A Cross-Country Investigation of the Determinants

of the Happiness Gender Gap
Sabrina Vieira Lima∗
June 6, 2011

Abstract
This paper investigates the existence and determinants of the happiness
gender gap across countries, using a large sample of individuals of around 80
countries, built with data from World Values Survey, World Development Indicators and CIRI Human Rights. We examine the role of rights, achievement
and beliefs for female gender, in economical, political and social context. We
find that women are happier than men in most African and many developing
countries, and less happy in around 15 European and other industrialized countries. Overall, the results indicate that women happiness have a paradoxical
component, where better objective conditions do not grant them happiness.
Keywords: Subjective well-being, happiness gender gap, women’s rights.

PhD student at University of Milan-Bicocca, Economics Department, Piazza dell’Ateneo Nuovo
1, 20126 Milan, Italy. Telephone: #39 2 6448 3052. E-mail: s.vieiralima@campus.unimib.it

1

1

Introduction

In the last two decades, major developments of the neurosciences provided valuable
information about differences on male and female brain. This evidence yields new
biological support for the conventional belief that men and women are different, both
for cognitive processes and behavioural outcomes. This also shapes different paths
of social interactions, and, as such, the entire social process might also result in a
different gender happiness path.
On the literature of Economics and Happiness, gender is considered generally as
a socio-demographic control variable. Few authors have given a bit closer attention
to this topic along the way (Easterlin, 2003; Blanchflower and Oswald, 2004; Marcelli
and Easterlin, 2005; Bjrnskov et al., 2007; Plagnol and Easterlin, 2008; Stevenson
and Wolfers, 2008, 2009; Guven et al, 2009), while the latest contributions focused
mostly in the work and leisure spheres1 (see, for example, Berger, 2009; Gash et al.,
2010; Gimenez-Nadal and Sevilla-Sanz, 2010; Trzcinski and Holst, 2010). To our
knowledge, little has been done to frame the study in more encompassing terms, or
in a worldwide perspective. A first contribution of our paper is an attempt to fill
this gap of the literature.
Moreover, evidences so far show that on average the two genders do not share
the same subjective levels of happiness; but there is no consensus on who is the
happiest (see, for example, Bjrnskov et al., 2007; Blanchflower and Oswald, 2004).
Nonetheless, the works of Stevenson and Wolfers2 (2008, 2009) and Plagnol and
Easterlin (2008), respectively, indicate shifts in the happiness gender gap: from
favouring women to favouring men, or disfavouring women later in life. Following
this state of art, we do not form a specific hypothesis on who is the happiest. However,
it was women’s life that historically suffered major changes in economic, political and
social spheres. Therefore we want to investigate these aspects. They are noticeable
instances where women conquered a right that previously were exclusively of male
domain (the right to work, to vote and engage in social activities outside the family
context). Moreover, regardless any (feminist) theory or efforts for gender equality,
to this day women’s life is subjected to important social changes, which impact
naturally both genders3 .
1

An exception (both methodological and thematically) is the work on fertility of Pakaluk and
Burke (2010), which investigates whether changes in the contraceptive technology may have negatively impacted female happiness.
2
Stevenson and Wolfers’ (2009) paper is noteworthy since, using different sources of data and
measures of subjective well-being for the USA and 12 European countries, they show that female
happiness has declined absolutely and relative to male’s over the past 35 years.
3
Some recent news (from small to front page articles) can help to illustrate it. See, for example,

Another important contribution of our paper is its test of the women’s rights
contribution to the happiness gender gap, besides assessing the role of institutional
variables. A noteworthy approach to the issue of reaching a good life (or human
flourishing), which gives a particular attention to the female context, was developed
by Martha Nussbaum (Nussbaum and Glover, 1995 (eds); Nussbaum, 1993, 2000,
2001). In particular, Nussbaum (1995a) departs from evidence that no country in the
world treats women as well as men, in different respects: longevity, health, education,
employment, political participation. Like Sen, she argues that an adequate theory
of gender (and social) justice is possible if we take into consideration individual
fundamental entitlements. She defends that the best approach to this end is to
center in the human capabilities, which means to focus in what people are effectively
able to do and to be. She argues that it is necessary to guarantee to the each person
the material and social conditions that make possible for each one to opt for what
one thinks is the best for him/herself. Alongside these conditions, she understands
that the internal capabilities of persons should be (institutionally) promoted. This
is equivalent of saying that is not only the objective conditions that matters but
also the internal perception of what one can do or be in his life. At this point the
government role would be to help its citizens to pursue their internal capabilities as
well as the necessary resources and social conditions to act in the way they choose for
themselves. Therefore, Nussbaum’s theoretical work confirms our hypothesis that
institutions, achievements and the system of beliefs that surrounds one’s context
would be of valuable role to people’s well-being.
In what regards institutions, the literature of gender and happiness has not much
specifically explored their role4 . Institutions though, have been shown to play an
important role on individual subjective well-being. A seminal contribution in this
direction was given by Frey and Stutzer (2000a, 2000b, 2002). Although not focusing
in gender aspects, their work provides us with empirical evidence that shows the
importance of democracy to individual well-being. They find that the greater the
degree of democracy, the happier the citizens are. We try to capture these sources
of happiness to the gender context by using variables of female political rights.
Other contributions consider the varying impact of civil rights on happiness.
Veenhoven (2000), for instance, provides intriguing results when exploring the role
on The Economist: Gendercide The worldwide war on baby girls (March, 2010); What happens
when women are over half the workforce (January, 2010). On The New York Times: The Stigma
of Being a Housewife (July, 2010); Italian Women Rise Up (August, 2009). They talk about
homicide and gender preference in millenary cultures; female work and criticism to the feminism;
the controversies of being housewife in the XXI century; sexism in modern, democratic Europe.
4
The work of Bjornsvok et al (2007) is an exception. They account also for institution when
incorporating indicators of women rights as a measure of discrimination.

3

of freedom in subjective well-being of 46 nations. He finds that freedom is not
always positively related to happiness: it contributes positively to happiness in rich
countries, but not in poor ones. On the contrary, he uncovers that economic freedom
is positively related to happiness in poor countries but not in rich ones. Similarly,
we try to test these relations in a gender context, using measures of female economic
and social rights.
Within this background we form our hypothesis that rights, achievements and
beliefs may respond for the happiness gender gap across countries. We measure it
by variables of female rights, female actual achievements and on the national beliefs’
environment, all captured in economic, political and social terms.
For the purpose of this study we apply an empirical strategy similar to Stanca’s
(2010) methodology. It consists on first, determining the subjective well-being at
individual level; and second, using macro level conditions to explain the differences
on the variable-matter (in our case, differences in gender premium on happiness)
across countries. For the first stage we use data from the World Values Survey
(WVS), which contemplates 94 countries, from 1980 to 2008, divided in five waves.
For the indicators of macro conditions we use additionally two other sources: World
Development Indicators (WDI) and Cingranelli-Richards (CIRI) Human Rights data
set.
This work contributes to the literature mainly in three ways. First, it helps to
deepen the scanty literature on gender and happiness based in a methodology5 so far
unused for gender issues; second, it introduces and tests institutions (understood as
female economic, politic and social rights) as determinants of the gender happiness
gap; third, it investigates the happiness gender gap in an encompassing way, by
taking into account women’s life as a whole instead of focusing in single aspects, as
for example age and work. We pursue the following questions: 1) is there a happiness
gender gap at all? and 2) what factors explain it across countries?
The paper is structured as follows: sections 2 and 3 present the methodology
and the data used for the empirical analysis, respectively; the results are presented
in section 4; section 5 concludes with a discussion of the main implications of the
5

We highlight the contribution of this method in at least two ways: it permits to control for
gender invariants, and allows the right variance for the country level explanatory variables, which
enhances its explanatory power comparing to other methods. In studies that put together in the
same regression equation individual and country level variables (as, for example, Bjornskov et al,
2007), the country level variables have minor variance with respect to the individual levels (especially
the dependent variable), what would, probably, transform some explanatory variables into controls.
In our work (as can be seen in the following section), we start at individual level and then, in a
second stage, we move to country level; and in this case, the explanatory country level variables
are indeed explanatory.

