Authors:
John C Scott
, University of North Carolina, Department of Public Policy
Yung-Ping (Bing) Chen
, University of Massachusetts—Boston
Date: 4/21/2008
Basic Concepts & Definitions
There is no statutory or legal definition of phased retirement, and, to the best of our knowledge, there is no consensus definition of phased retirement among scholars or practitioners. The practice that we are calling ‘phased retirement’ in this entry has been and is associated with other terms, such as ‘partial retirement,’ ‘gradual retirement,’ ‘Deferred Retirement Option Plans’ (DROP), and ‘retire and rehire’ arrangements, among others. We will discuss these other terms and practices, but first a working definition of phased retirement may be appropriate. Phased retirement is the reduction of work with a long-term employer as an older employee approaches full retirement. This definition has three parts or concepts that need to be considered in concert: ‘reduction of work’; ‘with a long-term employer’; and ‘approaches full retirement.’ Under this framework, phased retirement is part of the broader context of part-time work [see Reduced Hours Work/Part-Time Work and Flexible Work Arrangements].
The first part of our definition is the ‘reduction of work’, and this is usually expressed in terms of hours worked per week. The U.S. Bureau of Labor Statistics definition (1995) of full-time work is 35 or more hours a week such that fewer than 35 hours a week (but not 0) would be considered part-time work. There are exceptions to this standard, of course, depending on industry and occupation-specific standards. For example, among elementary school teachers, 32 hours per week is considered full-time (Barnett, 2003). In any event, for phased retirement purposes, a reduction in work is usually taken to mean a level of work that is less than the relevant standard for full-time work. There is no requirement as to how the reduction occurs. A reduction in work can be gradual (e.g., 35 hours a week reduced to 30 hours a week in the first year, 30 hours a week reduced to 25 hours a week in the second year, etc.). A reduction in work also could be more abrupt as in a one-time reduction (e.g., from full-time to half-time).
We want to stress that, in theory at least, a reduction in work could be based on criteria other than hours worked, such as a reconfiguration of job responsibilities. For example, a manager nearing retirement might transfer to a position that does not require supervision responsibilities and is similar to a consultant within a firm. However, most commentators and scholars consider phased retirement in terms of reduced hours at a job (ERISA Advisory Council, 2000; Allen, Clark, & Ghent, 2001; Ehrenberg, 2001; Chen & Scott, 2003; Hutchens, 2003; Lord, 2005). In this conceptualization, phased retirement is often contrasted with the classic or stereotypical case of an immediate change from working full-time to full-time retirement. Moreover, research into phased retirement tends to measure the phenomenon in terms of quantifiable measures like hours worked per week.
The second phrase in the definition--‘with a long-term employer’--suggests a state of full-time work from which a reduction in hours worked begins. However, there is not a consensus on this position in general or on what ‘long-term’ means. For Gustman and Steinmeier (1984), phased retirement, which they called ‘partially retired in the main job’, was a reduction in hours worked from the full-time job at age 55. Ruhm (1990) expressed the concept as partial retirement ‘in the career job.’ Quinn (1996) was more specific in denoting “a ‘full-time career job’ as one which a worker has held for at least 10 years and on which s/he is working full time (at least 1600 hours per year)” (1996: 12). A worker could work full-time with a series of different employers over a lifetime, for the last of which the worker goes from working full-time to phased retirement. While possible, we would not expect this to be the case because phased retirement programs are usually informal arrangements (Hutchens, 2003) and such high-turnover workers would have little bargaining power to negotiate a phased retirement program.
By this definition, the requirement of phasing down work with a long-term employer makes phased retirement distinct from what has been called ‘bridge jobs’ (Ruhm, 1990; Quinn, 1998). For example, the worker who ‘retires’ from a career employer and then takes a part-time job with a different employer or in a different line of work would not be considered a phased retiree but a worker in a bridge job.
Thus, the definition of phased retirement segments part-time work at the older ages into two categories based on who the employer is. Why do we make this distinction? First, phased retirement has the potential to provide the better pay, benefits, and working conditions that bridge jobs, as the ‘bad jobs’ literature suggests, does not (Kalleberg, Reskin, & Hudson, 2000). A worker who has accrued any substantial amount of tenure is likely to have access to health insurance, pensions, will be higher up on the salary schedule, and the like. Second, the better working conditions and benefits likely found in phased retirement reflects greater control over the terms of work by the employee. That control may stem from seniority in the firm, the attainment of a management position, or from an institution-wide program that provides phased retirement to employees who satisfy certain age and/or service criteria. This distinction also makes sense in terms of the ‘good jobs - bad jobs’ literature which has seen the creation of good part-time jobs for professionals and managers (Tilly, 1992).
