PSYCHOLOGICAL SCIENCE

Research Article
ZEROING IN ON THE DARK SIDE OF THE AMERICAN DREAM:
A Closer Look at the Negative Consequences of the Goal for
Financial Success
Carol Nickerson,1 Norbert Schwarz,2 Ed Diener,1,3 and Daniel Kahneman4
1
University of Illinois at Urbana-Champaign, 2University of Michigan, 3The Gallup Organization, and 4Princeton University

Abstract—Recent research has demonstrated that aspiring to the search. First, cross-sectional studies have nearly always found a posi-
American Dream of financial success has negative consequences for tive, albeit small, relation between an individual’s income and his or
various aspects of psychological well-being. The present longitudinal her subjective well-being or happiness (see Diener, 1984, and Diener,
study examining the relation between the goal for financial success, Suh, Lucas, & Smith, 1999, for reviews). These studies, however, have
attainment of that goal, and satisfaction with various life domains not considered the goal for financial success.
found that the negative impact of the goal for financial success on Second, the consumer research literature includes a number of
overall life satisfaction diminished as household income increased. studies that have examined the relation between materialism and sub-
The negative consequences of the goal for financial success seemed to jective well-being or perceived quality of life. Materialism is defined
be limited to those specific life domains that either concerned relation- as an individual differences variable reflecting a high relative impor-
ships with other people or involved income-producing activities, such tance placed on the acquisition and possession of income, wealth, and
as one’s job; satisfactions with two of those life domains, however, material goods (Sirgy, 1998); it is usually considered a personality
were among the strongest predictors of overall life satisfaction in this trait or a value, rather than a goal or an aspiration. For example, Belk
sample of well-educated respondents in their late 30s. The negative (1984, 1985), who considered materialism to be a trait, developed
consequences were particularly severe for the domain of family life; measures tapping three aspects of that trait—possessiveness, nongen-
the stronger the goal for financial success, the lower the satisfaction erosity, and envy—and found that all three had negative correlations
with family life, regardless of household income. with happiness and life satisfaction in a sample of respondents that in-
cluded both college students and working people. Using Belk’s mea-
sures, Wachtel and Blatt (1990) subsequently found a positive correlation
They’ve bought the BMW, and they have the $3 million Mill Valley house. And between the three aspects of materialism and self-criticism in college
they still wake up in the morning and say, “I don’t feel good about myself.” students. Richins and Dawson (1992), who conceptualized material-
—San Francisco Bay Area psychologist Stephen Goldbart, co-founder of the ism as a value and developed an alternative measure, found that mate-
Money, Meaning, and Choices Institute (quoted in Perspectives 2000, 2000/ rialism was negatively correlated with satisfaction with five life domains—
2001, p. 96) life as a whole, fun, family life, income or standard of living, and rela-
tionships with friends—in a sample of consumers. Materialism was
Anecdotes like this one suggest that aspiring to and achieving the negatively correlated with self-esteem in a second sample of consumers.
American Dream of financial success may have a dark side. Instead of Perhaps because they considered materialism to be a personality
leading to the expected increase in subjective well-being, aspiring to trait or a value, rather than a goal, neither Belk (1984, 1985) nor Rich-
and achieving financial success may have negative psychological con- ins and Dawson (1992) considered the possible effect of goal attain-
sequences, such as depression, anxiety, lessened self-esteem, and dis- ment on subjective well-being. On the basis of aspiration theory,
satisfaction with life. These consequences would come as no surprise which asserts that subjective well-being is a reflection of the discrep-
to humanistic psychologists (e.g., Fromm, 1976; Maslow, 1956, 1970; ancy between aspirations and attainments (Andrews & Withey, 1976;
Rogers, 1961, 1963), whose theories suggest that pursuing goals based Campbell, Converse, & Rodgers, 1976; McGill, 1967; Michalos, 1980,
on extrinsic rewards, the contingent approval of other people, and 1985; Wilson, 1960/1961), La Barbera and Gürhan (1997) hypothe-
“having” instead of “being” distract the individual from the meaningful sized that highly materialistic people with lower incomes have lower
aspects of life, hinder the individual from achieving his or her inherent levels of subjective well-being than highly materialistic people with
potential as a human being, and lead to psychological distress. These higher incomes. Materialism was assessed using Belk’s (1984, 1985)
consequences would not be predicted, however, by cognitive-behav- nongenerosity and envy measures and Ward and Wackman’s (1971)
ioral psychologists (e.g., Bandura, 1977, 1989; Carver, Lawrence, & measure of orientation toward money and possessions; subjective
Scheier, 1996; Locke & Latham, 1990; Seligman, 1991), whose theo- well-being was assessed using an index of “general affect.” All three
ries emphasize that the setting and pursuit of goals, regardless of the measures of materialism had negative correlations with general affect;
nature of those goals, lead to satisfaction, happiness, and well-being. income had a positive correlation with general affect. Income moder-
To our knowledge, no empirical study has examined the relation ated the effect of nongenerosity and the effect of orientation toward
between the goal for financial success, attainment of that goal, and money and possessions on general affect in the manner hypothesized.
subjective well-being. There are, however, three related lines of re- Third, research on motivation has recently begun to consider the
effects on subjective well-being of having and pursuing various kinds
of goals, including the goal for money or, more generally, financial
Address correspondence to Norbert Schwarz, Institute for Social Research, success. For example, Kasser and Ryan (1993) found that the relative
University of Michigan, 426 Thompson St., Ann Arbor, MI 48106-1248; importance of the goal for financial success was negatively correlated
e-mail: [email protected]. with self-esteem, vitality, and self-actualization and positively corre-

