Institutions and Human Capital Accumulation: The Effect of Economic Policies and Government Stability

Kézia de Lucas Bondezan1

Joilson Dias2

Francisco José Veiga3

Abstract

The objective of this paper is to test the influence of the institutions quality on human capital accumulation process. This paper follows the theoretical developments developed by Dias and McDermott (2006) and Dias and Tebaldi (2012) that establishes a micro-foundation link between human capital accumulation and institutions quality. Using a panel data series from 1960 to 2010 we observe that the economic policies and institutions quality do affect long-term human capital accumulation process in the economy. Better institutions do foster human capital growth.

Key words: Human Capital; Institutions; Growth Economic

Resumo

O objetivo desse artigo é testar a influência da qualidade das instituições no processo de acumulação de capital humano. O artigo segue o modelo teórico desenvolvido por Dias e McDermott (2006) e Dias and Tebaldi (2012) que estabelecem os microfundamentos que ligam o capital humano a qualidade das instituições. Usando uma série de dados em painel para o período de 1960 a 2010 observamos que a economia política e a qualidade das instituições afetam o processo de acumulação de capital humano no logo prazo. Melhores instituições tornam mais rápido o crescimento do capital humano.

Palavras chave: Capital humano; Institutições; Crescimento Econômico

Área: 02 JEL: 043

1 Professora adjunta – Departamento de Economia da Universidade Estadual de Maringá. klbondezan@uem.br

2 Professor titular – Programa de Pós graduação em Economia da Universidade Estadual de Maringá. jdias@uem.br

3 Professor catedrático – Escola de Economia e Gestão – Universidade do Minho – Braga Portugal. fjveiga@eeg.uminho.pt

1- Introduction

The literature about the role of institutions starts as the causality running from education to institutions quality as proposed by Lipset (1960). The empirical test of this idea was reported in the paper by Gyimah-Brepong et al. (1998). The authors investigated the role of human capital on political instability in Latin America. They found a positive effect that increasing nations human capital increases political stability. However, they also reported a regression from political instability causing human capital finding a negative and significant coefficient. Therefore, what they found was an existing relationship between the two, but not causality.

The paper by Hall and Jones (1999) took a step further on this matter by considering the social infrastructure of the country (institutions) as the main cause in explaining productivity levels. The authors conceptualized the social infrastructure as the institutions and government policies that act in the economy and determine the pattern of productivity by fostering physical, human and technological development. Using instrumental variables, they concluded that institutions can robustly explain the differences in income levels between countries by means of the levels productivity of human and physical capital as well as technology.

However, the empirical test done by Glaeser et al. (2004) found that causality may indeed run from education to democracy as the best measure for institution quality of any country. Again the causality issue between education and institutions returned to the main scene.

Acemoglu et al. (2005) took the Glaeser et al. (2004) empirical tests to a step further. They run several regressions where causality runs from education to institutions. According to the authors once you control for fixed effects the causality running from education to democracy fail to be true. As a result, the authors concluded that the omitted variables present in the fixed effects may be the cause of both democracy and education changes in the economies. Their main claim was that the historical development path might be cause for both improvements as predicted by Acemoglu et al. (2005 paper.

Following the same line on the importance of institutions for knowledge creation we have the paper by Coe et al. (2009) which find evidence that better institutions increase the returns to R&D investments. According to them it also increases the benefit from international R&D spillovers and human capital formation. Moreover, Seck (2011) showed that countries with strong institutions experience a significant increase in the absorption of the international R&D spillovers. Recently, Aisen and Veiga (2013) estimated the effect of political instability on economic growth and found an inverse relationship between political instability and economic growth. When investigating the transmission channels of political instability, the authors found that it affects productivity growth and the accumulation of physical and human capital. Regarding the impact of democracy on economic growth, the authors found a small negative effect.

The problems with the above views were that they do not offer consistent economic model with a micro-foundation that show the relationship between institutions and human capital and their causality. Plus, their empirical test did

not attempt to measure the direct effect of institutions on overall society´s human capital as done by Acemoglu et al. (2004).

The first paper to build a micro-foundation showing the role of institutions on human capital accumulation process was the one written by Dias and McDermott (2006). Their economic model proved the existence of a two-step process. The quality of institutions, given by their economic policies, influences the level of entrepreneurs in the economy. Hence, the level of entrepreneurs generates a higher or lower demand for human capital depending upon such economic policies. The empirical test in the paper indeed showed this to be case.

