CHANGE PROCESS
T h e J o u r n a l o f D e v e l o p i n g A r e a s Volume 48 No. 1 Winter 2014
GENDER-BASED WAGE AND
OCCUPATIONAL INEQUALITY IN THE NEW
MILLENIUM IN EGYPT
Fatma El-Hamidi
University of Pittsburgh, USA
Mona Said
American University in Cairo, Egypt
ABSTRACT This paper adds to the existing literature on the Egyptian labor market by examining the extent to which the treatment of women in the Egyptian private labor market has evolved, and if occupational segregation has affected gender wage gaps in the newly transformed Egyptian economy. The paper tracks the current trend of gender based wage gaps, and evaluates the role of occupational segregation in explaining these gaps in the Egyptian private labor market. Comparing the years 2000 and 2004, and arranging occupations in three broad categories, findings of this study
point to a wider occupational segregation and increased crowding of women in few jobs, which are becoming a more serious issue in pay differences than pure pay discrimination for both professional and blue collar women. Pay discrimination for white collar workers is not as severe as in professional and blue collar jobs. Therefore, policies that target inter-occupational components to close the wage gap may have far-reaching effects on professional and blue-collar workers, whereas policies targeting equal pay for equal jobs will have a greater success for white-collar workers. JEL Classifications: J71, J16, J24, J31 and C20
Keywords: Occupational segregation; wage decomposition; gender wage gap; Egypt Corresponding Author’s Email Address: mona_said@aucegypt.edu
INTRODUCTION
It is well-documented in the literature that labor market outcomes differ significantly
along gender lines in most, if not all, developing economies. Differences range from
participation rates in various economic activities, to occupational choice and sectoral allocation, to unemployment, as well as wages, an important indicator of the economic
well-being and of personal success. Thus, the level of gender pay gap is a revealing
indicator about women’s progress in the labor market and their status in the household.
International reports place Egypt low on gender related indexes. Out of 130
countries, Egypt ranks 105 in educational attainment, 124 in gender gap index, and
ranks129 in terms of labor force participation (Hausmann et al, 2008). In fact, the general
picture emerging from the recent literature on the Egyptian labor market renders
significant barriers to paid work for women in the private sector. Compared with the
private sector, there are numerous grounds for which the public sector, which has a
positive discrimination in favor of women, would be more appealing for female
employment once the decision to join the labor market is made. Given the recent
22
diminishing role of the public sector as the employer of last resort, and the increasingly
key role the private sector is set to play in employment generation, prospects of future
employment for women are uncertain.
Moreover, since becoming a member of the WTO in 1995, Egypt has pledged to
be in full compliance with its trade commitments by 2005. To further open its economy,
Egypt has signed several unilateral and multilateral trade agreements, liberalized its tariff structure, and reduced tariff rates for most imports. Becker’s theory of “employer taste
for discrimination” (1957) established a link between inequity and free trade. In a
noncompetitive market, excess profits allow employers to “purchase” or practice
discrimination. Free trade, on the other hand, brings in competitiveness and exposes firms
to a wider market, forces out discriminating employers, increases competition to cut
costs, and eventually reduces gender wage gaps. Yet, others such as Berik et al. (2004)
suggest an increase in trade can in fact increase gender wage premia in regions where
female workers may have lower bargaining power and /or segregated into lower-paying,
lower-status jobs.
Such outcomes seem contingent on level of development of the country in
question. Rising wage inequality in general has become increasingly evident in
industrialized countries over the past few decades, and has come to be associated with the stark decline in collective wage bargaining. However, small reductions in the gender
wage gap continue to be observed at the bottom of the wage distribution due to declines
in male wages in countries such as Germany for example (Antonczyk et al. 2010). The
predominant explanation advanced in the literature is the fall in demand for unskilled
labor on account of skill biased technological change, increased international trade and
outsourcing.
In the developing world, where labor market institutions are generally weak,
increased competition from trade is often accompanied by increased wage discrimination
in both tradable and non-tradable sectors. Oostendorp’s (2009) large cross country study
of the impact of globalization on the gender wage gap, also finds that increasing trade and
FDI reduce the occupational gender wage gap in richer countries, but little evidence that the same is true in poorer countries. Thus, although trade liberalization has the ability to
enhance the efficiency and competitiveness of host economies, it also has the potential
risk of creating or worsening inequities for the poor and for women.
The existing literature on the Egyptian labor market is largely confined to recent
available data collected in the 1980s and 1990s. Several studies have looked at the effects
of structural adjustment on gender wage gaps in the Egyptian labor market: Nassar
(1998); Assaad and Arntz (2005); Assaad and El-Hamidi (2001 and 2008); Said (1999
and 2003) and El-Hamidi (2003 and 2006). Some of these studies documented that paid
female employment in the private sector lags drastically behind the growth of the female
labor force and well behind the growth of male employment in that sector as well, which
suggests that there are considerable barriers to women employment in the private sector.
Assaad and Arntz (2005) identify a total of just nine job types constituting 95 percent of female nongovernmental paid work. Such overcrowding of female
employment in a limited number of work fields causes a downward pressure of wages. A
comparison between 1988 and 1998 data indicated that these few limited employment
fields for women are being further de-feminized.
23
In the decade between 1988 and 1998, Said (2003) demonstrates that gender
wage inequalities have risen for women. Most of the gender wage differential in Egypt is
explained by intra-occupational segregation and women’s segregation to less profitable
jobs (Said 2003; World Bank 2004). By the early 1990s, there existed a substantial wage
differential between men and women, ranging from 7 percent of female wages in the
public enterprise sector to 39 percent in the private sector. Not only are differentials bigger in the private sector, but studies of their composition draw attention to the
presence of substantial unjustified discrimination (Said 1999; 2003). Ninety percent of
the differential in the public sector is attributable to differences in productivity, whereas
this is true of less than 10 percent in the private sector. The remaining portion of the wage
premium includes intra-occupational pay discrimination (i.e., different pay for a
comparable job) and segregation that leads women to more likely work in sectors with
low productivity.
More recent evidence suggest further widening in the gender gap in the era of
trade liberalization in Egypt, with the an increase in the unjustified wage differential that
is especially pronounced for high earners in 2006 (El-Haddad, 2009; Kandil, 2009).
Education and skill accumulation of females are integral factors in determining the
impact of trade on women’s employment and the gender wage gap. The findings of the above studies indicate that, in Egypt, the increase in skills of women in the labor force
did not lead to a relative or absolute decrease in discrimination. This raises the question
of whether implicit barriers to mobility continue to hinder the entry of women into higher
paying jobs.
