University of Groningen Faculty of Economics and Business

University of Groningen

Faculty of Economics and Business

M.Sc. International Economics and Business Master Thesis

June 2016

The Impact of Trade Liberalisation on Total Factor Productivity:

The Case of the Central and Eastern European Countries

Laimdota Jarmusevica s2966530

2

Abstract

This thesis investigates the impact of joining the EU and trade liberalization on total factor productivity in 8 European Union countries that joined the European Union in 2004. We find the determinants of total factor productivity growth in Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak Republic and Slovenia and their impact of TFP growth.

The empirical results indicate that trade openness has a positive effect on TFP growth. However, joining the EU did not influence TFP growth.

3

Table of Content

3. Theoretical and Empirical Review ... 12

3.1. Total Factor Productivity ... 12

3.2. Trade Liberalisation and TFP growth ... 13

3.3. Other determinants of TFP Growth ... 14

4

List of Figures

Figure 1. GDP per capita in current $US ... 7

Figure 2. Teriff rate, applied, weighted mean, all products (%) ... 8

Figure 3. Exports of goods and services in billions of current $US. ... 9

Figure 4. Imports of goods and services in billions of current $US ... 10

Figure 5. Defining Output Growth ... 13

Figure 6. TFP growth % change ... 20

Figure 7. Manufacturing share of value added ... 34

List of Tables

Table 2. List of countries ... 25

Table 3. Summary statistics ... 27

Table 4. Normality test ... 28

Table 5. Correlation matrix of variables ... 29

Table 6. Test for multicollinearity ... 30

Table 7. Breusch-Pagan/Cook-Weisber test for heteroskedasticity ... 30

Table 8. Hausman test ... 31

Table 9. Pesaran's test of cross sectional independence ... 31

5

1. Introduction

In recent years European Union (EU) member states have been experiencing weak productivity growth. Between 2000 and 2012 total factor productivity (TFP) growth has been only 0.2% average per year (McQuinn and Whelan, 2013). The trend of decreasing growth may be a negative predictor of economic performance among the EU member states and poor innovation and technology development. It harms competitiveness of the European Union in the world. Especially, that TFP growth is the sustainable source of long-term economic growth. It does not include limited components as other two production factors as labour and capital, and it is mainly technology-driven. Moreover, technological progress contributed to TFP growth has been an important factor behind a convergence of the OECD countries in the post-war period (Nadiri and Kim, 1996).

In 2004, 10 new member countries joined the European Union and its market. Since new countries had the opportunity to join the free trade area within the borders of the union, surprisingly, influence of trade liberalization on the total factor productivity of the Eastern European countries has not been addressed before. Despite, available literature and empirical studies does not include new EU member states in research groups.

The year 2004 started a new integration and developments stage for each of 10 the new EU member states. Countries joined the free market that gave access to free movement of goods, services, labour and capital. It is important to distinguish a direct impact of trade on the TFP growth in Central and Eastern European countries. Increasing international trade has direct and indirect influence on the TFP growth through variety of channels described in the thesis. Besides free market, the EU has particular geographic, economic and political environment that includes factors that may play a role as drivers of the TFP growth among new member states. The main aim of the research is to investigate determinants of TFP in Central and Eastern European countries that joined the EU in 2004, and to reveal, if joining the EU and trade liberalization had an influence on the TFP growth of 8 European countries.

6

identify the main TFP growth determinants that had enhancing or reducing effects on TFP growth.

This paper is related to a well-established literature of total factor productivity growth, its determinants and the effect of trade liberalization on TFP. The thesis is related to Anandhika (2013), Jajri (2007) and Khan (2005) who performed analysis of TFP determinants in ASEAN, Malaysia and Pakistan, respectively.

7

2. Background

On 1 May 2004, the European Union (EU) experienced its largest enlargement. 10 European countries joined the EU, among them where 8 post-communist states. The enlargement drew re-unification and integration of the Europe that occurred after disintegration of the iron curtain. However, transition of the countries from planned to market economies started with a deep recession in early 1990’s, growth in Central and Eastern Europe exceeded 5,1% comparing 1,9% in the old Europe (Burda andSevergnini, 2009).

One of the most commonly used development indicators is GDP per capita. Figure 1 represents GDP per capita growth in Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland Slovak Republic and Slovenia from 1995 up to 2014. GDP per capita is important indicator of growth. Its value has more than doubled for all observed countries. It is hard to acknowledge convergence of GDP per capita between countries, because there is larger deviation in 2013 and 2014 than it was in 1995 at least for 7 countries, but definitely there is a presence of rapid increase in GDP in early 2000’s.

Figure 1. GDP per capita in current $US

The World Databank, World Development Indicators

Integration within the EU includes political integration, institutional development, trade integration, financial integration and free mobility of capital, labour and knowledge (Becker, 2010). In 1900’s, when Eastern Europe got rid of command economy and started preparation for integration with other European countries, trade and capital market integration has been established as a priority of the EU Eastern politics (Boeri and Brucker, 2001). The candidate countries signed membership related agreements that prepared the countries for economic

8

integration. Firstly, it started a gradual removal of tariffs on manufactured goods. Since 2004, customs policies of the new member states have been regulated by EU legislations. The common customs tariff (CCT) has been adopted by European institutions in 1968 and it is proclaimed as a major achievement of European integration (Europedia). The new member states of the EU by joining the EU market cancel import tariffs and quotas for import from other EU member countries and imply the common import tariff from the third part countries.