The third component to the definition, which is that phased retirement occurs as the worker “approaches full retirement,” may seem tautological, but it serves the purpose, already discussed, of distinguishing those employees who work part-time as a way to transition to full retirement from workers who just happen to work part-time as a work style of choice or circumstance. This clearly rules out younger workers who work part-time. Moreover, even older workers may be working part-time involuntarily and planning or hoping to resume working full-time in the future. Thus, phased retirement requires some intention or expectation on the part of the worker that the process of phased retirement will lead to full-time retirement.
We want to stress again that there is no consensus on all of these points. Phased retirement is a term that can be defined in different ways, but different approaches would have different implications for both research and policy. The larger lesson from this discussion is that in the absence of a consensus of a specific definition from a policy change, researchers should be explicit in defining phased retirement.
There are also some terms related to phased retirement, in addition to ‘bridge jobs’ mentioned above, that should be reviewed. First, there are ‘retire and rehire’ arrangements, also known as ‘reverse retirement’, which is an increase in work activity following full or partial retirement (Chan & Stevens, 2007). DROP plans (Deferred Retirement Option Plans) are in essence phased retirement programs in state and local governmental pension plans in which workers can continue working following the normal retirement date but their pension distributions are deposited into accounts that are held until the workers actually retire (Calhoun & Tepfer, 1998). The terms ‘partial retirement’ and ‘gradual retirement’ have also been used, but generally there is no significant difference between these terms and the more widely used term of phased retirement.
Importance of Topic to Work-Family Studies
Several demographic and economic factors are driving the development of phased retirement. These factors include the increasingly long work weeks of highly educated workers, which may make full-time work schedules less appealing for older workers. The growing numbers of women in the workforce may also be generating part-time work, whether voluntary or involuntary, for older women. The demand for phased retirement by employers may be caused in part by the increasing scarcity of younger workers due to population aging. The increasing numbers of older Americans who are available to work and who want to work, in part due to the large Baby Boom cohort and in part due to health and longevity gains, could also be responsible for increased employee interest in phased retirement. Increased work at the older ages, including phased retirement, could also be a function of changes in the nature of the economy that de-emphasize physical labor. The increasing importance of defined contribution plans, such as the 401(k) plan, and the decline of traditional defined benefit pensions as a source of retirement income to current and future retirees may also be a factor: Without the guaranteed income from defined benefit (DB) plans, workers may need to continue working to save more for retirement. Aside from a source of savings, phased retirement may also delay or slow withdrawals from pension plans, thereby preserving retirement assets. The growth of medical and pharmaceutical costs and the decline of employer-provided health insurance plans that augmented Medicare may encourage continued work as workers seek to pay for costs and ensure continued access to insurance coverage. As Barnett (2003) notes, scholars are also calling for reduced-hours options as a means to increase employee flexibility, reduce stress, and improve the general quality of life.
Employers may be developing phased retirement policies as a way to retain employees with special skills or experience, as these individuals may be resources to other workers (Hutchens, 2007). Phased retirement could also reduce the high costs of recruiting and training new employees. Moreover, phased retirement may be appealing to employers and institutions because it allows for better and more orderly succession plans, minimizes unfulfilled positions, and avoids costly buyouts in order to encourage retirement (Lord, 2005). Finally, a broad-based phased retirement policy may alleviate the pressures on national and corporate pension programs as workers stay employed longer and draw less from their pensions (Hutchens, 2007).
There has been a great deal of activity at the public policy level in phased retirement. Section 905 of the Pension Protection Act of 2006 permits in-service distributions from defined benefit plans when a participant reaches age 62 even if the normal retirement age is after 62. Prior to the change in the law, the Treasury Department and the Internal Revenue Service issued proposed regulations that generally provide for defined benefit plan distributions that are made as part of a “bona fide” phased retirement program. These proposed regulations have not been finalized as of this writing (May, 2008). Further developments at the national level are expected as the new law and regulations are clarified.
State of the Body of Knowledge
To date, there has been relatively little research on the number of people taking part in phased retirement in the United States. Gustman and Steinmeier (1984) made one of the first distinctions between phased retirement and part-time work in retirement. They found that nearly 3 percent of their sample were what they termed “partially retired in their main job” versus 11 percent who were partially retired outside of their main job. Subsequently, Ruhm noted that 6.2 percent of his sample of workers “were partially retired and working for their career employer” (Ruhm, 1990: 492). More recently, Even and Macpherson (2004) estimated the percentage of phased retirees at 3.9 percent of those aged 50-54; 2.7 percent of those aged 55-59; 3.5 percent of those aged 60-61; 8.3 percent of those aged 62-64; and 14.3 percent of those aged 65 and older.
