Content uploaded by James Thomas Bang
Author content
All content in this area was uploaded by James Thomas Bang on Apr 15, 2015
Content may be subject to copyright.
GENDER EQUALITY AND ECONOMIC GROWTH:
IS IT OPPORTUNITY OR OUTCOMES?
[PRELIMINARY DRAFT – DO NOT CITE]
ABSTRACT
This article explores the impact of gender equality on economic growth. In particular, we
focus on the multidimensional nature of gender equality with the object of identifying the
relative salience of different aspects of equality. Using exploratory factor analysis on five
measures of gender equality, we identify two distinct dimensions: equality in the access to
economic opportunities and equality in economic and political outcomes. Regression analysis
conducted on an unbalanced panel of 101 countries taken over non-overlapping five-year
periods from 1990-2000 reveals that a standard deviation improvement in opportunities
increases growth by 1.3 percentage points and a corresponding improvement in equality of
outcomes improves growth by an average of about 1.2 percentage points. However, this
impact is contingent on a country’s stage of development: While developing economies
experience significant improvements in growth from greater equality in opportunities,
developed societies see significant improvements resulting from greater equality in
outcomes.
INTRODUCTION
The impact of gender disparity on economic growth has emerged as an important area of inquiry
in the last two decades and there is considerable evidence that gender inequality in education (M.
Anne Hill and Elizabeth King, 1995; David Dollar and Roberta Gatti, 1999; Stephan Klasen,
1
1999, 2002; Stephen Knowles, Paula K. Lorgelly, and P. Dorian Owen, 2002; Stephan Klasen
and Francesca Lamanna, 2009), has acted as a significant impediment to economic growth.
Perhaps for lack of reliable data, the growth impact of gender inequality in employment remains
less explored in a cross-national context. In particular, there is a lack of consensus on whether
such inequality, especially considered in the form of the gender wage gap, constitutes an
impediment to growth: while Stephanie Seguino (2000) finds a positive impact of gender wage
inequality in the manufacturing sector on economic growth in a sample of semi-industrialized
export- oriented economies, Thomas Schober and Rudolph Winter-Ebmer (2011) based on a
meta-analysis of micro-level wage data fails to confirm the result.
Moving beyond the context of semi-industrialized economies, Klasen and Lamanna
(2009) use multiple indicators such as the female share of the labor force and the labor force
participation gap to capture gender inequality in unemployment and present compelling evidence
that inequality takes a significant toll on growth. To sum up, therefore, existing evidence on the
question appears to be sensitive both to the choice of indicator for gender inequality in
employment and to the construction of the sample.
This paper investigates the impact of gender equality on economic growth in an
unbalanced sample of 101 countries observed over the period 1990-2000. Our analysis is based
on the premise that gender equality is multidimensional and the various aspects may well differ
in their consequences for economic growth. Hence, the dominant convention of focusing on the
growth impact of a specific aspect of gender inequality, such as education or employment, is
likely to subject the estimates to omitted variable bias. At the same time, any exploration of the
consequences of multiple dimensions of gender equality in a unified empirical model is
challenging given the degree of collinearity exhibited by various indicators of gender equality.
2
Compounding the problem, the potentially differential impacts of various dimensions of gender
equality render the project of aggregating them into a composite index such as the Global Gender
Gap Index (GGI) of Ricardo Hausmann, Laura D. Tyson and Saadia Zahidi (2012) subject to
measurement error.
To address the multidimensionality of gender equality, we conduct an exploratory factor
analysis (EFA) on five distinct indicators, namely, the fertility rate; the percentage of women in
parliament; and the gender gaps in literacy, secondary enrolment and labor force participation.
The procedure reveals two latent dimensions of gender equality that we interpret as equality in
access to economic opportunity and equality in economic and political outcomes. Since the
factors are, by construction, free of high degrees of collinearity, we are able to include both
dimensions of equality simultaneously in our growth regressions, thereby avoiding omitted
variable bias, multicollinearity, and measurement error.
Our regression analysis reveals that the two dimensions of gender equality differ in their
impacts on economic growth: on average, a standard deviation improvement in the equality of
opportunity increases growth by 1.30 percentage points, while a corresponding improvement in
the equality of outcomes improves growth by an average of 1.19 percentage points. However,
these impacts are contingent on a country’s stage of development: while developing economies
see significant improvements in growth from greater equality in opportunities, developed
societies experience significant improvements from greater equality in outcomes.
3
CONCEPTUAL FOUNDATIONS
We devote this section to a brief account of why an improvement in gender inequality may
constitute a significant impetus to growth. To do so, we first focus on the impact of gender
inequality in education.
Gender Inequality in Education
As observed by Klasen (1999, 2002), gender inequality in education reduces the average quality
of human capital in an economy. Assuming that the intrinsic cost of skill investment is
identically distributed over both sexes, restrictions on female educational opportunities substitute
females of low intrinsic cost by males of relatively higher cost. As a result, the average cost of
skill investment rises for any given level of human capital, indicating a decline in the average
quality of human capital relative to the state of equal opportunity. Hence, an improvement in
gender inequality in education is predicted to improve the quality of human capital available and
consequently, the rate of growth.
Further, if we consider male and female education to be separate entities and assume that
the marginal impact of both types of human capital on economic growth is subject to diminishing
returns, gender inequality means that marginal returns to education are greater for women than
they are for men. Hence, an improvement in gender equality that increases the educational
attainment of women should increase growth (Knowles, Lorgelly and Owen, 2002).
Critically, an increase in female educational attainment due to improving gender equality
provides a significant impetus to growth via an interdependent set of externalities, namely, an
improvement in various indicators of child health; a reduction in fertility; and an improvement in
the level and quality of human capital of future generations.
4
Consider first the impact on child health. Intuitively, the impact of maternal education is
not hard to see: a more educated mother is expected to be aware of the best nutritional and
medical regime for her and her children, be more likely to adhere to it, and be able to access
modern medical services. If we consider child health as the output of a production function, this
yields two complementary impacts of maternal education on the health of children: a more
educated mother will utilize a given level of health inputs more efficiently (Michael Grossman,
2000), and she will be able to avail her family of the most efficient combination of inputs (Mark
R. Rosenzweig and T. Paul Schultz, 1982).
There is a wealth of evidence that identifies the health of children as a key determinant of
educational attainment in a society (see Paul Glewwe and Edward A. Miguel, 2008 and
references therein). Thus, the causal impact of female education on growth via the accumulation
of human capital is not hard to establish. Reducing child mortality should also reduce the
incentive to have a large number of children. Indeed, there is a great deal of evidence that female
education plays a critical role in reducing fertility by decreasing the desired family size; reducing
the need to plan a greater number of births to achieve the desired family size; and increasing the
ability to implement the planned number of births (Mamta Murthi, Anne-Catherine Guio and
Jean Drèze, 1995; Jean Drèze and Mamta Murthi, 2001; Una Okonkwo Osili and Bridget Terry
Long, 2008).
This again should enhance growth by increasing the amount of capital per worker (Oded
Galor and David N. Weil, 1996); reducing the youth dependency ratio (Allen C. Kelley and
Robert M. Schmidt, 2001); and improving the productivity of the labor force by allowing parents
to devote a greater fraction of familial resources to each child (Gary S. Becker, Kevin M.
Murphy and Robert F. Tamura, 1994).
