Do European structural funds have an impact on economic growth in Central and Eastern Europe (2001 – 2013)

(1)

UNIVERSITEIT VAN AMSTERDAM

Do European Structural Funds Have an

Impact on Economic Growth in Central and

Eastern Europe (2001 – 2013)

Petar Chavdarov (10602976)

petar.chavdarov@student.uva.nl

Dr. Dirk Veestraeten (Supervisor)

Dr. Konstantinos Mavromatis (Second Reader)

18.08 2014

(2)

Abstract

The current paper analyzes the impact of the European Structural Funds on economic growth in Central and Eastern European member states in the period 2001-2013 by using panel-data regression with country and time fixed effects. The paper also investigates the individual effect of the three funds that comprise the total, namely the European Regional Development Fund, the European Social Fund and the Cohesion Fund. The results suggest that there is no significant relationship between the funds received by countries and economic growth. The paper also examines if funds have a medium-term effect on growth but the 2 up to 5 year lagged fund variables are also not significant. A further examination of the issue includes analyzing the effect of the funds conditional on several proxies of institutional quality such as economic freedom, corruption, government efficiency, regulatory quality and trade openness. Neither of the interaction terms between funds and institutional quality are found to significantly contribute to economic growth contrary to the findings of Ederveen, de Groot and Nahuis(2011). The results across the various models indicate that funds received did not boost economic growth which can be attributed to the fact that they may have crowded-out more economically beneficial projects. However, it can be argued that longer historical data should be employed when it becomes available and that the long-term impact of Structural Funds should be studied.

(4)

Section 1 – Introduction

European regional policy is the main investment mechanism of the European Union to bring convergence among member states and to boost growth and job creation. Regional policy aims at reducing the economic, social and territorial disparities by transferring funds from the EU budget to support specific projects. The rationale behind it is based on the solidarity principle of the Union as well as on the fact that vast economic disparities between regions would undermine the functioning of the single market and the common currency. The regional policy is implemented by national and regional bodies alongside the European Commission for a programme period of 7 years.

The effectiveness of the EU’s regional policy is of great importance to policy makers, beneficiaries and researchers as the funds allocated to regions represent a great share of the EU budget and involve a broad range of stakeholders. It is argued that funds should bring convergence among EU countries and should positively impact on economic growth and employment as they are channeled to specific projects which in most cases also yield economic gains. The analysis also takes into account whether corruption and poor governance may hamper the effective utilization of the funds. The current state of knowledge on the topic of how foreign funds affect economic growth can be traced back to papers analyzing the growth impact of aid received by developing countries (Burnside and Dollar 2000). The paper argues that aid had little effect unless conditional factors are present such as good policy environment and governance. The same conclusion is reached on the EU level analyzing the impact of structural funds received by member states by Ederveen, de Groot and Nahuis (2006; henceforth EGN) and Beugelsdijk and Eijffinger (2005). Structural funds are concluded to have positive effect on growth only for countries with strong institutions. Introducing conditional factors such as a corruption index, level of governance, openness of country etc. has proved to be an important milestone for the regression analysis on the topic and helped explain why structural funds in certain countries do not bring positive economic growth.

The precise research aim of the current paper is to estimate to what extent EU Structural Funds affect economic growth in Central and Eastern European countries. The paper will extend the literature by performing a regression analysis assessing funds’ impact on economic growth in Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia. The time period in focus comprise the last two programming periods, namely 2000-2006 and 2007-2013. This is a contribution to the literature since few papers have empirically tested the last programme period of 2007 – 2013. The choice of member states is determined by the fact that all Central and Eastern European countries’ regions are eligible for Convergence objective funds (except Praha region and Budapest region which receive funds under the Regional competitiveness and employment objective). Another strong argument in favour of the choice of countries is that they are all similar in terms of economic structure and development through the years before and after

(5)

joining the European Union. That motivates the choice of excluding older member states such as Portugal, Greece and Cyprus from the analysis even though they are also major recipients of EU funds. The paper will study the impact of structural funds on economic growth on a country level unlike many regionally disaggregated analyses performed by Becker et al. (2010, 2012a and 2012b) and Cappelen (2003). The main reason for focusing on a country level analysis is that many institutional and educational variables are not available on a regional level. Also, when a certain region receives funds there are potentially huge spill-over effects that attribute to growth of neighboring regions or to the country as a whole. Data availability of funds received on a regional level is also limited and proposes difficulties for the regression analysis.

The regression analysis will be based on the neoclassical growth model extended by Mankiw, Romer and Weil (1992; henceforth MRW) and additional variables are added such as the explanatory variable of interest – amount of Structural Funds received related to GDP. The value of the total funds comprises the sum of the European Regional and Development Fund, the European Social Fund and the Cohesion Fund. Additional variables are added such as trade openness and corruption. As suggested by Rapacki and Próchniak (2009) there are specific variables that affect economic growth in Central and Eastern European member states such as the flow of Foreign Direct Investment (FDI) as a share of GDP and economic freedom, so that they are also included in the regression analysis.

The paper has the following structure. Section II explains what are the main EU funds, how are they allocated and what purpose do they serve. It also provides an overview of the economic development of the Central and Eastern European countries after the move to democracy. Section III will present a literature review and outline the main findings on the topic and how the current paper addresses and extends the research performed so far. Section IV is devoted to the empirical research, presenting an overview of the data used, the model and the results of the regression analysis. Section V draws conclusions and gives recommendations for further analysis.

Section 2 – Structural Funds and Economic Development of CEECs

In order to effectively analyze whether the Structural Funds have effect on Central and Eastern European member states’ economic growth it is essential to outline what is the purpose of the different funds and what are the criteria the EU Commission applies in order to determine which countries are eligible for receiving the funds. Section 2 also includes an overview of the economic development of Central and Eastern Europe after the democratic changes since it is important to point out what drives economic growth in the region.

(6)

2.1 Overview of Structural Funds

The Structural and Investment Funds comprise five funds working together to bring economic and social development across the Union. It includes the European Regional Development Fund (ERDF), the European Social Fund (ESF), the Cohesion Fund (CF), the European Agricultural Fund for Rural Development (EAFRD) and the European Maritime and Fisheries Fund (EMFF). These funds need to address the three main objectives of the EU regional policy – the Convergence Objective (formerly Objective 1), the Competitiveness and Employment (formerly Objective 2) and the European Territorial Cooperation (formerly Objective 3). The Convergence objective is considered the most important and receives the greatest share of the funds. Its aim is to accelerate economic development in less developed regions, namely in those with a GDP per capita under 75% of the EU average. The funds through which the convergence objective is initiated are ERDF, ESF and CF. Тhe main aim of the European Regional Development Fund is to promote public and private investments in order to help reduce regional disparities and help convergence among the member states. The priorities of the fund include research, innovation, environmental protection and support for small and medium-sized enterprises. All those priorities help regions to achieve higher economic growth and to speed up the process of catching up. The European Social Fund aims at promoting employment and supporting labour mobility. Additionally, the ESF supports projects related to education and lifelong learning courses thus directly having an impact on human capital development.

