See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/388200762

Is High-Tech Trade a Threat to Employment in Brazil?

Article in International Journal of Economics and Finance · January 2025

DOI: 10.5539/ijef.v17n3p1

CITATIONS READS

0 41

3 authors:

Rhemanuérick Silva Queirós Elano Ferreira Arruda

Federal University of Ceará Federal University of Ceará
1 PUBLICATION 0 CITATIONS 47 PUBLICATIONS 169 CITATIONS

SEE PROFILE SEE PROFILE

Felipe De Sousa Bastos

Federal University of Ceará
17 PUBLICATIONS 22 CITATIONS

SEE PROFILE

All content following this page was uploaded by Elano Ferreira Arruda on 21 January 2025.

The user has requested enhancement of the downloaded file.

 International Journal of Economics and Finance; Vol. 17, No. 3; 2025
ISSN 1916-971X E-ISSN 1916-9728
Published by Canadian Center of Science and Education

Is High-Tech Trade a Threat to Employment in Brazil?

Rhemanuérick Silva Queirós1, Elano Ferreira Arruda2 & Felipe de Sousa Bastos2
1
PhD Student in Economics, Federal University of Ceará(CAEN/UFC), Fortaleza, Brazil
2
PhD in Economics, Federal University of Ceará(CAEN/UFC), Fortaleza, Brazil
Correspondence: Elano Ferreira Arruda, Avenida da Universidade, 2700 – 2ºFloor. Zip code: 60020-181.
Benfica – Fortaleza/CE, Brazil. E-mail: elano@caen.ufc.br

Received: December 1, 2024 Accepted: January 15, 2025 Online Published: January 20, 2025
doi:10.5539/ijef.v17n3p1 URL: https://doi.org/10.5539/ijef.v17n3p1

Abstract
Unemployment is one of the primary economic and social challenges, particularly for developing regions. The
automation of processes and production techniques, along with the digitalization of services, has long instigated
fear and uncertainty due to their potential to induce what is commonly referred to as ―technological
unemployment.‖ From the perspective of international trade, the trade in high-tech products can lead to
unemployment regardless of the factor endowments prevalent in the region in question. This paper analyses the
effects of traditional and high-tech trade on the Brazilian state’s labor market. We utilized a panel data
framework that includes information for the Brazilian states, covering the period from 2000 to 2020, to estimate
dynamic labor demand equations using the System-GMM estimator. The results indicate a positive relationship
between traditional trade and employment, regardless of the trade proxy adopted and the economic sector
investigated. However, this relationship becomes negative when high-tech trade variables are employed. This
paper examines how high-tech trade in Brazil impacts employment within a regional and sectoral context,
providing important insights for the formulation of public policies aimed at mitigating unemployment.
Keywords: high-tech trade, employment, sectoral impact
1. Introduction
Brazilian economic literature and history extensively discuss the changes in the Brazilian economy due to trade
liberalization since the early 1990s. We have observed these changes in activities related to the external sector
and the domestic market.
Trade liberalization in Brazil led to increased efficiency and productivity but also had negative consequences for
the labor market. While it boosted economic growth by allowing the import of previously prohibited goods and
financial liberalization, it also contributed to rising informality and unemployment. Studies by Moreira (1999),
Markwald (2001), Soares et al. (2001), Moreira and Najberg (2000), and Rossi Júnior and Ferreira (1999)
support these findings, highlighting the trade-off between economic efficiency and labor market outcomes at that
period.
This trade liberalization has led to several changes in the labor market, such as changes in the returns to skilled
labor (Gonzaga et al., 2006), worker transitions to other sectors, unemployment or labor force exit
(Menezes-Filho & Muendler, 2011), changes in wages (Kovak, 2013), and the labor market’s response to this
liberalization may take a long time, reducing potential welfare gains (Dix-Carneiro, 2014; Dix-Carneiro &
Kovak, 2015b; Dix-Carneiro & Kovak, 2017), as well as leading to changes in the skill premium (Dix-Carneiro
& Kovak, 2015a) and reductions in gender inequality (Juhn, Vjhelyi, & Villegas-Sanchez, 2014).
With trade liberalization, new production structures are developed, new products are imported and exported, and
more sophisticated sectors, with high technological content, tend to grow. With this trade openness,
technological diffusion becomes more evident, and innovations are shared among industries (Gordon &
Gramkow, 2018).
Automation and digitalization of the economy, as well as the process of robotization, impact the labor market,
creating new jobs, requiring new skills and qualifications, and affecting the number of jobs in some sectors,
either positively or negatively, depending on the industry (Stryzhak, 2023; Tu & Pham, 2021).
Thus, in aggregate terms, it is expected that the export and import of high-tech products will affect the demand

1
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

for labor differently from what is observed in usual trade. Traditionally, exports create jobs in labor-abundant
countries, while imports reduce employment through import competition. In the case of high-tech exports, they
may not increase employment, even in labor-abundant countries. This may happen because the expansion of
high-tech exports parallels technological development. High-tech exports may employ more machines than
workers to perform tasks. Moreover, the impact of high-tech imports depends on the nature of the products.
Labor-saving final goods may replace workers in production, while intermediate goods may generate assembly
jobs (Idris et al., 2020).
In Brazil, trade liberalization opened the market for products not produced domestically and enabled national
production to flow abroad in the most diverse sectors. However, according to the Organization for Economic
Cooperation and Development (OECD) classification, when examining the level of export and import by
technology intensity, between 1991 and 2011, the aircraft sector in Brazil was the central high-tech export sector
(do Carmo Hermida et al., 2015).
Regarding BRICS countries, the United States, and the European Union, Brazil mainly exports low - and
medium-tech products and imports medium - and high-tech products. This differs from trade with Mercosur
countries, where medium-high-tech products are imported and exported (Martins, 2017).
Figure 1 shows a notable increase in exports from the uncategorized sector (represented by agriculture and
commodities) and low-tech products. From the perspective of imports, Brazil’s trade portfolio has a more
significant share of products classified as medium-high-tech and high-tech. Overall, high-tech exports showed a
downward trend, while high-tech imports remained relatively constant, around 20%. Therefore, it can be
concluded that technology-intensive, more complex sectors needed to receive the necessary attention and
consequently lost their place in Brazil’s industrial structure (Erber, 2000).

Figure 1. Share of exports and imports in trade by technology level in Brazil (%)

Although Brazil’s foreign trade portfolio predominantly consists of low-technology products, as noted by
Teixeira et al. (2021) and Santos et al. (2020), there is evidence that investments in high-technology sectors are
essential for the country’s economic growth (Morceiro & Guilhoto, 2020). For developing countries, sustained
long-term growth tends to be associated with high-technology productive structures, which are characterized by
more excellent learning benefits and spillover effects, with a higher capacity to induce growth and can serve as a
driving force for economic dynamism (Araujo & Peres, 2023; Marconi et al., 2014; Dasgupta & Singh, 2005).

Figure 2. Share of high-tech exports and imports in GDP

2
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

Figure 2 shows the share of high-technology exports and imports in the Gross Domestic Product (GDP).
Throughout the analyzed period, the volumes of both imported and exported high-tech products showed steady
growth, resulting in a progressive increase in their contribution to the Gross Domestic Product.
In terms of the labor market, when analyzing the composition of jobs in Brazil, Figure 3 shows that the industrial
sector accounts for the most significant number of jobs, followed by the agricultural sector, with the service
sector having the smallest share.

