Table of Contents
Economics Research International
Volume 2014 (2014), Article ID 709863, 20 pages
Research Article

Public Investments, Human Capital, and Political Stability: The Triptych of Economic Success

Department of Home Economics and Ecology, Harokopio University, 70 El. Venizelou, Kalithea, 17671 Athens, Greece

Received 15 March 2014; Revised 10 August 2014; Accepted 14 August 2014; Published 9 September 2014

Academic Editor: Jean Paul Chavas

Copyright © 2014 Ioannis Kostakis. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


This study assesses the effects of fiscal policy on economic growth in a sample of 96 countries from 1990 to 2010. Ordinary Least Squares (OLS) and Extreme Bound Analysis are mainly estimated in order to investigate whether public investments, human capital, and political stability affect growth controlling for initial output and human capital levels. Furthermore, in this empirical research four subsets of independent variables were used: (a) demographic factors, (b) political determinants, (c) region variables, and (d) variables regarding macroeconomic policy. Empirical results suggest that there is an important difference in the impact of public and private sector investments on the growth of per capita income. Moreover, political indicators such as corruption control, rule of law, and government effectiveness have a high impact on economic growth. Demographic factors, including fertility rate and mortality growth, as well as several macroeconomic variables, like inflation rate index and government consumption, were estimated to be statistically significant factors of economic performance. Fiscal volatility may also be a new possible channel of macroeconomic instability that leads to lower growth. Policy implications of the findings are discussed in detail.

1. Introduction

A question of considerable empirical and theoretical interest is whether countries reach growth rate at the same steady-state [1]. More specifically, one of the most discussed fields in economic literature is economic growth and convergence. On the basis of neoclassical growth model, exogenous parameters, such as technological progress and growth of population, constitute the major drivers of the growth rate at the steady-state, while its shift path to that state is affected by fiscal policy [2]. However, convergence is not always confirmed. That is, although economies are converging, the steady-state level is not always common; therefore, countries may converge to different level of steady-states [3]. Solow’s steady-state growth model of neoclassical general equilibrium was, for over thirty years, the dominant model in the theory of economic growth [4]. Based on this model, conditional convergence can be observed due to the diminishing returns of capital.

A plethora of empirical surveys on the understanding and explanation of the different rates of output growth across the countries has been conducted since mid-1980s, inspired by the endogenous so-called new growth theory. In this theory, the diminishing returns to capital assumption are relaxed and it is visible that, by increasing or constant returns, no presumption of convergence of per capita income across countries can be taken achieving the long-run steady-state growth equilibrium at the natural rate [5].

Simultaneously, several empirical works in the economic literature have analyzed the role of fiscal policy [613]. Empirical research has tried to interpret whether several macroeconomic variables including investment, consumption, output, and growth are influenced positively or negatively by fiscal consolidation and decomposition [14]. Additionally, several researches indicate that fiscal measures could stimulate economic growth under specific circumstances. A significant part of these studies is the role the composition of fiscal policy plays on whether its parameters lead to a higher growth rate and sustainability over time [12, 15]. Nonetheless, it is difficult to cut off the causal effect of fiscal policy, due to the correlation between fiscal policy variables and the level of initial income [12].

The purpose of this paper is to investigate, for a large set of developing and developed countries, public expenditure growth effects, public and private investments, and other impacts of exogenous variables on economic growth, taking conditioning variables and initial conditions into consideration. Utilizing cross-section methodologies and applying several OLS and Extreme Bound Analysis model estimations, we examine whether fiscal policies exert a causal influence on economic performance in the case of 96 countries over the period 1990–2010. We also test the possible interactions among demographic parameters and fiscal variables and look into the extent to which the public sector contributes to the countries’ economic development process. Evidence points out the need to improve productivity investments in the public sector, acknowledging types with higher returns and more complementary to private investments. This means that increases in public investments, which lead to increases in the marginal product of private capital, can enhance the economic performance of the countries [16].

Nevertheless, a key policy contribution of this work concerns the role of public investments across the countries. Thus, it examines a number of key issues, concerning to which extent public investments exert a considerable effect on long-run growth of countries. More specifically, the paper’s contribution to the existing literature is as follows. First, the purpose of the analysis is to extend our understanding of the potential determinants, economic and demographic of public choice by examining empirically the validity of the proposition that economic performance is a function of fiscal and demographic variables. Second, using average cross-section data modeling for 96 developing and developed economies for the period 1990–2010, we show that a long-run relationship exists between per capita economic growth and economic and demographic variables. Third, the model reestimated employing several econometric techniques showing the importance of public investments and human capital on the economic performance of countries in the long run. We also show that in modern economies, where the economic gap among them seems to be huge, high level of political stability and educational skills could give opportunities of growth convergence. Finally, the empirical findings support the proposition that public policy can be associated with higher economic growth since productive expenditures dominate the negative effect of financial gap between rich and poor countries.

The survey is organized as follows. Section 2 describes previous theoretical background. In Section 3 the data and econometric methodology are explained and analyzed. In Section 4 the empirical analysis is presented and the results are being discussed. Finally in Section 5 the main findings are provided, while possible limitations and policy implications are suggested.

2. Previous Literature

Over the last decades, several economic studies tried to recognize the effects monetary policy has on the economic performance of countries. On the other hand, fiscal policy has received less attention in economic studies despite the fact that many economic scientists have argued on it as an effective instrument for stabilizing economic performance [17, 18]. One of the frequently quoted issues in public economics is fiscal consolidation; concerning the long-run trend, public expenditures have to grow relatively to other macroeconomic variables (Wagner’s law). Public expenditures can be treated as an endogenous determinant rather than as a cause of economic growth. On the contrary, Keynesian economists argue that “public expenditures” are an exogenous factor possibly characterized as a policy instrument. Therefore, on the one hand, causality runs from the national income to public expenditures, whereas on the other hand, based on Keynesian proposition, causality runs from public expenditures to national income. However, founded on this topic, empirical evidence is not conclusive [12, 18, 19].

Linkages between economic growth and fiscal policy are attempted to be understood and verified by growth theory though recent revival of interest. Economic growth is also associated with the identification of the elements of public expenditures. Furthermore, researchers have included in their interests the effect of individual decomposed elements of private and public expenditures on per capita economic growth. Particularly, Barro [20] and Romer [21] underline the important role public expenditures play on economic growth and emphasize the research and development of expenditure relevance. What is more, the bond among the formation of public expenditures, the finance of fiscal deficit, and growth is a significant issue [15]. Simultaneously, the direction of causality between government expenditures and government revenues has received considerable attention in many scientific surveys [22].