Therefore, it is of great concern to learn to what extent the treatment of women
in the Egyptian private labor market has evolved and if occupational segregation has
affected gender wage gaps in the newly transformed Egyptian economy. Using micro
data for years 2000 and 2004, the purpose of this study is to: 1- track and evaluate the
status of the gender-based wage gaps in the Egyptian private labor market; 2- examine
the role of occupational segregation in explaining gender wage gap in the Egyptian
private labor market. The rest of this paper is organized as follows. Section II describes the data used
and provides sample characteristics. Section III outlines the estimation methodology and
empirical findings. Section IV Concludes and outlines some policy implications.
DATA AND SAMPLE CHARACTERISTICS
To consider the role of occupational segregation in explaining wage differentials we use
recent Labor Force Sample Surveys 1 (LFSS) from the Central Agency for Public
Mobilization and Statistics (CAPMAS), on the labor market for the years 2000 and 2004.
The data is very rich and unique in its coverage and consistency, with a large sample size
that ranges between 170,000 and 340,000 individuals. The sample includes 360 urban
clusters and 240 rural clusters, a total of 600 clusters, each containing 70 households. A simple analysis of the data reveals a key framework of the Egyptian labor
market. By and large, in 2004, the formal private sector has employed 68 percent of the
labor force; the government sector contracted 28 percent; whereas the public sector
secured a meager 5 percent of wage workers. The majority of women in 2004 (76%) were
recorded as private sector agriculture workers. Since a sizeable number worked as non -
24
paid family workers, and due to agriculture sector’s seasonal nature, in addition to the
recent retrenchments of public and government sector employment, the analysis is
restricted to non-agriculture private sector workers.
A simplistic approach in analyzing the occupational distribution is to group
employment by occupation into three categories, with the assumption that skills are
relatively homogenous within each category: professional workers (i.e., legislators, managers, health professionals and educators); white-collar workers (i.e., technical
assistance, clerks and sales and services) and blue-collar workers (i.e., vocational,
production workers and others) 2 . On average, women seem to be equally distributed
across the three occupational categories as follows: 32 percent in professionals, 38
percent in white collars, and 30 percent in blue collar jobs.
Table (1) presents the share of non-agriculture private employment by gender
and occupational categories. The overall employment of non-agriculture private sector
increased by 11% between 2000 and 2004, with almost equal growth for men and women
(12% for women and 11% for men). Women’s employment gain benefited professional
and blue collar workers at the expense of white collars. Men, on the other hand, gained
employment in professional categories and lost employment in white and blue collar jobs.
TABLE 1. SHARE OF NON-AGRICULTURE PRIVATE EMPLOYMENT BY
GENDER AND OCCUPATIONAL CATEGORIES (%)
Occupational
Categories Females Males Females Males
Professionals 2.50 24.55 2.59 27.10
White Collar 3.13 17.27 3.00 16.73
Blue Collar 2.31 50.24 2.40 48.18
Sub-Total 7.93 92.07 8.00 92.00
Total 100 100
2000 20 04
Source: Authors’ own calculations; LFSS 2000-2004.
According to Assaad and Arnetz (2005), prior to 2000, the Egyptian economy
witnessed a decline or de-feminization trend in private sector employment, which was
attributed to the falling share of foreign trade in the economy, and in particular, to the
declining share of labor-intensive manufacturing exports, especially textiles and
garments. Preliminary findings of a recent study utilizing 2006 ELMPS data in Egypt
show that after a period of de-feminization (1988-1998), some sectors operated as
magnets to women during the period between1998 and 2006. The most dramatic increase in the female share was in the textiles and garments sector, where the female share
doubled from 15 percent in 1998 to 30 percent in 2006, followed by food manufacturing.
This development has helped compensate for the declining share of female employment
in both government and public sectors (Assaad and El-Hamidi, 2008).
In fact, around year 2000, three main developments in the Egyptian macro scene
may have contributed to increased employment of women, especially in blue collar
occupations. First, the Egyptian economy observed a shift in its internal structure of the
GDP between 2000 and 2004, with both manufacturing and transportation increasing
25
their share of the GDP by 11 percent. In fact, seven major industries accounted for over
80 percent of growth in the manufacturing sector, most importantly textile, food and
beverage.
Second, following the introduction of partial floating of the exchange rate, the
Egyptian pound depreciated by 25 percent against the US Dollar, and the economy was
said to be driven by export revenues. Food industries alone experienced a growth of its values of exports reached 43% between 2000 and 2004, whereas textile industries gained
22% of its value of exports. ( World Bank Trade Stats). Third, the adoption of a package
of trade liberalization and policies that ranged from reduction in import tariffs to reducing
trade barriers through agreements such as QIZ (Qualified industrial zones), COMESA,
GAFTA and Euromed partnership.
How has this, albeit slight, increase in women’s employment in the private
sector affected gender wage differentials between 2000 and 2004, if any? To answer this
question, we look at Table (2) with some characteristics of the working sample. Table (2)
shows that in occupations where women increased their share of employment during the
2000-2004, such as professionals and blue collar jobs, the gender wage gap declined, and
the decline was greater in blue collar than professionals’ jobs. For example, women in
blue collar jobs earned 71% of median males’ wages in 2004, compared with 67% in 2000. Professional women earned 75% of median male wages in 2000 and 78% in 2004.
Contrarily, gender wage gap has widened for white collar employees. What’s interesting
to note is that while real wages of both gender have dropped during the five year span in
all occupations, the drop in women’s real wages was lower than that of men in
professional and blue collar jobs, whereas the opposite occurred in white collar jobs. In
other words, it seems that wages of both genders are getting closer to each other at the
expense of men’s wages in professionals and blue collar occupations, as opposed to at
women’s expense in white collar occupations. Some may argue that wage differences
may be due to differences in human capital (i.e. education and/or experience). The table
shows that males, on average, have more years of experience 3 than females, which is
accredited to the fact that women experience frequent detachment from the labor market for reasons related to marriage, childbearing, and/or caring for the elderly. Nevertheless,
a closer look at these figures reveals that women’s earnings are still lower than their
males’ counterparts regardless of the occupation, even when women accumulate more
years of education, as indicated in professional jobs.
Here, economic theory and econometric analysis come handy in explaining
gender wage differentials. The literature on wage discrimination explains how wage
differentials can be reduced to a part that is attributed to differences in human
productivity related factors such as education, experience, age …etc, and a part that is
attributed to occupational segregation. A detailed methodology and analysis are presented
in the following section.
