Economic Policy in Latin AmericaEconomic Policy in Latin America

Convergence and Convergence and Economic Growth Economic Growth considering Human considering Human Capital and R&D Capital and R&D Spillovers in MexicoSpillovers in Mexico

Alejandro Díaz-Bautista, Alejandro Díaz-Bautista, Ph.D.Ph.D.

Professor of Economics and Researcher at COLEFVisiting Research Fellow and Guest Scholar, Center for U.S.-Mexican Studies, University of California San Diego (UCSD). Graduate School of International Relations & Pacific Studies IR/PS University of California, San Diego

In the recent growth literature, the In the recent growth literature, the accumulation of human capital and R&D have accumulation of human capital and R&D have gained a central role. This study tries to gained a central role. This study tries to narrow the bridge between the fields of narrow the bridge between the fields of regional convergence theory, economic growth regional convergence theory, economic growth and human capital. Unlike traditional economic and human capital. Unlike traditional economic growth theories, which tend to focus on growth theories, which tend to focus on exogenous comparative advantage or exogenous comparative advantage or technological differences among regions as technological differences among regions as causes for growth, regional economic growth causes for growth, regional economic growth emphasizes the roles of increasing returns to emphasizes the roles of increasing returns to scale in production, human capital and R&D in scale in production, human capital and R&D in determining the growth of economic activities. determining the growth of economic activities.

IntroductionIntroduction

IntroductionIntroduction

The study considers the interaction of regional The study considers the interaction of regional human capital and R&D economics following human capital and R&D economics following the recent work in economic growth and the recent work in economic growth and convergence. Using the recent developments convergence. Using the recent developments in economic growth, the study centers on the in economic growth, the study centers on the regional convergence pattern in Mexico regional convergence pattern in Mexico emphasizing the effects of human capital, R&D emphasizing the effects of human capital, R&D and interregional spillovers on growth. and interregional spillovers on growth.

The findings suggest the existence of some The findings suggest the existence of some human capital and bounded knowledge human capital and bounded knowledge spillovers across regional states in Mexico. spillovers across regional states in Mexico.

SolowSolow

Robert "Bob" Solow is an American Robert "Bob" Solow is an American economist particularly known for his work economist particularly known for his work on the theory of economic growth. He was on the theory of economic growth. He was awarded the John Bates Clark Medal (in awarded the John Bates Clark Medal (in 1961) and the 1987 Nobel Prize in 1961) and the 1987 Nobel Prize in Economics.Economics.

Solow is Professor of Economics at MIT.Solow is Professor of Economics at MIT. Robert Solow won the Nobel Prize for his Robert Solow won the Nobel Prize for his

analysis of economic growth.analysis of economic growth.

Solow’s Economic Growth ModelSolow’s Economic Growth Model

Solow extended the Harrod-Domar Growth Solow extended the Harrod-Domar Growth model by:model by:

Adding labor as a factor of production. Adding labor as a factor of production. Requiring diminishing returns to labor and Requiring diminishing returns to labor and

capital separately, and constant returns to capital separately, and constant returns to scale for both factors combined. scale for both factors combined.

Introducing a time-varying technology Introducing a time-varying technology variable distinct from capital and labor. variable distinct from capital and labor.

The capital-output and capital-labor ratios are The capital-output and capital-labor ratios are not fixed as they are in the Harrod-Domar not fixed as they are in the Harrod-Domar model. These refinements allow increasing model. These refinements allow increasing capital intensity to be distinguished from capital intensity to be distinguished from technological progress.technological progress.

Solow’s Economic Growth ModelSolow’s Economic Growth Model

The basic Solow (1956) model explains The basic Solow (1956) model explains economic growth as a function of labor economic growth as a function of labor augmenting technological progress, population augmenting technological progress, population growth and the saving rate. growth and the saving rate.

It shows that the capital stock per effective It shows that the capital stock per effective unit of labor, k, converges towards a steady unit of labor, k, converges towards a steady state k* at which actual investment is equal to state k* at which actual investment is equal to break-even investment. break-even investment.

The neoclassical Solow model implies that the The neoclassical Solow model implies that the steady state income per capita, (Y/L), depends steady state income per capita, (Y/L), depends positively on the saving rate and negatively on positively on the saving rate and negatively on the population growth and depreciation rate. the population growth and depreciation rate.

MarshallMarshall Although classical and neoclassical Although classical and neoclassical

economists treated knowledge as a economists treated knowledge as a disturbance category in their model disturbance category in their model specifications, Marshall (1965) was specifications, Marshall (1965) was among the first neoclassical among the first neoclassical economists to state explicitly the economists to state explicitly the importance of knowledge in economic importance of knowledge in economic affairs. Marshall (1965) states that affairs. Marshall (1965) states that capital consists in a great part of capital consists in a great part of knowledge and organization, as knowledge and organization, as knowledge is our most powerful engine knowledge is our most powerful engine of production. of production.

