Social Psychology Quarterly
2004,Vol. 67, No. 1, 103–114
Status As a Valued Resource*
BERNARDO A. HUBERMAN
Hewlett Packard Laboratories
CHRISTOPH H. LOCH
INSEAD and Hewlett Packard Laboratories
AYSE ÖNÇÜLER
INSEAD
Delivered by Ingenta to
The striving for status has long been recognized in sociology and economics. Extensive
INSEAD LIBRARY; Mrs LE CAMUS (cid 35002204)
theoretical arguments and empirical evidence propose that people view status as a sign
IP: 194.57.233.136
of competence and pursue it as a means to achieve power and resources.A small literaDate:from
2004..08..17..05..26..
ture, however, based on arguments
biology and evolutionary psychology, proposes that people pursue status as an (emotional) goal in itself, independent of competence
and expressed by culturally flexible symbols. We present results of an experiment with
human subjects from five different national cultures.We found that the subjects valued
status independently of any monetary consequence and were willing to trade off some
material gain to obtain it. Although this result was stable across the five cultures, the
intensity of the striving for status and the desirability of a public display of status varied
strongly: the intensity of the status motive corresponded to Hofstede’s power distance
index of the respective culture. Finally, the amount of status seeking observed differed
for men and for women, a preliminary but intriguing observation that deserves further
study.
Humans strive not only for access to
resources and material benefits but also for
intangibles such as status, which is characterized by a rank-ordered relationship among
people associated with prestige and deference behavior (e.g., Ridgeway and Walker
1995). Much of the research in sociology
views status as a means to obtain future
resources (e.g., Lin 1990, 1994); status also
leads to power (e.g., Thye 2000). This implies
that people should pursue status “rationally,”
as a symbol of ability and as a means to
obtain resources.
It has also been suggested, however, that
status may be not only a means to an end but
* Bernardo A. Huberman wants to thank INSEAD
for the kind hospitality provided during the course of
this work. We thank Denise Salin of the Swedish
School of Economics in Helsinki for help with the
experiment and for valuable suggestions. We also
thank three anonymous referees for very helpful
comments, from which we learned a great deal.
Address correspondence to Christoph H. Loch,
IN SEA D, Bouleva rd de Constance, 77 305
Fontainebleau Cedex, France; christoph.loch@
insead.edu.
also an end in itself, an intrinsic component of
an individual’s utility function in addition to
the pursuit of resources (e.g., Barkow 1989;
Emerson 1962; Frank 1988). In this paper, we
test experimentally whether status may represent an end in itself by showing that subjects are willing to trade off an ephemeral
status symbol (which will not carry over into
their lives after the experiment, and thus will
not lead to resources) against an immediate
resource (money). Thus status itself can be a
valuable resource that generates direct utility
for individuals. Intrinsic status seeking by
individuals has important implications for
social and economic systems because it can
provide a powerful motivation to perform; it
also can lead to unproductive competitions
with no obvious social value, such as in the
overconsumption of positional goods (Frank
1985; Loch, Huberman, and Stout 2000).
After briefly reviewing previous work on
status-seeking behavior, we introduce the
experimental model and hypotheses. Next,
we discuss the results, and we conclude with
suggested topics for further research.
103
104
SOCIAL PSYCHOLOGY QUARTERLY
OVERVIEW OF RELATED WORK
tions (Ball and Eckel 1996) and market interchanges (Ball et al. 2001), and were credited
In sociology, status structures are defined with different personality traits by their coas “rank-ordered relationships among actors subjects (Gerber 1996). In Ridgeway’s work
describing the interactional inequa lities on status construction theory (Ridgeway and
formed from actors’ implicit valuations of Balkwell 1997 , Ridgeway et al. 199 8;
themselves and one another according to
Ridgeway and Erikson, 2000) in encounters
some shared standard of value. Status refers
between experimental subjects with different
to one’s standing in a social hierarchy as status levels, the lower-status subjects were
determined by respect, deference and social willing to adopt beliefs that disfavored theminfluence” (Ridgeway and Walker 1995:281).
selves in relation to their higher-status counEarly social scientists were quite willing
terparts,toand such beliefs could spr ea d
Delivered by Ingenta
to see status as an intrinsically valued social
through a Mrs
population
via observation.
INSEAD
LIBRARY;
LE CAMUS
(cid 35002204)
1
resource (e.g., Veblen 1899; Weber 1964).
In summary, this work is consistent with
IP:
194.57.233.136
Similarly, Emerson (1962, 1972) viewed stathe view of status as based on competence
Date: 2004..08..17..05..26..
tus recognition as an “ego-reward,” a highly
and serving as a means to an end such as
valued (emotional) good that could be given
power or resources (although not necessarily
by a lower-powered partner in an exchange
consciously). Some researchers, however,
to increase the higher-po wered partner’s
have called for acknowledgement of the role
“emotional investment” and make the power
of emotions in group behavior: “Sociologists
balance more equal.
have unwisely elevated the rational over the
Expectations state theory concentrates
emotional in attempting to understand and
on how regularly observed inequalities (and
explain human behavior. It’s not that human
associated expectations) govern opportunibeings are not rational—we are. The point is
ties, participation, evaluation, and influence
that we are not only rational. What makes us
in groups with a common goal (e.g., Bales
human is the addition of a rational mind to a
1953; Berger, Fisek, et al. 1977; Blau 1964); it
preexisting emotional base” (Massey 2002:2;
does not focus on what motivates status strivalso see Runciman 1998 ; Tooby and
ing. Although the theory is not inconsistent
Cosmides 1990).
with the idea of status as an intrinsic emoIn line with this argument, work in ecotional goal (Berger, Wagner, et al. 1985), it
nomics has formalized status as a component
has not examined how individual status strivof the utility function that people maximize
ing may pit group members against a com(e.g., Bolton and Ockenfels 2000; Coelho and
2
mon group goal.
