International trade, female labor, and entrepreneurship in MENA countries.
Middle Eastern and
A region of northern Africa generally considered to include the modern-day countries of Morocco, Algeria, Tunisia, and Libya.
North African adj. & n.
(MENA) countries stand out in
international comparisons of
obstacles to female employment and
entrepreneurship. These obstacles manifest themselves in low rates of
female labor participation, entrepreneurship, and ownership. Recent
research suggests a connection between international trade and female
labor participation. In this article, the authors focus on the
relationship between international trade and gender in the MENA
countries. They first analyze female labor as a production factor and
then focus on female entrepreneurship and firm ownership. The authors
use country- and industry-level data to identify countries and
industries characterized by a comparative advantage in female labor.
They find evidence suggesting a strong link between a country’s
and its measures of female labor participation consistent
with theories of brain-based technological bias and factor endowments
trade theories. Using firm-level data, the authors then study whether
trade empowers female entrepreneurs in country/industry pairs that
exhibit comparative advantage. They conclude that the evidence supports
the view that exposure to trade
Out of proportion, as in size, shape, or amount.
affects firms in
country/industry pairs with a comparative advantage in female
labor–both in terms of female employment and female entrepreneurship
and ownership–for the MENA countries and the period they study. (JEL
F11, F14, F16, J82)
The relationship between trade and gender has recently emerged as
an important theme Tin the international economics and development
literature. The United Nations’ Millennium Development Goal No. 3
is to promote gender equality and empower women, a broad goal that can
cover many areas of economic and non-economic activity. The entire World
Development Report 2012: Gender Equality and Development (
WDR World Development Report
WDR Wide Dynamic Range
Bank, 2011) is devoted to the study of gender issues, and its chapter 5
focuses specifically on the relationship between trade and gender,
highlighting the main conceptual issues and presenting several
interesting research avenues. In fact, this explicit effort is
reinforcing research on gender at both the
n. (used with a sing. verb)
The study of the overall aspects and workings of a national economy, such as income, output, and the interrelationship among diverse economic sectors.
n. (used with a sing. verb)
The study of the operations of the components of a national economy, such as individual firms, households, and consumers.
Macroeconomic evidence suggests that female labor participation
decisions have important aggregate consequences and are
v. cor·re·lat·ed, cor·re·lat·ing, cor·re·lates
1. To put or bring into causal, complementary, parallel, or reciprocal relation.
certain forms of technological change that may affect women differently
than men; specifically, although women are on par with men in
“brain-intensive” skills and abilities, women tend to have a
comparative disadvantage in “brawn-intensive” activity. When
technological change favors brain-intensive activities, women can
exploit their comparative advantage in these occupations. Microeconomic
evidence affords a better understanding of some of the mechanisms that
“empower” women–for example, within the household in both
advanced economies and in a developmental context. In this article, we
attempt to analyze the relationship between international trade and
gender–defined here as both female labor participation and female
ownership and entrepreneurship–in the Middle Eastern and North African
(MENA) countries. This region is interesting in a cross-country
perspective because international comparisons of de jure indicators
suggest that MENA countries are characterized by more marked gender
discrimination in female labor participation and entrepreneurial
activity than most other regions of the world.
To this end, we adopt a factor-endowment perspective and construct
measures of female labor use (or intensity) at the country/industry pair
level and match them to manufacturing trade data to determine the female
labor content of exports. Next, we identify country/industry pairs
characterized by comparative advantage in female labor for five
countries for which aggregated data are available. We then compare our
classification of country/industry pairs based on comparative advantage
in aggregated data with a similar classification based on firm-level
data from the World Bank Enterprise Surveys (WBESs). We find that the
two datasets track each other quite well, which allows us (and
potentially other researchers using the same procedure) to exploit the
firm-level data for other MENA countries for which aggregated data are
Finally, we study whether exposure to trade in comparative
advantage country/industry pairs empowers women by increasing the
probability of female entrepreneurship and ownership in firms belonging
to country/industry pairs with comparative advantage. We find some
support for the (theory-free) hypothesis that comparative advantage
empowers female entrepreneurs in country/industry pairs with comparative
The article is organized as follows. The next section provides an
overview of female
A place where labor is exchanged for wages; an LM is defined by geography, education and technical expertise, occupation, licensure or certification requirements, and job experience
and entrepreneurship in MENA countries.
We then discuss the relevant literature and describe the data used in
the empirical analysis detailed in the following sections.
AN OVERVIEW OF FEMALE LABOR MARKETS AND ENTREPRENEURSHIP IN MENA
In addition to
n information obtained from personal accounts, examples, and observations. Usually not considered scientifically valid but may indicate areas for further investigation and research.
of gender bias, MENA countries
stand out in international comparisons of de jure indicators as being
characterized by more marked gender discrimination in female labor
participation and entrepreneurial activity than other countries. This
position is clearly represented in Figure 1, which shows a summary
measure of female discrimination by groups of countries based on the
World Bank’s 2010 Women, Business and the Law (
WBL Wide Band Link
WBL Wideband Limiting
) dataset. (1)
The WBL is a cross-country dataset reporting information on
differential legal treatment experienced by women with regard to
business-related activities. We use this information and plot the number
of differential treatments by both aggregate geographic regions and
individual countries. The top panel of Figure 1 reports the statistics
by geographic regions; the number of countries in each region is listed
in parentheses. In particular, the top graph is constructed using a
of questions specifically focused on entrepreneurship and
business. (2) Each question in the WBL dataset highlights whether women
in business face restrictions in a specific activity (for example,
starting a business) relative to men. For each group of countries, each
bar in the graphs represents the average count of differential treatment
instances. Similar to sub-Saharan Africa, the MENA countries clearly
stand out for the high number of reported de jure discrimination
episodes against women. The bottom panel focuses on only the MENA region
and indicates that gender discrimination is particularly severe in
Jordan, Syria, and Yemen.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
the importance of de jure discrimination in
the MENA region and suggest that female labor participation and
entrepreneurship may be negatively affected but do not address whether
de jure obstacles translate directly into
whether this is then reflected in women’s labor and/or
entrepreneurship decisions. Figure 2 uses data from the World
Bank’s World Development Indicators for 2011 to plot male and
female labor participation for the world and the MENA region at both
regional and country levels. The figure shows that while male labor
force participation in the MENA countries is aligned with the world
level, female labor force participation is substantially lower. There is
The quality or state of being heterogeneous.
the state of being heterogeneous.
within the MENA group: Turkey and–to a lesser
extent–Morocco stand out for the high incidence of female workers in
the economy. Even within the rough picture provided by aggregate data,
substantial heterogeneity exists both within the group of countries and
within individual countries.
