INTERPERSONAL RELATIONS AND GROUP PROCESSES

Wealth and Happiness Across the World: Material Prosperity Predicts Life
Evaluation, Whereas Psychosocial Prosperity Predicts Positive Feeling

Ed Diener Weiting Ng
University of Illinois at Urbana–Champaign and Singapore Institute of Management
The Gallup Organization, Omaha, Nebraska

James Harter and Raksha Arora

The Gallup Organization, Omaha, Nebraska

The Gallup World Poll, the first representative sample of planet Earth, was used to explore the reasons why
happiness is associated with higher income, including the meeting of basic needs, fulfillment of psychological
needs, increasing satisfaction with one’s standard of living, and public goods. Across the globe, the association
of log income with subjective well-being was linear but convex with raw income, indicating the declining
marginal effects of income on subjective well-being. Income was a moderately strong predictor of life
evaluation but a much weaker predictor of positive and negative feelings. Possessing luxury conveniences and
satisfaction with standard of living were also strong predictors of life evaluation. Although the meeting of
basic and psychological needs mediated the effects of income on life evaluation to some degree, the strongest
mediation was provided by standard of living and ownership of conveniences. In contrast, feelings were most
associated with the fulfillment of psychological needs: learning, autonomy, using one’s skills, respect, and the
ability to count on others in an emergency. Thus, two separate types of prosperity— economic and social
psychological— best predict different types of well-being.

Keywords: income, subjective well-being, money, wealth, happiness

Does money increase happiness? Behavioral scientists have
studied this popular question, and several conclusions have
emerged. For example, people in wealthy nations on average are
considerably happier than those in very poor nations, and there is
a smaller but positive association within nations between income
and happiness (see Diener & Biswas-Diener, 2002, for a review).
There is much conjecture but little scientific understanding of why
income and well-being are associated.
In this study, we provide initial answers to several fundamental
questions about the relations of money and happiness. First, we
examined whether income is equally related to different types of
well-being. Although there are clear demarcations between life
evaluations, positive emotions, and negative emotions, past work on
income and well-being has focused mostly on global evaluations of
life, whereas positive and negative feelings have been largely ignored.
Ed Diener, Department of Psychology, University of Illinois at Urbana–
Champaign, and The Gallup Organization, Omaha, Nebraska; Weiting Ng,
Psychology Programmes, Singapore Institute of Management, Singapore;
James Harter and Raksha Arora, The Gallup Organization.
James Harter is an employee and stockholder of The Gallup Organiza-
tion, whose data are used in this article.
Correspondence concerning this article should be addressed to Ed Die-
ner, Department of Psychology, University of Illinois, 603 East Daniel
Street, Champaign, IL 61820. E-mail: ediener@uiuc.edu
Journal of Personality and Social Psychology, 2010, Vol. 99, No. 1, 52– 61
© 2010 American Psychological Association 0022-3514/10/$12.00
Second, we analyzed possible mediators of the relation between
income and well-being, which also have been largely ignored in
large-scale surveys. For example, we examined whether the associa-
tion of income with well-being is mediated by the fact that wealthy
people are more likely to have their physical and/or psychological
needs met or whether wealthy people are more satisfied because they
have more of their material desires fulfilled.
Another purpose of the present study is to analyze how universal
the relation between money and well-being is by determining the
consistency of the relation across various groups. If the relations
vary, it would suggest that context and culture can influence the
relation between income and well-being. For example, is income
more important to well-being in poor nations where basic need
fulfillment is a larger issue?
Fourth, we explored at the societal level whether citizens of
wealthy nations are happier than those of poor countries because
more people have high incomes or because of public resources
such as better health facilities and education. The Gallup World
Poll (GWP) is the first representative survey of the globe. Thus, we
explored the four questions about income and well-being in a
diverse and representative sample of the world.
Types of Well-Being
In describing types of subjective well-being, Kahneman (1999)
drew a distinction between the online experience of utility and
DOI: 10.1037/a0018066