Generally, employees who are contemplating retirement respond favorably to the option of phased retirement (AARP 2005; Abraham and Houseman 2004; ERISA Advisory Council 2000). Employee responses to phased retirement can vary, however, depending on how the program operates (see, e.g., Bertelsen, 1983; Berry, 1998).* In an AARP study (2005), 48 percent of workers said that, if being a phased retiree meant that they would have to hold a different job with the same employer, this would make phased retirement less attractive to them and 63 percent of all workers indicated that phased retirement would be less attractive to them if it reduced the amount of pension benefits. Likewise, Even and Macpherson (2004) reported that workers covered by defined benefit pension plans are less likely to transition to part-time work than those not covered by pension programs, and of those that do make the switch to part-time work, those with pension coverage are more likely to change employers in the process.
In addition, Watson Wyatt Worldwide (2004) found that phased retirement might influence the timing of retirement: Nearly 25 percent of phased retirees expected to work past age 65 and another 20 percent did not plan to retire at all. This finding comports with surveys of older workers not in phased retirement programs who answered that they would continue working longer than otherwise planned if their employer offered a phased retirement program. In the AARP survey (2005), 78 percent of older workers who expressed an interest in phased retirement anticipated that the availability of phased retirement would encourage them to work past their expected retirement age. In a different survey of employees in a state public school system, 44 percent of respondents indicated that they would consider delaying full-time retirement if a phased retirement option were made available (Bartle, 1989).
Regarding employer views on phased retirement programs, one survey of 600 large private firms found that 16 percent of employers provide a formal phased retirement program (Watson Wyatt, 1999). In a study of 950 public and private organizations that employ 20 or more workers, employers were asked whether they would be willing to permit an older full-time white-collar worker to take phased retirement (Hutchens, 2003). The survey found that although 73 percent of surveyed employers permitted an employee to reduce hours before official retirement, only 14 percent of those employers had a formal, written phased retirement policy that applied broadly to employees. What most employers have in mind is an informal arrangement that would largely depend on their staffing needs. In addition, the survey found that most formal phased retirement programs were more common among large employers than small employers. This is not surprising given that larger establishments have human resource policies and information systems that could support a formal program for phased retirement. However, the survey also found that informal arrangements for phased retirement dominated over formal policies among larger employers.
Why are formal, broad-based phased retirement programs less common? Significant legal, cultural, and institutional barriers stand in the way of the implementation of broad-based phased retirement programs and/or in the use of phased retirement by workers. A critical issue is how a phased retirement program would mesh with existing employee benefit programs, in particular retirement plans such as traditional defined benefit pensions and savings plans such as the 401(k). Researchers have found that employers with savings or defined contribution plans (DC) were more likely to offer phased retirement than employers with traditional defined benefit plans (DB) (Chen & Scott, 2003; Lord, 2005; Ehrenberg, 2000). Complicated tax rules on distributions from, and benefit accruals under, DB pension plans may be preventing employers from coordinating pension benefits with a phased retirement program, although recent changes in the pension laws, noted above, may alleviate some of these complexities. DC plans have more flexibility in terms of the timing of distributions. Traditional DB pension plans also have built-in incentives for retiring such as when a monthly benefit is maximized when an employee reaches the age of full retirement (e.g., age 65). In addition, benefits in a DB plan are often tied to the salary of an employee (such as when a monthly pension is calculated as a percentage of last year’s pay or the average of the last three years of salary) so that phased retirement, with a consequent drop in salary, might cause a reduction in pension benefits.
Moreover, it is unclear how age discrimination laws would apply to phased retirement programs, if at all. Employers may also be concerned about employees drawing down benefits, particularly in defined contribution plans. Employers are unlikely to institute phased retirement programs, particularly broad-based programs, without greater clarity in the law and without the flexibility to adapt such programs to their own needs (Penner, Perun, & Steuerle, 2002; Chen & Scott, 2003).
Perhaps as a result of the prevalence of informal phased retirement programs rather than formal, broad-based plans, phased retirement appears to be unevenly distributed. For example, phased retirement is much more prevalent in white collar skilled positions, and those older workers who engage in phased retirement are generally better educated and have higher levels of household wealth and income than those who are not phased retirees (Chen & Scott, 2006). Phased retirement, in its most common form, may be more of a retention tool for key older workers and less of a broader approach for dealing with labor shortages due to population aging.