26
TABLE 2. SELECTIVE CHARACTERISTICS OF NON-AGRICULTURAL
PRIVATE SECTOR WORKERS BY OCCUPATIONAL GROUPS
Occupational Categories Female-2000 Male-2000 Female-2004 Male-2004
Professionals
Median Real Hourly Wage (EGP) 1.88 2.50 1.79 2.31
Mean Age (Years) 36.54 42.15 37.46 43.18
Mean Years of Experience 19.66 20.13 21.01 20.61
Mean Years of Schooling 10.57 8.22 11.18 8.25
% (Females/Males) Wages 75% 78%
White Collars
Median Real Hourly Wage (EGP) 1.25 1.75 1.03 1.64
Mean Age (Years) 31.55 32.91 30.94 33.78
Mean Years of Experience 18.62 16.97 20.41 20.44
Mean Years of Schooling 6.86 7.04 8.20 7.67
% (Females/Males) Wages 71% 63%
Blue Collars
Median Real Hourly Wage (EGP) 1.25 1.88 1.23 1.74
Mean Age (Years) 31.18 33.35 34.14 34.02
Mean Years of Experience 21.23 24.28 23.68 23.77
Mean Years of Schooling 3.89 4.10 3.85 4.76
% (Females/Males) Wages 67% 71% Source: Authors’ own calculations; LFSS 2000-2004.
ESTIMATION METHODOLOGY
Occupational Attainment
The previous section offered a conventional analysis which returned several conclusions.
These key results, however, are not informative about the actual occupational and gender
differentials as they do not take account of differences in individual and job
characteristics. Determining the role of occupational segregation in explaining gender based wage gap deserves close examination. In reality, there is growing awareness,
supported by empirical evidence in many countries, that the pay in occupations
dominated by women is lower even when the effect of variables such as the different
levels of education or years of experience required are taken into account. Hirsch (2007)
argues that differences in gender specific mobility patterns caused, e.g., by domestic
responsibilities can generate less elastic female labor supply at the firm level, than men.
Bowlus (1997) also demonstrates that a significant part of the gender pay gap can be
explained in terms of women facing higher search frictions in labor markets.
The empirical analysis in this section follows Brown et al (1980) and proceeds
at two levels. First, a reduced form multinomial logit model is utilized to determine
occupational attainment and capture how certain variables affect the probability of an
individual working in a specific occupation. The model assumes that individuals determine their choice of occupational attainment, and so they face a number of mutually
27
exclusive alternatives when it comes to joining the labor market. As Meng (1998)
explains, the individual’s occupational attainment is a function of an employer's
willingness to hire that person (which is determined by the individual’s human
endowment) and the individual's desire to work in a particular occupation. The
determinants of occupational attainment are estimated separately 4 for male and female
non-agriculture private sector workers for the years 2000 and 2004, taking into consideration selection bias by using Heckman (1979) two step selection model
5 . This
step allows determining if the degree of occupational segregation is due to differences
between men and women in personal characteristics by predicting the distribution of
women across occupations if they were treated in the same way as men.
The functional form of the multinomial logit model is employed as follows 6 :
i=1, …, N; j = 1, …,J (1) where N= sample size
J= number of occupational groups
x= a vector of independent variables affecting occupational attainment, and
j is the parameter vector for occupational choice (j).
This step is followed by constructing the selection term as follows:
j = (Hj) / (Hj) (2)
where Hj = -1
(Pj); and are the standard normal density and distribution functions respectively.
Explanatory variables that enter into the multinomial logit model include
variables that determine the reservation wage such as: experience 7 , education dummies
and age. Levels of education are captured by six dummy variables: Illiterate (base), read
and write, primary, middle, secondary and university. It is assumed that higher
educational attainment imply selecting a professional occupation, followed by white
collar jobs. Regional differences in occupational characteristics are captured by regional
dummies as follows: Metropolitan (base), lower urban, upper urban, lower rural, upper
rural. Other control variables (affecting the supply side) are family background
characteristics such as: marital status, number of children below six years of age, the size
of the household. It is expected to find the presence of young kids and the greater the size
of the household to be associated with choosing a less risky/less demanding type of job.
In addition, the proportion of workers in the public and government sectors by governorate is added as a proxy for the size of demand on employment from
public/government sector side. Relative risk ratios (RRR) of the multinomial logit model
of occupational attainment are reported in the Appendix, Table (A-1), where blue collars
occupation is the base category, against which professionals and white collars groups are
compared.
28
Findings and Discussions
Table (A-1) in the appendix reveals that almost all coefficients are statistically
significant at least at the 10% level. Figure (A-1) in the Appendix summarizes the results
of table (A-1). The figure presents a simple display of the likelihood of obtaining a certain occupation relative to the base category (i.e. blue collars). The line extending
from the center of the radar to the outer surface represents a positive likelihood of
obtaining a professional or white collar occupation relative to blue collar occupation. No
line signifies a less likelihood of obtaining a professional or white collar job relative to a
blue collar one.
As panel A (professional vs. blue collar occupations) of figure (A-1) shows, the
occupational attainment pattern for women changes slightly between 2000 and 2004. For
example, in 2000, older, married women, living in any region, with a large household
size, carrying a university education, were more likely to obtain a professional
occupational job. By 2004, Married women, living in upper rural regions were less likely
to work in professional occupations. Having children in the age group between zero and
six was also positively associated with obtaining a professional job for both gender in 2004.Panel B (white collar vs. blue collar occupations) of the same figure displays the
following: For white collar workers, in 2000, older married women with a large
household and university degree were more likely to obtain a white collar job. By 2004,
women with intermediate and secondary education, along with those with young children
were more likely to work in white collar jobs as well. For males, in 2004, and in addition
to the significance of the share of public and government employment as a determining
factor in obtaining a white collar job, the older the worker, with secondary or university
education, and a sizeable household, the more likely he obtains a white collar job relative
to a blue collar one. In fact, these findings are in agreement with preliminary findings of
2006 ELMS study by Assaad and El-Hamidi (2008) pointing to the fact that most of the
female workers who have joined the manufacturing sector are young unmarried secondary school graduates still living with their parents.
The structural difference in occupational allocation between men and women,
reported in the Table (1) above, suggests they may be treated differently in the private
labor market. To evaluate the degree to which women are treated differently with regards
to occupational allocation, we predict the distribution for women using the estimated
parameters of the occupational attainment model for men. That is, the occupational
attainment that women would have realized if their personal characteristics were treated
the same way as their male counterparts. In other terms, controlling for individual
characteristics in occupational attainment allows to measure the difference between
actual and predicted women’s occupational attainment, that is a measure of the portion of
the occupational segregation due to differences between men and women in the way their
personal characteristics are treated or valued, and in turn is considered a measure of discrimination in occupational attainment. Table (3) provides predicted probabilities of
women’s occupational attainment.
Table (3) indicates that in 2004 if women were to face the same occupational
structure as that estimated for men, about 4 percent less women would have obtained
professional jobs; roughly 8 percent more women would have acquired white-collar jobs
29
and almost 4 percent fewer women would have been engaged in blue-collar jobs. In other
terms, it appears that women are overrepresented in professional and blue-collar jobs, and
underrepresented in white collar jobs.
