Globalization and Technological Growth: A Model for Technology Diffusion

in Mexico, 1996-2000

Dr. Jorge Eduardo Mendoza1 El Colegio de la Frontera Norte

Mtro. Victor Hugo Torres Universidad Autonóma de Guadalajara

Abstract

The study shows that the generation of innovations presented a moderate increasing trend in Mexico during the period 1996-2000. However, there was a divergence process between national and foreign innovation. The results of a panel data model showed a positive effect of the existing human capital within each state on the innovation process. The variables representing the impact of out-of–the-state innovation and the stock of foreign knowledge showed a negative effect on the process of innovation of the states. When including both the stock of foreign innovation and the out-of-the-state stock of innovations, the productivity of human capital decreased.

Resumen El estudio muestra que el proceso de innovación en México tuvó un crecimiento moderado durante el periodo 1996-2000. Sin embargo, la participación de las innovaciones nacionales decreció y la de las extranjeras se elevó. Se aplicó un modelo de datos de panel, cuyos resultados muestran un efecto positivo del capital humano disponible al nivel estatal. Las variables que representan las patentes de otros estados y las patentes extranjeras mostraron signos negativos respecto a la generación de patentes. Cuando se incluyen ambas variables el efecto de la productividad del capital humano para generar innovaciones decreció. 1. Introduction

Several theoretical and empirical studies have analyzed the problem of

technological change. The analysis of the determinants of innovation has

1 Director del Departamento de Estudios Económicos en El Colegio de la Frontera Norte. México. Dirección: Blvd. Abelardo L. Rodríguez 2925, Zona del Río, BC, 22320, México. US Address : P.O. Box “L”, Chula Vista, CA, 91912-1257, USA. Email: emendoza@colef.mx

1

1

developed based on economic growth theory. Solow (1956) pointed out the

importance of technological change in the rate of growth of the economy.

Following that approach, Romer (1990) set up a model including technological

innovation as endogenous variable, by assuming that it was a function of human

capital devoted to research and development and of the stock of knowledge.

Other studies that have contributed to the theoretical aspects of the

analysis of technological change are Jones (1995) and Aghion y Howitt (1998).

However, only a few studies have focused on the empirical analysis of

technological change. Among those papers two stand out. The first, by Lucas

(1988), introduced the concept of human capital, although he did not present any

specific conclusions regarding the relationship technology-growth. The second,

by Barro and Sala i Martín (1996), presented a growth model with human capital

as an explanatory variable in the convergence process.

Recently, Porter et. al. (1999) and Stern et. al. (1999) have focused on

the empirical analysis of technological change. The first paper estimated the

production function of innovation proposed by Romer, with the objective of

understanding the positive influence of both the human capital devoted to

research and the technological spillovers on economic growth. Stern et. al.

proposed the concept of national capacity as a synthesis of the theoretical work

of Porter, Nelson (1993) and Romer. Under this approach microeconomic factors

emerge from the cluster activities.

The studies regarding technological globalization have mainly focused on

the empirical process, with few contributions from the theoretical perspective

2

2

(Estay, 1995). Studies have been published on the process of globalization and

its influence on the regions within a country (Dávila, 1995) and (Manchón, 1995)

and its relationship with the process of technological innovation (Dávila, 1995)

and (Váldez, 1995). There are also studies related to the fragmentation of

technological innovation among different countries, according to the strategies of

the multinational firms located there.

During the decade of the eighties, several papers focused on the evolution

of science and technology in Latin America. They pointed out the dispersion of

research and development spending and patents production among developed

and developing economies (Sagasti, 1987). Other studies researched

technological capacity, considering that technology transference is a factor that

increases the national capacity to generate technology and stressed the need for

developing a broad base of researchers who can contribute to the production of

patents (Segal, 1987). Finally, another paper has stressed the importance of

inventions as a dynamic engine of economic activity (Tsur, 1989). Inventions are

the result of applied technology, which is derived from scientific knowledge. It

also underlined the potential of patents as instruments for developing

technology. It is important to stress that those papers acknowledge the

relevance of technology in regards to economic growth by regions. However,

they do not analyze the phenomenon with a formal and consistent theoretical

framework.

For the case of Mexico, there is a tendency to produce empirical studies

regarding technological change. Aditionally, those studies have not address

3

3

innovation activity as a crucial element for the economic growth of the country.

One such study focuses on innovation at the national level, by analyzing

statistical information about patents (Aboites, 1999).

The present paper seeks to analize national innovation activity from a

regional perspective, within the context of the globalization of technology. When

analyzing the technological innovation in Mexico, three characteristics stand out:

a) There is a high sensibility of both the foreign and domestic innovations

to the business cycle.

b) Foreign innovation showed an increasing trend during the decade of

the nineties.

c) There is a decline in the rate of growth of national innovations, in the

same period.

Taking into consideration the characteristics of technological innovation in

Mexico, the present paper analyzes the interrelationship of both the foreign and

domestic innovations and estimates the impact of technological globalization on

the domestic capacity to produce innovations. The context of analysis

incorporates regional dynamics in the process of innovation by assuming that

technological innovation is developed under a framework of regional

relationships.

2. The globalization of the technological Innovation 2.2.1 Innovation and foreign patents in Mexico The liberalization of the Mexican economy was characterized by the rapid growth

of foreign patents (see Graph 1). The annual average growth of foreign patents in

4

4

Mexico from 1980 to 2000 was 5.5% (Table 1). In particular, during the period

1995-2000 the foreign patents requested increased to an average growth rate of

16.8%, as a result of the technological globalization process and the increasing

presence of direct foreign investment in the manufacturing activities (Table 2).

Graph 1

Source: Own elaboration with data fro Table 12.

Patents of US origin make up the largest share of those arriving to Mexico.

In 1980 the US participated with 48.3% of the foreign innovations. Moreover, in

1994 that share increased considerably to 62.3% of the total patents (Table 3).

From a legal point of view, an important factor encouraging the flow of

foreign patents to Mexico was its incorporation into the World Intellectual

Property Organization (WIPO). That institution implemented a patent cooperation

treaty (PCT) in 1995. That agreement allows foreign patents to enter all signing

countries within a more secure legal framework. As a result, the multinational

enterprises of the industrialized countries have developed strategies for

Patents requested by foreigners in Mexico

1980-2000

0 2000 4000 6000 8000

10000 12000 14000

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Years

Num

ber o

f pat

ents

5

5

extending legal and market protection to their technological innovations in order

to prevent the expansion of new local competitors.