(7)

Fig 1: Eligibility of Regions under the Convergence Objective 2007 - 2013

Source: Eurostat

Eligible under Convergence objective

Phasing out eligibility under Convergence objectives

Eligible under Regional competitiveness and employment objective

Phasing in eligibility under Regional competitiveness and employment objective.

The Cohesion Fund, on the other hand, is available to a broader group of countries, namely to those that have GNI per capita less than 90% of the EU average. For the period of 2007 – 2013 eligible for receiving transfers from the Cohesion Fund are Bulgaria, Croatia, Cyprus, the Czech Republic, Estonia, Greece, Hungary, Latvia, Lithuania, Malta, Poland, Portugal, Romania, Slovakia and Slovenia. The fund supports environmental and infrastructural projects thus improving the trans-European transport network, so these funds contribute greatly to economic development and convergence. The three funds ERDF, ESF and the

(8)

Cohesion Fund will be included in the analysis and their individual effect will be estimated as well. It is interesting to look at those separate effects since the funds are designed to serve different purposes and magnitude of their effect on GDP is expected to be different.

The amount of funds that was available for the programme period 2007 – 2013 was €347 billion and it was allocated to the different funds that have to serve the EU objectives for the period.

Table 1: Structural Funds available for the period 2007-2013

Available Budget (EUR bn.) 2007-2013

Per Fund European Fund for Regional Development 201

European Social Fund 76

Cohesion Fund 70

Total 347

Per Objective

Convergence 283

Competitiveness and Employment 55 European Territorial Cooperation 9

Total 347

Source: European Commission

2.2 Economic Development of Central and Eastern European Countries

Countries in Central and Eastern Europe underwent significant political, economic and social changes during the long process of transition from centrally planned economies to free market systems. For the purpose a number of economic reforms have been initiated alongside massive privatization of state owned enterprises in the years after 1989. Since then most of the CEE countries have started a gradual revival and began the process of economic convergence towards their western counterparts, especially after 1996. GDP per capita in the region represented only 21.6% of the EU average in 1996 and 17 years later it stood at 43.6%. The fast pace of convergence of CEECs can be attributed mainly to the growing integration of their economies to the single market of the European Union and increased volumes of trade and FDI. Most countries from the region became members of the EU in 2004 (Czech Republic, Poland, Hungary, Latvia, Lithuania, Estonia, Slovakia, Slovenia) and in 2007 Bulgaria and Romania joined. However, countries have been receiving EU Structural Funds since 2001 as part of the pre-accession procedure of the EU. The Structural Funds received have been increasing through the years due to the expanding EU budget especially in the sphere of regional development and convergence. Structural Funds received by CEE

(9)

countries in the period 2001-2013 averaged 1.14% of their GDP. Through the years the funds allocated have been steadily increasing and in 2013 averaged 2.3% of GDP.

The common strength of the CEE region is that the workforce is highly educated but still labour costs are low. Levels of tertiary education among the labour force are comparable to those in western economies, however hourly wages average 75 percent less than in the EU-15. Also, worth mentioning is the fact that countries from the region maintain a stable macroeconomic environment. The CEE economies can be characterized by a relatively low public debt level which is below 60% of GDP. Exchange rates have been relatively stable and some of the countries like Slovakia, Slovenia, Estonia and Latvia already adopted the euro thus further integrated their economies with their western counterparts.

Also, countries in the region are characterized by high labour productivity growth rates which alongside the above mentioned characteristics evidently led to increasing FDI, especially during the years prior to the world financial crisis. Western European financial institutions were among the most active players moving in the region. In the years 2004-2008 one fifth of the USD 220 billion of net FDI inflows in the region was attributed to the financial sector. Automakers also stepped into the region by either acquiring local companies or setting up productive facilities. Countries in the region also developed a strong outsourcing industry attracting a great deal of foreign capital due to the highly educated and low-cost labour force. In the years 2000-2013 the FDI inflow as a share of GDP averaged 5% in the countries in question only mildly decreasing when the financial crisis hit.

Prior to the years of becoming members of the EU the countries from the region adopted a number of reforms strengthening the level of governance and improving the quality of public institutions. According to the World Governance Indicators compiled by the World Bank, the Government Efficiency Index marked a twofold increase in 2013 compared to 2000. However, the biggest challenges in front of the Central and Eastern European member states are still the poor public governance and corruption that hampers economic development.

Section 3 – Structural Funds and Their Impact on Economic Growth

Section 3 outlines the findings of the empirical papers addressing the question of what drives economic growth. Furthermore, the impact of Structural Funds on output growth is reviewed. Several papers and their results are presented in order to give an overview of the current state of knowledge in the literature regarding the effectiveness of the Structural Funds.

3.1 Drivers of Economic Growth

There is a broad discussion in the empirical literature on what are the main determinants of economic growth. A famous paper by Barro (1996) performs a panel data regression of 100 countries aiming to outline which are the variables that have a significant impact on

(10)

economic growth. His findings suggest that growth is enhanced by higher initial human capital (level of schooling), life expectancy, lower fertility, lower government consumption, better rule of law and government efficiency, lower inflation and terms of trade. However, the sample of countries used is too broad ranging from underdeveloped to advanced economies located in different parts of the world. Therefore, the variables suggested should be considered as universal determinants of growth. However, economic growth in developing countries depends on a unique set of economic variables, so it is important to turn the analysis into an European perspective.

A paper by Próchniak (2011) presents an empirical analysis of the determinants of economic growth for 10 Central and Eastern European member states for the period 1993 - 2009. The study suggests that the most important determinants are the investment rate alongside FDI, human capital accumulation, financial sector development, balanced public finances, high share of services in GDP, big private sector and favourable institutional environment (economic freedom and quality of governance). Low interest rates, low inflation and development in the ICT sector also positively affect economic growth in the countries in question. The findings are in line with a similar paper by Rapacki and Próchniak (2009) that takes into consideration also the Structural Funds received by the Central and Eastern European member states.

There is also a vast literature on the topic of how foreign funds affect economic growth in recipient countries. The oldest contributions on the topic can be found in papers analyzing foreign aid to developing countries. A prominent paper that deals with the issue is the Burnside and Dollar (2000). In their paper they use panel data regression analysis on 56 developing countries. Their finding is that foreign aid has a positive impact on growth only in countries with stable fiscal balance, sound monetary policy and high degree of trade openness. So their main conclusion is that aid is only effective in the presence of good institutions.