4

analysis.

2

Methods

We apply an empirical strategy similar to Stanca’s (2010) methodology6 . It is composed of two empirical steps. In the first, we are interested on performing subjective
well-being equations at individual level for a large range of countries globally. At
this stage, the effect of gender on subjective well-being is estimated for each country
singly. They consist on conditional estimations as we control for a set of factors (economic (ECO), social (SOC), regional (REG) and demographic (DEMO) aspects) that
is considered in the current literature to influence individual happiness7 . In the second step, we are interested on estimating the happiness gender gap across countries.
The dependent variable now are the effects of gender on subjective well-being (from
the first step). This analysis is done taking into account as explanatory variables a
set of macroeconomic, institutional and social conditions at country level.
The individual level equations (in the first step), where we estimate the effect of
gender on individual well-being, is given by:
SW Bij = α+β1j GENij +β2j ECOij +β3j SOCij +β4j DEM Oij +β5j REGij +εij (1)
where W Bij is the well-being of individual i in country j and GEN stands
for gender, that is set to capture the effects for females. Subjective well-being is
proxied by a life satisfaction variable, which is measured in a scale from 1 to 10.
For this reason we chose to use ordered probit model, to properly account for the
cardinal nature of the dependent variable. The variables that account for economic
conditions are income and unemployment; social factors are given by the degree
of education, marital status and a subjective perception of freedom; age and gender
form the demographic variables (although we preferred to isolate gender in the above
equation to emphasize its role in our study); national regional aspects were included
as a measure to capture individual external environment.
In the second step, the effect of gender on determining the individual happiness,
is assumed to depend linearly on macroeconomic, institutional and social conditions,
6

In a step further his work contemplates spatial patterns to capture the effects of economic
conditions on happiness, while here we are interested only in the country level effects themselves
on gender.
7
Present in a variety of research themes, for instance Blanchflower e Oswald (2000), Alesina et
al. (2001), Wolfers (2003), Frey e Stutzer (2004), Di Tella and MacCulloch (2006). Blanchflower
(2008), Clark et al (2008).

5

put in terms of female rights (RIGH) achievements (ACHI) and beliefs (BEL). The
country level equation is given by:

GAPj = α + β1 RIGHj + β2 ACHIj + β3 BELj + εj

(2)

where GAPj is the gender gap for country j.
In order to estimate equation (2), the gender gap is proxied by the βˆ1j of gender from equation (1). In other words, our dependent variable, gender gap, is the
estimated coefficient of gender in equation (1), and is estimated by OLS (in step 2).
The set of rights contemplates women economic, political and social rights; achievements accounts for women outcomes in the society and include female labor force
participation, number of seats of women in parliament, school attainment and life
expectancy ratios (women/men), and fertility rate; values (or beliefs) of the countries
are captured by a set of assertives.
In the following section, we detail the data sources and variable definitions.

3

Data

In this study we use three data sources: the World Values Survey (WVS), the World
Development Indicators (WDI), and the Cingranelli-Richards (CIRI) Human Rights
data. They are used jointly in the second step. For the individual level, in the first
step, we use only the WVS data. These data provide us sources to asses the impact
of female rights, achievements and values to explaining the happiness gender gap.
The World Values Survey is an association that promotes worldwide surveys
about socio-cultural and political change, including values, beliefs, economic and
environmental aspects. It contemplates 94 countries8 , from 1981 to 2008, divided into
five waves (1981-1984, 1989-1993, 1994-1999, 1999-2004, 2005-2008). The countries
investigated were not the same in all the waves9 . The first wave contemplates 20
8

Up to the fourth wave it had investigated over 80 independent countries which covers around
85% of the global population.
9
More specifically, a country belonging to a certain wave is surveyed within a given year among
the years corresponding to that wave. The time span of each wave implies that those were the years
where the survey was started and concluded for a group of countries (e.g.: Albania and Venezuela
were surveyed in the third wave (1994-1999); Albania in the year 1998 and Venezuela in 1996). For
the terms in comma and further details, please see Inglehart (2000).

6

countries, the second 43, the third 55, the fourth 71, and the fifth 57, totalizing
around 257,000 observations.
We use life satisfaction information to proxy the individual subjective well-being.
It is measured in a scale of integer values from 1 to 10, being 1 the most dissatisfied
and 10 the most satisfied, as the answer for the question: All things considered,
how satisfied are you with your life as a whole these days?. The income variable
is obtained in a scale from 1 to 10, where these boundaries work as references for
individual subjective perception of his personal position in the countries’ distribution
of income; they are not expressed in monetary terms, so that they are comparable
between countries. Unemployment is a dummy variable from the employment status
that includes the options retired, student, housewife, part-time, full-time and selfemployment. We work with three levels of education - low, medium and high. Our
measure of trust consists in a dummy variable assuming the value equal to 1 for
those who think that in general people can be trusted”, or 0 if the person cannot
be too careful when dealing with people”10 . Table 2 provides summary statistics for
the variables used in the individual level analysis, in the first step.
The first dataset in the country-level analysis is the CIRI Human Rights. This
is a widely used database by political economists and other social scientists. Its
advantage is that their evaluation distinguishes the human rights polices from its
practices11 . This is worthy of notice, since usually there is a difference on what
governments say they are going to do and what they actually put in practice. This
also converges to what we want to focus on: objective indicators of the degree of
rights women effectively enjoy in each country. The CIRI database uses as source of
its information the annual United States Department of State’s Country Reports on
Human Rights Practices, and the Amnesty International’s Annual Report12 . They
consider physical integrity rights, civil liberties, worker’s rights and women’s rights
to equal treatment, for 200 countries in a time span that varies from 1981 to 2007.
From this database we use the following three variables: women’s economic rights,
10

For the terms in comma and further details, please see Inglehart (2000).
For practices it is understood all the government human rights-related actions. By government
it is included any and all of its agents, such as police and paramilitary forces (Cingranelli and
Richards, 2008). Since they consider only practices, human rights conditions are not accounted
for. Conditions would enclose all that can be happening within a country besides the government’s
action, such as guerrilla groups. Although it seems that conditions would be more informative
than practices, practices give very reliable information, since it transcends the policies showing the
enforcement (or not) of the laws.
12
In case of disagreements among the reports, the Amnesty International report is taken as
authoritative, in order to avoid any US American bias. (Despite that it is highlight that the two
reports joy of high accordance please, see Poe et al (2001).
11