A somewhat comparable
1. One who lives near or next to another.
2. A person, place, or thing adjacent to or located near another.
3. A fellow human.
4. Used as a form of familiar address.
region consists of the 27
transition economies in
The countries of eastern Europe, especially those that were allied with the USSR in the Warsaw Pact, which was established in 1955 and dissolved in 1991.
and Central Asia (
). Figure 3
considers the firm-level
in chemistry, mixture in which fine particles of one substance are scattered throughout another substance. A dispersion is classed as a suspension, colloid, or solution.
of the use of female labor and plots
the frequency of female workers as shares of total workers in surveyed
firms in the MENA and ECA regions in firm-level data collected from
WBESs, as described in the next section. The histograms clearly indicate
1. Heavier, larger, or higher on one side than on the other.
2. Sagging or leaning to one side.
distribution of this measure in the MENA countries, with
a large mass at 0 to 4 percent. By comparison, firms in the ECA region
have a more uniform distribution that is closer to similar plots for
advanced economies. Figure 4 further divides this
individual countries and suggests cross-country heterogeneity in female
labor as an input in each country’s firms. The distributions in
Morocco and Turkey are closer to those in the ECA region, while the
histograms for other countries show approximately a negative
, which is extremely lopsided in the case of Yemen.
[FIGURE 3 OMITTED]
Finally, Figure 5 uses the same data source to consider de facto
indicators of female entrepreneurship participation, particularly the
incidence of female ownership and female management. While Lebanon and
Turkey show measures of the incidence of female owners that are
comparable to the world in general, Algeria, Jordan, Morocco, Syria, and
Yemen contribute to the low incidence of female owners in the MENA
region overall. In contrast, the incidence of female managers in top
management is higher than in the rest of the world for most countries in
the MENA group for which data are available–that is, Egypt, Lebanon,
and Syria–but extremely low in Yemen.
This article contributes to an emerging literature that highlights
the role of gender differences in microeconomic data and macroeconomic
models. In quantitative macroeconomic and labor economics, the
availability of data on the composition of occupations has generated a
small segment of the literature centered on the concept of
“brain-based technological change” Male and female labor are
not perfect substitutes because individuals are
tr.v. en·dowed, en·dow·ing, en·dows
1. To provide with property, income, or a source of income.
brain abilities (mental labor) but different
1. Solid and well-developed muscles, especially of the arms and legs.
2. Muscular strength and power.
3. Chiefly British The meat of a boar.
labor) that favor men (Galor and Weil, 1996). As long as brawn skills
have a positive marginal return, men enjoy a
1. Advantageous; helpful:
2. Encouraging; propitious:
gender wage gap.
However, when technological change is biased in favor of brain-intensive
activity, women can
1. To limit one’s profession to a particular specialty or subject area for study, research, or treatment.
2. To adapt to a particular function or environment.
1. As stated or indicated by; on the authority of:
2. In keeping with:
their comparative advantage.
The related increase in female labor demand then pushes wages up and
contributes to reducing the wage gap with men, as observed in the U.S.
data (see Rendall, 2010, and Keller, 2012).
[FIGURE 4 OMITTED]
While this theoretical argument holds for the U.S. economy, the
evidence from international data is less clear. On the one hand,
cross-country comparisons show a consistent picture for richer but not
poorer countries (Oostendorp, 2009). In the former, the wage gap tends
to decrease with increasing trade and foreign direct investment (
well as economic development. In the latter, trade and FDI do not appear
to reduce the occupational gender wage gap. These findings are
consistent with the possibility that brain-based technological change
may have a smaller impact in poorer countries.
On the other hand, Aguayo-Tellez, Airola, and Juhn (2010) reach a
different conclusion in their study of the impact in Mexico of
v. lib·er·al·ized, lib·er·al·iz·ing, lib·er·al·iz·es
To make liberal or more liberal:
measures resulting from the
North American Free Trade
(NAFTA), accord establishing a free-trade zone in North America; it was signed in 1992 by Canada, Mexico, and the United States and took effect on Jan. 1, 1994.
. Women’s relative wages increased, and both between- and
within industry shifts favored female workers since tariff reductions
expanded sectors that were initially female intensive. Women also gained
intrahousehold bargaining power as documented by the change in
expenditure toward goods associated with female preferences (for
example, clothing and education).
[FIGURE 5 OMITTED]
Do, Levchenko, and Raddatz (2012) offer a different perspective
that reconciles these findings. Improvement in trade openness changes
the opportunity cost of women staying out of the formal labor market and
therefore the trade-off between work and fertility decisions, a point
also made by Rees and Riezman (2011). In particular, Do, Levchenko, and
Raddatz (2012) develop a specific factor model of trade in which male
and female labor are combined with capital in two manufacturing sectors.
When trade increases the demand for female labor in country/industry
pairs with comparative advantage with intensive use of female labor,
trade also induces increases in female labor wages, which in turn affect
fertility decisions. Their model is supported by empirical evidence in a
large cross section of countries.
Saure and Zoabi (2009) study similar relationships but focus on the
on labor by gender. These authors argue
improves work opportunities for women, female
labor participation may drop if international specialization promotes
sectors that use female labor intensively. This effect arises because
expansions of the former sectors are accompanied by contractions of
others that induce male workers to move to the expanding sectors,
driving female workers out of formal employment. Thus, a country that is
exporting female labor content may actually be substituting male labor
Finally, there is a
Of considerable size; fairly large.
literature on female labor
participation and job mobility that includes Gayle, Golan, and Miller
(2012); their paper is important for our analysis of entrepreneurship
and management because it focuses on job progression and selection,
particularly the reasons why fewer women than men become executive
managers, earn less over their careers, hold more junior positions, and
exit the occupation at a faster rate. The authors find that, controlling
for executive rank and background, women earn higher compensation than
men, experience more income uncertainty, and are promoted more quickly
but also that these differences are related to the difficulties of
surviving in the organization. Among survivors, being female increases
the chance of becoming
because survival is rewarded with promotion
and higher compensation.
Because extensive coverage of the literature is beyond the scope of
this article, we refer interested readers to the WDR 2012 (World Bank,
2011), which summarizes the debate on the relationship between gender
inequality and development. The portions of the report most relevant to
our article are chapter 5 (on gender differences in employment) and
chapter 6 (on the relationship between globalization and gender
inequality). The data in these chapters support the arguments that (i)
developing countries are experiencing reductions in male/female wage
gaps and (ii) part of these reductions may be related to a
country’s openness. We contribute to this debate by providing
further evidence on the degree of gender labor participation in the MENA
region and linking it to international trade.
We maintain a flexible definition of the MENA countries because of
differences in the availability of data across countries; thus, we
combine multiple data sources. The largest group of countries (Algeria,
Egypt, Jordan, Lebanon, Morocco, Turkey, Syria, and Yemen) is
tr.v. con·strained, con·strain·ing, con·strains
1. To compel by physical, moral, or circumstantial force; oblige: See Synonyms at force.
by the availability of individual WBESs. In addition, we use
United Nations Industrial Development Organization
(UNIDO), specialized agency of the United Nations. Headquartered in Vienna, it was organized in 1966 and made a specialized UN agency in 1985. UNIDO’s mission is to promote industrial progress in developing nations.
