Content uploaded by Elaine Wethington
Author content
All content in this area was uploaded by Elaine Wethington on Dec 19, 2015
Content may be subject to copyright.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Motivation and Emotion, Vol. 24, No. 2, 2000
Expecting Stress: Americans and
the “Midlife Crisis”1
Elaine Wethington2,3,4
Despite frequent debunking of the inevitability of the midlife crisis in the research
literature (e.g., D. A. Chiriboga, 1997; R. McCrae & P. Costa, 1990), the term
remains a media staple, implying that midlife is a time of stress and difficulties
brought about by turning 40. A recent review of midlife crisis research (O. G. Brim,
1992) concluded that midlife is not universally stressful and estimated that roughly
only 10% of American men might undergo a midlife crisis. This paper examines
the disjunction between popular and researcher views of midlife and its “crisis.”
Using semistructured telephone survey techniques, this study of 724 participants
explores the definitions that Americans hold of the “midlife crisis” and analyzes
self-reportsofmidlife crises. MostAmericans (over 90%)could provide a definition
of the midlife crisis, and these definitions roughly coincide with the definitions
used in psychological and psychoanalytic theories of the midlife crisis. Twenty-six
percent of Americans reported that they had a midlife crisis. Qualitative analyses
showed that Americans use a much wider definition of what constitutes a midlife
crisis than that used by researchers. Despite the identification of this term with
male personality development, women were as likely as men to report having had a
midlife crisis. In addition, crises occurring well before age 40 and well after age 50
were frequently nominated as midlife crises. Most participants did not attribute
their self-reported midlife crises to aging, but rather to major life events that posed
a severe threat and challenge during a very broadly-defined period of “midlife.”
1The collection of the data was supported by the John D. and Catherine T. MacArthur Foundation
Research Network on Successful Midlife Development, directed by Dr. Orville Gilbert Brim. The
author is also supported by National Institutes of Aging Program Project 2P50 AG11711-06, project 3
(E. Wethington, Principal Investigator). Allison Kavey, Alexis Krulish, Nina Delligatti, and Melissa
Treppicione provided excellent research assistance.
2Department of Human Development, Cornell University, Ithaca, New York.
3Department of Sociology, Cornell University, Ithaca, New York.
4Address all correspondence to Elaine Wethington, Department of Human Development, Cornell Uni-
versity, G52 MVR, Ithaca, New York 14853; e-mail: ew20@cornell.edu.
85
0146-7239/00/0600-0085$18.00/0 C
°2000 Plenum Publishing Corporation
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
86 Wethington
For many Americans, the term “midlife crisis” is familiar as a description of expe-
riences thought to be prevalent in midlife. This term connotes personal turmoil and
sudden changes in personal goals and lifestyle, brought about by the realization
of aging, physical decline, or entrapment in unwelcome, restrictive roles. Inten-
sive studies of self-perceived personality change in adult life (e.g., Rosenberg,
Rosenberg, & Farrell, 1999) have suggested that the term midlife crisis is used by
many American men as a metaphor for the physical and psychological changes
they perceive taking place as they age.
It thus comes as a surprise to many that several major academic studies, using
representativesamples of thepopulation ratherthan selectedcase studiesof midlife
experiences, dispute the inevitability of the midlife crisis for people in their 40s. A
recent review of research studies (Brim, 1992) found that about 10% of adult males
experience the intense period of turmoil, called the midlife crisis, and concluded
that serious emotional disturbance is the exception rather than the rule for most
middle-aged men (and women). Yet the midlife crisis remains a current media and
literary staple, indicating that journalists, novelists, movie makers, and advertisers
expect a large number of middle-aged baby boomers to respond knowingly to the
term (Brandes, 1985).
One possible explanation for the persistence of belief in the midlife crisis,
despite numerous studies that have called it into question (e.g., Chiriboga, 1997),
is that researchers and middle-aged Americans hold very different notions of what
midlife crisis means. Another complication is that researchers and theorists have
proposed and used different definitions of midlife crisis. The simplest definition,
both academic and popular, is that a midlife crisis is a difficult transition occurring
atabout theage of40. Brim (1976,p. 4)characterized the midlife crisis asinvolving
a “dislocation” or change in basic personality, manifest behavior, or sense of
identity. Resolving the crisis generates psychological distress.
There are also varying academic theories for what provokes a personality
or behavioral change at midlife. Jaques (1965) suggested that the midlife crisis
was a response to the realization of approaching death. Erikson (1963), using
a developmental task perspective, saw midlife as the period when there would
be a “struggle” between generativity and stagnation. Levinson and colleagues
(Levinson, Darrow, Klein, Levinson, & McKee, 1978) theorized that the midlife
transition (not “crisis”) was one in a set of changes over time, related to stages of
personality development. Levinson and Levinson (1996) extended the same model
to women, making several adjustments to accommodate the typically different
social and developmental life paths of women. From the Levinsonian perspective,
problems occur at the midlife transition when a person around the age of 40
perceivesthatpersonal growth has beenstymied or thwarted.Thisdistance between
current achievement and aspirations arises from personal reflection at reaching a
symbolic (or physical) marker of age.
In contrast, other researchers have portrayed midlife psychological change
as driven by outside events, and to an extent, chance and individual experience.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 87
Lowenthal, Thurnher, and Chiriboga (1975) linked the midlife crisis to passage
through socially normative events associated with midlife, with the degree or
unpleasantness of psychological adjustment and change being dependent on the
challenge of the experiences.
