Conference Proceedings
COOPERATION CHALLENGES AFTER THE EU
ACCESSION OF CROATIA
April 23, 2015, Opatija, Croatia
Edited by:
Andrej Kumar
Chair Jean Monnet, ECSA Slovenia
Katja Zajc Kejžar
University of Ljubljana, Faculty of Economics
Publisher: ECSA Slovenia
URL: http://konference.ef.uni-lj.si/ecsa/publications/
Ljubljana, September 2015
Co-funded by the Lifelong Learning programme of the European Union
The European Commission support for the production of this publication does not constitute an endorsement of
the contents which reflects the views only of the authors, and the Commission cannot be held responsible for any
use which may be made of the information contained therein.
CIP - Kataložni zapis o publikaciji
Narodna in univerzitetna knjižnica, Ljubljana
339.94(497.5)(0.034.2)
COOPERATION challenges after the EU accession of Croatia [Elektronski vir] :
conference proceedings, April 23, 2015, Opatija, Croatia / edited by Andrej Kumar, Katja
Zajc Kejžar. - El. knjiga. - Ljubljana : ECSA Slovenia, 2015
ISBN 978-961-281-965-1 (pdf)
1. Kumar, Andrej
281112320
Program Committee
Vinko Kandžija
University of Rijeka, Faculty of Economics, Croatia
Andrej Kumar
Chair Jean Monnet, ECSA Slovenia
Katja Zajc Kejžar
University of Ljubljana, Faculty of Economics
Organizing Committee
Marko Tomljanović
University of Rijeka, Faculty of Economics
Tadeja Žabkar
University of Ljubljana, Faculty of Economics
Dejan Guduraš
University of Ljubljana, Faculty of Economics
Reviewers
Andrej Kumar
Chair Jean Monnet, ECSA Slovenia
Katja Zajc Kejžar
University of Ljubljana, Faculty of Economics
Sonja Šlander Wostner
University of Ljubljana, Faculty of Economics
I
CONTENTS
PREFACE ................................................................................................................................................................. III
OPENING LECTURE ................................................................................................................................................. 1
Andrej Kumar
EU TRADE STRATEGY AND THE BALKANS ................................................................................................ 2
SECTION I: IMPLICATIONS OF EUROPEAN INTEGRATION FOR CROATIA: PRE-ACCESSION EVIDENCE AND EARLY
MEMBERSHIP EXPERIENCES ................................................................................................................................. 12
Marino Golob, Martin Golob, Tomislav Kandžija
CROATIAN INSURANCE MARKET OVERVIEW AFTER EU ACCESSION .................................................... 13
Valerija Botrić
INDUSTRY WAGE PREMIUM AND EU TRADE EFFECTS IN CROATIAN MANUFACTURING SECTOR ...... 26
Vinko Zaninović, Katja Zajc Kejžar
INTENSIVE AND EXTENSIVE MARGINS OF CROATIAN MANUFACTURING EXPORTS: EVIDENCE FROM
2000-2012 PERIOD .................................................................................................................................. 37
SECTION II: EU MEMBER STATES’ EXPERIENCES IN DIFFERENT POLICY AREAS .................................................. 48
Jiří Dušek, Lubomír Pána
THE USE OF PPP PROJECTS ON THE LEVEL OF STATES, REGIONS AND MUNICIPALITIES UNDER THE
CONDITIONS OF THE CENTRAL EUROPEAN REGION ............................................................................. 49
Jiří Dušek
THE PROBLEM AREA OF CZECH REPUBLIC'S USE OF EU STRUCTURAL FUNDS IN THE PROGRAMME
PERIOD OF 2007/2013 ............................................................................................................................ 56
Andrej Kumar, Sonja Šlander Wostner
EU COHESION POLICY AND ABSORPTION IN SLOVENIA ........................................................................ 65
Sonja Šlander Wostner, Tej Gonza, Katja Zajc Kejžar
EVALUATION OF EU COHESION POLICY: LESSONS FROM SLOVENIAN CASE ........................................ 84
Ernest Dautović, Lucia Orszaghova, Wilem Schudel
CONVERGING IN DIVERGENT WAYS: EXPLAINING TRADE INTEGRATION BETWEEN CESEE COUNTRIES
AND THE EU-15 ..................................................................................................................................... 103
SECTION III: THE CHALLENGES OF OTHER WBCs IN THE PROCESS OF EUROPEAN INTEGRATION ................... 126
Branka Topić-Pavković
FISCAL AND MONETARY ASPECTS OF ACCESSION OF BOSNIA AND HERZEGOVINA TO THE MONETARY
UNION ................................................................................................................................................... 127
II
Preface
After the accession of Croatia to the EU in July 2013 new opportunities have been created for more
active co-operation among researchers, institutions and firms from Slovenia and Croatia and as wel
from other Western Balkan and EU countries. As part of its Jean Monnet project ECSA Slovenia
organized the international conference on Cooperation chal enges after the EU accession of Croatia,
on April 23, 2015 in Opatija, Croatia. With the goal of promoting further cooperation in the field of
research, institutions and business the conference aimed at (i) evaluating economic, political, legal
and other implications of the Croatian accession to the EU, (ii) identifying changes in business
environment for Croatian firms after joining the EU internal market and (iii) facilitating the transfer of
experience of “older” EU member states to the to the Croatian academic, institutional and business
community.
This volume presents the proceedings of this conference with nine peer reviewed papers and notes
on the opening lecture by ECSA Slovenia President, Professor Andrej Kumar on EU trade strategy and
the Balkans. Papers in this volume address the impacts, challenges and opportunities for the
cooperation among Croatia, other EU member states and countries of Western Balkan region in the
context of the European integration processes. They have been divided into three sections.
Papers in the first section deal with different aspects of pre-accession economic development of
Croatia and the early evidence on the impacts of the Croatian membership in the EU. The section
starts with the industry perspective where Golob, Golob and Kandžija analyse the development of
the Croatian insurance market after the EU accession. The paper finds that real liberalisation only
started in 2013 after the EU accession and resulted in increased competition pressure, leading to,
among others, overall decline of gross written premiums. Botrić estimates the wage equation for
Croatian manufacturing sector in the 10 year period and confirms the industry-specific wage
premiums pointing towards rigidities in labour market and low inter-sectorial labour mobility. While
Zaninović and Zajc Kejžar analyse trade patterns of Croatia in the recent decade and decompose
them into intensive and extensive margins. The gravity model results show that Stabilization and
Association Agreement and the introduction of the diagonal cumulation of rules of origin significantly
affect intensive export margin (average firm exports across industry divisions), specially exports of
consumption goods, while the global trade col apse affected negatively both intensive and extensive
trade margin (number of firms and as well average firm exports).
The second group of papers review and evaluate experiences of selected EU member states in
different policy areas. Dušek and Pána compare approaches to the projects based on partnership
between private and public sector (PPP) on various levels (state, regional, municipalities) between
Czech Republic and Austria with the suggestion to improve practises in such projects. A series of
three papers provides lessons related to the EU cohesion policy. Dušek identifies problematic aspects
of the use of structural funds in the programme period 2007/2013, while Kumar and Šlander Wostner
expose the importance of the absorption capacity for successful use of the EU cohesion funds for the
regional and national economic growth improvements. Šlander Wostner, Gonza and Zajc Kejžar focus
on the Slovenian experiences with triangulation-based Cohesion policy evaluation process. Further,
III
Dautović, Orszaghova and Schudel assesse real, nominal and institutional determinants of intra-
industry trade between EU 15 and new EU member states, EU candidates and potential candidates
showing that even though determinants for new EU member states deviate considerably from those
of candidate and potential candidate countries, there exist common factors promoting intra-industry
trade across the CESEE region, such as the corporate tax competitiveness, the flexibility of exchange
rate regimes and lower levels of corruption.
The third section addresses the challenges of other WBCs in the process of European integration.
Topić-Pavković analyses expected fiscal and monetary implications of joining the monetary union for
BiH and discusses necessary reforms before entering the E(M)U. She suggests that a rational solution
for BiH, after joining the EU, would be gradual process of monetary integration with stable monetary
policy, effective management of public finances and careful management of public debt.
This conference would not have been possible without the support and assistance of the University
of Rijeka, Faculty of Economics, especially professor Vinko Kandžija and his group of researchers. We
would also like to thank Sonja Šlander Wostner for her help in referring the papers. We are very
grateful to Dejan Guduraš and Tadeja Žabkar for editorial assistance.
We hope these proceedings bring important and useful new insights, evidence and policy
implications needed to foster future cooperation among the countries in this region and their
economic success within the context of European integration processes.
Andrej KUMAR
Katja ZAJC KEJŽAR
IV
OPENING LECTURE
EU TRADE STRATEGY AND THE BALKANS
1
International conference Opatija, April 2015
ECSA Slovenia; „COOPERATION CHALLENGES AFTER THE EU ACCESSION
OF CROATIA“
EU TRADE STRATEGY AND THE BALKANS
Prof. dr. Andrej Kumar
Chair Jean Monnet
The discussion TOPICS
• The discussion conceptually connects two issues:
- The First Issue; The perspectives of the Western Balkan States (WBSs) to
successfully conclude the accession process to the EU soon.
- The second issue; The characteristics of the EU Trade Strategy and Policy and
its recent specifics,
Both issues together create the following QUESTION:
Does the contemporary EU‘s Trade Strategy reduces the actual dynamics and
perspectives of the WBSs to successfully realize the EU accession process soon?
2
The structure of the presentation
A. WBSs specifics and relations with the EU; WBSs changing membership, the EU
attitude towards WBSs accession, the accession process, WBSs‘ challenges and
obstacles.
B. EU Trade Strategy concepts and orientation; Europe 2020, general EU Trade
Strategy orientation, Strategy implementation and its contemporary focus, the EU
trade interests towards WBCs.
C. Evaluating the impacts of the EU Trade Strategy on the WBSs accession
perspectives.
The EU and its relations with the WBSs
•
The EU declared its willingness and determination in accepting all states of the
Balkan region into the EU. Such EU determination was expressed in the EU-
Western Balkans Summit Declaration (http://europa.eu/rapid/press- release_PRES-03-163_en.htm ) agreed in Thessaloniki on June 21, 2003.
•
The Thessaloniki Declaration was based on decision taken by the European
Council “recalling its conclusions from Copenhagen (December 2002) and
Brussels (March 2003), repeated its determination to fully and effectively support
the European perspective of the Western Balkan Countries, which will become
an integral part of the EU, once they meet the established criteria.
3
Who are the WBSs in 2015?
• When in 2003 Thessaloniki Declaration was accepted none of the states from the
Balkans was part of the EU.
• The region got its geographic name from the Balkan mountains. Balkan
Peninsula is another name for this region of the South East Europe (see:
http://en.wikipedia.org/wiki/Balkans )
• The combination of geographic and political aspects created the EU used
regional name. In 1999 the EU initiative was called the Stability Pact for South
Eastern Europe. The EU defined the "Western Balkans" as the south-east
European area that include states not being the EU members yet.
• Croatia as the 2013 EU member isn’t any more the WB state. In 2015 the
WBSs are by the EU definition; Serbia, B&H, Montenegro, Macedonia,
Albania and Kosovo.
The EU Trade with WBSs*; 2003-2013
•
EEC/EU member states‘ trade has
always been the major driving
force for their improved economic
growth and for new jobs creation.
WBSs accession to the EU is
similarly DEPENDING on the
trade growth potential.
•
EU trade with WBSs has been
growing but is not large – only
1%** of the EU external trade in
2014.
•
After 2013 the WBSs trade
potential for the EU was further
reduced when Croatia joined the
EU.
* in 2013 the data still include Croatia
* * see:
http://trade.ec.europa.eu/doclib/docs/2006/septe
mber/tradoc_122530.pdf
4
WBSs‘ trade potential and interests to join EU
•
Only Serbia has some, although
Client and Supplier Countries of the EU28
small, potential for larger trade
in Merchandise Trade (value %)
with EU. For the EU Albania, B&H,
(2014, excluding intra-EU trade
and Macedonia present highly limited
trade potentials. Kosovo and
as WBSs
Value of EU Value of EU Value of EU
Montenegro are even not „visible“ EU
total trade (%)
M (%)
X (%)
trade partners. Further
Serbia
0.5
0.4
0.6
•
Serbia might have reservations to
FYR Macedonia
0.2
0.2
0.2
speed up EU accession considering
B&H
0.2
0.2
0.3
presently growing positive impacts of
Albania
0.1
0.1
0.1
its trade agreement with Russia.
Montenegro
0.0
0.0
0.1
•
All WBSs might have similarly
Kosovo
0.0
0.0
0.0
certain reservations to join realizing
1.0
0.9
1.3
that by accession they will lose FTA
or other arrangements‘ advantages
Source: Author‘s calculation from;
that they presently have with non
http://trade.ec.europa.eu/doclib/docs
EU states.
/2006/september/tradoc_122530.pdf
WBSs‘ trade shares with CEFTA partners
as % of total trade in 2013;
http://www.unwe.bg/uploads/Alternatives/9_Moraliyska.pdf
•
For WBSs trade with CEFTA is rather substantial. Data not include Serbia trade
with Russia or other trade of the WBSs with other preferential partners.
5
The EU and Global Trade
•
In 2015 the European Union (EU) seen as a SINGLE ENTITY is still the largest
player in global trade and investments.
•
As economic, and even more complex integration of 28 states, the EU’s impact
on global trade and investments is realized through the economic size and
specifics of each EU member state, and by their ability to create and implement
common external trade policy.
•
TRADE POLICY IS A CORE COMPONENT OF THE EU’S 2020
STRATEGY(see;COM(2010)612)
EU Trade Strategy orientation & specifics
• The EU‘ general development strategy „Europe 2020“ is based on the triple
objectives;
- smart,
- inclusive and
- sustainable growth of all EU member states.
•
The external dimension of the „Europe 2020“ strategy specify how trade and
investment policies should support the three Strategy‘s objectives realization.
•
The EU Trade Strategy sets the economic framework of the EU interests
towards the WBSs accession process.
6
Why Trade is so vital for the EU development strategy realization?
•
The theory suggests; more trade creates additional economic growth, more jobs,
better standard of living and better market choices. The dilemma is whether that
is always true? The trade growth combined with national growth into the
misery is one of such potential problems.
•
The entire EU history is about enhancing trade growth. First among the
member states; starting from customs union to the internal market and economic
and monetary union. The internal trade growth concept is continuously
implemented by the EU enlargement and deepening processes.
•
From 1968 the EU common Trade Policy has been focused on increasing
EXTERNAL trade based on market opening with required reciprocity. In the
specific cases - G 77, ACP countries, or to be potential EU member countries - the
EU market opening could be asymmetric in time or in functional specifics.
The EU external trade some facts
•
The EU estimated (COM(2010)612) that about 90 % of world economic growth will
be generated outside Europe already in 2015 and after, with one third from
China alone.
•
In the years to come the EU intend to seize the opportunity of higher levels of
economic growth abroad, especially in East and South Asia.
•
The EU envisage (COM(2010)612) that the developing and emerging states are
likely to account close to 60 % of world GDP by 2030, compared to less than 50
% today.
7
EU trade expansion to the third markets
http://trade.ec.europa.eu/doclib/docs/2012/november/tradoc_150129.pdf
•
The key priority for the Eu is to open up more market opportunities for
European business by negotiating new FTAs or other agreements with the key
countries. If EU complete all its current free trade talks tomorrow, (see next slide)
it could add 2.2% to the EU's GDP or €275 billion. This is equivalent of adding a
country as big as Austria or Denmark to the EU economy.
•
In terms of employment, these agreements could generate 2.2 million new jobs or
additional 1 % of the EU total workforce. At present 14% of EU workforce
directly depend on trade with the third countries.
•
Realistically the new jobs vreated by the EU oppening UNFORTUNATELY
will not be evenly distributed among the all 28 EU member states.
EU‘s future FTAs and expected – 2014
http://trade.ec.europa.eu/doclib/docs/2012/june/tradoc_149622.jpg
8
EU external Trade Strategy and the WBSs
•
The WBSs are included into the EU external trade strategy based on asymmetric
trade openning defined by the Stabilization Association Partnership
agreements (http://eur-lex.europa.eu/legal-
content/EN/TXT/HTML/?uri=URISERV:r18008&rid=1 and http://eur-lex.europa.eu/legal-
content/EN/TXT/?uri=uriserv:r18003 )
•
The present focus of the EU external trade strategy on negotiating different global
FTAs including among others the one with USA (TTIP), and Canada (CETA),
reduces the ability and interest of the EU to create new and enhancing trade
and investment opportunities for the WBSs.
EU and WBSs‘ accession future
• Based on EU Trade Strategy
(see; http://eur-
lex.europa.eu/summary/chapter/external_trade.html?root_default=SUM_1_CODED=07), the WBSs are not in the main focus of the EU Trade Strategy, neither they are among
the top strategic development goals of the EU up to 2020,
• Such facts limit the EU future WBSs‘ support for faster realization of the
market efficiency and of the other accession criteria requirements.
• The WBSs accession to the EU, based on the EU strategic trade orientation, is
in fact moved into the foggy distant future.
9
Concluding remarks
• At present the EU doesn't shows any specific eagerness to enlarge
towards the WBSs.
• The limited EU enlargement interest towards the WBSs is caused by the
EU‘s Foreign trade strategy orientation and by WBSs‘ limited trade
potential.
• EU Trade strategy is focused on East and South Asia and USA with
Canada. The WBSs are not an evident part of the EU trade strategy.
• Among the WBSs are substantial objective differences and in some cases
eventual national political or economic reservations towards their
interests and abilities to join the EU by fulfilling the accession criteria.
Concluding remarks (2)
•
WBSs have different reasons for slow realization of the accession process. The
WBSs limitations are:
-
Internal difficulties to speed up the fulfilling of the accession criteria - case of
B&H,
-
The limits in overcoming the external obstacles to proceed with negotiations -
case of FYR Macedonia,
-
Specific internal and external impacts - case of Serbia,
-
Limited economic potential relative to the EU – case of Montenegro
-
Not entire EU confirmed national sovereignty – case of Kosovo
•
The EU economic and political tensions with Russia lead to the new challenges for
Serbia on the bases of the Serbia-Russia customs union treaty. Similarly
Montenegro is strongly influenced by Russia FDI flows.
10
Concluding remarks (3)
•
WBSs general dilemma is about the actual content and type of the EU fictional
framework that they will join in the future. The varies are enhanced by the size and
dynamic of changes in the EU external trade environment that will occur before
their membership. WBs are not part of the ongoing EU external trade environment
change but will have to join and accept all impacts of its enlarged trade openness.
•
The EU trade strategy to open its market to Worldwide competition, creates
potential dangers for WBSs . As transitional states, with limited economic support
from the EU, they might enter the EU with low level of economic ability to resist
the increased global competition on national markets
•
The experiences gained after the past EU enlargements by the transitional
countries ( from 2004 to today 11 of 28 members) are at least partially
documenting different negative national economic impacts created after entering
the strongly open EU market.
•
When the WBSs will join the EU its global market openness will be much higher
and more complex as it is in 2015. That expectedly leads to substantially larger
development problems for the new EU members states from the WB region in the
future.
11
SECTION I
IMPLICATIONS OF EUROPEAN INTEGRATION FOR CROATIA:
PRE-ACCESSION EVIDENCE AND EARLY MEMBERSHIP EXPERIENCES
12
Marino Golob
Colegium Fluminense Polytechnic of Rijeka, Rijeka, Croatia
Martin Golob
Mara Mara d.o.o., Pazin, Croatia
Tomislav Kandžija
Primorsko−goranska County, Rijeka, Croatia
CROATIAN INSURANCE MARKET OVERVIEW AFTER EU ACCESSION
ABSTRACT
The aim of this paper is to examine Croatian insurance market after The Republic of Croatia EU accession in
July of 2013. Insurance has evolved as a process of safeguarding the interest of people from uncertainty and can
be described as a social device to reduce or eliminate risk of loss to life and property. Insurance industry
contributes to the general economic growth of the society, provides safety and security that reduces uncertainties
in business and human life, generates financial resources, encourages savings, etc. Thus far, it is safe to say that
the insurance industry is vital to any economy. In the past, the insurance market in The Republic of Croatia was
characterized with state-owned monopoly that only slightly changed during the last decades and after the EU
accession and the market liberalization, market conditions are changing rapidly every day. The aim of this paper
is to give an overview of the main key indicators on the Croatian insurance market, including the amount of
premiums, the scale of investment and the essential social and economic role the insurance market operators
play on personal and business risk coverage on the Croatian market; but more importantly to give an overview
of the market liberalization effects in the past year and a half after The Republic of Croatia EU accession and a
perspective for the future.
JEL classification: G22
Keywords: liberalization, insurance market, Croatia, effects, EU accession
1. INTRODUCTION
Insurance companies are important participants of the financial markets and represent an important factor of
economic development of each country. The primary function of insurance is to provide security to individuals
from the dangers of an uncertain future. In economic terms, insurance is an instrument which an individual uses
to pay a relatively small amount of insurance premium to gain an "upper hand" in case of a relatively large and
uncertain financial loss that would be possible if there were no insurance present to protect this individual from
his loss. Insurance industry has its own characteristics; Insurance is based on the Law of Probability, the Law of
Large Numbers and the Dispersion of risks. Insurance business process begins with sales and the conclusion of
insurance contracts. Competitive advantages in the Insurance industry are achieved through greater
specialization of offers, in creating new and improving existing insurance services, in providing wider choice in
the selection of coverage, in the use of new sales channels, in managing a consistent business policy as well as
creating a positive self-image with the use of wide spectrum of promotional activities (The Geneva Association,
2012). It is a known fact that European insurance law advocates a free market competition in all areas. All
European Union Member States must adapt their legislation in this area and strongly comply with all relevant
laws of The European Union. This adjustment is done during the process of negotiations and the process of
13
adjusting the regulations of each state with EU regulations. The Republic of Croatia has gone through some of
those adjustments in the years prior to EU accession and the Financial and Insurance services sector is now still
under the influence of the global financial crisis. Combination of unstable economic conditions and rapid
changes in the competitive environment due to EU accession are forcing some companies to face a very
challenging future. The main aim of this paper is to give an overview of the Croatian Insurance Market in the
past year and a half after EU accession and a perspective for the future.
2. MARKET OVERVIEW
The Croatian insurance industry in comparison with other countries of the European Union shows visible signs
of an industry still in development. This is firstly visible in the basic division of insurance on Life and Non-life
in The Republic of Croatia, whereas in the developed European markets, Life insurance has an approximate
share of 60%. The most dominant countries in the case of Life insurance are the Scandinavian countries, as well
as countries that are carriers of the European industry, firstly United Kingdom, France and Italy, with the
exception of Germany where the relationship is much more balanced. In the lower part, the share of Life
insurance around 30 percent, are developing countries such as Romania, Bulgaria, Croatia, Slovenia and the
Baltic countries. Croatian insurance industry can be described as small and emerging with a high potential for
further growth and development in the future (Filipović, 2014).
The Republic of Croatia, as an EU member, has a harmonized national insurance regulation according with the
EU insurance directives and its industry shareholders strongly adhere to the international insurance standards and
core principles. Foreign ownership of insurance companies in The Republic of Croatia is still dominant and a
year and a half after Croatia's EU accession insurance companies are more than forced to constantly innovate and
design new insurance products in a market that is clearly getting more competitive with every passing year. The
importance of insurance industry in The Republic of Croatia can be drawn from the share of total assets of
financial institutions. The share was 6.49% in 2013. Commercial banks occupied a high share of 73.95% in the
same year and mandatory pension funds took up 10.68%. The structure of financial institutions hasn't changed
largely regarding previous years. The insurance industry's share rose from 5.92% in 2008. to 6.41% in 2012
(HUO, 2014).
Table 1. Number of insurance and reinsurance companies
Type of Insurance
2008
2009
2010
2011
2012
2013
2014
Life insurance
8
8
6
7
7
7
7
Non - Life insurance
9
10
10
10
10
10
10
Composite
10
10
10
10
10
9
8
Reinsurance
2
2
1
1
1
1
1
Total
29
30
27
28
28
27
26
Source: HANFA, 2015; HANFA, 2014
There were 26 insurance companies operating in The Republic of Croatia in 2014. There was only 1 company
providing reinsurance services while 10 companies engaged in Non-Life insurance services. Number of
companies providing only Life insurance declined by 1 and in 2014. there were 7 Life insurance companies.
Composite insurance companies provided Life and Non-Life services and there were 8 of them operating the
Croatian market in 2014. The overall number of business entities declined in the observed period from 29 to 26.
Croatian insurance industry in the past two years has undergone a significant restructuring in the market. The
largest company in the industry went from state owned to private ownership and there were several acquisitions
of smaller companies.
14
Table 2. Gross Written Premium in 000 HRK
Non-Life
Year
Life Insurance
% of Total
% of Total
Total
Index
Insurance
2003
1.349.981
22,25
4.717.061
77,75
6.067.042
108,8
2004
1.569.421
23,68
5.057.446
76,32
6.626.867
109,2
2005
1.895.769
25,79
5.454.305
74,21
7.350.074
110,9
2006
2.165.061
26,47
6.015.094
73,53
8.180.156
111,3
2007
2.482.743
27,39
6.582.189
72,61
9.064.932
110,8
2008
2.545.775
26,28
7.140.327
73,72
9.686.102
106,9
2009
2.488.675
26,44
6.922.661
73,56
9.411.336
97,2
2010
2.457.683
26,58
6.787.860
73,42
9.245.543
98,2
2011
2.431.268
26,59
6.713.977
73,41
9.145.245
98,9
2012
2.461.154
27,23
6.577.321
72,77
9.038.475
98,8
2013
2.538.414
27,97
6.538.186
72,03
9.076.600
100,4
Source: HUO, 2014
Gross written premium was showing strong and consistent growth starting from 2003. and up to 2008. when
premium started to decline due to the current global economic developments. In 2013., after four years of
negative growth rates, total premium recorded a mild positive growth compared to 2012. Total premium in 2013
amounted to 9,08 billion HRK. This stopped the decline in insurance premiums, which from 2009 to 2012
ranged between -2,8% and -1,1%. As previously stated, share of Life insurance premium in total written
premium in the observed period was ranging from 22% to 28%. The share has been showing positive trends
throughout the period which indicates a slow but consistent direction of Croatian life insurance segment to life
insurance segments existing in more developed insurance markets. Non-life insurance has dominated the
Croatian insurance market from its beginnings and it remains so to this day. Still, the share of Non-life insurance
premium has been showing a reverse trend from Life insurance premium and from 2003., when the share was
77,75%, it decreased to 72,03% in 2013. On 1st of July 2013. The Republic of Croatia accessed the EU and
further liberalization of the insurance market could not be stopped. It took some time for insurance companies to
adapt and prepare for a now truly free market. By the end of 2013. two insurance companies applied their own
commercial tariffs and soon every insurance market operator had to follow. The next table compares Life, Non-
life and total premiums in 2013. and 2014 (Svijet osiguranja, 3/2015).
Table 3. Grow Written Premium in 2014 in 000 HRK
Gross Written
Gross Written
INSURANCE
%
%
Change 14/13
Premium in kn
Premium in kn
I.-XII./2013
2013
I.-XII./2014
2014
Aps.(HRK)
Relat.(%)
NON-LIFE INSURANCE
6.538.186.057 72,03
5.923.573.258 69,19
- 614.612.799
-9,4
LIFE INSURANCE
2.538.414.004 27,97
2.637.784.389 30,81
99.370.385 3,91
TOTAL
9.076.600.061 100
8.561.357.647 100
- 515.242.414 -5,68
Source: HUO, 2014
As evidenced in the table, Non-life premium declined from 6,53 billion HRK in 2013. to 5,93 billion HRK in
2014. which makes a staggering decline of 614 million HRK or 9,4%. This certainly was a significant impact for
the insurance industry and is a direct result of lowering the compulsory motor liability premiums during the last
15
year and a half. This negative impact was somewhat mitigated with the rise of Life insurance premium for 99
million HRK or 3,91%. The overall written premium a year and a half after the start of the real liberalization in
July of 2013. is characterised with a decline of 5,68% or 515 million HRK. The share of Non-life insurance is
further declining in 2014, from 72,03% to 69,19% thus marking the point of the lowest market share for Non-life
insurance premium in the observed period. On the other hand, Life insurance premium is rising to a highest share
recorded of 30,81% in 2014. Insurance companies continue to take measures to stabilize the total portfolio of
Life insurance through regular activities concerning the collection of due premium, reducing the number of buy-
outs along with the possibility of changes (the amount of premium and life insurance duration) or giving loans to
clients with favourable interest rates. Gross amount of settled claims in 2013. amounted to 4.68 billion HRK
with a growth rate of 1% compared to 2012. The rate of growth was affected by the high growth rate of Life
insurance settled claims of 10.1%, while Non-life insurance settled claims growth rate had decreased by -3.1%.
The growth rate of Life insurance settled claims was always higher than Non-life claims in all of the observed
years except for the 2008. This growth rates can be attributed to the expiration of Life policies made in large
numbers during 1995. after the stabilization of Croatian currency when a faster growth of Life insurance
premium had started (HUO, 2014).