New Endogenous Growth New Endogenous Growth ModelsModels

In the eighties, the new endogenous In the eighties, the new endogenous growth theories were developed on growth theories were developed on the assumptions of imperfect the assumptions of imperfect competition between firms, the role competition between firms, the role of history, ideas and accidents, and of history, ideas and accidents, and the appearance of multiple the appearance of multiple equilibria in the markets. The equilibria in the markets. The existence of increasing returns for existence of increasing returns for explaining sustained growth is explaining sustained growth is supported and influenced by the supported and influenced by the research progress made in trade research progress made in trade theory and industrial organization. theory and industrial organization.

New Endogenous Growth New Endogenous Growth ModelsModels

Paul Romer (1986) defends the Paul Romer (1986) defends the endogenous economic growth and endogenous economic growth and increasing returns to scale view from increasing returns to scale view from location and knowledge location and knowledge accumulation perspectives. Some accumulation perspectives. Some theoretical models of economic theoretical models of economic growth, such as Lucas (1988), growth, such as Lucas (1988), Becker, Murphy and Tamura (1990), Becker, Murphy and Tamura (1990), Rebelo (1991) , Barro and Lee Rebelo (1991) , Barro and Lee (1993) and Mulligan and Sala-i-(1993) and Mulligan and Sala-i-Martin (1992), emphasize the role of Martin (1992), emphasize the role of human capital in the form of human capital in the form of educational attainment. educational attainment.

New Endogenous Growth ModelsNew Endogenous Growth Models In economics, endogenous growth theory or In economics, endogenous growth theory or

new growth theory was developed in the 1980s new growth theory was developed in the 1980s as a response to criticism of the neo-classical as a response to criticism of the neo-classical growth model. In neoclassical growth models, growth model. In neoclassical growth models, the long-run rate of economic growth is the long-run rate of economic growth is exogenously determined by either assuming a exogenously determined by either assuming a savings rate (the Solow model) or a rate of savings rate (the Solow model) or a rate of technical progress. technical progress.

Endogenous growth theory tries to overcome Endogenous growth theory tries to overcome this shortcoming by building macro models out this shortcoming by building macro models out of microeconomic foundations. Households are of microeconomic foundations. Households are assumed to maximize utility subject to budget assumed to maximize utility subject to budget constraints while firms maximize profits. constraints while firms maximize profits. Crucial importance is usually given to the Crucial importance is usually given to the production of new technologies, R&D and production of new technologies, R&D and human capital.human capital.

New Endogenous Growth ModelsNew Endogenous Growth Models

During the last decade, several During the last decade, several models of economic growth tend to models of economic growth tend to emphasize the importance of emphasize the importance of investment in intangible assets as a investment in intangible assets as a major source of economic growth. major source of economic growth. Investment in research and Investment in research and development (R&D) and human development (R&D) and human capital has been ascribed to yield capital has been ascribed to yield high social returns for the economy. high social returns for the economy.

New Endogenous Growth ModelsNew Endogenous Growth Models

Economies of scale exist in relation to Economies of scale exist in relation to capital, more specific in the capital, more specific in the production of human capital or production of human capital or knowledge and technology as in knowledge and technology as in Romer (1986) and Krugman (1991). Romer (1986) and Krugman (1991). The marginal product of capital grows The marginal product of capital grows as the stock of capital expands. Put as the stock of capital expands. Put simply, the more we invest in simply, the more we invest in knowledge the more the economy knowledge the more the economy grows. grows.

Economic Growth in MexicoEconomic Growth in Mexico

Consequently, an important topic Consequently, an important topic for Mexican economists who study for Mexican economists who study the interaction of convergence and the interaction of convergence and economic growth, is the economic growth, is the interaction of R&D and human interaction of R&D and human capital in regional growth. capital in regional growth.

Economic ConvergenceEconomic Convergence

The catch-up effect, also called the The catch-up effect, also called the theory of economic convergence, theory of economic convergence, mentions that poorer economies tend mentions that poorer economies tend to grow at faster rates than richer to grow at faster rates than richer economies. economies.

The theory states that all economies The theory states that all economies should in the long run converge in should in the long run converge in terms of per capita income and terms of per capita income and productivity.productivity.

Regional Convergence in Regional Convergence in the U.S.the U.S.

Recent research in the economic Recent research in the economic growth literature, such as Barro and growth literature, such as Barro and Sala-i-Martin (1991) focused on the Sala-i-Martin (1991) focused on the convergence of regional income in convergence of regional income in developed economies. They address developed economies. They address the question of whether poor regions the question of whether poor regions tend to converge toward rich ones. tend to converge toward rich ones. For the U.S. states, they estimate For the U.S. states, they estimate the rate of convergence of per the rate of convergence of per capita personal income from 1880 to capita personal income from 1880 to 1988 to be around 2 percent per 1988 to be around 2 percent per year by looking within or across four year by looking within or across four geographical regions. geographical regions.