5
Gender Inequality in Employment
As previously mentioned, the growth impact of the employment dimension of gender equality
may be sensitive to both the choice of indicator and the economy in question. Consider first an
improvement in the gender gap in labor force participation. Analogous to the argument made for
gender inequality in education (Klasen, 1999; 2002), the systemic exclusion of women from the
labor market reduces the productivity of the labor force by substituting more productive female
workers with male workers of relatively lower productivity. As such, an improvement in female
labor force participation should increase labor productivity (Berta Esteve-Volart, 2004) and
hence, the rate of growth. Further, improved prospects for female employment should reduce
fertility by increasing the opportunity cost of childbearing (Galor and Weil, 1996), which, in
turn, should enhance growth for the reasons stated earlier.
A key impact of increasing labor force participation on economic growth by women
operates via reducing the gender asymmetries in education. Note that an increase in female
employment in the economy should, on average, lead to a rise in household income. This should
reduce the incentive to prioritize the male child in the resource allocation decisions of the
household.
1
As importantly, increasing the expected returns from educating the girl child should
reduce gender asymmetries in familial investment on children (Gary S. Becker, 1985).
The increased probability of employment increases the bargaining power of women
within the household and facilitates the erosion of traditional norms that legitimize male
dominance over labor and sexuality (Ester Boserup, 1970; Claudia Goldin, 1990; Stephanie
Seguino 2007). There is evidence that this renegotiation of bargaining power in favor of women
leads to reduced fertility, greater household saving, a greater fraction of income being diverted to
6
investment in health and education; and reduced exclusion of the girl child from familial
investment (Esther Duflo, 2012). Each of these provides an impetus to growth.
However, it is not clear if an improvement in gender equality in employment will
necessarily stimulate growth once we measure it with the gender wage gap. On one hand, a
reduction in the gender wage gap may stimulate female labor force participation (Galor and
Weil, 1996) and increase the rate of growth for reasons described above. On the other hand, such
an improvement may impede growth for export-oriented semi-industrialized economies that rely
on low-paid female labor in export sectors to acquire a competitive edge in the international
market (Robert Blecker and Seguino, 2002). Indeed, the influential study by Seguino (2000)
presents compelling evidence supporting this argument.
Schober and Winter-Ebmer (2011) have criticized the results obtained by Seguino (2000)
on the grounds that the aggregate earnings data used to construct the gender wage gap does not
accurately measure wage discrimination, a proper identification of which requires micro-level
wage data. Relying on a cross-national dataset constructed from a meta-analysis of Blinder-
Oaxaca decomposed micro data, Schober and Winter-Ebmer find no evidence of a negative
impact of wage equality on growth. The debate, however, is far from resolved, since the
rejoinder by Stephanie Seguino (2011) has questioned both the applicability of the meta-
regression methodology Schober and Winter-Ebmer use and their decision to extend the data
coverage beyond the manufacturing sector, given Seguino’s specific objective of estimating the
growth impact of wage inequality in semi-industrialized economies.
Gender Inequality in Political Participation
7
There is an emerging consensus that increasing political participation by women may have
significant benefits, though it should be acknowledged that most empirical studies, including our
own, focus on a specific aspect of political outcomes, namely, female presence in government.
Various mechanisms have been unearthed, not the least of which being the fact that
increased female presence in government can, at least partially, remedy gender bias in public
policy. In a panel study of American states, Timothy Besley and Anna Case (2003) find that
greater presence of women in state legislatures increases public expenditure on family assistance
programs and leads to greater enforcement of child support laws. Critically for our purpose, this
is not exclusively a feature of female political outcomes in developed societies: Raghabendra
Chattopadhyay and Esther Duflo (2004) find that reserving a third of the seats in self-
governments at the village level significantly increased infrastructural investment specifically
needed by women in the Indian states of West Bengal and Rajasthan.
There is also reason to believe that greater female presence in legislative bodies may alter
the composition of public expenditure in favor of investment in health and education.
2
M. Marit
Rehavi (2008) finds that the increase in the number of women in American state legislatures
from 1970-2000 accounts for a modest but significant share of the increase in state health
expenditures. Helena Svaleryd (2009) finds that female representation in Swedish local councils
increases the expenditure on education and child care relative to that on caring for the elderly.
3
Remarkably, Irma Clots-Figuerasa (2012) links female presence in government with educational
outcomes: in a study of district-level data from India, increasing female representation in urban
districts increases the probability with which an individual from that district attains primary
education.
8
Finally, a key impact of female political representation on economic growth operates via
improving the quality of domestic institutions. There is evidence that gender equality in political
representation reduces corruption (David Dollar, Raymond Fisman and Roberta Gatti, 2001;
Anand Swamy, Stephen Knack, Young Lee and Omar Azfar, 2001). There is also evidence from
Klaus Deininger, Songqin Jing, Hari K. Nagarajan, and Xia Fang (2011) that it improves popular
perception of government accountability at the local level, and, consequently or otherwise, the
willingness to contribute to public goods. Further, it has been documented that political
representation for women improves the quality of the judicial system in terms of increasing the
ability of women to report crimes (Lakshmi Iyer, Anandi Mani, Prachi Mishra, and Petia
Topaleva, 2011). Finally, there is increasing evidence that women in government reduces the
likelihood of collective violence in the form of state-sponsored human rights abuses (Eric
Melander, 2005a), civil war (Eric Melander 2005b), and interstate conflict (Patrick M. Regan
and Aida Paskeviciute, 2003).
DATA AND METHODOLOGY
We test the impact of gender equality on economic growth in the context of a standard
neoclassical model (Robert J. Barro, 1991; Robert J. Barro and Xavier Sala-i-Martin, 1995)
augmented with measures of gender equality:
(1)
To calculate the dependent variable, , we first calculate the annual growth rate
of gross domestic product (GDP) per capita in each country for each year in our sample.
4
We
then calculate the average of this variable over five-year intervals. We correspondingly calculate
9
the five-year averages for each of the explanatory variables on the right-hand side of equation (1)
and keep only the non-overlapping five-year averages corresponding to the years that are evenly
divisible by 5.
5
Our final sample covers a sample of 101 countries forming an unbalanced panel
of 228 observations for the five-year periods ending in 1990, 1995, and 2000. Table 1 reports
summary statistics for our sample.
6
Standard Correlates of Economic Growth
The first set of controls includes the natural logarithm of per capita GDP at the end of the
previous period; the growth rate of per capita GDP over the previous period; investment,
measured by gross capital formation (GCF) as a percentage of GDP; government consumption
net of defense and education expenditure as a percentage of GDP; the consumer price index
(CPI) inflation rate; and the secondary completion rate from Robert J. Barro and Jong-Wha Lee
(2001) to capture the average level of human capital in a country.
Measures of Gender Equality
We utilize information from five distinct indicators of gender equality.
7
The gender gaps in
literacy and secondary enrollment from the World Development Indicators (WDI) capture
constraints on skill investment arising from a social ethos that prioritizes the male child in
education (Jean Drèze and Geeta G. Kingdon, 2001; Geeta G. Kingdon, 2005; Monazza Aslam
and Geeta G. Kingdon, 2008). The gender gap in labor force participation, also taken from the
WDI, captures the consequences of biased resource allocation in addition to restricted female
employment opportunities (Jean Drèze and Amartya Sen, 1995; Lourdes Benería, 2001),
discrimination in the labor market (Berta Esteve-Volart, 2004), reduced mobility of female
workers, and social norms that prioritize fertility over professional attainment (Kaivan Munshi
10
and Jacques Myaux, 2006). The inverse of the adolescent fertility rate, also from WDI, explicitly
captures the fertility aspect of gender bias.
8
Lastly, the percentage of women in parliament from
the Women in National Parliaments Dataset released by the Inter-Parliamentary Union captures
the voice of women in the design and implementation of social policy.
