REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
ISSN: 1576-0162
DOI: http://dx.doi.org/10.33776/rem.v0i68.8293
FORECASTING THE GENDER WAGE GAP IN ARGENTINA
FROM A PRODUCTIVITY-BASED APPROACH
PROYECCIONES SOBRE LA BRECHA SALARIAL DE GÉNERO EN ARGENTINA
A PARTIR DE UN ENFOQUE BASADO EN LA PRODUCTIVIDAD
Diana Suárez
dsuarez@campus.ungs.edu.ar
Instituto de industria - Universidad Nacional
de General Sarmiento (IDEI/UNGS)
Florencia Fiorentin
ffiorentin@campus.ungs.edu.ar
Instituto de industria - Universidad Nacional de General Sarmiento (IDEI/
UNGS) / Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
Florencia Barletta
fbarlett@ungs.edu.ar
Instituto de industria - Universidad Nacional
de General Sarmiento (IDEI/UNGS)
Recibido: mayo 2024; aceptado: noviembre 2024
ABSTRACT
This paper analyzes the association between the gender wage gap (GWG)
and productivity at the industry level in Argentina from 2002 to 2022, drawing
on gender studies and literature on structural change. Using dynamic OLS
regression with random effects across the 1st to 4th quantiles, we examine
the relationship between the GWG, female labor participation, and industry
productivity. Results indicate that increased female participation is positively
associated with reductions in the wage gap in the least productive industries.
Additionally, female education rates contribute positively to reducing the GWG.
The findings also reveal that productivity gains in the most productive sectors
help narrow the gender wage gap, suggesting that structural shifts toward
higher productivity levels may play a role in closing this gap.
Keywords: Gender wage gap; productivity; structural change.
RESUMEN
Este artículo analiza la asociación entre la brecha salarial de género
(GWG, por sus siglas en inglés) y la productividad a nivel sectorial para el caso
argentino (2002-2022), con base en los estudios de género y la literatura
sobre cambio estructural. A partir de una regresión dinámica con efectos
aleatorios por cuartiles, se testea la relación entre la GWG, la participación
femenina y la productividad sectorial. Los resultados muestran que el aumento
de la participación femenina se asocia positivamente con reducciones en la
brecha salarial en los sectores menos productivos. La tasa de graduación
femenina impacta positivamente en la reducción de la GWG. Estos resultados
sugieren que el cambio estructural hacia niveles más altos de productividad
podría contribuir a cerrar la brecha salarial.
Palabras clave: Brecha salarial de género; productividad; cambio
estructural.
JEL Classification/ Clasificación JEL: O11; J16; O47.
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
1. INTRODUCTION
The gender wage gap (GWG) refers to the disparities in earnings between
genders and is a topic of significant and growing concern in both the
academic literature and policy discussions (Jayachandran, 2015; UN-Women,
2022; WEF, 2023). In this context, several empirical studies have sought to
understand the underlying sources of this gap (Allen and Sanders, 2002;
Olivetti and Petrongolo, 2016; Redmond and McGuinness, 2019, among
others). These studies are framed within the concepts of endowment effects
and discrimination effects.
The endowment effect is grounded in human capital theory and posits
that differences in earnings are related to salary determinants such as
individual productivity levels, skills, education, and personal predispositions.
Consequently, this perspective suggests that the salary gap is likely to
decrease as women attain higher levels of education and pursue better-paying
jobs (Abegaz and Nene, 2018; Brynin and Perales, 2016; Redmond and
McGuinness, 2019; Si et al., 2021).
The discrimination effect, based on feminist economic theory, argues that
even when women and men possess equal qualifications, women tend to receive
lower wages than their male counterparts due to systemic discrimination within
the labor market. Most of these studies conclude that while endowment factors
have contributed to the reduction of the GWG, a substantial portion of the gap
remains attributable to unobservable factors that account for discrimination
against women (Ahmed and McGillivray, 2015; Redmond and McGuinness,
2019; Sin et al., 2022).
To shed some light on the GWG, several studies indicate that the gap
varies significantly across different industries (Durán Lima and Galván, 2023;
Olivetti and Petrongolo, 2014). In this context, one manifestation of structural
heterogeneity that impacts Latin American countries is the presence of wage
differentials among industries, which reflect substantial productivity gaps
(Cimoli and Porcile, 2016; Fagerberg, 2018; Montobbio, 2002). Consequently,
the characteristics and trends of the GWG may be intricately linked to the
structure of productivity within these sectors.
We assert that this structural heterogeneity has important implications
for the gender wage gap, particularly given the well-documented unequal
distribution of female workers across various industries. Moreover, the ways
in which structural heterogeneity influences both participation rates and the
50 Diana Suárez · Florencia Fiorentin · Florencia Barletta
GWG remain largely underexplored. Therefore, this paper seeks to contribute
to this debate within the context of Latin American countries.
Therefore, policies aimed at diversified economic development in
Latin American countries must take the GWG into account, alongside the
implementation of measures to address the reduction of this disparity.
The GWG continues to affect thousands of female workers, and given that
endowment effects are insufficient to eliminate the gap and discrimination
persists, its impact is likely to endure in the future. According to various studies,
the estimated time required to close the GWG ranges from 169 years (WEF,
2023) to 300 years (UN-Women, 2022) to close the GWG.
The objective of this paper is to examine the gender wage gap within
the productive structure of Argentina, a Latin American country that has
shown minimal progress in reducing gender disparities (CEP XXI, 2021). We
aim to illuminate the characteristics of this wage gap, providing insights that
policymakers—particularly those involved in gender policy and structural
change—can utilize. Additionally, this study seeks to contribute to the fields of
human capital and feminist economics through the lens of structural change.
