Pursuing Happiness: The Architecture of Sustainable Change
Sonja Lyubomirsky
University of California, Riverside
Kennon M. Sheldon
University of Missouri—Columbia
David Schkade
University of California, San Diego
The pursuit of happiness is an important goal for many people. However, surprisingly
little scientific research has focused on the question of how happiness can be increased
and then sustained, probably because of pessimism engendered by the concepts of
genetic determinism and hedonic adaptation. Nevertheless, emerging sources of opti-
mism exist regarding the possibility of permanent increases in happiness. Drawing on
the past well-being literature, the authors propose that a person’s chronic happiness
level is governed by 3 major factors: a genetically determined set point for happiness,
happiness-relevant circumstantial factors, and happiness-relevant activities and prac-
tices. The authors then consider adaptation and dynamic processes to show why the
activity category offers the best opportunities for sustainably increasing happiness.
Finally, existing research is discussed in support of the model, including 2 preliminary
happiness-increasing interventions.
The pursuit of happiness holds an honored
position in American society, beginning with
the Declaration of Independence, where it is
promised as a cherished right for all citizens.
Today, the enduring U.S. obsession with how to
be happy can be observed in the row upon row
of popular psychology and self-help books in
any major bookstore and in the millions of
copies of these books that are sold. Indeed,
many social contexts in the United States have
the production of happiness and positive feel-
ings as their primary purpose, and questions
such as “Are you happy?” and “Are you having
fun?” fit nearly every occasion (Markus &
Kitayama, 1994). Not surprisingly, the majority
of U.S. residents rate personal happiness as very
important (Diener, Suh, Smith, & Shao, 1995;
Triandis, Bontempo, Leung, & Hui, 1990) and
report thinking about happiness at least once
every day (Freedman, 1978). Furthermore, the
pursuit of happiness is no longer just a North
American obsession, but instead it is becoming
ever more global as people seek to fulfill the
promises of capitalism and political freedom
(Diener et al., 1995; Freedman, 1978; Triandis
et al., 1990). It seems that nearly all people
believe, or would like to believe, that they can
move in an “upward spiral” (Sheldon & Houser-
Marko, 2001) toward ever greater personal
well-being.
Is the pursuit of happiness merely a bourgeois
concern, a symptom of Western comfort and
self-centeredness, a factor that has no real im-
pact on psychological adjustment and adapta-
tion? The empirical evidence suggests that this
is not the case. Indeed, a number of researchers
and thinkers have argued that the ability to be
happy and contented with life is a central crite-
rion of adaptation and positive mental health
(e.g., Diener, 1984; Jahoda, 1958; Taylor &
Brown, 1988). Bolstering this notion, Ly-
ubomirsky and her colleagues recently com-
Sonja Lyubomirsky, Department of Psychology, Univer-
sity of California, Riverside; Kennon M. Sheldon, Depart-
ment of Psychology, University of Missouri—Columbia;
David Schkade, Rady School of Management, University of
California, San Diego.
This work was supported in part by grants from the
Positive Psychology Network. We are grateful to Linda
Houser-Marko, Kathleen Jamir, and Chris Tkach for con-
ducting library research and to Shelley Taylor, David Sher-
man, and the other members of Psychology 421 for valuable
comments on a draft.
Correspondence concerning this article should be ad-
dressed to Sonja Lyubomirsky, Department of Psychology,
University of California, Riverside, CA 92521, or Kennon
M. Sheldon, Department of Psychological Sciences, 112
McAlester Hall, University of Missouri, Columbia, MO
65211. E-mail: sonja@citrus.ucr.edu or sheldonk@missouri
.edu
Review of General Psychology Copyright 2005 by the Educational Publishing Foundation
2005, Vol. 9, No. 2, 111–131 1089-2680/05/$12.00 DOI: 10.1037/1089-2680.9.2.111
111
piled evidence showing that happiness has nu-
merous positive byproducts that appear to ben-
efit individuals, families, and communities
(Lyubomirsky, King, & Diener, 2004; see also
Fredrickson, 2001). Furthermore, Lyubomirsky
et al.’s analysis revealed that happy people gain
tangible benefits in many different life domains
from their positive state of mind, including
larger social rewards (higher odds of marriage
and lower odds of divorce, more friends, stron-
ger social support, and richer social interac-
tions; e.g., Harker & Keltner, 2001; Marks &
Fleming, 1999; Okun, Stock, Haring, & Witter,
1984), superior work outcomes (greater creativ-
ity, increased productivity, higher quality of
work, and higher income; e.g., Estrada, Isen, &
Young, 1994; Staw, Sutton, & Pelled, 1995),
and more activity, energy, and flow (e.g., Csik-
szentmihalyi & Wong, 1991).
