Gender differences in altruism: Expectations, actual behaviour and accuracy of beliefs

12 Pages·2016·0.1 MB·English

Gender differences in altruism:
Expectations, actual behaviour and accuracy of beliefs
Pablo Brañas-Garza∗, Valerio Capraro†, Ericka Rascón-Ramírez∗
June 16, 2016
6
1
0
2

n
u Abstract
J

5 Previous research shows that women are more altruist than men in dictator game ex-
1 periments. Yet,little isknownwhetherwomenareexpected tobemorealtruist thanmen.

Hereweelicitthird-parties’ beliefsaboutdictators’ donations conditional onknowingthe
]
h
gender of the dictator. Our data provide evidence of three main findings: (i) women are
p
expectedtobemorealtruistthanmen;(ii)bothmenandwomenhavecorrectbeliefsabout
-
c
the level of altruism among men; and (iii) both men and women overestimate the level
o
s of altruism among women. In doing so, our results uncover a perception gap according
.
s to which, although women are more altruist than men, they are expected tobe even more
c
altruistthantheyactually are.
i
s
y
h Keywords: dictator game,expectations, accuracy ofbeliefs, genderdifferences.
p
[

1 1 Introduction
v
0
0
9 The exploration of gender differences in social preferences, results are more mixed:
4
decision making has a long tradition in be- while previous research has not uncovered
0
. havioural economics and other social sci- anyobviousgenderdifferenceincooperative
6
0 ences, and has touched several research ar- behaviour (see Croson and Gneezy (2009)
6 eas, including risk-aversion, competitive be- for a review), experimental studies have re-
1
: haviour,and socialpreferences. peatedly found that women are, on average,
v
more altruist than men (Bolton and Katok
i For example, a classical study by Eckel
X
(1995); Eckel and Grossman (1998);
andGrossman(2002)hasshownthatwomen
r
a Andreoni andVesterlund(2001);Dufwenberg and Muren
are more risk averse than men, while an
(2006); Houserand Schunk (2009);
equally classical study by Niederle and
Dreber et al.(2014);Capraro and Marcelletti
Vesterlund (2007) has shown that women
(2014); Capraro (2015)). More recently,
are less competitive than men. In terms of
∗EconomicsDepartment,MiddlesexUniversityLondon,BusinessSchool,TheBurroughs,LondonNW44BT,
UnitedKingdom.
†CenterforMathematicsandComputerScience,SciencePark123,1098XG,Amsterdam,TheNetherlands.
1
Rand et al. (2016) have extended this line difference in stereotypes. We move a first
of research by showing, through a meta- step into this research area by starting from
analysis of 22 studies, that promoting intu- asimplequestion: arewomen expectedtobe
itionversus reflection increases altruisticbe- morealtruisticthanmen?
haviour among women, but not among men, To the best of our knowledge, only a
suggesting that women but not men have handfulofpapershaveapproachedthisques-
internalisedaltruismastheirspontaneousre- tion, and mostly did so from a psychologi-
action. cal perspective. For example, Heilman and
Other studies have shown that women Chen (2005) showed that work-related altru-
tend to be more altruistic than men when in- ism isless optionalforwomenthan for men,
vesting in human capital for children. For and Heilman and Okimoto (2007) showed
example, women allocate more resources that penalties for women’s success in male
for women’s and children’s clothing relative domains result from the perceived violation
to men’s clothing (Lundberg et al. (1997)), of gender-stereotypicprescriptions. From an
invest more in health and nutrition for economic perspective, we are aware of only
children (Duflo (2000)), and spend more one study devoted to eliciting participants’
on child goods and small scale livestock beliefs about the level of altruism in men
(Rubalcavaet al. (2009))than men. and women (Aguiaret al. (2009)). In this
Despite the vast research in gender dif- labexperiment,subjectswerepresentedwith
ferences, the literature has mostly neglected two boxes, A and B, where A contained do-
the inverse question of whether people have nations left by men and B contained dona-
gender stereotypes in specific decisional tionsleftbywomen. Subjectswereinformed
settings. Understanding whether people that they could choose only one of the two
have correct beliefs about others’ behaviour boxes and one donation would be taken at
is an important question per se, because random from the selected box and used to
the standard equilibrium analysis assumes pay them. Results showed that subjects
that people strategise on their beliefs about weremorelikelytoselectdonationsfromthe
their counterparts’ behaviour (Camerer et al. "women" box, indicating that women were
(2004));anditbecomesevenmoreimportant indeed expected to be more generous than
when it comes to gender differences, since men.