On the other side, endogenous growth models recommend the enhancement or deceleration of economic growth by fiscal policy. Changes in public expenditures result in effects on economic growth which are central in development economics and in macroeconomics, with their size and sign being controversial [23]. In his research, Barro [24] investigated the effects public expenditures have on economic growth for 98 countries between 1960 and 1985, paying attention to human and physical capital. He concluded in a negative relationship between the ratio of real public consumption to real GDP and the economic growth and investments. The argument is that private productivity in not affected directly by government consumption. On the contrary, through the distorting effects of the way of financing (taxation) or public expenditure programs, it lowers savings and growth. Nevertheless, the decomposition of public expenditure is crucial; specifically, education or defense expenditures seem to affect productivity, therefore influencing private investments. Alesina et al. [25], confirming these results, indicated a sizeable negative effect of public spending on economic growth. Grossman [8] supports the nonlinear relationship between growth of government size and total economic growth. Furthermore, Gupta et al. [15] underlined a strong association between fiscal consolidation and economic growth. Particularly, by showing that countries whose expenditures were concentrated on nonproductive activities (wages), they tended to have lower economic growth, whereas governments distributing higher shares to nonwage goods and productive activities enjoyed faster output expansion.

Growth and employment papers give special interest to government consumption, budget constraint, cash surplus or deficit, the kind and quality of public finances, and government size, measured with either total government expenditures or revenues [69, 12, 26]. Nonproductive expenditures, such as welfare expenditures and social security, wages and economic services expenditures, tend to reduce income per capita growth. On the other hand, productive expenditures on human capital, health, defense, and physical infrastructure may ensure the sustainability of public finances [9, 12, 15, 17, 19, 27].

Expenditures side and fiscal policy volatility appear to affect long-run growth over the business cycles constantly, both of them having negative effect on output growth [7, 19, 2729]. On the revenue side, economic models predict that there is a positive effect on growth by shifting towards nondistortionary forms of taxation rather than shifting away from them [2, 27]. Furthermore, several kinds of taxations create distortionary effects in the market, reducing the incentive for private investments and affecting negatively GDP per capita growth rate.

Recently, in politics and academia, discussions have taken place according to the role private and public investments play in the growth process [6, 9, 12, 16]. The majority of previous empirical studies argue on the positive impact of investments as ratio of GDP on growth [16, 30] whereas, concerning the public investments, they can increase the productivity of private capital. Easterly and Rebelo [12] emphasized the positive correlation of general government investments with both economic growth and private investments, thus being beneficial for economic growth. On the other hand, they can also exclude private investments by corruption and instability in the political system or through the use of scarce resources.

According to recent growth theory models, the income growth per capita and the investments ratio have a strong correlation; therefore, growth rate increases by an exogenous improvement in productivity [9]. Nonetheless, description evidence suggests that changes in fiscal policy can influence private investments. Alesina et al. [25] observed that when public spending increases, labor costs are increased while profits and investments are reduced. Additionally, taxation plays a key role in investments and growth [31]. Particularly, distortions in the market are being created by increases in taxes, thus decreasing profits and private investments. Khan and Kumar [16] pointed that public sector can produce services substitutes with private ones and goods. Moreover, increasing public investments may cause an adverse effect on the private sector implicitly through budget constraints. For instance, if increasing taxes finances the public investments, the costs of inputs may increase, reducing the level of private investments and profits. Finance by borrowing leads to a negative effect on the availability of credit. Furthermore, inflation tax for financing public investments creates high risks and uncertainties regarding the expected returns from investments. However, the marginal productivity of private capital can be increased by public investments and can lead to potential crowding-in of private investments. Nevertheless, positive effects and private investments could be diminished by weak institutions and distortionary finance of public investments [26, 32].

Based on recent literature, human capital also plays an emphasized role in economic development. More specifically, either educational quality or quantity seems to be major determinants of economic growth, whereas in the existing economic growth models, human and physical capital conduct a major role [9, 26, 33, 34]. Empirical surveys identified the productive contribution of infrastructure components and human capital of public capital on per capita income growth [10, 14, 20, 21, 24, 27, 30, 32, 3537]. There are two different channels through which a higher human capital generates higher economic growth: firstly through facilitating the absorption of advanced technologies from advanced economies. Additionally, human capital tends to be less easily adjustable than physical capital within the countries [24]. Nonetheless, the proxy variables of enrolment-ratios and average years of schooling measure only the quantity of human capital. Thereafter, they are not perfect measures of the educational component of human capital [36]. That is the reason why many recent studies use international data of schooling quality such as tests scores, pupils-teacher ratio, repetition rates, and dropout rates. In other studies, the effects that government expenditures on education and health have on economic growth are being investigated [26, 30]. Whereas these measures may not be perfect for productive expenditures (there is a different impact on economic growth when governments invest on building school-houses or hospitals and when they pay wages on public employees), it has been noted that public health expenditures improve citizens’ health, while education expenditures promote cost efficiency and meliorate skillful personnel.

In the same category of productive government expenditures with a direct effect on growth rate is public capital [2]. Shioji [33] points out that public capital has a direct effect on subsequent growth through the enhancement of private capital accumulation. For the improvement of the standards of living as well as long lasting growth, both public and private capital accumulation are a necessary condition [37]. Particularly, long-run growth is affected positively by infrastructure components of public capital, the quality of public capital (paved roads, telecommunication services, and ports), and the relationship between private and public capital matters for the economic growth [12]. For instance, good marginal productivity is increased by institutions and access to international credit markets, leading to an increase in economic growth [32] which is also achieved by technology spillovers and learning-by-doing factors [38].

Several studies have also focused on public sector efficiency and also on the role of good governance, political and macroeconomic stability, export growth, and market distortions [9, 24, 26, 37, 3941]. The most recent studies have emphasized the role of good governance to economic growth, whereas poor management seems to be the main reason of the difficulty public sector faces in developing countries upon the transition of government expenditures into effective services [26]. Moreover, major negative growth factors are the number of political murders annually, plus the number of revolutions and coups. High public investments can also be associated with bad institutions. Furthermore, high corruption can also be related to lower human capital, while public investments are more penetrable to corruption and bad governance. Consequently, “bad public sector” reduces private productivity directly by decreasing the expected returns of investors and by reducing the effectiveness of public investments [32].

Public expenditures often do not yield the expected improvements in human development outcomes, possibly due to high level of corruption and low level of quality of bureaucracy [26, 42]. Additionally, aimless uncertainty in the market economy which increases the investments risk and growth risk is being created by poor macroeconomic stability and management. The government’s ability to minimize the destabilizing impact of several shocks and uncertainties is an important issue [40], along with the relationship between income distribution and economic growth. Adding to that, there is a negative correlation of inequalities and long-term growth [29].

One more controversial factor of growth is openness [43]. On the one hand, trade openness has a positive relationship with per capita growth, through the channel of investments, since more open economies could take advantage of returns to benefit and to scale from the embodied technology. On the other hand, openness is related in a negative way to economic growth due to an inverse connection operating directly on total factor productivity. Action forces related to resource endowments and changes in export diversification may influence this negative association [37]. Furthermore, openness may curtail government expansion within economies [43]. Simultaneously, more macroeconomic indicators including initial income per capita (beta or sigma-convergence hypotheses), inflation, debt or deficit-ratio, and Gini index have been mentioned as powerful factors of economic growth across the countries [6, 9, 13, 29, 40].

