required for different jobs.

Most respondents finished working between 63 and 65.

Generally, occupations with greater physical labor are associated with earlier work-force departure, while jobs associated with white collar professionals tend to be associated with longer work.

“We find some interesting exceptions, however. For example, ‘Taxi cab drivers and chauffeurs’ are among the most likely to be working to older ages.”

Other longer-working careers: gardeners and groundskeepers; musicians and composers; management analysts; postsecondary teachers; social workers; clergy and religious workers; lawyers and judges; writers, authors, and technical writers; designers; real estate sales; messengers; guards, watchmen and doorkeepers; and other protective services.

“It is important to recall that our analyses exclude respondents who were retired or worked part-time at the time of their initial HRS interview. This surely introduces some selection biases, so these initial results should not be construed as representative of older adults as a whole.” Helppie McFall went on to explain that the authors consider their findings to be preliminary steps pointing the way toward future research. |•|