TABLE 3. ACTUAL AND PREDICTED PROBABILITIES OF FEMALE
OCCUPATIONAL DISTRIBUTION
Occupational Categories Predicted 2000 Actual 2000 Predicted 2004 Actual 2004
Professional 28 33 30 34
White Collar 49 40 46 38
Blue Collar 23 27 24 28 Source: Authors’ own calculations; LFSS 2000-2004.
Wage Determination and Occupational Segregation
After estimating the predicted probabilities of female occupational distribution, the
selection term constructed from the first step is entered linearly into the second step: the wage equation. The dependent variable in the wage equation is the log real hourly
wages 8 . The model therefore is:
LnWij = Xijj+ ij j +uij (3)
Where Xij represents the characteristics (e,g. education, experience, and region) of the
individual, 0 and j are parameters to be estimated and u is a random error term. Following Mincer (1974), we use level of education and years of experience
(EXP) as the main explanatory variables. Levels of education are captured by six dummy
variables: Illiterate (base), read and write, primary, middle, secondary and university.
Regional differences are captured by five dummies as follows: Metropolitan (base), lower
urban, upper urban, lower rural, upper rural 9 . Experience variables are included in the
model since workers with more years of job experience are likely to earn more. A firm is
likely to offer higher wages to induce experienced workers to stay on in their jobs, as the
cost of training new workers could be very expensive. The experience squared variable is included to capture the possibility of a non-linear relationship between experience and
earnings. We expect a positive sign of the experience variable for the reason that working
experience is likely to contribute to enhancement of individual’s human capital, and
negative coefficient of experience square as marginal returns from experience tend to
decline over the lifetime. Wage equations are estimated separately for males and
females 10
.
Next, we follow the decomposition model developed by Brown et al (1980)
which accounts for occupational distribution and distinguishes between across-
occupation and within-occupation wage differences in the analysis of wage differentials.
The model can be written as follows:
30
(WJ) (WD)
Within/Intra-occupational
(BJ) (BD)
Across/Between/Inter-occupational (4)
Where: A bar over a variable denotes the mean value, Superscripts m and f refer to male worker and female worker, respectively.
and
are the observed proportion of male and female workers in
occupation j.
measures the proportion of the sample of female workers who would be in occupation j if female workers were allowed the same occupational choice as
male employees.
W refers to Within; and B refers to Between; D refers to Discrimination; and J refers to Justified; WJ and WD are the explained and unexplained within-occupation wage
differences respectively. BJ and BD represent the explained and unexplained portions of the inter-
occupational wage differences respectively.
Overall, the mean log wage difference shown in the previous equation consists
of four distinct components. The ‘justified” or explained term, refers to wage differentials
resulting from gender differences in productivity-related characteristics. The ‘unjustified”
or unexplained term, refers to wage differentials that cannot be accounted for on the basis
of productivity endowments, and is commonly interpreted as a measure of labor market
discrimination 11
.
The analysis is done across the three occupational categories: professional,
white-collars, and blue-collars. This allows for differences in wage setting in the three
aggregate occupations and for differences in parameter estimates by gender. As a reminder, professional women earned 78% of professional men in 2004 (up from 75% in
2000); white collar women, being the worst in terms of wage gap, earned 63% of their
male counterparts in 2004 (down from 71% in 2000); and in 2004, blue collar women
earned 71% of blue collar males’ earnings (up from 67% in 2000).
Findings and Discussions
Table (4) summarizes decompositions results for males and females by occupational
categories. The positive sign in the explained column indicates that men enjoy a
productivity wage advantage over women by the amount indicated. In other words, men
31
have higher levels of education and/or experience, in addition to residing in regions of
high demand on their labor than women, therefore, the difference in the wage gap is
justified in accordance with human capital theory predictions. Conversely, a negative
sign in the explained (endowment) portion indicates that the labor market exhibits some
favoritism towards men vis-à-vis women. In other words, women on average have higher
endowments in terms of levels of education and/or experience, and should have earned more than what they are currently paid
12 .
These results also reveal that the bulk of the observed gap between men and
women’s wages is almost exclusively due to intra-occupational differentials, or unequal
treatment of male and female productivity-related characteristics within occupations. A
smaller portion of the wage gap is attributed to inter occupation differentials between
genders in their occupational allocation. In fact, the few negative values of the explained
terms are offset by the unexplained component, which implies that women face
unjustified treatment in occupational allocation.
Between 2000 and 2004, and while the level of intra-occupation discrimination
of professionals and blue collar women has dropped, it still constitutes a large portion of
wage differences. Contrarily, while women working in white collar occupations were
treated fairly and in accordance with their endowments (relative to professionals and blue collar women) they faced greater intra-occupation discrimination in 2004 if compared
with 2000. For example, in 2000, differences in productivity-related endowments
between men and women explained about 69% of intra occupation differential, compared
with 56% in 2004. Furthermore, while the degree of discrimination in occupational
allocation in professional and blue collar occupations accounted for a tiny portion of the
wage differences, they appear to worsen especially for blue collar women. This finding is
in agreement with Assaad and Arnetz’s (2005) finding of increased de-feminization of
blue collar jobs in the Egyptian economy in addition to the fall of Egypt’s share of
foreign trade in late 1990s. The same conclusion is also confirmed by El-Hamidi’s (2008)
findings of increased gender wage discrimination particularly in tradable sectors between
1998 and 2006. A result that is in part was due to restraint female labor movements between industries, and lack of transferable skills.
Clearly, professional and blue collar jobs are becoming harder for women to get
in, but once entered, they face a second wave of discrimination that is practiced within
the occupation, unlike the situation for white collar jobs where women have increased
access to work and have lower levels of intra-occupational wage discrimination, which is
not surprising since close to 40% of them are classified as working in that occupation.
32
TABLE 4. DECOMPOSITION OF WAGE DIFFERENTIALS BY GENDER AND
OCCUPATIONAL CATEGORIES BETWEEN AND WITHIN OCCUPATIONS
Inter Occupation Intra Occupation
Occupational % Wag e Gap Due to % Wage Gap Due to
Categories Endowments Discrimination Endowments Discrimination
Professionals
2000 -17 .45 5.82 -24.83 136.46
2004 -26 .74 10.03 1.73 114.98
White Collars
2000 2 .38 -30.92 69.10 59.44
2004 -3 .82 -29.29 55.84 77.27
Blue Collars
2000 -11 .11 -26.67 4.39 133.40
2004 1 .43 2.60 3.23 92.75 Source: Authors’ own calculations; LFSS 2000-2004. Simply put, most of the differences in the overall gender wage gap in the Egyptian
private labor market cannot be explained by the differences in workers’ productivity-
related characteristics. The differences are due to labor market discrimination, especially
discrimination with regard to occupational allocation. And, while the part of the wage
gap attributed to pay discrimination has declined for both professionals and blue collar
women, the portion of the wage gap that is credited to occupational allocation has
widened in the five year span.