VOL. 14, NO. 6, NOVEMBER 2003 Copyright © 2003 American Psychological Society 531
 PSYCHOLOGICAL SCIENCE

Dark Side

lated with depression and anxiety in college students. In a more het- ing the respondent’s educational history, employment history, retro-
erogeneous group of adolescents drawn from the community, the spective views of college, civic activities, satisfaction with various
relative importance of the goal for financial success was negatively aspects of life, and sociodemographic characteristics. The American
correlated with global functioning and social productivity and posi- Freshman database (Astin, King, & Richardson, 1976) was developed
tively correlated with behavior disorders. Similarly, Chan and Jo- by the Higher Education Research Institute at the University of Cali-
seph’s (2000) study of college students in England found that the fornia at Los Angeles, which administers the Cooperative Institutional
importance of the goal for financial success was negatively correlated Research Program. This program surveys entering freshmen at several
with happiness, self-actualization, and self-esteem. Diener and Oishi’s hundred colleges and universities across the United States each year,
(2000) study of college students from 41 nations found that the impor- collecting information regarding respondents’ educational and career
tance of money was negatively correlated with life satisfaction. plans, life goals and values, self-ratings of abilities and personality
None of these researchers, however, considered whether achieving traits, opinions on various social and educational issues, and sociode-
financial success might moderate the generally negative consequences mographic characteristics.
of the goal for financial success, perhaps because of the difficulty of The College and Beyond survey sample consisted of all members
obtaining the data to do so. Achieving financial success is a long-term of the 1976 entering freshman classes of the 17 private colleges and
process for most individuals, and so an examination of the relation be- universities and a sample of about 2,000 members of the 1976 enter-
tween the goal for financial success, attainment of that goal, and ing freshman classes of the 4 public universities; each public-univer-
subjective well-being requires a longitudinal study of a duration suffi- sity sample consisted of all known minority students, all college
cient to allow individuals to achieve or to make substantial progress athletic-letter winners, all students with a composite Scholastic Apti-
toward that goal. All three studies were concurrent. The study by La tude Test score of 1350 or greater, and a random sample of the other
Barbera and Gürhan (1997), in which income was found to moderate members of the class. The total number of individuals in the survey
the effect of materialism on subjective well-being, was also concur- sample equaled 23,597. Of these, 16,632 (70%) completed the Col-
rent, and so its results, although suggestive, are not as convincing as lege and Beyond survey. Through the linking mechanism, it was possi-
those of a longitudinal study would be. ble to identify 17,734 (75%) of the individuals in the survey sample as
We are fortunate to have access to a longitudinal data set that con- having completed The American Freshman survey at college entry. In-
tains information about the importance of financial success for indi- formation from both surveys was available for 12,894 (55%) of the in-
viduals at the time of college entry, coupled with information about dividuals in the survey sample. These 12,894 individuals constituted
their goal attainment and subjective well-being about 19 years later. the sample used in this study, although any particular analysis may
The data set is large, containing information for more than 12,000 in- have been based on fewer individuals because of missing values for
dividuals, a considerable advantage given the difficulty of detecting one or more of the variables in the analysis.
moderator effects in field studies (McClelland & Judd, 1993). The in-
formation about subjective well-being consists of a set of measures as-
sessing overall life satisfaction and satisfaction with six specific life Variables
domains. Most studies examining the effects of materialism or the Predictor variables
goal for financial success have focused on global aspects of subjective
well-being, such as happiness, depression, and self-esteem. Accord- One of the predictor variables, the goal for financial success
ingly, in order to zero in on the negative consequences of the goal for (financial goal), was drawn from The American Freshman survey. At
financial success found by previous researchers, we first determined college entry in 1976, respondents rated “the importance to you per-
whether goal attainment—household income—moderated those conse- sonally of being very well off financially” on a 4-point scale where
quences for overall life satisfaction, as suggested by La Barbera and 1  not important, 2  somewhat important, 3  very important, and
Gürhan’s (1997) study of materialism. We then investigated whether 4  essential.1 The other predictor variable, household income, was
the goal for financial success had negative consequences for more nar- drawn from the College and Beyond survey. Respondents indicated
rowly defined life domains, as suggested by Richins and Dawson’s
(1992) study of materialism, and if so, whether goal attainment mod-
erated those consequences.
1. Materialism and the goal for financial success are usually measured with
multi-item scales (e.g., Belk, 1984, 1985; Kasser & Ryan, 1993; Richins &
METHOD Dawson, 1992; Ward & Wackman, 1971). In this study, financial goal was as-
sessed with only a single survey item, raising concerns about the psychometric
properties of this variable. The American Freshman database provides no di-
Study Sample
rect evidence for the reliability or validity of financial goal; however, there is
The individuals in this study were drawn from the 1976 entering some indirect evidence. For example, respondents were asked to indicate the
freshman classes of 21 academically selective institutions, including 4 importance of each of 11 reasons for their decisions to attend college. The cor-
relation between financial goal and the reason “to be able to make more
large public universities and 17 private colleges and universities. In-
money” equaled .51, indicating reasonable convergent validity given that one
formation about these individuals is available in two linked databases.
might attain this goal by inheriting money, marrying money, investing wisely,
The College and Beyond database (Bowen & Bok, 1998) was devel- and so on, as well as by making more money as a result of attending college.
oped by the Andrew W. Mellon Foundation to facilitate study of the Moreover, the correlations between financial goal and the other 10 reasons
long-term consequences of attending academically selective institu- were much lower (e.g., “to make me a more cultured person,” .04; “to learn
tions in the United States. This database contains the responses to a more about things that interest me,” .09; “to meet new and interesting peo-
survey conducted in 1995 through 1997 to collect information regard- ple,” .01), suggesting reasonable divergent validity as well.