Recently, many economists have researched on the existing relationship between sociodemographic variables and economic growth. Most importantly, Malthus [44] first proposed that the growth of population is a potential factor of economic performance. Thereafter, classical and neoclassical growth models [45] underlined that economic growth is an endogenous variable dependable upon population growth, whereas fertility is an exogenous factor. Nowadays, however, the major tendency in economic literature is the development of a dynamic model in which population growth and fertility are determined as endogenous variables [46, 47]. The effect of fertility on per capita income growth is indirect to the increase of the value of parents’ time [20]. Additionally, population growth, mortality rate, and life expectancy are related to the economic performance within the countries. In particular, low fertility may exist in most advanced economies, due to low infant mortality rate and high opportunity cost of having children [48]. At the same time mortality rate could indicate an index of human capital within the countries [26]. Nevertheless, significant advances in life expectancy remain a priority.

3. Data and Methodological Issues

Initially, this section includes the summary statistics of the data which are used in the empirical part. Furthermore, the components of government policy that may affect economic growth across the countries are investigated. A cross-country dataset is used for the empirical analysis. More specifically, our sample includes 96 countries from 1990 to 2010 and annual data are taken from various sources. Specifically, real per capita GDP in 2005 international purchasing power (PPP) dollars (chain index), population growth, and openness to trade are taken from the Penn World Tables. Additionally, data on human capital, measured mainly as the average years of schooling and pupils-teacher ratio, are retrieved from Barro and Lee database, whereas political variables are retrieved from the Worldwide Governance Indicators (WGI). Investment variables are constructed from World Bank (World Development Indicators) and AMECO dataset. Finally, the rest of the variables used in our models are retrieved from World Bank dataset. The dataset is summarized in Table 10 and can be found in Appendix B, providing the definition and the source of all key variables.

Thereafter, econometric and statistical analyses results for the estimation of the “profile” of the countries, with respect to their fiscal policy actions, are being presented. The research provides insights on the determinants affecting GDP per capita growth, for a cluster of developing and developed countries. More specifically, in this research, for the evaluation of the impact of fiscal variables, same approach as adopted in many previous studies is followed [9, 10, 49] and is based on the following regression: where is the per capita GDP growth rate of a cross-section of the countries over a specific period of time; is the constant term; is a set of variables always used in this regression; is a matrix of explanatory control variables proposed as potential significant factors of growth and suggested by previous empirical studies; and is the error term [5] (the theory underlying (1) could express the neoclassical model with capital (physical and human) and labor as the determinants of production if equation was presented as ; according to new growth theory models, human capital could be treated to other factors as a facilitator of technological spill-over results [40]). Alternatively, we can say that the universe of regressors could be divided firstly into free variables (variables of interest) which, as theory dictates, should be in the regression, and secondly into focus and doubtful variables (other control variables) which, as theories suggest, might be important [50]. More specifically, in the case of this research, the base set of variables of interest, the so-called I-variables, consists of the initial (1990) real GDP per capita, the initial human capital (as proxy of human capital either the average years of schooling or the pupils-teacher ratio is used), the government consumption as a ratio of GDP, and the ratio of public and private investments as a percent of GDP.

Other determinants, which could be significant factors of growth such as openness, fertility or mortality rate, inflation index, and debt- or deficit-ratio to GDP, could be included in the set of independent variables. Extreme Bound Analysis, which is explained below, [10, 51] chooses M-variables from demographic variables sets, such as mortality growth and fertility rate, population growth, and life expectancy; political indicators, such as political stability, government effectiveness, rule of law, and control of corruption; and economic factors, such as inflation, Gini index, public debt and deficit-ratio, openness, government consumption, and the volatility of government consumption. Evaluating the robustness of the variables of interest, this pool of M-variables is restricted, eliminating possible measurement errors or econometric issues.

Several econometric methods can be used to generate estimates of the parameters of economic relationships; however, this research follows the tradition of the growth regression literature, OLS estimator [52]. If coefficients are consistently significant and of the same sign, then the result is called “robust,” whereas in any other case “fragile.” More specifically, we are checking whether the coefficients of the main regressors remain significant as additional relevant regressors are included. The main idea of this procedure (EBA) is to examine the beta-coefficients distribution rather than presenting an absolute criterion of robustness [53]. However, following this analysis, the degree of confidence and correlations between GDP growth and individual policy measures is clarified. If a policy measure is correlated robustly with long-term growth, its association with growth seems more confident than a policy measure with a fragile relationship [54].

The results gained by empirical methods are based on the estimation of Ordinary Least Squares (OLS) regression models using cross-country data set, as mentioned above. Thus, according to this procedure, the capture between countries variation is achieved. An additional advantage of this methodology is the controlling for unobserved heterogeneity fixed for each country. What is more, the robustness of the results including control variables is examined, with a focus on the variables of interest related to growth in a significant matter. However, Easterly and Rebelo [12] indicate that the relation between growth and many fiscal variables is statistically fragile. In particular, the statistical importance of fiscal variables in a cross-section regression is largely dependable on the other control variables which are included in the model. The most significant cause of this phenomenon is the existence of multicollinearity between fiscal variables. An additional possible issue of this specific analysis is the problem of endogeneity, facing a difficulty on the determination upon the causality direction (is it causality or reverse-causality?) and to find valid instruments. Nonetheless, the purpose of this work is to provide a summary of the statistical association between economic growth and fiscal policy that will be comprehensive. Table 1 summarizes the expected sign for coefficients based on the majority of previous theoretical and empirical studies (more details acquired from significant studies on economic growth are given in Appendix C).

Table 1: Expected sign of the factors defined in the empirical analysis.

It is assumed that growth rate is related negatively to the initial level of GDP per capita controlling for other variables and positively related to initial human capital. Furthermore, previous studies have highlighted the existence of crowding-in effects of higher public and private investments, especially when linked with good institutions. Fiscal policy volatility may also be a different channel of negative relationship between macroeconomic situation and growth. Good governance proxied by several indices seems to affect the real per capita income growth positively. Last but not least, the effects of inflation and government consumption on economic growth are not conclusive even if the majority of surveys indicate the negative relation to economic performance.

4. Empirical Results and Robustness Checks

4.1. Descriptive Statistics

Descriptive statistics for the data set of 96 countries are found in Table 2 where many differences across the countries can be noticed. Some possibly important characteristics are observed and different growth rates of income per capita across the economies may be explained.

Table 2: Descriptive statistics.

It is obvious that African countries encounter severe economic problems regarding first statistical results. For instance, bad political indices, low levels of physical and human capital, and harmful macroeconomic and demographic indicators concentrate on African countries, possibly, being related to some initial reasons why these countries have such dismal growth performance. That gargantuan failure is the most significant question economists face nowadays [13].

More generally, the trend of average growth rates across the countries or cluster of countries is of considerable empirical interest. Thereafter, Figure 1 illustrates the average growth rates for advanced and developing countries (the distinction between developing and developed economies was based on IMF’s data base), respectively, which are recorded for 5-year period between 1990 and 2010.

Figure 1: Average growth rates for advanced and developing countries, recorded for each 5-year period.