52
 HAPPINESS AND MONEY
global evaluations of life. Whereas broad evaluations such as life
satisfaction judgments involve a reflective analysis of a person’s
life, online reports of experience comprise a person’s positive and
negative momentary feelings over time. Supporting this distinc-
tion, Lucas, Diener, and Suh (1996) found that life satisfaction,
pleasant emotions, and unpleasant emotions formed distinct fac-
tors. Therefore, in the present study, we analyze how various
potential mediators of the income–well-being relation are associ-
ated with each form of subjective well-being. In this study, we
examined three types of well-being that were included in the
GWP: Cantril’s (1965) Self-Anchoring Striving Scale, which as-
sesses global life evaluations; positive feelings (enjoyment and
smiling or laughing); and negative feelings (anger, sadness, worry,
and depression). On the basis of Scitovsky (1978), we predicted
that life satisfaction might be more strongly related to income,
positive feelings to social psychological rewards, and negative
emotions inversely to the fulfillment of basic needs.
Mediators of the Income and Well-Being Relation
Why is income associated with subjective well-being? Although
many studies have been conducted that establish the correlation
between income and subjective well-being, little research has
analyzed the mediating mechanisms for this relation. In the current
article, we examine the fulfillment of physical and psychological
needs as possible mediators, as well as the fulfillment of material
aspirations. Is it the fulfillment of basic needs such as the need for
food that underlies the income and well-being association, or is it
learned desires for material goods that cause the relationship?
Many people who write about money and well-being implicitly
assume that the relation exists because income allows people to
fulfill their physical needs for food and shelter and, perhaps to
some degree, their psychological needs as well, for example, for
respect. Previous researchers suggested that the stronger associa-
tion between money and well-being at lower income levels indi-
cates the importance of basic needs to the income and well-being
relation (Howell & Howell, 2008). This interpretation is consistent
with Maslow’s (1943) theory that basic biological needs are pre-
potent in that they emerge first in people’s motivations. Thus, we
expected that the fulfillment of basic needs might increase well-
being, although lack of fulfillment might be most associated with
negative emotions. The GWP included measures of not having
enough money for food and for shelter and therefore allowed us to
examine the path from income through basic needs to the various
forms of subjective well-being.
In recent decades, a number of theories have fleshed out psy-
chological needs beyond basic physical ones. For example, Ryan
and Deci (2000) argued that there are three universal psychological
needs—for autonomy, competence, and social relationships—and
Ryff (1989) suggested six psychological needs—for self-
acceptance, positive relations with others, autonomy, environmen-
tal mastery, purpose in life, and personal growth. The GWP
included a number of measures reflecting psychological needs. In
the social realm, the GWP asked respondents whether they feel
respected and whether they have family and friends they can count
on in the case of an emergency. In the psychological domain, the
GWP included questions about feeling free to choose one’s daily
activities, learning new things, and having opportunities to do
“what one does best.” We combined these three questions with the
53
two social questions to assess the fulfillment of social psycholog-
ical needs. Thus, we were able to assess the mediation of both
physical and psychological needs on the income and well-being
relation.
Another reason that income might be related to well-being is
that people learn to desire material goods, and the fulfillment of
desires leads to feelings of well-being, regardless of whether needs
are met. Emmons (1986) found that achieving goals leads to higher
subjective well-being. If people want money because of social
learning, the influence of advertising, and so forth, then obtaining
it could be linked to feelings of well-being.
The possibility of goal fulfillment as a mediator is suggested by
the findings of Biswas-Diener, Vittersø, and Diener (2005), who
found that some groups without much income have reasonably
high subjective well-being. The researchers found that the Maasai
report high life satisfaction even though they live in dung houses
without running water, electricity, or modern conveniences. Sim-
ilarly, the Amish are reasonably happy even though they live
without electricity and automobiles. In addition, Biswas-Diener
and Diener (2001) showed that even some extremely poor people
in the slums of Calcutta report subjective well-being above neutral.
These findings might be explained by the fact that the Amish and
traditional Maasai have relatively low material aspirations, al-
though their basic material needs are largely met. Conversely,
Graham and Pettinato (2002) found that people with rapidly rising
incomes sometimes feel frustrated because their aspirations have
outpaced their incomes. These findings suggest that it may not be
only the fulfillment of actual needs that connects income to well-
being, but it could be that people’s material desires are important
as well. Thus, we analyze whether fulfilling goals versus needs is
most associated with well-being, with which forms of well-being,
and whether they play a mediational role between income and
well-being.
In this study, we had several proxy measures to assess fulfill-
ment of material desires. There was a measure of whether people
felt satisfied with their standard of living. If respondents’ satisfac-
tion with their standard of living predicts well-being beyond the
effects of need fulfillment, it will suggest that desires are an
important mediator. Second, we had measures of whether people
own luxury conveniences that are widely desired but not necessary
for survival, such as televisions, computers, and the Internet. If we
find that possessing these items predicts well-being, it will suggest
that desires can influence well-being beyond the fulfillment of
basic needs.
In addition to the effects of household income, we examined the
impact of people’s relative income positions within their societies.
A potential reason for the income and happiness association is that
money might bestow status and respect on people. Earning more
money than others could have value above and beyond the goods
that money can buy because a person’s relative position regarding
money might be a cause of status. Furthermore, people might
evaluate their material lives largely by judging how they compare
with those around them. In the GWP, we analyzed whether relative
income—income compared with the income of others in one’s
own society—predicts well-being beyond the effects of absolute
income.
Another purpose of the present study was to analyze how
consistent the relation between money and well-being is by deter-
mining the consistency of the association across various groups.
54 DIENER, NG, HARTER, AND ARORA
For example, we examined the relation between income and well-
being in poor versus wealthy nations, predicting that the associa-
tion would be stronger in poor nations.
Why Is Mean Well-Being Higher in Wealthy Nations?
Whereas correlations of mean income and mean well-being are
often .70 –.80 across countries, associations within nations are
often closer to .20 (Diener & Biswas-Diener, 2002). One plausible
explanation is that there is much more error and random variation
in the individual figures, whereas extraneous factors are averaged
out when societal figures are analyzed. For example, individual
differences might largely be averaged out with societal correla-
tions, making the error term smaller. An alternate explanation is
that some nations have societal or public goods that, when it comes
to well-being, are more important than individual income. We are
able to determine the effects of national wealth after first control-
ling for the fulfillment of needs, the ownership of modern conve-
niences, and individual income. Thus, we can determine whether
there is a societal effect after controlling for individual consump-
tion and income.
Method
Sample
The Gallup Organization conducted representative surveys of
132 countries from 2005 to 2006. These nations include about 96%
of the world’s population and varied greatly in terms of economic
development, political structure and stability, and culture. A few
nations such as North Korea were not accessible to the poll, and
most small nations such as Andorra and Vanuatu were not in-
cluded. There was representative sampling of the entire adult
population within each nation. In wealthy nations, this was
achieved through telephone surveys based on random-digit dial-
ing; in poorer nations, where telephones are less ubiquitous, this
was accomplished by door-to-door interviews, with residences
selected randomly within randomly selected geographical sam-
pling units.
The final sample size for the present study consisted of a maximum
of 136,839 respondents (aged 15 years or older); the mean sample size
of each nation was 1,061 respondents. The numbers of individuals and
nations for specific analyses were often lower because of missing
data, although the numbers were always very large. Respondents were
weighted within nations to further refine the samples to represent the
correct demographic proportions within societies.
Subjective Well-Being
Three types of well-being were selected from the survey to
represent it. The global life evaluation measure, Cantril’s (1965)
Self-Anchoring Striving Scale, asked respondents to evaluate their
current life on a ladder scale, which ranged from 0 (worst possible
life) to 10 (best possible life). Questions about positive feelings and
negative feelings assessed the experience of positive and negative
emotions “yesterday.” Respondents were asked whether they ex-
