Are traditional retirements a thing of the past? new evidence on retirement patterns and bridge...

This article is reprinted by permission of the authors and the U.S. Bureau of Labor Statistics. This article is part of the U.S. Bureau of Labor Statistics Working Paper Series. All views expressed in this article are those of the authors and do not necessarily reflect the views or policies of the

Introduction

Are permanent, one-time retirements coming to an end just as the trend towards earlier and earlier retirements among older American men did nearly twenty years ago? The leading edge of the Baby Boomers, those born between 1946 and 1964, will approach traditional retirement age in less than five years and, along with the transition, retirement as we know it will be redefined. The reason is simple: fundamental changes in Social Security, private pensions, health, and longevity mean that many Americans will be unable to finance twenty or more years of leisure later in life without a significant reduction in living standards. Some who can afford it may prefer not to retire in the stereotypical manner. Others might not have a choice. Whatever the reason, many older Americans may delay complete retirement and exit the labor force gradually.

Since the 1960s, increasing prosperity and more generous public and private retirement programs have meant that many older workers could retire early, supported by Social Security, defined-benefit pension plans, and savings. Each of these sources, however, may look much different in the future. Social Security is facing financial shortfalls over the second half of the traditional seventy-five year window, which will likely lead to some combination of lower benefits and replacement rates, later ages for normal and perhaps early retirement benefit eligibility, and higher Social Security taxes or other government revenues. Traditional defined benefit (DB) pensions are being replaced by defined-contribution (DC) plans like 401(k)s, with their attendant investment risks. Finally, private savings, as reported in the U.S. National Income and Product Accounts, are at their lowest levels since the Great Depression. (1) With reductions in these income sources, traditionally considered the three pillars of retirement income, and possible cutbacks in Medicare, many Americans will have to choose between working longer or enduring a lower standard of living during their retirement years.

Older Americans appear to be adjusting to the new retirement landscape already. A near century-long trend towards earlier retirement among men came to a halt in the mid-1980s. Labor force participation rates among older men have been relatively flat since then, and have increased in recent years. Among women, the decline in earlier retirements and the increase in women entering the labor force in the post-war period have resulted in large increases in labor force participation among older women over the past twenty years. (2) Retirement is also becoming more of a process than a single event. A significant number of individuals, previously estimated at one-third to one-half of retirees, take on short-duration or part-time jobs after leaving full-time career (FTC) employment. These jobs bridge the gap between FTC employment and complete labor force withdrawal, and are aptly called "bridge jobs."

While the existence of bridge job employment is not news, the degree to which it is utilized by today's older workers is. Much of the recent research on bridge jobs is based on data from the late 1990s, just before some fundamental changes in the retirement environment occurred. The stock market decline of 2000, combined with a continuing shift towards 401(k) pension plans, led to an unexpected decline in older Americans' retirement assets. Moreover, the Social Security normal retirement age began its legislated gradual increase from age 65, for those who turned 62 before 2000, to age 66, for those who turned 62 in 2005. This is equivalent to an across-the-board benefit cut, providing an additional incentive to remain in the labor force longer. Private savings rates also continued their decline, albeit with a slight uptick in the last year or so. Anecdotal evidence suggests that these events led to an increase in the labor force participation rates of older Americans, in the form of delayed retirement or, in some cases, labor market reentry after retirement. (3)

Recent waves of the Health and Retirement Study (HRS), the premier data set for the study of older Americans' retirement behavior, now enable an up-to-date analysis of the status of bridge job activity. The HRS now contains information on the retirement patterns of individuals aged 51 to 61 in 1992 for a full ten years, spanning the run up in the stock market in the late 1990s, its subsequent decline, and the continuing recovery. Using these data we explore whether the majority of retirees still exit the labor force in the stereotypical fashion, or if the majority now retire gradually, in stages, utilizing bridge jobs on the way out.

This article is structured as follows. The next section discusses how the retirement landscape has changed over the past two decades, most notably, the abrupt end of the trend towards earlier and earlier retirement among older men. We then document the extent of bridge job employment among today's retirees and those on the cusp of retirement, to see if their behavior resembles that of their predecessors. We conclude that changes in the retirement income landscape since the 1980s are beginning to have an effect and, as a consequence, traditional retirements in the form of a one-time, permanent event have become the exception rather than the rule.

Background

The average retirement age, defined here as the youngest age at which half of the population is out of the labor force, declined dramatically among men during the 20th century, from age 74 in 1910 to age 70 in 1950, age 65 in 1970 and age 62 by 1985. (4) The decline is predominantly a result of increasing prosperity over the past century and the growth of public and private retirement programs. As productivity and real wages increased, workers spent a portion of their increased wealth on leisure, including earlier retirement. Since the mid-1980s, however, the average retirement age for American men has remained relatively unchanged. While there is some debate over the cyclical or permanent nature of this break in trend, it is clear that the retirement landscape has changed. The end of mandatory retirement for most workers in 1986, the shift away from traditional DB pension plans towards employee-controlled DC plans, improvements in health and longevity, and changes in the physical nature of jobs have all created incentives for workers to stay in the labor force longer, either by remaining on their career jobs, by taking on bridge jobs, or both.