It is realized that growth rate of advanced economies increases during the 1990s, while the opposite trend exists during the decade of 2000. Particularly, between 1990 and 1995 advanced economies have approximately 1% growth rate while in 1995–2000 its growth rate triples. Nonetheless, growth rate declines after 2000, while financial crisis seems to have influenced the growth rate of developed economies; average growth rate is negative between 2005 and 2010. On the other hand, the average growth rate in developing countries increases during the period of the investigation. More specifically, average growth rate increased from 0.25% in 1990–1995 to 2.89% in 2005–2010.

More interestingly, Figure 2 presents the average growth rate for all the period of research per continent. Asian economies have revealed very high rates of economic growth, with contrast to the relatively poor growth performances of the group of European and mostly Sub-Saharan African counties over the whole recorded period [29].

Figure 2: Average growth rate per continent, 1990–2010.

Asian economies enjoy approximately 3.11% average growth over the period of interest; excluding China the average growth rate is 2.52% and excluding China and India it is 2.30%, while Sub-Saharan African countries present only almost 1.14% average growth rate. At the same period, European countries, American and Latin American countries, and Middle-East countries display approximately 1.57%, 1.88%, and 2% average growth rates, respectively. Previous surveys have focused on the determinants of the mentioned results, indicating that Asian countries might preserve stable and sustainable fiscal policies, while countries of other regions repeatedly adopt unsustainable fiscal policies [14, 15, 26, 30, 32, 42].

A fiscal policy analysis issue of an outmost importance is the dynamic structure of an economy, characterized by the growth of some major fiscal elements within countries. Thus, this analysis outlines the growth of the elements of public expenditures including expenditures on education, health, and military services. Additionally, the growth of mortality rate could be expressed as an index of the level of human capital, related to the health system in an economy [26]. Figure 3 illustrates the noteworthy dynamic situation of our sample of countries, with respect to public expenditures and quality of health system. These specific components of a country consist of significant factors of the trend for an economy. Despite the fact that the majority of surveys analyze them as static measures, taking their average level into consideration, in this research, the dynamic structure (growth) of these fiscal components is considered and used.

Figure 3: Average growth rates of mortality rate and public expenditures on education, health, and military services [%].

It is noticeable that for public governments there is a trend to increase the expenditures on education and health system, while the proportion of military services seems to decrease over time. Moreover, it is noteworthy that the average growth rate of mortality is negative, indicating a possible improvement of one index of standard of living across countries. Furthermore, it is of high interest to evaluate the averages of important variables across the fast-growers and slow-growers countries of our sample [10]. Table 3 illustrates these findings of this current research.

Table 3: Cross-country averages, 1990–2010.

Noticeably, countries that are growing faster than average over the years 1990–2010 tend to have a higher average share of public and private investments, lower government consumption, larger openness and initial primary-enrollment rate, lower initial primary pupils-teacher ratio, and lower average debt as a percent of GDP. More interestingly, fast-growers countries have both higher total factor productivity growth and higher growth of employment in industry. It is noteworthy that slow-growers countries have negative average growth of these measures, which seems to agree with Lewis model and the crucial role of the capitalist surplus in the development process. On the one hand, agriculture is an activity subject to diminish returns owned by the fixity of the supply of land, whereas, on the other hand, industrial sector leads an economy to high economic growth [5]. Nevertheless it is considerably questionable why public expenditures on health and education are higher for slow-growers countries, while expenditures on military are higher on fast-growers countries. Simultaneously, deficits seem to enhance the growth of per capita GDP. Empirically, it can be noted that the kind of finance of deficits is more significant than the level of deficits, whereas expenditures on health or education could not be associated with higher growth in case these expenditures concentrate on wages or on nonproductive activities [15].

4.2. Correlations

This section mainly illustrates the simple and partial associations between many variables of interest and growth of per capita GDP rate. More specifically, Table 4 presents the simple correlations (the calculation of the coefficients is done using the Spearman formula in order to eliminate the effect of outliers) of variables of interest and GDP growth per capita rate during the period of 1990 to 2010.

Table 4: Simple correlations of fiscal variables and GDP growth per capita rate, 1990–2010.

The analysis starts with the calculation of simple correlations between long-run growth and several variables of interest. One of the most cited issues in economic theory of growth is the analysis of convergence. According to Solow’s model, convergence is predicted only after controlling for the steady-state determinants. Additionally, augmented Solow model indicates that accumulation of human capital is associated with savings and growth of population, simultaneously reducing their estimated effects on growth. Endogenous growth models make different predictions; no steady-state level of income is found here. In these models there is a strong assumption of nondecreasing returns and there is an explanation for the differences among countries in GDP per capita, which are found even in the case of these countries having the same savings and population growth rates [11]. Moreover, Barro [9] explains that it is inconsistent with the cross-country evidence that rich countries grow lower than poor countries, due to the small correlation with the initial level of per capita income. That is known as the hypothesis of beta-convergence. (Apart from the beta-convergence, the issue of sigma-convergence is one of a high interest due to the fact that the level of inequality across economies is seen. It is measured as the variance of the log of GDP per person. However it is not explained in detail here because it is beyond the aim of this survey.) In Figure 4 the simple relation between initial income per capita (1990) and average growth rate from 1990 to 2010 for 96 countries is visible.

Figure 4: Per capita growth versus initial per capita GDP. (Correlation matrix between per capita GDP growth (Growth) and initial per capita GDP (Initial). Obs. = 96. Growth = 1.000 (Growth). Initial = −0.094 (Growth).)

A small negative relation between the aforementioned variables is visible, with this relationship not being affected by some outliers like China or Congo. The average growth rate of per head real income is not significantly related to the 1990 value of real GDP per capita; the correlation is −0.094 (Pearson correlation).

Besides that, human capital contributes significantly to a number of endogenous economic growth models. Human capital can be the channel of generation of new ideas and the accumulation of technological progress. Countries with higher initial human capital stock have a higher income per capita growth, while these countries also tend to be the technological leaders. Thus, poor countries for a given quantity or quality of human capital tend to have higher growth than rich ones [9]. Figure 5 illustrates the correlation between per capita income growth and human capital measures in different countries.

Figure 5: Per capita GDP growth versus initial average years of schooling and initial primary pupils-teacher ratio. (Correlation matrix between per capita GDP growth (Growth) and initial human capital (Inhuman) and initial pupil-teacher ratio (Inpupil). Obs. = 86. Growth = 1.000 (Growth). Inhuman = 0.123 (Growth), Inpupil = −0.260 (Growth).)

Simple correlation between initial human capital and growth is expected to exist, according to previous studies [9, 10, 24]. More specifically, based mainly on new growth theory economists (Barro, Romer, Lucas, etc), human capital is the main positive parameter of economic success for an economy. Human capital is not subject to diminishing return scales versus physical capital giving to this the advantage to boost economy. Here, in our case, the correlation between growth and initial average years of schooling is 0.12, while the correlation between per capita income growth and the initial primary pupils-teacher ratio is −0.26. In this research two different measures of human capital are used, due to the desire to capture both impact of quality and quantity of human capital of education on economic growth across the countries. Nevertheless, based on Barro [9], the most interesting question is the partial correlation between growth of GDP per capita and initial income (1990) controlling for human capital. Figure 6 plots per capita GDP growth rate versus initial income net of the value predicted by all other explanatory variables; that is, the graph amplifies the partial association between growth of per capita income and initial GDP per capita.