532 VOL. 14, NO. 6, NOVEMBER 2003

 PSYCHOLOGICAL SCIENCE

C. Nickerson et al.

their household incomes for 1995 using the following categories: 1  of the six domain-specific life satisfactions. The analysis for job satis-
less than $1,000; 2  $1,000 to $9,999; 3  $10,000 to $19,999; 4  faction used as a predictor only the portion of household income at-
$20,000 to $29,999; 5  $30,000 to $49,999; 6  $50,000 to $74,999; tributable to personal earned (job) income and included only respondents
7  $75,000 to $99,999; 8  $100,000 to $149,999; 9  $150,000 to working full time for all of 1995, so that personal earned income repre-
$199,999; and 10  $200,000 or more. The midpoint of each category sented remuneration for a comparable amount of work across respon-
was used as the dollar value of household income. The midpoint of the dents. For personal earned income, the midpoint of the open-ended
open-ended 10th category was assigned the value $290,000, estimated 10th category was assigned the value $265,000, estimated from 1990
from 1990 U.S. Census data (Bowen & Bok, 1998, pp. 357–358). U.S. Census data (Bowen & Bok, 1998, pp. 357–358). Respondents
Household income was defined as the total 1995 household income be- with a personal earned income in the 1st category (less than $1,000;
fore taxes, including personal earned income from a job; net business, n  27) were excluded from the analysis for job satisfaction because it
farm, or rent income; pensions or Social Security payments; dividends was not clear how a respondent working full time for all of 1995 could
or interest; and income earned by the respondent’s spouse or partner. have had such a low income.
Ordinary least squares regression was used in all analyses. Previ-
ous research has suggested that the relation between income and sub-
Criterion variables
jective well-being may be curvilinear, with increases in income being
The seven criterion variables—overall life satisfaction and six do- associated with increases in subjective well-being at lower but not at
main-specific life satisfactions—were drawn from the College and Be- higher incomes (see Diener, 1984, and Diener et al., 1999, for re-
yond survey. Respondents indicated, “in general,” how satisfied they views). Logarithmic transformations of household income, personal
were with “life right now” on a 5-point scale where 1  very satisfied, earned income, and parental income were used to linearize this rela-
2  somewhat satisfied, 3  neither satisfied nor dissatisfied, 4  tion.
somewhat dissatisfied, and 5  very dissatisfied. Respondents also All predictor and control variables were centered about their means
used this scale to rate their satisfaction with the current (1995) job. prior to analysis. The reported effect of each predictor variable is a unique
Additionally, respondents indicated how much satisfaction they ob- effect, with all other effects controlled. All predictor variables, control
tained from the five other specific life domains—(a) the city or place variables, and interactions were retained in the analyses, regardless of
of residence, (b) nonworking activities such as hobbies, (c) family life, their statistical significance; that is, step-down procedures were not