The increasing flow of patents into the Mexican economy has resulted in a

higher technological capacity because the introduction of new products has

enlarged the stock of knowledge available to research and development

activities. However, the rapid growth of foreign patents has also determined a

gap between the Mexican economy and the economies that have developed

technology innovations. This technological gap decreases as a result of the

positive impact of technological spillovers on other economic sectors, and on the

other hand, increases because of difficulties in adopting foreign knowledge.

Finally, it is important to point out that, in 1991, the foreign patents in

Mexico were concentrated on the chemical and metallurgic industries (35.7%).

Other important receptors of technology were the industrial and consumer goods,

and the electricity sector, with shares of 18.7%, 16.7% and 8.8% (Table 4). In

1999, the shares of patents were concentrated in the electricity and consumer

goods sectors.

2.2.2 National innovations and patents In Mexico, the domestic generation of technology, as measured by the number of

requested patents, decreased during the last decade. The aggregate production

of innovations presented an annual average rate of growth of 2% for the period

1980-2000. The number of patents requested by Mexican nationals decreased

from 665 in 1980 to 431 in 2000 (Table 5).

It is worth noting that during the period 1985-1990 the rate of growth of

patents was positive (1.3%). In contrast, from 1990 to 1995 the rate of growth of

6

6

patents became negative at –6.8 (Table 1). However, the next 5 year period

(1995-2000) was characterized by a rapid process of economic liberalization and

reactivation, but even so, the rate of growth was stagnant, while foreign patents

experience very rapid growth ( see Graph 2).

Graph 2

Source: Own elaboration with data from Table 12.

In order to evaluate the interaction of the inputs required to develop

technological change, it is important to analyze the impact of human capital. In

the present study, this variable is specified as the number of researchers that

belong to the National Research System (SNI).

In 1991 and 1992 the rates of growth of members of the SNI were 7.9% and

7.3% respectively. However, during the period 1993-1995 negative rates of

growth were experienced, probably impacted by the economic recession. During

1997-1999 the growth rate recovered its dynamism to levels above 7%. Even

though its growth trend showed clear fluctuations, the positive growth of

researchers for the period 1991-1999 could be positively affecting the innovation

process (Table 6).

Patents requested by residents of Mexico

1980-2000

0

200

400 600

800

1980 19821984198619881990199219941996 1998 2000 Years

Num

ber o

f pat

ents

7

7

Patents are important indicators of technological change. The

accumulation of the stock of innovations can be mainly observed through the

accumulated stock of patents. For the case of Mexico the growth of the patent

stock has been decreasing as a result of the diminishing number of national

patents requested. Thus, the annual rate of growth of patents decreased from

38.4% in 1982 to 3.7% in 2000 (see Table 6). It is assumed that the larger the

stock of patents the greater is the access to knowledge by researchers and their

productivity in the innovation process.

2.3 Regional innovation structure and dynamics The structure of the regional production of innovations reflects the diversity of this

activity in the Mexican economy and the existence of different conditions under

which the innovation process is undertaken. Its implications are exhibited in the

heterogeneous contribution of the states to the innovation process.

In the year 2000, the regions that had a larger participation in the process

of national innovation were Distrito Federal (with its greater technological

infrastructure), state of Mexico, Jalisco, Nuevo Leon, Puebla and Queretaro, with

shares greater than 4%. The states with lower shares were Aguascalientes, Baja

California, Baja California Sur, Campeche, Chiapas, Durango, Guerrero, Hidalgo,

Michoacan, Nayarit and Oaxaca (see Table 7).

The importance of the diversity of innovations among states is related to

the fact that regions with weaker technological innovations can be affected by

those regions with greater technological innovation. On one hand, the possibility

of positive effects of technological spillovers among states could encourage the

production of innovation in comparatively weaker regions. However, the disparity

8

8

in the technological change at the regional level suggests the lack of interaction

among innovation activities in the states of Mexico. This phenomenon could be

the result of the restrictions to inter-regional technological spillovers.

Apparently, the innovation process in Mexico has been characterized by a

limited relationship between the availability of human capital potentially able to

participate in research and development activities and the production of patents.

Therefore, the higher number of researchers (SNI) has not implied, in many

cases, a greater rate of technological innovation growth.

According to information on SNI researchers, there exists a high possibility

that those researchers have a tendency to move to regions with better conditions

in terms of technological infrastructure and access to knowledge. The highest

concentration of researchers for the year 2000 have been located in Distrito

Federal, state of Mexico, Morelos, Puebla, Jalisco, Guanajuato and Baja

California. Since 1996, Baja California, Estado de Mexico, Morelos and Puebla

have increased their stock of human capital (see Table 8).

3. Theoretical aspects of the innovation process under the globalization of technology This approach to the evaluation of the impact of the globalization of technology

on the production of innovation in Mexico is based on the endogenous economic

growth model developed by Romer (1990). This approach considers that

technological innovation is a function of new knowledge or ideas that, in turn, are

9

9

endogeneized, since they depend on the human capital devoted to research and

development and on the existing stock of knowledge2. Therefore, according to

this model, there are spillover effects from the stock of ideas, which can be used

by the human capital in the production of new ideas.

The conclusions derived from the Romer model are that, in equilibrium,

the rate of growth of per capita production yg , the rate of growth of capital per

worker kg , and the rate of growth of ideas Ag , increase at the same rhythm.

Therefore, production growth depends on the production of ideas, which in turn

depends on the number of researchers. The other important implication is related

to the positive effect of ideas spillovers on the average productivity of

researchers (−

δ ), which can allow for a growing expansion of innovation even if

the number of researchers remains constant.

Porter et. al. (1999), developed a model based on Romer’s approach to

evaluate the effect of domestic innovation spillovers (within each country) and

international spillovers (from the rest of the world). He relates the international

patents to a national production function of ideas and considers that there are

few national patents that, in a given point in time, are new or at the cutting edge

at the international level, and only the ones that go beyond the knowledge

frontier are relevant at the international level.

2 AδHA A

.= is thecnological change, where

.A ≈ A∆ , depends on the flow of new ideas, A is the stock

of ideas. The parameterδ represents the human capital research productivity. A non-rivalry environment is assumed (A can be used by all the researchers).