Regional Convergence in Regional Convergence in Europe Europe

Barro and Sala-i-Martin also find Barro and Sala-i-Martin also find a rate of economic convergence a rate of economic convergence of about 2 percent per year for of about 2 percent per year for per capita GDP across 73 regions per capita GDP across 73 regions of seven European countries of seven European countries from 1950 to 1985. from 1950 to 1985.

Professor’s Barro 1984 Macroeconomics Professor’s Barro 1984 Macroeconomics textbook remains a standard for explaining textbook remains a standard for explaining the subject, and his 1995 book, with the subject, and his 1995 book, with Columbia University economist Xavier Sala-Columbia University economist Xavier Sala-i-Martin, on Economic Growth is a widely i-Martin, on Economic Growth is a widely cited and read graduate-level textbook on cited and read graduate-level textbook on the theory and evidence concerning long-the theory and evidence concerning long-run economic growth. run economic growth.

Barro's research in the 1990s was mainly Barro's research in the 1990s was mainly focused on the theoretical and empirical focused on the theoretical and empirical determinants of economic growth: he gave determinants of economic growth: he gave fundamental contributions to the theory of fundamental contributions to the theory of endogenous growth, and was a pioneer in endogenous growth, and was a pioneer in the econometric analysis of the main the econometric analysis of the main factors associated with growth in the factors associated with growth in the modern era.modern era.

Economic Convergence and Economic Convergence and Divergence in MexicoDivergence in Mexico

Mallick and Carayannis (1994), Juan Ramón and Mallick and Carayannis (1994), Juan Ramón and Rivera Bátiz (1996), Esquivel (1999), Messmacher Rivera Bátiz (1996), Esquivel (1999), Messmacher (2000), Garcıá Verdú (2002) and Rodrıg and (2000), Garcıá Verdú (2002) and Rodrıg and Sánchez-Reaza (2002) studied absolute Sánchez-Reaza (2002) studied absolute convergence across Mexican states. They all tend convergence across Mexican states. They all tend to conclude that convergence was a feature across to conclude that convergence was a feature across Mexican states up to the mid-eighties, but that this Mexican states up to the mid-eighties, but that this pattern broke down afterwards. Most of these pattern broke down afterwards. Most of these studies, however, used samples including regional studies, however, used samples including regional growth rates at most up to 1996, so that the post-growth rates at most up to 1996, so that the post-NAFTA growth experience is not fully taken into NAFTA growth experience is not fully taken into account nor analyzed explicitly. Moreover, a account nor analyzed explicitly. Moreover, a conditional convergence analysis is not conducted conditional convergence analysis is not conducted in any of these studies, even when this would be in any of these studies, even when this would be the natural step to follow given the findings the natural step to follow given the findings concerning an apparent breakdown of absolute concerning an apparent breakdown of absolute convergence. convergence.

Díaz-Bautista (2003) Díaz-Bautista (2003) Regional Growth Process in MexicoRegional Growth Process in Mexico

What is the role of human capital and What is the role of human capital and R&D in the regional growth process of R&D in the regional growth process of Mexico? Does human capital support Mexico? Does human capital support the transmission of knowledge and the transmission of knowledge and therefore promote future growth? The therefore promote future growth? The empirical work attempts to provide empirical work attempts to provide new insights on the regional pattern new insights on the regional pattern of the interaction of the Mexican of the interaction of the Mexican states over the 1970-2000 period. states over the 1970-2000 period.

For Mexico, there have been some For Mexico, there have been some advances in terms of education and advances in terms of education and R&D. Between 1970 and 1995; the R&D. Between 1970 and 1995; the illiterate population below 15 years of illiterate population below 15 years of age diminished from 26 to 11 percent age diminished from 26 to 11 percent in Mexico. However, the human in Mexico. However, the human capital indicators for Mexico show a capital indicators for Mexico show a low level of matriculation in the low level of matriculation in the elementary and high school age elementary and high school age groups compared to other countries, groups compared to other countries, as well as a low number of as well as a low number of researchers per million inhabitants, researchers per million inhabitants, which are shown in the next table. which are shown in the next table.

Table 1. Education in Mexico, 2000 Cost by primary and secondary student as (% of the GDP per capita) Mexico 11 and 16 OECD Countries 19 and 24 Coverage of university education (%) Mexico 21 Argentina, Chile and Uruguay 28 Average of students by classroom in basic education Mexico 27.2 OECD Countries 18 Elementary, Junior High, baccalaureate Matriculation (%) Mexico 71 Colombia 73 Panama 74 Uruguay 79 Peru and Brazil 80 Researchers per million inhabitants Japan 5,130 Mexico 97 Sources: Secretaría de Educación Pública (2000) and OECD (2000).