perienced certain feelings (positive ⫽ enjoyment and smiling or
laughing; negative ⫽ worry, sadness, depression, and anger) a lot
in the previous day (1 ⫽ yes, 0 ⫽ no). “Yes” responses to the two
positive feelings were averaged into a positive feelings score, and
“yes” responses to the four negative feelings were averaged into a
negative feelings score. Thus, scores could vary from 0 to 1.0 for
feelings. One advantage of using measures of emotions during the
previous day is that it reduces the extent of memory bias. Thus, the
reports are more likely to be anchored in actual experience than are
reports of long-time periods or reports of emotions in general
(Robinson & Clore, 2002). However, anchoring the reports on a
single day will introduce random error into the measures in that
short-term fluctuating influences are likely to be related to a single
day’s feelings. Thus, we expected that the correlations for the
affect measures would be lower than those for the evaluation
measure.
Economic Predictors
Various measures were selected to assess economic conditions.
For individuals, annual household income was used as an indicator
of personal income. Because of the diminishing marginal utility of
money, the logarithm (log 10) of reported household income was
used. We created another income measure by adjusting household
income for household size. However, household size data were not
available for many nations and individuals, thus the adjusted
measure was dropped so as not to lose a large portion of the
sample. Preliminary analyses with both measures showed similar
patterns of findings. Because the effects of income may not only be
due to the absolute amount of money that one has but might also
be due to one’s relative standing within the nation, we also
calculated each respondent’s relative income by standardizing
incomes within nation.
A measure assessing the modern luxury household conve-
niences that respondents possessed was derived to reflect the
fulfillment of material desires that go beyond basic needs. Respon-
dents were asked whether their home has a television, a computer,
and access to the Internet (1 ⫽ yes, 0 ⫽ no). The number of “yes”
responses to the three items was averaged to give an overall score.
Respondents were also asked whether there were times in the past
year when they did not have enough money for food or for shelter.
The “yes” responses to the two items were averaged, yielding an
indicator for unfulfilled basic needs. Satisfaction with standard of
living also was reported on a yes–no scale.
National wealth was assessed using the 2005–2007 per capita
gross domestic product (GDP; in purchasing power parity) data of
each nation in constant 2,000 international dollars, computed by
the World Bank (2010). The logarithm of the GDP data was then
taken in log 10 units, using a common method of economists,
which has the effect of compressing scores as one moves progres-
sively up the income ladder.
Social Psychological Predictors
Five additional items from the survey were included as social
psychological measures. As an indicator of respect, respondents
reported whether they were treated with respect on the previous
day (1 ⫽ yes, 0 ⫽ no). Another question inquired as to whether the
respondent had family or friends he or she could count on in an
emergency. The questions used to assess the fulfillment of psy-
chological needs were all asked in reference to “yesterday”:
learned something new, did what you do best, and chose how your
 HAPPINESS AND MONEY
time was spent. We also examined the number of hours that
participants worked the previous day, on a scale that could range
from zero to 24, to analyze whether long work hours reduces the
positive association between income and well-being.
The questionnaire was translated into the major languages of
each country. The translation process started with an English, a
French, or a Spanish version, depending on the region. A translator
who was proficient in the original and target languages translated
the survey into the target language. A second translator reviewed
the language version against the original version and recom-
mended refinements.
Although many of the measures from the survey are dichoto-
mous items, this type of measure is quicker for use in very large
surveys. Furthermore, as the survey is administered in many na-
tions across the world, dichotomous items are less likely to be
influenced by number-use response sets across nations. The large
number of respondents ensures that individual reliability is not a
major concern and that the results can be interpreted, although the
dichotomous items are likely to reduce the size of correlations
compared with correlations of data gathered using scales with
more response options. Unreliability and translation discrepancies
would serve to reduce the effects we found rather than enhance
them, especially in light of the very large number of nations and
respondents included. Therefore, our estimates might be conser-
vative.
Results
The basic descriptive statistics for the well-being variables and
predictors are shown in Table 1. As can be seen, the average
person in the world was a relatively happy respondent who judged
his or her current life to be slightly above neutral in the balance
between positive and negative feelings, experienced frequent pos-
itive feelings, and infrequently felt negative feelings. Although the
majority of people had their psychological needs met to some
extent, basic needs were unmet in about one quarter of respon-
dents. Women made up 51.1% of the sample and men 48.9%, with
a small number not identified. The marital status of the samples
was 49.3% married, 30.6% single, 4.9% divorced or separated,
6.1% widowed, and 2.1% in domestic partnership; the remaining
Table 1
Descriptive Statistics for the Sample
Statistic M SD
Log household income 3.82
Cantril’s Self-Anchoring Striving Scale 5.36
Positive feelings 71
Negative feelings 23
Basic needs unmet 26
Psychological needs met 69
Luxury conveniences 40
Satisfaction with standard of living 59
Free time 4.2
Note. Log income is household income from the Gallup World Poll,
whereas national per capita income comes from the World Bank (2010).
Free time is in hours, and Cantril’s Self-Anchoring Striving Scale is a 0 –10
scale. The other figures represent first averaging the items composing the
scale, then averaging all respondents, and finally multiplying by 100.
55
respondents did not answer this question. In terms of residential
location, 25.5% were in a rural area, 30.4% lived in a village,
31.4% lived in a city, and 8.6% lived in a suburb; data were
missing regarding the residential locations of the rest of the re-
spondents. Thus, 58.3% of the sample for whom we had residential
data lived rurally or in a village. In terms of education, data were
available for only 19.6% of the respondents; of these, 25.9% had
a primary education or less, 50.0% had some secondary and/or
some tertiary education, and 19.6% had four or more years of
tertiary education.
The correlations between well-being and its predictors are
shown in Table 2. Because of the very large sample sizes, all of the
correlations are significantly different from zero, and most differ
significantly from the other correlations in the table. We found that
income was positively associated with Self-Anchoring Striving
Scale scores and positive feelings and negatively associated with
negative emotions. Ownership of luxury conveniences and satis-
faction with standard of living showed patterns of relations with
well-being similar to that of income with well-being, whereas
unmet basic needs showed an inverse relation with well-being. As
shown, income was much more strongly associated with the Self-
Anchoring Striving Scale score than with the other measures ( p ⬍
.001). For the Self-Anchoring Striving Scale, the strongest corre-
lates were income and national income, satisfaction with standard
of living, and ownership of luxury conveniences, whereas for
feelings, the fulfillment of psychological needs was the strongest
predictor. Satisfaction with standard of living was a significantly
stronger predictor of life evaluation than of feelings, whereas
psychological need fulfillment significantly more strongly pre-
dicted positive feelings than evaluations of life. It is interesting that
societal income predicted life evaluations as well as individual
income. The number of hours worked did not have a sizeable
association with individual well-being and is not included in the
remaining analyses.
Another approach to understanding the relation between income
and well-being is to examine the regression slopes between the
two, and we present these in Figure 1. Each of the three well-being
measures is plotted against log income categories, and for both
feelings and life evaluations, the range shown is approximately
half of the full range of the respective scales. As can be seen, the
slope is much steeper for income and the Self-Anchoring Striving
Scale than for income and the two measures of feelings. Thus, not
only is there a higher correlation for the Self-Anchoring Striving
Scale and income, but variations in income are associated with
change across a wider portion of the possible range of scores.
0.61 Strength of the Predictors
2.24
39 To explore the associations more systematically, we conducted
30 hierarchical linear regression analyses separately on each of the
37 three types of well-being, and the results are shown in Tables 3
28 and 4. In Table 3, we present the amount of variance that was
34
49 accounted for by each predictor as it was added to the linear model,
4.5 as well as the total amount of variance predicted at each step. The
variance figures give a clear picture of the power of the predictors.
The amount of variance predicted is especially helpful in this case,
where virtually all statistics are significantly different because of
the large sample sizes. Luxury conveniences and satisfaction with
standard of living were added at the same step because they were
56 DIENER, NG, HARTER, AND ARORA