10

10

For that reason, the stock of international ideas has a double effect. On

one hand, it has a diffusion effect on the national growth of technology, and on

the other hand, it also functions as a barrier or negative diffusion due to the

existence of technological competition to be on the cutting edge of technology.

However, the assumption of non-rivalry and partial exclusivity remain, which

imply that national reserachers have access to the international knowledge in

their own country.

By the same token, this assumption is also applied to the world stock of

knowledge. The increasing stock of knowledge acquired within a country implies

that another country could not patent it, although it can have access to it. In that

sense there is a competition to patent year by year and there are different

propensities to patent.

3.1 The model to evaluate the impact of international patents at the regional

level

The model for the evaluation of the impact of foreign patents on the national

production of patents at the state level is based on a national production function

of ideas:

)1.........(........................................AAδHA sAλ

.αψβψϕ −−

−= xs A

Where:

.A = growth of new ideas, including the ones discovered in other countries

and that have not been used in country j

AλH = human capital devoted to research and development

11

11

ϕsA = accumulated production of national ideas in the state s

βψ −

−sA = stock of ideas discovered in other states of the country that have

not been discovered in the state s (out-of-state innovation)

αψ −

xA = stock of foreign ideas

Therefore, the production of innovation results from the interaction

between the innovation within each individual state and from out of the states.

When the parameters show the following characteristic ψ > β , the impact is

positive, and it would be negative if ψ < β . In order to take into account the

effects of the globalization of innovations, a variable representing the stock of

international innovations is included. Its impact can be divided into two parts. The

first one is positive and it derives from the effect of diffusion that the stock of

international ideas has in *.

A . However, there is a barrier effect or negative

diffusion when the new ideas in the country are not new at the international level

(α > ψ ).

It is assumed that there is a non-rival or partial exclusivity environment,

which implies that the human capital has access to the stock of international

knowledge both in the state sAφ and in the rest of the regions that make up the

national economy βψ -s-A .

The present study is based on the following hypotheses:

• The regional human capital devoted to research and development at the state

level (HAs) has a positive influence on the production of regional patents. This

12

12

implies that the research effort undertaken in Mexico is an important factor

leading to the increase in technological innovation.

• The stock of regional individual patents (As) has a positive effect on the

regional production of individual patents. Therefore, there are innovation

spillovers from knowledge acquired previously in each region, which is used

by researchers to generate new ideas.

• The stock of out-of-state patents (A-s) has a negative impact on the

production of regional individual patents. This is caused by barriers to the

diffusion of knowledge acquired by the states, which in turn reveals a weak

link between the innovation activities among the states and problems of

adaptation and technology transfer.

• The stock of foreign patents (Ax) has a negative effect on the production of

regional individual patents. Thechnological globalization, expressed by the

increasing production of foreign technological innovation, seems to be related

to limited national innovation activities and problems in the diffusion of

international knowledge.

4. The empirical model of technological change The theoretical model presented above is based on the use of patents as a proxy

for innovation activities. The data from patents has frequently been used as an

economic indicator in applied economic research (Grilichez, 1990). Moreover, the

use of patents as a proxy for technological knowledge has been accepted by

empirical researchers such as Romer (1990), Jones (2000), Porter et. al. (1999)

13

13

and Stern et. al. (2001). However, it must be acknowledged that patents are an

imperfect indicator of technological innovation, since not all innovation is

patented.

Since the states of Mexico showed different rates of growth in the

production of innovation, it is important to be able to estimate the impact of

innovation activity at the regional level. The estimation of the econometric model

will be based on a fixed effects panel regression applied to a cross section

database that encompasses several years. By estimating a production function

using states as decision units, we will be able to evaluate the national production

of innovation as a result of the state innovation activity.

The empirical model extends the methodology proposed by Porter (1999)

to analyze the evolution of regional technological innovation in Mexico by

estimating the following two empirical equations:

1) st stAstY ηφlnAλlnHγc

sδ tlnPatsst ++++=

Where sPats represents the innovation produced and patented in the period t in

all the states of Mexico in the year t, AstH is the human capital devoted to the

production of ideas (research and development) for each of the states of Mexico

in the year t, AstH is the stock of innovation accumulated (requested patents) that

have been discovered in other states of Mexico and that have not yet been

discovered in the state s, in the year t (out-of-state patents).

There are two dummy variables to control for state and year. They are

introduced to take into account the differences in the propensity to patent

14

14

resulting from the different industrial composition of states γcj , and also to

evaluate the variation of that propensity over the yearsδ tY ,

2) ηAXα)(ψAβ)(ψφlnAλlnHγcsδ tPatsln st ststststAst

Y +−−++++= −

Where AXst is the stock of foreign innovation (foreign patents requested)

that flow to Mexico and affect the state s in the year t. The impact of this variable,

at the regional level, is calculated as the ratio of foreign patents requested to the

total of all patents requested in the state.

In order to be able to evaluate the effects of the globalization of

technology on the innovation growth at the state level over a period of years, we

chose to apply a fixed effects panel economteric model to a cross sectional and

times series database.

4.1 Data Base

The number of patents by state was obtained from the matrix of patents

requested from 1991 to 2000, published by El Consejo Nacional de Ciencia y

Tecnología (CONACYT). The stock of patents at the state level and the stock of

patents of the rest of the states were constructed from the data of patents at the

state level. The proxy of human capital devoted to research and development

was obtained from the data of the Sistema Nacional de Investigadores (SNI) by

state, for the year 1996 and 20003.

3 The SNI stands for the National Researchers System and it was created by CONACYT in order to organize the researchers in the country and to strengthen the scientific and technological research of the country.

15

15

The SNI classifies its researchers into the following groups: physics and

mathematics, medicine and health sciences, biology and chemistry, humanities

and behavioral sciences, agricultural sciences and engineering.

Even though it is not possible to disaggregate human capital by activity

and region at the same time, it is important to stress that the researchers

dedicated to research and development in those sciencies (excluding social

sciences), make up a large proportion of the members of the SNI (Table 9).

Finally, the foreign patents requested information is published by CONACYT and

includes data beginning with1963.

5. Results The innovation activity showed a continuous decline in the period studied, in a

context of globalization of technological innovations. By estimating the

parameters of the production function of innovation, this decline can be explained

at the regional level, according to the endogenous growth theory.