Industrial Statistics databases (UNIDO INDSTAT2) and the United Nations
Commodity Trade Statistics (UN
) as data sources. (3)
The WBESs collect firm-level data from a selected number of
countries to provide a representative picture of the population of firms
in the countries’ economies. Table 1 shows the frequency of firms
by country and industry. From the available data we select country/year
pairs for the years between 2006 and 2009 to maintain a cross-country
perspective as broad as possible (Table 2). Each survey questionnaire is
constructed to capture the economic and institutional environment in
addition to the main business constraints faced by firms within the
country. This information is captured by quantitative and qualitative
indicators used in the questionnaire to measure, for example, the level
of the firm’s sales, the amount of export, FDI intensity, and
qualitative perceptions about the business environment. We use all
available information regarding the gender dimension of the firm’s
activity. In addition to standard firm demographic information, the
datasets include information on whether any of the firm’s owners,
top managers, or largest shareholders are women and the total percentage
of women employees in the firm.
The UNIDO INDSTAT2 database contains industry-level data organized
by country, year, and industry and collects information for a large
cross section of countries. We focus on the countries in the MENA region
during the period the WBESs were collected (see Table 2). The UNIDO
INDSTAT2 database contains seven indicators: number of establishments,
employment, wages and salaries, output,
, and number of female employees. We focus on the total
employment (number of employees) and the female labor composition of the
workforce by industry (number of female employees). The database adopts
the International Standard Industrial Classification (
ISIC Immediate Superior In Command
) code system,
which classifies industries broadly along product lines (such as food,
textiles, iron, and steel). The ISIC code covers all areas of economic
constrains us to the manufacturing
sector. We use a fairly aggregate level of industry classification at
the two-digit level of ISIC Revision 3. (4)
The information from UNIDO INDSTAT2 is matched with that from UN
COMTRADE, which details annual international trade statistics data by
commodities and partner countries. Data availability from the UNIDO
INDSTAT2 and UN COMTRADE databases reduces the set of matchable MENA
countries to Algeria, Egypt, Jordan, Morocco, and Turkey.
THE FEMALE LABOR CONTENT OF EXPORTS IN AGGREGATE DATA
Our first step is to construct a country-level measure of the
female labor content of exports following the work of Do and Levchenko
(2007), which is based on Almeida and Wolfenzon’s (2005) approach
to measuring the external finance content in production and exports. Do
and Levchenko’s (2007) measure is based on a model in which a
country’s comparative advantage contains both a Ricardian component
Because of; on account of:
prep → ,
total factor productivity differences across countries) and a
factor proportions component (owing to different country-level
endowments of male and female workers related to differences in labor
We follow this approach and
a sector-level measure of
.sub.cs]) that is the share of female labor
force at the country-industry level:
[FLI.sub.cs] = [FL.sub.cs]/[TL.sub.cs]
where [FLI.sub.cs], corresponds to the number of female employees
in country c and sector s and [TL.sub.cs], represents the number of
total workers in country c and sector s as reported in the UNIDO
INDSTAT2 database. (5) We then match each sector with the UN COMTRADE
data and compute the female labor content in export production as
[FLNX.sub.cs] = [X.sub.cs]/[X.sub.c][FLI.sub.cs],
where [X.sub.cs]/[X.sub.c] is the share of sector s in total
manufacturing exports to the rest of the world by
country c. Therefore, [FLNX.sub.cs] gives a measure of female labor
content of exports from each country/sector pair. Similarly, a
country-level measure summarizes the content of female labor in exports
for the entire manufacturing sector:
[FLNX.sub.c] = [s.
n. the final argument of an attorney at the close of a trial in which he/she attempts to convince the judge and/or jury of the virtues of the client’s case. (See: closing argument)
Conceptually, the [FLNX.sub.c] measure captures the female labor
content of exports, but it also reveals the comparative advantage of
individual countries in certain industries. The model developed by Do,
Levchenko, and Raddatz (2011) delivers the Heckscher-Ohlin-style
prediction that countries with relatively abundant female labor–as
measured by female labor force participation-should be observed as
exporting relatively larger shares of the goods that make intense use of
female labor. Do, Levchenko, and Raddatz (2011) use cross-sectional
country data and find support for this theoretical prediction. Countries
with higher female labor force participation show larger export shares
in sectors that use female labor intensively, after controlling for
country and industry fixed effects. Here, we take a less formal approach
to verify whether this fact is confirmed in our sample of countries and
years. Figure 6 plots a
with the product of each
country’s female labor participation (
FLP Front de Libération de la Palestine
FLP Fasting Lipid Profile
) times the female share
of total labor for each country/industry pair (FL) on the x-axis and the
level of export from country/industry
pair cs as a share of country c’s total
exports([[omega].suc.cs] = [X.sub.cs]/[X.sub.c]) on the y-axis. (6) The
solid line indicates
1. To cause to separate and go in different directions.
2. To separate and go in different directions; disperse.
3. To deflect radiation or particles.
dots and shows the positive
relationship predicted by the theory. Consistent with the findings of
Do, Levchenko, and Raddatz (2012), the solid line in Figure 6 shows a
positive relationship between,
, female labor
participation and exports.
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
Now a new question arises: Which industries in each country have a
comparative advantage in female labor? To answer this question, we pool
countries and consider the cross section of sectors and compute the
female labor content of exports at the country level (Figure 7). The
ranking of industries is determined by pooling all MENA countries for
which data are available in the UNIDO INDSTAT 2 database. Figure 8
indicates which sectors are female (male)
average–that is, the sectors that display a higher-than-average share
of female (male) labor of the total labor used by industry. (7)
[FIGURE 8 OMITTED]
The ranking of industries according to female labor intensity by
sector and the ranking of countries according to the female labor
content of exports at the country level also help us to define the
industries in which individual countries have a comparative advantage in
female labor relative to the MENA group overall. Morocco and Jordan have
a comparative advantage in female-labor-intensive industries, while
Egypt and Turkey have a comparative advantage in male-labor-intensive
industries. We also consider Algeria as having a comparative advantage
in male-labor-intensive industries, although the UNIDO data were
available only for 1996; therefore, this country’s measures should
be considered carefully.