Table 2 all of the well-being variables but was never the most powerful
Zero-Order Correlations Between Types of Well-Being explanatory variable.
and Predictors Table 4 presents the outcome of regression analyses when all
predictors were entered together. For the Self-Anchoring Striving
Predictor Striving Scale PF NF Scale score, national income and satisfaction with standard of
Log household income .44 .17 ⫺.09 living were the strongest predictors when all were entered together,
Relative income (within-nations) .21 .11 ⫺.11 whereas for feelings, the psychological needs fulfillment index
National log GDP per capita .44 .10 ⫺.03 was by far the strongest predictor. Relative income was significant
Basic needs unmet ⫺.32 ⫺.16 .19 for the Self-Anchoring Striving Scale and negative feelings but not
Psychological needs met .25 .45 ⫺.28
for positive feelings. Although life evaluation was significantly
Luxury conveniences .39 .11 ⫺.05
Satisfaction with standard of living .40 .24 ⫺.20 associated with relative income within one’s society, other predic-
Work hours .03 ⫺.02 .03 tors were stronger. The strong association of national income with
Positive feelings .23 — well-being even with many individual resources entered suggests
Negative feelings ⫺.19 ⫺.38 — the importance of societal circumstances for people’s well-being.
Note. PF ⫽ positive feelings; NF ⫽ negative feelings; GDP ⫽ gross The substantial correlation between the log income of individ-
domestic product. For all correlations, p ⬍ .001. uals and their Self-Anchoring Striving Scale scores across a large
number of nations implies that a cross-societal standard is being
used by respondents, especially in light of the fact that the effect
persisted for two of the three well-being variables even when
used together to reflect the fulfillment of material desires. Overall, within-nation income was controlled. It is striking how large the
negative feelings can be predicted least and life evaluation best, effect of societal income was, even when individual income was in
with positive feelings intermediate between them. For the Self- the equation. This suggests that people’s satisfaction with life is
Anchoring Striving Scale, the step where luxury conveniences and strongly influenced by the general circumstances of their societies.
satisfaction with standard of living were added accounted for by
far the most variance, although both physical and psychological
Consistency of Associations
needs, as well as income and national income, added additional
variance. In contrast, for both positive and negative feelings, the In Table 5, we present the correlations of income and well-being
step with psychological needs added most to the prediction. In- for various categories of respondents. The correlations for income
deed, for positive feelings, the social psychological needs ac- and life evaluations were relatively consistent across groups, vary-
counted for the vast majority of the explainable variance. Basic ing from a low of .31 to a high of .47. Similarly, there was a
needs fulfillment accounted for a nontrivial amount of variance for moderately strong association in virtually all groups between pos-

