Economic Decision-Making in Tribal Societies: Three Field

Gender and Economic Behavior:
Insights from Experiments
Gender and Economic Behavior
Gender an important factor in many economic
decisions and outcomes.
Gender differences in wages, prospects for
advancement, occupational choice (Blau and
Kahn (2000), Blau, Ferber and Winkler (2010)).
Bertrand & Hallock (2001): Only 2.5% of US
executives are women, and they earn 45% less
than male counterparts.
Much fewer women in leadership positions in
firms, in politics etc.
Gender Differences in:
•
•
•
•
Risk-taking
Altruism
Beliefs (e.g. overconfidence)
Competitiveness (in performance tasks, or
negotiations/bargaining)
Etc.
Experimental Literature on Gender
• Documents differences
• Explores the source of differences
• Studies the role of institutions in influencing
these differences (e.g. feedback, single-sex
schooling etc.)
Different explanations: ability differences, familycareer balance, discrimination…
Another explanation: men and women might have
different responses to and attitudes toward
competition.
Competitive career path for high-level positions,
self-selection.
Gender and Competitiveness
1. Gender differences in the response to
competition
Gneezy, Niederle & Rustichini (2003):
Men are more motivated by competition than
women.
Piece-rate vs. tournament
Single-sex vs. mixed groups (results mixed)
Gneezy (2004), in Israel
•
•
•
•
Physical ed. Class
Running
When ran alone, no gender difference
Boys’ performance improves in competition
Self-Selection
Gender differences in choice of incentive
schemes:
Niederle & Vesterlund (2007):
Piece-rate, tournament, choice, submit to
tournament.
Women shy away from competition. Piece-rate
vs. tournament: 73% of males, only 35% of
females.
Why?
Nature vs. Nurture:
Is it innate, or shaped by society?
“Sex vs. gender”?
Important for policy issues. Suppose you want to
close the gap:
• If nature, may need to make education
system less competitive.
• If nurture, socialize girls in a different way,
keep education system competitive.
Gneezy, Leonard, List (2009):
Compare a matrilineal society (India) with an
extremely patriarchal society (Tanzania).
• “Men treat us like donkeys”
• --A Maasai woman (Hodgson, 2001)
• “We are sick of playing the roles of breeding
bulls and baby-sitters.”
• --A Khasi man (Ahmed, 1994)
Matrilineal society: Khasi in India
The Khasi Society
• Females are holders of property
• Lineage descends through the mother
(matrilineal)
• Youngest daughter is the heir, and does not
leave the mother’s home.
• Men live in their mother’s or wife’s home.
• Women more involved in economic activity.
• “Men’s rights to property” movement in
Shillong city.
• The return to investment in the human capital
of girls is retained within the household.
• Khasi families can choose to raise the
daughter they would like to keep in their
household, not the daughter most likely to be
preferred by other households.
Gneezy, Leonard & List (2009)
Massai (Tanzania) vs. Khasi (India)
Fraction compete
0.6
0.5
0.4
Tanzania
0.3
India
0.2
0.1
0
Men
Women
• 50% vs. 26% in Massai, 54% vs. 39% in Khasi.
• The “men compete more” result is not universal, and
societal factors can be important.
• One interpretation: Khasi society may remove social
barriers that prevent naturally competitive women from
expressing their true personalities.
• Or, the structure of the Khasi society may allow competitive
women to earn greater rewards for their effort and to pass
on wealth and competitive tendencies to their daughters,
both of which increase the fecundity of competitive genes
[gene-culture co-evolution]
Gender and Socialization
• Data from adults do not tell us at what stage
of the socialization process the difference
starts.
• Learning how competitiveness evolves
through age can be important for policy.
=>Study kids of different ages.
• Matrilineal (Khasi) vs. patriarchal (Kharbi)
=>Both in Northeast India
Very close in terms of location, so can keep
other factors constant and have a cleaner
comparison
Experiments on Children & Teens
Booth and Nolen (2009)—the effect of single-sex
education on risk-taking and competitiveness
Gneezy and Rustichini (2004)—Boys perform better
under competition.
Dreber, Essen & Ranehill (2009)—No difference in
7-10 year-olds’ reaction to competition.
Cardenas et al. (2011)—9-12 year-olds, mixed
results.
Khachatryan (2011)—Girls as competitive, but less
risk-taking around puberty.
Sutter and Ruetzler (2011)—3-18 year-olds, boys
more competitive.
Procedures
• Use children aged 7-15, elicit information on age,
grade in school.
• Task: Throwing tennis balls into a bucket from a
distance (Gneezy et al. find no significant
difference in performance across gender).
• Subject selects between two incentive schemes:
*Piece-rate: 10 Rupees per successful ball.
*Tournament: 30 Rupees per ball, only if you
are better than opponent. 0 if worse, 10 if tie.
• Opponent is randomly selected—subject never
learns whom she is competing with.
Results: Across Gender, Within Society
Younger kids: No significant difference in matri,
no significant difference in patri
Older kids: Still no significant difference in matri,
significant difference emerges in patri
Girls compete less than boys (p<0.01)
Results: Within Age & Gender, Across
Society
• Younger boys in matri vs. patri: No significant
difference
• Younger girls in matri vs. patri: No significant
difference
• Older boys in matri vs. patri: Patriarchal more
competitive but only significant at 10% in a
one-sided test
• Older girls in matri vs. patri: Matri more
competitive, p=0.06
Within Society & Gender, Across Age
Approaching puberty:
No significant difference for either girls or boys in
matrilineal
Girls’ competitiveness declines in the patriarchal
society (p=0.07)
Boys’ competitiveness increases in the patriarchal
society (only significant at 10% in a one-sided
test)
Diff-in-diff regression: For boys, significant at
p=0.03, for girls, significant at p=0.08.
Are Choices Ex-Post Optimal?
• Risk-neutral subject should compete if
3Pr(win)+Pr(tie)>1
• For each performance level, randomly draw 1000
opponents and calculate the win and tie
probabilities
• Check whether choices are optimal, given the
empirical distribution of performance.
• Older girls in the patriarchal society underenter 62.5% of the time, never over-enter.
• In contrast, older girls in the matrilineal
society under-enter 28.6% of the time.
Discussion
Socialization might act along with biological forces
to create the difference around puberty
“Gender-intensification theory” in psychology (Hill
and Lynch (1983))—with puberty, more pressure
for sex-typed behavior (e.g. precarious
manhood)
An interesting question: Are the societies
biologically different as well?
Further dimensions: Self-confidence, response to
performance feedback, self-selection in other
domains.