The contributions to Insights in the Economics of Aging uncover how financial, physical, and emotional well-being are integrally related. The authors consider the interactions between financial circumstances in later life, such as household savings and home ownership, physical circumstances such as health and disability, and emotional well-being, including happiness and mental health.
|Publisher:||University of Chicago Press|
|Series:||National Bureau of Economic Research Conference Report|
|Sold by:||Barnes & Noble|
|File size:||28 MB|
|Note:||This product may take a few minutes to download.|
About the Author
Read an Excerpt
Insights in the Economics of Aging
By David A. Wise
The University of Chicago PressCopyright © 2017 National Bureau of Economic Research
All rights reserved.
Trends in Pension Cash-Out at Job Change and the Effects on LongTerm Outcomes
Philip Armour, Michael D. Hurd, and Susann Rohwedder
Promoting financial security in retirement is a major objective of US policies governing employer-provided pensions. To encourage workers and employers to participate, legislation mandates very large tax advantages for private-pension savings. These effectively represent "tax expenditures" to the federal government in the form of forgone tax revenues.
US policymakers have a substantial interest in the results of these large expenditures for promoting financial security in retirement. Is the private-pension system effectively enhancing financial security in retirement? What are the barriers or impediments to achieving economic security for old age among US workers? Which groups of workers are at greatest risk of falling short?
One feature of the US pension system in particular may jeopardize the objective of promoting retirement-income security: the ability of workers to cash out (i.e., withdraw funds from) their private-pension plans upon job separation. Federal rules aim to discourage such preretirement cash-outs. For example, the Tax Reform Act of 1986 introduced a 10 percent tax penalty on withdrawals from tax-advantaged accounts prior to the age of 59.5. Burman et al. 2012 showed that this tax penalty reduced preretirement cash-out of pension balances and increased rollovers into individual retirement accounts that preserve the tax-advantaged status of the pension balances. They also found reductions of cash-outs in response to a 1992 reform that imposed 20 percent tax withholding (without affecting the total tax liability).
These policy changes have reduced, but not eliminated, early withdrawals. As we document below, following workers in their early fifties in 1992 through subsequent job separations, 13.6 percent of those with a defined-contribution (DC) plan cashed out all or part of their plan; among workers with defined-benefit (DB) plans, 18.9 percent cashed out. For later cohorts the percent cashing out was substantially higher, even exceeding a 50 percent increase for the latest cohort in our study.
Several studies have investigated the causes of these early pension withdrawals now subject to withholding. It appears that a significant portion of these are made by households facing liquidity constraints and experiencing financial shocks (Amromin and Smith 2003; Scherpf 2010). Still, according to Butrica, Zedlewski, and Issa (2010), about half of early withdrawals from 401(k) defined-contribution pension accounts and individual retirement accounts (IRAs) could not be attributed to the events observed in the data, possibly indicating "unnecessary loss of retirement savings."
This chapter uses the long panel of data collected in the Health and Retirement Study (HRS), spanning up to twenty years for the earliest cohorts, to add new insights to prior research findings on this topic. Analyses in the current study addressed trends in pension cash-outs among older workers, cohort differences, and retirement-income security metrics at later years or ages and their relations to earlier job and cash-out choices. We did not restrict ourselves to looking at single cash-out actions, but incorporated cumulative measures of pension cash-out decisions. The chapter includes analyses of precipitating events that shed light on determinants of cash-out behavior and how it may have changed over time. We were especially interested in how the Great Recession affected cash-out choices. The HRS data allowed relation of variation in cash-out choices of older workers to a variety of outcomes observed in panel up to twenty years later, including assets, income, and health.
