SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research

This research extends the literature on regional inequalities in the European Union in two directions. First, it confirms the importance of institutions for regional inequalities among 18 EU member countries, and second, indicates which dimension of the institutional quality is important by using the six dimensions of the Worldwide Governance Indicator. The results support standard “rules of engagement” that reduce transaction costs by lowering uncertainty and facilitating the mutual trustworthiness of individual economic agents that spread equally among regions in specific countries. In addition, we interpret empirical evidence to determine that institutions also shape regional inequalities indirectly through political channels.

1 Introduction

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research

The European Union (EU), as one of the world's most important economic areas, has a distinct role in the global economy and, subsequently, problems in the EU cause significant effects at a global level. Decreasing regional spatial inequalities among EU member countries has been recognized as an important goal for the EU (Art. 158 Treaty on European Union), and this has been additionally stressed as an essential challenge after recent additions, which progressively increased the EU from 15 to 27 member states. Moreover, accession of new countries to the single market and their lower income level, compared to older member states, have had significant effects on regional distribution of economic performance in the EU (Arabia et al., 2010).

Investigating the dynamics of the convergence process has provided contradictory results (Mohl & Hahen 2010; Le Pen 2011). Some authors find a positive trend in the convergence process (Le Pen 2011), some weakly positive (Fingleton 1999), and some even a negative trend (Pittau & Zelli 2006). Several studies agree that Europe, as a whole, is experiencing a downward trend in the level of inequality (Petrakos et al., 2005), while inequalities between regions within the same country have tended to increase (Rodríguez-Pose & Gill 2006; Brakman & van Marrevijk 2008; Arabia et al., 2010). Thus, policymakers are especially motivated to analyze the convergence or divergence processes among national inter-regional economic systems.

To understand growing regional imbalances it is necessary to stress that doubt on regional convergence is also grounded in economic theory. Arguments in favor of convergence have been recognized in neoclassical growth theory using an exogenous technological change model (e.g. 1956, 1994). Depending on the assumptions made on preferences and demography, this model predicts unconditional or conditional convergence (Le Pen 2010). On the other hand, divergence is initiated in the theory of endogenous growth (Romer 1986, 1990) and the “new” theory of international trade triggered by Krugman (1991) and Krugman and Venables (1995). In the presence of increasing returns, economic activity is expected to concentrate geographically in a few areas and economic disparities at the regional scale would be more pronounced. For example, most industrial production, skilled labor, and higher wages tend to agglomerate in urban areas where geographical proximity between economic agents facilitates communication and creates an environment which favors frequent interactions and the flow of ideas (Ascani et al., 2012). Following this line of reasoning, we found it worthwhile to analyze the factors that have an impact on these processes.

The absence of empirical and theoretical confirmation of the convergence process may be impacted by the use of imprecise data, different methods for testing the convergence process (Mohl & Hahen 2010; Le Pen 2010), or not taking into consideration significant determinants (Rodriguez-Pose 2013).

As the use of imprecise data and different methods have been in focus of numerous papers (see Mohl & Hahen 2010; Le Pen 2010, Becker et al., 2010; Azomahou et al., 2011), this paper attempts to investigate the role of the institutions on national regional inequalities among member countries in the EU. Researchers across a wide range of social science disciplines are increasingly resorting to analyzing institutions in order to have a better grasp of how economic development takes place (Rodriguez-Pose 2010, p. 5) and, although institutions have been selected as important factors for economic activity on a regional level (Farole et al., 2011; Ascani et al., 2012; Rodriguez-Pose 2013), the mechanisms of the institutional influence on regional inequalities have not been properly challenged.

The remainder of the paper is organized as follows. In the next section, we give a brief summary of previous research on regional inequalities among European regions, define the institution, and discuss the role of the institution on (regional) economic growth. In Section 'Data and Methodology', we present data, empirical methodology, and discuss the empirical results. Section 'Empirical Results and Analyses' concludes with a comment on how further investigation may be relevant.

2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research

The concepts of convergence among economies and disparities are intrinsically linked together (Arabia et al., 2010, p. 3). Therefore, in the first part of this section, we identify relevant empirical facts concerning the convergence process.

Three broad sets of approaches have been used to test for convergence. A first approach is based on the beta (β) and sigma (σ) convergence criteria. The second is the intra-distribution dynamics approach, and the third is based on time series econometrics.

The first approach includes two measures of convergence, which are conceptually different, the sigma and the beta convergence (see Barro & Sala-í-Martin 1995). The sigma convergence describes how the distribution of cross-section incomes evolves over time, whereas the beta convergence emphasizes income mobility. These methods have been modified to include spatial interactions, panel methods, and different time periods (Fingleton 1999; López-Bazo et al., 2004; Ertur et al., 2006, Le Gallo & Dall'Erba 2006), and results of these modifications offer conditional confirmation of the convergence process in the EU. Quah (1993b) shows that a decrease in standard deviation may be compatible with a nonconvergent distribution of the per capita income and that the beta convergence can be consistent with cases where countries do not converge in the long run. Thus, Quah (1993a,1993b, 1996, 1997) proposes the intra-distribution dynamics approach, which uses the stochastic kernel and Markov chain to estimate the shape of the income distribution and intra-distribution mobility, and its change through time. Recent examination has provided ambiguous results. Pittau and Zelli (2006) conclude that the process of integration is characterized by divergence, Ezcurra et al. (2006) show a reduction in polarization, and Hierro and Maza (2009) find evidence of structural shifts in the dynamics of the European income distribution between 1980–1993 and 1993–2005. They conclude “high relative persistence” in this income distribution after the mid-1990 s. The third approach to convergence is based on time series econometrics. The stochastic convergence criterion is proposed by Bernard (1991), Quah (1990), Bernard and Durlauf (1995), and Evans and Karras (1996), and it implies that convergence between two per-capita outputs is accepted if their log-difference is a zero-mean stationary process. An application by De Siano and D'Uva (2006) confirmed the existence of convergence clubs in Europe.

All of these approaches yield inconclusive results and, therefore, do not establish a definitive conclusion. Thus, further investigation should look for the missing piece of the regional growth puzzle pattern, which we now attempt to identify under institution framework.

Under a neoclassical growth framework, achieving economic development is mainly a matter of investing in physical capital (Solow 1956). After development of the endogenous growth theory, two additional factors emerged: (i) innovation (Romer 1986), and (ii) education (Lucas 1988). Hence, the recipe to generate greater economic and social cohesion seemed rather straightforward: greater investment in infrastructure, in education and training, and in the promotion of innovation and industrial activities channeled to lagging regions. However, the adoption of such a policy did not provide expected results (Rodríguez-Pose 1999, p. 3).

Obviously, (European) regional strategies have overlooked an additional determinant of the growth process. But, which one? We suggest that the missing determinant is institutions.

The link between institutions and economic growth had been fundamentally overlooked by mainstream economic theory in general, and growth theory in particular (Rodriguez-Pose 2013). However, in the last 20 years, researchers have increasingly resorted to analyzing institutions in order to have a better grasp of how economic growth takes place. They have made considerable progress in showing that institutions “matter” more for economic growth than traditional factor-endowments (Hall & Jones 1999; Acemoglu et al., 2001; Rodrik et al., 2004; Beugelsdijk et al., 2004; Beugelsdijk & van Schaik 2005; Rodrik 2008).