7

women’s political rights, women’s social rights13 . They account for a series of internationally recognized rights. Each variable assume a value that ranges from 0 to 3,
where 0 is assigned if there is none of the (specific) right for women in law and that
sex discrimination might have been brought into law; 1 if women have some rights
under law but not effectively enforced; 2 if women grant of some rights under law
effectively enforced allowing low level of discrimination; and 3 if women have all or
almost all rights under law, fully enforced in practice. Additionally, we created a
fourth variable which is an average of these three, called simply women’s rights. In
our work the three of them are not treated categorically (by dummies), but as the
mean of each variable for all the years corresponding to a given country (the same
for women’s rights variable). The expected sign for their coefficient is positive, since
they reflect provision of means women have to be equal to men in economic, political and social lives, and we assume that this can impact positively the happiness of
women.
The other source of data is the World Development Indicators, provided by the
World bank, with data for 234 countries, covering the period 1960 to 200914 . This is
our source of female achievement (or outcome) in the society. As Nolen-Hoeksema
and Rusting (1999, p. 341) affirm, inequities of power leave women feeling helpless
and devalued and contribute to their higher rates of internalizing disorders, therefore
we expect that their achievement in the society would contribute to enhance female
happiness. We use five variables to capture this achievement. The first is female
labor force participation, which is the female percentage of the total labor force.
The second is the percentage of seats held by women in national parliaments. The
third is a ratio of the educational gender gap, that serves as the female achievement
in educational terms with respect to male; it is given by the ratio of girls to boys in
primary and secondary education. The fourth is the life expectancy ratio, used as a
proxy for the female health outcome in comparison to male’s; the WDI provides us
with the life expectancy at birth (in number of years) for females and males, that
13

Women’s economic rights goes from the right to equal pay for equal work, to job security,
right to work or exercise a profession without male’s consent, and to be free of sexual harassment.
Women’s political rights accounts for right to vote, to run for political office, join political parties
among others. Women’s social rights consider right to inheritance, enter marriage in equality with
men, participate in socio-cultural and community events, freedom from genital mutilation (see
Cingranelli and Richards, 2008c).
14
For the WDI, as for CIRI data, the number of observations and the time series perspective
are disadvantaged since the informations are not available equally to all the countries within all
the years. The further junction of the three data sources constrained us to use the information
as long-run averages (for the maximum time span available in each case), which, nonetheless, is a
procedure used in the literature.

8

permits us to constructing the ratio of female to male life expectancy at birth. The
fifth variable (that can be interpreted by a control) is the fertility rate, given by the
national number of births given by women15 .
All these variables were averaged out in the available period and the percentage
rates were transformed in decimals. For all of them we expect positive coefficients as
we assume that they enhance women’s feeling of participation in the social/national
life and as so would help to increase their happiness within a country.
Lastly, we select a set of variables that can serve as a proxy for the system of
beliefs women face within each country. We use WVS data for the two genders,
averaging it out to transform the data from individual to national levels. They are
obtained in base on statements for which individuals are asked to express agreement
or disagreement choosing from an ordering that indicates 1 for strongly agreement,
up to 4 strongly disagreement. We rescaled these variables so then they express an
agreement ordering16 . The statements are: when jobs are scarce, men should have
more right to job than women; men make better political leaders than women do;
university education is more important for a boy than for a girl; being a housewife
is just as fulfilling as working for pay. We assume that the disagreement to these
beliefs would enhance women happiness17 .
Table 3 provides summary statistics to all the variables used in the macro level
analysis.
15

Fertility indicators are the core of development economics and demography, and it may also
work as a synthetic measure of women status within a society, since fertility choices are correlated
with several domains (women’s role in the family, labour force participation etc.). It is important
to recall also that one of the reasons women quit job and give up their career is exactly due to
motherhood. If having and bearing children would not be important to women’s satisfaction this
trade off would not be made elsewhere (even with different degrees across women and countries).
Hence fertility may be a proxy for women’s achievement on being able to express their maternity
(specially in the countries were female work is more noticed). Our hypotheses is that higher levels
of national fertility is positively related to female happiness.
16
All these variables were rescaled to express an agreement ordering, varying now from -1 (agree
strongly) to -4 (strongly disagree). Exception for one variable which have only a binary response
(agree or disagree) when jobs are scarce, men should have more right to job than women.
17
With respect to the statement being a housewife is just as fulfilling as working for pay we
are a bit dubious on its effect. If it is understand as a discriminatory belief that would prevent or
discourage women from working despite their personal will, then a disagreement with this statement
would contribute to female happiness. But, if it can be understood as a belief that is supportive
to women’s choice of working or not, than a disagreement could represent a decrease in female
happiness. So, we depart somehow neutral regarding this variable.

9

4

Results

The results of the estimation of equation (1) for the individual level are shown in table
4. To synthesize them, the variables that provide a strong and significant impact
on subjective well-being (at the 1% significance level) are income, unemployment,
gender, education, married status, freedom and trust. The increase of one decile in
the income scale represents a strongly significant increment in life satisfaction (0.07).
In the opposite direction, being unemployed has a wide strong negative significant
impact on subjective well-being (-0.19). Education (middle and high education,
compared to lower levels) has a positive effect on individual life satisfaction (0.04 and
0.03 respectively). With respect to singles, being married reveals a strong positive
effect to individual well-being (0.10). Self perception of freedom (0.16) and trust
(0.10) are also positive and strongly significant to individual life satisfaction. Lastly,
compared to females, males are less happy, presenting negative significant coefficient
(-0.03).

4.1

The happiness gender gap

Analysing the estimate of equation (1) we can answer our first question, affirming
that indeed there is a global happiness gender gap and that it favours women. This
result is shown in detail in Table 1, that presents for the country specific regressions
the estimated coefficients of being female on life satisfaction. They range from 3.57 for Trinidad e Tobago, to 8.41 to Tanzania. Out of 94 countries, 59 (almost
63%) present a positive gap (favouring women) and 35 (37%) show a negative gap
(favouring men).
Surprisingly, the first five countries in the ranking are Tanzania, Uganda, Algeria,
Zimbabwe and Jordan, which are developing countries where women’s rights are not
well developed18 . This is the pattern that the majority of the continent follows - the
exception is Rwanda whose happiness gap does not favour women (-0.74).
On the other edge of the ranking, were women are less happy than men, we
cannot verify a particular pattern of any continent; instead, countries of all sorts
of socio-economic and geographic characteristics are represented: Singapore (-0.01),
Burkina Faso (-0.30), Norway (-0.40), Israel (-0.95), Portugal (-1.23), South Korea
18
Indeed, using the CIRI Human Rights data, we calculated the women rights syntactic indicator
for these countries (that consists on an average of the three CIRI indicators for economic, political
and social rights see section 3), which are: 1.27 for Tanzania, 0.97 for Uganda, 0.91 for Algeria,
1.24 for Zimbabwe, and 0.89 for Jordan (in a scale from 0 to 3, where the closer the country score
is to zero, lower level of rights women have, and the closer to 3, higher level of rights women have).

