Content uploaded by Andrew Clark
Author content
All content in this area was uploaded by Andrew Clark
Content may be subject to copyright.
PARIS-JOURDAN SCIENCES ECONOMIQUES
48, BD JOURDAN – E.N.S. – 75014 PARIS
TEL. : 33(0) 1 43 13 63 00 – FAX : 33 (0) 1 43 13 63 10
www.pse.ens.fr
WORKING PAPER N° 2006 - 29
The curved relationship between
subjective well-being and age
Andrew E. Clark
Andrew J. Oswald
JEL Codes : C23, I31, J10
Keywords : Happiness, ageing, well-being
CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE – ÉCOLE DES HAUTES ÉTUDES EN SCIENCES SOCIALES
ÉCOLE NATIONALE DES PONTS ET CHAUSSÉES – ÉCOLE NORMALE SUPÉRIEURE
halshs-00590404, version 1 - 3 May 2011
1
The Curved Relationship Between
Subjective Well-Being and Age
Andrew E. Clark
PSE and IZA, Paris, France
Email: andrew.clark@ens.fr
Andrew J. Oswald
University of Warwick, UK
Email: andrew.oswald@warwick.ac.uk
halshs-00590404, version 1 - 3 May 2011
2
The Curved Relationship Between
Subjective Well-Being and Age
October 10, 2006
Abstract
This article is concerned with a body of work on happiness and age represented by
important papers such as Mroczek and Kolarz (1998) and Mroczek and Spiro (2005).
Using a large British data set, the paper presents new longitudinal evidence. It also
points out that, perhaps unknown to many psychologists, a parallel literature on this
topic exists in economics journals. The paper shows that subjective well-being follows
a U-shape through the life course. We argue that eventually the two literatures will
have to be made consistent with one another, and suggest that, although it is not easy to
live in both worlds, with their different styles and conventions, economists and
psychologists still have much to learn from one another.
Keywords: Happiness; ageing; well-being
Corresponding author: andrew.oswald@warwick.ac.uk.
Address: Department of Economics, University of Warwick, Coventry CV4 7AL,
United Kingdom.
Telephone: (+44) 02476 523510
Acknowledgements: Paris-Jourdan Sciences Economiques (PSE) is a Joint Research
Unit CNRS-EHESS-ENPC-ENS. The second author’s work was funded by an
ESRC professorial research fellowship
halshs-00590404, version 1 - 3 May 2011
3
The Curved Relationship Between
Subjective Well-Being and Age
The relationship between happiness and age is important but still imperfectly understood. On
page 1336 of an interesting article in this journal, Mroczek and Kolarz (1998) state that, to
their knowledge, “no previous study has considered the possibility of nonlinear relationships
between age and happiness.” A follow-up study by Mroczek and Spiro (2005) is in the same
spirit but uses different methods. The former article finds that positive affect has an
upwardly curved component; the latter suggests that subjective well-being follows an
inverted U-shape and peaks at around age 60. On our reading of these two papers, they are
important work. They are currently acquiring high numbers of citations in the Social Science
Citation Index. A literature in psychology journals is growing up around them.
We present new longitudinal evidence on this issue. Its chief finding does not seem to exist
in the psychology journals. A second aim of our article is to alert psychologists to a parallel
literature on the nonlinear relationship between happiness and age. It is research published
mainly in economics journals. This stream of work may have originated from Clark and
Oswald (1994). Using a 12-item measure of psychological well-being, and a cross-section of
6100 randomly sampled Britons from the first wave of the British Household Panel Survey,
that paper found mental well-being to be U-shaped in age. The authors’ data were for 1991.
The paper drew the conclusion that the happiness minimum along the U is reached around a
person’s mid-30s.
Since then, a body of work by economists and statisticians has suggested, across a range of
settings, that the relationship between subjective well-being and age is curved. Almost every
study replicates the basic conclusion: that happiness is U-shaped in age, with a minimum
halshs-00590404, version 1 - 3 May 2011
4
point in a range from the mid-30s to approximately the late-40s. The result has now been
demonstrated for perhaps fifty nations. It has been found especially in data from Western
countries such as the United States, the United Kingdom, Australia, France, Italy, Belgium,
Germany, Spain, and Switzerland.
The U-shape in age is documented in, for example, Gerlach and Stephan (1996), Oswald
(1997), Theodossiou (1998), Winkelmann and Winkelmann (1998), Di Tella et al (2001),
Frey and Stutzer (2002), Helliwell (2003), Blanchflower and Oswald (2004), Graham (2005),
Frijters et al (2004, 2005), Senik (2004), Van Praag and Ferrer-i-Carbonell (2004), Clark
(2005), Long (2005), Shields and Wheatley Price (2005), Propper et al (2005), Powdthavee
(2005), Bell and Blanchflower (2006), and Uppal (2006). Blanchflower (2001) documents
the U-shape in 23 East European transition nations. Di Tella et al (2003) provides the same
for 12 separate Western industrial nations. Clark et al (1996) makes a similar argument for
job satisfaction equations, and also gives some results for mental well-being equations.
