This conference is supported by Grant #G-2017-8870 from the Alfred P. Sloan Foundation
Contract Work and Labor Force Participation at Older Ages
This paper was distributed as Working Paper 26612, where an updated version may be available.
Over the past several decades, private sector workers in the U.S. with employed-sponsored pensions have experienced a dramatic shift from DB to DC type plans. However, public sector workers have not experienced a similar shift. In this paper Coile and Stewart use (primarily) HRS data to explore changes in the retirement incentives and retirement behavior of public and private sector workers over the past quarter-century. They find that trends in retirement behavior are largely similar across the two groups. They also find that both groups have become less likely to report having a DB pension or any pension. Public sector workers have a higher level of retirement wealth and a larger financial gain from continued work at older ages than do private sector workers, and these differences by sector are growing over time. Both worker types are similarly responsive to retirement incentives.
Working consistently through one's fifties and early sixties is a key to attaining retirement security. However, workers also need access to retirement plans - so they can continue to accumulate resources - and health insurance - so they can avoid withdrawing assets in the event of a health shock. Yet, despite the fact that a large literature focuses on nontraditional jobs (i.e., jobs that often lack these benefits), it is unclear how older workers use them. If some older workers end up in nontraditional work for much of their later careers, then they likely will end up worse off financially. If, instead, older workers use nontraditional jobs only temporarily before returning to traditional work or as a bridge to retirement, then it is unlikely that their financial situation will substantially change. Munnell, Sanzenbacher, and Walters use the Health and Retirement Study to identify nontraditional jobs and relies on sequence analysis to explore how workers ages 50-62 use them. The results suggest that the majority of nontraditional jobs are used by workers consistently, and that fewer workers use these jobs briefly or as a bridge to retirement. In the end, these consistently nontraditional workers end up with less retirement income than other workers, even controlling for characteristics like education. Policymakers may want to consider policies like Auto-IRAs that ensure retirement savings vehicles are readily available even to workers in nontraditional jobs.
Engaging in paid employment after claiming retirement benefits may be an important avenue for individuals to work longer as life expectancies rise. After
separating from one's career employer, individuals may engage in paid work to stay active or to supplement their current level of retirement savings or both. Individuals who choose not to work after claiming may be expressing their preference to stay retired, perhaps because their retirement income is sufficient. However, the decision to work after claiming may be driven by the lack of retirement planning and insufficient savings, while the lack of post-claiming work may reflect the inability to find adequate employment opportunities. Clark, Hammond, and Liu use administrative records merged with panel data from several surveys of public employees in North Carolina to study the decision to engage in paid work after claiming retirement benefits. More than 60 percent of active workers plan to work after claiming benefits, while only around 42 percent of the same sample of individuals have engaged in post-claiming paid work in the first few years after leaving public sector employment. Despite this gap, stated work plans are strongly predictive of actual post-claiming work behavior. The researcher's final analysis uses selfreported measures to gauge the financial well-being of their sample in the early years after leaving career employment.
Although the labor market is characterized by long-term associations between employers and employees, the labor supply literature frequently models employer-employee interactions in the context of competitive spot markets. Patki exploits firm's decisions to renege on implicit pension promises to study the labor supply behavior of older workers in the context of long-term employer-employee relationships. Combining rich administrative data and survey data, he finds an extensive margin labor supply elasticity of 0.47, which is two times larger than existing estimates. These findings illustrate the importance of long-term contracts in explaining labor supply behavior and provide evidence about the effectiveness of policies aimed at lengthening the working lives of older Americans.
How the CEA and the Trump Administration Addressed Work and Retirement Policies
There is great interest among researchers and policymakers in understanding how economic, social, and other factors affect the retirement age of older workers. Hudomiet, Hurd, Parker, and Rohwedder present results based on a recent survey fielded in the RAND American Life Panel that queried older workers about their current, desired, and expected job characteristics; as well as about how certain job characteristics would affect their retirement decisions. Having access to flexible work hours was found to be the most consistent predictor of retirement preferences and expectations. For example, they estimated that the fraction of individuals working after age 70 would be 32.2% if all workers had flexible hours, while the fraction working would be 17.2% if none had the option of flexible hours. The researchers further found that job stress, physical and cognitive job demands, the option to telecommute, and commuting times were also strong predictors of retirement preferences and expectations. By comparing workers' current job characteristics with those that individuals desire in their jobs preceding retirement, Hudomiet, Hurd, Parker, and Rohwedder show that people would like preretirement jobs to be less cognitively and physically demanding and more sociable compared to their current jobs. They also find that most workers worry about their health and the demands of their jobs when they think about their future work trajectory, but relatively few worried about labor demand. Finally, the researchers found that having access to part-time jobs, and expected longevity were less important predictors of retirement.
Insuring retirement security is an important challenge for our aging society, and many policymakers are seeking ways to help individuals save more for retirement. The state of Oregon recently launched an auto-enrollment retirement savings program for private sector workers who lack access to workplace retirement plans; many of these workers are lower-paid employees working at smaller firms. Chalmers, Mitchell, Reuter, Sanzenbacher, and Zhong's paper investigates early results from the OregonSaves program using data through June 2019. They find that OregonSaves is serving firms across many industries, including food services, health care, retail trade, and agriculture. In June 2019, approximately 24,000 contributing participants deposited an average of $110 per month, or about 5% of their pay, which is the default savings rate. To date, over 40,000 individuals have accumulated combined assets over $22.7 million. The researchers also find that OregonSaves has provided access to workplace retirement accounts for employees of small to mid-sized firms (average firm size 36 employees), with participating employees' earning an average of $2,182 per month.
Maurer and Mitchell design and implement an experimental module in the 2014 Health and Retirement Study (HRS) to measure older persons' willingness to defer claiming of Social Security benefits. Under the status quo of the current system, where delaying claiming boosts eventual benefits, 46% of the respondents would delay claiming and work longer. If respondents were instead offered an actuarially fair lump sum payment instead of higher lifelong benefits, about 56% indicate they would delay claiming. Without a work requirement, the average amount needed to induce delayed claiming is only $60,400, while when part-time work is stipulated, the amount is slightly higher, $66,700. This small difference implies a low utility value of leisure foregone, of under 20% of average household income.
This paper was distributed as Working Paper 22942, where an updated version may be available.