one of the dominant explanations for gender Although it represents an important first
differences in decision-making relies on the step towards understanding whether women
assumption that the behaviour of men and are expected to be more altruistic than men,
women is governed by stereotypes regard- the work by Aguiar et al. (2009) has two
ing theirsocial roles (Eagly (1987); Brescoll important limitations. First of all, while it
(2011)). In sum, understanding whether shows that women are expected to be more
there is a correspondence between stereo- generousthanmen,itdoesnotshowwhether
typesofmenandwomenandtheiractualbe- people have correct beliefs about the be-
haviourisanimportantquestion,withpoten- haviourofmenandwomen. Thus,itremains
tial consequences in economic and psycho- unclear whether people have correct stereo-
logicalmodelling. types regarding each gender’s level of altru-
Given the aforementioned literature ism. Second, it is only one study: the re-
showingthat women are more altruisticthan centoutbreakofthereplicabilitycrisis(Open
men, here we ask whether this gender dif- Science Collaboration, 2015) calls for more
ferenceinbehaviourcorrespondstoagender studies.
2
In the current work, we wish to fill these
gapsby: (i)replicatingtheresultthatwomen Dictators: They were given $0.20 and were
are expected to be more generous than men; asked to decide how much, if any, to give
and (ii) giving a quantitative version of this to the receiver. Before making their deci-
result. This allows to answer the question: sion, dictators were asked two comprehen-
do men and women fulfil people’s expecta- sion questions. Specifically, theywere asked
tionsaboutaltruisticbehaviour? which choice would maximise their payoff
The rest of the paper is organised as fol- and which choice would maximise the re-
lows. Next section is devoted to methods; ceiver’s payoff. Subjects failing any com-
section 3 focuses on results and discussion; prehension question were automatically ex-
lastsectionconcludes. cluded from the survey. This screening pro-
cedure had the effect that we had fewer dic-
tators (N = 456) than receivers (N = 530).
2 Method
Thus, the computation of receivers’ payoffs
is not straightforward, since there is no one-
2.1 Subject pool to-onecorrespondencebetweendictatorsand
receivers. To address this problem, receivers
Subjects were living in the US at the time
weresequentiallypairedwitharandomlyse-
of the experiment and were recruited using
lecteddictator;incaseadictatorwasalready
Amazon Mechanical Turk (Paolacci et al.
used to pay another receiver, we paid the
(2010);Hortonet al.(2011);MasonandSuri
current recipient ‘out ofour pocket’, and not
(2012); Paolacci and Chandler (2014))
using the donation of that dictator, because
to play a standard Dictator Game
that donation had already been used. This
(Kahneman etal. (1986); Forsytheet al.
procedure is doable on Amazon Mechanical
(1994)).
Turk, because participants are matched only
In the Dictator Game, one player acts after theendoftheexperiment.
in the role of the dictator and the other
one in the role of the receiver. Dictators Receivers: Apartfrompotentiallyreceiving
are given a certain amount of money and money from dictators, receivers played also
are asked how much, if any, they want to as guessers. Specifically, they were asked
give to the receiver. Receivers have no to predict the donation that another dicta-
choice and only get what the dictators de- tor would make to another receiver. They
cide to give. Since dictators have no in- would receive, on top of the actual dona-
centivestogivemoney,apayoff-maximising tion, $0.20 reward for correct guesses. Since
dictator would donate nothing. For this theydonotguesstheirowndonationthereis
reason, dictators’ donations are taken as no opportunitytohedge (Brañas-Garza et al.