USING FIRM-LEVEL DATA TO IDENTIFY COMPARATIVE ADVANTAGE IN
In previous sections, comparative advantage at the country level
was defined on the basis of the factor content of trade, particularly
focusing on female and male labor as inputs. We continue to assume these
are the only two factors of production and abstract from the measurement
of capital, given the available data. Our next step is to link our
analysis to the firm-level implications of recent theories of
comparative advantage in international trade theory that introduce firm
heterogeneity in more standard factors. Typical Heckscher-Ohlin models
in factor intensity. “New trade
theory” approaches with firm heterogeneity usually use models with
only labor as a production factor, therefore abstracting from factor
intensities. However, the evidence shows that factor intensities vary
greatly within industries. For example, Leonardi (2007) documents (i)
the wide capital-to-labor ratio dispersion within U.S. industries using
Compustat-reported firms and (ii) its upward trend over time and relates
the dispersion to the increase in residual wage inequality. Figure 9,
for which we use firm-level data, shows the within-industry dispersion
in the female-to-male labor ratio across MENA countries; a vast
dispersion is seen both within and between industry sectors. The results
hold true when the graphs for each country are plotted separately (not
[FIGURE 9 OMITTED]
The next step in our analysis considers the following two
definitions consistent with Crozet and Trionfetti (2012): (i)
Female-labor-intensive industries are those in which the female-to-male
labor ratio of the industry is larger than the female-to-male labor
ratio for the region. (ii) Female-labor-abundant countries are those in
which the female-to-male labor ratio of the country is larger than the
female-to-male labor ratio for the region as a whole.
Making these two definitions operational with firm-level data can
be challenging because there are multiple ways to measure the
female-to-male labor ratio within a country/industry pair. The two most
intuitive choices are within-country and within-industry averages and
median female-to-male labor ratios. However, the
within-country/within-industry distributions of firm size are very
, identify country/industry pairs with comparative advantage
at odds with the procedure used in the previous sections based on
aggregate data. Therefore, we rely on a third option and compute
country/industry pair measures of female-to-male labor by first summing
workers by gender within a country/industry group and then measuring the
tr.v. de·not·ed, de·not·ing, de·notes
1. To mark; indicate:
measures constructed using micro data with lowercase
letters and aggregate data with capital letters:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN
or a set of codes used to represent letters, numbers, a few symbols, and control characters. Originally designed for teletype operations, it has found wide application in computers.
where i, c, and s denote firms, countries, and industries,
respectively. We have country/industry pairs for 8 sectors times 5
countries and corresponding measures of the female-to-male labor ratio
[fmlr.sub.cs]. We then sum the firm-level female employment and male
employment within a country across industries [fmlr.sub.c] and across
countries in the same industry [fmlr.sub.s]. We obtain the
female-to-male labor ratio by summing female and male workers across
sectors and countries. A sector is female labor intensive if
[fmlr.sub.s] >fmlr, while a country is female labor abundant if
[fmlr.sub.c] > fmlr.
These definitions identify the country/industry pairs with
comparative advantage in either female or male labor (Table 3). Panel A
of the table reports the female-to-male labor ratio measures and
identifies the country/industry pairs with comparative advantage in
female labor based on (i) firm-level data in Panel B (identified by a
called DV-CA micro) and (ii) aggregate data and the
measure of the female labor content of exports (Panel C) developed by
Do, Levchenko, and Raddatz (2012). The table shows that the
countries using micro and aggregate data delivers the same
classification of countries in female-labor-abundant and
male-labor-abundant industries except for the chemicals and
pharmaceutical industry. It should be noted that this industry measure
of female labor intensity is particularly close to the total
manufacturing measure (see Figure 8) based on aggregate data; this
finding suggests that the
between classifications may be
influenced by small differences between the firm-level and aggregate
More generally, our match shows that the firm-level surveys could
be used to define comparative advantage country/industry pairs even when
the aggregate data are not available, a result we plan to exploit in
future research. In this article, we use this methodology to add another
country for which we have firm-level data to our set of countries for
the remainder of the analysis.
Before we use these classifications of country/industry pairs in
our firm-level analysis, we discuss some evidence on hiring practices
related to the gender of workers and managers based on the Jordanian
for 2006 in the boxed insert.
We investigate the main factors that
v. hin·dered, hin·der·ing, hin·ders
1. To be or get in the way of.
2. To obstruct or delay the progress of.
a firm’s operation
and growth to better grasp the business environment faced by firms. WBES
respondents identify “the biggest obstacle for the operation and
growth of [the] establishment” from an extensive country-specific
list of possible constraints to business. (9) Figure 11 plots the survey
answers according to whether the firm has any female managers.
Interestingly, access to capital is the most frequent issue raised by
firms regardless of the gender composition of management. Political
instability, high tax rates, and illegal or unfair competition from the
informal sectors (such as
illegal transport across state or national boundaries of goods or persons liable to customs or to prohibition. Smuggling has been carried on in nearly all nations and has occasionally been adopted as an instrument of national policy, as by Great Britain
or dumping) are also cited by both
types of firms. However, there is more heterogeneity in the relative
importance of these issues across firm types. Firms with women in
management place more importance on political instability and less on
the competition from the informal sector or high tax rates.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
[FIGURE 12 OMITTED]
The WBESs provide a representative sample of firms within but not
necessarily across countries (see Table 2 for a cross-country comparison
of the representativeness of our data). Figure 12 plots the two most
frequent answers by country and gender of the
to uncover the
possible presence of country-specific differences masked by the
aggregates. The main
A restriction on the natural degrees of freedom of a system. If n and m are the numbers of the natural and actual degrees of freedom, the difference n – m is the number of constraints.
to business is the same for male- and
female-managed firms in all MENA countries except Syria and Yemen; in
Syria women suffer more because of competition from the informal sector,
and the Yemenite complicated taxation policies are the most-reported
barrier by female-managed firms. The second most important constraint is
somewhat different across countries and gender composition of a
firm’s management. Algeria and Lebanon are the only exceptions
where access to capital and electricity, respectively, is always
We test this descriptive evidence to gain further insights.
Respondents to the WBESs not only identified the two most important
constraints to business, but they also ranked all possible business
constraints in order of importance. In particular, they were asked to
determine whether each constraint was “not an obstacle” or a
“minor, moderate, severe, or very severe obstacle.” We used
the answers to these qualitative questions to construct an index ranging
from 0 to 4 to indicate the importance of each business constraint. (10)
We used these answers and added firm-level controls (such as the firm
age, labor productivity, and size) to determine whether there are
significant systematic differences across gender (Tables 4 and 5). Firms
with female managers appear to have fewer telecommunication problems but
face greater macroeconomic uncertainty (i.e., uncertainty about
inflation and exchange rate fluctuations) and have more problems in
obtaining access to financing (e.g., because of insufficient
collateral); the latter result is likely related to the fact that firms
managed by women tend to be smaller. While macroeconomic uncertainty
harms exporters as expected, access to capital is more problematic for
non-exporters; such lack of access is consistent with the notion of
selection into export. Access to capital remains more difficult for
enterprises owned by women than those owned by men. Macroeconomic
conditions lose their relevance, but obtaining business licensing
becomes more difficult for women and access to electricity gains
importance for male owners.