Figure 6: Partial association between per capita growth and 1990 GDP per capita controlling for initial average years of schooling (a) and for initial average years of schooling and initial primary pupils-teacher ratio (b). (Correlation matrix between per capita GDP growth (Growth) and initial per capita GDP (Initial) controlling for initial average years of schooling (a) and for initial average years of schooling and initial primary pupils-teacher ratio (b). Obs. = 96. Growth = 1.000 (Growth). Inhuman = −0.301 (Growth). Inpupil = −0.375 (Growth).)

Specifically, in contrast with simple correlation (Figure 4), it is now more negative; the correlation between per capita GDP growth rate and initial income, with controlling only for initial average years of schooling (quantitative measure of education), is −0.31, while with a controlling for initial average years of schooling and initial primary pupils-teacher ratio (qualitative measure of education) it is −0.37. That is, the results indicate that, by holding the proxies for initial level of human capital constant, higher initial income per capita is mainly correlated in a negative way with posterior per capita growth.

However, the main objective of the current dissertation is to investigate the linkages between public investments and economic growth. That is the more important hypothesis of our research. The role of investments has been stressed in the recent literature on long-run growth and convergence issues [9, 11]. Figure 7 illustrates the expected relation between these variables.

Figure 7: Per capita GDP growth versus public investments. (Correlation matrix between per capita GDP growth (Growth) and public investment (invest). Growth = 1.000 (Growth), invest = 0.375 (Growth).)

Figure 7 shows a positive simple correlation (0.37) between public investments and economic growth for all countries. Increases in average public investments are related to higher per capita GDP growth. Our hypothesis is to check if public investments remain statistically significant for economic growth after controlling for several interesting variables. (Table 12 in Appendix D illustrates more partial correlations between public investments and per capita GDP growth controlling several variables.)

Finally at this step, the focus turned to government consumption and its fiscal volatility mechanism. Econometric evidence endeavors to shed light on another relation between macroeconomic volatility and growth [14, 29]. Figure 8 affiliates the relationships between government consumption, government consumption volatility, and GDP per capita growth rate.

Figure 8: Per capita growth rate versus government consumption and consumption volatility. (Correlation matrix between per capita GDP growth (Growth) and government consumption (Gov) and government consumption volatility (Volat). Obs = 95. Growth = 1.000 (Growth). Gov = −0.037 (Growth). Volat = −0.159 (Growth).)

In Figure 8 a negative relation between both of the variables and per capita income growth is visible. More specifically, the correlation between government consumption and GDP per capita growth and government consumption volatility and GDP per capita growth is −0.04 and −0.16, respectively. This results in countries with higher government consumption ratio or higher government consumption volatility having lower per capita GDP growth. Previous empirical studies have supported that government consumption or fiscal volatility could be powerful channels of economic performance and could affect indirectly growth via the higher risk and the reduced incentives to invest [9, 14, 29].

4.3. Econometric Results

Thereafter, econometric results are presented. More specifically, the relation between economic performance and fiscal policy measures can be calculated through the regressions of the annual rate of real GDP per capita growth on a set of independent control variables. This empirical analysis uses a cross-section dataset covering for the years from 1990 to 2010. The outcome of regressions is presented in detail on Tables 5 and 6, including the estimated coefficients and their respective White Heteroskedasticity corrected -values for the explanatory variables. Particularly, the regressions for annual average growth rates of real GDP per capita are shown. Table 5 illustrates the results by adding several variables of specific groups at a time, while Table 6 includes the main regressions containing all statistically significant variables from the previous procedure (we cannot use many variables of the same category—demographic, educational or political variables—in the same equation/regression due to the existence of high level of multicollinearity issue that leads to biased results) for the robustness check of the independent variables of interest. The standard errors are based on White’s Heteroskedasticity covariance matrix, because Heteroskedasticity can be important across the countries, and are presented in parentheses.

Table 5: Regressions for per capita GDP growth, 1990–2010.
Table 6: Main regressions for per capita GDP growth, 1990–2010.

Previous empirical results investigate upon the importance of public investments and their contribution to long-run growth of developing and developed countries. This dissertation takes also the importance of human capital into account and embodies various measures of it in estimating equations. Prior to the examination of the significant impact of public investments, obtained by the results of the main regressions (Table 6), the focus is on the overall benchmark regression of Table 5. (As we can see from Table 6 includes as proxies for human capital regarding the education both the initial average years of schooling and the initial pupils-teacher ratio in order to control quantity and quality of human capital. Results remain almost similar; however, due to the small observations numbers for pupils-teacher ratio variables, the rest of the equations include only the initial average years of schooling as a proxy of human capital.) This gives an explanation to around one-third of the variation of the depended variable. The F-test statistic for the analysis (F = 5.50) rejects the null hypothesis of no explanatory power of the independent variables at 1% level.

As regards the individual coefficients, the main variable of interest is public investments-ratio. The calculated coefficient from benchmark regression indicates that an increase of one percentage point in public investments-ratio across the countries is related to an increase in per capita income of approximately one-fifth of a percentage point. Moreover, the role of both public and private investments in determining per capita income growth should be further considered. Specifically, as indicated in almost all columns, both types of investments have a positive impact on per capita growth but its magnitude and significance differ confoundedly, with public investments having a much stronger impact than private ones. One possible explanation of this difference could be the fact that the stock of public infrastructural capital, such as railways, paved roads, and public transportation services, is higher than the previous decades, leading to higher returns from such investments.

The estimated coefficient of initial income per capita is negative, as known from neoclassical growth theory. However, it is not always significant. Nevertheless, for better understanding on the magnitude of this coefficient, the following analysis should be considered. Due to the fact that GDP per capita is measured in thousands of 2005 ($), the results reveal that a per capita initial real GDP increase by $1,000 decreases the real per capita growth rate by 0.004 percentage points annually. For instance, if you take into consideration the 1990 per capita GDP of Iceland ($35,758) and Kenya ($1,181), the difference of $34,577 would be related to around 0.14% greater growth rate for Kenya than for Iceland ceteris paribus.

The coefficient of initial human capital (average years of schooling) is significantly positive as expected under previous new growth theory models analysis. An extra year of stock of human capital (initial average years of schooling) is associated with an increase by 0.284 percentage points for per capita income growth. Lastly, government consumption is statistically significant with a negative sign, meaning that an increase in average consumption-ratio by one percentage point leads to a decrease in per capita income growth by 0.086 percentage points.