But how do they impact economic growth patterns?

To deal with the aforementioned issues, we must first define what “institutions” are. Defining an institution is notoriously difficult and current literature cannot agree on a common definition. However, the most commonly cited definition describes institutions as “the rules of the game in a society; (and) more formally, (as) the humanly devised constraints that shape human interaction” (North 1990, p. 477).

Based on this definition, many researchers have tried to establish a link between institutions and economic performance, indicating that a market economy cannot function properly without institutions. They imply, among numerous other channels of influence, that institutions generate trust among economic actors and reduce transaction costs (North 1992), provide collective goods (Streeck 1991), foster transparency (Storper 2005), promote entrepreneurship, and grease the functioning of labor markets (Giddens 1990).

The complexity of institutions is confirmed by the existence of multiple types. A first approach includes formal and informal institutions. As Amin (1999, p. 4) indicates, any economy is molded by “enduring collective forces,” which include “formal institutions such as rules, laws, and organization, as well as informal or tacit institutions such as individual habits, group routines, and social norms and values.” A second approach recognizes political (constitutions, governance structures, checks, and balances), economic (property rights, markets, regulatory structures), and social (formal groups and associations, norms) dimensions of institutions (Farole et al., 2011). A third distinguishes three different processes important for the quality of an institution (Kaufmann et al., 2010). The first represents the process by which governments are selected, monitored, and replaced. The second process boosts the capacity of the government to effectively formulate and implement sound policies, and the third process elevates the respect of citizens and the state for the institutions that govern economic and social interactions among them. Emphasizing the process dimensions of the institution, relevant literature (e.g. Glaeser et al., 2004) favors the third approach. These three different patterns of processes are, therefore, discussed further.

What should we expect?

As has been previously mentioned, the relevant literature on a national level identifies the main channels through which “human devised constraints” might enter into the production function to shape patterns of economic growth. But is this global pattern applicable for a regional level perspective?

Regarding beliefs of institutionalists and their idea that markets are “social constructs made and reproduced through frameworks of socially constructed institutions and conventions” (Pike et al., 2006, p. 91), institutions become much more than simple regulators of economic activity on regional level. They determine the level of activity and its efficiency. But, what are the channels of influence of institutions?

Recent regional literature has indicated different forms of institutions and possible explanations of influence on regional growth process (e.g. Farole et al., 2011; Ascani et al., 2012; Rodriguez-Pose 2013).

Farole et al. (2011) emphasize that an institutional lack of quality, defined as institutional sclerosis, clientelism, corruption, and pervasive rent-seeking by durable local elites (Acemoglu & Robinson 2001), can block local innovation. Farole et al. (2011) also suggest that dysfunctional informal institutions effect lower levels of trust, subsequently declining associative capacity for effective collective actions. Weak institutions may also have a negative influence on the provision of public goods or other determinants of regional economic growth. As a final point, they underline that such inappropriate conditions have cumulative effects that can block convergence among regions.

For a second group of authors, the starting point is the paper by Rodríguez-Pose (1999), in which he indicates that variations in different institutional settings are crucial for regional economic performance. Ascani et al. (2012) argue that physical proximity and co-location between economic agents are not relevant for regional growth. By following other researchers (Knack & Keefer 1997; Zack & Knack 2001; Boschma 2005), they argue that cognitive, organizational, and social proximity are crucial for innovation to take place. Finally, institutional proximity, defined as mechanisms of coordination of the economy, ranging from the legal and regulatory system, to informal cultural norms and habits, has also been recognized as crucial for regional economic activity.

Rodriguez-Pose (2013) presents a variety of possible explanations. He recognizes institutions that can, by lowering uncertainty and information costs, smooth the process of knowledge and innovation transfer within and across regions, and improve the conditions for the development of economic activity. At the same time, Rodriguez-Pose (2013), following a paper by North (1995) indicates that institutions can shape the sets of incentives and disincentives that contribute to establish balance between coordination and competition among local economic actors, therefore, facilitating the learning process. He also indicates the importance of institutional “adaptive efficiency,” representing the help that institutions offer to the territories to adjust and react to change.

In sum, relevant literature stresses that institutions matter for economic activity on a regional level. However, it does not provide a clear answer as to which type of institutional framework is relevant for regional inequalities. Moreover, different forms of institutions are in constant interaction and tend to affect one another in different ways, which makes the process of indentifying the important form and channel even more difficult.

3 Data and Methodology

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research

In this section we explore the impact of institutional framework on regional inequalities on national levels among EU member countries. By choosing the national (macro) level of our investigation, we do not deny the importance of analysis on a lower (micro) level, but we find the national level more appropriate for our research. An explanation for a national (macro) approach starts from the definition of institutions. More precisely, if we follow the most commonly cited definition that describes institutions as “the rules of the game in a society” (North 1990), then we should analyze the subjects that define said rules, which are, in most cases, defined at a national (macro) level. Moreover, regional data on these aspects are limited and/or of poor quality. Thus, at least for the institutional processes that we analyze in this paper, the most appropriate level for research is national.

The dataset allows us to work with a sample of 18 EU member countries (see Appendix I). This study represents a unique attempt to test the impact of different institutional dimensions on regional imbalances in a sample of countries engaged in a one of the most important process of economic integration in the world, the EU.

Before explaining our methodology and the results of the regression analysis, we will first describe the data set. Data has been collected from the Eurostat database and from the Worldwide Governance Indicators (WGI) database for the period 2000–2007. Two groups of variables are particularly important for this study: (i) the measure of regional inequality, and (ii) instruction measures.

Regional inequalities are measured by the sum of the absolute difference between the regional nomenclature of territorial units for statistics (NUTS) II level and national gross domestic product (GDP) per inhabitant, weighted with the share of population, and expressed in percent of the national GDP per inhabitant. The indicator is calculated from regional GDP figures based on the European System of Accounts (ESA95). The dispersion of regional GDP is zero when the GDP per inhabitant in all regions of a country is identical, and rises if there is an increase in the distance between a region's GDP per inhabitant and the country mean. This measure of regional inequality has been chosen for two reasons. Firstly, it is the most important measure of regional inequality in the EU considering that it is the single available measure on the official statistical web page of the EU, Eurostat. Second, and even more important, is the fact that this measure is in line with standards introduced by Portnov and Felsenstein (2010), which are used to test the sensitivity of commonly used income inequality measures to changes in the ranking, size, and number of regions into which a country is divided.

Taking into consideration that defining institutions is notoriously difficult, it is even more challenging to choose an adequate measure for the institutions, or more precisely, an appropriate proxy variable.

As we have previously mentioned, we have focused on the national level of the institutional framework and, therefore, choose a variable that has been proposed by group of authors (see Kaufman et al., 2010) for measuring the quality of governance, the WGI. There are several reasons for choosing the WGI.

Firstly, there is a resemblance between the WGI definition of governance and institutions. Although there is no strong consensus around a single definition of governance, the definition offered by Kaufmann et al. (2010, p. 3) defines governance as “rules, enforcement mechanisms, and organizations,” which covers an essential part of the institutional definition and, therefore, the WGI indicator is the most appropriate proxy variable for institutional quality.