The Figure 2 represents the applied weighted mean tariff rate of all products imported by 8 countries. The import rate in the World Bank Databank appears as identical for all sample countries. In early 2000’s it starts to decrease rapidly that draws a trade liberalization experienced by those countries.

Figure 2. Tariff rate, applied, weighted mean, all products (%)

World Bank Databank

As mentioned earlier joining the EU opened the new market without custom duties. Therefore, in this case is also important to look at trade related indicators as export and import that provide evidence of trade volume changes in the countries.

Figure 3 examines export change in Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland Slovak Republic and Slovenia from 1995 up to 2014. The rapid growth of export started in early 2000’s before the countries joined the EU. All countries experienced the peak of export before the crisis in 2008 and the dip in 2009. Later almost all countries kept increase of export value, with few exceptions that were related with Eurozone crisis in 2010 and that touched Eurozone countries. To sum up, the volume of export has been increasing since joining the EU in 2004 with few fluctuation related to the global crisis and the Eurozone crisis.

0 0,5 1 1,5 2 2,5 3 3,5 4 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

9

Figure 3. Exports of goods and services in billions of current $US.

The World Databank, World Development Indicators

The Figure 4 represents import change in Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland Slovak Republic and Slovenia from 1995 up to 2014. Likewise, in the case of export growth, import growth started rapidly in early 2000’s. It faced a dip after the crisis in 2009, and followed growth later on, with fluctuations in Eurozone countries due to Eurozone crisis. To sum up, majority of countries experienced more than 4 times increase of export and

10

import between 1995 and 2014. Although, there are observed fluctuations related to the global crisis and the Eurozone crisis.

Figure 4. Imports of goods and services in billions of current $US

11

We observe increase in volume of trade in all 8 countries that provides evidence that trade increase occurred shortly before joining the EU and experiences rapid growth after joining the EU.

Nevertheless, the results proposed by Becker et al (2010) show that the main resource of growth was capital accumulation. Between 1995 and 2012 the Central and Eastern European countries have followed an extensive and investment-oriented growth model. Growth of labour productivity and multifactor productivity had marginal effect only. During this period the region countries became the most dynamically developing in the world. Becker et al (2010) revealed common aspects of the pre-crisis economic growth of 10 countries in addition to the main source of growth. Foreign direct investment level was reached 8-10% of GDP before crisis.

12

3. Theoretical and Empirical Review

In this section we discuss the theoretical literature and empirical researches about growth and TFP and its growth, and determinants. In the end of the section the research questions and hypothesis are proposed.

3.1. Total Factor Productivity

Total factor productivity (TFP) measurement history started with Solow (1957) when the author introduced additional production factor. At that moment other two factors – labour and capital, were not able to explain output produced, and the new component of the equation was explained as measure of efficiency of factors used in production. Therefore, a positive TFP growth verifies efficient factors use, when higher TFP results in output growth. Bosworth and Collins (2007) approve by the statement that TFP measures “gains in the efficiency with which inputs are used, including technical progress as well as myriad other determinants”1. Despite, TFP growth does not include only technology side of the production, it relates to both technology improvements and human capital improvements that lead to better technology use. Moreover, it may be influenced through various channels described later.

Importance of TFP is reflected in overall output growth that is determined by labour productivity and number of workers and working hours involved in production. Labour productivity reflects increase in contribution of capital, increase in contribution of human capital and increase in overall efficiency of production - total factor productivity (see Figure 5). Despite, TFP growth measure is not influenced by other production input factors, and it exists by itself.

1

13 Figure 5. Defining Output Growth

The Conference Board, 2015

Mainly contribution of productivity of the factor endowments as labour and capital drives TFP growth that may occur as spillover result from use of better technology and equipment or improved management and human capital input. Both production factors physical and human capital are important and are influenced by free trade that will be discussed in the next section.

3.2. Trade Liberalisation and TFP growth

It is difficult to provide a direct theoretical link between trade and growth, especially, between economic integration that occurs in the EU and growth of the country. There are different channels of influence besides trade liberalisation. However, there is a literature that assumes free trade expected to be a significant factor for economic growth and prosperity of a country, because free trade policy does not restrict imports and exports of goods that bring into the market new goods, services and knowledge about them. Economidou and Murshid (2007) propose that trade is a vehicle for transmitting technologies and knowledge.

Jang and Kim cited in Anandhika (2013) provide two theoretical reasons. First comes from the neoclassical theory and Heckscher-Ohlin model, when a country has to enrich its comparative advantage to compete in international trade. It may occur through the inter-sectoral

Output Growth

Increase in output per worker or worker hour (labour

produtctivity)

Increase in contribution of capital (machinery, equipment, structures)

Increase in contribution of human capital (skills, managemnet

competencies)

Increase in overall effieciency of production (total factor

producitivity)

14

redistribution of factors, as well, as through the external restructuring of productivity. Increasing trade implies openness to new techniques of production and their import that leads to faster TFP growth. Moreover, trade between EU member states imply less trade costs, therefore it reveals the comparative advantage and lead to the production specialization introduced by Ricardo. The second opportunity is restructuring of productivity. In this situation more-productive firms can resist harsh competition, while less-productive ones face competition and decide to exit a market. After liberalisation of trade production resources will be allocated into more productive firms (Melitz, 2003). Melitz model determines that openness of trade brings competition to the domestic market, besides, provides export opportunities for more-productive firms. Free trade increases the number of export-oriented domestic firms and enhances their productivity. Factors move to the best uses. It leads to more efficient use of the production factors.