10

(-1.51), Brazil (-2.67), among others. Nonetheless, interestingly there is a considerable representation of developed countries such as Japan (-0.04), United States
(-0.28) and 14 European countries (including Switzerland (-0.31), Luxembourg (0.36), Austria (-0.39) and Italy (-0.47)), although the majority of these countries
present low happiness gender gap, given their small coefficients.

4.2

The determinants

We now turn to the country level estimates, analysing the potential determinants of
the gender gap. Tables 5 to 8 show the results of estimating the effects of aggregate
conditions on the happiness gender gap. We recall from section 2 that our dependent
variable, the happiness gender gap, is the estimated coefficient of the gender dummy
featuring in equation (1) (that measures the country-specific sensitivity of happiness
to the female gender with respect to male).
The main results obtained were initially suspicious. Differently from our initial
hypotheses, some variables of women achievement and rights presented negative significant coefficients (female labor force participation and three out of four measures
of women rights). In the same way, two variables that account for social beliefs are
instead surprisingly related to the gender premium in a positive way (they are: when
jobs are scarce, men should have more right to job than women and men make better
political leaders than women do). These paradoxical results will be discussed now in
detail.
Table 5 provides the results of regressing the gender gap on women’s rights variables. Contrary to our hypothesis, women’s economic rights are not significant to
explain the happiness gender gap. This result is very intriguing. In the introduction,
we affirmed that one of the aspects for which women’s life suffered a major change
was, indeed, the economic sphere. It is one of the first instances where women conquered a right that previously were exclusively of male domain. The three other
measures of rights possess a significant negative impact. We offer some possible
explanation for these facts in the sequence.
The effect of women’s political rights is stronger. An increment of one percentage
point in political rights19 represents a decrease of 0.48 percentage points on the happiness gender gap, whereas the effect of women’s social rights and of the synthetic
indicator of women’s rights are respectively, a decrease of 0.22 and 0.33 percentage
points. These results are significant only if the happiness gender gap is regressed unconditionally on the female rights. Their estimated impact, in fact, is not maintained
19

The original discrete variables of women rights have been rendered continuous once averaged
across the years.

11

when we add the variables of achievement and beliefs (table 8).
The results for the effect of achievement on the happiness gender gap are reported
in Table 6. Except for female labor force participation (FLFP) and fertility rate, the
other variables are not statistically significant. FLFP has a strong and significant
negative effect on women’s well-being; an increase of one percentage point reduces
women’s happiness of 3.38 percentage points. Also when regressing conditionally
on all achievement variables, FLFP marginal effect remains strong and significant
(equal to -3.22, column 6). This result is robust also in the specification where all
the variables of rights, achievements and beliefs are included (table 8).
Finally, the results for the analysis of the role of the national context of beliefs
on the happiness gender gap is reported in table 7. Among them, the ideas that
men should have more right to job and that men make better political leaders are
significant to explain the happiness gender gap. Surprisingly, they show that living
in a country where the social context is in favour of men’s job over that of women
provides significant positive marginal effect (1.12) on female life satisfaction. Although the significance and the absolute value are reduced, the variable men should
have more rights to job continues to be significant also in the regression conditioned
to other belief variables (column 5); the variable men make better political leaders,
this time, loses its significance.
When first thinking on what could explain the eventual difference on happiness
among men and women across the world, we focused on the female case, since it is
the gender which still suffers from inequalities in many domains, and the one that
had major changes in life-style in the past half century. Many of these changes were
promoting women’s rights and their effective achievements in the society. Nations,
even if in different degrees, have decreased the gender discrimination in many fronts,
specially in the developed and more industrialized countries (for recent years, see
for instance, the Gender Gap Index Reports, 2006 to 2010). Nonetheless, all these
advantages seem to be not corroborated by our finding women’s happiness. This
is the reason why Blanchflower and Oswald (2004) referred to this phenomenon as
a paradoxical view, and why Stevenson and Wolfers (2009) coined the expression
the paradox of (declining) female happiness20 . Our results show that most of the
variables that would hypothetically contribute to their happiness, instead seem to
produce a diminishing effect. This paradox needs further investigation and evokes
some hypotheses for its existence. We highlight four possible explanations: (i) social costs involved in the transition; (ii) self-perception of freedom/autonomy; (iii)
20

Stevenson and Wolfers (2009) analysed women happiness along the time, and this is why they
can point a declining pattern. Our analysis do not account for the time domain, but still confirms
the paradoxical issue of women happiness highlighted by these two works.

12

relational crowding out effect; and (iv) rights-expectations treadmill effect.
For hypotheses (i) and (ii) we recall Nussbaum’s work. The first, refers to the
intrinsic dynamic nature of the processes of (gender and) social change, highlighting
an alternative view of the role of social (hence women’s) rights and their impact on
individual happiness. For example, Nussbaum highlights that the conquest of higher
rights involve conflicts and social tensions (what I refer to as social costs), which
might eventually decrease the net benefit of acquiring them at least in the short
and medium run21 . Following this line of reasoning, one could collect the subjective
benefit of social activism only in the long run, perhaps after a generation and this
requires a proper account of the time dimension of the underlying rights-happiness
relation, which is severely constrained by the predominant cross-sectional nature of
our estimation exercise.
The second hypothesis suggests the importance of accounting for the self-perception
of freedom/autonomy. Nussbaum argues that well-being depends on material and social conditions as well as on internal capabilities. Without these external conditions
and the internal capability it would not be possible to anyone to choose what to do or
to be according to what each one considers valuable for himself. Our results seem to
show that an improvement in external materials and social conditions (represented
by female rights and achievements in economic, political and social aspects) in reality produce a negative impact on female life satisfaction. Instead, it might simply
mean that, without being accompanied by internal personal conditions, rights and
achievements themselves are not sufficient to increase female happiness relative to
male’s. Therefore, we might be missing out the role of internal personal capabilities.
This is an interesting issue to be further investigated. Indeed, Bavetta and Navarra
(2011), show empirical evidence that the individual perception of own freedom or
autonomy is complementary to the countrys economic freedom in enhancing one’s
well-being.
The third hypothesis refers to the possibility of a relational crowding out effect,
where the impact of higher rates of female achievement (we can focus on female
labor force participation) might be offsetting the consumption of relational goods22
(Bruni and Stanca, 2008). The second shift theory (Hochschild and Machung, 1989),
21

Easterlin (2009) provides empirical evidence on this respect. He shows that Eastern Europe’s
transition from socialism to capitalism in the 1990’s was not followed instantaneously by an increase
in life satisfaction. Instead, around 15 years later subjective well-being records were bellow pretransition levels. More specifically, he argues that the increased satisfaction with material living
level occurred at the cost of satisfaction with work, health and family life.
22
Bruni and Stanca (2008) investigates the role of relational goods to individual happiness, finding
not only a significant relation, but also suggesting a relational treadmill, where the effect of higher
income on happiness would be offset by a lower consumption of relational goods.

