Análisis espacial del empleo manufacturero en México, 1984-2013 - Núm. 84, Enero 2020 - Revista Desarrollo y Sociedad - Libros y Revistas - VLEX 840033091

Análisis espacial del empleo manufacturero en México, 1984-2013

AutorLeobardo de Jesús-Almonte, Roldán Andrés-Rosales, Yolanda Carbajal-Suárez
Páginas91-129
91
desarro. soc. 71, primer semestre de 2013, pp. x-xx, issn 0120-3584
Revista
Desarrollo y Sociedad
84
Primer cuatrimestre 2020
PP. 91-129, ISSN 0120-3584
E-ISSN 1900-7760
Spatial analysis of manufacturing employment
in Mexico, 1984-2013
Análisis espacial del empleo manufacturero
en México, 1984-2013
Leobardo de Jesús-Almonte1
Roldán Andrés-Rosales2
Yolanda Carbajal-Suárez1
DOI: 10.13043/DYS.84.3
Abstract
The slow economic growth that characterized the Mexican economy during the
1984-2013 period has had a differential influence on the dynamics of manu-
facturing employment in the country’s 32 states. We estimate an employment
function from census information using a spatial panel technique and show
that both the endogenous and exogenous factors of the variables included in
the model act as triggers for employment growth in the manufacturing sec-
tor. Results confirm that states’ internal factors determine the trajectory of
employment growth, but so do the factors of neighboring states, which are
analyzed through direct, indirect and total impacts. The estimation of short
and long-term impacts allows the improvement of economic policy recom-
mendations regarding employment.
Keywords: Economic analysis, employment, econometrics, México.
JEL Classification: C51,O4, R1, R12.
1 Centro de Investigación en Ciencias Económicas, Facultad de Economía, Universidad Autónoma del Estado
de México. Correo electrónico: ldejesusa@uaemex.mx y ycarbajals@uaemex.mx, respectivamente.
2 Facultad de Estudios Superiores Cuautitlán-UNAM. Correo electrónico: roldandres@apolo.acatlan.
unam.mx.
Este artículo fue recibido el 24 de enero de 2019, revisado el 8 de abril de 2019 y finalmente aceptado
el 8 de julio de 2019.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
92
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Resumen
El lento crecimiento económico que ha caracterizado a la economía mexicana
de 1984-2013 ha influido, de forma diferente, en la dinámica del empleo manu-
facturero en las 32 entidades federativas del país. A partir de la información
censal y con la técnica de panel espacial, se estima una función de empleo.
Se muestra que tanto los factores endógenos como exógenos de las variables
incluidas en el modelo actúan como detonantes del crecimiento del empleo en
el sector manufacturero. Los resultados permiten confirmar que los factores
internos presentes en una entidad determinan la trayectoria de crecimiento
del empleo, pero también lo hacen los factores de las entidades vecinas, los
cuales son analizados mediante los impactos directos, indirectos y totales.
Los impactos de corto y largo alcance permiten mejorar las recomendaciones
de políticas económicas en el rubro del empleo como se muestra en el trabajo.
Palabras clave: Análisis económico, empleo, Econometría, México.
Clasificación JEL: C51,O4, R1, R12.
Introduction
Many of Mexico’s major current national problems derive from the varying
economic growth levels of each of its states. The growth rate since the early
eighties of the last century has been well below the historical norm for the
previous four decades (Ros, 2013). This has hindered job creation in the mag-
nitude demanded by the size of the economy and population growth.
In this sense, there has been a significant decrease in the secondary sector,
specifically in manufacturing as a generator of employment and a pole of
absorption of the labor force as it was in past decades (de Jesús and Carba-
jal, 2017). Now, the tertiary sector, also called the services sector, has been
strengthened as the main generator of employment and of the absorption of
the employed population, which in Kaldor’s (1984) view is known as the pre-
mature maturity of the manufacturing sector. This process itself would not
be a problem since productive tertiarization is a constant at the internatio-
nal level (see Pérez-Campusano et al., 2018, Maroto-Sánchez and Cuadrado-
Roura, 2009 and Cuadrado-Roura, 2016), except that the working conditions
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
93
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
of these sectors are not adequate. Labor precarization has increased due to
the fact that the highest growth has occurred in activities with low added
value and low productivity, as well as to labor market flexibility in the coun-
try (Andrés-Rosales, Czarneski and Mendoza, 2019).
The process of economic opening in Mexico that began in the mid-1980s has
accentuated the disparities between Mexico´s states and regions. Everything
indicates that even in this scenario of growth, states with activities predomi-
nantly in the secondary sector are those that have shown a greater dynamism
and growth due to their commercial links with the international market. In this
regard, Asuad and Quintana (2010) argue that regional inequality gap bet-
ween Mexican states has widened, at least for the years 1986-2008. Germán-
Soto and Escobedo (2011) coincide with this, they show that the states of the
center and north border are distancing themselves from the southern states,
mainly since the beginning of the commercial opening process. In addition to
the above, labor precariousness in the manufacturing sector is higher in the
country´s center and south of the country than it is in the north, as shown by
Andrés-Rosales et al. (2019).
In this context, the aim of this paper is to analyze the states’ endogenous
factors that influence the manufacturing employment, considering that large
companies are the main drivers of labor conditions in Mexico, not the micro,
small and medium enterprises (SMEs). In line with Andrés-Rosales, Quintana-
Romero, Namkwon and Mendoza-González (2019a), we find more micro and
SMEs spread out throughout national territory, while large companies are
concentrated in only a few states. Through a spatial panel, we measure the
impacts of these endogenous and exogenous factors. The hypothesis is that
the economic growth of the 32 states of Mexico has led to weaker employ-
ment endogeneity. In other words, factors endogenous to the territory are not
significant, but the exogenous variables generated within a state do influence
employment (endogeneity of the variable) as do the exogenous variables of
neighboring states (exogeneity of the variable).
The importance of employment is that it is “a key indicator of the performance
of an economy, not only because it is closely related to aggregated production
but also because it is a fundamental determinant of the level of well-being of
the population” (Mejía, Mejía and Rendón, 2017, p. 43).
Spatial analysis of manufacturing employment in Mexico, 1984-2013
94
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
The work is divided in three sections; The first section presents a review of
the literature that analyzes Mexico’s slow growth. Section two reviews the
performance of manufacturing in recent decades and places an emphasis on
employment in the states. The third section shows the results of a spatial panel
estimation for the manufacturing sector. Finally, in the last section, conclu-
sions are drawn.
1. The context of employment, manufacturing
and endogeneity in Mexico
In the mid-1980s, the Mexican economy began a process of economic libe-
ralization intensified with the beginning of the North American Free Trade
Agreement in 1994. It was believed that free trade would increase exports
and consolidate the long-term sustained increase in economic activity, with
positive effects on the country´s economic growth and job creation. Howe-
ver, results show that the balance of free trade, in terms of employment and
growth, was not as expected.
In fact, specialists agree that in recent years “manufacturing has had the worst
performance in terms of job creation: in 16 years -for the period 1994-2011-
it barely generated just over 500,000 jobs, at an average annual growth rate
of 1.0, while during 2000-2011 its average annual growth rate was negative
(-1.7%), which represents more than half a million jobs lost. No other sec-
tor has suffered such a poor performance in terms of employment” (Dussel
Peters, 2011, p. 20).
Sectors without links to foreign trade were expected to not generate emplo-
yment. Even though manufacturing is strongly linked to export dynamics, it
has not been able to significantly influence the process of job creation (Dus-
sel Peters, 2003a).
This behavior is largely due to the slow growth of the Mexican economy during
the last thirty-years (Calderón and Sánchez, 2012, Loría, 2009, Ros, 2013,
2010a, 2010b and 2008). The effects of this include the accumulation of unem-
ployment, idle productive capacity, and social backwardness. During the past
three decades, economic growth has suffered a severe slowdown compared
to the previous forty years. Between 1981 and 2005, Mexico’s GDP per capita
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
95
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
grew at an average annual rate of 0.5% compared to 3.2% in the 1940-1981
period (Ros, 2008). As stated by Ros (2013, 2010a and 2010b), this perfor-
mance is the result of the fall in income per capita from 1982 to 1989, a period
characterized by strong external shocks, acute macroeconomic instability and
a continuous transfer of resources abroad in a scenario of severe rationing
of external credit. Moreover, Ros (2013) reports that Mexico’s slow economic
growth is the result of the poor performance of the total factor productivity.
This aggravates the situation even more; if labor productivity does not grow,
salaries cannot follow. Munnell (1990, p. 3) argues: “productivity growth is
the major determinant of the future living standard. If the efficiency in the use
of resources rises at 2.5 percent per year, people can expect their real wages,
and their living standards to double every 28 years or roughly once a genera-
tion. In contrast, productivity growth of 0.5 percent means that children can
expect living standard only 15 percent higher than those of their parents.”
In such circumstances, the empirical evidence shows that the export-led growth
strategy has had favorable but limited effects on production, which have been
offset by imports of manufactures, displacing domestic production. It also
shows that the positive effect of exports on direct and indirect employment
is not as important as the effect of national production (Ruiz-Nápoles, 2004).
The above is remarkable given that manufacturing has historically been a ben-
chmark by which to understand the quality of employment in Mexico.This is
mainly due to the fact that economic orientation towards a model in which
exports increase in importance for global demand, had a significant effect
on employment, especially in the 1993-2000 period. A period in which trade
liberalization led to the restructuring of employment towards a greater con-
centration in the manufacturing sectors of unskilled labor in activities that do
not necessarily generate good quality jobs (Cervantes and Fujii, 2012).