Many studies have explored the factors that affect retirement decisions, such as age, health and health insurance status, and Social Security and private pension eligibility and incentives. (5) Other studies have focused on retirement patterns. Honig and Hanoch (1985) investigated partial retirement trends using Retirement History Study (RHS) data from 1967 to 1973. They found that the decision concerning whether or not to participate in the labor market is of primary importance to older workers, while the choice between part-time or full-time work is secondary. Ruhm (1990) used the same source to analyze partial retirement and found that the majority of workers leave career jobs for partial retirement at some point in their working lives. He argued that pension status and Social Security regulations have little effect on this decision and that uncertainty concerning retirement income and institutional job restrictions (e.g. requiring full-time employment for most jobs) are potential causes. (6)

Quinn (1999) studied retirement patterns and bridge jobs in the 1990s. Using the first three waves of the HRS, Quinn estimated that between one-third and one-half of older Americans would take on bridge jobs before exiting the labor force completely. Quinn also found that age, health status, type of pension, and pension eligibility are all important determinants of whether an individual is employed in either a FTC job or a bridge job, or is retired. He concludes that "retirement patterns in America are much richer and more varied than the stereotypical one-step view of retirement suggests." (7)

Chen and Scott (2002) and Purcell (2005) both noted that several forms of gradual retirement through bridge jobs are currently available to workers, including job sharing, reducing work schedules, and the rehiring of retired workers as temporary employees. These studies conclude that older workers are remaining in the labor force longer, that retirement patterns are diverse, and that financial incentives, health, and uncertainty are key explanatory factors. Maestas (2004) observed that more than one-third of retirees in their 50s go back to work after retirement, defined as complete withdrawal from the labor force. Most notably, Maestas found that returning to the labor force is often anticipated prior to retirement, and that non-traditional retirements are often not associated with negative outcomes.

This article focuses on the importance of gradual retirement--bridge job behavior--in the retirement patterns of today's older American workers.

Design and Methods

An ideal data set for this research is the Health and Retirement Study. The HRS is a nationally representative panel data set designed to understand the antecedents and consequences of retirement, to monitor work disability, and to examine the complex relationships among labor supply, health, income and wealth, and saving and consumption over time. (8) As shown in Table 1, the HRS core sample consists of about 12,600 persons (in over 7,600 households)--respondents aged 51 to 61 in 1992 and their spouses, who could be older or younger. Oversamples are taken of Hispanics, Blacks, and Florida residents. The first wave of the HRS is based on in-home interviews in 1992. Subsequent waves are based on extensive follow-up interviews conducted every other year.

The longitudinal nature of the HRS allows us to examine each respondent's work history and identify bridge jobs. For the purposes of this paper, we define a full-time career job as one that consists of at least 1,600 hours per year ("full time") and that lasts ten or more years ("career"). Jobs that follow FTC jobs and precede labor force withdrawal are considered bridge jobs. (9) These definitions are consistent with earlier studies investigating bridge job behavior.

We begin our analysis by focusing on individuals who have had work experience since age 49, and obtain a sample of 10,540 HRS respondents. Table 1 describes the entire sample by work status and gender. We find that 91 percent of age-eligible men and 77 percent of age-eligible women have worked since age 49. Next, we focus on respondents who have had an identifiable FTC job in their work history. Most of the male sample is retained after this restriction, while a larger portion of the female sample is dropped. In total, 80 percent of males and 51 percent of females had worked since age 49 and had an identifiable FTC job in their work history. In later analysis, we study just those with a FTC job in 1992, which reduces the sample further.

Table 2 illustrates the impact of the definition of a fulltime career job on the sample sizes of those with a FTC job ever, and those with a FTC job in 1992. While we think that five years is too short to be considered a career job and requiring twenty years is too restrictive, eight, ten, and fifteen years are all reasonable choices. If the length of tenure required to be on a FTC job is reduced from ten to eight years, there is about a 2 percent increase in the number of men and women with a FTC job in their employment history, and a 7 percent increase in the number on a FTC job in 1992. Conversely, an increase in time required from ten years to fifteen years results in about a 6 to 20 percent reduction in the number men and women with a FTC job and on one in 1992, respectively. The absence of larger fluctuations in the sample implies that the ten-year tenure requirement for a FTC job is reasonable. These definitional changes alter somewhat the number of career vs. bridge jobs, but not the qualitative conclusions discussed below.

Results

Descriptive Statistics

We begin with a cross-sectional discussion of retirement patterns. The first step is to identify each individual's status throughout the retirement process, at every survey point. How many individuals remain on a FTC job and how many move to a bridge job? How many have already exited the labor force and how many return? Table 3 presents the labor force status of HRS males and females from 1992 to 2002 for those who have had a FTC job in the past and work experience since age 49. In 1992, 66 percent of the men were still on their FTC jobs, while 15 percent were employed on bridge jobs and 19 percent were not in the labor force. Among women, 73 percent were still on FTC jobs, while the remaining respondents were divided almost equally between bridge job employment and absence from the labor force.

Ten years later, only 14 percent of the (now much older) male sample was still on a FTC job and 56 percent had exited the labor force. One-quarter of the male sample was on a bridge job in 2002. The story is similar for females. The actual importance of bridge jobs is greater than observed in these cross sections, however, because many of the individuals who are out of the labor in a given wave transitioned through a bridge job before exiting.

The longitudinal nature of the HRS allows us to track individuals' labor force withdrawal patterns over time. To do so, we examine all jobs in a person's work history that have lasted five or more years prior to the wave one (1992) interview. We then examine employment in each wave and construct the path from employment to labor force withdrawal for each respondent. Table 4 demonstrates the extent of bridge job activity estimated in 1996, when the most recent comparable study on bridge job behavior was performed, and then in 2002, based on the most recent data. (10) The percentage of men either working at a bridge job or who had last worked at a bridge job before leaving the labor force increased from 33 percent in 1996 to 50 percent in 2002. Women experienced a similar increase in the number who were holding or had held a bridge job, from 28 to 45 percent. Note that in both cases in 2002, about half of the sample had already utilized a bridge job before exiting from the labor force.

If we concentrate just on those men who had left their FTC jobs by 2002 and whose subsequent status we know 27.8 + 25.3 + 22.0 = 75.1% of the male sample), we find that over 65 percent of them (27.8 + 22.0 = 49.8%; 49.8 / 75.1 = 66.3%) were then employed or were last employed on a bridge job. For women who had left FTC jobs by 2002, the estimate is slightly lower, 62 percent (45.3 / 73.0 = 62.0%).