McClure 1993; Congleton 1989; Kahneman
Other research has approached status
and Thaler 1991; Konrad and Lommerud
simply as a rational tool: it can provide a sig199 3; Robson 19 92). Kem per (1991)
nal about product quality and thus can help
observed that status assigned to experimenguide a consumer’s choice of product
tal subjects systematically causes positive
(Podolny 1997). At the same time, status can
emotions (if the subject is accorded status) or
be used as a means to gain valuable resources
negative emotions (if status is lost or given to
via a better hierarchical position in society
another subject). In keeping with Emerson’s
(Lin 1990, 1994). Status leads to power (by
power depend ence theory, Frank (1984)
increasing the value of a high-status individshowed that workers’ pay scales sometimes
ual’s goods; e.g., Lovagli a 1994; Lovagli a
are compressed because low-status workers
1997; Thye 2000). This point is consistent with
must be “paid” for giving high-status workers
studies in which people with higher status
the pleasure of prestige. Frank (1988) argued
were conceded higher benefits in negotiathat striving for status as a valuable resource
could lead to strong motivational effects if
1 Social relationships could rest on various affectustatus depends on workers’ relative producal, emotional, or traditional bases (Weber 1964:137).
tivity (or performance). Loch et al. (2000)
2 In our experiment, participants pursue a status
extended Frank’s model to show that status
symbol that pits them against the group. There is no
also may reduce productivity if it rests on
performance toward any group goal (see next sec“political” criteria.
tion).
STATUS AS A VALUED RESOURCE
105
Some of this work draws on biology and ticipants try to win a risky all-or-none payanthropology and on the fact that status off.3 In the first stage, four players in a group
behavior is a general tendency among pri- compete against one another to win the right
mates. Evolution favors efficient competition to participate in a lottery held in Stage 2 of
among group members, to be performed with the game. They do so by allocating to that
as little risk of injury as possible. goal a certain amount of their game card
Determining which of two competing indi- endowments, which are fixed and equal for
viduals would be likely to win in an all players (this is the Stage 1 “expenditure”
encounter, without actual fighting, leads to a or “investment”). From all the game cards
status hierarchy in primate groups. Human allocated by the players in one group toward
prestige has developed from the primate sta- winning the first stage, the experimenter ranDelivered
Ingenta
to one card; the owner of that card
tus tendency but has become symbolic
andby domly
pulls
INSEAD
Mrs
CAMUS
flexi ble (because ancien t humans
oftenLIBRARY;
is the winner
of LE
the first
stage. (cid 35002204)
IP: 194.57.233.136
migrated to different habitats). Status
criteria
In the status condition of the experiment
Date:exam2004..08..17..05..26..
are largely determined culturally—for
(but not in the no-status condition), the Stage
ple, by skills, knowledge, or the control of 1 winner receives public recognition from all
resources, or even arbitrarily (Barkow 1989). other players (announced by the facilitator at
At the same time, the primate status bias the outset) in the form of applause. Status
favoring phys ical size still exists, though (resting on prestige rather than on a formal
weakly: tall men tend to be listened to more position) is conveyed by an expression of
than short men, are granted more respect,
recognition and approval from others, such as
and, on average, enjoy better career progress
a standing ovation or applau se (Barkow
(e.g., Cialdini 1993:181– 82). Moreover, tall
1989:20 3; Frank 1995:11 3; Ridgeway and
men tend to have more reproductive success:
Walker 1995). Applause is a universally recwomen actively select for stature in male
ognized way of expressing such approval.
partners (Pawlowski, Dunbar, and Lipowicz
In both conditions of the experiment the
2000).
winner of Stage 1 is permitted to take part in
People in all cultures crave respect and
a lottery (the second stage). In Stage 2 the
recognition (e.g., Cialdini 1993, 2001; Güth
experim enter replaces the cards that the
and Tietz 1990). This suggests that status
Stage 1 winner spent in the first round with
sometimes may be based not merely on performance but on any culturally accepted sym- “loser cards,” and then draws one card from
bol, and that peo ple may pursue such a the total combined stack of game cards and
symbol not to gain resources but as an end in loser cards. If the player wins the lottery (that
is, if one of his or her remaining game cards is
itself.
In summary, we find a strong theoretical drawn rather than a loser card), the player
basis as well as empirical support for the fact receives the prize for the game. Thus the
that status signals competence, provides chance of winning the prize decreases with an
access to power and resources, and therefore increase in the investment made to win the
is pursued consciously in many situations. first stage. The participants therefore are
The social sciences, however, still lack empir- forced to make a trade-off between allocatical evidence for the pursuit of status as an ing more of their budget to influence the out(emotional) end in itself (an exception is come of Stage 1 and increasing their chances
Bakshi and Chen 1996 in the context of stock of winning in Stage 2.
market prices). Our experiment attempts to
address this situation.
3 The first stage modeled a contractor’s investment
A STATUS-SEEKING EXPERIMENT:
MODEL AND HYPOTHESES
The experimental model consists of a
two-stage game based on Tullock’s (1980)
“rent seeking” formulation, in which the par-
in an effort to win a government contract; the second
stage, the execution of the contract. Competition
forces the contractors to overinvest (e.g., engage in
influencing) in the first stage in order to win the contract; the overinvestment in turn, forces them to cut
corners in the second stage so as not to lose money.