The nineties experience provides evidence that the manufacturing sector has
not been an important generator of formal employment. This may be associa-
ted with the slow growth rate of the economy in recent decades, which has
been widely studied by various authors with different hypotheses, and from
different perspectives, ranging from the historical linkage of Mexico’s eco-
nomic growth to the productive performance of the United States, and the
evolution of terms of trade (Mejía-Reyes, Rendón-Rojas and Vergara-Gonzá-
Spatial analysis of manufacturing employment in Mexico, 1984-2013
96
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
lez, 2018, Guerrero, 2006). Other arguments state that the establishment of
rules of conduct and behavior of macroeconomic policy have helped to reduce
volatility and macroeconomic instability, but at the expense of lower growth
(Esquivel, 2010); or that the obstacles to growth are found in an environment
adverse to productive investment (Perrotini, 2004) and in the low investment
rate (Ros, 2008 and Ibarra, 2008).
International processes of global integration play an important role in the
growth of the Mexican economy, and specifically in the performance of the
manufacturing sector. In this sense, the competitiveness of the regions and
countries must be analyzed from a systemic competitiveness approach, in which
approaches related to the micro, meso, macro and metaeconomic levels are
fundamental as a whole to understand the new global trends followed since
the 1970s. Analysis from a purely microeconomic or macroeconomic pers-
pective is not enough (Dussel Peters, 2003d). We must look at the regions’
conditions of territorial endogeneity (Dussel Peters, 1997, 2000, 2002, 2003a,
2003b and 2003c, De Jesús-Almonte, 2019).
This research upholds within this context. The issue of territorial endogeneity
becomes relevant if it is understood as the possibility that based on their endo-
genous characteristics, regions have to take advantage of the virtues of the
global process that economies and territories experience, and of the benefits
of the market size that comes with the economic opening, favoring regions’
specialization and competitiveness. This refers to the degree of integration
of specific territories in the value chain based on their endogenous potential,
measured in terms of geographical conditions and the conditions of the fac-
tors of production (De Jesús-Almonte, 2019).
A proposed element of analysis to explain the differences in the growth dyna-
mics of economic sectors, regions, and countries is territorial endogeneity. For
Mexico and other Latin American countries, Dussel Peters (1997, 2000a and
2000b, 2002, 2003a, 2003b and 2003c), among others have conducted stu-
dies under this approach for manufacturing and maquila. These works argue
that an important condition in the systemic vision of competitiveness is the
degree of territorial endogeneity.
Finally, it can be argued that territorial endogeneity poses elements to analyze
the role played by the territory, understood as economic space, in the explana-
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
97
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
tion of growth. Hence, territorial endogeneity, seen as the degree of integration
of specific territories in the value chain, means much more. The endogenous
nature of territories can be expressed in many ways, ranging from the con-
ditions and provisions of human capital that strengthen the production pro-
cesses of the different manufacturing branches or divisions, to the possibility
of developing regional infrastructure using the same logic of strengthening
their productive capacity. These would make such territories susceptible to
developing efficient productive linkages that promote regional growth and,
in this case, growth in the manufacturing industry.
However, the Mexican economy has focused on the international market,
neglecting the local market, so territorial endogeneity, understood as the
local capacity to boost economic growth, employment and production, is weak
because domestic demand has contracted. This, in turn, influences the con-
traction of the domestic market. In this sense, considering the fact that only
20 percent of the world’s gross product is commercialized at an internatio-
nal scale (Zevallos, 2003), the domestic market remains vitally important for
domestic companies. In this regards, Quintana, Andrés and Mun (2013) argue
that exporting companies are important for employment, growth and the
country’s development; nonetheless, to rely only on export activity to sustain
growth, and development at a national scale can have serious consequences.
Based on this argument, this paper builds a model that includes proxy varia-
bles of territorial endogeneity of the states and regions of Mexico.
1.1. Theoretical arguments about the
determinants of employment
Keynes’s first criticism of the neoclassical model is precisely the existence
of the labor market. Keynes (1939) considers that employment is not deter-
mined in the labor market but in the goods market influenced by effective
demand. In addition, according to Keynes, workers are not hired based on the
real wage but on the nominal wage, hence, full employment is only a specific
point in the neoclassical analysis, and economies operate below that point of
full occupation. Some Keynesian assumptions are idle capacity of the labor
factor and of productive resources, and unemployment. Therefore, workers
have no control over the real salary they receive; they can only establish agre-
ements on nominal wages.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
98
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Within the Keynesian analysis, the salary is recognized as a function of the
marginal productivity of labor, as in the neoclassical approach. The core of
the theory proposed by Keynes is based on admitting the possibility that,
for different reasons, the community as a whole does not spend the entire
income. He also observes that when the level of occupation increases, so does
the community’s global income. However, consumption grows by a smaller
proportion due to psychological reasons. Hence, as employment increases, a
growing gap between consumption and income is generated, which needs to
be compensated with a greater volume of investments. It is only in this way
that the increase in occupation can be sustained (Keynes, 1939). As such,
one of his recommendations is precisely to encourage investment as a way
to reduce unemployment, and if there is a lack of private investment, public
investment must increase. This is contrary to the policies applied to macro-
economic stability in Mexico, where the free market forces have been left to
determine both employment and production.
The focus of this paper is based on Keynesian arguments of effective demand
to explain the determinants of employment. The fact that employment depends
on production is based on Keynesian approaches, in the sense that when a
company satisfies the demand for its product, it uses exactly the amount of
labor needed to produce the quantity demanded. If more work is needed to
produce a larger amount, companies must employ more workers when pro-
duction demand is higher (Abel and Bernanke, 2004).
The empirical evidence for different economies has provided elements to accept
this relationship (see, for example, Lanteri (2013) for the case of Argentina;
Dixon, Freebairn and Lim (2004) for the case of Australia; and Lechuga and
Varela (2001) for the case of Mexico). From this perspective, it is important
for an employment function to consider wages as one of its determinants. In
this regard, there are two theoretical arguments that explain this relations-
hip between employment and wages. First, the neoclassical argument consi-
ders that wages act as a cost variable with an inverse relationship, so that if
wages increase, employment tends to decrease. However, for some economies,
there is evidence that the relationship between manufacturing employment
and wages is positive. The theoretical explanation for this is based on the effi-
ciency wages of the New Keynesian Economy (NEK). The central hypothesis is
that although the payment of a higher salary generates higher costs for the
company, it also provides more benefits because of its positive effect on the
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
99
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
workers effort and, ultimately, on productivity. Efficiency wages would be the
amount that firms pay above the value of market equilibrium wage to avoid
falls in productivity (Malcomson, 1981 and Gordon, 1990).
Within post-Keynesianism developed by Lavoie and Stockhammer (2012), it
is considered that the wage decline plays a fundamental role in the unequal
growth of economies and polarization over the distribution of income. From
the authors’ point of view, the essential cause of the long-term evolution of
income distribution, and its increasing dispersion, is the change in economic
policies and the institutional and legal environment, which in recent decades
has been more favorable for capital and for high-level supervisors in general.
Thus, the authors Lavoie and Stockhammer (2012, p. 1) do not consider that
salary improvement is an endogenous phenomenon, but rather “the distri-
bution of income can be modified or influenced by appropriate government
policies that act as much in the distribution of primary income, for example,
reinforcing the power of negotiation of unions or ensuring low real interest
rates, as in the distribution of secondary income, modifying the tax legislation.”
The authors Lavoie and Stockhammer consider that it is time to change the
“distribution policies in favor of capital (pro-capital distribution)” to policies in
favor of a more equitable distribution of income through wage improvement
(pro-labor distribution), or what they call “a wage-based strategy” that could
generate more stable long-term growth than the currently prevailing model.
It stems that the wage share of workers will have negative consequences on
both the consumption of goods (which reduces effective demand), and on the
level of savings (Lavoie and Stockhammer, 2012). This leads us to reflect on
the structure of the domestic market within post-Keynesian analysis, and on
government policies that are of the utmost importance to enter the path of
equitable development.
In Mexico, the macroeconomic policies of stability have focused more on bene-
fiting national and international capital. Following Shaik (1983), with each
economic crisis, international investors place greater conditions on the govern-
ment, and have a higher level of negotiation to impose the salaries to be paid.
To overcome this problem, the authors believe that expanding the country’s
wage base will not only reactivate the domestic market, but it will also reduce
impacts of international shocks if the domestic market is strong. These argu-
ments are used in this paper to analyze employment and to highlight the rela-
Spatial analysis of manufacturing employment in Mexico, 1984-2013
100
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
tionship between production and real wages with employment. The analysis is
complemented by the inclusion of the weight of the endogenous conditions of
the territories to boost employment, as suggested by Jesús-Almonte (2019).
2. The manufacturing scenario in Mexico
The experience since the nineties reveals that the manufacturing sector has
not been an important generator of formal employment. Moreover, Mexican
manufacturing has been characterized in recent years as a sector that is expe-
lling work and where job creation has been reduced, even in those branches
where exports are growing significantly.
The explanation for this performance could be associated with a relatively high
capital intensity of manufacturing, and a relatively low absorption of emplo-
yment, particularly in the more modern and productive sectors of manufac-
turing (Cárdenas and Dussel Peters, 2007). It could also be associated with
the important changes that have occurred in the structure of the productive
sectors during the last decades; in contrast with the industrial and agricultu-
ral sectors, the service sector has become increasingly important (see Garza,
2008; González and López, 2007). In this regard, in 1980, the service sector
generated 56.3% of the national total GDP, while the secondary sector con-
tributed 39.5% and the manufacturing sector, specifically, contributed 18.1%,
while the primary sector only contributed 4.2%. However, by 2015, the con-
tribution of the tertiary sector to total GDP was 62.7%; the contribution of
the secondary sector was 34.2%; manufacturing contributed 17.3%; and the
primary sector, 3.2% (Carbajal and Carrillo, 2017).
On the employment side, in 1980 the tertiary sector employed 34.1% of the
total employed personnel, while the secondary sector did so with 29.1%, manu-
facturing participated with 18.0% and the primary sector with 36.7%. For 2015,
this data changed substantially, 61.4% of the total national employed popula-
tion is occupied in the tertiary sector, 24.3% in the secondary sector, 16.0% in
manufacturing, and 13.3% in primary activities (Carbajal and Carrillo, 2017).