3.2 Impact of Structural Funds on Economic Growth

There are several theoretical considerations for the channels through which Structural Funds should affect economic growth. The basic hypothesis is that the funds should have a positive effect on national output since they increase the rate of investment in a country. Funds can be considered as foreign aid received for specific projects that in most cases aim to generate economic profits. Therefore, financing economic projects should increase employment and output growth. The Structural Funds are also expected to boost the aggregate demand in the recipient countries. It is also assumed that the increased expenditures towards economically viable projects will improve the human capital and therefore will boost productivity levels. It must be taken into consideration that ESF finances projects related to promoting employment, improving education and offering lifelong learning opportunities for the elderly thus having a direct impact on the development of human capital.

(11)

However, it should be noted that not all projects that qualify for European funding are growth promoting, for example financing cultural or environmental projects is also common in the EU agenda. Therefore, these initiatives might not have effect on the GDP growth. Moreover, since the projects eligible for EU financing require factors such as human and physical capital that otherwise could be employed into other potentially more growth enhancing activities there may be an adverse impact on growth or no effect at all.

Additionally, the projects receiving Funds must be co-funded by domestic tax revenues, so that it can be argued that an increase in taxation in order to finance projects may have a highly distortionary effect and lead to negative growth. Having no effect at all is also a possible scenario since corruption and poor governance may take away Structural Funds from productive activities. The basic idea is that the effectiveness of the funds depend on the institutional quality in the countries and more specifically on the ability of the public authorities to transparently and adequately utilize the funds. If for example, the level of corruption in the public sector is high the funds will be most likely transferred to rent-seeking activities that may have a negative impact on economic growth. Poor government efficiency and regulatory quality also undermine the effective utilization of Structural Funds. Following these assumptions, the current paper will assess the effectiveness of the funds but taking into consideration several proxies of institutional quality used as conditioning variables.

There are three main approaches on how to assess the impact of the Structural Funds on economic growth. In the current literature the question is addressed either by model simulation, case studies or econometric evaluation. The literature review that follows covers papers focusing on the latter approach.

Econometric evaluation focuses on analyzing either regional or country specific data. On the regional level a great contribution to the literature is the paper of Becker, Egger and von Ehrlich (2010) focusing on how Convergence funds (formerly Objective 1) affect economic growth in recipient regions. The results of the paper suggest that there is a positive effect of the Convergence funds on economic growth and the authors conclude that the funds received raise real GDP per capita growth by 1.6 percentage points within the same programming period.

Country level analysis on the other hand, offers the opportunity to include various institutional variables such as quality of governance, economic and structural reforms, corruption index etc. which are not available on a regional level. A paper that investigates the effect of the Structural Funds interacted with a set of institutional quality variables is EGN (2006). The authors perform panel data analysis based on the neoclassical growth model but extended to include Structural Funds received and institutional variables such as corruption, inflation and trade openness. The results suggest that Structural Funds solely included do not contribute to higher growth. However, the interaction term between the level of institutional quality and the Structural Funds is found to be positively related to economic growth. The conclusion of the paper is that Structural Funds promote economic growth only when the level of corruption is low and trade openness is high . Similarly, Beugelsdijk and Eijffinger (2005) focus on country level data and perform GMM estimations finding a positive

(12)

relationship between funds received and economic growth. Their paper, however, does not suggest that higher corruption leads to less efficient use of the Structural Funds received. A major critique on the paper of EGN (2006) is Bradley and Untiedt (2008). They argue that the results presented by the former are biased and unstable thus questioning the validity of the conclusions and policy recommendations made. There are several points made why the results may not be credible. First, it is argued that the data sample from 1960 to 1995 used in EGN (2006) is not accurate due to the fact that in 1989 a major reform of the Structural Funds was initiated, increasing considerably the amounts transferred and changing the purpose the funds need to serve. Secondly, Bradley and Unitiedt (2008) suggest that including only the interaction term of the level of institutional quality and Structural Funds may not be sufficient to explain the data. Including the level of institutional quality as a separate variable has a significant positive impact, although it does not change the coefficient results.

The above papers focus on all countries that received Structural Funds in the periods under consideration. Thus, it is required to further narrow down the question and focus on the impact of Structural Funds in Central and Eastern European countries. An interesting paper on the topic is Rapacki and Próchniak (2009) which studies the impact of joining the European Union on economic growth in CEECs. Although, the focus of their work is quite broad compared to the current analysis, their paper provides useful insights on which variables proved to be important determinants of economic growth in the countries in question. Several variables are outlined such as Foreign Direct Investment, economic freedom, structural reforms and Structural Funds received (in their analysis denoted by aid). All four variables were found to be positive and statistically significant. The impact of aid per capita on economic growth have values between 0.0652 and 0.1032 depending on the econometric approach used.

The current paper will extend the research performed by EGN (2006) and take into account the critiques proposed by Bradley and Unitiedt (2008). Therefore, a neoclassical growth model extended with the inclusion of human capital will be used. Following the work of Rapacki and Próchniak (2009) additional explanatory variables will be added such as Structural Funds related to GDP, FDI/GDP, quality of institutions, economic freedom and corruption. The time frame in focus is 2001-2013, covering the last two programme periods. Last two programme periods share the same characteristics in terms of aims and eligibility criteria thus can be analyzed together. Additionally, taking into consideration one specific region, namely Central and Eastern Europe, the current paper will not face the problem of huge outliers in terms of institutional quality and other variables of interest. This also allows for the addition of independent variables specific for the countries in question. It is important to be added that the paper of EGN (2006) analysed only ERDF funds, while in the current analysis the European Social Fund and the Cohesion Fund will be included as well because they may also account for economic growth. The impact of the total amount of funds will be estimated as well as their separate effect on GDP growth.

(13)

Section 4 - Methodology and Estimation

The current paper investigates the relationship between Structural Funds received by Central and Eastern European member states and the GDP growth rate in a sample of 10 countries, namely Bulgaria, the Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Romania, Slovenia and Slovakia. The estimation is based on annual data for the period 2001-2012. The analysis also presents the separate effect the different types of Structural Funds have on per-capita GDP growth. These are ERDF, ESF and the Cohesion Fund which are found to be highly correlated with each other (Table C in the Appendix). This can be explained by the fact that even though they serve different purposes the amounts transferred depend on the ability of the countries eligible for EU financing to successfully utilize the funds. If they are accused of misallocation of the funds this leads in most cases to suspending the access to all the three funds. Moreover, the allocated funds depend also on the budget of the EU Commission. Therefore, since the total budget of the EU has been mostly increasing, the amounts distributed via the three funds also grew.