Recently, Blanchflower and Oswald (2006), using data on approximately half a million
Europeans and Americans, argues that allowance for the inclusion of cohort effects makes
little difference to the finding of a U-shape.
These papers’ methods vary slightly. Nevertheless, each of them, as in the work of Mroczek
and Kolarz (1998), draws upon multiple-regression models. In this way, factors other than
age are held constant.
We here report briefly an updated version of the results of Clark and Oswald (1994). Since
then, the British Household Panel Survey has happened annually, so it is possible to re-do
those calculations on a longitudinal data set now exceeding 100,000 observations. A new
halshs-00590404, version 1 - 3 May 2011
5
test can now be done, in which both fixed-effects and age can be allowed for in a multiple
regression equation framework.
One variable we use is life satisfaction. It is measured on a seven point scale from 1 (not
satisfied at all) to 7 (completely satisfied). Over all 14 waves of the BHPS, this has a mean
of 5.19 with standard deviation 1.26.
Another is a General Health Questionnaire (GHQ) score. This is a questionnaire-based
method of measuring psychological well-being and has been used widely in the health and
epidemiological research literature (for recent examples, see Cardozo et al 2000, Martikainen
et al 2003, and Pevalin and Ermish 2004). It amalgamates answers to the following list of
twelve questions:
Have you recently:
1. Been able to concentrate on whatever you are doing?
2. Lost much sleep over worry?
3. Felt that you are playing a useful part in things?
4. Felt capable of making decisions about things?
5. Felt constantly under strain?
6. Felt you could not overcome your difficulties?
7. Been able to enjoy your normal day-to-day activities?
8. Been able to face up to your problems?
9. Been feeling unhappy and depressed?
10. Been losing confidence in yourself?
11. Been thinking of yourself as a worthless person?
12. Been feeling reasonably happy all things considered?
halshs-00590404, version 1 - 3 May 2011
6
Responses are made on a four-point scale of frequency of a feeling in relation to a person's
usual state: "Not at all", "No more than usual", "Rather more than usual", and "Much more
than usual".
The between-item validity of the GHQ-12 is high in the BHPS, with a Cronbach alpha of
0.89. For simplicity, we use the so-called Caseness GHQ score, which counts up the number
of questions for which the response is in one of the two "low well-being" categories. This
count is then reversed, to convert it into a well-being rather than ill-being score, which thus
runs from a possible low of 0 (all twelve responses indicating poor psychological health) to a
high of 12 (no responses indicating poor psychological health). The well-being GHQ
measure has mean of 10.14 and standard deviation of 2.92.
Life-satisfaction and GHQ measures of subjective well-being seem valuably complementary.
They capture different aspects of affect. However, as we show below, the two exhibit a
similar pattern over the life course.
Although the size of the BHPS data set has expanded since the initial research was done, and
it is now possible to control for people’s unchanging dispositions (fixed effects), the main
conclusion remains approximately the same as in the simpler evidence of Clark and Oswald
(1994). Both wellbeing-GHQ and life satisfaction are U-shaped across age groups. In the
full longitudinal data set here from 1991 to 2004, it reaches a minimum in the age range 40-
49, which is a little later than the early estimate that came out of our single-year of 1991 data.
The coefficients on the other age-bands reveal a fairly smooth curve -- down and then up --
through the lifespan. This finding is not sensitive to the exact specification of the well-being
equations. The pattern in the raw data is illustrated in Figure 1 for the latest wave at our
halshs-00590404, version 1 - 3 May 2011
7
disposal, which is 2004. As is clear from the Figure, it is not necessary to use regression
equations to see the approximate U-shape across age-groups.
It is known that many variables shape subjective well-being (Diener et al 1999, Easterlin
2003, Lucas et al 2004, and Gilbert 2006). If we control in the regression equations for
region, year, income, the work status of the individual (ie. whether employee, self-employed,
unemployed, retired from work), the number of children, gender, marital status, whether a
renter or home-owner, the level of education, and subjective physical health, then, once
again, well-being reaches a minimum in the age band 40-49.