In this context, the manufacturing situation is not encouraging since this sec-
tor has been considered the driving force of the economy, and its importance
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
101
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
lies not only in its direct contribution to output and employment (for Mexico,
manufacturing has contributed significantly to the growth of the economy
for a long time) but also in finding the indirect effects that these exogenous
variables have on the neighboring regions. Manufacturing also articulates, and
complements itself with other productive activities and, above all, as mentio-
ned by Sánchez (2012), it is a sector with increasing returns, a high income
elasticity of demand, and an important source of innovation and technologi-
cal diffusion. Manufacturing presents unique characteristics, such as its bac-
kward and forward linkages, its potential for the generation of added value
and employment, and its capacity for technological diffusion and productivity
growth (Dussel Peters, 1997). Thanks to its dynamism, its effects are trans-
mitted to other sectors of activity inducing innovative behavior in economic
agents (Garduño, 2009).
2.1. Growth and employment in the states of Mexico
In recent years, specifically from the beginning of NAFTA, gross value added
(VACB) and employed population in manufacturing show a tendency towards
concentration in production and employment growth in states that are more
closely linked to the external sector, specifically those located on the country’s
northern border. In order to highlight the importance of the northern border,
which has a geographical proximity to the United States of America (USA), the
32 states were grouped into 5 regions: Northern Border, North, Central West,
Central East, and South. According to the data on gross value added (VACB) and
employed population, in 1988, the Northern Border - which includes the states
of Baja California, Coahuila, Chihuahua, Nuevo Leon, Sonora and Tamaulipas -
generated 26.0% of total national manufacturing value added, and employed
27.7% of total employed population. By 2013, these variables have increased
significantly in the region, generating 35.6% of the VACB, and concentrating
35.9% of the employment generated by the sector in the country (see Table 1).
The North region (which groups Aguascalientes, Baja California Sur, Durango,
San Luis Potosi, Sinaloa and Zacatecas), provides the lowest contributions to
the country’s manufacturing product. In 1988, it generated 4.2% of the VACB
and employed 6.5% of the employed personnel. By 2013, its contribution to the
total manufacturing product grew to 6.7% and employed 7.8% of personnel.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
102
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Table 1. Manufacturing sector
Gross value added (VACB)* and employment by regions.
Percentage structure, 1988-2013
Employed population Gross value added
1988 1993 1998 2003 2008 2013 1988 1993 1998 2003 2008 2013
Northern
Border
27.7 29.8 34.3 35.3 33.9 35.9 26.0 23.8 30.8 33.1 32.9 35.6
North 6.5 7.1 6.9 7.3 7.3 7.8 4.2 5.1 6.6 6.2 6.7 6.7
Central
West
13.9 14.5 15.7 15.6 16.9 17.2 12.0 13.5 15.8 14.3 14.2 14.5
Central
East
43.0 39.4 34.7 32.7 31.9 29.6 49.3 47.0 39.4 35.3 34.1 30.4
South 9.0 9.1 8.5 9.0 9.9 9.5 8.5 10.7 7.4 11.1 12.1 12.7
* Constant prices, base year 2013.
Source: Authors based on INEGI (1989, 1994, 1999, 2004, 2009 and 2014).
In the case of the Central West region, production represented 12% of total
gross value added in 1988 and employed 13.9% of the total personnel emplo-
yed in manufacturing. By 2013, it showed a slight increase with 14.5% of VACB
and 17.2% of employed personnel.
The Central East region (Mexico City (formerly DF), Hidalgo, Mexico, Morelos,
Puebla, Queretaro and Tlaxcala) is noteworthy, as it has considerably reduced
its participation both in its contribution to added value and in manufacturing
employment. From 1988 to 2013, the region went from 49.3, and 43.0% to
30.4, and 29.6%, respectively (see Table 1). This is explained by the fact that
this region incorporates the State of Mexico and Mexico City, which for several
decades have been characterized by their growth dynamics in the manufactu-
ring sector, but in recent years have significantly decreased their dynamism.
In 1988, these two states contributed with 33.4% of the employed popula-
tion and with 38.2% of VACB generated of the total national manufacturing.
However, these data decreased substantially in 2013, when both states emplo-
yed only 17.6% of total employed population in the manufacturing sector and
contributed 18.3% of total VACB generated by this sector nationwide. Never-
theless, the manufacturing sector continues to be an important determinant
of the economic activities of these two states.
Table 2 shows the performance of the states led by the manufacturing sec-
tor. To highlight some important changes, the 1988-2013 period was divided
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
103
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Table 2. Manufacturing sector
Gross value added (VACB)* and employment by regions.
Annual average growth rate by periods, 1988-2013
Period 1988-2013 1988-1993 1993-2013 2008-2013
Region/State VACB Emp VACB Emp VACB Emp VACB Emp
Northern Border
Baja California 8.7 5.8 18.0 11.8 6.5 4.3 12.1 2.7
Coahuila 5.1 4.0 -8.1 2.5 8.6 4.4 12.4 7.6
Chihuahua 5.9 3.2 0.1 4.9 7.4 2.7 9.0 1.6
Nuevo León 6.8 2.9 -1.3 3.8 8.9 2.7 22.6 3.4
Sonora 9.9 4.0 7.0 5.2 10.7 3.8 22.9 1.8
Tamaulipas 8.1 3.5 5.7 7.4 8.6 2.5 18.1 -0.1
Region 7.4 3.9 3.6 5.9 8.5 3.4 16.2 2.8
North
Aguascalientes 9.0 3.8 13.6 6.6 7.9 3.1 8.6 3.5
Baja California
Sur 14.1 3.2 6.2 12.3 16.1 1.1 68.1 -0.6
Durango 7.7 2.0 2.9 1.9 9.0 2.0 21.9 4.2
Sinaloa 10.9 3.3 9.4 6.9 11.3 2.4 37.4 1.3
San Luis Potosí 6.5 3.6 1.6 4.2 7.7 3.4 15.8 2.9
Zacatecas 17.6 6.4 18.6 17.5 17.4 3.8 38.0 2.5
Region 11.0 3.7 8.7 8.2 11.6 2.6 31.6 2.3
Central West
Colima 12.9 4.9 5.8 9.7 14.7 3.7 58.9 3.5
Guanajuato 6.5 4.3 -4.0 5.4 9.3 4.1 17.5 5.0
Jalisco 6.7 3.3 8.2 4.6 6.3 2.9 17.5 0.6
Michoacán 7.5 2.6 5.6 2.9 8.0 2.5 16.1 -1.1
Nayarit 8.6 1.6 2.6 5.3 10.1 0.7 47.8 0.5
Region 8.4 3.3 3.7 5.6 9.7 2.8 31.6 1.7
Central East
Distrito Federal 8.7 -1.3 0.8 -0.3 10.8 -1.5 57.4 -2.3
Hidalgo 4.5 3.1 -0.8 5.0 5.9 2.6 8.9 1.0
México 3.4 1.4 1.1 1.8 4.0 1.2 10.0 0.1
Morelos 2.0 2.5 -5.2 5.2 3.9 1.8 18.5 0.7
Puebla 7.4 3.2 0.6 7.3 9.1 2.2 16.4 0.4
Querétaro 6.7 5.0 0.5 4.7 8.3 5.0 11.6 6.6
Tlaxcala 4.8 3.5 1.0 5.6 5.8 3.0 5.1 0.6
Region 5.4 2.5 -0.3 4.2 6.8 2.0 18.3 1.0
South
Campeche 30.5 4.2 3.0 9.6 38.4 2.9 225.2 -0.2
Chiapas 11.5 4.4 -6.2 6.3 16.4 3.9 29.6 3.2
Guerrero 12.3 5.8 11.1 12.9 12.6 4.0 65.9 0.2
Oaxaca 5.4 4.1 8.2 6.3 4.7 3.6 27.8 2.9
Quintana Roo 18.0 4.0 14.8 8.2 18.8 3.0 66.3 -1.5
Tabasco 16.3 2.5 2.3 3.4 20.1 2.3 63.0 2.0
Veracruz 7.3 0.6 7.0 -2.1 7.3 1.3 13.4 -0.9
Yucatán 9.3 4.2 3.9 11.0 10.7 2.5 35.4 -0.4
Region 13.8 3.7 5.5 7.0 16.1 2.9 65.8 0.7
National 7.7 2.6 1.7 3.8 9.2 2.4 27.1 0.3
* Constant prices, base year 2013.
Source: Authors based on INEGI (1989, 1994, 1999, 2004, 2009 and 2014).
Spatial analysis of manufacturing employment in Mexico, 1984-2013
104
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
into three sub periods: a) 1988-1993, the years prior to NAFTA, b) 1993-2013
to identify the evolution of the sector during the NAFTA; and c) 2008-2013,
which includes the recovery phase of the great recession that began in 2008,
also because from the first years of the 21st century the specialization of
manufacturing has intensified in export intensive and capital intensive acti-
vities, but was reduced by the recession in 2008-2009.
An important regularity in this period is the interruption of the integration to
NAFTA process, and the growing competition with Asia and China, specifica-
lly in manufacturing.
In this regard, it has been documented that NAFTA has gone through two
different stages (Dussel Peters and Gallagher, 2013). The first stage, from its
beginning in 1994 to 2000, was characterized by a process in which the region
was largely integrated, due to the increase in trade flows, investments, rules
of origin, and specific segments of industrial chains (such as the automotive-
auto chain and the yarn-textile-clothing chain). NAFTA was meeting expec-
tations when GDP, trade, investment, employment, and wages grew in the
region and intra-industry trade increased significantly. However, in the case
of Mexico, this integration did not translate into a broader process that pro-
moted forward and backward linkages (Dussel Peters and Gallagher, 2013),
which implied that some sectors were integrated within regions, but also that
sectors of other regions could be integrated into exporting dynamics. This
caused spillovers from one region to others, and the lagging regions could be
integrated into the dynamic companies within NAFTA.