Thus two-stage contracts (“rent seeking”) may cause
wasteful activity.
106
SOCIAL PSYCHOLOGY QUARTERLY
Our experiment shows that people are
willing to overinvest in Stage 1, thus making a
monetary trade-off for an arbitrary status
symbol that carries no resource significance
for the future. As a result, the status symbol
reduces the group’s performance by diminishing everyone’s chances of winning the payoff. This behavior would be irrational unless
people cared about it intrinsically.
To model a player’s decision, suppose
et al. 1977). This point implies that the presence of status does not interact with risk preferences over money.
Formal Result
This game contains a unique Nash equilibrium that is characterized by
xi = –
Sx +
j¹i
j
xj
US
S
Delivered
by Ingenta to x 2 w
j¹i
(2)
S
j
+ i { S xj 1 + U ,
INSEADofLIBRARY; Mrs
LE CAMUS
(cidP35002204)
j¹i
j¹i
is the aggregate rent-seeking expenditure
IP: 194.57.233.136
the opponents in Stage 1, wi is player
i’s initial
for i = 1,...,n.
Date:left
2004..08..17..05..26..
endowment, w i – x i is the amount
for
that xi is player i’s Stage 1 expenditure,
=
Stage 2, and UP is the utility derived from the
predetermined and publicly known prize
P > 0.
In the status condition of the game, winning the first stage brings recognition and
thus confers status. If status carries value, we
would expect individuals to gain utility from
achieving it, in addition to the utility derived
from the final prize. We call this utility
derived from status US. This is a special case
of the utility function used in Loch et al.
(2000): status is evaluated as per the definition relative to the other players; here, only
being first (winning the stage) brings recognition, while no other ranks bring any. As the
experimenter draws from all the game cards
allocated by the players for the first round,
the probability that player i wins the first
stage is xi / xi +
Sx
j¹i
j
.
Given that the player has succeeded in
reaching the second stage, the chance of winning the prize is expressed by the probability
that one of the player’s remaining cards is
drawn, or (wi – xi) / wi. Thus each player faces
the following first-stage problem:
Max EUi(xi, x–i) =
xiÎ[0,wi]
xi
xi +
Sx
j¹i
j
wi – x i
UP + US .
wi
(1)
In this utility function, status and monetary utility are additive, in keeping with previous models (e.g., Loch et al. 2000) and with
empirical results in sociology (Berger, Fisek,
The following argument proves the formal result. Taking the actions xj of the other
players as given, the first derivative of player
i’s objective function is [(wi – xi) Sj¹i xj – xi
(xi + Sj¹i xj)] UP + (xi + Sj¹i xj) wi US. Setting
the first derivative to zero for all players
yields Eq. (2). Furthermore, it is a straightforward matter to show that the second derivative is negative; thus the objective function is
strictly concave for each player. Therefore
there exists a unique Nash equilibrium that is
characterized by the first-order conditions
(Harker and Pang 1990; Nash 1950). Note
that in the no-status condition, Eq. (2)
reduces to
xi = –
Sx
j¹i
j
+
=
Sx
j¹ i
2
j
+ wi {
Sx
j¹i
j
for i = 1,...,n. This proves the formal result.
If the opponents’ equilibrium expenditures in Stage 1 are positive, inspection of Eq.
(2) shows immediately that the equilibrium
first-stage expenditures xi must increase with
the size of the status utility US relative to the
monetary utility UP. The main proposition,
that status is an end in itself, translates into
the following hypothesis in terms of the
model:
Hypothesis 1: Status recognition via the
applause symbol carries a positive intrinsic
value, US; thus the first-round expenditure is
significantly higher in the status condition of
the experiment than in the nonstatus condition.
We performed the experiment in five countries: the United States, Turkey, Hong Kong,
and Germany, and with Swedish-speaking
Finns at the Swedish School of Economics in
STATUS AS A VALUED RESOURCE
107
Helsinki. Hypothesis 1 claims that the desire tive purchasing power or utility. Therefore we
for status and prestige is a universal human cannot compare absolute values, but only the
characteristic (e.g., Barkow 1975, 1989) ; incremental changes in the Stage 1 expenditherefore we expect to find significantly high- tures when a status condition is imposed. This
er expenditures in the first round in all cul- yields our second hypothesis:
tures when a status symbol is available.
Hypothesis 2: The intrinsic value of status is
The relative importance of status, howevranked across the countries as follows: U S
er, is influenc ed by culture, particularly
(Hong Kong) > U S (Turkey) > U S (United
national culture. A proxy for the importance
States) > U S (Germany ) > U S (Sweden/
of status in national cultures that is available
Finland). Thus the incremental difference in
for many countries is Hofstede’s power disthe first-round
expenditures between the stato
tance index (Hofstede 1997). 4 Delivered by Ingenta
tus
condition
and
no-status(cid
condition
(xi
CAMUS
35002204)
The power distance index isINSEAD
defined asLIBRARY; Mrs LE the
(status)
–
x
(no
status))
is
ranked
in
the
same
IP: 194.57.233.136
i
the extent to which the less powerful
memorder.
