Content uploaded by Xin Meng
Author content
All content in this area was uploaded by Xin Meng on Mar 23, 2014
Content may be subject to copyright.
1
Title: Little Emperors: Behavioral Impacts of China’s One-Child Policy
Authors: L. Cameron
1
, N. Erkal
2
, L. Gangadharan
3
, and X. Meng
4∗
Abstract: We document that China’s One-Child Policy, one of the most radical approaches to
limiting population growth, has produced significantly less trusting, less trustworthy, more risk-
averse, less competitive, more pessimistic, and less conscientious individuals. Our data were
collected from economics experiments conducted with 421 individuals born just before and just
after the One-Child Policy’s introduction in 1979. Surveys to elicit personality traits were also
used. We use the exogenous imposition of the One Child Policy to identify the causal impact of
being an only child, net of family background effects. The One-Child Policy thus has significant
ramifications for Chinese society.
One Sentence Summary: China’s One-Child Policy has produced less trusting, less trustworthy,
more risk-averse, less competitive, more pessimistic and less conscientious individuals.
Main Text: China’s One-Child Policy (OCP) restricts the number of children that urban couples
can have to one, with exceptions for those from ethnic minorities or with a severely disabled
child. The policy has given rise to a land of “little emperors” whose parents dote on them
exclusively. This has led to widespread concern within China about the social skills of this
generation and the observation that these children tend to be more self-centred and less
cooperative (1-3). This can be seen in developments such as employers including phrases like
“no single children” in job advertisements (4). In March 2007, 30 delegates in the Chinese
People’s Political Consultative Conference (CPPCC) called on the government to abolish the
1
Department of Econometrics, Monash University, Clayton, Vic., 3800, Australia.
2
Department of Economics, University of Melbourne, Vic., 3010, Australia.
3
Department of Economics, Monash University, Clayton, Vic., 3800, Australia.
4
Research School of Economics, College of Business and Economics, Australian National University, ACT 0200
Australia.
∗
Correspondence to: lisa.cameron@monash.edu.
2
policy (5). Their concerns centred on “social problems and personality disorders in young
people.”
The question of the behavioral consequences of not having siblings has been of interest in
developmental psychology for many decades. China’s One-Child Policy provides us with a
natural experiment that allows us to identify the causal impact of being a single child. Pro-social
development is believed to be shaped by parents as well as by social interactions with peers,
including siblings (6). Different relationships with parents and lack of interaction with siblings
thus have been identified in the psychology literature as two reasons why only children may
develop differently from their counterparts with siblings (7). For example, parents of only
children may be more responsive to their needs which may produce a greater sense of security,
confidence and intellectual competence (8,9). Parents of only children may also be more able to
interact with their children in ways that promote desirable development (10). More attention
from parents can however come with downsides in terms of higher expectations and pressure to
succeed in life (11). Only children are often viewed as disadvantaged as a result of “sibling
deprivation,” which may lead to their being self-centered, less cooperative, and less likely to get
along with peers (12).
In this paper, we use techniques from experimental economics to measure behavioral
differences between the pre- and post-OCP generations. Behavior in economic games has been
widely shown to be correlated with actions outside the experimental setting (13-18). We
investigate the impact of the OCP on altruism, trust, trust-worthiness, risk attitudes and
competitiveness. The OCP can be thought of as a natural experiment which enables us to
separate out the effect of being an only child from the effect of family background. In addition to
3
our experimental results, personality survey questions reveal that the OCP cohorts are also
substantially more pessimistic, less conscientious, and possibly more neurotic.
Participant Sample: We conducted experiments with participants recruited from the general
population of Beijing where the policy is strictly enforced (19). The OCP was introduced in
1979. We required participants to be born in either 1975 or 1978 (our pre-OCP cohorts) or in
1980 or 1983 (our post-OCP cohorts). Participants were also required to have both parents with
urban residency status (hukou) at the time of the participant’s birth. This sampling ensures that
those in the post-OCP cohorts were subject to the strictly enforced policy and that all the cohorts
are similar with respect to their parental hukou status. The final sample consists of 421 subjects
spread evenly across the birth cohorts 1975, 1978, 1980 and 1983, with gender balance within
these cohorts.