From 2000 to date, NAFTA has been facing a process of deterioration in terms
of trade, investment and intra-industry trade. Apparently, Mexico is losing
ground to countries like China (Dussel Peters and Gallagher, 2013). Cárdenas
and Dussel Peters (2007) point out that since the 2000-2002 period, Mexico’s
foreign trade and, particularly, intra-industry trade have undergone signifi-
cant changes that, in the case of the United States, have been reflected in an
important process of commercial disintegration, both in aggregate terms and
in its main value chains.
An important feature that can be identified from the data in Table 2 is that
the average growth rates of total manufacturing for the 1988-1993 and 1993-
2013 periods report higher rates for added value, and lower rates for emplo-
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
105
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
yment. This would imply that there is growth in the sector but that it is not
reflected in an increase in employment since the treaty makes its production
intensive in capital and not in labor. This contrasts with the international trend
of transnational companies that decide to settle in the Mexican territory dri-
ven by low wage costs. We should observe an increase in employment in all
sectors, but instead, there is a decreasing trend in the potential or attracti-
veness of large companies.
Between 1988 and 1993, the average growth of VACB was 1.7%, while emplo-
yment grew 3.8%. For the 1993-2013 period, an important recovery was
reported: added value grew by an annual average of 9.2, reflecting an impor-
tant recovery in manufacturing production. However, it did not favor growth
in employment in the same proportion, resulting in a 2.4% annual average
growth, lower than the growth during the previous period.
The performance of manufacturing in the second period may be explained by
the recovery of the sector after the Great Recession in 2008-2009, which led
to significant growth in VACB (27.1% annual average) but was not reflec-
ted in similar magnitudes in employment (the average annual growth rate of
employment was 0.3%) (see Table 2).
The states that are supposed to be more closely linked to the world economy,
via trade, had greater resentment for the great crisis that began in 2008
(including all the states in the Northern Border region), but they are also the
ones that recovered the fastest. It has been documented that the recessions
of 1995, 2001-2002 and 2008-2009, seem to be increasingly more severe for
richer states compared to poorer ones, and that the states of the Northern
Border are increasingly less resistant to recessions (Mejía and Erquizio, 2012).
2.2. The structure of employment and production of manufacturing by state
Although the manufacturing sector in Mexico has been considered one of the
engines of economic growth and a bastion of industrial development in the
country, since the 1980s, it has undergone significant changes in its productive
structure. Several authors have discussed these changes; specifically, Mariña
(2005) argues that the opening process has not led to a substantial increase
in formal employment and better working conditions.
From the description and analysis of production and employment in manufac-
turing by states and regions, we have attempted to identify some regularity
Spatial analysis of manufacturing employment in Mexico, 1984-2013
106
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
in the evolution of both variables to provide evidence, and to understand the
dynamics of employment growth in the manufacturing sector in recent deca-
des, specifically in the 1989-2014 period. The statistical information reveals
that the difficulties in the manufacturing sector to make employment growth
respond to the slow growth of production. We aim to identify whether fac-
tors related to the endogenous conditions of states, in addition to the growth
dynamics of manufacturing production can be central factors in the determi-
nation of manufacturing employment, as referred by De Jesus-Almonte (2018),
and to identify possible spatial effects associated with neighboring states.
3. An estimate of manufacturing employment
in Mexico with spatial effects
3.1. Data description
The spatial panel was estimated using information from the Economic Censuses
of years 1989, 1994, 1999, 2004, 2009 and 2014, published by the National
Institute of Statistics and Geography (INEGI). Data of all years were consoli-
dated for the total manufacturing of Mexico’s 32 states. These are real data
at constant 2013 prices. Based on the definition of the variables, as indicated
below, the empirical employment model and the aggregate level for the total
Mexican manufacturing are estimated.3
3.2. Spatial effects model
From economic theory, we have a priori knowledge that the signs for the para-
meters to be estimated for total manufacturing were the following (the sign
+/- represents the expected sign effect for the coefficients of each explana-
tory variable):
+ - + + + + +
PO= f [c, VACBR, REMR, FBKFR, H, IE, IDE, MEXP] [1]
3 Manufacturing GDP aggregated deflator was used to convert gross censal added value to real values,
while the consumer national price index was used to deflate earnings. The reason for using the aggre-
gated deflator is that there is no availability of deflators for each state since 1989. The bias for this
form of deflation is minimal since differentials of state inflation are not considered.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
107
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
The definition of the variables are presented in the following table:
Dependent variable
PO Employed population for the total manufacturing and by state (employment).
Independent variables
VACBR Real gross census value added of manufacturing. A positive correlation with employment is
assumed due to the effects of aggregate demand.
REMR Real wages of manufacturing. A negative correlation with employment is assumed since
salaries act as a cost variable.
FBKFR Gross formation of fixed capital of manufacturing. Its correlation with employment is
assumed to be positive because of the beneficial effects that investment has on production
and employment.
HA proxy variable for human capital, is defined as the level of education of each state,
measured with population by the levels of qualification of the workforce as the proportion of
the economically active population (PEA) with technical training, high school or professional
studies. The assumption is that the skill levels of the labor force are decisive in explaining
regional differences and, consequently, employment levels. It is expected that the higher the
qualification of the workforce, the greater the possibility of generating jobs, thus a positive
correlation is expected.
IEaManufacturing specialization index for each state. The specialization pattern of a region will
make it possible to recognize the activities that lead the economic process, and to identify
which region would have comparative advantages. Hence, a positive correlation between
specialization and employment is assumed.
IDEbThe Economic Diversification Index represents an empirical measure of agglomeration of
economies to analyze the local economic structure. It makes it possible to measure the
type of industrial structure based on diversification or concentration (see Sobrino, 2003). A
positive correlation between IDE and employment is expected. The greater the diversification,
the higher the employment, especially due to the heterogeneity in the qualification of the
regions’ labor force.
MEXP Temporary imports for export. This variable captures the impact of the policy to boost exports
from favoring the importation of inputs for production, which in the long run, can become
an obstacle to growth.
a The specialization index or location quotient (Qij) is defined as in Boisier (1980): Q
V
V
V
V
ij
ij
e
ij
e
i
ij
n
i
ij
n
ii
=
where Vij
e is total gross production (PBT) of activityi in region j; Vij
e
i
å
is manufacturing PBT of region
j; Vij
n
i
å
is total gross production of activity i of national manufacturing, and Vij
n
ji
åå
stands for
national manufacturing PBT.
This index compares the relative share of regional activities to national level activities, which makes it
possible to measure the concentration of economic activities in the region, compared to those in the country
as a whole. Thus, we can determine inter-regional specialization based on the following criteria: if Qij = 1,
then the relative size of activity i in region j is identical to the relative size of the same sector throughout
Spatial analysis of manufacturing employment in Mexico, 1984-2013
108
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
the country. In this case, it cannot be said that there is a regional specialization in this activity. If Qij
in region j, then the relative size of activity i is lower than in the country; indicating that there is no
specialization. Finally, if Qij > 1 in region j, then the relative size of activity i is greater than in the country;
indicating regional specialization in this activity.
b The Economic Diversification Index (IDE) is formally expressed as follows (see Sobrino, 2003):
IDEn
n
p
pn
i
t
=
1
1
22
, where n is the number of sectors, groups or branches of economic activity, pi
is production of activity i, and pt is total local production. The IDE is a positive value between 0 and 1: values
closer to zero indicate greater diversification of the activity, and values closer to one indicate concentration.
Variables H, IE, and IDE are proposed as proxies of territorial endogeneity
conditions of the states and regions of Mexico, as suggested by De Jesús-
Almonte (2019), because they are variables that allow us to consider factors
that are apparently not tangible but that are influential in a region’s develo-
pment process.
The variable MEXP (imports for export) for each Mexican state is built from
official information from the National Institute of Statistics and Geography
(INEGI). This variable is part of the Statistics on Manufacturing and Export
Maquiladora. The indicator of Total imported input by manufacturing esta-
blishments in the IMMEX program is in thousands of pesos. The inclusion of
MEXP seeks to provide empirical evidence on the effect that export promo-
tion programs have on job creation. The impact of such programs is central
to the analysis because manufacturing in Mexico has not achieved a pro-
cess of integration with the rest of the productive system, since, due to its
production structure, it requires an increase in imports in order to export. In
this context, the great importance of temporary imports for export, at least
nationally, is noteworthy. Some authors agree that as of 1993, about 80%
of Mexican exports depended on temporary importation processes for export,
which expresses their low level of local integration (Dussel Peters, 2009 and
Capdevielle, 2007 and 2005)4.
The idea that the more efficient a territory is, the greater the possibilities for
investment and production flows to converge to it, would affect not only the
growth of economic activity, but also the creation of new formal jobs, and even
4 The IMMEX Program seeks to strengthen the competitiveness of the Mexican export sector by providing
the possibility of temporarily importing -free of import taxes and VAT- the goods necessary to be used
in an industrial or service process for elaboration, transformation or repair of goods of foreign origin
temporarily imported for export or to provide export services (SE, 2014).
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
109
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
of higher added value because the structure of its population, the infrastruc-
ture, and its human capital enables innovation under the assumptions of the
endogenous growth model. Hence, the regions’ endogenous growth requires the
promotion and development of comparative advantages and competitiveness.
As argued previously, territorial endogeneity must be seen as the degree of
integration of specific territories in the value chain. That is, the endogenous
nature of the territories can be expressed from the conditions, and dispositions
of human capital, and from the possibility of developing regional infrastructure
to strengthen the production processes of the different branches or divisions
of manufacturing, and the regional productive capacity. This development will
make regions potentially susceptible spaces to foster efficient linkages that
boost the growth of the region and, in our case, of the manufacturing industry
(see Dussel Peters, 2007 and 2004 and De Jesús-Almonte, 2019).
The model in equation [1], is expressed as a linear regression model with panel
data of the use of manufacturing for each pair of state-year in equation [2].
Variables with the prefix L, indicate logarithmic transformation.