8.0 95

7.5 90

85
7.0
PF and No-NF (% of people)

80
6.5

75
Current Ladder

6.0
70
5.5
65
5.0
Ladder
60
PF
4.5
No NF 55

4.0 PF
50
No NF
3.5 Ladder 45

3.0 40
2.60 2.76 2.92 3.08 3.24 3.40 3.56 3.72 3.88 4.04 4.20 4.36 4.52 4.68 4.84 5.00

Log Income

Figure 1. Plots of log income versus mean subjective well-being, as well as the best-fit linear trend lines, for
each type of well-being. PF ⫽ positive feelings; NF ⫽ negative feelings.
 HAPPINESS AND MONEY 57

Table 3
Variance Predicted in Hierarchical Regression Analyses by Potential Mediators

Cantril’s Self-Anchoring
Striving Scale Positive feelings Negative feelings

Predictors R 2
⌬R 2
p R 2
⌬R2
p R 2
⌬R2 p

Step 1: Basic needs unmet .088 .088 .001 .029 .029 .001 .035 .035 .001
Step 2: Add psychological needs .125 .037 .001 .213 .184 .001 .091 .056 .001
Step 3: Add conveniences and satisfaction
with standard of living .265 .140 .001 .223 .010 .001 .099 .007 .001
Step 4: Add log income .288 .023 .001 .224 .001 .001 .099 .000 .031
Step 5: Add comparative income .288 .000 .001 .224 .000 .002 .102 .003 .001
Step 6: Add national income (log GDP/capita) .305 .016 .001 .225 .001 .001 .104 .002 .001

Note. R2 values are adjusted for the number of predictors. GDP ⫽ gross domestic product.

itive feelings and social psychological resources, varying from .39
to .47. Income accounted for 19% of the variance in life evaluation
but only 3% of the variance in positive feelings, for which social
psychological wealth accounted for 20% of the variance.
The enormous sample sizes result in many small differences
between correlations being significant. However, several of the
differences are systematic and account for nontrivial amounts of
variance. For instance, correlations for the predictors of well-being
were higher for less educated and older respondents compared
with the corresponding correlations for more educated and
younger respondents. In contrast, there were virtually no differ-
ences in the correlations for men and women. But the pattern of
correlations is consistent, with income substantially predicting life
evaluations in every group and social psychological needs sub-
stantially predicting positive feelings in every group.
Prosperity of Nations
Does the prosperity of nations differ depending on what type of
quality-of-life outcome is considered? In Table 6, we present the
rank order for selected nations of the world. Whereas Northern
European and Anglo societies are currently most successful in the
economic area, Latin societies appear to be relatively high in social
psychological well-being. Although certain nations such as Den-
mark achieve consistently high scores across measures, no nation
is happiest in the sense of being first on all measures of prosperity
and well-being. Sierra Leone’s scores were consistently low, but
other nations showed divergent rankings across the measures. For
instance, Russia and South Korea are substantially lower in social
psychological prosperity and in positive feelings than they are in
income. In contrast, Costa Rica ranks much higher in social

Table 4
Hierarchical Regression Analyses at Step 6 With All Predictors

Predictors B SE ␤ t p

Cantril’s Self-Anchoring Striving Scale

Basic needs unmet ⫺0.47 0.024 ⫺.077 ⫺19.53 .001
Psychological needs met 0.99 0.034 .110 29.53 .001
Luxury conveniences 0.63 0.034 .095 18.50 .001
Satisfaction with standard of living 0.94 0.018 .208 52.42 .001
Log income 0.25 0.025 .066 9.76 .001
Relative income 0.20 0.012 .089 17.01 .001
National income (log GDP/capita) 1.01 0.028 .221 35.73 .001
Positive feelings
Basic needs unmet ⫺0.059 0.004 ⫺.057 ⫺13.77 .001
Psychological needs met 0.625 0.006 .409 104.18 .001
Luxury conveniences 0.018 0.006 .016 3.04 .002
Satisfaction with standard of living 0.074 0.003 .096 23.02 .001
Log income 0.047 0.005 .074 10.40 .001
Relative income ⫺0.004 0.002 ⫺.010 ⫺1.74 ns
National income (log GDP/capita) ⫺0.039 0.005 ⫺.051 ⫺7.79 .001
Negative feelings
Basic needs unmet 0.109 0.004 .138 30.75 .001
Psychological needs met ⫺0.266 0.005 ⫺.225 ⫺53.27 .001
Luxury conveniences ⫺0.002 0.005 ⫺.002 ⫺0.34 ns
Satisfaction with standard of living ⫺0.056 0.003 ⫺.094 ⫺20.87 .001
Log income 0.001 0.004 .003 0.38 ns
Relative income ⫺0.009 0.002 ⫺.030 ⫺5.14 .001
National income (log GDP/capita) 0.047 0.004 .079 11.27 .001

Note. GDP ⫽ gross domestic product.

58 DIENER, NG, HARTER, AND ARORA

Table 5
Potential Moderators in Predicting Well-Being (Unstandardized Regression Slopes)

Log income predicting

Positive feelings and social Negative feelings and unmet
Moderator Striving Scale PF NF psychological needs basic needs

Entire sample .44 .17 ⫺.09 .45 .19

Residential area
Urban .39 .16 ⫺.12 .44 .21
Rural .44 .17 ⫺.08 .46 .18
Sex
Male .44 .16 ⫺.09 .45 .19
Female .44 .18 ⫺.08 .45 .19
Education
Primary only .45 .26 ⫺.20 .48 .26
4⫹ years tertiary .40 .12 ⫺.11 .39 .20
Age (in years)
Early adult (15–25) .39 .14 ⫺.04 .42 .16
Young (26–40) .45 .18 ⫺.07 .45 .18
Middle age (41–55) .47 .17 ⫺.11 .46 .19
Young-old (56–70) .45 .22 ⫺.17 .47 .23
Old-old (71–99⫹) .45 .28 ⫺.22 .46 .27
National income (GDP/capita)
⬍10,000 .31 .17 ⫺.08 .45 .18
ⱖ10,000 .40 .16 ⫺.13 .39 .18

Note. Income is reported in international dollars. PF ⫽ positive feelings; NF ⫽ negative feelings; GDP ⫽ gross domestic product.

psychological prosperity and positive feelings than it does on
income. In general, the former Soviet Union bloc nations have
greater economic than social psychological prosperity, whereas
Central American and certain African nations have greater social
psychological than economic prosperity. Discrepancies between
positive and negative feelings can also be seen in the table, for
example, in the rankings of Japan, Nepal, and India. The United
States was at the top in income but ranked below the middle of the
list in negative feelings. Thus, it is oversimplified to capture the
well-being of a nation with a single score, whether it be income or
life satisfaction.
Discussion
We addressed several questions in this study and uncovered
relatively unequivocal answers. First, we found that income does
indeed have a stronger association with some types of well-being
than with others. Namely, the relation of income and well-being