A post-experiment participant survey collected demographic and socio-economic
information. We test the representativeness of our sample by comparing it with the Beijing
subsample of the National Bureau of Statistics (NBS) 2009 Urban Household Survey (UHS)
data. Our sample is better educated than the general population but otherwise similar (Table S1).
We carefully control for education in the empirical specifications below.
Table 1 shows the proportion of only children in each birth cohort. This increased from
27% in the 1975 cohort to 91% for the cohort born in 1983. The average number of siblings
decreased from 0.97 to 0.12 over the same period (20). Note that the proportion of single
children had increased prior to the 1979 introduction of the policy as a result of other non-
compulsory population growth policies that were pre-cursors to the OCP (Figure S1, 21). In this
4
paper, we thus identify the behavioral impact of the OCP relative to the non-compulsory fertility
policies implemented prior to the OCP.
Although in later years the policy undoubtedly reduced the size of extended families,
extended family size differs only slightly between the pre- and post-OCP cohorts we study. The
number of cousins declines slightly from an average of 7.4 to 7.0 cousins (Table 1). We are thus
identifying the impact of growing up as an only child, not the impact of having a smaller
extended family.
Measurement of Behavior - Experiments: The experiments we conduct are standard games
from the economics literature. The Dictator Game (22) is designed to elicit the extent of altruism
amongst participants. The Trust Game
In the
elicits the extent to which participants are able to trust one
another and the extent of their trust-worthiness (23). These games are explained in the
supplementary materials.
Risk Game
In the
(24), each participant is given 100 yuan (which is approximately
US$15) and the choice to put any amount between 0 and 100 yuan into an ‘investment”, which
yields triple the amount invested with 50% probability and 0 with 50% probability. More risk-
averse participants will invest less in the risky option. The outcome of the investment is decided
by the flip of a coin at the completion of the session.
Competition Game, participants are asked to add up as many sets of five two-digit
numbers as possible in five minutes (25). The numbers are randomly generated and presented in
rows. Participants write the total in a blank box provided at the end of the row. Calculators
cannot be used, but scrap paper is provided for hand-written calculations.
5
Participants are asked to choose between two different payment schemes. Option 1 is a
piece-rate which pays 5 yuan for every sum correctly completed. In Option 2, payments are
determined in a competitive way. Each participant is randomly and anonymously paired with
someone else in the room. S/he is paid 10 yuan for every sum correctly completed if s/he
completes more sums correctly than the person with whom s/he is paired, 5 yuan if both
participants complete the same amount of correct sums, and 0 yuan if s/he loses the competition.
Results: Impact of Being Born under the OCP Figure 1 presents the unconditional differences
in behavior between participants born before and after the OCP, for each of the games. The
underlying data are presented in Table S2 in the supplementary materials. Those born under the
OCP share slightly less of the endowment in the dictator game with the other player (40.1% of
the endowment on average as compared to 43.4% sent by those born before the policy’s
introduction). The t-test of difference in means is statistically insignificant (p=0.11). OCP
participants on average were less trusting, sending less to the other player (46.1% versus 50.6%)
and returning less than those not born under the policy (30.4% versus 35.4%). Both of these
differences are statistically significant.
OCP participants invested significantly less in the riskier investment (58.1% versus
66.4%). In the competition game, many fewer OCP participants chose to compete than those
born before the policy (44.2% versus 51.8%). This difference is substantial but not statistically
significant (p=0.12).
Differences in competitiveness may reflect beliefs in one’s ability. Participants were
asked in which performance quintile they expected themselves to be relative to others in the
room. There is no significant difference in this self-reported ranking between the pre- and post-
6
OCP cohorts although participants born under the OCP completed significantly more sums
correctly than their counterparts (Table S2). This is consistent with the findings in the literature
that only children perform better academically (26).
The effects discussed above are simple mean differences. These could be due to
differences in the demographic backgrounds of the pre- and post-OCP samples. To examine the
OCP effects net of these observable characteristics, we estimate an econometric model where we
control for participants’ gender and education, maternal education, and whether the individual
was born in Beijing, in addition to the main variable of interest, an indicator of whether the
participant was born under the OCP. The summary statistics for the control variables are
presented in Table S3. Table 2 reports the estimation results. We find that controlling for the
demographic and family background variables, the unconditional effects we observed in Figure 1
above persist in terms of signs, magnitudes, and significance levels. The only difference from the
unconditional estimates is that the regression results indicate that the lesser propensity for those
who were born after the OCP to choose to compete becomes statistically significant once one
controls for demographic characteristics.