LPOLVACBRLREMR LFBKFR LH
it it it it it,,
,,
,
=+ ++
++
ββ βββ
01 234
ββ βε
56 7
IE IDEMEXP
it it it it,, ,,
++ + [2]
For i = 1,…, 32 states of Mexico, and t = 1989, 1994, 1999, 2004, 2009 y 2014,
years of analysis.
The variables that measure the impact of territorial endogeneity on employ-
ment are LH, IE, and IDE.
Although there is important evidence in the literature to explain the deter-
minants of employment,5 only a few studies have measured or provided any
empirical regularity regarding the weight of the endogenous conditions of the
territories to boost employment. The notion of territorial endogeneity is assu-
med in a strictly functional sense of the territories, so that production rises in
the value chain from its endogeneity conditions and generates positive effects
on the creation of employment.
5 In applied literature, the problem of manufacturing employment is usually explained from traditional
functions that include only manufacturing GDP, productivity of the manufacturing sector, and gross
fixed capital formation as explanatory variables (see, for example, Lechuga and Varela, 2001).
Spatial analysis of manufacturing employment in Mexico, 1984-2013
110
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Initially, the estimation was made using panel data because, following Wool-
drige (2002), the technique makes it possible to capture unobservable hete-
rogeneity, either between economic agents or over time. This technique offers
several advantages in that given its ability to control heterogeneity, it reduces
multicollinearity and provides more degrees of freedom (Baltagi, 2005). Howe-
ver, as Anselin (1988) observes, some cross-sectional variables can present
spatial autocorrelation, hence the decision was made to spatialize the tradi-
tional panel model –as suggested by Anselin (1988), Lesage and Pace (2009),
Giuseppe (2014) and Elhorst (2014)–, which leads to equation [3]:
LPOwLPOXwX
id tist,,
ε++ + [3]
where the error term is as stated in equation [4]:
ελ
tt
wu u
=+
[4]
The only difference with previous approaches is q, which captures the spatial
behavior of the exogenous variables, r the spatial autoregressive term, and
the scalar l that captures local spatial autocorrelation.
Equations 3 and 4 incorporate a spatial weights matrix (W), which is a positive
square matrix with its dimension dependent on the size of the data sample
that describes the grouping of possible interactions between the spatial units.
Elements wij are space weights, which equal zero when there is no neighbor-
hood. The elements of the main diagonal are zero because the possibility of
self-proximity is excluded. Finally, the contiguity matrix was standardized,
implying that the sum of each row equals one and represents a smoothing of
the neighbors’ impacts.
In previous equations, wYt and wXt represent spatial weights of dependent
and independent variables respectively, and following authors such as Para-
july and Haynes (2017), the values of r, l and q can generate the following
models that we use in this research:
If l = 0 and q = 0 then a spatial autoregressive model (SAR) is generated.
If l = 0 then a spatial Durbin model (SDM) is generated.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
111
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
The spatial Durbin model (SDM) makes it possible to determine the influence
of the exogenous variables generated within a region, known as direct impacts,
while the indirect impacts measure the influence of the exogenous variables
of the neighbors on the endogenous variable of a given region. With the direct
and indirect impacts, as with the totals, the “circular feedback process” can
be obtained in the spatial estimates (Anselin, 1988, Lesage and Pace, 2009,
Giuseppe, 2014, Quintana and Andrés-Rosales, 2014; Fisher and Getis, 2010).
3.3. Results and discussion
In panel data regressions, the fixed-effect model is generally preferred when
working with cross-sectional data, because generally short-term rather than
long-term information is available. However, the decision on this type of models
is based on the Hausman test. The fixed-effect model is used if the p-value
of this test is less than 5%.
The best model should be selected among spatial panel models with fixed
effects. There are several criteria that authors like Andrés-Rosales et al.
(2018) and Parajuli and Haynes (2017) recommend, including the following:
a) cross-sectional estimates of the different periods focusing on the values of
the Lagrange Multiplier and its robustness, as well as the special lag with its
robustness, as suggested by Anselin (1988), Moreno and Vayá (2000); b) con-
sidering the statistical significance of the coefficients of the indirect impacts
measured by q, as well as the statistical significance of the spatial autocorre-
lation measures (r and l); and finally c) the different types of models such
as the spatial autoregressive (SAR) and the Durbin spatial model (SDM) are
calibrated using the log-likelihood statistics, the Akaike Criterion (AIC) and
the Bayesian Information Criterion (BIC). The Durbin Spatial Model (SDM) was
chosen in both cases, according to the criteria of the aforementioned statistics
and because the values of Rho (r) are greater and more significant.
The main findings of the estimates are reported in Table 3. Columns (a) and (b)
are estimates without temporary imports for export (LMEXP), while columns
(c) and (d) include LMEXP. This criterion seeks to evaluate whether export pro-
motion programs are factors that significantly influence job generation in the
context of the country’s commercial opening.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
112
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Due to methodological issues, it is important to clarify that direct impacts
are variables that are exogenous to the model, but are generated within each
state; that is, they are endogenous to that state. While indirect impacts are
variables that are exogenous to the model, but are not determined within a
region; instead, they are the exogenous variables of neighboring states, but
have an effect on the employment of a particular region. On the other hand,
total impacts capture the sum of the direct and indirect impacts of the exo-
genous variables.
Direct, but endogenous impacts to the state
The results of the estimation showed the following (see Table 3):
a). Remuneration (LREM), sector production (LVACB), gross capital formation
(LFBK), human capital index (LH), and economic diversification (IDE) have
a positive and significant direct impact on job creation. Although in the
case of capital formation, and productive diversification this impact is not
important, given that the value of the coefficients is very small.
b). The most important impacts are generated by the sector’s wages, and by
production growth. Here, it is striking that the impact of wages is positive
and significant, as this goes against the neoclassical approaches in the
sense that the higher the remuneration, the lower the demand for workers,
but what is observed here is that labor demand in the different states has
increased along with the increase in wages.
c). Productivity has a negative impact on the increase of employment in the
manufacturing sector, since fewer workers are required when there is an
increase in productivity; that is, more goods are produced with fewer but
more productive workers. This is a trend observed in the Mexican economy,
which means that companies have increased their production but have
not influenced the demand for workers in the same proportion, probably
because large companies are more capital-intensive than labor-intensive.
d). Human capital (LH), the specialization index (IE) and the diversification
index (IDE) measure the importance of territorial endogeneity. We find
that human capital has positively, and significantly influenced job creation
(coefficients with values of 0.08, and 0.07, respectively), while manufac-
turing specialization has done so negatively, which means that a greater
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
113
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
specialization in the sector tends to hinder job creation. The rationale
behind this is that as the sector in the different Mexican regions specia-
lizes, it does so with greater technological content, and not so much by
increasing its use of employed personnel.
e). Temporary imports for export (LMEXP) have not influenced employment
growth. This means that sectors linked to the export sector do not have
much influence on the manufacturing employment of the different Mexican
regions. These sectors are more technified than traditional sectors and
than micro, small and medium enterprises.
Table 3. Estimation of the spatial panel in the manufacturing sector, 1989-2014
a) SAR b)SDM c) SAR d) SDM
LREMAN -0.03 (-0.76) 0.22 *** (3.95) -0.02 (-0.5) 0.23 *** (3.5)
LVACBMAN 0.06 (1.22) 0.18 *** (3.9) 0.04 (0.88) 0.17 *** (4.52)
LFBKMAN 0.00 (0.68) 0.01 ** (1.92) 0.00 (0.56) 0.01* (1.81)
IPRODMAN -0.18*** (-2.82) -0.18 *** (-3.88) -0.15 *** (-2.61) -0.17 *** (-4.66)
LH 0.17*** (5.59) 0.08* (1.66) 0.15 *** (5.26) 0.07* (1.72)
IE -0.10*** (-4.64) -0.07*** (-5.26) -0.1*** (-4.68) -0.07*** (-4.26)
IDE 0.02 *** (5.52) 0.01 *** (6.41) 0.02 *** (5.79) 0.01 *** (4.13)
LMEXP 0.00 ** (2.14) 0.001 (0.75)
W*LREMAN -0.27 *** (-3.95) -0.23 *** (-3.5)
W*LVACBMAN -0.08 (-1.35) -0.15** (-2.05)
W*LFBKMAN -0.02 (-1.23) -0.01 (-0.94)
W*IPRODMAN -0.03 (-0.39) 0.00 (0.04)
W*LH 0.00 (0.00) 0.00 (0.16)
W*IE 0.00 (-0.33) 0.00 (-0.06)
W*IDE 0.00 (0.48) 0.00 (0.41)
W*LMEXP 0.009** (2.29)
Rho (r) 0.55*** (8.98) 0.68*** (12.26) 0.54*** (9.5) 0.66*** (13.18)
Hausman test 29.3 *** 24.05 * 42.28 *** 33.02 **
AIC -169.49 -255.96 -173.61 -258.25
BIC -140.17 -203.84 -141.03 -199.62
Log-pseudo likelihood 93.74 143.98 -1.36 -1.36
***, **, * are significant at 1%, 5%, and 10% levels, respectively.
T-stats are shown in parentheses.
Source: Authors based on INEGI (1989, 1994, 1999, 2004, 2009 and 2014).
Spatial analysis of manufacturing employment in Mexico, 1984-2013
114
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
f). The values of r are positive and significant, which suggests that manu-
facturing employment is spatially endogenous with respect to its neigh-
bors. This variable measures global spatial autocorrelation generated by
employment, which implies that as employment increases in neighboring
regions, a particular region can potentially increase its employment. In
economic terms, if a particular state attracts manufacturing companies, the
employment that a company will offer will not only benefit the workers of
that state but also the neighboring states. This is very recurrent in states
such as Mexico City, State of Mexico, Puebla, and Tlaxcala, among others.
Finally, according to Andrés-Rosales et al. (2019a), between 1999 and 2014,
employment growth rates for large enterprises (LEs) increased by 0.73%, and
employment in small and medium-sized enterprises (SMEs) increased by 0.95%.