While the theory supposes output growth through increase of efficiency of resource allocation the empirical evidences provide several channels of an influence that include more determinants as foreign direct investment (FDI), research and development (R&D) or education. It is useful to examine all of them.

3.3. Other determinants of TFP Growth

Nevertheless, besides the trade openness, there are other factors described in empirical researches that directly or indirectly influence TFP growth as FDI, human capital, R&D, institutional framework, capital deepening, etc.

15

Empirical studies usually propose that increasing FDI results in R&D not only through international cooperation and spillovers, but due to rising competition in internal market that forces technologies to innovate. Grossman and Helpman (1991) argue that foreign competition forces domestic firms to innovate. Innovation-based growth model discussed by Grossman and Helpman (1991) suggests that R&D leads to innovations, and that results in increase of output. FDI has been admitted as important source of productivity due to its effect on spillovers between domestic and foreign firms. Borensztein et al (1998) considers that FDI influences economic growth through improvements of technology and productivity. FDI leads to technology and knowledge transfer. However, there may take place a negative effect. Previous study of Djankov and Hoekman (2000) showed Czech firms have experienced a negative effect of FDI, because they did not have foreign partnership, but faced harsh competition between domestic and foreign-owned firms. R&D includes international spillovers that are influenced not only by FDI and direct foreign research or management, but also by international cooperation, international trade (Nadiri and Kim, 1996). In this situation, trade openness has its own share, it results in better ability to benefit from technologies through technology import.Madsen (2007) found that 93% increase in TFP among OECD countries has been due to imports of knowledge.

Stated by Jajri (2007) TFP determinants can be organized into two conceptual variable groups – economic restructuring and education and training. The first as discussed before is caused by allocation of resources from less productive sectors to more productive ones. The latter includes education of workers, upgrade of skills that result in efficient work and better-quality products and services. The model used in the study of Jajri (2007) defines determinants of TFP – investment rate, trade ratio to GDP, percentage of foreign-owned companies, annual manufacturing sector output growth, and percentage of employed persons acquired tertiary education.

In turn, Danquah, Moral-Benito, and Outtara (2011) include total 19 TFP determinants. Among them is not only trade openness and investment share, but such human capital variables as population, population under 15, labour force, secondary education and life expectancy that indirectly influence TFP growth. But Danquah, Moral-Benito, and Outtara (2011) findings admit that the most robust TFP growth determinants are unobserved heterogeneity, initial GDP, consumption share and trade openness.

16

output. Country with better labour quality is more able to benefit from openness of trade and FDI.

Previous studies also focused on political and institutional framework as possible productivity driver. These factors are important in the EU due to its political integration and common governance practices among the members. Holmes et al (2011) pays attention to need of regulatory institution that establishes rules and reduces uncertainties, for example, through legal enforcement of contracts and property rights. The presence of such institutions and corruption control enhance activity of investors. For example, Becker et al (2010, 2012) found a positive impact of the Structural Funds on output growth. Majority of Eastern European countries used an opportunity of the EU funds to catch up developed union members. Moreover, efficient and productive institutions have a positive impact on aggregate productivity through boosting investments. The study of Doucouliagos and Ulubasoglu (2004) present that there are more indirect effects of institutional framework that are constantly ignored. The empirical results show that democracy has positive effect on TFP growth. There are a lot of specific influence channels that are captured under the institutional framework as political situation, regulations and implimitations systems that need to be addressed as possible TFP determinants. The significance of such determinant as institutional framework is expresses by Boeri and Brucker (2001). The authors argue that there is a case if short-term institutional changes do not lead to TFP growth, Eastern European countries will remain “second-class members for a long time”.

All factors described above are as important as trade flows, because the EU has not only economic, but political and social impact on the member countries. Moreover, some of them, for example, R&D may be influenced by trade openness as well. Integration of the countries that join the EU is deeper than just usual free trade agreement example, and its influence on TFP growth has to be investigated including all possible influence factors. The Table1 represents TFP growth determinants discussed in literature review. We will check these determinants in empirical part.

Table 1. Determinants of TFP growth

Determinants Literature

Trade openness Danquah et al (2011);

De Hoyos and Iacovone (2011);

Export Anandhika (2013);

17

Nadiri and Kim (1996);

Education Barro (2001)

FDI Borensztein et al (1998)

Djankov and Hoekman (2000);

Institutional framework Becker et al (2010, 2012)

Concerning the rest of the world, there are several research papers observing integration and free trade impact on NAFTA countries. For example, Schiff and Wang (2003) found that NAFTA agreement has increased TFP in manufacturing of Mexico by more than 5.5%. Despite the positive effect, the controversial result showed Segerstrom and Sugita (2015). The authors found that industrial productivity increased in non-liberalized industries than in liberalized ones in Canada after joining NAFTA. The results of Segerstrom and Sugita (2015) reject the Melitz model that assumes countries involved in free trade experience an increase of productivity due to competition and resource reallocation (Melitz, 2003).