The OCP indicator is defined according to birth cohort and hence is highly correlated
with age. There is no evidence that age has a strong systematic effect on behavior in experiments
of the type we conduct, particularly over the relatively small age range in our sample (24, 27-32).
Age is not a determinant of behavior within the pre- and post-OCP cohorts (Table S8). Our
results are also largely unaffected when we limit the potential for age effects by focusing on just
the 1978 and 1980 cohorts (Table S9). We thus conclude that age effects are not driving our
results. Consistent with this result, marital status, being a parent, age of parents, and people
7
potentially becoming more capitalistic over time do not seem to explain our results either (Tables
S10 and S11).
Identifying the Causal Impact of Growing up Without Siblings Although the main effect of
the OCP is to grow up as a single-child, the coefficients we estimate in Table 2 are not estimates
of the effect of being a single child as being a single child is not a unique characteristic of the
after cohort. Many before cohort individuals are also single children and some children born
after the policy are not single children (see Table 1). The coefficient on the OCP indicator is
thus the average behavioral effect of the OCP across the population.
If being a single child were exogenous, then the effect of growing up without siblings
would be estimated consistently from:
Y
i
= α + βX
i
+ φSingle
i
+ υ
i
(1)
where Y
i
is the behavior of interest, Single
i
is an indicator for being an only child, α, β, and φ are
coefficients to be estimated and ν
i
is a random error term. The coefficient of interest would be φ.
Being a single child in the pre-policy cohort was however largely a choice of parents.
Thus, the coefficient on Single
i
in Equation (1) would pick up not only the effect of being a
single child but also the effect of any omitted family background variables which influence the
probability of being an only child. This is a problem if the unobserved parental characteristics
which make parents choose to have one child also affect individuals’ behavioral outcomes via
genetic or “nurture” channels, which is likely. The variable Single
i
is thus endogenous.
8
We can however exploit the exogeneity of the imposition of the OCP and use the
indicator of whether one was born under the policy, D
i
, to instrument for the endogenous
variable Single
i
. We can thus identify the causal effect of growing up without siblings as a result
of the OCP. That is, we estimate the equations below using an instrumental variables (IV)
approach. Equation (2b) is the first-stage regression.
Y
i
= α + βX
i
+ φSingle
i
+ υ
i
(2a)
Single
i
= η + δX
i
+ κD
i
+ u
i
(2b)
where η, δ, and κ are coefficients to be estimated, u
i
is a random error term and the other terms
are defined as above.
Provided the instrument satisfies a few assumptions (see section 2.1 in the
supplementary materials), the IV estimate of φ can be interpreted as the Local Average
Treatment Effect (LATE) of growing up as only children because of the OCP (33).
The IV results are reported in Table 3. In every case where we identified a difference in
behavior between the pre- and post-OCP cohorts, being an only child as a result of the OCP is
also shown to have a significant causal impact on behavior. The coefficients on being an only
child are over double that of the coefficients on the indicator of whether one was born under the
OCP reported in Table 2. Individuals who are only children as a result of the policy sent on
average 16 percentage points less of the endowment to the other player and returned 11
percentage points less of what they received in the trust game; invested 19 percent less of the
endowment in the risky option in the risk game; and were 20 percentage points less likely to
choose the competitive option in the competition game. In terms of standard deviations of the
dependent variable, being a single child is associated with a decrease of 0.58 standard deviations
9
of the percentage sent in the trust game, 0.44 standard deviations of the percentage returned in
the trust game, 0.75 standard deviations of the percentage invested in the risky option, and 0.41
standard deviations of the probability of competing. These are thus sizeable effects.
Personality Traits: In the post-experiment survey, participants were asked a number of
questions that seek to assess their personality type and outlook on life.