Further supporting the argument that subjec-
tive happiness may be integral to mental and
physical health, happy people are more likely to
evidence greater self-control and self-regulatory
and coping abilities (e.g., Aspinwall, 1998;
Fredrickson & Joiner, 2002; Keltner & Bon-
anno, 1997), to have a bolstered immune system
(e.g., Dillon, Minchoff, & Baker, 1985; Stone et
al., 1994), and even to live a longer life (e.g.,
Danner, Snowdon, & Friesen, 2001; Ostir,
Markides, Black, & Goodwin, 2000). Also,
happy people are not just self-centered or self-
ish; the literature suggests that happy individu-
als instead tend to be relatively more coopera-
tive, prosocial, charitable, and “other-centered”
(e.g., Isen, 1970; Kasser & Ryan, 1996; Wil-
liams & Shiaw, 1999).
In summary, happy individuals appear more
likely to be flourishing people, both inwardly
and outwardly. Thus, we argue that enhancing
people’s happiness levels may indeed be a wor-
thy scientific goal, especially after their basic
physical and security needs are met. Unfortu-
nately, however, relatively little scientific sup-
port exists for the idea that people’s happiness
levels can change for the better. For example,
the happiness-boosting techniques proposed in
the self-help literature generally have limited
grounding in scientific theory and even less
empirical confirmation of their effectiveness
(Norcross et al., 2000). Consider a representa-
tive best seller, You Can Be Happy No Matter
What: Five Principles for Keeping Life in Per-
spective, by Carlson (1997). Do the five princi-
ples work? Do some work better than others?
Do the principles work better for some people
than for others? Are any positive effects of the
principles due, ultimately, to placebo effects? If
the book actually helps people “get happier,”
does the happiness boost last? Although it is
possible that some of the advice given in this
and other similar books could well be appropri-
ate and effective, the authors provide almost no
empirical research in support of their claims.
One receives little more guidance from con-
temporary academic psychology. Of course, re-
search psychologists have identified many pre-
dictors of people’s happiness or subjective well-
being. For example, well-being has been shown
to be associated with a wide variety of factors,
including demographic status (e.g., Argyle,
1999; Diener, Suh, Lucas, & Smith, 1999; My-
ers, 2000), personality traits and attitudes (e.g.,
Diener & Lucas, 1999), and goal characteristics
(e.g., McGregor & Little, 1998). However, a
limitation of previous research is that the vast
majority of studies have been cross sectional
and have reported between-subjects effects
rather than investigating well-being longitudi-
nally and examining within-subject effects. In
addition, very few happiness intervention stud-
ies have been conducted. Thus, researchers still
know surprisingly little about how to change
well-being, that is, about the possibility of “be-
coming happier.” Doubtless, part of the reason
for this neglect is the difficulty of conducting
longitudinal and intervention studies. The prob-
lem is further compounded by the tendency of
applied mental health researchers to focus on
pathology rather than on positive mental health
(Seligman & Csikszentmihalyi, 2000) and by
the thorny issues raised when theorists speculate
on how people “should” live their lives to max-
imize their potential for happiness (Schwartz,
2000). However, we believe the principal rea-
son for the neglect of this question is the con-
siderable scientific pessimism over whether it is
even possible to effect sustainable increases in
happiness.