Secondly, these indicators are based on several hundred variables obtained from 31 different data sources, such as nongovernmental organizations, commercial business information providers, and public sector organizations, which allow us to incorporate different types of institutional framework in our research.

Moreover, the six dimensions of the WGI dataset (Voice and Accountability [VA], Political Stability and Absence of Violence/Terrorism [PV], Government Effectiveness [GE], Regulatory Quality [RQ], Rule of Law [RL], and Control of Corruption [CC]) allow us to explore our investigation on particular institutional processes important for regional inequalities. The first two indicators represent the process by which governments are selected, monitored, and replaced. The third and fourth indicators represent the process which boosts the capacity of the government to effectively formulate and implement sound policies, and the last two indicators represent the process which elevates the respect of citizens and the state for the institutions that govern economic and social interactions among them.

Finally, the WGI dataset enables users to avoid over-interpreting small differences between countries and over time that are unlikely to be statistically or practically significant, which separates the WGI from other possible variables for measuring institutional frameworks (Kaufmann et al., 2010).

An assumption that only the institution influences regional inequalities is rather restrictive, and results can potentially suffer from the omission of other (possibly) significant determinants of regional inequalities. Thus, we test whether the relationship between regional inequalities and institutions holds when including additional explanatory variables. To the best of our knowledge, this paper represents a unique empirical study, which addresses different dimensions of institutional framework important for regional growth inequalities; therefore, it was not possible to follow a similar variable selection procedure. However, we have considered relevant papers, such as Barrios and Strobl (2009), Rodriguez-Pose and Ezcurra (2010), Lessman (2012), and Ezcurra and Rodríguez-Pose (2013).

The first explanatory variable considered is a measure of national trade openness. This is an important inclusion given that technological spillovers, which have been found to be important in the literature for regional inequalities, are related to trade intensity (Coe & Helpman 1995; Gianetti 2002). The models of new economic geography indicate the relevance of the link between trade and spatial inequality and, depending on the theoretical assumptions employed in each case, they assume different outcomes (Krugman & Livas Elizondo 1996; Puga & Venables 1999; Paluzie 2001). The empirical literature uses the ratio of total trade (import + export) to GDP in order to measure trade openness. However, Alcalá and Ciccone (2004) have criticized the use of this index and instead propose alternative indices: the real openness index, which is the sum of imports plus exports expressed in common currency (here the Euro), relative to the GDP expressed in purchasing poverty parity (PPP)1 terms.

The second variable to be considered is a measure of fiscal decentralization, which also may have been a cause of regional divergence in the EU (Rodríguez-Pose 1996, Rodríguez-Pose & Gill 2003; Rodriguez-Pose & Ezcurra 2010; Lesmann 2012). The starting position in the literature points to Oates theorem on fiscal decentralization according to which differences in preferences about public goods across regions will require decentralized provision of such goods in order to improve regional economic performance. In contrast, several authors have had rather contradictory results, finding little evidence for a significant positive effect of fiscal decentralization on regional growth. Tanzi (1996) stresses that decentralization might cause coordination problems, excessive regulation, higher administrative costs, or a poor quality of local bureaucrats. Lessmann and Markwardt (2010) indicate that decentralization might increase corruption and cronyism, undermining potential efficiency gains. Despite theoretical doubts, in papers by Rodríguez-Pose and Ezcurra (2010), Lessmann (2012), and Ezcurra and Rodríguez-Pose (2013), the authors find empirical evidence for the significant influence of decentralization on regional inequalities. Thus, in order to control for the possible influence of fiscal decentralization, we use the indicator expressed as the sum of the shares of local and state revenues as a percentage of national GDP.

The third additional explanatory variable tries to catch the impact of EU regional policy on regional inequalities among EU member states. Although evidence of the effective impact of EU structural funds is controversial (see Boldrin & Canova 2001; Beugelsdijk & Eijffinger 2003; De la Fuente 2002; Becker et al., 2010), our dataset covers 18 EU member states in order to incorporate EU efforts to decrease regional imbalances among EU regions. Thus, we use the level of structural funds per capita as a control variable.

The final variable to be considered is a measure of regional industrial specialization. The use of this index was made popular after the study by Kenen (1969), indicating that sectoral specialization may play an important role in determining regional economic growth patterns. The idea is that two regions with dissimilar industrial structures should also have differing growth experiences. Moreover, regions highly specialized in agriculture and/or traditional manufacturing activities should be less prone to adopt new technologies during periods of innovation-led economic growth and should experience less growth than more technologically advanced regions (Barrios & Strobl 2009). Gianetti (2002) showed that European regions with similar technological capabilities (directly linked to the specialization of regions in traditional sectors) have converged substantially, while the others have displayed some tendency to diverge over the period considered. We use the country/year average of the so-called adjusted Krugman specialization index (AK). The Krugman index (K) corresponds to the expression:

  • display math

where Xsjt is the share of sector s in total employment of region j at a given year t, and Xskt is the share of sector s in total employment of region k at a given year t. Considering our focus on regional inequalities at a national level, we have used the adjusted Krugman index, which corresponds to the following expression:

  • display math

where AKi is the value of the adjusted Krugman index for country i, inline image is the share of sector s in the total employment of region j in country i at year t, and inline image is the share of sector s in the total employment of country i at a given year t, while Nji is the number of regions in country i and Ns is the number of sectors.

Thus, the starting point for empirical confirmation is the model with the structure:

  • display math(1)

where Yit is a measure of regional inequalities, measured by the sum of the absolute differences between regional (NUTS II level) and national GDP per inhabitant, weighted with the share of the population and expressed in percent of the national GDP per inhabitant in country i for period t. inline image is 1×K matrix of control variables, where K is the number of control variables in model and variable Iit institution quality in country i for a period t. β is K×1 vector of parameters and δ is parameter. μ is a constant term. It is assumed that ϵit are inline image, identically and independently distributed error terms. We also assume the country specific part of error term αi are inline image, identically and independently distributed error terms.

In the process of choosing control variables, we have considered related literature (Barrios & Strobl 2009; Rodriguez-Pose & Ezcurra 2010; Lessmann 2012; Ezcurra & Rodríguez-Pose 2013), keeping in mind all of the specific needs of this unique empirical investigation. Thus, national trade openness is measured by real trade index, regional industrial specialization is measured by an adjusted Krugman index, fiscal decentralization is measured by the sum of the shares of local and state revenues in GDP, EU regional policy is measured by amount of structural funds per capita, and institution quality Iit is measured by the WGI indicator, which includes six dimensions of indicator quality (VA, PV, GE, RQ, RL, and CC) in country i for period t.