Table 6
Rankings of Selected Nations on Types of Prosperity (Out of 89)

Social psychological Self-Anchoring Low negative

Nation GDP/capita prosperity Striving Scale Positive feelings feelings

United States 1 19 16 26 49
Denmark 5 13 1 7 1
The Netherlands 7 36 4 3 26
Japan 14 50 24 44 6
Italy 18 33 20 67 63
Israel 20 56 11 61 64
New Zealand 22 12 9 1 21
South Korea 24 83 38 58 77
South Africa 35 42 53 29 44
Russia 36 72 56 79 42
Mexico 39 22 23 17 28
Costa Rica 41 6 18 4 38
Indonesia 59 63 57 24 43
India 61 85 46 63 22
Ghana 68 51 70 68 20
Nepal 76 88 69 50 10
Sierra Leone 87 80 87 87 86
Tanzania 89 58 86 52 32

Note. GDP ⫽ gross domestic product.

 HAPPINESS AND MONEY
was strong for life evaluations but weaker for feelings. These
findings are consistent with the conclusions of Diener, Kahneman,
Tov, and Arora (2009), who found that long-term income changes
are more related to changes in life evaluations than to changes in
positive and negative feelings. We found the strongest mediation
between income and life evaluations by satisfaction with standard
of living. There were weaker income pathways for all of our
mediating variables for feelings, in part because feelings were not
highly associated with income in the first place. Instead, feelings
were most strongly associated with social psychological prosper-
ity. In addition, we found that national income predicted life
evaluations even after controlling for individual income, indicating
that there are societal circumstances related to income that might
substantially influence people’s life satisfaction.
We started with the question of why money is associated with
well-being. In retrospect, it is difficult to imagine it otherwise for
life evaluations in the modern world of market economies. Money
is an object that many or most people desire and pursue during the
majority of their waking hours. Thus, it would be surprising if
success at this pursuit had no influence whatsoever when people
are asked to evaluate their lives. Thus, the more interesting ques-
tions might be why income is so weakly associated with feelings
of well-being and ill-being and why social psychological success is
so predictive of positive feelings.
One possibility is that there is reverse causality in terms of
positive feelings leading to social psychological success. People
and societies with frequent positive feelings might develop stron-
ger social relationships, in the manner described by Fredrickson
and Joiner (2002). Similarly, people who are enjoying themselves
might feel that they are autonomous, learning new things, and
doing their best. Indeed, Fredrickson’s broaden-and-build model
suggests that it is when people feel positive that they are most
likely to develop resources for the future, for example, by learning
new things and by strengthening social ties. We cannot, however,
discount the other causal direction going from social psychological
resources to feelings. For example, Pavot, Diener, and Fujita
(1990) found that people experience more positive feelings on
average when they are with others than when alone. Csikszentmi-
halyi’s (1990) theory of flow suggests that people are pleasantly
engaged when they are using their skills in activities that are
optimally challenging and require their skills. Ryff (1989) and
Ryan and Deci (2000) suggested that there are basic and universal
human psychological needs, and these needs are highly associated
with the social psychological variables that predicted well-being in
our study. Learning new things, using one’s skills, feeling re-
spected, and having others to count on are reflective of meeting
needs, and thus it is not unexpected that fulfilling these needs
would enhance positive feelings.
Tibor Scitovsky (1978) in The Joyless Economy offered several
ideas that help to explain our pattern of findings. He differentiated
comfort and pleasure. It may be that comforts increase life eval-
uations whereas pleasures increase reports of positive feelings.
Comfort comes from having one’s needs and desires continuously
fulfilled, whereas pleasures come from fulfilling unmet needs and
from stimulating and challenging activities. One source of pleasure
according to Scitovsky is social stimulation, which he suggested
lies largely outside the realm of economics. Novelty and learning
can be sources of pleasure too. Thus, Scitovsky’s reasoning is in
accord with our findings that wealth predicts life satisfaction, and
59
social relationships and learning new things predict positive feel-
ings. In Berridge’s (2004) terminology, money might have more to
do with people having what they want than liking what they have,
and social psychological prosperity produces the feelings they like.
The correlations we found for income and life evaluations were
larger than in many past studies. One reason for these larger
associations is that our sample encompassed a much wider range
of income and included more poor and rural nations than did
previous studies. The correlations we found are likely to be larger
than those found within populations with a more restricted range of
income. A second possible reason is that we examined the income
associations separately for the various types of well-being. Our
results showed that the Self-Anchoring Striving Scale evaluation
of life yielded the strongest associations with income, whereas
affective assessments of well-being yielded smaller correlations. In
past studies, measures might have been used that were about
feelings or that mixed feelings and judgments (e.g., reports of
happiness).
Might the Self-Anchoring Striving Scale score produce spuri-
ously high correlations with income compared with other mea-
sures? There are several arguments against this possibility. First,
Helliwell, Barrington-Leigh, Harris, and Huang (2009) found that
income correlated similarly with a traditional life satisfaction
measure. Second, even if the Self-Anchoring Striving Scale’s
correlation with income were higher than its correlation with other
measures of life evaluation, it would not mean that this is spurious.
Instead, it would point to the fact that different measures might
prime somewhat different information to be evaluated, just as
recent events might prime certain information.
It is possible that response biases influenced our results, for
example, respondents might differ by culture or region in their
tendency to avoid scale extremes. Although this might have had
some influence, we have reason to believe that the majority of the
patterns we observed are substantive. For example, the patterns
repeat across cultures and groups, as well as across many lan-
guages. Several of our predictors are subjective, such as having
friends to count on and learning new things, meaning that several
interpretations are possible for associations using these variables.
Nevertheless, Diener et al. (2009) found similar patterns at the