(d) friendships, and (e) health and physical condition—on a 7-point employed (Aiken & West, 1991; Cohen & Cohen, 1983, pp. 301–350;
scale where 1  a very great deal, 2  a great deal, 3  quite a bit, Jaccard, Turrisi, & Wan, 1990; Judd & McClelland, 1989, pp. 239–280).
4  a fair amount, 5  some, 6  a little, and 7  none. The scales
for all criterion variables were reversed for use in the analyses. Overall Life Satisfaction
The regression (n  10,659) showed that, in general, the higher the
Control variables household income, the higher the overall life satisfaction (b 
0.30456, semipartial correlation, sr  .26, t  27.73, p  .0001). At
Because there may be a temperamental predisposition for happi- the mean of financial goal, overall life satisfaction at the lowest and at
ness (see Diener, 1984, and Diener et al., 1999, for reviews), and be-
cause the goal for financial success may be influenced by the
socioeconomic status of the family in which an individual was raised
(Kasser, Ryan, Zax, & Sameroff, 1995), two control variables—cheer-
Table 1. Overall life satisfaction by household income and
fulness and parental income—were selected from The American financial goal
Freshman survey. At college entry in 1976, respondents rated their
cheerfulness relative to that of the average student of their own age on Financial goal
a 5-point scale where 1  lowest 10%, 2  below average, 3  aver- Household-
income Not Somewhat Very
age, 4  above average, and 5  highest 10%. Respondents also esti-
midpoint important important important Essential Range
mated their parents’ total 1975 income before taxes using the following
categories: 1  less than $3,000; 2  $3,000 to $3,999; 3  $4,000 to $500 3.05 2.77 2.49 2.21 –0.84
$5,999; 4  $6,000 to $7,999; 5  $8,000 to $9,999; 6  $10,000 to $5,500 3.62 3.45 3.28 3.11 –0.51
$12,499; 7  $12,500 to $14,999; 8  $15,000 to $19,999; 9  $15,000 3.86 3.73 3.61 3.48 –0.37
$20,000 to $24,999; 10  $25,000 to $29,999; 11  $30,000 to $25,000 3.98 3.88 3.78 3.67 –0.30
$40,000 4.09 4.01 3.93 3.85 –0.24
$34,999; 12  $35,000 to $39,999; 13  $40,000 to $49,999; and
$62,500 4.20 4.14 4.08 4.02 –0.18
14  $50,000 or more. The lower limit of each category was used as the $87,500 4.28 4.23 4.19 4.14 –0.13
dollar value of parental income because the data necessary to estimate $125,000 4.36 4.33 4.31 4.28 –0.08
the midpoint of the open-ended 14th category could not be obtained. $175,000 4.44 4.43 4.42 4.40 –0.04
$290,000 4.56 4.57 4.59 4.59 0.03
RESULTS Note. Overall life satisfaction ranged from 1, very dissatisfied, to 5,
The results of the analyses are presented as a two-part sequence. very satisfied. The values of overall life satisfaction shown were
generated from simple-slope regression equations (Aiken & West,
The primary analysis predicted overall life satisfaction from financial 1991, pp. 12–22). Although the table shows the household-income
goal, household income, and their interaction. This analysis was then midpoint in dollars, the natural logarithm of the household-income
repeated controlling for cheerfulness and for parental income. The midpoint was used in the analysis. n  10,659.
secondary analysis repeated the three-step primary analysis for each