A characteristic the differing midlife crises perspectives share is that they
assume the passage through midlife is apt to be both eventful and stressful. Another
shared characteristic is that they seek to define the predictable crisis or expected
stress that midlife poses to an average person in contrast to other periods of life.
Another assumption is that aging itself is apt to be perceived as stressful, for
existensial reasons. Midlife symbolizes a more drastic decline to come.
A transformation seems to have taken place between the midlife crisis as a
theoretical and research concept, and the midlife crisis as a contemporary folk be-
lief (e.g., Brandes, 1985). The folk beliefs tend to overestimate the risk of having
a midlife crisis and to overstate the relative stressfulness of midlife in comparison
to other periods of life. Epidemiological study of psychological distress in adult-
hood does not suggest that midlife is a time of out-of-the-ordinary distress, for
either men or women (e.g, McCrae & Costa, 1990). Nor do stressors associated
in popular lore with the midlife crisis peak in midlife. Career crises and decisions
may be more prevalent in early career than in mid-career (Levinson et al., 1978).
Marital disruption is more characteristic of the early years of marriage than of
the longer-lived marriages of the majority of middle-aged men and women (e.g.,
Cherlin, 1992).
Research on self-perceived turning points in life (e.g., Clausen, 1995;
Thurnher, 1983), defined as periods of time when a person perceives that life
has taken a different direction, has also not found self-reported psychological
change to be more common in midlife, in comparison to other periods. Clausen’s
study (Clausen, 1995) found that the majority of self-perceived “most important”
turning points in life were reported as taking place in early adulthood, or even
adolescence. Only a minority of important turning points in life involved a midlife
change. The most profound changes took place early in life when the groundwork
was being laid for the career and relationship trajectories of adult life. Big changes
in trajectories were relatively rare in midlife.
One major reason for undertaking this study was to explore what is believed
about psychological change in adulthood. The world has changed a great deal
since the original theorists of the midlife crisis and psychological change across
life described normative adult development. For example, Jaques (1965), who
is generally given credit for coining the term “midlife crisis,” believed that the
realization death was closer shaped the midlife transition. Given the longevity rev-
olution of the twentieth century, the notion that age 40 is the time people begin
thinking about death, may be outdated. Another factor is that the life courses of
American men, and women, have become more varied (e.g., Wheaton & Gotlib,
1997). Delays in childbearing, postponement of marriage, and shifts in employ-
ment and career patterns (e.g., Cherlin, 1992) have made the life course of many
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
88 Wethington
Americans much less predictable and sure. No longer is early adulthood the pre-
scriptive time for childbearing, and the forties and fifties the only time for consol-
idation of family and career. These changes may have led to a shift in how middle
age is perceived and when its beginning and end are dated (Moen & Wethington,
1999).
The term “midlife crisis,” however, remains ubiquitous in popular culture
(Wethington, Cooper, & Holmes, 1997), rather than the more neutral term “midlife
transition.” Why the emphasis on “stress” rather than on “growth”? The analysis of
the American “culture of fear” by Glassner (1999) suggests several explanations.
The first is the sheer size and influence of the baby boom cohort, now in the
midst of middle age, and seeking to find meaning in their experiences (Rosenberg
et al., 1999). The second is media attention. What people hear and read may be
as influential as personal experience in creating beliefs about the world. Glassner
(1999, p. xxviii) argued that social commentators and media pundits, in search of
opportunities to market information as products, arouse “moral insecurities” that
evoke a culture of fear. These fears create panic over aging, even when life is going
well. Uncertainty about what constitutes a standard yardstick or proper time for
achievements symbolic of successful adulthood may in itself generate feelings that
a crisis is approaching, even when conditions of life are relatively good.
Cultural and social uncertainties may manifest themselves as concerns over
one’s personal well being and fears for the future. The paper reports findings re-
lated to three research questions on self-perceived psychological changes in adult-
hood and how they are perceived. First, it describes the beliefs average American
adults hold regarding the midlife crisis. Second, it attempts to estimate how many
Americans believe they have had a midlife crisis. Third, it examines whether re-
ports of having a midlife crisis are related to several specific types of life events or
transitions associated in popular belief with the midlife crisis, particularly events
symbolic of life collapse.
Although the findings presented are primarily descriptive and exploratory, it
is also possible to test several specific hypotheses about the midlife crisis. The
hypotheses, based on previous research and theory, are as follows:
1. Americans will report, on an average, that the midlife crisis takes place
during the 40s.
2. Given the emphasis on men, in the theoretical literature on the midlife
crisis, significantly more men than women will report having a midlife
crisis.
3. Reports of having a midlife crisis will be associated with awareness of
impending mortality or shortened future.
4. Reports of having a midlife crisis will be associated with life events and
transitions symbolic of radical life changes in middle age, such as divorce
and job loss.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 89
METHODS
Sample and Procedures
The data for this study are intensive interviews with a sample of 724 adults
aged 28–78. The Psychological Turning Points Study (PTP) was a follow-up study
to the MacArthur Foundation National Study of Midlife (MIDUS; Mroczek &
Kolarz, 1998). The 724 PTP respondents were randomly selected from the national
random digit dial (RDD) sample of MIDUS (N=3,032).
The PTP collected data on self-reports of the midlife crisis and other types
of psychological changes. The interviews took place in 1997 and 1998. The inter-
views took on an average 43 min. The response rate was 82.7%. The sample had
49.4% males and 50.6% females (Wethington, Kessler, & Pixley, in press).