The empirical results allowing differences between clusters of variables are visible in Table 5. Columns 3–7 show the results controlling for demographic variables, columns 8–11 for political variables, column 12 for regional slope dummies, and columns 13-14 for economic variables. (We could not include more variables of the same category in the same regression due to the high correlation among them leading to the problem of multicollinearity [12]. For example, correlation between initial life expectancy and initial fertility rate is −0.89 while the correlation between macroeconomic stability and government effectiveness is 0.77.) More specifically, as it can be noted, all demographic variables have the expected sign. However, only the initial fertility rate and mortality growth (under 5 years old) are statistically significant at 5% level. In particular, a country with one percentage point higher fertility rate leads to a lower growth rate of per capita income for 0.352 percentage points. In addition, mortality growth is a variable that is also associated negatively with economic performance. Countries with lower mortality growth enjoy faster per capita income growth through time. As far as political variables are concerned, it is really interesting that political variables play a very important role in economic performance across the countries. More specifically, indices of government effectiveness, control of corruption, and rule of law result in a faster economic growth. That is, empirical results show that an increase in index of mentioned political variables by one unit associated with increases by 1.88, 1.41, and 1.41 percentage points, respectively. These results confirm the significant importance of political determinants in economic life within and across the countries. Dummy variables for Asia and Africa are not found statistically significant. Finally, several macroeconomic variables such as inflation rate index, debt and deficit-ratio, openness, Gini index, and government consumption volatility through time are added as control variables in 13th and 14th regressions. All variables have the expected sign. However, only inflation index and the channel of volatility of government consumption are statistically significant at 1% level. More specifically, one unit of index increase in inflation is associated with a decrease of 0.002 percentage points in per capita income growth rate. Similarly, an increase in volatility of government consumption by one standard deviation is associated with a decrease of 0.366 and 0.491 percentage points in per capita GDP growth (columns 13 and 14, resp.). Moreover, a joint test for the importance of the entire number of macroeconomic control variables leads to the statistic equal to 12.30 and 12.38, respectively (for columns 13 and 14).

A further step in this research is to check the robustness of the results by observing whether the above estimates for the variables of interest remain statistically significant with the same sign, in the case of including all significant control variables of different clusters in the same regression (Table 6). A major question raised is the examination of public investments variable. First three columns (1–3) include the same mixed control variables but for different indices of political factors and controlling for deficit-ratio, while the following three columns (4–6) contain debt-ratio instead of deficit-ratio. (We cannot include all control variables in a simple regression because of the high multicollinearity issue.)

Generally, it could be argued that the share of public investments and human capital may simply reflect a higher growth rate of per capita income growth, resulting in evidence consistent with our basic and more important hypothesis; public investments matter. All columns show robustly correlation of the share of public investments with growth per capita income in our cross-country section model, once controlling for the slew of high-cited variables in previous studies. The statistical importance of public investments variable is not dependable upon what other variables are included in the regression, thus indicating the robustness of it. Apart from investments, initial human capital coefficient remains positive and robust, confirming the new-growth theory hypothesis that human capital is an indicator of economic growth of significant importance. In the case of initial income per capita, the result is fragile while significance is sensitive to the control variables that are included in the regressions. It indicates very little importance in the explanation of economic growth of per capita income. The ratio of government consumption to GDP is related negatively to economic growth, as assumed, but the result does not indicate robustness; however it may shed little light on this hypothesis. The hypothesized positive effect of a good political situation in countries on economic performance receives, at this point, high support in our data. Finally, the coefficient of inflation is found to be negatively correlated with economic growth. Specifically, an increase in inflation index by one unit leads to a decrease in growth by 0.001 percentage points at 1% level.

Thereafter, it is also interesting to understand roughly the contribution of each regressor explaining the variation in the dependent variable. Thus, following the methodology of Kormendi and Meguire [6] the change in is computed, obtaining by selectively omitting each variable from benchmark equation. More specifically, marginal contribution could be defined as the proportion reduction in the variation of dependent variable achieved by the introduction of the entire set of independent variables [55] with each regressor in the first column deleted in turn.

Thus, Table 7 illustrates a previous analysis in which public investments variance itself explains approximately 45.9% while initial human capital variance explains almost 24.2% of the observed variation in economic per capita growth. Based on our data, the procedure of contributions to also confirms the importance of public investments and human capital on economic growth based on our data.

Table 7: Marginal contribution to based on the benchmark regression, 1990–2010.

Levine and Renelt [10] proposed a more empirical application of EBA test. Below, this test, for our variables of interest which were found formerly robust, is presented. In particular, the robustness of the variables of public investments and initial human capital separately is tested, due to the fact that the specific variables remained robust based on previous regressions. Thereafter, the base model (benchmark) is run and the regression result is computed, for all possible linear combinations of up to three macroeconomic (the regressions include only the M-variables of the economic group of variables because combinations of either demographic or political variables suffer from high collinearity) M-variables. Then, the lowest and highest coefficients are identified. More specifically, many regressions were run keeping the variables of benchmark equation stable and up to three different economic control variables in each regression were added. Actually, the distribution of beta-coefficients of the robust variables of interest is presented. Table 8 consists of the aforementioned results.

Table 8: Sensitivity results for public investments and initial human capital.

As it can be noticed, this procedure also confirms previous results indicating that both public investments and initial average years of schooling coefficients are positive and robust. These robust positive relations between public investments and growth of per capita GDP and between initial human capital and growth of per capita GDP are consistent with a large collection of growth surveys.

5. Discussion and Conclusions

This research provides insights upon the factors that affect the economic growth of a country. Empirical results suggest that government policy may affect the performance of an economy. As Easterly and Rebelo have shown, the development of neoclassical model provided public finance enhances thinking about the growth effects of fiscal policy. Several methodologies and techniques have been developed trying to explain the dynamic impact of public policy on economic performance among economies. On the other hand, endogenous growth models tend to transform the temporary growth effects of fiscal policy (neoclassical model) into permanents growth effects. In addition, the recognition that the determinants of long-term economic growth in a central macroeconomic problem led to the development of several methodological approaches that endeavor to decompose the effect of each fiscal policy variable on per capita economic growth. Thus, there is a good deal of empirical estimation of growth models using cross-country and cross regional data [9, 9, 24, 36]. The majority of aforementioned research has shown the positive and robust impact of public investments and human capital (productive expenditures) on economic growth. We also followed similar econometric estimations. In general, our research results agree with previous literature. However, very few economic variables were found to be robustly associated with cross-country growth rates. Furthermore, some robust correlations that could slight alterations in the list of regressors were identified. More specifically, using data from a sample of 96 countries, the research measured if certain variables were related to economic performance, motivated by previous theoretical assumptions. The findings of our research are briefly presented below.

Initially, the variables which were found statistically significant usually appear with the expected sign. Thus, it was noted that public and private investments, initial average years of schooling, government effectiveness, control of corruption, and rule of law tend to have a positive impact on GDP per capita growth while initial pupils-teacher ratio, initial fertility rate, mortality growth, government consumption, volatility of government consumption, and inflation rate index have a negative effect on growth. Secondly, a positive and robust correlation between average share of public investments on GDP and the average growth rate of per capita income was found. That was the most meaningful hypothesis of our research. Moreover, a positive and robust association between initial human capital (average years of schooling) and average growth rate was found. As far as government consumption is concerned, the research investigated a negative correlation with GDP per capita growth. However, this result does not hold with the inclusion of some control variables indicating the nonrobustness of this correlation. Initial per capita income has negative sign but it is not always significant. Thus, it can be concluded that conditional convergence is not clear or conclusive. Finally, a large set of other economic and political indicators such as inflation index, government effectiveness, control of corruption, and other variables are correlated significantly with economic growth.