national level using objective measures taken from sources such as
the World Bank. Just as we found that the GDP per capita predictor
we used here, which comes from outside our data set and is not
subject to response artifacts from the GWP, predicts our well-
being variables quite well, other studies we have conducted on
societal well-being show that variables such as the rated corruption
of nations also predicts well-being. A concern is that because
questions about feelings were focused on “yesterday,” believing
that one learned things, is respected, and so forth could come from
short-term moods rather than long-term circumstances. Further-
more, the psychosocial variables and positive and negative feelings
were both measured on yes–no formats, and this might have
increased the association between them. Thus, it will be important
for future researchers to assess feelings over longer periods of time
so that short-term and long-term processes can be distinguished; it
will also be helpful for researchers to use varying response for-
mats.
Can other research artifacts explain our findings? It is unlikely
that item-order priming could have led to our results because the
income questions came at the end of the survey, far removed from
60 DIENER, NG, HARTER, AND ARORA
the subjective well-being questions, which were toward the begin-
ning of the interview. It is also unlikely that the results are due to
language translation differences, because the survey was adminis-
tered in a very large number of diverse languages. In addition, the
wording used in the survey was carefully developed in each
location in close collaboration with local researchers.
In recent years, a number of authors have questioned whether
the effects of income or subjective well-being are absolute or
relative (e.g., Clark, Frijters, & Shields, 2008; Frank, 2005; Hsee,
Yang, Li, & Shen, 2009; Johnson & Krueger, 2006). We found
effects of income relative to others in people’s own nations, but
they were smaller than the associations for absolute income. This
might be due to the fact that people compare themselves with local
others or those in their profession but do not often make nation-
wide comparisons. More probably, it may be that the standard for
material well-being is now largely worldwide rather than defined
within nations. This is consistent with the conclusions of Diener et
al. (2009), who found steeper income and well-being slopes be-
tween nations than within them. Thus, our findings do not dem-
onstrate that people do not use social comparison to evaluate their
incomes; they only suggest that the social comparisons are not
heavily focused at the nation level. Our findings clearly indicate
that people evaluate their incomes relative to a global standard. If
there was no global standard, we could not have found that
people’s life evaluations correlated so highly with their incomes
across nations with very diverse incomes. Thus, social comparison
might be a significant force, but the within-nation level of analysis
might not strongly capture these effects.
The analyses revealed that the income and well-being correla-
tions were fairly consistent across groups. An unexpected finding
was that the relation between income and life evaluations was
slightly stronger in wealthy nations than in poor ones, because this
is counter to the idea that money is most important when it helps
meet basic needs. However, this finding is in accord with the idea
that the income and life evaluation association rests on the fulfill-
ment of material aspirations. This was supported by the finding
that satisfaction with standard of living and the overall evaluation
of life were more highly correlated in wealthy than in economi-
cally underdeveloped nations.
Because of the declining marginal utility of money, differences
in raw absolute income produce larger effects for the poor than for
the rich. Many past studies in psychology used raw income as a
measure. However, an increase from $10,000 to $20,000 will have
more of an impact than an increase from $80,000 to $90,000. In
using log income as our measure, we examined increases in terms
of the percentage of increase at different levels of income. Thus, an
increase from $10,000 to $20,000 dollars is equivalent in log terms
to an increase from $80,000 to $160,000 —in both cases a dou-
bling of one’s income. Viewed from this perspective, income has
an effect on life evaluations even in wealthy nations, although the
raw amount of income required to produce noticeable change is
much larger.
Which is the correct way to analyze income data, given that
economists prefer log income and psychologists prefer raw in-
come? In our data, the log income correlations were higher than
those measuring raw income. Thus, one can argue for smaller or
larger income effects depending on which scaling of income is
used. Some argue that log income effects are misleading because
people do not think of their income in these terms. The important
point is to interpret the log findings correctly.
Log income is preferred by economists because it is truer to the
statistical requirements of linear statistics. An additional reason
that log data are desirable is that they reflect the psychological fact
that the perception of stimuli is most frequently based on the log
of the intensity of those stimuli. Increases in light, sound, weight,
and so forth produce increases in perception that are generally in
line with log increases in the stimuli, and this is likely to be true
of income as well. After all, a few dollars will not mean the same
thing to a millionaire and to a pauper, any more than a candle will
produce the same perception of light on a bright beach and in a
dark cave.
The difficulty for psychologists in using log income comes at
the level of communication with the public, because people often
do not consider their income changes in these terms. For example,
people would not think of an increase from $10,000 to $100,000
when they think of their incomes rising, although this is only one
unit increase on the log 10 scale. The important points to keep in
mind are to correctly interpret the conclusions and to use the right
form of description when scales of either type are used.
The present findings on the relation of income to different types
of well-being might partly resolve the Easterlin paradox debate, in
which some maintain that there has been no change in well-being
as nations have become wealthier over time (e.g., Easterlin, 1995,
2005; Myers, 2000), whereas others have claimed the opposite
(e.g., Stevenson & Wolfers, 2008; Veenhoven & Hagerty, 2006).
If income has a low association with certain forms of well-being,
such as feelings, it would be likely that increases in income would
have little effect on them. However, improvements in other types
of factors such as the social environment might influence these
forms of well-being. This is an important avenue for future re-