9
Methodological Concerns
Estimating equation (1) confronts us with a number of concerns, the first relating to the
measurement of gender equality. Most cross-national studies on the impact of gender inequality
on economic growth focus on a specific aspect of such inequality, such as education (Hill and
King, 1995; Dollar and Gatti, 1999; Klasen, 2002; Knowles, Lorgelly and Owen, 2002; Mina
Baliamoune-Lutz and Mark McGillivray, 2009) or employment (Seguino, 2000; Schober and
Winter-Ebmer, 2011). While there are contributions that explore the consequences of more than
one aspect of inequality (Klasen, 1999; Mark Blackden, Sudharshan Canagarajah, Stephan
Klasen and David Lawson, 2006; Klasen and Lamanna, 2009), these dimensions are generally
considered in separate regressions.
However, if one believes that an ethos of gender bias in society manifests itself along
multiple dimensions, then focusing on one dimension is likely to understate the impact of gender
stratification. In particular, restricting the set of measures of gender disparity is likely to subject
the estimated impacts to omitted variable bias (James T. Bang and Aniruddha Mitra, 2011b).
Accounting for multiple aspects of gender inequality in a single empirical model faces
the problem that these variables exhibit a high degree of collinearity, for example, the correlation
between the fertility rate and the gender gap in education. As mentioned in the last section, there
is evidence that increasing female educational attainment reduces fertility. At the same time, a
decline in fertility may reduce the gender gap in education. Rodrigo R. Soares and Bruno L. S.
11
Falcão (2008), for example, consider the impact of a reduction in mortality due to technological
progress. On one hand, this should increase the expected returns to human capital investment for
both genders. On the other hand, the reduced incentive to have a large family and the associated
decline in fertility should mobilize women into the labor force. The resultant increase in the
bargaining power of women should then reduce the gender gap in education.
One may, of course, address the problem of multicollinearity by aggregating various
indicators of gender equality into a unidimensional index, as has been the practice in the
literature on formal institutions (Stephen Knack and Philip Keefer, 1995; Alberto Alesina and
Roberto Perotti, 1996; Roberto Perotti 1996).
10
Indeed, existing indices such as the Gender
Empowerment Measure (GEM), the Gender-Related Development Index (GDI) and the GGI are
based on precisely this convention.
However, as observed by Richard Jong-A-Pin (2009) in the context of political instability
and Bang and Mitra (2011a) in the context of institutions, the limitation of this method this is
that the aspects of gender equality being aggregated may differ in their impacts on economic
growth. Hence, the composite index will likely be subject to measurement error.
11
As such, we
follow Bang and Mitra (2011b) in conducting an EFA on the set of gender variables. This allows
us to identify two distinct dimensions of gender equality that are essentially uncorrelated. These
dimensions are included in the vector Zit. We provide a brief description of the EFA and
robustness checks we perform on it in the next section.
Turning to the estimation procedure, any growth regression must address the fact that the
classical least squares estimator is likely to be biased if any of the explanatory variables are
endogenously determined by the same factors that determine growth. Further, any panel study
12
faces the problem of serial correlation within panels and unobservable entity-specific
heterogeneity across panels.
The problem of endogeneity is particularly relevant in our context since there is a
substantial literature that investigates the role of development in promoting gender equality. To
review the proposed mechanisms, recall that the rise in household income due to development
should reduce the incentive to prioritize the male child in the allocation of household resources.
Further, the improvement in employment prospects for women that accompanies development
increases the returns to female education and reduces gender bias in familial investment in
education (Becker, 1985).
Improved prospects for employment and the diffusion of productivity-augmenting
technology that allows women to devote less time to domestic duties and seek formal
employment (Roger D. Clark, Thomas W. Ramsbey and Emily S. Adler, 1991) also allows
women to renegotiate power relations within the family (Boserup, 1970; Goldin, 1990). The
literature identifies this increase in bargaining power of women as a key contributor to
empowerment.
Finally, it has been argued that as economic development expands education, it also
transforms a society from a traditional culture that emphasizes physical and economic security in
favor of one that promotes postmaterialist values that encourages gender equality (Ronald
Inglehart, 1997; Ronald Inglehart and Pippa Norris, 2003).
The traditional response to endogeneity in growth regressions is to employ an
instrumental variables technique, using geographic or institutional variables as instruments.
However, this method has recently come under criticism based on the fact that many of the
13
commonly-used instruments are of dubious strength or validity (Michael P. Murray, 2006;
Samuel Bazzi and Michael A. Clemens, 2013).
To address the problem of endogeneity simultaneously with that of unobserved
heterogeneity, we employ the difference-GMM estimator of Manuel Arellano and Stephen Bond
(1991), which estimates the dynamic model in first differences, instrumenting for current-period
differences in the endogenous variables with their lagged values. While the estimator does have
its own caveats (David Roodman, 2009; Bazzi and Clemens, 2013), the fact that it utilizes a
greater number of exclusion restrictions compared to the two stage least squares model has made
it a staple in empirical studies on economic growth.
12
Our response to the problem of serial correlation within panels employs a twofold
strategy. First, we consider non-overlapping five-year averages of our variables in order to filter
out the serial correlation in growth rates arising from short-term fluctuations attributable to
changes in the business cycle. Second, we include lagged values of the growth rate of per capita
GDP to account for autocorrelation in the dependent variable as well as lagged values of log per
capita GDP to account for the convergence hypothesis of the neoclassical growth model (Barro,
1991; N. Gregory Mankiw, David Romer, and David N. Weil, 1992).
Finally, since the gender variables of interest are generated factor scores, the standard
errors of our coefficients may also be biased (Adrian Pagan, 1984). A common strategy to deal
with this problem is to employ a bootstrap technique. However, given the dynamic structure of
the model used to calculate the difference-GMM estimator, generating bootstrap samples that
can be considered to be random presents a challenge (Roodman, 2003). Hence, we calculate
heteroskedasticity-consistent robust standard errors using a jackknife technique (Russell
14
Davidson and James G. McKinnon, 1993) that includes clustering by country to account for
heteroskedasticity across panels.
MULTIDIMENSIONALITY OF GENDER EQUALITY
The methodology of EFA is based on the assumption that each of a set of potentially correlated
variables is generated by a linear combination of a smaller set of latent factors and an error term.
The hypothesized latent factors include common factors that impact more than one observed
variable and specific factors that are unique to each variable. Hence, variation in each observed
variable can be decomposed into the part caused by variation in the common factors and the part
unique to the variable in the form of specific factors and measurement error.
13
The value of EFA lies in its ability to explore a theoretical structure underlying
multivariate data, since the common factors identified by the method ideally lend themselves to
theoretical interpretation. Further, since the factors emerge from an optimization process, they
are less susceptible to measurement bias than indices constructed on the basis of subjective
assignment of weights to the constituent variables. In addition, being extracted by identifying
common sources of variation in the observed variables, the factors are free of high degrees of
multicollinearity. This allows us to include multiple dimensions of gender equality
simultaneously in the growth regression, thus avoiding the problem of omitted variable bias.
The EFA conducted on the gender variables employs the principal factor extraction
method with a promax rotation procedure and factor loadings from the exercise are reported in
Table 2.
14
Two common factors underlying the observed variables emerge from the analysis. Of
these, the first is interpreted as reflecting equality in economic opportunity and the second as
15
capturing equality in economic and political outcomes.
15
In the remainder of this section, we
clarify the interpretations of these factors.