We assume that regional inequalities represent a dynamic relationship, thus, their current value depends upon past values. We, therefore, introduce a modified equation, which includes the dynamic behavior of the dependent variable, characterized by the presence of a lagged dependent variable among the regressors:

  • display math(2)

In this case, the recently used fixed effect estimator2 is biased and inconsistent because with the inclusion of the lagged dependent variable inline image in the model, it becomes correlated with αi. Even though the fixed effect estimator becomes consistent when T increases, bias does not vanish as the number of individuals increases (Nickel 1981). Therefore, in our case, a fixed effect estimator is not appropriate. A random effect estimator is also biased and inappropriate for the estimation of Equation (2).3

Therefore, we have used an estimator from the dynamic panel model. Anderson and Hsiao (1981) proposed first altering the model in order to exclude αi. Additionally, they proposed inline image or inline image as an instrument for inline image. Their estimation method is consistent, but is not an efficient method. Arellano and Bond (1991) proposed a new estimator for the dynamic panel model. To solve the problem of correlation between inline image and αi Arellano and Bond (1991) proposed taking Equation (2) in first differences:

  • display math(3)

In the equation with first differences αi is excluded, but a new problem is observed. Namely, in Equation (3) inline image is correlated with inline image. To solve this problem, instrumental variables are included in the model. Valid instruments for inline image are lagged values of the dependent variable in level inline image. Furthermore, if some of inline image is endogenous in the sense that inline image for s > t and inline image otherwise lagged values of the independent variable inline image are valid instruments4 for inline image. This estimator outperformed previous estimators in terms of bias for large N and fixed T (Arellano & Bond 1991; Judson & Owen 1999; Santos & Barrios 2011). It also revealed weaknesses when the dependent variable is high persistent, and when the ratio of the individual effect variance and the remainder error variance (inline image) increases. Blundell and Bond (1998) followed the idea advanced by Arellano and Bover (1995) and proposed a new system, the generalized method of moments (GMM) estimator. Instead of the equation in the first differences (3) system, GMM5 uses an equation in level (2). This estimator showed better properties than that used by Arellano and Bond, as well as estimators used in numerous other studies (Blundell & Bond 1998, 2000; Bond 2002; Soto 2009; Bun & Windermeijer 2010). However, both of the estimators mentioned are suitable for data sets with a large number of individuals and a small number of periods. Our data set contained a moderately large number of individuals (18) and a relatively small number of periods (8). Considering that both estimators use instrumental variables for a lagged dependent variable and for endogenous “independent” variables, the problem of too many instruments can occur. There is convincing evidence that too many instrument conditions introduce bias while increasing efficiency (Baltagi 2008). Simply by being numerous, instruments can over fit instrumented variables, filing to expunge their endogenous components and biasing coefficient estimates toward those from non-incrementing estimators (Roodman 2009a). Roodman (2009b) recommended that the number of instruments should be lower than the number of cross-sections. To avoid the problem of too many instruments, we used Arellano and Bond's estimator. In our case, the number of instruments was between 16 and 20 in all estimated models. If we had used the Blundell and Bond estimator, the number of instruments would be significantly higher (minimally for six instruments). So the number of instruments used in each model was marginal and did not produce significant bias or significantly reduce the quality of the Sargan test.6 We used only two or three lags of the dependent variable as instruments. We did not use instruments for independent variables, because the Sargan test indicated that there was no problem with endogeneity.

The two-step Arellano and Bond GMM estimator is used because a one step estimator assumes the error terms to be independent and homoscedastic across countries and over time. A two-step estimator relaxes the assumption of independence and homoscedasticity by using the residuals obtained from the first step estimation to construct a consistent estimate of the variance–covariance matrix.

The validity of the instruments used in the model is tested using the Sargan test. The results of the Sargan test show that the instruments are valid and that in a specified model there is no problem of endogenity. Testing the autocorrelation in residuals is also performed using the m1 and m2 tests. The null hypothesis of the m1 test assumes the absence of a first-order autocorrelation between the differenced residuals, and the null hypothesis of the m2 test assumes the absence of a second-order autocorrelation between the differenced residuals.

Before turning to econometric testing of the hypothesis, we provide some descriptive statistics regarding the regional GDP per capita inequalities for the period 2000–2007 for the selected 18 EU countries (Table 1).

Table 1. Dispersion of regional (NUTS II level) gross domestic product per capita for the period 2000–2007
CountryRegional inequalities
  • *

    Available data for the year 2000 and the period 2005–2007.

Source: Eurostat.
Belgium25.3
Bulgaria25.5
Czech Republic24.7
Denmark*15
Germany17.5
Greece24.6
Spain19.2
France20.4
Hungary35.1
Netherlands11.2
Austria17.4
Poland18.7
Portugal22.7
Romania25.1
Slovakia28.8
Finland16.2
Sweden15.2
United Kingdom22.3

According to Table 1, Hungary has the highest regional inequalities, while the Netherlands has the lowest level. The table shows that new member state countries display, on average, higher regional inequalities than older member states. There is also a huge variation in the levels of regional imbalances among EU member countries, which makes this investigation even more attractive. As a more appropriate introduction to our empirical investigation, we offer summary statistics for all relevant variables in Table 2. Along with the enormous difference among regional imbalances (disp), it is interesting to note the scope of our proxy variable WGI (represented in the table with symbol I).

Table 2. Summary statistics
 MeanOverall standard deviationBetween standard deviationWithin standard deviationMinimumMaximumObs
  1. AK, adjusted Krugman index; CC, control of corruption; disp, measure for regional inequalities; fiscal decent, fiscal decentralization; GE, government effectiveness; I, institutions; obs –observations; PV, political stability and absence of violence; RL, rule of law; RQ, regulatory quality; Sf, structural funds; VA, voice and accountability.

Disp21.5755.85485.77441.678710.637.8140
AK0.07660.04860.46470.01430.00940.3023175
Sf per capita59.667875.575265.085640.21460392.5347270
Fiscal decen11.98517.47327.49371.25380.636.6270
Real openness1043.214084.4734095.599642.810323.698926547.55269
I1.11740.46610.46720.0787−0.07591.900243
CC1.11540.74900.75060.1274−0.34302.4665243
PV0.85370.36510.33580.1559−0.17981.5768243
VA1.17510.31390.30910.07810.34401.8266243
GE1.20360.60280.59330.1515−0.12662.124243
RL1.13160.56900.57050.0949−0.15551.9640243
RQ1.2250.40080.39210.1094−0.10452.0120243

The units of our proxy indicators for institutional environment range from a minimum of approximately −2.5 to a maximum of 2.5. The value of the WGI variable (I) demonstrated is larger between the standard deviation than within the standard deviation, which is in line with the expectation that the quality of institutions is a relatively high time-consistent variable for all countries. A similar situation exists for all six dimensions of the WGI variable, showing diversity among them, mainly between standard deviations. This is especially the case for the CC variable, which is twice as large as the VA or PV variations.

After introducing descriptive statistics, we can now present econometric methods for challenging the two key dimensions of our research. The first tackles the importance of institutional quality for regional inequalities in EU countries, the second attempts to discern which dimension of the institutional quality is important for the pattern of regional imbalance.

4 Empirical Results and Analyses

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research

All econometric details integrated into our analysis are presented in Table 3. The coefficients of control variables are all statistically significant. The only exception is the adjusted Krugman index; however, this is in line with results found by Barrios and Strobl (2009). The deviation of the coefficients of control variables from expected signs can be reasonably explained and do not influence the value of the coefficient, which is our main interest. For example, the statistically significant positive effect of structural funds for regional inequalities is a direct consequence of propositions and a short period of implementation. However, the coefficient on real openness does not have the expected result, because the small value (influence) on regional inequalities could not be recognized as statistically significant.