The guiding questions of this research are as follows: To what extent do trends
in the GWG vary according to the productivity levels of different industries?
How does the relationship between wage gaps and labor force participation
vary across sectors? Furthermore, does the enhancement of female capabilities
play a role in narrowing the GWG?
These questions are relevant in the context of Latin American countries,
and particularly in the case of Argentina, where a heterogenous sectoral
structure prevails in terms of productivity, among many other dimensions (Cao
and Vaca, 2006; Katz, 2018; Niembro and Starobinsky, 2021). Unlike the
existing literature on sectoral differences in the gender wage gap (GWG), our
approach is grounded in the examination of varying levels of productivity that
affect wages across industries, rather than relying on a sector-based definition
of technological intensity. This methodological choice is informed by the critical
importance of enhancing productivity in all sectors, particularly in developing
countries like Argentina, where many industries remain significantly below
the international frontier (Saviotti and Frenken, 2008). The central question
we explore is whether advancements to higher productivity levels exert any
influence on the gender wage gap. Ultimately, our aim is to provide evidence
that contributes to a deeper understanding of the characteristics of gender
disparities within economies marked by structural heterogeneity.
The remainder of the paper is structured as follows. After this introduction,
the second section presents the conceptual framework, which combines
studies on the gender gap with the economic view related to structural change,
from where the research questions are derived. The third section presents
the empirical strategy, in terms of the construction of the database and
econometric tests. Section four is centered on the results and the answer to
the research question. The fifth and last section is devoted to the conclusions.
51
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
2. CONCEPTUAL FRAMEWORK
Studies on the wage and participation gap have proliferated in recent
years (Abegaz and Nene, 2018; Allen and Sanders, 2002; Angelov et al.,
2016; Azmat and Ferrer, 2017; Brynin and Perales, 2016; Redmond and
McGuinness, 2019; Si et al., 2021; Sin et al., 2022). This body of literature
primarily seeks to clarify the factors contributing to labour gaps and to analyze
potential solutions for narrowing these disparities. In this paper, we focus
specifically on the gender wage gap (GWG) and its relationship with the labor
force participation gap (PG) among women.
The GWG refers to the lower salaries women receive compared to men,
even when they are in the same position, in the same industry (Blau and Kahn,
2017). At the macro level, the GWG is estimated as the difference in average
salaries between men and women in the whole economic structure. Studies
focusing on understanding the causes and determinants of the gender gaps
wonder how much of the gap is due to differences between female and male
workers, and how much is due to unobservable factors (Abegaz and Nene,
2018; Allen and Sanders, 2002; Angelov et al., 2016; Azmat and Ferrer, 2017;
Brynin and Perales, 2016; Redmond and McGuinness, 2019; Si et al., 2021;
Sin et al., 2022). Studies focusing on observable differences in the gender wage
gap (GWG) are primarily grounded in human capital theory. These studies argue
that the GWG tends to diminish as female workers acquire increasingly valued
skills in the labor market, particularly through higher levels of education. This
phenomenon is referred to as the “endowment effect” (Ahmed and McGillivray,
2015; Anker, 1997; Redmond and McGuinness, 2019). Research that
explores additional unexplained factors influencing the GWG is situated within
the framework of feminist economics. These studies assert that such factors
are primarily linked to discrimination against women, a concept termed the
“discrimination effect.” Overall, both effects have been shown to significantly
impact the GWG (Ahmed and McGillivray, 2015; Sin et al., 2022).
The human capital literature converges on several key factors that
contribute to the endowment effect.
Firstly, one critical aspect is the recognition that women’s education and
skill acquisition play a significant role in reducing wage disparities (Abegaz and
Nene, 2018; Brynin and Perales, 2016; Redmond and McGuinness, 2019; Si et
al., 2021). In addition, there are female and male characteristics that cause the
GWG. Different explanations were found. Women tend to prioritize employment
that is geographically convenient, offers job security, and provides flexible
hours, among other non-wage-related factors (Redmond and McGuinness,
2019; Sin et al., 2022). In contrast, men often select jobs primarily based
on salary rather than additional benefits (Redmond and McGuinness, 2019).
Moreover, it has been demonstrated that women possess less bargaining power
when it comes to wages, which accounts for a portion of the overall gender
wage gap (GWP) observed in New Zealand (Sin et al., 2022). Additionally,
evidence indicates that men typically receive higher wages due, in part, to
52 Diana Suárez · Florencia Fiorentin · Florencia Barletta
greater productivity, suggesting that the wage gap is also linked to disparities
in productivity (Abegaz and Nene, 2018; Brynin and Perales, 2016; Sin et al.,
2022). For instance, Abegaz and Nene (2018) study the relationship between
gender wage and productivity gaps in the Ethiopian manufacturing sector
from 1996 to 2010. Their findings indicate that when average productivity is
controlled for, the gender wage gap is diminished but remains significant. They
further identify an aspect of occupational segregation, noting that a higher
proportion of women are employed in lower-paying firms (Abegaz and Nene,
2018).
Secondly, despite the robust evidence regarding the determinants of the
GWG, existing studies consistently indicate that a portion of this gap remains
unexplained and is thus unrelated to observable worker characteristics. The
factors identified in previous research do not fully account for the disparity. For
instance, Sin et al. (2022) report that the gender wage gap ranges from 13%
to 17%. Of this gap, the productivity differential accounts for 4.5 percentage
points (p.p.), while women’s attitudes toward salary negotiation contribute an
additional 5 p.p. Consequently, between 3.5 and 7.5 p.p. remain unexplained.
Furthermore, Redmond and McGuinness (2019) reveal that only 13% of
the gender wage gap can be attributed to identifiable factors. Ahmed and
McGillivray (2015) demonstrate that in Bangladesh, the reduction in the GWG
is primarily attributable to discrimination effect –less discrimination against
women in the labour market-, rather than endowment effect.