Endogenous Growth Productive CapacityEndogenous Growth Productive Capacityin Mexicoin Mexico

Capacity for Innovation in Mexico and South Korea

Mexico S. Korea

Patent applications by National Firms/total patent applications 4.90% 51%R&D expenditure/GDP 0.36% 2.60%Scientists and Engineers per million persons 225 2152Science and Technological Journal Articles 2024 5219Technicians in R&D per million persons 172 576

(average, 1995 to 2000)

Economy based on knowledgeEconomy based on knowledge

Żółkiewski (1999), has mentioned that the Żółkiewski (1999), has mentioned that the XXI Century will be the century of economy XXI Century will be the century of economy based on knowledge, where expenditures for based on knowledge, where expenditures for future development of scientists and future development of scientists and researchers which includes expenditure on researchers which includes expenditure on R&D and expenditure on education will drive R&D and expenditure on education will drive economic growth. economic growth.

When making international comparisons of When making international comparisons of R&D activities, Mexico ranks among the R&D activities, Mexico ranks among the lowest in OECD countries. lowest in OECD countries.

Table 2 .Gross domestic expenditures on R&D as % of GDP (1994 – 1999). as % of GDP USD per capita (current PPP) Countries

1994 1999 Countries

1994 1999

Sweden 3.27 c) 3.70 b) USA 650.3 892.5

Finland 2.29 3.09 Sweden 571.6 c) 773.8 b)

Japan 2.84 3.04 a) Japan 602.5 732.6 a)

USA 2.42 2.65 Finland 381.9 707.1

Korea 2.44 2.55 a) Germany 454.8 563.0

Germany 2.26 2.38 Denmark 344.5 c) 520.6

France 2.34 2.18 a) Iceland 270.6 495.7

Denmark 1.74 c) 1.99 Norway 370.4 c) 480.9

Netherlands 2.00 c) 1.95 a) Netherlands 382.3 470.8 a)

Iceland 1.38 1.88 France 446.8 461.6 a)

Belgium 1.74 1.84 b) Austria 312.8 450.7

United Kingdom 2.07 1.83 a) Belgium 360.3 419.5 b)

Austria 1.53 1.82 Canada 348.2 419.0

Norway 1.73 1.73 United Kingdom 372.7 395.8 a)

Canada 1.67 1.58 Korea 286.1 365.7 a)

Australia 1.61 1.49 a) Australia 312.9 360.9 a)

Ireland 1.31 1.39 b) Ireland 209.2 296.1 b)

Czech Republice) 1.10 1.27 Italy 198.3 231.0

New Zealand 1.02 c) 1.13 b) New Zealand 153.0 c) 199.4 b)

Italy 1.05 1.04 Czech Republice) 120.9 167.4

Spain 0.85 0.90 Spain 115.5 163.5

Poland 0.78 0.75 Portugal 78.1 d) 95.2 b)

Hungary 0.89 0.68 Hungary 74.0 75.9

Portugal 0.57 d) 0.62 b) Greece 52.5 c) 68.7 b)

Greece 0.48 c) 0.51 b) Poland 44.6 63.6

Turkey 0.36 0.49 b) Turkey 19.1 31.3 b)

Mexico 0.29 0.34 b) Mexico 20.5 26.0 b)

Switzerland 2.73 e) Switzerland 685.1 e) .

European Union 1.84 1.81 a) European Union 339.9 385.1 a)

OECD 2.10 2.18 a) OECD 399.7 469.4 a) a) 1998 b) 1997 c) 1993 d) 1995 e) 1996 Source: Main Science and Technology Indicators, OECD.

In terms of gross domestic In terms of gross domestic expenditures on R&D, Mexico expenditures on R&D, Mexico ranks as one of the last countries ranks as one of the last countries within OECD in terms of the ratio within OECD in terms of the ratio of domestic expenditures on R&D of domestic expenditures on R&D activity to GDP and in terms of activity to GDP and in terms of expenditures per capita.expenditures per capita.

intraregional human capital and R&D spillovers in the growth process. In this study, the basic

model of human capital, R&D spillovers and regional economic growth is based on Romer’s

(1996) endogenous model of technological change and Aghion and Howitt (1998) growth model.

The model starts with the simple aggregate production function in region i that is given by the

following equation:

Yi = K H (Ly) 1- H (LA 1- ) (1)

Where K is capital and H (L) is the state specific human capital stock (regional raw labor force) or

the amount of human capital (raw labor) available in all regions, and 0 < , < 1, and > 0. The

regional raw labor force is used to produce either output or ideas ( L = LY + LA ), where LY is

the labor used to produce output and LA is the labor used to produce ideas. The production

process is assumed to generate positive knowledge externalities.

