As life expectancies of the American and other high-income country populations have increased, the need for retirement savings has increased concomitantly, which has been one factor behind increased labor force participation at older ages. This increased participation, however, has not necessarily taken the form of working longer at one's full-time career job, but rather has taken diverse forms, from moving to full-time or part-time employment at another firm or to various forms of contract work, including staying on as an independent contractor at one's former firm (Quinn et al., Reference Quinn, Cahill and Giandrea2019). Furthermore, both the pattern and timing of these transitions have differed across individuals, with some workers moving immediately from their career job to a new job with others leaving the labor force for a period of time and then returning.
A special issue of the Journal of Pension Economics and Finance (October 2019) containing nine papers examined various aspects of labor force participation at older ages as well as the adequacy of retirement savings (Bronshtein et al., Reference Bronshtein, Scott, Shoven and Slavov2019; Clark et al., Reference Clark, Hammond and Vanderweide2019; Coe Reference Coe2019; Fitzpatrick, Reference Fitzpatrick2019; Horneff et al., Reference Horneff, Maurer and Mitchell2019; Hsu et al., Reference Hsu, Newhouse, Overhage and Zuvekas2019; Morrill and Westall, Reference Morrill and Westall2019; Papke, Reference Papke2019; Quinn et al., Reference Quinn, Cahill and Giandrea2019). The set of six papers in this special issue continue to explore those topics, with four of them reporting results based on new survey data collected by the authors.
Four of the papers in this special issue focus on the pattern and timing of the transition from leaving a career job to full retirement. A fifth compares the behavior of public and private employees with respect to leaving a career job with somewhat surprising results. The final paper explores the timing of claiming of Social Security benefits.
Employment transitions at older ages
Two papers take up the growing proportion of the labor force who work as independent contractors rather than as traditional employees. Munnell et al. (Reference Munnell, Sanzenbacher and Walters2020) start from the observation that working in one's 50s and early 60s is critical, at least for the great majority of people, if their income in retirement will suffice to maintain their standard of living. But if their jobs at those ages do not come with pension and health insurance benefits, their retirement income could well be insufficient.
Munnell and her colleagues use the Health and Retirement Survey (HRS) data to explore both the various patterns of job holding and the nature of those jobs for persons between the ages of 50 and 62 using data from the HRS. They find that only a quarter of their sample has the ideal pattern of work during these ages, namely working continuously at a job with benefits. Of the remainder, 11% work consistently in non-traditional jobs without benefits; this group ends with about 20% less retirement wealth at age 62 than those with the ideal pattern of work. Munnell and her colleagues suggest that provision of auto-IRA's for this group, such as the OregonSaves program could address the shortfall in retirement wealth experienced by those in jobs without benefits.
Abraham et al. (Reference Abraham, Hershbein and Houseman2020) start from the known fact that self-employment rises markedly at older ages, a trend that has been assumed to stem from persons who, having left their career job, value the greater flexibility of self-employment (). They point out that self-employment can take a variety of forms, not all of which imply greater flexibility. A prime example of less flexibility is starting a new business. Therefore, Abraham et al. ask the question: What types of jobs do older self-employed workers have?
Standard public data sources such as the Current Population Survey (CPS) and the HRS do not provide sufficient detail to address this question, so Abraham et al. mounted their own survey which was conducted by Gallup. Although their data replicate the finding of the CPS and HRS that self-employment rises with age, they find much higher self-employment rates than the CPS and HRS, which they attribute to the CPS’ and, to a lesser degree, the HRS’ miscoding some independent contractors as employees. They also believe their survey better identifies persons with low hours of work, who are not identified as working in the CPS and HRS data.
They also find that working as an independent contractor is the most prevalent form of work among those who have left their career job. Being an independent contractor is disproportionately found among the better educated. Moreover, of those working as independent contractors, a sizeable minority are motivated to do so primarily to earn money, despite the self-employed group's being on average better educated and so presumptively having less need to work. Furthermore, a quarter of those working as independent contractors work for their former employer. Although a relatively small percentage of older workers are currently engaged in platform work, Abraham et al. believe this type of work is likely to grow and will be important to monitor in ongoing surveys.
Hudomiet et al. Reference Hudomiet, Hurd, Rohwedder and Parker2020 ask about the current and desired job characteristics of older workers and how they affect retirement behavior. They use data from a battery of questions that they asked of 50–79 year-olds who participated in the RAND American Life Panel Survey and who had with some attachment to the labor force (perhaps in the past if over age 65). Of the characteristics they asked about, the quantitatively most important to the respondents was whether a job had flexible hours (note this is not the same as access to part-time work). If all jobs had flexible hours compared with no jobs having flexible hours, 15 percentage points more workers over 70 said they would work (32% vs. 17%). Workers also emphasized on jobs being less physically and cognitively demanding and more sociable than their current jobs. Perhaps surprisingly, the respondents did not place much emphasis on access to part-time work.