Table 4. Settled Claims Gross Amount in 000 HRK
Year
Life Insurance
Non-Life Insurance
Total
Index
2003
173.422
2.791.330
2.964.752
108,1
2004
259.748
2.951.202
3.210.950
108,3
2005
315.131
3.139.855
3.454.986
107,6
2006
421.048
3.510.062
3.931.110
113,8
2007
636.639
3.634.697
4.271.336
108,7
2008
682.594
3.909.271
4.591.865
107,5
2009
931.253
3.849.595
4.780.848
104,1
2010
1.038.460
3.357.310
4.395.770
91,9
2011
1.298.669
3.269.099
4.567.768
103,9
2012
1.420.631
3.214.206
4.634.837
101,5
2013
1.564.285
3.115.561
4.679.846
101,0
Source: HUO, 2014
The trend evidenced in premium data can be observed as well in gross amount of settled claims. Non-life claims
declined 7,49% regarding the previous year (233 million HRK) and Life insurance premium declined 2,28% (35
million HRK) which totals of 5,75% of overall decline in settled claims (268 million HRK).
Table 5. Settled Claims Gross Amount in 2014 In 000 HRK
Gross Claims
Gross Claims
INSURANCE
%
%
Change 14/13
Paid in HRK
Paid in HRK
I.-XII./2013
2013
I.-XII./2014
2014
Aps.(kn)
Relat.(%)
NON-LIFE INSUR.
3.115.890.824
66,58
2.882.571.493
65,35
- 233.319.331
-7,49
LIFE INSURANCE
1.564.284.852
33,42
1.528.664.126
34,65
- 35.620.726
-2,28
TOTAL
4.680.175.676
100
4.411.235.620
100
- 268.940.057
-5,75
Source: HUO, 2014
Throughout the observed period, share of Life insurance premium in GDP was averaging from 0,60% to 0,78%.
Share of Non-life premium in GDP shows a higher range of percentage, but a declining trend from 2,06% in
16
2003. to 1,97% in 2011. and 2,00% in 2013. From Table 6. can be observed that in the years of GDP growth the
share of insurance premiums in GDP followed that growth, in the years of the global economic crisis the share of
premium was showing a declining trend alongside with the declining GDP.
Although the crisis was mostly a banking crisis, insurance companies in The Republic of Croatia were not
directly threatened and remained fairly solvent. The overall decline can be attributed directly to the reduced
investment portfolio, reduced economic activity and reduced purchasing power as a consequence of the crisis
(HUO,2014).
Table 6. Share of Gross written premium in GDP (%)
Life
Non-Life
Year
Insurance %
Insurance %
Total
of GDP
of GDP
2003
0.59
2.06
2.65
2004
0.63
2.04
2.67
2005
0.71
2.05
2.76
2006
0.74
2.07
2.81
2007
0.78
2.07
2.85
2008
0.74
2.07
2.81
2009
0.74
2.07
2.81
2010
0.73
2.03
2.76
2011
0.71
1.97
2.68
2012
0.75
1.99
2.74
2013
0.78
2.00
2.78
Source: HUO, 2014
Total premium per capita in 2013. amounted to 2.127 HRK, 1.532 HRK for Non-life insurance and only 595
HRK for Life insurance. Compared to the previous year there was a slight increase recorded but regarding this
indicator, The Republic of Croatia is far behind the average of developed countries. In 2012 an average of 1 843
EUR per capita was spent on insurance in European union`s full member countries. Of this insurance amount, 1
083 EUR was spent on life insurance and the remaining 760 EUR on non-life insurance, of which 190 EUR was
on health insurance. These figures were broadly stable compared to the previous year of 2011 (Insurance Europe,
2014).
Even when comparing The Republic of Croatia with neighbouring Slovenia, which has a smaller insurance
market, Croatia is still lagging. For reference, The Republic of Slovenia has an average insurance premium per
capita of 960 EUR in 2013. (268 EUR for Life insurance and 691 EUR for Non-life insurance) whereas Croatia
has an average of 279 EUR (Ivanušič, 2014).
17
Table 7. Premium per capita in HRK
Life
Non-Life
Total
Insurance Insurance
2003
304.00
1062.00
1366.00
2004
354.00
1139.00
1493.00
2005
427.00
1228.00
1655.00
2006
488.00
1335.00
1823.00
2007
560.00
1484.00
2044.00
2008
574.00
1610.00
2184.00
2009
562.00
1563.00
2125.00
2010
556.00
1534.00
2090.00
2011
552.00
1525.00
2077.00
2012
574.00
1535.00
2109.00
2013
595.00
1532.00
2127.00
Source: HUO, 2014
Premiums per employee grew from 2003. to 2005., in the period between 2006. and 2013. premiums per
employee were shaped by a constant rate of decline. The decline is caused by the rapid employment of
employees in insurance companies. At the moment there are 11,500 employees working in the insurance
industry. This declining trend also showcases the fact that recent use of different distribution channels, especially
internet, does not necessarily mean downsizing of employees in the industry. In The Republic of Croatia,
insurance is still mainly distributed internally, followed by agency and broker distribution channels (HUO,
2014).
Table 8. Premium per employee in HRK
Total
Insurance
Premium
Year
Industry
per
Employees
Employee
2003
6059,00
1001,00
2004
6485,00
1022,00
2005
6970,00
1055,00
2006
7984,00
1025,00
2007
9360,00
968,00
2008
10544,00
919,00
2009
11184,00
841,00
2010
11145,00
830,00
2011
11288,00
810,00
2012
11616,00
778,00
2013
11533,00
787,00
Source: HUO, 2014
In the analysis of business performance of insurance companies, indicators specific to this industry were used,
such as: claims ratio, costs ratio and combined ratio (HUO, 2014). Claims ratio, which is calculated as the ratio
of the sum of claims paid, changes in claims reserves and changes in other technical reserves and earned
premium (multiplied by 100), in 2013 amounted to 61.1% Costs ratio is calculated as the ratio of the sum of
operating expenses (reserves and administrative costs), other technical expenses and gross written premium
reduced by premium ceded to reinsurance (multiplied by 100) in 2013 amounted to a high 47.6%. Normal range
for the indicator within the insurance industry ranges between 20% to 30%. Combined ratio is calculated as the
sum of the claims ratio and costs ratio, and it shows operating results before inclusion of income from
investments, in 2013 is as high as 108,7%.
18
Table 9. Basic insurance indicators
Claims
Costs
Combined
Year
Ratio
Ratio
Ratio
2003
69.60
34.10
103.70
2004
71.00
36.10
107.10
2005
70.50
37.80
108.30
2006
70.80
39.40
110.30
2007
73.90
40.20
114.10
2008
68.40
39.40
107.80
2009
69.80
44.50
114.30
2010
67.60
44.30
111.90
2011
64.40
45.70
110.10
2012
63.00
46.40
109.40
2013
61.10
47.60
108.70
Source: HUO, 2014
Combined ratio is calculated as the sum of the claims ratio and costs ratio, and it shows operating results before
inclusion of income from investments, in 2013 is as high as 108,7%.
3. COMPULSORY MOTOR LIABILITY INSURANCE MARKET
Croatian compulsory motor liability insurance market has been chosen for a detailed analysis due to the
importance of this insurance segment in The Republic of Croatia and because the effects of the liberalization,
upon accessing the European Union, have been very visible from the start, given the short amount of time (only
year and a half) in which the effects could be observed.
Croatian compulsory motor liability market can historically be divided into several periods. The first period
lasted until the 1st of January 2008. when regarding the compulsory insurance segment there was, on a
regulatory level regulated by HANFA (Croatian Agency for Supervision of Financial services), an administrative
determination of the insurance conditions and tariff systems for all insurance companies operating the market.
Companies were required to obtain authorization from HANFA prior to the application of insurance conditions
and tariff system. The conditions and tariff system approved by HANFA were common and were used by all
companies on the market. HANFA had legal power to independently adopt binding common conditions and
tariff systems with unique functional bases of premiums, if such was neccessary based on the technical results of
the insurance companies. HANFA determined, after the given permission, even the day from which the
conditions and the tariff system was applicable. So, it can easily be concluded that the State owned agency used
the system of prior control of conditions and tariff systems, and the procedural approval of conditions and tariffs
was only a formality. The insurance market in The Republic of Croatia was administratively controlled up to
2008 (Ćurković, 2014).
Along with the legislative change, after 1st of January 2008., the market should have been fully liberalized. The
objective of reporting to supervisory body was not, like it was up to 2008., getting an approval for the change of
conditions and tariff systems, but only to enable the supervisory authority to check whether the conditions and
premiums were according to regulations, actuarial principles and other rules of the profession. Lack of
conditions and tariff systems transparency was still evident. All the companies operating the market actually
continued to use the same insurance conditions and tariff systems (HANFA approved) that were already used on
the market. There was no real competitiveness on the market and competition was reduced to a slightly
decreased expenses loading with (secretly, and this necessarily meant unlawful) offer of benefits to clients such
as free technical inspections, free gift certificates, gas vouchers and other. The role of the supervisory authorities
was thereon reduced to a relatively strict control of application of the bonus-malus system. Few insurance
19
companies (foreign owned) tried to apply their own new insurance conditions and tariff systems, but these efforts
ended unsuccessful as the supervisory body objected the aforementioned conditions and systems as being
inadequate due to being based on a insufficiently broad statistics base. The period from 2008. up until the
accession of The Republic of Croatia to the European Union could truly be called a quasi-liberalised market
(Ćurković, 2014).
Gross written premium of compulsory motor liability insurance during the observed period was always
maintaining a relatively steady share in Total gross written premium. The share ranged from 32,26% in 2003. to
its lowest share of 29,96% in 2006. The same share of compulsory motor liability insurance was 32,81% in
2013. Given the fact that Non-life insurance segment dominated the Croatian insurance market from its
beginnings, Compulsory motor liability insurance has and it still is an important segment of it as evidenced from
the Table 10. below. Share od compulsory motor liability insurance had a dominant and steady market share of
40% to 42% of Non-life insurance premium up to 2008. After 2008. a steady rise can be observed in the Table
10. Reaching up to 45,55% in 2013. Gross written premium amounted to 2.978.147.000 HRK in 2013. Reaching
its higher number so far.
Table 10. Gross Written Premium of Compulsory Liability Insurance for Motor Vehicles in 000 HRK
% of
% of Non
Total
Year Gross Written Premuim
Index
Life
Gross
Insurance
Written
Premium
2003
1.957.116
110,00 41,49 32,26
2004
2.111.470
107,90 41,75 31,86
2005
2.246.038
106,40 41,18 30,56
2006
2.450.936
109,10 40,75 29,96
2007
2.721.082
111,00 41,34 30,02
2008
2.922.728
107,40 40,93 30,17
2009
2.922.648
100,00 42,22 31,05
2010
2.890.062
98,90
42,58 31,26
2011
2.935.198
101,60 43,72 32,10
2012
2.939.904
100,20 44,70 32,53
2013
2.978.147
101,30 45,55 32,81
Source: HUO, 2014
Settled claims of Compulsory liability insurance reached its lowest share of 23,19% in 2013. thus trending a
steady decline in the years after 2008. as evidenced in Table 11.
20
Table 11. Settled Claims of Compulsory Liability Insurance for Motor Vehicles in 000 HRK
% of Non
% of
Life
Total
Year Settled Claims
Index
Insurance Settled
Claims
Claims
2003
1.286.947 109,80 46,11 43,41
2004
1.327.199 103,10 44,97 41,33
2005
1.385.872 104,40 44,14 40,11
2006
1.590.194 114,70 45,30 40,45
2007
1.581.392 99,40
43,51 37,02
2008
1.634.874 103,40 41,82 35,60
2009
1.422.808 87,00
36,96 29,76
2010
1.202.030 84,50
35,80 27,35
2011
1.195.476 99,50
36,57 26,17
2012
1.112.080 93,00
34,60 23,99
2013
1.085.247 97,60
34,83 23,19
Source: HUO, 2014
After 1st of July 2013., real liberalization and deregulation of the market could finally start. HANFA can now
only ask for a premium tariff system, technical and other elements of it, while prior it was an automatic
obligation of insurance company to deliver the conditions and tariff system for approval. Insurance companies
could now sell insurance based on their own insurance terms & conditions and tariff systems. Two insurers
started applying their own and new conditions and tariffs at the end of 2013. and every other insurance company
had to follow. New compulsory motor insurance premiums are now based on an „individualised” tariff. This
tariff is based on periods without damages/accidents, age of the insured, other family vehicles insured, the
existence of other types of insurance with the same insurer, vehicle mileage, bonus points given in conjunction
with banks and other enterprises. Along with new conditions & terms and an individualized tariff system came a
significant lowering of the compulsory motor liability insurance premium. The effect of it can be observed in the
following tables.
Table 12. Gross Written Premium of Compulsory Liability Insurance for Motor Vehicles in HRK
Gross
Gross
Written
Written
Change
RISK
%
%
Premium in
Premium in
14/13
HRK
HRK
I.-XII./2013
2013
I.-XII./2014
2014
Relat.(%)
Third Party
2949920371 99,38 2357112780 99,22
-20,1
Public
4814775
0,16
4853072
0,2
0,8
Transportation
Air Vessels
1582928
0,05
1471100
0,06
-7,1
Marine
12125030
0,41
12396938
0,52
2,2
Total
2968443104
100
2375833889
100
-20
Source: HUO, 2015
As evidenced in the table above, third party liability insurance comprises more than 99% of gross written
premium of compulsory motor liability insurance, which makes this segment of insurance market in the Republic
of Croatia very important for insurance companies. There has been only a slight change in market share of this
type of insurance, going from 99,38% in 2013. to 99,22% in 2014. Other types of compulsory insurance like
public transportation liability, air vessels liability and marine vehicles liability comprise only a smaller share,
21
and compared to 2013. these types show a slight rise in percentages regarding market share. Apart from marine
vehicles and public transportation liability which recorded a relative positive change, +2,2% and +0,8%
respectively; air vessels liability recorded a negative relative change of -7,1%. Main focus is on third party
liability insurance that recorded truly staggering -20,1% in 2014. regarding to 2013. This is directly connected
with transfer from administrative (common) to commercial (“individualised”) tariffs and the liberalization of the
insurance market. Average compulsory motor liability premium went from 1500 HRK in 2013. to 1196 HRK in
2014 (HUO, 2015). An overall decline of 20% is evidenced in this type of insurance.
Table 13. Settled Claims of Compulsory Liability Insurance for Motor Vehicles in HRK
Gross
Gross
Claims
Claims
Change
RISK
Settled in
Settled in
14/13
kn
kn
I.-
I.-XII./2013
Relat.(%)
XII./2014
Third Party
1065888100 984631420
-7,6
Public
436825
328962
-24,7
Transportation
Air Vessels
65729
1000
-98,5
Marine
1075190
264010
-75,4
Total
1067465844 985225393
-7,7
Source: HUO, 2015
Settled claims for the same type of insurance shows an overall decline of 7,7%. Public transportation liability
claims are declined for 24,7%, while air vessels show a big drop of 98,5%, marine vehicles liability recorded a
decline of 75,4%.
After 1st of July 2013. insurance companies had to increase the minimum of principal sum insured regarding
compulsory motor liability insurance. Minimum formerly in force was 3.500.000 HRK for persons (460.000 €)
and 1.500.000 HRK for property (200.000 €) (HUO, 2014). Current minimum amounts to 5.600.000 € for
persons and 1.120.000 € for property. It is an enormous one-time increase which was positive news for
consumers, but there is a possibility that some smaller insurance companies will bear some consequences in the
long run. New minimum of sum insured means increased outflow of domestic capital accumulation for
reinsurance mainly to foreign reinsurers. Another result of the liberalization are certainly new coverages and
commercial insurance products (riders) with compulsory insurance and certain other novelties:
• long-term period contracts/policies are now permitted,
• coverage of legal protection is included in the compuslory insurance,
• compulsory casualty insurance now covers 24hrs,
• new benefits for drivers regarding coverage,
• benefits for combined motor insurance (compulsory + motor hull),
• free road assistance is attached with the compulsory insurance,
• replacement vehicle coverage (is now cheap or free),
• loyalty bonus is given to consumers, as well as,
• family bonus (if more family members are insured with the same company),
• discounts for cash payments (enterprises are now included),
• lower premiums for certain kinds of vehicles (leasing, taxi, dangerous cargo transport)
• Bonus protection options (are now cheap or free).
22
All the above mentioned changes on the compulsory insurance market, along with the lower average compulsory
premium, introduction of new terms & conditions and tariff systems that differ from insurer to insurer, are a
direct effect of the liberalization. Combined with constantly rising competitiveness levels among insurance
companies and number of insurance companies that operate on the compulsory motor liability insurance market
make little room left for any new company to enter the market given that there are 15 insurance companies
providing such services for 1.884.000 motor vehicles in the Republic of Croatia (HANFA, 2014).
4. LEGISLATIVE AND REGULATORY OVERVIEW
The biggest possible obstacles for Croatian insurance companies definitively represent the possibilities of
difficulties in business operations regarding implementation of "Solvency II" framework. In May 2012. a
working group was formed among HANFA (Croatian Financial services Regulatory Agency), HUO (Croatian
Insurance Bureau) and HAD (Croatian Actuarial Association) to conduct a QIS Study (Qualitative Impact Study)
to gather market operators insight regarding implementation of "Solvency II" framework. Majority of
participants of the QIS study reported that they are not fully prepared for the implementation of the "Solvency II"
framework. According to data from the questionnaire, participants in majority felt that they don’t have all the
available resources and the implementation plan of the "Solvency II" framework has not yet been completed in
their companies. "Solvency II" implementation in the Republic of Croatia starts with 1st of January 2016
(HANFA, 2014).
New insurance law is currently being in development and will enter into force on 1st of January 2016. This new
legislative should improve the existing one and fully adjust it with the European insurance law. New insurance
law (NN 30/15) will enable insurance companies to sell investment fund shares and offer different retirement
programs to their clients. In addition, insurance companies will be able to represent business interest and sell
insurance products and services for other insurance industry companies. Adjustments will also include some new
prospects for insurance agencies. After 1st of January 2016, insurance agencies will be able to provide different
kinds of intellectual and technical services to their clients regarding insurance. Also, insurance agencies will be
able to sell investment fund shares and retirement programs (Gajski, 2014). Insurance agents will no longer have
to have 300 ECTS accompanied with a 3-year working experience, but 180 ECTS and a 3-year working
experience to provide intermediary services on their own (Gajski, 2015). New category of insurance agent
Assistant is being introduced with the implementation of the new law and assistants will be able to conduct a part
of insurance agentsòperations without the required license issued by HANFA (Gajski, 2014).
5. CONCLUSIONS
Before EU accession, despite legally declared and regulated liberalization and deregulation of the Insurance
Market (especially in the Compulsory Motor Liability Insurance), the expected liberalization was not achieved.
The real liberalisation of the insurance market started from 1st of July 2013. The role of HANFA is now, after
the accession to the European Union, reduced to sufficiency control of capital coverage regarding obligations of
each Insurer and insistence on transparency for additional benefits that are given to policyholders. Each insurer
can now operate the market with its own terms and conditions and tariff system. Resulted freedom of insurance
companies in designing their own tariffs and with no further obstacles regarding the implementation of
commercial tariffs directly led to create a significant overall decline in gross written premium (Total) as well as
Non-life written premium (mostly due to liberalisation of the compulsory motor liability insurance market).
Along with the commercialized tariffs, insurance companies started to discount compulsory motor liability
premium for 10%, 20, even 30% thus accumulating a bigger client base. Bigger client portfolio also means
bigger payments of claims, which could, in the long run, confront some smaller or capitally insufficient insurers
with serious operating difficulties (possible bankruptcy). Significant decline in premium for an insured
23
individual followed the before mentioned development (from average 1500 HRK in 2013. to 1196 HRK in
2014.). The accession also obliged insurers to increase Insured Sums (in Compulsory Motor Liability insurance)
and to provide equal premium for men and women in all types of insurance services and products. Final
adjustments are being prepared for law implementation to fully adjust Croatian insurance laws with the European
insurance laws and certain new provisions are being introduced that will largely advance and benefit insurance
companies and insurance intermediaries.
Further decline of Compulsory motor liability premiums can be expected, as well as an overall decline in
premium, at least for the foreseeable future. Possible disappearance of insurers that are less capitally secured is
to be expected to some extent, but eventual bankruptcy of certain insurers still cannot jeopardize the insurance
market due to the Guaranty fund. Mergers & Acquisitions of smaller insurers had already occurred on the market
and similar development can be expected in the future. Some difficulties are expected for insurers regarding the
implementation of “Solvency II” framework. Further increase of competitiveness is eminent, which will lead to
further development of new and innovative insurance products, especially in health and life insurance segment
which is considered as a market for further progress within the insurance industry. Re-designing of existing
insurance services and products is currently an on-going process on the Croatian insurance market. In the long
period, new technology risks will inevitably produce new insurance coverage that will be offered on the market.
All the above mentioned development will certainly force greater segmentation of insurance products.
Croatian economy is still feeling the effects of the financial crisis and it will take more time to recover to the
level of economy which will have a significant impact on the further growth of written premiums. Because of the
overall decline in premiums, insurers will try to improve their business results by lowering claims handling
costs, they will try to enhance detecting and preventing of frauds, rationalize internal costs, which will in the
long run have a positive effect on most of the Non-life insurance sector.
Greater use of information technologies and internet by insurance companies is to be expected. Social networks
and internet distribution will certainly be an important asset in improving insurance companies` business results.
Financial literacy and education is an important issue in the European Union. The European insurance sector
recognizes the importance of financial education of consumers and strives toward awareness by supplying
simple and user-friendly access to information that will equip them with basic knowledge about finance.
Croatian insurance regulatory body, as well as other stakeholders on the market are hosting public events,
issuing publications and brochures, conducting and publishing research and other studies and surveys, consulting
consumer services, media activities & campaigns and similar activities, but further efforts will be neccessary to
successfully educate wider Croatian public on matters of insurance.
REFERENCES
Ćurković, M., (2014), “Liberalizacija tržišta obveznog osiguranja od automobilske odgovornosti – hrvatsko iskustvo“, 25. Susret osiguravača i reosiguravača 2014., Sarajevo
Filipović, H., (2014), “Dohodovna elastičnost tržišne penetracije odabranih kategorija osiguranja”, Zbornik radova: Dani hrvatskog osiguranja 2014., Opatija
Gajski, Z., (2014), “Kako zaboraviti 2014. godinu”, Svijet osiguranja, 12/2014, Zagreb, p. 13, 14
Gajski, Z., (2014), “Milijarda manje nego 2008.”, Svijet osiguranja, 12/2014, Zagreb, p. 28, 29
Gajski, Z., (2015), “Prilagodba novom Zakonu već od 1. travnja”, Svijet osiguranja, 3/2015, Zagreb, p. 7
HANFA (2014), “Statistika osiguranja u Republici Hrvatskoj za 2014. godinu”, Zagreb
HANFA (2015), “Statistika osiguranja u Republici Hrvatskoj za 2013. godinu”, Zagreb
24
Hrvatski ured za osiguranje (2014), “Tržište osiguranja u Hrvatskoj”, Hrvatski ured za osiguranje, Zagreb
Insurance Europe, (2014), „Statistics: European Insurance in Figures“, Brussels
Ivanušič, Z., (2014), “Slovenija – pregled tržišta osiguranja u 2013. godini“, 25. Susret osiguravača i reosiguravača 2014., Sarajevo Letica, G., (2014), “ Tržište osiguranja u Republici Hrvatskoj godinu nakon ulaska u Europsku uniju", presentation, Dani hrvatskog osiguranja 2014., Opatija
The Geneva Association, (2012), “Social and Economic Value of Insurance”, Brussels
25
Valerija Botrić
The Institute of Economics, Zagreb, Zagreb, Croatia
INDUSTRY WAGE PREMIUM AND EU TRADE EFFECTS IN CROATIAN
MANUFACTURING SECTOR
ABSTRACT
Public debates and previous studies in Croatia emphasize different adjustment mechanisms in private and public
sector in terms of wage corrections during the recent economic downturn. The general conclusion is that the
public sector, mostly due to the collective bargaining procedures, enabled the employees to enjoy both relatively
more secure and better paid jobs. The aim of this paper is to investigate the parallel processes within
manufacturing sector, in particular the segment expected to compete on the international market. The initial
hypothesis is that two aspects have shaped the wage dynamics of manufacturing during the recent period – crisis
and EU integration. By relying on the Labour Force Survey (LFS) data, and restricting the analysis to the
manufacturing sector, we explore the development of the industry wage premium in the analysed segment of the
Croatian economy. Furthermore, the identified industry wage premiums are analysed with respect to the
international trade pressures indicators. Specifically we investigate whether the intra-industry trade with
European Union had impact on wages in Croatia’s tradable sector. In order to empirically investigate this
relationship, we match the Eurostat COMEXT with LFS data.
Key words: intra-industry trade, industry wage premium, Croatia, integration.
JEL classification: F14, F15, F16
1. INTRODUCTION
Croatia is a small open economy, recently under the dominance of two powerful external factors – global
economic crisis and EU accession process. The latter process entails complete liberalization of trade with EU
countries and expected successful integration of the domestic producers on the wider common market. The
process could also incur costs, which could manifest themselves on the labour market. As Brülhart and Elliot
(2002) explain, the size of the costs are assumed to be in line with smooth adjustment hypothesis, which states
that they will be lower if trade is mostly intra-industry in nature. So the trade with European Union and specific
pattern of trade play important role in the success of the integration process, but could also be significant for the
local labour market developments.
Public debates and previous studies in Croatia emphasize the different adjustment mechanisms in private and
public sector in terms of wage (and employment) corrections during the recent economic crisis. The general
conclusion is that the public sector, mostly due to the collective bargaining procedures, enabled the employees to
enjoy both relatively more secure and better paid jobs. The aim of this paper is to investigate the parallel
processes in manufacturing sector, in particular the segment expected to compete on the international market.
The initial hypothesis is that two aspects have shaped the wage dynamics of manufacturing during the recent
period – crisis and EU integration.
26
The integration process and its effects are dynamic in nature. To assess the overall impact of the integration
process on labour market adjustment would consequently require building and estimating a model in a dynamic
framework. Due to the fact that there are no prior estimates for Croatia, we focus on the relatively simple
estimation strategy in order to provide first insights. Naturally, the wages and their dynamics do not depend only
on trade patterns. In addition to personal characteristics of workers, labour market factors – including wage
bargaining process, tax policy, strength of the unions, skills demand and supply, etc. – are the most important
determinants of final wage determination. However, in order to fill in the gap in the existing literature, we want
to focus on specific industry features and trade patterns and abstain from other possible determinants.
Structure of the paper is following. The next section briefly summarizes the main findings from the literature in
order to provide theoretical framework for the empirical analysis. Section 3 discusses data sources and provides
preliminary insights on the subject. Section 4 presents empirical strategy, while results are laid out in Section 5.
The last section offers conclusions.
2. THEORETICAL FRAMEWORK
The idea that labour markets (wages) are under the influence of trade patterns, and that different segments of the
labour force (skilled vs. unskilled) are expected to have different consequences accordingly, is standard textbook
case of trade economics. The traditional models of Heckscher-Ohlin and famous Stolper-Samuelson theorem are
frequently used to analyse the effects of trade liberalization. One of the issues is that in the long run, when
factors of production are mobile across industries, standard Heckscher-Ohlin’s theory predicts that factor prices
will be equalised across industries and any differences in wages for similar types of work will eventually
disappear. The empirical studies have usually not been able to find this long-run relationship. Another point can
be attributed to Krugman (2008) who states that the nature of trade has significantly changed during the past
decades and this is not frequently taken into account in the empirical studies.
Relying on theoretical models, we can foresee benefits from increased integration-related trade related to product
variety. This love for variety increases consumers’ utility, but on the other hand produces new competitiveness
pressures for the domestic firms. One assumption is that, as a result, domestic firms will adopt more efficient
behaviour (Helpman and Krugman, 1985). If the trade is more intra-industry (defined as intensive trade of
similar products within the same industry) than inter-industry (when the division of trade products is more clear,
implying trade of products with different quality) it is assumed that the consequence will be relatively low
adjustment costs of production factors reallocation through smooth adjustment process. Such success stories are
more likely in case of developed economies integration. Whether integration induces low adjustment costs in
case of transition economies is a question that deserves empirical verification.
In general we can assume several adjustment mechanisms of labour markets to trade. The first one is related to
the increased variety gains as previously described (Krugman, 1981). It could be foreseen that the internal
restructuring due to increased competition on the domestic market will result in closing down of low competitive
firms (Melitz, 2003). We can also assume the case when the effect will be entirely shifted to the reduction of
labour costs, without closing down of enterprises (Davis and Harrigan, 2011). Both adjustment mechanisms have
been documented in Croatia on a case-level basis.
The focus of this paper is on the trade patterns at the level of economic activity, and in particular the links to
labour market indicators. Revealing the trade patterns on the level of economic activities is important in order to
enhance the discussion of competitiveness. The attention to the latter issue has been frequently drawn within the
analysis of EU accession process of transition economies, related to the smooth adjustment hypothesis. The
hypothesis states that if intra-industry trade (IIT) has higher share in the overall trade between the countries, the
integration associated adjustment costs will be less severe than in cases when the share of inter-industry trade is
27
relatively higher. Azhar and Elliot (2008) offer following explanation for this argument. The increases in trade
will result in changes in imports and export on a sector/product level. If the trade patterns are for the most part
inter-industry in nature, than these sector changes will be reflected in transferring production resources between
industries, from contracting to expanding industries. If there are large differences in relative production factor
endowments of the two trading countries, the costs of adjustments from one industry to another will be higher.