This new micro-foundation venue was further developed by Dias and Tebaldi (2012) in a more detailed model. The authors´ model showed a micro-foundation that institutions quality might influence the human capital expected rate of return. Therefore, they offered an explanation of a more long run view of the human capital accumulation process or how societies create its own historical development path in accordance to Acemoglu et al. (2004) paper findings. However, the authors did not test directly the institutions quality on human capital accumulation, but rather on explaining long run economic growth.

In this paper we follow the Dias and Tebaldi (2012) model; however, we test the direct effect of institutions quality on human capital accumulation process. So, this paper differs from the other in three aspects. First, our empirical model try to find a more stable development path that explains long-term growth through the influence of institutions on human capital accumulation process; second, the empirical definition of institutions encompasses political stability and country risk; third, we test an empirical growth model using institutions as instrumental variables for human capital and a broader definition of structural human capital.

The remaining of de paper is organized as follow: Section 2 develops an economic model that shows how institutions operate at a micro-level in the economy; the model was developed by Dias and Tebaldi (2012). Section 3 presents an econometric analysis that uses cross-country panel data from 1960 to 2010. Section 4 present results and discussion, following the conclusion.

2- Theoretical Model

The theoretical model to be developed in this paper follows the proposal of Dias and Tebaldi (2012), in which they emphasized the importance of the educational sector in the economy. Following the models of Uzawa (1965) and Lucas (1988), the authors have created a function of human capital accumulation based on the following assumptions. The models assume that population N grows at the constant rate n. The population is comprised of educated (h ) and non-educated people(n), that is,N=h+n. There are two main sectors in the economy: final goods and education

An important consideration in this model is that the final goods sector demands labor of educated and non-educated workers, who are paid according to their marginal product. Due to this fact, educated workers have higher income and are more productive. There is an incentive for the non-educated to invest in education in order to obtain higher wages.

The education sector also uses the work of educated and non-educated workers to create human capital. In this sector, the work is remunerated according to the social return. The main contribution of this model is the addition of the education sector that elevates the product of the economy to generate income. The derivation of the model can be expressed as follows:4

The goods sector

The production function of the goods sector depends on educated and non-educated labor:

y (g )=A (an)1−β(ah)β=aAn1−βhβ (1)

Where: y (g )is the final product; n is non-educated labor ; h is educated labor and A technology.

The real wages of educated employees who work in the final goods sector is: wh

g=(W hg/AP), where W h

g,is the nominal wages for education and P the price level.

The firm's profit function is given by:

π=an1−βhβ−whgh−wn

gn ,(2)

Where: whandwnare the effective real wages of the educated and non-educated.

Whereas the technological level A is given, we have the following wage equation:

whg=βan1−βhβ−1(3)

wng=(1−β )an− βhβ (4 )

The income is distributed between educated and non-educated.

(wh

wn)=( β

1−β ) nh (5)

Equation (5) shows that when uneducated workers become educated, there is a continuous reduction in the wage rate. As Dias and Tebaldi (2012 p. 305) “that equation also suggests that a historical sef-reinforcing mechanism explaining income distribution across qualified and non-qualified labor could exist because the higher the initial stock of qualified individuals, the higher the probability that a person will be included in the human capital stock rather than in the non-qualified work-force, and this trend may persist over time.”

4 For or a more detailed description see Dias and Tebaldi (2012).

The educational sector

The model assumes that the non-educated workers can be trained and become educated workers. This sector’s production function is given by:

y (e )=γ [ (1−a )n ]1−β [ (1−a )hβ ] ,(6)Where: 0≤ γ ≤1 is institutional quality, so that a larger γ implies better

institutions. In this specification, we can say that institutions affect the productivity of educated workers in the process of transferring knowledge to the uneducated.

Combining equations (6) and (1) gives:

y (e )=γ ( 1−aa ) y (g)A

.(7)

This implies that technological advancement makes the process of creation of human capital more complex, since it requires a larger amount of product to create increased human capital.

The model also assumes that whe= y (e)/h, i.e., the return on human

capital employed in the education sector is the average cost of producing effective human capital. It also implies an important role of the quality of institutions in determining social return.