The increase from 1996 to 2002 does not represent an increase in bridge job behavior over these years, but rather it provides a new and better estimate of the extent of bridge job activity. In 1996, none of the many respondents still on FTC jobs was counted as having a bridge job, even though, as we knew then and now see, many would later take a bridge job before exiting the labor force. Between 1996 and 2002, the percentage still on a FTC job dropped from 42 to 14 percent among men and from 51 to 17 percent among women. The 2002 observation of actual bridge job activity (nearly 50 percent) is still an underestimate, because of those respondents who have yet to leave their FTC jobs. The larger estimates in the paragraph above (62 to 66%) assume that those still on FTC jobs and those for whom bridge job status could not be determined in 2002 will behave, on average, like those who have already left their FTC job.

Non-traditional retirements are even more common than these numbers suggest, since some individuals who do transition directly out of the labor force from a FTC job may re-enter at a later date. We estimate that about 9 percent of individuals who exited the labor force directly eventually reentered the labor force at a later date by 2002. (11) Taken together with bridge job activity, this implies that traditional one-time, permanent retirements are definitely in the minority.

A key determinant of labor force withdrawal patterns is wage-and-salary employment versus self-employment. We find that a larger percentage of self-employed workers, relative to wage-and-salary workers, made the first step from a FTC job to a bridge job rather than to no job by 2002. Among those who left their FTC jobs, over three-quarters of self-employed workers moved to a bridge job, compared to only half of those who were wage-and-salary workers. This may speak to the relative ease with which the self-employed can decrease hours or shift to a new job.

Self-employed workers also remained working longer and switched job types more often later in life compared to wage-and-salary workers. Nearly half of self-employed FTC workers who took on a bridge job were still employed in 2002 compared to only 37 percent of wage-and salary workers. In addition, while only 15 percent of wage-and-salary workers who took a bridge job have taken on a job in self-employment, 40 percent of self-employed FTC workers took on a wage-and-salary bridge job. A similar story is observed with those who had taken bridge jobs but were out of the labor force by 2002.

Covariates of Bridge Job Activity

The first wave of the HRS provides detailed information about health status, pension status, wealth, and many other characteristics related to retirement decisions. Because some of these variables are time variant, we cannot rely on these measures to examine exits from FTC jobs that took place prior to 1992. Therefore, we base our analysis of bridge job determinants on those individuals who were on FTC jobs in 1992, and we follow their exit patterns through 2002.

Table 5 presents the first transitions from FTC jobs by the age of the respondent. Older workers who leave their career jobs are less likely to transition to bridge job employment, as indicated in the last column by the ratio of bridge job employees to all individuals who made a transition. In 2002, of those who moved from a career job, 44 percent of males aged 65 or older had moved to a bridge job, compared to 59 percent of those aged 62 to 64 and 63 percent of men under age 62. The difference is even more pronounced among women, where the difference ranges from 78 percent for females under age 60, about 60 percent for those aged 60 to 64, and only 41 percent among women aged 65 years and older. Bridge job behavior is much more common among those initiating the retirement process at a younger age.

Table 6 examines the health status in 1992 of HRS respondents who were on FTC jobs in 1992, and stratifies the results according to their first transition from the FTC job. Three measures of health status are shown in the table: the presence of a work-limiting condition, a subjective health assessment, and a summary of activities of daily living for which respondents have had a lot of difficulty. Men and women who left their career jobs and were in excellent or very good health took on bridge jobs in 55 to 60 percent of cases (see last column.) Those who departed with self-assessed fair or poor health made the transition to bridge jobs less than 40 percent of the time.

Both men and women who had FTC jobs in 1992 and who reported a work-limiting health condition are much more likely to have exited the labor force by 2002 than those without such conditions. They are also less likely to remain on their FTC job, and less likely to have moved to a bridge job if they did leave their career job. Respondents also reported the number of daily activities for which their health condition limits them from completing. Not surprisingly, men and women with two or fewer ADL limitations are more likely to remain in their FTC job or move to a bridge job, and less likely to have left the labor force altogether.

These estimates underestimate the importance of health in job transition late in life, since they include only those still on a FTC job in 1992. Many of those with serious health issues would have left their career jobs and likely the labor force before 1992.

Earlier studies have shown the impact of the availability of health insurance coverage beyond the FTC job on retirement patterns. (12) Table 7 reports first transitions of respondents by health insurance status on their FTC jobs. Because individuals aged 65 and older are eligible for Medicare, the sample in Table 7 includes only those who are younger than 65 in 2002. We find that men and women who have no health insurance on their FTC job are most likely to take a bridge job if they leave their career job, as 78 percent of these uninsured men and 74 percent of the women did. Also, both men and women who lose health insurance coverage when leaving a FTC job are more likely to exit the labor force completely than other workers in the sample. (13)

Men with defined-benefit (DB) pension plans, which often contain age-specific financial incentives to leave the job, are less likely to remain on their FTC jobs in 2002 than others, are more likely to have left the labor force, and are least likely to move to a bridge job when a transition is made. As seen in Table 8, men with no pension or with DC plans only on their FTC jobs are more likely to remain on that job than are men with a DB plan. While only 41 percent of males who left a career job with a DB pension take a bridge job, nearly 60 percent of their counterparts with no pension and 54 percent with a DC pension do so. Among women, those with DB pension plans only are the most likely to have left the labor force, and the least likely to utilize a bridge job on the way out.

Finally, bridge jobs appear to be more common at both ends of the wage distribution. As shown in Table 9, low-wage and high-wage individuals are more likely to take on bridge jobs when making a transition than are those in the middle of the wage distribution (see last column). This result is consistent with expectations. Among low-wage individuals, many take on bridge jobs out of financial necessity while individuals at the upper end of the wage distribution, many of whom could afford to retire, may choose bridge jobs for quality of life reasons. The same conclusion appears when we look at the occupational status of the FTC job. Both individuals in white collar, highly-skilled jobs and individuals in blue collar, non-highly-skilled jobs are more likely to take on bridge jobs after leaving FTC employment.