Participants were asked “What do you think are the chances that it will be sunny
tomorrow? Please write a number from 0 to 100, where 0 means ‘absolutely no chance’ and 100
means ‘absolutely certain’. Responses to this question are widely used as an indicator of
optimism (34). We find that those born under the OCP and those who grew up as single children
as a result of the OCP are significantly less optimistic than others (Table S15).
The post-experiment survey also implemented the “Big Five Inventory” (BFI), which
contains 44 questions designed to categorize people in terms of openness (inventive/curious
versus consistent/cautious); conscientiousness; extraversion; agreeableness; and neuroticism
(sensitive/nervous versus secure/confident). The five broad factors together encompass most
known personality traits and the 44 questions are used to construct scores for each of these traits
(35). We examine to what extent these traits are determined by the OCP and single-child status.
Being born under the OCP and growing up as a single-child as a result of the OCP are associated
with a higher neuroticism score and a lower conscientiousness score (Table S15).
These effects are not small: 0.69 standard deviations of our optimism indicator, 0.52
standard deviations of the conscientiousness score and 0.71 standard deviations of the
neuroticism score. The result for neuroticism is consistent with the finding that positive sibling
relationships moderate the relationship between stressful life events and internalizing behaviors
10
(36). The finding that conscientiousness is lower is counter to the argument in the empirical
literature in psychology that single children have a greater motivation to achieve, but it is
consistent with Chinese parents’ views of their only children (37).
Like the experimental results, our results for conscientiousness and optimism are
qualitatively similar when estimated using just the 1978 and 1980 cohorts, and age is not a
significant determinant within pre- and post-OCP periods. The results for neuroticism are less
robust to checks for age effects (Table S16).
Implications: Previous research has shown that non-cognitive attributes like conscientiousness,
neuroticism, and optimism are important determinants of educational attainment, labor market
outcomes, health and marriage and divorce (38-40). Pro-social behavior is consistently seen to
be an important determinant of social capital and plays a role in institutional development (41).
A willingness to take risks is an important component of entrepreneurship (17). Our data show
that being an only child as a result of the OCP is associated with taking less risk in the labor
market (Table S19).
While our findings are obtained from a comparison of cohorts in Beijing born directly
around the time of the policy’s introduction, our results are generalizable to other urban areas of
China where the OCP was strictly implemented. Previous work suggests that differences
between only children and others in Beijing is similar to that in other urban areas (26). The effect
of the policy on the behavior of people born long after the policy’s introduction may however
differ from what we find here as later cohorts will have grown up with very limited extended
family and in a society dominated by only children. Under such circumstances, we would expect
that the policy’s effect would, if anything, be magnified.
11
References and Notes:
1. L. Lee in Child Care in Context: Cross Cultural Perspectives, M. E. Lamb, K. Sternberg,
Eds. (Lawrence Erlbaum, Hillsdale, NJ, 1992), pp. 355-392.
2. C. Fan, C. Wan, G. Lin and Q. Jin. A comparative study of personality characteristics
between only and nononly children in primary schools in Xian. Psychological Science,
China Academic Journal Electronic Publishing House, 17(2), 70-74 (in Chinese) (1994).
3. Q. Wang, M. D. Leichtman, S.H. White, Childhood memory and self-description in young
Chinese adults: the impact of growing up an only child. Cognition, 69, 73-103 (1998).
4. L. Chang, Factory Girls: From Village to City in a Changing China (Spiegel and Grau, New
York 2008).
5. "Consultative Conference: “The government must end the one-child rule”". AsiaNews.it.
2007-03-16. http://www.asianews.it/index.php?l=en&art=8757&size=A.
6. J. Piaget, The Moral Judgment of the Child (Free Press Paperbacks, New York, 1997).
7. L. Levy-Garboua, C. Meidinger, B. Rapoport in Handbook of the Economics of Giving,
Altruism and Reciprocity, S.C. Kolm, J. Mercier-Ythier Eds. (Elsevier B.V., 2006), Vol. I,
pp. 545-613.
8. J. Bowlby, Attachment and Loss (Basic Books, New York, 1982). Vol. 1, 2
nd
edition.
9. J. Blake, Family size and the quality of children. Demography, 18, 421–442, (1981).
10. M. Lewis, M., C. Feiring in Families as Learning Environment for Children, L. M. Laosa, I.
E. Sigel Eds. (Plenum: New York, 1982), pp. 115-145.
12
11. L. Roberts, P. Blanton, I always knew mom and dad loved me best. Journal of Individual
Psychology, 57, 125-140 (2001).