Smooth adjustment hypothesis has been frequently assessed and confirmed or refuted in empirical studies. Part
of the differences in results could certainly be attributed to the different measures of intra-industry trade and
labour cost changes. However, the precise measurement issues related to the appropriate intra-industry trade
dynamics and/or those related to the adequate labour market changes remain unresolved. Brülhart, Elliott and
Lindley (2006) suggest individual employees sectoral and occupational distance indicator within the
manufacturing sector. Earlier studies have used industry employment change as an indicator of adjustment cost
(Brülhart and Elliott, 1998; Greenaway et al, 1999), while others made use of job turnover indicator (Brülhart,
2000; Andersson, Gustafsson and Lundberg, 2000). Over the years more consensual tone has been achieved for
the measurement of intra-industry trade, where researchers mostly agree that marginal intra-industry trade is
more appropriate for dynamic analysis of the changes in the labour market. Another frequently used indicator of
intra-industry trade - Grubel-Lloyd index has been challenged in the literature (Brülhart and Elliot, 1998) for its
ability to disentangle trade patterns especially in the cases of transition countries, which usually have large trade
disbalances as well as structural changes.
3. DATA SOURCES AND PRELIMINARY ANALYSIS
The nature of the analysis is empirical, making clear presentation of the data used in the estimates provided
below important. For the labour market data, we rely on the most frequently used data source for this type of
analysis – Labour Force Survey (LFS) data. Individual LFS data without identifier has been used in empirical
estimation. Since 2007, LFS methodology includes panel component. However, the data used was not actually
anonymised, so the panel component could not be utilised for the research purposes. In order to avoid double-
counting the same respondent, the individual data have been used only when they appeared first time in the
analysis (Drinkwater and Robinson, 2011).
In order to provide industry perspective, some indicators had to be aggregated to relevant NACE classification.
This has been done both in the case of labour market and trade data. The LFS data prior and including 2008
relied on an earlier NACE classification version (in Croatia referred to NKD2002) in comparison to more recent
data (NKD 2007). Fortunately, the data for 2009 included information on both classifications, so matching could
have been performed to ensure the comparability for longer time period.
To produce IIT indicators, Eurostat COMEXT data has been used. Estimates were made on the most detailed
level of aggregation (CN8), which enables correspondence between CN-PRODCOM-NACE classifications.
Using the available Eurostat correspondence procedures, the data were aggregated to the most recent NACE 2-
level classification (NKD2007) throughout the analysed period.
After presenting the data sources, we provide some initial trade indicators. Trade with EU countries presents a
large part of overall Croatian trade, which is one of the arguments behind integration process. However, the
question is whether this trade resembles more North-South pattern or the pattern which develops between
similarly developed economies. To provide some insights, we present the intra-industry trade indicators. The
methodology applied has been previously frequently used in the literature (Abd-el-Rahman, 1991; Fontagné and
Freudenberg, 1997; Freudenberg and Lemoine, 1999). IIT can be estimated following the concept of trade
overlap:
28
𝑀𝑀𝑛(𝑇𝑒𝑜𝑜𝑇𝑒𝑒, 𝑀𝑖𝑜𝑜𝑇𝑒𝑒)
𝑇𝑇𝑇𝑇𝑇 𝑜𝑜𝑇𝑇𝑜𝑇𝑜 = 𝑀𝑇𝑒(𝑇𝑒𝑜𝑜𝑇𝑒𝑒,𝑀𝑖𝑜𝑜𝑇𝑒𝑒)
The expression is evaluated at the disaggregated level of product classification. If it is above certain threshold,
then it is assumed that significant trade overlap exists and the trade is considered to be two-way (or IIT).
Threshold of 10 percent, frequently used in the literature, is applied in order to avoid the possible sensitivity of
the results to this parameter.
Figure 1 IIT with EU-15 and industrial production (1998=100) in Croatia
140
120
100
80
60
40
20
1998199920002001200220032004200520062007200820092010201120122013
Two way trade
Industrial production
Source: Central Bureau of Statistics and author’s estimates based on COMEXT.
The previous data shows that the share of two-way trade (IIT) between Croatia and EU-15 is relatively low, but
it seems to be increasing in the last few years. The industrial production pattern, on the other hand, reveals the
severe impact the crisis had on Croatian economy. Since we are analysing labour market effects, we cannot
assume that all of the changes in specific industries could be attributed to trade effects. Clearly, specific
industries have followed the defensive restructuring through shedding labour (Botrić, 2012). It does not
necessarily imply that retained workers have suffered from wage cuts or were able to gain additional wage
increases. Thus, the overall effect on the industry level cannot be assumed in advance.
The intra-industry trade varies significantly among specific industries. Also, trade patterns might be quite
different across time. To illustrate this, we present the shares of intra-industry trade in Croatian trade with EU-15
in two specific years – 2000 and 2010. The results are presented in Figure 2.
The data clearly shows that intra-industry trade shares in the overall trade are not the same through time. It might
be suspected that integration process in general increases the share of IIT, however there are examples where the
trend is reversed. In Croatian case, there is a sharp decline in IIT in leather industries, but some other industries
have also recorded decline. On the other side of the spectrum seem to be wearing apparel and rubber
manufacturing, which have recorded increase in IIT. One of the arguments behind these data could be attributed
to restructuring of specific enterprises. However, we might also argue that these data are year-specific, since it
has been frequently argued in the public debates that Croatian exports and imports dynamics is erratic due to the
lack of consistent economic policies.
29
Figure 2 IIT shares in total trade across industries
90
80
70
60
50
40
30
20
10
0
10 13 14 15 16 17 20 21 22 23 24 25 26 27 28 29 30 31 32
2000
2010
Source: author’s estimates based on COMEXT data.
NACE codes refer to the manufacture of: 10 - food products; 13 – textiles; 14 – wearing apparel; 15 – leather and related products; 16 –
wood and products of wood and cork, except furniture; 17 – paper and paper products; 20 – chemicals and chemical products; 21 – basic pharmaceutical products; 22 – rubber and plastic products; 23 – other non-metallic mineral products; 24 – basic metals; 25 – fabricated metal products, except machinery and equipment; 26 – computer, electronic and optical products; 27 – electrical equipment; 28 – machinery and equipment n.e.c.; 29 – motor vehicles, trailers and semi-trailers; 30 – other transport equipment; 31 – furniture; 32 – other manufacturing.
The dynamics of the intra-industry trade in time is more appropriately explored with marginal intra-industry
trade (MIIT) indicators, which capture the relative changes in trade between two periods. Similar to IIT, the
literature proposes various indicators. We follow the methodology proposed by Brülhart (1994) and calculate
MIIT based on following expression:
|∆𝑋 − ∆𝑀|
𝑀𝑀𝑀𝑇 = 1 − |∆𝑋| + |∆𝑀|
Where X refers to exports and M refers to imports, both of which are on a detailed level of aggregation. This
index varies between 0 and 1, where 0 indicates marginal trade in the particular industry to be completely of the
inter-industry type, and 1 represents marginal trade to be entirely of the intra-industry type. Specifically this
index has been used in the empirical estimates further discussed in subsequent sections.
4. EMPIRICAL STRATEGY
Our basic empirical strategy is to estimate the wage equation, which includes following traditional labour market
variables:
Age and age-squared. The persons can expect relatively different wages with respect to their age. It could be
argued that older persons have important experience, which cannot be measured directly with other observable
variables. However, there are arguments that diminishing returns are associated with age, so in order to capture
this effect all the specifications include age-squared.
Gender. It has been frequently addressed in the literature, even in case of Croatia (Nestić, 2010) that women
obtain on average lower wages than men. Consequently, we include dummy variable - which takes value 1 if a
person is male - into our specification.
30
Living in urban areas. It is frequently argued that urban areas offer wider variability in jobs, and consequently
also that important business centres are frequently located in such areas. Wage patterns are related to the
urbanisation degree. To capture this effect, we include a dummy variable which has value 1 if a person lives in
urban or semi-urban area.
Education is measured by the qualifications obtained and aggregated to the three levels – lower secondary,
upper secondary and tertiary. Due to the fact that the classification has changed during the analysed period, the
categories within each segment are not the same. Prior and including the year 2009, as lower secondary
education, categories »No school«, »1-3 basic school grades«, »4-7 basic school grades« and »Basic school« are
considered. As upper secondary education, categories »School for skilled and highly skilled workers«,
»Vocational secondary schools« and »Grammar school« are included. As tertiary education, categories from
»Non-university college« to »Doctorate« are considered. From the year 2010, the classification is as following.
Lower secondary includes three categories up to basic school. Upper secondary includes all the varieties of high
school education in Croatia, including short specialised after high school courses that enable students for certain
activities (like craftsmanship certificates). Tertiary starts with short university programmes (2 or 2.5 years) and
finish with doctorate. In order to avoid multicolinearity, upper secondary has been excluded from estimation.
Occupation in the analysis is defined as the occupation of the main job listed by the employed person.
Following occupation-dummies have been included in the specifications: Armed forces occupations; Managers;
Professionals; Technicians and associate professionals; Clerical support workers; Service and sales workers;
Skilled agricultural, forestry and fishery workers; Craft and related trades workers; Plant and machine operators,
and assemblers; Elementary occupation.
There are two sets of estimates. The first one is concentrated on the issue of industry wage premium. To that end,
previous list of variables is augmented with dummy variables for each NACE2 industry. Since we are interested
only in manufacturing sector, workers from other economic activities are not included in the sample. In order to
avoid multicolinearity, we have excluded activity NACE 19 – manufacture of coke and refined petroleum
products because the total trade with European Union in this segment was negligible throughout the analysed
period.
The first specification, consequently, has the following form:
ln 𝑤𝑇𝑔𝑇 = 𝛼 + 𝛽1𝑇𝑔𝑇 + 𝛽2𝑇𝑔𝑇2 + 𝛽3𝑖𝑇𝑜𝑇 + 𝛽4𝑢𝑇𝑢𝑇𝑛 + 𝛽5𝑜𝑜𝑤𝑇𝑇 + 𝛽6𝑢𝑜𝑜𝑇𝑇
14
32
+ � 𝛽𝑖𝑜𝑜𝑜𝑢𝑜𝑇𝑒𝑀𝑜𝑛𝑖−5 + � 𝛿𝑗𝑇𝑜𝑒𝑀𝑜𝑀𝑒𝑎𝑗 + 𝜖
𝑖=7
𝑗=10
Where all the variables have been previously explained and the estimates have been repeated for each year in the
period 2004-2012. In this case we are interested in the delta-coefficients and in order to save space, only these
are presented in Table 1.
In case of alternative specification, most of the variables are the same, but instead of the dummy variables for
economic activity, MIIT indicator has been used for the NACE-2 level activity a worker is employed in. In that
case we have specific coefficient related to that variable, and these results are presented in Table 2. Both results
are presented and discussed in following section.
31
5. RESULTS
The results of the estimation in Table 1 reveal that there is an industry wage premium within Croatian
manufacturing sector. Relative to the sector that had the lowest share of trade with the EU countries throughout
analysed period, some industries had consistently lower wages. This implies that the openness to competition of
those industries and orientation towards the foreign markets results in relatively lower wages (after controlling
for education, age, sex, occupation and living area of their workers). Important fact is that we were not able to
find any positive significant coefficient in the analysed period. Thus, those industries that are competing on the
international market are not able to compensate their workers in a same way that those oriented towards the local
market were. Not only that we can see negative wage premium for the manufacturing sector vs. for example,
public sector and other non-tradables, we have also detected tradable-non tradable pattern within the
manufacturing sector.
It is interesting to notice that traditional labour-intensive industries - such as food, textiles, wearing apparel,
leather – have consistently significant negative wage premium, even after controlling for worker-specific
characteristics. This suggests that labour intensive industries continue to compete on the international market
with relatively lower labour costs, even though the competition from Asian markets has significantly increased
during the last decades.
Another interesting point is that, even during this relatively short timeframe, we can notice that changes occur.
The relative wage premiums are not the same through time.
32
Table 1 Estimated industry wage premium coefficients
Estimated coefficients (standard errors) across years
NACE activity
2004
2005
2006
2007
2008
2009
2010
2011
2012
-0,20*** -0,27*** -0,24*** -0,23*** -0,13*
-0,23**
-0,24**
-0,18**
-0,20***
10
(0,05)
(0,06)
(0,06)
(0,09)
(0,07)
(0,10)
(0,11)
(0,08)
(0,07)
-0,30*** -0,13**
-0,26*** -0,15
-0,03
-0,17
-0,20
-0,17*
-0,21**
11
(0,06)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,13)
(0,10)
(0,10)
-0,20*
-0,15
-0,22*
-0,20
0,41***
0,01
-0,14
-0,03
0,11
12
(0,11)
(0,11)
(0,12)
(0,14)
(0,13)
(0,15)
(0,17)
(0,13)
(0,14)
-0,56*** -0,44*** -0,65*** -0,43*** -0,15*
-0,40*** -0,58*** -0,33*** -0,39***
13
(0,06)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,13)
(0,10)
(0,10)
-0,58*** -0,50*** -0,68*** -0,50*** -0,40*** -0,46*** -0,45*** -0,42*** -0,46***
14
(0,05)
(0,06)
(0,07)
(0,09)
(0,07)
(0,10)
(0,11)
(0,08)
(0,08)
-0,62*** -0,45*** -0,60*** -0,30*** -0,23*** -0,40*** -0,42*** -0,34*** -0,34***
15
(0,05)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,12)
(0,09)
(0,08)
-0,49*** -0,42*** -0,56*** -0,40*** -0,22*** -0,36*** -0,40*** -0,28*** -0,38***
16
(0,05)
(0,06)
(0,07)
(0,09)
(0,07)
(0,10)
(0,12)
(0,08)
(0,08)
-0,35*** -0,40*** -0,28*** -0,25*** -0,18**
-0,15
-0,22*
-0,26*** -0,27***
17
(0,06)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,12)
(0,09)
(0,09)
-0,33*** -0,25*** -0,32*** 0,02
-0,28*** -0,31*** -0,35*** -0,33*** -0,32***
18
(0,06)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,13)
(0,10)
(0,10)
-0,29*** -0,21*** -0,24*** -0,33*** -0,13*
-0,18
-0,25**
-0,26*** -0,21**
20
(0,05)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,12)
(0,09)
(0,09)
0,05
0,19**
-0,48*** 0,12
-0,15*
-0,00
0,11
0,09
0,11
21
(0,07)
(0,08)
(0,08)
(0,10)
(0,09)
(0,14)
(0,14)
(0,10)
(0,11)
-0,24*** -0,24*** -0,31*** -0,21**
-0,28*** -0,17
-0,40*** -0,29*** -0,20**
22
(0,05)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,12)
(0,09)
(0,09)
23
-0,27*** -0,17*** -0,23*** -0,39*** -0,00
-0,17*
-0,25**
-0,13
-0,13
33
(0,05)
(0,06)
(0,07)
(0,09)
(0,07)
(0,10)
(0,12)
(0,08)
(0,08)
-0,49*** -0,44*** -0,51*** -0,44*** -0,30*** -0,26**
-0,45*** -0,33*** -0,28***
24
(0,06)
(0,07)
(0,07)
(0,09)
(0,08)
(0,11)
(0,12)
(0,09)
(0,09)
-0,29*** -0,22*** -0,35*** -0,25*** -0,11
-0,15
-0,21*
-0,24*** -0,25***
25
(0,05)
(0,06)
(0,07)
(0,09)
(0,07)
(0,10)
(0,11)
(0,08)
(0,08)
-0,09
-0,12
-0,28*** -0,09
-0,08
-0,16
-0,15
-0,20*
-0,22**
26
(0,06)
(0,07)
(0,08)
(0,11)
(0,09)
(0,13)
(0,13)
(0,11)
(0,10)
-0,20*** -0,17**
-0,32*** -0,23**
-0,15**
-0,17
-0,11
-0,12
-0,18**
27
(0,05)
(0,07)
(0,07)
(0,09)
(0,08)
(0,11)
(0,12)
(0,09)
(0,09)
-0,31*** -0,26*** -0,47*** -0,23**
-0,16**
-0,27**
-0,25**
-0,30*** -0,20**
28
(0,05)
(0,07)
(0,07)
(0,10)
(0,08)
(0,11)
(0,12)
(0,09)
(0,08)
-0,07
0,04
-0,30*** -0,34*** -0,10
-0,17
-0,17
-0,32*** -0,19*
29
(0,06)
(0,07)
(0,08)
(0,10)
(0,08)
(0,11)
(0,13)
(0,11)
(0,11)
-0,16*** -0,09
-0,19*** -0,12
-0,03
-0,10
-0,15
-0,12
-0,07
30
(0,05)
(0,06)
(0,07)
(0,09)
(0,08)
(0,10)
(0,12)
(0,08)
(0,08)
-0,30*** -0,48*** -0,56*** -0,86*** -0,41*** -0,47*** -0,46*** -0,51*** -0,55***
31
(0,05)
(0,06)
(0,07)
(0,09)
(0,07)
(0,10)
(0,11)
(0,08)
(0,08)
-0,51*** -0,31*** -0,34*** -0,45*** -0,23**
-0,10
-0,30**
-0,49*** -0,42***
32
(0,07)
(0,07)
(0,08)
(0,10)
(0,09)
(0,12)
(0,14)
(0,10)
(0,09)
N
3371
3134
3182
2986
2798
1434
1327
1045
1009
Adjusted R2 (%) 44,85
46,88
44,42
56,95
51,90
47,70
44,94
48,83
50,63
Source: author’s estimates based on LFS and COMEXT data.
Notes: *** denotes significance at 1 level, ** at 5 and * at 10 percent.
NACE codes refer to the manufacture of: 10 - food products; 11 – beverages; 12 – tobacco products; 13 – textiles; 14 – wearing apparel; 15 – leather and related products; 16
– wood and products of wood and cork, except furniture; 17 – paper and paper products; 18 – printing and reproduction of recorded media; 20 – chemicals and chemical products; 21 – basic pharmaceutical products; 22 – rubber and plastic products; 23 – other non-metallic mineral products; 24 – basic metals; 25 – fabricated metal products, except machinery and equipment; 26 – computer, electronic and optical products; 27 – electrical equipment; 28 – machinery and equipment n.e.c.; 29 – motor vehicles, trailers and semi-trailers; 30 – other transport equipment; 31 – furniture; 32 – other manufacturing.
34
To further elaborate the issue of trade pressures on the wages, we have explicitly included marginal intra-
industry trade estimated on the level of NACE2 activity into the equation. Controlling for individual labour
market indicators (education, age, gender, occupation and living area), we focus on the relationship between
intra-industry trade and wages. Specifically, we analyse whether the industries in which the intra-industry trade
with European Union have on average higher or lower wages. The results for the analysed period are presented
in following Table 2. All of the estimated coefficients from the wage equations are not presented in order to save
space, but could be available from the author upon request.
Table 2 Estimated MIIT coefficients in wage equations
Year
Estimated coefficient(standard error)
N
Adjusted R2 (%)
2004
-0,19*** (0,06)
3371
34,57
2005
-0,35*** (0,07)
3134
37,73
2006
-0,80*** (0,06)
3182
36,08
2007
-0,81*** (0,09)
2986
45,84
2008
-0,41*** (0,07)
2798
45,35
2009
-0,29*** (0,09)
1434
41,48
2010
-0,14* (0,07)
1327
38,91
2011
-0,81*** (0,12)
1045
43,41
2012
0,08 (0,10)
1009
40,79
Source: author’s estimates based on LFS and COMEXT data.
As the results of the estimation show, until 2011 the higher marginal intra-industry trade in the activity was
associated with significantly lower wages. This means that the more industry actively tried to integrate into the
European market by trading products of similar value, the lower average wage it was able to offer to its workers.
The accession period in Croatian industry was consequently associated with increased pressures on its workforce
in tradable sector.
6. CONCLUSIONS
The paper has addressed the issue of industry wage premium and trade pressures on wages in Croatian
manufacturing sector. The estimates have revealed that within manufacturing sector there is an industry wage
premium, which remains for some industries active throughout the period. Specifically, relative to the economic
activity that virtually had no trade with EU-15 during the 2004-2012 period, all other activities had negative
wage premiums. In case of labour intensive activities, those negative wage premiums were consistently
significant.
To further investigate the issue, wage equation has been re-specified in order to explicitly include the marginal
intra-industry trade with EU-15. The analysis has confirmed that the higher the marginal intra-industry trade in
specific economic activity, the lower the relative wage of the workers in that industry. This implies that the more
specific industry is integrated in the common EU market, the more it tries to compete with relatively cheaper
labour force.
The analysis presented in the paper points to the conclusion that there is an additional tradable vs. non-tradable
wage policy issue within the manufacturing sector itself. It has been frequently emphasized that this Dutch
decease has important consequences for the overall Croatian competitiveness position. However, previous
analysis in the literature did not go beyond the public-private gap or the manufacturing-services gap. The
analysis in this paper implies that the effects are possibly even deeper.
35
The notion that there are industry wage premiums is of particular importance for Croatian labour market policy.
It has been frequently emphasized that the labour market in Croatia is rather rigid and suffering from low
occupational and any sort of mobility. This implies that workers “stuck” in low-wage industry are most likely to
have less prospects to move to other industries. Without increased mobility, however, there are even less chances
for decreasing wage premiums in the future.
REFERENCES
Abd-el-Rahman, K. (1991), Firms’ competitive and national comparative advantages as joint determinants of trade composition, Weltwirtschaftliches Archiv, Vol. 127, No. 1, 83–97.
Andersson, L., Gustafsson, O. and Lundberg, L. (2000) Structural Change, Competition and Job turnover in the Swedish Manufacturing Industry 1964-96, Review of International Economics, Vol. 8, 566-582.
Azhar, A.K.M. and Elliott, R.J.R. (2008), On the Measurement of Changes in Product Quality in Marginal Intra-Industry Trade, Weltwirtschaftliches Archiv, Vol. 144, No. 2, 225-247.
Brülhart, M. (1994), Marginal intra-industry trade: measurement and relevance for the pattern of industrial adjustment, Weltwirtschaftliches Archiv, 130(3): 600–613.
Brülhart, M. (2000), Dynamics of Intra-Industry Trade and Labour-Market Adjustment, Review of International Economics, Vol. 8, 420-435.
Brülhart, M. and Elliott, R.J.R. (1998), Adjustment to European single market: inferences from intra-industry trade patterns, Journal of Economic Studies, Vol. 25, 225-247.
Brülhart, M. and Elliott, R.J.R. (2002), Labour-Market Effects of Intra-Industry Trade: Evidence for the United Kingdom, Weltwirtschaftliches Archiv, Vol. 138, No. 2, 207-228.
Brülhart, M., Elliott, R.J.R. and Lindley, J. (2006), Intra-Industry Trade and Labour-Market Adjustment: A Reassessment Using Data on Individual Workers, Review of World Economics, Vol. 142, No. 3, 521-545.
Botrić, V. (2012), Mehanizmi prilagodbe hrvatske prerađivačke industrije: analiza s aspekta tržišta rada, Ekonomski vjesnik, Vol. 25, No.1, 121-132.
Davis, D. R. and Harrigan, J. (2011), Good Jobs, Bad Jobs, and Trade Liberalisation, Journal of International Economics, Vol. 84, 26 – 36.
Drinkwater, S. and Robinson, C. (2011), Welfare Participation by Immigrants in the UK, IZA Discussion Paper Series No. 6144.
Fontagné, L. and Freudenberg, M. (1997), Intra-industry trade: methodological issues reconsidered, CEPII Working Paper, No. 97-01, January, Paris: CEPII.
Freudenberg, M. and Lemoine, F. (1999), Central and Eastern European Countries in the International Division of Labour in Europe, CEPII Working Paper, No. 99-05, April, Paris: CEPII.
Greenaway, D., Hine, R., Milner, C. And Wright P. (1999), An Empirical Assessment on the Impact of Trade on Employment in the United Kingdom, European Journal of Political Economy, Vol. 15, 485-500.
Helpman, E. and Krugman, P.R. (1985) Market Structure and Foreign Trade: Increasing Returns, Imperfect Competition, and the International Economy, Cambridge MA: MIT Press.
Krugman, P. R. (1981), Intra-industry Specialization and the Gains from Trade, Journal of Political Economy, Vol. 89, No. 5, 959 – 973.
Krugman, P. (2008), Trade and wages, reconsidered, Brookings Papers on Economic Activity, No.1, 103-154.
Melitz, M. J. (2003), The Impact of Trade on Intra-industry Reallocations and Aggregate Industry Productivity, Econometrica, Vol.71, No.
6, 1695 – 1725.
Nestić, D. (2010), The Gender Wage Gap in Croatia – Estimating the Impact of Differing Rewards by Means of Counterfactual Distribution, Croatian economic survey, Vol. 12, No.1, 83-119.
36
Vinko Zaninović
University of Rijeka, Faculty of Economics, Rijeka, Croatia
Katja Zajc Kejžar
University of Ljubljana, Faculty of Economics, Ljubljana, Slovenia
INTENSIVE AND EXTENSIVE MARGINS OF CROATIAN MANUFACTURING
EXPORTS: EVIDENCE FROM 2000-2012 PERIOD
ABSTRACT
This paper investigates the development of intensive and extensive export margins using augmented gravity
model on the industry level data for Croatia during the 2000-2012 period. During this period important events
on national and global level occurred with the significant implications for trade flows. On the national level:
signing of Stabilization and Association Agreement (SAA) in 2001 and Croatia’s signing of the Central
European Free Trade Agreement 2006 (CEFTA). Moreover, in 2011, Croatia began applying protocols on the
rules of origin providing diagonal cumulation (DC) between CEFTA parties involved in the Stabilisation and
Association Process (SAP) and European Union.
Also, during observed time period, financial crisis originating in the United States, spread to the rest of the
world and the sovereign debt crisis in European Union resulted in low economic growth in the whole Europe.
Sharp drop in economic activity was particularly noticed in Croatia, where manufacturing sector lost more than
50,000 workers (around 17% of total workforce in manufacturing sector) and the number of firms decreased by
2,650 (around 11% of all firms in manufacturing sector). We use augmented gravity model, where apart from
standard variables used in gravity equations, dummy variables for trading partners under preferential trading
systems are included to explain changes in export activity of Croatian firms across industrial divisions during
observed period. Separate analysis is carried out for exports of intermediate, consumption and capital goods,
defined using Broad Economic Categories classification. Our results show that SAA and DC significantly
affected intensive export margin (average exports across industry divisions), specially exports of consumption
goods, while the global trade collapse affected negatively intensive margin and the extensive trade margin
(number of firms across industry divisions exporting to foreign markets).
Keywords: economic crisis, manufacturing sector, intensive and extensive export margins, gravity model
JEL classification: F10, F12, F15
37
1. INTRODUCTION
Gravity equation, introduced into international trade by Tinbergen in 1962, became standard tool in the analysis
of bilateral trade flows due to its excellent explicative power. Standard gravity equation assumes that trade
between countries is determined by the economic masses of trading partners, proxied by gross domestic product
(GDP), and by the distance between trading partners, distance being proxy variable for the bilateral trade costs.
Gravity trade model is generally used for ex post estimation of impacts of different trade policies and trade
distortions on bilateral trade. Subsequently, so-called augmented gravity model emerged in the literature with the
primary aim of capturing ever-wider set of trade policy options effects.
Since its introduction, gravity model was estimated on the country level, and from 1990s, analysis performed on
firm level and transaction level data (in 2000s) emerged. Mainly, focus of these papers was estimation of
elasticities of bilateral trade with respect to trade costs. Trade costs include costs incurred from engaging in
international trade such as transportation costs, tariffs, non-tariff barriers, etc. Aim of the papers was to
understand the factors that affect trade costs, so that the welfare implications of their change can be given. One
of the main issues in this research field is the problem of measure, since the “direct measures are remarkably
sparse and inaccurate” (Anderson & van Wincoop, 2004).
In this paper we estimate augmented gravity model on industry level data for Croatia with the aim of estimating
effects of the change of macroeconomic surroundings on Croatian manufacturing export. We estimate how
signing of Stabilization and Association Agreement (SAA) in 2001 and Croatia’s admission into the Central
European Free Trade Agreement 2006 in 2006 (hereinafter CEFTA) affected industry export flows. Moreover,
we are interested to see whether enabling of diagonal cumulation of origin between CEFTA parties involved in
the Stabilisation and Association Process (SAP) and European Union significantly affected exports of Croatian
manufacturing sector. Also, we are interested to see whether so-called trade collapse in 2009 had major impact
on exports and the number of firms exporting. Trade collapse in 2009 was induced by the financial crisis
originating in the United States, that spread to the rest of the world and the sovereign debt crisis in European
Union resulted in low economic growth in the whole Europe. Sharp drop in economic activity was particularly
noticed in Croatia, where manufacturing sector lost more than 50,000 workers (around 17% of total workforce in
manufacturing sector) and the number of firms decreased by 2,650 (around 11% of all firms in manufacturing
sector) (CBS, 2015). By using country-time dummy variables we take into account time specific effects for each
destination market, so we can monitor changes in trade margins by year. Moreover, separate analysis is carried
out for exports of intermediate, consumption and capital goods, defined using Broad Economic Categories
(BEC) classification. BEC classification was used because we wanted to test whether vertical specialization and
supply chains in general play important part of Croatian manufacturing exports.