The first econometric model showed an acceptable goodness of fit of the

adjusted R square. The DW statistic did not show autocorrelation, and the

coefficients were statistically significant at the 99% level of confidence.

Additionally, an F test was applied to the models, with a null hypothesis that fixed

effects are equal among states4, and that there are no differences among states.

The results rejected that hypothesis (Table 10).

According to Romer (1990) the parameter of human capital devoted to the

production of ideasλ and the parameter of the stock of knowledge φ are

4 Using the sum of the squared residuals we tested the null hypothesis that α1= α2 = α3 …= αn, the estimated F was higher that the F statistic in the tables, and the null hypothesis was rejected.

16

16

assumed to be equal to 1. Therefore, an increment in the stock of knowledge

determines a proportional increment in the productivity of the researcher.

For the case of Mexico, the regression results showed that both

coefficients were statistically significant, and φ presented a value of 1.07 and λ a

value of 0.16. Hence, the coefficient φ suggests that there are conditions within

the states that allow for a proportional increase in research productivity.

The positive effect of a value greater than 1 of the coefficient φ implies

that within each state there are positive effects of the stock of patents that

encourage innovation activity. The effect of stock spillovers is an externality in

each state, resulting from innovations and experiences that were developed in

some activities, and that are used by other innovation efforts. Therefore, within

each state there is a process of sharing of information. Since the study is using

patents as a proxy of innovation, the diffusion effect consists of the use of

previous patents as technological information to create new patents.

On the other hand, the coefficient of human capitalλ had a value less than

1, implying the existence of a duplication of research effort (Jones, 2000); this

effect reduces the growth of ideas resulting from an increase in the number of

researchers. Therefore, according to Jones (1995) and (2000), the value of this

parameter (λ < 1), suggests the possibility of having a weak but positive growth

in the production of ideas.

Therefore, a positive value of φ implies that there are internal conditions

within the states that encourage the development of technological innovation. It is

possible that there are also difficulties in the use of those favorable conditions,

17

17

which in turn are restricting the growth of innovation. In the case of Mexico, the

low value of the human capital coefficient suggests that external effects such as

the national competition from the rest of the states and the foreign competition

are reducing the possibility of increasing the stock of patents at the state level.

Besides, due to the heterogeneity in the characteristics of production activities

and industrial structure, it is possible that the states are not able to incorporate

the stock of knowledge from other states in the innovation process.

A second model was estimated in order to quantify the effect of innovation

in other states on the growth of the technological innovation of each state. The

impact of the innovation from the rest of the states also explains the difficulties in

the research effort (λ < 1) and the decreasing participation in the national

innovations at the state level.

The results are consistent with the first model. The coefficient φ is

statistically significant at the 99 level of confidence and had a value of 1.60. The

coefficient of the variable rest of the states is expressed as (ψ-β), and it can have

a positive range when ψ > β (diffusion effect), and a negative range when ψ < β

(barrier effects from competition). Table 15 shows a negative coefficient of 1.29,

which is statistically significant at the 99% level of confidence. Therefore, the

results suggest the possibility of barriers to technological spillovers. In fact

according to the value of the coefficient (close to 1), the increase of the

innovation activity in the rest of the states is negatively affecting the innovation

process at the single state level. Therefore, the coefficient of the variable rest of

the states helps to explain the decreasing trend of the national innovation activity.

18

18

The analysis of the coefficients representing the fixed effects of the

regression allow the estimation of the impact of the barrier effect on human

capital and on the existent stock of knowledge, at the state level. The states

showing a higher impact from technological competition were Quintana Roo,

Campeche, Chiapas, Durango, Tabasco, Colima, Nayarit, Baja California Sur,

Chihuahua and San Luis Potosi. It is worth mentioning that these states are

characterized by having low levels of technological innovation.

The 10 states with lower barrier effects were Puebla, Coahuila, Queretaro,

Hidalgo, Morelos, Veracruz, Yucatan, Jalisco, Nuevo Leon and Distrito Federal.

With the exception of Veracruz and Yucatan, these states are characterized by

relatively higher levels of technological innovation (Table 11).

These results suggest that the states with a higher production of

innovation are experiencing lower barrier effects and that those states negatively

impact the technological innovation of the states with low innovation activity. This

negative impact could be derived from competition (patents can not be registered

twice), the Law of Patents in Mexico and the technical difficulties of absorbing the

knowledge from the more developed states. Therefore, the existence of negative

diffusion effects are related to the process of convergence in the production of

innovations in Mexico, which is is based on the high rates of innovation growth of

a few states, Mendoza and Torres (2002).

In order to estimate the impact of technological globalization on the

production of national innovations in Mexico, a third model including the foreign

patents requested was estimated. According to Table 5 the parameter φ

19

19

exhibited a value of 1.66, the coefficient λ was 0.07, the coefficient ψ-β had a

value of -11.27 (higher than the regression without foreign patents) and the

coefficient showing the effect of the foreign patents (ψ - α) presented a negative

value of -0.11. All of the coefficients shown were significant at the 99% level of

confidence. The R square has an acceptable goodness of fit and the Durbin

Watson did not point out any autocorrelation problem.

It is important to point out that when incorporating the variable foreign

patents, the previous coefficients increased their values. Therefore, the results

suggest that the availability of foreign patents supports the existence of regional

conditions that increase the productivity of human capital and that increase the

innovation activities at the state level in Mexico. However, the value of the

coefficient of human capital (λ < 1) again shows difficulties for the human capital

in the development of innovation activities at the state level. From that

perspective, the results suggest the existence of barrier effects determined not

only by the lack of technological spillovers among states but also among the

domestic and the foreign patents.

As a result, innovation activity in Mexico, evaluated at the state level, has

been relatively weak, and in the long run, is translated into a reduction of the

national accumulated knowledge. The models estimated provide information

about the trend of the innovation process in Mexico. In particular, since 1995

foreign innovations have shown an accelerated growth while national innovation

activities have been relatively stagnant.

20

20

The effects of the foreign patents on the coefficients of the fixed effects by

states showed a barrier to diffusion for the states of Campeche, Durango,

Tabasco, Colima, Nayarit, Baja California Sur, Chihuahua, Guanajuato, Sonora

and San Luis Potosi. As it was mentioned above, a low level of innovation

characterizes these states. With respect to the states with less negative impact,

the same pattern emerges as the previous model. The states were Puebla,

Veracruz, Querétaro, Guerrero, Coahuila, Morelos, Hidalgo, Nuevo León and

Distrito Federal (Table 12).