VOL. 14, NO. 6, NOVEMBER 2003 533

 PSYCHOLOGICAL SCIENCE

Dark Side

the highest household income equaled 2.64 and 4.58, respectively, a hold income increased. This pattern was not altered substantially when
difference of 1.94 on a 5-point scale. either cheerfulness or parental income was controlled.
The regression also showed that, in general, the stronger the finan-
cial goal, the lower the overall life satisfaction (b  0.04346, sr 
.04, t  4.39, p  .0001). At the mean of household income, the
Domain-Specific Life Satisfactions
overall life satisfaction at the weakest and at the strongest financial The regression results for the six domain-specific life satisfactions
goal equaled 4.28 and 4.15, respectively, a difference of 0.13. are presented in Table 2. For none of these satisfactions was the pat-
However, the generally negative effect of financial goal on overall tern of findings substantially altered when either cheerfulness or pa-
life satisfaction was moderated by household income (b  0.04562, rental income was controlled.
sr  .03, t  3.56, p  .0004). As Table 1 shows, at lower household Table 2 shows that the higher the household income, the higher the
incomes, respondents with a stronger financial goal had lower overall satisfaction with all of the specific life domains except nonworking ac-
life satisfaction than respondents with a weaker financial goal, tivities; similarly, the higher the personal earned income, the higher
whereas at higher household incomes, there was little difference in the job satisfaction.
overall life satisfaction between respondents, regardless of financial Financial goal had either no significant effect or a significant posi-
goal. Overall life satisfaction equaled 3.05 for respondents with the tive effect on satisfaction with three of the six specific life domains—
lowest household income and the weakest financial goal and 2.21 for satisfaction with city or place of residence, satisfaction with nonwork-
respondents with the lowest household income and the strongest finan- ing activities, and satisfaction with health and physical condition—
cial goal, a difference of 0.84. Overall life satisfaction equaled 4.56 and a significant negative effect on the other three—satisfaction with
for respondents with the highest household income and the weakest fi- family life, satisfaction with friendships, and job satisfaction.
nancial goal and 4.59 for respondents with the highest household in- As shown in Table 3, personal earned income moderated the effect
come and the strongest financial goal, a difference of 0.03. This of financial goal on job satisfaction in a manner similar to that
pattern of results indicated that the negative consequences of the goal whereby household income moderated the effect of financial goal on
for financial success for overall life satisfaction diminished as house- overall life satisfaction. That is, at lower incomes, respondents with a

Table 2. Regression results for the six domain-specific life satisfactions

Financial Household
Domain-specific life satisfaction Statistica goal income b Interaction
City or place of residence (n  10,666) b 0.03288 0.20825 —
sr .02 .12 —
t 2.12 12.10 —
p .0341 .0001 —
Nonworking activities (n  10,662) b — — —
sr — — —
t — — —
p — — —
Family life (n  10,651) b 0.07509 0.48595 —
sr .05 .26 —
t 4.85 28.24 —
p .0001 .0001 —
Friendships (n  10,669) b 0.04342 0.06881 0.06556
sr .03 .04 .03
t 2.81 4.01 3.28
p .0050 .0001 .0011
Health and physical condition (n  10,665) b 0.03527 0.16658 0.06281
sr .02 .09 .03
t 2.13 9.08 2.94
p .0331 .0001 .0033
Jobc (n  7,782 ) b 0.03179 0.25081 0.04173
sr .03 .18 .03
t 2.49 16.24 2.33
p .0129 .0001 .0200
Note. Dashes indicate effects that were not statistically significant.
a
b is the unstandardized regression coefficient, sr is the semipartial correlation coefficient (magnitude of
effect), t is the t-test statistic, and p is the significance level.
b
The natural logarithm of the midpoint for household or personal earned (job) income was used in the
analysis.
c
Only the portion of household income attributable to personal earned (job) income was used as the
predictor variable.

534 VOL. 14, NO. 6, NOVEMBER 2003

 PSYCHOLOGICAL SCIENCE

C. Nickerson et al.

Table 3. Job satisfaction by personal earned (job) income and Table 4. Satisfaction with friendships by household income
financial goal and financial goal

Personal- Financial goal Financial goal

earned- Household-
income Not Somewhat Very income Not Somewhat Very
midpoint important important important Essential Range midpoint important important important Essential Range

