Deciding to Retire From the Federal Service
| Author | Gregory B. Lewis,David Pitts |
| Published date | 01 March 2018 |
| Date | 01 March 2018 |
| DOI | http://doi.org/10.1177/0734371X16639110 |
| Subject Matter | Articles |
research-article2016
Article
Review of Public Personnel Administration
2018, Vol. 38(1) 49 –82
Deciding to Retire From the
© The Author(s) 2016
Reprints and permissions:
Federal Service
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0734371X16639110
journals.sagepub.com/home/rop
Gregory B. Lewis1 and David Pitts2
Abstract
The swelling wave of federal retirements heightens the importance of understanding
the factors that influence retirement decisions. We examine both behavior and
intentions to retire using two large data sets on federal white-collar employees aged
50 and above. Analysis of personnel records for 1979-2009 shows that several of the
usual factors—in particular, age, federal experience, and pension design—have strong
effects on when employees retire. Perhaps more revealing is our analysis of the 2012
Federal Employee Viewpoint Survey (FEVS), which indicates that satisfaction with
agency leadership and with one’s own job appear to encourage postponing retirement
plans, whereas satisfaction with supervisors and with recognition received on the job
do not. These relationships—especially those for pension design, satisfaction with
leadership, and satisfaction with one’s job—produce interesting implications for both
theory and practice.
Keywords
aging workforce issues, employee attitudes, behavior, and motivation, federal
government HRM, turnover/organizational mobility, workplace environment/culture
After a brief dip during the Great Recession, the number of federal employees retiring
annually rose nearly 50% between 2009 and 2014 (Table 1), surpassing the pre-
recession peak by 10%. Numbers of retirements rose although retirement rates in most
age groupings were markedly lower in 2014 than in 2005. The steady rise in the per-
centage of civil servants who are eligible to retire (Figure 1) means that the Great
Recession has delayed but not prevented the long-anticipated wave of retirements,
even if retirement rates do not return to their 2005 levels. The especially high
1Georgia State University, Atlanta, GA, USA
2University of California, Irvine, CA, USA
Corresponding Author:
Gregory B. Lewis, Georgia State University, P.O. Box 3992, Atlanta, GA 30302-3992, USA.
Email: glewis@gsu.edu
50
Review of Public Personnel Administration 38(1)
Table 1. Recent Trends in Federal Retirements, by Fiscal Year.
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Total workforce
3.4
3.3
3.4
3.1
2.3
2.5
3.0
3.3
3.2
3.4
Age group
50-54
2.3
2.0
1.9
1.6
1.0
1.0
1.3
1.5
1.2
1.3
55-59
9.6
9.2
9.2
8.2
5.9
6.0
7.1
7.5
6.6
6.6
60-64
17.0
16.2
16.2
14.4
10.7
11.5
13.4 14.1
13.3
13.7
65+
19.0
18.6
18.9
16.5
12.1
13.7
16.2 17.7
18.1
19.8
Total retirements 61,860 60,253 62,366 58,913 46,100 52,660 64,238 69,320 65,258 68,616
Figure 1. Age distribution of federal civil service.
concentration of Baby Boomers in the federal service (Lewis & Cho, 2011; Tobias,
2001) means that the federal government is facing the challenges of an aging work-
force sooner than most employers, making this a particularly important topic for
public-sector human resources management.
This anticipated increase in retirements should both offer benefits to and impose
costs on the federal service. The government will have an opportunity to hire
younger employees (only 7% of federal workers were under 30 in 2013; Feintzing,
2014) with skills that are perhaps more aligned to the government’s needs. However,
because retirement-eligible employees typically receive higher performance ratings
than those in their first 10 years of service (Oh & Lewis, 2013), federal managers
Lewis and Pitts
51
may prefer to retain older employees who seem to perform better. Unfortunately,
research has offered little practical advice to public managers who would like to
encourage employees to delay retirement. Although scholarly interest in public-
sector employee turnover has grown in recent years (e.g., Cho & Lewis, 2012;
Lewis & Cho, 2011; Llorens & Stazyk, 2011; Moynihan & Landuyt, 2008;
Moynihan & Pandey, 2008; Pitts, Marvel, & Fernandez, 2011), research on retire-
ments is sparse. Because retiring and quitting are fundamentally different decisions
(Topa, Moriano, Depolo, Alcover, & Morales, 2009; Y.-D. Wang, Yang, & Wang,
2012), research on the factors driving turnover likely tells us little about the deter-
minants of retirement.