To shed light on the unexplained aspects of the wage gap, a limited
number of studies have incorporated the industry affiliation of workers. This
approach aims to identify the industry-specific sources of both the wage and
participation gaps through decomposition and/or multilevel analyses (Olivetti
and Petrongolo, 2014, 2016; Sin et al., 2022). This literature finds that
regardless of workers’ attributes, the characteristics of the industry also play
a crucial role in explaining the gap. As previously noted, women are often
employed in sectors that offer lower salaries, a phenomenon that may be
associated with reduced levels of productivity within those industries (Abegaz
and Nene, 2018). Durán Lima and Galván (2023) demonstrate that the gap is
higher in exporting compared to those that do not engage in export activities.
Similarly, Sin et al. (2022) identify that a portion of the wage gap is attributable
to industrial characteristics, noting that women are often employed in less
productive firms. Ahmed and McGillivray (2015) reveal that the GWG varies
across the wage distribution, suggesting that this variability may be influenced
by industry-specific factors, as the gap is not uniform across different sectors.
Nevertheless, the impact of productive structures and industry-specific
characteristics on the wage gap is less addressed by the literature.
In the framework of human capital literature, Olivetti and Petrongolo
(2016) investigate the extent to which the closure of the participation
gap can be attributed to shifts in industry composition and/or changes in
female participation within each sector. Their analysis utilizes a database
encompassing nine high-income countries. Starting from a binary classification
53
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
of the economy into goods and services, they propose that the increasing
participation of women in the labour market is, in part, a consequence of
the expansion of the service sector. This sector has generated employment
opportunities that align more closely with the preferences and household roles
of female workers. Consistent with human capital theory, this phenomenon
is particularly evident in the production of services that are less reliant on
physical skills and more dependent on “soft” skills. Consequently, this evidence
has resurged the longstanding debate regarding the division of the workforce
between “hard” and “soft” capabilities.
In this context, economic literature on structural change has largely proved
that industry structure is not neutral in the explanation of economic growth
and incomes (Cimoli and Porcile, 2016; Fagerberg, 2018; Montobbio, 2002).
In turn, innovation theory has historically recognized the existence of sectoral
patterns of technological intensity, which impact both productivity and wage
levels (Malerba et al., 1997; Pavitt, 1984). If industry impacts on incomes and
wages, derived from technological intensity; and if there are patterns of female
participation in employment, which tends to be higher in low-tech industries
such as social services, hotels, and business services industries (Allen and
Sanders, 2002); then there are good reasons to expect some association
between productivity and the gender wage gap.
In this context, evidence suggests that structural change processes,
characterized by an increasing emphasis on knowledge-intensive sectors,
facilitate job creation that prioritizes intellectual attributes over physical
capabilities. Consequently, following the main hypothesis of human capital
literature, it can be expected that structural change will contribute to a
reduction in gender wage gaps. Thus, countries that exhibit specialization
patterns biased toward natural resources and less knowledge-intensive
sectors are likely to experience larger gender wage disparities, while those
where high-tech industries have gained prominence should exhibit narrower
gaps1. For instance, Baum and Benshaul-Tolonen (2021) empirically examine
the impact of natural resource dependence on gender equality. They found
that countries in which natural resources rents -from oil, gas, coal, minerals,
and forests- account for a greater share of GDP have higher levels of gender
inequality. Some specificities of those sectors in terms of activities, type of
work, geographical location and organizational culture and their impact on
the gender gap have been also addressed by literature (Aragón et al., 2018;
Argoitia et al., 2023; Kotsadam and Tolonen, 2016). In turn, Rendall (2013)
shows evidence about the importance of structural change towards “brain-
intensive” sectors in reducing gender disparities, by decreasing the labour
demand for physical attributes. His results suggest a positive impact of this type
of transformation on reducing gender inequality in wages and employment
shares in five countries – USA, Brazil, Mexico, Thailand and India-.
1 This does not mean that the wage gap disappears with the level of education. On the contrary, it
generally persists, although on a smaller scale.
54 Diana Suárez · Florencia Fiorentin · Florencia Barletta
Based on the literature reviewed, three hypotheses lead our empirical
exercise in the following section. Firstly, since literature has found evidence
that positively associate female participation with low-productivity firms and
industries, we expect increases in female participation in the labour force to
be positively associated with increases in the GWG. This way, H1 states that a
reduction in the gender participation gap (female to male participation) leads
to an increase in the gender wage gap. Following the literature, the idea behind
the hypothesis is that women incorporate (or are accepted) in the labour
market in less productive activities and worse paid jobs
The second and third hypothesis are based on human capital theory. The
second one is derived from the impact of productivity on the gender wage gap.
According to the literature reviewed, movements to higher levels of productivity
are associated with more brain-intensive jobs, as opposite to brawn-intensive,
given the implementation of technological innovation. Literature has also
shown the relative advantages of women in the first case, and the negative
effect in the second one. Hence, H2 states that productivity gains positively
impact the reduction in the gender wage gap.
The third hypothesis is derived from the human capital theory statement
about the relative advantage of women in industries with a higher demand
of skills, or brain-intensive. If true, higher levels of capabilities among women
would put them into similar levels of employability, then a shorter gap should
be expected. Accordingly, H3 states that higher levels of female capabilities
positively impact on the reduction of the gender gap.