Historical Sources of Pessimism
Three considerations serve to illustrate the
depth of this pessimism. First is the idea of a
genetically determined set point (or set range)
for happiness. Lykken and Tellegen (1996)
have provided evidence, based on twin studies
112 LYUBOMIRSKY, SHELDON, AND SCHKADE
and adoption studies, that the heritability of
well-being may be as high as 80% (although a
more widely accepted figure is 50%; Braungart,
Plomin, DeFries, & Fulker, 1992; Tellegen et
al., 1988; cf. Diener et al., 1999). Whatever the
exact coefficient, its large magnitude suggests
that for each person there is indeed a chronic or
characteristic level of happiness. Consistent
with this idea, Headey and Wearing (1989)
found, in a four-wave panel study, that partici-
pants tended to keep returning to their own
baselines over time (see also Suh, Diener, &
Fujita, 1996). Thus, although there may be sub-
stantial variation around this baseline level in
the short term, in the long term people perhaps
cannot help but return to their set point, or to the
middle of their set range: “What goes up must
come down” (a more detailed description of the
happiness set point is provided later).
A second and closely related source of pes-
simism comes from the literature on personality
traits. Traits are cognitive, affective, and behav-
ioral complexes that are, by definition, consis-
tent across situations and across the life span
and therefore may account for part of the sta-
bility of the set point. In support of the latter
assumption, McCrae and Costa (1990) have
shown impressive long-term stability for the
“Big Five” traits, including the two traits most
closely related to well-being: neuroticism and
extraversion. Specifically, people tend to main-
tain the same rank ordering in their levels of
worry, rumination, and guilt, as well as in their
levels of social engagement, enthusiasm, and
self-confidence. Because of the close relation
between psychological well-being and these
personality characteristics, McCrae and Costa
argued that people also tend to maintain the
same relative level of happiness over time (see
also Costa, McCrae, & Zonderman, 1987; Die-
ner & Lucas, 1999).
A third source of pessimism arises from the
concept of the hedonic treadmill (Brickman &
Campbell, 1971), which suggests that any gains
in happiness are only temporary, because hu-
mans so quickly adapt to change (see also Kah-
neman, 1999; Tversky & Griffin, 1991). Thus,
although new circumstances may temporarily
cause people to become happier or sadder, they
rapidly adjust, and the effect of these new cir-
cumstances on happiness then diminishes
quickly or even disappears entirely. For exam-
ple, Brickman, Coates, and Janoff-Bulman
(1978) showed that, after 1 year, lottery winners
were no happier than controls, and furthermore
recent paralysis victims were not as unhappy as
one would expect. Further evidence of hedonic
adaptation comes from findings of remarkably
small correlations between happiness and
wealth (Diener & Lucas, 1999) and Myers’s
(2000) observation that while U.S. citizens’ per-
sonal income has more than doubled in the
past 50 years, their happiness levels have re-
mained the same. The notion of an individual
fighting against the effects of adaptation brings
to mind an image of a pedestrian walking up a
descending escalator. Although the improving
circumstances of her life may propel her upward
toward ever greater happiness, the process of
adaptation forces her back to her initial state.
Together, these concepts and findings suggest
that trying to become happier may be as futile as
trying to become taller (Lykken & Tellegen,
1996). Indeed, some have argued that pursuing
happiness may backfire altogether, if the pursuit
becomes a conscious “extrinsic” goal that dis-
tracts people from enjoying the moment
(Schooler, Ariely, & Loewenstein, in press; see
also Sheldon, 2004). Moreover, striving for
happiness may inevitably result in deep disap-
pointment for many people. From this perspec-
tive, rather than seeking an upward spiral,
maybe people would be better off simply ac-
cepting their current personality and happiness
levels (McCrae & Costa, 1994). In Zen terms,
perhaps one should try to transcend the pursuit
of happiness rather than trying to maximize it
(Gaskins, 1999). Indeed, a number of philo-
sophical traditions embrace the notion that hap-
piness should not be increased beyond an ideal
level, one akin to a “Golden Mean” (Aristotle,
1974) between agony and ecstasy. To be sure,
most people would undoubtedly reject an unre-
strained, ceaseless pursuit of well-being.