Table 3. The results of the two-step Arellano-Bond dynamic panel estimator
VariableModel
Const.10.7997***
Lagged disp20.6918***
AK−4.0518
Sf per capita0.0071***
Fiscal decen−0.1565***
Real openness−0.0001**
I−1.6761**
Number of observations84
Sargan test (P-value)0.4512
m1-test (P-value)0.1081
m2-test (P-value)0.7611
Correlation coefficients
VariableAKSf per capitaFiscal decenReal opennessI
  1. *, **, *** Indicate significance at 10%, 5%, and 1% level.

  2. AK, adjusted Krugman index; Const, Const-constant term; fiscal decen, fiscal decentralization; Sf, structural funds.

AK1    
Sf per capita0.14781   
Fiscal decen−0.33−0.1691  
Real openness−0.0999−0.04340.03061 
I−0.5393−0.01750.513−0.071

Finally, Table 3 offers negative and significant coefficients for institutional quality, which could be interpreted as evidence for the first contribution of this paper, that is, for the hypothesis that institutional quality has an important negative influence on the level of regional inequality.

It should be stressed that diagnostic (the Sargan and m2 statistics) tests for the estimated model in Table 3 are satisfying at a five percent confidence level and, therefore, the proposed model is well specified. The correlation coefficients between each of the variables are reported in the lower part of Table 3. The highest coefficient of correlation is –0.5393, indicating that we should not expect a high risk of multicollinearity problems between the variables of our interest.

The literature indicates that institutions are an important force in the regional growth process (e.g. Farole et al., 2011; Ascani et al., 2012; Rodriguez-Pose 2013). But how do institutional factors contribute to regional inequalities; or how do institutions shape the ability of an economy to use and develop its resources in particular ways? These questions still require attention.

Therefore, we investigate which processes of institutional quality are important to regional inequalities. The relevance of this effort is based on three notions: (i) the literature indicates that institutions matter for economic activity on regional levels by proposing various possible explanations of the influence on the regional growth process (Farole et al., 2011; Ascani et al., 2012; Rodriguez-Pose 2013); (ii) relevant literature (Glaser 2004) emphasizes the process dimension of the institution; and (iii) the aggregate WGI with all six indicators can be recognized, not only as a proper proxy for institution, but also for the process dimension of the institution.

We next focus our attention on the six indicators included in the WGI.

  1. VA: Captures perceptions of the extent to which a country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media.
  2. PV: Captures perceptions of the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means, including politically motivated violence and terrorism.
  3. GE: Captures perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies.
  4. RQ: Captures perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development.
  5. RL: Captures perceptions of the extent to which agents have confidence in and abide by the rules of society, and, in particular, the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence.
  6. CC: Captures perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as “capture” of the state by elites and private interests.

The indicators are based on several hundred variables obtained from 31 different data sources, which allow them to be comprehensive and representative pointers of the different dimensions of the institutional frame. Moreover, indicating the most significant indicator(s) from the aggregate WGI allows us to focus our efforts in right direction. Namely, each of the six dimensions of the WGI is based on different data sources. Therefore, if, for example, the CC indicator shows the most significant level of regional inequality, researchers have the opportunity to focus on only 22 data sources (Kaufmann et al., 2010). Because the 22 data sources provide 1950 country/indicator data points, further investigation can be more effective.

The next step of our research commences by testing the influence of various institutional quality indicators on regional inequalities in the EU. We estimate six versions of the model given in Equation (2), with the exception that variable Iit now represents n dimensions of institutional quality measured by the different indicators (n = 1 to 6, where 1 stands for VA, 2 for PV, 3 for GE, 4 for RQ, 5 for RL, and 6 for CC).

The results presented in Table 4 indicate that we should expect a high risk of multicollinearity problems between these variables, and that we should test these indicators separately. Therefore, there are six models, each including a control variable vector and a different indicator for institutional quality. Thus, the first model includes the VA indicator, the second the PV indicator, the third the GE indicator, the fourth the RQ indicator, the fifth the PL indicator and the sixth model the CC indicator.

Table 4. Correlation coefficients of institutional quality indicators
VariableCCPVVAGERLRQ
  1. CC, control of corruption; GE, government effectiveness; PV, political stability and absence of violence; RL, rule of law; RQ, regulatory quality; VA, voice and accountability.

CC1     
PV0.58381    
VA0.90960.641   
GE0.9490.60630.91421  
RL0.9480.64760.92820.94221 
RQ0.87940.58230.86640.90230.89671

As previously explained, we have used an adjusted two-step Arellano and Bond GMM estimator. Table 5 encompasses the results of testing these six models.

Table 5. The results of the two-step Arellano-Bond dynamic panel estimator for the six indicators of the Worldwide Governance Indicator (WGI)
VariableModel 1Model 2Model 3Model 4Model 5Model 6
Const.9.6130***9.0647***9.9428***13.26931***12.3103***12.5813***
Lagged disp20.6877***0.6520***0.7329***0.5383***0.6479***0.5929***
AK−7.33425−18.72206**2.379746−38.40294***−2.9778−11.71705
Sf per capita0.0817***0.0077***0.00871***0.011351***0.0085***0.00689***
Fiscal decen−0.0981*−0.0683−0.2045***−0.1112−0.1248**−0.1237**
Real openness−0.00007*−5.0000−0.00017***−0.00014***−0.00008−0.00004
Voice and Accountability (VA)−1.0111***     
Political Stability and Absence of Violence/Terrorism (PV) 0.377292*    
Government Effectiveness (GE)  −1.142454*   
Regulatory Quality (RQ)   0.2635324  
Rule of Law (RL)    −2.721868*** 
Control of Corruption (CC)     −1.364107***
Number of observations848484848484
Sargan test (P-value)0.53730.57210.35840.17960.33040.4064
m1-test (P-value)0.08950.09740.12360.19770.09250.1141
m2-test (P-value)0.88350.93440.81040.69360.80920.5929
Correlation coefficients
VariableCCPVVAGERLRQ
  1. *, **, *** Indicate significance at 10%, 5%, and 1% level.

  2. AK, adjusted Krugman index; Const, Const-constant term; decen, decentralization; Sf, structural funds.

AK−0.4594−0.3888−0.55−0.532−0.4975−0.6299
Sf per capita−0.0647−0.05440.0434−0.01270.02050.0043
Fiscal decen0.57150.20880.51410.54190.46280.4463
Real openness−0.12370.0263−0.0334−0.0914−0.0711−0.0169

Before we introduce our main results it should be stressed that diagnostic (the Sargan and m2 statistics) tests for estimated models in Table 5 are satisfying at a five percent confidence level and, therefore, the proposed model is well specified. The correlation coefficients between each of the variables are reported in the lower part of Table 5. The highest coefficient of correlation is −0.6299 for the adjusted Krugman index (AK) and RQ and it may explain the fact that the coefficient of RQ is not statistically significant. The other coefficients indicate that we should not expect a high risk of multicollinearity problems between variables of our interest.

The interpretation of this result should start with indicating the value of the coefficient in front of the lagged dependent variable (Lagged disp2). The value of the coefficient, either close to or higher than 0.6, confirms that the dynamic panel model suits our investigation.