Finally, we investigate whether trade plays a role in empowering
women. Specifically, we analyze (i) whether women are more likely to be
business owners or managers of a company (Table 6) and (ii) whether the
existence of comparative advantage in a sector affects the likelihood of
women being business owners in that sector (Table 7). We find evidence
suggesting women are likely to advance professionally in sectors in
which their presence is stronger. We find evidence that suggests women
are likely to climb professional ladders in sectors where their presence
is stronger, a fact that is consistent with the evidence analyzed in
Golan, Gayle, and Miller (2012). In our analysis, this empowerment
through employment narrative is motivated by the positive and
significant correlation (the point estimate is 0.21 with a corresponding
p-value of 0.0000) between the average share of female employment and
the average share of female-owned firms by sector and country that
results after controlling for other observable characteristics of the
firm (see Table 7).
Female owners are more likely to be observed in industries with a
higher presence of female workers or in female-labor-intensive
industries (with a larger-than-the-median share of female workers).
These results, however, are not economically or statistically
intr.v. con·versed, con·vers·ing, con·vers·es
1. To engage in a spoken exchange of thoughts, ideas, or feelings; talk. See Synonyms at speak.
, we find that women are significantly more
likely to be managers of firms with lower labor productivity.
Table 6 also accounts for the role of the de jure constraints
analyzed in a previous section. We construct an index between 0 and 1
where 1 (0) corresponds to the maximum (minimum) number of de jure
constraints faced by women in a given country. As expected, women are
more likely to become business owners in female-labor-abundant countries
when they face fewer de jure constraints and operate in industries with
a higher concentration of female workers. The negative effect of de jure
constraints is reduced for women when they are employed in a
This article contributes to the small but growing literature on the
relationship between female labor participation and openness to
globalization. We focus on two aspects of this relationship: the
analysis of comparative advantage in female labor using both aggregate
and firm-level data and the role of trade openness in favoring female
entrepreneurship and ownership.
We focus specifically on MENA countries because they represent a
somewhat extreme case in international comparisons of de jure obstacles
to female employment and entrepreneurship and have low female labor
participation and low female entrepreneurship and ownership rates. We
use a novel approach to match the classification of country/industry
pairs with comparative advantage based on aggregate data with a
corresponding definition in firm-level data from the WBESs. Our approach
a. Producing fruit.
b. Conducive to productivity; causing to bear in abundance:
avenue for research when aggregate data are not
available and researchers are interested in identifying country/industry
pairs with and without comparative advantage in certain factors that may
be available in micro data (for example, high- and low-skilled labor).
We provide informal evidence of a link between a country’s
specialization and its measures of female labor participation consistent
with theories of brain-based technological bias and comparative
Finally, we use the classification of countries and industries
according to their comparative advantage to test a form of female
empowerment through export orientation. That is, we test whether women
are more likely to be entrepreneurs in industries characterized by
comparative advantage in female labor in countries where female labor is
relatively abundant. We find that women are more likely to be business
owners in female-labor-abundant countries when they face fewer de jure
constraints and operate in industries with a higher concentration of
female workers. The negative effect of de jure constraints is dwarfed by
being employed in a female-labor-intensive sector.
We leave to future researchers the broader extension of our results
to a model of international trade with a specific role for female labor
in countries beyond the MENA region.
Hiring Practices Related to the Gender of Workers and Managers in
According to our classification of countries based on the aggregate
data, Jordan is abundant in male labor relative to other countries in
the MENA group (Figure 10). The classification also reveals a relatively
high number of legal differentiations affecting female workers and
entrepreneurs (see Figure 1), a low female labor participation rate (see
Figure 2), and a low incidence of female ownership among entrepreneurs.
These findings are highly consistent with the firm-level data from the
Several indicators suggest that firms’ hiring preferences are
biased in favor of male workers. Here we exploit a unique section of the
Jordanian WBES that further details the respondents’ explanations
for such bias. They were directly asked whether they prefer to hire male
or female workers; the number of firms that preferred to hire men as
production workers is about 4.4 times larger than those that preferred
to hire women. This measure is in line with the country-level data
showing that the participation rate of male workers is about 4.3 times
larger than that of female workers (see Figure 2). More interestingly,
this bias varies among exporters and non-exporters (Figure B1) and
establishments with female ownership and those without female ownership
(Figure B2). As Figure B2 shows, exporters and establishments with
female ownership have a higher tendency to hire women and a higher
percentage of reporting managers
about gender preferences in
[FIGURE B1 OMITTED]
[FIGURE B2 OMITTED]
In addition, the surveyed firms reported the reasons for their
preferences for male or female workers; the major reasons tend to be
similar. They also list the three most important reasons for their
preference to hire men (or alternatively, women) after answering the
question on their gender-based recruiting preferences. Figure 133 plots
the percentage of answers to “the most important reason”
question. The top-rated reason is “productivity” Somewhat
surprisingly, the second most important reason is listed as
“other–specify” which once deconstructed consists essentially
of answers that can be combined as “nature of the
job/business” followed by “flexibility” and “lower
rates of absenteeism.” The fact that productivity ranks high for
both male and female workers supports the sector-specific technological
bias in favor of either male or female workers.
[FIGURE B3 OMITTED]
[FIGURE B4 OMITTED]
[FIGURE B5 OMITTED]
“Cultural reasons” play a less important role in
influencing the hiring process compared with other factors as this is
listed as the third most important reason to hire men as production
workers. When we pool the entire set of the three most important
reasons, “cultural reasons” ranks fifth among the major
reasons (Figure B4), accounting for 13 percent of total responses. Among
the cultural reasons, “Men have to support families” accounts
for 76 percent of the responses after pooling the three most important
reasons. This finding is highly consistent with the answers to the
binary gender differentiation question–“Can a married woman be
‘head of a household’ or ‘head of a family’ in the
same way as a man?”–in the WBL data.
Finally, the survey supports the view that technology and
industries may be brawn-labor biased (Figure B5) even if there is some
discrimination because of cultural values. After pooling the reasons,
“nature of the job/business,” which usually means that the job
requires physical effort, ranks third overall (see Figure B4) and is
second in “the most important reason” category (see Figure B3)
for firms that prefer to hire men as production workers. Figure B5 shows
that “nature of the job/business” is one of the leading
factors in male-labor-intensive industries (such as food, metals and
1. Not metallic.
2. Chemistry Of, relating to, or being a nonmetal.
and plastic materials, and other manufacturing)
in which female-to-total labor ratios are below average (see Figure 8)
in a male-abundant country such as Jordan. Therefore, the low female
participation rates may be partially explained by the higher demand for
brawn labor than brain labor in these industries.
Aguayo-Tellez, Ernesto; Airola, Jim and Juhn, Chinhui. “Did
Trade Liberalization Help Women? The Case of Mexico in the 1990s.”