search.
Take-Home Message
Contrary to both those who say money is not associated with
happiness and those who say that it is extremely important, we
found that money is much more related to some forms of well-
being than it is to others. Income is most strongly associated with
the life evaluation form of well-being, which is a reflective judg-
ment on people’s lives compared with what they want them to be.
Although statistically significant, the association of income with
positive and negative feelings was modest. Furthermore, we found
that societal income has a substantial influence on life evaluations
beyond the effects of personal income, indicating that it is very
desirable for life satisfaction to live in an economically developed
nation. However, we also discovered that social psychological
prosperity is very important to positive feelings. Some nations that
do well in economic terms do only modestly well in social psy-
chological prosperity, and some nations that rank in the middle in
economic development are stars when it comes to social psycho-
logical prosperity.
If replicated, our findings have profound implications for both
psychological theories of well-being and for societal policies. At
the theoretical level, our results indicate that different types of
well-being can be influenced by very different predictors. It is
important to note that social psychological well-being is shown to
be an important correlate of feelings across the globe. At the policy
 HAPPINESS AND MONEY
level, our findings indicate that more than money is needed for
quality of life, and the social psychological forms of prosperity
correlate only moderately with economic development. This
means that societies must pay careful attention to social and
psychological variables, not simply to enlarging their economies.
Our findings indicate that it is important for societies not only to
measure economic variables but to measure social psychological
well-being variables as well.
References
Berridge, K. C. (2004). Motivation concepts in behavioral neuroscience.
Physiology & Behavior, 81, 179 –209.
Biswas-Diener, R., & Diener, E. (2001). Making the best of a bad situation:
Satisfaction in the slums of Calcutta. Social Indicators Research, 55,
329 –352.
Biswas-Diener, R., Vittersø, J., & Diener, E. (2005). Most people are pretty
happy, but there is cultural variation: The Inughuit, the Amish, and the
Maasai. Journal of Happiness Studies: An Interdisciplinary Periodical
on Subjective Well-Being, 6, 205–226.
Cantril, H. (1965). The patterns of human concern. New Brunswick, NJ:
Rutgers University Press.
Clark, A. E., Frijters, P., & Shields, M. (2008). Relative income, happiness
and utility: An explanation for the Easterlin paradox and other puzzles.
Journal of Economic Literature, 46, 95–144.
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience.
New York, NY: Harper & Row.
Diener, E., & Biswas-Diener, R. (2002). Will money increase subjective
well-being? A literature review and guide to needed research. Social
Indicators Research, 57, 119 –169.
Diener, E., Kahneman, D., Tov, W., & Arora, R. (2009). Income’s asso-
ciation with judgments of life versus feelings. In E. Diener, D. Kahne-
man, & J. F. Helliwell (Eds.), International differences in well-being
(pp. 3–15). Oxford, United Kingdom: Oxford University Press.
Easterlin, R. A. (1995). Will raising the incomes of all increase the
happiness of all? Journal of Economic Behavior and Organization, 27,
35– 48.
Easterlin, R. A. (2005). Feeding the illusion of growth and happiness: A
reply to Hagerty and Veenhoven. Social Indicators Research, 74, 429 –
443.
Emmons, R. A. (1986). Personal strivings: An approach to personality and
subjective well-being. Journal of Personality and Social Psychology, 51,
1058 –1068.
Frank, R. (2005). Does absolute income matter? Economics and Happi-
ness, 27, 65–91.
Fredrickson, B. L., & Joiner, T. (2002). Positive emotions trigger upward
spirals toward emotional well-being. Psychological Science, 13, 172–
175.
Graham, C., & Pettinato, S. (2002). Frustrated achievers: Winners, losers
61
and subjective well-being in new market economies. Journal of Devel-
opment Studies, 38, 100 –140.
Helliwell, J. F., Barrington-Leigh, C., Harris, A., & Huang, H. (2009).
International evidence on the social context of well-being. In E. Diener,
D. Kahneman, & J. F. Helliwell (Eds.), International differences in
well-being. Oxford, UK: Oxford University Press.
Howell, H., & Howell, C. (2008). The relation of economic status to
subjective well-being in developing countries: A meta-analysis. Psycho-
logical Bulletin, 134, 536 –560.
Hsee, C. K., Yang, Y., Li, N., & Shen, L. (2009). Wealth, warmth and
well-being: Whether happiness is relative or absolute depends on
whether it is about money, acquisition or consumption. Journal of
Marketing Research, 46, 396 – 409.
Johnson, W., & Krueger, R. F. (2006). How money buys happiness:
Genetic and environmental processes linking finances and life satisfac-
tion. Journal of Personality and Social Psychology, 90, 680 – 691.
Kahneman, D. (1999). Objective happiness. In D. Kahneman, E. Diener, &
N. Schwartz (Eds.), Well-being: The foundations of hedonic psychology
(pp. 3–25). New York, NY: Russell Sage Foundation.
Lucas, R., Diener, E., & Suh, E. (1996). Discriminant validity of well-
being measures. Journal of Personality and Social Psychology, 71,
616 – 628.
Maslow, A. H. (1943). A dynamic theory of human motivation. Psycho-
logical Review, 50, 370 –396.
Myers, D. (2000). American paradox: Spiritual hunger in an age of plenty.
New Haven, CT: Yale University Press.
Pavot, W., Diener, E., & Fujita, F. (1990). Extraversion and happiness.
Personality and Individual Differences, 11, 1299 –1306.
Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evidence for
an accessibility model of emotional self-report. Psychological Bulletin,
128, 934 –960.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the
facilitation of intrinsic motivation, social development, and well-being.
American Psychologist, 55, 68 –78.
Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the
meaning of psychological well-being. Journal of Personality and Social
Psychology, 57, 1069 –1081.
Scitovsky, T. (1978). The joyless economy. Oxford, United Kingdom:
Oxford University Press.
Stevenson, B., & Wolfers, J. (2008). Economic growth and subjective
well-being: Reassessing the Easterlin paradox. Third Annual Conference
on Empirical Legal Studies Papers. Abstract retrieved from the Social
Science Research Network website: http://ssrn.com/abstract⫽1121237
Veenhoven, R., & Hagerty, M. (2006). Rising happiness in nations 1946 –
2004. Social Indicators Research, 79, 421– 436.
World Bank. (2010). World development indicators online database.
Available from the World Bank website: http://www.worldbank.org/
Received September 10, 2008
Revision received July 9, 2009
Accepted August 13, 2009 䡲