Date:within
2004..08..17..05..26..
bers of institutions and organizations
a country expect and accept the fact that
power is distributed unequally (Hofstede
DESIGN AND RESULTS
1997:28). Of course, power is not the same as
status, but Hofstede’s results showed (consis- Procedure
tent with Lovaglia 1994 and Thye 2000) that a
The game was implemented as follows.
country’s power distance is correlated highly The subjects had no a priori knowledge that
with the use, expectation, and approval of sta- they were being studied to clarify the nature
tus differences (Hofstede 1997:37). Because of status. They were assigned randomly to the
this is the only measure availab le across status/no-status conditions and, within condimany countries, it is justifiable to use power tions, to groups of four. Each player received
distance as a proxy for the importance of sta- 30 cards, which represented the endowment
tus in a culture. Out of 53 countries, Hong to be used throughout the game.
Kong ranked sixteenth (index 68), Turkey
First, each player was allowed to choose
nineteent h (66), the United States thirty- how many of the cards he or she would spend
eighth (40), Germany forty-second (35), and in Stage 1 by sending the appropriate number
Finland /Sweden 5 forty-sixth/forty-eighth of cards to the experimenter. The experi(index 33/31) (Hofstede 1997:26).
menter then mixed together all the cards for
We were unable to control the financial each group and chose one. This step detersituation of the students who participated in mined the winner of the rent-seeking game in
five countries. We used US$20 as the lottery each group: cards sent to the experimenter
prize, translated at current exchange rates were the Stage 1 investment that increased
into the appropriate national currency. This the chance of reaching Stage 2.
approach, however, leaves open the possibiliIn the lottery of Stage 2, the winning
ty that the amount of money as weighed player’s remaining cards were mixed with
against status was not comparable in effec- blank “loser” cards to total the origin al
4
Hofstede’s influential study of managers from
more than 50 countries, conducted for the first time in
1980 and repeated since, found statistical factors for
cultural characteristics, including the power distance
index (PDI), individualism,the differentiation of gender roles, and uncertainty avoidance. The PDI captures fear of disagreement with superiors, autocratic
management style, and preference for autocratic style
by subordinates. (The measures cluster tightly.)
5 Swedish-speaking Finns, a minority of about eight
percent in Finland, regard themselves as Finnish
nationals but also have maintained a significant
amount of Swedish culture.
endowment. If the lottery manager drew a
card belonging to this player, he or she won
$20. The game ended after this drawing; participants were paid privately and were asked
to leave the room.
In the first condition (no status), the
game was conducted exactly as described
above. In the second version (status), we
introduced a status condition by stating at the
beginning of the game that the winner of
Stage 1 would be announced publicly, given a
small tag saying “Winner,” and congratulated
108
SOCIAL PSYCHOLOGY QUARTERLY
by all participants with applause. The second should observe a higher (to the point of ineffistage of the experiment remained the same ciency) Stage 1 investment when the status
as in the no status treatment. In both treat- condition is introduced into the game. To test
ments, participants were asked in a postgame this claim we compared the Stage 1 investquestionnaire to explain their rationale for ments across the two conditions in each counthe decision. In the status condition, they also try. As see n in the table, the average
were asked what they felt when the winner investments are systematically higher in the
status condition in four countries (Hong
was announced.
The experiments were conducted with Kong, Turkey, the United States, and
undergraduate students in four major univer- Germany). The Kruskal-Wallis test shows that
sities in the United States, Turkey, Hong this increase is statistically significant (p < .05)
except fortoGermany (p = .22). The lower sigKong, and Germany, and at the
Swedis hby Ingenta
Delivered
nificance in
Germany
can be explained
by the
School of Economics in Helsinki,
Finland. In
INSEAD
LIBRARY;
Mrs
LE CAMUS
(cid 35002204)
the first four countries, subjectsIP:
were
paid a fact that the subject pool was considerably
194.57.233.136
show-up fee equivalent to $5; theDate:
Finnish
stu- smaller than elsewhere, which resulted in a
2004..08..17..05..26..
dents received attendance credit in a man- higher p value. Even though the statistical sigagement class. In all cases, the subjects were nificance is not very strong, the difference still
paid what they earned in the experiment. The points in the right direction.6
Thus the results for the first four counexperiment used a between-subjects design,
so that no subject participated in more than tries support our first hypothesis, while the
one type of game. The games were conducted investments for Swedish/Finnish Stage 1 stay
in a classroom in such a way that the subjects constant over the two conditions. (The difference is slightly negative but is not statistically
could not communicate with one another.
different from zero, p = .81. On the basis of
Hypothesis 2, we expected the status effect to
The Intrinsic Value of Status
be weakest in this treatment.)
Figure 1 shows the distribution of Stage 1
Because the only difference between the
expenditures in the two conditions (status and two experimental conditions is the presence
no-status) for all treatments. The Shapiro- or absence of a status component, our results
Wilk test statistic indicates that the distribu- show clearly that participants in the first four
tions in the German treat ments are not countries valued status independently of any
normal. To compare the distributions of Stage monetary consequence and were willing to
1 expenditures between status and no-status trade off some material gain to obtain it.
conditions, we employ a nonparametric statisThe higher spending in the first round of
tical test (Kruskal-Wallis) in the subsequent the game did not help to increase the expectanalysis.
ed payoff at the end. If participants spent too
The mean values (and standard deviations) of rent-seeking expenditures in each
6 One could ask whether the variance of the status
treatment are presented in Table 1, with n
condition might increase because of added consideradenoting the number of participants. If indition for participation, resulting in an increase in confusion. This turned out not to be the case.
viduals view status as a valuable resource, one
Table 1. Average Stage 1 Investment
Experiment
No Status
—Average investment
——(game cards)
—Standard deviation
—n
Status
—Average investment
——(game cards)
—Standard deviation
—n
U.S.