$500 — — — — — $500 5.63 5.25 4.87 4.48 1.15

$5,500 3.87 3.73 3.60 3.46 0.40 $5,500 5.57 5.34 5.12 4.89 0.68
$15,000 4.06 3.96 3.87 3.78 0.28 $15,000 5.54 5.38 5.22 5.06 0.48
$25,000 4.15 4.08 4.01 3.94 0.21 $25,000 5.53 5.40 5.28 5.15 0.38
$40,000 4. 24 4.19 4.14 4.09 0.16 $40,000 5.52 5.42 5.32 5.23 0.29
$62,500 4.33 4.29 4.26 4.23 0.10 $62,500 5.50 5.44 5.37 5.30 0.20
$87,500 4.39 4.37 4.35 4.33 0.06 $87,500 5.50 5.45 5.41 5.36 0.13
$125,000 4.46 4.45 4.45 4.44 0.01 $125,000 5.49 5.47 5.44 5.42 0.06
$175,000 4.52 4.53 4.54 4.55 0.03 $175,000 5.48 5.48 5.48 5.48 0.00
$265,000 4.62 4.65 4.68 4.71 0.09 $290,000 5.46 5.50 5.53 5.57 0.10
Note. Satisfaction with friendships ranged from 1, none, to 7, a very
Note. Job satisfaction ranged from 1, very dissatisfied, to 5, very
great deal. The values of satisfaction with friendships shown were
satisfied. Respondents with a personal earned income of less than
generated from simple-slope regression equations (Aiken & West,
$1,000 were excluded from the analysis because it was not clear how a
1991, pp. 12–22). Although the table shows the household-income
respondent working full time all year for 1995 could have had such a
midpoint in dollars, the natural logarithm of the household-income
low income. The values of job satisfaction shown were generated from
midpoint was used in the analysis. n  10,669.
simple-slope regression equations (Aiken & West, 1991, pp. 12–22).
Although the table shows the personal-earned-income midpoint in
dollars, the natural logarithm of the personal-earned-income midpoint
was used in the analysis. n  7,782.

faction with nonworking activities, .08; and satisfaction with friendships,

.07—indicating that satisfaction with family life and job satisfaction
stronger financial goal had lower job satisfaction than respondents were by far the two strongest predictors of overall life satisfaction.
with a weaker financial goal, whereas at higher incomes, financial
goal had little effect on job satisfaction, indicating that the negative
consequences of the goal for financial success for job satisfaction di- DISCUSSION
minished as income increased. The results of the present longitudinal study suggest that although
The overall pattern of results obtained therefore suggests that the there are indeed negative consequences of the goal for financial suc-
negative consequences of financial goal may be limited to those life cess, these consequences are neither as deleterious nor as pervasive as
domains that either concern relationships with other people or involve previous studies have suggested. For example, although this goal had
incoming-producing activities, such as one’s job. In regards to the negative consequences for overall life satisfaction, these consequences
former, the negative consequences seemed particularly severe for the diminished as household income increased, a finding in accord with
domain of family life. As shown in Table 4, household income moder- that of La Barbera and Gürhan’s (1997) concurrent study of material-
ated the effect of financial goal on satisfaction with friendships, such ism. Moreover, overall life satisfaction was on the positive side of the
that at lower household incomes, respondents with a stronger financial scale for nearly all respondents, regardless of the importance of the
goal had lower satisfaction with friendships than respondents with a goal for financial success; that is, although respondents with a stronger
weaker financial goal, whereas at higher household incomes, financial financial goal tended to have lower overall life satisfaction than those
goal had little effect on satisfaction with friendships. There was no with a weaker financial goal, especially at lower household incomes,
such moderating effect of household income on satisfaction with fam- nearly all respondents were satisfied. Finally, the positive effect of
ily life; the stronger the financial goal, the lower the satisfaction with household income on overall life satisfaction was stronger than the
family life, regardless of household income. At the mean of household negative effect of the goal for financial success.
income, the difference in satisfaction with family life between respon- Examination of the effects of the goal for financial success on sat-
dents with the weakest financial goal and respondents with the stron- isfaction with six specific life domains suggests that these negative
gest equaled about 0.23 on a 7-point scale. consequences may be limited to those life domains that either concern
Although the negative consequences of financial goal may be lim- relationships with other people (friends, family) or involve income-
ited to certain life domains, a dominance analysis (Budescu, 1993) producing activities, such as one’s job. However, satisfactions with two
showed that two of those three domains had the most influence on of those life domains—family life and job—were the strongest predic-
overall life satisfaction in this sample of well-educated respondents in tors of overall life satisfaction for this sample of respondents. More-
their late 30s. Specifically, the dominance analysis partitioned the over, although personal earned income moderated the effect of the
variance in overall life satisfaction accounted for by the six domain- goal for financial success on job satisfaction, there was no such mod-
specific life satisfactions (40%) as follows: satisfaction with family erating effect of household income on satisfaction with family life; the
life, .37; job satisfaction, .25; satisfaction with city or place of resi- stronger the goal for financial success, the lower the satisfaction with
dence, .12; satisfaction with health and physical condition, .11; satis- family life, regardless of household income. These findings differ