Figure 3. Employed population (thousands) in Brazil - Sectors

However, these occupations are primarily concentrated in industries with low to medium technological
dynamism. The agricultural sector consists of simple, labor-intensive products, and its competitive advantage
comes from the local availability of natural resources. The service sector comprises jobs that do not require high
levels of qualification and are not considered capable of driving economic dynamism (Lall, 2000; da Cruz et al.,
2007).
In Brazil, exports increased following trade liberalization, primarily in agrobusiness and activities related to
extractivism. In these low-tech complexity sectors, job creation is a trend. However, in high-tech sectors, where
changes in the technological structure of production have occurred, employment levels have decreased with
increased productivity (Perobelli et al., 2016).
Considering this, this work aims to analyze the effect of total trade and high-tech products trade on the labor
market in Brazilian states. Specifically, it investigates the relationship between total trade and high-tech trade at
state-level employment volumes using dynamic labor demand equations proposed by Idris et al. (2020).
Therefore, this study fills a gap in the literature regarding the impact of high-technology trade on the Brazilian
labor market, using 20 years of data on employment and the volume of this type of trade. Additionally, these
effects on the labor market will also be evaluated at the sectoral level, involving non-manufacturing sectors such
as agriculture and services, as well as at the aggregate level. For this purpose, dynamic labor demand equations
will be estimated for the agriculture, manufacturing, and services sectors. All these equations will be estimated
using System-GMM.
The results indicate that total trade variables show a positive relationship with employment levels at the
aggregate and sectoral levels. In contrast, when examining high-tech trade variables, a negative relationship with
employment levels is observed at both the aggregate and sectoral levels, regardless of the measure of high-tech
trade used.
This paper is organized as follows: In addition to this introduction, the second section will cover the literature
review on the topic, followed by a presentation of the methodology used. The fourth section will present the data
and variables utilized, the fifth section will present the results, and finally, the sixth section will offer the
concluding remarks.
2. Literature Review
This section will review literature on the impact of automation, trade liberalization, and high-tech trade on the
labor market.
2.1 Automation and Labor Market
When analyzing the impact of the high-tech market on employment levels, it becomes evident that both
automation and trade liberalization play significant roles. The rise of Industry 4.0 and the increasing
digitalization of the economy have brought about profound changes in the labor market.

3
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

Automation has led to the creation of new jobs, while simultaneously demanding new skills. Moreover, the
digitalization processes have affected countries and sectors in distinct ways, resulting in varying impacts on
employment levels (Stryzhak, 2023; Hepaktan & Şimşek, 2022; Tu & Pham, 2021; Kergroach, 2017; Arntz et al.,
2016).
Over the years, researchers have found that technological advancements harm the labor market, particularly in
areas with well-established routines and procedures that are replaced by computational algorithms (Autor &
Dorn, 2013).
Therefore, numerous studies have been conducted to analyze the impact of automation on the global labor
market. Frey and Osborne (2017) investigated the impact of automation on the labor market and its prospects,
identifying occupations susceptible to replacement by robots. This spread of robotics is associated with an
increase in unemployment, primarily due to the loss of jobs in the manufacturing industry that relied on manual
labor (Acemoglu & Restrepo, 2019).
The increasing adoption of robots in industries worldwide is associated with gains in productivity and labor. Fu
et al. (2021), using a panel data, found that increased robot adoption in 74 countries between 2004 and 2016 led
to higher earnings, especially for skilled workers in developing countries, without causing significant job losses.
The same pattern was observed in studies conducted in Emerging Countries, Latin America, and Brazil.
For developing countries, Pavez and Martínez-Zarzoso (2024) conducted a panel study of 16 sectors across 10
emerging countries from 2008 to 2014, examining the impact of domestic and foreign robots on the labor market.
Through an instrumental variables approach and a shift-share exposure index, they found that using foreign
robots negatively affected employment levels in emerging countries. For Latin America and Brazil, studies also
show a high probability of job transformation or disappearance due to automation (Brambilla et al., 2021;
Albuquerque et al., 2019).
Using Brazil as a reference, Stemmler (2019) investigated how automation affects an emerging economy. With
panel data and a shift-share approach, from 2000 to 2014, the author found that domestic automation decreases
the proportion of unskilled workers in the industry and increases the proportion of skilled workers. However, the
wage gap between these groups widens, increasing wage disparity.
To understand whether the growth of high-tech industries benefits low-skilled workers, Lee and Clarke (2019)
investigated the development of innovative industries in the labor market and their impact on low- and
medium-skilled workers. The authors found that the high-tech industry has a positive multiplier effect, creating 7
out of 10 jobs in the service sector; however, six are for low-skilled workers.
While most studies examine the formal labor market, Ottoni et al. (2022) analyze the impact of automation on
Brazil’s informal labor market, estimating that 58.1% of existing jobs, and 62% of informal jobs, could be at risk
in the next 10-20 years.
Thus, the literature shows that while automation benefits productivity, it can lead to job losses, whether through
unemployment, job transformation, or changes in job requirements.
2.2 Trade Liberalization and Labor Market
Trade liberalization has impacted various sectors of the economy in both developed and developing countries,
stimulating a wide range of studies, particularly regarding its impact on the labor market. This subsection will
explore empirical works on the subject.
A wide range of studies investigating the impacts of trade liberalization on the labor market can be found
worldwide. For Sri Lanka, Herath (2014) used OLS estimations to investigate the impact of trade liberalization
on employment levels between 1990 and 2012. The author found a positive relationship between export intensity
and employment levels. Conversely, import penetration harmed employment levels.
In a similar study, Rajesh and Sasidharan (2015) explored the potential impact of international trade on
manufacturing employment and wages in India from 1980 to 2005. Through panel data estimations, the authors
found that import penetration had a detrimental effect on job creation, particularly in export-intensive sectors.
For Vietnam, Van Ha and Tran (2017) used a quantile regression with fixed effects and data from 2010 to 2015.
The authors found that international trade is negatively associated with employment in low-employment firms
and positively associated with employment in high-employment firms.
Brummund and Connolly (2018), using panel data from 2004 to 2013, analyzed the impact of increased trade
between Brazil and China on the labor market. They found that exposure to exports reduced unemployment and
increased job transitions within the formal sector, while exposure to imports led to job transitions between the

4
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

industrial sector and unemployment.

Feenstra et al. (2019) investigated employment responses to Chinese import competition and global exports in
the United States. Using instrumental variable estimations from 1991 to 2007, the authors found a net loss of
jobs in the country.
Consequently, the effects of trade liberalization can be observed in a wide range of countries, regardless of their
level of development.
Specifically addressing Brazil, Gonzaga et al. (2006) employed several empirical exercises to examine the
traditional transmission mechanism of trade on the labor market. The authors used decomposition analysis of
changes, a panel regression of skilled labor prices, and a wage equation analysis, finding that skilled workers’
wage differentials decreased, and employment shifted from sectors intensive in skilled workers to sectors
intensive in unskilled workers.
Kovak (2013), using a model with region-specific factors, found that industries experiencing larger tariff cuts
had a greater negative impact on the labor market (job losses), while industries with smaller tariff cuts resulted in
a smaller negative impact on the labor market (job gains).
Dix-Carneiro (2014) argues that the Brazilian labor market’s responses to trade liberalization may take years and
be costly, as potential welfare gains are reduced due to this delayed adjustment. Using a dynamic equilibrium
model, the author further emphasizes that less educated and older workers face higher costs of sector switching,
in terms of average wages.
Dix-Carneiro and Kovak (2015a) extend Kovak’s (2013) model to include two types of labor (skilled and
unskilled) in Brazil. The authors found that a portion of the average wage premium for skill decreased due to
trade liberalization.
In another study, Dix-Carneiro and Kovak (2015b) examined the long-term impact of trade liberalization in
Brazil. Local negative trade shocks induced workers to move from the formal trade sector to the formal
non-trade sector; moreover, in the long run, unemployed workers ended up finding reemployment in the informal
sector. In Dix-Carneiro and Kovak (2017), regions with larger tariff declines experienced a deterioration of
formal labor market outcomes compared to other regions.
To complement the existing literature, Dix-Carneiro and Kovak (2019) conducted a study using cross-sectional
data from Brazilian household surveys to identify labor market adjustments post-liberalization in the early 1990s.
The authors reported that in regions experiencing more significant tariff reductions, formal sector workers
displaced by trade often shifted to the informal sector after periods of unemployment.
Regarding the relationship between trade liberalization and the labor market in Brazil, Borges et al. (2023)
conducted a study to reassess the effect of 1990s liberalization on labor productivity and employment in Brazil.
They used a method to decompose labor productivity before and after trade liberalization. Results suggest that,
although there was an increase in average intersectoral productivity, there was also an increase in unemployment
and job displacement, with labor reallocating to lower productivity sectors.
A substantial amount of research highlights the effects of trade liberalization on labor markets, utilizing diverse
methodologies and datasets. Consistently, these studies indicate adverse consequences for wages and
employment.
2.3 High-Tech Trade and Labor Market
Regarding the high-tech market, there is still a limited body of research analyzing its impact on the labor market.
However, some studies have highlighted this issue and demonstrated its significance in this field of inquiry.
Idris et al. (2020) investigated whether a country’s involvement in high-tech trade would destroy or create jobs,
considering different possible effects on employment. Using a dynamic labor demand model to analyze the top
twenty high-tech exporting countries from 2007 to 2016, the study found a negative relationship between
high-tech trade and employment, highlighting a risk associated with technology-driven growth strategies.
Idris et al. (2021) examined the impact of engagement in high-tech trade on national competitiveness. Using
panel data from 20 countries between 2007 and 2016, they observed that high-tech exports drive technological
development in these countries and enhance national competitiveness.
To assess the effect of high-tech trade on economic growth in Malaysia, Lam et al. (2023) used quarterly data
from 1990 to 2018 and ARDL models. The authors found that high-tech trade, foreign direct investment, and
physical capital stock are determinants of economic growth.