Consider the effects of human capital in the model, in the case where there is no physical capital.

The higher the average level of human capital in the economy, the greater the incidence of

knowledge spillovers to raise the marginal productivity of human capital across regions. In other

words, the external human capital produces a public good effect that adds to the region-specific

stocks of L and H. Human capital in region i is paid its private marginal product (MPH), thus

MPH = (Hi/ Li) -1 (H/L) (2)

In equilibrium, Hi/Li = H/L, so the marginal product of human capital can be rewritten as

MPH = (H/L) -(1- - ) (3)

The regional aggregate production function can be written by substituting (3) in (1) as:

LHY )1()( (4)

Defining regional production per capita as y Y/L and regional production per capita as h H/L,

allows us to define the regional production function in intensive form given by:

hy (5)

Taking the time derivatives of both sides of h yields the following equation:

nhshh (6)

The equation gives the standard result of the Solow model, where s is the level of savings in the

economy. The economy will converge to a situation where actual investment per capita is equal to

break-even human capital investment per capita. The steady state human capital stock h* is then

given by:

ns

h1

1

(7)

The steady state per capita human capital stock h* is an increasing function of and therefore the

size of the regional spillover effects. The regional spillover externality implies that, absent

government intervention, private yi* = f(hi*) will be suboptimally low because individual regions

do not internalize the learning by doing externalities that their investment produces for other

regions.

The Empirical AnalysisThe Empirical Analysis

In this study, the analysis is focused on the In this study, the analysis is focused on the regional interaction and convergence of regional interaction and convergence of human capital and R&D. The identification of human capital and R&D. The identification of the regional interaction structure is based on the regional interaction structure is based on a generalized growth regression analysis a generalized growth regression analysis that focuses on the relationship between that focuses on the relationship between regional income per capita growth and the regional income per capita growth and the human capital and R&D activity. The human capital and R&D activity. The dependent variable is the average annual dependent variable is the average annual income per capita growth rate between 1970 income per capita growth rate between 1970 and 2000 obtained from INEGI. and 2000 obtained from INEGI.

-2,5

-2,0

-1,5

-1,0

-0,5

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

5.000 10.000 15.000 20.000 25.000 30.000 35.000

Per cápita Income

Rate

of

Gro

wth

(%

)

1. Aguascalientes2. B. California3. B. Calif. Sur4. Campeche5. Coahuila de Z.6. Colima7. Chiapas8. Chihuahua9. D. F.10. Durango11. Guanajuato12. Guerrero13. Hidalgo14. Jalisco15. México16. Michoacán de O.17. Morelos18. Nayarit19. Nvo. León20. Oaxaca21. Puebla22. Querétaro de A.23. Quintana Roo24. San Luis Potosí25. Sinaloa26. Sonora27. Tabasco28. Tamulipas29. Tlaxcala30. Veracruz31. Yucatán32. Zacatecas

País

1

6

3

4

5

2

7

8

9

10

11

12

13 14

15

16

17

18

19

20

21

22

24

25

26

27

28

29

30

3132

23

Sector 2:Dynamic StatesWith low per capita income

Sector 1: Dynamic States With high income per capita

Sector 3: Static States with Low per capita income

Sector 4: Static States with high per capita income

Regional per Capita Rate of Growth in México (1994-2000),and per capita Income

Fuente: Informe sobre Desarrollo Humano México 2002

Regional Disparities in DevelopmentRegional Disparities in Development

México’s Regional HDI:México’s Regional HDI:Contrasts between the North and SouthContrasts between the North and South

Fuente: Informe sobre Desarrollo Humano México 2002

To test the convergence hypothesis of income per capita and to estimate the annual

velocity of convergence, the following non lineal model is derived from the aggregate

production function:

(1/T) Ln (Yi, t+T / Yi t ) = - (Ln (Y it ) ) ( 1 - e - t) (1/T) + educational variables +

R&D variable + uit (8)

where:

Yi t is the level of income per capita in region i.

t is the initial time.

T is a period of time.

is the velocity of convergence or the average annual rate at which economies get closer to the steady state. The higher the , the lower the period of time necessary for the system to reach the long run equilibrium. If < 1 the system diverges.

The data on education used in the present study The data on education used in the present study comes from the Secretaría de Educación Pública comes from the Secretaría de Educación Pública (2000), INEGI and the Ministry of Education website. (2000), INEGI and the Ministry of Education website. The educational variables used to condition the The educational variables used to condition the data are the following: data are the following:

Illiterate, refers to the percentage of people that Illiterate, refers to the percentage of people that know how to read and write.know how to read and write.

Elementary is a variable that show the percentage Elementary is a variable that show the percentage of people with elementary studies completed.of people with elementary studies completed.