CONCLUSION AND POLICY RECOMMENDATIONS
This paper attempts to add to the current literature on the Egyptian labor market by tracking the recent trends of both gender pay gaps and occupational allocation. It also
evaluates the role of occupational segregation in explaining gender wage gap in the
Egyptian private labor market. Our findings show that occupational segregation are
becoming a more serious issue in pay differences than pure pay discrimination for both
professional and blue collar women. These occupations have experienced deterioration in
the magnitude of discrimination between 2000 and 2004. Pay discrimination for white
collar workers is not as serious as in the other two occupations. Therefore, policies that
target inter-occupational components to close the wage gap may have far-reaching effects
on professional and blue-collar workers, whereas policies targeting equal pay for equal
jobs will have a greater success for white-collar workers.
We recognize that changing the gender structure of employment along with
eliminating barriers to a free and informed choice is likely to be slow process. However, clear government goals and public support from men as well as women have been
demonstrated to achieve sustainable results in many parts of the world.
33
Results also indicate that while women may hold more years of education than
men, they have fewer years of experience than men. Therefore, a strategy that aims at
increasing women's participation, simply by raising their educational achievements is
going to be counterproductive, as this has been shown to lead to an excess supply of
educated women. More attention should be paid to the burden that social reproduction
places on women, hence resulting in interrupted careers. In this regard, one can suggest family policies that emphasize parental leave schemes, reduced working hours and
flexible working and distance working arrangements. Such policies have been
instrumental in eliminating obstacles to the individual's free and informed choices of
occupation, regardless of what stereotyped or actual characteristics one has (Malkas and
Anker, 2003).
Based on our findings for professional and blue collar jobs, where inter-
occupational segregation remains prevalent, it is particularly important to eliminate
implicit mobility barriers that create working conditions that are not ‘gender-friendly.’
The introduction of flexible arrangements and part-time work that aim at making work
conditions more compatible with married life, has historically been the driver for
increased female participation amongst married women in non-traditional jobs in
industrialized countries. More specific equal opportunity programs such as the currently piloted World Bank project ‘Equity Model Egypt’ (GEME) that promoted gender equity
in large private firms should be expanded to small and medium-sized enterprises as well.
Such programs aim to influence employer incentives by certifying firms after auditing
their existing policies towards female employment and effort to promote gender equity
vis-à-vis staff recruitment, training, career development (World Bank, 2010). For lower
skilled (blue collar) workers, there remains a need for continued awareness-raising effort
and media campaigns that contribute to the break down of stereotypes portraying women
as subordinates and men as decision-makers.
As for white collar occupations, where pure pay discrimination remains an issue,
Egypt can learn from the experiences of other countries, such as Turkey, who passed the
Equal Pay Act in 2003 because there remained a substantial gap despite the trend towards decreasing gender differentials. Such legislative intervention was documented to
contribute significantly to the prohibition of discrimination in job titles, job ranks, and
pay scales according to gender at labor market entry points. However, at the same time, it
is important to not excessively exercise the law to impose social entitlements and fringe
benefits for women because doing so can discourage employers to create new female jobs
(El-Haddad, 2009). Another source of worry is that the overall quality of jobs held by
women might deteriorate as employers resort to hiring women without contracts to avoid
the additional cost associated with such entitlements.
Overall, policies that promote the gender integration of occupations, and equal
pay for equal job, need to be carefully balanced with the creation of incentives that
encourage private firms to employ women, especially married women. Thus, current
proposals in Egypt for the payment of maternity leave pay out of social security pool, or by providing tax incentives that lessen their burden, aim at equalizing the costs of hiring
men and women for private employers. This is definitely a step in the right direction as
women are encouraged to enter the private sector labor market in larger numbers in the
new millennium.
34
APPENDIX
TABLE A-1: ESTIMATES OF MULTINOMIAL LOGIT MODEL OF
OCCUPATIONAL ATTAINMENT LEVEL IN THE NON-AGRIGCULTURAL
PRIVATE SECTOR, 2000 AND 2004 Relative Risk Ratios (RRR) of the Multinomial Logit Model of Occupational Attainemnt (Base Category=Blue Collar)
2000 2004
Females Males Females Males
RRR Std. Err. z RRR Std. Err. z RRR Std. Err. z RRR Std. Err. z
Professionals
Age 1.631 0.010 83.310 1.406 0.002 229.370 1.817 0.010 105.090 1.361 0.002 202.200
Age Sq. 1.017 0.009 1.790 1.070 0.002 32.690 0.831 0.007 -23.250 1.105 0.002 48.490
Region of Residence (Metro=base)
Lower Urban 1.673 0.028 30.610 1.177 0.004 42.750 1.072 0.018 4.230 1.090 0.004 25.080
Lower Rural 1.613 0.024 32.580 0.807 0.003 -60.310 1.544 0.022 30.630 0.814 0.003 -62.460
Upper Urban 1.233 0.029 8.930 0.790 0.004 -47.690 1.140 0.022 6.690 1.062 0.004 15.310