Multivariate Analysis

The cross-sectional analyses above indicate that bridge job behavior is very common among older Americans, especially among those who are relatively young and healthy, who are without health insurance, and who are at the lower or upper ends of the wage or occupation spectrum. Bridge jobs are less common among those with DB pension plans on their career jobs. This section examines bridge job determinants in a multivariate context. (14) We model the transitions from FTC jobs among HRS respondents who were on a FTC job in 1992 and for whom later employment status could be identified. Each worker faces three choices between 1992 and 2002; he or she can continue on the career job, can leave the career job for a bridge job, or can exit the labor force. The coefficients of the model are estimated for men and women separately using multinomial logistic regression. Table 10 presents marginal effects estimated at sample means.

As expected, younger men and women are more likely than others to remain on their FTC jobs (the negative and significant coefficients increase with age) as are individuals in excellent or very good health. Bridge job behavior is also more common among the youngest men and women and declines with deteriorating health status. Men and women with dependent children (and college tuitions ahead?) are more likely to remain on their career jobs than others (about 6 percentage points more likely), and women (but not men who do transition are more likely to move to a bridge job.

Another important finding is that men with a DB pension plan are significantly less likely to remain in a FTC job and less likely to take a bridge job. This FTC job effect, oddly, is not found among the women, though the bridge job effect is. Self-employed men and women are much more likely than others to remain on their FTC jobs, and self-employed men are more likely to take on a bridge job when a job transition is made. The "u-shaped" pattern of bridge job behavior by occupational status is flattened at one end in the multivariate context, as white collar, high-highly skilled men are more likely than others to take on a bridge job, while blue collar, not-highly-skilled men are not. Occupational status is not statistically significant among women. Both men and women who maintain their health insurance status with a transition from their FTC job, either because health insurance is portable or because they are not covered, are more likely to take on a bridge job compared to those who would lose their health insurance coverage. The impact, however, is only statistically significant among the uninsured, with a marginal effect of more than 10 percentage points.

Owning one's own home has no discernible effect on the decision to remain on the FTC job, but it does reduce the chances of taking a bridge job when one leaves the career job. Wealth effects are inconsistent, as is often the case in such empirical work, perhaps because those with high wealth also have an unmeasured taste for work, which is one of the reasons that they have accumulated wealth.

The multivariate analysis confirms the cross-sectional descriptive results. Bridge job behavior is more common among younger retirees, among the healthy and the self-employed and among those without DB pension plans. One prediction is that, as DB plans continue to decline in importance, bridge job behavior will become more common, as already appears to be the case.

Discussion

Research from the 1990s demonstrated that, at a minimum, about one-third to one-half of older workers utilize bridge jobs before completely withdrawing from the labor force. Six more years of data are now available to estimate the actual extent of bridge job employment. Using ten years of HRS data, we conclude that about one-half of respondents with FTC jobs have already taken bridge jobs. Of the vast majority who had already left their career jobs by 2002, between 62 and 66 percent moved to a bridge job rather than directly out of the labor force.

What does the importance of bridge job employment mean for individuals and the country as a whole? Overall, continued work is generally good news, as more individuals remain productive late in life, fewer are dependent on public programs, and the nation has more goods and services to distribute among an aging population. From an individual perspective, however, the story is more complicated. For some, bridge job employment and work later in life create the potential for more satisfying years than complete labor force withdrawal would provide. People remain engaged in the world of work and with their former or, more often, with new colleagues. Bridge jobs can provide older Americans with the opportunity to try something different, earn wages and experience in a different line of work, and generate discretionary income, augmenting Social Security, pension income and returns from savings. For others, however, at the lower end of the socio-economic scale, bridge jobs are more likely to be an unfortunate financial necessity, and may involve physically demanding work that signals a bleak and undesirable situation at the twilight of their work lives.

Nonetheless, bridge job employment has the potential to be greatly advantageous to many older Americans, to the firms wise enough to tap these productive and experienced workers, and to the country as a whole. The key is to understand better how and why older Americans choose to leave the labor force, and how best to harness the desire of many of them to continue working beyond traditional retirement ages. This paper takes a first step and assesses just how prevalent these nontraditional retirement patterns are today. We find that traditional retirement has become the exception, not the rule. and that for a majority of older Americans, retirement is a process, often over many years, and not a single event.

U.S. Department of Labor Bureau of Labor Statistics http://www.bls.gov

References

Anderson, P.M., Gustman, A. L., & Steinmeier, T. L. 1997. "Trends in Male Labor Force Participation and Retirement: Some Evidence on the Role of Pensions and Social Security in the 1970s and 1980s. Journal of Labor Economics, 17(4), 757-783.

Blau, D. M., & Gilleskie, D. B. 2001. "Retiree Health Insurance and the Labor Force Behavior of Older Men in the 1990s." Review of Economics and Statistics, 83(1), 64-80.

Chen, Y., & Scott, J. C. 2003. "Gradual Retirement: An Additional Option in Work and Retirement." North American Actuarial Journal, 7(3), 62-74.

Coile, C. C., & Gruber, J. 2001. "Social Security Incentives for Retirement." In D. A. Wise (Ed.), Themes in the Economics of Aging (pp. 311-341). NBER Conference Report series. Chicago and London: University of Chicago Press.

Coile, C. C., & Levine, R B. 2004. "Bulls, Bears, and Retirement Behavior." (National Bureau of Economic Research, Inc., NBER Working Papers). http://www.nber. org/papers/wl0779.pdf

Gruber, J., & Madrian, B. C. 1994. "Health Insurance and Job Mobility: The Effects of Public Policy on Job-Lock." Industrial and Labor Relations Review, 48(1), 86-102.

Gruber, J., & Madrian, B. C. 1995. Health-Insurance Availability and the Retirement Decision." The American Economic Review, Vol. 85, No. 4. pp. 938-948.

Herz, D. E. 1995. "Work After Early Retirement: An Increasing Trend Among Men." Monthly Labor Review, 118(4), 13-20.

Honig, M., & Hanoch. G. 1985. "Partial Retirement as a Separate Mode of Retirement Behavior." Journal of Human Resources, 20(1), 21-46.

Juster, F. T. & Suzman, R. 1995. "An Overview of the Health and Retirement Study." Journal of Human Resources, 30 (Supplement), S7-S56.