Data on the amounts of Structural Funds is obtained from the European Commission website, where a detailed breakdown of funds received by countries is publicly available. It is important to be noted that the amounts of funds taken into consideration are those that have been actually paid to the member states in contrast to funds being planned to be received in the beginning of every programme period. There is a discrepancy between the money spent and the initially programmed funds to be allocated to member states resulting from the limited capacity of countries to successfully and transparently utilize the funds. The data source for the other dependent variables is mostly Eurostat and World Bank. A description of the data sources is presented in Table A in the Appendix.

The empirical growth model employed is based on the Solow model as extended by Mankiw, Romer and Weil (1992) to include human capital. The Solow model addresses economic growth by assuming a standard neoclassical production function with decreasing returns to capital and labour. The model implies that GDP per capita growth is positively related to either investments or savings as a share of GDP. On the other hand it assumes that the higher the population growth and initial level of GDP per capita the lower the income growth rate is. The rate of investments and population growth are assumed to be exogenous in the model and these two variables determine the steady-state level of income per capita. Because investment and population growth rates vary across countries, different countries reach different steady states. The Solow model is further extended by Mankiw, Romer and Weil (1992) to include human capital as a determinant of economic growth. They argue that the increasing level of education boosts the productivity of workers, which translates into higher GDP growth. The MRW paper also suggest that human-capital accumulation is correlated with investment and population growth rates which implies that omitting it will lead to biases in the estimated coefficients. Additionally, the growth rate of technological progress and depreciation rate are included in the model and are assumed to be constant across countries. Technological progress is considered as the general worldwide advancement of knowledge which is not country-specific. Also it is assumed that depreciation rates hardly vary across countries. As

(14)

suggested by MRW (1992) the sum of technological progress and depreciation growth rate is equal to 5% for all countries and time-periods.

This paper modifies the above model by including the Structural Funds received as a share of GDP. The panel data regression equation has the following form, where 𝑔_𝑎 and 𝛿 represent the exogenous rate of technological progress and the rate of depreciation, both adding to 0.05. Subscripts 𝑖 and 𝑡 account for country and time, while µ_𝑖 and 𝜈_𝑡 represent country and time fixed effects.

𝐺𝑑𝑝𝑔𝑟𝑜𝑤𝑡ℎ𝑖,𝑡 = 𝛽0+ 𝛽1𝐺𝑑𝑝𝑔𝑟𝑜𝑤𝑡ℎ𝑖,𝑡−1+ 𝛽2𝐿𝑛𝑔𝑑𝑝𝑖,𝑡−1+ 𝛽3𝐼𝑛𝑣𝑔𝑑𝑝𝑖,𝑡−1+ 𝛽4𝐻𝑢𝑚𝑐𝑎𝑝𝑖𝑡𝑎𝑙𝑖,𝑡−1+ 𝛽5�𝑃𝑜𝑝𝑔𝑟𝑜𝑤𝑡ℎ𝑖,𝑡−1+𝑔𝑎+ 𝛿� + 𝛽6𝑆𝐹𝑔𝑑𝑝𝑖,𝑡−1+ 𝛽7𝐶𝑜𝑟𝑟𝑢𝑝𝑡𝑖𝑜𝑛𝑖,𝑡−1+ 𝛽8𝐹𝐷𝐼𝑔𝑑𝑝𝑖,𝑡−1+

𝛽9𝑇𝑟𝑎𝑑𝑒𝑜𝑝𝑒𝑛𝑛𝑒𝑠𝑠𝑖,𝑡−1+ 𝛽10𝐸𝑐𝑓𝑟𝑒𝑒𝑑𝑜𝑚𝑖,𝑡−1+ µ𝑖+ 𝜈𝑡+ 𝜀𝑖,𝑡

GDP growth is calculated as the annual percentage change of real GDP per capita in constant 2005 prices added in decimal form while Lngdp represents the natural logarithm of the initial level of per capita real GDP. Invgdp is the ratio of gross fixed capital formation to GDP in decimal terms. As a proxy of human capital the share of total population with upper secondary or tertiary education attainment is used. There is a broad discussion on the topic of which might be the best human capital proxy and as suggested by MRW (1992) school enrollment rates are found to be significant. The motivation behind the choice of proxy is also because the data is available in Eurostat which makes it consistent and comparable across the countries in the sample. Population growth is the annual change of population in the respective countries. In the regression equation the exogenous rate of technological progress and the rate of depreciation both adding to 0.05 is added to the decimal population growth rate defined as Popgrowth005. These variables follow the neoclassical growth model described above and are widely used in the empirical growth literature.

The model is extended to include specific variables that have an impact on economic growth in the countries in question. Since the CEECs have been a favourable investment destination by western companies and attracted a great deal of foreign investments, FDI as a percentage of GDP is included. Additionally, a trade openness variable is added, calculated as the sum of export and import as a share of GDP. Both FDI to GDP ratio and trade openness expand the model to account for an open economy. Economic Freedom and Corruption are also added since it is argued that they also have an impact on economic growth, especially in the new EU member states as suggested by Rapacki and Próchniak (2009) and Próchniak (2011). There are a number of channels through which high corruption negatively affects growth as discussed by Tanzi (1997) and Gupta (2000) among many others. It is argued that high level of corruption distorts incentives thus leading to misallocation of resources and inefficiency of public and private expenditures which directly relate to the effective utilization of Structural Funds. Corruption also acts as a tax on business so that production costs rise. The variable is obtained from the World Bank’s World Governance Indicators database. It reflects the perceptions of the extent to which public power is exercised for private gain. It ranges from -2.5(high corruption) to 2.5 (low corruption). The Economic Freedom Index is compiled by the Heritage Foundation and combines four broad quantitative indices such as rule of law, limited government, regulatory efficiency and open markets. These four categories are graded

(15)

on a scale of 0 (weak) to 100 (strong) and the overall country’s score is derived by their average with equal weights being given to each. It is argued that economic freedom increases prosperity and allows for better allocation of resources which makes it a useful variable when studying economic growth. Moreover, Central and Eastern European countries’ economic development depends largely on a favourable institutional environment and as suggested by Próchniak (2011) economic freedom is found to have a significant effect on GDP growth in the region.

A panel-data regression analysis is employed including country/entity (µi) and fixed time (νt)

effects. In this way the model takes into account specific-country effects which are constant over time and time-varying effects which are common for all countries in the sample, therefore, the negative effects of the financial and the European sovereign debt crises on economic growth are captured in the model. A modified Wald statistic is calculated in order to test for group wise heteroskedasticity in the residuals of a fixed effect regression model, following Greene (2000, p. 598). This suggests that there exists heteroskedasticity and therefore country clustered standard errors are used. Also, in the specification of the model all explanatory variables are included with a lag of one year in order to take into consideration the time it takes of the variables to affect economic growth. One year lag of GDP per capita growth is also included as an explanatory variable in order to account for the sluggishness and path dependency and also mitigates autocorrelation problems.

