Beijing females growing up during the communist regime are more competitively inclined than males, females growing up during the market regime, & Taipei females

Gender Differences in Willingness to Compete: The Role of Culture and Institutions. Alison Booth, Elliott Fan, Xin Meng, Dandan Zhang. The Economic Journal, February 2 2018. https://doi.org/10.1111/ecoj.12583

Abstract: Our Beijing‐based laboratory experiment investigated gender differences in competitive choices across different birth‐cohorts experiencing – during their crucial developmental‐age – different institutions and social norms. To control for general time trends, we use Taipei counterpart subjects with identical original Confucian traditions. Our findings confirm that exposure to different institutions/norms during crucial developmental‐ages significantly changes individuals’ behaviour. In particular, Beijing females growing up during the communist regime are more competitively inclined than their male counterparts; their female counterparts growing up during the market regime; and Taipei females. For Taipei, there are no statistically significant cohort or gender differences in willingness to compete.

4 Mechanisms

Alesina and Fuchs-Schundeln (2007) discussed the endogenous role political regimes play
in forming people’s tastes for public social policies. They provide strong evidence that
there is a feedback effect from the regime on people’s attitudes. In our experiment, the
finding that women in the 1958 Beijing cohort are more willing to compete may well be
related to this feedback effect. In this section we further explore the mechanisms through
which women growing up under different regimes form different preferences.

To examine if there is a long lasting communist indoctrination effect, we follow Alesina
and Fuchs-Schundeln (2007) and check what other evidence there is in our data that
indicate attitudes towards government intervention in social and economic affairs differs
across cohorts and regions. In our exit-survey, we asked subjects if they believe that
government should play a role in reducing income inequality in general and if they support
the view that the less intervention from the government to the economy the better. The
answers are given on a scale of 1 to 5, with 1 being strongly disagree and 5 being strongly
agree. Using these data as the dependent variables we estimate using OLS a modified
version of Equation 2 without the game controls and added in log individual income into
the vector Xij . The selected results are reported in the first two columns of Table 7, where
we report three panels: the top panel reports the selected estimated coefficients from the
regressions of modified Equation 2, the middle panel reports the implied differences across
cohorts and regions based on the top panel coefficients, and the bottom panel reports
differences between Beijing women/men and Taipei women/men based on results from
the regressions that look only at aggregated differences and disregard cohort variations.

The results for the state-intervention preference variables suggest that, in general, the
Mainland Chinese female cohorts, relative to the Taiwanese female cohorts, are much more
likely to support the view that the government should try to reduce income inequality and
much less likely to think that the less government intervention in the economy the better.
On a scale of 1 to 5 points, the three female cohorts are 1.4 to 1.9 points more likely
to support government reducing inequality and 1.6 to 1.8 points less likely to think that
less government intervention in the economy the better. These are very large differences,
between 28% to 38% of the total scores. These results are largely consistent across the
three female cohorts.

If we examine the bottom panel, which treats male and female subjects in Beijing
and Taipei as single groups, we see that this conclusion applies to males as well. Thus,
after 40 years of market oriented economic reform, the society is still largely geared towards equality and government intervention. This finding is in line with that of Alesina
and Fuchs-Schundeln (2007) – that communist propaganda has a long lasting effect. In
addition, our results indicate that individuals who grew up subject to the heavy dose of
indoctrination, and individuals who largely grew up with new ideology, all seem to prefer
more state intervention in social and economic affairs. This latter finding differs slightly
from that of Alesina and Fuchs-Schundeln (2007) which revealed that in East Germany
those born after 1975 have a much weaker preference for government intervention relative
to their older counterparts. This difference between their results and ours may relate to
two broad sets of factors. First, after German reunification Communism was discredited
and there was a large migration from East to West Germany. Second, in mainland China
the Communist Party is still the ruling party, even though the economy largely operates
under market economic rules. One other interesting issue related to our finding is the
following issue. According to our results, only the behaviour of the cohort which grew
up during the communist regime is in line with the communist indoctrination, while the
behaviour has changed for cohorts growing up in the economic reform era. Yet the attitudes difference across cohorts seems to be limited. This may be related to the lack of
continuation in propaganda on gender issues which affect gender willingness to compete as
we will discuss below,13 whereas government role in the economy persisted in the reform
era.