The MIDUS data, and the PTP, were known to not represent the U.S. popula-
tion accurately (Mroczek & Kolarz, 1998). To increase the national representative-
ness of the PTP sample, weights were created for analyses. The weights adjusted
for differences in (1) the probability of selection and (2) differential nonresponse
by SES, race, age, gender, and other factors.
Measures
Life Events
Interviewersinquiredabout eventsand transitions overthepast5 years, probed
responses for clarity and meaning and dated them to month and year of occur-
rence. Semistructured probing was used to encourage complete and detailed re-
sponses. The interview included questions about significant types of major life
events known to be associated with psychological distress, including divorce, sep-
aration, job loss, marriage and remarriage, serious health problems, and other
major life crises. These questions were excerpted from the Structured Life Events
Interview (Wethington, Brown, & Kessler, 1995). The interview also included a
screen for depression episodes over the same 5 years, using the short form screen-
ing scale for DSM depression, developed from the World Health Organization’s
CompositeInternational Diagnostic Interview(Kessler,Andrews,Mroczek,Ustun,
& Wittchen, in press).
The Midlife Crisis
Respondents were also asked five questions about the midlife crisis. The
questions were placed after other questions about psychological experiences
(Wethington et al., in press) and before the questions about life events and
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
90 Wethington
depression. The five questions about the midlife crisis, including probes for pos-
itive responses, are reported in Appendix. The questions about the midlife crisis
were piloted in two preliminary studies (see Wethington et al., 1997).
Responses to the midlife crises questions were coded for content. First, study
staff coded them for beliefs about what constitutes a midlife crisis. Second, staff
codedthe actual experienceofthe midlifecrisisfor those whoreported that theyhad
one.All entrieswere codedindependently bytwo trained coders, and discrepancies
were resolved by the author. Interrater agreement between the coders was .84.
Age Groups for Analysis
Reports of the midlife crisis were analyzed by gender and age group. The age
groups, which varied across specific analyses, were designed to capture critical
transition points between Levinson’s life stages (Levinson et al., 1978; Levinson
& Levinson, 1996), as well as the relatively quiet periods between transitions.
Levinson’s theory of life stages is widely known to both researchers and the public
(e.g., Sheehy, 1976). Crisis periods are believed to occur at the time of transition
from one to another period of life.
RESULTS
Self-Reported Midlife Crises
The raw data from the intensive study (PTP) indicate that 26% of respondents
(25.4% of the men and 26.3% of the women) report having had a midlife crisis
in the past. Reports of having had a midlife crisis increase across the life span.
Among respondents who have reached the age of 50, 35.2% (34% of the men,
36.1% of the women) reported that they had experienced a midlife crisis.
The number of self-reported midlife crises is much higher than previous
estimates of the prevalence of the midlife crisis (Brim, 1992), all of which used
investigator definitions of the crisis rather than self-report. One likely reason for
the wide discrepancy is that study participants were using a wider, less precise
definition of the midlife crisis than what researchers typically use. The next section
examines characteristics of the midlife crisis, as reported by study participants.
Americans’ Descriptions of the Midlife Crisis
The findings reviewed in the introduction to this paper suggest that the midlife
crisis is not as prevalent as contemporary folk beliefs predict. Narrative data from
the PTP showed that Americans have complex, varied beliefs about the midlife
crisis. Most saw it as a time of stress and confusion. Others seemed to anticipate
it as a gift of adult life, and a few even planned to welcome it (Rosenberg et al.,
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 91
Table I. Beliefs About the Midlife Crisis, Percent Reported by Gender (PTP, N=724)
Men (%) Women (%)
Awareness of aging/time passing 20.9 19.7
Life review or reevaluation 14.2 11.2
Change in personal approach to life 14.0 14.2
Events and transitions associated with middle age/aging
Family changes 4.5 6.0
Job loss/career/disappointment 4.7 2.2
Health crises 2.0 7.4
Other life events 5.3 4.6
Mental health problems 2.0 1.6
Overtly skeptical view (e.g.,“does not exist”) 16.8 21.9
Positive growth 0.8 1.6
Other 2.2 3.6
Don’t know 12.6 6.0
1999). But still others held skeptical views, saying that the midlife crisis is just a
reflection of the national tendency to whine and complain even when things are
going well (Samuelson, 1996).
Most Americans readily recognized the term midlife crisis. In PTP, only 9.2%
(12.6% of men and 6% of women) were unable to provide a definition (see Table I).
They were also in consensus that it occurs during the 40s, for most people. Re-
spondents were asked, “At what age do you believe someone might have a midlife
crisis?”Men andwomenreported that themidlife crisis wouldoccur,on an average,
atabout theage of 46.Men reported, onan average,thatage 46.1 isthe expectedage
forthe midlifecrisis, women age46.7, on anaverage(F=1.056,df =1,690, ns).
When dating their own reported midlife crises, however, men reported a slightly
lower age than women did (Men =45.0, Women =47.6), although the difference
was not statistically significant (F=2.32,df =1,190), ns).
To summarize, 26% reported having had what they consider to be a midlife
crisis, a figure much higher than the 10% estimated by Brim (1992). Because
this number relied on respondent self-report rather than clinical or investigator
judgment,these reports wereexaminedmore closelytosee whethertheexperiences
reported matched previous theoretical descriptions of the midlife crisis.