Madrian, B. C. 1994. "The Effect of Health Insurance on Retirement." Brookings Papers on Economic Activity, 0(1), 181-232.

Maestas, N. (2004). Back to work: Expectations and realizations of work after retirement. (Rand Working Paper WR-196). Retrieved July 29, 2005 from http://www. rand.org/publications/WR/WR196/

Munnell, A. H., & Cahill, K. E., & Jivan, N.A. 2003. "How Has the Shift to 401(k)'s Affected the Retirement Age?" Issue in Brief 13 (September). Chestnut Hill, MA: Center for Retirement Research at Boston College.

Purcell, P. J. 2005. "Older Workers: Employment and Retirement Trends." CRS Report for Congress, Congressional Research Service. The Library of Congress. Washington, DC.

Quadagno, J., & Quinn, J. F. 1997. "Does Social Security Discourage Work?" In E. Kingson & J. Schulz (Eds.), Social Security in the 21st Century (pp. 127-146). New York: Oxford University Press.

Quinn, J. F. 1977. "Microeconomic Determinants of Early Retirement: A Cross-Sectional View of White Married Men." Journal of Human Resources, 12(3), 329-346.

Quinn, J. F. 1999. "Retirement Patterns and Bridge Jobs in the 1990s." EBRI Issue Brief 206 (February). Washington, DC: Employee Benefit Research Institute. 1-23.

Quinn, J. F. 2002. "Changing Retirement Trends and Their Impact on Elderly Entitlement programs. In S. H. Altman & D Shactman (Eds.), Policies for an Aging Society. (pp. 293-315). Baltimore and London: Johns Hopkins University Press.

Ruhm, C. J. 1990. "Bridge Jobs and Partial Retirement." Journal of Labor Economics, 8(4): 482-501.

Ruhm, C. J. 1995. "Secular Changes in the Work and Retirement Patterns of Older Men." Journal of Human Resources, 30(2), 362-385.

Samwick, A. A. 1998. "New Evidence on Pensions, Social Security, and the Timing of Retirement." Journal of Public Economics, 70(2), 207-236.

Smith, R. E. 2004. "Disability and Retirement: The Early Exit of Baby Boomers from the Labor Force." The Congress of the United States: Congressional Budget Office. Retrieved July 29, 2005 from http://www.cbo.gov/ showdoc.cfm?index=6018&sequence=0

Stock, J. H., & D. A. Wise. 1990. "Pensions, the Option Value of Work, and Retirement." Econometrica, 58(5), 1151-1180.

U.S. Department of Commerce, Bureau of Economic Analysis. 2005. National Income and Product Accounts. Retrieved July 29, 2005 from http://www.bea.doc.gov/ bea/dn/nipaweb/SelectTable.asp?Selected=N

(1) U.S. National Income and Product Accounts, Table 5.1, Saving and Investment, line 33.

(2) See Quinn (2002). Labor force participation rates of American men aged 65 to 69 declined from 43 percent in 1965 to 25 percent in 1985 (a nearly 60 percent decline in only two decades) and then increased back to 33 percent by 2004, with half of that increase occurring during the last 5 years. Participation rates for women aged 60 to 64 for the same three years are 33 percent, 33 percent and then 45 percent and rising in 2004.

(3) Coile and Levine (2004) challenge the notion that the decline in stock market value in 2000 explains rising labor force participation rates of older Americans.

(4) See Quinn (2002).

(5) Anderson, Gustman, and Steinmeier (1997), Coile and Gruber (2000), Herz (1995), Munnell, Cahill, and Jivan (2003), Quadagno and Quinn (1997), Quinn (1977), Ruhm (1995), Samwick (1998), and Stock and Wise (1990), have studied the relationships among financial variables and labor force participation. Likewise, Blau and Gilleskie (1997), Gruber and Madrian (1994), Madrian (1994), and Smith (2004) have considered how health status and health insurance status affect the retirement decisions of older American workers.

(6) Ruhm used a different definition of career job than that used in this paper. A career job according to Ruhm is the longest spell of employment with a single firm.

(7) Quinn (1999), p. 1.

(8) See luster and Suzman (1995) for a detailed description of the HRS.

(9) One concern with this methodology is that respondents might not yet have enough tenure in a given wave for a job to be considered a career job, even though the respondent may still be working and thereby increasing tenure. In some instances, when the individual remains on the job for ten years, these jobs will in fact turn out to be career jobs. When subsequent waves do not cover work status through age 62, we assume that the respondent will work on the job until age 62. This assumption implies that our estimates of bridge job activity are conservative, since some of these individuals will leave their jobs before age 62.

(10) See Quinn (1999).

(11) For a respondent to exit from the labor force he or she must not have a job for at least 2 waves, i.e., a minimum of 2 years.

(12) See Gruber and Madrian (1995) and Madrian (1994). Health insurance status is based on three sources of coverage: government-provided health insurance, such as Medicare and Medicaid, private health insurance, and employer-provided health insurance, based on the individual's employer or a spouse's employer. Portability requires that health insurance coverage is maintained after an individual leaves his or her current job. Government-provided insurance, private health insurance, and insurance through a spouse's employer are all unaffected by the respondent's employment status and are considered portable. Health insurance through an individual's employer is only considered portable if the respondent indicates that the employer-provided coverage will be maintained in retirement. Respondents with no health insurance are captured with a separate variable indicating whether a respondent has health insurance.

(13) Health insurance may be a proxy for other personal or job characteristics. There is much that is not being held equal here, especially age.

(14) The empirical specification is based on the framework of Quinn (1999). Details are available from the authors.