They show that the share of economic units engaged in the service sector rela-
tive to the national total was 36.7%, and observed a similar percentage in
the employed population. In the trading sector, these values were 49.9% and
30.5%, respectively. Meanwhile, the manufacturing sector share is 11.7% of
the economic units, and 23.2% of national employment. Also, only in the manu-
facturing sector of all economic units in the sector, microenterprises (MEs)
account for 92.5%, SMEs 6.7%, and LEs 0.7%. The total employed population
in manufacturing is distributed as follows: 23.2% in microenterprises, 27.1%
in SMEs, and 49.7% in LEs. Regarding total gross production, manufacturing
LEs represent 77.1% of the output, SMEs 20.6%, and MEs 2.4%. Large com-
panies are determinants of production, employment and wages.
Direct, indirect and total impacts
It is noteworthy that the estimated SDM (spatial Durbin) models are not inter-
preted as partial derivatives as in the case of the traditional regression (see
Lesage and Pace, 2009, and Parajuli and Haynes, 2017). In this regard, Lesage
and Pace (2009) mention that a simple change in a region associated with any
of the explanatory variables could directly affect a particular region (direct
impact) and potentially affect other regions indirectly (indirect impact), which
is one of the benefits of using spatial methodology.
In this sense, one of the advantages of spatial panel models is that we can
obtain the direct impacts that influence employment in a specific region,
which are endogenous to the same region where they are generated, but we
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
115
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
can also obtain the indirect impacts that the neighboring regions generate
for a specific region. In our case study, this implies that salaries paid within a
region, as well as economic growth in the manufacturing sector (VACBMAN),
the gross formation of capital, and human capital (H) have positive and sig-
nificant impacts in the short and long term, while productivity (IPRODMAN)
and specialization (SDI) have a negative impact on employment growth in the
short and long term (see Table 4).
Regarding indirect impacts generated in the manufacturing sector, it is repor-
ted that the wage paid in the neighboring regions has a negative effect on
employment growth in a certain region, given the negative coefficient obtai-
ned in columns 1 and 2 (-0.31 and -0.52). On the other hand, gross capital
formation in neighboring regions tends to affect employment positively in a
certain region, given the coefficient obtained in the short and long term (0.05
and 0.08). Hence, the direct impact of remunerations on employment is posi-
tive, while the indirect impact is negative.
The coefficient of exports is significant and positive for employment deter-
mination from total impact, because the indirect impact of this variable is
greater than the direct impact. This means that an increase in exports does
not unfold as employment in the region where this sector is located, but if
neighbors with these sectors increase their exports, they can influence emplo-
yment in the given region.
The total impact of wages on employment growth in the manufacturing sec-
tor is negative (-0.25 and -0.41), in contrast with the direct impacts descri-
bed above. This means that an increase in salaries within a region has positive
impacts on employment in the same region, but a negative impacts one on
its neighbors.
Conclusions
One of the consequences of the breakup of productive linkages in different
regions of the country is precisely that the high-tech sectors or exporters do
not influence or determine employment in the regions where they have been
concentrated. This limits their growth and also restricts their integration. In
this way, if a state does not obtain great benefits due to the dynamism of
Spatial analysis of manufacturing employment in Mexico, 1984-2013
116
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Table 4. Direct, indirect and total short and long-term impacts of the manufacturing
sector
Manufacturing Manufacturing
Without LMEXP-SDM With LMEXP-SDM
SR_Direct LR_Direct SR_Direct LR_Direct
LREMAN 0.05 (1.4) 0.10 (1.54) 0.14*** (3.25) 0.14*** (3.05)
LVACBMAN 0.11*** (3.28) 0.2*** (3.23) 0.12*** (3.47) 0.12*** (3.28)
LFBKFMAN 0.02** (2.28) 0.03** (2.19) 0.02*** (2.69) 0.02*** (2.73)
IPRODMAN -0.16*** (-4.88) -0.28*** (-4.85) -0.14*** (-3.94) -0.14*** (-3.87)
LH 0.08** (2.19) 0.15** (2.18) 0.07* (1.74) 0.07* (1.83)
IE -0.05*** (-3.05) -0.08*** (-3.07) -0.04*** (-2.5) -0.04*** (-2.45)
IDE 0.00*** (2.42) 0.01*** (2.45) 0.01*** (2.96) 0.01*** (2.89)
LMEXPMAN 0.00** (2.18) 0.00*** (2.47)
SR_Indirect LR_Indirect SR_Indirect LR_Indirect
LREMAN -0.31*** (-3.19) -0.52*** (-3.08) -0.22*** (-2.46) -0.23** (-2.14)
LVACBMAN 0.09 (1.01) 0.14 (0.86) -0.12 (-1.32) -0.12 (-1.08)
LFBKFMAN 0.05* (1.82) 0.08* (1.68) 0.03 (1.42) 0.04 (1.54)
IPRODMAN -0.13 (-1.58) -0.19 (-1.28) 0 (-0.03) -0.02 (-0.31)
LHD1 -0.01 (-0.23) -0.04 (-0.35) 0.1 (1.5) 0.13* (1.63)
IED1 0.01 (0.38) 0.03 (0.49) 0.01 (0.4) 0 (0.2)
IDED1 0 (-0.38) 0 (-0.47) 0 (-0.1) 0 (0.12)
LMEXPMAN 0.02*** (4.06) 0.02*** (3.94)
SR_Total LR_Total SR_Total LR_Total
LREMAN -0.25** (-2.24) -0.41** (-2.11) -0.07 (-0.67) -0.08 (-0.66)
LVACBMAN 0.21** (1.92) 0.34* (1.81) 0.00 (0.02) 0.00 (0.01)
LFBKFMAN 0.07** (2.16) 0.11** (2.04) 0.06** (2.08) 0.07** (2.06)
IPRODMAN -0.29*** (-2.91) -0.47*** (-2.67) -0.14 (-1.53) -0.17 (-1.52)
LHD1 0.07 (0.82) 0.11 (0.79) 0.17** (2.3) 0.21** (2.24)
IED1 -0.03 (-0.76) -0.05 (-0.73) -0.02 (-0.72) -0.03 (-0.71)
IDED1 0 (0.5) 0 (0.48) 0.01 (1.07) 0.01 (1.06)
LMEXPMAN 0.02*** (4.27) 0.02*** (4.08)
***, **, * are significant at 1%, 5%, and 10% levels, respectively.
T-stats are shown in parentheses.
Source: Authors based on INEGI (1989, 1994, 1999, 2004, 2009 and 2014).
exports, a state with export sectors in the neighboring region has even less
opportunities to benefit. What we find is that factors that could boost endo-
geneity in the states are very weak.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
117
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
Among the benefits of investment are that it generates a boost in economic
activity and, consequently, in job creation. In this case, the result for the LFBK-
MAN coefficient shows a limited effect on employment for total manufactu-
ring, although positive and significant, given a coefficient of 0.02 and 0.03 in
the short and long terms.
In relation to qualification levels of available labor (LH), it is assumed that they
correspond to development levels in each state, so that education levels are
associated with regional endogenous conditions, from income levels of fami-
lies to infrastructure conditions and the role of governments.
Specialization and diversification in manufacturing are important because
they reflect the maturity of the sector to a certain degree. In this sense, the
results of the specialization index (IE), and the economic diversification index
(IDE) show that the levels of specialization and diversification have not been a
sufficient condition to generate growth effects in manufacturing employment.
Given the slow growth observed in the macroeconomic environment of the
Mexican economy, added to the renegotiation of the North American Free
Trade Agreement (NAFTA), we find that, during the period of study, wages
have positively influenced employment in the sector. Now, with the new con-
ditions of the trade agreement and with the new government, we hope that
the salary increase will also lead to the creation of new jobs. This would imply
that the internal market would improve as workers’ income and consumption
would increase. The only risk we perceive is that various studies, along with
the empirical evidence, show that Mexico imports more goods than it produ-
ces. In addition, exports have a large component of imports, since there are no
solid productive linkages. Under these circumstances, the increase in wages can
lead in the long term to an economic imbalance given that the high rates of
growth in wages and employment will lead to demand for either more impor-
ted goods or domestic goods with a high content of imported inputs, that will
eventually detonate economic crises.
We conclude that the international trend, in which Mexico has been inser-
ted, has shown that worker protection has declined, and the flexibility of
the labor market has led to greater job precariousness. This implies that the
employment and remunerations generated are not enough. Hence, as men-
tioned by Lavoie and Stockhammer (2012), the government must modify its
policies and its growth model to one that is pro-labor or pro-capital policy, or
Spatial analysis of manufacturing employment in Mexico, 1984-2013
118
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
in their words, a “wage-led and profit-led economic regime,” which is actua-
lly more appropriate.
It has been shown that good labor relations improve the propensity of wor-
kers to contribute to the production process. Naastepad (2006) found that a
1% increase in real wages leads to a 0.52% increase in labor productivity in
the Netherlands. Storm and Naastepad (2009), and Vergeer and Kleinknecht
(2010-11) investigated OECD countries from 1960-1984 until 2004. They found
that regulated and robust labor market institutions led to faster growth in
productivity. Hein and Tarassow (2010) investigated the link between income
distribution and productivity growth in six OECD countries and reached the
same conclusions.
Thanks
We appreciate the comments and recommendations of the Journal’s anony-
mous referees of, who contributed greatly to improve our work.
This research is part of the project Wage Inequality in manufacturing. An
analysis for the regions of Mexico, 1994-2014. Registration 4565/2018 / CIV,
of the Secretariat of Research and Advanced Studies, Autonomous University
of the State of Mexico.
Bibliography
1. Abel, A. B. and Bernanke, B. S. (2004). Macroeconomía. (4a ed.). España:
Pearson-Addison Wesley.
2. Alcaraz, C. and García, R. (2006). Cambios en la composición del empleo
y evolución de la productividad del trabajo en el sector formal de la
economía mexicana: 2000-2005, Documentos de Investigación del Banco
de México, núm. 2006-3, Banco de México, México.