Dias and Tebaldi (2012) also consider that there is perfect mobility across sectors, so that workers can move from the goods sector to the education sector and vice versa. Using this condition together with equation (3), we arrive at the equation (8).

a= γγ+ β

(8)

Substituting equation (6) to (7), we obtain

y (e )=( γβγ+β )n1−β hβ (9)

This equation implies that improvements in the quality of institutions increases the productivity of inputs allocated to the education sector, i.e.,

(∂ y ( e )∂ y

)>0

The decision to accumulate human capital

Dias and Tebaldi (2012) develop a link between the individual decision to accumulate human capital and market conditions. The representative agent decides whether or not to invest in human capital and that decision depends on the costs incurred in the investment of that capital and on expected income flows, i.e., on the expectation of future gains.

W=∫t

∞

whg e

−( rγ )(s−t)ds=∫

t

∞

( γβγ+β )n1−βhβ−1e

−( rγ )(s−t )ds, (10)

Where: r is the rate of return of the market; r / γis the effective discount set by institutional inefficiency created by institutional arrangements. As r / γ is the investment in education, its inverse can be interpreted as the actual rate of return of education.

The opportunity cost required for n to become h is also affected by time (t−T ). Whereas the costs are updated from time to time, at the rate φ, then:

C=∫T

t

[( γγ+β ) (1−β )n−βhβ+( γβ

γ+β )n−βhβ ] x eφ ( s−t ) ds=¿

∫T

t

[( γγ+β )n− βhβ ]eφ(s−t)ds (11)

The individual chooses to accumulate human capital if the flows of discounted future income is > or = to the cost of human capital accumulation. Assuming that, at the margin, non-educated individuals choose to acquire skills to become educated, then:

∫t

∞

( γβγ+ β )n1− βhβ−1 e

−( rγ )( s−t )ds

¿∫T

t

[( γγ+β )n−βhβ]eφ (s−t ) ds(12)

Integrating both sides of the equation regarding the s, assuming T→−∞

hn=(φβγr )(13)

This equation implies that there is an optimal ratio of educated to non-educated labor and that this ratio depends on: the quality of institutions (γ); the share of human capital in the economy (β); and the discount rate and the rate assigned to the cost return of capital, (φ, r).

Good institutions are associated to the ratio of educated to non-educated, this is, to the share of the most educated population in the economy.

Substituting (3) into (6):

wh

wn=( β

1−β ) rφβγ

(14 )

Equation (14) suggests that improvements in institutions reduce the wage and income inequality between educated and non-educated.

The dynamic path of capital accumulation can be obtained by solving equation (13) for n and introducing it into the equation (9).

h= y (e )=((γβ )β

γ+β )( r∅ )1− β

h(15)

The interpretation for this equation is that the balanced growth rate of human capital depends on deeper parameters such as the quality of the institutions (γ ¿, the share of human capital in the economy (β ¿and the effective rate of return on physical capital ¿). The close to one is the parameter of the quality of the institutions (γ ) more human capital gets accumulated in the economy. So, overtime lower quality institutions harm the historical development path by creating a low incentive for knowledge accumulation. As a result the economy shows a historical low level of human capital compared to the others.

Unlike in Lucas (1988) where human capital accumulation growth rate is an exogenous here the growth rate of human capital accumulation function is endogenous and depends upon the quality of institutions.

The model also implies that income can be generated both in the final good and education sector. The production function is:

Y= y (g )+ y (e )=ω (β+A )h(16)

Where ω=¿

Solving the optimization problem in a balanced growth path, output per capita and consumption per capita must grow at the same rate. The condition implies that:

gy=yy= cc=1σ

(ωrv−η−ρ )(17)

Again the economy balanced growth rate depends on the same parameter as the human capital accumulation function, equation (13).

Our focus on the empirical will be equations (13) and (17). The idea is to test

3- Data and Empirical Model

This section presents a model that examines an empirical relationship among institution, input accumulation (human and physical capital) and economic growth. The major economic problems this relationship is endogeneity and heterogeneity in data. This can solve through a dynamic panel data model, where differences among countries are captured across over time. This model also allows obtaining the long-term average growth rate as the constant in the model. (see Kenny and Williams, 2001; Dias and Tebaldi 2012, Tebaldi and Elmslie 2013).