a measure of individual’s altruism, or in- (2016)). To elicit recipients’ expectations,
equity aversion (Fehr and Schmidt (1999); we designedfourtreatments:
Boltonand Ockenfels (2000); Brañas-Garza
O : recipients were presented with the
(2006, 2007); Charness and Gneezy (2008)). n
same screenshots shown to dictators
and they were asked to guess the dic-
2.2 Protocol tator’sdecision(N = 134);
In our experiment, subjects were randomly O : was identical to O with the only dif-
mow n
dividedbetween dictatorsand receivers. ference that recipients were informed
3
that the dictator was either a man or a 3.2 Gender framed vs non-
woman(N = 140). framed treatments
As a preliminary step we start by looking at
O : was identical to O with the only dif-
m n framing effects on recipients’ beliefs. Both
ference that recipients were informed
treatments O and O report similar aver-
thatthedictatorwasaman(N = 124); n mow
ages(2.79and3.16,resp.). Table1showsno
significant differences between O and O
n mow
O : was identical to O with the only dif-
w m (t-test, p = 0.24; z-test, p = 0.23). Sim-
ference that recipients were informed
ilarly, we do not find significant differences
that the dictator was a woman (N = betweenO ∪O andO (t-test,p = 0.89;
m w mow
132). z-test, p = 0.84). Hence, the sum of ‘men’
and‘women’framesisequaltothetreatment
in which‘both’gendersare mentioned.
We need both O and O baselines for
n mow Therefore, we may conclude that men-
two reasons: on the one hand, by comparing
tioning ‘gender’ does not frame recipients
O and O with O , separately, we may
m w mow expectations.
investigate the effect of making one partic-
ular gender salient versus making both gen-
derssalient;ontheotherhand,bycomparing 3.3 Are women expected to be
dictators’donationswith O , wecan explore more generous than men?
n
whetherpeoplehavecorrectbeliefsaboutthe
To answer this question we compare treat-
levelofaltruisminanonymousstrangers.
ment O with O and treatment O with
m mow w
O . Figure 1 shows the distributionof be-
mow
liefs by treatment. Figures 1a, 1b and 1c
3 Results
show the histograms for O , O and O ,
m mow w
respectively. While the modal values for ex-
3.1 Descriptive statistics pected behavior of males is 0 (giving noth-
ing) the modal for women (i.e., O ) and,
w
A total of 986 subjects (56% men, mean age to a lesser extent, for women or men (i.e.,
= 34.5 years) participated in our experiment. O ) is the equal split. Average values re-
mow
The average donation was 27.3% of the to- flect the same result: the mean expected al-
tal endowment, which is very close to the truisminO is2.33,whilethemeanforO
m mow
average donation reported in Engel’s meta- is 3.18 (t-test, p = 0.01; z-test, p = 0.01).
analysis of 616 Dictator game experiments Conversely, when the dictator is a ‘woman’,
conducted in the standard physical labora- the mean expected generosity in O is 4.05,
w
tory (28.3%, Engel (2011)). This confirms which is significantly larger than the mean
the reliability of data collected on Amazon for O (t-test, p = 0.01; z-test, p = 0.00).
mow
Mechanical Turk using very small stakes, a Comparing the expected level of generosity
fact that was already observed in the con- among males and females, O vs O , we
w m
text of the Dictator Game by d’Adda et al. observe than the average and median differ-
(2015). Althoughthepiesizewas$0.20data ences are 1.72 and 4 units, respectively. The
arenormalisedsuchthatthedonationscorre- top part ofTable1 showstherelevanttests.
spond to 0-10. Next we pass to the analysis Figure 1ato Figure1c providevisualev-
oftreatmenteffects. idence that we can reject the hypothesis that
4
menareexpectedtobeasaltruisticastheav- 3.5 Do subjects have correct be-
eragepersonandthatwomenareexpectedto liefs about each gender’s av-
beas altruisticas theaverageperson.
erage level of generosity?