In an antecedent to this chapter, Hurd and Panis (2006) analyzed HRS data on cash-outs and other dispositions of pension entitlements among workers over the age of fifty who left their jobs between 1992 and 2000 (five waves of biennial HRS data). In this study, they found 13 percent of pension entitlements were cashed out, representing 5.3 percent of entitlement dollars. Among plans with a lump sum option, 20 percent were cashed out. However, their study highlighted an issue that had been underappreciated in prior research: whether a lump-sum distribution (LSD) harms retirement preparation depends critically on what the worker does with the money, and whether these cash-outs represent "leakage" from wealth available to finance consumption in retirement. Some LSDs may be rolled into an IRA, some may be annuitized, and some may be cashed out. Only the last of these may harm retirement preparation, and even then some uses may function as savings. Hurd and Panis use the following graphic to clarify the situation. In the cash-out branch, some of the funds may be invested or saved directly and some may be invested in the home, which is a form of saving. While bringing such funds out of tax-sheltered accounts may not be optimal tax management, it is primarily spending for current consumption among those not facing binding credit constraints that poses the greatest harm to economic preparation for retirement (see figure 1.1).
Hurd and Panis established several facts that are important for understanding the causes and consequences of LSD decisions. Not all plans allow an LSD on job separation. In fact, the availability of LSDs varies dramatically across types of plans: a little over 80 percent of DC-plan participants report an LSD option versus just 42 percent of DB-plan participants.
Besides looking at the fraction of workers who cashed out their pensions, Hurd and Panis examined the implications of cash-outs for aggregate pension balances and net wealth, including nonretirement wealth. They identified two factors that implied a limited overall impact of cash-outs on retirement and total household wealth. First, cashed-out plans had lower average value than other plans, especially among those holding DC plans. Second, over 75 percent of cashedout funds were either invested or used to pay off debt. Hurd and Panis (2006, 2226) conclude that "among workers that are within roughly ten years of retirement, only a small fraction of pension plan dollars is consumed immediately after job separation and that the vast majority is preserved for retirement income security."
While the Hurd and Panis paper provided a useful perspective up through the year 2000, the demographic and pension landscape has changed considerably with the decreasing importance of DB plans, the increasing pension entitlement of women, and changing trends in marriage and divorce. Furthermore, the Great Recession may have led to more cash-outs, harming particular segments of the population. These changes in the landscape warrant revisiting the Hurd and Panis analysis, which is the objective of this chapter.
The HRS is a biennial longitudinal survey of persons at least fifty years of age. Since its launch in 1992, the HRS has gathered data on income, work assets, pension plans, health insurance, disability, physical health and functioning, cognitive functioning, and health-care expenditures, among other topics. Periodic additions of cohorts ensure the HRS remains representative of the population at least fifty years of age.
The analyses in this chapter are focused on several key variables. We analyzed self-reported data on employer-provided pensions for HRS respondents. The HRS asks whether respondents own such a pension, and whether it is a defined-benefit (DB) or a defined-contribution (DC) plan. It also asks respondents whether the pension plan allows for a lump-sum distribution. They are asked about the disposition of the pension plan at job separation or retirement: whether it was left with the former employer to accumulate; whether a full or partial LSD was taken; whether DB holders started drawing benefits on separation or chose to await future, larger benefits; whether the pension plan was lost with separation (likely where there is lack of vesting); or whether some other disposition occurred. For those who took an LSD, the survey asks whether the money was rolled into an IRA, converted to an annuity, or cashed out. For those who cashed out their pension plan, the HRS asks whether the money was saved or invested, whether it was used to pay off debt or to purchase durable goods or a home, or whether it was used for nondurable consumption.
This research updates and expands that of Hurd and Panis in several directions. First, more waves of the HRS are now available. Hurd and Panis used five waves of HRS data from 1992 through 2000. Since then, six more waves of HRS surveys — from 2002 through 2012 — have been conducted and the data made available for analysis, bringing not only an increase in sample size, but also an expansion in the types of analyses that could be conducted. In particular, because additional waves of data became available, differences across cohorts (e.g., those born before World War II and postwar "baby boom" cohorts) could be analyzed. A growing number of DC plans is also available for analysis, partly because of the time elapsed since 2000, but perhaps more importantly because DC plans have become increasingly prevalent in the US pension system, so workers in more recent cohorts are more likely to have them.
More recent cohorts are also likely to consist of more women who have earned pension entitlements at work. Their decisions regarding pension wealth may differ from those of men and merit additional analysis. Indeed, within a household, the behavior of both spouses is important in determining use or disposal of pension assets. The incorporation of more waves of HRS data with more female respondents who hold pension wealth promotes the analysis of pension wealth and its use or disposal from a household perspective.