5
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

Concerning technological trade, Ayhan and Elal (2023), using panel data from OECD countries and fixed-effect
estimations, found that GDP and trade liberalization positively impact the labor market. Bouattour et al. (2023)
provided empirical evidence of a non- linear effect of imported technology on industrial employment. Using
panel data and nonlinear threshold regression, covering both developed and developing countries from 2000 to
2019, the authors highlighted that this nonlinear effect depends on the country’s level of economic development.
The positive impact on industrial employment is weak but more pronounced in developed countries due to more
efficient compensation mechanisms compared to developing countries.
Bouattour et al. (2024) analyzed the impact of imported technology on employment in developed and developing
countries between 2000 and 2019. They found that while higher levels of imported technology can boost
industrial employment in developed countries, the opposite is true for developing countries.
Hence, more research is needed to understand the impact of high-tech trade on a country’s development. This
study aims to fill this gap in the literature by providing a Brazilian perspective and examining the impact of
high-tech trade on overall employment levels, as well as on the agricultural, industrial, and service sectors.
3. Method
3.1 Data Source and Descriptions
A panel data framework was utilized to verify high-tech trade’s impact on the Brazilian state’s labor market. This
panel includes data from the twenty-six federation states plus the Federal District, covering the period from 2000
to 2020.
The data were sourced from various institutions: the National Household Sample Survey (NHSS or PNAD) and
the Continuous PNAD (PNADC) for occupation variables; the Ministry of Development, Industry, Commerce,
and Services (MDIC) for foreign trade indicators; and the Brazilian Institute of Geography and Statistics (IBGE)
for GDP data.
Data on the employed population, both aggregated and sectoral — agriculture, manufacturing, and services —
were obtained from PNAD and PNADC. PNADC was prioritized and used from 2012 to 2020 as it better and
more accurately reflects the population profile (Ottoni & Barreira, 2016). PNAD was used to provide a sample
extending back to 2000, when the first year of high-tech trade data was available in Brazilian states. Both PNAD
surveys were harmonized according to the recommendations of Ottoni and Barreira (2016) and Veloso et al.
(2019).

Chart 1. Description of the variables used in the estimated models

Variable Description Source
Dependent
PO Occupied Population. PNAD and PNADC
Sectoral Occupied Population. Subscript j refers to the sectors of
POj PNAD and PNADC
agriculture, manufacturing, and services.
Explanatory
RM Average Labor Income PNAD and PNADC
Sectoral Average Labor Income. Subscript j refers to the sectors of
RMj PNAD and PNADC
agriculture, manufacturing, and services.
Gross Domestic Product in 2019 reais, deflated by the General Price
GDP IBGE
Index - Domestic Availability (IGP-DI).
GA The ratio of the sum of exports and imports to GDP. MDIC and IBGE
GAx Share of total exports in GDP MDIC and IBGE
GAm Share of total imports in GDP MDIC and IBGE
GAh The ratio of the sum of high-tech exports and imports to GDP. MDIC and IBGE
GAxh Share of high-tech exports in GDP. MDIC and IBGE
GAmh Share of high-tech imports in GDP. MDIC and IBGE
FCh
Share of high-tech trade flow (FCh) in total trade flow (FC) MDIC
FC
Xh
Share of high-tech exports in total trade flow. MDIC
FC
Mh
Share of high-tech imports in total trade flow. MDIC
FC
MTPh High-tech import penetration rate. MDIC
Source: Own elaboration.

6
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

As explanatory variables, average labor income at an aggregated and sectoral level (agriculture, manufacturing,
and services) will be used. Harmonizing data from PNAD and PNADC also obtained this information. The state
Gross Domestic Product (GDP) was appropriately deflated using the General Price Index–Internal Availability
(IGP-DI).
The first is the ratio of exports and imports to GDP - GA. The second is the ratio of total exports to GDP - GAx.
The third is the ratio of total imports to GDP - GAm.
A second group of variables relates to the volume of high-tech trade in relation to GDP, thus creating three
variables: total volume, export volume, and import volume of high-tech products relative to GDP – GAh , GAxh ,
and Gamh.
The third group of variables concerns the volume of high-tech products exported, imported, and total relative to
the trade flow. This results in three more variables: the share of high-tech trade flow in total trade flow (FCh /FC),
the share of high-tech exports in total trade flow (Xh ⁄FC), and the share of high-tech imports in total trade flow
(Mh ⁄FC).
Lastly, a variable is created to measure the penetration rate of high-tech imports, calculated as the total value of
high-tech imports as a percentage of total domestic demand. Domestic demand equals GDP minus the trade flow.
A summary description of the variables can be found in Chart 1.
The Ministry of Development, Industry, Commerce, and Services (MDIC) provides data on exports and imports
categorized by technological level in Brazil. This ministry offers a manual methodological way of organizing
trade data to align with the OECD’s definitions of technological levels.
The estimations will be conducted at four levels: one aggregate and three sectoral levels (agriculture,
manufacturing, and services). The trade variables will be adjusted to assess the impact on employment at both
the aggregate and sectoral levels, resulting in 40 estimated models.
3.2 Econometric Approach
The dynamic labor demand equations employed in this study were developed under the premise that high-tech
trade generates greater efficiency in productivity. Thus, it follows the approach proposed by Greenaway et al.
(1999) regarding using the dynamic labor demand model. Its augmented model will include a high-tech trade
variable in the state-level dynamic panel specification. The dynamic labor demand equation that defines the
estimated model will have the following form:
lnPOit =α0 + α1 lnPOit-1 +α2 lnPIBit +α3 lnRMit +α4 lnCit +fi +εit (1)
Where POit represents the employment level in state i at time t; PIBit represents the real GDP in state i at time
t; RMit represents the real average wage in state i at time t; Citrepresents a variable related to international
trade. The trade and high-tech trade variables used as proxies for Cit are described in Table 1; fi represents an
individual fixed effect.
The variable POit-1 characterizes a dynamic panel. In this case, using Ordinary Least Squares (OLS) leads to
biased estimates for the coefficient α1 due to endogeneity, known as dynamic panel bias. Even within-group
estimation is insufficient to eliminate this bias. One option to mitigate endogeneity is to use a first-difference
transformation and estimate it using GMM (Generalized Method of Moments).
The equation to be estimated can be defined as:
∆lnPOit =α1 ∆lnPOit-1 +α2 ∆lnPIBit +α3 ∆lnRMit +α4 ∆lnCit +∆fi +∆εit (2)
However, despite this transformation attempting to eliminate endogeneity, there will still be a correlation
between ∆lnPOit-1 and ∆εit . Therefore, to obtain unbiased results despite the endogeneity problem, the
approach proposed by Arellano and Bond (1991) will be used. In their proposal, the authors suggest using the
first difference of (POit-1) with lags more significant than one to address this endogeneity issue.
To mitigate the endogeneity problem in dynamic panels, Arellano and Bover (1995) and Blundell and Bond
(1998) proposed System-GMM, a more suitable technique for cases with high persistence in the dependent
variable. System-GMM combines Equation (1) and Equation (2) and uses lagged first-differenced and lagged
level variables as instruments for those level variables.
The Arellano and Bond test for first and second-order autocorrelation is applied to investigate the results’
adequacy, as dynamic panels are sensitive to autocorrelation in residuals. For consistency in the estimates, the
null hypothesis of no first-order autocorrelation should be rejected, and the null hypothesis of no second-order