12. J. Blake, The Only Child in America: Prejudice versus Performance. Population and
Development Review,7(1), 43-54, (1981).
13. E. Fehr, A. Leibbrandt, A Field Study on Cooperativeness and Impatience in the Tragedy of
the Commons. Journal of Public Economics, 9-10, 1144-1155 (2011).
14. E. Liu, Time to change what to sow: risk preferences and technology adoption decisions of
cotton farmers in China. Princeton University Industrial Relations Section Working Paper
No 526. (2008).
15. E. Liu, J. Huang, “Risk preferences and pesticide use by cotton farmers in China,”
University of Houston Working Paper, (available at
http://www.class.uh.edu/Faculty/emliu/index.html).
16. D. Karlan, Using experimental economics to measure social capital and predict financial
decisions. American Economic Review, 95(5), 1688-1699 (2005).
17. M. Benz, S. Meier, Do people behave in experiments as in the field? – evidence from
donations. Experimental Economics, 11, 268-281 (2008).
18. M. Castillo, R. Petrie, M. Torero, On the preferences of principals and agents. Economic
Inquiry, 48(2), 266-273 (2010).
19. Materials and methods are available in the supplementary materials.
20. Some of the increase in the proportion of children without siblings in the 1978 cohort is due
to parents not having had enough time to conceive a second child by the time of the
introduction of the OCP in mid-1979. This issue is discussed in detail in the supplementary
materials.
13
21. W. Lavely, R. Freedman, The Origins of the Chinese fertility decline. Demography, 27(3),
pp357-367 (1990).
22. R. Forsythe, J. Horowitz, N. Savin, M. Sefton, Fairness in simple bargaining games. Games
and Economic Behavior, 6, 347-369 (1994).
23. J. Berg, J. Dickhaut, K. McCabe, Trust, reciprocity, and social history. Games and Economic
Behavior, 10, 122-142 (1995).
24. U. Gneezy, K. L. Leonard, J. A. List, Gender differences in competition: evidence from a
matrilineal and a patriarchal society. Econometrica, 77(5), 1637-64 (2009).
25. M. Niederle, L. Vesterlund, Do women shy away from competition? Do men compete too
much? Quarterly Journal of Economics, 1067-1101 (2007).
26. T. Falbo, D. Poston, The academic, personality, and physical outcomes of only children in China.
Child Development, 64(1), 18-35 (1993).
27. L. Schechter, Traditional trust measurement and the risk confound: an experiment in rural
Paraguay. Journal of Economic Behavior and Organization, 62(2), 272-292 (2007).
28. M. Sutter, M. Kocher, Trust and trustworthiness across different age groups. Games and
Economic Behavior, 59, 364-382 (2007).
29. G. Charness, M. C. Villeval, Cooperation, competition and risk attitudes: an
intergenerational field and laboratory experiment. American Economic Review, 99(3), 956-
978 (2009).
30. S. Andersen, G. Harrison, M. Lau, E. Rutstrom, Preference Heterogeneity in Experiments;
Comparing the Field and Lab. Journal of Economic Behavior and Organization, 73(2), 209-
224 (2010).
14
31. G. Harrison, S. Humphrey, A. Verschoor, Choice under uncertainty: evidence from Ethiopia,
India and Uganda. Economic Journal, 120,80-104 (2010).
32. C. Engel, Dictator games: a meta study. Experimental Economics, 14(4), 583-610 (2011).
33. G. Imbens, J. Angrist, Identification and estimation of local average treatment effects
Econometrica, 62(2), March, 467-475 (1994).
34. See, for example, the U.S. Health and Retirement Survey.
35. For details of the questions used and how the index is calculated,
see http://www.ocf.berkeley.edu/~johnlab/bfi.htm.
36. K. Gass, J. Jenkins, J. Dunn. Are sibling relationships protective? A longitudinal study.
Journal of Child Psychology and Psychiatry, 48(2), 167-175 (2007).