Departing from a growth regression:y¿− y i ,t−τ+w i ,t δ+ηi+ξ t+ε¿(18)

We can re write the equation in a way that clearly shows that we have a dynamic model:

y¿=~β y i , t−τ+wi , t δ+η i+ξt+ϵ ¿,~β=1+β (19)

Estimations of this model by OLS, even with fixed or random effects, are inconsistent, since the errors term is correlated with the individual effect. Following Arellano and Bond (1991), we can eliminate the individual effect by taking first differences:

y¿− y i ,t−τ=~β ¿¿

(20)

The error term is correlated with the lagged dependent variable and some variables in vector W may be endogenous or pre-determined, the model is estimated by instrumental variables.

Arellano and Bond (1991) suggest the use of the following instruments: Lagged levels (2 or more periods) of the dependent variable and of the endogenous explanatory variables; Lagged level (1 or more periods) of the pre-determined explanatory variables and the exogenous variables can be used as their own instruments. This model is knows as Difference-GMM and will be used in this article.

The data sample covers 128 countries, using consecutive and non-overlapping periods of five years, for the period 1960-2010. The descriptive statistics used in this article are presented in Table 1.

Table 1 – Descriptive Statistics 

Variable Obs MeanStd. Dev. Min Max Source

Growth of GDP per capita 1579 0.002 0.004 -0.025 0.026 PWT 8.1GDP per capita (log) 1746 8.316 1.225 5.031 11.576 PWT 8.1Human capital (log) 1898 3.440 4.244 0.000 26.364 PWT 8.1Growth of Human Capital 1181 0.012 0.015 -0.048 0.103 PWT 8.1Growth of Physical Capital 1569 0.023 0.032 -0.084 0.385 PWT 8.1Government (% GDP) 1746 0.200 0.118 -0.152 0.927 PWT 8.1Investment 1746 0.203 0.116 -0.107 1.577 PWT 8.1

(%GDP)Exports - Imports (%GDP) 1746 -0.069 0.309 -6.393 1.090 PWT 8.1Population Growth 1588 0.002 0.001 -0.005 0.015 PWT 8.1Cabinet Change 1916 0.456 0.397 0.000 3.000 CNTSIndex Chain 939 6.035 1.332 1.782 9.141 EFWEconomic Freedom 893 5.551 2.005 1.143 9.625 EFWPolity2 1653 0.678 7.335 -10.000 10.000 POL. IVEthnic Tensions 797 3.902 1.429 0.000 6.000 ICRGReligion in Politics 797 4.565 1.337 0.000 6.000 ICRGBureaucracy Quality 797 2.129 1.185 0.000 4.000 ICRGCorruption 797 3.044 1.346 0.000 6.000 ICRGGovernment Stability (% GDP) 797 7.536 2.069 1.000 11.625 ICRGStructural Institutions 1898 5.806 40.335 0.000 1390.80 BLStructural Institutions2 1898 19.505 245.835 0.001 10306.5 BLStructural Institutions3 1898 16.813 228.154 0.000 9604.98 BLStructural Institutions4 1898 0.541 2.624 0.000 48.438 BL

Sources: BL: Barro and Lee (2013); CNTS: Cross-National Time Series database (Databanks International, 2009); EFW: Economic Freedom of the World (Gwartney and Lawson, 2009); Polity IV: Polity IV database (Marshall and Jaggers, 2009); PWT: Penn World Table Version 8.1; ICRG: International Country Risk (2013).

4- Result and Discussion  4.1 Political instability, institutions and Human Capital

Tables 2 and 3 present estimates of the impact of institutions on human capital. The variable that represent the human capital were taken from PWT 8.1 In all models presented we used the following explanatory variables: initial human capital, investment, trade openness, government consumption, and population growth. Institutional variables were also used in order to study their impact on human capital: cabinet change, index chain, economic freedom and polity2. This procedure was done because the institutions can affect economic growth directly (with an impact on the product), but also indirectly (with impact on human capital).