7
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

autocorrelation should not be dismissed. Finally, Hansen’s (1982) J-test will be conducted to examine the
validity of the instruments.
4. Results and Discussions
4.1 Pre-Estimation Results
A panel with information from all states of the federation and the Federal District between 2000 and 2020 was
used to compare the relationships investigated in the study. Table 1 summarizes the descriptive statistics of the
variables under study in their original form.

Table 1. Descriptive statistics

Variables Mean Standard Deviation Minimum Maximum
PO 3,226,577 4,041,832 80,019.63 22,737,451.25
POa 426,222.6 401,985.7 3,319.46 2,093,371.62
POi 708,981 1,008,398 12,276.71 5,746,945.00
POs 1,943,500 2,713,933 55,781.89 16,016,778.50
RM 1,674.06 438.87 709.06 2,720.77
RMa 1,106.45 554.72 208.88 2,705.21
RMi 1,529.64 438.87 629.82 2,739.73
RMs 1,636.58 408.55 773.65 2,794.09
GDP 58,200.00 112,000.00 819.00 858,000.00
GA 0.17 0.13 0.00 0.64
GAx 0.10 0.09 0.00 0.61
GAm 0.07 0.08 0.00 0.54
GAh 0.02 0.05 0.00 0.34
GAxh 0.00 0.01 0.00 0.10
GAmh 0.01 0.04 0.00 0.25
FCh
0.07 0.11 0.00 0.62
FC
Xh
0.01 0.02 0.00 0.20
FC
Mh
0.05 0.09 0.00 0.52
FC
MTPh 0.02 0.07 0.00 0.54
Source: Own elaboration.

Descriptive statistics indicate, on average, a higher concentration of the employed population in the services
sector, followed by the industrial sector and, to a lesser extent, the agricultural sector. The services sector
reported the highest variability in this indicator. Regarding average income, the services sector continues to have
the highest value (R$ 1,636.58), followed by the industrial and agricultural sectors, with R$ 1,529.64 and
R$ 1,106.45, respectively.
Regarding trade variables, there is a higher participation of the volume of exports relative to GDP compared to
imports, with values of 0.10 and 0.07, respectively. However, when considering high-tech trade flow, high-tech
imports have a higher participation relative to GDP than exports, with values of 0.01 and 0.00, respectively. This
same trend is observed when examining the average participation of high-tech imports and exports in the total
trade flow, with higher involvement of high-tech imports compared to exports, with values of 0.05 and 0.01,
respectively. It is essential to note the wide range observed in these indicators, reflecting significant differences
among Brazilian states in this type of trade.
4.2 Empirical Results
To examine the impact of trade, and more specifically, high-tech trade, on aggregate and sectoral employment
levels in Brazilian states, dynamic labor demand equations were estimated using the System-GMM estimator as
previously described. The discussion in this section will focus primarily on the coefficients of the trade variables.
However, the complete results of the estimated equations and post-estimation tests are available in the Appendix,
tables A1, A2, A3, and A4.
Figures 4, 5, 6, and 7 present the effects of trade and high-tech trade on the labor market of Brazilian states. They
present the impact of openness, export and import participation, and import penetration variables on state-level
employment, that is, on the employed population at the aggregate level and in the agricultural, industrial, and
service sectors.

8
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

First, the effects of trade on aggregate employment in Brazilian states are discussed based on the results shown
in Figure 4. When analyzing trade variables that do not focus solely on high-tech trade - GA, GAx, and GAm -
a consistently positive and significant impact is observed. These results indicate that Brazilian states are
abundant in labor, and the hypothesis that trade can reduce occupied population through increased trade
competition is not verified, results that align with those found by Ayhan and Elal (2023).

Figure 4. Effects of trade and high-tech trade on aggregate levels of employment

Note. Confidence interval presented at 95% level; *Significant at 10% level.

On the other hand, high-tech trade variables - GAh and FCh ⁄FC - report a negative and significant effect.
Similarly, the participation of exports, used as proxies for high-tech exports - GAxh and Xh ⁄FC - and the
proxy variables for high-tech imports - GAmh , Mh ⁄FC, and MTPh - also negatively and significantly impact
aggregate state-level employment. All these findings corroborate the evidence observed by Idris et al. (2020),
Brambilla et al. (2021), and Bouattour et al. (2024). This evidence indicates that the production process of
high-tech exported goods has intensified the substitution of workers for machinery. Additionally, it may reflect
that high-tech imports are concentrated in labor-saving final goods, leading to unemployment.
For the agricultural sector, the effects of trade and high-tech trade are highlighted in Figure 5. For this sector,
trade variables GA and GAx positively impact the occupied population. The variable GAm was not found to
be significant. The high-tech trade variables GAxh and Xh ⁄FC were also not found to be significant at any
relevant level. From variables GAh and FCh ⁄FC, with a 1% change in these variables leading to a reduction of
approximately 0.05% and 0.04%, respectively, in the agricultural sector’s occupied population.

Figure 5. Effects of trade and high-tech trade on employment in the agricultural sector
Note. Confidence interval presented at 95% level; *Significant at 10% level.

Similarly, variables associated with high-tech imports - GAmh , Mh ⁄FC, and MTPh - had a negative impact of
approximately 0.03%, 0.03%, and 0.02% on the sector’s employed population, respectively. Despite Brazil’s
abundance of natural resources, Brazilian agriculture has been characterized by high productivity and the use of
technology, and more significant high-tech trade, primarily through imports, seems to foster a labor substitution
effect, reducing the number of workers employed. This productivity gain is due to a greater use of machinery
and equipment with a certain degree of technology, innovations in production processes, and investments in
research and innovation in this sector. It is also possible to highlight the role of Embrapa in this process despite a
reduction in the employed population, as raised by Gasques et al. (2010) and Garcia et al. (2013).

9
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

For the manufacturing sector (Figure 6), only the participation of high-tech exports in GDP, GAxh , was not
statistically significant. The variables GA, GAx, and GAm positively affected the employed population. The
most significant impacts can be observed in variables GAh and FCh ⁄FC , with a negative impact of
approximately 0.03% if there is a positive 1% change in these high-tech trade variables, a level like that
observed in the agricultural sector.

Figure 6. Effects of trade and high-tech trade on employment in the manufacturing sector
Note. Confidence interval presented at 95% level; *Significant at 10% level.