Implications for Practice and Research
There are a number of avenues for extending this research. A large question remaining for any research on phased retirement is how to define the phenomenon. In Chen and Scott’s study (2006), the definition of phased retirement combined changes in hours worked with self-report of work-retirement status. However, definitions of phased retirement could use other employment-related or personal characteristics, as discussed above.
How phased retirement is defined also impacts a second line of inquiry, which is the scope of phased retirement activity. How do different definitions of phased retirement change the number of phased retirees? With the maturation of the Health and Retirement Study since its inception in the early 1990s, has the number of phased retirees changed over time? As the topic of phased retirement has received more attention by management consulting firms (see, e.g., Lord, 2005; Watson Wyatt Worldwide, 2004; William M. Mercer, 2001) as well as by policymakers and academics over the past ten years, do we see a concurrent rise in phased retirees and/or employers offering phased retirement to their workers?
It would also be helpful to know more specific motivations behind entering phased retirement. Do workers reduce their hours of work because they want more time in leisure, because they cannot afford full-time retirement, because of health issues, or because of some other reasons? In addition, future work should examine combinations or sequences of work-retirement statuses as the transition from work to retirement is not always orderly or linear. For example, workers could retire, work part-time, return to full-time work for their old employer before retiring again. Additional research might examine not merely one particular state like phased retirement but specific combinations or patterns of different work-retirement states over time.
Moreover, there is some anecdotal evidence that some workers enter into phased retirement involuntarily (see, e.g., Watson Wyatt, 2004, as discussed above). Therefore, a broadly based phased retirement that would be implemented by a policy change may look different from one that is currently produced by an ad hoc process of negotiation between employer and employee.
As noted above, the Pension Protection Act of 2006 may possibly expand opportunities for combining continued work and the receipt of pension distributions. An interesting question is whether these policy changes will affect behavior at both the employer and worker levels. A change in public policy encouraging phased retirement may not achieve the results desired by policymakers. While we do believe that there is a ‘market’ for a policy that accommodates continued work at older ages, how workers respond to a policy change will depend on the type of incentives offered. Pension benefits are a critical inducement for leaving the workforce. Any proposed policy change encouraging phased retirement should be considered in terms of its effect on pension plans, and any change to the pension laws, such as the timing or form of distributions, is likely to affect phased retirement policies. In addition, given that such policies are implemented via employers, how a policy change promotes phased retirement will also depend on how such a policy fits within the human resource and financial structures of the adopting employers.
* In a study of university faculty, the overall rate of workers leaving full-time employment increased significantly, but the increase in full retirement was small. Based on observable characteristics such as age, salary, years of service, and job characteristics, employees entering a formal phased retirement program offered by an employer more closely resembled those remaining in full-time jobs than those entering full retirement. The probability of entering into phased retirement was also related to job performance, workload, and maximization of personal income (Allen, Clark, and Ghent 2001). One study of a large state university system found that a phased retirement program raised the odds that low-performing faculty would start the retirement process earlier (Allen 2004).
References
AARP. (2005). Attitudes of individuals 50 and older toward phased retirement. Washington, DC: AARP.
Abraham, K. G., & Houseman, S. N. (2004). Work and retirement plans among older Americans. Employment Research, 11(4), 1-4.
Allen, S. G. (2004). The value of phased retirement. Working Paper 10531. Cambridge, MA: National Bureau of Economic Research.
Allen, S. G., Clark, R. L., & Ghent, L. S. (2001). Phasing into retirement. Paper presented at the Society of Labor Economists annual meeting in Austin, Texas.
Barnett, R. (2003). Reduced hours work/part-time work. Work and Family Encyclopedia, Retrieved on April 21, 2008, from: http://wfnetwork.bc.edu/encyclopedia_entry.php?id=252&area=academics
Bartle, L. (1989). An analysis of the attitudes and perceived needs of a population of public school system employees and retirees regarding retirement and pre-retirement planning. Ph.D. Dissertation. College Park, MD: University of Maryland.
Berry, L. J. (1998). Faculty retirement turnover in community colleges: A real or imagined problem? Ph.D. Dissertation. Little Rock, AR: University of Arkansas.
Bertelsen, K.H. (1983). A study of phased retirement programs at land-grant colleges and universities and the development of phased retirement models for the University of Tennessee. Ph.D. Dissertation. Knoxville, TN: University of Tennessee.