Table 2 - Institutions and Human Capital Growth(1) (2) (3) (4)

VARIABLES hcap_gr hcap_gr hcap_gr hcap_gr

L. Human capital (log) -0.00719* -0.0146* -0.0210** -0.0147***

(-1.683) (-1.907) (-2.404) (-3.054)Investment (%GDP) 0.0316** 0.00427 0.00943 0.0393***

(2.209) (0.188) (0.497) (3.227)Exports - Imports (%GDP)

0.00132 0.00312 0.000686 0.00350

(0.302) (0.548) (0.0995) (0.663)Government (% GDP) -0.0136 -0.00692 -0.0145 -0.00189

(-1.144) (-0.521) (-1.229) (-0.156)Cabinet Change -0.0102***

(-2.769)Population Growth 1.256 1.079 0.0361 2.650**

(1.275) (0.632) (0.0198) (2.299)Index Chain 0.00502**

*(2.873)

ChainArea2 0.00134(1.439)

Polity2 0.000435*(1.972)

Constant 0.0104** -0.0224* 0.00586 -0.00187(2.087) (-1.860) (0.934) (-0.399)

Observations 1,132 823 784 1,093Number of id 130 108 108 124hansenp 0.631 0.243 0.180 0.837ar1p 2.33e-07 5.85e-06 1.45e-05 4.32e-07ar2p 0.849 0.873 0.845 0.893

Notes:-System-GMM estimations for dynamic panel-data models. Sample period: 1960–2010;-All explanatory variables were treated as endogenous. Their two period lagged values were used as instruments in the first-difference equations and their once lagged first-differences were used in the levels equation; -t-Statistics are in parentheses. Significance levels at which the null hypothesis is rejected: ***, 1%; **, 5%, and *, 10%.

Table 2, the dependent variable is the growth rate of human capital measured by PWT 8.1, models 2, 3 and 4 the results indicate that there is conditional convergence in human capital, as found for output. The investment share was statistically significant in models 1 and 4, indicating that human capital grows faster when the economy's investment rate is higher. This result makes sense from an economic standpoint, as increased investment translates into higher employment, which can generate incentives for the acquisition of qualifications because of the better performance of the labor market. The other explanatory variables do not seem to have an impact on human capital.

Regarding the institutional variables, we observe that political instability tends to reduce the growth rate of human capital. Countries with greater economic freedom tend to have greater human capital accumulation. This result also makes sense because economic freedom translates into greater mobility of

factors and less regulated labor markets creating incentives for human capital accumulation. The variable democracy was also significant, showing that more democratic countries accumulate more human capital.

Table 3 uses variables of International Country Risk for explain the growth of human capital; the other variables are the same of Table 2.

Table 3- International Country Risk and Human Capital Growth(1) (2) (3) (4) (5)

VARIABLES hcap_gr hcap_gr hcap_gr hcap_gr hcap_gr

L. Human capital (log) -0.00950 -0.0159** -0.0203** -0.0187* -0.00950(-1.399) (-2.115) (-2.565) (-1.780) (-1.191)

Investment (%GDP) 0.0259 0.0249 0.0350* 0.0281 0.0214(1.274) (1.178) (1.663) (1.449) (1.051)

Exports - Imports (%GDP)

0.00330 0.00661 0.00632 0.00603 0.00362

(0.642) (1.079) (1.132) (0.976) (0.707)Government (% GDP) 0.00403 0.00166 0.00312 0.0115 -

0.000339(0.283) (0.130) (0.205) (0.712) (-0.0255)

Ethnic Tensions 0.000957(-0.640)

Population Growth 2.553* 2.725* 1.674 1.679 2.113(1.652) (1.712) (1.126) (0.831) (1.332)

Religion in Politics 0.00331**(2.554)

Bureaucracy Quality 0.00333**(1.990)

Corruption 0.00103(0.509)

Government Stability (% GDP)

0.00112 0.000314

(1.057) (0.388)Constant 0.00740 -0.0108 -0.00372 -

0.008660.00370

(0.881) (-1.192) (-0.475) (-0.739) (0.412)

Observations 643 643 643 643 643Number of id 114 114 114 114 114hansenp 0.0765 0.118 0.0764 0.0506 0.0886ar1p 2.93e-05 4.60e-05 3.60e-05 7.62e-

053.40e-05

ar2p 0.736 0.837 0.875 0.950 0.757Notes:-System-GMM estimations for dynamic panel-data models. Sample period: 1960–2010;-All explanatory variables were treated as endogenous. Their two period lagged values were used as instruments in the first-difference equations and their once lagged first-differences were used in the levels equation; -t-Statistics are in parentheses. Significance levels at which the null hypothesis is rejected: ***1%; **, 5%, and *, 10%.