The coefficient of the adjusted Krugman index (AK) indicates that regional specialization can bring regional inequalities on lower levels. Statistically significant positive effects of structural funds (SF per capita) for regional inequalities can be interpreted as direct consequences of propositions and a short period of implementation. Fiscal decentralization (fiscal decen) seems to have an ambiguous effect on regional inequalities, in line with the literature, which offers little evidence for a significant positive effect of fiscal decentralization on regional growth (Tanzi 1996, Lessmann & Markwardt 2010). Openness (real openness) of the economy does not show the expected result, but because of the small value (influence) of regional inequalities, this could not be recognized as statistically significant.

The most important result in Table 5 is the fact that the model revealed the role of different dimensions of institutional quality to regional inequalities, recognizing key dimensions presented by the VA, RL, and CC indicators.

These results show that for regional growth, inequalities are the most important processes that elevate the respect of citizens and the state for the institutions that govern economic and social interactions among them and, to some extent, the process by which governments are selected, monitored, and replaced. The process, which boosts the capacity of the government to effectively formulate and implement sound policies, is not significant to the regional growth pattern.

What it is additional value of these results?

The first important dimension of these results can be recognized by the interpretation of empirical evidence. The results emphasize two indicators (RL and CC), which can be interpreted as support for the standard “rules of engagement” that reduce transaction costs by facilitating the mutual trustworthiness of individual economic agents and lower the uncertainty that spreads equally among regions in specific countries.

Additionally, the significant indicators in our analysis signal the functioning of the political frame demonstrating that institutions also shape regional inequalities indirectly through political channels. This point expands on the literature by indicating that this fundamental channel through which institutions determine economic outcomes is not only relevant at the aggregate national level (e.g. Tabellini 2005; Acemoglu et al., 2004), but also at a regional level. This is a presumption that has been represented in the literature (La Porta et al., 1999; Stasavage 2000; Aghion & Howitt 2005) by showing that openness of political and economic participation, political competition, and the existence of “checks and balances” are critical for the link between institutional quality and economic growth and obviously have significant spatial dimension.

Our results indicate a pathway for further research. Each of the six dimensions of the WGI is based on different data sources. In highlighting the most significant indicator(s) from the aggregate WGI, our effort was focused on finding the mechanism of institutional influence on regional imbalances. More precisely, by emphasizing the role of “processes that elevate the respect of citizens and the state for the institutions that govern economic and social interactions among them,” researchers have the opportunity to focus on concepts measured especially for these two indicators (RL and CC).

5 Conclusion

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research

In this paper we have examined the link between institutions and regional inequalities among EU member countries and found evidence of a relationship between these two variables.

Our paper offers rare empirical confirmation that institutional quality is relevant to regional inequality among EU member countries, by determining which processes are important for the link between institutions and regional inequalities. Our results are in line with relevant literature, indicating that institutions are an important force in the regional growth process (Farole et al., 2011; Ascani et al., 2012; Rodriguez-Pose 2013).

Additionally, we have analyzed how institutional factors contribute to regional inequalities, or which channels are used.

The relevance of this effort is based on three notions: (i) the literature indicates that institutions matter for economic activity on regional levels, but do not propose explicit explanations of their influence on the regional growth process (Farole et al., 2011; Ascani et al., 2012; Rodriguez-Pose 2013); (ii) the literature (Glaser 2004) emphasizes the process dimension of the institution; and (iii) the aggregate WGI indicator with all six dimensions can be recognized not only as a proxy for institutions, but also for different aspects of the institutional influence. Therefore, we investigated which processes of institutional quality are important to regional inequality by testing the six dimensions of the aggregate WGI.

Importantly, our results emphasize the two indicators that represent the processes that elevate the respect (of citizens and the state for the institutions that govern economic and social interactions among them). We, therefore, interpret this as support for the standard “rules of engagement” that reduce transaction costs by facilitating the mutual trustworthiness of individual economic agents and lowering the uncertainty that spreads equally among regions in specific countries.

By offering empirical support for the standard rules of engagement and demonstrating the importance of reducing transaction costs to lower the level of regional inequalities, this paper offers direction for further research. Firstly, it shows policymakers that only greater investment in infrastructure, in education and training, and in the promotion and innovation of industrial activities will generate lower regional inequalities. Moreover, our results demonstrate that interdependence between social and institutional proximity (represented in the paper by Ascani et al. [2012]) indicates the necessity of combining different institutional factors for harmonious regional development.

In addition, our results show that institutions also shape regional inequalities indirectly through political channels. This expands on the literature by indicating that these channels are not only relevant at the aggregate national level (Tabellini 2005; Acemoglu et al., 2005), but also on a regional level.

These results contain a strong message, not only for policymakers on national level, but also for those on local and regional levels. They stress importance of a bottom-up and coordinated approach in order to decrease levels of regional inequalities in EU member countries.

Notes
  1. 1

    Defintion: Purchasing power parities (PPPs) are indicators of price level differences across countries. They indicate how many currency units a particular quantity of goods and services costs in different countries. PPPs can be used as currency conversion rates to convert expenditures expressed in national currencies into an artificial common currency (the Purchasing Power Standard, PPS), thus eliminating the effect of price level differences across countries. In particular, PPPs can be used to compare the Gross Domestic Product (GDP) of different countries without the figures being distorted by differing price levels in those countries. For details please see: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2007:336:0001:0024:EN:PDF.

  2. 2

    The fixed effect estimator is the Least Squared Dummy variable estimator.

  3. 3

    For more information about the bias of the random effect estimator, see Baltagi (2008).

  4. 4

    In empirical research, the subset of instruments is usually used.

  5. 5

    For a detailed explanation of this estimator, see Blundell and Bond (1998).

  6. 6

    Bowsher (2002) showed that merely keeping the instrument count below N does not safeguard the Sargan test (the test for the validity of instrumental variables).