According to Behrens et al. (2013), who did analysis of the global trade collapse effects on the firm level data for
the Belgium, trade fall affected strongly intensive margin (exports per firm), while the extensive margins
(number of firms, average number of destination and origin markets countries per firm) were relatively
unaffected. Our results confirm their findings for intensive margin (average exports across sectors per country),
but in the case of Croatia’s industry sector, extensive margin was affected negatively as well (although we
follow only one definition of extensive margin, i.e. number of exporters per country).
In second chapter we provide brief literature review of the gravity model in general and empirical papers using
industrial level data. Third chapters contain description of the database used and explanation of the methodology
applied. Fourth chapter is reserved for the estimations results and discussion. The final chapter concludes.
38
2. LITERATURE REVIEW
Gravity model of international trade by Tinbergen (1962) became the corner stone of the bilateral trade flows
research, and was continually methodologically updated, since the original model lacked microfoundations and
was not consistent with prevailing theories of that time (like Heckscher-Ohlin theory). First, so-called augmented
gravity model, was developed by Linnemann (1966), which included population as a measure of country size.
Also, other papers included different right hand side variables (RHS) like per capita income, common language,
common currency, historical ties (like colonial links, wars, etc.). Each specification of the gravity model had
primary aim of capturing supply structure of the exporter and demand structure of the importer country.
Seminal papers by Bergstrand, in 1985 and 1990 introduced price indexes and exchange rate variables in the
gravity equation (1985) and monopolistic competition (1990) by assuming that countries completely specialize
in different product varieties. One the most important methodological contributions came from Anderson and
van Wincoop (2003) and are discussed in chapter 3 (see equation [2]).
From the econometric point of view, standard estimation of cross-sectional data that prevailed in 1990s was
substituted for panel data analysis using fixed effects estimation (Cheng & Wall, 2005). Aim of the change (apart
from the fact that panel data become more available) was to control for heterogeneity between trading pairs and
to allow for unobserved and/or misspecified factors that explain trade flows (see term Ωijt in equation [1]).
When it comes to the gravity model estimation on industry level, paper from Chen and Novy (2011) measures
trade integration across manufacturing industries in EU using modern gravity model setting (baseline is
Anderson and van Wincoop’s contribution). They use data for 163 manufacturing industries and find substantial
degree of heterogeneity across industries for substitution elasticities and the degree of trade integration (degree
of trade integration is connected with industry-specific characteristics). Moreover, they find that cross-country
trade integration is lower for new EU member states (EU-10) and that cross-border trade significantly depends
on transportation costs, proxied by c.i.f./f.o.b. ratios.
Paper from Sohn (2005) uses gravity model in order to explain change in bilateral trade flows of South Korea,
using industry level data. His findings are that APEC variable (Asia-Pacific Economic Cooperation) has
significant effects of Korea’s trade volume.
Aforementioned papers present theoretical and empirical baseline for our paper in which we estimate gravity
model (using theoretically valid gravid model setting) on industry level data for Croatia between 2000 and 2012.
3. DATA AND METHODOLOGY
3.1. Data
Trade data was obtained from Croatian Bureau of Statistics and includes firm-level data on exports and imports
of goods (8-digit Combined Nomenclature codes) for Croatian firms from 2000 up to 2012. We then aggregated
the data on industry level according to 2-digit National Classification of Activities (2007 version). Sample
includes bilateral trade flows between Croatian manufacturing sector (from division 10 to division 33) and 39
countries and represents more than 90% of total export from Croatia during observed period (EU27, CEFTA
countries, Unites States, Turkey, China, Switzerland, South Korea and Japan).
From 8-digit CN codes, we aggregated products into three BEC product categories, namely intermediate,
consumption and capital goods (BEC1, BEC2 and BEC3, respectively). Nominal GDP data for the destination
countries were taken from Eurostat, while dummy variables for free trade agreements and application of the
39
protocols on rules of origin providing diagonal cumulation (DC) between Croatia and other CEFTA countries
with EU were taken from official journals.
Next chapter explains methodology used in order to explain changes in trade margins of Croatian industrial
sector during observed period.
3.2. Gravity model specification
General formulation of the gravity equation in multiplicative form is the following:
X
Ω
ijt = GSitMjt
ijt , [1]
where Xijt is the monetary value of exports (or imports or total trade) from country (or firm or industry) “i” to
country “j” in time “t”. Sit includes exporter-specific factors (usually gross domestic product) that effectively
presents supply of exports (in general equilibrium context of the gravity model), Mjt comprises importer-specific
factors (again, gross domestic product) that effectively present demand for imports of destination market “j” in
time “t”. Last term, Ωijt denotes the ease of access to market “j” for exporter “i”. Equation [1] is so-called naive
form of gravity equation. Modern approach is to include fixed effects for exporter and importer (in the case of
panel data, exporter-year and importer-year effects, i.e. exporter and importer fixed effects are time varying).
We loosely follow approach Anderson and van Wincoop (2003) and include multilateral resistance terms (MRT)
which take into consideration trade resistance between a country and all other trading partners (their original
approach is technically demanding and is very rarely followed empirically – we used country-year fixed effects
to account for MRT). Main idea is that bilateral trade flows between trading partners “i ” and “j” are depending
on the multilateral resistance, i.e. they are depending on all other trading partners of those two countries. Their
formulation of the gravity equation, which is the basis for almost all subsequent papers that use gravity models in
order to explain bilateral trade flows is the following:
1−𝜎
𝑋𝑖𝑗𝑖 = 𝑌𝑖𝑖𝑌𝑗𝑖 � 𝑖𝑖𝑗𝑖 � , [2]
𝑌𝑖
𝜋𝑖𝑖𝑃𝑗𝑖
where Yit and Yjt stand for particular countries and Yt stands for world GDP, while tijt stands for tariff equivalent
of overall trade costs. Elasticity of supstitution between goods is represented with 𝜎, while 𝜋𝑖𝑖 and 𝑃𝑗𝑖 represent multilateral resistance terms (in another words – exporter and importer ease of market access). In practice,
importer and exporter fixed effects (dummy variables) are usually used in order to capture multilateral resistance
terms. Since we have panel data, we use country-year dummies in order to avoid “gold medal mistake” in
estimating gravity model, as suggested by Baldwin and Taglioni (2006). Since one part of our dyadic
relationships is always fixed (exports from Croatia) we only use destination country-year fixed effects.
So, the gravity model on industry-country level is the following:
𝑧𝑘𝑗𝑖 = 𝛼0 + 𝛽1𝑔𝑇𝑜𝑗𝑖 + 𝛽2𝑇𝑀𝑒𝑒𝑗 +𝛽3𝑜𝑜𝑛𝑒𝑀𝑔𝑗+ 𝛽4𝑒𝑇𝑇𝑗𝑖 + 𝛽4𝑜𝑇𝑐𝑒𝑇𝑗𝑖 + 𝛽5𝑇𝑀𝑇𝑔_𝑜_𝑤𝑢𝑗𝑖+ 𝛽𝑗𝑖+ 𝛽𝑖+ 𝜀𝑘𝑗𝑖 , [3]
where the dependent variable 𝑧𝑘𝑗𝑖 can be decomposed into the extensive and intensive trade margins:
𝑇𝑒𝑜𝑜𝑇𝑒𝑒𝑘𝑗𝑖 = 𝑛𝑘𝑗𝑖 + 𝑒𝑒𝑒𝑒𝑒𝑖𝑒𝑘𝑗𝑖 , [4]
𝑛𝑘𝑗𝑖
where values are expressed in logs and 𝑇𝑒𝑒𝑒𝑒𝑖
stands for industry division “k” (“k” being divisions from 10 to
𝑘𝑗𝑖
33) exports to country “j” in year “t”, 𝑛𝑘𝑗𝑖 is a number of firms in industry division “k” exporting to the
40
𝑒𝑒𝑒𝑒𝑒𝑖𝑒
destination j,
𝑘𝑗𝑖 are average exports per firm in industry division “k” to country “j”. For the other
𝑛𝑝𝑗𝑖
variables we have the following notation - 𝑔𝑇𝑜𝑗𝑖 stands for GDP of the trading partner, distj stands for distance
between capital cities between Croatia’s capital city and capital cities of partner countries, while contigj stands
for contiguity and has value 1 if the partner country shares land border with Croatia and 0 otherwise. Terms
𝑒𝑇𝑇𝑗𝑖, 𝑜𝑇𝑐𝑒𝑇𝑗𝑖, 𝑇𝑀𝑇𝑔_𝑜_𝑤𝑢𝑗𝑖 are dummy variables with the value of 1 if the free trade agreement is implemented between Croatia and EU, CEFTA and whether Croatian firms can cumulate origin with firms from
other CEFTA members and EU (value 1 if the trading partner is one of the WBC), respectively.
We use two free trade agreements (FTA) as proxies for variable trade costs of exporting, therefore we expect
positive signs of the parameters for 𝑒𝑇𝑇𝑗𝑖 and 𝑜𝑇𝑐𝑒𝑇𝑗𝑖. Inclusion of the dummy variable for the adoption of
protocols enabling diagonal cumulation should generally also affect positively exports, since theoretical
prediction and empirical findings prove that inclusion into the system of diagonal cumulation leads to a trade
creation effect, i.e. switch from less efficient domestic sources towards imports (in our case, we predict that other
CEFTA members will increase their import demand from Croatia, since we analyze only exports flows of
Croatian industrial sector). Diagonal cumulation should also contribute to trade reorientation – in our case from
EU towards other partners in the diagonal cumulation system. The study on the economic impact of extending
the pan-European system of cumulation of origin to the Mediterranean countries in the Barcelona process made
by The Sussex European Institute (2003) found supporting empirical evidence on cumulation based on a gravity
model estimates. They found that the trade actually taking place between partners that are not part of a system
allowing diagonal cumulation of origin with the European Economic Area is lower by up to 40-45% compared to
trade flows between countries which do allow for diagonal cumulation.
All continuous variables are in logs, so the estimated coefficients can be interpreted as elasticities. The
coefficients for the dummies need to be transformed in order to be interpreted as elasticities with the following
transformation – [exp(a)-1] (multiplied by 100 in order to get the percentage change) – where “a” is the
estimated coefficient of the dummy variable. Also, estimations of [2] were done with clustering on the panel
variable (nkd_iso, i.e. industry division and trading partner), which produces estimates that are robust to cross-
sectional heteroscedasticity and serial correlation.
Among different econometric estimators used in gravity model estimations, two are the most prominent - fixed
effects estimator (FE) and random effects estimator (RE). The estimation techniques allow us to control for
partner country and time-specific effects and to thereby control for economic and other country-pair-specific
factors that are constant over time and are not explicitly represented in the model. Usually, Hausman’s test is
used to test the specification for the fixed versus the random effects model. High values of Hausman χ2 statistics
reject the null hypothesis that individual specific effects are not correlated to the explanatory variables, which is
the assumption of REM. Low values of Hausman’s statistics thus favour REM. Since, the Hausmans’s test is
valid only under homoscedasticity, we use test of overidentyfing restrictions in order to see which of the two
estimations methods is more suitable. Logic of the test is the following:
a) The FE estimator used ortogonality conditions that the regressors (Xit) are uncorrelated with the
idiosyncratic error term – eit , so expected value of Xit*eit is equal to zero.
b) The RE estimator uses additional (overidentifying restrictions with respect to FE) orthogonality
conditions that the regressors are uncorrelated with the group-specific-time-invariant error term - ui , so
expected value of Xit*ui
The test is implemented in Stata 13 statistical package (as all estimations in this paper) and follows artificial
regression approach described by Arellano (1993) and Woolridge (2002).
41
Based on this test, we decided for fixed effects estimator, so the basic structure of our model is the following:
𝑎𝑖𝑖 = 𝛼 + 𝑒𝑖𝑖𝛽𝑘+ 𝑧𝑖𝛿 + 𝑢𝑖+ 𝜀𝑖𝑖 , [5]
Where, by the assumptions of the model, individual-specific (and time-invariant) effect (ui) is potentially
correlated with the regressors. When estimating, time-invariant variables (zi) like distance and contiguity are
removed together with time-invariant characteristics due to the demeaning, but we obtain estimates δ using
residuals from fixed effects estimations and regress them on distance and contiguity. Term 𝜀𝑖𝑖 stands for
idiosyncratic error term.
In the context of our research we find that using by using fixed effects we partially (only partially because we are
accounting for time fixed effects, while time variant effects for each of the industry division are not accounted
for) overcome the lack of other variables on the industry level (only variable at the industry level are exports by
industry divisions).
4. RESULTS OF THE EMPIRICAL ANALYSIS AND DISCUSSION
Results obtained from estimation of equation [2] can be explained as the elasticity of average exports across
industry divisions (for intensive margin see decomposition of total exports under [3]) and number of firms across
industry (for extensive margin, see [3]) with respect to the change in one of the RHS while holding other RHS
variables constant. Results for estimation of FE model are shown in tables 1a, 2a and 3a. Estimates of the time-
invariant variables, obtained after regressing FE residuals on distance and contiguity dummy are shown in
appendix (tables 1b, 2b and 3b).
Table 1a shows that aggregate exports across industries are significantly affected by the rise in foreign demand,
as well as by the enabling of diagonal cumulation between CEFTA members and EU. Parameters for two FTAs
are negative, but after estimating with the one-year lags (results not shown, but available upon request),
parameters are in line with theoretic predictions. When looking across different product groups, enabling of
diagonal cumulation affected significantly exports of intermediates.
Table 1a Gravity regression results for intensive export margin (total exports) on industry level, total and
by BEC products groups, 2000-2012, fixed effects estimation
(1)
(2)
(3)
(4)
VARIABLES
lnexp
lnexp
lnexp
lnexp
Total
Intermediate
Consumption
Capital
goods
goods
goods
gdp
1.580**
1.129
0.206
-0.541
(0.697)
(0.820)
(0.916)
(1.630)
saa
-1.795*
0.807
1.856**
-1.921
(0.954)
(1.026)
(0.905)
(1.834)
cefta
-1.634*
-1.647
-0.317
-2.944
(0.964)
(1.178)
(1.349)
(2.126)
diag_c_eu
2.567**
2.519**
2.364
0.246
(1.080)
(1.068)
(1.708)
(2.999)
Constant
-27.15
-17.50
5.874
24.11
(17.80)
(20.92)
(23.54)
(41.60)
Observations
9,686
8,638
6,570
5,303
R-squared
0.162
0.149
0.144
0.169
Number of
906
871
823
711
nkd_iso
*Country-year and year effects included
**Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Source: Authors’ estimations
42
Results from Table 2a indicate that average exports per firm across industry divisions reacted positively for
exports of intermediates and consumption goods as a results of the SAA, but the impact on exports of
consumption goods is stronger that for intermediates, which is in line with theory, i.e. substitution elasticities are
higher for consumption that for intermediate goods (Broda & Weinstein, 2006). Same is for the impact of
diagonal cumulation.
Table 2a Gravity regression results for intensive export margin (average exports per firm) on industry
level, total and by BEC products groups, 2000-2012, fixed effects estimation
(1)
(2)
(3)
(4)
VARIABLES
lnavg
lnavg
lnavg
lnavg
Total
Intermediate Consumption
Capital
goods
goods
goods
gdp
1.294**
0.948
0.511
-0.391
(0.650)
(0.731)
(0.934)
(1.469)
saa
-1.843**
1.252
1.756*
-1.750
(0.932)
(0.920)
(0.961)
(1.699)
cefta
-1.547
-1.338
-1.296
-3.168
(0.956)
(1.140)
(1.377)
(2.003)
diag_c_eu
2.315**
2.457***
3.095*
1.173
(1.000)
(0.928)
(1.769)
(2.698)
Constant
-20.30
-14.06
-2.767
19.44
(16.58)
(18.62)
(23.94)
(37.40)
Observations
9,686
8,638
6,570
5,303
R-squared
0.139
0.096
0.112
0.136
Number of
906
871
823
711
nkd_iso
*Country-year and year effects included
**Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Source: Authors’ estimations
Table 3a shows that the number of exporters across industry divisions was positively affected by the increase of
foreign demand in general, while the SAA affected negatively. Reason for the significant negative effects of
SAA could be increased competition from the EU. CEFTA membership affected positively number of firms
exporting consumption goods that indicates (estimation using one-year lags also showed positive effects on total
exports as well) that Croatian industry sector adapted quickly to tariff reduction/removal.
43
Table 3a Gravity regression results for extensive export margin (number of firms in industry) on industry
level, total and by BEC products groups, 2000-2012, fixed effects estimation
(1)
(2)
(3)
(4)
VARIABLES
lnid
lnid
lnid
lnid
Total
Intermediat
Consumption
Capital
e goods
goods
goods
gdp
0.428***
0.155
-0.194
-0.0444
(0.130)
(0.172)
(0.119)
(0.177)
saa
-0.370**
-0.354*
0.0842
-0.182
(0.187)
(0.204)
(0.143)
(0.177)
cefta
-0.118
-0.240
0.687***
0.0865
(0.248)
(0.260)
(0.189)
(0.185)
diag_c_eu
0.211
0.0393
-0.469**
-0.591*
(0.205)
(0.296)
(0.191)
(0.303)
Constant
-9.084***
-2.385
6.268**
2.427
(3.321)
(4.419)
(3.017)
(4.564)
Observations
9,686
8,639
6,571
5,303
R-squared
0.387
0.361
0.299
0.324
Number of
906
872
823
711
nkd_iso
*Country-year and year effects included
**Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Source: Authors’ estimations
Estimation results for our gravity model show dependence of Croatian exports at the industry level with respect
to foreign demand for both trade margins. This is logical, since more that 90% of Croatian exports are covered
with in our data set and exports of goods makes around 20% of the national GDP on average during observed
period. FTAs variables (SAA and CEFTA), that are proxies for variable trade costs have negative effects on total
exports flows (although, as already mentioned, estimations with one-year lag show positive signs on the
parameters), while when looking at the different product groups, we see that exports of consumption goods was
positively affected by trade integration. Enabling of the diagonal cumulation of origin affected positively both
the total exports and export by product groups which testifies to importance of the rules of origin rules to
regional trade flows.
When looking at the country-dummy variables (not presented, available upon request), global trade collapse
from 2009 (started at Q4 of 2008) affected negatively both intensive and extensive margins. Results from the
Tables in the Appendix all confirm negative effect of distance and positive effects of common land border on
total exports and exports across product groups as well.
5. CONCLUSION
The results based on gravity model estimations confirm that enabling of diagonal cumulation of rules of origin
had significantly positively impact on Croatian export across industry divisions (on both trade margins), in
particular for trade with intermediates and consumption goods. This is in line with theoretical predictions and
proves that the introduction of a system of diagonal cumulation of origin between the European Union, the
Western Balkan countries participating to the Stabilisation and Association Process and Turkey is contributing to
reduction of regional trade costs and enhance the trade performance of, in our case Croatia. On the other hand,
estimates of the impact of the trade part of the Stabilization and Association Agreement as well as Central
European Free Trade Agreement 2006 are ambiguous and therefore it is impossible to conclude whether they
had significant impact of exports performance of Croatian manufacturing sector.
44
REFERENCES
Anderson, J. E., van Wincoop, E. (2003), Gravity with Gravitas: A Solution to the Border Puzzle, The American Economic Review, 93(1): 170-192.
Anderson, J. E., van Wincoop, E. (2004), Trade costs, Journal of Economic Literature, 42(3), 691-751.
Arrelano, M. (1993), On the testing of correlated effects with panel data, Journal of Econometrics, 59(1-2): 87-97.
Augier, P., Evans, D., Gasiorek, M., Holmes, P. and Lai-Tong, C. (2003), Study on the economic impact of extending the pan-European system of cumulation of origin to the Mediterranean partners’ part of the Barcelona process.
Baldwin, R. and Taglioni, D. (2006), Gravity for dummies and dummies for gravity equations, National Bureau of Economic Research Working Paper 12516, NBER.
Behrens, K., Corcos, G., and Mion, G. (2013), Trade Crisis? What Trade Crisis?, The Review of Economics and Statistics, 95(2): 702-709.
Broda, C., and Weinstein, D. E. (2006), Globalization and the Gains from Variety, Quarterly Journal of Economics, 121(2): 541–585.
Chen, N., and Novy, D. (2011), Gravity, trade integration, and heterogeneity across industries, Journal of International Economics, 85(2): 206-221.
Cheng, I. H., and Wall, H. J. (2005), Controlling for Heterogeneity in Gravity Models of Trade and Integration, Federal Reserve Bank of St.
Louis Review, 87(1): 49-63.
Clark, T. S., and Linzer, D. A. (2015), Should I use Fixed or Random Effects?, Political Science Research Methods, 3: 399-408.
Linnemann, H. (1966), An Econometric Study of International Trade Flows, Amsterdam: North-Holland.
Tinbergen, J. (1962), Shaping the World Economy – Suggestions for an International EconomicPolicy, New York: The Twentieth Century Fund.
Woolridge, J. M. (2002), Econometric Analysis of Cross Section and Panel Data, The MIT Press, Cambridge, Massachusetts.
45
APPENDIX
Table 1b Gravity regression results for intensive export margin (total exports) on industry level, total and
by BEC products groups, 2000-2012, OLS estimation
(1)
(2)
(3)
(4)
VARIABLES
lnexp
lnexp
lnexp
lnexp
Total
Intermediate Consumption
Capital
goods
goods
goods
dist
-3.149***
-2.032***
-0.462***
-0.0372
(0.121)
(0.120)
(0.138)
(0.197)
contig
1.129***
2.257***
1.822***
-1.780***
(0.314)
(0.339)
(0.346)
(0.396)
Constant
21.46***
13.50***
2.781***
0.608
(0.860)
(0.867)
(0.971)
(1.366)
Observations
9,686
8,638
6,570
5,303
R-squared
0.565
0.449
0.124
0.048
Number of
906
871
823
711
nkd_iso
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Source: Authors’ estimations
Table 2b Gravity regression results for intensive export margin (average exports per industry) on
industry level, total and by BEC products groups, 2000-2012, OLS estimation
(1)
(2)
(3)
(4)
VARIABLES
lnavg
lnavg
lnavg
lnavg
Total
Intermediate Consumption
Capital
goods
goods
goods
dist
-2.652***
-1.401***
-0.527***
0.169
(0.110)
(0.107)
(0.134)
(0.155)
contig
0.798***
1.538***
1.448***
-1.626***
(0.282)
(0.273)
(0.319)
(0.277)
Constant
18.09***
9.310***
3.279***
-0.785
(0.782)
(0.770)
(0.938)
(1.059)
Observations
9,686
8,638
6,570
5,303
R-squared
0.526
0.330
0.123
0.093
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Source: Authors’ estimations
46
Table 3b Gravity regression results for extensive export margin (number of firms in industry) on industry
level, total and by BEC products groups, 2000-2012, OLS estimation
(1)
(2)
(3)
(4)
VARIABLES
lnid
lnid
lnid
lnid
Total
Intermediate
Consumption
Capital goods
goods
goods
dist
-0.951***
-0.553***
-00157
-0.220***
(0.0351)
(0.0351)
(0.0504)
(0.0429)
1.contig
1.068***
0.549***
0.464***
0.0266
(0.138)
(0.138)
(0.127)
(0.132)
Constant
6.388***
3.680***
0.0251
1.447***
(0.259)
(0.258)
(0.364)
(0.304)
Observations
9,686
8,638
6,570
5,303
R-squared
0.608
0.389
0.039
0.066
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Source: Authors’ estimations
47
SECTION II
EU MEMBER STATES’ EXPERIENCES IN DIFFERENT POLICY AREAS
48
Jiří Dušek, Lubomír Pána
College of European and Regional Studies, Ceske Budejovice, Czech Republic
THE USE OF PPP PROJECTS ON THE LEVEL OF STATES, REGIONS AND
MUNICIPALITIES UNDER THE CONDITIONS OF THE CENTRAL EUROPEAN
REGION
ABSTRACT
In the last 20 years, there has been a revival in Central Europe of the idea of including private capital in
financing construction and operating public infrastructure in the form of partnership between public and private
sector (so-called PPP). In the Czech Republic, several projects have been selected, so-called pilot projects. On
these projects, legislative conditions and the ability of both sectors to cooperate in this area should have been
tested. One of these pilot projects should have been the D3 motorway between Prague and České Budějovice. Up
until today, however, not a single PPP has reached the implementation stage and all remain more or less in the
area of ideas and preliminary discussions. In contrast, in the neighbouring Austria, a number of PPPs have been
completed, albeit not without problems. One of the consequences of this experience in Austria with PPPs was a
decision of ending these projects on federal level and leaving them in the competence of individual ministries
and federal states. Big infrastructure projects are then realised with the help of a “private partner” which is a
company founded and fully owned by a state.
This contribution thus focuses on an analysis of selected private partnership projects on various levels of state
and on a comparison of conditions for their implementation in the Czech Republic and Austria, which will show
the historically different development of implementation and inclusion of PPP projects in both countries.
However, the objective of the contribution is not only to show the diametrically different experience of the two
countries, but also to lay out suggestions which could be useful for successful implementation and use of this
method of cooperation. A major problem encountered during work on the contribution is the fact that the
situation around the PPP projects is highly politicized and unclear, in neither of the two countries there is a fully
functional and to the public available central register of projects which would transparently declare a useful use
of public finances. The topic of PPPs is frequently discussed. However, almost exclusively this discussion
remains a theoretical discussion without practical application. Under certain conditions, PPP projects seem to
be a promising form of cooperation suitable for both the public and the private sector. However, it still remains
a form of cooperation which is negatively influenced by many half-truths, myths and political unwillingness,
which are only some of the obstacles which hinder its spread in Central Europe.
Keys words: Austria, Czech Republic, PPP projects, private sector, public sector, regional policy
JEL classification: R580 Regional Development Policy
49
1. INTRODUCTION
The need to include private capital into construction of public infrastructure already occurred in the 19th century
when states issued to private companies temporary licenses for transport constructions (especially canals and
railways) under the condition that the private investor builds the construction at its own cost and will have the
right to operate it for a defined time. In cases where it was obvious that the profit from operating such a
construction would not cover the costs or generate profit, a guarantee of some profit was a part of the licence
(local railways guaranteed by the Municipal Committee of the Czech Kingdom and others). In the 20th century,
after most licenses ran out, these constructions were taken over by the state and further development of
infrastructure was financed solely by the state.
In the last 20 years, there has been a revival in Central Europe of the idea of including private capital in
financing of construction and operation of public infrastructure in the form of partnership between public and
private sector (so-called PPP). In the Czech Republic, several projects have been selected, so-called pilot
projects. On these projects, legislative conditions and the ability of both sectors to cooperate in this area should
have been tested. One of these pilot projects should have been the D3 motorway between Prague and České
Budějovice. Up until today, however, not a single PPP has reached the implementation stage and all remain more
or less in the stage of ideas and preliminary discussions.
At the same time, in the years 2007-2010, the Austrian government decided to extend and modernize
infrastructure and to facilitate application of the PPP model. In this context, it is important to stress that in
Austria, there is no solid legal framework for implementation of PPP projects and there is no single state-wide
concept for PPP projects. As a result, there is no central register of projects. Although many projects were
successfully implemented, the overall situation makes a very uncoordinated impression (EIB, 2009, p. 4). Up
until now, only one major transport infrastructure project has been successfully implemented in Austria on
federal level. The project in question is a PPP project called PPP-Projekt Ostregion, i.e. the connection of Vienna
with the northern part of the country via a motorway and the follow-up by-pass roads and motorways around
Vienna. Responsible for financing of such highly important transport infrastructure projects is the state-owned
company ASFINAG (Autobahn- und Schnellstraßen-Finanzierungs-Aktiengesellschaft). Within the PPP project
Ostregion, a part of the planning, construction, operation and maintenance was transferred to a private partner -
Bonaventura Consortium through a licence contract. This project was implemented in the so-called Design-
Build-Finance-Operate (DBFO) form and the contract was signed on 12th December 2006. Work on the
construction was completed in January 2010 and since February 2010, the motorway has been in operation.
Notably, gradual take-over and putting into operation of parts of the motorway was taking place since the year
2009. However, apart from this Ostregion PPP project, there is no other PPP project under way on federal level
in Austria.
In the area of railway infrastructure, an analysis was carried out in the past using an external advisor in regard to
the possibility of including private entities in the construction of planned railway tracks. However, it has shown
that for this area, the use of a PPP model is not suitable. The reasons are as follows: firstly, it is very difficult in
railway transport to clearly define and mark routes and areas which should be operated separately. Secondly, the
size of the projects and at the same time the slow return of investment would mean a big financial burden for a
possible private investor. In this context, it would be very difficult to find somebody who would be interested in
such an investment.