Turkey
Hong Kong
Germany
Sweden/Finland
16.09
16.17
14.32
10.85
14.51
3.67
44
5.27
36
3.39
28
6.45
20
5.06
76
17.72
19.32
18.17
13.26
13.38
3.41
36
5.19
32
4.65
24
5.81
19
3.75
80
STATUS AS A VALUED RESOURCE
50%
109
Hong Kong
40%
30%
20%
10%
0%
50%
Turkey
40%
30%
Delivered by Ingenta to
INSEAD LIBRARY; Mrs LE CAMUS (cid 35002204)
IP: 194.57.233.136
Date: 2004..08..17..05..26..
Percentage of Individuals
20%
10%
0%
50%
United States
40%
30%
20%
10%
0%
50%
Germany
40%
30%
20%
10%
0%
50%
Finland
40%
30%
20%
10%
Stage 1 Investment
Status
28-30
25-27
22-24
19-21
16-18
13-15
10-12
7-9
4-6
0-3
0%
No status
Figure 1. Distributions of Stage 1 Investments
much in the first round, they reduced the
expected final payoff by not leaving enough
cards for the lottery in the second round.
Participants in the status condition of the
gam e recognized this trade-off. In the
postgame questionnaire, several participants
explained it as follows: “On the one hand,
you want to increase the chances of winning
the money. On the other hand, you want to
get recognition from your peers since you
won’t get anything by losing in the first game.
If you pass, you get the applause and might
also get the cash.” Nor did the status-seeking
behav ior serve other purposes, such as
110
SOCIAL PSYCHOLOGY QUARTERLY
increasing a player’s future reputation: the
participants separated after the game was
over, and the result did not exert any further
influence on their lives. Most participants
returned the “winner” tag after the game
because it carried no further value for them.
Therefore these results show that the participants valued a generally recognized status
symbol, such as applause or being acknowledged.
duct a Chow test between the three pairs of
slopes closest together.
The Chow test shows whether two models differ structurally from one another or
whether they could be combined into a single
regression equation: in other words, whether
the regression parameters are statistically
different. When Hong Kong is compared with
Turkey, Turkey with the United States, and
the United States with Sweden/Finland, the
respective slopes are significantly different (p
< .05). This
Cross-Cultural Differences in Status-Seeking
Delivered by Ingenta
to finding strongly supports our
second hypothesis:
the intrinsic
valuation
of
Behavior.
INSEAD LIBRARY;
Mrs LE CAMUS
(cid
35002204)
status
(as
weighed
against
a
monetary
beneIP: 194.57.233.136
Our second hypot hesis concerns the fit) in different countries is related to the
Date: 2004..08..17..05..26..
cross-cultural comparison of status-seeking
countries’ power distance indices.
behavior. We expect to see a decrease in levWe now discuss a possible explanation
els of status seeking from Hong Kong to for the “disappearan ce” of status in the
Turkey, the United States, Germany, and Finnish sample. This is related to the highly
Sweden/Finland. To analyze this cross-cultur- egalitarian nature of Scandinavian culture,
al effect, we conducted an ordinary least which includes a feature known as “the laws
squares regression analysis for each country, of Jante” (janteloven in Norwegian, jantelagen
where Stage 1 expenditure was the depen- in Swedish). This norm strongly encourages
dent variable (y) and the presence of status individuals to be modest, to keep a low prothe explanatory variable (x, where x = 0 in file, and not to act superior to others. The
the no-status condition and x = 1 in the status term was invented for a Norwegian novel
condition). 7 In the regression model for each describing a fictitious town (Sandemose
country, the parameter estimate of the slope 1933) and quickly entered the general vocabcorresponds to an estimate for the incremen- ulary across Scandinavia because it struck an
tal increase in the Stage 1 expenditure when
a status symbol is introduced (Table 2). By
comparing the slopes of the models, we check Table 2. Stage 1 Investment as a Function of Status
for Five Countries
whether the relative effect of status symbols
Country
Regression Output
is different for each of the four countries.
The statistical significance of the regres- Hong Kong
ŷHK = 14.32 + 3.85xHK
sion coefficients repeats the results of the test —Regression output
1.12
of Hypothesis 1. The results for Germany are —Standard error
52
inconclusive because of the small sample —n
Turkey
size; thus we do not include Germany in the —Regression output
ŷTR = 16.18 + 3.14xTR
subsequent analysis. We include Sweden/
—Standard error
1.13
Finland, howeve r, where status does not —n
68
seem to increase the Stage 1 expenditures.
U.S.
ŷUS = 16.09 + 1.63xUS
To test Hypothesis 2, we now compare —Regression output
—Standard
error
.96
the slope estimates across the four remaining
80
countries. The average expenditure increase —n
associated with the introduction of status is Germany
—Regression output
ŷGER = 10.85 + 2.41xGER
highest in Hong Kong, followed by Turkey
—Standard error
1.97
and the United States, and lowest (about —n
39
zero) in Sweden/ Finland. To check whether
Sweden/Finland
ŷFIN = 14.51 – 1.14xFIN
the slopes are statistically different, we con- —Regression output
7
The comparative analysis results also hold in a
log-linear regression model, which we do not report
here.
—Standard error
—n
.71
156
Note: y = Stage 1 expenditure; x = 1 for status, x = 0
otherwise.