Some junior high is a variable that refers to the Some junior high is a variable that refers to the percentage of people that have more that percentage of people that have more that elementary studies, but haven’t finished junior high. elementary studies, but haven’t finished junior high.

Finished Junior High refers to the percentage of Finished Junior High refers to the percentage of people that completed a level of education in the people that completed a level of education in the National Educational System.National Educational System.

High School refers to the percentage of people that High School refers to the percentage of people that completed a level of education in the National completed a level of education in the National Educational System.Educational System.

College refers to the percentage of people that are College refers to the percentage of people that are in college or completed a level of education in in college or completed a level of education in University studies.University studies.

R&D activity is measured by the number of R&D activity is measured by the number of Researchers per state at the SNI (Sistema Nacional Researchers per state at the SNI (Sistema Nacional de Investigadores) in the year 2000 provided by de Investigadores) in the year 2000 provided by CONACYT. CONACYT.

Table 3. Estimated Results of the Velocity of Convergence in Income Per Capita. Conditional to R&D and human capital variables in Mexico Dependent Variable: Average annual income per capita growth rate. Period 1970-1993 1970-1985 1970-1980 1970-2000 1985-1993 1970-2000

-0.008 (-0.6671)

0.031 (2.0441)

0.046 (1.9122)

-0.023 (0.405)

-0.010 (-.1548)

15.98 (1.98)

0.014 * (1.628)

0.031* (2.6696)

0.018 (1.2769)

0.032 * (2.546)

-0.014 (-0.619)

0.45* (2.38)

Elementary .10572 (.04575)

0.026328 (0.0545)

-0.08076 (0.0886)

0.017* (2.26)

-0.8378 (0.1759)

0.150 (1.78)

Some Junior High -0.36605 (.15247)

-0.176821 (.18183)

-0.199977 (.29560)

0.0015 (0.141)

0.1792 (0.3266)

-0.149 (1.78)

Finished Junior High

6.01545 (1.8517)

2.43713 (2.1678)

5.120347 (3.5241)

0.0028 (0.186)

-0.4447 (0.7874)

0.162 (1.98)

High School 1.8517 (.56078)

0.704419 (0.6687)

0.860725 (1.0871)

0.006 (0.372)

0.11467 (0.4351)

0.145 (1.98)

College -0.79644 (0.3394)

-0.187491 (0.4048)

-0.456506 (0.6581)

0.225 (1.495)

-0.2228 (0.4026)

0.147 (1.71)

Illiterate 0.017202 (0.0098)

-0.002294 (0.0117)

0.012342 (0.0191)

0.003 (0.365)

0.00067 (0.0056)

0.161 (1.962)

R&D Activity

0.003

(1.29)

Half life

50.4 22 37.8 31.7 49.8 32.5

R2 adjusted 0.401

0.440 0.221 0.072 0.068 0.320

T (years) 23 15 10 30 8 30

*Significance at 5% level. T stats in parenthesis.

Empirical ResultsEmpirical Results The results are based on generalized least The results are based on generalized least

squared regressions with and without R&D squared regressions with and without R&D activity and by level of human capital. For the activity and by level of human capital. For the period 1970 to 1993, the conditional period 1970 to 1993, the conditional convergence parameter is the 1.4% positive convergence parameter is the 1.4% positive and significant, while for the period 1970 to and significant, while for the period 1970 to 2000 the parameter is close to 3.2% without 2000 the parameter is close to 3.2% without considering the R&D activity. The results considering the R&D activity. The results indicate that the states of Mexico can reach a indicate that the states of Mexico can reach a level of long run income. For the period 1985-level of long run income. For the period 1985-1993, the convergence parameter is negative 1993, the convergence parameter is negative 1.4%, but not significant indicating that states 1.4%, but not significant indicating that states of Mexico were diverging in this short period of Mexico were diverging in this short period from the long run level in income per capita. from the long run level in income per capita.

Empirical ResultsEmpirical Results

The analysis made suggests that The analysis made suggests that the convergent behavior across the convergent behavior across Mexican states’ per-capita output Mexican states’ per-capita output that was observed before 1985 that was observed before 1985 broke down afterwards. As a broke down afterwards. As a consequence, a process of consequence, a process of increasing economic increasing economic differentiation across regions differentiation across regions seems to have characterized the seems to have characterized the country from 1985 to 2000. country from 1985 to 2000.

Contribution to National GDP in Contribution to National GDP in Mexico 1993 and 2002 (%)Mexico 1993 and 2002 (%)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1940 1950 1960 1970 1980 1988 1990 1993 1995 2000

Chiapas

Guerrero

Oaxaca

Nuevo León

Sonora

Chihuahua

Per Capita Income Divergence between Border States and rest of Mexico

(% relation to Mexico City)

Empirical ResultsEmpirical Results Only for the regression with the R&D, no Only for the regression with the R&D, no

regional significance of human capital can be regional significance of human capital can be detected. The result seems to be generated detected. The result seems to be generated by the concentration of R&D in the Mexico by the concentration of R&D in the Mexico City area, where more than 50% of all R&D City area, where more than 50% of all R&D activity is concentrated. activity is concentrated.