3. DATASET AND ESTIMATION STRATEGY
3.1. DATASET AND DESCRIPTIVE STATISTICS
The empirical analysis relies on a dataset that aggregates data from the
Observatory of Employment and Entrepreneurial Dynamics (OEDE, by its
acronym in Spanish), from the Argentinean Human Capital Ministry, along
with data from national accounts sourced from the National Institute of
Census and Statistics (INDEC, by its acronym in Spanish) –hereinafter OEDE/
INDEC database. This database encompasses information about employment,
salaries, economically active population (EAP), tertiary graduates, and
productivity, all viewed through a gender lens. Encompassing the period from
2004 to 2022, it covers all economic activities at the two- and three-digit
levels of the International Standard Industrial Classification (ISIC), resulting in a
balanced panel comprising 810 observations across 45 industries2.
The selection of Argentina for our empirical analysis is twofold. On the one
hand, evidence of the gender gaps in its labour market shows that they are
decreasing very slow, and that in some industries the gap has even increased
(CEP XXI, 2021). On the second hand, Argentina has been largely studied in
2 Description of the variables and sources is presented in Appendix 1.
55
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
terms of its productive structure, and a lot has been written related to the
need to increase the level of productivity among all the sectors (Niembro and
Starobinsky, 2021).
Therefore, Table 1 presents the descriptive statistics of the database. For
the total sample, female labour force participation increased from 30.2%
in 2004 to 33.2% in 2022, reflecting only a 3 percentage points (p.p.) rise
over nearly 20 years. Conversely, female employment-to-EAP ratio remained
relatively stable, ranging from 43.5% to 44.2% over the same period. Notably,
the disparity in monthly average salaries3 between genders persists, although
it reduced between 2004 and 2022 in 5 p.p.: while women earned 24.5 p.p.
less than men in 2004, this gap narrowed to 19.4 p.p. by 2022. In addition,
women’s representation among tertiary graduates increased from 59.3% to
64.2% of the total graduates during this period. In summary, there has been
progress in female labour market participation, wage equality, and educational
attainment, albeit at a gradual pace over the past two decades.
TABLE 1. DESCRIPTIVE STATISTICS
Total Female (% to total)
2004 2022 2004 2022
Labour force (1000 persons) 4105 6487 30.2 33.2
Average wage (US$) 930 2827 75.50 80.64
Economically active population (1000 persons) 485400 622397 43.5 44.2
Tertiary graduates (1000 persons) 75 140 59.3 64.2
Source: Own elaboration based on OEDE/INDEC database. Obs. 810.
Following Badel and Goyal (2023), the gender participation gap (GPG)
was defined as the difference between the ratio female to male employment
and female to male participation in the economically active
population, at the industrial level. Formally:
(1)
The GWG was estimated analogously although in absolute terms. That is,
the ratio of the difference between male average wages ( and female average
wages ( to male wages (, at the industry level. Formally:
(2)
This way, both gaps account for percentage point differences between men
and women, where higher numbers mean higher gaps for women.
3 It is important to note that the database provides data about salary per worker, not about salary
per worker hours.
56 Diana Suárez · Florencia Fiorentin · Florencia Barletta
Given the definition of gender gaps, various methodologies in the literature
have been employed to forecast their evolution using deterministic trend methods,
which provide time-to-gender-equality estimations. Based on Badel and Goyal
(2023), we will estimate the time-to-gender-equality based on two methods: the
logarithm and the raw one. The logarithmic method consists of regressing the
logarithm of the gender gap against the time variable, specifically the year in this
case. In contrast, the raw method entails regressing the raw gender gap against
the same time-variable. Results yield insights into the temporal trends of the
gap and allow for extrapolation until the gap approaches zero. In the logarithmic
method, a negative resulting coefficient indicates the annual percentage rate
of reduction in the gender gap, whereas in the raw method, it represents the
average yearly reduction in percentage points (p.p.). Both methods are utilized
by international organizations such as the IMF (Badel and Goyal, 2024), WEF
(2023) and UN-Women (2022), particularly for monitoring progress toward the
5th Sustainable Development Goal of United Nations.
The third dimension under analysis is productivity levels. Table 2 illustrates
the GPG and GWP at the end of the period (2022), alongside productivity
levels categorized by quantile distribution. The relationship exhibits an inverted
U-shape, where a negative coefficient in the first quantile indicates that female
participation exceeds that of males. This is followed by an increase in the second
and third quantiles, culminating in a decrease in the final quantile, although
the positive coefficient persists. In the case of GPG the relationship shows a
U-shaped curve; the gap starts at a level of 0.3008 in the first quantile, declines
between the second and third quantiles, and subsequently rises again in the
fourth. Consistent with the findings presented in Section 2, the descriptive
statistics indicate that female participation is higher at lower productivity levels,
which is accompanied by a larger wage gap. However, this relationship becomes
less clear across the remaining productivity levels. These heterogeneous results
account for the relevance of a more complex estimation.
TABLE 2. GPG, GWG AND PRODUCTIVITY (2022)
Quantiles Productivity GPG GWG
1 0.3836 -0.0085 0.3008
2 0.6603 0.3231 0.2807
3 1.0669 0.2476 0.0024
4 3.7455 0.2098 0.2426
Source: Own elaboration based on OEDE/INDEC database. Obs. 810.
3.2. ESTIMATION STRATEGY
In order to test the hypotheses presented in section 2, a dynamic OLS
regression with random effects was estimated. The model tests the association
between the GPG, GWG and variations in productivity growth, at the industry
level. Formally:
57
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
(3)
Where the wage gap of industry i at time t depends on a yearly trend (Year),
the GPG, the productivity growth and the female graduation rate ), to account for
the impact of the accumulation of skills. and are the usual random effects and
error terms, respectively. Hausman (1978) tests for efficiency between random
and fixed effects, RE and FE; Wooldridge (2010) Serial Correlation Test and Wald
Test for heteroscedasticity were estimated afterwards, in order to assess the
specification of the model. P-values are reported after the estimation results.
Robust standard errors were set in the estimation.