The Brazilian industrial structure is composed mainly of low and medium-low- technology industries,
accounting for 70% of the employed population in the country. High-tech industries employ only 5% of the
employed population in Brazil. However, in Brazil, production process innovations predominately result from
high-tech trade. Thus, this type of innovation reduces employability, substituting labor for technology, as studied
by Mérida et al. (2022).
Finally, the results for the service sector can be verified in Figure 7. In this case, GAxh was the only variable
that was not statistically significant. The aggregate trade variables GA, GAx, and GAm were significant and
had a positive effect on the sector’s employed population. High-tech trade variables had a negative impact on
employment. GAh and FCh ⁄FC reported the largest impacts, around 0.02%.
The aggregate trade openness variables showed a positive and significant effect on state employment in most
cases, corroborating Ayhan and Elal’s (2023) argument that trade openness positively influences the labor market.
Overall, and in an asymmetric manner, the results also indicate a negative relationship between high-tech trade
and employment, in line with the findings of Idris et al. (2020), Brambilla et al. (2021), and Bouattour et al.
(2024). Specifically, for high-tech imports, this negative relationship between trade and employment
corroborates the results of Herath (2014), Rajesh and Sasidharan (2015), Van Ha and Tran (2017) and Feenstra et
al. (2019).

Figure 7. Effects of trade and high-tech trade on employment in the service sector
Note. Confidence interval presented at 95% level; *Significant at 10% level.

When analyzing the other estimation coefficients, such as those for GDP and Average Income, the results have
the expected sign and significance. There is a negative relationship between Average Income and the Employed
Population; that is, an increase in this variable leads to a reduction in the levels of the Employed Population. The
average income coefficients oscillate between -0.41% and -0.016% for the aggregate-level estimation. For the

10
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

agricultural and industrial sector estimations, the average income coefficients for these sectors oscillate between
-0.20% and 0 and -0.11% and 0, respectively. These coefficients oscillate between -0.28% and -0.13% for the
service sector.
An analysis of the remaining variables in the estimated model reveals the following results: regarding Real
Product, the results show a positive relationship between this variable and employment levels. The Real Product
coefficients oscillate between 0.16% and 0.31% for aggregate-level estimations. Regarding the agricultural
sector, this variation of estimated coefficients ranges from 0.06% to 0.6%. For the manufacturing sector, the
coefficients vary between 0.01% and 0.16%. Finally, for the service sector, the estimated coefficients for Real
Product vary between 0.14% and 0.22%. In other words, increased output is associated with increased
employment levels, corroborating the results obtained by Idris et al. (2020). As for the results for the lagged
variable of the employed population, since all estimated coefficients are within the interval [0,1], they indicate
that the volume of employment converges to a stationary equilibrium; that is, Brazilian state employment does
not exhibit explosive behavior.
Regarding the robustness of the estimations, the Arellano-Bond tests show that the null hypothesis of no
second-order autocorrelation in the residuals is not rejected at the 5% significance level in all estimations.
Similarly, when analyzing the Hansen test, it is found that the null hypothesis that the instruments are valid
cannot be rejected for all estimated models.
Finally, suppose Brazil aims to become a major player as an exporter or importer in the high-tech trade, given its
negative impact on employment levels. In that case, there is an urgent need to implement public policies to
mitigate these adverse effects, primarily through improving human capital and expanding the productive
structure’s economic complexity.
5. Closing Remarks
This study analyzed the effects of total and high-tech trade on the Brazilian labor market at the aggregate and
sectoral levels, using System-GMM estimation of dynamic labor demand equations in a panel of Brazilian states
between 2000 and 2020.
By analyzing the aggregate trade openness variables, a positive relationship with the level of employed
population was generally found, corroborating Ayhan and Elal (2023) and Feenstra et al. (2019). On the other
hand, when investigating high-tech trade, a negative impact on job creation was found at both the aggregate and
sectoral levels, confirming the findings of Idris et al. (2020), Pavez and Martí
nez-Zarzoso (2024) and Bouattour
et al. (2024). The literature shows that the relationship between high-tech trade and the labor market in
developing countries, such as Brazil, tends to be negative, contrary to what happens in developed countries
(Bouattour et al., 2023).
Therefore, while the development of high-tech trade is essential for a country’s growth, it can cause some
problems. One of these problems is unemployment, an adversity most developing countries face. Thus, in
addition to incentivizing this type of trade, accompanying policies should reduce the unemployment that this
trade may cause, either through labor reallocation or through incentives for education and changes in the
employment sector.
In future research, it could be verified whether these results hold for other developing countries, such as other
Latin American countries, and the effect of this type of trade on other socioeconomic variables, such as GDP per
capita and the Gini index. Finally, it may also be interesting to study the short-term and long-term effects of
high-tech trade on the labor market using time series econometric methodologies.
Acknowledgments
The authors would like to thank the Foundation for Support of Scientific and Technological Development of
Ceará(FUNCAP) for funding this research, project ITR-0214-00123.01.00/23
References
Acemoglu, D., & Restrepo, P. (2019). Automation and New Tasks: How Technology Displaces and Reinstates
Labor. Journal of Economic Perspectives, 33(2), 3-30. https://doi.org/10.1257/jep.33.2.3
Albuquerque, P. H., Saavedra, C. A. P. B., de Morais, R. L., Alves, P. F., & Peng, Y. (2019). Na era das máquinas,
o emprego é de quem? Estimação da probabilidade de automação de ocupações no Brasil. Texto para
Discussão, No. 2457.
Araujo, E., & Peres, S. (2023). Estrutura produtiva e crescimento econômico no Brasil. In Da Vitoria (coord.) &
Francisco Jr.; technical team: Pedro Garrido da Costa Lima (org.) et al., Retomada econômica e geração de

11
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

emprego e renda no pós-pandemia (1st ed., p. 403). Brasí

lia: Câmara dos Deputados, Edições Câmara.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an
application to employment equations. Review of Economic Studies, 58, 277-97.
https://doi.org/10.2307/2297968
Arellano, M. and Bover, O. (1995). Another look at the instrumental variable estimation of error-components
models. Journal of Econometrics, (1), 29-51. https://doi.org/10.1016/0304-4076(94)01642-D
Arntz, M., Gregory, T., & Zierahn, U. (2016). The Risk of Automation for Jobs in OECD Countries: A
Comparative Analysis. OECD Social, Employment and Migration Working Papers, No. 189. Paris: OECD
Publishing.
Autor, D. H., & Dorn, D. (2013). The growth of low skill service jobs and the polarization of the US labor
market. American Economic Review, 103(5), 1553-1597. https://doi.org/10.1257/aer.103.5.1553
Ayhan, F., & Elal, O. (2023). The IMPACTS of technological change on employment: Evidence from OECD
countries with panel data analysis. Technological Forecasting and Social Change, (190).
https://doi.org/10.1016/j.techfore.2023.122439
Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models.
Journal of Econometrics, 87(1), 115-143. https://doi.org/10.1016/S0304-4076(98)00009-8
Borges, T., Romero, J. P., & Silveira, F. (2023). Abertura comercial, produtividade e emprego no Brasil.
Brazilian Journal of Political Economy, 3(4), 914-935. https://doi.org/10.1590/0101-31572023-3450
Bouattour, A., Kalai, M., & Helali, K. (2023). The nonlinear impact of technologies import on industrial
employment: A panel threshold regression approach. Heliyon, 9(10).
https://doi.org/10.1016/j.heliyon.2023.e20266
Bouattour, A., Kalai, M., & Helali, K. (2024). Threshold effects of technology import on industrial employment:
A panel smooth transition regression approach. Journal of Economic Structures, 13(1).
https://doi.org/10.1186/s40008-023-00322-x
Brambilla, I., César, A. M., Falcone, G., & Gasparini, L. (2021). New technologies and the future of jobs in Latin
America. In Cracking the future of work: Automation and labor platforms in the Global South. CIPPEC,
Buenos Aires.
Brummund, P., & Connolly, L. (2019). Labor Market Adjustments to Trade with China: The Case of
Brazil. Universidad de Alabama, Tuscaloosa, AL. Retrieved from
http://barrett.dyson.cornell.edu/NEUDC/paper_189.pdf
Charles, K. K., Erik, H., & Notowidigdo, M. (2013). Manufacturing decline, housing booms, and
non-employment. NBER Working Paper, No. 18949. https://doi.org/10.2139/ssrn.2273684
da Cruz, M. J. V., Nakabashi, L., Porcile, G., & Scatolin, F. D. (2007). Uma análise do impacto da composição
ocupacional sobre o crescimento da economia brasileira. Revista Economia, 8(4), 55-73.
https://doi.org/10.5380/ret.v4i3.27373
Dasgupta, S., & Singh, A. (2005). Will services be the new engine of Indian economic growth? Development and
Change, 36(6), 1035-1057. https://doi.org/10.1111/j.0012-155X.2005.00449.x
Dix-Carneiro, R. (2014). Trade liberalization and labor market dynamics. Econometrica, 82(3), 825-885.
https://doi.org/10.3982/ECTA10457
Dix-Carneiro, R., & Kovak, B. K. (2015a). Trade liberalization and the skill premium: A local labor markets
approach. American Economic Review, 105(5), 551-557. https://doi.org/10.1257/aer.p20151052
Dix-Carneiro, R., & Kovak, B. K. (2015b). Trade reform and regional dynamics: Evidence from 25 years of
Brazilian matched employer-employee data (No. w20908). National Bureau of Economic Research.
https://doi.org/10.3386/w20908
Dix-Carneiro, R., & Kovak, B. K. (2017). Trade liberalization and regional dynamics. American Economic
Review, 107(10), 2908-2946. https://doi.org/10.1257/aer.20161214
Dix-Carneiro, R., & Kovak, B. (2019). Margins of labor market adjustment to trade. Journal of International
Economics, 117, 125-142. https://doi.org/10.1016/j.jinteco.2019.01.005
do Carmo, H. C., de Avellar, A. P. M., Xavier, C. L., dos Reis, A. B. M. et al. (2015). Desempenho e
fragmentação da indústria de alta tecnologia do Mercosul. Brazilian Journal of Political Economy, 35(3),