This article makes two important contributions to the literature in public-sector
human resources management. First, it focuses explicitly on retirement and differenti-
ates the decision to retire from the decision simply to leave. As noted in the previous
paragraph, this is not just a semantic difference and we should not expect the same
factors to affect both decisions. Second, it uses two large data sets to shed light on
retirement action and intention. We analyze a 1% sample of federal personnel records
for white-collar employees in 1979-2009 to see the impact of age, federal experience,
pension design, and other variables on retirement. We pair this with analysis of the
2012 Federal Employee Viewpoint Survey (FEVS) to examine how satisfaction with
various aspects of the federal work environment influences retirement intention. In
both cases, we limit our analysis to employees aged 50 and above, and although we
include all relevant independent variables in our models, we focus our discussion on
factors that managers can influence.
We begin by reviewing the literature on retirement, focusing on three sets of factors
that appear to correspond to one’s decision to retire. We follow with a discussion of
our two data sources and a description of our method. We then move to a discussion of
findings and conclude with implications for research and practice.
Which Factors Influence the Decision to Retire?
Research has shown the impact of three sets of factors on an individual’s decision to
retire: (a) pension structure, age, and work experience; (b) demographic characteris-
tics; and (c) work-related attitudes. We describe research in each of these three areas
in the sections below.
Pension Structure, Age, and Work Experience
The structure of one’s pension plan plays a crucial role when employees retire.
Federal benefits vary strongly with age and experience (Asch, Haider, &
Zissimopoulos, 2002, 2004; Asch & Warner, 1999; Cowen, 2011; Lewis & Stoycheva,
forthcoming). Most federal employees hired before 1984 fall under the Civil Service
Retirement System (CSRS), whereas most of those hired since 1984 fall under the
Federal Employees Retirement System (FERS). We describe the basic components of
each below.1
52
Review of Public Personnel Administration 38(1)
CSRS is a defined-benefit pension plan. Retirees initially receive 2% of their final
average salary (abbreviated as FAS, this is the mean of their three highest years of sal-
ary) for each year they worked for the federal government, including military service.
For example, an individual who worked for the federal government for 25 years and
had an FAS of $100,000 would receive an annual pension of $50,000.2 That amount
increases annually due to a cost-of-living adjustment based on the consumer price
index, but CSRS retirees do not collect Social Security. Those with 30 years of service
can retire at 55, those with 20 years of service can retire at 60, and those with at least
5 years of service can retire at 62. Asch and Warner (1999) conclude that these ages are
also typically the normal retirement ages (NRA) for federal employees, that is, the
ages at which their expected lifetime pension benefits reach their maximum.
(Employees working past the NRA receive larger pensions but for fewer years.)
FERS, in contrast, is a “three-legged stool” pension plan. It combines Social
Security, a defined-benefit component, and a defined-contribution component. The
defined-benefit portion is roughly half as generous as that offered under CSRS:
Retirees receive 1% instead of 2% of their FAS per year of service. The defined-
contribution portion of the plan allows employees to contribute to a Thrift Savings
Plan (TSP), and the government matches contributions up to 5% of an employee’s
annual salary. The NRA for the defined-benefit portion of FERS retirees’ benefits is
the same as for CSRS (Asch & Warner, 1999), but FERS employees also need to con-
sider how age affects the other two legs of their retirement plan. Social Security allows
early retirement at 62 and retirement with full benefits at 66 or 67. Retirees can begin
drawing out of their TSP accounts without penalty at 59½ and typically qualify for
Medicare at age 65. Calculating the payout for employees under FERS is much more
complicated than doing so for those under CSRS.3
Given the structures of these two plans, we expect the probability of retirement to
increase with both age and experience, but in nonlinear ways, with jumps as employ-
ees hit their 55th, 60th, 62nd, and 65th birthdays and as they hit retirement thresholds:
30 years of service if they are at least 55, 20 years if they are at least 60, and 5 years if
they are at least 62.
Demographics
Age has attracted far more attention than other demographic characteristics, for which
findings are much more tentative. Women tend to retire later than men (Han & Moen,
1999; Honig, 1996, 1998; Quinn & Burkhauser, 1994; Talaga & Beehr, 1995), just as
they are no more likely than men to quit their jobs (Griffeth,...
To continue reading
Request your trial