Respondents tended to define the midlife crisis in a way generally consistent
with the research definitions of the midlife crisis, albeit varying definitions. More
respondents described it as internally motivated (Levinson & Levinson, 1996)
rather than externally caused (Lowenthal et al., 1975). A large number of respon-
dents (20.9% of the men, 19.7% of the women) said that awareness of aging and
time passing was the cause of the crisis, more colloquially as the feeling that
“time is passing you by” or “you’re at the halfway point” (see Table I). Only a
small number of respondents explicitly connected the midlife crisis to feelings of
impending mortality or approaching death, as Jaques (1965) defined the crisis.
Rather, they saw midlife as a time to catch-up to where they would like to be,
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
92 Wethington
or expected to be when they were younger, a formulation more consistent with
Levinson’s stages theory of adult development (Levinson & Levinson, 1996), as
popularized by Sheehy (1976).
About 14.2% of men and 11.2% of women defined the midlife crisis as a time
of life review or reevaluation. These mentions reflected uncertainty about future
direction in life. “You question everything,” said one respondent. Other responses
were more positive, or neutral, such as feeling that a new part of life is starting,
that midlife is a time for reassessment or taking stock, or that midlife involves
continual learning and analysis. Others characterized the midlife crisis as a time
of difficult decisions.
Another 14%, both men and women, defined the midlife crisis as a period
of time when someone makes a major change in personal approach to life. Some
respondents explicitly defined the change as having a second childhood or adoles-
cence. Other respondents believed that the midlife crisis involves doing something
out of character, such as (jokingly) “wearing your shirt wide open.” Many of these
responses condemned people who have midlife crises, criticizing those who make
big life changes without “taking the feelings of others into account.” Others crit-
icized the concept of the midlife crisis as a justification for making bad choices,
not dealing with reality, living in a fantasy world, abandoning moral values, acting
childishly, or justifying selfish actions that hurt their families.
Although most definitions emphasized internal motivation or aging as the
cause for midlife crisis, a number of respondents defined the midlife crisis as a
crisis in response to life events and transitions (Lowenthal et al., 1975). Most
of these events are associated with middle age or aging. About 4.5% of men
and 6% of women believed that the midlife crisis comes about in reaction to
changes in a family normatively associated with middle age. These events include
death of parents, siblings, and other relatives; increasing illness of similar aged
others; marital difficulties leading to separation or divorce; extramarital affairs;
children leaving home; feeling less emotionally close to children; disappointment
in children’s achievements; and feeling “sandwiched” between older and younger
dependents. More women than men reported events involving children (see also
Ryff, Schmutte, & Less, 1996).
Among men, 4.7% believed that the midlife crisis is provoked by job loss and
career disappointments (only 2.2% of women defined the midlife crisis this way).
Such events included being blocked from further promotions, forced retirements,
disappointment with job duties, excessive stress at work, and career changes.
Other nominated events were personal health crises. Two percent of men
and 7.4% of women attributed the midlife crises to increasing health problems.
Surprisingly, only a handful (<1%) mentioned menopause or sexual dysfunction
as the cause of the midlife crisis.
It should be noted, however, that some respondents simply attributed the
midlife crisis to vague causes, such as “increasing stress” and “the stress of mid-
dle age,” or to idiosyncratic events they believed (from personal or observational
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 93
experience) caused what they call a midlife crisis in themselves or others. Ex-
amples of such unique events are legal troubles or moving, events drawn from
personal experience.
A few respondents (2% of men, 1.6% of women) defined the midlife crisis as
mental health problems. Some respondents specifically mentioned that they (or a
friend, or their spouse) developed mental health problems, anxiety, or depression
because of severe events.
The majority of respondent definitions emphasized the normality, or expect-
edness of midlife stress, and following that, a crisis. Yet a sizable proportion of the
participants had very skeptical views of the midlife crisis, criticizing it as an idea,
or being critical of people who felt that they were having one; 16.8% of men and
21.9% of women were skeptics. A number of respondents refused to define the
term, saying, “that term doesn’t mean anything,” “it’s just an excuse, “it’s a crutch,”
or “it’s stupid and immoral.” “Some people are always in crisis, that’s what I think
it is,” one respondent said. Some of the responses took on a very condemnatory
tone, such as “the midlife crisis is about not growing up,” or it is about “not tak-
ing responsibility” for one’s actions or “copping out when your family needs you
the most.”
Finally, a few respondents (<1% of men, 1.6% of women) had a very positive,
growth-oriented view of the midlife crisis. Sample responses included, “You don’t
get older, you get better,” “It’s a learning experience,” and “It’s freeing.” In sum,
most beliefs about the midlife crisis were negative.
Self-Reported Age at Midlife Crisis
Given the variety of definitions that Americans give for the midlife crisis,
it is perhaps not surprising that those who reported having had a midlife crisis
themselves would have a tendency to place it in a decade other than the 40s.
“Midlife” is an elastic term, encompassing the thirties and the sixties, as well as
the forties and fifties. In fact, most self-reported midlife crises reported in the study
occurred before the age of 40, and after the age of 50 (see Fig. 1). The youngest
age reported for the crisis was 17, and the oldest 75 (see Fig. 1).
The age of self-reported midlife crisis was significantly related to age of the
participant. Respondents younger than 40 on an average reported crises occurring
before 40, and respondents older than 60 on an average reported crises later in life
(see Fig. 2).
“Off-Time” Midlife Crises
The reports of “off-time” midlife crises were real events or crises, a great
majority of them (in the judgment of the investigator) very severe and threatening.
However, from a “life stages” theoretical standpoint (Levinson & Levinson, 1996)
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
94 Wethington
Fig. 1. Distribution of self-reported midlife crises, by gender and age at crisis.