Upper Rural 1.498 0.037 16.310 0.542 0.003 -119.820 1.027 0.021 1.280 0.722 0.003 -79.360
Experience 0.734 0.002 -110.420 0.760 0.001 -348.280 0.712 0.002 -125.460 0.783 0.001 -317.750
Experience Sq. 0.824 0.006 -28.850 0.928 0.001 -47.550 0.929 0.005 -13.230 0.903 0.001 -67.110
Levels of Education (Illit=base)
Read & Write 0.412 0.006 -56.630 0.622 0.002 -139.250 0.722 0.010 -22.900 0.646 0.002 -129.960
Primary 0.861 0.028 -4.590 0.568 0.004 -91.480 0.289 0.009 -39.840 0.693 0.004 -64.370
Intermediate 1.742 0.073 13.190 0.780 0.007 -28.980 0.270 0.010 -34.810 0.685 0.006 -45.400
Secondary 4.905 0.304 25.680 1.435 0.020 26.350 1.084 0.062 1.410 1.282 0.017 18.870
University+ 113.356 9.469 56.630 12.631 0.230 139.250 5.702 0.433 22.900 10.270 0.184 129.960
Married 1.213 0.013 18.110 1.663 0.007 117.400 0.882 0.009 -12.110 1.816 0.008 138.460
Childern 0-6 0.712 0.007 -33.030 1.052 0.003 17.900 1.465 0.015 37.230 1.019 0.003 7.110
Size of Household 1.032 0.001 26.290 0.992 0.000 -26.290 1.023 0.001 23.590 0.998 0.000 -8.420
Share of Pub/Gov Emp 0.897 0.003 -34.610 0.921 0.001 -109.460 0.910 0.003 -32.350 0.995 0.001 -7.460
White Collars
Age 1.075 0.005 16.260 1.098 0.001 76.480 1.174 0.005 37.780 1.094 0.001 71.980
Age Sq. 1.237 0.009 27.740 1.151 0.002 76.320 1.061 0.007 9.170 1.146 0.002 73.860
Region of Residence (Metro=base)
Lower Urban 0.788 0.010 -18.340 0.817 0.003 -50.830 0.945 0.012 -4.450 0.752 0.003 -75.780
Lower Rural 0.528 0.006 -58.220 0.713 0.002 -96.760 0.585 0.006 -48.510 0.627 0.002 -134.880
Upper Urban 1.193 0.022 9.490 1.103 0.005 19.920 0.922 0.015 -5.060 0.955 0.004 -11.400
Upper Rural 0.537 0.011 -31.540 0.633 0.003 -87.740 0.340 0.006 -62.060 0.492 0.002 -159.210
Experience 0.897 0.002 -45.330 0.854 0.001 -221.950 0.872 0.002 -59.490 0.866 0.001 -209.460
Experience Sq. 0.849 0.005 -27.650 0.951 0.001 -34.270 0.886 0.004 -25.340 0.937 0.001 -45.420
Levels of Education (Illit=base)
Read & Write 0.374 0.005 -71.070 0.816 0.003 -55.460 0.712 0.010 -25.090 0.733 0.003 -79.630
Primary 1.793 0.049 21.470 0.882 0.006 -18.780 0.848 0.023 -6.190 1.002 0.007 0.230
Intermediate 14.362 0.530 72.160 1.726 0.016 57.330 2.901 0.104 29.580 2.309 0.024 82.050
Secondary 37.882 2.059 66.860 3.646 0.053 89.560 7.295 0.399 36.300 5.225 0.079 108.830
University+ 190.359 14.060 71.070 2.960 0.058 55.460 6.136 0.444 25.090 5.230 0.109 79.630
Married 1.434 0.013 39.400 0.925 0.003 -20.860 1.224 0.012 21.070 0.937 0.004 -16.930
Childern 0-6 0.931 0.008 -8.450 1.059 0.003 20.650 1.264 0.011 27.000 0.988 0.003 -4.500
Size of Household 1.029 0.001 29.760 0.993 0.000 -25.930 1.022 0.001 27.760 1.001 0.000 3.570
Share of Pub/Gov Emp 0.995 0.002 -1.930 1.019 0.001 25.850 0.922 0.002 -35.400 1.013 0.001 19.810
No. of Observations 496021 5752017 552175 6370338
Pseudo R 2 : 0.2295 0.1946 0.2829 0.1940
LR chi2(32) 248460 2230938 342089 2492853
Source: Authors’ own calculations; LFSS 2000-2004.
35
FIGURE A-1. PANEL A: MORE LIKELY TO OBTAIN A PROFESSIONAL
OCCUPATION RELATIVE TO A BLUE COLLAR OCCUPATION
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Females-2000
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Females-2004
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Males-2000
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Males-2004
Source: Authors’ own calculations; LFSS 2000-2004.
PANEL B: MORE LIKELY TO OBTAIN A WHITE COLLAR OCCUPATION
RELATIVE TO A BLUE COLLAR OCCUPATION
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Females-2000
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Females-2004
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Males-2000
Age
Lower Urban
Lower Rural
Upper Urban
Upper Rural
Experience
Read & Write Primary Intermediate
Secondary
University+
Married
Childern 0-6
Size of HOUSEHOLD
Share of Pub/Gov …
Males-2004
Source: Authors’ own calculations; LFSS 2000-2004.
36
TABLE A-2. ESTIMATES OF SELECTIVITY CORRECTED WAGE
EQUATIONS, WORKING PRIVATE SECTOR WORKERS (15-64), 2000
Variables
Female Male Female Male Female Male
Experience 0.044*** 0.014*** 0.009*** 0.011*** 0.052*** 0.024***
(0.002) (0.001) (0.001) (0.000) (0.002) (0.000)
Experience Sq. -0.055*** -0.025*** -0.016*** -0.015*** -0.082*** -0.047***
(0.003) (0.002) (0.002) (0.001) (0.005) (0.001)
Region of Residence (Metro=base)
Lower Urban -0.775*** -0.367*** -0.563*** -0.209*** -0.138*** -0.231***
(0.010) (0.006) (0.007) (0.003) (0.009) (0.002)
Lower Rural -1.143*** -0.201*** -0.711*** -0.281*** -0.114*** -0.433***
(0.012) (0.006) (0.010) (0.002) (0.008) (0.002)
Upper Urban 0.276*** -0.011** -0.364*** -0.119*** -0.343*** -0.162***
(0.008) (0.004) (0.006) (0.003) (0.011) (0.002)
Upper Rural -1.016*** -0.177*** -0.239*** -0.171*** -0.255*** -0.548***
(0.029) (0.009) (0.016) (0.004) (0.012) (0.002)
Levels of Education (Illiterate=base)
Read & Write -0.454*** -0.602*** -0.108*** 0.025*** 0.395*** 0.117***
(0.032) (0.016) (0.013) (0.004) (0.010) (0.001)
Primary -2.114*** 0.099*** 0.580*** -0.148*** 0.098*** 0.076***
(0.045) (0.022) (0.017) (0.005) (0.015) (0.002)
Intermediate -1.245*** -0.527*** 0.368*** -0.208*** -0.270*** -0.163***
(0.026) (0.015) (0.016) (0.005) (0.009) (0.002)
Secondary -3.574*** -1.573*** 0.122*** 0.113*** -0.103*** -0.229***
(0.037) (0.018) (0.011) (0.006) (0.018) (0.006)
University+ 4.709*** 2.594*** 0.421*** 0.373*** 0.278*** -0.056***
(0.049) (0.021) (0.011) (0.004) (0.021) (0.007)
Lambda 5.152*** 3.570*** -0.365*** 0.991*** -0.542*** -0.695***
(0.050) (0.022) (0.032) (0.013) (0.026) (0.005)
Constant 0.772*** 0.651*** 0.358*** 0.375*** -0.015 1.011***
(0.031) (0.017) (0.015) (0.006) (0.022) (0.005)
Observations 48870 176734 71634 398135 35244 1154552
R-squared 0.50 0.20 0.26 0.15 0.23 0.15
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Prof. White Collar Blue Collar
Source: Authors’ own calculations; LFSS 2000-2004.