Plenty of studies are more focused on productivity enhancing in NAFTA or ASEAN, and the old EU countries. Opposed results of those studies indicates a need to find how trade liberalisation influenced productivity growth among the new EU members. Previous studies by Bretschger and Steger (2004) and Gao (2005) found effects of the European integration process on growth in general, but it is also important to provide new evidences of joining to the EU and its market impact on TFP.

This paper will identify possible determinants of TFP growth in 8 EU countries – Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak Republic and Slovenia. We will check if joining the EU market had an influence on TFP growth.

After examining theoretical and empirical literature on determinants of TFP growth and trade liberalisation we can propose following hypotheses:

Hypothesys 1: Trade openness is positively associated with the TFP growth in the Central and

Eastern European countries.

Hypothesis 2: The EU membership is positively associated with the TFP growth in the Central

18

19

4. Methodology

4.1. TFP growth

There are 3 contributing elements in growth accounting: labour input, capital input and TFP. Labour and capital are known as “factor of production” that represent workforce and capital goods as building, machines, vehicles, tsc. TFP refers to the rest that is known, for example, as methods employed by labour, capital that lead to more efficient and faster production, but TFP may refer to productivity through R&D, education, government efficiency, etc. (Jajri, 2007)

Cobb-Douglas production function, where TFP is captured by the variable A:

(1)

In this function measures technology contribution to output ( ) in country c at time t. The function may be As into log-linear Cobb-Douglass form, it is transformed into the ‘stochastic frontier model’ as described as:

(2) or

(3)

Residual approach (Solow’s approach) calculates TFP through growth accounting equation:

(4)

TFP value determines that the same level of inputs (labour, capial) produces higher output value. Our task is to recognize TFP determinants among the Eastern and Central EU member countries and find out the effect of joining the EU on the TFP growth.

4.1.1. TFP growth in 8 EU countries

20

only positive or only negative TFP growth during this period. Almost all countries experienced negative TFP growth in 2008 and 2009 that is a result of global crisis in 2008. TFP growth in Poland is captured with huge fluctuations. The lowest TFP growth among all countries refers to Lithuania. It was -12.43% in 2010. The highest TFP growth was in Estonia in 1998 – 9.77%. The effect of the crisis was short, therefore in 2010 all countries showed positive or 0 TFP growth. However, in 2013, 6 out of 8 countries had negative or 0 TFP growth. Overall, TFP growth follows the same path in majority of countries.

Figure 6. TFP growth % change

21 Conference Board Total Economy Database

4.2. Determinants of TFP

The next step is to consider possible determinants of TFP growth. As considered by Danquah et al (2007) there is a lack of a theoretical guidance and plenty of empirical studies on the choice of TFP determinants. Such TFP determinants as openness of trade, cover ratio, FDI share, labour share with tertiary education have been taken from the researches of Anandhika (2013), Danquah et al (2007), Jajri (2007) and Khan (2006). Additional control and dummy variables have been used by the initiative of the author.

To find all possible effects that can enhance TFP growth besides joining the EU the model consists of 15 more independent variables. Some of them belong to trade liberalisation consequences, for example, variables connected to trade flows, FDI, R&D, and institutional variables may reflect political effects of joining the EU. TFP growth and labour productivity from previous year have been taking to prevent possible spillover between years, when the productivity growth from t-1 year influences productivity growth it the year t.

1 . Openness

is a share of trade (export plus import) with respect to GDP. Openness of economy represents country’s engagement in trade. It may enhance TFP growth through export and import. In the export case, TFP is increasing due to international competition faced by domestic firms and their willingness to keep market share. Respectively, in the import case, TFP

22

is increasing due to import of new goods and technologies that are involved in production of new products.

Cover Ratio

variable represents cover ratio. It captures export divided by import. The ratio is higher if a country exports more than imports. It reflects the production for domestic and foreign markets. Higher export encourages a rapid technological progress due to increasing number of exporters and international competitiveness.

FDI share

variable represents foreign direct investment inflow % of GDP. FDI share reflects a presence of foreign companies and foreign ownership that leads to potential technology and knowledge transfer. In the result, it has positive influence on TFP.

Labour with tertiary education

variable represents a labour force with tertiary education in % of total labour in the country. Higher education level of labour results in higher productivity, better quality products and services. It determines better ability of workforce to use and master advanced technologies.

Manufacturing share

variable represents manufacturing share of value added. The importance of

manufacturing share is due to its connection with TFP growth that is faster in manufacturing industries. Including manufacturing share of value added as independent variable helps to examine its connection to aggregate TFP growth.

R&D

variable represents research and development expenses with the respect to GDP.

23 Labour productivity level

variable represents labour productivity per person employed in 2014 $US. TFP is a key determinant of labour productivity growth. Labour productivity growh may be expressed as contribution of TFP growth and “capital deepening” (McQuinn, 2015).