Presented models it is observed that there is conditional convergence in human capital, as found for output as in Table 2. The institutional variables religion and bureaucracy was statistically significant and their signs were positive but The Hansen’ test wasn’t significant. Joined all variables used of International Country Risk in a one model (Model 6), so it was possible to find, positive results of religion and bureaucracy and Hansen’ test was acceptable. We can conclude that the institutional variables used in this study affect the accumulation of human capital.

4.2 Structural Institution

In this section, we study the effects on GDP and human capital growth rates of structural institutional variables which represent the ration of educated to non-educated people. Four variables were created:

Structural Institution higher education / no educationStructural Institution 2 (higher education + completed secondary

education) / no educationStructural Institution3 (complete higher education + completed

secondary education) / no educationStructural Institution4 complete higher education / (no education +

incomplete primary education).Estimates for the effects of four measures of structural institutions on the

growth rate of GDP per capita are presented in Tables 6.

Table 4- Structural Institutions and Economic Growth(1) (2) (3) (4)

VARIABLES gdp_gr gdp_gr gdp_gr gdp_gr

L. Growth of GDP per capita -0.000978*** -0.000950*** -0.000918*** -0.000868***(-3.293) (-3.322) (-3.304) (-3.072)

Growth Human Capital

0.0400** 0.0230** 0.0242* 0.0229

(1.900) (1.210) (1.265) (1.188)Government (% GDP) -0.00821*** -0.00750** -0.00724* -0.00817**

(-2.294) (-2.011) (-1.926) (-2.344)Investment (%GDP) 0.00575 0.00633 0.00665* 0.00570

(1.451) (1.613) (1.738) (1.502)Exports - Imports (%GDP) -0.000734 -0.000883 -0.000974 -0.000628

(-0.707) (-0.834) (-0.919) (-0.601)Population Growth -0.982** -0.933*** -0.941*** -0.925***

(-4.158) (-4.107) (-3.949) (-3.842)Cabinet Change -0.00279** -0.00237** -0.00224* -0.00214*

(-2.495) (-2.563) (-2.500) (-2.309)Structural Institutions 6.08e-06*

(1.670)L.Structural Institutions2 1.83e-07***

(2.153)

L.Structural Institutions3 1.77e-07**(2.023)

L.Structural Institutions4 2.41e-05(0.788)

Constant 0.0132*** 0.0127*** 0.0123*** 0.0121***(4.521) (4.231) (4.355) (4.322)

Observations 1,124 1,124 1,124 1,124Number of id 128 128 128 128hansenp 0.181 0.622 0.582 0.578ar1p 0.0152 0.0188 0.0193 0.0194ar2p 0.505 0.562 0.556 0.555Notes:-System-GMM estimations for dynamic panel-data models. Sample period: 1960–2010;-All explanatory variables were treated as endogenous. Their two period lagged values were used as instruments in the first-difference equations and their once lagged first-differences were used in the levels equation; -t-Statistics are in parentheses. Significance levels at which the null hypothesis is rejected: ***, 1%; **, 5%, and *, 10%.

When we use variables structural institution the impact of human capital on economic growth is significant and positive in all regressions. We also observed the influence of structural institution on economic growth this result consistent with the expectation of the model i.e. , with the idea that change in structural institutions do affect growth.

Model 2 we can observe the impact of variable institutional structural2 i.e., the ratio higher education + completed secondary education and no education. Thus, expanding the structural institution when we consider people with secondary education the result of the product keeps it positive. Model 3 we can see again the influence of institution, i.e., institution estrutural3, ratio complete higher education + completed secondary education and no education.

Summarizing the models provide strong evidence that structural institutions matter and affect long-term economic growth. This results implies that structural institutions affect productivity and determine an economy’s growth path.

These results agree with those presented by Dias and Tebaldi (2012), but extends the analysis to include other way of measuring the structural institutions is being more widely and can better capture the direction of this variable especially in less educated countries.