37
TABLE A-3. ESTIMATES OF SELECTIVITY CORRECTED WAGE
EQUATIONS, WORKING PRIVATE SECTOR WORKERS (15-64), 2004
Variables
Female Male Female Male Female Male
Experience 0.061*** -0.005*** 0.051*** 0.041*** 0.046*** 0.036***
(0.001) (0.001) (0.001) (0.000) (0.001) (0.000)
Experience Sq. -0.113*** 0.009*** -0.067*** -0.071*** -0.078*** -0.064***
(0.004) (0.002) (0.004) (0.001) (0.004) (0.000)
Region of Residence (Metro=base)
Lower Urban -0.095*** -0.117*** -0.217*** -0.041*** -0.103*** -0.081***
(0.009) (0.004) (0.006) (0.002) (0.007) (0.001)
Lower Rural -0.358*** -0.121*** -0.273*** -0.103*** -0.220*** -0.088***
(0.008) (0.005) (0.008) (0.002) (0.006) (0.002)
Upper Urban -0.067*** -0.038*** 0.077*** 0.021*** -0.004 -0.039***
(0.007) (0.003) (0.006) (0.002) (0.008) (0.001)
Upper Rural -0.439*** -0.054*** -0.343*** -0.043*** 0.217*** -0.104***
(0.014) (0.006) (0.012) (0.003) (0.010) (0.002)
Levels of Education (Illiterate=base)
Read & Write -0.382*** 0.313*** 0.087*** 0.097*** 0.289*** 0.111***
(0.036) (0.016) (0.011) (0.003) (0.007) (0.001)
Primary -0.900*** 0.382*** 0.235*** 0.040*** 0.185*** 0.120***
(0.055) (0.017) (0.013) (0.004) (0.008) (0.002)
Intermediate -0.167*** 0.685*** 0.466*** 0.151*** 0.178*** 0.141***
(0.027) (0.014) (0.014) (0.004) (0.007) (0.001)
Secondary -0.442*** 0.181*** 0.429*** 0.288*** 0.022 0.174***
(0.033) (0.016) (0.012) (0.005) (0.016) (0.004)
University+ 0.575*** 0.707*** 0.709*** 0.563*** 0.602*** 0.222***
(0.046) (0.020) (0.010) (0.004) (0.016) (0.004)
Lambda 0.836*** 2.526*** -0.209*** -0.412*** -0.371*** -0.078***
(0.048) (0.020) (0.019) (0.010) (0.016) (0.004)
Constant 0.150*** -0.407*** -0.311*** -0.041*** -0.022* 0.292***
(0.027) (0.014) (0.010) (0.004) (0.012) (0.003)
Observations 92118 314806 153603 810057 77283 2199802
R-squared 0.24 0.15 0.16 0.23 0.15 0.26
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Prof. White Collar Blue Collar
Source: Authors’ own calculations; LFSS 2000-2004.
ENDNOTES
1 Data is made possible by a grant from the Fulbright Association in 2005.
2 We recognize that results depend on the fineness of the occupational classifications. Since our
estimates are at the one digit level, occupational segregation within each occupational category is ignored. 3 Since the reported figures for experience were inconsistent, a potential experience has been
created as follows: (pot exp=age –school-6). Where “age” refers to the current age of the worker; and “school” is the
number of years of schooling.
38
4 To do so, we first conduct an F-test to see if there is a structural difference. The test suggests that
there is a statistically significant difference between males and females in the equations explaining occupational attainment. The calculated F-statistic is 9.53, which is greater than the critical value at the 1 percent significance level. Hence, the null hypothesis of no structural difference can be rejected. 5 According to Heckman (1979), estimating the wage equation may produce biased estimates:
Those who reported wages at the time of the survey are entered into the analysis while the ones who were not working were not. This is a problem of sample selection bias since the subsample used for determining wages represents a non-random sample of the population. In order to correct
for this bias, we followed Heckman’s (1979) suggestion where two equations are estimated: 1) the selection/occupational attainment equation, which models the probability of engaging in a particular occupation for all observations in the sample (working and non working). This step estimates a selection variable (the inverse mills ratio) which is entered linearly into the second (wage) equation. 2) The wage equation applies only to those who are observed in paid work, and includes the selection term yielding consistent estimates of the coefficients. 6 Heckman’s solution suggested using probit analysis to model the participation decision (i.e.
participate/no participate). We broadened this solution and used multinomial logit to model the
occupational attainment of three distinct categories (professionals, white collars, and blue collars). 7 Since the reported figures of actual work experience were inconsistent, we used potential
experience as follows: Pot. Exp=Age-School-6. Nevertheless, results did not vary much when using actual vs. potential experience. 8 Log hourly wages are used (instead of hourly or weekly wages) because they reduce the effects of
wage outliers. 9 In this regard, it is important to keep in mind that the wage equation is built on a number of
limiting assumptions. For instance, it assumes that workers have equal abilities and confront equal opportunities. Second, there is the problem of “ability” and the associated difficulty of measuring the quality of education. Human capital theory suggests that ability is likely to be positively correlated with schooling. Therefore, neglecting the ability factor from the regression equation may very well result in upward bias in the estimated returns to schooling. As a result, and because the survey data does not include variables that could be used as a proxy of ability, this problem is ignored in our estimations. A large portion of wage differences that cannot be explained by differences in human capital measured by educational attainment and experience highlights the importance of other unobservables such as firm size or firm profitability. A high paying industry
(i.e. finance) is high paying because it attracts the most skilled and simply because it pays a premium to its employees. Other high paying industries (i.e. gas) offers high wages merely because the entire industry pays above average wages. Wage equations also disregard direct costs of schooling and overlooks earnings while at school. Besides, it assumes a fixed yearly return of schooling. 10
Results of the selectivity corrected wage equations reported in the Appendix Tables (A-2) and (A-3). 11
We grouped differences due to discrimination and differences due to selection bias in one
“unexplained” 11
factor. We recognize that the unexplained term may have a problem of omitted variables, including attachment to the labor force, lack of specific training, tastes, personality and/or interrupted careers. 12
Large components of decomposition results are not unique to this study. Considerable figures have been reported for cases in developing countries which results from omitted variables problems and distortions in the labor markets. Moreover, the unexplained components include selection errors (resulting from occupational attainment module).
39
REFERENCES
Antonczyk, Dirk and Fitzenberger, Bernd and Sommerfeld, Katrin, “Rising
Wage Inequality, the Decline of Collective Bargaining, and Gender Wage Gap,” Labor
Economics, 2010, Vol. 17, pp. 835-847.