12
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

622-644. https://doi.org/10.1590/0101-31572015v35n03a13
Erber, F. S. (2000). Perspectivas da América Latina em ciência e tecnologia. Parcerias Estratégicas, 8, 181-200.
Feenstra, R. C., Ma, H., & Xu, Y. (2019). US exports and employment. Journal of International Economics, 120,
46-58. https://doi.org/10.1016/j.jinteco.2019.05.002
Frey, C. B., & Osborne, M. A. (2017). The future of employment: How susceptible are jobs to computerisation?
Technological Forecasting and Social Change, 114, 254-280. https://doi.org/10.1016/j.techfore.2016.08.019
Fu, X., Bao, Q., Xie, H., & Fu, X. (2021). Diffusion of industrial robotics and inclusive growth: Labour market
evidence from cross country data. Journal of Business Research, 122, 670-684.
https://doi.org/10.1016/j.jbusres.2020.05.051
Garcia, J. G., Bastos, E. T., Valdes, C., & Bacchi, M. R. P. (2013). Produtividade e crescimento: algumas
comparações. In E. de A. Alves, G. da S. Sousa, & E. G. Gomes (Eds.), Contribuição da Embrapa para o
desenvolvimento da agricultura no Brasil (pp. 140-155). Embrapa, Brasí
lia.
Gasques, J. G., Bastos, E. T., Valdes, C., & Bacchi, M. R. P. (2010). Produtividade total dos fatores e
transformações da agricultura brasileira: análise dos dados dos censos agropecuários. In J. G. Gasques, J. E.
R. V. Filho, & Z. Navarro (Eds.), A agricultura brasileira: desempenho, desafios e perspectivas (pp. 19-44).
IPEA, Brasilia.
Gonzaga, G., Menezes, F. N., & Terra, C. (2006). Trade liberalization and the evolution of skill earnings
differentials in Brazil. Journal of International Economics, 68(2), 345-367.
https://doi.org/10.1016/j.jinteco.2005.07.008
Gordon, J. L., & Gramkow, C. L. (2018). As características estruturais recentes da inserção externa brasileira e
suas principais implicações–2000/2010. Cadernos do Desenvolvimento, 6(9), 93-118.
Greenaway, D., Hine, R. C., & Wright, P. (1999). An Empirical Assessment of the Impact of Trade on
Employment in the United Kingdom. European Journal of Political Economy, 15(3), 485-500.
https://doi.org/10.1016/S0176-2680(99)00023-3
Hansen, L. (1982). Large sample properties of generalized method of moments estimators. Econometrica, 50,
1029-1054. https://doi.org/10.2307/1912775
Hepaktan, C. E., & Şimşek, D. (2022). Industry 4.0 and the Future of the Labor Market. İzmir Sosyal Bilimler
Dergisi, 4(2), 80-88. https://doi.org/10.47899/ijss.1174005
Herath, M. S. (2014). The Impact of International Trade on Employment Generation: A Sri Lankan Experience.
International Journal of Economics, Commerce and Management, II.
Idris, Z. Z., Ismail, N. W., Saifuzzaman, I., Hamzah, H. Z. et al. (2020). The impact of high-technology trade on
employment. The Journal of International Trade & Economic Development, 30, 512-529.
https://doi.org/10.1080/09638199.2020.1852301
Idris, Z. Z., Ismail, N. W., Saifuzzaman, I., Hamzah, H. Z. et al. (2021). High-Technology trade: Does it enhance
national competitiveness?. Jurnal Ekonomi Malaysia, 55(3), 35-48.
https://doi.org/10.17576/JEM-2021-5503-03
Juhn, C., Ujhelyi, G., & Villegas-Sanchez, C. (2014). Men, women, and machines: How trade impacts gender
inequality. Journal of Development Economics, 106, 179-193.
https://doi.org/10.1016/j.jdeveco.2013.09.009
Kergroach, S. (2017). Industry 4.0: New Challenges and Opportunities for the Labour Market. Foresight and STI
Governance, 11(4), 6-8. https://doi.org/10.17323/2500-2597.2017.4.6.8
Kovak, B. K. (2013). Regional effects of trade reform: What is the correct measure of liberalization?. American
Economic Review, 103(5), 1960-1976. https://doi.org/10.1257/aer.103.5.1960
Lall, S. (2000). The Technological structure and performance of developing country manufactured exports,
1985-98. Oxford Development Studies, 28(3), 337-369. https://doi.org/10.1080/713688318
Lam, L. F. L., Law, S. H., Azman-Saini, W. N. W., & Ismail, N. W. (2023). High technology trade and the roles
of absorptive capabilities in Malaysia. Millennial Asia, 14(3), 315-340.
https://doi.org/10.1177/09763996211065565
Lee, N., & Clarke, S. (2019). Do low-skilled workers gain from high-tech employment growth? High-technology
multipliers, employment and wages in Britain. Research Policy, 48(9).