Following Badel and Goyal (2024), regressing the year against the gap
accounts for institutional, regulatory, and other political measures aimed at
closing the gender gap at the national level. Given the national average and
the overall trend of the gap in Argentina, we expect a negative and significant
association, suggesting that an additional year positively contributes to the
reduction of the gap. Additionally, we incorporate the effect of the GPG,
which accounts for the relative increase in the number of women in the total
labour force of the industry, compared to male participation. As discussed in
Section 2, it is important to note that women tend to be overrepresented in
low-productivity industries. In alignment with H1, we expect a positive and
significant relationship, indicating that reductions in the GPG lead to increases
in the GWG. The association between productivity growth and the GWG is
addressed by H2. Based on the literature reviewed in Section 2, we expect
that increases in average wages, which are typically associated with higher
levels of productivity and skill, will have a positive and significant impact on
the reduction of the gap, resulting in a negative coefficient. Table 4 outlines the
primary variables and their respective characteristics. Finally, in accordance
with H3, we expect to observe a negative association between graduation
rates and the gender gap
To address the presence of heterogeneity, the aforementioned hypotheses will
be tested across different levels of productivity. The distribution of productivity
levels is established through a quantile classification, ranging from low to high
labor productivity. The sample is classified according to the interannual variation
in productivity, thereby allowing shifts between quantiles over time.
It is important to note that this estimation does not encompass the
determinants of the gender wage gap. Our primary objective is to test and
measure the association between the GWG and three key factors that differentiate
men and women in the labour market: female participation in the workforce,
labour productivity, and graduation rates, as well as the relationship between
varying productivity levels and the gap. Nonetheless, we aim to contribute to
gender studies that examine the determinants of the wage gap, recognizing that
certain aspects of the explanation is still a puzzle. We consider that this puzzle
can be partially addressed considering differences in productivity gains at the
industry level.
58 Diana Suárez · Florencia Fiorentin · Florencia Barletta
TABLE 3. DEFINITION OF VARIA BLES
Variable Definition Values
WG Gender wage gap. -1 to 1
GPG Gender participation gap. Interannual variation. -1 to 1
Prod Gross productive value to total workforce. US$ dollars. Interannual variation. 0 to ∞
F_Grad Female tertiary graduation rate, relative to female labour force participation. 0 to 1
Case identifiers
i Industry. Industry on two- and three-level ISIC Rev. 3. 45 industries
T Time. 2004 to 2022
Source: Own elaboration based on OEDE/INDEC database.
4. RESULTS
4.1. FORECASTING OF THE GENDER GAP
Table 4 presents the results of the time-to-gender-equality estimations,
based on national averages for the period 2004-2022. The GPG decreased
from 0.128 p.p. in 2004 to 0.109 p.p. in 2022. This means that female labour
participation was 0.109 p.p. lower than female participation within the total
EAP in 2022. Therefore, while women constituted approximately 44% of the
EAP, they accounted for around 33.2% of the total labour force. Regarding the
evolution of the gap, this represents a yearly reduction rate of -1.04% and an
average reduction rate of -0.0013 p.p. per year. At this pace, and assuming all
else remains constant, the GPG would close within 90 to 230 years.
The evolution of the GWG shows fewer promising results, given the slower
rhythm of reduction. Between 2004 and 2022, it dropped from 0.331 to
0.271, that is a -0.0013 p.p. of average annual reduction and a -0.45% average
interannual variation. This means that it will be closed sometime between the
years 2225 and 2752, ceteris paribus. In other words, it shows a stable tendency
which can be marginally reduced only over the centuries.
TABLE 4. PARTICIPATION AND WAGE GAP – NATIONAL ESTIMATES AND FORECASTING
Gap Forecast (ln) Forecast (raw)
2004 2022 Coeff ln Years Coeff raw Years
GPG 0.128 0.109 0.0104 230 -0.0013 90
GWG 0.331 0.271 0.0045 730 -0.0013 203
Source: Own elaboration based on OEDE/INDEC database. Obs. 810.
As we shall see, average national values hide high levels of heterogeneity at
the industrial level and show a convergence pattern towards the disappearance
of the gender gap where in fact there are industries where female participation
is higher than men’s one and industries where the gap is still on the rise. Graphs
1.a. and 1.b. depict the relationship between the GPG in 2004 (abscissas) versus
2022 (ordinates) at the industrial level. Out of 45 industries, 29 reduced the
59
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
GPG (those below the 45° line), in 10 industries it widened (the ones over),
and the gap is non-existent in 5 industries (below 0). In the case of the GWG,
similar heterogeneity is observed: 35 out of 45 industries reduced the gap, in 7
industries is widened, and 3 industries do not exhibit a WG.
The combined analysis of the gaps shows that except for fishing, all industries
closed at least one of the gaps and 24 industries closed both. Among the
industries that closed both gaps it is worth mention those usually classified as
knowledge intensive, such us medical equipment and business activities. Among
the ones that increased the GPG although with a reduction in the GWG there are
chemicals and communications.
GRAPH 1.A GENDER PARTICIPATION GAP 2004-2022 – INDUSTRY LEVEL
Source: Own elaboration based on OEDE/INDEC database. Obs. 810.
GRAPH 1.B GENDER WAGE GAP 2004-2022 – INDUSTRY LEVEL
Source: Own elaboration based on OEDE/INDEC database. Obs. 810.
60 Diana Suárez · Florencia Fiorentin · Florencia Barletta
4.2. DYNAMIC OLS REGRESSION WITH RANDOM EFFECTS
Table 5 depicts estimation results for the total sample (column 1) and the
selected productivity levels (column 2 to 6). Results show a negative association
between year and the gender wage gap, confirming the average trend observed
in section 3: it decreases 0.00496 p.p. per year, on average for all industries.