6. Concluding Remarks Technological globalization is characterized by a rapid growth of foreign

innovation and an increasing share in the total creation of technology of the

receptor countries. This phenomenon is experienced in both the more

technologically advanced countries and in the developing countries. Mexico is a

net receptor of foreign technology, which represents a larger share of the total

national innovation.

The analysis of the evolution of innovation in Mexico, within the context of

the technological globalization experienced during the decade of the nineties,

shows two different results. On one hand, there is an important diminishing trend

in the process of innovation at the national level, expressed in the decline of the

national patents requested. However, the foreign patents requested exhibited an

increasing rate of growth.

In order to provide explanations of this dichotomy in the process of

innovation within the Mexican economy, a model of technological growth at the

21

21

regional level was established. The different econometric models presented

interesting findings with respect to the specific conditions, at the state level, in

which the human capital devoted to research and development, the regional and

international technological competition, and the stock of innovations are affecting

technology growth in Mexico.

The most important results are the following:

1. The coefficient of the impact of the stock of patents on the productivity

of research and development activity (φ ), supports the conclusion of the

existence of available knowledge that facilitates the generation of new patents.

2. The duplication effect (λ <1), implies that there are difficulties in using

the stock of patents available. An increase of the stock of knowledge generates

an increase less than proportional to the potential. Therefore, this result explains

part of the problems causing slow growth of domestic patents at the state level.

3. The variable out-of-state innovation helps to estimate the effect that the

other states have on the individual states in regards to innovation. The estimation

(ψ-β< 0) shows negative effects among states due to the technological

competition (the patents requested in one state precludes that other states

request a patent for the same technological development). Therefore,

technological innovation depends increasingly on the innovation activities

developed within each state,

4. The model, which includes the variable of the stock of foreign patents

available in Mexico, suggests the existence of a negative effect on national

innovation activity, derived from the competition that blocks the creation of new

22

22

national patents, and the resulting diminishing average productivity of the

research and development activities −

δ . In addition, the negative effect of foreign

patents reflects the weak technological integration of economic activities carried

out by national and foreign firms, and the wide technological gap between local

and foreign innovation.

Bibliography Aboites A., Jaime (1995). Cambio Institucional e Innovación Tecnológica. UAM-

Xochimilco. México.

23

23

Aboites A., Jaime y Soria L., Manuel (1999). Innovación, Propiedad Intelectual y Estrategias Tecnológicas, la Experiencia de la Economía Mexicana. UAM-Xochimilco. México.

Acs, Zoltan, de la Mothe, John and Paquet, Gilles (2000). “Regional Innovation: In

Search of an Enabling Strategy”, in Zoltan J. (comp.), Regional Innovation, Knowledge and Global Change. Series in Science, Technology and International Political Economy. Great Britain.

Aghion, Philippe and Howitt, Peter (1998). Endogenous Growth Theory, The MIT

PRESS. USA. Archibiugui, Daniele and Michie, Jonathan (1997). “The Globalization of Technology: A

New Taxonomy”, in Technology, Globalization and Economic Performance, Cambridge University Press. Great Britain.

Arrow, K. (1962). “The Economic Implications of Learning by Doing”, Review of

Economics Studies, 29, pp. 155-173. Barro, Robert and Xavier Sala-i-Martin (1996). “Convergence Across States and

Regions”, Brookings Papers on Economic Activity, 1, pp. 107-182. Cantwell, J. (1989). Technological Innovation and Multinational Corporatio”, Basil

Blackwell. Capdevielle, M., Corona, J.M. and Hernández C. (2000). “Production System and

Technological Patterns”, in Mario Cimoli (comp.), Developing Innovations Systems: Mexico in a Global Context. London.

Cimolli, Mario (2000). “Macroeconomic Setting and Production System”, in Mario Cimoli

(comp.) Developing Innovations Systems: Mexico in a Global Context. London. Consejo Nacional para la Ciencia y Tecnología (1999). Indicadores de Actividades

Científicas y Tecnológicas, México. Dávila Aldás, Francisco R (1995). “Cambios Tecnológicos, Globalización Económica y Regionalización”, in Globalización y Bloques Económicos: Realidades y Mitos,

U.A.P. and U. de G, México. Dunning, H. John (1993). Multinational Enterprises and The Global Economy, Addison-

Wesley Publishing Company, Great Britain. Estay, Jaime (1995). “Globalización y sus Significados”, in Globalización y Bloques

Económicos: Realidades y Mitos,U.A.P. and U. de G. México.

24

24

Griliches, Zvi (1990). “Patent Statistics as Economic Indicators: A Survey Part I”. NBER, Working Paper 330, Cambridge.

Jones I., Charles (1995). “R&D-Based Models of Economic Growth”, Journal of Political

Economy, 103, pp. 759-784. Jones I., Charles. (1999). Sources of U.S. Economic Growth in a World of Ideas,

Stanford. USA 1999. Judge, George G., Griffiths, W.E., Carter, Hill R., Lutkepohl, Helmuth and Lee, Tsoung

Chao (1985). The Theory and Practice of Econometrics, John Wiley and Sons, inc.

Lucas, Robert (1988). “On the Mechanics of Development Planning”, Journal of

Monetary Economics, 22, pp. 3-42 . Manchón, Federico (1995). “Globalización, Regionalización y Comportamiento

Financiero”, in Globalización y Bloques Económicos: Realidades y Mitos, U.A.P. and U. de G, México.

Mendoza Eduardo y Torres Victor (2002), “Innovación tecnológica y crecimiento regional en México”, Revista Mexicana de Economía y Finanzas, Centro de Finanzas, Instituto Tecnológico de Estudios Superiores de Monterrey, vol. 1, no. 3, pp. 187-201.

Nelson, Richard (1993). National System of Innovation: A Comparative Analysis,

University Press. New York. Organization for Economic Cooperation and Development (1998, 2000, 2001). Main

Economic Indicators. Porter E. Michael and Stern Scott (1999). “Measuring The Ideas Production Function:

Evidence from International Patent Output”. NBER, Working Paper 7891. Cambridge.