Consequently, based also on previous studies, it is noticeable that there is no unique model or formula to estimate the economic growth. Cross-country growth findings are sensitive to conditional information, difficult to be interpreted, and based on a fragile statistical basis, therefore not easily isolating a strong econometric relationship between an indicator and long-run per capita income growth. However, it may have several implications especially for those who design economic policies. In particular, governments should assess their choices even if some measures of interest are not statistically significant; it does not mean that nonstatistically significant variables do not matter for the social economy. Mixed fiscal policy can enhance economic performance in a country. Nevertheless, further research is needed. Particularly, extended analysis should be done to investigate how several kinds of public expenditures, such as government wages and nonwage components, transfers, and subsidies, or revenues like direct and indirect taxes or social contributions can influence per capita income growth rate. More importantly, there must be an emphasis on the issues of endogeneity and reserve causality. The regressions allow us to find correlations between variables but without being sure whether it is causality or not. Generally, public polices seem to be a complicated issue; therefore future research must focus on regimes in the policy sector and also on the interactions between policies as opposed to the independent effect of any particular policy.


A. List of 96 Advanced and Developing Countries

See Table 9.

Table 9: List of 96 advanced and developing countries.
Table 10: Data description and sources.

B. Data Description and Sources

See Table 10.

C. Empirical Studies on Economic Growth

See Table 11.

Table 11: Empirical studies on economic growth.
Table 12: Partial associations.

D. Partial Associations

See Table 12.

Conflict of Interests

The author declares that there is no conflict of interests regarding the publication of this paper.


The author would like to express his great appreciation to his supervisor Dr. Keisuke Otsu for his patient guidance and his constructive suggestions for this paper. Special thanks are due to Professor George Hondroyiannis and Dr. Eleni Sardianou, who provided him with the econometric background that he extensively exploited to prepare the present paper. The author would like to thank the anonymous referees for providing him with constructive comments and suggestions.