STATUS AS A VALUED RESOURCE
111
important chord in Scandinavian culture. intriguing observation: males and females
Jantelagen expresses the idea that responded differently to status. Table 3 shows
Scandinavians feel very uncomfortable when the average Stage 1 investment in the no-stanamed publicly in front of a crowd. For exam- tus condition by country, separately for men
ple, the Swedish press has ascribed to the and for women, as well as the increase in
Wallenbergs, the richest and most powerful Stage 1 expenditure in the status condition.
family in Sweden, the motto “Be but do not In the no-status treatment, the average Stage
be seen!” 8
1 expenditure is higher for males than for
The responses in the Scandinavian ques- fem ales in all countries (p < .05). (The
tionnaire support this interpretation. Twenty- Germany treatment is not included in the
five percent of the winners refused to answer analysis because of the low level of female
Ingenta to three in the no-status condition
the question about their feelingsDelivered
when they by participation:
were announced as winners (although
theyLIBRARY;
INSEAD
LE CAMUS
(cid 35002204)
and four inMrs
the status
condition).
were not shy about discussing their
IP:rationale
194.57.233.136
This finding is consistent with observafor playing as they did). Thirty percent
the tions in anthropology and psychology (e.g.,
Date:of2004..08..17..05..26..
winners emphasized that they were lucky. Barkow 1989; Campbell 2002; Maccoby 1998;
Among all participants, not a single person Pawlowski et al. 2000; Wilson and Daly 1985)
mentioned the status symbol (the applause).
that males across situations are systematicalRather, thirty-one percent of the participants
ly more aggressive than females, and seek
commented on winners’ luck or the fairness status more intensely.
of the game.
In our experiment, men reacted more
Thus Scandinavians are strongly averse strongly to a salient status symbol than did
to being acknowledged in public. Yet this women in all countries (even in the small
does not necessarily imply that their desire Ger man sample), ex cept in the
for status disappear s, but only that Swedish/Finnish sample. In the other counScandinavians prefer to satisfy it in a small tries, the comments in the postgame quescircle of people they know rather than in tionnaires support the statistical results. In
public. This interpretation would suggest that the United States treatment, for instance,
culture determines not only the weight of the three of the five first-round male winners
status consideration but also the acceptable commented “It’s nice to get applause”; and
status symbols. This hypothesis must be test- none made a negative comment. Of the four
ed in further research.
female first-round winners, only one made a
positive comment; another stated “I was a bit
Observations on Gender Differences in
embarrassed.”
Status-Seeking Behavior
It remains unexplained why the women
in the Swedish/Finnish sample seemed to be
In addition to the effect of status on rentless averse to public acknowledgment than
seek ing activiti es and the cross-cultural
the men. On the one hand, this agrees with
implications, our results prompted another
earlier findings in gender resear ch, that
behavioral gender differences are purely sit8 Anecdotal conversations with American manuational (Deaux and Lafrance 1998). On the
agers who have moved to Sweden suggest that
other hand, it may indicate interesting inter“employee of the month” schemes backfire there
actions between underlying gender differbecause the employees are embarrassed to be singled
ences and national culture. Although we had
out, even when the intent is positive.
Table 3. Average Stage 1 Expenditures in “No Status” Condition and Average Increase in “Status” Condition
Men
Hong Kong
Turkey
U.S.
Sweden/Finland
Women
No Status
Status Increase
No Status
Status Increase
15.78
20.17
16.70
16.19
04.16
01.30
03.00
–2.82
11.70
15.75
15.31
12.33
2.16
0.79
–.06
1.05
112
SOCIAL PSYCHOLOGY QUARTERLY
not designed our experiment to examine gender differences (sample selection bias may
exist because of differences in recruitment to
academic programs for men and for women),
we believe this topic deserves further study.
nificant implications for designing more efficient institutional arrangements. If individuals seek status as a value in itself, regardless
of subsequent payoffs, an organization cannot be governed with material rewards and
incentive systems alone; one also must consider emotional rewards such as recognition.
CONCLUSION
Under which circumstances may an indiDoes status have an intrinsic value that vidual perceive status as a means or as an
influences one’s preferences? We have shown end? One might reasonably hypothesize that
empirically that humans pursue status as an both mechanisms are at work simultaneously
end in itself across cultures: status
is worth aby Ingenta
all the time.
Delivered
to Which one is more important at
positive amount of material gain.
Cross-culany givenMrs
pointLE
probably
depends
strongly
INSEAD LIBRARY;
CAMUS
(cid 35002204)
tural differences come into playIP:
in the
rela- on the situation: for example, the size of the
194.57.233.136
tive importance of status,Date:
predicted
rationally recogniz able rewards and the
2004..08..17..05..26..
accurately by Hofstede’s cultural power dis- salience and nature of the status symbol may
tance index, and in the relative willingness to influence what is included in a decision to act.
display status publicly. Thus our study of a This topic would be highl y releva nt for
“second ary” or “higher social” emotion understanding when one can motivate peo(Damasio 2000) corresponds to previous ple with incentives as opposed to emotions,
cross-cultural studies of the so-called “basic but no theory currently addresses this quesemotions” (joy, anger, fear, sadness, disgust, tion; it requires further research.
guilt) that have been significantly stable
across cultures, with some important “moduREFERENCES
lations, or variations around a univer sal
Bakshi, Gurdip and Zhiwu Chen. 1996. “The Spirit
theme”. (See the experiment and overview
of Capitalism and Stock Market Prices.”
reported in Scherer 1997:146.)
American Economic Review 86:133–57.