NBER Nittany and Bald Eagle Railroad Company
Working Paper No. 16195, National Bureau of Economic Research, July
2010; www.nber.org/papers/w16195.pdf?new window=l.
Almeida, Heitor and Wolfenzon, Daniel. “The Effect of External
Finance on the Equilibrium Allocation of Capital” Journal of
Financial Economics, January 2005, 75(1), pp. 133-64.
Bernard, Andrew B.; Jensen, J. Bradford;
city (1990 pop. 66,462), seat of Shasta co., N central Calif., on the Sacramento River; inc. 1872. A principal tourist center for a mountain and lake region, it also has lumbering, food-processing, and diverse manufacturing.
, Stephen J. and
Schott, Peter K. “Firms in International Trade’ Journal of
Economic Perspectives, Summer 2007, 21(3), pp. 105-30.
Burstein, Ariel and Vogel, Jonathan. “International Trade,
Technology, and the Skill Premium” NBER Working Paper No. 16459,
National Bureau of Economic Research, 2012.
Crozet, Matthieu and Trionfetti, Federico. “Firm-Level
Comparative Advantage.” Working paper,
Paris School of Economics
November 2012; http://ces.univ-parisl.fr/membre/crozet/FirmComparativeAdv.pdf.
Do, Quy-Toan and Levchenko, Andrei A. “Comparative Advantage,
Demand for External Finance, and Financial Development” Journal of
Financial Economics, December 2007, 86(3), pp. 796-834.
Do, Quy-Toan; Levchenko, Andrei A. and Raddatz, Claudia.
“Engendering Trade.” Working Paper No.
5777, World Bank,
August 2011; www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2011/08/23/ 000158349_20110823155547/Rendered/PDF/WPS5777.pdf.
Do, Quy-Toan; Levchenko, Andrei A. and Raddatz, Claudia.
“Comparative Advantage, International Trade, and Fertility”
Discussion Paper No. 624,
University of Michigan
Research Seminar in
International Economics, February 15, 2012;
Galor, Oded and Well, David N. “The Gender Gap, Fertility, and
Growth.” American Economic Review, June 1996, 86(3), pp. 374-87.
Gayle, George-Levi; Golan, Limor and Miller, Robert A. “Gender
Differences in Executive Compensation and Job Mobility.”
, October 2012, 30(4), pp. 829-71.
Hallward-Driemeier, Mary and Aterido, Reyes. “Comparing Apples
with … Apples: How to Make (More) Sense of Subjective Rankings of
Constraints to Business “Working Paper No. WPS5054, World Bank,
September 2009; http://elibrary.worldbank.org/docserver/download/5054.pdf?expires=1354644540&id=id&accname=guest&
Harrigan, James and Reshef, Ariel. “Skill Biased Heterogeneous
Firms, Trade Liberalization, and the Skill Premium” NBER Working
Paper No. 17604, National Bureau of Economic Research, November 2011 ;
Keller, Elisa (2012). “Experience and the Gender Wage Gap
across Occupation” Unpublished manuscript,
University of Iowa
Leonardi, Marco. “Firm Heterogeneity in Capital-Labour Ratios
and Wage Inequality” Economic Journal, March 2007, 117(518), pp.
Melitz, Marc J. “The Impact on Trade on Intra-Industry
Reallocations and Aggregate Industry Productivity.” Econometrica,
2003, 71(6), pp. 1695-725.
Oostendorp, Remco H. “Globalization and the Gender Wage
Gap” World Bank Economic Review, January 2009, 23(1), pp. 141-61.
Rees, Ray and Riezman, Raymond G. “Globalization, Gender and
Growth” Review of Income and Wealth, March 2012, 58(1), pp. 107-17.
Rendall, Michelle. ”
Brain Versus Brawn
: The Realization of
Women’s Comparative Advantage “Working Paper No. 306,
, September 2010;
Romalis, John. “Factor Proportions and the Structure of
Commodity Trade” American Economic Review, 2004, 94(1), pp. 67-97.
Saure, Philip and Zoabi, Hosny. “Effects of Trade on Female
Labor Force Participation.” Working Paper No. 2009-12,
World Bank. World Development Report 2012: Gender Equality and
Development. Washington, DC: World Bank, 2011.
(1) These data are available at http://wbl.worldbank.org/.
(2) The survey questions reflected in these graphs are (i) Can a
married woman apply for a passport in the same way as a man? (ii) Can a
married woman travel outside the country in the same way as a man? (iii)
Can a married woman travel outside her home in the same way as a man?
(iv) Can a married woman get a job or pursue a trade or profession in
the same way as a man? (v) Can a married woman sign a contract in the
same way as a man? (vi) Can a married woman register a business in the
same way as a man? (vii) Can a married woman be “head of
household” or “head of family” in the same way as a man?
(viii) Can a married woman confer citizenship on her children in the
same way as a man? (ix) Can a married woman open a bank account in the
same way as a man? (x) Can a married woman choose where to live in the
same way as a man? Similar questions are available for unmarried women
but are not shown here.
(3) The databases are available at www.enterprisesurveys.org
(WBES), www.unido.org/Datal/Statistics/Databases/ISAV.cfm?dig=2 (UNIDO
INDSTAT2), and http://comtrade.un.org/ (UN COMTRADE).
(4) Industry classification changes across countries, although it
is fairly consistent for the five MENA countries available.
(5) Data at the industry level are available and comparable for
Egypt, Jordan, Morocco, Turkey, and–with some limitations-Algeria.
(6) The interaction between FLP and FL provides a measure of
comparative advantage. Similar indicators are used in Romalis (2004) and
Do, Levchenko, and Raddatz (2011).
(7) The female-labor-intensive industries are chemicals and
chemical products; leather, leather products, and footwear; textiles;
medical, precision, and optical instruments; electrical machinery and
apparatus; radio, television, and communication equipment; and wearing
apparel, fur. The male-labor-intensive sectors are basic metals; wood
products (excluding furniture);
tr.v. fab·ri·cat·ed, fab·ri·cat·ing, fab·ri·cates
1. To make; create.
2. To construct by combining or assembling diverse, typically standardized parts:
metal products; other
transport equipment; nonmetallic mineral products; furniture,
manufacturing not elsewhere classified; motor vehicles, trailers,
semi-trailers; machinery and equipment not elsewhere classified; rubber
and plastics products; paper and paper products; coke, refined petroleum
products, and nuclear fuel;
the process of recovering and reusing waste products—from household use, manufacturing, agriculture, and business—and thereby reducing their burden on the environment.