Fig. 2. Age differences in reported age at midlife crisis.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 95
they did not meet a basic criterion of what constitutes a midlife crisis, which is a
crisis brought on by turning 40.
To explore differences between “on-time” and “off-time” midlife crises, all
reported crises were classified into categories, those taking place between ages 38
and 50 (n=90), and those that took place “off-time” from the life stages perspec-
tive, before age 38 (n=34), and after age 50 (n=66). In all, 52.6% of reported
midlife crises were classified as “off time.” The age boundaries chosen as “on-
time” for the midlife crisis were generous. They assume that some people might
have midlife crises in anticipation of turning 40, and that others had crises that
took many years to resolve. The large number of reported “off-time” midlife crises
were likely to reflect uncertainty about exactly what type of crisis should count as
a midlife crisis, and also lack of consensus about what years of life comprise the
midlife transition. The age boundaries thus imposed yielded a conservative, but
more theoretically defensible estimate of the prevalence of midlife crises. They
eliminated a number of crises that may be more appropriately classified as “age 30
transition” or “age 50 transition” crises (Levinson & Levinson, 1996).
Figure 3 reports the percentage of men and women of each age group who
believe that they experienced a midlife crisis, at ages 38 through 50, both inclusive.
The age groups were chosen to capture Levinson’s stages of life schema (Levinson
et al., 1978). Respondents under the age of 39 were not included, because they
were not “eligible” for experiencing a midlife crisis. Overall 14.4% of respondents
aged 39 and over reported having had a midlife crisis that met the investigator age
criterion. The traditional view of the midlife crisis would predict that more men
Fig. 3. Percent reporting midlife crisis that occurred between ages 38 and 50, by age and gender of
respondent.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
96 Wethington
than women would report believing that they had a midlife crisis. This was not
the case. The data showed that 15.5% of men, and 13.3% of women, reported
having had a midlife crisis from ages 38 through 50. The gender difference was
not significant.
The probability of having had a midlife crisis was expected to increase across
the 40s. For women aged 39–53, this was what happens. For men, however, those
aged 44–48 were slightly less likely to report having had a midlife crisis than those
aged 39–43. Reports of having had a midlife crisis peaked in the 49–53 age group,
but declined thereafter. Because these were cross-sectional data, it is impossible
to determine whether the decline across age groups was due to forgetting (falloff),
denial, lack of knowledge of the term midlife crisis (likely in older age cohorts),
or historical factors.
To summarize the major trends, women were as likely as men to report a
midlife crisis. People who report having had a midlife crisis at any time during
their lives were slightly more likely to be presently in decade transition points
in the life course (age groups 39–43, 49–53), but this pattern is not statistically
significant. Finally, relatively few people reported having a midlife crisis in the
period of time suggested by the developmental stage theory. The final statement
was all the more notable because the midlife crisis is self-defined, and the age
boundaries used as a minimum criterion are very generous.
Age Differences in Reports of “On-Time” Midlife Crises
Qualitative analyses presented earlier implied that the definition of the midlife
crisis is very elastic in the minds of everyday Americans. Participants reported
events and crises that did not fit research criteria for midlife crises.
It was also likely that reports of midlife crises were affected by retrospective
recall error. Younger participants were asked to recall events and crises in the rel-
atively recent past. Older participants were asked to recall events that may have
occurred in the very distant past. Research on the recall of life events (e.g., Brown,
Sklair, Harris, & Birley, 1973) has long noted that the recall of events falloff over
time as specific details of the event fade in memory. Research on autobiographi-
cal memory has shown that people reinterpret past events to fit social expectations
(and survey questions; McFarland, Ross, & Giltrow, 1992; Rubin, Rahhal, & Poon,
1998). Taken together, research on recall of life events and autobiographical mem-
ory implies that reports on midlife crises that took place 20 years ago should be
less valid than reports of midlife crises that took place in the last few years. In
addition, this research also implies that more recent life events may be reported as
“midlife crises” in a survey interview even if they did not occur in midlife.
To examine the impact of memory biases on reports of the midlife crises,
reports from three age groups of respondents were compared. The first was the
group aged 39–43. This group was judged to be the most likely to report having
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 97
had a midlife crisis that fit theoretical criteria. First, this group was likely to be
familiarwith theterm, andto consider it personally salient, given their age. Second,
this group was being asked to recall relatively recent events. The second group
examined was aged 49–53. This group was overall the most likely to have reported
having had a midlife crisis. The third group consisted of participants aged 64 and
over, who were recalling a crisis that took place 15–30 years before. If there were
significant autobiographical memory biases reflected in the recall and reporting of
midlife crises, then there should be predictable differences between what people
of various ages recounted as a midlife crisis.
Specifically, we predicted that those aged 39–43 would report situations that
were more like the traditional midlife crisis. They were likely to interpret the
midlife crisis question as “how are you coping with becoming 40?” We predicted
that respondents aged 64 and over would more likely report major life events
that in retrospect they believe triggered a “crisis in midlife.” They were likely
to interpret the midlife question as “What event during your middle age was the
biggest crisis?” The group aged 49–53 was predicted to display both tendencies
because they were recalling events that occurred from the previous month up to
nearly 12 years before.
The numbers involved here were small, but the results were suggestive, and
consistent with the memory bias perspective. In the 39–43 age group, 67% of
the reports described a crisis brought about by approaching or turning age 40.