References

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Related Literature on Regional Inequality and the Role of the Institution in the Regional Economic Growth Process
  5. 3 Data and Methodology
  6. 4 Empirical Results and Analyses
  7. 5 Conclusion
  8. Acknowledgment
  9. References
  10. Appendix I: List of the countries included in the research
  • Acemoglu, D. and Robinson, J.A. 2001. A Theory of Political Transitions. American Economic Review, American Economic Association 91(4), 938963.
  • Acemoglu, D., Johnson, S. and Robinson, J.A. 2001. The Colonial Origins of Comparative Development: An Empirical Investigation. American Economic Review, American Economic Association 91(5), 13691401.
  • Acemoglu, D., Johnson, S.H. and Robinson, J.A. 2004. Institutions as the Fundamental Cause of Long-Run Growth. NBER Working Paper No. W10481, Available from SSRN: http://ssrn.com/abstract=541706.
  • Aghion, P. and Howitt, P. 2005. Joseph Schumpeter Lecture Appropriate Growth Policy: A Unifying Framework. Journal of the European Economic Association, MIT Press, 4(2–3), 269314 04-05.
  • Alcalá, F. and Ciccone, A. 2004. Trade and Productivity. Quarterly Journal of Economics 129(2), 613646.
  • Amin, A. 1999. An Institutionalist Perspective on Regional Development. International Journal of Urban and Regional Research 23, 365378.
  • Anderson, T.W. and Hsiao, C. 1981. Estimation of Dynamic Models with Error Components. Journal of the American Statistical Association 76(375), 598606. DOI: 10.1080/01621459.1981.10477691
  • Arbia, G., Battisti, M. and Di Vaio, G. 2010. Institutions and Geography: Empirical Test of Spatial Growth Models For European Regions. Economic Modelling 27(1), 1221.
  • Arellano, M. and Bond, S. 1991. Some Test of Specification for Panel Data, Monte Carlo Evidence and Application to Employment Equations. Review of Economic Studies 58(2), 277297. DOI: 10.2307/2297968
  • Arellano, M. and Bover, O. 1995. Another Look at the Instrumental Variable Estimation of Error-Components Models. Journal of Econometrics 68(1), 2952. DOI: 10.1016/0304-4076(94)01642-D
  • Ascani, A., Crescenzi, R. and Iammarino, S. 2012. Regional Economic Development: A Review. SEARCH Working Paper WP1/03 http://www.ub.edu/searchproject/wp-content/uploads/2012/02/WP-1.3.pdf (accessed 15 June 2013).
  • Azomahou, T.T., El ouardighi, J., Nguyen-Van, P. and Pham, T.K.C. 2011. Testing Convergence of European Regions: A Semiparametric Approach. Economic Modelling 28(3), 12021210.
  • Baltagi, B.H. 2008. Econometric Analysis of Panel Data. fourth edition. John: Wiley & Sons, New York.
  • Barrios, S. and Strobl, E. 2009. The Dynamics Of Regional Inequalities. Regional Science and Urban Economics 39(5), 575591.
  • Barro, R.J. and Sala-í-Martin, X. 1995. Economic Growth. McGraw-Hill, New York.
  • Becker, S.O., Egger, P.H. and von Ehrlich, M. 2010. Going NUTS: The Effect Of EU Structural Funds On Regional Performance. Journal of Public Economics 94(9–10), 578590.
  • Bernard, A.B. 1991. Empirical Implications of the Convergence Hypothesis. Working Paper 3717. Standford University, Center for Economic Policy Research.
  • Bernard, A.B. and Durlauf, S. 1995. Convergence in International Output. Journal of Applied Econometrics 10, 97108.
  • Beugelsdijk, M. and Eijffinger, S.C.W. 2003. The Effectiveness of Structural Policy in the European Union: An Empirical Analysis for the EU-15 in 1995–2001. Journal of Common Market Studies, Wiley Blackwell 43(1), 3751.
  • Beugelsdijk, S. and van Schaik, A. 2005. Differences in Social Capital Between 54 Western European Regions. Regional Studies 39, 10531064.
  • Beugelsdijk, S., de Groot, H.L.F. and van Schaik, A. 2004. Trust and Economic Growth. Oxford Economic Papers 56, 118134.
  • Blundell, R. and Bond, S. 1998. Initial Conditions and Moment Restrictions in Dynamic Panel Data Models. Journal of Econometrics 87(1), 115143. DOI: 10.1016/S0304-4076(98)00009-8
  • Blundell, R. and Bond, S. 2000. GMM Estimation with Persistent Panel Data: An Application to Production Functions. Econometric Reviews 19(3), 321340. DOI: 10.1080/07474930008800475
  • Boldrin, M. and Canova, F. 2001. Europe's Regions: Income Disparities and Regional Policies. Economic Policy 16, 206253.
    Direct Link:
  • Bond, S.R. 2002. Dynamic panel data models: a guide to micro data methods and practice. Portugese Economic Journal 1, 141162.
  • Boschma, R. 2005. Proximity and innovation: A critical assessment. Regional Studies 39, 6174.
  • Bowsher, C.G. 2002. On testing overidentifying restrictions in dynamic panel data models. Economics Letters 77(2), 211220.
  • Brakman, S. and van Marrewijk, C. 2008. It's a Big World After All: On the Economic Impact of Location and Distance. Cambridge Journals of Regions, Economy and Society 1, 411437.
  • Bun, M.J.G. and Windemeijer, F. 2010. The Weak Instrument Problem of the system GMM Estimator in Dynamic Panel Data Models. Econometric Journal 13(1), 95126.
  • Coe, D.T. and Helpman, E. 1995. International R&D Spillovers. European Economic Review 39(5), 859887.
  • De la Fuente, A. 2002. Does Cohesion Policy Work? Some General Considerations and Evidence from SPAIN, IAE- CSIC Barcelona, mimeo.
  • De Siano, R. and D'Uva, M. 2006. Club Convergence in European Regions. Applied Economics Letters 13, 569574.
  • Ertur, C., Le Gallo, J. and Baumont, C. 2006. The European Regional Convergence Process, 1980–1995: Do Spatial Regimes and Spatial Dependence Matter? International Regional Science Review 29, 334.
  • Evans, P. and Karras, G. 1996. Convergence Revisited. Journal of Monetary Economics 37, 249265.
  • Ezcurra, R. and Rodríguez-Pose, A. 2013. Political Decentralisation, Economic Growth and Regional Disparities in the OECD. Regional Studies 2013 (forthcoming).
  • Ezcurra, R., Pascual, P. and Rapún, M. 2006. Regional Polarization in the European Union. European Planning Studies 14, 459484.
  • Farole, T., Rodríguez-Pose, A. and Storper, M. 2011. Cohesion Policy in the European Union: Growth, Geography, Institutions. Journal of Common Market Studies 49(5), 10891111.
  • Fingleton, B. 1999. Estimates of Time to Economic Convergence: An Analysis of Regions of the European Union. International Regional Science Review 22, 535.
  • Gianetti, M. 2002. The Effects of Integration on Regional Disparities: Convergence, Divergence or Both? European Economic Review, 46(3), 539567.
  • Giddens, A. 1990. The Consequences of Modernity. Stanford University Press, 1990.
  • Glaeser, E.L., La Porta, R., Lopez-de-Silanes, F. and Shleifer, A. 2004. Do Institutions Cause Growth? Journal of Economic Growth, Springer, 9(3), 271303.
  • Hall, R.E. and Jones, C.I. 1999. Why Do Some Countries Produce So Much More Output Per worker Than Others? Quarterly Journal of Economics 114(1), 83116.
  • Hierro, M. and Maza, A. 2009. Structural Shifts in the Dynamics of the European Income Distribution. Economic Modelling 26, 733739.
  • Judson, R.A. and Owen, A.L. 1999. Estimating Dynamic Panel Dana Models: A Guide for Macroeconomists. Economics Letters 65(1), 915.
  • Kaufmann, D., Kraay, A. and Mastruzzi, M. 2010. The Worldwide Governance Indicators: Methodology And Analytical Issues. Policy Research Working Paper Series 5430, The World Bank.