Kevin E. Cahill, Ph.D., Analysis Group, Inc., Michael D. Giandrea, Ph.D., U.S. Bureau of Labor Statistics, Office of Productivity and Technology, and Joseph F. Quinn, Ph.D., Dean, College of Arts and Sciences, Boston College

Table 1. Sample Size by Gender and Work Status

                                        Men       Women      Total

Participated in Wave 1
  n                                   5,869      6,783      12,652

Worked Since Age 49
  n                                   5,344      5,196      10,540
  % of HRS Core                          91.1%      76.6%       83.3%

Worked Since Age 49 and Had FTC Job
  n                                   4,695      3,472       8,167
  % of HRS Core                          80.0%      51.2%       64.6%

On FTC in 1992
  n                                   3,056      2,514       5,570
  % of HRS Core                          52.1%      37.1%       44.0%

Note: Calculations based on the Health and Retirement Study.

Table 2. Percent of Sample with Full-Time Career Jobs
by FTC Definition and Gender

                                 Males
                        Tenure Required for FTC
                           Designation (Years)

                         5          8          10

Worked Since Age 49
  and Had FTC Job
  n                   4,817      4,760      4,697
  % of HRS Core          82.1%      81.1%      80.0%

On FTC Job in 1992
  n                   3,488      3,277      3,056
  % of HRS Core          59.4%      55.8%      52.1%

                              Males
                      Tenure Required for FTC
                        Designation (Years)

                         15         20

Worked Since Age 49
  and Had FTC Job
  n                   4,537      4,466
  % of HRS Core          77.3%      76.1%

On FTC Job in 1992
  n                   2,439      1,986
  % of HRS Core          41.6%      33.8%

                                Females
                        Tenure Required for FTC
                          Designation (Years)

                        5          8          10

Worked Since Age 49
  and Had FTC Job
  n                   3,666      3,578      3,470
  % of HRS Core          54.0%      52.7%      51.2%

On FTC Job in 1992
  n                   2,857      2,710      2,514
  % of HRS Core          42.1%      40.0%      37.1%

                              Females
                      Tenure Required for FTC
                        Designation (Years)

                        15         20

Worked Since Age 49
  and Had FTC Job
  n                   3,160      2,935
  % of HRS Core          46.6%      43.3%

On FTC Job in 1992
  n                   1,959      1,428
  % of HRS Core          28.9%      21.1%

Note: Calculations based on the Health and Retirement Study.

Table 3. Labor Force Status by Year and Gender, Individuals
with a Full-Time Career Job in Their Work History and Work
Experience Since Age 49

                         Males

               Full-
                Time             Not in
               Career   Bridge   Labor    Don't
Year     n      Job      Job     Force    Know

1992   4,695   65.9%    14.5%    19.0%    0.7%

1994   4,353   52.6%    18.5%    28.3%    0.6%

1996   4,044   42.4%    22.2%    34.7%    0.7%

1998   3,833   27.0%    29.3%    42.7%    1.0%

2000   3,573   18.0%    33.0%    47.7%    1.3%

2002   3,414   14.4%    25.3%    55.6%    4.7%

                        Females

               Full-
                Time             Not in
               Career   Bridge   Labor    Don't
Year     n      Job      Job     Force    Know

1992   3,472   72.7%    14.1%    12.7%    0.5%

1994   3,272   62.0%    17.4%    20.2%    0.5%

1996   3,102   51.4%    19.2%    28.9%    0.5%

1998   2,984   32.6%    29.9%    36.3%    1.3%

2000   2,855   19.4%    38.3%    40.7%    1.7%

2002   2,795   17.4%    27.7%    49.7%    5.2%

Source: Calculations based on the Health and Retirement Study.

Table 4. Current Employment Status by Gender, Individuals with a
Full-Time Career Job in Their Work History and Work
Experience Since Age 49 (Horizontal Percentages),
1996 and 2002

                                                  1996

                                          Full-Time            Not in
                                           Career     Bridge   Labor
                                    n        Job       Job     Force

Men, Working                      2,625     42.4%      22.7%     0.7%
Men, Nonworking, Last Job Was     1,419     22.2%      10.4%     1.5%
Total                             4,044     64.6%      33.1%     2.2%

Women, Working                    2,203     51.4%      18.7%     0.5%
Women, Nonworking, Last Job Was     899     19.2%       8.8%     1.3%
Total                             3,102     70.6%      27.5%     1.8%

                                                  2002

                                          Full-Time            Not in
                                           Career     Bridge   Labor
                                    n        Job       Job     Force


Men, Working                      1,522     14.4%      27.8%     4.7%
Men, Nonworking, Last Job Was     1,892     25.3%      22.0%     5.8%
Total                             3,414     39.7%      49.8%    10.5%

Women, Working                    1,392     17.4%      24.9%     5.2%
Women, Nonworking, Last Job Was   1,403     27.7%      20.4%     4.4%
Total                             2,795     45.1%      45.3%     9.6%

Note: Calculations based on the Health and Retirement Study.

Table 5. First Transitions from Career Jobs by 2002,
Those with Full-Time Career Jobs in 1992, by
Gender and Age (Horizontal Percentage and Ratio)

                                                            Ratio of
                  Still on   Moved to                        Bridge/
Age in             Career     Bridge    Moved to   Don't   (Job Bridge
 2002       n       Job        Job       No Job    Know    Job+No Job)

Men
    <60      92    28.5%      37.9%      22.2%     11.4%      63.1%
  60-61     461    24.7%      43.5%      24.8%      7.1%      63.7%
  62-64     819    18.6%      43.0%      30.2%      8.3%      58.7%
    65+   1,348     9.7%      37.5%      47.5%      5.3%      44.1%
  Total   2,720    15.5%      40.2%      37.6%      6.7%      51.7%

Women
    <60     542    26.5%      52.8%      15.0%      5.6%      77.9%
  60-61     402    26.9%      39.2%      25.6%      8.3%      60.5%
  62-64     596    15.3%      46.0%      32.4%      6.4%      58.7%
    65+     770     8.1%      35.9%      52.6%      3.5%      40.6%
  Total   2,310    17.6%      43.1%      33.7%      5.6%      56.1%

Note: Calculations based on the Health and Retirement Study.