Calhoun, C. V., & Tepfer, A.H. (1998). Deferred retirement option plans (“DROP” plans). Pension & Benefits Week, October 13. Retrieved May 14, 2008, from: http://benefitsattorney.com/modules.php?name=Content&pa=showpage&pid=14
Chan, S., & Stevens, A. H. (2007). Is retirement being remade? Developments in labor market patterns at older ages. Pension Research Council Working Paper CP-2007-04-23-Chan. Philadelphia, PA: Wharton School of the University of Pennsylvania. Retrieved April 21, 2008, from: http://www.pensionresearchcouncil.org/publications/document.php?file=312&download=1
Chen, Y.P., & Scott, J. C. (2003). Gradual retirement: An additional option in work and retirement. North American Actuarial Journal, 7(3), 62-74.
Chen, Y.P., & Scott, J. C. (2006). Phased retirement: Who opts for it and towards what end? AARP Research Report 2006-01. Washington, DC: AARP.
Ehrenberg, R. G. (2001). Career’s end: A survey of faculty retirement policies. Academe, July-August. Retrieved April 21, 2008, from: http://www.dol.gov/ebsa/publications/phasedr1.htm
ERISA Advisory Council (2000). Report of the working group on phased retirement. U.S. Department of Labor. The Advisory Council on Employee Welfare and Benefit Plans. Washington, DC: U.S. Department of Labor. Retrieved April 21, 2008, from: http://ideas.repec.org/p/iza/izadps/dp1353.html
Even, W. E., & Macpherson, D. A. (2004). Do pensions impede phased retirement? Discussion Paper No. 1353. Bonn, Germany: Institute for the Study of Labor. Retrieved April 21, 2008, from: http://www.ilr.cornell.edu/extension/files/20031219112155-pub1251.pdf
Gustman, A. L., & Steinmeier, T. L. (1984). Partial retirement and the analysis of retirement behavior. Industrial and Labor Relations Review, 37(April), 403-415.
Hutchens, R. M. (2003). The Cornell study of employer phased retirement policies: A report on key findings., Ithaca, New York: School of Industrial and Labor Relations, Cornell University. Retrieved April 21, 2008, from: http://www.tiaa-crefinstitute.org/research/trends/docs/tr060105.pdf
Hutchens, R. M. (2007). “Phased retirement: Problems and prospects.” Issue Brief, Series 8, February 2007. Chestnut Hill, MA: Center for Retirement Research, Boston College.
Kalleberg, A. L., Reskin, B. F., & Hudson, K. (2000). Bad jobs in America: Standard and nonstandard employment relations and job quality in the United States. American Sociological Review, 65(2), 256-78.
Lord, M. (2005). Phased retirement: Often a win-Win situation. TIAA-CREF Institute, Trends and Issues Paper. New York: TIAA-CREF. Retrieved April 21, 2008, from:
http://escholarship.bc.edu/cgi/viewcontent.cgi?article=1226&context=econ_papers
Penner, R. G., Perun, P., & Steuerle, E. (2002). Legal and institutional impediments to partial retirement and part-time work by older workers. Washington, DC: The Urban Institute.
Quinn, J. F. (1998). New paths to retirement. Working Papers in Economics, Economics Department, Boston College. Retrieved April 21, 2008, from: http://escholarship.bc.edu/cgi/viewcontent.cgi?article=1226&context=econ_papers
Ruhm, C. J. (1990). Bridge jobs and partial retirement. Journal of Labor Economics, 8(4), 482-501.
Schopp, W. K. (2000). Testimony on behalf of the Association of Private Pension and Welfare Plans before the U.S. Senate Special Committee on Aging, April 3, 2000, Committee on Aging Print: 106-26. Washington, DC: United States Senate.
Tilly, C. (1992). Two faces of part-time work: Good and bad part-time jobs in U.S. service industries. In B. Warme, K. Lundy, and L. Lundy (Eds.), Working Part-Time: Risks and Opportunities. New York: Praeger.
U.S. Bureau of Labor Statistics (1995). Labor force statistics: Annual average data 1994. Washington, DC: U. S. Bureau of Labor Statistics.
Watson Wyatt Worldwide (1999). Phased retirement survey. Washington, DC: Watson Wyatt Worldwide.
Watson Wyatt Worldwide (2004). Phased retirement: Aligning employer programs with worker preferences. Washington, DC: Watson Wyatt Worldwide.
William M. Mercer, Inc. (2001, May). Phased retirement and the changing face of retirement. William M. Mercer, Inc.
Locations in the Matrix of Information Domains of the Work-Family Area of Studies
The Editorial Board of the Teaching Resources section of the Sloan Work and Family Research Network has prepared a Matrix as a way to locate important work-family topics in the broad area of work-family studies. (More about the Matrix...)
Note: The domain areas most closely related to the entry's topic are presented in full color. Other domains, represented in gray, are provided for context.