13

indeed, helps to point out this issue. The fact that women work outside home in the
market and then face a second shift at home, leaves little space to opportunities for
a richer social life, entertainment or relaxing activities. Despite some evidence that
the time dedicated by men in home production is increasing (Blau, 1998), culturally
perhaps, home making is still womens role, will or responsibility. We should also
comment that the strong significant and positive effect that when jobs are scarce,
men should have more right to a job than women puts in evidence once again the
relational character of female kind. Again, a possible explanation is that women
are able to give up something to the other gender if they judge this is important or
necessary, or that they prefer to balance out the professional life with other familiar
and social relations, if they were given the choice.
The fourth hypothesis suggests the existence of a rights-expectations treadmill,
that would follow in a certain way the income treadmill, which means that some
beneficial effect of rights can be offset by aspirations of more/better rights. Someone
who lives in a developed country, with her rights guaranteed by the democratic
State, may feel tempted of judging insufficient what is already available23 . This
line of thinking can be also translated as the more you have, the more perfect you
want it to be, showing a not-satisfying pattern that intensifies as the consumption
of the rights grows. In the other side, there are many women in the world that
probably do not know (or understand) the existence of rights (we can think of the
least developed African countries, for example). It is worth noticing, in this case,
that when they answer how satisfied they are with their lives as a whole (which
originates our dependent variable), they are considering what effectively exists in
what they conceive their universe is24 .
Regarding the relation between fertility and happiness, it is interesting to notice
from the literature that, at the individual level, there is usually a negative association
between own fertility and subjective well-being, unless other issues are taken into
consideration, such as financial satisfaction (Stanca, 2009), or that country level
23

For instance, we can mention the German case that offers part-time child care, or the French
case that provides a certain amount of money for mothers who decide to stay at home to bring up
their children. In the first case, a part of the mothers who make use of this right are not satisfied as
they judge insufficient the amount of time made available by the welfare State. In the second case,
some women may judge as not fair or insufficient the amount of money reserved for the finality of
raising their children.
24
If a woman lives within a burka, and she conceives her life as doing what that restriction permits
her (which means basically having a life limited to the family and the house), then, the right of
living without it and doing other things may not be part of what she foresees for herself. Moreover,
it can also be damaging for her, whatever external decision aimed at granting women more rights
a similar episode occurred recently in France, where the burka was publicly banned.

14

indicators are incorporated in the analysis (Margolis and Myrskyla, 2010). At this
point, the negative relation may vanish, and turn to positive. Our result, presenting
a positive relation between fertility and women’s happiness, matches this second line
of evidence (though using another estimation strategy). In this case, the different
level of analysis and aggregation might capture other dynamics. In fact, we might
suppose that, while at the individual level to decide to have a numerous family
involve substantial costs and trade-offs, at the aggregate level a high fertility rate
might signal a virtuous environment, enhancing women happiness.

5

Concluding remarks

This paper was interested in finding and explaining a possible gap between female and
male subjective well-being across countries. For this purpose, it laid an encompassing
view, considering female rights and achievements as well as the context of national
beliefs, all in different domains (economic, political and social).
Combining an original dataset from the World Values Survey, World Development
Indicators and CIRI Human Rights, a two-step estimation strategy was used. This
found a happiness gender gap, favouring women throughout the world. Surprisingly,
we found that many African and developing countries belong to the positive side of
the female happiness distribution, while many European and industrialized countries
are responsible for the negative indices of women happiness.
We obtained some paradoxical results, at least at first sight: objective matters
of female rights and achievements display a negative impact on women’s happiness,
whereas national beliefs that would favour men at the expense of women in economic
and political terms would grant them happiness. Beyond these findings, fertility has
a positive effect on their happiness.
First, our results show that the work of Stevenson and Wolfers (2009) cannot
be generalized to other countries, with respect to those for which their results hold
(USA and 12 European countries). In fact, while they affirm that female happiness
has been declining both in absolute terms and relatively to men in their sample, we
found that, in a larger domain, women are, on average, happier than men.
Concerning our estimates, for interpreting the main apparently counter-intuitive
results (the greater their rights and objective conditions, the less happy women are),
we formulated a few possible hypotheses.
Specifically, our apparent paradox of women happiness can first be solved going
back to Nussbaum’s theory. The extensive work on the subject of this scholar highlights a few possible lines of explanation, mostly complementary. First, Nussbaum’s
work clarifies the personal and social costs involved in acquiring higher rights for
15

women, which might eventually decrease the net individual pay-off at least in the
short and medium run. However, the nature of our data, at the moment, does not
allow an easy temporal test of this hypothesis.
Next, Nussbaum emphasizes the role of internal and external capabilities for
achieving well-being, and argues that objective conditions are just a facilitator for
individual flourishing. Last, the second shift theory and the possible existence of a
rights-expectations treadmill particularly for women - might add further pieces of
sense to the initially puzzling evidence on the main determinants of the happiness
gender gap.
Our findings, then, require a more complex discussion than that initially expected
on the basis of the existing literature, and are worthy of further investigations. The
concerned social phenomena, in fact, involve non trivial methodological complexities.
At the moment, this task was severely constrained by the predominant nature of our
cross-sectional estimates, while a more refined test would require additional and
longitudinal evidence, better able to capture the intrinsically dynamic nature of the
process of socio-economic and political change.
Last, but not least, a certain aggregate evidence of a positive role of fertility
for women happiness emerges, together with a call for appropriate public policy
programs.

16

6

References

Alesina, A., Di Tella, R., MacCulloch, R. (2001). Inequality and happiness: Are
Europeans and Americans different? NBER Working Paper n. 8198. National
Bureau of Economic Research, Inc.
Altindag, D. T., Xu, J. (2009). The Impact of Institutions and Development on
Happiness. Working Paper n. 2009-17, Louisiana State University.
Bavetta, S.; Navarra, P. (2011). Economic Freedom and the Pursuit of Happiness.
In: Miller, T., Holmes, K. R. (Eds.). 2011 Index of Economic Freedom. The
Heritage Foundation.
Berger, E. M. (2009). Maternal Employment and Happiness: The Effect of NonParticipation and Part-Time Employment on Mothers’ Life Satisfaction. SOEPpapers 178, DIW Berlin.
Bjrnskov, C., Dreher, A., Fischer, A. V. J. (2007). On Gender Inequality and Life
Satisfaction: Does Discrimination Matter?. Economics Discussion Paper, n.
2007-07, University of St Gallen.
Blanchflower, D. G., Oswald, A. J. (2000). Well-Being Over Time in Britain and
the USA. NBER Working Papers 7487, National Bureau of Economic Research,
Inc.
Blanchflower, D. G., Oswald, A. (2004). Well-Being Over Time in Britain and the
USA. Journal of Public Economics, 88(7-8), 1359-1386.
Blanchflower, D. G. (2008). International Evidence on Well-being, IZA Discussion
Paper 3354, Institute for the Study of Labor (IZA).
Blau, Francine D. (1998). Trends in the Well-Being of American Women, 19701995.
Journal of Economic Literature. 36(1), 11265.
Bruni, L., Stanca, L. (2008). Watching alone: Relational goods, television and
happiness, Journal of Economic Behavior & Organization, Elsevier, 65(3-4),
506-528.