Artecona, R. and W. Cunningham, “Effects of Trade Liberalization on the
Gender Gap in Mexico”, Policy Research Report on Gender and Development, Working
Paper 21, 2002, World Bank, Washington D.C.
Assaad, R., and F. El-Hamidi, Is all work the same? A comparison of the
determinants of female participation and hours of work in various employment states in
Egypt,” Research in Middle East Economics, Vol. 4. The Economics of Women and Work
in the Middle East and North Africa, 2001. Mine Cinar ed, Assaad, R. and F. El-Hamidi, Forthcoming. “Assessing female labor force in
Egypt then 1998, and now 2006,” ERF Working Paper, 2008
Assaad, R., and M. Arntz, Constrained geographical mobility and gendered
labor market outcomes under structural adjustment: Evidence from Egypt. Canada:
World Development, 2005, Vol. 33 (3), pp. 432-454.
Assaad, R., “Informalization and De-feminization: Explaining the Unusual
Pattern in Egypt,” Rethinking Informalization: Precarious Jobs, Poverty and Social
Protection. 2005, Neema Kudva and Lourdes Beneria (Eds.)
Assaad, R., “Why Did Economic Liberalization Lead to Feminization of the
Labor Force in Morocco and De-Feminization in Egypt?” Gender Impact of Trade
Liberalization in the MENA Region., 2006, Tunis: Center of Arab Women for Training
and Research (CAWTAR). Bassusi Al., N., and S. Amin, “Wage work to prepare for marriage: Labor
force entry for young women in Egypt.” 2002, Paper presented at the annual meeting
of the Population Association of America. Atlanta.
Becker, G., The Economics of Discrimination, 1957, University of Chicago
Press
Berik et al, “International Trade and Wage Discrimination: Evidence from East
Asia”, World Bank, Poverty Reduction and Economic Management Network, Policy
Research Working Paper 3111, 2004.
Bowlus, A. J., A Search Interpretation of Male-Female Wage Differentials.
Journal of Labor Economics, 1997, Vol.15 (4), 625-657.
Brown, R. S., M. Moon, and Zoloth, B.S., “Incorporating occupational attainment in studies of male-female earnings differentials,” The Journal of Human
Resources, 1980, Vol. 15 (1):3–28.
Brown (1980), “Occupational attainment and segregation by sex,” Industrial
and Labor Relations Review, July 1980, Vol. 33, pp.506-517.
El-Haddad, Amirah, “Labor Market Gender Discrimination under Structural
Adjustment: The Case of Egypt,” The American University in Cairo, Social Research
Center, Canadian International Development Agency, Working Paper #003, Nov., 2009,
pp. 23-25.
40
El-Hamidi, F., “Labor supply of Egyptian married women when self-
employment is an option: Participation and hours of work.” Paper presented at the annual
meeting of the Middle East Economic Association (MEEA), in collaboration with the
Allied Social Science Associations (ASSA), 2003. Washington, D.C.
El-Hamidi, F., “General or vocational? Evidence on school choice, returns, and
sheep skin effects from Egypt 1998.” Journal of Policy Reform. 2006, Vol. 9 (2), pp.
157-176.
El-Hamidi, F.. “Trade Liberalization, Gender Segmentation, and Wage
Discrimination: Evidence from Egypt,” June 2008, Working Paper 414, Economic
Research Forum, pp. 15.
Feliciano, Zadia M., “Workers and Trade Liberalization: The Impact of Trade
Reforms in Mexico on Wages and Employment,” Industrial and Labor Relations Review, 2001, Vol.55, No.1, pp 95-115.
Hausmann, R., L. Tyson, S. Zahidi, The Global Gender Gap Report 2007.
Geneva: World Economic Forum, 2008.
Hasan and Chen, “Trade and Workers: Evidence from the Philippines.” East-
West Center Working Paper (Economic Series), 2003, No. 61
Heckman, J. J., “Sample selection bias as a specification error,” Econometrica,
1979, Vol. 47 (1), pp. 153-161.
Hirsch, B., “Joan Robinson Meets Harold Hotelling: A Dyopsonistic
Explanation of the Gender Pay Gap,” Bavarian Graduate Program in Economics, 2007,
BGPE Discussion Paper No. 24.
Jacob, Marilyn,“The Impact of Trade Liberalization on Gender and Caste Groups in India”. Paper presented at the Annual Meeting of Population Association of
America, 2007.
Kandil, Lamia E., “Gender Wage Discrimination in Egypt: A Quintile
Regression Analysis”. Paper presented to the European Association of Labour
Economists Annual Meetings , September, 2009.
Kumar, Utsav and Prachi Mishra, “Trade Liberalization and Wage Inequality:
Evidence from India”, IMF Working Paper # WP/05/20, forthcoming, Review of
Development Economics, 2006.
Melkas, H., and R. Anker, Towards gender equity in Japanese and Nordic labor
markets: A tale of two paths. 2003, Geneva: ILO.
Mincer, J., Schooling, experience and earnings. 1974, New York: National
Bureau of Economic Research. Moghadam, V., Women, work and economic reform in the Middle East and
North Africa. 1998, Boulder, London: Lynne Rienner Publishers.
Nassar, Heba, Structural adjustment and women’s employment in Egypt.
Workshop on Economic Restructuring and Women in the Middle East: Gender, Jobs and
Activist Organizations. 1998, Brown University, Thomas Watson Jr. Institute for
International Studies.
Oostendorp, R. H., “Globalization and the Gender Wage gap”. The World
Bank Economic Review, 2009, Vol. 23, No. 1, pp.141-161. Washington, D.C.: World
Bank.
41
Pavcnik, N. et al , Trade Liberalization and Industry Wage Structure: Evidence
from Brazil. World Bank Economic Review, 2004, Vol. 18, No. 3., pp. 319.
Reilly, Barry and Puja Vasudeva Dutta, “The Gender Pay Gap and Trade
Liberalization: Evidence for India”, 2005, PRUS Working Paper No. 32.
Said, M., “The distribution of gender and public sector pay premia: Evidence
from the Egyptian organized sector,” 2003, SOAS Economics Department Working
Paper. London, UK: School of Oriental and African Studies, University of London.
Said, M., “The distribution of gender differentials & public sector wage
premia in Egypt,” Economic Research Forum for the Arab Countries, Iran & Turkey,
Working Paper 9925, 1999, Cairo, Egypt.
Wood, A., “North-south trade and female labor in manufacturing: An
asymmetry,” Journal of Development Studies. January 1991. World Bank, Gender and development in the Middle East and North Africa.
2004, Washington D.C.: World Bank.
World Bank, Narrowing the Gap: Improving Labor Market Opportunities for
Women in Egypt, Arab Republic of Egypt: Gender Assessment 2010, 2010, Washington,
D.C.: World Bank.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.