37. Y. Zhang, G. Geldolph, P. C. Cheung, S. Lau, A new look at the old “little emperor”:
developmental changes in the personality of only children in China. Social Science
and Personality, 29(7), 725-732, (2001).
38. J. Heckman, J. Stixrud, S. Urzua, The effects of cognitive and noncognitive abilities on labor
market outcomes and social behavior. Journal of Labor Economics, 24(3), 411-482 (2006).
39. L. Borghans, A. Duckworth, J. Heckman, B. ter Weel, The economics and psychology of
personality traits. Journal of Human Resources, 43(4): 972-1059, (2008).
40. S. Lundberg, Psychology and family economics. Perspektiven der Wirtschaftspolitik, 12, 66-
81 (2011).
41. E. Ostrom, J. Walker. Trust and Reciprocity: Interdisciplinary Lessons for
Experimental Research. (Russell Sage Foundation, 2003).
15
S1. D. Downey, D. Condron, Playing well with others in kindergarten: the benefit of siblings at
home. Journal of Marriage and Family, 66 (2), 333-350, (2004).
S2. K. Kitzmann, R. Cohen, R. Lockwood, Are only children missing out? Comparison of the
peer-related social competence of only children and siblings. Journal of Social and Personal
Relationships, 19(3), 299-316, (June, 2002).
S3. T. Dohmen, A. Falk, D. Huffman, U. Sunde, The intergenerational transmission of risk and
trust attitudes. Review of Economic Studies, 79(2), 645-677, (2012).
S4. T. Falbo, D. F. Polit., Quantitative review of the only child literature: research evidence and
theory development. Psychological Bulletin, 100, 176–189, (1986).
S5. D. F. Polit, T. Falbo, Only children and personality development: a quantitative review.
Journal of Marriage and the Family, 49, 309–325, (1987).
S6. R. X. Liu, W. Lin, Z. Chen, School performance, peer association, psychological and
behavioral adjustments: a comparison between Chinese adolescents with and without
siblings. Journal of Adolescence, 33, 411-417, (2010).
S7. A. Mancillas. Challenging the stereotypes about only children: a review of the literature and
implications for practice. Journal of Counseling and Development, 84, 268–275. 2006
S8. J. Chen and L. T. Goldsmith, Social and behavioral characteristics of Chinese only children:
a review of research. Journal of Research in Childhood Education, 5(2), 127-139, (1991).
S9. J. Shen, B. J. Yuan. Moral values of only and sibling children in mainland China. Journal of
Psychology: Interdisciplinary and Applied, 1331), 115-124, (Jan, 1999).
S10. D. Wang, N. Kato, Y. Inaba, T. Tango, Y. Yoshida, Y. Kusaka, Y. Deguchi, F. Tomita, Q.
Zhang, Physical and personality traits of preschool traits in Fuzhou, China: Only children vs
sibling. Child: Care, Health and Development, 26(1), 49-60, (2000).
16
S11. C. Liu, T. Muakata, F. Onuoha, Mental health condition of the only-child: a study of urban
and rural high school students in China. Adolescence, 40, 831-45, (2005).
S12. D. Davis, C. Holt, Experimental Economics (Princeton University Press: Princeton, NJ,
1993), pp xi+571.
S13. J. Banister, China’s Changing Population (Stanford Univ Press: Palo Alto, CA, 1987).
S14. K. Ying, The birth control movement on the Chinese mainland. Issues and Studies, 10(5),
80-9, (1974).
S15. X. Peng, China’s demographic history and future challenges. Science, 333, 581-587, (29
July, 2011).
S16. Center for Population Studies, Chinese Academy of Social Sciences & Editorial Board of
China's Population Yearbook, "The major events of China's population activities," in China's
Population Yearbook 1985, Chinese Social Sciences Press: Beijing, 1986), pp. 1263-1288.
S17. X. Peng, Demographic Transition in China (Clarendon: Oxford, 1991).
S18. G. Feeney, F. Wang, Parity progression and birth intervals in China: the influence of policy
in hastening fertility decline. Population and Development Review 19 (1), 61–101, (1993).
S19. Y. Cai, China’s below-replacement fertility: government policy or socioeconomic
development? Population and Development Review 36(3), 419-440, (2010).
S20. A. Ebenstein, The “missing girls” of China and the unintended consequences of the one
child policy. Journal of Human Resources, 45(1), 87-115, (2010).