This prospective result coincides with other studies that have focused on the
tendency of the wage gap both at the national and international level (Badel
and Goyal, 2024; UN-Women, 2022).
Regarding the association between productivity levels and the GWG,
results show that the gap is being closed faster at the extreme values of the
productivity variation, meaning a higher coefficient at the 1st and 4th quantile,
and even a larger one in the latest (0.00502 and 0.00629, respectively). Based
on the average gender wage gap presented in table 3, and once that gap has
been controlled by the gender participation gap, the productivity levels and
the rate of female graduation, these coefficients mean that the gap will be
closed in 66 years, instead of the 203 observed in raw values.
The other way around, the positive coefficient of the gender participation
gap means that the impact is direct: reductions in the GPG leads to reductions
in the GWG. However, this impact is significant only for the 1st and 2nd quantiles,
with a decreasing impact (0.0339 and 0.0055, respectively). In the case of
the 3rd and 4th the association is not significant. Therefore, H1 is rejected,
as no inverse relationship exists between the two gaps. On the contrary, the
relationship is reversed at the lowest levels of productivity.
Regarding H2, results show that an interannual increase in the productivity
levels leads to a shorter wage gap at the 4th quantile (0.01860). For the rest
of the quantiles the impact is not significant. Therefore, the hypothesis is only
true for the most productive industries, while it is not for the whole industry.
Finally, regarding H3, it is verified at all productivity levels except for the
middle-low one (2nd). The impact of an increase in the relative participation of
graduated women leads to reductions in the wage gap in the case of the 1st, 3rd
and 4th quantiles (0,01233, 0.01491 and 0.02460, respectively).
TABLE 5. ESTIMATION RESULTS – DEP VAR.: WAGE GAP
(1) (2) (3) (4) (5)
VARIABLES Total 0.25 0.50 .75 .1
Year -0.00496** -0.00502** -0.00359** -0.00486** -0.00629**
(0.001) (0.001) (0.001) (0.001) (0.001)
GPG 0.00367** 0.03390* 0.00557** 0.00069 -0.00432
(0.001) (0.017) (0.002) (0.001) (0.007)
Prod -0.00288 0.08440 -0.38653 0.03500 -0.01860*
(0.010) (0.079) (0.290) (0.308) (0.009)
F_grad -0.01207 -0.01233** -0.00867 -0.01491* -0.02460*
(0.009) (0.004) (0.009) (0.007) (0.011)
Constant 10.17884** 10.19211** 7.79720** 9.97299** 12.94819**
(Continue)
61
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
(1) (2) (3) (4) (5)
VARIABLES Total 0.25 0.50 .75 .1
(1.878) (1.710) (2.104) (2.008) (2.924)
Observations 810 203 202 203 202
# of industries 45 40 43 45 42
Hausman Test p-value (H0 ad-
equately modeled by RE) 0.4323 0.3317 0.0655 0.2666 0.7134
Wald Test p-value (H0: heterosce-
dasticity) 0.0000 0.0003 0.0000 0.0000 0.0000
Wooldridge Serial Correlation Test
p-value (H0: no serial correlation) 0.6048 0.3611 0.5131 0.7804 0.3133
Robust standard errors in parentheses. ** p<0.01, * p<0.05, + p<0.1. Source: Own estimations
based on OEDE/INDEC database.
When examining the quantiles, the annual coefficients reveal that the
66 years required to close the GWG at average levels conceal significant
heterogeneity in the productive dynamics of various industries. In the first
quantile, the trend suggests that the wage gap will close in 60 years (from
0.3008 to 0.00502). In the second quantile, this duration increases to 78
years, while it is less than one year for the third quantile. The estimation for
the fourth quantile is 38 years. These heterogeneous trends underscore the
importance of adopting an industry-specific perspective when implementing
policies aimed at closing the GWG. Furthermore, contrary to existing literature,
we did not identify a direct relationship between productivity levels and the
GWG. Consistent with the observations made in the descriptive statistics, we
found a U-shaped relationship between productivity and the rate of closure of
the GWG.
When analyzing the industries within each quantile, significant heterogeneity
becomes evident. While the industries represented in the first quantile
predominantly belong to low-tech sectors (such as textiles, construction, and
wholesale and retail trade), the third and fourth quantiles encompass both
traditional high-tech industries (including chemicals, vehicles, rental and
business activities, and communications) and sectors typically classified as
low-tech. Notably, in the context of Argentina, these low-tech industries play
a substantial role in both domestic and international activities, as exemplified
by agriculture. Furthermore, consistent with findings in the literature (Durán
Lima and Galván, 2023), the dynamics of high-profile exports and proximity
to international markets appear to be positively correlated with the enhanced
impact of productivity gains on the closure of the GWG.
Another finding that aligns with previous studies is the relationship between
the two gaps: the participation gap (PG) and the wage gap (WG). The association
between these gaps is significant only at the lowest levels of productivity—
specifically in the 1st and 2nd quantiles—where the gender participation gap
is notably low. This observation is consistent with the industries represented in
these quantiles, which tend to have high female participation due to historical
patterns of labor distribution (e.g., education, health, and social work).
62 Diana Suárez · Florencia Fiorentin · Florencia Barletta
Moreover, these sectors exhibit wage gaps that substantially exceed the national
average, coupled with average salaries that are also significantly higher than the
national average.
This aligns with evidence from human capital literature suggesting that women
are more likely to be employed in firms that offer lower wages (Abegaz and Nene,
2018), particularly as many of these industries in Argentina are predominantly
public. Furthermore, this prompts the need for future research examining the
differences in gender gaps between the public and private sectors. Consequently,
the sector-specific characteristics and historical patterns contribute to the gender
gap in these fields, indicating that targeted policies are necessary to achieve equity.