13
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

https://doi.org/10.1016/j.respol.2019.05.012
Marconi, N., Magacho, G. R., & Rocha, I. L. (2014). Estratégias de desenvolvimento nos BRICs: Uma análise
insumo-produto. Revista Economia Ensaios, 29(1), 119-134.
Markwald, R. A. (2001). O impacto da abertura comercial sobre a indústria brasileira: Balanço de uma década.
Revista Brasileira de Comércio Exterior, 68(3), 7-36.
Martins, I. O. (2017). Análise do componente tecnológico das importações e exportações brasileiras para os
BRICS, Estados Unidos, Mercosul e União Europeia: 2007-2010. Debate Econômico, 5(2).
MDIC - Ministério do Desenvolvimento, Indústria, Comércio e Serviços. (2020). Manual for the Use of
Brazilian Foreign Trade Statistical Data. Brasília. Retrieved from
https://balanca.economia.gov.br/balanca/manual/Manual.pdf
Menezes-Filho, N. A., & Muendler, M. A. (2011). Labor reallocation in response to trade reform (No. w17372).
National Bureau of Economic Research. https://doi.org/10.3386/w17372
Mérida, S. M. A., Hasenclever, L., & de Carvalho, M. M. (2022). Inovação e emprego no Brasil Innovation and
employment in Brazil. Brazilian Journal of Development, 8(3), 17539-17539.
https://doi.org/10.34117/bjdv8n3-136
Morceiro, P. C., & Guilhoto, J. J. M. (2020). Adensamento produtivo e esgarçamento do tecido industrial
brasileiro. Economia E Sociedade, 29(3), 835-860. https://doi.org/10.1590/1982-3533.2020v29n3art07
Moreira, M. M. (1999). A Indústria Brasileira nos Anos 90. O que jáse pode dizer? In F. Giambiagi, & M. M.
Moreira (Eds.), A Economia Brasileira nos Anos 90, BNDES, Rio de Janeiro.
Moreira, M. M., & Najberg, S. (2000). Trade Liberalization in Brazil: Creating or Exporting Jobs? Journal of
Development Studies, (3), 78-99. https://doi.org/10.1080/00220380008422629
Ottoni, B., & Barreira, T. (2016). Metodologia de Retropolação da Pesquina Nacional por Amostra de
Domicí lios Contínua de 1992 a 2012. Technical Note from IBRE.
Ottoni, B., Oliveira, P. R., Estrela, L., Santos, A. T., & Barreira, T. (2022). Automation and job loss: The
Brazilian case. Nova Economia, 32(1), 157-180. https://doi.org/10.1590/0103-6351/6367
Pavez, L. R. D., & Martí nez-Zarzoso, I. (2024). The impact of automation on labour market outcomes in
emerging countries. The World Economy, 47, 298-331. https://doi.org/10.1111/twec.13523
Perobelli, F. S., de Andrade, B. S. Q., & Pereira, M. Z. (2016). Decomposição estrutural do emprego por grau de
instrução: Uma análise de insumo-produto para o perí odo pós-abertura (1990 a 2005). Nova Economia,
26(3), 909-942. https://doi.org/10.1590/0103-6351/2403
Rajesh, R., & Sasidharan, S. (2015). Impact of foreign trade on employment and wages in Indian manufacturing.
South Asia Economic Journal, 16(2), 209-232. https://doi.org/10.1177/1391561415598457
Rossi Júnior, J. L., & Ferreira, P. C. (1999). Evolução da produtividade industrial brasileira e abertura comercial.
Pesquisa e Planejamento Econômico, 29(1).
Santos, F. A. V., Silva, D. S., de Santana, J. R., & Ribeiro, L. C. D. S. (2021). Grau de especialização e
decomposição do crescimento das exportações brasileiras por intensidade tecnológica. RDE-Revista de
Desenvolvimento Econômico, 3(47).
Soares, S., Servo, L. M. S., & Saba, J. A. (2001). O que (não) sabemos sobre a relação entre abertura comercial e
mercado de trabalho no Brasil. Texto para Discussão, 843.
Stemmler, H. (2019). Does Automation Lead to De-Industrialization in Emerging Economies? - Evidence from
Brazil. CEGE Discussion Papers Number, 382. https://doi.org/10.2139/ssrn.3453313
Stryzhak, O. (2023). Analysis of Labor Market Transformation in the Context of Industry 4.0. Studia
Universitatis Vasile Goldiş, Arad-Seria Ştiinţe Economice, 33(4), 23-44.
https://doi.org/10.2478/sues-2023-0017
Teixeira, F. O., Freitas, C. A. D., & Coronel, D. A. (2021). Grau de especialização da cesta de exportações e sua
relação com a complexidade econômica das unidades da federação brasileiras. Pesquisa e Planejamento
Econômico (PPE), 51(3), 75-116. https://doi.org/10.38116/ppe51n3art3
Tu, A. T., & Pham, T. X. (2021). Vietnam’s national human resource development strategy in response to
industry 4.0: An analysis from a labor market perspective. Human Resource Development in Vietnam:

14
ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

Research and Practice, 69-96. https://doi.org/10.1007/978-3-030-51533-1_3

Van Ha, H., & Tran, T. Q. (2017). International trade and employment: A quantile regression approach. Journal
of Economic Integration, 32(3), 531-557. https://doi.org/10.11130/jei.2017.32.3.531
Veloso, F., Matos, S., & Peruchetti, P. (2019). Nota metodológica dos indicadores anuais de produtividade do
trabalho setorial no Brasil. Technical Note from IBRE.

Appendix
Empirical Results
Table A1. Estimations for the dependent variable Employed Population (aggregate level)
Dependent variable Total Employed Population (aggregate level)
Trade variable used (C):
Explanatory variables
GA GAh FCh/FC GAx GAxh Xh GAm GAmh Mh/FC MTPh
0.6586452* 0.7160739* 0.773301* 0.737342* 0.729608* 0.726146* 0.807908* 0.828032* 0.744536* 0.760986*
Ln(PO)t-1
-0.056686 -0.0624542 -0.268826 -0.036787 -0.058234 -0.056729 -0.035625 -0.027386 -0.04181 -0.046853
-0.405474* -0.3027066* -0.2392* -0.30988* -0.34132* -0.34755* -0.19901* -0.16376* -0.26879* -0.24542*
Ln(RM)
-0.0873661 -0.0742879 -0.054443 -0.054459 -0.073551 -0.073332 -0.054439 -0.044502 -0.055626 -0.05712
0.3038227* 0.2816459* 0.220996* 0.23737* 0.269485* 0.271736* 0.164332* 0.172528* 0.245834* 0.236617*
Ln(GDP)
-0.0584561 -0.0605382 -0.274525 -0.034772 -0.055566 -0.054678 -0.029832 -0.027561 -0.042371 -0.046922
0.0389838** -0.0178206* -0.01857* 0.026057** -0.0049*** -0.00544*** 0.021766*** -0.01307** -0.01258** -0.01117**
Ln (C)
-0.0143401 -0.0050931 -0.006671 -0.010489 -0.002717 -0.00269 -0.010612 -0.004888 -0.004912 -0.005212
0.3206495 -0.8956181* -0.62125** 0.15054 -0.43791 -0.39361 0.159094 -0.75177* -0.59743** -0.79059*
Constant
-0.3640304 -0.2458204 -0.292393 -0.32282 -0.264074 -0.269651 -0.299134 -0.235048 -0.220413 -0.278828
Arellano-Bond test for
first-order autocorrelation 0.002 0.001 0 0 0.018 0.019 0 0 0 0

Arellano-Bond test for

0.624 0.393 0.387 0.554 0.287 0.27 0.709 0.619 0.337 0.415
second-order autocorrelation
Hansen's J-test 0.188 0.161 0.175 0.128 0.276 0.278 0.129 0.175 0.19 0.17
Number of instruments 24 21 23 23 9 9 23 23 24 24
Prob>F 0 0 0 0 0 0 0 0 0 0
Number of observations 442 388 366 416 446 446 416 416 441 441
Number of groups 26 26 26 26 26 26 26 26 26 26
Note. In parentheses, standard error of the estimate; * significant at 1% level; ** significant at 5% level; *** significant at 10% level.