A Hausman test is performed to investigate whether random or country and time fixed effects model better fits the data. It shows whether the unique errors are correlated with the regressors, while the null hypothesis suggests that they are not. The null hypothesis is not rejected thus prompting that both fixed and random effects models can be used compared to the alternative that strongly suggests that the random effects model is inappropriate. The analysis that follows employs the fixed effects model because it is assumed that GDP growth depends also on country and time specific characteristics. Country fixed effects that are time-invariant and may affect growth area the political system, culture, risk-perception, geographic location and abundance of natural resources. By using fixed effects the model removes the effect of those time-invariant characteristics that may bias the parameters. It controls for all differences between the countries that do not change over time so the estimated coefficients of the fixed effects model cannot be biased because of omitted time-invariant characteristics. The model also includes time fixed effects in order to control for unexpected events in time that may affect the outcome variable such as the financial crisis and the European sovereign debt crisis.

Table 2 presents the results of the panel-data regression accounting separately for entity, time and both fixed effects in order to find the best model. Inclusion of both country and time fixed effects gives R-squared value of 74% which is the highest among all three models. Also, a joint test is performed estimating that the year dummies are significantly different than 0, therefore should be included in the regression analysis (Table B in the Appendix). The paper will focus on the model including both entity and time-fixed effects.

(16)

Table 2: Fixed Effects Regressions

(Panel data regression with separate entity and time fixed effects using country clustered standard errors, dependent variable: growth of GDP per capita, All explanatory variables are included with a lag of one year, variable of

interest: Structural Funds)

--- Entity Effects Time Effects Both --- Gdpgrowth 0.447*** 0.347*** 0.359** (0.136) (0.0954) (0.122) Lngdp -0.253*** -0.0458*** -0.257** (0.0749) (0.0155) (0.105) Invgdp -0.274 -0.0843 0.0553 (0.183) (0.0924) (0.216) Humcapital 1.069*** 0.223** 0.620** (0.289) (0.0902) (0.241) Popgrowth005 0.444 -0.372 -0.0237 (0.809) (0.511) (0.635) SFgdp 0.35 -0.0803 0.439 (0.827) (0.548) (0.772) Corruption 0.137*** 0.0255* 0.0692** (0.0278) (0.0153) (0.0257) Tradeopenness -0.0874** 0.00383 0.0171 (0.0318) (0.0117) (0.0198) FDIgdp 0.250** -0.0342 0.0217 (0.0824) (0.0788) (0.086) Ecfreedom 0.000617 -0.000654 0.000659 (0.00115) (0.000758) (0.00109) _cons 1.491** 0.326*** 1.688* (0.482) (0.116) (0.792) --- N 120 120 120 adj.R-sq 0.424 0.715 0.756 ---

Standard errors in parentheses * p<0.10, ** p<0.05, *** p<0.01

(17)

The three models show that the initial level of GDP is negatively related to economic growth thus suggesting that countries with lower initial level of GDP per capita tend to have higher economic growth meaning that they are converging to their Western counterparts. The variable is found statistically significant at the 5% level. GDP growth in the previous year also significantly affects the current year growth rate suggesting for the path dependency of the dependent variable. Also as suggested in the Solow growth model an increase in population has a negative impact on per capita GDP growth. Human capital has a significant positive effect on GDP growth. This can be explained by the fact that the increase in the share of people with upper secondary and tertiary education boosts productivity levels and hence economic growth. The positive impact of human capital on economic growth is in line with the findings of MRW (1992). Investments as a share of GDP also have a positive effect on output growth, although not significantly. As expected higher FDI share of GDP has a positive sign in the model including only country-fixed effects and the one that comprise both country and time fixed effects. The reasoning behind the results is that foreign direct investments are mostly engaged in productive activities that increase economic growth. The trade openness variable is also found positively related to GDP growth because it allows for absorption of technological advances generated in developed countries as suggested by Barro and Sala-i-Martin (1995). Other main reason is that higher trade openness increases the foreign competition a country faces, therefore, domestic producers need to become more efficient and to boost productivity levels which in turn translates into higher economic growth. The corruption variable is found significant at the 5% level, meaning that less corruption leads to higher economic growth.

The effect of the Structural Funds on GDP growth appears to be insignificant across all three models. Country and time fixed effects regression estimating the separate impact of the ERDF, ESF and CF on GDP growth also shows that the impact is insignificant (Table 3). This can be attributed to various factors including the institutional capacity of the countries to effectively utilize the funds, the time it takes for the funds to start reap economic benefits, the short time span of the regression analysis, etc.

(18)

Table 3: Effect of the Different Structural Funds

interest: ERDF, ESF and CF)

--- ERDF ESF CF --- Gdpgrowth 0.382** 0.340** 0.331** (0.126) (0.123) (0.121) Lngdp -0.259** -0.250** -0.241** (0.104) (0.105) (0.103) Invgdp 0.0753 0.0477 0.0397 (0.202) (0.223) (0.203) Humcapital 0.632** 0.612** 0.621** (0.242) (0.247) (0.242) Popgrowth005 -0.0462 -0.0219 -0.0798 (0.685) (0.62) (0.633) Corruption 0.0700** 0.0706** 0.0752** (0.0236) (0.0261) (0.0251) Tradeopenness 0.0187 0.0207 0.0273 (0.0209) (0.0211) (0.0209) Fdigdp 0.0108 0.0271 0.0317 (0.0821) (0.0866) (0.0763) Ecfreedom 0.00079 0.000432 0.000122 (0.00111) (0.00107) (0.000979) ERDFgdp 1.234 (1.244) ESFgdp 0.816 (2.621) CFgdp -0.991 (1.535) _cons 1.688* 1.648* 1.579* (0.786) (0.787) (0.775) --- N 120 120 120 adj.R-sq 0.759 0.755 0.756 ---

(19)

4.1 Lagged Effect

It can be assumed that Structural Funds need several years to affect GDP growth since they are mostly engaged in initiatives that take time to be developed and to bring returns. Therefore, more than one year is required for the effect on GDP to take place. A regression analysis is run with Structural Funds lagged from 2 up to 5 years. The results are presented in Table 4. All of the included lags do not yield significant effect on GDP growth. Tables from D.1 to D.3 in the Appendix show the results of the different funds namely ERDF, ESF and CF but included with a lag. The effect on GDP growth is also found to be insignificant.