VOL. 14, NO. 6, NOVEMBER 2003 535

 PSYCHOLOGICAL SCIENCE

Dark Side

somewhat from those of Richins and Dawson (1992), who obtained norms for fall 1976. Los Angeles: Cooperative Institutional Research Program of
the American Council on Education and the University of California at Los Angeles,
negative correlations between materialism and satisfaction with all four Graduate School of Education, Laboratory for Research in Higher Education.
specific life domains that they examined: fun, family life, income or Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psycho-
standard of living, and relationships with friends; the negative relation logical Review, 84, 191–215.
Bandura, A. (1989). Self-regulation of motivation and action through internal standards
between materialism and satisfaction with family life was the weakest. and goal systems. In L.A. Pervin (Ed.), Goal concepts in personality and social psy-
Three additional comments are in order. First, the respondents in chology (pp. 19–85). Hillsdale, NJ: Erlbaum.
this study were students who matriculated at selective colleges and Belk, R.W. (1984). Three scales to measure constructs related to materialism: Reliability,
validity, and relationships to measures of happiness. In T.C. Kinnear (Ed.), Ad-
universities; most of them were from affluent families with well-edu- vances in consumer research (Vol. 11, pp. 291–297). Provo, UT: Association for
cated parents, and 22% had household incomes of $150,000 or more. Consumer Research.
Belk, R.W. (1985). Materialism: Trait aspects of living in the material world. Journal of
Thus, the results obtained may not be generalizable to individuals who Consumer Research, 12, 265–280.
matriculated at less selective institutions, or who did not attend college Bowen, W.G., & Bok, D. (1998). The shape of the river: Long-term consequences of con-
at all. In particular, the intellectual abilities and superb educations of sidering race in college and university admissions. Princeton, NJ: Princeton Univer-
sity Press.
the respondents in this study may have afforded them more opportuni- Budescu, D.V. (1993). Dominance analysis: A new approach to the problem of relative im-
ties to achieve the goal to be very well off financially and hence miti- portance of predictors in multiple regression. Psychological Bulletin, 114, 542–551.
gated the negative consequences of that goal. Campbell, A., Converse, P.E., & Rodgers, W.L. (1976). The quality of American life: Per-
ceptions, evaluations, and satisfactions. New York: Russell Sage.
Second, the results of the present study do not provide unequivocal Carver, C.S., Lawrence, J.W., & Scheier, M.F. (1996). A control-process perspective on
support for either the humanistic theories or the cognitive-behavioral the origins of affect. In L.L. Martin & A. Tesser (Eds.), Striving and feeling: Inter-
actions among goals, affect, and self-regulation (pp. 11–52). Mahwah, NJ: Erl-
theories. For example, the humanistic theories would not have pre- baum.
dicted that overall life satisfaction increased as household income in- Chan, R., & Joseph, S. (2000). Dimensions of personality, domains of aspiration, and sub-
creased. The fact that overall life satisfaction increased as household jective well-being. Personality and Individual Differences, 28, 347–354.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the be-
income increased for respondents for whom the goal to be very well havioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
off financially was very important or essential accords with the cogni- Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542–575.
tive-behavioral theories, but the fact that these respondents were no Diener, E., & Oishi, S. (2000). Money and happiness: Income and subjective well-being
across nations. In E. Diener & E.M. Suh (Eds.), Culture and subjective well-being
more satisfied at high household incomes than were respondents for (pp. 185–218). Cambridge, MA: MIT Press.
whom this goal was not important or somewhat important does not; Diener, E., Suh, E.M., Lucas, R.E., & Smith, H.L. (1999). Subjective well-being: Three
decades of progress. Psychological Bulletin, 125, 276–302.
respondents having a stronger financial goal who have made substan- Fromm, E. (1976). To have or to be? New York: Harper & Row.
tial progress toward that goal would have been expected to be more Jaccard, J., Turrisi, R., & Wan, C.K. (1990). Interaction effects in multiple regression
satisfied than respondents with an equivalent income but a weaker fi- (Sage University Paper series on Quantitative Applications in the Social Sciences