Table 6. First Transitions from Career Jobs by 2002,
Those with Full-Time Career Jobs in 1992, by Gender
and Health Status in 1992 (Horizontal Percentage and Ratio)

                                               Moved
                                    Still on     to
                                     Career    Bridge
                              n       Job       Job

Health Conditions
That Limit Work
  Men
   No                       2,811    14.6%     36.8%
   Yes                        268     8.4%     35.7%

  Women
   No                       2,482    16.5%     40.8%
   Yes                        182    10.7%     36.1%

Subjective Health
  Men
   Excellent or Very Good   1,771    16.8%     39.0%
   Good                       936    10.3%     34.9%
   Fair or Poor               375     9.5%     28.5%

  Women
   Excellent or Very Good   1,639    17.9%     42.9%
   Good                       742    14.3%     38.2%
   Fair or Poor               283     8.5%     29.4%

Number of ADLs' with Lots
of Difficulty
  Men
   0                        2,306    14.9%     37.7%
   1-2                        679    12.3%     33.9%
   3-4                         72     6.0%     28.2%
   5+                          25     2.4%     36.6%

  Men
   0                        1,712    16.2%     41.3%
   1-2                        797    16.8%     40.1%
   3-4                        111    11.3%     33.0%
   5+                          44     9.9%     31.4%

                                               Ratio of
                            Moved              Bridge/(Job
                            to No    Don't     Bridge Job+
                            Job       Know     No Job)

Health Conditions
That Limit Work
  Men
   No                       36.1%    12.5%     50.5%
   Yes                      43.0%    12.9%     45.4%

  Women
   No                       31.9%    10.9%     56.1%
   Yes                      44.2%     9.0%     45.0%

Subjective Health
  Men
   Excellent or Very Good   32.5%    11.7%     54.5%
   Good                     40.9%    13.9%     46.0%
   Fair or Poor             47.8%    14.1%     37.4%

  Women
   Excellent or Very Good   29.0%    10.2%     59.7%
   Good                     36.6%    10.9%     51.1%
   Fair or Poor             48.0%    14.1%     38.0%

Number of ADLs' with Lots
 of Difficulty
  Men
   0                        34.5%    12.9%     52.2%
   1-2                      41.7%    12.1%     44.8%
   3-4                      54.7%    11.1%     34.0%
   5+                       53.7%     7.3%     40.5%

  Men
   0                        31.0%    11.5%     57.1%
   1-2                      34.3%     8.8%     53.9%
   3-4                      42.0%    13.7%     44.0%
   5+                       50.4%     8.3%     38.4%

(a) Activities of Daily Living.

Note: Calculations based on the Health and Retirement Study.

Table 7. First Transitions from Career Jobs by 2002,
Those with Full-Time Career Jobs in 1992,
By Gender and Health Insurance Status on 1992 Job,
Age < 65 in 2002 (Horizontal Percentage and Ratio)

                                              Still on    Moved to
                                               Career      Bridge
Health Insurance Status                 n       Job          Job

Men
  Not Covered on Career Job            147     25.1%        50.6%
  "Covered and Would Maintain"         551     22.2%        43.0%
    Some Coverage
  "Covered and Would Lose" Coverage    686     20.3%        41.1%

Women
  Not Covered on Career Job            162     14.6%        14.6%
  "Covered and Would Maintain"         883     22.6%        22.6%
    Some Coverage
  "Covered and Would Lose" Coverage    606     23.2%        23.2%

                                                          Ratio of
                                      Moved                Bridge/
                                      to No    Don't     (Job Bridge
Health Insurance Status                Job      Know     Job+No Job)

Men
  Not Covered on Career Job           14.3%    10.1%        78.0%
  "Covered and Would Maintain"        25.9%     8.9%        62.4%
    Some Coverage
  "Covered and Would Lose" Coverage   31.7%     6.9%        56.5%

Women
  Not Covered on Career Job           21.5%     4.1%        73.6%
  "Covered and Would Maintain"        24.0%     7.2%        65.8%
    Some Coverage
  "Covered and Would Lose" Coverage             7.4%        62.2%

Note: Calculations based on the Health and Retirement Study.

Table 8. First Transitions from Career Jobs by 2002,
Those with Full-Time Career Jobs in 1992, By Gender
and Pension Status on 1992 Job (Horizontal Percentage and Ratio)

                                        Still on    Moved to
                                         Career      Bridge
Health Insurance Status           n       Job          Job

Men
  No Pension                      962    17.6%        40.1%
  Defined Benefit Plan Only     1,345    10.5%        31.5%
  Defined Contribution
    Plan Only                     602    16.9%        38.4%
  Defined Benefit and Defined     173    12.2%        50.5%
    Contribution Plan

Women
  No Pension                      911    13.6%        46.8%
  Defined Benefit Plan Only     1,056    17.2%        34.7%
  Defined Contribution
    Plan Only                     615    19.0%        37.3%
  Defined Benefit and Defined
    Contribution Plan              82     7.1%        68.2%

                                                    Ratio of
                                Moved                Bridge/
                                to No    Don't     (Job Bridge
Health Insurance Status          Job      Know     Job+No Job)

Men
  No Pension                    28.0%    14.3%        58.9%
  Defined Benefit Plan Only     45.5%    12.5%        40.9%
  Defined Contribution
    Plan Only                   32.3%    12.4%        54.3%
  Defined Benefit and Defined   31.3%     6.0%        61.7%
    Contribution Plan

Women
  No Pension                    30.2%     9.5%        60.8%
  Defined Benefit Plan Only     36.6%    11.6%        48.7%
  Defined Contribution
    Plan Only                   32.3%    11.4%        53.6%
  Defined Benefit and Defined
    Contribution Plan           16.1%     8.6%        80.9%

Note: Calculations based on the Health and Retirement Study.