It is clear from what has been said above that in regard to financing transport infrastructure, financing from
public money is without doubt still predominant. Alternative forms of financing can, in special cases, represent a
kind of additional financing. In the latest Austrian government programme for the period of 2013-2018, we can
read in regard to financing of transport infrastructure projects that a new analysis will be carried out with the
objective of determining where and to which extent it would be possible and suitable to use alternative sources
of financing such as PPP projects. This analysis is especially important in the context of the planned project of a
broad gauged railway which is supposed to connect Moscow, Bratislava and Vienna and thus open up a railway
connection to China. It is obvious from the project plan that financing will be more than difficult and for this
50
reason it is necessary to take into consideration all available options since the possible future railway connection
Vienna-Bratislava-Moscow and further east is economically very interesting. However, in view of the currently
uneasy relationship of the EU with Russia, it is a question whether this project will be implemented.
2. MATERIAL & METHODS
This contribution focuses on an analysis of selected private partnership projects on various levels of state and on
a comparison of conditions for their implementation in the Czech Republic and Austria, which will show the
historically different development of implementation and inclusion of PPP projects in both countries. However,
the objective of the contribution is not only to show the diametrically different experience of the two countries,
but also to lay out suggestions which could be useful for successful implementation and the use of this method of
cooperation. A major problem encountered during work on the contribution is the fact that the situation around
the PPP projects is highly politicized and unclear; in neither of the two countries there is a fully functional and to
the public available central register of projects which would transparently declare a useful use of public finances.
The topic of PPPs is frequently discussed. However, almost exclusively this discussion remains a theoretical
discussion without practical application.
The methodology of this contribution is in compliance with methods usually used in scientific research; it is
based on the use of the latest theoretical knowledge gained from specialised literature, specialised research and
studies, newspapers and materials published by individual participants in regional development. Also, the
methodology is based on looking for and assessment of mutual relationships which contribute to the clarification
of the problems solved and to a deduction and formulation of adequate conclusions which can be derived from
such an analysis. The analysis done in the contribution is based on data from the European Commission, the
Ministry of Finances of the Czech Republic and the Ministry of Regional Development of the Czech Republic.
The main methods used in the article are analysis, description and comparison.
3. FORMS AND CHARACTERISTICS OF VARIOUS PPP MODELS
In practical reality, there are a huge number of various forms of contract arrangements related to PPP projects.
The differences lie mainly in the form of spreading risks between the private and the public investor and in
division of benefits created as a result of the cooperation. The state is looking for a private license holder to
which it could transfer the right to build, finance and operate an infrastructure project using a so-called license
contract. After expiry of a period defined in this contract, i.e. after 25 to 30 years, the thus constructed project
goes into the possession of a state (Mittendorfer, Weber, 2004, p. 36).
This model most resembles contractual agreements known as concession models, although it is also possible to
use other forms of agreements, specifically the mixed or the leasing model. All these models have in common
that the licensee usually commits to financing, construction and maintenance of an infrastructure project for the
whole period of a contract.
Operation model – A private company commits to design, construct, finance and operate an
infrastructure project at its own risk. Mostly, so-called one-purpose companies are founded for this
purpose, the only purpose of which is the above mentioned activity.
Licence model – An ownership company is founded which subsequently rents a facility to another
company which operates it. The reward for this activity is paid out from taxpayer money, usually in
annual instalments in the form of instalments or directly by enabling to collect tolls.
Mixed model – Depending on the purpose of a construction, a “reward” is paid out from public money,
i.e. an amount defined in a licence treaty.
Leasing model – Regular annual leasing payments are paid from public sources and the value of an
investment is paid back in the form of leasing.
51
A different classification of PPP projects according to types and forms is provided by the European Commission
(2014): DBB (Design-Bid-Build), OM (Operation and Maintenance), BOT (Build-Operate-Transfer), DBFO
(Design-Build-Finance-Operate), BOO (Build-Own-Operate).
Similarly as the classification of PPP projects differs, differ the opinions about them. It is thus not possible to
clearly and with general validity state all advantages and disadvantages because every involved party approaches
these projects differently, from its own angle.
PPP projects are usually characterized by the following features (Ministry of Regional Development of the
Czech Republic, 2006, ASPI, 2003):
The commissioner is always a public entity which defines requirements for a public service and remains
responsible for provision of this service,
The role of a private partner is to provide as effectively as possible public infrastructure and/or service
according to requirements defined by the commissioner,
The commissioner transfers to the private partner some risks which it usually bears alone when
implementing a public project (i.e. the risk of demand, availability and the construction risk),
The project is usually implemented using a purpose-founded company,
In a number of projects, the private partner builds, operates, maintains and finances an infrastructure
project alone. This makes control and easier planning of overall costs of a project possible,
The public commissioner pays to a private partner in the course of a project regular payments or enables
to a licence holder to use an infrastructure project or a service (the collection of payments from users)
or both, there are a number of payment mechanisms available that can be used,
The projects are characterized by relatively long duration of contract regarding cooperation between
public and private partner on different aspects of a project, the licence contracts are usually signed for
25-30 years,
The method of financing a project, partially by the private sector, sometimes using complex agreements
among various parties which define the transfer of level of risk and responsibility among individual
partners,
The significant role of the economic operator which takes part in various stages of a project (design,
implementation, financing). The public partner mainly focuses on defining aims which have to be
reached in public interest, the quality of the services provided and the price policy. It is also responsible
for overseeing the fulfilment of these aims,
The division of risks between a public partner and a private partner, to which risks usually born by the
public sector are transferred. However, PPP does not necessarily have to mean that a private partner
bears all or most risks connected with a project. Division of risks is defined according to possibilities of
parties involved to evaluate, control and confront this risk.
4. RESULTS AND DISCUSSION
Experience with PPP projects abroad varies greatly. They are the most popular in Anglo-Saxon countries where
their use was to a great extent connected with the government of neo-conservative political forces and the arrival
of a theory called New Public Management (Nemec, Wright, Stillman, 2002). Within this theory, or better said
in its British form, the presupposition is accentuated that the private sector is more effective in production of
goods and services than the public sector, especially due to hard budget limitations (and thus also the possibility
to go bankrupt) and the presence of a profit motive. According to proponents of this philosophy, it is thus
necessary to leave the maximum possible volume of production of good and services to the private sector and to
try to use successful methods of management from the private sector in the public sector. One of the possible
tools is then seen in the use of PPP projects. However, the use of PPP projects is not limited to Anglo-Saxon
countries. They are also used in countries of Continental Europe, although to a lesser extent. Moreover, not all
existing mutations of these projects are used in Continental Europe, only the basic forms.
52
There is also historical experience with PPP projects in the region of Central Europe where first projects were
already implemented in the 19th century. In most cases in the Czech Republic and Austria, the projects were
nationalized in subsequent decades, in some cases without any form of compensation having been paid. Under
current circumstances, it can be assumed that the society has moved so far forward that this kind of treatment
caused by possible political changes is unlikely, nevertheless this fact can represent a certain risk for private
partners. Especially in ex-communist countries where political culture and the party system is still developing we
cannot speak of stability.
In the following table, the current situation in the Czech Republic and in Austria is summed up. We can see clear
differences in the understanding and use of PPP projects. Also, the number of successfully implemented projects
is different. In the Czech Republic, special legislation is in place which defines the area of PPPs, i.e. the already
mentioned licence law; in Austria, only already existing laws are used, i.e. the Federal Law on Public Orders
from the year 2006 into which amendments required by the EU were incorporated. In the Czech Republic, so-
called pilot projects were selected; in Austria there were no official pilot projects, although sometimes the
Ostregion project is considered as such, especially as it was the first project in the area of transport
infrastructure. In the Czech Republic, there is not a single project implemented on the level of state or even
region, whereas in Austria there are several big infrastructure projects implemented on federal level, as well as
on the level of ministries and federal states. Similarly, we can find many more projects on the level of
municipalities in Austria than is the case in the Czech Republic, very different are also the areas in which PPPs
are used the most often. Due to missing or unavailable information, it is not possible to compare the financial
volume of successfully implemented projects in both countries.
Table 1 - Comparison of conditions for the implementation of PPP projects in the Czech Republic and
Austria (own analysis, 2014)
Area
Czech Republic
Austria
Law 137/2006 Code on Public There is no legal framework intended
Legal
Orders and Law 139/2006 Code exclusively for PPP projects, contracts
framework
on Licence Treaties and Licence are awarded according to existing laws.
Procedures.
Only one project is considered as a
9 so-called pilot projects were pilot/trial – the construction of 52 km of
Pilot projects
selected, none was implemented
motorways and express ways as a part
for various reasons.
of the Ostregion project. The project
was fully implemented.
Projects
In the Czech Republic, no project Big transport infrastructure projects
awarded by
announced by a ministry has been have been implemented (motorways,
ministries
implemented.
railways).
Projects on the
According to available
Dozens of projects have been
level of regions
information, no project has been implemented, a big part in the area of
/ federal states
implemented.
healthcare and transport infrastructure.
Almost two hundred projects have been
Projects on the
Dozens of projects have been implemented. However, due to lack of
level of towns
implemented, some cases more central evidence of public orders, data
and
similar to outsourcing than PPPs is only hard to obtain or completely
municipalities
in their classic sense.
unavailable.
Areas where
Most projects in terms of volume Most projects in terms of quantity are in
PPPs are used
of signed licence treaties are in the area of sport and leisure time.
the most often
the area of water management.
Value of
EUR 1075.3 mil. (only projects
implemented
CZK 4358 mil. (municipalities).
implemented on the federal level or the
projects
level of ministries).
It is obvious from the comparison that although the two countries have common initial experience in the area of
cooperation of the private and the public sector (the construction of railways in the Austro-Hungarian
Monarchy), the current functioning of this cooperation in both countries is diametrically different. In both
countries, centres were founded the task of which was to promote, support and monitor the functioning of PPPs.
In the Czech Republic, this centre was the now defunct PPP Centre (activity of the PPP Centre was stopped in
53
2012, since then PPP projects are fiscally managed from the Ministry of Finances of the Czech Republic and
legislatively by the Ministry of Regional Development of the Czech Republic, which was partially replaced by
the PPP Association (in Czech: Asociace PPP)), in Austria by the so-called PPP Forum, which has not been
replaced by any other institution. Exact data and detailed information about projects are not available because
there is no centralized database of projects as there is no central institution managing the projects. The once
functioning PPP Forum.at, i.e. a kind of equivalent of the Czech Asociace PPP has its web pages continually
unavailable. The whole PPP Forum.at project was cancelled due to lack of finances and individual projects are
managed by the involved institutions themselves, i.e. by municipalities, towns, federal states and ministries on
federal level – especially the Austrian Ministry of Finances and the Austrian Ministry for Transport, Innovation
and Technology.
Both countries also differ in the area of selection of a private partner. In Austria, in case of big infrastructure
projects, there is an affiliated company founded and fully owned by the state, or as the case may be, a
corporation of several big companies, e.g. the Bonaventura consortium. Except for the D47 motorway project, it
was not possible to find out from available sources which private investor was selected in so-called pilot projects
in the Czech Republic, but it can be said that most of these pilot projects were stopped before a private partner
was selected.
Also in the number of successfully implemented and functioning projects, Austria is ahead of the Czech
Republic, even though there are problems with availability of data necessary for exact monitoring of projects.
4. CONCLUSION
Worldwide, the process of including private sector in the provision of public services only reached maturity at
the end of 1980s. The first country where this concept started to be applied on a bigger scale and which first
acquired experience in this area was Great Britain. In the last decade, the role of the private sector in financing
and operating infrastructure increased significantly. PPP projects proved useful in Ireland, the Netherlands,
Portugal, Spain, France, Austria, USA, Canada, Japan and Australia, but also in a number of developing
countries such as Chile and South Africa.
The study of experience with PPP projects in individual countries and sectors enables to acquire certain know-
how and avoid the repetition of mistakes and methods which proved as wrong. When evaluating projects on both
national and international level, it is necessary to abstain from specific problems or successes of individual
projects and to focus on general features which make PPPs a desirable method of operating public services or
providing public goods (Ministry of Finance of the Czech Republic, 2004).
The necessity to finance public interest projects and to preserve and develop the quality of life of citizens is
obvious. However, not only in the Czech Republic but also in the neighbouring Austria, questions are asked
where to get finances for these expensive investments. Especially in the area of transport infrastructure, it is
necessary to further invest into development of roads and railways, to enable better transport connections and
thus create better business environment. This can subsequently contribute to economic growth and thus new
finances for the state. One of the ways is the above described cooperation of the private and the public sector. If
we look at the situation in the Czech Republic, it is obvious that this form of cooperation has not been accepted
enthusiastically and the effort to implement it failed from the start, notably on projects which were intended to
test its functionality. Most of so-called pilot projects turned out to be expensive for the state, without even
reaching the stage of implementation. However, it has to be said that the situation is much better on the level of
towns and municipalities and if we find any successfully implemented projects in the Czech Republic, then it is
on this level.
In Austria, the situation was handled completely differently. In case of big projects, several stable and strong
partners joined or a state-owned company was founded (however, the question here is to what extent we can
speak of a private partner).
54
Also, areas in which PPPs are used in both countries are different. In the Czech Republic, on local level, we see
projects which are trying to solve urgent needs of a town or municipality. This can be the necessity to increase
the number of parking spaces in a town, selection of a new partner for the operation of a school canteen or a
sewer system. Only rarely can we see projects which are related to free time of citizens. Unfortunately, in the
Czech Republic, the meaning of PPPs was not fully understood, since even from the handful of implemented
projects, the majority belong more into the category of outsourcing.
If representatives from Czech municipalities, i.e. elected representatives whose duty is to work in public interest
claim, among other things, that PPP projects are not satisfactory due to their transparency and their almost zero
corruption potential, it will be very difficult to implement these projects on a bigger scale in the Czech Republic.
In the current situation, it would be more suitable to introduce a system similar to the one applied in Austria, i.e.
to accept in case of big infrastructure projects a consortium of several big stable companies as a private partner
which will be a guarantee of a successful implementation of a project. In the opinion of the authors, Czech public
administration has so far not shown the ability to correctly manage big infrastructure projects in which the
private partner bears some risks. The contracting authorities are failing not only in negotiation of conditions, but
also in safeguarding transparent selection of suppliers. All this confirms the generally known facts concerning
the poor quality of public administration in the Czech Republic.
REFERENCES
Austria - PPP Units and Realated Institutional Framework (2009), EIB, Luxembourg.
Guidelines for succesful Public-Private Partnership (2003) [online]. Luxembourg : European Commision, 2003 [cit. 2013-12-21]. Available at WWW: .
Metodika k zákonu 139/2006 Sb. o koncesních smlouvách a koncesním řízení (2006) , Ministry of Regional Development of the Czech Republic, Praha.
Mittendorfer, F., Weber, S. (2004), Public Private Partnerships: Gestaltung aus ökonomischer und juristischer Sicht, LexisNexis ARD
ORAC, Wien.
Nemec, J., Wright, G., Stillman, R. J. (2002), Public Management in the Central and Eastern European Transition: Concepts and Cases, NISPAcee Press, Bratislava.
Public private partnership - partnerství soukromého a veřejného sektoru (2003) [online]. Praha : ASPI, 2003 [cit. 2015-01-16]. Available at WWW: .
Zahraniční zkušenosti v oblasti Partnerství veřejného a soukromého sektoru (2004) [online]. Praha : Ministry of Finance of the Czech Republic, 2004 [cit. 2014-10-10]. Available at WWW: .
55
Jiří Dušek
College of European and Regional Studies, Ceske Budejovice, Czech Republic
THE PROBLEM AREA OF CZECH REPUBLIC'S USE OF EU STRUCTURAL
FUNDS IN THE PROGRAMME PERIOD OF 2007/2013
ABSTRACT
Regional politics only started to develop in the Czech Republic in the second half of 1990s. Until then, the
government was more focused on problems related to economic transformation. Moreover, differences between
regions were not so big in the Czech Republic. However, the second half of the 1990s brought significant socio-
economic differences between individual regions, which manifested themselves e.g. in a different level of
employment. Significantly, increased interest in regional politics was not only caused by economic problems, but
also by the approaching entry of the Czech Republic into the EU, which brought the possibility of using
significant financial means from EU structural funds.
In 2015, the Czech Republic finds itself in the third programme period and it is thus possible to assess not only
the impact of EU regional politics on the Czech Republic, but also the degree of success of the Czech Republic in
using these funds. In view of the fact that the Czech Republic belongs among the worst performing EU countries
in this regard in the programme period of 2007-2013, this contribution focuses on identification of the main
causes and barriers. Also dealt with will be examples of "bad” practice, on which systemic failures in drawing
EU structural funds can be shown. What we are witnessing is an interesting paradox in which financial means
which should have contributed to moving the Czech economy towards advanced European economies are in
many cases becoming a problem both for the submitter and the Czech state which due to possible non payment
increases its deficit in the range of single and double digit billions of Czech crowns. In its concept, however, the
contribution does not “only” focus on assessing the problems related to drawing EU funds in the Czech
Republic in the programme period of 2007-2013, it also has the ambition to show on concrete examples selected
problems in using EU structural funds, which does not have to be specific to the Czech Republic but can occur
any time in any other old or new EU member state. Among the most significant problems of the last programme
period are e.g. bureaucracy, insufficient administrative capacity, wrongly set up system of checks, frequent and
not systemic changes in legislation, corruption and many other factors. Unfortunately, we are in the year 2015,
the second year of the programme period 2014-2020 and instead of really starting the new programme period,
drawing of funds from the last period is still not completed, without calls from new operational programmes
having been announced. It is obvious that if the Czech Republic was 1.5-2 years behind in using EU structural
funds in 2007, the situation is unfortunately repeating, with all the related consequences for all parties involved.
Keys words: Czech Republic, EU funds, operational programme, project, regional policy
JEL classification: R580 Regional Development Policy
56
1. INTRODUCTION
One of the basic and significant policies of the EU is regional politics. The main objective of this policy is
elimination of differences between the levels of development in individual regions of EU member states. The
gap in development among individual regions has increased manifold after the accession of new members in
2004 and 2007. The significance of applying these principles within the EU is manifested by the fact that
regional politics represents more than 40% of EU budget. Typical for this policy is mainly construction and
repair of roads and motorways, railway tracks, airports, mainly connecting remote regions with main centres of
economic growth. The policy of solidarity is thus not only an empty word. It helps people in individual
countries, regions, towns and municipalities to find work and live better lives.
Regional politics belongs among so-called community or coordinated kinds of politics. This means that its focus
and implementation lies in individual member states while EU authorities are responsible for coordination and
correct implementation. Objectives and priorities of regional politics of the EU are changing with the
development of the European Union and are always newly defined for the next programme period. For the
period of 2007-2013, three main objectives were defined for the area of regional politics, which were
subsequently implemented with the use of so-called operational programmes. These objectives were common for
all EU member states and in the middle-term fiscal budget, 347 billion euro were reserved for these objectives.
According to the Ministry of Regional Development of the Czech Republic – National Coordination Authority,
2008, the objectives are as follows:
Convergence,
Regional competitiveness and employment,
European regional cooperation.
Table 1 - Division of EU structural funds among objectives of economic and social cohesion policies in the
period of 2007-2013 (Ministry of Regional Development of the Czech Republic, 2006)
Objective
Funds for EU 27
Funds for Czech Republic
Convergence
283 billion €
81.54 %
25.88 billion €
96.98 %
Regional competitiveness and
54.96 billion €
15.95 %
419.09 billion €
1.56 %
employment
European regional cooperation
8.72 billion €
2.52 %
389.05 million €
1.46 %
Total
347 billion €
100 %
26.69 billion €
100 %
The funds come from three different sources, based on which area in which country or region is being financed:
European Fund for Regional Development (EFRD) – innovations, investments and general
infrastructure,
European Social Fund (ESF) – projects helping in the area of employment and programmes for the
creation of jobs,
Cohesion Fund – research of renewable energy production, projects related to protection of the
environment and to traffic infrastructure. 15 countries are drawing finances from this fund. A
prerequisite is that economic output has to be lower than 90% of EU average (Portugal, Greece plus 12
new EU member states).
Generally speaking, in the last few years, the most investment is taking place into projects situated in the
countries of Central and Eastern Europe, including the Czech Republic. The most finances are used for support
of innovations, research and sustainable development and to create favourable conditions for small companies
which are the backbone of European economy. Part of the funds is also used for cross-border and inter-regional
cooperation projects.
57
Table 2 - Division of EU funds among EU states in the period of 2007-2013 (Dušek, 2011, own adaptation)
Country
Accession
Inhabitants
Area
GDP (PPP)
Allocation
(2012)
(km²)
per capita
of finances
(2012)
2007-2013
eur / % EU
in billion €
Austria
1995
8,443,018
83,871
33,600
131
1.20
Belgium
1952
11,094,850
30,528
30,500
119
2.06
Bulgaria
2007
7,327,224
110,879
12,100
47
6.67
Croatia
2013
4,398,150
56,594
15,600
61
-
Cyprus
2004
862,011
9,251
23,200
91
0.61
Czech Republic
2004
10,236,445
78,867
24,590
96
26.53
Denmark
1973
5,573,894
43,094
32,000
125
0.51
Estonia
2004
1,294,486
45,228
17,500
68
3.40
Finland
1995
5,401,267
338,145
29,400
115
1.60
France
1952
65,327,724
643,801
27,500
108
13.45
Germany
1952
81,843,743
357,022
31,100
121
25.49
Greece
1981
11,290,067
131,957
19,200
75
20.21
Hungary
2004
9,932,000
93,028
16,800
66
24.92
Ireland
1973
4,582,769
70,273
33,100
129
0.75
Italy
1952
59,394,207
301,340
25,200
98
27.96
Latvia
2004
2,041,763
64,589
14,700
62
4.53
Lithuania
2004
3,003,641
65,300
17,800
70
6.78
Luxembourg
1952
524,853
2,586
69,400
271
0.05
Malta
2004
417,520
316
22,000
86
0.84
Netherlands
1952
16,730,348
41,543
32,900
128
1.66
Poland
2004
38,538,447
312,685
16,800
66
67.19
Portugal
1986
10,542,398
92,090
19,200
75
21.41
Romania
2007
21,355,849
238,391
12,600
49
19.21
Slovakia
2004
5,465,311
49,036
25,300
75
11.50
Slovenia
2004
2,055,496
20,273
21,000
82
4.10
Spain
1986
46,196,276
505,370
24,900
97
34.66
Sweden
1995
9,482,855
450,295
32,800
128
1.63
United Kingdom
1973
63,256,141
243,610
28,400
110
9.89
EU total
28
506,820,764
4,381,376
25,600
100
347
members
billion €
Thanks to accession into the EU, the Czech Republic has the possibility of using EU funds devoted to policies of
economic and social cohesion in underdeveloped regions of the EU, with the objective of increasing the
competitiveness of these regions. In this context we have to see the EU programme periods of 2007-2013 and
2014-2020 as a wholly unique chance, since the Czech Republic will, for the first and probably also the last time,
have a chance to use a huge amount of financial means based on defined objectives and priorities set by the
National Development Plan of the Czech Republic for the programme period of 2007-2013, in the framework of
which so-called National Strategic Framework has been developed (Blažek, 2006). A New Strategy of Regional
Development in the Czech Republic for the period of 2014-2020 was approved on 15th May 2013 by
Government of the Czech Republic Resolution No. 344 (see Markl, 2013). However, the key question is whether
and how the Czech Republic will be able to effectively use these funds.
In total, since its accession into the EU on 1st May 2004 until 31st December 2013, the Czech Republic paid into
the EU CZK 342.8 billion and received CZK 676.2 billion. The total positive balance of the Czech Republic in
relation to the EU budget thus reached CZK 333.4 billion. Behind the markedly positive balance of the Czech
Republic in the last year lies especially drawing of the Czech Republic from the Cohesion Fund and the EU
Common Agricultural Policy, payments of the Czech Republic into the EU budget are growing only gradually
(Zeman, 2014).
The difference between income from the EU budget and payments into it, i.e. the net position of the Czech
Republic for the year 2007 amounted to CZK 15.2 billion. The Czech Republic reached this positive result even
despite the fact that in 2007, it already lost the right to draw finances from so-called budget compensations. In
2008, the net balance of the Czech Republic in relation to EU budget amounted to CZK 23.8 billion and even
CZK 42.3 billion in 2009. The balance further improved in 2009 when it reached CZK 47.9 billion. On the other
58
hand, the situation got significantly worse in 2011 because of problems with drawing funds from the operational
programmes Education for competitiveness, Environment and Transport, so there was a drop in the net balance
to CZK 30.8 billion. At the end of the programme period there was significant improvement and the net balance
reached CZK 73.1 billion in 2012 and CZK 84.1 billion in 2013. However, this did not change the fact that the
Czech Republic became one of the worst performing countries in terms of drawing finances from EU funds.
Figure 1 - Development of the net balance of the Czech Republic in relation to EU budget in the years
2004-2013 (Zeman, 2014)
2. MATERIAL & METHODS
The contribution focuses on an analysis of using/drawing of EU financial means in the programme period of
2007-2013 with focus on the Czech Republic and its comparison with other EU countries. In its concept,
however, the contribution does not only assess problems related to drawing from EU funds in the Czech
Republic in the programme period of 2007-2013, it also has the ambition to show on concrete examples selected
problems in using EU structural funds, which does not have to be specific to the Czech Republic but can occur
any time in any other old or new EU member state.
The methodology of this contribution is in compliance with methods usually used in scientific research; it is
based on the use of the latest theoretical knowledge gained from specialised literature, specialised research and
studies, newspapers and materials published by individual participants in regional development. Also, the
methodology is based on looking for and assessment of mutual relationships which contribute to the clarification
of the problems solved and to a deduction and formulation of adequate conclusions which can be derived from
such an analysis. The analysis done in the contribution is based on data from the European Commission, the
Ministry of Finances of the Czech Republic and the Ministry of Regional Development of the Czech Republic.
The main methods used in the article are analysis, description and comparison.
3. RESULTS AND DISCUSSION
The reasons behind inability to draw finances from EU funds or to draw only slowly are obvious. Excessive
fragmentation of supportive programmes markedly increases their costs, both on the side of the providers (costs
related to administration – from text of programme announcement to elaboration of request forms and evaluation
of received project proposals) and the side of the proposer (especially the necessity to acquire information about
the existence of individual programmes and detailed information about conditions that have to be met to be able
59
file a request for support in a project related to a specific programme). It is a sad fact that these extra costs lower
the volume of financial means available for implementation of the actual projects. The low transparency of a
whole array of support programmes also leads to a situation where some weaker starting entities do not manage
to reach any support at all because they do not have time to study the possibilities on offer or do not possess the
specific knowledge necessary to be able to propose a project meeting all formal criteria which are different from
programme to programme. It can thus happen that support is given to stronger entities which do not always need
it. Another problem (despite increased level of co-financing from the EU budget of up to 85% of legitimate
costs) is the necessity to provide the remaining 15% from Czech public sources (Blažek, 2006). In this context,
the following risks have been identified (Ministry of Regional Development of the Czech Republic – National
Coordination Authority, 2014):
non-use of allocations due to returning of the finances back to the programme (savings, corrections,
decision of beneficiaries not to implement projects etc.). At the same time, time available for
implementation of new projects is getting shorter, including timely completion of implemented projects
or non-implementation of some projects (e.g. for reasons of delays in public orders)
insufficient absorption capacity (i.e. inability to find high-quality projects and to manage their
implementation by the end of 2015, so-called n+2 rule),
high error rate resulting from audits carried out (up to double digit percentage points!).
The author has been dealing with the problem area of drawing finances from EU funds for several years and is of
the opinion that one of major problems is certain non-transparency with which data regarding absorption
capacity of individual countries are presented. For this reason, this area is, at least in the Czech Republic,
connected with a whole number of half-truths and myths. One of the most frequently occurring myths in the
media is the claim that the Czech Republic is the absolutely worst performing country in the area of drawing
finances from EU funds. This claim is disproved by data acquired by the author, according to which the Czech
Republic is “only” the 4th worst performing country with 63.2% (see figure 2). However, it is very difficult to
acquire up to date data because in the Czech Republic, the Ministry of Regional Development does not publish
the data regularly and e.g. DG Regional and Urban Policy prefers to publish data in its annual reports (Annual
Activity Report) in blank graphs so it is only possible to guess concrete numbers. This attitude of national and
European authorities logically makes any attempts to carry out regular comparative analyses which would
increase the pressure on relevant ministries and governing bodies more difficult.