Is the difference in gender willingness to compete and in the preference for government
intervention for social and economic issues really a reflection of indoctrination? There is
a small literature that discusses intergenerational transmission of cultural norms (see for
example, Bisin and Verdier, 2000; Tabellini, 2010; and Nunn and Wantchekon, 2011).14 In
the post-experiment survey, we follow their ideas and ask respondents to pick, from a list
of qualities or attributes, those that their parents and school encouraged when they were
young. We report here the results from our examination of three attributes that are related
to the competitive inclination and preference for inequality in general: 1. Encourage to
believe that men and women are equal; 2. Encourage to be competitive; and 3. Encourage
to be unselfish. We generate two dummy variables for each of these qualities: whether
mother or school encouraged individuals to have these qualities (yes=1, 0 otherwise). The
modified version of Equation 2 is estimated using a linear probability model (LPM) for
ease of interpretation of the coefficients. The results for mothers’ encouragement and
schools’ encouragement are reported in columns [3] to [8] of Table 7. The three panels
separately report the regression coefficients; the implied differences across gender, cohorts
and region; and the aggregated differences between Beijing and Taipei for females and
males.

The results on gender equality (both from mothers’ and schools’ teaching) show that
there is no consistent gender difference within each Beijing cohort. Relative to the 1958
Beijing females, those who were born later are consistently less likely to have either
mother or school encourage them to believe in gender equality. Although the size of the
differences are large (9% to 13%), none of the differences are precisely estimated. Relative
to Taipei women, however, Beijing women in all three cohorts are more likely to have their
parents or school to encourage them to have the view of gender equality. The difference
in probability ranges from 70% to 74% for mothers’ encouragement and 59% to 82% for
schools’ encouragement. This is not only the case for females: the bottom panel shows
that Beijing males were equally more likely to have their mothers and schools encourage
them to adopt the view of gender equality than their Taipei counterparts.

Regarding being competitive, mainland parents do not seem to have prepared their
children any more than their Taiwan counterparts. However, schools in Beijing did,
but this seems to be more so for the later cohorts than for the 1958 cohort. Mainland
parents and schools are more likely to teach their children to be unselfish than their
Taiwan counterparts—further evidence of the communist indoctrination. This is true
for all female and male cohorts. It is also true that the probability is higher for the
1958 Beijing cohort than that for the later Beijing cohort, but the differences are not
statistically significant.

How did propaganda affect individuals’ behaviour and preferences 40 years after the
change of the regime? Psychologists have long been discussing how social norms about
sex roles may affect children’s personality characteristics and behavioural competencies
to prepare them to fulfil the societal expectations so that they can perform those roles
(see, for example, Horner, 1972; Fitzgerald and Betz, 1983).
Our post-experiment survey implemented the “Big Five inventory” (BFI), which consists of 44 questions designed to elicit individual’s personality traits. The psychological
literature has identified the overlap between extroversion and competitiveness (Hogan,
1986; Digman, 1990; Chen et al., 2011). Recent empirical studies on online game players
also identified that individuals who have higher scores on openness, extroversion, and
conscientiousness are more likely to be players (Teng, 2008). We examine across gendercohort-region differences in the ‘Big Five’ personality by estimating Equation 2 without
controlling for Wij and Xij .

We find that females in all Beijing cohorts are more extroverted than their male counterparts, though only the 1977 cohort exhibit statistically significant difference. More importantly, all three cohorts Beijing women are statistically significantly more extroverted than their Taipei counterparts. The aggregated estimation (the bottom panel of the table) also confirms that Beijing females are more extroverted than Taipei males and
females and they are more extroverted than Beijing males. However, it does not seem to
be the case that there are significant cohort difference in extroversion. If anything, the
1958 cohort seems to be slightly less extroverted than their younger counterparts but the
differences are not statistically significant.

With regard to openness, for the 1958 and 1977 birth cohorts we observed statistically
significant gender differences within the Beijing sample. Women are more open than
men. Beijing women are also more open than Taipei women in every cohort, and the
size of the difference is the largest for the 1958 cohort. Further, at the aggregated level,
Beijing women have higher openness scores than their Beijing male, Taipei male, and
Taipei female counterparts. Once again, we fail to detect across-cohort variations among
Beijing females which may help to shed light on why the 1958 Beijing women are more

For agreeableness, neuroticism, and conscientiousness we find similar but less statistically significant patterns. The Beijing women seem to be more agreeable, less neurotic, and more conscientious than Beijing males, Taipei females and Taipei males.

These results provide some weak evidences that perhaps indoctrination at young age could affect an individual’s personality and subsequently affect individuals’ behaviour. To this end, more research is needed.