The first factor is primarily composed of the inverse of the fertility rate, which acquires a
rotated factor loading of 0.663, and the secondary enrolment gap, with a loading of 0.652. The
gender gap in secondary enrollment directly reflects constrained female opportunities for skill
investment while the fertility rate implies the existence of cultural norms that limit female access
to education and employment opportunities. As such, we interpret this factor to represent gender
equality in economic opportunity. The second factor is determined by the percentage of women
in parliament (0.566) and the gender gap in labor force participation (0.563). These two variables
may be regarded as capturing complementary dimensions of outcome equality. Hence, we
interpret this factor as representing gender equality in economic and political outcomes.
Two points bear clarification in this context. First, in observing that a dimension of
gender equality is primarily composed of a set of variables, we are not asserting that this factor is
solely composed of these variables. The methodology of factor analysis is based on the premise
that each of the variables reflects an impact from all of the underlying latent factors. Hence, the
factor loading of 0.566 on the percentage of women in parliament in the outcomes factor is
capturing the impact of this variable that is uncorrelated with the opportunity factor. Second, it
may appear surprising that the gender gap in literacy does not contribute significantly to either of
the factors. However, definitions of literacy vary widely across countries, and even within the
same country, have evolved over time. As such, we are unsure if we would have expected any
meaningful contribution from this variable.
16
Finally, Table 3 provides an idea of how countries in our sample rank with respect to the
two dimensions of gender equality in 2000. Given the legacy of the universal education policy
16
under State Socialism, it is not surprising that Bulgaria, Estonia, and Latvia rank among the top
five in equality of opportunity. However, despite the history of state commitment to guaranteed
employment for both genders, none of these countries appear to have performed as satisfactorily
in terms of gender equality in economic and political outcomes. Indeed, a 2010 report by the
Confederation of Independent Bulgarian Unions (KNSB) finds that the average salary of women
is likely to be 15.7% lower than that of men. Further, the employment rate for women aged 20-
64 stands at only 58.3%, in contrast to 66.9% for men. The example of Bulgaria highlights the
need to distinguish between different aspects of equality. More importantly, it adds a note of
caution to the discourse on empowerment by indicating that progress on the path to gender
equality has not been uniform with respect to the different dimensions of equality in many
societies.
RESULTS
The difference-GMM estimation of equation (1) reported in column 1 of Table 4 confirms our
hypothesis that various aspects of gender equality have a significant impact on economic growth:
a standard deviation improvement in economic opportunities increases the rate of growth by 1.30
percentage points and the impact is significant at the 0.05 level. Similarly, a standard deviation
improvement in the outcomes dimension of gender equality increases economic growth by 1.19
percentage points and the impact is again significant at the 0.05 level.
The standard correlates of growth have the signs predicted by theory. Consistent with the
existing literature and the convergence hypothesis, the lagged log of per capita GDP negatively
and significantly impacts growth at the 0.01 level. Also consistent with the existing literature, the
lagged growth rate of per capita GDP has a negative and significant impact, and investment has a
17
positive and significant impact. We also observe a negative and significant effect for net
government consumption that has been documented by Barro (1991), but the lack of consensus
regarding the impact of this variable (Niloy Bose, M. Emranul Haque and Denise Osborn, 2007)
prevents us from reading too much into this result.
17
Finally, the inflation rate also exhibits the
predicted negative sign but fails to achieve statistical significance.
In addition, the various specification tests for the Arellano-Bond model remain with the
bounds necessary to conclude that the assumptions of no second-order serial correlation in the
errors and that of the validity of the instruments hold. To investigate the first assumption, we
report the Arellano-Bond test statistic for first-order serial correlation in the first-differenced
errors. In each case, this statistic fails to reject the null hypothesis of zero first-order correlation
at the 0.05 level. Note, however, that since our panel is only three periods long, testing for
second order serial-correlation is not possible. However, given the fact that there is such weak
evidence of first-order serial correlation, we find it very unlikely that second-order correlation
could exist. Secondly, since Arellano-Bond estimators generate large numbers of instruments, it
is often the case that these models result in over-identification. To test for this, we report the
Hansen J statistic for over-identification. In each of our specifications, this statistic passes the
criteria for no over-identification problem (fails to reject the null). Hence, we conclude that the
problem of excessive instruments and over-identification is not too serious in our specification.
At first blush, the results presented in column 1 may appear to imply that both
dimensions of gender equality are of approximately equal importance for economic growth. To
explore this further, we now test whether the impacts of various dimensions of gender equality
on economic growth depend on the stage of development of a society. To do so, we distinguish
between countries that are members of the Organization for Economic Cooperation and
18
Development (OECD) and ones that are not, the rationale being that membership in the OECD is
likely to correlate with a relatively higher stage of economic and institutional development.
18
Accordingly, we first estimate our model for the subsample of countries that belong to
the OECD. As seen from column 2 in Table 4, it is the outcomes dimension of gender equality
alone that has significantly impacted economic growth over the period in question: while a
standard deviation improvement in the outcomes factor increased the rate of growth by about
1.51 percentage points at the 0.10 level, equality in economic opportunities did so by about 0.55
percentage points only and the impact is statistically insignificant.
Interestingly, the relative salience of the two aspects of gender equality appears to reverse
for countries that are not members of the OECD: as seen from column 3, a standard deviation
increase in the equality of opportunity improves growth by about 1.38 percentage points and the
impact is significant at the 0.10 level. By contrast, an improvement in the outcomes dimension
has a small positive impact of about 0.31 percentage point but fails to acquire significance to any
acceptable level.
The last exercise in this set addresses the dichotomy between the opportunity and
outcome dimensions in a unified model by introducing a dummy variable which takes the value
1 if a country is a member of the OECD and 0 if it is not and considering interaction effects of
the dummy with the two dimensions of gender equality.
19
As before, the impact of outcome
equality is only significant for the subset of OECD members, with a standard deviation
improvement in equal outcomes increasing growth by 1.91 percentage points on average and the
impact being significant at the 0.05 level. At the same time, the opportunity dimension of gender
equality significantly impacts growth for non-OECD countries exclusively, with a standard
deviation increase in the factor resulting in a 1.26 percentage point increase in the rate of growth.
19
It, therefore, appears that the opportunity dimension of gender equality is of greater
salience as a determinant of economic growth for developing nations while it is the outcomes
dimension that plays a greater role in stimulating growth for societies that have attained a
threshold level of economic and institutional development. However, it should be clarified that in
emphasizing the relative salience of the opportunity dimension of gender equality in developing
countries, we are not claiming that variables such as the gender gap in labor force participation
or the percentage of women in parliament, which contribute significantly to the outcomes
dimension, do not have an important impact on growth. Rather, insofar as these variables
positively contribute to both dimensions of gender equality, our point is that these variables
stimulate growth more by improving female access to education in particular and economic
opportunity in general than by improving outcomes in the economic and political sphere.
ROBUSTNESS
Since conventional gender-aggregated measures of human capital are strongly correlated with
our variables of interest, it is natural to ask if our results are being confounded by the eventuality
that female access to education is too highly correlated with the overall level of opportunity and
hence, the gross secondary completion rate. To address this concern, we follow Klasen and
Lamanna (2009) in replacing the latter with the secondary completion rate for males.
Interestingly, the male secondary completion rate fails to achieve statistical significance at the
5% level or better in the results of this exercise.
20
We have also re-estimated the model by
excluding the secondary completion rate altogether. With respect to the gender equality
variables, the results of each of these robustness checks confirm the basic flavor of the analysis
conducted in the previous section.
21
20
We now return to our initial specification to explore whether our results are robust to our
choice of controls. Accordingly, we begin by testing whether the observed impacts of gender
equality are robust to the consideration of a nonlinear quadratic impact of inflation (Jenny
Minier, 2007; Robert Pollin and Andong Zhu, 2006) and trade openness, defined as the sum of
exports and imports expressed as a percentage of GDP (Jeffrey D. Sachs and Andrew M.