These included concern over changing physical appearance and stamina, dissatis-
faction with achievements so far in life, and expectations of fewer opportunities for
achievement in the future. The remaining reports involved life events and mental
illness. In contrast, 89% of the midlife crises reported by respondents aged 64 and
older described life events, many of them severe. For men, these events tended
to involve either job or marriage. For women, the events involved health, family
deaths, and marriage. Only 11% of this group reported a midlife crisis provoked
byturning 40.As predicted,the reports inthe 49–53age groupwere almost roughly
divided between reports of crises attributed to turning 40 and reports of crises at-
tributed to life events. Forty-four percent reported psychological crises brought
about by perceived physical decline, changing priorities, blocked achievement,
and shortened futures. Forty percent reported severe life events as midlife crises.
The remaining 16% reported life events that could have been brought about by
changes in appraisal caused by turning 40, specifically being trapped in unsatis-
factory jobs, and feeling the need to make career changes before it was too late.
Midlife Crises and Major Life Events
Our fourth hypothesis predicted that people in middle age who experience
events associated with the midlife crisis should be more likely to report a midlife
crisis. As described earlier, many respondents connected the midlife crisis to life
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
98 Wethington
Table II. Recent Life Events, Recent Depression, and Risk of Reporting a Midlife Crisis,
Respondents Aged 38–55, Logistic Regression (n=324)
BSE(B) Wald df Exp(B)
Age in years 0.07∗0.03 1.55 1 1.07
Femalea−0.51 0.30 2.97 1 0.60
Some collegeb−0.14 0.29 0.22 1 0.87
History of depressionc0.90∗0.38 5.55 1 2.46
Divorced or separated, past 5 yearsd1.88 1.25 2.26 1 6.56
Job loss, past 5 yearsd−0.22 0.48 0.20 1 0.81
Death of close friend, family, past 5 yearsd−0.27 0.30 0.81 1 0.77
Other major life crisis, past 5 yearse−0.27 0.32 0.70 1 0.76
Note. Constant =−2.65; χ2=17.77.
aMale =0, female =1.
bAt least some college =1, all others =0.
cProbable episode of depression in the past =1, all others =0.
dEvent occurred =1, all others =0.
eEvents occurred =1, all others =0. Examples are health problems, legal difficulties, major financial
losses, and crime victimization.
∗p<.05.
events such as job loss or forced unemployment, early retirement, extramarital
affairs, divorce, separation, deaths of close friends or family members, and other
major life crises, such as health problems.
The data allowed a stronger, although far from comprehensive, test of this
notion. The interview asked participants if they had lost their job or become un-
employed, gotten divorced or separated, experienced the death of a close friend or
family member, or had any other major life crisis in the past 5 years. To evaluate
this prediction, the odds of reporting a midlife crisis, after experiencing events
like these, was calculated, using multivariate logistic regression. To simplify in-
terpretation, the calculation includes respondents aged 38–55 only. The results are
reported in Table II.
Only two predictors in the logistic equation were significant. Older partici-
pants were more likely to report a midlife crisis (B=0.07,SE(B)=0.03,p<
.05). Those with a history of depression episodes were significantly more likely
as well (B=0.90,SE(B)=0.38,p<.05). Overall, having experienced a recent
event symbolically associated with midlife crises was not significantly related to
reporting one. For example, a person who had experienced job loss or unemploy-
ment was just 0.81 times more likely to report a midlife crisis (Exp(B)=0.81).
A person who was divorced or separated in the past 5 years was 6.56 times more
likely to report a midlife crisis than someone who was not recently divorced or
separated; however, the estimate was not significant (only eight people reported
being divorced or separated in the last 5 years). The model overall fits the data
very poorly. Only 8.3% of instances of midlife crises were correctly predicted by
this model.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 99
DISCUSSION
This study tested four specific hypotheses relating to self-reported midlife
crises, using a combination of qualitative and quantitative analyses. The first
hypothesis was that Americans would report that the midlife crisis takes place
in the fourth decade of life. This hypothesis was supported, at least on an average.
Americans believe that, on an average, other Americans have midlife crises in
their 40s. But when asked about their own midlife crises, Americans were more
inclusive, describing personal crises that began anytime from age 17 to 75. In
fact, the majority who reported midlife crises described crises outside the midlife
transition age boundary (ages 38 to 50, both inclusive).
The second hypothesis was that more men than women would report having
had midlife crises. This was not the case. Men and women were almost equally
likely to report having had midlife crises.
The third hypothesis was that reports of having a midlife crisis would be
associated with the awareness of impending death or shortened future. In general,
Americans’ definitions of the midlife crises were much more diverse than this con-
ceptualization implies. Only a few responses explicitly mentioned the awareness
of impending death, although many emphasized an awareness of aging, declining
health and stamina, and not enough time to accomplish goals.
In contrast, many Americans associated the midlife crisis with normal aging,
not imminent death. Their general descriptions of the crisis reflected popular belief
about what happens as you transit middle age. When describing their own crises,
however, respondents applied the term midlife crisis to a variety of past, significant
events and crises that occurred in the course of life, any time from age 30 to 70.
A parsimonious explanation for why the self-reported prevalence of the midlife
crisisis sohigh isthat respondentswere reinterpretingpast experienceto answer an
unusual question (cf. Ross & Newby-Clark, 1998). It also implies a parsimonious
explanation for why beliefs that the midlife crisis is a common risk of aging are
so persistent. Almost any event or feeling socially symbolic of aging can qualify
as a midlife crisis, if the definition is very elastic.