Table A2. Estimations for the dependent variable Employed Population (agriculture sector)
Dependent variable Total Employed Population (agriculture sector)
Trade variable used (C):
Explanatory variables
GA GAh FCh/FC GAx GAxh Xh GAm GAmh Mh/FC MTPh
0.7929876* 0.9508507* 0.937926* 0.810167* 0.93771* 0.849482* 0.834435* 0.949034* 0.942947* 0.929136*
Ln(PO)t-1
-0.0833537 -0.0304987 -0.02712 -0.046533 -0.031697 -0.070141 -0.055935 -0.022152 -0.022992 -0.028494
-0.19933* -0.0001059 -0.03642 -0.20248* -0.03729 -0.12156 -0.13691** -0.01335 -0.02931 -0.02319
Ln(RM)
-0.095152 -0.0440713 -0.038118 -0.05646 -0.036136 -0.066495 -0.056576 -0.030686 -0.362517 -0.024764
0.1395266 0.0814703*** 0.064655*** 0.133878* 0.595976*** 0.107339** 0.146792** 0.069372** 0.060734** 0.085052**
Ln(GDP)
-0.0845692 -0.0421843 -0.032325 -0.044293 -0.031786 -0.046791 -0.070645 -0.026725 -0.272526 -0.031498
0.0807273*** -0.0469432** -0.03976* 0.055158** -0.00298 0.005388 -0.00882 -0.02797** -0.0268* -0.02559**
Ln (C)
-0.0437184 -0.0175665 -0.007861 -0.025973 -0.007602 -0.010132 -0.031113 -0.0119 -0.009256 -0.012047
0.5727799 -1.759972** -7920854** 0.513654 -0.53958 0.213465 -0.75215 -1.23501** 0.765377* -1.30035**
Constant
-0.8553046 -0.7287104 -0.306165 -0.603817 -0.48965 -0.612002 -1,003,159 -0.474937 -0.268839 -0.481174
Arellano-Bond test for
0.002 0.001 0 0.013 0.018 0 0.014 0.003 0.003 0.006
first-order autocorrelation
Arellano-Bond test for
0.624 0.393 0.387 0.086 0.287 0.114 0.143 0.196 0.185 0.195
second-order autocorrelation
Hansen's J-test 0.188 0.161 0.175 0.308 0.276 0.357 0.273 0.175 0.371 0.507
Number of instruments 24 21 23 23 24 22 23 21 21 25
Prob>F 0 0 0 0 0 0 0 0 0 0
Number of observations 442 388 366 494 389 408 494 441 441 390
Number of groups 26 26 26 26 26 26 26 26 26 26
Note. In parentheses, standard error of the estimate; * significant at 1% level; ** significant at 5% level; *** significant at 10% level.
Source: Author’s own work.

15
 ijef.ccsenet.org International Journal of Economics and Finance Vol. 17, No. 3; 2025

Table A3. Estimations for the dependent variable Employed Population (manufacturing sector)
Dependent variable Total Employed Population (manufacturing sector)
Trade variable used (C):
Explanatory variables
GA GAh FCh/FC GAx GAxh Xh GAm GAmh Mh/FC MTPh
0.9499331* 0.9404013* 0.985683* 0.944137* 0.929493* 0.863647* 0.959896* 0.943095* 0.924876* 0.925209*
Ln(PO)t-1
-0.0400404 -0.0394313 -0.042566 -0.026802 -0.031988 0.048574 0.032314 0.02178 0.028911 0.033579
-0.092208 -0.0025905 0.030739 -0.10194* -0.07646 -0.06592 -0.05216 -0.06106 -0.08371 -0.04218
Ln(RM)
-0.0712486 -0.0791422 -0.067105 -0.045133 -0.045315 0.047025 0.063119 0.041316 0.051798 0.062859
0.0254869 0.0738703*** 0.014188 0.049343* 0.070955*** 0.156173* 0.011676 0.071269* 0.081164** 0.101094*
Ln(GDP)
-0.0512791 -0.0418981 -0.047778 -0.031686 -0.037758 0.05551 0.044944 0.022674 0.030363 0.034596
0.0493565** -0.0292242** -0.03219* 0.022484** -0.00023 -0.01117** 0.036798** -0.01557*** -0.01473** -0.02593*
Ln (C)
-0.0180799 -0.0137758 -0.007721 -0.00912 -0.005738 0.004659 0.014095 0.007635 0.006154 0.008746
0.7876942 -1.257652* -0.51264 0.283641 -0.33028 -183,822 0.729507 -0.72088** -0.53051** -1.4507*
Constant
-0.5818453 -0.4458481 -0.428529 -0.317593 -0.421095 0.735413 0.502667 0.345315 0.249635 0.353968
Arellano-Bond test for
0 0 0 0 0.001 0 0 0 0 0
first-order autocorrelation
Arellano-Bond test for
0.421 0.485 0.945 0.465 0.347 0.974 0.259 0.471 0.462 0.877
second-order autocorrelation
Hansen's J-test 0.152 0.161 0.158 0.179 0.119 0.26 0.164 0.157 0.19 0.196
Number of instruments 24 23 24 24 23 23 24 24 25 25
Prob>F 0 0 0 0 0 0 0 0 0 0
Number of observations 442 366 388 442 366 300 442 441 467 390
Number of groups 26 26 26 26 26 26 26 26 26 26
Note. In parentheses, standard error of the estimate; * significant at 1% level; ** significant at 5% level; *** significant at 10% level.
Source: Author’s own work.

Table A4. Estimations for the dependent variable Employed Population (service sector)
Dependent variable Total Employed Population (service sector)
Trade variable used (C):
Explanatory variables
GA GAh FCh/FC GAx GAxh Xh GAm GAmh Mh/FC MTPh
0.8156885* 0.8487687* 0.839519* 0.841318* 0.82074* 0.821833* 0.740039* 0.843311* 0.831846* 0.84435*
Ln(PO)t-1
-0.0236569 -0.0225664 -0.025011 -0.01865 -0.018821 -0.026407 -0.041325 -0.017133 -0.015909 -0.01761
-0.207089* -0.1313029* -0.16076* -0.17838* -0.20242* -0.19565* -0.27221* -0.16043* -0.17606* -0.15843*
Ln(RM)
-0.0348339 -0.335503 -0.032118 -0.025505 -0.031596 -0.049603 -0.056323 -0.031233 -0.027771 -0.031475
0.1611605* 0.163178* 0.160157* 0.146388* 0.181395* 0.180504* 0.214517* 0.159317* 0.164899* 0.158353*
Ln(GDP)
-0.0203337 -0.0222743 -0.023769 -0.017917 -0.020456 -0.027633 -0.037185 -0.017134 -0.015657 -0.017822
0.029844** -0.0243525* -0.01636* 0.013851** -0.00358 -0.00557*** 0.049969* -0.01018* -0.01107* -0.00974*
Ln (C)
-0.126677 -0.0053989 -0.004327 -0.005389 -0.004148 -0.003002 -0.017524 -0.003166 -0.00286 -0.003116
0.0873188 -1.197395* -0.69211* -0.1324 -0.63323** -0.68566** 0.362847 -0.72509* -0.57259* -0.725*
Constant
-0.2496119 -0.241169 -0.209023 -0.139492 -0.248407 -0.26928 -0.412368 -0.136172 -0.149542 -0.135061
Arellano-Bond test for
0 0 0 0 0 0 0 0 0 0
first-order autocorrelation
Arellano-Bond test for
0.466 0.393 0.374 0.421 0.316 0.802 0.638 0.427 0.337 0.423
second-order autocorrelation
Hansen's J-test 0.164 0.165 0.174 0.149 0.148 0.158 0.28 0.212 0.19 0.207
Number of instruments 24 24 24 24 22 22 24 24 24 24
Prob>F 0 0 0 0 0 0 0 0 0 0
Number of observations 442 388 388 442 408 345 442 441 441 441
Number of groups 26 26 26 26 26 26 26 26 26 26
Note. In parentheses. standard error of the estimate; * significant at 1% level; ** significant at 5% level; *** significant at 10% level.
Source: Author’s own work.

Copyrights
Copyright for this article is retained by the author(s), with first publication rights granted to the journal.
This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution
license (http://creativecommons.org/licenses/by/4.0/).

16

View publication stats