Figure 2 - Level of absorption capability of EU countries as of 31.12.2014 (European Commission, 2015)
70% 60% 30% 90% 20% 30%
100%
30% 30% 80%
93, 92, 92, 89, 89, 88,
20% 40% 70% 70%
90%
85,
80% 60%
84,
70%
83, 83,
90%
82, 81, 81,
80, 80,
30% 30%
79,
78,
40% 10% 80%
76, 76,
80%
73, 73, 72, 50%
30% 20%
70%
65,
10%
63, 63,
30%
60, 56,
60%
10%
50%
45,
40%
30%
20%
10%
0%
a
a
g
a
ni
ia
ia
ria
m
ria
ly
ic
tia
oni
rus
ur
um
ark
ary
eden
any
en
ndsla and
alta
do
Spain
Ita
ania
Finland
Greece
Poland
Cyp
bo
Latv
nm
Irel
Aust
France
M
ing
epubl
m
Croa
Lithua
Portugal
Est
Sw
em
Germ
Belgi
Slov
De
Hung
Bulga
Slovaki
Ro
ether
Lux
N
ted KniU
Czech R
60
In view of the unflattering standing of the Czech Republic in the programme period of 2007-2013, it is striking
that at the beginning of 2015, the Czech Republic did not learn a lesson from the previous periods of 2004-2006
and 2007-2013 and again did not even start drawing finances in the new period due to non-approval of
operational programmes. Let's characterize the biggest problems related to Czech Republic's drawing of finances
from EU funds in the period of 2007-2013:
political changes – there were changes on government level at the beginning of individual programme
periods (2006 – start of conservative government, 2009 – fall of government, 2013 – government
resignation, start of interim government, early general elections, 2014 – start of socialist government),
which lead to mistakes and slowdown/stop of the process of preparation for a new programme period; it
is an interesting fact that during the interim governments, the Czech Republic drew finances the fastest,
high fluctuation of employees both in control committees and individual operational programmes
(connected with political changes),
big complexity of operational programmes – a total of 26 operational programmes (Goal Convergence –
8 thematic operational programmes and 7 regional operational programmes, Goal Regional
Competitiveness and Employment – 2 operational programmes, Goal European Territorial Cooperation
– 9 operational programmes),
bureaucracy → complexity of methodology, which was not even available at the beginning of the
programme period, inconsistency of interpretations of controlling committees etc., non sticking to
deadlines from the side of controlling authorities, inclusion of appendices to requests which already are
available to state administration,
frequent changes in legislation – especially Law on Public Orders (19 amendments in the period of
2007-2013), e.g. amendment to Law No. 73/2011 Code made conditions for public orders so much
stricter that their number dropped by 50% annually. Also, further executive legislative directives are
often missing after amendments,
monitoring system of EU funds → failure of national audit authorities – a check made by the European
Court of Auditors in 2012 identified a significant risk that the central audit authority of the Czech
republic systematically edits audit results so that in the annual control report the number of errors is
reduced under or to a 2% threshold. A good example is report on transport subsidies for the year 2011.
The audit authority found out that 1.85 percent of EU subsidies were allocated wrongly. This was a
negligible number of under two percent which proved that the system is functioning well. However, an
audit from Luxembourg checked the findings of the audit authority and found out that in transport
subsidies, 41.82 percent of the total amount were allocated wrongly. In transport subsidies in the year
2010, Czech audit authorities did not find any mistakes whereas the European Court of Auditors found
out that almost 5% of subsidies were spent in a breach of law. This number shows that serious mistakes
occurred in the control system, although it is a much better result compared with the year 2011 when
the Czech audit “oversaw” a mistake of 40%. The control report identifies as the most frequent problem
manipulation in public tenders – “concrete lobby” – this term refers to interests of construction
companies which, in cooperation with politicians, push through various unnecessary and overpriced
construction projects (e.g. the construction of a draw-bridge in Kolín which does not draw, deepening
of the Vltava River near Hluboká nad Vltavou as a tourist facility, construction of the D8 motorway
from Lovosice to Řehlovice etc.) – for more see e.g. Vlček, 2014 or Hradilek, 2014,
corrupt environment, methods of projects evaluation, fraud – in case of many implemented projects, the
final costs were several times higher than the originally envisaged costs – a good example is a three-
kilometre long bicycle path Rokytka in Prague Libeň which cost 5.2 million EUR, which is ten times
more than one kilometre of an average bicycle path in the German federal states of North Rhine-
Westphalia or Brandenburg where the costs are 180 thousand EUR per kilometre. The most expensive
German bicycle path is a luxurious route through the Hamburg neighbourhood Wilhelmsburg where
one kilometre cost € 550 thousand. The price of some domestic investments financed by the EU
exceeds the German level five to ten times – the most expensive domestic railway track is Votice-
Benešov; one kilometre of this fast train track for a speed of 160 km/hour cost 14 million euro, the same
61
as cheaper tracks for high speed trains, e.g. Madrid-Barcelona, Cologne-Aachen or Rennes-Connerré
where the trains reach a speed of over 300 km per hour (Holub, 2014),
fragmented communication and promotion of structural funds – paradoxically, the most massive
investments are taking place at the end of programme periods because of effort to spend money
allocated for technical support,
stopping of funds for operational programmes from the side of the European Commission,
requirements for co-financing or complete pre-financing of projects etc.
Table 3 - Level of absorption capability of EU countries in the years 2007-2014 (European Commission,
2015, own calculations)
Member state
2007
2008
2009
2010
2011
2012
2013
2014
Austria
2.00%
5.10% 19.00% 28.90% 39.60% 52.80% 67.80% 78.90%
Belgium
1.70%
5.00% 18.10% 23.20% 32.20% 49.20% 68.80% 82.40%
Bulgaria
2.20%
5.50%
9.50% 15.50% 23.60% 36.20% 49.50% 65.50%
Croatia
0.00%
5.00%
5.00%
5.50%
7.40% 10.30% 18.30% 45.10%
Cyprus
2.20%
5.50% 15.20% 26.20% 37.40% 44.30% 61.30% 84.30%
Czech Republic
1.40%
5.50% 12.10% 20.20% 26.50% 38.40% 51.90% 63.20%
Denmark
2.00%
5.00% 11.50% 19.70% 38.30% 45.30% 54.40% 80.80%
Estonia
2.20%
5.50% 19.50% 35.00% 42.00% 61.30% 81.30% 92.30%
Finland
2.00%
5.00% 16.50% 25.80% 40.90% 54.70% 75.70% 89.20%
France
1.60%
5.00% 13.60% 23.60% 34.50% 43.00% 59.90% 76.30%
Germany
2.00%
5.20% 17.40% 28.60% 41.20% 54.10% 70.80% 83.20%
Greece
2.00%
5.00% 10.60% 21.90% 34.90% 49.20% 69.60% 88.30%
Hungary
2.20%
5.60% 13.20% 21.30% 35.30% 44.20% 59.30% 76.30%
Ireland
2.00% 11.10% 23.30% 36.20% 48.30% 60.30% 70.10% 79.70%
Italy
1.70%
5.00%
9.80% 14.90% 21.70% 30.80% 50.10% 63.30%
Latvia
2.20%
5.50% 14.90% 25.10% 36.40% 52.20% 66.00% 81.70%
Lithuania
2.20%
5.50% 21.30% 34.10% 48.00% 62.90% 78.80% 93.70%
Luxembourg
1.00%
5.00% 10.10% 16.10% 40.60% 51.80% 67.80% 83.80%
Malta
2.20%
5.50%
9.70% 17.60% 27.30% 37.20% 50.30% 73.40%
Netherlands
2.00%
5.00%
8.30% 17.40% 33.60% 45.60% 63.80% 80.60%
Poland
2.10%
5.30% 13.00% 23.20% 37.20% 52.30% 67.90% 85.30%
Portugal
2.00%
5.00% 13.00% 25.20% 37.80% 59.20% 78.70% 92.60%
Romania
2.20%
5.60% 10.30% 13.00% 16.60% 22.60% 37.80% 56.30%
Slovakia
2.10%
5.50% 10.00% 18.90% 27.80% 41.10% 52.70% 60.10%
Slovenia
2.20%
5.50% 13.50% 24.80% 37.00% 50.30% 62.90% 81.70%
Spain
2.00%
5.00% 10.60% 22.40% 36.60% 51.70% 62.80% 72.80%
Sweden
2.00%
5.00% 16.20% 26.90% 46.50% 53.30% 68.70% 89.90%
United Kingdom
2.00%
5.00% 13.60% 27.70% 38.80% 50.90% 56.70% 73.10%
EU's absorption capacity
76.77%
4. CONCLUSION
According to Postránecký, 2010, regional politics in the Czech Republic was newly constituted after the year
1989 as one of major tools contributing to elimination of regional differences in social and economic
development of the Czech Republic. Whereas during the 1990s most policies of regional politics reacted mostly
non-systematically to the newly created regional disparities as a consequence of transformation of the Czech
economy, in the first decade of the 21st century, all pillars necessary for a systemic attitude to regional politics
were created as a consequence of external and internal factors.
Having said that, the current situation is still far from ideal, especially in regard to drawing financial means in
individual EU programme periods or the different views of the Ministry of Regional Development and the
62
individual players of regional development when solving various regional problems. When evaluating the
individual programme periods - 2004-2006, 2007-2013, 2014-2020, we can divide problems related to drawing
of financial means into three areas:
recurring problems – e.g. bureaucratic burden, political changes, failure of audit systems, delayed
beginnings of drawing from structural funds etc.,
problems already solved – with small exceptions, the Czech law is compatible with Acquis
communautaire (for more see e.g. Pána, 2010), there has also been professionalization of entities
involved both on the side of submitters and processors of projects,
completely new problems – an example is criminal prosecution of several directors of regional
programmes → temporary stop in drawing of financial means, sophisticated manipulations with public
orders or audit results, devaluation of the Czech crown by the Czech National Bank in the year 2013 →
changes in allocations in CZK etc.
Problems related to using EU funds are best demonstrated by the fact that from 6,253 Czech municipalities in the
Czech Republic, more than 20% were never involved in any operational programme, which is an alarming
number. As is the fact that the EU funds should have increased the GDP of the Czech economy and the standard
of living of Czech citizens to Western European level, whereas the reality is that due to breaching conditions
related to drawing funds for operational programmes or even pushing through projects not compatible with goals
of the operational programmes, the European Commission is not going to pay some implemented projects which
leads to an increase in state budget deficit and the follow-up fiscal measures – e.g. the rapid decrease in drawing
of funds by the Czech Republic in 2011 (see figure 2) is sometimes called as Řebíček’s Tax (minister of
transport and hidden owner of a major construction company called Viamont who pushed through several
expensive and later not co-financed construction projects). In the long term, however, there is gradual
improvement in conditions for drawing and using of EU funds, although not as much due to initiative from the
side of the Czech Republic but more due to pressure from the European Commission and other EU authorities
(see e.g. the Czech Law on Public Service and the threat of stopping all operational programmes).
REFERENCES
Abeceda fondů Evropské unie 2007-2013 (2008), Ministry of Regional Development of the Czech Republic – National Coordination Authority, Praha.
Analýza čerpání evropských fondů a krizové plány (2014), Ministry of Regional Development of the Czech Republic – National Coordination Authority, Praha.
Annual Activity Report 2012 (2012), DG Regional and Urban Policy, Brussels.
Blažek, J. (2006), Jakou regionální politiku potřebujeme po vstupu do EU? In Dočkal, V ., Šest let regionální politiky v ČR – Šance a limity, Masarykova univerzita v Brně – Mezinárodní politologický ústav, Brno. ISBN 80-210-3983-3.
Dušek, J. (2011), Historie a organizace Evropské unie, College of European and Regional Studies, České Budějovice. ISBN 978-80-86708-95-9.
EU cohesion funding – key statistics (2015) [online]. Brussels : European Commission, 2015 [cit. 2015-01-16]. Available at WWW:
.
Holub, P. (2014), Proč se zdvihací most postavený z peněz EU nezdvihá, diví se Němci [online]. Praha : MAFRA, 2014, 2.4.2014 [cit. 2015-01-16]. Available at WWW: .
Hradilek, L. (2014), Fotogalerie projektů [online]. Praha : Economia, 2014, 30.4.2014 [cit. 2015-01-16]. Available at WWW:
.
63
Markl, J. (2013), Strategie regionálního rozvoje ČR 2014-2020 [online]. Praha : Ministry of Regional Development of the Czech Republic, 2013 [cit. 2015-01-16]. Available at WWW: .
Pána, L. (2010), Normotvorná činnost obcí v ČR na příkladu měst České Budějovice a Příbram. In Decentralizácia a efektívnosť verejnej správy v podmienkach regiónov EÚ, UMB Banská Bystrica a Univerzita Hradec Králové, Banská Bystrica. ISBN 978-80-557-0088-5.
Postránecký, J. (2010), Regionální politika a regionální rozvoj v České republice. In Urbanismus a územní rozvoj [online]. Vol. XIII, 2010, No. 5. ISSN 1212-0855, 10-16 [cit. 2015-01-16]. Available at WWW: .
Regionální politika EU (2006) [online]. Praha : Ministry of Regional Development of the Czech Republic, 2006 [cit. 2015-01-16]. Available at WWW: .
Vlček, F. (2014), Fotogalerie projektů [online]. Praha : MAFRA, 2014, 2.4.2014 [cit. 2015-01-16]. Available at WWW:
.
Zeman, M. (2014), Čistá pozice České republiky vůči rozpočtu EU za rok 2013 [online]. Praha : Ministry of Finance of the Czech Republic, 2014, 12.2.2014 [cit. 2015-01-16]. Available at WWW: .
64
Andrej Kumar
Faculty of Economics, University of Ljubljana
Sonja Šlander
Faculty of Economics, University of Ljubljana
EU COHESION POLICY AND ABSORPTION IN SLOVENIA
ABSTRACT
The European Union cohesion policy helps to shape and improve the economic performance of member
countries from its earliest beginnings. The initial concept and orientation of the cohesion policy were formulated
already in the Treaty of Rome (signed 1957) introducing the European Economic Community (EEC) and later
the EU by the Maastricht Treaty (signed February 2, 1992). Similar to formal and functional changes of the
entire integration, the objectives, methods, terminology, and resources of the cohesion policy were evolving and
changing. The past changes of the EU structure and functioning, together with the cohesion policy changes, have
evolved from the ongoing EU enlargement and deepening processes. Discussing cohesion policy requires
understanding and analyzing of the EU achievements, together with the analyzing and understanding of the
cohesion policy specific implementation and impacts on the level of the individual EU member states.
On the EU level, analyzing of the cohesion policy and searching for the new concepts, objectives, resources, and
arguments supporting its active use is the most heated before accepting each financial perspective. On member
states level, the intensity of debates and arguments about the cohesion policy coincide with that on the EU level.
Additionally on member states level, discussions are more continuous because of the linking of cohesion
financial resources with the national economic development plans and achievements. The debates on national
levels are generally related to the assessments of potentials and realized levels of absorption for the allocated
EU cohesion funds. Assessments and analysis investigate the national adequacy of the cohesion funds. The issue
is linked to evaluating of the ability and conditions on the side of the member state to actually use the EU
cohesion funds successfully for the regional and national economic growth improvements.
The paper discusses a selection of aspects of the EU cohesion policy economic rationale. Some formal and
conceptual changes of the cohesion policy from the past are described with the idea to assess potential
difficulties in the use of the cohesion policy instruments on the level of eligible subjects. Major focus of the paper is on describing and critical evaluating of the two membership periods of the EU cohesion policy
implementation in Slovenia. The analyzing of the specifics in using the cohesion policy instruments in Slovenia is
based on the concept of the national absorption capacity and practical obstacles in achieving better economic
cohesion development results.
Keywords: The EU cohesion policy, cohesion, absorption capacity, economic growth, cohesion policy financial
instruments, Slovenia, National Strategic Reference Framework
JEL classification: O52, R10, N94, R58
65
1. INTRODUCTION
In the EU, different forms of policies and financial instruments are used to support and help regions and member
states to realize more even economic development. The background of such policies and instruments dates back
to the Treaty of Rome (1957). In the Treaty, the EU founding nations in the preamble decided their aim of
“reducing the differences existing between the various regions and the backwardness of the less favoured
regions”. Further the Article 2 specified that the Community has a task to promote a “harmonious development
of economic activities” and “a continuous and balanced expansion”. In the main body of the Treaty, the issue promoting more even level of economic development among member states or their regions was largely
addressed indirectly. The issue was linked to a series of provisions concerning specific sectoral policies such as agriculture, transport, and state aid. The only financial instrument created to directly promote regional economic
development was the European Investment Bank (EIB). EIB got the task of granting loans “which facilitate the
financing of projects for developing less developed regions”.
From such initial orientation and description of the EU policy, which is today officially called the “EU Cohesion
Policy”, some useful conclusions might be drawn. The conclusions should lead to a better understanding of
different concept, supports and rejections of the policy, terminology related to the policy and some other issues
which are especially heated when new financial perspective of the EU is in preparation or when general
changing of the members of the EU institutions is relevant. In debates around the EU cohesion policy there are
always groups of strong supporters and of strong adversaries. One of the biggest problems for supporters of the
EU Cohesion policy is the difficulty in generating a credible economic case for the policy. Proofs of the positive
policy results, based on conclusive statistical evidences, are difficult to obtain or create. After more than fifty
years of intervention policy focused on reducing regional development differences and increasing economic
cohesion, its actual contribution to economic development, growth and economic cohesion remains a constant
element of debates and uncertainty (Bachtler J., and Gorzelak G., 2007). Some theoretical and factual results
created by the trade liberalization practices and in the economic integration offer at least some additional
reasonable arguments to the supporters of the cohesion policy. The Cohesion Policy issues are not relevant only
for the assessments of its economic relevance and efficiency, they are strongly affecting political positions and
interests of the member states and their representatives. When the European Regional Development Fund
(ERDF) was first created in the mid-1970, the German Chancellor Helmut Schmidt commented that there was
nothing much of Community interest in the policy and that it “lay firmly in the hands of national governments”
(Bulmer S., and Paterson W., (1987), p. 202).
In this paper we will present a theoretical discussion on why cohesion policy in the EU is in fact necessary in
spite of some general rejections based on beliefs in the efficiency of unregulated economies and markets. The
free trade effects confirm the need for cohesion policy on a theoretical background, with already mentioned
difficulty to offer a completely credible statistical proof. As is evident from the content of the Treaty of Rome,
the Cohesion policy has to serve different objectives. The objectives range from regional development to
structural economic changes and are combined by the task of reducing the economic development differences. A
broad specter of the policy objectives created different understanding of the policy nature in the past. That led to
different names used to describing the policy and to different financial instruments used in the policy realization.
However the Cohesion policy affects the economic structure of the member states, changes the cohesion level,
and helps regions to improve economic structure and reduce development level differences. All of these impacts
are part of economic development process in each national economy. The EU policy funds used – absorbed – on
the member’s economy level represent an important support to structural change, general development – growth
and improved cohesion with the other EU regions and member states. The case of Slovenia using the EU
cohesion funs gives some evidences of the effects and problems related to the EU cohesion policy realization.
66
2. TRADE LIBERALIZATION, BENEFITS AND STABILITY OF THE ECONOMIC INTEGRATION
ENVIRONMENT
The majority of international trade theories prove that international trade in reality does not and cannot create
equal size of economic benefits to all nations involved. Furthermore international trade creates uneven
distribution of actual welfare benefits among different groups of the economic subjects belonging to each
national economy (Leamer, 1995). Let us leave the concept of “welfare trade benefits” unspecified, but accept
theoretically backed suggestion that such welfare benefits based on trade actually exist. The actual uneven
welfare distribution among nations and among people (subjects) and companies, suggests that those who are
“granted” by smaller level of welfare trade benefits, will try to change trade effects into their favour. We know
that such reasoning leads to trade regulation, protectionism or some other way of trade limitation.
Traditionally the economic integration agreements (Regional Trade Agreement – RTA - by the WTO
terminology), which started to be more used only after the WW II (Fig. 1.), and which “exploded” in their
numbers after the last decade of the twentieth century, as a matter of fact increased “selected” trade
liberalization, created faster and more visible differentiation of welfare trade benefits distribution among partner
nation and within the individual nation.
Figure 1: Evolution of Regional Trade Agreements in the world, 1948-2014
Source: WTO Secretariat; The WTO Chart; ▬ cumulative RTAs notification, ▬ cumulative active RTAs
Too fast and too big differences in the actual creation and distribution of the trade benefits, following the
implementation of any economic integration agreement, might lead to the destabilization or even to the
destruction of such agreements among the participating states. In the EU enlargement practice, the increasingly
long accession periods with asymmetric trade liberalization process are used, to offer possibilities of reducing
the problem of too large uneven trade benefits distribution.
In the world economic history the cases where the trade benefits among integration nations were too unequally
distributed are evidenced. On one side such situation leads to a relative small size of actual trade among
integrated partner. Or on the other side, uneven trade benefits distribution led to breakups of the integration
67
agreements and in some extreme cases – in Africa for example – to wars among previously integration-partner
states. The description of reasons to relative limited success of the economic integrations in Africa (De Melo, J,
2013) supports the reality of uneven trade benefits distribution within the economic integration. Beside general
reasons for unequal benefits distribution there are some other specific (as in the case of African states) that can
create additional inequalities in the benefits distribution:
- large cost differences among integrated states can result in welfare-decreasing trade diversification,
- low trade complementarity between members of an economic integration (RTA) generally leads to low
level of trade among integrated partners,
- prevalence of high non-tariff barriers among the integrated countries create high trade costs and reduce
trade volumes and values. Such negative trade impacts are often enhanced by transport infrastructure
under development and by related high cost,
- high degree of differences in economic development levels combined with cultural and other diversities
among members of economic integration results in strongly diverse interests to trade and cooperation.
Some of the above mentioned specifics which further support uneven trade benefits distribution among
integrated states, together with the trade growth limitations, might be seen even in the EU of the 28 member
states. It is therefore not the most suitable environment for the efficiency of economic integration economies,
which creates different pressures among member states, especially during the periods of non-prosperous
economic growth. To prevent negative developments among integrated nations in case of highly uneven
distribution of trade and integration effects and benefits, the implementation and use of a specific compensating
mechanism might be one of the acceptable solutions.
Among other evidences of recent negative developments in relations among integrated nations is the growing
skepticism among EU nations and citizens. The integration efficiency skepticism could be at least partly due to
the negative impacts created by unequal distribution of trade benefits. The forecasts for this year’s European
Parliament elections further show the growth of EU-skepticism. Present developments of EU-skepticism
obviously could not be attributed solely to differences and problems created by unequal distribution of the trade
benefits. However reality of the uneven trade benefits distribution effectively adds to the general enhancing and
visibility of other EU problems and negative economic developments. Combination of uneven trade benefits
distribution and of other EU effective functioning problems no doubt negatively affect EU nations and their
citizens especially in the last five years of economic crises.
Today the EU is one of the largest and most efficient economic integrations globally. As in the theory, similarly
in the case of the EU, the distribution of benefits created by free internal trade among 28 member states is not at
all symmetric. Asymmetry of the free trade benefits distribution among member states is at least partially caused
by impacts that are explained in the majority of trade theories known today. Indirect proof of such asymmetries
existence, at least for a limited period, is shown by the accession period practice where EU opens its market at
the start of the process and candidate countries reciprocate only gradually. It is obvious that some part of
different trade benefits distribution within the economic integration is based on other and not on trade effects. As
shown above, partially differentiated integration benefits realization is based on specifics in national
productivity, market size, price and income elasticity, domestic markets supply and demand characteristic, etc.
The point however is that a part of the benefits distribution differences is directly created by the trade
liberalization. Accession period of the new member, with all asymmetric integration specifics, could not entirely
neutralize the actual difference of trade benefits distribution contained in the essence of the new liberalized trade
developments. Similarly for the “old” EU members the trade benefits are unequally distributed from the
beginning of the economic integration. That is reflected in the Treaty of Rome preamble and some articles
referring to the issue of cohesion of the regions.
68
In the interest of the economic integration stability the following question is relevant. What to do so that the
inequalities in benefits distribution generated by trade liberalization within the integration, would not lead to
instability or even destruction of the economic integration agreement? The trade benefits are materialized
through differentiated achievements like; higher economic growth, additional GDP per capita increase, higher
employment level, etc. The solution of the unequal benefits distribution problem, caused by liberalization of
trade among integrated nations, is in fact conceptually rather simple. Specific compensation mechanism within
the economic integration has to be developed and implemented. Such mechanism should help those with less
trade benefits, to improve their economic potential to grow faster and to gain more trade and economic benefits
within the economic integration. The compensation mechanism should help in reducing large disparities between
trade benefits distribution. The reduced disparities support process of increasingly similar economic growth
levels among integrated nations or in other words lead to higher level of cohesion among regions and member
nations. In the EU such compensation mechanism has developed in the form of the EU Cohesion Policy. EU
Cohesion policy from its early beginnings to today is focused on increasing cohesion. In the essence Cohesion
Policy has to compensate, among other, for the integration’s impacts, as well for the impacts of the unequal trade
benefits distribution.
“The objective of reducing disparities between development levels across the EU's various regions, which is a
key characteristic of economic and social cohesion policy, first appeared as early as the Preamble to the Treaty
of Rome (1957). Yet it was not until almost thirty years later, in the Single European Act (1986), that economic
and social cohesion was finally included as a specific objective in itself along with the objective of achieving the
single market. This policy area was formally institutionalized in the Treaty of Maastricht (1992).” (Subsidiarity,
p.1)
Economic and social cohesion policy in the EU was and should be the mechanism, which among other
integration’s impacts, compensates for the trade-based unequal economic distribution of benefits, together with
economic development differences from the period before the integration. Economic development differences in
fact alone might destabilize the economic integration cooperation and further accelerate uneven distribution of
integration-induced trade growth benefits.
The cohesion policy is no doubt necessary for the economic integration sustainability. But in spite of such a firm
statement, there are certain dilemmas. One dilemma is how big the EU cohesion support should be that it would
not cause negative impacts for competition and efficiency. Further the dilemma is how much and who is going to
give up part of the benefits – financial resources – to support those with lower development and benefits levels.
And the last big dilemma is how the cohesion supports will be used – absorbed – by member states with less
trade benefits and with lower level of economic development. Above all of these issues is a major doctrinal
dilemma whether any kind of economic intervention in the market economy could or could not be acceptable.
The dilemma in the sense of the EU economic efficiency, stability, and sustainability is the following. Is it better
to refrain from intervention to alleviate market’s determined unequal distribution of trade benefits and other
economic integration’s benefits, or it is better to intervene and so to implement the cohesion policy measures.
The dilemma concerning the options of using or not the regulation – compensation – EU mechanism(s) for
successful integration’s functioning could be solved by comparing the actual and expected national key
economic and social objectives realization results. The first objective of the EU is the stability and functioning of
the integration agreement in the way that it creates and sustains peaceful relations among nations in Europe.
With such accepted political and economic primary objective of the EU, the cohesion policy is acceptable and
necessary. It helps to improve the stability of the economic integration, which is endangered by the actual and
unavoidable uneven distribution of trade liberalization, and other benefits among member states and within the
member states societies. The uneven distribution of benefits within the EU members societies created by trade
liberalization and other integration effects, and especially in the periods of economic crisis, similar to the present
69
one, calls for more solidarity and on the EU level for more intensive and broader use of the cohesion policy
attitudes and instruments.
Different implementation and efficiency dilemmas related to the EU cohesion policy could at least partially be
described by presenting the cohesion policy implementation specifics and its results in the case of the economic
development in Slovenia after 2004. Additionally and overview of some EU cohesion policy implementation
specifics might help to enhance the understanding of the concepts and procedures of the EU cohesion policy
realization.
3. TERMINOLOGY AND INSTRUMENTS OF THE EU COHESION POLICY – AN OVERVIEW
This paper addresses a selection of arguments for the EU Cohesion Policy implementation with description of
the specifics related to the economic and business projects of EU co-financing practice in the member states. The
EU co-financing is implemented based on relevant EU policies realization and according to the abilities of the
member states, their regions and of other subjects to utilize the potentially available – allocated - EU financial
resources. Relation between the amounts of EU allocated financial resources for a members states and its ability
to utilize –use – the allocated amounts is often referred as the national „absorption capacity “. Understanding of
the concept and results related to the absorption capacity issue requires understanding of the economic and
political concepts, together with the sources and terminology used in the process of the EU funds allocation and
use. On the other side, the level of absorption in the member states depends on national abilities and specifics to
provide required resources and organization necessary to establish the optimal absorption capacity for the EU
allocated funds. On the national level, absorption of the EU allocated funds has to be, by the definition, based on
the national and regional economic development objectives. Such aspects of the absorption capacity will be
linked further to the case of Slovenia as the EU member country from 2004.
The EU funds' support for the regional and national economic development of the member states have long
conceptual and practical implementation history. The structure, terminology and all other aspects of the
instruments used to support cohesion growth as described in the probable of the Treaty of Rome (1957) had to
evolve and change, following the requirements created by the ongoing deepening and enlarging processes of the
EU. On such basis it is understandable that one might not always see a clear and consistent use of the terms
describing activities that support the EU regional and national cohesion growth. A short overview of the
terminology and concepts used in the area of the cohesion policy helps to reduce possible misunderstandings and
mistakes important for the nation’s absorption capacities of the relevant EU funds. In such perspective the
relevant questions are:
1. Which EU policy (policies) is (are) constituting the framework that defines the maximal absorption amount
for the member state?
2. Which EU resources – financial instruments – through the allocation processes create the potential size –
financial amount - of a Member state‘s potential „absorption capacity“?
In relation to the first question we might use the following EU description of the cohesion policy which was
created by the European Parliament (EP); “The EU’s Cohesion Policy provides the framework for a wide range
of programmes aimed at increasing economic growth and social cohesion and reducing disparities among the
Member States (MS) and their 270 regions. Perceived weaknesses of the current arrangements include
complexity, inadequate integration with other policies, and low absorption rates in many MS.” (EU Cohesion
Policy 2014-20). Referring to the complexities and inadequate integration with other EU policies when EP
speaks about Cohesion policy might be equally observed in some other EU institutions’ statements published in
different EU information sources. Due to the complexity of the issue, the terminology and explanations
describing the EU financial resources and objectives of the cohesion policy could often not be consistently
70
applied or explained. Inconsistent use of terms and explanations of the cohesion policy could sometimes be
observed even in different statement of the same EU institution, as in the case of the EP itself. Such
inconsistencies - although understandable due to the complexities of the issues and due to constant changes of
the EU - should however be kept to the minimum, otherwise they serve to support further doubts of the positive
perception of the EU impacts among member states and among their populations.