Warner, 1995; Jeffrey Frankel and David Romer, 1999; David Dollar and Aart Kraay, 2004).
22
Once again, the basic results with respect to relative impacts of equal opportunity and equal
outcomes prove to be robust to the inclusion of these two variables.
23
Finally, note that the results reported so far have been obtained on the basis of an
unbalanced panel. As such, it may be asked if they are biased since the estimation weighs
countries with a full complement of observations over the entire time series two or three times as
heavily as the countries for which we only observe one time period. To allay these concerns, we
re-estimate the model with a balanced panel of 45 countries, including 22 OECD countries and
23 non-OECD countries.
24
If anything, the results of this exercise sharpen the dichotomy
between the opportunity and outcomes dimensions of gender equality, and also between OECD
and non-OECD members.
25
CONCLUSION
This article investigated the consequences of gender equality on the growth experience of
nations. An EFA conducted on five indicators of gender equality revealed two latent dimensions
of gender equality, namely, equality in the access to economic opportunity and equality in
economic and political outcomes. Confirming our hypothesis on the differential impacts of
various dimensions of gender equality on economic growth, subsequent regression analysis
21
found a robust positive impact of the opportunity dimension for developing societies, while the
impact of the outcomes dimension was found to be significant for countries that had already
attained a threshold level of development.
In addressing the multidimensionality of gender equality, the paper provides a more
nuanced analysis of the role of the gender equality as a determinant of economic growth. No
study on the topic of gender can understate the importance of attaining greater equality in the
opportunity to education and this is indeed a conclusion that emerges from our analysis. Yet a
key implication of our study is that this is not all that needs to be done. In fact, the focus of
policy intervention to address the problem of gender inequality should be contingent on the stage
of development of a society: a developing nation embarking on a growth trajectory will benefit
more from policy intervention directed at improving female access to education and economic
opportunity. Once a threshold level of economic and institutional development has been
achieved, however, it would benefit more from policies directed at ensuring equality in outcomes
in economic activity and in the political sphere.
REFERENCES
Alesina, Alberto and Roberto Perotti. 1996. "Income Distribution, Political Instability and
Investment." European Economic Review 40(6): 1203-1228.
Arellano, Manuel, and Stephen Bond. 1991. "Some Tests of Specification for Panel Data: Monte
Carlo Evidence and an Application to Employment Equations." Review of Economic
Studies 58(2): 277-297.
Aslam, Monazza and Geeta G. Kingdon. 2008. "Gender and Household Education Expenditure
in Pakistan." Applied Economics 40(20): 2573-2591.
22
Baliamoune-Lutz, Mina and Mark McGillivray. 2009. "Does Gender Inequality Reduce Growth
in Sub - Saharan African and Arab Countries?" African Development Review 21(2): 224-
242.
Bang, James T. and Aniruddha Mitra. 2011a. "Brain Drain and Institutions of Governance:
Educational Attainment of Immigrants to the US 1988-1998." Economic Systems 35(3):
335-354.
----. 2011b. "Gender Bias and the Female Brain Drain." Applied Economics Letters 18(9): 829-
833.
Barro, Robert J., 1991. "Economic Growth in a Cross Section of Countries." Quarterly Journal
of Economics 106(2): 407-443.
Barro, Robert J. and Jong-Wha Lee. 2001. "International Data on Educational Attainment:
Updates and Implications." Oxford Economic Papers 53(3): 541-563.
Barro, Robert J. and Xavier Sala-i-Martin. 1995. Economic Growth, (New York: McGraw-Hill).
Bazzi, Samuel and Michael A. Clemens. 2013. "Blunt Instruments: Avoiding Common Pitfalls in
Identifying the Causes of Economic Growth." American Economic Journal:
Macroeconomics 5(2): 152-186.
Becker, Gary S. 1985. "Human capital, effort, and the sexual division of labor." Journal of Labor
Economics 3(1): S33-S58
Becker, Gary S., Kevin M. Murphy, and Robert Tamura. 1994. "Human Capital, Fertility, and
Economic Growth." In Human Capital: A Theoretical and Empirical Analysis with Special
Reference to Education (3rd Edition), pp. 323-350. The University of Chicago Press, 1994.
Benería, Lourdes. 2001. "Shifting the Risk: New Employment Patterns, Informalization and
Women’s Work." International Journal of Politics, Culture and Society 15(1): 27-53.
23
Besley, Timothy and Anne Case. 2003. "Political Institutions and Policy Choices: Evidence from
the United States. Journal of Economic Literature 41: 7-73.
Blackden, Mark, Sudharshan Canagarajah, Stephan Klasen and David Lawson. 2006. "Gender
and Growth in Sub-Saharan Africa: Issues and Evidence" Research Paper No. 2006/37,
UNU-WIDER, United Nations University.
Blecker, Robert and Stephanie Seguino. 2002. "Macroeconomic Effect of Reducing Gender
Wage Inequality in an Export Oriented, Semi-Industrialized Economy." Review of
Development Economics 6(1):103-119.
Bose. Niloy, M. Emranul Haque, and Denise R. Osborn. 2007. "Public Expenditure and
Economic Growth: A Disaggregated Analysis for Developing Countries." Manchester
School 75(5): 533-556.
Boserup, Ester. 1970. Women’s Role in Economic Development (New York City: St. Martin’s
Press).
Bruno, Michael and William Easterly. 1998. "Inflation Crises and Long-Run Growth." Journal
of Monetary Economics 41(1): 3-26.
Caselli, Francesco, Gerardo Esquivel, and Fernando Lefort. 1996. "Reopening the Convergence
Debate: A New Look at Cross-Country Growth Empirics." Journal of Economic Growth
1(3): 363-389.
Chattopadhyay, Raghabendra and Esther Duflo. 2004. "Women as Policy Makers: Evidence
from a Randomized Policy Experiment in India." Econometrica 72(5): 1409-1443.
Clark, Roger D., Thomas W. Ramsbey and Emily S. Adler.1991. "Culture, Gender, and Labor
Force Participation: A Cross-National Study. " Gender and Society 5(1): 47–66.
24
Clots-Figuerasa, Irma. 2012. "Are Female Leaders Good for Education?" American Economic
Journal: Applied Economics 4(1): 212-244.
Costello, Anna B. and Jason W. Osborne. 200. "Best Practices in Exploratory Factor Analysis:
Four Recommendations for Getting the Most from your Analysis." Practical Assessment,
Research and Evaluation 10(7): 1-9.
Coulombe, Serge and Jean-François Tremblay. 2006. "Literacy and Growth." B.E. Journal of
Macroeconomics 6(2): Article 4.
Davidson, Russell, and James G. MacKinnon. 1993. Estimation and inference in econometrics
(Oxford: Oxford University Press).
Deininger, Klaus, Songqin Jing, Hari K. Nagarajan and Xia Fang. 2011. "Does Female
Reservation Affect Long-Term Political Outcomes? Evidence from Rural India." Policy
Research Working Paper 5708 (Washington, DC: The World Bank).
Dijkstra, A. Geske. 2002. "Revisiting UNDP's GDI and GEM: Towards an Alternative." Social
Indicators Research 57(3): 301-338.
Dollar, David and Roberta Gatti. 1999. "Gender Inequality, Income and Growth: Are Good
Times Good for Women?" Policy Research Report on Gender and Development Working
Paper No. 1 (Washington, D.C.: World Bank).
Dollar, David, Raymond Fisman and Roberta Gatti. 2001. "Are Women Really the 'Fairer' Sex?