17

Cingranelli, D. L., Richards, D. L. (2008a). The Cingranelli-Richards (CIRI) Human Rights Data Project Coding Manual Version 2008.3.13. Available at
< http : //ciri.binghamton.edu/documentation/ciric odingg uide.pdf >
Cingranelli, D. L., Richards, D. L. (2008b). The Cingranelli-Richards Human Rights
Dataset Version 2009.02.20. < http : //www.humanrightsdata.org >
Cingranelli, D. L., Richards, D. L. (2008c). The Cingranelli-Richards Human Rights
Project: Short Variable Descriptions for Indicators in the Cingranelli-Richards
(CIRI) Human Rights Dataset. < http : //www.humanrightsdata.org >
Clark, A., Frijters, P., Shields, M.A. (2008). A survey of the income happiness
gradient, Journal of Economic Literature, 46(1), 95-144.
Crocker, D. A. (1995). Functioning and Capability: The Foundation of Sen’s and
Nussbaum’s Development of Ethics. In: Nussbaum, M. C., Glover, J. (Eds.)
(1995). Women, Culture and Development: A Study of Human Capabilities;
Delhi: Oxford University Press.
Di Tella, R., MacCulloch, R.J., Oswald, A.J. (2001). Preferences over inflation
and unemployment: evidence from surveys of happiness. American Economic
Review, 91, 335-341.
Di Tella, R., MacCulloch, R. (2006). Some Uses of Happiness Data in Economics,
Journal of Economic Perspectives, 20(1), 25-46.
Easterlin, R. A. (2003). Happiness of Women and Men in Later Life: Nature,
Determinants, and Prospects. In: Sirgy, M. J., Rahtz, D. and Samli, A. C.
(Eds.). Advances in Quality-of-Life Theory and Research. Dordrecht, The
Netherlands: Kluwer Academic Publishers, 13-26.
Easterlin, R. A. (2009). Lost in Transition: Life Satisfaction in the Road to Capitalism. Journal of Economic Behavior and Organization. 71(1), 130-145.
Frey, B. S., Stutzer, A. (2000a). Happiness, Economy and Institutions. Economic
Journal. 110(466), 918-38.
Frey, B. S., Stutzer, A. (2000b). Happiness Prospers in Democracy. Journal of
Happiness Studies. 1(1), pp.79-102.
Frey, B S.; Stutzer, A. (2002). Happiness and Economics. Princeton and Oxford:
Princeton University Press.
18

Frey, B S.; Stutzer, A. (2004). Does marriage make people happy, or do happy
people get married? The Journal of Socio-Economics, v. 35, p. 326347.
Gash, V., Mertens, A., Romeu Gordo, L. (2010). Women between Part-Time and
Full-Time Work: The Influence of Changing Hours of Work on Happiness and
Life-Satisfaction. SOEPpapers 268, DIW Berlin.
Gimenez-Nadal, J.I.; Sevilla-Sanz, A. (2010). The Time-Crunch Paradox, forthcoming in: Social Indicators Research. Online first http://dx.doi.org/10.1007/s11205010-9689-1
Guven, C., Senik, C., Stichnoth, H. (2009). You Can’t Be Happier than Your Wife:
Happiness Gaps and Divorce. IZA Discussion Papers 4599, Institute for the
Study of Labor (IZA).
Inglehart, R. (2000). World Values Surveys and European Values Surveys, 19811984, 1990-1993, and 1995-1997; Ann Arbor, MI: Interuniversity Consortium
for Political and Social Research.
Luechinger, S., Meier, S., Stutzer, A. (2008). Bureaucratic Rents and Life Satisfaction. Journal of Law, Economics and Organization 24(2), 476-88.
Hochschild, A., Machung, A. (1989). The Second Shift. New York: Viking.
Margolis, R., Myrskyla, M. (2010). A Global Perspective on Happiness and Fertility.
Working Paper n. 2010-025, MPIDR. Max Planck Institute for Demographic
Research, Rostock, Germany.
Nolen-Hoeksema, S., Rusting, C. L. (1999). Gender Differences in Well-Being.
In: Kahneman, D., Diener, E., Schwarz, N. (Eds.) (1999). Well-Being: The
Foundations of Hedonic Psychology. New York: Russel Sage Foundation.
Nussbaum, M. C., Sen. A. (1993), The Quality of Life, Oxford: Clarendon Press.
Nussbaum, M. C., Glover, J. (Eds.) (1995). Women, Culture and Development: A
Study of Human Capabilities; Delhi: Oxford University Press.
Nussbaum, M. C. (1995a). Introduction. In: Nussbaum, M. C., Glover, J. (Eds.)
(1995). Women, Culture and Development: A Study of Human Capabilities;
Delhi: Oxford University Press.

19

Nussbaum, M. C. (2000), Woman and Human Development. The Capabilities Approach (Italian translation: (2001), Diventare persone. Donne e universalita
dei diritti, Bologna: Il Mulino).
Nussbaum, M.C. (2001), Upheavals of Though. The Intelligence of Emotions (Italian translation: L’intelligenza delle emozioni, Bologna: Il Mulino).
Pakaluk, C., Burke, J. (2010). The New Battle of the Sexes: Understanding the
Reversal of the Happiness Gender Gap. Working Papers 1004, Ave Maria
University, Department of Economics.
Plagnol, A., Easterlin, R. (2008). Aspirations, Attainments, and Satisfaction: Life
Cycle Differences Between American Women and Men. Journal of Happiness
Studies. 9, 601619.
Poe, S. P., Carey, S. C., Vazquez, T.C. (2001). How are these pictures Different?
A quantitative comparison of the US State Department and Amnesty International human rights reports, 1976-1995. Human Rights Quarterly. 23(3),
650-677.
Stanca, L. (2009). Suffer the Little Children: Measuring the Effects of Parenthood
on Well-Being Worldwide. Working Paper n. 173, Economics Department,
University of Milan Bicocca.
Stanca, L. (2010). The Geography of Economics and Happiness: Spatial Patterns in
the Effects of Economic Conditions on Well-being. Social Indicators Research,
99(1), 115-133.
Stewart, F. (2001), Women and human development: the capabilities approach,
by MARTHA NUSSBAUM (Cambridge University Press, Cambridge: 2000).
Journal of International Development. 13, 11911192.
Stevenson, B., Wolfers J. (2008). Happiness inequality in United States. NBER
Working Paper n. 14220. National Bureau of Economic Research, Inc.
Stevenson, B., Wolfers. J. (2009). The Paradox of Declining Female Happiness.
American Economic Journal: Economic Policy. 1(2), 190-225.
Trzcinski, E., Holst, E. (2010). Gender Differences in Subjective Well-Being in and
out of Management Positions. SOEPpapers 299, DIW Berlin.

20

Veenhoven, R. (2000). Freedom and Happiness: A Comparative Study in Forty
Four Nations in the Early 1990s. In: Diener, E., Suh, E. M. (Eds.). Culture
and Subjective Well Being. Cambridge, MA: MIT Press. 257-88.
Wolfers, J. (2003). Is Business Cycle Volatility Costly? Evidence from Surveys
of Subjective Well-being, NBER Working Papers 9619, National Bureau of
Economic Research, Inc.
World Bank. (2009). Data retrieved March 23, 2010, from World Development
Indicators Online (WDI) database.