Posada, Pedro de la Concha (1998). El Sistema de Patentes en México, Consejo

Nacional de Ciencia y Tecnología, Dirección Ajunta de Política Científica y Tecnológica. Serie Documentos de Trabajo. México.

Romer, P. (1990), “Endogenous Technological Change”, Journal of Political Economy,

vol. 98, pp. S71-S102. Romer, Paul (1986). “Increasing Returns and Long Run Growth”, Journal of Political

Economy, vol. 94, no. 5, pp. 1002-1037.

25

25

___________(1989). “Increasing Returns and New Developments in The Theory of Growth”, NBER, Working Paper 3098. Cambridge.

___________(1990), “Endogenous Technological Change”, Journal of Political

Economy, vol. 98, pp. S71-S102. ___________(1994). “The Origins of Endogenous Growth”, Journal of Economic

Perspectives, 8, pp. 3-22. Sagasti, Francisco R. y Cook, Cecilia (1987). “La Ciencia y la Tecnología en América

Latina durante los Ochenta”, Comercio Exterior, vol. 37, no. 12. Scott Stern, Michael E. Porter, y Jeffrey L. Furman (2000). “The Determinants of

national Innovative Capacity”. National Bureau of Economic Research, Working Paper 7876.

Segal, Aarón (1987). “De la Transferencia de Tecnología a la Institucionalización de la

Ciencia y Tecnología”, Comercio Exterior, vol. 37, núm. 12. Shaiken, Harley (1990). “México in the Global Economy, High Technology and Work

Organization in Export Industries”, Monograph Series, 33. Center for U.S.-Mexican Studies, University of California.

Solow, R. M. (1956): “A Contribution to the Theory of Economic Growth”, Quarterly

Journal of Economics, vol. 70, no. 1, S. 65-94. Solow, Robert M (1957). “Technical Change and The Aggregate Production Function”,

Review of Economics and Statistics, 39, 312-320. Stern, S. M.E. Porter and J.L. Furman (1999). “The Determinants of National

Innovative Capacity”, HBS Working Paper. Tsur, Yoel (1989). “Las Patentes: Reflexiones sobre Tecnología y Desarrollo”,

Comercio Exterior, vol. 39, no. 4. Unger, Kurt y Oloriz, M (2000). “Globalization of Production and Technology”, in Mario

Cimoli (comp.), Developing Innovation Systems: Mexico in a Global Contex. Unger, Kurt y Saldaña, Luz Consuelo (1984). “México, Transferencia de Tecnología y

Estructura Industrial”, Centro de Investigación y Docencia Económica (CIDE), México.

Váldez Fernández, Ma. Teresa (1995). “Cambio Tecnológico, Globalización y Nuevo

Orden Mundial”, in Globalización y Bloques Económicos: Realidades y Mitos, UAP, U. de G, México.

26

26

World Intellectual Property Organization (1999). Industrial Property, Statistics B, Part 1, Switzerland. Electronic addresses http://www.conacyt.mx http://www.siicyt.com http://www.inegi.gob.mx

27

27

ANNEX Table 1

Annual Average Rate of Growth of Requested Patents in México

Period National Foreign 1980-1985 -1.4 -3.9 1985-1990 1.3 5.2 1990-1995 -6.8 2.0 1995-2000 0.0 16.8 1980-2000 -2.0 5.5 Source: Own elaboration with data from CONACYT. Indicadores de Actividades Científicas y Tecnológicas, 1999.

Table 2

Foreign Investment in the Manufacturing Industry in Mexico (Millions of dollars)

Years 1988 1994 1997 AARG 88-97 Food and beberages 125 700 347 11.0%Textiles and Apparel 45 151 595 29.0%Oil, Chemicals and Plastic 394 871 1459 14.0%Metallic products 109 339 807 22.0%Computers 48 165 145 12.0%Cars and trasportation equipment 243 2754 1890 23.0%Subtotal 964 4980 5243 18.0%AARG: Annual average rate of growth Source: OECD. 1998. Main Economic Indicators.

Table 3

Foreign Patents Requested in Mexico 1980-2000 Percentage Structure (%)

Years Mexico Germany USA France Italy Japan UK Spain Other 1980 13.86 8.11 48.28 5.44 2.71 3.59 3.88 1.40 12.741983 17.07 5.01 55.90 4.37 1.78 2.30 2.98 1.17 9.431985 15.83 5.30 56.27 3.54 2.35 2.72 3.10 0.85 10.011990 13.06 6.88 55.80 3.93 1.82 2.25 4.25 0.63 11.381994 5.01 7.46 62.26 2.82 1.57 2.63 3.91 0.71 13.631995 8.01 9.51 58.24 4.95 1.65 3.89 1.28 1.02 11.441999 3.76 9.53 56.68 5.15 1.31 3.27 3.40 0.77 16.13

Source:Own elaboration with data from: CONACYT. Indicadores de Actividades Científicas y Tecnológicas 1999.

28

28

Table 4

Domestic and Foreign Patents Requested by Sector 1991-1999 Percentage Structure

Year Origin

Consumption goods

Industrial goods

Chemical and

MettalurgyTextile

and paper Construction Transportation Scientific equipment Electricity

1991 Domestic 27.84 21.10 15.96 1.60 12.77 9.75 6.91 4.08 Foreign 16.72 18.70 35.71 3.04 3.82 7.63 5.59 8.801995 Domestic 21.53 20.60 16.67 1.62 14.12 12.50 5.09 7.87

Foreign 14.86 21.83 26.51 2.60 2.78 8.83 8.45 14.151999 Domestic 22.64 18.90 15.16 1.76 10.11 14.07 9.01 8.35

Foreign 21.62 17.17 31.14 2.47 2.65 5.62 7.30 12.03Source: Own elaboration with data form: CONACYT. Indicadores de Actividades Científicas y Tecnológicas. 1999.

Table 5

Domestic and Foreign Patents Requested in Mexico 1980-2000Years Domestic Foreign 1980 665 4,132 1981 704 4,624 1982 526 4,280 1983 699 3,396 1984 642 3,361 1985 612 3,253 1986 629 3,071 1987 742 3,509 1988 652 3,748 1989 757 3,817 1990 661 4,400 1991 564 4,707 1992 565 7,130 1993 553 7,659 1994 498 9,446 1995 432 4,961 1996 386 6,365 1997 420 10,111 1998 453 10,440 1999 455 11,655 2000 431 12,628

TMCA -2.0 5.5 Source: Own elaboration with data from CONACYT, Indicadores de Actividades Científicas y Tecnológicas, 1999.