S21. F. Wang, The future of a demographic overachiever: Long-term implications of the
demographic transition in China. Population and Development Review, 37(supplement),
173-190, (2011).
17
S22. J. Zhang, R. Sturm, When do couples sign the one-child certificate in urban China?",
Population Research and Policy Review, 13 (1), 69-81, (1994).
S23. See http://rse.anu.edu.au/rumici/.
S24. P. Kane C. Choi, China’s one child family policy. British Medical Journal, 319(7215), 992-
994, (1999).
S25. N. Frohlich, J. Oppenheimer, J. B. Moore. Some doubts about measuring self-interest using
dictator experiments: the costs of anonymity,” Journal of Economic Behavior &
Organization, 46, 271-90, (2001).
S26. L. Cameron, N. Erkal, L. Gangadharan, M. Zhang, Cultural integration: experimental
evidence of the evolution of immigrant’s preferences”, working paper, University of
Melbourne, (2011). (available
at: http://www.economics.unimelb.edu.au/staff/nisvan/research.htm.)
S27. S. Burks, J. Carpenter, E.Verhoogen, Playing both roles in the trust game. Journal of
Economic Behavior and Organization, 51(2), 195-216 (2003).
S28. Y. Tang, Choices: the Chinese reform since 1978 (in Chinese), (Economics Daily
Publishing House: Beijing, 1998).
S29. S. Johansson, O. Nygren, The missing girls of China: A new demographic account.
Population and Development Review, 17(1), Mar, 35-51, (1991).
S30. P. Costa, R. McRae, Personality in Adulthood: A Six-Year Longitudinal Study of Self-
Reports and Spouse Ratings on the NEO Personality Inventory. Journal of Personality and
Social Psychology, 54(5), 853-863 (1988).
S31. D. Cobb-Clark, S. Schurer, The stability of Big-Five personality traits,” Economics Letters,
115(1), 11-15 (2011).
18
S32. B. Roberts, K. Walton, W. Viechtbauer, Patterns of mean-level change in personality traits
across the life-course: a meta-analysis of longitudinal studies. Psychological Bulletin, 132(1),
1-25, (2006).
Acknowledgements: We gratefully acknowledge funding from the Australian Research Council.
The data used in this paper will be made available from an approved database. The Rumici data
is available at rse.anu.edu.au/rumici. All authors contributed equally to this paper.
19
Figure 1: Behavioral Consequences of the One Child Policy. Mean differences in
behavior between participants born before and after the One Child Policy. The p-values
reported in parentheses are from t- tests of differences in means. Error bars are mean+/-SEM.
0
10
20
30
40
50
60
70
80
90
100
%SentinDictatorGame %SentinTrustGame %returnedinTrustGame %investedinriskyoption %choosingtocompete
BornBeforeOneChildPolicy BornAfterOneChildPolicy
Altruism
(p=0.11)
Trust
(p=0.09)
Trustworthiness
(p=0.02)
Risk‐Taking
(p=0.001)
Competitiveness
(p=0.12)
20
Table 1: Demographic Composition by Cohort.
We present means of variables by birth
cohort, and pre- and post-One Child Policy.
a
Column (2) reports means for those born in 1978 but
excludes those born in 1978 who report they are an only child as a result of the policy.
(1)
(2)
(3)
(4)
(5)
(6)
(7)
1975
1978
a
1978
1980
1983
Pre-
OCP
Post-
OCP
% only children
27.0
23.7
60.9
81.8
90.7
45.1
86.4
% with one sibling
56.0
62.7
32.2
17.2
5.6
43.3
11.2
% with two siblings
10.0
8.5
4.4
1.0
1.9
7.0
1.5
% with three sibling
4.0
3.4
1.7
0.0
0.9
2.8
0.5
% with four siblings
2.0
1.7
0.9
0.0
0.0
1.4
0
Average # of siblings
0.97
0.97
0.50
0.19
0.12
0.72
0.16
% being the first born
52.0
35.6
67.0
85.9
93.5
60.0
89.8
Average # of cousins
6.93
7.95
7.78
7.14
6.90
7.4
7.0
Average # aunts and uncles
5.15
5.42
5.81
5.91
6.27
5.5
6.1
Age (years)
34.7
31.8
31.7
29.7
26.7
33.1
28.2
N
100
59
115
99
107
215
206
21
Table 2: Estimation results.