6.1

Further Readings

Piore, M.J., Sabel, C. F. (1984). The Second Industrial Divide. Possibilities for
Prosperity, Basic Books.
Moller Okin, S. (1999), Is Multiculturalism Bad for Women? (Italian translation:
(2007) Diritti delle donne e multiculturalismo, Milano: Raffaello Cortina).
Nussbaum, M.C. (2000), Duties of Justice, Duties of Material Aid. Cicero’s Problematic Legacy, (Italian translation: (2008), Giustizia e aiuto materiale, Bologna:
Il Mulino).
Nussbaum, M.C. (2002), Giustizia sociale e dignita umana, Bologna: Il Mulino.
Sen, A. K. (1980). Equality of What? In: McMurrin, S. (Ed.), Tanner Lectures on
Human Values, i. Cambridge: Cambridge University Press.
Sen, A. (1993). Capability and Well-Being. In: Nussbaum, M. C., Sen. A. (1993),
The Quality of Life, Oxford: Clarendon Press.
Tran-Nguyen, A-N, Zampetti, A.B, (2004). Trade and Gender: Opportunities and
Challenges for Developing Countries. UNCTAD. United Nations, New York
and Geneva.

21

Table 1: Gender gap (Female-Male), Life Satisfaction (OLS), by country
Rank Country
Coeff. t-stat Rank Country
Coeff. t-stat
1
Tanzania
8.41
3.33
48 El Salvador
0.40
0.22
2
Uganda
7.59
4.30
49 Venezuela
0.35
0.24
3
Algeria
6.83
3.54
50 Egypt
0.30
0.21
4
Zimbabwe
6.73
3.05
51 Canada
0.23
0.40
5
Jordan
5.99
3.70
52 Australia
0.21
0.36
6
Iraq
4.38
2.04
53 Germany
0.21
0.46
7
Morocco
3.21
3.75
54 Denmark
0.16
0.22
8
Nigeria
2.83
3.79
55 Albania
0.08
0.05
9
Malaysia
2.53
2.29
56 Sweden
0.06
0.11
10
Taiwan
2.42
3.15
57 Argentina
0.04
0.06
11
Uruguay
2.34
1.41
58 Georgia
0.04
0.05
12
Hong Kong
2.09
1.64
59 Singapore
-0.01 -0.01
13
Netherlands
2.02
3.33
60 Russian Fed.
-0.03 -0.05
14
Bangladesh
1.99
1.35
61 Iceland
-0.03 -0.04
15
Iran
1.86
2.01
62 Japan
-0.04 -0.05
16
Poland
1.82
2.19
63 North Korea
-0.07 -0.05
17
Czech Republic 1.68
2.13
64 Chile
-0.09 -0.11
18
India
1.68
2.01
65 Malta
-0.22 -0.14
19
China
1.67
2.31
66 United Kingdom -0.24 -0.43
20
Ghana
1.66
1.19
67 United States
-0.28 -0.57
21
France
1.63
2.47
68 Burkina Faso
-0.30 -0.18
22
Hungary
1.60
1.77
69 Switzerland
-0.31 -0.40
23
Turkey
1.58
1.71
70 Slovakia
-0.33 -0.33
24
Mali
1.50
0.76
71 Luxembourg
-0.36 -0.28
25
Croatia
1.47
1.39
72 Bulgaria
-0.39 -0.44
26
South Africa
1.32
2.48
73 Austria
-0.39 -0.29
27
Serbia
1.31
1.54
74 Norway
-0.40 -0.57
28
Ethiopia
1.20
1.17
75 Cyprus
-0.41 -0.29
29
Macedonia
1.09
0.95
76 Italy
-0.47 -0.71
30
Finland
1.08
1.39
77 Romania
-0.50 -0.56
31
New Zealand
1.06
1.33
78 Ukraine
-0.53 -0.64
32
Slovenia
1.06
1.49
79 Thailand
-0.73 -0.80
33
Belgium
0.97
1.57
80 Rwanda
-0.74 -0.58
34
Armenia
0.82
0.67
81 Peru
-0.79 -0.91
35
Philippines
0.81
0.71
82 Azerbaijan
-0.93 -0.74
36
Latvia
0.77
0.76
83 Israel
-0.95 -0.61
37
Moldova
0.73
0.83
84 Belarus
-1.10 -1.39
38
Montenegro
0.73
0.54 22 85 Colombia
-1.13 -2.29
39
Mexico
0.69
0.78
86 Pakistan
-1.15 -0.74
40
Estonia
0.63
0.71
87 Portugal
-1.23 -1.31
41
Ireland
0.63
0.76
88 South Korea
-1.51 -1.51
42
Zambia
0.62
0.40
89 Lithuania
-1.59 -1.41
43
Greece
0.61
0.42
90 Puerto Rico
-1.69 -1.49
44
Viet Nam
0.57
0.72
91 Dominican Rep. -1.69 -0.55
45
Indonesia
0.53
0.56
92 Andorra
-2.06 -2.08
46
Bosnia-Her.
0.48
0.54
93 Brazil
-2.67 -3.23
47
Spain
0.47
1.20
94 Trinidad-Tobago -3.57 -2.28
Note: Source: World Values Survey (ICPSR).

Table 2: Descriptive statistics, individual level
Variable
Mean Std. Dev. Min. Max.
Life satisfaction
66.42
24.58
10
100
Income
4.69
2.48
1
10
Unemployed
0.08
0.28
0
1
Empl. Full time
0.38
0.48
0
1
Empl. Part time
0.07
0.26
0
1
Empl. Self
0.09
0.29
0
1
Empl. other
0.02
0.13
0
1
Retired
0.13
0.34
0
1
At home
0.14
0.35
0
1
Student
0.07
0.26
0
1
Education, lower
0.27
0.45
0
1
Education, middle
0.45
0.5
0
1
Education, upper
0.28
0.45
0
1
Married
0.59
0.49
0
1
As married
0.05
0.22
0
1
Divorced
0.03
0.18
0
1
Separated
0.02
0.13
0
1
Widowed
0.07
0.25
0
1
Single
0.25
0.43
0
1
Number of children
2.15
1.68
0
8
Age
40.96
16.4
15
101
Male
0.48
0.5
0
1
Freedom
6.72
2.43
1
10
Importance of religion
2.94
1.07
1
4
Trust in others
0.29
0.45
0
1
Honesty
8.59
2.35
1
10
Survey wave 1
0.09
0.28
0
1
Survey wave 2
0.17
0.38
0
1
Survey wave 3
0.24
0.43
0
1
Survey wave 4
0.28
0.45
0
1
Survey wave 5
0.22
0.41
0
1

23

N
340504
294490
330385
330385
330385
330385
330385
330385
330385
330385
345102
345102
345102
341234
341234
341234
341234
341234
341234
282635
337669
341580
325257
304050
330569
321290
346324
346324
346324
346324
346324