Table 4: Effect of Lagged Structural Funds

(Panel data regression with separate entity and time fixed effects using country clustered standard errors, dependent variable: growth of GDP per capita, All explanatory variables are included with a lag of one year except the

variable of interest: Lagged SFgdp)

---

2-Years 3-Years 4-Years 5-Years

--- Gdpgrowth 0.348** 0.368** 0.342** 0.379*** (0.127) (0.135) (0.119) (0.109) Lngdp -0.300* -0.355** -0.442** -0.513** (0.135) (0.156) (0.174) (0.181) Invgdp 0.0976 0.0994 0.215 0.287 (0.278) (0.324) (0.299) (0.291) Humcapital 0.854** 0.913** 1.356*** 1.376*** (0.288) (0.324) (0.346) (0.372) Popgrowth005 0.0824 -0.955 -0.465 0.315 (0.699) (0.974) (0.875) (0.901) Lagged SFgdp 0.662 -0.473 -0.979 1.631 (0.617) (1.044) (1.059) (1.913) Corruption 0.0675** 0.0828** 0.103** 0.0755 (0.0277) (0.0339) (0.0377) (0.0473) Tradeopenness 0.00816 0.0161 0.0271 0.00119 (0.0343) (0.05) (0.0444) (0.0383) Fdigdp 0.0196 -0.0245 0.0197 0.0479 (0.0908) (0.0946) (0.0934) (0.0949) Ecfreedom 0.000434 -0.00113 -0.00178 -0.00238* (0.00147) (0.00194) (0.00144) (0.00121) _cons 1.908 2.496* 2.903* 3.530** (1.122) (1.246) (1.384) (1.4) --- N 110 100 90 80 adj. R-sq 0.769 0.781 0.802 0.819 ---

(20)

4.3 Effect of the Structural Funds Conditional on Institutional Quality

A broad discussion on the topic focuses on the importance of the institutional quality for the effective utilization of the Structural Funds. It is argued that the higher the quality of governance and institutional capacity the greater the funds’ impact will be. Therefore, in the empirical literature on the topic the conditional effectiveness of the funds is studied by including an interaction term between the Structural Funds and a set of variables on institutional quality. The current paper uses several variables provided by the World Bank and the results are summarized in Table 5. The first variable used as a proxy for institutional quality and the general business environment is the index of economic freedom compiled by the Heritage Foundation. It is expected that higher economic freedom will result in a better allocation of resources, hence the Structural Funds will be more effectively utilized. Other proxies used are Government Efficiency, Corruption, Regulatory Quality and Trade Openness. Description of the variables can be found in Table A in the Appendix.

The interaction term between SF and the conditioning variable (COND) is included in the analysis alongside the separate institutional quality variable. As suggested by the paper of Bradley and Untiedt (2008) these variables by itself have an explanatory power for the growth of GDP. Therefore, inclusion of the separate institutional quality variables in the regression equation appears warranted.

Another proxy for the institutional quality is the degree of government efficiency in the countries in question. It reflects the quality of the public services and the degree of independence from political pressures which have a direct impact on the effective utilization of Structural Funds since they are distributed by the different governmental agencies and local governmental bodies. Regulatory quality addresses the perception of the ability of government to implement sound policies and to promote private sector development while Corruption reflects the extent to which public power is exercised for private gain. The three variables have a direct impact on how the Structural Funds are distributed since the approval of projects is done by the public authorities. Another proxy for institutional quality is the trade openness of a country as it represents the foreign competition the country faces and captures the pressure to efficiently use Structural Funds in order to become more competitive. The coefficients on Structural Funds interacted with the conditional variable remain statistically insignificant across every institutional quality variable used. This prompts that even when we account for the level of governance the effect remain insignificant and cannot be concluded that Structural Funds have a positive impact on growth.

The results from Table 5 strongly suggest that the effect of the Structural Funds remains insignificant across the use of various institutional quality proxies. The same holds true when we estimate the separate effect of the ERDF, the ESF and the Cohesion Fund.

(21)

Table 5: Conditional Effectiveness of Structural Funds

interest: SFgdp_COND)

Ec.freedom Government Eff. Regulatory quality Corruption Openness

Gdpgrowth 0.371** 0.344** 0.358** 0.392*** 0.359** (0.121) (0.119) (0.119) (0.114) (0.127) Lngdp -0.261** -0.277** -0.256** -0.298** -0.257** (0.102) (0.114) (0.11) (0.11) (0.108) Invgdp 0.0644 0.104 0.032 0.0538 0.0552 (0.213) (0.234) (0.212) (0.22) (0.217) Humcapital 0.759*** 0.681* 0.588* 0.431 0.616* (0.231) (0.303) (0.286) (0.255) (0.281) Popgrowth005 0.0106 0.046 -0.0391 0.0588 -0.0243 (0.636) (0.58) (0.635) (0.609) (0.642) SFgdp -3.954 0.474 0.806 1.104 0.482 (2.562) (1.535) (1.605) (0.954) (1.491) SFgdp_COND 0.0616 0.05 -0.278 -1.417 -0.0285 (0.0352) (1.5) (1.44) (0.881) (0.784) Corruption 0.0705** 0.0509 0.0625* 0.0909** 0.0690** (0.0269) (0.0311) (0.0341) (0.0289) (0.0233) Tradeopenness 0.0171 0.0267 0.0229 0.0127 0.0177 (0.0184) (0.0253) (0.0249) (0.0223) (0.0179) Fdigdp 0.0221 0.026 0.0206 0.0411 0.0212 (0.0817) (0.0899) (0.0883) (0.0857) (0.0859) Ecfreedom 0.00054 0.000912 0.000482 0.000386 0.000664 (0.00113) (0.00124) (0.00126) (0.00112) (0.0011) Govefficiency 0.0412 (0.0442) Regquality 0.0183 (0.0349) N 120 120 120 120 120 adj. R-sq 0.755 0.755 0.752 0.759 0.754

(22)

4.3 Instrumental Variables

In the Solow growth model further extended by MRW (1992) the independent variables are considered exogenous in the model. However, the current paper extends it to include Structural Funds in order to estimate the effect they have on the GDP growth rate in Central and Eastern European member states. There is a strong evidence to believe that Structural Funds are endogenous in the model thus correlated with the residuals. Simultaneous causality bias may exist in the model because high GDP growth may provide incentives for utilization of larger amounts of Structural Funds. GDP growth is characterized by increasing demand and higher prices which stimulate new business ventures, therefore more projects requiring EU funding are submitted. In order to control for the simultaneous causality bias instrumental variables are introduced in the model. The Structural Funds are instrumented by their 4-year lagged value and the 4-year lag of government efficiency index. Both variables positively explain the movements of the Structural Funds and are believed to be uncorrelated with the GDP growth rate in the current period. Results of the 2-SLS regression are provided in Table 5 alongside the tests employed in the analysis. The coefficients of the different funds are reported in the first row.