For the period 1970-1993 the half-life was For the period 1970-1993 the half-life was 50.4 years. This is the number of years to 50.4 years. This is the number of years to cover half the distance of the logarithms of cover half the distance of the logarithms of income per capita. However, the regressions income per capita. However, the regressions yields non-significant coefficient for all yields non-significant coefficient for all explanatory variables, except for the explanatory variables, except for the convergence coefficient. The convergence convergence coefficient. The convergence coefficient confirms the findings of previous coefficient confirms the findings of previous studies on conditional ß-convergence done in studies on conditional ß-convergence done in Mexico, which include Esquivel (1999), Díaz-Mexico, which include Esquivel (1999), Díaz-Bautista (2000) and Messmacher (2000).Bautista (2000) and Messmacher (2000).

Empirical ResultsEmpirical Results

The empirical analysis provides empirical The empirical analysis provides empirical evidence for the hypothesis that R&D evidence for the hypothesis that R&D spillovers are regionally bounded and do not spillovers are regionally bounded and do not constitute a significant source of regional constitute a significant source of regional economic growth. The huge agglomeration in economic growth. The huge agglomeration in Mexico City can be assessed as the main Mexico City can be assessed as the main origin of the non-existent R&D spillovers at a origin of the non-existent R&D spillovers at a regional scale in Mexico. As for human regional scale in Mexico. As for human capital, elementary education is starting to be capital, elementary education is starting to be an important factor that explains regional an important factor that explains regional economic convergence in Mexico due to the economic convergence in Mexico due to the similar coverage at the state level. similar coverage at the state level.

Maquiladoras and the Export Maquiladoras and the Export Oriented Northern Border Oriented Northern Border

RegionsRegions

A Maquiladora is an export oriented A Maquiladora is an export oriented assembly plant in Mexico, especially assembly plant in Mexico, especially one along the border between the one along the border between the United States and Mexico, to which United States and Mexico, to which foreign materials and parts are shipped foreign materials and parts are shipped and from which the finished product is and from which the finished product is returned to the original market.returned to the original market.

A maquiladora or maquila is a factory A maquiladora or maquila is a factory that imports materials and equipment that imports materials and equipment on a duty-free and tariff-free basis for on a duty-free and tariff-free basis for assembly or manufacturing and then assembly or manufacturing and then re-exports the assembled product, re-exports the assembled product, usually back to the originating country.usually back to the originating country.

% of employment in export oriented industry by state

Number of employees in export oriented industry by states (thousands)

B.C. 38% (216)

N.L. 18% (166)

Tamps 30% (142)

Coah. 26% (125)

Son. 23% (75)

Chih. 40% (251)

Tlax. 25% (19)

Q.R. 3% (4)

Ags. 19% (30)

Gto. 19% (63)

Edo. Mex. 19% (188) Mor. 14% (18)

D.F. 13% (282) Oax. 8% (8)

Ver. 10% (37)

Pue. 21% (82)

Qro. 20% (86)

Zac. 11% (10)

Nay. 5% (3)

Mich. 9% (19)

Dgo. 12% (20)

Sin. 9% (24)

Jal. 11% (94)

B.C.S. 12% (9)

Yuc. 18% (35)

S.L.P. 20% (37)

Tab. 7% (6)

Camp. 8% (7)Gro. 7% (8)Chis. 5% (5)

Col. 4% (3)

Hgo. 9% (13)

THE IMPORTANCE OF THE EXPORT THE IMPORTANCE OF THE EXPORT ORIENTED INDUSTRY BY STATES IN ORIENTED INDUSTRY BY STATES IN MEXICOMEXICO National Average 15% (2,045)

SOURCES: SE e INEGI (2003)

Border Average 29.1% (975)

ConclusionsConclusions The results confirm the empirical evidence on The results confirm the empirical evidence on

bounded non-existent knowledge spillovers, bounded non-existent knowledge spillovers, shown by the increased concentration of R&D shown by the increased concentration of R&D activities in Mexico City. In other words, the activities in Mexico City. In other words, the findings suggest that regional growth is not findings suggest that regional growth is not determined by regional R&D activity, due to determined by regional R&D activity, due to the lack of expenditures on R&D at the the lack of expenditures on R&D at the national and regional level, and to the national and regional level, and to the increase concentration of R&D in the Mexico increase concentration of R&D in the Mexico City Metropolitan Area. City Metropolitan Area.