The fourth hypothesis was that reports of having a midlife crisis would be
associated with life events and transitions symbolic of maturity and aging. This
hypothesis received mixed support. On the one hand, many respondents believed
that the midlife crisis was provoked by events symbolic of middle age in the United
States today, such as career dissatisfaction, forced early retirement, and extramar-
ital affairs. Many respondents also attributed their own self-reported midlife crises
to their divorces, career disasters, health problems, and the empty nest. But on
the other hand, middle-aged people who recently underwent some of these same
crises were not that much more likely to label them as midlife crises. The report
of having a midlife crisis did not reduce to experiencing a negative event symbolic
of middle age.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
100 Wethington
The term midlife crisis remains a powerful metaphor. Its ubiquity in de-
scriptions of middle-aged men, in television advertisements for sports cars, and
in literary fiction (e.g., Roth, 2000) speaks to its standing as a staple of popular
culture. There is considerable evidence in the data that people at midlife use it as
a tool for constructing meaning in their lives. For example, a majority of people
aged 38–43 who believed they had had a midlife crisis produced textbook descrip-
tions of midlife crises, perhaps based on textbooks used in their college psychology
courses. A great majority (around 90% of the PTP sample) could provide a def-
inition for the crisis, and these definitions correspond fairly well with research
definitions, albeit differing ones.
The preceding data suggest, consistent with previous research, that the expe-
rience of the midlife crisis as a result of turning 40 (or 30, or 50) is not a universal
factor of adult life, for either men or women. Why do so many Americans be-
lieve that they have experienced a midlife crisis? It is notable that a majority of
people who reported having had a midlife crisis mentioned crises that took place
outside the investigator-marked boundary of ages 38 through 50. Almost cer-
tainly some of these reports came about because the boundary of what constitutes
midlife has now stretched to include the years leading up to retirement (Moen &
Wethington, 1999). Another reason respondents may have reported midlife crises
outside the 38–50 age boundary is that self-perceived psychological changes are
prevalent across the life course (Wethington et al., in press).
Limitations of the Study
The data utilized in this paper were relatively straightforward, based pri-
marily on self-report. Self-report data on the midlife crisis are seriously limited.
Time was short even in the intensive interviews, and as a consequence probing
was standardized rather than open-ended. Most previous studies of the midlife
crisis used clinical judgments, which are far superior. Not everyone could present
an articulate narrative of the midlife crisis. More highly skilled interviewers, or
clinical investigators, would have improved the quality of the data collected. On
the other hand, use of a telephone survey made it possible to collect data from
a representative sample of Americans, rather than a local sample available to a
clinical interviewing team.
Others may argue that the study underestimated the real prevalence of the
midlife crisis. The questioning may not have been thorough enough to evoke
reliableand valid reports.Anotherpossibility isthatmany people, particularlymen,
may deny having had a crisis (Rosenberg et al., 1999). The state of crisis implies
thatone maybe mentallyill orunable to cope effectivelywith thechallenges oflife.
On the other hand, the content coding scheme we applied to the midlife crisis
relies very heavily on the life stage paradigm. Other interpretations of these data,
based on other paradigms, are possible.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 101
Social Implications and Conclusions
The data interpretations presented here, although preliminary, have numerous
implications for studies of perceived psychological change, including studies of
the midlife crisis. The findings suggest a number of useful ways to analyze beliefs
about psychological change.
Although researchers make fine distinctions between different types of psy-
chological experiences and changes related to aging, those who are less familiar
with the theories justifying the concepts are more inclusive about “what counts.”
Another possibility is that crises occurring earlier in life and later in life pose
greater emotional and practical difficulties than crises in midlife (cf. Clausen,
1997; Erikson, 1963; Thurnher, 1983). The dataset, although very rich, did not
provide all of the measures necessary to come to a firm conclusion whether any of
these explanations is correct (probably all are correct to some extent).
Should we be worried that many Americans believe that there is a predictable
crisis lurking for them in their middle years? It is possible to take several perspec-
tives on this issue. One prominent sociologist (Glassner, 1999) has recently decried
“the culture of fear,” which he defined as overestimation of the risks of danger in
the context of objectively favorable, even improving social circumstances. For ex-
ample, Americans believe that crime is a very serious issue, even as it is declining
at unprecedented levels (Glassner, 1999). From this perspective, overestimation of
risk and danger is problematic, because beliefs in the population affect the alloca-
tion of resources. Big, intractable problems are overlooked, in favor of devoting
public resources to less likely ones. Does such reasoning apply to beliefs about
the midlife crisis? One way in which it could is by the diversion of scarce mental
health treatment resources toward socially-created and mild ailments, instead of
toward treatment of real and much more serious mental illnesses.
Finally, many responses hinted that the myth of the midlife crisis has its good
side. The term midlife crisis gives meaning to experiences that may be unusual,
and may provide helpful ways to appraise and cope with events. The idea of an
expectable crisis (“everyone has one”) may provide comfort for those who are
having a difficult time in their middle years (Rosenberg et al., 1999). The use of
the term “crisis” may encourage those having a particularly difficult time to seek
help and social support.
APPENDIX: MIDLIFE CRISIS QUESTIONS
1. Peopleoftenuse theterm“midlifecrisis” todescribeimportantexperiences
during their middle years. What does that term mean to you?
2. At what age do you believe someone might have a midlife crisis?
3. Have you ever have something you would consider a midlife crisis?
4. How old were you when this happened?
5. Briefly, what was that about?
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
102 Wethington
REFERENCES
Brandes, S. H. (1985). Forty: The age and the symbol. Knoxville: University of Tennessee Press.