Present Sources of Optimism
Is the pursuit of happiness futile? We believe
not. Despite the seemingly compelling reasons
we have listed for pessimism regarding attempts
to elevate levels of well-being, there are also
compelling reasons for optimism. In the follow-
ing, we briefly describe four sources of opti-
mism, returning to consider some of them in
greater detail later. First, some researchers have
had success, albeit limited and short term, in
113SPECIAL ISSUE: ARCHITECTURE OF SUSTAINABLE HAPPINESS
using interventions to increase happiness (e.g.,
Fava, 1999; Fordyce, 1977, 1983, Lichter,
Haye, & Kammann, 1980; Sheldon, Kasser,
Smith, & Share, 2002). The potential of happi-
ness-enhancing interventions is further reflected
in emerging research in the positive psychology
tradition demonstrating that practicing certain
virtues, such as gratitude (Emmons & McCul-
lough, 2003), forgiveness (McCullough, Parga-
ment, & Thoresen, 2000), and thoughtful self-
reflection (King, 2001; Lyubomirsky, Sousa, &
Dickerhoof, 2004), can bring about enhanced
well-being. Furthermore, research documenting
the long-term effectiveness of cognitive and
behavioral strategies to combat negative affect
and depression has encouraging implications
for the possibility of elevating long-term happi-
ness (e.g., Gloaguen, Cottraux, Cucherat, &
Blackburn, 1998; Jacobson et al., 1996).
Second, many different motivational and at-
titudinal factors have been linked to well-being,
factors that are presumably amenable to some
volitional control. Examples of possible moti-
vational factors include the successful pursuit of
life goals that are intrinsic in content (e.g.,
Kasser & Ryan, 1996); concordant with a per-
son’s interests, motives, and values (Brunstein,
Schultheiss, & Grassman, 1998; Sheldon & El-
liot, 1999; Sheldon & Kasser, 1995); and inter-
nally consistent (e.g., Emmons & King, 1988;
Sheldon & Kasser, 1995). Examples of poten-
tially controllable attitudinal factors include the
tendency to take an optimistic perspective on
one’s life situations (e.g., DeNeve & Cooper,
1998; McCrae & Costa, 1986), the inclination
to avoid social comparisons and contingent self-
evaluations (e.g., Lyubomirsky & Ross, 1997),
and the tendency to feel a sense of optimism or
efficacy regarding one’s life (Bandura, 1997;
Scheier & Carver, 1993; Seligman, 1991; Tay-
lor & Brown, 1988).
A third reason for optimism is provided by
recent findings that older people tend to be
somewhat happier than younger people
(Charles, Reynolds, & Gatz, 2001; Diener &
Suh, 1998; Roberts & Chapman, 2000; Sheldon
& Kasser, 2001). Specifically, both cross-sec-
tional and longitudinal work has shown that
older persons report higher life satisfaction and
lower negative affect. Although these main ef-
fects do not always emerge, they are observed
frequently enough to suggest that greater hap-
piness can indeed be achieved over time, not
just by a few people but perhaps by the majority
of people. Indeed, Carstensen’s (1995) socio-
emotional selectivity theory suggests that older
people learn to structure their lives and pursue
particular goals that maximize positive emo-
tions, consistent with the proposal that people
can learn to sustainably increase their well-
being. Further supporting this notion are Shel-
don and Kasser’s (2001) results, which showed
that age-related increases in well-being are in
part mediated by volitional changes, including
older people’s ability to select more enjoyable
and self-appropriate goals.
Yet another reason why genes are not neces-
sarily destiny is that they appear to influence
happiness indirectly, that is, by influencing the
kinds of experiences and environments one has
or seeks to have. Thus, unwanted effects of
genes could be minimized by active efforts to
steer oneself away from situations that detract
from well-being or by avoiding being enticed
toward maladaptive behaviors (Lykken, 2000;
Lyubomirsky, 2001). In addition, it is worth
noting that heritability coefficients describe co-
variations, not mean levels. Furthermore, even a
high heritability coefficient for a particular trait
(such as happiness) does not rule out the possi-
bility that the mean level of that trait for a
specific population can be raised. Under the
right conditions, perhaps anyone can become
happier, even if her or his rank ordering relative
to others remains stable.
To summarize, it appears there is a paradox:
Some theoretical perspectives and empirical
data suggest that happiness can be increased,
whereas other theories and data imply that it
cannot. How can these conflicting perspectives
on the possibility of happiness enhancement be
resolved? Also, if enhanced happiness is indeed
possible, what kinds of circumstances, activi-
ties, or habits of mind are most likely to bring
gains, especially gains that can be maintained?