Corruption and Women in Government." Journal of Economic Behavior and Organization
46(4): 423-429.
Dollar, David and Aart Kraay. 2004. "Trade, Growth, and Poverty." Economic Journal 114(493):
F22-F49.
25
Drèze, Jean and Amartya Sen. 1995. India: Economic development and Social Opportunity
(Oxford and New Delhi: Oxford University Press).
Drèze, Jean and Geeta G. Kingdon. 2001. "School Participation in Rural India." Review of
Development Economics 5(1): 1-24.
Drèze, Jean and Mamta Murthi. 2001. "Fertility, Education and Development: Evidence from
India." Population and Development Review 27(1): 33-63.
Duflo, Esther. 2012. "Women Empowerment and Economic Development." Journal of
Economic Literature 50(4): 1051–1079.
Esteve-Volart, Berta. 2004. "Gender Discrimination and Growth: Theory and Evidence from
India." DEDPS 42 (Suntory and Toyota International Centres for Economics and Related
Disciplines, London: London School of Economics and Political Science).
Ferreira, Fernando and Joseph Gyourko. 2010. "Do Political Parties Matter? Evidence from U.S.
Cities." Quarterly Journal of Economics 124(1): 399-422.
Forbes, Kristin J. 2000. "A Reassessment of the Relationship Between Inequality and Growth."
American Economic Review 90(4): 869-887.
Frankel, Jeffrey and David Romer. 1999. "Does Trade Cause Growth?" American Economic
Review 89(3): 379-399
Galor, Oded and David. N. Weil. 1996. "The Gender Gap, Fertility, and Growth." American
Economic Review 86(3): 374-387.
Glewwe, Paul and Edward A. Miguel. 2008. "The Impact of Child Health and Nutrition on
Education in Less Developed Countries," in T. Paul Schultz and John A. Strauss, eds.
Handbook of Development Economics 4, pp. 3561–3606. Amsterdam: North Holland.
26
Grossman, Michael. 2000. "The Human Capital Model," in Mark V. Pauly, Thomas G. Mcguire
and Pedro Pita Barros, eds. Handbook of Health Economics, 1, pp. 347–408. Amsterdam:
Elsevier.
Goldin, C. 1990. Understanding the Gender Gap: An Economic History of American Women.
New York: Oxford University Press.
Hausmann, Ricardo, Laura D. Tyson and Saadia Zahidi. 2012. The Global Gender Gap Report
2012. Geneva: World Economic Forum.
Hill, M. Anne and Elizabeth King. 1995. "Women’s Education and Economic Well-Being."
Feminist Economics 1(2): 21-46.
Inglehart, Ronald. 1997. Modernization and Post-Modernization: Cultural, Economic, and
Political Change in 43 Societies. Princeton: Princeton University Press.
Inglehart, Ronald and Pippa Norris. 2003. Rising Tide: Gender Equality and Cultural Change
around the World. New York: Cambridge University Press.
Iyer, Lakshmi, Anandi Mani, Prachi Mishra and Petia Topaleva. 2011. "The Power of Political
Voice: Women’s Political Representation and Crime in India." BGIE Unit Working Paper
No. 11-092, Harvard Business School.
Jong-A-Pin, Richard. 2009. "On the Measurement of Political Instability and its Impact on
Economic Growth." European Journal of Political Economy 25(1): 15-29.
Kelley, Allen C. and Robert M. Schmidt. 2001. "Economic and Demographic Change: A
Synthesis of Models, Findings, and Perspectives," in Nancy Birdsall, Allen C. Kelley and
Steven Sinding, eds. Population Matters: Demographic Change, Economic Growth, and
Poverty in the Developing World, pp. 67-105. New York: Oxford University Press.
27
Kingdon, Geeta G. 2005. "Where has All the Bias Aone? Detecting Gender Bias in the
Intrahousehold Allocation of Educational Expenditure." Economic Development and
Cultural Change 53(2): 409-431.
Klasen, Stephan. 1999. "Does Gender Inequality Reduce Growth and Development? Evidence
from Cross-Country Regressions." Policy Research Report on Gender and Development
Working Paper Series No. 2 Washington, D.C.: The World Bank.
----. 2002. "Low Schooling for Girls, Slower Growth for All? Cross-Country Evidence on the
Effect of Gender Inequality in Education on Economic Development." World Bank
Economic Review 16(3): 345-373.
Klasen, Stephan and Francesca Lamanna. 2009. "The Impact of Gender Inequality in Education
and Employment on Economic Growth: New Evidence for a Panel of Countries." Feminist
Economics 15(3): 91-132.
Knack, Stephen and Keefer, Philip. 1995. "Institutions and Economic Performance: Cross-
Country Tests Using Alternative Institutional Measures." Economics and Politics 7(3):
207-227.
Knowles, Stephen, Lorgelly, Paula K., and P. Dorian Owen. 2002. "Are Educational Gender
Gaps a Brake on Economic Development? Some Cross-Country Empirical Evidence."
Oxford Economic Papers 54(1): 118-149.
Langbein, Laura and Stephen Knack. 2010. "The Worldwide Governance Indicators: Six, One,
or None?" Journal of Development Studies 46(2): 350-370.
Lott Jr., John R. and Lawrence W. Kenny. 1999. "Did Women’s Suffrage Change the Size and
Scope of Government?" Journal of Political Economy 107(6): 1163-1198.
28
Mankiw, N. Gregory, Romer, David, and David N. Weil. 1992. "A Contribution to the Empirics
of Economic Growth." Quarterly Journal of Economics 107(2): 407-437.
Melander, Eric. 2005a. "Political Gender Equality and State Human Rights Abuse." Journal of
Peace Research 42(2): 149-166.
----. 2005b. "Gender Equality and Intrastate Armed Conflict." International Studies Quarterly
49(4): 695-714.
Minier, Jenny. 2007. "Nonlinearities and Robustness in Growth Regressions." American
Economic Review 97(2) 388-392.
Munshi, Kaivan and Jacques Myaux. 2006. "Social Norms and the Fertility Transition." Journal
of Development Economics 80(1): 1-38.
Murray, Michael P. 2006. "Avoiding Invalid Instruments and Coping with Weak Instruments."
Journal of Economic Perspectives 20(4): 111-132.
Murthi, Mamta, Anne-Catherine Guio, and Jean Drèze. 1995. "Mortality, Fertility and Gender
Bias in India: A District Level Analysis." Population and Development Review 21(4): 745-
82.
Okonkwo Osili, Una and Bridget Terry Long. 2008. "Does Female Schooling Reduce Fertility?
Evidence from Nigeria." Journal of Development Economics 87(1): 57-75.
Pagan, Adrian. 1984. "Econometric Issues in the Analysis of Regressions with Generated
Regressors." International Economic Review 25(1): 221-247.
Perotti, Roberto. 1996. "Growth, Income Distribution and Democracy: What the Data Say."
Journal of Economic Growth 1(2): 149–187.
Pollin, Robert and Andong Zhu. 2006. "Inflation and Economic Growth: a Cross-Country
Nonlinear Analysis." Journal of Post-Keynesian Economics 28(4): 593-614.
29
Regan, Patrick M. and Aida Paskeviciute. 2003. "Women’s Access to Politics and Peaceful
States." Journal of Peace Research 40(3): 287–302
Rehavi, M. Marit. 2008. "Sex and Politics: Do Female Legislators Affect State Spending?"
mimeo, University of Michigan.
Rodríguez, Francisco and Dani Rodrik. 2001. "Trade Policy and Economic Growth: A Skeptic's
Guide to the Cross-National Evidence," in Ben S. Bernanke and Kenneth Rogoff, eds.