29

29

Table 6

Evolution of the Accumulated Patents and SNI Members Year Accumuted Patents Rates of Growth (%) Researchers Rates of Growth (%)1980 665 1981 1,369 105.9% 1982 1,895 38.4% 1983 2,594 36.9% 1984 3,236 24.7% 1985 3,848 18.9% 1986 4,477 16.3% 1987 5,219 16.6% 1988 5,871 12.5% 1989 6,628 12.9% 1990 7,289 10.0% 5,704 1991 7,853 7.7% 6,155 7.9%1992 8,418 7.2% 6,602 7.3%1993 8,971 6.6% 6,233 -5.6%1994 9,469 5.6% 5,879 -5.7%1995 9,901 4.6% 5,868 -0.2%1996 10,287 3.9% 5,969 1.7%1997 10,707 4.1% 6,278 5.2%1998 11,160 4.2% 6,742 7.4%1999 11,615 4.1% 7,252 7.6%2000 12,046 3.7% 7,466 3.0%

Source: Own elaboration with data from: CONACYT, Indicadores de Actividades Científicas y Tecnológicas 1999.

30

30

Table 7

Requested Patents by States, 1994-2000

Percentage Structure and Annual Average Rate of Growth 1994 1995 1996 1997 1998 1999 2000 AARG 94-00Aguascalientes 0.2% 0.9% 1.0% 1.9% 0.9% 0.7% 0.7% 17.0% Baja California 1.4% 2.1% 1.6% 1.4% 2.0% 2.4% 0.9% -7.7% Baja California Sur 0.8% 0.0% 0.3% 0.2% 0.2% 0.2% 0.0% -100.0% Campeche 0.0% 0.5% 0.3% 0.0% 0.0% 0.2% 0.2% - Coahuila 1.8% 2.3% 3.1% 3.8% 3.1% 2.9% 1.6% -3.5% Colima 1.2% 0.2% 0.0% 0.5% 0.9% 0.2% 1.2% -2.6% Chiapas 0.0% 0.5% 0.3% 0.0% 0.2% 0.0% 0.0% - Chihuahua 0.4% 1.6% 0.5% 0.5% 1.1% 2.0% 2.1% 24.0% Distrito Federal 34.7% 44.9% 42.7% 34.0% 32.7% 39.8% 38.5% -0.6% Durango 0.6% 1.2% 0.5% 0.0% 0.0% 0.2% 0.2% -14.5% Guanajuato 2.0% 3.2% 1.8% 2.1% 2.2% 3.3% 2.8% 2.6% Guerrero 0.4% 0.0% 0.8% 0.2% 0.0% 0.2% 0.2% -9.4% Hidalgo 0.4% 0.2% 0.5% 1.4% 2.6% 0.2% 0.5% 0.0% Jalisco 7.0% 7.6% 6.7% 5.0% 5.5% 7.3% 9.0% 1.6% Edo. Mexico 14.7% 5.6% 7.8% 11.7% 15.5% 9.7% 14.4% -2.3% Michoacan 1.6% 0.5% 1.8% 0.2% 1.1% 0.4% 0.7% -13.1% Morelos 2.8% 2.5% 2.3% 6.4% 3.3% 3.1% 2.6% -3.4% Nayarit 0.2% 0.0% 0.3% 0.5% 0.2% 0.0% 0.2% 0.0% Nuevo Leon 9.4% 12.5% 11.1% 13.3% 10.4% 8.4% 6.3% -7.6% Oaxaca 0.4% 0.2% 1.6% 0.5% 0.7% 0.7% 0.5% 0.0% Puebla 4.6% 1.6% 2.6% 2.6% 3.3% 3.5% 4.4% -2.7% Queretaro 4.4% 2.5% 4.1% 4.0% 3.3% 5.3% 4.4% -2.1% Quintana Roo 0.2% 0.2% 0.0% 0.0% 0.0% 0.4% 0.2% 0.0% San Luis Potosi 0.8% 0.5% 0.8% 0.7% 0.9% 2.2% 1.9% 10.4% Sinaloa 0.4% 0.9% 1.0% 0.7% 1.3% 0.7% 1.9% 21.9% Sonora 1.8% 0.2% 1.0% 0.7% 0.4% 0.7% 0.9% -10.9% Tabasco 0.4% 0.7% 0.0% 0.7% 0.7% 0.2% 1.2% 14.0% Tamaulipas 0.6% 1.4% 1.0% 1.7% 1.3% 0.9% 0.7% 0.0% Tlaxcala 1.4% 0.0% 0.3% 0.0% 0.0% 0.0% 0.0% -100.0% Veracruz 1.8% 1.9% 1.3% 2.4% 2.0% 1.1% 0.9% -10.9% Yucatan 1.0% 1.6% 1.0% 0.7% 0.7% 0.9% 0.9% -3.1% Zacatecas 0.2% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% -100.0% Total 97.8% 98.1% 98.2% 98.1% 96.7% 97.6% 100.0% Sin Clasificar 2.2% 1.9% 1.8% 1.9% 3.3% 2.4% 0.0% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Source: Own elaboration with data from: CONACYT, Indicadores de Actividades Científicas y Tecnológicas.