These results have implications beyond Bales, Robert F. 1953. “The Equilibrium Problem
determining the nature of status. For examin Small Groups.” Pp. 111–61 in Working
ple, status forms a basis for the existence of
Papers in the Theory of Action edited by
positiona l goods, serving as an additional
Talcott Parsons, Robert F. Bales, and Edward
A. Shils. Glencoe, IL: Free Press.
source of value to those goods. For some
positional goods, status may be the only value Ball, Sheryl B. and Catherine C. Eckel. 1996.
“Buying Status: Experimental Evidence on
(like the applause in our experime nt). In
Status in Nego tiatio n.” Psych ology and
organizations, status symbols may be an
Marketing 13:381–405.
important determinant of organizational per- Ball, Sheryl B., Catherin e C. Eckel, Philip J.
formance (e.g., Loch et al. 2000), and statusGrossm an and Willia m R. Zame. 20 01.
seeking efforts are not always efficient. More
“Status in Markets.” Quarterly Journal of
Economics 116:161–88.
specifically, if status symbols are political,
they may encourage office politics, whereas Barkow, Jerom e H. 1975. “Strategie s for Self
Est eem a nd P res ti ge in Mar ad i, Nig er
merit-based symbols place value on contribuRepublic.” Pp. 373–88 in Psych ological
tion and group performance.
Anthropology edited by T.R. Williams. The
The experim ents reported here imply
Hague: Mouton.
that people tend to overinvest resources ———. 1989. Darwin, Sex, and Status, Toronto:
whenever “winning against other s” is
University of Toronto Press.
involved, because winning confers status. Our Be rge r, Jo se ph, M. Ha mit Fisek, Rob ert Z.
Norman, and Morris Zelditch Jr. 1977. Status
comparative analysis suggests some condiCharacterist ics and Social Interaction : An
tions under which this inefficient status-seekExpectation States Approach, New York:
ing behavior is aggravated, namely cultural
Elsevier.
and gender differences. Further empirical Berger, Joseph, David G. Wagner, and Morris
studies of different factors that influence staZelditch Jr., 1985. “Expectation states theotus-seeking activities would provide further
ry: Review and assessment . “ Pp. 1–72 in
insight into the problem and would have sigStatus, Rewards and Influence, edited by
STATUS AS A VALUED RESOURCE
113
Joseph Berger and Morris Zelditch Jr. San
“Economic Analysis and the Psychology of
Francisco: Jossey Bass.
Utili ty: Ap pli cati ons to C ompensati on
Blau, Peter M. 1964. Exchange and Power in Social
Policy.” Ameri can Eco nomic Revi ew
Life. New York: Wiley.
81:341–46.
Bolton, Gary E. and Axel Ockenfels. 2000. “ERC:
Kemper Theodore D. 1991. “Predicting Emotions
A The ory of Equity, Re cip ro cit y a nd
From Social Relations.” Social Psychology
Competition.” American Economic Review
Quarterly 54:330–42.
90:166–93.
Konrad, Kai A. and Kjell E. Lommerud. 1993.
Campbell, Anne. 2002. A Mind of Her Own: The
“ Relat ive Stand ing Compari sons, Risk
Evolutionary Psychology of Women. Oxford:
Taking, and Safety Regulations.” Journal of
Oxford University Press.
Public Economics 51:345–58.
Cialdini, Robert B. 1993. Influence. New York:
Lin, Nan 1990. “Social Resources and Social
Morrow.
Mobility: A Structural Theory of Statu s
Delivered by Ingenta
to
———. 200 1. “The S ci ence o f Per sua si on.”
Attainment.” Pp. 247–71 in Social Mobility
INSEAD
LIBRARY;
Mrs LE
CAMUS
35002204)
Scientific American, February, pp: 62–67.
and Social
Structure,
edited(cid
by Ronald
L.
IP: 194.57.233.136
Coelho, Philip R.P., and James E. McClure.
1993.
B rei ge r. C ambrid ge, UK: Cambridge
“Toward an Economic TheoryDate:
of Fashion.”
University Press.
2004..08..17..05..26..
Economic Inquiry 31:595–608.
———. 1994. “Action, Social Resources and the
Congl eton, Roger D. 1989. “Efficie nt Statu s
Emergence of Social Structure.” Pp. 67–85 in
Seeking.” Journal of Economic Behavior and
Advances in Group Processes, vol.11, edited
Organization 11:175–90.
by Barry Markovsky, Edward J. Lawler, Jodi
Damasio, Antonio R. 2000. The Feeling of What
O’Brien, Karen Heimer. Greenwich, Conn:
Happens. New York: Vintage Books.
JAI Press.
Dea ux , Kay, and Mari ann e Laf ran ce. 1 998 .
Loch, Christoph H., Bernardo A. Huberman, and
“Gender.” Pp. 788–827 in The Handbook of
Suzanne K. Stout. 2000. “Status Competition
Social Psychology, vol.1, 4th ed., edited by D.
and Performance in Work Groups.” Journal
T. Gilbert, Susan T. Fiske, and Gardner
of Economic Behavior and Organization
Lindzey. Oxford: Oxford University Press.
43:35–55.
Emerson, Richard M. 1962. “Power-Dependence
Lovaglia, M. J. 1994. “Relating Power to Status.”
Relations.” American Sociolgical Review
Pp. 87–111 Advances in Group Processes,
27:31–40.
vol.11, edited by Barry Markovsky, Edward
———. 1972. “Exchange Theory II.” Pp. 58–87 in
J. Lawler, Jodi O’Brien, Karen Heimer.
Sociological Theories in Progress, vol. 2, editGreenwich, Conn: JAI Press.
ed by J. M. Berger , M. Zelditch , and B.