The analysis has been updated to provide insights on the effects of the Great Recession on pension behavior, particularly on cash-outs. The earlier work by Hurd and Panis studied a period of relatively low unemployment and high stock market and housing returns. The years since then, particularly those surrounding the Great Recession that began in 2008, have not been as favorable. Unemployment in 2009 reached 10 percent, more than 2 percentage points higher than it was at any point between 1992 and 2000, and more than double what it was in the late 1990s. Though eventually recovering, the US stock market lost about half its value during the Great Recession, and housing values decreased by more than one-third, representing a large shock to wealth that may have led some workers to cash out their pensions. Indeed, using tax data on preretirement withdrawals, Argento, Bryant, and Sabelhaus (2015) verified that workers substantially increased withdrawal rates between 2004 and 2010, especially after 2007.
The long HRS panel supports analyses of retirement-security outcomes at later years or ages and how they relate to earlier job and cash-out choices. For example, consider a fifty-seven-year-old worker who cashed out a pension between 1992 and 1994. We have been able to observe that worker's subsequent economic position at age seventy-five in 2012, and we could then compare that worker with otherwise similar workers who did not cash out.
By gaining access to more years of data, we were able to analyze and compare a broader array of events precipitating cash-out, including whether different precipitating events led to differences in subsequent events. We could, for example, analyze and compare cash-outs resulting from adverse health changes, unemployment, shocks to household wealth caused by the Great Recession, marital disruption, and extractions to buy real estate during the housing bubble of 2004 to 2008 and the subsequent loss of equity and, possibly, home ownership during the Great Recession.
1.2.1 Changes in the Macroeconomic Environment, 1992–2012
We begin with an overview of the contextual changes occurring over the period 1992–2012. The first half of that period covers the HRS waves available to Hurd and Panis in conducting their analysis, and the second half folds in the years covered by the current work. We specifically focus on labor force participation (LFP) and the recessions that characterized the macro-economy near the beginning and toward the end of the period of interest.
Labor Force Participation
Using Current Population Survey (CPS) data, we examined trends in LFP by sex. As shown in figure 1.2, between the early 1990s and the early 2010s, LFP among males ages sixty-five to sixty-nine increased substantially, whereas LFP among males forty to fifty-four decreased slightly. Men of intermediate age (fifty-five to sixty-four) increased their LFPs modestly, if at all. The LFPs among older women (figure 1.3), ages fifty-five to sixty-nine, increased at rates matching those of the oldest men in the analysis, although there appears to have been a leveling off following the start of the Great Recession. The LFPs among women in their forties exhibited a slight increase or stasis until around the turn of the century, and a slight downward trend thereafter.
Clearly, the most dramatic trends are the LFP increases among older men and women. These increases reflect trends toward later retirement. In the descriptive analyses, which compare cohorts over eight years, we thus expect to see trends toward relatively fewer separations due to retirement, which may alter the frequency of pension cash-outs.
We are here concerned with the recession of 1991 and the Great Recession beginning in 2008. They are of interest because a recession is characterized by unemployment and adverse financial outcomes — loss of income and loss of assets, including the value of stocks and real property. Involuntary job losses could trigger pension cash-out particularly when accompanied by wealth losses.
Recession of 1991. Unemployment, which had been falling in the late 1980s from around 7.5 percent to 5 percent, turned around with the recession to exceed 7 percent again in 1992 (all figures seasonally adjusted). Stocks simultaneously dropped in value; the Standard & Poor's 500 index lost some 15 percent of its worth in 1991. Value of housing was not so dramatically affected. The Case-Schiller house price index had been falling for several years and bottomed out in 1991. (The Federal Housing Finance Agency's house price index showed no movement, but had just been established.)
Great Recession. While changes in macroeconomic indexes were noticeable in 1991, they were much more dramatic for the Great Recession that began in 2008 (see figure 1.4). The unemployment rate had been falling for several years to 4.3 percent, or down about 20 percent since 2002. In the second half of 2007 it began rising and continued doing so very rapidly until the end of 2009, when it topped out at more than 10 percent.