; food and beverages; office,
machinery; tobacco products; and printing and
(8) There is little research explaining this within-industry
dispersion of factor intensities in a trade context. At least two
factors of production and heterogeneous productivity must be
tr.v. pos·tu·lat·ed, pos·tu·lat·ing, pos·tu·lates
1. To make claim for; demand.
2. To assume or assert the truth, reality, or necessity of, especially as a basis of an argument.
to generate within-industry dispersion of factor intensity. In an open
economy, Bernard et al. (2007), Harrigan and Reshef (2011), and Burstein
and Vogel (2012) model heterogeneous firms that use multiple inputs in
the production function, thereby moving beyond the standard Melitz
(2003) approach. Both papers focus on two inputs–high- and low-skilled
labor–and discuss the relationship between factor intensity and trade.
Bernard et al. (2007) show that standard Heckscher-Ohlin results carry
over to a framework with heterogeneous total factor productivity.
Burstein and Vogel (2012) show how heterogeneous firms’ decisions
shape the factor content of trade, the changes in relative factor
prices, and between-sector factor allocations as a response to trade
liberalization. However, neither of these papers applies the theory to
the data. Crozet and Trionfetti (2012) also develop a
of trade with heterogeneous firms in which the
individual factors (capital and labor) is heterogeneous, and, therefore,
unlike the model of Bernard et al. (2007), the two-factor model
translates to Hicks-biased technology. This heterogeneity can generate a
within-industry dispersion that is then measured in firm-level data from
the AMADEUS database.
(9) Problematic access to capital, competition from the informal
sector, high tax rates, political instability, the lack of skilled or
educated workers, access to electricity, and corruption account for
almost 90 percent of all problems reported. The remaining 10 percent of
problems include difficulty in registering the firm and obtaining
licenses, labor regulations, access to land, customs regulations,
macroeconomic instability, tax administration, transportation, economic
uncertainty, difficulties in dealing with the legal system, conflict
resolution, and crime.
(10) Recent research shows that qualitative self-reported measures
are representative of objective conditions of the firms. For example,
Hallward-Driemeier and Aterido (2009) show that subjective measures are
significantly correlated with objective measures of problematic access
to electricity, skills shortage, and other variables.
Silvio Contessi is an economist and Li Li is a research associate
at the Federal Reserve Bank of St. Louis. Francesca de Nicola is a
Of, relating to, or engaged in academic study beyond the level of a doctoral degree.
fellow at the
International Food Policy Research Institute
The authors are grateful to Ali Bayar, Subhayu Bandyopadhyay, Maria
Canon, Elisa Keller, Marcella Nicolini, Jeff Nugent, and participants to
the MEDalics workshop on “Multinational Strategies, Economic
Integration, and Knowledge Transfer in the Mediterranean Area” for
Table 1 Distribution of Firms (by Industry and MENA Country) Country Sector Algeria Egypt Jordan Morocco Textiles 12.47 16.91 2.85 11.60 Garments 12.21 24.50 22.98 Food 31.43 26.50 26.26 Metals and machinery 28.05 19.89 6.27 0.66 Electronics 2.34 12.01 1.99 0.44 Chemicals and pharmaceuticals 8.05 7.62 10.26 20.79 Nonmetallic and plastics 0.00 11.66 9.12 0.00 Other manufacturing 5.45 31.90 18.52 17.29 Total 100 100 100 100 Country Sector Turkey Yemen Total Textiles 19.15 17.77 Garments 16.00 10.04 15.25 Food 13.29 29.28 14.16 Metals and machinery 18.48 16.13 18.2 Electronics 1.39 0.09 1.67 Chemicals and pharmaceuticals 4.87 3.30 5.15 Nonmetallic and plastics 11.38 25.60 11.86 Other manufacturing 15.45 15.57 15.93 Total 100 100 100 NOTE: MENA data refer to individual countries' fiscal years (specified in Table 2). Figures indicate percentages. SOURCE: World Bank Enterprise Surveys. Table 2 Summary Information on the WBES Survey Fiscal No. of Country year year observations Algeria 2007 2006 423 Egypt 2008 2007 1,156 Jordan 2006 2006 352 Lebanon 2009 2008 140 Morocco 2006 2005 466 Syria 2009 2008 349 Turkey 2008 2007 896 Yemen 2009 2009 241 Total MENA 8 4,023 Percent Percent of total Population of total Country MENA 8 (millions) MENA 8 Algeria 10.51 36 12.84 Egypt 28.73 82.5 29.43 Jordan 8.75 6.2 2.21 Lebanon 3.48 4.3 1.53 Morocco 11.58 32.3 11.52 Syria 8.68 20.8 7.42 Turkey 22.27 73.4 26.19 Yemen 5.99 24.8 8.85 Total MENA 8 100 280.3 100 SOURCE: World Development Indicators and World Bank Enterprise Survey data. Table 3 Determination of Country/Industry Pairs: Comparative Advantage Industries Using Firm-Level Data Country Sector Turkey Egypt Morocco Jordan Algeria Panel A: fmlr Textiles 0.39 0.21 1.71 0.72 0.57 Garments 1.02 2.96 0.94 0.72 Food 0.25 0.41 0.08 0.20 Metals and machinery 0.11 0.08 0.06 0.03 0.09 Electronics 0.19 0.79 1.11 0.07 0.19 Chemicals and pharmaceuticals 0.19 0.45 0.56 0.17 0.09 Nonmetallic and plastic materials 0.14 0.20 0.02 Other manufacturing 0.06 0.27 0.42 0.11 0.11 [fmlr.sub.c] 0.303 0.256 1.070 0.431 0.199 Country classification ML-AB ML-AB FL-AB FL-AB ML-AB Panel B: DV-CA micro Textiles 0 0 1 1 0 Garments 0 1 1 0 Food 1 0 0 1 Metals and machinery 1 1 0 0 1 Electronics 0 0 1 1 0 Chemicals and pharmaceuticals 1 1 0 0 1 Nonmetallic and plastic materials 1 1 0 Other manufacturing 1 1 0 0 1 Panel C: DV-CA macro Textile 0 0 1 1 Garments 0 0 1 1 Food 1 1 0 0 Metals and machinery 1 1 0 0 Electronics 0 0 1 1 Chemicals and pharmaceuticals 0 0 1 1 Nonmetallic and plastic products 1 1 0 0 Other manufacturing 1 1 0 0 Country classification ML-AB ML-AB FL-AB FL-AB ML-AB Country Industry Sector Yemen fmlr classification