It is worthy of a final mention on the relationship between productivity and
the gender wage gap. The literature indicates a negative correlation between
these two variables (Abegaz and Nene, 2018; Brynin and Perales, 2016; Sin et al.,
2022), which aligns with our initial expectations regarding the results. However, this
study’s industry-level focus is predicated on the assumption that this relationship
may vary across different sectors, albeit without specific hypotheses regarding the
nature of these differences. Our findings reveal that women employed in industries
with the lowest levels of productivity are not impacted by this negative relationship.
Contrary to the expectations set forth by human capital literature, productivity
gains do not result in improvements in women’s relative wages. In contrast, within
the industries situated in the highest quantiles, the closure of the gender wage gap
is positively associated with productivity gains and skill development, while it does
not correlate with the participation gap.
This observation does not diminish the importance of addressing the
participation gap; rather, it underscores the necessity for policies aimed at
increasing women’s participation to also encompass wage equity. Increased
participation alone does not automatically translate into equitable income
distribution. Therefore, both types of policies—those focused on enhancing
participation and those addressing wage disparities—are essential for achieving
gender equity.
5. CONCLUSIONS
This paper analyzes the gender wage gap through the lens of industrial
heterogeneity. A significant body of theoretical and empirical research seeks to
explain the existence, persistence, and reduction of these gaps. Most studies on the
wage gap derive their explanations from human capital theory, taking into account
individual worker characteristics such as education, productivity, age, experience,
and bargaining power. From this perspective, researchers aim to identify how these
factors contribute to the reduction of the gap. Typically, education is identified as
an “equalizing” factor that helps diminish the gender wage gap.
Research examining the gender wage gap in relation to different industries is
scarce but consistent in its findings: industrial characteristics significantly impact
the gender wage gap. Such analysis is particularly crucial for Latin American
countries, which exhibit heterogeneous productive structures. Therefore, any
63
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
investigation into gender gaps must consider the economic structure, as it plays
a pivotal role in wage determination. Additionally, analyses of economic structure
that yield recommendations for development must account for gender gaps, given
that these gaps reflect broader structural imbalances. Consequently, closing the
gender wage gap is intricately linked to the evolution of the productive structure,
particularly in terms of the distribution of economic industries concerning value-
added output and employment.
In light of this context, we analyzed the trends and characteristics of the
gender wage gap in Argentina for the period 2004-2022 at the industry level.
Our results indicate the presence of industry-specific factors that influence the
gender wage gap. First, we demonstrate that varying timeframes are required
for different industries to close the gender gap, depending on their productivity
levels. Furthermore, the relationship between wage gaps and participation gaps
varies across industries, with a non-significant association observed in the most
productive sectors. Finally, female university graduates play a crucial role in
narrowing the gender gap across nearly all industries, aligning with expectations
derived from human capital theory. This evidence enriches the literature on
the gender wage gap, particularly from an industrial perspective, by illustrating
how the characteristics of the gap differ across types of activities. Notably, this
paper also provides new insights into the positive relationship between female
university graduates and the wage gap while highlighting its limitations when
industrial characteristics are overlooked. Further research is warranted to explore
the intensity and determinants of this association, such as fields of study and
educational attainment levels.
Several limitations affect our research and warrant acknowledgment. Our
analysis is conducted at the meso-level, lacking detailed information regarding
workers’ positions and working hours. Consequently, we cannot definitively
ascertain whether the wage gap stems from differences in hourly wages or pay-
per-job, or whether it is related to variations in working hours. This challenge
remains unresolved due to the absence of hourly-level databases or position-
specific data in Argentina. Nevertheless, the consistency of our findings with prior
studies indicates that our dataset is adequate for analyzing the gender wage
gap. We anticipate that future research will address these issues by integrating
more granular data, potentially through the merging of various administrative
records.
Despite these limitations, this research elucidates the characteristics of the
gender wage gap and the implications of structural heterogeneity. We aim to
contribute to the discourse on how structural change can adversely affect female
workers if gender inequalities are not addressed. Moreover, we emphasize the
significance of public policy in this context, as our primary conclusion highlights
the necessity for a policy mix that combines objectives of structural change and
gender equity. Any industrial policy, or initiative aimed at promoting female labor
force participation, must carefully consider its potential effects on the gender
wage gap.
64 Diana Suárez · Florencia Fiorentin · Florencia Barletta
REFERENCES
Abegaz, M., and Nene, G. (2018). Gender wage and productivity gaps in the
manufacturing industry. The case of Ghana. Economic Papers: A Journal of
Applied Economics and Policy, 37(3), 313–326.
Ahmed, S., and McGillivray, M. (2015). Human capital, discrimination, and the
gender wage gap in Bangladesh. World Development, 67, 506–524.
Allen, J., and Sanders, K. (2002). Gender gap in earnings at the industry level.
European Journal of Women’s Studies, 9(2), 163–180.
Angelov, N., Johansson, P., and Lindahl, E. (2016). Parenthood and the gender
gap in pay. Journal of Labor Economics, 34(3), 545–579.
Anker, R. (1997). La segregación profesional entre hombres y mujeres. Revista
Internacional Del Trabajo, 16(3), 343–370.
Aragón, F. M., Rud, J. P., and Toews, G. (2018). Resource shocks, employment,
and gender: evidence from the collapse of the UK coal industry. Labour
Economics, 52, 54–67.
Argoitia, J. M., Warszawski, T. F., de las Nieves Puglia, M., and Erlauer, M. F.
(2023). Ser y parecer minera: inserción sostenible de las trabajadoras
mineras.
Azmat, G., and Ferrer, R. (2017). Gender gaps in performance: Evidence from
young lawyers. Journal of Political Economy, 125(5), 1306–1355.