No. 07–072). Newbury Park, CA: Sage.
nancial goal. Both approaches need to be elaborated to account for the Judd, C.M., & McClelland, G.H. (1989). Data analysis: A model-comparison approach.
pattern of results obtained in the current study. San Diego, CA: Harcourt Brace Jovanovich.
Third, several researchers have noted that because their studies Kasser, T., & Ryan, R.M. (1993). A dark side of the American Dream: Correlates of finan-
cial success as a central life aspiration. Journal of Personality and Social Psychol-
were cross-sectional, concurrent, and correlational, no inferences about ogy, 65, 410–422.
the causal relation between materialism or the goal for financial suc- Kasser, T., Ryan, R.M., Zax, M., & Sameroff, A.J. (1995). The relations of maternal and
social environments to late adolescents’ materialistic and prosocial values. Develop-
cess and subjective well-being could be drawn (e.g., Belk, 1984, 1985; mental Psychology, 31, 907–914.
Kasser & Ryan, 1993; La Barbera & Gürhan, 1997). The present study La Barbera, P.A., & Gürhan, Z. (1997). The role of materialism, religiosity, and demo-
was longitudinal, but we echo a similar caveat, despite the fact that graphics in subjective well-being. Psychology & Marketing, 14, 71–97.
Locke, E.A., & Latham, G.P. (1990). A theory of goal setting and task performance. En-
we, like others, have written about the “consequences” of the goal for glewood Cliffs, NJ: Prentice Hall.
financial success. Inferring cause from correlational studies, even lon- Maslow, A.H. (1956). Defense and growth. Merrill-Palmer Quarterly, 3, 36–47.
gitudinal ones such as this, is risky. That one variable—goal for finan- Maslow, A.H. (1970). Motivation and personality (2nd ed.). New York: Harper & Row.
McClelland, G.H., & Judd, C.M. (1993). Statistical difficulties of detecting interactions
cial success—precedes another variable—subjective well-being—is a and moderator effects. Psychological Bulletin, 114, 376–390.
necessary but not sufficient condition for the inference of cause. There McGill, V.J. (1967). The idea of happiness. New York: Praeger.
Michalos, A.C. (1980). Satisfaction and happiness. Social Indicators Research, 8, 385–
is always the possibility of another variable being the causal mecha- 422.
nism. For example, the true cause of the psychological distress result- Michalos, A.C. (1985). Multiple discrepancies theory (MDT). Social Indicators Research,
ing from the goal for financial success perhaps could be found in the 16, 347–413.
Perspectives 2000. (2000, December 25/2001, January 1). Newsweek, 136/137, 79–107.
circumstances that led to the development of that goal. Richins, M.L., & Dawson, S. (1992). A consumer values orientation for materialism and
its measurement: Scale development and validation. Journal of Consumer Research,
19, 303–316.
Acknowledgments—We gratefully acknowledge the Andrew W. Mellon Rogers, C.R. (1961). On becoming a person: A therapist’s view of psychotherapy. Boston:
Foundation for providing the data and the funding for this study, Alexander Houghton Mifflin.
Astin for permission to use The American Freshman survey database, Rogers, C.R. (1963). The actualizing tendency in relation to “motives” and to conscious-
Stephanie Olson and Abigail Tolhurst for research assistance, and Gary ness. In M.R. Jones (Ed.), Nebraska Symposium on Motivation (Vol. 11, pp. 1–24).
McClelland for statistical advice. Lincoln: University of Nebraska Press.
Seligman, M.E.P. (1991). Learned optimism. New York: Knopf.
Sirgy, M.J. (1998). Materialism and quality of life. Social Indicators Research, 43, 227–
260.
Wachtel, P.L., & Blatt, S.J. (1990). Perceptions of economic needs and of anticipated fu-
ture income. Journal of Economic Psychology, 11, 403–415.
REFERENCES Ward, S., & Wackman, D. (1971). Family and media influences on adolescent consumer
learning. American Behavioral Scientist, 14, 415–427.
Aiken, L.S., & West, S.G. (1991). Multiple regression: Testing and interpreting interac- Wilson, W.R. (1961). An attempt to determine some correlates and dimensions of hedonic
tions. Newbury Park, CA: Sage. tone (Doctoral dissertation, Northwestern University, 1960). Dissertation Abstracts,
Andrews, F.M., & Withey, S.B. (1976). Social indicators of well-being: Americans’ per- 21(09), 2814.
ceptions of life quality. New York: Plenum Press.
Astin, A.W., King, M.R., & Richardson, G.T. (1976). The American freshman: National (RECEIVED 4/15/01; REVISION ACCEPTED 11/25/02)

536 VOL. 14, NO. 6, NOVEMBER 2003

This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy.
Users should refer to the original published version of the material.