Brim, O. G. (1976). Theories of the male midlife crisis. The Counseling Psychologist, 6, 2–29.
Brim, O. G. (1992). Ambition. New York: Basic Books.
Brown, G. W., Sklair, F., Harris, T. O., & Birley, J. L. T. (1973). Life-events and psychiatric disorders.
Part I: Some methodological considerations. Psychological Medicine, 3, 74–87.
Cherlin, A. (1992). Marriage, divorce, remarriage. Cambridge, MA: Harvard University Press.
Chiriboga, D. A. (1997). Crisis, challenge, and stability in the middle years. In M. E. Lachman &
J. B. James (Eds.), Multiple paths of midlife development (pp. 293–322). Chicago: University of
Chicago Press.
Clausen, J. A. (1995). Gender, contexts, and turning points in adults’ lives. In P. E. Moen, G. Elder, &
K. Luscher (Eds), Examining lives in context: perspectives on the ecology of human development
(pp. 365–389). Washington, DC: American Psychological Association.
Clausen, J. A. (1997). Life reviews and life stories. In J. Z. Giele & G. H. Elder (Eds.), Methods of life
course research: Qualitative and quantitative approaches (pp. 189–212). Thousand Oaks, CA:
Sage.
Erikson, E. H. (1963). Childhood and society. New York: Norton.
Glassner, B. (1999). The culture of fear: Why Americans are afraid of the wrong things. New York:
Basic Books.
Jaques, E. (1965). Death and the midlife crisis. International Journal of Psychoanalysis, 46, 502–514.
Kessler, R. C., Andrews, G., Mroczek, D., Ustun, B., & Wittchen, H. -U. (in press). The World Health
Organization Composite International Diagnostic Interview Short-Form (CIDI-SF). International
Journal of Methods in Psychiatric Research.
Levinson, D. C., Darrow, E. B., Klein, M. H., Levinson, M. H., & McKee, B. (1978). The seasons of
a man’s life. New York: Knopf.
Levinson, D., & Levinson, J. (1996). The seasons of a woman’s life. New York: Knopf.
Lowenthal, M. J., Thurnher, M., & Chiriboga, D. A. (1975). Four stages of life. San Francisco: Jossey-
Bass.
McCrae, R., & Costa, P. (1990). Personality in adulthood. New York: Guilford.
McFarland,C., Ross, M., & Giltrow, M.(1992). Biased recollections in olderadults:The role of implicit
theories of aging. Journal of Personality and Social Psychology, 57, 1069–1081.
Moen, P. E., & Wethington, E. (1999). Midlife development in a life course context. In S. L. Willis &
J. D. Reid (Eds.), Life in the middle (pp. 3–23). San Diego: Academic.
Mroczek, D., & Kolarz, D. (1998). The effect of age on positive and negative affect: A developmental
perspective on happiness. Journal of Personality and Social Psychology, 75, 1333–1349.
Rosenberg, S. D., Rosenberg, H. J., & Farrell, M. P. (1999). The midlife crisis revisited. In S. L. Willis
& J. D. Reid (Eds.), Life in the middle (pp. 47–73). San Diego: Academic.
Ross, M., & Newby-Clark, I. (1998). Construing the past and the future. Social Cognition, 16, 133–150.
Roth, P. (2000). The human stain. Boston: Houghton Mifflin.
Rubin, D. C., Rahhal, T. A., & Poon, L. W. (1998). Things learned early in adulthood are remembered
best. Memory and Cognition, 26, 3–19.
Ryff, C. D., Schmutte, P. S., & Less, Y. H. (1996). How children turn out: Implications for parental self-
evaluation. In C. D. Ryff& M. M. Seltzer (Eds.), The Parental experience in midlife (pp. 383–422).
Chicago: University of Chicago Press.
Samuelson, R. (1996). The good life and its discontents: The American Dream in the age of entitlement
1945–1995. New York: Times Books.
Sheehy, G. (1976). Passages. New York: E. P. Dutton & Co.
Thurnher, M. (1983). Turning points and developmental change: Subjective and “objective” assess-
ments. Journal of Orthopsychiatry, 53, 52–60.
Wethington, E., Brown, G. W., & Kessler, R. C. (1995). Interview measurement of stressful life events.
In S. Cohen, R. C. Kessler, & L. U. Gordon (Eds.), Measuring stress: A guide for health and
social scientists (pp. 59–79). New York: Oxford University Press.
Wethington, E., Cooper, H., & Holmes, C. S. (1997). Turning points in midlife. In I. H. Gotlib &
B. Wheaton (Eds.), Stress and adversity across the life course: Trajectories and turning points
(pp. 215–231). New York: Cambridge University Press.
P1: FTK
Motivation and Emotion [me] PP007-290057 November 2, 2000 9:28 Style file version Nov. 19th, 1999
Crises and Turning Points 103
Wethington, E., Kessler, R. C., & Pixley, J. E. (in press). Psychological turning points and the “midlife
crisis.” In C. D. Ryff & R. C. Kessler (Eds.), Midlife in the United States. Chicago: University of
Chicago Press.
Wheaton, B., & Gotlib, I. N. (1997). Trajectories and turning points across the Life Course: Concepts
and Themes. In I. N. Gotlib & B. Wheaton (Eds.), Stress and adversity across the life course:
Trajectories and turning points (pp. 1–25). New York: Cambridge University Press.