Badel, A., and Goyal, R. (2024). When will global gender gaps close? Economics
Letters, 237, 111544. https://doi.org/10.1016/J.ECONLET.2024.111544
Baum, S., and Benshaul-Tolonen, A. (2021). Extractive industries and gender
equality. Review of Environmental Economics and Policy, 15(2), 195–215.
Blau, F. D., and Kahn, L. M. (2017). The gender wage gap: Extent, trends, and
explanations. Journal of Economic Literature, 55(3), 789–865.
Brynin, M., and Perales, F. (2016). Gender wage inequality: The de-gendering of
the occupational structure. European Sociological Review, 32(1), 162–174.
Cao, H., and Vaca, J. (2006). Desarrollo regional en la Argentina: la centenaria
vigencia de un patrón de asimetría territorial. EURE (Santiago), 32(95), 95–
111.
CEP XXI. (2021). Empleo femenino y composición sectorial en Argentina 2007-
2021. Ministerio de Desarrollo Productivo.
Cimoli, M., and Porcile, G. (2016). Latin American structuralism: the co-evolution
of technology, structural change and economic growth. Handbook of
Alternative Theories of Economic Development, 228–239.
Durán Lima, J. E., and Galván, M. (2023). América Latina: Brechas salariales en
sectores exportadores y no exportadores según regresiones por cuantiles,
2011-2018. Brechas de Género En Las Cadenas Globales de Valor de
América Latina y El Caribe: Nuevos y Viejos Desafíos En Un Contexto de
Incertidumbre. Santiago: CEPAL, 2023. LC/TS. 2023/17. p. 75-121.
Fagerberg, J. (2018). Technological progress, structural change and productivity
growth: a comparative study. In Innovation, Economic Development and
Policy (pp. 214–232). Edward Elgar Publishing.
65
FORECASTING THE GENDER WAGE GAP IN ARGENTINA FROM A PRODUCTIVITY-BASED APPROACH
REVISTA DE ECONOMÍA MUNDIAL 68, 2024, 47-66
Hausman, J. A. (1978). Specification tests in econometrics. Econometrica:
Journal of the Econometric Society, 1251–1271.
Jayachandran, S. (2015). The roots of gender inequality in developing countries.
Economics, 7(1), 63–88.
Katz, J. (2018). Las cuatro argentinas que conviven... pero no conversan. Una
historia de éxitos, fracasos y desencuentros. Boletin Informativo Techint, 356.
Kotsadam, A., and Tolonen, A. (2016). African Mining, Gender, and Local
WORLDDEV.2016.01.007
Malerba, F., Orsenigo, L., and Peretto, P. (1997). Persistence of innovative
activities, sectoral patterns of innovation and international technological
specialization. International Journal of Industrial Organization, 15(6), 801–
826. https://doi.org/10.1016/S0167-7187(97)00012-X
Montobbio, F. (2002). An evolutionary model of industrial growth and structural
change. Structural Change and Economic Dynamics, 13(4), 387–414.
Niembro, A. A., and Starobinsky, G. (2021). Sistemas regionales de ciencia,
tecnología e innovación en la periferia de la periferia: un análisis de las
provincias argentinas (2010-2017).
Olivetti, C., and Petrongolo, B. (2014). Gender gaps across countries and skills:
Demand, supply and the industry structure. Review of Economic Dynamics,
17(4), 842–859. https://doi.org/10.1016/J.RED.2014.03.001
Olivetti, C., and Petrongolo, B. (2016). The evolution of gender gaps in
industrialized countries. Annual Review of Economics, 8, 405–434.
Pavitt, K. (1984). Sectoral patterns of technical change: Towards a taxonomy and
a theory. Research Policy. https://doi.org/10.1016/0048-7333(84)90018-0
Redmond, P., and McGuinness, S. (2019). The gender wage gap in Europe: Job
preferences, gender convergence and distributional effects. Oxford Bulletin
of Economics and Statistics, 81(3), 564–587.
Rendall, M. (2013). Structural change in developing countries: has it decreased
gender inequality? World Development, 45, 1–16.
Saviotti, P., and Frenken, K. (2008). Export variety and the economic performance
of countries. Journal of Evolutionary Economics, 18(2), 201–218. https://doi.
org/10.1007/s00191-007-0081-5
Si, C., Nadolnyak, D., and Hartarska, V. (2021). The gender wage gap in
developing countries. Applied Economics and Finance, 8(1), 1–12.
Sin, I., Stillman, S., and Fabling, R. (2022). What drives the gender wage gap?
examining the roles of sorting, productivity differences, bargaining, and
discrimination. Review of Economics and Statistics, 104(4), 636–651.
UN-Women. (2022). Progress on the sustainable development goals. The gender
snapshot. 2022.
WEF. (2023). Global Gender Gap Report 2023.
Wooldridge, J. (2010). Econometric Analysis of Cross Section and Panel Data
(2nd ed.). MIT Press.
66 Diana Suárez · Florencia Fiorentin · Florencia Barletta
APPENDIX 1
TABLE A. DEFINITION OF VARIABLES
Variable Definition Values Source
LTotal registered employment (except public employment).
Number of persons. 1996-2022. 0 to ∞OEDE
W Average salary per employee. US$ dollars. 1996-2022. 0 to ∞OEDE
EAP
Economically active population. People who have an occupa-
tion or who, without having one, are actively looking for one.
Number of persons. 2003-2022.
0 to ∞OEDE
GPV Gross productive value. US$ dollars. 2004-2022. 0 to ∞INDEC
Grad Graduation. Total number of graduated persons. 1999-2022. 0 to 1 Min. Educ.
F Female. Based on national ID. 1 if yes; 0 otherwi-
se. OEDE
S Industry on two- and three-level ISIC Rev. 3. 45 industries OEDE