  1. N. Islam, “Growth empirics: a panel data approach,” Quarterly Journal of Economics, vol. 60, pp. 1128–1170, 1995. View at Google Scholar
  2. R. Kneller, M. F. Bleaney, and N. Gemmell, “Fiscal policy and growth: evidence from OECD countries,” Journal of Public Economics, vol. 74, no. 2, pp. 171–190, 1999. View at Publisher · View at Google Scholar · View at Scopus
  3. A. Gaspar, “Convergence analysis: a new approach,” in Munich Personal RePEc Archive, vol. 9, pp. 382–390, 2012. View at Google Scholar
  4. G. Grossman and E. Helpman, “Endogenous innovation in the theory of growth,” The Journal of Economic Perspectives, vol. 8, no. 1, pp. 23–44, 1994. View at Google Scholar
  5. A. Thirlwall, Economics of Development: Theory and Evidence, Palgrave Macmillan, New York, NY, USA, 9th edition, 2011.
  6. R. C. Kormendi and P. G. Meguire, “Macroeconomic determinants of growth: cross-country evidence,” Journal of Monetary Economics, vol. 16, no. 2, pp. 141–163, 1985. View at Publisher · View at Google Scholar · View at Scopus
  7. D. L. Landau, “Government expenditure and economic growth in the developed countries: 1952–76,” Public Choice, vol. 47, no. 3, pp. 459–477, 1985. View at Publisher · View at Google Scholar · View at Scopus
  8. P. J. Grossman, “Government and economic growth: a non-linear relationship,” Public Choice, vol. 56, no. 2, pp. 193–200, 1988. View at Publisher · View at Google Scholar · View at Scopus
  9. R. J. Barro, “Economic growth in a cross section of countries,” The Quarterly Journal of Economics, vol. 106, no. 2, pp. 407–443, 1991. View at Publisher · View at Google Scholar · View at Scopus
  10. R. Levine and D. Renelt, “A sensitivity analysis of cross-country growth regressions,” American Economic Review, vol. 82, no. 4, pp. 942–963, 1992. View at Google Scholar · View at Scopus
  11. N. G. Mankiw, D. Romer, and D. N. Well, “A contribution to the empirics of economic growth,” The Quarterly Journal of Economics, vol. 107, no. 2, pp. 407–437, 1992. View at Google Scholar
  12. W. Easterly and S. Rebelo, “Fiscal policy and economic growth,” Journal of Monetary Economics, vol. 32, no. 3, pp. 417–458, 1993. View at Publisher · View at Google Scholar · View at Scopus
  13. X. Sala-i-Martin, 15 Years of New Growth Economics: What have We Learnt?Discussion Paper Series, Columbia University, New York, NY, USA, 2002.
  14. D. Furceri, “Is government expenditure volatility harmful for growth? A cross-country analysis,” Fiscal Studies, vol. 28, no. 2, pp. 103–120, 2007. View at Publisher · View at Google Scholar · View at Scopus
  15. S. Gupta, B. Clemennts, E. Baldacci, and C. Mulas-Granados, “Fiscal policy, expenditure composition, and growth in low-income countries,” Journal of International Money and Finance, vol. 24, no. 3, pp. 441–463, 2005. View at Publisher · View at Google Scholar · View at Scopus
  16. M. S. Khan and M. S. Kumar, “Public and private investment and the growth process in developing countries,” Oxford Bulletin of Economics and Statistics, vol. 59, no. 1, pp. 69–88, 1997. View at Google Scholar · View at Scopus
  17. A. Afonso and R. M. Sousa, “The macroeconomic effects of fiscal policy,” Applied Economics, vol. 44, no. 34, pp. 4439–4454, 2012. View at Publisher · View at Google Scholar · View at Scopus
  18. H. Hamdi and R. Sbia, “Fiscal policy and economic growth in PIIGS countries: an empirical assessment,” The Journal of Applied Business Research, vol. 29, no. 5, pp. 1343–1352, 2013. View at Google Scholar
  19. A. Afonso and D. Furceri, “Government size, composition, volatility and economic growth,” European Journal of Political Economy, vol. 26, no. 4, pp. 517–532, 2010. View at Publisher · View at Google Scholar · View at Scopus
  20. R. Barro, “Government spending in a simple model of endogenous growth,” Journal of Political Economics, vol. 98, pp. 103–125, 1990. View at Google Scholar
  21. P. Romer, “Endogenous technological change,” Journal of Political Economy, vol. 98, pp. 71–102, 1990. View at Google Scholar · View at Scopus
  22. G. Hondroyiannis and E. Papapetrou, “An examination of the causal relationship between government spending and revenue: a cointegration analysis,” Public Choice, vol. 89, no. 3-4, pp. 363–374, 1996. View at Publisher · View at Google Scholar · View at Scopus
  23. D. Hatzinikolaou, “Government spending, taxation of interest, and private saving: an empirical analysis using Greek aggregate data,” Journal of Development Economics, vol. 52, no. 2, pp. 317–330, 1997. View at Publisher · View at Google Scholar · View at Scopus
  24. R. J. Barro, “Human capital and growth,” The American Economic Review, vol. 91, no. 2, pp. 12–17, 2001. View at Publisher · View at Google Scholar · View at Scopus
  25. A. Alesina, S. Ardagna, R. Perotti, and F. Schiantarelli, “Fiscal policy, profits, and investment,” American Economic Review, vol. 92, no. 3, pp. 571–589, 2002. View at Publisher · View at Google Scholar · View at Scopus
  26. A. S. Rajkumar and V. Swaroop, “Public spending and outcomes: does governance matter?” Journal of Development Economics, vol. 86, no. 1, pp. 96–111, 2008. View at Publisher · View at Google Scholar · View at Scopus
  27. D. Romero-Ávila and R. Strauch, “Public finances and long-term growth in Europe: evidence from a panel data analysis,” European Journal of Political Economy, vol. 24, no. 1, pp. 172–191, 2008. View at Publisher · View at Google Scholar · View at Scopus
  28. E. Bairam, “Government size and economic growth: the African experience, 1960–85,” Applied Economics, vol. 22, no. 10, pp. 1427–1435, 1960. View at Google Scholar
  29. J. Woo, “Growth, income distribution, and fiscal policy volatility,” Journal of Development Economics, vol. 96, no. 2, pp. 289–313, 2011. View at Publisher · View at Google Scholar · View at Scopus
  30. N. Bose, M. E. Haque, and D. R. Osborn, “Public expenditure and economic growth: a disaggregated analysis for developing countries,” The Manchester School, vol. 75, no. 5, pp. 533–556, 2007. View at Publisher · View at Google Scholar · View at Scopus
  31. A. Johansson, C. Heady, J. Arnold, B. Brys, and L. Vartia, “Taxation and Economic Growth,” Working Papers 620, OECD Economic Department, 2008. View at Google Scholar
  32. E. Cavallo and C. Daude, “Public investment in developing countries: a blessing or a curse?” Journal of Comparative Economics, vol. 39, no. 1, pp. 65–81, 2011. View at Publisher · View at Google Scholar · View at Scopus
  33. E. Shioji, “Public capital and economic growth: a convergence approach,” Journal of Economic Growth, vol. 6, no. 3, pp. 205–227, 2001. View at Publisher · View at Google Scholar · View at Scopus
  34. C. Bjornskov and P. Guillaume, “Is trust the missing root of institutions, education and development?” Public Choice, vol. 157, pp. 641–669, 2013. View at Google Scholar
  35. R. Barro and X. Sala-i-Martin, “Public finance in models of economic growth,” Review of Economic Studies, vol. 59, no. 201, pp. 645–661, 1992. View at Google Scholar
  36. J. W. Lee and R. Barro, “Schooling quality in a cross-section of countries,” Economica, vol. 68, pp. 465–488, 2000. View at Google Scholar
  37. P. Astorga, “A century of economic growth in Latin America,” Journal of Development Economics, vol. 92, no. 2, pp. 232–243, 2010. View at Publisher · View at Google Scholar · View at Scopus
  38. J. W. B. Bos, C. Economidou, M. Koetter, and J. W. Kolari, “Do all countries grow alike?” Journal of Development Economics, vol. 91, no. 1, pp. 113–127, 2010. View at Publisher · View at Google Scholar · View at Scopus
  39. J. R. Lott Jr., “Political cheating,” Public Choice, vol. 52, no. 2, pp. 169–186, 1987. View at Publisher · View at Google Scholar · View at Scopus
  40. M. F. Bleaney, “Macroeconomic stability, investment and growth in developing countries,” Journal of Development Economics, vol. 48, no. 2, pp. 461–477, 1996. View at Publisher · View at Google Scholar · View at Scopus
  41. A. Williams and A. Siddique, “The use (and abuse) of governance indicators in economics: a review,” Economics of Governance, vol. 9, no. 2, pp. 131–175, 2008. View at Publisher · View at Google Scholar · View at Scopus
  42. R. Hodler and A. Dreher, “Development (paradigm) failures,” Journal of Development Economics, vol. 101, no. 1, pp. 63–74, 2013. View at Publisher · View at Google Scholar · View at Scopus
  43. A. M. Wu and M. Lin, “Determinants of government size: evidence from China,” Public Choice, vol. 151, no. 1-2, pp. 255–270, 2012. View at Publisher · View at Google Scholar · View at Scopus
  44. T. R. Malthus, An Essay on Population, M. Deut, London, UK, 1933.
  45. R. M. Solow, “A contribution to the theory of economic growth,” Quarterly Journal of Economics, vol. 70, pp. 65–94, 1956. View at Google Scholar
  46. G. Hondroyiannis and E. Papapetrou, “Fertility choice and economic growth: empirical evidence from the U.S,” International Advances in Economic Research, vol. 5, no. 1, pp. 108–120, 1999. View at Publisher · View at Google Scholar · View at Scopus
  47. G. Hondroyiannis and E. Papapetrou, “Demographic changes, labor effort and economic growth: empirical evidence from Greece,” Journal of Policy Modeling, vol. 23, no. 2, pp. 169–188, 2001. View at Publisher · View at Google Scholar · View at Scopus
  48. G. Hondroyiannis, “Fertility determinants and economic uncertainty: an assessment using European panel data,” Journal of Family and Economic Issues, vol. 31, no. 1, pp. 33–50, 2010. View at Publisher · View at Google Scholar · View at Scopus
  49. X. X. Sala-i-Martin, “I have just run two million regressions,” The American Economic Review, vol. 87, no. 2, pp. 178–183, 1997. View at Google Scholar · View at Scopus
  50. K. D. Hoover and S. J. Perez, “Truth and robustness in cross-country growth regressions,” Oxford Bulletin of Economics and Statistics, vol. 66, no. 5, pp. 767–798, 2004. View at Google Scholar · View at Scopus
  51. E. E. Leamer, “Let’s take the con out of econometrics,” American Economic Review, vol. 73, pp. 31–43, 1983. View at Google Scholar
  52. J. M. Wooldridge, Introductory Econometrics, South West, Nashville, Tenn, USA, 4th edition, 2009.
  53. S. Fölster and M. Henrekson, “Growth effects of government expenditure and taxation in rich countries,” SSE/EFI Working Paper Series in Economics and Finance 301, The Economic Research Institute, Stockholm School of Economics, 2000. View at Google Scholar
  54. R. Levine and S. J. Zervos, “What we have learned about policy and growth from cross-country regressions?” The American Economic Review, vol. 83, pp. 426–430, 1993. View at Google Scholar
  55. U. Desai, “Determinants of educational performance in India: role of home and family,” International Review of Education, vol. 37, no. 2, pp. 245–265, 1991. View at Publisher · View at Google Scholar · View at Scopus
  56. R. J. Barro and J.-W. Lee, “International comparisons of educational attainment,” Journal of Monetary Economics, vol. 32, no. 3, pp. 363–394, 1993. View at Publisher · View at Google Scholar · View at Scopus