Table 3: Descriptive statistics, country level
Variable
Mean Std. Dev. Min. Max. N
Gender gap (female)
0.09
0.68
-1.42
5.99 94
Women Economic Rights
1.47
0.52
0
2.5
92
Women Political Rights
1.89
0.36
0.54
3
92
Women Social Rights
1.48
0.78
0
3
92
Female Labor Force Part
0.41
0.08
0.14
0.53 89
Woman in Parliament
0.16
0.09
0.02
0.44 89
School Enrollment Ratio
0.98
0.07
0.71
1.07 86
Life Expec. Ratio
1.08
0.03
1
1.19 90
Fertility Rate
3.18
1.57
1.57
7.24 93
Men more right to job
-1.8
0.23
-2.4
-1.19 79
Men better P. Leaders
-2.49
0.43
-3.43 -1.43 79
Univ. more imp. boys
-3.02
0.29
-3.7
-2.31 79
Housewife fulfilling
-2.25
0.31
-3.16 -1.71 79

Table 4: Determinants of subjective well-being, individual level
LS (OP)
Regressors
Income
Unemployed
Empl. Part time
Empl. Self
Empl. other
Retired
At home
Student
Education, middle
Education, upper
Married
As married
Divorced
Separated
Widowed
Number of children
Age
Male
Honesty
Trust in others
Freedom
Importance of religion
Pseudo R2 / R2
Number of obs.

0.07**
-0.19**
-0.01
-0.02*
-0.03
0.08**
0.09**
0.09**
0.04**
0.03**
0.10**
0.02
-0.13**
-0.21**
-0.04**
-0.01*
-0.00*
-0.03**
0.01**
0.10**
0.16**
0.06**
0.08
185573

LS (OLS)
(55.83)
(-17.71)
(-1.06)
(-2.47)
(-1.78)
(8.35)
(9.93)
(7.11)
(5.51)
(3.73)
(11.51)
(1.44)
(-9.22)
(-10.13)
(-3.00)
(-2.50)
(-2.41)
(-5.96)
(12.49)
(18.59)
(115.14)
(22.43)

1.42**
-4.15**
-0.29
-0.52**
-0.85*
1.33**
1.64**
1.75**
0.92**
0.94**
1.97**
0.39
-2.67**
-4.25**
-0.85**
-0.12**
-0.02**
-0.63**
0.27**
2.15**
3.20**
1.14**
0.31
185573

(60.52)
(-19.47)
(-1.51)
(-2.96)
(-2.26)
(6.99)
(9.30)
(7.12)
(6.76)
(6.11)
(11.30)
(1.47)
(-9.32)
(-10.29)
(-3.06)
(-3.02)
(-3.62)
(-5.90)
(11.70)
(19.93)
(123.60)
(20.11)

Note: Dependent variable: Life satisfaction. Estimator: Ordered Probit (column 1) or OLS
(column 2). t-statistics reported in brackets (heteroskedasticity-robust standard errors). *
indicates p<0.05, ** indicates p<0.01.

24

Table 5: Determinants of the happiness gender gap: rights
(1)
Regressors
Women Economic Rights

(2)

(3)

-0.23
(-1.64)

Women Political Rights

-0.48*
(-2.49)

Women Social Rights

-0.22*
(-2.48)

Women Rights
Constant

0.43
(1.98)
0.03
92

R2
R2

(4)

1.00**
(2.70)
0.06
92

0.43**
(2.82)
0.06
92

-0.33*
(-2.40)
0.62**
(2.69)
0.06
92

Note: Dependent variable: Effect of on Life Satisfaction. t-statistics reported in brackets
(heteroskedasticity-robust standard errors). * indicates p<0.05, ** indicates p<0.01.

Table 6: Determinants of gender gap: achievements
(1)
Regressors
Female Labor Force Part

(2)

(3)

(4)

-3.38**
(-3.85)

Woman in Parliament

-1.47
(-1.72)

School Enrollment Ratio

-0.71
(-0.66)

Life Expec. Ratio

-4.24
(-1.95)

Fertility Rate
Constant
R2
R2

1.46**
(4.04)
0.15
89

0.34*
(2.13)
0.03
89

0.79
(0.75)
0.01
86

4.69*
(1.99)
0.04
90

(5)

-3.22**
(-2.87)
-0.66
(-0.69)
2.18
(1.59)
-0.08
(-0.03)
0.09*
0.10
(2.10) (1.55)
-0.21
-0.85
(-1.31) (-0.27)
0.05
0.19
93
82

Note: Dependent variable: Effect of on Life Satisfaction. t-statistics reported in brackets
(heteroskedasticity-robust standard errors). * indicates p<0.05, ** indicates p<0.01.

25

(6)

Table 7: Determinants of gender gap: beliefs
(1)
Regressors
Men more right to job

1.12**
(3.35)

Men better P. Leaders
Univ. more imp. boys
Housewife fulfilling
Constant
R2
R2

2.12**
(3.49)
0.13
79

(2)

(3)

(4)

(5)

1.04*
(2.07)
0.43*
0.31
(2.27)
(0.78)
0.23
-0.58
(0.81)
(-1.24)
0.35
0.19
(1.28) (0.73)
1.17*
0.80
0.88
1.42
(2.45) (0.92) (1.43) (1.19)
0.06
0.01
0.02
0.15
79
79
79
79

Note: Dependent variable: Effect of on Life Satisfaction. t-statistics reported in brackets
(heteroskedasticity-robust standard errors). * indicates p<0.05, ** indicates p<0.01.

26

Table 8: Determinants of the happiness gender gap: all
(1)
Regressors
Women Economic Rights

-0.12
(-0.55)

Women Rights

School Enrollment Ratio
Life Expec. Ratio
Fertility Rate
Men more right to job
Men better P. Leaders
Univ. more imp. boys
Housewife fulfilling
Constant
R2
R2

(4)

-0.79
(-1.81)

Women Social Rights

Woman in Parliament

(3)

-0.09
(-0.30)

Women Political Rights

Female Labor Force Part

(2)

-3.14*
(-2.27)
-0.60
(-0.44)
2.95
(1.95)
1.89
(0.54)
0.09
(1.02)
1.08
(1.70)
0.07
(0.14)
-0.60
(-1.22)
0.07
(0.19)
-3.17
(-0.76)
0.29
71

-0.31
(-0.87)
-2.55
-3.00* -2.91*
(-1.84) (-2.12) (-2.07)
0.79
-0.47
-0.20
(0.51) (-0.33) (-0.14)
3.29*
2.91
2.97
(2.22) (1.93) (1.98)
1.10
1.84
1.74
(0.32) (0.53) (0.50)
0.08
0.08
0.07
(0.94) (0.86) (0.72)
0.99
1.10
1.10
(1.61) (1.72) (1.74)
0.04
0.03
0.01
(0.08) (0.06) (0.01)
-0.59
-0.60
-0.63
(-1.27) (-1.23) (-1.31)
0.19
0.08
0.12
(0.54) (0.22) (0.32)
-1.58
-3.09
-2.84
(-0.38) (-0.74) (-0.68)
0.33
0.29
0.30
71
71
71

Note: Dependent variable: Effect of on Life Satisfaction. t-statistics reported in brackets
(heteroskedasticity-robust standard errors). * indicates p<0.05, ** indicates p<0.01.

27

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.