Government effectiveness

variable represents the index of government effectiveness creates by the World Bank. As stated by the World Bank government effectiveness “captures perception 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”2

. The government is one of the main driving forces in the economy, especially, its importance is observed in factor accumulation. If the government’s consumption is unproductive it does not enrich growth, and legislations proposed by the government create sustainable environment for productivity growth.

Corruption Control

is a World Bank indicator “Control of Corruption that “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”3. Ugur and Nandini (2011) gathered 72 studies about corruption together and concluded that corruption had negative effect on growth. Moreover, they found evidences that corruption has stronger negative effect in middle and high-income countries.

Dummy of EU Membership

is a dummy variable that appears as 1 for the years up from 2004 when countries joined the European Union, and 0 for years to 2003 including when they did not. The dummy variable allows us to see the EU membership and its free market contribution to TFP growth.

Dummy of Crisis in 2008

is a dummy variable for the global crisis in the year 2008. All observed countries have a value of dummy variable 1 in the 2008, and 0 for the rest years. From

2 _{Citation of the World Bank, available online: http://info.worldbank.org/governance/wgi/pdf/ge.pdf} 3

24

the background section we saw that crisis in 2008 had disruptive impact on GDP per capita, export and import. Therefore, we decided to include the crisis dummy variable into the regression due to expectation of its negative impact of TFP growth.

TFP growth and Labour productivity level from previous year

and are TFP growth and labour productivity from previous years are included as variables to prevent convergence of productivity growth due to growth in previous year.

4.3. Econometric Issues

25

5. Data Sources

The paper involves a panel dataset of 8 Central and Eastern European countries – Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak Republic and Slovenia. The countries are listed in the Table 2.

Table 2. List of countries

Country Number in the panel dataset Czech Republic 1 Estonia 2 Hungary 3 Latvia 4 Lithuania 5 Poland 6 Slovak Republic 7 Slovenia 8

The data on the aggregate TFP growth and GDP growth is taken from Conference Board Total Economy Database. It covers the period from 1997 to 2013 necessary for analysis, and includes data for all 8 countries. The more detailed industry level evidence would be better to provide TFP growth analysis in the EU. Unfortunately, the industry level TFP growth data for the sample states is not observed in EU KLEMS Growth and Productivity Accounts. Therefore, we use aggregate economy data for understanding impact of joining the EU on the aggregate TFP growth. As mostly observed countries are more developed in service sector and manufacturing takes not a big part of the production, the aggregate TFP growth is relevant indicator of productivity growth that captures not only manufacturing sector, but others as well, that may provide reliable results due to specific production condition.

26

Data on labour productivity per person employed in 2014 $US is obtained from the Conference Board Total Economy Database.

27

6. Results

Before estimating the model some descriptive statistics of the variables are presented in the section 6.1. The section 6.2. provides econometric issues. The section 6.3. includes estimate results of regressions.

6.1. Summary statistics

The Table 3 provides summary statistics of variables used in the empirical research. We use one dependent variable – TFPGrowth, and 15 independent variables explained before. The highest number of observations is 136. RD variable has missing values for Lithuania the years from 1997 to 2003. Later we provide regression with variables that observations count 136, and one additional regression that include RD variable that has less observations.

Two of the variables are dummy variables that have 0 or 1 value for respective years. We use natural logarithms of two variables (labour productivity levels in going and past year) to prevent an abnormal distribution.

Table 3. Summary statistics

Variable Obs Mean Std. Dev. Min Max

TFPGrowth 136 .9651627 3.412738 -12.42735 9.766198 Openness 136 1.170032 .3057729 .509017 1.834276 CoverRatio 136 .9343227 .1213317 .0342167 1.092492 FDIShare 136 .0507314 .0674671 .1609109 .5078472 TertEduc 136 .2260221 .0888634 0 .44 GovEffect 136 74.99587 5.40644 60 84.95145 CorrCntrl 136 68.84124 7.741607 48.29268 85.85366 lnLabProdLevel 136 10.79196 .2312605 10.11171 11.13073 lnLabProdLevelPreYear 136 10.75972 .2426017 10.11171 11.10877 TFPGprev 136 1.141097 3.465351 -12.42735 9.766198 ManufShare 136 20.03938 3.81308 10.78377 26.15517 DummyEU 136 .5882353 .4939724 0 1 DummyCrisis2008 136 .0588235 .236164 0 1 RD 129 .009576 .0047791 .0032 .0259 6.2. Econometric Issues

28

CoverRatio, FDIShare, DummyCrisis2008, TFPGprev, GovEffect, lnLabProdLevel, lnLabProdLevelPreYear and ManufShare are normaly distributed can be rejected. It has to be admited as one of the main limitations of the data sample. At least, skeweness has not been detected for DummyEU and ManufShare. But kurtosis has not been detected in the sample of GovEffect, CorrCntrl, lnLabProdLevel and lnLabProdLevelPreYear.

Table 4. Normality test

Skewness/Kurtosis test for

Normality

joint

Variable Obs Pr(Skewness)

Pr

(Kurtosis) adj chi2(2) Prob>chi2

29 Table 5. Correlation matrix of variables

The Table 5 indicates correlation table of dependent variable and independent variables. Correlation matrix is an option how we can detect multicollinearity in the case if several variables are correlated. The correlation table reveals a problem of multicollinearity among the variables.