As mentioned above, the concept, terminology, and resources for the EU cohesion policy have been
continuously changing. This continues to be true for the period of 2014-2020. In the information on the Cohesion
Policy for the period 2014-2020 is the following observation that proofs continuity of the changes in the EU
environment defining the framework for the national absorption capacity. “The new legislative package (for the
EU Cohesion Policy 2014-20 – authors addition) includes a new single overarching regulation setting out
common rules for the European Regional Development Fund (ERDF), the European Social Fund (ESF), the
Cohesion Fund, the European Agricultural Fund for Rural Development, and the European Maritime and
Fisheries Fund. In addition, the package has specific regulations for the ERDF, the ESF, the Cohesion Fund, the
European Territorial Cooperation Goal, the European Grouping of Territorial Cooperation, the Globalization
Adjustment Fund, and the EU Programme for Social Change and Innovation.” (EU Cohesion Policy 2014-20).
Although we could understand the numerous changes in the cohesion policy scope, terminology and in its
financial resources in the past – based on enlargement and deepening processes - we are a bit confused by
looking at the terminology and financial instruments related to the actual cohesion policy description and
interpretation based on different EU institutions’ available contemporary information sources. The instruments –
resources - used to support the EU cohesion policy in the period of 2014-020 are, following the above EU
Parliament listing, different Funds, Goals, and Programs. The problem however is that another EP information
source - EP News - defines the cohesion policy differently and narrower in the sense of its resources. According
to EP News the cohesion policy can be described in the following form. “Cohesion policy is the EU's main
common investment policy tool. Often referred to as "regional policy" it provides vital basic financial support for investing in regions of the EU, thus helping to create jobs and boost economic growth”. (Cohesion policy
background note, p.1) After stating that the cohesion policy is the EU investment policy the same document
states that EU cohesion policy is simultaneously the EU regional policy. Eventual confusion made by using the
same instruments as part of the two different policies is eventually reduced by using a quote. But further the EP
News information reduces the number of financial instruments that are – or can be used to realize the cohesion
policy according to the above quoted description of the new EU legislative package covering the resources to be
used in that scope. Opposite to a broad definition of the cohesion policy resources the EP News information
reduces the scope of the potential resources. “The money for cohesion policy projects comes from three funds,
also called "European Structural and Investment Funds" (ESIF). These are the European Regional Development
Fund (ERDF), the European Social Fund (ESF) and the Cohesion Fund.” (Cohesion policy background note,
p.1). The finances contained in only three EU funds, as suggested by the EP News information statement, thus in
reality constitute the total amount for the potential member state’s absorption capacity of the EU allocated
cohesion resources. The finances available to the Member States in Funds like Agriculture or Maritime and
Fisheries Fund, together with some other EU resources mentioned above, and including the European Investment
Bank (EIB) loans, are not part of the available finical resources that can be used for the cohesion policy
realization. Only the funds allocated to member states through the mentioned three EU Funds are creating the
potential for the absorption capacity realization. In the further text absorption capacity will be based on utilizing
the allocated financial resources through the mentioned three EU Funds, which are, based on definitions the only
available finances for the cohesion policy realization on the national bases.
The three EU Funds providing the financial resources for cohesion policy and the framework for the maximal
potential level of absorption, are related to another development issue. The three Funds together are called
European Structural and Investment Funds (ESFI). The idea within such description of the Funds is related to the
change of the economic structure in the member states based on investments coming from the three cohesion EU
funds. Analyzing absorption capacity of member states in Section 4, specifically for Slovenia, is going to leave
71
aside the confusion created by the fact that the same EU Funds, according to the quoted EP News information,
could serve for three different EU Policies – the cohesion policy, the regional policy and the structural policy.
The absorption capacity of the EU nations is depending on the EU conditions, allocation, and rules, and
additionally on a number of national economic, financial, and organizational specifics. National specifics make
the relations between available EU funds and effective increase of cohesion following the absorption of the EU
funds, somehow more complex and not so straightforward.
Among the issues relevant for absorption complexities is the dilemma about the adequate size of the EU
resources available for individual member’s absorption. Simple suggestion could be that more EU resources
could lead to higher absorption and to increased cohesion. Discussing the absorption abilities and effects of
absorbing the EU Funds on the member state’s level so often leads to ideas that increased levels of available
financial amounts will automatically lead to higher level of the absorption and further again automatically to the
higher cohesion. Further such ideas suggest that with more of financial resources available through the EU Funds
could support the member state’s structural changes, based on realized regional and general cohesion policy
objectives, more effectively. In reality utilization of finances available through the three EU Structural and
Cohesion funds as a whole, shows that more funds do not always mean more growth. „For EU Structural Funds
as a whole, more funds do not mean more growth. A point is reached where returns begin to decline and
additional funds do not lead to higher growth. Transfers to regions should therefore not exceed maximum
desirable levels if inefficiency and misuse are to be avoided“(Becker, 2012, p. 1). The question related to the
issue of adequate levels of financial resources for cohesion policy objectives realization obviously is not easy to
be answered. Except for the theoretical knowledge, the data on absorption capacities are not an indicative to
where available amounts of cohesion resources reach the growth efficiency turning point. Due to big differences
among the GDP per capita levels of EU member states, the issue of adequate amounts of cohesion resources is
further complicated. Relatively high national contribution to the EU budget in relation to the GDP per capita
especially in less developed EU member states (Figure 2) create an environment where in fact it is impossible to
judge what could be the appropriate absorption level to create optimal growth rates.
Figure 2: The uneven GDP per capita “burden” of the national EU budget contributions in 2011
Source: Becker, 2012, p.3
72
On the EU budget revenue side, contributions of the member states are non-progressive, i.e. poorer, and richer
member states alike contribute roughly 1% of their GNI to the EU budget, as shown in Figure 2 for the budget
year 2011. The idea of the “burden” is that 1% of contribution from a low level of the GDP per capita is a bigger
burden for the economy as 1% in the case of a state with higher GDP per capita level. On the EU budget
expenditure side, poorer countries overall receive more – have relative more allocated funds - than richer
countries. However there are two problems. Firstly, contributions to the EU budget although roughly on the 1 %
levels of GNI represent a relatively bigger “burden” in relation to the GNI per capita level of the poorer member
states. Poorer nation receive overall more allocated EU funds, which should probably compensate for the non-
proportional “burden” created by budget contributions and as well for unequal distribution of the benefits within
the integration. The second problem of the poorer nations is related to actual ability of utilizing the allocated EU
funds. The larger amount of EU funds allocated in cases of poorer nations in practice directly does not provide
equally higher actual funds’ utilization level. The absorption – utilization of the allocated resources - depends on
the ability of a member state to create proper absorption conditions.
The national absorption conditions for the allocated EU funds depend on a number of elements and only a part of
them is in the member state’s control. Among those out of the member’s control are the national co-financing
ratios, the allocation of the EU funds and decisions about the EU general cohesion policy objectives. Formally it
is true that all member states participate in the process of deciding about the EU general cohesion objectives,
about the procedures to be followed for the allocated funds absorption realization, and about other issues of the
EU cohesion policy realization framework. In the reality the smaller member states have a limited impact in the
process of accepting the decisions shaping the objectives, procedures, and allocation of the financial amounts
within the EU cohesion policy and are forced to form alliances with other countries to increase their bargaining
power. A limited influence on the external elements of the EU funds national absorption conditions creates
certain difficulties in the actual absorption of the cohesion funds at the national level. Absorption capacity
conditions for the cohesion funds on the national level are further limited by the impacts of the national
economic, political, and other internal specifics. Some of such specifics, which reduce the level of the cohesion
funds national absorption capacity, are explained and documented for the EU cohesion funds absorption
realization in the next section for the case of Slovenia.
4. COHESION POLICY IN SLOVENIA
Slovenia has been the recipient of pre-accession assistance from as early as 1992, while full access to cohesion
policy was gained after full membership, first in the 2004-2006 period and later in 2007- 2013 financial
perspective.
4.1. The 2004–2006 programming period
Slovenia became eligible for cohesion policy programmes with full EU membership in 2004, when the amount
of cohesion funds, negotiated for the 2004-2006 period, was €458 million - annually €136 million, of which 52%
was allocated to the Structural funds, 42% to the Cohesion Fund and 6% on the two Community initiatives,
INTERREG and EQUAL. Figure 3 gives allocation of funds.
73
Figure 3: EU cohesion funds in Slovenia for the period 2004-2006 in current EUR and in % of total value
EUR 190.6 m io.
42%
Community Initiatives
EUR 23.7 m io.
30,1
5%
7%
EUR 6.4 m io.
1%
EUR 237.5 m io.
52%
Structural Funds
Cohesion Fund
Community Initiative Interreg
Community Initiative Equal
Source: Wostner, 2004.
The result was not entirely satisfactory since the allocated funds represented only 0.6% GDP, which was the
lowest share among all EU member states (the average per year share of allocations in GDP among the new
member states was 1.93%) and it has also received the lowest per capita aid intensity in objective one – the
convergence regions. Such unfavorable division of assets was the result of separate financial negotiations for the
new Member States whose total available amount of funds has been identified in advance and was therefore
fixed. Among the new Member States, Slovenia was the most developed and consequently it was allocated
comparatively lower aid intensity, which was particularly evident in relation to the EU15 countries.
The eligible use of the Structural Funds in the amount of €237.5 million was determined in the Single
Programming Document 2004-2006, which contained three priorities: (1) promoting entrepreneurship and
competitiveness (52% of funds; intended for facilitating the development of innovative environment, tourism
and entrepreneurship as well as business zones and associated infrastructure); (2) knowledge, human resource
development and employment (29% of funds; to the development of education and training of adults, the
unemployed and employed individuals, to the promotion of social inclusion and also to the improvement of
education and training system), and (3) the restructuring of agriculture, forestry and fisheries (15% of funds; for
the investments in the food processing industry, farms and forests, as well as for rural development in the context
of alternative income sources, healthy diet, and fisheries and aquaculture).
Compared to convergence (Objective 1) regions in other countries, Slovenia's investment in human resource
development was above average (30.6% compared to an average of 23.1% and 20.5% in the old Member States
and new Member States, respectively) and quite comparable with Ireland's (35%), where the focus on human
capital was considered the key to their economic successs at that time. Above average was also the investment
share in the productive sector without basic infrastructure (35% compared to an average of 20.1% and 17.6%),
but below average investment in basic infrastructure (approximately 19% compared to an average of 41.3% and
19.4%) and agriculture (10 7% compared with an average of 13.7% and. 16.2%). Cohesion Fund resources were
used for co-financing of projects in the fields of environment and trans-European networks, with the ratio 50:50
between the two areas already set by the EC.
Despite the relatively limited resources, in absolute terms cohesion funds represented more than 18% of the total
Slovenian budget expenditures for subsidies to businesses and private individuals, including investment
expenditures and transfers. Taking into account the national public co-financing (on average 25%), cohesion
74
funds represented roughly a quarter of »development« expenditures of the national budget, which means that
cohesion policy has actually constituted an important part of development policy in Slovenia at that time.
4.2. The 2007–2013 programming period
In December 2005, the European Council passed the agreement on the Financial Perspective 2007-2013.
Slovenia was still considered as one region and since its development level (GNI per capita) was just below the
75% of EU average, it managed to negotiate €4.2 billion of cohesion funds, which was a substantial increase
relative to the 2004-06 period (even on comparable terms). This meant an average allocation of €600 million a
year or rather between 1.6 and 1.7% of Slovenia's GDP, which represents 5.5% of total gross investment into
fixed assets of Slovenia, by adding the own participation this is further increased by one percentage point. From
the perspective of the national budget, cohesion funds represent between 6.2 and 6.7% of total revenues, but the
true relevance becomes evident on the expenditure side. With regards to public investments, capital transfers
and subsidies, cohesion policy funding accounted for somewhere between 30 and 33% of total expenditures in
the 2008 and 2009 budgets and even increased to 50% by 2012, which means that the cohesion policy actually
became a key actor in the development policy in the Republic of Slovenia.
The comparison of national allocations of cohesion funds is presented in Figure 4. Although with regards to the
share of GDP, the amount of allocations in Slovenia is strictly speaking relatively low (the twelfth highest
share), aid intensity per capita (based on purchasing power parity) is actually the fifth-highest. This difference
occurs due to the relatively high levels of GDP in Slovenia, which means that despite its high intensity the aid it
is not so high in terms of share in GDP. In absolute terms, the biggest recipients of cohesion policy funds are
Poland with €67 billion, next are Spain (€35 billion), Italy (€29 billion), Czech Republic (€27 billion), Germany
(€26 billion), Hungary (€25 billion) followed by Portugal and Greece with the €22 and €20 billion of eligible
spending. In terms of absolute amounts, Slovenia ranks as 16th regarding the largest amount of resources.
Figure 4: National allocations of cohesion policy funds for the period 2007-2013, in EUR (in purchasing
power parity for 2004) per capita and. as a share of GDP
Source: OECD, 2007, p.144
75
In the 2007-2013 period the eligible use of funds is set by the National Strategic Reference Framework
(hereinafter NSRF) which is prepared by Member States and confirmed by the Commission. The more thematic
operational programs (hereinafter OPs) were then prepared which are used as a basis for direct use of the
cohesion funds. Slovenian NSRF targets to "improve the welfare of the Republic of Slovenia by facilitating
economic growth, creating jobs and strengthening human capital as well as ensuring a balanced and harmonious
development, particularly between regions" (Government Office for Local and Regional Development, 2007, p.
72). Special attention is therefore given to promoting growth and job creation (which are also the two key
objectives of the Lisbon Strategy) and to sustainable development.
The commitment to Lisbon expenditures thus represent more than 60% of all available cohesion funds in
Slovenia, which is one of the highest shares among the cohesion policy recipients. Specific objectives of NSRF
are: (1) to encourage entrepreneurship, innovation and technological development; (2) improve the quality of
education and also research and development activities; (3) improve labor market flexibility while ensuring job
security, in particular by creating jobs and promoting social inclusion; (4) to provide conditions for growth by
providing sustainable mobility, to improve the quality of the environment and adequate infrastructure, and (5)
balanced regional development. The basis for the implementation of these objectives are three operational
programs (OPs), the Operational Programme for Strengthening Regional and Development Potentials, the
Operational Programme for Human Resource Development and the Operational Programme for Environmental
and Transport Infrastructure Development. Breakdown of available cohesion funds to individual OP is presented
in Table 1.
Table 1: Breakdown of 2007-2013 cohesion policy funds in Slovenia, by Operational Programme for the
Convergence and European Territorial Cooperation objectives
Amount in EUR
Operational Programme
Fund
(current prices)
%
%
%
OP for Strenghthening Regional
and Development Potentials
ERDF
1,709,749,522 40.7
41.7
63.6
OP for Human Resource
Development
ESF
755,699,370 18.0
18.4
28.1
OP for the
Development of
Environmental and Transport CF
1,411,569,858 33.6
34.4
Infrastructure
ERDF
224,029,886 5.3
5.5
8.3
Trans-border and inter-regional
OPs
ERDF
96,941,042 2.3
Transnational OPs
ERDF
7,315,278 0.2
Total
4,205,304,956 100.0 100.0 100.0
Legend: ERDF - European Regional Development Fund; ESF – European Social Fund; CF – Cohesion Fund
Source: Government Office for Local and Regional Development,, 2007, p. 74.
One third of the total cohesion policy funds (€1412 million) is provided by the Cohesion Fund, the remaining
part is financed by the European Regional Development Fund (€2038 million or 48%) and by the European
Social Fund (€756 million or 18%). The aggregate use of funds by theme for the Convergence objective is
presented in Figure 5. €104 million is intended for cross-border, transnational and interregional cooperation,
51% of the remaining €4.1 billion is allocated to infrastructure development (including economic infrastructure),
around 30% to productive investments and 18% allocated to human resource development.
76
Figure 5: Thematic breakdown of the convergence objective in Slovenia (as % of all allocated cohesion
resources)
Sustainable energy ;
Efficient p ublic sector;
4,0%
2,4%
Education and training
sy stems; 4,1%
Transp ort
Infrastructure; 22,8%
Tourism, culture and
sp orts; 6,6%
Human resources -
active emp loy ment
p olicy ; 11,6%
Entrep reneurship and
Environmental
comp etitiveness; 19,9%
Infrastructure; 13,2%
Develop ment of
Regions; 15,4%
Source: Wostner, 2013.
Below we briefly present the basic logic of Objective 1 Operational Programmes in Slovenia.
Operational Programme for Strengthening Regional Development Potentials (OP RD) aims to create an
»innovative, dynamic and open Slovenia, with developed regions and competitive, knowledge-based economy«
(NSRF, 2007, p. 84) by financing investments in several priority areas (»thematic concentration«): promotion of
entrepreneurship, innovation and technological development as well as balanced regional development. It is
particularly focused on increasing and improving investments in R&D activities as well as the education system.
OP RD finances mainly the productive investments, in particular to enhance the competitiveness of the
Slovenian economy in terms of achieving the Lisbon goals - the promotion of entrepreneurship, innovation and
technological development that would translate into job creation, one of the key goals of the OP. The planned
activities of the OP RD include developmental investment projects, centers of excellence, subsidies and other
forms of financial assistance for small and medium-sized enterprises, in particular the purchase of technological
equipment (€402 million or more than 23.5% of funds allocated to OP RD was planned for the priority
»competitiveness and research excellence), development of economic, information and developmental-
educational infrastructure (23% or €397 million), for priority »integration of natural and cultural resources«,
which were mainly the development of tourism, cultural and sports infrastructure (15.4% or €263 million) and
regional development, to which 36% (€619 million) was allocated. The latter refers particularly to the
construction of infrastructure (economic, transport, educational, environmental, tourism and partly also to social
infrastructure and urban development) on regional level based on the "bottom-up" initiative.
The aim of the Operational Programme for Human Resource Development (OP HR) is to "invest in people
whose capital will secure a higher level of innovation, employability and economic growth, which is the best
way to ensure high employment, social inclusion, reduction of regional differences and high living standard"
(NSRF, 2007, p.93). The programme is focused on strenghthening human capital, creating jobs, encouraging
employment and employability, promoting innovation and thus the competitiveness of the economy by investing
77
into research and other personnel, life-long learning, promoting social inclusion and equal opportunities and also
to increase the effectiveness of the public sector (through projects such as e-government, e-justice and e-health).
The biggest share of resources, almost €262 million (35% of resources for OP HR) aim at promoting
entrepreneurship and adaptability (e.g. young researchers, scholarship schemes, self-employment, co-financing
of company training), followed by human resource development and life-long learning which aims to modernize
the educational system and training (22% or €165 million of OP HR funds), while the promotion of
employability of job seekers and inactive have been allocated 18.5% (140 million). Additional €64 million were
dedicated to the promotion of equal opportunities and reinforcing social inclusion and 13% (€97 million) to
enhancing institutional and administrative capacity, especially in public, but also in the non-governmental sector.
The aim of the Operational Programme of the Environmental and Transport Infrastructure Development
(OP ETID) is "to ensure conditions for growth by providing sustainable mobility, improving environment quality
and appropriate infrastructure." (NSRF) OP ETID almost exclusively finances the construction of infrastructure
in the field of environment and transport, which is related to the promotion of sustainable development and job
creation as well as ensuring high quality of living. The transport part has €915 million (56% of funds for OP
ETID) at its disposal for the purposes of constructing railway and port infrastructure, highways and state roads.
A small part of the funding is reserved for the aviation and airport infrastructure and also for the public transport.
As for the the environmental section, projects with a clear benefit for the environment are supported: €531
million (33%) is forseen for projects related to municipal waste management, disposal and treatment of
municipal waste water, drinking water supply and reducing the water damages. Sustainable use of energy is the
sixth priority of OP ETID, with €160 million (10% of resources) assigned, intended for energy rehabilitation
and sustainable construction of buildings, efficient electricity use, innovative measures for local energy supply
and demonstration projects.
4.3. Absorption capacity in Slovenia
Absorption capacity can be defined as »the extent to which a Member State and its regions are able to spend the
financial resources allocated from the Structural and Cohesion Funds in an effective and efficient manner”
(Theurer, 2011).
Similar to other new member state, Slovenia performed well with regards to absorption in the 2004-06
programming period, achieving practically full absorption with average absorption rate of 99.3% for the
Structural funds and 104% for Cohesion fund at the end of the period (cumulative payment execution of the
2004-2006 allocations reached 100% in 2012). In part, the success was due to the government decision to
overcommit in 2006, which means that contracted grants exceeded the funds allocated to Slovenia, thus
providing “reserve” projects to ensure utilization of all funds in case implementation of some projects fails.
For the 2007-2013 programming period, the budgetary execution data is provided by the managing authority,
Government Office for Development and European Cohesion Policy (2014): on 31.3.2014, approved instruments
in Slovenia reached 119.26% of available funds while the share of signed contracts was 95,42 % of available
funding. €2.75 billion (95.98% of available funds) were payed out to beneficiaries from the national budget
while the EC-certified expenditures amounted to €2.55 billion (62.2 % of allocated funds). The absorption
performance varies substantially among the funds, with 77.5% of funds claimed back from the Commission from
ERDF, 70% ESF and 41% from CF.
78
Figure 6 presents the progress in absorption by intervention type at the end of 2012. As reported by KPMG
(2013, p.43), with the exception of the technical assistance–related operations, the contracting ratio (ratio
between the signed contracts and allocated funds) for Slovenia ranks above the CEE1 average in human capital (at 90%) and also in R&D and ITC (at 87%). The share of signed contracts is lowest for infrastructural
operations (46%) but was still slightly above half of the CEE average. Although absorption for infrastructural
projects seems to the main cause of concern both in Slovenia and in the entire region, Slovene payment rate of
27% for infrastructure-related operations ranks only second-to-last among the CEE countries. A brighter picture
emerges in R&D and ITC operations, where Slovenia has the highest payment ratio in the entire CEE region.
Figure 6: Contracted grants by intervention type (in %) in Slovenia for the 2007-2013 programming
period
Source: KPMG, 2013, p.48.
4.4. Absorption problems
Absorption of structural and cohesion funds seems to be a long-standing concern among the member states,
translating in the pre-2007-accession period to a fear that the new member states will lack both the required
administrative capacities as well as enough high-quality projects to be able to use much of the allocated funds.
In practice, however, it turned out that new member states fared relatively well in terms of absorbing cohesion
funds during the 2004-06 period. Some »teething problems« were recognized at the beginning but countries were
quick in adapting to the more successful practices, learning, according to Rosenberg and Sierhej, 2007, p.11, that
initial frameworks were over-regulated, often to prevent the misuse of EU funds and also that absorption seemed
to be faster in countries with a strong central managing authority. By the end of 2012, the cumulative allocation
for EU-10 at the global level reached 99.3% of their 2004-06 allocation while the EU-15 rate stood at 97.4% of
their 2000-2006 allocation (European Commission, DG Budget, 2013, p.5).
1 For the purposes of the report, CEE region refers to countries which are both part of Central and Eastern European region and EU members states: Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, Slovenia.
79
During the 2007-13 period, member states continue to strive at absorbing as much financial support as possible,
where the absorption capacity can be determined by (Theurer, 2011, p.14):
1. Ability by project applicants to generate acceptable projects (the demand side);
2. Macroeconomic absorption capacity in terms of GDP;
3. National financial absorption capacity to co-finance the programs suppported by the EU
4. National administrative capacity of central and local authoroties.
By April 2014, Slovenia's share in 2007-13 allocations payed by the Commission rose to 63.8%, which puts
Slovenia in the 14th place among the Member States (with best performers being Estonia, Portugal and Lithuania
(European Commission, DG Regio, 2014), nevertheless, the gap between the expenditures from the national
budget and the amount claimed from the Commission is still a big concern and it is not difficult to relate, at least
to some extent, to most, if not all, reasons behind the absorption problems on Theurer's list (2011, p.14-16):
1. Initial problems at the beginning the programming period applied to all member states due to: i)
parallel implementation of two programming periods and ii) member states having difficulties over
completing the compliance assessment procedures concerning the new management and control system.
Slovenia was no exception with low expenditures from the national budget and practically no
intermediate payments from the Commission during 2007 and 2008. This problem was especially felt in
policy areas where there were implementation difficulties in the first period, leaving larger amounts of
funds to be spent at the end of the period – in case of Slovenia ESF was such an example.
2. Financial problems were in many member states caused or at least exaggerated by the global economic
recession. It became not only more difficult for countries to find resources to co-finance projects due to
budgetary restraint measures applied to many public budgets, but also for firms to find liquidity
financing. A far as the former is concerned Slovenia needed to introduce new flexibility instruments in
the national budget to accommodate for the increased cofinancing pressure, while for the latter advance
payments were newly allowed after the outbreak of the crisis.
3. Regulatory requirements present an important burden for member states due to incompatibility of EC
requirements with the existing national arrangements as well as in coping with the changes and
interpretations of the regulations.
4. Organisational requirements can cause difficulties for the member states in the sense of hierarchy,
cooperation and communication problems between institutions, difficulties over the allocation of tasks
and responsibilities, the need to establish new institutions. Slovenia has, from the beginning, adopted a
centralised approach for the implementation of cohesion policy, with a central managing authority to
coordinate between the ministries involved in the implementation system (implementing instruments in
their specific fields), but level and quality of coordination varies with the political cycle. As reported in
Wostner (2013, p.8), Slovenia has failed to develop a strong managing authority due to staff turnover as
well as weak political positioning within the government, which often prevented effective intervention.
The frequent moving of the coordinating body within the government was certainly not helpful in
improving the efficiency of CP management but rather increased costs in terms of time and resources.
5. Human resources. Theurer (2011, p.7) reports limited staff numbers, inadequately trained staff at the
national and regional level, and difficulties with staff retention as one of key reasons for absorption
problems in member states. Slovenia is no exception with relatively high staff turnover, limited staff
numbers, frequent changes of the responsible ministers. The staff turnover was even shown to be
directly assotiated with the absorption rate by Wostner, 2013, p.12-13.
6. Information technology systems pose problems in many member states, including Slovenia. In 2012, the
Court of Auditors of the Republic of Slovenia assessed the information system introduced by the
managing authority as one of the key problems in the implementation of the Cohesion Policy. The data,
which the managing authority retrieved from the information system, were not complete and sometimes
80
also incorrect. The court concluded that such data cannot represent adequate basis for sound
management and monitoring of the Cohesion Policy implementation.
7. Control requirements represent a significant burden, especially taking into account scale of the projects
with smaller ones being disproportionately affected. Due to risk-aversion, national procedures tend to
be over-complicated and over-strict and thus “deflect attention from content and impact”. Beneficiaries
can even be deterred by control requirements, which can be very time-consuming, especially when they
are considered as being unnecessarily introduced due to national considerations. On the other hand
Slovenia has experienced interruption of payments from the Commission, which points to problems in
ensuring legality and regularity of cohesion policy implementation, in turn resulting in an even higher
administrative burden for the beneficiaries, with a wide-ranging system of supervision and control.
5. CONCLUSION
With 28 member states, today European Union is one of the largest and most efficient economic integrations
globally. As in the theory, similarly in the case of the EU, the distribution of benefits created by free internal
trade among 28 member states is not at all symmetric. Indirect proof of such asymmetries existence, at least for a
limited period, is shown by the accession period practice where EU opens its market at the start of the process
and candidate countries reciprocate only gradually. But this could not entirely neutralize the actual difference of
trade benefits distribution contained in the essence of the liberalized trade developments. Specific compensation
mechanisms must therefore be in place to help those with less trade benefits, to improve their economic potential
to grow faster and to reap more trade and economic benefits within the economic integration. In the EU such
compensation mechanism has developed in the form of the EU Cohesion Policy. In the essence Cohesion Policy
has to compensate, among other, for the integration’s impacts, as well for the impacts of the unequal trade
benefits distribution.
“The objective of reducing disparities between development levels across the EU's various regions, which is a
key characteristic of economic and social cohesion policy, first appeared as early as the Preamble to the Treaty
of Rome (1957). Yet it was not until almost thirty years later, in the Single European Act (1986), that economic
and social cohesion was finally included as a specific objective in itself along with the objective of achieving the
single market. This policy area was formally institutionalized in the Treaty of Maastricht (1992).” (Subsidiarity,
p.1).
Slovenia has been the recipient of pre-accession assistance from as early as 1992, while full access to cohesion
policy was gained after full membership, first in the 2004-2006 period and later in 2007- 2013 financial
perspective.
For the 2004-06 perspective, Slovenia was allocated €458 million, which was, with the ratio of 0.6% GDP per
annum, the lowest share among all EU member states. Compared to convergence regions in other countries, with
30.6% Slovenia's investment in human resource development was above average and quite comparable with that
of Ireland, where the focus on human capital was considered the key to their economic successs at that time.