Figure 1d focuses on the CDFs (cumu-
lative distribution functions). While males’ In the previous subsections, we have shown
CDF is closer to the top right – more selfish that women are expected to be more altru-
– Females CDF is closer to the bottom left istic than men and that this expectation is
–more generous. It is easy to check that O grounded, in the sense that women are actu-
w
stocastically dominates O which is consis- ally more altruistic than men. Now, we ask
m
tentwiththetestshowninTable1. Theentire whether people have correct beliefs about
distribution of O is always toward the right each gender’saverage levelofgenerosity.
w
of the rest of distributions. In sum, women We begin by observing that subjects
are expectedtomemorealtruisticthan men. have, on average, correct beliefs about the
average level of altruism. Specifically, the
mean levelofaltruismacross theexperiment
(both males and females) is 2.735, while the
Result1 Womenareexpectedtobemoreal-
mean level of expected generosity in the O
truisticthanmen. n
condition is 2.798 (t-test, p = 0.81; z-test,
p = 0.83), see Table 1 bottom). Hence sub-
jectshavecorrectbeliefsaboutaveragelevel
3.4 Are women actually more of generosity, which in turn means that we
generous than men? donotobserveeitherwishfulthinkingorpes-
simism.
Next we analyse whether subjects have
In the previous subsection, we have shown
correct beliefs about men’s average level of
that women are expected to be more altru-
altruism. Figures3aanalysesaccuracyofbe-
istic than men in a Dictator Game. Is this
liefs for men and shows that there is no dis-
expectationgroundedornot?
crepanciessincebothexpectationsandactual
Figure 2a and Figure 2b respectively
behavior are almost identical (CDFs are on
compare the distribution of donations for
parallel).
both men and women, and provides visual
Controlling for the gender of the recipi-
evidence that women are, on average, more
ent, we also find that both men and women
altruist than men (means: 3.04 vs 2.49; t-
have,onaverage,correctbeliefsaboutmen’s
test, p = 0.03; z-test, p = 0.01). In fact,
level of altruism (t-test, guess by men p =
giving nothing is the modal value for males
0.21 and guess by women p = 0.60; z-test,
(49.6% gave 0) while giving the equal split
guess by men p = 0.22and guess by women
is the modal value for women (48.3% gave
p = 0.44,seeTable1 bottom).
half).
In sum, not only women are expected to
Result 3 Both men and women have correct
be more generous than men, but they are de
beliefs about average level of generosity in
factomoregenerousthan men.
men.
HoweverFigure3bshowsstrongdiscrep-
Result2 Women are more altruistic than anciesbetweencurrentbehaviourandexpec-
men. tations for women: females are not as gen-
5
erous as they are expected to be (O CDF (t-test, both p-values < 0.03; z-test, both p-
w
dominatestheD CDF). values< 0.02,seeTable1).
w
This remains true also after controlling
for gender. Both men and women overes- Result 4 Bothmenandwomen overestimate
timate women’s average level of generosity thelevel ofgenerosityinwomen.
Table 1: Hypothesis testing
Hypothesis ParametricTests Non-ParametricTests
Difference T-test P-Value Difference Z-tests P-value
inMeans inMedians
O =O -0.38 -1.17 0.24 -2 -1.19 0.23
n mow
O ∪O =O 0.04 0.14 0.89 0 0.20 0.84
m w mow
O =O -0.85 -2.58 0.01 -4 -2.59 0.01
m mow
O =O 0.87 2.77 0.01 0 2.89 0.00
w mow
O =O 1.72 5.50 0.00 4 5.51 0.00
m w
D=O -0.06 -0.25 0.81 0 -0.22 0.83
n
D =Om 0.44 1.26 0.21 1 1.24 0.22
m m
D =Ow -0.20 -0.53 0.60 -1 -0.78 0.44
m m
D =Om -0.81 -2.25 0.03 0 -2.28 0.02
w w
D =Ow -1.27 -3.92 0.00 0 -3.57 0.00
w w
D =D 0.55 2.21 0.03 4 2.36 0.02
w m
Note: t-tests assume unequal variances per treatment and normality of the distribution of
differencesin means; z-tests correspond to Mann-Whitney-Wilcoxonnon-parametrictests.