  • Kenen, P.B. 1969. The Theory of Optimum Currency Areas: An Eclectic View. In Mundell, R. and Swoboda, A.K. (eds). Monetary Problems of the International Economy. University of Chicago Press, Chicago, IL 4160.
  • Knack, S. 2003. Groups, Growth and Trust: Cross-Country Evidence on the Olson and Putnam Hypotheses. Public Choice 117, 341355.
  • Knack, S. and Keefer, P. 1997. Does Social Capital Have an Economic Payoff? A Cross-Country Investigation. Quarterly Journal of Economics 112(4), 12511288.
  • Krugman, P. 1991. Geography and Trade. Leuven University Press, Leuven; MIT Press, Cambridge, MA.
  • Krugman, Paul R. and Venables, Anthony J. 1995. Globalization and the Inequality of Nations. The Quarterly Journal of Economics, MIT Press, 110(4), 857880.
  • Krugman, Paul. and Elizondo, R.L. 1996. Trade policy and the Third World metropolis. Journal of Development Economics, 49(1), 137150.
  • La Porta, R., Lopez-de-Silane, F., Shleifer, A. and Vishny, R. 1999. The Quality of Government. Journal of Law, Economics and Organization, 15(1), 222279.
  • Le Gallo, J. and Dall'Erba, S. 2006. Evaluating the Temporal and Spatial Heterogeneity of the European Convergence Process, 1980–1999. Journal of Regional Science 46(2), 269288.
  • Le Pen, Y. 2011. A Pair-Wise Approach to Output Convergence between European Regions. Economic Modelling 28(3), 955964.
  • Lessmann, C. 2012. Regional Inequality and Decentralization. An Empirical Analysis. Environment and Planning A 44, 13631388.
  • Lessmann, C. and Markwardt, G. 2010. One Size Fits All? Decentralisation, Corruption and the Monitoring of Bureaucrats. World Development 38(4), 631646.
  • López-Bazo, E., Vayá, E. and Artís, M. 2004. Regional Externalities and Growth: Evidence from European Regions. Journal of Regional Science 44, 4373.
  • Lucas, R.E. 1988. On the Mechanics of Economic Development. Journal of Monetary Economics 22, 342.
  • Mohl, P. and Tobias, H. 2010. Do EU Structural Funds Promote Regional Growth? New Evidence from Various Panel Data Approaches. Regional Science and Urban Economics 40, 353365.
  • Nickel, S. 1981. Biases in Dynamic Model with Fixed Effect. Econometrica 49(6), 14171426.
  • North, D.C. 1990. Institutions, Institutional Change and Economic Performance. Cambridge University Press: New York.
  • Paluzie, E. 2001. articles: Trade policy and regional inequalities. Papers in Regional Science, Springer, 80(1), 6785.
  • Petrakos, G., Rodriguez-Pose, A. and Rovolis, A. 2005. Growth, Integration and Regional Inequality in Europe. Environment and Planning A 37, 18371855.
  • Pike, A., Rodriguez-Pose, A. and Tomaney, J. 2006. Local and Regional Development. Routledge: London, New York.
  • Pittau, M.G. and Zelli, R. 2006. Empirical Evidence of Income Dynamics Across EU Regions. Journal of Applied Econometrics 21, 605628.
  • Portnov, B.A. and Felsenstein, D. 2010. On the Suitability of Income Inequality Measures for Regional Analysis: Some Evidence from Simulation Analysis and Bootstrapping Tests. Socio-Economic Planning Sciences 44(4), 212219.
  • Puga, D. and Venables, A.J. 1999. Agglomeration and Economic Development: Import Substitution vs. Trade Liberalisation. Economic Journal, Royal Economic Society 109(455), 292311.
  • Quah, D.T. 1990. International Patterns of Growth: I. Persistence in Cross- Country Disparities. Working Paper. MIT.
  • Quah, D.T. 1993a. Galton's Fallacy and Tests of the Convergence Hypothesis. Scandinavian Journal of Economics 95, 427443.
  • Quah, D.T. 1993b. Empirical Cross-Section Dynamics in Economic Growth. European Economic Review 37, 426434.
  • Quah, D.T. 1996. Empirics for Economic Growth and Convergence. European Economic Review 40, 13531375.
  • Quah, D.T. 1997. Empirics for Growth and Distribution: Polarization, Stratification, and Convergence Clubs. Journal of Economic Growth 2, 2559.
  • Rodríguez-Pose, A. 1996. Growth and Institutional Change: The Influence of the Spanish Regionalisation Process on Economic Performance. Environment and planning c: government and policy 14(1), 7187.
  • Rodríguez-Pose, A. 1999. Innovation Prone and Innovation Averse Societies: Economic Performance in Europe. Growth and Change 30, 75105.
  • Rodríguez-Pose, A. 2013. Do Institutions Matter for Regional Development? Regional Studies 2013, (forthcoming) previously available as Working Papers 2010-02, InstitutoMadrileño de EstudiosAvanzados (IMDEA) Ciencias Sociales.
  • Rodríguez-Pose, A. and Ezcurra, R. 2010. Does Decentralization Matter for Regional Disparities? A Cross-Country Analysis. Journal of Economic Geography 10(5), 619644.
  • Rodríguez-Pose, A. and Gill, N. 2006. How Does Trade Affect Regional Disparities? World Development 34(7), 12011222.
  • Rodríguez-Pose, A. and Gill, N. 2003. Is There a Link Between Regional Disparities and Devolution? Environment and planning A 36(12), 20972117.
  • Rodrik, D. 2008. Second-Best Institutions. American Economic Review. American Economic Association 98(2), 100104.
  • Rodrik, D., Subramanian, F. and Trebbi, F. 2004. Institutions Rule: The Primacy of Institutions Over Geography and Integration in Economic Development. Journal of Economic Growth 9, 131165.
  • Romer, P.M. 1986. Increasing Returns and Long-Run Growth. Journal of Political Economy 94, 10021037.
  • Romer, P.M. 1990. Endogenous Technical Change. Journal of Political Economy 98, S71S102.
  • Roodman, D. 2009a. Practioners' Corner: A Note on the Theme of Too many Instruments. Oxford Bulletin of Economics and Statistics 71(1), 135158.
  • Roodman, D. 2009b. How to Do xtabond2: An Introduction to Difference and System GMM in Stata. Stata Journal 9(1), 86136.
  • Santos, L. and Barrios, E. 2011. Small Sample Estimation in Dynamic Panel Data Models. American Open Journal of Statistics 1(2), 5873.
  • Solow, R.M. 1956. A Contribution to the Theory of Economic Growth. Quarterly Journal of Economics 70, 6594.
  • Solow, R.M. 1994. Perspectives on Growth Theory. Journal of Economic Perspectives 8(1), 4554.
  • Soto, M. 2009. System GMM Estimation with Small Sample. Barcelona Economics Working Paper Series, WP. 395., str. 1-32.
  • Stasavage, D. 2000. Private Investment and Political Uncertainty. STICERD—Development Economics Papers 25, Suntory and Toyota International Centres for Economics and Related Disciplines, LSE.
  • Storper, Michael. 2005. Society, community, and economic development In Paula, S. de., and Dymski, G. A. (eds). Reimagining Growth: Towards a Renewal of Development Theory. Zed Books: London, UK, 198229.
  • Streeck, W. 1991. On the Institutional Conditions of Diversified Quality Production. In Matzner, E. and Wolfgang, S. (eds). Beyond Keynesianism: The Socio-Economics of Production and Employment. Edward Elgar: London, 2161.
  • Tabellini, G. 2005. Culture and Institutions: Economic Development in the Regions of Europe. Working Papers 292, IGIER (InnocenzoGasparini Institute for Economic Research), Bocconi University, June 2005.
  • Tanzi, V. 1996. Fiscal Federalism and Decentralisation: A Review of Some Efficiency and Macroeconomic Aspects, in Annual World Bank Conference on Development Economics 1995, World Bank, Washington D.C., 1996, 295–316.
  • Zak, P.J. and Knack, S. 2001. Trust and Growth. Economic Journal, Royal Economic Society 111(470), 295321.