For the formal analysis, as say reported using simple methods in Table 1, it makes no
substantive difference whether we use logistic regression or simple methods like Ordinary
Least Squares (OLS), or random-effects or fixed-effects models. The third column of Table
1 follows the same individuals through time -- annually from 1991 to 2004 -- whilst holding
constant their unchanging fixed-effects, and allowing for many other characteristics such as
marital status. By using banded age dummy-variables, in a way that to our knowledge has
not been done before in longitudinal work of this sort, in either the psychology or economics
literatures, Table 1 provides clear evidence of a U-shape.
Our finding of a U-shape using longitudinal data does not appear to exist in the psychology
literature. It is not clear why the results of this economics research literature differ in some
significant ways from, say, those in Mroczek and Spiro (2005), which finds almost the
reverse -- an inverted U-shape in age. It seems important that future work explore these
interesting issues more fully. A deeper theoretical understanding of the empirical happiness-
age curve will also be required.
halshs-00590404, version 1 - 3 May 2011
8
Psychologists were working on well-being data some decades before researchers in
economics (Bradburn 1969, Diener 1984). We hope, nevertheless, that this short article
might be a useful guide to economists’ writings on happiness and age.
Over recent decades, economists -- including ourselves -- have been guilty of not reading the
psychology journals sufficiently. Psychologists, similarly, may not have read the economics
journals as much as might be ideal. Despite the methodological differences, both sides
would, arguably, benefit if we could learn to communicate more effectively.
halshs-00590404, version 1 - 3 May 2011
9
Figure 1. Mean Life-Satisfaction and Wellbeing-GHQ in British Data. BHPS Wave 14 (2004)
Mean Life Satisfaction. BHPS 2004
4.7
4.8
4.9
5
5.1
5.2
5.3
5.4
5.5
5.6
16-19 20-29 30-39 40-49 50-59 60-69
Mean GHQ. BHPS 2004
9.85
9.9
9.95
10
10.05
10.1
10.15
10.2
10.25
10.3
10.35
16-19 20-29 30-39 40-49 50-59 60-69
halshs-00590404, version 1 - 3 May 2011
10
Table 1. Well-Being Regressions. BHPS Waves 1-14.
No controls Demographic
controls Demographic
controls plus
Individual Fixed
Effects
Age 20-29 -0.154** -0.230** -0.161**
(0.021) (0.023) (0.025)
Age 30-39 -0.224** -0.380** -0.170**
(0.023) (0.029) (0.034)
Age 40-49 -0.315** -0.487** -0.187**
(0.024) (0.031) (0.042)
Age 50-59 -0.155** -0.334** -0.165**
(0.026) (0.032) (0.050)
Age 60-69 0.188** -0.053 -0.061
(0.028) (0.038) (0.059)
Constant 5.339** 5.023** 4.938**
(0.019) (0.036) (0.496)
Observations 90567 87797 87797
R-squared 0.014 0.146 0.025
No controls Demographic
controls Demographic
controls plus
Individual Fixed
Effects
Age 20-29 -0.093** -0.223** -0.228**
(0.036) (0.040) (0.045)
Age 30-39 -0.248** -0.399** -0.374**
(0.042) (0.053) (0.064)
Age 40-49 -0.347** -0.452** -0.464**
(0.045) (0.057) (0.081)
Age 50-59 -0.228** -0.186** -0.314**
(0.048) (0.059) (0.099)
Age 60-69 0.157** 0.197** -0.106
(0.049) (0.067) (0.119)
Constant 10.265** 9.204** 9.625**
(0.032) (0.067) (1.247)
Observations 145424 141875 141875
R-squared 0.003 0.119 0.032
Notes. The omitted age band is 16-19 years old. Standard errors in parentheses.
* significant at 5%; ** significant at 1%. The controls in columns 2 and 3 are for
income, labour force status, number of children, sex, marital status, education,
health,and wave and regional dummies.
Life Satisfaction
GHQ
halshs-00590404, version 1 - 3 May 2011
11
References
Blanchflower, D.G. (2001). Unemployment, well-being and wage curves in Eastern and
Central Europe. Journal of Japanese and International Economies, 15, 364-402.
Blanchflower, D.G., & Oswald, A.J. (2004). Well-being over time in Britain and the USA.
Journal of Public Economics, 88, 1359-1386.
Blanchflower, D.G, & Oswald, A.J. (2006). Is well-being U-shaped over the life cycle?
Unpublished paper, Dartmouth College, US.
Bradburn, N.M. (1969). The structure of psychological well-being. Chicago: Aldine.
Cardozo, B.L., Vergara, A., Agani, R., Gotway, C.A. (2000). Mental health, social
functioning, and attitudes of Kosovar Albanians following the war in Kosovo. Journal
of the American Medical Association, 284, 569-577.
Clark, A.E. (2005). Your money or your life: Changing job quality in OECD countries.