From the table we can see that the natural logarithm of labour productivity level is highly correlated to the natural logarithm of labour productivity level in previous year. We do not exclude any of those variables, because the correlation between them is assumed as possible. Both variables have tight relation because of similar values and their incontestable connection.

Surprisingly, both the natural logarithm of labour productivity level and the natural logarithm of labour productivity level in previous year are highly correlated with the EU dummy variable.

30

The Table 6 represents results of VIF test for multicollinearity. The results reveal multicollinearity in the model. VIF values are high for the natural logarithm of labour productivity level and the natural logarithm of labour productivity level in previous years. Mean VIF is higher than 1, and we prove the presence of multicollinearity.

Table 6. Test for multicollinearity

Variable VIF 1/VIF

lnLabProdL~l 69.50 0.014388 LabProdLev~r 67.36 0.014845 DummyEU 3.44 0.290687 GovEffect 3.18 0.314099 TertEduc 3.07 0.325771 CorrCntrl 2.54 0.394017 ManufShare 2.44 0.410465 Openness 2.22 0.451106 CoverRatio 1.83 0.546813 TFPGprev 1.21 0.826531 DummyCr~2008 1.21 0.829126 FDIShare 1.14 0.879914 Mean VIF 13.26

The next task is to check if the data sample is homoskedastic or there is a presence of heteroskedasticity. Table 7 shows the result of Breusch-Pagan/Cook-Weisberg test for heteroskedasticity. The chi-square value is small that shows there is not heteroskedasticity detected.

Table 7. Breusch-Pagan/Cook-Weisber test for heteroskedasticity Breusch-Pagan / Cook-Weisberg test for heteroskedasticity

Ho: Constant variance

Variables: fitted values of TFPGrowth

chi2(1) = 0.17 Prob > chi2 = 0.6790

31 Table 8. Hausman test

The Table 9 reveals results of Pesaran cross-sectional dependence test for fixed-effects. The test is used to test correlation of residuals across entities. Cross-sectional dependence detected from the test can lead to bias in results. To prevent possible bias in results we apply Driscoll and Kraay standard errors in regression analysis.

Table 9. Pesaran's test of cross sectional independence

Pesaran's test of cross sectional independence = 6.115, Pr = 0.0000 Average absolute value of the off-diagonal elements = 0.312

6.3. Estimation Results

32

productivity level on TFP growth. However, the standard error has the largest value among all variables. The natural logarithm of labour productivity level of previous year has highly significant negative impact on TFP growth. The labour share with tertiary education has positively influences TFP growth, but at the same time it is poorly significant. Manufacturing share impact on TFP growth is positive, tiny and highly significant In the Model 1 and Model 2 crisis dummy variable appears highly significant and, obviously, negative. The introduced variable confirmed to R&D expenses as a share of GDP has a high coefficient value, but at the same time it is insignificant and has the largest standard error. After introducing the variable other coefficients epxerience little changes. TertEduc doubles its value and becomes highly significant. Conversely, government effectiveness variable loses its significance. TFP growth in previous year has small and poorly significant influence on TFP growth in ongoing year.

Table 10. Estimate results

Model 1 Model 2

VARIABLES TFPGrowth TFPGrowth

33 (68.40) Constant 12.28 17.24 (30.19) (29.73) Observations 136 129 Number of groups 8 8 R-sq 0.76 0.77 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

7. Discussion

The regression analysis shows empirical findings regarding the sample of TFP growth determinants in the eight Central and Eastern European countries that joined the EU in 2004. Consistent with theory and empirical section we conclude that trade openess has a positive and significant impact on TFP growth. The results come along previous empirical findings of Anandhika (2013), Danquah et al (2011), De Hoyos and Iacovone (2011), Jajri (2007).

. If the trade volume with respect to GDP increases it enhances TFP. As refered in the literature review it might be caused by competition due to new export channels or imported substitutes that forces domestic firms to innovate and implement new technologies into production. The finding regarding the impact of openness of trade confirms the first hypothesis that a trade openness is positively associated with the TFP growth in the eight Central and Eastern European countries.

However, cover ratio does not reveal significant impact as the determinant of TFP growth. Therefore, we are not able to assume that only export or only import is a determinant of TFP growth.

Unfortunately, the second hypothesis that the EU membership is positively associated with the TFP growth in the eight Central and Eastern European countries is rejected, because the results show negative sign of coefficient of the variable that refers to the EU membership. It means that the membership in the EU reversely to expected results, might lead to a reducing effect of TFP growth, although, the coefficient does not appear as significant. The EU membership with opportunities of free market and the highest level of trade liberalisation within itsmember did not increase TFP growth during the time period from the year 1997 till 2013.

34

appear as significant forces that would influence TFP growth. R&D expenses as a share of trade have high value of coefficient, but, at the same time, it has high value of standard error.

The labour force with tertiary education is positively and significantly associated with TFP growth. Its significance level grows as we include an addition variable that refers to R&D expenses.

Government effectiveness is highly significant until we add a variable that refers to R&D expenses. Later it loses the significance level and its coefficient shows poor contribution to TFP growth.