3. Andrés-Rosales, R., Bustamante L., C. and Ramírez A., G.S. (2018). Social
exclusion and economic growth in the Mexican regions: A spatial
approach, Journal of Regional Research, 40, 57-78.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
119
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
4. Andrés-Rosales, R., Czarnecki, L. and Mendoza-González, M.A. (2019).
Spatial Analysis of labour market inequality by gender in Mexico, 2005-
2018, en dictamen en The Journal of Chinese Sociology.
5. Andrés-Rosales, R., Quintana-Romero, L, Namkwon, M. and Mendoza-
González, M.A. (2019a). SMEs´productivity in the Mexican regions: A
spatial panel data approach, 1999-2014, en dictamen en Contaduría
y Administración.
6. Anselin, L. (1988). Spatial econometrics methods and models. California,
Estados Unidos: Kluwer Academic Publishers.
7. Arriaga, R., Leyva, E. and Estrada, J. L. (2005). Perfil y estructura industrial
de Guanajuato y Querétaro: un análisis de la producción, el empleo y
los salarios, Análisis Económico, 20 (044), 135-189.
8. Asuad Sanén, N. and Quintana Romero, L. (2010). Crecimiento econó-
mico, convergencia y concentración económica espacial en las enti-
dades federativas de México 1970-2008, Investigaciones Regionales.
18, 83-106.
9. Baltagi, B.H. (2005). Econometrics analysis of panel data, West Sussex,
UK: John Wiley & Sons.
10. Belotti, F., Hughes, G., and Piano, A. (2016). Spatial panel data models
using state, CEIS, Research Papers, 14(5), 1-40.
11. Boisier, Sergio (1980). Técnicas de análisis regional con información
limitada. Serie II, 27, Cuaderno ILPES, Santiago de Chile, Chile: ILPES.
12. Calderón, C. and Martínez, G. (2005). La ley de Verdoom y la industria
manufacturera regional en México en la era del TLCAN, Frontera Norte,
17(34).
13. Calderón, C. and Sánchez, I. (2012). Crecimiento económico y política
industrial en México, Revista Problemas del Desarrollo, 170(43).
Spatial analysis of manufacturing employment in Mexico, 1984-2013
120
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
14. Capdevielle, M., Molina, T. and Rosado, G. (1999). Especialización produc-
tiva y comercio internacional en la industria manufacturera mexicana.
En F. Novelo y J. Flores (Coords.). Globalización, estado y actores sociales
en México. México: Universidad Autónoma Metropolitana Xochimilco.
15. Capdevielle, M. (2005). Globalización, especialización y heterogeneidad
estructural en México, en M. Cimoli (ed.). Heterogeneidad estructural,
asimetrías tecnológicas y crecimiento en América Latina. CEPAL-BID,
Santiago de Chile. 101-126. Recuperado de http://www.cepal.org/
publicaciones/xml/0/27320/LCW35e.pdf
16. Capdevielle, M. (2007). Heterogeneidad estructural y globalización de
la producción, Economía Informa, 345.
17. Calva, José Luis (2001). México más allá del neoliberalismo: Opciones
dentro del cambio global. México: Plaza y Janés.
18. Calva, José Luis (2007). Hacia una visión integral de estabilidad macro-
económica. En José Luis Calva (coordinador). Macroeconomía del creci-
miento sostenido (pp. 23-48). México: Porrúa-UNAM.
19. Carbajal, Y. and Carrillo, B. (2017). Producción y empleo manufacturero
en las entidades federativas de México, 1988-2013. En Yolanda Carbajal
y Leobardo de Jesús (Coord). Dinámica del empleo y la producción manu-
facturera en México (pp. 10-41). Toluca, México:EON-UAEMéx.
20. Cárdenas Castro, H. and Dussel Peters, E. (2007). México y China en
la cadena hilo-textil-confección en el mercado de Estados Unidos,
Comercio Exterior, 57(7).
21. Cervantes, R. and Fujii, G. (2012). The Mexican trade liberalization
process and its net effects on employment: 1988-2004, EconoQuantum.
Revista de Economía y Negocios, 9(2), 81-97.
22. Cuadrado-Roura, Juan R. (2016). Service industries and regional
analysis: New directions and challenges, Journal of Regional Research,
36, 107-127.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
121
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
23. De Jesús-Almonte, L. and Carbajal Suárez, Y. (2017). Empleo en el sector
terciario. Una estimación espacial para los municipios de la región centro
de México, 1999-2009, Región y Sociedad, 29(68), 76-114. https://dx.doi.
org/10.22198/rys.2017.68.a208
24. De Jesús-Almonte, L. (2019). Lento crecimiento y empleo manufacturero
en México: Un análisis de endogeneidad territorial. México: Eón editores-
UAEM. En prensa.
25. De León, A. (2002). El Tratado de Libre Comercio en América del Norte y
el crecimiento económico en las manufacturas mexicanas: una perspec-
tiva regional. México: Universidad de Guadalajara, división de Gestión
Empresarial.
26. Dixon, R., Freebairn, J. and Lim, G. C. (2004). An Employment Equation
for Australia: 1966-2001. Research Paper. Number 892. Department
of Economics, University of Melbourne. Recuperado de https://www.
researchgate.net/publication/5177684.
27. Dussel Peters, E. (1995). El reto del empleo en México. Cambio estruc-
tural en el empleo durante 1982-1992, Investigación Económica,
LV(212),135-176.
28. Dussel Peters, E. (1997). Economía de la polarización: Teoría y evolución
del cambio estructural de las manufacturas mexicanas (1988–1996).
JUS-UNAM, México.
29. Dussel Peters, E. (2000). Condiciones y retos de la industria electrónica
en Jalisco, El Mercado de Valores. Octubre.
30. Dussel Peters, E. (2002). La estrategia de desarrollo mexicana desde
los noventa: ¿un modelo a seguir para el resto de América Latina?, en
C. Maggi y D. Messner (eds.). Modelos de desarrollo nacionales y desa-
fíos de la globalización: Los casos de Chile, Cuba y México. Institut für
Entwicklung und Frieden der Gerhard-Mercator-Universität Duisburg.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
122
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
31. Dussel Peters, E. (2003a). Perspectivas y retos de la competitividad en
México. México: Centro de Desarrollo Empresarial, Facultad de Economía,
UNAM, Canacintra.
32. Dussel Peters, E. (2003b). Endogeneidad territorial y polarización, en
M. E. Negrete Salas; S. Levi y J. Page (coords.). Entre fenómenos físicos
y humanos. México: El Colegio de México.
33. Dussel Peters, E. (2003c). Ser maquila o no ser maquila, ¿es ésa la
pregunta?, Comercio Exterior, 53(4).
34. Dussel Peters, E. (2003d). The Maquila Export Industry and Productive
Integration, Voices of Mexico, 62, 31-35.
35. Dussel Peters, E. (2008). GCCs and Development: A Conceptual and
Empirical Review, Competition & Charge, 12(1).
36. Dussel Peters, E. (2009). Crisis de la industrialización orientada hacia
las exportaciones en México. Condiciones y opciones. En A. Oropeza
García (coord.). El comercio exterior aduanal y la gestión aduanal en el
siglo XXI (pp. 61-96). México: AAADAM-UNAM.
37. Dussel Peters, E. (2011). La manufactura en México: condiciones y
propuestas en el corto, mediano y largo plazo, en J. L. Calva (editor).
Nueva política de industrialización. Vol. 7 de Análisis Estratégico para
el Desarrollo. México: Juan Pablos Editor.
38. Dussel Peters, E. (2013). Retos para una política de competitividad
industrial en México: prioridades e instrumentos. En A. Oropeza (coord.).
México frente a la tercera revolución industrial: Cómo relanzar el proyecto
industrial de México en el siglo XXI. México: Instituto de Investigaciones
Jurídicas.
39. Dussel Peters, E. and Gallagher, K. P. (2013). El huésped no invitado del
TLCAN: China y la desintegración del comercio en América del Norte,
Revista CEPAL, 110 .
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
123
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
40. Dussel Peters, E. and Ortiz Velásquez, S. (2013). Monitor de la manu-
factura mexicana, 9(10) UNAM. México. Recuperado de http://www.
economia.unam.mx/cechimex/
41. Dussel Peters, E. and Ortiz Velásquez, S. (2015). Monitor de la manu-
factura mexicana, 10(11). Recuperado de http://www.economia.unam.
mx/cechimex/
42. Elhorst, J.P. (2014). Spatial Econometrics: From cross-sectional data to
spatial panels. The Newtherlands: Springer.
43. Esquivel, G. (1999). Convergencia regional en México, 1940-1995, El
Trimestre Económico, LXVI(264).
44. Esquivel, G. (2010). De la inestabilidad macroeconómica al estanca-
miento estabilizador: El papel del diseño y la conducción de la política
económica. En N. Lustig, IX Crecimiento económico y equidad. México:
El Colegio de México.
45. Fisher, M. M. and Getis, A. (2010). Handbook of applied spatial analysis,
Springer
46. Fragoso, E. C. (2003). Apertura comercial y productividad en la industria
manufacturera mexicana. Economía Mexicana, 12(001), 5-38.
47. Fujii, G. (2010). La lógica laboral del modelo de exportaciones inten-
sivas en trabajo no calificado: El caso de México, Revista de Economía
Mundial, 24, 67-102.
48. Fujii, G. (2011). Empleo y modalidad de crecimiento económico en
México. Madrid, España: Fundación Ortega-Marañon-Ediciones Cinca.
49. Fujii, G. and Cervantes, R. (2008). Apertura comercial y empleo en
México, 1988-2004. Trabajo presentado en el VI Congreso Nacional de
la Asociación Mexicana de Estudios del Trabajo, mayo de 2008, Queré-
taro, México.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
124
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
50. Fujii, G., Candaudap, E. and Gaona, C. (2005). Exportaciones, industria
maquiladora y crecimiento económico en México a partir de la década
de los noventa, Investigación Económica, LXIV(254), 125-156.