NBER Macroeconomics Annual 2000, pp. 261-338 Cambridge, MA: National Bureau of
Economic Research.
Roodman, David. 2003. "XTABOND2: Stata Module to Extend XTABOND Dynamic Panel
Data Estimator." mimeo, Boston College Department of Economics.
----. 2009. "A Note on the Theme of Too Many Instruments." Oxford Bulletin of Economics and
Statistics 71(1):135–58
Rosenzweig, Mark R. and T. Paul Schultz. 1982. "The Behavior of Mothers as Inputs to Child
Health: The Determinants of Birth Weight, Gestation, and Rate of Fetal Growth," in Victor
R. Fuchs, ed Economic Aspects of Health, pp. 53-92. Chicago: University of Chicago Press.
Sachs, Jeffrey D. and Andrew M. Warner. 1995. "Economic Reform and the Process of Global
Integration." Brookings Papers on Economic Activity 1995(1): 1-118.
Sala-i-Martin, Xavier. 1997. "I Just Ran Two Million Regressions." American Economic Review
87(2): 178-83.
Schober, Thomas and Rudolph Winter-Ebmer. 2011. "Gender Wage Inequality and Economic
Growth: Is There Really a Puzzle? – A Comment." World Development 39(8): 1476-1484.
Seguino, Stephanie. 2000. "Gender Inequality and Economic Growth: A Cross-Country
Analysis." World Development 28(7): 1211-1230.
30
----. 2007. "Plus Ça Change? Evidence on Global Trends in Gender Norms and Stereotypes."
Feminist Economics 13(2): 1-28.
----. 2011. "Gender Inequality and Economic Growth: A Reply to Schober and Winter-Ebmer."
World Development 39(8): 1485-1487.
Soares, Rodrigo R. and Bruno L. S. Falcão. 2008. "The Demographic Transition and the Sexual
Division of Labor." Journal of Political Economy 116(6): 1058-1104.
Svaleryd, Helena. 2009. "Women's Representation and Public Spending." European Journal of
Political Economy 25(2): 186-198.
Swamy, Anand, Stephen Knack, Young Lee and Omar Azfar. 2001. "Gender and Corruption."
Journal of Development Economics 64(1): 25-55.
31
Table 1. Summary Statistics
Table 1a. Full Sample Summary Statistics
Number of
Observations
228
Number of Countries
101
Observations per Country
Minimum
1.0
Average
2.3
Maximum
3.0
Variable
Mean
Std. Dev.
Min
Max
Per Capita GDP Growth
2.081
2.534
-5.149
10.951
Per Capita GDP Growtht-1
1.472
2.882
-9.592
9.268
ln(GDP per Capita)t-1
7.741
1.627
4.902
10.486
Investment (% of GDP)
22.186
6.422
7.703
56.196
Net Government Spending
12.321
5.665
-3.330
25.508
Inflation
14.131
25.510
0.198
242.309
Secondary Enrollment
17.184
12.048
0.543
52.413
Opportunity Factor
-0.037
0.752
-1.727
1.891
Outcomes Factor
-0.020
0.644
-1.351
1.715
Table 1b. OECD Subsample Summary Statistics
Number of
Observations
63
Number of Countries
22
Observations per Country
Minimum
2.0
Average
2.9
Maximum
3.0
Variable
Mean
Std. Dev.
Min
Max
Per Capita GDP Growth
2.480
1.817
-1.156
8.577
Per Capita GDP Growtht-1
2.083
1.811
-1.156
8.577
ln(GDP per Capita)t-1
9.656
0.583
7.899
10.486
Investment (% of GDP)
22.282
4.165
16.725
37.475
Net Government Spending
15.747
4.688
4.389
25.508
Inflation
7.157
14.828
0.305
79.302
Secondary Enrollment
27.414
10.935
1.905
47.092
Opportunity Factor
0.589
0.350
-0.853
1.287
Outcomes Factor
0.335
0.632
-0.841
1.715
32
Table 1c. Non-OECD Subsample Summary Statistics
Number of
Observations
165
Number of Countries
79
Observations per Country
Minimum
1.0
Average
2.1
Maximum
3.0
Variable
Mean
Std. Dev.
Min
Max
Per Capita GDP Growth
1.928
2.748
-5.149
10.951
Per Capita GDP Growtht-1
1.238
3.171
-9.592
9.268
ln(GDP per Capita)t-1
7.010
1.261
4.902
10.146
Investment (% of GDP)
22.149
7.108
7.703
56.196
Net Government Spending
11.012
5.466
-3.330
23.760
Inflation
16.794
28.139
0.198
242.309
Secondary Enrollment
13.278
10.007
0.543
52.413
Opportunity Factor
-0.277
0.727
-1.727
1.891
Outcomes Factor
-0.155
0.597
-1.351
1.431
33
Table 2. Factor Analysis
Factor analysis/correlation
Observations
203
Method: principal factors
Factors
2
Rotation: oblique promax (Kaiser off)
Parameters
9
Factor
Variance
Proportion
Equality in Opportunity
1.083
0.790
Equality in Outcomes
0.752
0.549
Rotated factor loadings (pattern matrix) and unique variances
Variable
Equality in
Opportunity
Equality in
Outcomes
Uniqueness
Labor Force Participation Gap
0.196
0.563
0.645
Percent of Parliament Women
0.300
0.566
0.590
Literacy Rate Gap
0.300
0.103
0.900
Secondary Enrollment Gap
0.652
0.146
0.554
Fertility Inverse
0.663
0.290
0.476
Factor rotation matrix
Equality in
Opportunity
Equality in
Outcomes
Equality in Opportunity
0.796
0.606
Equality in Outcomes
-0.606
0.796
Notes: Shading indicates a factor loading greater than 0.5.
Table 3. Selected Percentiles of Gender Factor Variables
Opportunity
Outcomes
Top Five
1
Bulgaria
Sweden
2
Latvia
Norway
3
Spain
Finland
4
Estonia
Denmark
5
Italy
Iceland
First Quartile
23
Korea
Malawi
24
Norway
Ghana
Middle Five
43
Sri Lanka
Senegal
34
44
Colombia
Gabon
45
Panama
Zambia
46
Jamaica
Thailand
47
Albania
Bolivia
Third Quartile
67
Algeria
Panama
68
Egypt
Malaysia
Bottom Five
84
Malawi
Iran
85
Cote d’Ivoire
Niger
86
Mali
Morocco
87
Niger
Jordan
88
Togo
Pakistan
35
Table 4. Unbalanced Panel Results (Dependent Variable = Per Capita Growth; Estimation
Method = Difference GMM)
(1)
(2)
(3)
(4)
VARIABLES
Full Sample
OECD Only
Non-OECD
Only
Interactions
GDPC Growtht-1
-0.549***
-0.582***
-0.450***
-0.509***
(0.0821)
(0.126)
(0.127)
(0.0844)
ln(GDPC)t-1
-7.051***
-4.592
-6.962**
-6.645***
(2.540)
(3.864)
(3.119)
(2.435)
Investment (% of GDP)
0.149***
0.170*
0.164**
0.168***
(0.0541)
(0.0915)
(0.0688)
(0.0562)
Government Expenditure (Net of
Education and Military, % of GDP)
-0.261***
-0.508**
-0.213*
-0.250***
(0.0982)
(0.220)
(0.119)
(0.0925)
Inflation (CPI, %)
-0.0101
-0.0375
-0.00564
-0.0117
(0.00953)
(0.0422)
(0.0140)
(0.0124)
Secondary Completion Rate
0.101*
0.117*
0.0435
0.0849
(0.0573)
(0.0669)
(0.101)
(0.0667)
Opportunity
1.735**
0.857
1.842*
1.677*
(0.827)
(1.264)
(1.073)
(0.993)
Outcomes
1.847**
2.348*
0.484
0.394
(0.895)
(1.210)
(1.577)
(1.403)
OECD X Opportunity
-0.493
(1.169)
OECD X Outcomes
2.577**
(1.280)
Constant
55.80***
47.87
48.74**
52.05***
(19.92)
(34.97)
(22.38)
(19.16)
Observations
228
75
153
228
Number of Countries
101
28
75
101
Arellano-Bond Statistic
-1.46
-0.46
-1.88
-1.56
Hansen Over-ID Statistic
26.87
20.71
21.56
25.18
Degrees of Freedom
20
20
20
20
P-Value
0.139
0.414
0.365
0.195
Instruments for first differences equation:
Standard: (Investment, Government, Inflation, Secondary Education, Opportunity, Outcomes,
Interactions)
GMM-type (separate instruments for each period): L(Growtht-1, ln(GDPC)t-2)
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
36
NOTES
1
This should also reduce the precautionary demand for children and the resultant decline in fertility should
stimulate growth.