———. 1 997 . “ Power and In flu en ce: A
Anderson. Boston: Houghton-Mifflin.
Theoret ica l Bri dg e.” So ci al Force s
Fra nk, Rob ert H. 19 84. “Inte rde pe nde nt
76:571–603.
Prefer ences and the Competitiv e Wage
Mac cob y, Ele ano r E. 19 98 . T he Two S exes :
Structure.” RAND Journal of Economics
G rowin g Up Ap ar t, C oming Tog et her .
15:510–20.
Cambridge, MA: Belknap.
———. 1985. “The Demand for Unobservable
Massey, Douglas S. 2002. “A Brief History of
and Other Nonpositional Goods.” American
Human Society: The Origin and Role of
Economic Review 75:101 –16.
E motion s in S ocial Lif e.” A mer ic an
———. 1988. Passions Within Reason. New York:
Sociological Review 67:1–29.
Norton.
Nash, John F., 1950. “Equilibrium Points in n———. 1995. The Winner-Take-All Society. New
Person Games.” Proceedings of the National
York: Free Press.
Academy of Sciences 36:48–49.
Gerber, Gwendolyn L. 1996 “Status in SamePawlowski, B., R.I.M. Dunbar, and A. Lipowicz.
Gender and Mixed-Gender Police Dyads.”
2000. “Tall Men Have More Reproductive
Social Psychology Quarterly 59:350–63.
Success.” Nature 403: 156.
Güth, Wer ner and R ein har d Tiet z. 1 990 .
Podolny, Joel M. 1997. “A Status-Based Model of
“Ultimatum Bargaining Behavior.” Journal
Market Competition.” American Journal of
of Economic Psychology 11:417–49.
Sociology 98:829–72.
Harker, Patrick T. and Jong-Sh i Pang. 1990.
Ridgeway, Cecilia L. and J. W. Balkwell. 1997.
“Finite-Dimensional Variational Inequality
“Group Process es and the Diffusion of
and
Non-Linea r Co mplemen ta ri ty
Status Beliefs.” Social Psychology Quarterly
Problem s.” Mathematical Program ming
60: 14–31.
48:161–220.
Ridgeway, Cecilia L., Elizabeth H. Boyle, Kathy J.
Hofstede, Geert 1997. Cultures and Organizations.
Kuipers, and Dawn T. Robinson. 1998. “How
New York: McGrawHill.
do Statu s Beliefs Develo p?” Amer ican
Kahneman, Daniel and Richard H. Thaler. 1991.
Sociological Review 63: 331–350.
114
SOCIAL PSYCHOLOGY QUARTERLY
Ridgeway, Cecilia L. and Kristan G. Erickson.
Power in Exchange Relations.” American
200 0. “Crea tin g and Spread in g Stat us
Sociological Review 65:407–32.
Beliefs.” American Journal of Sociology
Tooby, John and Leda Cosmides. 1990. “The Past
106:579–615.
Explai ns the Pres ent: Em oti on al
Ridgeway, Cecilia L. and Henri A. Walker. 1995.
Adaptations and the Structure of Ancestral
“Stat us Stru ctu re s.” Pp. 2 81– 31 0 in
Environments.” Ethology and Sociobiology
Socio log ic al Per spect iv es o n So ci al
11:375–424.
Psychology edited by K. Cook, G. Fine, and J.
Tullock, Gordon 1980. “Efficient Rent-Seeking.”
House. Upper Saddle River, NJ: Pearson
Pp. 97–112 and 269–82 in Toward a Theory of
Education.
the Rent-Seeking Society, edited by J. M.
Robson,Arthur J. 1992. “Status, the Distribution of
Buchanan, R. D. Tollison, and G. Tullock.
Wealth , Privat e and Social Attitudes to
College Station: Texas A&M Press.
Risk.” Econometrica 60:837–57.
Delivered
by
Ingenta
to
Veblen, Thorstein
1899. Theory of the Leisure
Runciman , Walter G. 1998. “The Select ionist
Class.
New
York:
Macmillan.(cid 35002204)
INSEAD
LIBRARY;
Mrs
LE
CAMUS
Par adigm an d It s Imp li cat io ns for
Weber, Max 1964. The Theory of Social and
IP: 194.57.233.136
Sociology.” Sociology 32:163–88.
Economic Organization. New York: Free
Sandemose, Aksel. 1933. En Flyktnyng
Date:
Krysser
2004..08..17..05..26..
Sitt
Press
Spor. Oslo: Tiden Norsk Førlag.
Wil son , Mar go, and Ma rt in Da ly. 1 98 5.
Scherer, Klaus R. 1997. “Profiles of Emotion“Competitiveness, Risk Taking and Violence:
antece de nt App ra isa l.” Cog nit io n a nd
the Young Male Syndrome.” Ethology and
Emotion 11(2):113– 50.
Thye, Shane R. 2000. “A Status Value Theory of
Sociobiology 6:59–73.
Bernardo A. Huberman is a senior fellow at Hewlett Packard and Director of the Information
Dynamics Laboratory at Hewlett Packard Laboratories in Palo Alto, California. His research
interests include the dynamics of information flow and aggregation among social groups, privacy, and large distributed informational systems.
Christoph H. Loch is a professor of technology management at INSEAD in Fontainebleau,
France. His research interests include innovation management, research and development management, and the motivation of professional personnel.
Ayse Önçüler is an assistant professor of decision sciences at INSEAD in Fontainebleau,
France. Her research interests include behavioral decision making and experimental economics.