The empirical results suggest that a manufacturing share of value added is important determinant of TFP growth. If the manufacturing share grows it enhances TFP growth in observed country sample. Unfortunately, manufacturing share of valued added is decreasing in Eastern European countries. The Figure 7 shows that all countries experienced shrink of manufacturing share of value added produced in the country.

Figure 7. Manufacturing share of value added

World Input-Output Database

It might become unpromising future due to important manufacturing sector role in output growth. Van Ark et al (2011) argue that manufacturing activities usually have been those that included innovation and technological change. Manufacturing production is more standardized that allows implementing technologies and observing actual results of technological change when services mostly focus on interaction with consumer and introduction of technology does not provide results quickly.

35

8. Conclusion

This research was conducted in a way to examine the effect of trade liberalization and membership in EU on the total factor productivity (TFP) growth in eight countries that joined the EU in 2004 – Czech Republic, Estoni, Hungary, Latvia, Lithuania, Poland, Slovak Republic and Slovenia. We observed time period from 1997 till 2014 that captures the EU enlargement in 2004 and global crisis in 2008.

As determinants of TFP growth were chosen and checked such variables as openness of trade, cover ratio, FDI share of GDP, labour force with tertiary education, government effectiveness, corruption control, R&D expenses as share of GDP and EU membership. We find a positive, significant influence of openness of trade on TFP growth. The share of labour force with tertiary education also is positively associated with TFP growth. Unfortunately, the empirical results do not find positive association between EU membership and TFP growth. Such variables as FDI share and R&D expenses that were expected to have positive influence do not appear as significant determinants of TFP growth.

Understanding the TFP growth determinants is important for future economic growth of the union, because the empirical results may not follow the theoretical framework. Moreover, the countries and county unions need to be addressed as heterogeneous formations.

8.1. Limitations

The thesis has a few important limitations.

Firstly, available databases do not include industry data on TFP growth for countries presented in the research. In the future, the data for those countries needs to be specified among the industries for detailed researches of TFP changes. It would help for new researches on TFP growth determinants and tendencies not only on the aggregate economy level, but as well within different industries, especially in manufacturing sector. Moreover, due to incomplete data collection in the sample countries, much of data for earlier years might be biased.

Secondly, the applicability of the results are restricted to eight EU countries. Future researches may assign the same TFP growth determinants to the whole union.

36

References

1. Anandhika M. R., “Total Factor Productivity Determinants on ASEAN Free Trade Ares: Free Trade Ares Channels Evaluation”, Research Paper IDEC8010, Crawford Schoold of Public Polic, 2013

2. Bah El-hadj M. , Brada Josef C., “Total Factor Producitivity Growth and Structural Change in Transition Economies”, Comparative Economic Studies, Vol. 51, Issue 4., 2009

3. Barro R.J., “Human Capital and Growth”, The American Economic Review, Vol. 91, No. 2, May 2001

4. Becker, S.O., Egger, P.H., von Ehrlich, I., “Too much of a good thing? On the growth effects of the EU’s regional policy”, European Economic Review, 2012

5. Becker, S.O., Egger, P.H., von Ehrlich, I., “Going NUTS: the effect of EU structural funds on regional performance.”, Journal of Public Economics, 2010

6. Ben-David D., Loewy M. B., “Free Trade, Growth, and Convergence”, Journal of Economic Growth, Volume 3, 1998

7. Benkovskis K., Fadejeva L., Stehrer R., Worz J., “How Important is Total Factor Productivity for Growth in Central, Eastern and Southeastern European countries”, Latvijas Banka, Working paper, 5, 2012

8. Boeri, T., H. Brucker et al., „The Impact of Eastern Enlargement on Employment and Labour Markets in the EU Member States”, Report for the European Commission, DG Employment and Social Affairs, Brussels, 2001

9. Borensztein E., De Gregorio J., Lee J-W., “How does foreign investment affect economic growth?”, Journal of International Economics, 45, 1998 10. Bosworth B., Collins S. M., “Accounting for Growth: Comparing China

and India”, Working Paper 12943, National Bureau of Economic Research, 2007

37

innovations: an empirical analysis for 13 OECD countries”, Applied Economics, 2013

13. Bretschger L., Steger T., “The Dynamics of Economic Integration: Theory and Policy”, WIF- Institute of Economic Research, Working Paper 04/32, 2004

14. Danquah M., Moral-Benito E., & Outtara B., ‘TFP growth and its determinants: nonparametrics and model averaging”, Bank of Spain 1104, 2011

15. Djankov S., Hoekman B., “Foreign Investment and Productivity Growth in Czech Enterprises”, World Bank Economic Review, Volume 14, Issue 1, 2000

16. Dovis M., Milgram-Baleix J., “Trade, Tariffs and Total Factor Productivity: the Case of Spanish Firms”, Instituto Valenciano de Investigaciones Economicas, S.A., 2008

17. Economidou C., Murshid A.P., “Testing the Linkages between Trade and Productivity Growth”, Utrecht School of Economics, Discussion Paper Series 07-22, 2007

18. Europeadia. The Common Customs Tariff of the EU (available online: http://europedia.moussis.eu/books/Book_2/3/5/2/1/index.tkl?all=1) 19. Fagerberg J., “Technology and International Differences in Growth