The study confirms the qualitative hypothesis The study confirms the qualitative hypothesis that R&D knowledge and human capital may that R&D knowledge and human capital may spill over, but the regional extent of such spill over, but the regional extent of such knowledge spillovers may be bounded in knowledge spillovers may be bounded in Mexico.Mexico.

ConclusionsConclusions The concept of economic The concept of economic

convergence is the tendency for convergence is the tendency for income differences to narrow over income differences to narrow over time. time.

It can inform policy makers of the It can inform policy makers of the need for development policies to need for development policies to promote equity and economic promote equity and economic growth. growth.

It also mentions that economic It also mentions that economic disparities between regions may disparities between regions may diminish naturally in the long run.diminish naturally in the long run.

ConclusionsConclusions For the last 50 years, the neoclassical For the last 50 years, the neoclassical

growth model remained the most important growth model remained the most important model of regional economic growth. Solow model of regional economic growth. Solow (1956) shows how growth in the capital (1956) shows how growth in the capital stock, labor force and exogenous advances stock, labor force and exogenous advances in technology interact and how they affect in technology interact and how they affect the growth of output. The Solow growth the growth of output. The Solow growth model shows that in the long run, an model shows that in the long run, an economy’s rate of saving determines the economy’s rate of saving determines the size of its capital stock and thus its level of size of its capital stock and thus its level of production. production.

Starting in the 1980’s, more sophisticated Starting in the 1980’s, more sophisticated growth models have been developed. Unlike growth models have been developed. Unlike the neoclassical model, technological the neoclassical model, technological change is not assumed to be exogenous. change is not assumed to be exogenous.

ConclusionsConclusions The new endogenous growth models explain The new endogenous growth models explain

the sources of technologically driven the sources of technologically driven productivity growth. In particular, the productivity growth. In particular, the accumulation of knowledge plays a key role in accumulation of knowledge plays a key role in driving productivity growth in these models. driving productivity growth in these models.

One important implication of the new studies One important implication of the new studies is that the location of research and is that the location of research and development (R&D) activity may matter. If development (R&D) activity may matter. If there are significant agglomeration effects there are significant agglomeration effects associated with R&D activity, then the associated with R&D activity, then the benefits of R&D are largely captured by the benefits of R&D are largely captured by the region in which R&D activity takes place. region in which R&D activity takes place.

Economic Policymakers need to have a Economic Policymakers need to have a greater appreciation for the role of ideas in a greater appreciation for the role of ideas in a region’s or nation’s economic development. region’s or nation’s economic development.

ReferencesReferences Barro R., Sala-i-Martin X (1991) Convergence across states and regions. Barro R., Sala-i-Martin X (1991) Convergence across states and regions.

Brookings Papers on Economic Activity 1991: 107–182Brookings Papers on Economic Activity 1991: 107–182

Barro R., Sala-i-Martin X (2004) Economic growth (2nd ed). The MIT Press, Barro R., Sala-i-Martin X (2004) Economic growth (2nd ed). The MIT Press, Cambridge, MA.Cambridge, MA.

Díaz-Bautista, Alejandro (2003), “Convergence and Economic Growth Díaz-Bautista, Alejandro (2003), “Convergence and Economic Growth Considering Human Capital and R&D Spillovers”, Mexican Journal of Economics Considering Human Capital and R&D Spillovers”, Mexican Journal of Economics and Finance, ITESM, Volume 2, Number 2.and Finance, ITESM, Volume 2, Number 2.

Romer, Paul M. (1986). Increasing Returns and Long-Run Growth , Journal of Romer, Paul M. (1986). Increasing Returns and Long-Run Growth , Journal of Political Economy, v. 94, pp. 1002-1037.Political Economy, v. 94, pp. 1002-1037.

Solow, Robert, "A Contribution to the Theory of Economic Growth." Quarterly Solow, Robert, "A Contribution to the Theory of Economic Growth." Quarterly Journal of Economics 70 (February 1956): 65-94.Journal of Economics 70 (February 1956): 65-94.

Solow, Robert, "Technical Change and the Aggregate Production Function." Solow, Robert, "Technical Change and the Aggregate Production Function." Review of Economics and Statistics 39 (August 1957): 312-20.Review of Economics and Statistics 39 (August 1957): 312-20.

Economic Policy in Latin AmericaEconomic Policy in Latin America

Convergence and Convergence and Economic Growth Economic Growth considering Human considering Human Capital and R&D Capital and R&D Spillovers in MexicoSpillovers in Mexico

Alejandro Díaz-Bautista, Alejandro Díaz-Bautista, Ph.D.Ph.D.

Professor of Economics and Researcher at COLEFVisiting Research Fellow and Guest Scholar, Center for U.S.-Mexican Studies, University of California San Diego (UCSD). Graduate School of International Relations & Pacific Studies IR/PS University of California, San Diego