British Journal of Industrial Relations, 43, 377-400.
Clark, A.E., & Oswald, A.J. (1994). Unhappiness and unemployment. Economic Journal,
104, 648-659.
Clark, A.E., Oswald, A., & Warr, P. (1996). Is job satisfaction U-shaped in age? Journal of
Occupational and Organizational Psychology, 69, 57-81.
Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542-575.
Diener, E., Suh, E.M., Lucas, R.E., & Smith, H.L. (1999). Subjective well-being: Three
decades of progress. Psychological Bulletin, 125, 276-302.
Di Tella, R., MacCulloch, R.J., & Oswald, A.J. (2001). Preferences over inflation and
unemployment: Evidence from surveys of happiness. American Economic Review, 91,
335-341.
Di Tella, R., MacCulloch, R.J., & Oswald, A.J. (2003). The macroeconomics of happiness.
Review of Economics and Statistics, 85, 809-827.
Easterlin, R.A. (2003). Explaining happiness. Proceedings of the National Academy of
Sciences, 100, 11176-11183.
Frey, B.S., & Stutzer, A. (2002). Happiness and economics. Princeton University Press:
Princeton.
Frijters, P., Haisken-DeNew, J.P., & Shields, M.A. Money does matter! (2004). Evidence
from increasing real income and life satisfaction in East Germany following
reunification. American Economic Review, 94, 730-740.
Frijters, P., Haisken-DeNew, J.P., & Shields, M.A. (2005). The causal effect of income on
health: Evidence from German reunification. Journal of Health Economics, 24, 997-
1017.
Gerlach, K., & Stephan, G. (1996). A paper on unhappiness and unemployment in Germany.
Economics Letters, 52, 325-330.
Gilbert, D. (2006). Stumbling on happiness. Alfred A Knopf: New York.
Graham, C. (2005). Insights on development from the economics of happiness. World Bank
Research Observer, 20, 201-231.
Helliwell, J. (2003). How’s life? Combining individual and national variables to explain
subjective well-being. Economic Modelling, 20, 331-360.
Long, A. (2005). Happily ever after? A study of job satisfaction in Australia. The Economic
Record, 81, 303-321.
Lucas, R.E., Clark A.E., Georgellis Y., & Diener E. (2004). Unemployment alters the set
point for life satisfaction. Psychological Science, 15, 8-13.
Martikainen, P., Adda, J., Ferrie, J.E., Smith, G.D., Marmot, M. (2003). Effects of income
and wealth on GHQ depression and poor self-rated health in white collar women and
halshs-00590404, version 1 - 3 May 2011
12
men in the Whitehall II study. Journal of Epidemiology and Community Health, 57,
718-723.
Mroczek, D.K., & Kolarz, C.M. (1998). The effect of age on positive and negative affect: A
developmental perspective on happiness. Journal of Personality and Social Psychology,
75, 1333-1349.
Mroczek, D.K., & Spiro, A. (2005). Change in life satisfaction during adulthood: Findings
from the Veterans Affairs Normative Aging study. Journal of Personality and Social
Psychology, 88, 189-202.
Oswald, A.J. (1997). Happiness and economic performance. Economic Journal, 107, 1815-
1831.
Pevalin, D.J., & Ermisch, J. (2004). Cohabiting unions, repartnering and mental health.
Psychological Medicine, 34, 1553-1559.
Powdthavee, N. (2005). Unhappiness and crime: Evidence from South Africa. Economica,
72, 531-547.
Propper, C., Jones, K., Bolster, A., Burgess, S., Johnston, R., & Sarker, R. (2005). Local
neighbourhood and mental health: Evidence from the UK. Social Science and
Medicine, 61, 2065-2083.
Senik, C. (2004). When information dominates comparison – Learning from Russian
subjective panel data. Journal of Public Economics, 88, 2099-2123.
Shields, M.A., & Wheatley Price S. (2005). Exploring the economic and social determinants
of psychological well-being and perceived social support in England. Journal of the
Royal Statistical Society (Series A), 168, 513-537.
Theodossiou, I. (1998). The effects of low-pay and unemployment on psychological well-
being: A logistic regression approach. Journal of Health Economics, 17, 85-104.
Uppal, S. (2006). Impact of the timing and severity of disability on the subjective well-being
of individuals with disabilities. Social Science and Medicine, 63, 525-539.
Van Praag, B., & Ferrer-i-Carbonell, A. (2004). Happiness quantified: A satisfaction
calculus approach. Oxford University Press: Oxford.
Winkelmann, L., & Winkelmann, R. (1998). Why are the unemployed so unhappy? Evidence
from panel data. Economica, 65, 1-15.
halshs-00590404, version 1 - 3 May 2011