5- Conclusion

The objective of this paper is to test the influence of the institutions quality on human capital accumulation process. This paper follows the theoretical developments developed by Dias and McDermott (2006) and Dias and Tebaldi (2012) that establishes a micro-foundation link between human capital accumulation and institutions quality. Using a panel data series from 1960 to 2010 we observe that the economic policies and institutions quality do affect long-term human capital accumulation process in the economy. Better institutions do foster human capital growth.

This paper analyzes the influence of the institutions quality on human capital accumulation process.The theoretical model was proposed by Dias and Tebaldi (2012) in which institutions can affect the economic performance through its impact about human capital. We find that economic, policy and structural institutions can affect economic growth, through their impact on human capital.

The data used in this paper beheld the years 1960-2010. The methodology was applied Dynamic Panels, aimed to correct the possible endogeneity of data.

In line with the literature, we find that political instability, economic freedom, democracy, ethnic tensions, government stability and structural institutions determining the economic growth.

It was also observed that the institutions act on the human capital, i.e., we observe that the economic policies and institutions quality do affect long-term human capital accumulation process in the economy. Better institutions do foster human capital growth.

This paper is in line with the proposed Dias and Tebaldi (2012) in which underscored the importance of structural institutions for economic growth. The major policy implication of this process is that countries with poor initial institutions accumulate less human capital because their returns to education tend to be smaller. This lower human capital accumulation will slow economic growth in the long term.

Dias and Tebaldi (2012) no found significant relationship between political institutions and economic growth. However, in this paper we use other institutional indicators both political and economic, and the results were significant and with expected signs on several variables such as: political instability, economic freedom, democracy, ethnic tensions, government stability.

References

Acemoglu, D., Johnson, S., Robinson, J.A., Yared, P., 2005. From education to democracy? The American Economic Review: Papers and Proceedings 95 (2), 44–49.

Aisen A., Veiga F. J., 2013. How does political instability effect economic growth? European Journal of Political Economy 29, 151-167.

Arellano, M., Bond, S.R., 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies 58, 277–297.

Arellano, M., Bover, O., 1995. Another look at the instrumental variable estimation of error-components models. Journal of Econometrics 68, 29–51.

Coe, D.T., Helpman, E., Hoffmaister, A.W., 2009. International R&D spillovers and institutions. European Economic Review 53, 723–741.

Dias, J., McDermott, J., 2006. Education, institutions, and growth: the role of entrepreneurs. Journal of Development Economics 80, 299–328.

Dias, J., E. Tebaldi, 2012. Institutions, human capital and growth: The institutional mechanism. Structural change and Economic Dynamics 23, 300-312.

Glaeser, E.L., La Porta, R., Lopes-de-Silanes, F., Shleifer, A., 2004. Do institutions cause growth? Journal of Economic Growth 9 (1), 271–303.

Gyimah-Brempong, K., Camacho, S., 1998. Political instability, human capital, and economic growth in Latin America. Journal of Developing Areas 32, 449–466.

Hall, R.E., Jones, C.I., 1999. Why some countries produce so much more output per worker than others? The Quarterly Journal of Economics 114 (1), 83–117.

Hansen, L., 1982. Large sample properties of generalized method of moments estimators. Econometrica, 50(3), 1029-1054.

Kenny, C., Williams, D., 2001. What do we know about economic growth? or, why don’t we know very much? World Development 29 (1), 1–22.

Lipset, S.M., 1960. Political Man. Doubleday Company, Inc., NY, Available at: http://www.questia.com/PM.qst?a=o&d=23560876.

Lucas, R., 1988. On the mechanics of economic development. Journal of Monetary Economics 22, 3–42.

Seck, A., 2011. International technology diffusion and economic growth: Explaining the spillover benefits to developing countries.Structural Change and Economic Dynamics doi: http://dx.doi.org/10.1016/j.strueco.2011.01.003.

Tebaldi, E., Elmslie, B., 2013. Does institutional quality impact innovation? Evidence from cross-country patent grant data. Applied Economics 45 (7), 887–900 (available online: 24 October 2011).

Uzawa, H., 1965. Optimum technical change in an aggregative model of economic growth? American Economic Review 6 (1), 18–31.

Wooldridge, J., 2002. Econometric Analysis of Cross Section and Panel Data. The MIT Press.