Model of Happiness
Accordingly, the primary question addressed
in this article is the following: Through what
mechanisms, if any, can a chronic happiness
level higher than the set point be achieved and
sustained? To this end, we describe the archi-
tecture of sustainable happiness. The integra-
tive model of happiness we present accommo-
dates the role of both personality/genetic and
114 LYUBOMIRSKY, SHELDON, AND SCHKADE
circumstantial/demographic factors in happi-
ness. However, it also goes beyond these cross-
sectional or concurrent factors to incorporate
dynamic, time-sensitive factors. This extension
allows the question of within-subject change in
well-being, and maintained change, to be ad-
dressed. Most important, the model incorpo-
rates the role of motivational and attitudinal
factors, consistent with the assumption that hap-
piness can be actively pursued. We attempt to
show that certain types of intentional activities
indeed offer ways to achieve sustainable
changes in well-being, despite the counteracting
effects of adaptation.
In the sections to follow, we first provide a
working definition of chronic happiness. Then
we define the three factors that affect it (genetic
set point, circumstances, and activities) and ar-
gue that intentional activities offer the best po-
tential route to higher and sustainable levels of
happiness. Subsequently, we consider some
more complex issues pertaining to the achieve-
ment of sustainable well-being via intentional
activity, such as the role of person–activity fit,
optimal timing and variety of activity, and the
supportive role of sustained effort and posi-
tive habits. Then, in the final section of the
article, we describe several preliminary efforts
to increase happiness, based on our model,
and discuss the nature of effective happiness
interventions.
Defining Happiness
Here we define happiness as it is most often
defined in the literature, that is, in terms of
frequent positive affect, high life satisfaction,
and infrequent negative affect. These three con-
structs are the three primary components of
subjective well-being, according to Diener and
colleagues (for reviews, see Diener, 1984, 1994;
Diener et al., 1999). Supporting the legitimacy
of considering them as indicators of the same
underlying construct, we find that the measures
are highly correlated and typically yield a single
factor after negative affect has been recoded
(Sheldon & Kasser, 1998, 2001; Sheldon &
Lyubomirsky, 2004). To refer to this group of
measures, we use the term happiness or subjec-
tive well-being, although we also discuss mood
and life satisfaction at times according to the
specific ideas and data being presented.
It is important to note as well that we use a
subjectivist definition of happiness, one that
commonly relies on people’s self-reports. We
believe this is appropriate and even necessary
given our view that happiness must be defined
from the perspective of the person. In other
words, happiness is primarily a subjective phe-
nomenon for which the final judge should be
“whoever lives inside a person’s skin” (Myers
& Diener, 1995, p. 11; see also Diener, 1994).
However, the fact that the judgment of happi-
ness is necessarily subjective does not mean that
influences on that judgment cannot be studied
empirically; for example, researchers might in-
vestigate the effects of factors such as a per-
son’s recent experiences of positive emotion
(Frijda, 1999), the frame in which the question
is presented (Larsen & Fredrickson, 1999), the
meaning that the person ascribes to the question
(Schwarz & Strack, 1999), and the person’s
sense of making satisfactory progress toward
life goals at the time of the judgment (Carver &
Scheier, 1990). We consider some of these fac-
tors in greater detail in a later section. Finally,
the fact that self-reported happiness is subjec-
tive does not mean that it is unrelated to rela-
tively more “objective” variables. For example,
research has shown significant convergence of
self-reported well-being with peer and spouse
reports of well-being (e.g., Lyubomirsky &
Lepper, 1999; Sandvik, Diener, & Seidlitz,
1993), with recall of particular types of events
(e.g., Seidlitz, Wyer, & Diener, 1997), with
smiling behavior (e.g., Harker & Keltner,
2001), and with physiological responses (e.g.,
Lerner, Taylor, Gonzalez, & Stayn, 2002).
Chronic Happiness Level
Our primary focus in this article is on a
person’s characteristic level of happiness during
a particular period in his or her life, which we
term the chronic happiness level. We define
happiness this way because we wish to identify
a quantity that is more enduring than momen-
tary or daily happiness but that is also somewhat
malleable over time and, thus, amenable to
meaningful pursuit. According to this defini-
tion, although it is possible to alter one’s
chronic happiness level, it is much more diffi-
cult to do so than to alter one’s happiness level
at a particular moment or on a particular day.
Operationally, one might define a person’s
115SPECIAL ISSUE: ARCHITECTURE OF SUSTAINABLE HAPPINESS