Rates”, Journal of Economic Literature, Volume 32, 1994

20. Frankel J.A., Romer D., “Does Trade Cause Growth?”, The American Economic Review, June 1999

21. Grosskopf S., Self S., “Factor Accumulation or TFP? A Reassessment of Growth in Southeast Asia”, Pacific Economic Review, 11: 1 2006

22. Grossman G.M., Helpman E., “Endogenous Innovation in the Theory of Growth”, The Journal of Economic Perspectives, Vol. 8, Issue 1, 1994 23. Gehringer A., Martinez-Zarzoso I., Nowak-Lehman Danziger F., “The

Determinand of Total Factor Productivity in the EU: Insights from Sectoral Data and Common Dynamic Processes”, available online: https://ideas.repec.org/p/ekd/004912/5343.html

38

http://www.bankofengland.co.uk/publications/Documents/quarterlybullet 25. Holmes, R. M., Miller, T. Hitt, M. A. and M. P. Salmador “The interrelationships among informal institutions, formal institutions and inward foreign direct investment”, Journal of Management, 2011

26. Iacovone L., De Hoyos R. E., “Economic Performance under NAFTA: A Firm-Level Analysis of the Trade-Productivity Linkages”, The World Bank Policy Research Working Paper 5661, 2011

27. Inklaar R., O’Mahony M., Timmer M., “ICT and Europe’s Productivity Performance Industry-level growth account comparions with the United States”, Review of Income and Wealth, Vol. 51., Issue 4, 2003

28. Isaksson A., “Determinants of total factor productivity: a literature review”, Research and Statistics Branch, United Nations Industrial Development Organization, 2007

29. Jajri I., “Determinants of Total Factor Productivity Growth in Malaysis”, Journal of Economic Cooperation, Volume 28, Number 3, 2007

30. Khan S. U., “Macro determinants of total factor productivity in Pakistan”, SPB Research Bulletin, Volume 2, Number 2 , State Bank of Pakistan, 2006

31. Madsen J.B., “Technology spillover through trade and TFP convergence: 135 years of evidence for the OECD countries”, Journal of International Economics, 72, 2007

32. McQuinn K., Whelan K., “Europe’s Growth Crisis”, preliminary draft: http://www.cepr.org/sites/default/files/WHELAN%20-%20McQuinn-Whelan-Modena.pdf

33. Melitz M., “The Impact of Trade on Intra-industry Reallocations and Aggregate Industry Productivity”, Econometrica, Volume 71, Number 6, 2003

34. Miller S. M., Upadhyay M. P., “The Effects of Openness, Trade Orientation, and Human Capital on Total Factor Productivity”, Journal of Development Economics, Vol 63., 2000

39

36. Nadiri M.I., Kim S., “International R&D Spillovers, Trade and Porductivity in Major OECD Countries”, NBER Working Paper Series 37. Panagariya A., “Preferential Trade Liberalization: The Traditional

Theory and New Development”, University of Maryland, 1999

38. Schiff M., Wang Y., “Regional Integration and Technology Diffusion: The Case of the North America Free Trade Agreement”, Policy Research Working Paper 3132, The World Bank, 2003

39. Schreyer P., “The Contribution of Information and Communication Technology to Output Growth”, OECD Science, Technology and Industry Working Papers STI Working Paper No.2, 2000

40. Segerstrom P. S., Sugita Y., “The Impact of Trade Liberalization on Industrial Productivity”, Journal of the European Economic Association, Volume 13, Number 6, 2015

41. Shiff M., Winters L. A., “Regional Integration and Development”, The International Bank for Reconstruction and Development, The World Bank, 2003

42. Solow R. M., “Technical Change and the Aggregate Production Function”, The Review of Economics and Statistics, Vol. 39, No. 3., 1957

43. The Conference Board, “Productivity Brief 2015”,

https://www.conference-board.org/retrievefile.cfm?filename=The-Conference-Board-2015-Productivity-Brief.pdf&type=subsite

44. The Conference Board Total Economy Database, available online:

https://www.conference-board.org/data/economydatabase/index.cfm?id=27762

45. Timmer M.P., Inklaar R., O;Mahony M., van Ark B., “Productivity and Economic Growth in Europe: A Comparative Industry Perspective”, Internationa Productivity Monitor, Springer, Issue 21, 2011

46. Ugur M., Nandini D., “Corrupiton and Economic Growth: A Meta-analysis of the Evidence on Low-income countries and beyond”, MPRA Paper No.31226, 2011

40

GDP, GERD per capita and GERD per researcher, available online: http://data.uis.unesco.org/Index.aspx

48. Van Ark B., Chen V., Colijn B., Jaeger K., Overmeer W., Timmer M., “Recent Changes in Europe’s Competitive Landscape and Medium-Term Perspectives: How the Sources of Demand and Supply Are Shaping Up”, European Commission, European Economy, Economic Papers 485, 2013 49. Van Ark B., “Total Factor Productivity – Lessons from the Past and

Directions for the Future”, 2014, available online: https://www.nbb.be/doc/ts/enterprise/speeches/presentations/presentat 50. World Input Output Database Socio Economic Accounts, available