The Standard & Poor's 500 index had been rising since 2003, making up some losses from 2002 and eventually reaching some 35 percent over the 2002 datum. It then plummeted through 2008, losing more than half its value. The Federal Housing Finance Agency (FHFA)'s US house price index had increased dramatically, by about 40 percent, between 2002 and the middle of 2007. It then began a long downturn that by early 2011, when it leveled off, it had lost almost half of the gain.
1.3.1 Cohort Comparisons
Table 1.1 defines the cohorts and shows sample sizes. For example, we follow as Group 1 the 5,355 people who entered the HRS with the 1992 wave of data collection. Of these 5,355 persons, 3,871 were working at entry, 2,161 were working with pension coverage, and 1,396 were working and covered by a pension plan allowing a lump-sum distribution (LSD). We follow these groups for eight years, as their participants age from fifty-one to fifty-six up to fifty-nine to sixty-four. Group 4 only entered in 2010, so insufficient time has elapsed for a longitudinal analysis; we use this group for baseline comparisons only.
Labor Force Status. Figure 1.5 shows labor force status at age fifty-one to fifty-six, as reported by the respondents in each group. Employment was lower in 1992 and 2010, and unemployment was higher particularly in 2010, reflecting the Great Recession.
Pension Coverage and Plan Type. Pension coverage improved modestly over the period of interest (see figure 1.6), increasing a few percentage points to a 60 percent coverage rate in 2010. There was a large change in the type of coverage, though. Most respondents who had pensions were covered by defined-benefit (DB) plans versus defined-contribution (DC) plans in 1992. By 1998, that pattern switched around. The trend from DB to DC still continues.
The great majority — over 80 percent — of persons having a DC pension plan are allowed by the plan to cash out via an LSD (see figure 1.7). The like percentage for DB plans is 47 in 2010, which represents a steady but modest increase since 1992.
Excerpted from Insights in the Economics of Aging by David A. Wise. Copyright © 2017 National Bureau of Economic Research. Excerpted by permission of The University of Chicago Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of ContentsPreface
David A. Wise and Richard Woodbury
1. Trends in Pension Cash-Out at Job Change and the Effects on Long-Term Outcomes
Philip Armour, Michael D. Hurd, and Susann Rohwedder
Comment: James M. Poterba
2. Liquidity in Retirement Savings Systems: An International Comparison
John Beshears, James J. Choi, Joshua Hurwitz, David Laibson, and Brigitte C. Madrian
Comment: Daniel McFadden
3. House Price Volatility and the Housing Ladder
James Banks, Richard Blundell, Zoë Oldfield, and James P. Smith
Comment: Steven F. Venti
4. What Determines End-of-Life Assets? A Retrospective View
James M. Poterba, Steven F. Venti, and David A. Wise
Comment: Brigitte C. Madrian
5. Understanding the Improvement in Disability-Free Life Expectancy in the US Elderly Population
Michael Chernew, David M. Cutler, Kaushik Ghosh, and Mary Beth Landrum
Comment: Jonathan Skinner
6. Are Black-White Mortality Rates Converging? Acute Myocardial Infarction in the United States, 1993–2010
Amitabh Chandra, Tyler Hoppenfeld, and Jonathan Skinner
Comment: David R. Weir
7. Measuring Disease Prevalence in Surveys: A Comparison of Diabetes Self-Reports, Biomarkers, and Linked Insurance Claims
Florian Heiss, Daniel McFadden, Joachim Winter, Amelie Wuppermann, and Yaoyao Zhu
Comment: James P. Smith
8. Challenges in Controlling Medicare Spending: Treating Highly Complex Patients
Thomas MaCurdy and Jay Bhattacharya
Comment: Hidehiko Ichimura
9. Movies, Margins, and Marketing: Encouraging the Adoption of Iron-Fortified Salt
Abhijit Banerjee, Sharon Barnhardt, and Esther Duflo
10. Suicide, Age, and Well-Being: An Empirical Investigation
Anne Case and Angus Deaton
Comment: David M. Cutler
11. Does Retirement Make You Happy? A Simultaneous Equations Approach
Raquel Fonseca, Arie Kapteyn, Jinkook Lee, and Gema Zamarro
Comment: Anne Case