51. Garduño, S. O. (2009). Ciclos económicos manufactureros en México,
Territorio y Economía, 27.
52. Garza, G. (2008). Macroeconomía del sector servicios en la Ciudad de
México 1960-2003. México: El Colegio de México.
53. Germán Soto, V. and Escobedo Sagaz, J. L. (2011). ¿Ha ampliado la
liberalización comercial la desigualdad económica entre los estados
mexicanos? Un análisis desde le perspectiva econométrico espacial,
Economía mexicana, XX(1), 37-77.
54. Giuseppe, Arbia (2014). A primer spatial econometrics with application
in R. Inglaterra: Palgrave Macmillan.
55. Guerrero de Lizardi, C. (2006). Determinantes del crecimiento económico
en México, 1929-2003: una perspectiva postkeynesiana, Investigación
Económica, LXV(255), 127-158.
56. Hein, E. and Tarassow, A. (2010), Distribution, aggregate demand and
productivity growth: theory and empirical results for six OECD countries
based on a post-Kaleckian model, Cambridge Journal of Economics,
34(4): 727-754.
57. Ibarra, C. (2008). La paradoja del crecimiento lento de México, Revista
de la CEPAL, 95.
58. INEGI (1989). Censo Económico. Instituto Nacional de Estadística,
Geografía e Informática, México: INEGI.
59. INEGI (1994). Censo Económico. Instituto Nacional de Estadística,
Geografía e Informática, México: INEGI.
60. INEGI (1999). Censo Económico. Instituto Nacional de Estadística,
Geografía e Informática, México: INEGI.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
125
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
61. INEGI (2004). Censo Económico. Instituto Nacional de Estadística,
Geografía e Informática, México: INEGI.
62. INEGI (2009). Censo Económico. Instituto Nacional de Estadística,
Geografía e Informática, México: INEGI.
63. INEGI (2014). Censo Económico. Instituto Nacional de Estadística,
Geografía e Informática, México: INEGI.
64. Gordon, R. (1990). What is new keynesian economics? Journal of
Economic Literature, 28(3), 1115-1171.
65. Kaldor, N. (1984). Causas del lento ritmo de crecimiento del Reino Unido,
Investigación Económica, XLIII(167), 9-24.
66. Keynes, J. M. (1939). Relative Movements of Real Wages and Output,
Economic Journal, 49, 34-51.
67. Lanteri, Luis N. (2013). Determinantes económicos del nivel de empleo:
Alguna evidencia para Argentina, Ensayos Revista de Economía. XXXII(1),
73-100.
68. Lavoie, M. and Stockhammer, E. (2012). Wage-led growth: concept,
theories and policies, Conditions of Work and Empyoyment Series 41,
Genová, pp.1-32.
69. Lechuga M., J. and Varela O., M. (2001). Empleo manufacturero en
México, 1990-1998, Análisis Económico, XVI(33).
70. Lesage, J. and Pace, K. R. (2009). Introduction to spatial econometrics.
London: CRC Press/Taylor and Francis Group.
71. Loría, E. (2009). Sobre el lento crecimiento económico de México. Una
explicación estructural, Investigación Económica, LXVIII(270).
72. Loría, E. (2001a). El desequilibrio comercial en México, o por qué no
podemos crecer al 7%, Momento Económico, 113 , 16-21.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
126
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
73. Loría, E. (2001b). La restricción externa dinámica al crecimiento de
México a través de las propensiones del comercio, 1970-1999, Estudios
Económicos, 16(2), 227-251.
74. Malcomson, J. (1981). Unemployment and the efficiency wage
hypothesis, The Economic Journal, 91(364), 848-866.
75. Mariña, A. (2005). Balance y perspectivas de la industria manufacturera
mexicana tras veinte años de reestructuración neoliberal: Integración
subordinada a Estados Unidos, desindustrialización y precarización del
empleo. México: Universidad Autónoma Metropolitana-Azcapotzalco.
76. Maroto-Sánchez, A. and Cuadrado-Roura, J. R. (2009). Is growth of
services an obstacle to productivity growth?: A comparative analysis,
Structural Change and Economic Dynamics, 1-24.
77. Mejía Reyes, P., Gutiérrez Alva, E. E. and Farías Silva, C. A. (2006). La
sincronización de los ciclos económicos de México y Estados Unidos,
Investigación Económica, UNAM, México, 45(258), 15-45.
78. Mejía Reyes, P. and Erquizio Espinal, A. (2012). Expansiones y recesiones
en los Estados de México. México: Pearson Educación-Universidad de
Sonora.
79. Mejía Reyes, P, Mejía Reyes, A. and Rendón Rojas, L. (2017). Dinámica
del empleo manufacturero despúes de la recesión en los estados mexi-
canos. En Y. Carbajal Suárez y L. de Jesús Almonte (Coord), Dinámica
del empleo y la producción manufacturera en México. México: UAEM.
80. Mejía Reyes, P, Rendón Rojas, L. and Vergara González, R. (2018).
International synchronization of the Mexican states business cycles:
Explaining factors, North American Journal of Economics and finance,
44, 278-288.
81. Montiel, J. Y., González, A. N. and López, A. (2007). Crecimiento del PIB,
desempleo y estructura del empleo en Estados Unidos, Francia, Japón y
México, Aportes, XII (34), 159-173.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
127
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
82. Moreno Serrano, R. and Vayá, E. (2000). Técnicas econométricas para
el tratamiento de datos espaciales: la econometría espacial. Barcelona,
España: Ediciones Universitat de Barcelona.
83. Munnel, Alicia H. (1990). Why Has Productivity Growth Declined?
Productivity and Public Investment, New England Economic Review,
1-22.
84. Naastepad, C.W.M. (2006). Technology, demand and distribution: A
cumulative growth model with an application to the Dutch productivity
growth slowdown, Cambridge Journal of Economics, 30(3), 403-434.
85. Ocegueda Hernández, J. M. (2000). La hipótesis del crecimiento restrin-
gido por la balanza de pagos: Una evaluación de la economía mexicana,
1960-1997, Investigación Económica, LX(232), 91-122.
86. Storm, S., Naastepad, C. W. M. (2009). Labor Market Regulation and
Productivity Growth: Evidence for Twenty OECD Countries (1984–2004),
Industrial Relations, 48(4), 629-54.
87. Parajuli, J. and Haynes, K. E. (2017). Panel data models of new firm
formation in new England, Region, 4(3), 65-76.
88. Pérez-Campuzano, E., Sánchez-Zárate, A. and Cuadrado-Roura, J.R.
(2018). Distribución espacial del sector servicios en México, 1999-2009:
Especialización y diversificación desde una perspectiva macrogeográfica,
Eure, 44(131), 215-238.
89. Perrotini, I. (2004). Restricciones estructurales del crecimiento en
México, 1980-2003, EconomíaUNAM, 1(1).
90. Quintana, L., Andrés, R. and Mun, N. (2013). Regional Growth and
Development in Mexico and South Korea: A Comparative Analysis of
Kaldor’s Law, Investigación Económica, 72(284), 79-105.
91. Quintana Romero, L. and Andrés-Rosales, R. (2014). Técnicas modernas
de análisis regional. México, D.F.: Plaza y Valdés.
Spatial analysis of manufacturing employment in Mexico, 1984-2013
128
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
92. Ros, J. (2008). La desaceleración del crecimiento económico en México
desde 1982, El Trimestre Económico, 75(299).
93. Ros, J. (2010a). Política fiscal, tipo de cambio y crecimiento en regí-
menes de alta y baja inflación: la experiencia de México: En N. Lustig
(coordinadora). Los grandes problemas de México. Vol. IX. Crecimiento
económico y equidad. México: El Colegio de México.
94. Ros, J. (2010b). Reformas microeconómicas, política macroeconómica
y crecimiento. El caso de México, EconoQuantum, 6(1).
95. Ros, J. (2013). Algunas tesis equivocadas sobre el estancamiento econó-
mico de México. México: El Colegio de México.
96. Ruiz Nápoles, P. (2004). Exports, growth, and employment in Mexico,
1978–2000, Journal of Post Keynesian Economics, 27(1), 105-124.
Recuperado de http:/www.jstor.org/stable/4538912.
97. Shaikh, A. (1983). La actual crisis económica mundial: causas e impli-
caciones, Investigación Económica, 42(165), 11-38.
98. Secretaría de Economía. (2014). Industria manufacturera, maquiladora
y de servicios de exportación (IMMEX). Secretaría de Economía. Estados
Unidos Mexicanos. Recuperado de http://www.economia.gob.mx/
comunidad-negocios/industria-y-comercio/instrumentos-de-comercio-
exterior/immex.
99. Sobrino, J. (2003). Competitividad de las ciudades en México. México:
El Colegio de México.
100. Torres, A. and Vela, O. (2002). Integración comercial y sincronización
entre los ciclos económicos de México y los Estados Unidos, Documento
de Investigación 2002-6, Banco de México.
101. Vergeer, R. and Kleinknecht, A. (2010-11). The impact of labor market
deregulation on productivity: a panel data analysis of 19 OECD countries
(1960-2004), Journal of Post Keynesian Economics, 33(2), 371-408.
Leobardo de Jesús-Almonte, Roldán Andrés-Rosales y Yolanda Carbajal-Suárez
129
desarro. soc. 84, bogotá, primer cuatrimestre de 2020, pp. 91-129, issn 0120-3584, e-issn 1900-7760, doi: 10.13043/dys.84.3
102. Wooldridge, J. (2002). Econometric Analysis of cross sections and panel
Data. London, England: MIT Press, Cambridge Massachuset.
103. Zevallos, E. (2003). Micro, pequeñas y medianas empresas en América
Latina, Revista de la Cepal, 79, 53-70.

VLEX utiliza cookies de inicio de sesión para aportarte una mejor experiencia de navegación. Si haces click en 'Aceptar' o continúas navegando por esta web consideramos que aceptas nuestra política de cookies. ACEPTAR