The Hansen J statistics tests whether the instruments are overidentified. The joint null hypothesis is that the instrumental variables are valid instruments, being uncorrelated with the error term, and that the excluded instruments are correctly excluded from the estimated equation. In the model the null hypothesis is far from being rejected thus providing greater confidence that the instrument set is appropriate. An endogeneity test of the Structural Funds variable is performed. Under the null hypothesis the variable under consideration is treated as exogenous as opposed to the alternative hypothesis that suggests that even after the variable has been instrumented it still continues to be endogenous. The test fails to reject the null hypothesis, therefore the Structural Funds can be treated as exogenous variables.

The tests performed clearly suggest that the instrumental variables introduced satisfy the conditions of relevance and exogeneity thus the results provided can be considered as reliable. The effect of the total Structural Funds variable is still insignificant which is consistent with the previous results. The same 2-SLS regression is performed for the other three funds and neither of the coefficients is found to be significant as well.

(23)

Table 6: Instrumental Variables

(2-SLS regression with entity and time fixed effects robust standard errors, dependent variable: growth of GDP per capita All explanatory variables are included with a lag of one year, variable of interest: Instrumented Fund variable - Fundsgdp,

IVs: 4-Year Lag of Funds and 4-Year Lag of Government Efficiency)

SF ERDF ESF CF Fundsgdp 2.786 1.576 52.13 4.694 (2.351) (2.317) (90.04) (8.221) Gdpgrowth 0.419*** 0.407*** -0.0559 0.372*** (0.118) (0.106) (0.796) (0.114) Lngdp -0.455*** -0.490*** -0.244 -0.485*** (0.123) (0.114) (0.479) (0.124) Invgdp 0.302 0.275 1.027 0.214 (0.209) (0.188) (1.416) (0.225) Humcapital 1.192*** 1.297*** 0.96 1.238*** (0.4) (0.307) (0.906) (0.445) Popgrowth005 -0.234 -0.0684 1.504 0.142 (1.032) (0.891) (2.821) (0.856) Corruption 0.0469 0.0778** -0.0762 0.0555 (0.0495) (0.0355) (0.295) (0.055) Tradeopenness 0.0192 0.0186 0.0946 -0.0108 (0.0531) (0.0463) (0.159) (0.0584) Fdigdp -0.0209 0.0203 -0.141 0.0765 (0.13) (0.105) (0.389) (0.0928) Ecfreedom -0.000294 -0.00221 0.00229 -0.00171 (0.00322) (0.00209) (0.0111) (0.00337) Hansen J 0.191 0.0950* 0.6907 0.0598* (P-val) Endogeneity Test 0.1283 0.5168 0.0727* 0.6709 (P-val)

(24)

Section 5 - Conclusion

Do Structural Funds promote economic growth in the Central and Eastern European member states? Using country and time-fixed effects panel data model the paper does not find a significant relationship between the total Structural Funds received and GDP growth in the period 2001-2013. The separate effects of the three funds that comprise the Structural Funds, namely the European Regional Development Fund, the European Social Fund and Cohesion Fund are also not statistically significant. Taking into account that funds may not immediately affect output growth, the paper analyzes the medium-term effect. For the purpose lagged total Structural Funds and separate funds from 2 up to 5 years are included in the analysis but still not found significant, thus suggesting that time-horizon for the effect to take place might be longer. The paper also analyzes the effect of the funds conditional on several institutional quality proxies such as economic freedom, corruption, government efficiency, regulatory quality and trade openness. The idea behind is to evaluate whether the presence of high quality institutional environment interacted with Structural Funds might positively affect GDP growth. The coefficients of the funds conditional on all of the institutional quality variables are found not statistically significant as well.

It can be concluded that Structural Funds have no effect on GDP growth in CEEC for the period 2001-2013. There are several reasons behind the results which should be outlined and taken into account in future research on the topic. Firstly, countries in question started receiving EU support only in 2001, namely before becoming EU members. Therefore, data availability is limited and in future research on the topic a longer time span should be adopted. Additionally, Structural Funds may contribute to the long-term economic growth which may also be taken into consideration in future research. A third argument is that Structural Funds indeed may not have boosted economic growth in the CEEC and instead of engaging in economically beneficial projects funds may have just crowded out more productive activities. Also, it should be pointed out that some EU financed projects do not yield economic benefits because they are related to more social and environmental issues.

(25)

Section 6 - Bibliography

Barro, R. (1996). Determinants of Economic Growth: A Cross-Country Empirical Study,

NBER Working Paper 5698

Barro, R. And Sala-i-Martin, X. (1995), Economic Growth, New York: McGraw-Hill Becker, S. O., Egger, P. and von Ehrlich, M. (2010), ‘Going NUTS: The Effect of EU Structural Funds on Regional Performance’, Journal of Public Economics 94 (9–10): 578– 90.–68.

Becker, S. O., Egger, P. and von Ehrlich, M. (2012), ‘Too Much of a Good Thing? On the Growth Effects of the EU’s Regional Policy’, European Economic Review 56 (4): 648–68. Beugelsdijk, M. and S.C.W. Eijffinger (2005). The Effectiveness of Structural Policy in the European Union: An Empirical Analysis for the EU-15 during the Period 1995–2001,

Journal of Common Market Studies 43: 37–51.

Bradley, J., Untiedt, G., (2008), EU cohesion policy and “conditional” effectiveness: What do cross-section regressions tell us?, GEFRA Working Paper: May 2008 – Nr. 4

Burnside, C. and D. Dollar (2000). Aid, Policies and Growth, American Economic Review 90: 847–68.

Ederveen, S., de Groot, H. and Nahuis, R. (2006), “Fertile Soil for Structural Funds? A Panel Data Analysis of the Conditional Effectiveness of European Cohesion Policy”, Kyklos 59: 17-42

Greene, W.H. (2000). Econometric Analysis . New York: MacMillan., page 598

Gupta, S., de Mello, L., Sharan, R., (2000), Corruption and Military Spending, IMF Working

Paper 00/23

Islam, N. (1995). Growth Empirics: A Panel Data Approach, Quarterly Journal of

Economics. 110: 1127–1170.

Mankiw, N.G., Romer, D. and Weil, D.N. (1992), ‘A Contribution to the Empirics of Economic Growth’, Quarterly Journal of Economics. 107: 407–437.