We estimate
Y
i
=
α
+
βX
i
+
γD
i
+
ε
i
where Y
i
is a behavioral
outcome observed in the experiments, and X
i
is a vector of control variables which includes participants’
gender and education, maternal education, and whether the individual was born in Beijing. The main
variable of interest is D
i
, which equals 1 if an individual is born after the introduction of the OCP and zero
otherwise. The coefficient γ identifies any differences between those born before and after the policy
and is our estimate of the behavioral impact of the OCP. Columns (1) –(4) report coefficients from tobit
estimation with lower censoring at 0 and upper censoring at 100. Column (5) presents marginal effects
from a probit on whether to compete or not. Robust standard errors are shown in square brackets. *, **,
*** indicate statistical significance at the 10%, 5% and 1% levels respectively.
Altruism
Trust & Trustworthiness
Risk-Taking
Competitiveness
Dictator Game
Trust Game
Risk Game
Competition
Game
% Sent
% Sent
% Returned
% Invested
Choose to
Compete
(1)
(2)
(3)
(4)
(5)
One-Child policy
-2.834
-6.980**
-5.024*
-7.987**
-0.090*
[2.275]
[3.116]
[2.728]
[3.213]
[0.050]
Male
-0.646
10.73***
0.694
3.086
0.072
[2.299]
[3.120]
[2.687]
[3.152]
[0.049]
University or above
-14.37***
1.496
-15.34**
-14.030*
0.149
[5.484]
[6.407]
[6.314]
[8.111]
[0.115]
3-year college
-13.28**
-3.989
-15.53**
-5.323
0.12
[5.538]
[6.429]
[6.290]
[8.143]
[0.118]
Born in Beijing
-5.319*
-11.02**
4.788
5.391
0.082
[2.977]
[4.656]
[4.485]
[4.096]
[0.075]
Mother with 3-year college
0.518
8.887*
4.594
8.123*
-0.009
[3.784]
[4.669]
[4.253]
[4.220]
[0.067]
Mother with university or above
-4.363
6.983
-2.612
5.004
0.091
[5.320]
[5.998]
[5.439]
[5.406]
[0.100]
Observations
410
408
398
416
419
22
Table 3: Causal Impact of Being an Only Child.
We present results from instrumental variables estimation. We use
an indicator of being born under the One Child Policy as an instrument for being an only child. Columns (1)-(4) present the results of IV
Tobit estimation. Column (5) presents marginal effects from IV Probit estimation.
Robust standard errors are shown in square
brackets. *, **, *** indicate statistical significance at the 10%, 5% and 1% levels respectively.
Altruism
Trust Trustworthiness
Risk-Taking
Competitiveness
Dependent Variables:
(Dictator Game
% sent)
(Trust Game
% sent)
(Trust Game
% returned)
(Risk Game
% invested)
(Competition Game
– competing.)
(1)
(2)
(3)
(4)
(5)
Only-child
-6.743
-15.998**
-11.34*
-19.002**
-0.203*
[5.454]
[7.219]
[6.199]
[7.920]
[0.108]
Male
-0.829
10.604***
0.584
2.641
0.067
[2.307]
[3.153]
[2.693]
[3.280]
[0.049]
University or above
-13.742**
3.064
-14.19**
-11.998
0.166
[5.657]
[6.593]
[6.680]
[8.788]
[0.114]
3-year college
-12.573**
-2.323
-14.49**
-2.93
0.138
[5.692]
[6.571]
[6.657]
[8.807]
[0.118]
Born in Beijing
-3.054
-5.914
8.442*
11.528**
0.142*
[3.486]
[4.981]
[4.606]
[4.797]
[0.075]
Mother with 3-year college
1.238
10.396**
5.642
10.045**
0.011
[3.837]
[4.721]
[4.251]
[4.605]
[0.069]
Mother with uni or above
-3.418
8.522
-1.571
7.361
0.109
[5.440]
[6.038]
[5.554]
[5.456]
[0.099]
Observations
410
408
398
416
419