31

31

Table 8

SNI Researchers by States, 1994-2000

Percentage Structure and Annual Average Rate of Growth 1996 1997 1998 1999 2000 AARG 97-00 Aguascalientes 0.00% 0.26% 0.29% 0.36% 0.70% 0.63% Baja California 3.02% 2.94% 3.03% 0.03% 3.27% 0.08% Baja California Sur 0.00% 1.18% 1.29% 1.36% 1.18% 0.06% Campeche 0.00% 0.08% 0.09% 0.11% 0.14% 0.22% Coahuila 0.00% 0.52% 0.63% 0.67% 0.84% 0.19% Colima 0.00% 0.63% 0.65% 0.62% 0.51% 0.00% Chiapas 0.00% 1.28% 1.28% 1.31% 1.45% 0.09% Chihuahua 0.00% 0.53% 0.50% 0.59% 0.62% 0.10% Distrito Federal 56.04% 54.91% 53.43% 52.42% 50.64% 0.04% Durango 0.00% 0.26% 0.24% 0.25% 0.34% 0.13% Guanajuato 2.97% 5.74% 5.67% 3.06% 3.23% -0.08% Guerrero 0.00% 2.97% 3.03% 0.17% 0.17% -0.48% Hidalgo 0.00% 0.18% 0.18% 0.36% 0.34% 0.13% Jalisco 3.70% 0.18% 0.26% 3.87% 3.66% 1.25% Edo. Mexico 5.46% 3.64% 3.75% 5.85% 6.45% 0.22% Michoacan 0.00% 1.71% 1.85% 1.97% 1.75% 0.06% Morelos 5.75% 5.79% 5.85% 5.85% 6.03% 0.07% Nayarit 0.00% 0.10% 0.12% 0.10% 0.10% 0.07% Nuevo Leon 2.55% 2.57% 2.67% 2.46% 2.36% 0.03% Oaxaca 0.00% 0.40% 0.41% 0.42% 0.40% 0.06% Puebla 4.32% 4.35% 4.62% 4.53% 4.47% 0.06% Queretaro 0.00% 1.76% 2.07% 2.17% 1.83% 0.07% Quintana Roo 0.00% 0.27% 0.27% 0.34% 0.31% 0.09% San Luis Potosi 0.00% 1.05% 1.14% 1.08% 1.08% 0.06% Sinaloa 0.00% 0.58% 0.62% 0.62% 0.68% 0.10% Sonora 0.00% 1.58% 1.53% 1.62% 1.52% 0.05% Tabasco 0.00% 0.06% 0.03% 0.03% 0.03% -0.16% Tamaulipas 0.00% 0.52% 0.42% 0.39% 0.44% 0.02% Tlaxcala 0.00% 0.19% 0.24% 0.24% 0.26% 0.14% Veracruz 1.74% 1.75% 1.68% 1.87% 1.75% 0.06% Yucatan 0.00% 1.58% 1.67% 1.80% 2.01% 0.12% Zacatecas 0.00% 0.44% 0.53% 0.53% 0.64% 0.16% Sin Clasificar 14.46% 0.00% 0.00% 0.00% 0.00% Total 100.00% 100.00% 100.00% 100.00% 100.00% 0.06%

Source: Own elaboration with data from: CONACYT, Indicadores de Actividades Científicas y Tecnológicas.

32

32

Table 9

Members of the National System of Researchers (SNI) by Area, 1991-2000 (percentage shares)

Years Physics ands Mathematics

Biology and Chemicals Medicine Humanities

Social Sciences Biotechnology Engeniering

Sub-total** Total

1991 17.06 19.12 7.17 12.42 8.39 20.26 15.57 79.19 1001992 16.65 20.65 7.97 12.86 8.71 18.45 14.72 78.43 1001993 18.74 22.09 8.45 14.66 9.56 13.41 13.08 75.77 1001994 20.84 21.76 9.58 16.16 10.04 9.73 11.91 73.81 1001995 21.83 21.05 9.99 17.42 10.69 7.92 11.11 71.90 1001996 22.27 20.89 10.15 17.99 11.11 7.15 10.44 70.90 1001997 22.87 20.93 10.35 17.81 10.72 7.37 9.94 71.47 1001998 23.30 20.85 10.43 17.38 10.01 7.86 10.16 72.60 1001999 22.35 19.79 9.94 17.46 10.18 8.85 11.43 72.37 1002000 21.02 19.22 10.25 17.00 10.85 9.38 12.30 72.15 100

Source: CONACYT. SNI data base. 1991-2000. ** The sub-total indicates the shares of the research areas with the exception of Humanities and Social and Behavioural Sciences. The medical research includes Health Sciencies; the area of biotechnology includes the Agricultural Sciences.

33

33

Table 10

Dependent Variable: Log PATSst Coefficients

LogH Ast 0.16* 0.09* 0.07* (2.02) (2.13) (4.40) LogAst 1.07* 1.60* 1.66* (12.29) (17.94) (10.61) LogA –st -1.29 -11.27 (-12.06) (-'8.30) Log AX st -3 -0.11 (-5.36)

Control Variables Trend -0.46* -0.10* 2.08*

(-9.53) (-2.61) (6.09) Log(Prod?) -0.21* 0.03 -0.23*

(-2.42) (0.41) (-6.84)

Statistics R2 0.94 0.95 0.99R2 adjusted 0.92 0.92 0.96Durbin-Watson 2.59 2.4 1.88Observations 114 114 50* Significative at 1% t statistics in paréntesis Log PATSst: annually requested patents by states

Table 11

Fixed Effects by Selected States *

States Effect Average Patents 1996-2000 States Effect Average Patents 1996-

2000 Quintana Roo 8.7 0.6 Puebla 5.7 14.2 Campeche 8.4 0.6 Coahuila 5.5 12.4 Chiapas 7.8 0.4 Querétaro 5.5 18.2 Druango 7.8 0.8 Hidalgo 5.5 28.8 Tabasco 7.5 2.4 Morelos 5.4 15.2 Colima 7.3 2.4 Veracruz 5.4 6.6 Nayarit 7.3 1 Yucatán 5.2 3.6 Baja C. Sur 7.2 0.8 Jalisco 5.0 51 Chihuahua 7.0 5.4 Nuevo León 4.6 42.2 San Luis P. 6.7 5.6 Distrito Federal 3.2 160.6 * When including the variable “out-of-the-state patents.”

34

34

Table 12

Fixed Effects by Selected States*

States Effect Average Patents 1996-2000 States Effect Average Patents

1996-2000

Campeche 76.0 0.6 Puebla 72.4 14.2 Druango 75.1 0.8 Veracruz 72.3 6.6 Tabasco 74.9 2.4 Guerrero 72.2 4.6 Colima 74.8 2.4 Querétaro 72.1 18.2 Nayarit 74.7 1.0 Coahuila 72.1 12.4 Baja C. Sur 74.6 0.8 Morelos 71.8 15.2 Chihuahua 74.2 5.4 Hidalgo 71.7 28.8 Guanajuato 74.0 1.2 Nuevo León 70.2 42.2 Sonora 73.8 3.2 Jalisco 69.9 51.0 San Luis P. 73.8 5.6 Distrito Federal 65.3 160.6 * When including the variable "foreign patents"