Próchniak, M. (2011), Determinants of Еconomic Growth in Central and Eastern Europe: the Global Crisis Perspective, Post-Communist Economies, 23:4, 449-468, DOI

Rapacki, R. and Próchniak, M. (2009), The EU Enlargement and Economic Growth In the CEE New Member Countries, Economic Papers 367, European Commission

Tanzi, V. and Davoodi, H., (1997), Corruption, Public Investment and Growth, IMF Working

(26)

Appendix

Table A: Data Description and Data Sources

Variable abbrev. Variable name/description Source

Gdpgrowth Annual growth rate of real GDP per capita in 2005 prices Eurostat Lngdp Natural logarithm of real GDP per capita in 2005 prices Eurostat Invgdp Gross Fixed Capital formation as a percentage of GDP Eurostat Humcapital Share of total population with upper secondary or tertiary

education attainment

Eurostat

Popgrowth005 Annual population growth rate + 0.05 Eurostat

Tradeopenness Export + Import as a percentage of GDP Eurostat

Fdigdp Foreign Direct Investments (Flow) as a percentage of GDP Eurostat

Ecfreedom Index of Economic Freedom Heritage

Foundation SFgdp Structural Funds received as a percentage of Nominal GDP European

Commission ERDFgdp European Regional Development Funds received as percentage of

GDP

European Commission ESFgdp European Social Funds received as a percentage of GDP European

Commission

CFgdp Cohesion Funds received as a percentage of GDP European

Commission

Govefficiency Government Effectiveness Index World Bank

Corruption Corruption Index World Bank

Regquality Regulatory Quality Index World Bank

Table B: Test of the Significance of Time Fixed Effects

testparm YEAR* ( 1) YEAR2 = 0 ( 2) YEAR3 = 0 ( 3) YEAR4 = 0 ( 4) YEAR5 = 0 ( 5) YEAR6 = 0 ( 6) YEAR7 = 0 ( 7) YEAR8 = 0 ( 8) YEAR9 = 0 ( 9) YEAR10 = 0 (10) YEAR11 = 0 (11) YEAR12 = 0 Constraint 2 dropped Constraint 10 dropped F( 9, 9) = 19.08 Prob > F = 0.0001

(27)

Table C: Correlation Table of the Different Funds | ERDFgdp CFgdp ESFgdp ---+--- ERDFgdp | 1.0000 CFgdp | 0.8040 1.0000 ESFgdp | 0.7939 0.7914 1.0000

Table D.1: Effect of Lagged ERDF

variable of interest: Lagged ERDFgdp)

---

--- Gdpgrowth 0.343** 0.372** 0.352** 0.373*** (0.134) (0.139) (0.124) (0.109) Lngdp -0.294* -0.359* -0.442** -0.506** (0.133) (0.159) (0.177) (0.18) Invgdp 0.0938 0.104 0.222 0.265 (0.266) (0.316) (0.294) (0.289) Humcapital 0.872** 0.935** 1.357*** 1.344*** (0.282) (0.322) (0.352) (0.375) Popgrowth005 0.0271 -0.839 -0.379 0.198 (0.707) (0.935) (0.881) (0.954) Lagged ERDFgdp 0.904 -0.149 -1.23 1.182 (0.788) (1.082) (1.441) (3.459) Corruption 0.0704** 0.0785** 0.0998** 0.0809 (0.0263) (0.0339) (0.0358) (0.0484) Tradeopenness 0.0111 0.00772 0.0249 0.00527 (0.0326) (0.0444) (0.0423) (0.0347) Fdigdp 0.0241 -0.0185 0.0185 0.0575 (0.0864) (0.0966) (0.0922) (0.0922) Ecfreedom 0.000386 -0.000887 -0.00176 -0.00288** (0.00156) (0.00195) (0.00141) (0.00109) _cons 1.846 2.504* 2.900* 3.545** (1.114) (1.269) (1.401) (1.401) --- N 110 100 90 80 adj.R-sq 0.769 0.781 0.802 0.817 ---

(28)

Table D.2: Effect of Lagged ESF

variable of interest: Lagged ESFgdp)

---

--- Gdpgrowth 0.333** 0.372** 0.359** 0.372*** (0.13) (0.143) (0.127) (0.112) Lngdp -0.297* -0.359* -0.446** -0.516** (0.137) (0.159) (0.179) (0.178) Invgdp 0.0993 0.104 0.215 0.277 (0.289) (0.326) (0.304) (0.286) Humcapital 0.828** 0.943** 1.327*** 1.369*** (0.28) (0.352) (0.352) (0.367) Popgrowth005 0.145 -0.794 -0.478 0.214 (0.719) (0.903) (0.923) (0.899) Lagged .ESFgdp 2.967 0.214 -5.406 5.463 (3.123) (3.345) (6.728) (6.27) Corruption 0.0673** 0.0771** 0.0981* 0.0801 (0.0253) (0.0339) (0.0457) (0.0469) Tradeopenness 0.015 0.00518 0.0157 0.00466 (0.034) (0.039) (0.0373) (0.0381) fdigdp 0.0177 -0.0148 0.0186 0.0463 (0.0894) (0.0973) (0.0914) (0.0926) Ecfreedom 0.000305 -0.000797 -0.00157 -0.00282** (0.00142) (0.00197) (0.00156) (0.00102) _cons 1.901 2.494* 2.958* 3.603** (1.119) (1.28) (1.438) (1.383) --- N 110 100 90 80 adj.R-sq 0.769 0.781 0.801 0.818 --- Standard errors in parentheses

(29)

Table D.3: Effect of Lagged CF

variable of interest: Lagged CFgdp)

--- 2-Years 3-Years 4-Years 5-Years --- Gdpgrowth 0.349** 0.343** 0.330** 0.375*** (0.122) (0.122) (0.129) (0.115) Lngdp -0.299** -0.331* -0.448** -0.507** (0.131) (0.149) (0.176) (0.182) Invgdp 0.0693 0.0685 0.219 0.305 (0.274) (0.325) (0.305) (0.296) Humcapital 0.845** 0.920** 1.358*** 1.392*** (0.286) (0.337) (0.337) (0.347) Popgrowth005 0.0398 -0.9 -0.507 0.569 (0.714) (0.967) (0.918) (0.86) Lagged CFgdp 0.588 -3.072 -2.267 5.453 (1.859) (2.978) (3.15) (5.475) Corruption 0.0717** 0.0859** 0.0976** 0.0755 (0.0283) (0.031) (0.0381) (0.0465) Tradeopenness 0.0139 0.0244 0.019 0.00348 (0.0349) (0.0463) (0.0469) (0.0365) Fdigdp 0.0276 -0.00222 0.018 0.0677 (0.0876) (0.102) (0.0911) (0.0807) Ecfreedom 0.000208 -0.00115 -0.00148 -0.00229 (0.0015) (0.00204) (0.0015) (0.00147) _cons 1.923 2.282* 2.955* 3.436** (1.082) (1.188) (1.395) (1.378) --- N 110 100 90 80 adj.R-sq 0.766 0.785 0.801 0.821 ---

Do European structural funds have an impact on economic growth in Central and Eastern Europe (2001 – 2013)