Clark et al. (Reference Clark, Hammond and Liu2020) explore working behavior and financial well-being of North Carolina public employees after retirement. To do so, they conducted three surveys of teachers and other state and local employees in North Carolina who were working full-time in 2014 and were aged 50 and over. Respondents were then re-interviewed in 2016 and 2018. For the subset of employees for whom they had at least one survey prior to retirement and at least one following retirement, they analyzed plans for retirement and the realization of those plans. They find that over 70% of the workers planned to engage in paid work following retirement; however, only 42% had actually worked for pay in the 2 years after retirement. Many of these retirees returned to work in the first year after leaving their career job. Among those who planned to work after retirement, 55% reported post-retirement work while only 21% of those who did not plan for work after retirement returned to the labor force between the two surveys.
Not surprisingly, younger retirees and those in better health are more likely to return to work, as are those with working spouses because of the complementarity of leisure within the household. Interestingly, teachers were more likely to work after retirement than other public employees. Clark et al. find some evidence that while on their full-time job, workers did not save as much as they wished, in part, because they planned to work after retirement. Self-reported well-being in retirement was lower for individuals who, late in their careers, reported that they planned to engage in post-retirement work.
Public and private employee contrasts
Public employees typically have defined benefit pension plans, whereas private employees who in earlier decades also mostly had a defined benefit plan but now mostly have defined contribution plans. It is well known that these two types of plans offer very different retirement incentives, namely workers with defined benefit pension plans that have vested have much less incentive to continue work at their career job than similar workers with defined contribution plans because they accrue substantially less increase in wealth from continued work (Stock and Wise, Reference Stock and Wise1990; Coile and Gruber, Reference Coile and Gruber2007). It is also well known that workers overall are now retiring later and it has been presumed that the shift in the private sector to defined contribution pension plans could be contributing to this trend. Because of the much greater dominance of defined benefit pension plans in the public sector, however, one might have expected public sector workers to retire earlier. The paper by Papke (Reference Papke2019) in last year's conference examined this question and found support for the hypothesis that the dominance of defined benefit pensions in the public sector could causally relate to earlier retirement there.
Coile and Stewart (Reference Coile and Stewart2020) used in their paper longitudinal data from the HRS to re-examine this question. Unlike Papke, they find that the changes over time in labor force behavior are similar for public and private sector workers, despite the marked difference in their pension plans. Coile and Stewart believe their differences with Papke's results are attributable to their more accurate measure of Social Security benefits at the household level. They posit that common changes in other factors explain the similar retirement patterns across sectors. Determinants of labor supply that seem to have followed similar trends across sectors include improving health and rising levels of education across cohorts.
The timing of social security claiming
Delaying claiming Social Security benefits increases an individual's real benefit 8% for every year of delay, yet substantial numbers of beneficiaries do not delay claiming. Over a third of beneficiaries claim benefits at the earliest possible age of 62. One potential explanation of the failure to delay claiming is that people place a low value for the increase in lifetime real income in the form of an annuity, i.e. monthly Social Security benefits. The paper by Maurer and Mitchell (Reference Maurer and Mitchell2020) examines whether a lump sum payment for delaying claiming would entice individuals to claim at older ages. They ask if individuals had the option to delay claiming 4 years in return for receiving a lump sum at the end of 4 years, how many would do so?
To obtain data to answer this question, Maurer and Mitchell implemented an experimental module in the HRS. They began by setting the lump sum offered for delaying claiming equal to an actuarily fair amount ($60,400) with no work required during the 4-year delay period. Relative to the status quo where a delay boosts the lifetime income stream, adding a lump-sum option increased the percentage delaying claiming from 46 to 56%. Retirement income would obviously be greater if persons continued to work during the 4-year delay period. Maurer and Mitchell then investigated how requiring part-time work during the 4-year delay period would change the required amount. They found the lump-sum would only have to be about 10% larger, implying a low utility of the foregone leisure.
Conclusion
Some of these papers reinforce prior findings in the literature. For example, working at jobs with pension and health insurance benefits into one's 60s is important for maintaining one's standard of living throughout the retirement years. Most studies also find that prior to retirement, many workers are excessively optimistic about the income they will have relative to the income they want to have after retirement and thus underestimate the need to save and to remain employed until older ages.
But with new data, these papers break new ground in demonstrating the variety of labor force participation patterns of older workers, especially the role of independent contractors. One paper, drawing on behavioral economics, proposes a nudge type intervention to reduce the number of persons who claim Social Security benefits at an early age, thereby depriving themselves of a considerably enhanced lifetime income stream. Another paper finds a puzzle: if defined benefit plans, which are much more prevalent in the public sector, lead workers to retire at an earlier age than defined contribution plans, as prior literature has suggested, why do public and private sector employees retire at about the same age? As the baby boomers move into older ages, it is important to have a better understanding of their labor force behavior and the ability of retirees to provide for themselves after leaving their full-time career jobs. The papers in this volume help to illuminate the critical policy debates on the ability of retirees to maintain their prior standard of living.