Panel A: fmlr Textiles 0.38 FLINT Garments 0.10 1.05 FL-INT Food 0.07 0.24 ML-INT Metals and machinery 0.01 0.10 ML-INT Electronics 0.00 0.36 FL-INT Chemicals and pharmaceuticals 0.53 0.28 ML-INT Nonmetallic and plastic materials 0.03 0.14 ML-INT Other manufacturing 0.04 0.10 ML-INT [fmlr.sub.c] 0.068 0.305 Country classification ML-AB Panel B: DV-CA micro Textiles Garments 0 Food 1 Metals and machinery 1 Electronics 0 Chemicals and pharmaceuticals 0 Nonmetallic and plastic materials 1 Other manufacturing 1 Panel C: DV-CA macro Textile Garments Food Metals and machinery Electronics Chemicals and pharmaceuticals Nonmetallic and plastic products Other manufacturing Country classification ML-AB NOTE: The three panels compare the determination of country- industry pairs with comparative advantage (identified as "1 ") using aggregate and firm-level data. FL-AB, female- labor-abundant relative to the group of countries; ML-AB, male-labor-abundant relative to the group of countries; FL-INT, female-labor-intensive relative to the manufacturing sector as a whole; ML-INT, male-labor-intensive relative to the manufacturing sector as a whole. Table 4 "Do Business Constraints Have a Differential Impact on Firms Managed by Women?" Constraints Macro Informal Controls uncertainty Tax rate Corruption practices Women manager? 0.19 ** 0.01 -0.01 -0.15 Yes = 1 (0.07) (0.15) (0.20) (0.27) Exporter? 0.21 ** 0.19 0.19 0.02 Yes = 1 (0.09) (0.14) (0.20) (0.16) No. 1,192 2,145 2,124 2,277 Not enough Controls skilled Electricity Tax women administration Women manager? -0.15 0.07 -0.21 Yes = 1 (0.28) (0.20) (0.27) Exporter? 0.1 -0.14 0.29 ** Yes = 1 (0.14) (0.18) (0.15) No. 2,334 2,303 2,140 Controls Access Labor Business Customs to capital regulations licensing regulations Women manager? 0.27 * 0.13 -0.05 -0.19 Yes = 1 (0.15) (0.26) (0.15) (0.20) Exporter? -0.33 ** 0.18 0.12 0.57 *** Yes = 1 (0.15) (0.15) (0.20) (0.19) No. 1,851 2,314 2,131 1,870 Controls Legal system Transportation Telecommunications Women manager? -0.13 -0.07 -0.22 ** Yes = 1 (0.28) (0.08) (0.09) Exporter? 0.29 -0.14 0.05 Yes = 1 (0.20) (0.17) (0.10) No. 1,980 1,657 1,161 Controls Crime Women manager? 0.16 Yes = 1 (0.20) Exporter? -0.01 Yes = 1 (0.17) No. 2,251 NOTE: Ordered probit, marginal effect. The regressions include the following variables (not displayed): dummy for exporter, age of the firm, total employment (log), total sales (log), and labor productivity (log). Robust standard errors are reported in parentheses. *, **, and *** indicate significance at the 10 percent, 5 percent, and 1 percent levels, respectively. Table 5 "Do Business Constraints Have a Differential Impact on Firms Owned by Women?" Constraints Macro Informal Controls uncertainty Tax rate Corruption practices Women owner? 0.01 -- 0.18 -0.08 Yes = 1 (0.14) -- (0.14) (0.16) Exporter? 0.06 -- 0.22 0.03 Yes = 1 (0.11) -- (0.19) (0.16) No. 605 -- 1,564 1,700 Constraints Not enough skilled Tax Access to Controls women Electricity administration capital Women owner? 0 -0.27 * -- 0.25 ** Yes = 1 (0.13) (0.14) -- (0.12) Exporter? 0.11 -0.17 -- -0.31 ** Yes = 1 (0.15) (0.18) -- (0.15) No. 1,776 1,772 -- 1,682 Constraints Business Customs Legal Controls regulations licensing regulations system Women owner? 0.04 0.32 ** -0.16 0.01 Yes = 1 (0.15) (0.14) (0.16) (0.15) Exporter? 0.2 0.14 0.60 *** 0.28 Yes = 1 (0.15) (0.20) (0.20) (0.20) No. 1,760 1,579 1,581 1,104 Constraints Controls Transportation Telecommunications Crime Women owner? 0.11 -0.19 -0.01 Yes = 1 (0.36) (0.15) (0.16) Exporter? -0.24 0.14 -0.01 Yes = 1 (0.25) (0.11) (0.17) No. 772 614 1,694 NOTE: Ordered probit, marginal effect. The regressions include the following variables (not displayed): dummy for exporter, age of the firm, total employment (log), total sales (log), and labor productivity (log). Robust standard errors are reported in parentheses. *, **, and *** indicate significance at the 10 percent, 5 percent, and 1 percent levels, respectively. Table 6 Does Trade Empower Female Managers or Owners? Regressors Women owner? Women manager? Yes = 1 Yes = 1 Firm-level 0.21 * 0.10 ** female labor (0.12) (0.04) share Female-labor- 0.04 -0.03 intensive (0.06) (0.03) sector Log -0.61 -0.62 -0.33 -0.38 (Employment) (0.63) (0.62) (0.26) (0.25) Age 0 0 0 0 (0.00) (0.00) (0.00) (0.00) Log (MPL) -0.67 -0.69 -0.33 -0.40 * (0.61) (0.60) (0.25) (0.24) Log (Sales) 0.65 0.68 0.31 0.38 (0.61) (0.60) (0.25) (0.24) Mean dependent 0.37 0.37 0.08 0.07 variable No. 2,185 1,921 2,336 2,076 NOTE: Probit, marginal effect. Robust standard errors are reported in parentheses. MPL, marginal productivity of labor. * and ** indicate significance at the 10 percent and 5 percent levels, respectively. Table 7 Trade or De Jure Constraints: What Empowers Female Managers or Owners? Female Owners? Yes = 1 Regressors FL-AB Countries (Morocco and Jordan) Index of de -0.09 * -0.12 ** 0 -0.05 jure constraints (0.05) (0.06) (0.04) (0.05) Female-to-male 0.05 *** -0.05 labor ratio (0.02) (0.06) Index of de jure 0.10 * constraints x (0.06) Female-to-male labor ratio Female labor 0.05 * -0.04 intensive (0.03) (0.05) Dejure constraints 0.13 ** x Female labor (0.06) intensive Log(Employment) 0.34 ** 0.33 ** 0.30 ** 0.31 ** (0.17) (0.16) (0.15) (0.15) Age 0 0 0 0 (0.00) (0.00) (0.00) (0.00) Log(MPL) 0.33 ** 0.32 * 0.28 * 0.29 * (0.17) (0.16) (0.15) (0.15) Log(Sales) -0.33 ** -0.32 ** -0.28 ** -0.29 * (0.16) (0.16) (0.14) (0.15) Mean dependent 0.14 0.14 0.14 0.14 variable No. 775 775 775 775 NOTE: Probit, marginal effect. Robust standard errors are reported in parentheses. FL-AB, female-labor-abundant relative to the group of countries; MPL, marginal productivity of labor. *, **, and *** indicate significance at the 10 percent, 5 percent, and 1 percent levels, respectively.