2
It should be mentioned that Fernando Ferreira and Joseph Gyourko (2010) fail to obtain any impact of gender on
expenditure by American city councils.
3
While Svaleryd (2009) investigates the consequences of female representation in government, John R. Lott, Jr. and
Lawrence W. Kenny (1999) consider the impact of women’s suffrage in America and identify it as a key
determinant of the increase in public expenditure on education observed from 1870-1940.
4
Annual percent growth in GDP per capita = (GDP p.c.it - GDP p.c.i,t-1)/GDP p.c.i,t-1.
5
Also note that in constructing the time dimension of the panel as five year intervals, the one-period lagged value of
any variable in our specification represents the five-year lagged value of the five-year average.
6
A list of countries included in our sample is also available online as a supplemental table.
7
We have not considered the gender wage gap due to the unavailability of reliable wage data for many of the
countries in our sample.
8
All gender gaps are defined as the ratio of female to male magnitudes of the relevant variables, where a higher
value indicates greater equality. To maintain parity with this convention, we take the inverse of adolescent fertility.
9
For countries with a bicameral legislature, we take the percentage of women in the lower chamber. The Inter-
Parliamentary Union (http://www.ipu.org/wmn-e/classif-arc.htm) does not provide data prior to 1997. We rely on
version 3.0 of the Democracy Time Series Data compiled by Pippa Norris (http://www.pippanorris.com) for the
missing years.
10
The most commonly-used aggregation procedure is to perform principal component analysis and consider the first
component as institutional quality (Knack and Keefer, 1995; Alesina and Perotti, 1996; Perotti 1996).
11
Highlighting this problem for institutional variables, Laura Langbein and Stephen Knack (2010) undertake a
confirmatory factor analysis of the World Governance Indicators and fail to confirm the hypothesis that these
measures are causally related to a single variable good governance. Also, unidimensional indices of gender equality
such as the GEM and GDI have been criticized on the grounds that they do not reflect gender equality per se (A.
Geske Dijkstra, 2002; Klasen and Lamanna, 2009).
37
12
A limitation of dynamic GMM panel estimators is that they assume that the lagged values of the endogenous
regressors are strong and only test their validity, using either a Sargan or Hansen statistic. Further, the finite-sample
properties of these estimators are not well-known (Bazzi and Clemens, 2013).
13
The unique part of the decomposed variance can be seen as a residual, consisting of a random component and
measurement error. The uniqueness factor reported in Table 2 consists of the total variability of each variable minus
the sum of its squared factor loadings.
14
In obtaining the underlying factors, one faces the choice between several extraction methods, including principal
component, principal factor, and maximum likelihood. Of these, the principal component extraction method is
inappropriate for our purpose since it seeks to explain all of the variance in the observed variables and not the
common variance, and hence leads to correlated errors. Maximum likelihood extraction requires the assumption of
multivariate normality. One advantage of principal factor extraction is that it requires no distributional assumption
on the observed variables. With respect to rotation, one faces the choice between orthogonal and oblique methods.
Orthogonal methods, such as orthomax or quartimax, force the assumption of orthogonality onto the factors, which
leads to loss of information if the factors are correlated. We have followed the prescription of Anna B. Costello and
Jason W. Osborne (2005) in choosing the oblique promax rotation method. We have replicated our analysis using
alternative extraction and rotation procedures and obtained virtually identical results, which are available on request.
15
It should be clarified that we have not restricted the number of factors. Rather, the process determined that two
was the appropriate number of factors, based on the proportion of common variance they explain.
16
Serge Coulombe and Jean- François Tremblay (2006) address this problem by considering results from the
International Adult Literacy Survey as the measure of literacy so as to standardize the definition of literacy across
countries. But this survey and other cross-national initiatives, such as the Adult Literacy and Lifeskills Survey and
the Programme for the International Assessment of Adult Competencies, only cover OECD countries.
17
As an example of the lack of clarity on the topic, while Barro (1991) finds a negative impact of net government
consumption on growth, Xavier Sala-i-Martin (1997) fails to find any robust association between the variables.
18
Of the 101 countries in our sample, 22 have been members of the OECD over the entire sample period, 76 have
been nonmembers for the entire time period, and three – Mexico, Hungary, and South Korea – joined the OECD in
1994, 1996, and 1996 respectively.
38
19
Recall that with interactions between a dummy and a continuous variable, the non-interacted coefficient on the
latter represents the impact of the continuous factor on the excluded group, here non-OECD countries. The impact of
the factor for the included group, here OECD countries, is the sum of the non-interacted and the interacted
coefficients and its standard error is calculated as (
). Note also that we have excluded
the non-interacted OECD dummy variable since its effects are almost perfectly correlated with country fixed effects.
20
This is consistent with results obtained by a number of studies on the topic (Hill and King, 1995; Dollar and Gatti,
1999; Kristin J. Forbes, 2000). Given that each of our regressions includes the gender gap in the access to education;
we interpret the insignificance of the male secondary completion rate for developing countries as indicating that the
importance of human capital as a determinant of growth depends critically on the level of equality allowed in its
acquisition (Francesco Caselli, Gerardo Esquivel and Fernando Lefort, 1996; Coulombe and Tremblay, 2006).
21
The results of these tests are available online as supplemental tables.
22
Since it is well documented that the volume of trade is correlated with the geographical area and population of a
country, we follow Barro (1991) in filtering our measure of openness for the impact of these variables.
23
Results are available online as a supplemental table. It is interesting to note that both inflation and trade openness
mostly fail to achieve statistical significance. The first result is consistent with Michael Bruno and William Easterly
(1998), who also find no impact of inflation on growth from 1960-1992, except in extreme episodes of inflation.
Also, the ratio of trade to GDP is not the only measure of openness. Despite critiques by Francisco Rodríguez and
Dani Rodrik (2001) and others, the literature has predominantly followed Frankel and Romer (1999) in filtering the
trade volume for the effects of geographical characteristics. Also, while our difference-GMM methodology
implicitly follows the prescription of Caselli, Esquivel and Lefort (1996) in addressing the endogeneity between
trade and growth by estimating the growth model in differences and using lags of the explanatory variables as
instruments (Dollar and Kraay, 2004) this is not the only way to address the endogeneity problem.
24
We have also reconstructed the predicted factors for gender equality on the basis of the balanced sample. Neither
the factor loadings nor the interpretations of the factors change significantly with the change in sample.
25
Results are available online as a supplemental table.