D (D )referstomen(women)dictators;Dtoanydictator.
m w
4 Conclusion
Here we have used Dictator Game ex- perception gap according to which, although
periments to measure people’s expectations women are more altruist than men, they are
about dictators’ level of generosity, condi- expected to be evenmorealtruisticthan they
tional on knowing the gender of the dicta- actually are. This result is particularly puz-
tor. Ourdataprovideevidenceofthreemajor zling since it regards also women: while
results: (i) women are expected to be more womenhavecorrectbeliefsaboutthelevelof
generous than men (replicating Aguiaret al. altruism in men, they overestimate the level
(2009) results); (ii) both men and women ofaltruisminotherwomen.
have correct beliefs about the mean level of We hope that future research can shed
generosity among men; (iii) both men and light on the ultimate origin of this percep-
women overestimate the level of generosity tion gap and on the potential psychological
amongwomen. andeconomicconsequencesthatcanhaveon
In doing so, our experiment uncovers a women’sand men’sbehaviour.
6
References
Aguiar, F., Brañas-Garza, P., Cobo-Reyes, R., Jimenez, N., and Miller, L. M. (2009). Are
womenexpected tobemoregenerous? ExperimentalEconomics,12(1):93–98.
Andreoni, J. and Vesterlund, L. (2001). Which is the fair sex? Gender differences in altruism.
TheQuarterlyJournalofEconomics,116:293–312.
Bolton, G. E. and Katok, E. (1995). An experimental test for gender differences in beneficent
behavior. EconomicsLetters, 48(3):287–292.
Bolton,G.E.andOckenfels,A.(2000). ERC:Atheoryofequity,reciprocity,andcompetition.
TheAmerican EconomicReview, 90:166–193.
Brañas-Garza, P. (2006). Poverty in dictator games: Awakening solidarity. Journal of Eco-
nomicBehavior &Organization,60(3):306–320.
Brañas-Garza, P. (2007). Promotinghelpingbehaviorwithframingindictatorgames. Journal
ofEconomicPsychology,28(4):477–486.
Brañas-Garza, P., Rodriguez-Lara, I., and Sanchez, A. (2016). Nobody expect selfishness.
Mimeo.
Brescoll, V. L. (2011). Who takes the floor and why: Gender, power, and volubilityin organi-
zations. AdministrativeScienceQuarterly,56:622–641.
Camerer, C. F., Ho, T.-H., and Chong, J. K. (2004). Behavioural game theory: Thinking,
learning and teaching. In Advances in Understanding Strategic Behaviour, pages 120–
180.Springer.
Capraro, V. (2015). Theemergence ofhyper-altruisticbehaviourin conflictual situations. Sci-
entificReports,4:9916.
Capraro, V.andMarcelletti,A.(2014). Dogoodactionsinspiregoodactionsinothers? Scien-
tificReports,4:7470.
Charness, G. and Gneezy, U. (2008). What’s in a name? Anonymity and social distance in
dictatorandultimatumgames. JournalofEconomicBehavior&Organization,68(1):29–
35.
Croson, R. and Gneezy, U. (2009). Gender differences in preferences. Journal of Economic
Literature,47:448–474.
d’Adda, G., Capraro, V., and Tavoni,M. (2015). Push, don’t nudge: Behavioral spilloversand
policyinstruments. Mimeo.
7
Dreber, A., von Essen, E., and Ranehill, E. (2014). Gender and competition in adolescence:
taskmatters. ExperimentalEconomics,17(1):154–172.
Duflo, E.(2000). Childhealth and householdresources in southafrica: Evidencefrom theold
agepensionprogram. The AmericanEconomicReview, 90(2):393–398.
Dufwenberg, M. and Muren, A. (2006). Gender composition in teams. Journal of Economic
Behavior& Organization,61(1):50–54.