Table 9. First Transitions from Career Jobs by 2002,
Those on Full-Time Career Jobs in 1992, by Gender,
Wage Rate, and Occupational Status
(Horizontal Percentage and Ratio)

                                   Still on     Moved to
                                    Career       Bridge
Wage Rate                  n         Job          Job

Men
  <$6/Hour                   266    15.6%        43.0%
  $6-$10/hour                576    14.3%        38.6%
  $10-$20/hour             1,394    12.4%        33.1%
  $20-$50/hour               704    15.0%        38.1%
  >$50/hour                   99    21.4%        38.9%

Women
  <$6/Hour                   437    14.1%        47.7%
  $6-$10/hour                921    18.0%        38.6%
  $10-$20/hour             1,035    15.3%        38.0%
  $20-$50/hour               218    16.4%        41.6%
  >$50/hour                   10     0.0%        58.8%

Men
  White Collar,
    Highly Skilled         1,046    14.6%        41.7%
  White Collar, Other        439    15.4%        35.3%
  Blue Collar,
    Highly Skilled         1,178    12.1%        32.3%
  Blue Collar, Other         405    17.4%        35.8%

Women
  White Collar,
    Highly Skilled           866    16.4%        43.9%
  White Collar, Other        999    17.3%        39.3%
  Blue Collar,
    Highly Skilled           312    12.8%        34.1%
  Blue Collar, Other         485    14.4%        40.2%

                                                Ratio of
                                              Bridge/(Job
                        Moved to    Don't     Bridge Job+No
Wage Rate                No Job      Know         Job)

Men
  <$6/Hour               30.9%      10.5%        58.2%
  $6-$10/hour            32.6%      14.4%        54.2%
  $10-$20/hour           42.1%      12.4%        44.0%
  $20-$50/hour           34.7%      12.2%        52.3%
  >$50/hour              24.7%      15.1%        61.2%

Women
  <$6/Hour               28.4%       9.8%        62.7%
  $6-$10/hour            31.7%      11.7%        54.9%
  $10-$20/hour           36.5%      10.3%        51.0%
  $20-$50/hour           32.1%       9.8%        56.4%
  >$50/hour              17.6%      23.5%        77.0%

Men
  White Collar,
    Highly Skilled       31.7%      12.0%        56.8%
  White Collar, Other    35.8%      13.6%        49.6%
  Blue Collar,
    Highly Skilled       43.3%      12.4%        42.7%
  Blue Collar, Other     33.5%      13.2%        51.7%

Women
  White Collar,
    Highly Skilled       30.8%       9.0%        58.8%
  White Collar, Other    32.5%      10.9%        54.7%
  Blue Collar,
    Highly Skilled       43.5%       9.6%        43.9%
  Blue Collar, Other     30.4%      14.9%        56.9%

Note: Calculations based on the Health and Retirement Study.

Table 10. Marginal Effects from Multinomial Logistic
Regression, (a) Dependent Variable: First Transition from
Full-Time Career Job, Men and Women on a Full-time Career
Job In 1992

                                     Full-Time
                                     Career Job         Bridge Job

Men                               Coef.    P-Value    Coef.    P-Value

Age in 1992
  51-54                            --        --        --        --
  55-59                          -0.1023    0.000    -0.0962    0.000
  60-61                          -0.1443    0.000    -0.0726    0.040
  62 or Older                    -0.2032    0.000    -0.0708    0.101

Health Status
  Excellent/Very Good             0.0547    0.001     0.0141    0.539
  Good                             --        --        --        --
  Poor                           -0.0034    0.896    -0.0943    0.008

Dependent Children                0.0585    0.000    -0.0141    0.592

Pension Status
  No Pension                       --        --        --        --
  Defined Benefit                -0.0370    0.041    -0.0565    0.041
  Defined Contribution            0.0127    0.521     0.0099    0.753
  Both                           -0.0497    0.140     0.0794    0.083

Self-Employed                     0.0940    0.000     0.0829    0.009

Occupational Status
  White Collar, Highly Skilled     --        --        --        --
  White Collar, Other             0.0177    0.437     -0.069    0.039
  Blue Collar, Highly Skilled     0.0028    0.888    -0.1256    0.000
  Blue Collar, Other              0.0427    0.095    -0.0976    0.009

Health Insurance
  Portable                       -0.0042    0.789     0.0322    0.162
  Not Portable                     --        --        --        --
  None                           -0.0168    0.522     0.1384    0.001

Own Home                          0.0246    0.271    -0.0966   0.0020

Wealth ($1,000)                   0.0028    0.023

Constant                         -0.1746    0.000     0.1764    0.001

                                     Full-Time
                                     Career Job         Bridge Job

Women                             Coef.    P-Value    Coef.    P-Value

Age in 1992
  51-54                            --        --        --        --
  55-59                          -0.1296    0.000    -0.0615    0.017
  62 or Older                    -0.2602    0.000    -0.0999    0.042

Health Status
  Excellent/Very Good             0.0308    0.086     0.0516    0.041
  Good                             --        --        --        --
  Poor                           -0.0412    0.213    -0.0904    0.034

Dependent Children                0.0598    0.002     0.1027    0.001

Pension Status
  No Pension                       --        --        --        --
  Defined Benefit                 0.0675    0.001    -0.0922    0.001
  Defined Contribution            0.0702    0.002    -0.0645    0.040
  Both                           -0.0383    0.495     0.2529    0.000

Self-Employed                     0.0849    0.005     0.0052    0.904

Occupational Status
  White Collar, Highly Skilled     --        --        --        --
  White Collar, Other             0.0097    0.642    -0.0269    0.359
  Blue Collar, Highly Skilled    -0.0129    0.672    -0.0412    0.330
  Blue Collar, Other              0.0091    0.743    -0.0408    0.293

Health Insurance
  Portable                       -0.0166    0.344     0.0306    0.215
  Not Portable                     --        --        --        --
  None                           -0.0821    0.020     0.1325    0.004

Own Home                         -0.0228    0.303    -0.0535    0.087

Wealth ($1,000)                  -0.0023    0.532     0.0043    0.292

Constant                         -0.1359    0.000     0.2099    0.000

(a) The regressions also control for college, ethnicity,
marriage status, wage, region, and whether a spouse works.

Note: Calculations based on the Health and Retirement Study.