Eagly,A.H.(1987). Sexdifferencesinsocialbehavior: Asocial-roleinterpretation. Mahwah,
NJ:Erlbaum.
Eckel, C. C. and Grossman, P. J. (1998). Are women less selfish than men?: Evidence from
dictatorexperiments. TheEconomicJournal,108(448):726–735.
Eckel,C.C.andGrossman,P.J.(2002). Sexdifferencesandstatisticalstereotypinginattitudes
towardfinancial risk. EvolutionandHumanBehavior,23(4):281–295.
Engel,C. (2011). Dictatorgames: A metastudy. ExperimentalEconomics,14(4):583–610.
Fehr, E. and Schmidt, K. M. (1999). A theory of fairness, competition, and cooperation. The
QuarterlyJournalof Economics,114:817–868.
Forsythe,R.,Horowitz,J.L.,Savin,N.E.,andSefton,M.(1994).Fairnessinsimplebargaining
experiments. Games andEconomicBehavior,6(3):347–369.
Heilman, M. E. and Chen, J. J. (2005). Same behavior, different consequences: Reactions
to men’s and women’s altruisitc citizenship behavior. Journal of Applied Psychology,
90:431–434.
Heilman, M. E. and Okimoto, T. G. (2007). Why are women penalized for success at male
tasks?: Theimpliedcommunalitydeficit. Journalof AppliedPsychology,92:81–92.
Horton, J. J., Rand, D. G., and Zeckhauser, R. J. (2011). The online laboratory: Conducting
experimentsin areal labormarket. ExperimentalEconomics,14(3):399–425.
Houser, D. and Schunk, D. (2009). Fairness, competitionand gender: Evidence from German
schoolchildren. JournalofEconomicPsychology,30:634–641.
Kahneman, D., Knetsch, J. L., and Thaler, R. H. (1986). Fairness and the assumptions of
economics. Journalof Business,59:S285–S300.
Lundberg, S. J., Pollak, R. A., and Wales, T. J. (1997). Do husbands and wives pool their
resources? Evidencefromtheunitedkingdomchildbenefit. JournalofHumanResources,
pages 463–480.
Mason, W. and Suri, S. (2012). Conducting behavioral research on Amazon’s Mechanical
Turk. Behavior ResearchMethods,44(1):1–23.
8
Niederle, M. and Vesterlund, L. (2007). Do women shy away from competition? Do men
competetoo much? The QuarterlyJournalofEconomics,122:1067–1101.
Open Science Collaboration (2015). Estimating the reproducibility of psychological science.
Science, 349.
Paolacci, G. and Chandler, J. (2014). Inside the Turk: Understanding Mechanical Turk as a
participantpool. Current DirectionsinPsychologicalScience, 23(3):184–188.
Paolacci, G., Chandler, J., and Ipeirotis, P. G. (2010). Running experiments on Amazon Me-
chanical Turk. JudgmentandDecision Making,5:411–419.
Rand, D. G., Brescoll, V., Everett, J. A., Capraro, V., and Barcelo, H. (2016). Social heuris-
tics and social roles: Intuition favors altruism for women but not for men. Journal of
ExperimentalPsychology: General.
Rubalcava, L., Teruel, G., and Thomas, D. (2009). Investments, time preferences and public
transfers paidto women. EconomicDevelopmentand CulturalChange, 57(3):507.
9
Figures
Figure 1: Expected behavior for men, women and both
.5 .5
.4 .4
nsity.3 nsity.3
e e
D D
.2 .2
.1 .1
0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Om Omow
(a)Expectations formen[O ] (b)Expectations foreithermenorwomen[O ]
m mow
.8 1
8
.
n
.6 utio
distrib.6
Density.4 mulative .4
u
C
2
.
2
.
0
0 2 4 6 8 10
0 Om Ow
0 1 2 3 4 5 6 7 8 9 10
Ow Owom
(c)Expectations forwomen[O ] (d)CDFsforO ,O andO
w m mow w
10