Regarding absorption capacity, Slovenia achieved practically full absorption with average absorption rate of
99.3% for the Structural funds and 104% for Cohesion fund at the end of the programming period.
For the 2007-2013 programming period, Slovenia managed to negotiate €4.2 billion of cohesion funds, which
meant an average allocation of €600 million a year or rather between 1.6 and 1.7% of Slovenia's GDP.
Compared to other member states, aid intensity per capita was the fifth-highest. With regards to public
investments, capital transfers and subsidies, cohesion policy funding accounted for around 50% by 2012, which
means that the cohesion policy actually became a key actor in the development policy in the Republic of
Slovenia. The current absorption rate (data for 15.4.2014) regarding payments from the Commision is 63.8%,
81
which puts Slovenia in the 14th place amongh the EU member states. The absorption performance varies
substantially among the Funds with the Cohesion fund performing the worst. For example, compared to
countries of the Central and Eastern European region, Slovene payment rate of 27% for infrastructure-related
operations ranks only second-to-last. A brighter picture emerges in R&D and ITC operations, where Slovenia
has the highest payment ratio in the entire CEE region.
A big concern regarding Slovenian absorption rate is the gap between the expenditures from the national budget
and the amount claimed from the Commission and it is not difficult to relate to the reasons behind the
absorption problems identified by Theurer (2011) fo the 2007-13 financial perspective: initial problems at the
beginning the programming period, financial problems, regulatory requirements, organisational requirements,
human resources, information technology systems, and control requirements. And even though the Court of
Auditors of the Republic of Slovenia, in its 2012 report, assessed the information system as one of the key
problems in the implementation of cohesion policy in Slovenia, we argue further that a stable implementation
system as well as stable staff structure are the two missing key elements to a cohesion policy success story of
Slovenia.
REFERENCES
Bachtler, J., and Gorzelak, G., (2007), Reforming EU Cohesion Policy, in Policy Studies, Vol. 28:4, p309.
Becker, Sascha O.; EU Structural Funds: Do They Generate More Growth? The CAGE–Chatham House Series, No. 3, December 2012,
http://www.chathamhouse.org/sites/default/files/public/Research/International%20Economics/1212bp_becker.pdf
Bulmer, S., and Paterson, W.; The Federal Republic of Germany and the European Community, London, Allen and Unwin, (1987) Cohesion policy background note; Regional policy − 08-07-2013 - 16:13, http://www.europarl.europa.eu/news/en/news-
room/content/20130708BKG16812/html/Cohesion-policy-background-note
Court of Audit of the Republic of Slovenia (2012). Audit of European Cohesion Policy implementation system in the Republic of Slovenia for the programming period 2007-2013. Ljubljana: Court of Audit of the Republic of Slovenia
De Melo, Jaime; Regional trade agreements in Africa: Success or failure? http://www.ideasforafrica.net/articles/regional-trade-agreements-
africa-success-or-failure#sthash.3AnWrCZ2.dpuf , 9th December 2013
EU Cohesion Policy 2014-20; Posted by EP Library ⋅ November 15, 2013, http://epthinktank.eu/2013/11/15/eu-cohesion-policy-2014-20/
European Commission, DG Budget (2013). Analysis of the budgetary implementation of the Structural and Cohesion Funds in 2012.
Brussels: European Commission.
European Commission, DG Regio (2014): EU Cohesion Funding –
Key Statistics
[http://ec.europa.eu/regional_policy/thefunds/funding/index_en.cfm]
Government Office for Development and European Cohesion Policy (2014): Poročilo o črpanju sredstev evropske kohezijske politike 2007-2013. Ljubljana, april 2014
Government Office for Local and Regional Development (2007): Nacionalni strateški referenčni okvir 2007–2013. Ljubljana: SVLR.
KPMG (2013): EU Funds in Central and Eastern Europe.
[http://www.kpmg.com/CEE/en/IssuesAndInsights/ArticlesPublications/Documents/eu-funds-in-central-and-eastern-europe-2012.pdf]
Leamer, E.E.; The H-O Model in Theory and Practice, Princeton Studies in International finance, No. 77, Feb. 1995.
OECD (2007): Economic Surveys of the European Union 2007. Paris: OECD.
Rosenberg, C.B., Sierhej, R.(2007). Interpreting EU Funds Data for Macroeconomic Analysis in the New Member States. IMF Working Paper no. 07/77.
82
Subsidiarity monitoring in the area of economic, social, and territorial cohesion;
https://portal.cor.europa.eu/subsidiarity/policyareas/Pages/EconomicandSocialCohesion.aspx
The WTO chart of the RTAs growth; http://www.wto.org/english/tratop_e/region_e/regfac_e.htm
Theurer, M. (2011). Report on absorption of Structural and Cohesio Funds: lessons learnt for the future cohesion policy of the EU. Brussels: European Parliament.
Wostner, P. (2009). Potrebnost in uspešnost kohezijske politike EU. Ljubljana: Ekonomska fakulteta.
Wostner, P. (2013). Cohesion policy in Slovenia over the last 20 years. [http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2261803]
Wostner, P. (ed.), Aljančič, M., Klužer, F., Juvančič, L., Medved, E., Kušer, N. (2004). Kako do denarja EU. Ljubljana: GV Založba.
83
Sonja Šlander1
Faculty of Economics, University of Ljubljana
Tej Gonza2
Erasmus University of Rotterdam
Katja Zajc Kejžar3
Faculty of Economics, University of Ljubljana
EVALUATION OF EU COHESION POLICY: LESSONS FROM SLOVENIAN CASE
ABSTRACT
The impact of Cohesion policy is being regularly evaluated by the European Commission, responsibility for
evaluation, however, also lies with the Member States. We present Slovenia as a successful case of such
triangulation-based Cohesion policy evaluation following a recommendation from the Commission that
“Whenever possible, evaluation questions should be looked at from different viewpoints and by different
methods”. Based on Slovenian CP evaluation experiences we argue further that the basic triangulation-based
process should be modified with an additional feed-in mechanism. For an exposition, results are presented from
the Slovenian case where the estimates on R&D spillover elasticities are obtained with microeconometric,
counterfactual methods and fed directly into DG Regio’s Slovenian model within CSHM system to obtain more
accurate CP impacts on macroeconomic variables.
Keywords: European cohesion policy, evaluation, triangulation, counterfactual methods, HERMIN
JEL classification: R10, C54, O52, R58
1. INTRODUCTION
Cohesion policy represents financially the second strongest policy area in the EU budget. Its objective is to
contribute to reduced development disparities and at the same time to promote growth across the European
Union. The impact of cohesion policy (CP, hereafter) is being regularly evaluated by the European Commission,
responsibility for evaluation, however, according to Council Regulation 1083/2006, also lies with the Member
States. As suggested in Barca & Bachtler (2008), “Cohesion policy has become one of the most widely reported
and evaluated policies in Europe”, yet Barca (2009, p. XV) reports that the state of its empirical results is »very
unsatisfactory«. Batterbury (2006) sees a narrow focus of CP evaluation framework which is restricted to three
core purposes: accountability, improved planning, and quality and performance as one of the reasons for limited
usefulness of evaluation outputs. Further, she identifies the lack of data comparability, rigidity of time-scales and
a focus on performance approaches as major obstacles to effective evaluation. On the other hand, Polverari,
Mendez, Gross, & Bachtler (2007) report favourable trend in the evolution of monitoring and evaluation of CP
since evaluation design is becoming more and more systematic and rigorous and availability of data has been
considerably improving. Recognising several deficiencies, the European Commission concludes its last Cohesion
Report by noting that “the monitoring and evaluation systems need to be improved across the EU to track
1 sonja.slander@ef.uni-lj.si
2 PhD Student. gonza.tej@gmail.com
3 katja.zajc@ef.uni-lj.si
84
performance and to help redirect efforts as necessary to ensure that objectives are attained.” (European
Commission, 2010, p. 257).
Among the important changes in the understanding and organization of CP evaluation planned for the
programming period 2014-2020, European Commission stresses first a need for clearer articulation of the policy
objectives to move away from an excessive focus on the absorption of funding towards a results oriented policy.
And secondly, calls for more methodological rigour in capturing the effects of CP interventions. Furthermore,
European Commission strongly encourages CP evaluation to follow principle of triangulation by explicitly
stating “Whenever possible, evaluation questions should be looked at from different viewpoints and by different
methods”.
This chapter deals with the choice of methods and approaches as a crucial part of CP evaluation process
following the triangulation principle and summarizes the resulting evidence on impacts of the European
Cohesion Policy. Based on the case of Slovenian CP evaluation practices, we aim to discuss the usefulness of
triangulation based approach as suggested by the European Commission and argue further that there is a need to
integrate more systematically different micro and macro methodological approaches in order to obtain more
accurate evaluation of aggregate CP impacts. We demonstrate in the case of Slovenia that basic triangulation-
based process may be enriched with an additional feed-in mechanism, i.e where the estimates on R&D spillover
elasticities are obtained with microeconometric, counterfactual methods and fed directly into DG Regio’s
Hermin model.
The rest of the chapter is organised as follows. The next section sets the evaluation framework with outputs,
results and impacts for the purposes of programming, monitoring and evaluation of CP. In section 3 we discuss
the choice of methods and approaches following the triangulation principle and summarize the evidence on
impacts of the European Cohesion policy. Section 4 critically assesses Slovenian CP evaluation practices and
evidences and proposes refinements of the CP evaluation, while the last section concludes.
2. EVALUATION FRAMEWORK: OUTPUTS, RESULTS AND IMPACTS
In the latest guidance document on monitoring and evaluation of European regional development fund and
Cohesion fund in the programming period 2014-2020 European commission defines a simplified logical
framework for the purposes of programming, monitoring and evaluation of CP (see Graph 1). Monitoring and
evaluation serve the management purpose of delivering the programme in an efficient manner, while evaluation
contributes also to the assessment whether a programme has produced the desired effects. This logical
framework has been changed from former guidance provided by DG Regional Policy to facilitate some
important changes in the understanding and organization of CP evaluation. First, a move away from an excessive
focus on the absorption of funding towards a results oriented policy based on a clearer articulation of the policy
objectives is proposed. Second, Commission sets out more clearly the different types of evaluation and calls for
more methodological rigour in capturing the effects of our interventions.
85
Graph 1: Outputs, results and impact in relation to programming, monitoring and evaluation
Source: European Commission (2014).
According to the European Commission’s 2014 guidance document, there are two key tasks of impact
evaluation: (i) to disentangle the effects of the intervention/policy measure from the contribution of other factors,
and (ii) to understand the functioning of a programme. The former task is addressed with counterfactual impact
evaluations while the latter with theory-based impact evaluations.
Theory-based evaluations can provide insights into why things work, or don’t and under what circumstances.
They mainly produce a narrative estimate of the impact rather than quantified. The main focus is thus not a
counterfactual (“how things would have been without”) rather a theory of change (“did things work as expected
to produce the desired change”). Typical methods include literature reviews, administrative data analysis, case
studies, interviews and surveys in order to reconstruct and verify the intervention logic. Often mentioned
approaches are realist evaluation, general elimination methodology, contribution analysis and participatory
evaluation. The ex-ante evaluation of programmes can be understood also as a theory-based analysis, assessing
the strength of the theory of change and the logical framework before the programme is implemented.
On the other hand, counterfactual impact evaluation aims to provide a credible answer to the question "Does it
work?". The central question of counterfactual evaluations is rather narrow — how much difference does a
treatment make. Is the difference observed in the outcome after the implementation of the intervention caused by
the intervention itself, or by something else? Evaluations of this type are based on models of cause and effect
and require a credible and rigorously defined counterfactual to control for factors other than the intervention that
might account for the observed change. The existence of baseline data and information on the situation of
supported and non-supported beneficiaries at a certain point in time after the public intervention is a critical
precondition for the applicability of counterfactual methods. Typical methods are difference-in-difference,
discontinuity design, propensity score matching, instrumental variables and randomised controlled trials.
3. COHESION POLICY EVALUATION
The impact of cohesion policy is being regularly evaluated by the European Commission, responsibility for
evaluation however, according to Council Regulation 1083/2006, also lies with the Member States. As suggested
in Barca & Bachtler (2008), “cohesion policy has become one of the most widely reported and evaluated policies
in Europe”. Nevertheless, the European Commission concludes its last Cohesion Report by noting that “the
86
monitoring and evaluation systems need to be improved across the EU to track performance and to help redirect
efforts as necessary to ensure that objectives are attained.” (European Commission, 2010, p. 257).
3.1. Evaluation approaches and methods
The decision on the choice of methods and approaches is one of the crucial steps in CP evaluation. A range of
methods and approaches is available but they all have their strengths and weaknesses. Since there is no ideal CP
evaluation approach guaranteeing valid answers for all circumstances, a choice and combination of methods
need to be decided on a case-by-case base.
However, when dealing with such complexity and multidimensionality of evaluation process as it is the case of
CP evaluation, usually the triangulation principle is called for. Triangulation, i.e. use of multiple methods mainly
qualitative and quantitative in studying the same phenomenon (Jick, 1979), has been increasingly recognised and
used among scholars and researchers in social sciences in recent years. One of advantages from the combination
of quantitate and qualitative methodological approaches usually stressed in the literature is increasing study
credibility and reliability by focusing on convergent information from different methods. By combining multiple
observers, theories, methods, and data, several intrinsic biases and the problems that come from single method,
single-observer and single-theory studies may be overcome by triangulation. Further, Bechara and Van de Ven
(2011) accentuate validity by discussing how divergent information from different methods reveals important
aspects and values of a complex phenomenon that often go unrecognized without triangulation. The principle of
triangulation is emphasized also in the above mentioned European Commission’s (2014) guide by specifying
“Whenever possible, evaluation questions should be looked at from different viewpoints and by different
methods”.
Following Fay (1996) and Alecke, Blien, Frieg, Otto and Untiedt (2010) a comprehensive, triangulation-based
evaluation process consists of:
a) estimation of the impacts of the measures on the individual firm (microeconometric evaluation);
b) examination whether this is the best outcome that could have been achieved for the money spent
(efficiency, e.g. case study based cost-benefit analysis);
c) examination whether the impacts are large enough to yield net social gains if all spillover effects and
side-effects are taken into account (macroeconomic evaluation).
These approaches and the resulting evidence on the impacts of CP are summarized in the following sections. In
this section we focus on quantitative evaluation studies and will therefore omit the discussion on the more
qualitative case studies (see for example Davies et al., 2007, for an overview).
3.2. Macroeconomic evaluation
Cohesion policy aims to promote productivity and economic growth, stimulate the creation of jobs and promote
investment in regions. Because the desired region- and economy-wide effects are mainly focused on enhancing
the long-run, supply side growth potential of the economy, it is important to be able to model them in a
macroeconomic setting.
3.2.1 Single-equation approach: The impact of cohesion policy funds on growth and convergence
Macroeconomic, single-equation approach has mostly been used to assess the impact of CP on economic growth
and real convergence. Most studies are set in the context of new (endogenous) growth theory (Romer, 1994)
employing time series, cross-section or preferably panel data estimators on structural funds dataset of
NUT2/NUTS 3 regions. However, the results are mixed and no general conclusion regarding the growth effects
of CP funds can be drawn. On one hand, some studies find evidence for a positive relation between SF and
87
economic growth (Beugelsdijk and Eijffinger, 2005; Pellegrini, Terribile, Tarola, Muccigrosso, & Busillo, 2013),
while on the other hand others failed to find significant positive impact (Dall’erba and Le Gallo, 2007).
As summarized by Hagen and Mohl (2009, 2010), the differences in the impact of SF on economic growth can
be attributed to several factors: (i) to differences in the choice of units (countries versus regions), (ii)
methodological approaches (panel versus cross-section; endogeneity problems), (iii) time horizons, and (iv) to
the lack of high quality SF data (e.g., some authors use SF commitments instead of payments). They conclude
European structural funds are only conditionally effective provided a good quality of the institutional setup
(Ederveen, de Groot, and Nahuis, 2006), decentralized governmental structures (Bähr, 2008) or conditionally on
which Objective is analysed (Mohl and Hagen 2008, 2010).
With regard to the convergence effects, Becker, Egger & Von Ehrlich (2012) analyse not only whether CP funds
contribute to fostering growth in the target regions but also whether or not more transfers generate stronger
growth effects. They show that EU transfers enable faster growth in the recipient regions as intended, but some
reallocation of the funds across target regions would lead to higher aggregate growth in the EU and could
generate even faster convergence than the current scheme does. Namely, in 36% of the recipient regions the
transfer intensity exceeds the aggregate efficiency maximizing level and in 18% of the regions a reduction of
transfers would not even reduce their growth.
3.2.2. Macroeconomic simulation models for cohesion policy impact evaluation
Macroeconomic models allow us to take account of spillovers in the economy, and also provide the
counterfactual, baseline (no-cohesion-funds) development of the economy against which we can compare the
economic development in the presence of cohesion policy.
The first macroeconomic models to simulate the impact of cohesion policy date back to the 1990s: a macro-
econometric model, based on the complex, multi-sectoral HERMES was applied to Ireland (Bradley et al., 1992)
and then evolved into HERMIN model (i.e. Bradley et al., 1995); a two-sector endogenous growth model was
applied to Greece, Ireland and Portugal (Gaspar and Pereira, 1995, Pereira, 1997); a CGE model was applied to
simulate the effects of cohesion funds in Greece by Lolos et al. (1995); an input-output model by was applied to
Objective 1 regions by Beutel (1993), and Goybet and Bertoldi (1994) applied a dynamic general equilibrium
model to the same group of regions.
The European Commission puts a strong emphasis on macroeconomic simulations to estimate the cohesion
policy impacts4 and presents them in the Report on economic and social cohesion every three years. The last report (DG Regio, 2010) includes results from two models5:
1. QUEST II is a micro-founded neo-Keynesian dynamic stochastic general equilibrium model with endogenous
growth, developed by DG-ECFIN (Varga, J., In’t Veld, J. (2010)). The QUEST model has been used to estimate
the net effects of Cohesion funds. The cumulative net effect on the GDP of the EU-25 of the 2000–2006
programmes expenditure is estimated at 0.7% in 2009 (i.e. GDP was higher to this extent as a result of policy),
this was estimated to rise to 4% by 2020. In the EU-15 alone, the numbers show a cumulative net effect on GDP
of just over 3% by 2020 and in the EU-10, the effect of cohesion funds on GDP is estimated to be
15.9%,compared to non-cohesion policy baseline (DG Regio, 2010, p254).
2. HERMIN is a macroeconometric model with neoclassical features on the supply side and is now the most
4 It should be noted that these models do not measure the impact/ or effectiveness of cohesion policy - rather, they model it, and the differences in their results arise from their different theoretical underpinnings about the workings of the economy.
5 The 4th Report on economic and social cohesion also reported the results from a CGE model EcoMod, developed by the EcoMod Network/Free University of Brussels (for a basic presentation of the model see Bayar, 2007), but the results diverged widely from those of the other two macroeconomic models (see Bradley, Untiedt, 2008 for a discussion).
88
widely applied framework to Structural Fund analysis at the national and macro-regional level. It has been
developed in Ireland in to the late 1980s to specifically to evaluate the medium- to long-run macroeconomic
impacts of structural funds and taking into account the limited data availability in the less-developed EU
Member States and regions (i.e. Greece, Ireland, Portugal, Spain, the Italian Mezzogiorno, East Germany and
Northern Ireland). The design of HERMIN model is based on a small open economy model and incorporates
mechanisms which ar based on the endogenous growth literature which allow it to capture the long-run supply
side impacts of Structural Funds along with the short-run Keynesian impacts.
Currently, the European Commission is using a Cohesion System of HERMIN Models (CSHM), which was
developed for DG Regio6 to be, in the first instance, applied to the sixteen Objective 1 member states7, regions of the former East Germany and the Mezzoiorno region of southern Italy, to permit inter-country and
interregional comparisons. The newest, 2012 revision of the CSHM system now includes models for the 27 EU
member states (as of yet without Croatia but plus Turkey) in order to study the spillover impacts of Structural
Funds on the so-called “net donor” states.
The HERMIN framework includes five key production sectors: manufacturing (a largely traded sector), market
services (a mainly non-traded sector), building and construction, agriculture, and government (non-market)
services. The total of cohesion funds is disaggregated into three main economic categories: physical
infrastructure, human resources, and direct aid to the productive sector. The impact of cohesion policy on the
economy is then modelled through a mix of supply- and demand-side factors. The short-run, multiplier
(Keynesian) demand effects come directly from the increased CP expenditures, but these are not its raison
d’être. The essence of CP is to improve the economy’s competitiveness – its long-run supply-side potentials, by
providing finance for the access to better infrastructure, more educated workforce and research aid for firms.
While the structure and scope of the majority of macroeconomic models emphasize the short-run effects of
economic shocks on the demand side of the economy, the model to evaluate CP must also be able to focus on its
ability to support the development on the supply-side.
HERMIN captures the supply-side effects of cohesion policy through spillover (externality) mechanisms.
Parameters of ‘externality elasticities8’ are used, which are positive when the programmes of education, infrastructural investments and direct aid to firms have been carefully planned and successfully executed. The
model includes two types of spillovers-externalities:
1. Production externalities: directly increase the economy’s output – any rise (relative to the no-cohesion
policy baseline) in the stock of infrastructure, human capital and direct aid to productive sector will
induce in a direct rise in output of manufacturing and market services (for given inputs), where the
responsiveness depends on the size assumed for the spillover elasticity;
2. Externalities of factor productivity: are the increase in total factor productivity in the manufacturing and
9
market services sectors due to cohesion policy investment in human capital, infrastructure and R&D .
The model handles them by endogenizing the Hicks neutral technological progress (A) in the CES
production function and thereby making it dependent on the changes in the stock of physical and human
capital as well as R&D. This type of externalities also brings along a negative side-effect where an
improvement in TFP can induce a decrease in jobs unless output is increased enough to offset the loss.
The estimates of macroeconomic impact of cohesion funds (2000-2006 programmes) produced by the HERMIN
system of models show an increase in GDP of 11% by 2009 in the main recipient Member States (compared to
the counterfactual of no cohesion policy), while the number of employed is estimated to be higher by 5.6 million
(European Commission, 2010, p 250-253).
6 Bradley, Untiedt, 2007.
7Member states that are recipients of development assistance under the »convergence« criterion: Greece, Ireland, Portugal, Spain, Estonia, Latvia, Lithuania, Poland, Czech Republic, Slovakia, Hungary, Slovenia, Cyprus, Malta, Bulgaria, and Romania.
8 Values of spillover elasticities are usually calibrated based on literature review but can be estimated empirically (see the case of Slovenia below).
9 Only the R&D part of direct aid to productive sector is assumed to have a long term effect on the economy.
89
3.3. Microeconometric approach
In contrast to the contribution of macroeconomic studies to understanding of aggregate effects, micro-level
studies are used to estimate the impacts of interventions on the individual firm. The appeal of micro studies is the
flexibility to address the specific goals of the projects and programmes. This is particularly valuable since such
approach reduces the mismatch between operational logic of the evaluation system and the project/programme
logic. Most importantly, micro-econometric studies exploiting the rich firm-level datasets allow counterfactual
impact assessment of CP funds on various firm performance measures, e.g. employment, investment, R&D
activity, productivity, etc.
Despite widely recognised advantages of micro-econometric counterfactual impact evaluation the attempts at
evaluating CP impacts based on this approach are relatively rare and have started to appear only recently. The
summary of studies based on this approach is presented in Table 1 below.
Table 1: Summary of a counterfactual impact evaluation studies based on micro data
Study
Region/
Programme/
Country
Measure
Impact
Methodology
Findings
Bondonio and
northern
“Objective 2” area employment
difference in
Business incentives promoted employment
Greenbaum
and central
business
impact
difference model
growth in the target areas. They were most
(2006)
and Italy
incentives
effective when targeting production in
province-industry pairs that had the least
severe declines during the years prior to the
program intervention.
Alecke, Blien,
East German direct investment
investment
propensity score
Grants induced both strong investment and
Frieg, Otto &
Länder,
and R&D grants
and R&D
matching,
R&D effects: (i) an average support of 8,000
Untiedt (2010)
2000-2006
via the ERDF Obj.
behaviour
difference in
EUR per employee led to 11,000-12,000 EUR
1
difference,
of extra investment per employee, and (ii)
IV
R&D grants of roughly 8,000 EUR led to an
additional 8,000 EUR of investment.
A rough calculation of the direct
employment effect from investment grants
was some 27,000 extra jobs.
Hart and Bonner
Northern
public sector
Firm
2-stage Heckman
A significant positive impact on productivity
(2011)
Ireland,
financial
performance
model
and turnover, but insignificant on
2001 -2008
assistance to
(productivity,
(robustness checks employment.
private firms
employment,
DiD and PSM)
Employment in non-assisted manufacturing
sales)
firms fel by 3.9% per annum - for assisted
firms the drop was only 1.9%. While in the
business service sector non-assisted firms
grew by about 4.9% per annum, while
assisted firms grew by 6.9%.
Czarnitzki, Lopes
Czech
innovation
impact on
difference-in-
In Czech republic the treated firms suffered
Bento and
republic and
support
innovation
difference
less from a reduction in patenting during the
Doherr (2011)
Germany
activity
estimator,
financial crisis than the non-recipients.
(patents,
nearest neighbor
In Germany the representative firm would
R&D
matching
have had R&D expenditure of 213,000 EUR
investment,
without ERDF. The treatment effect in terms
innovative
of EUR amounts to 87,000 EUR, on average,
behaviour)
for a typical grant size of up to 51,000 EUR.
CP recipients also score higher on a range of
innovation indicators.
ASVAPP (2012)
Italy and
enterprise
employment
Conditional
Smaller grants proved to be much more
region of
support:
impact (cost
difference in
cost-effective than larger ones: cost per jobs
Piemonte
investment grant
per job, job
difference, control averaged €79,000 for the smallest grants
and various SME
quality)
group selected by
(less than €125,000), rising to €489,000 for
schemes
matching using a
the largest grants (above €500,000). The
stratification &
loans had a cost per job around half that of
reweighting
grants plus a surprisingly high impact on
approach
investment – EUR 5 per euro of gross grant
equivalent. The quality of the jobs created is
usually similar to average jobs in the
enterprises concerned.
90
All of the studies presented in Table 1 deal with the direct impacts of the selected business incentive schemes on
performance of the incentive recipients. In most cases the studies find more or less pronounced positive impacts
on various aspects of recipient’s performance. Furthermore, this approach can be also used to assess indirect
effects, i.e spillover effects on other non-recipient firms. For instance, Basile, Castellani and Zanfei (2008)
examined one of such indirect impacts of CP funds on attracting multinationals. They find evidence that
Structural and Cohesion funds allocated by the EU to laggard regions have contributed to attracting foreign
investors from both within and outside Europe.
It can be established based on comparison of different evaluation approaches and the resulting evidence on the
impacts of CP that there are strong complementarities among different approaches contributing to the better
fulfilment of the two essential tasks of managing authorities when running a programme, i.e. to deliver the
programme in an efficient manner and guarantee that programme has produced the desired effects. In the
following chapter we deal with the concrete evaluation of CP in Slovenia.
4. COHESION POLICY EVALUATION IN SLOVENIA
Slovenia has been the recipient of pre-accession assistance from as early as 1992, while full access to cohesion
policy has been possible after full membership, first in 2004-2006 and later in 2007- 2013 period. In this
financial perspective, Slovenia has €4.2 billion of available commitment appropriation at its disposal until 2015,
which, given its size, puts the question of the effects of resources at the forefront of our research interest. In the
following sections, we present the current state of cohesion policy evaluation in Slovenia.
4.1. Microeconomic evidence
4.1.1 Survey of cohesion policy evaluations in Slovenia for the 2007-2013 financial perspective
We begin with an exposition of case-study evaluations of the three Slovenian Operational programmes for 2007-
2013 financial perspective.
4.1.1.1. Evaluations within the Operational Programme for Human Resource Development for the Period 2007 –
2013 (OP HR)
Evaluation of the instrument financed under OP HR, Priority Axis »Equal opportunities and reinforcing social
inclusion«, Activity Field »Increased employability of vulnerable groups in the field of culture and support to
their social inclusion« (Pitija, 2009b)
The evaluation assessed 17 operations (involving 500 participants), focused on two main activities: (i)
training and (ii) employment of members of vulnerable groups. The biggest share of the subsidies
(36.6%) were approved for operations for disabled people, 21.35% of the funds were allocated to projects
for »other ethnic groups and immigrants«, 15.8% were allocated for the Roma community. The report
concluded that the instrument was useful in both increasing the employability of vulnerable groups in the
field of culture and in supporting their social inclusion. The great majority (above 80%) of participants
agreed that inclusion in the projects helped them to re-establish social contacts, improved their self-
esteem and gain new knowledge and skills. Around 20 % of the participants reported that the programme
helped them with employment and improved their financial status. The report nevertheless argues that
despite identifying specific social groups (unemployable youth, elderly and women) as vulnerable groups,
they have not been included in the programme.
91
Evaluation of the public tender for co-financing horizontal NGO networks and regional centres for the year
2008 (Pitija, 2009c)
The subjects to evaluation were eight projects co-financed under the Operational Programme
Development of Human Resources 2007-2013, Priority »Encouraging development of non- government
organizations, social and civil dialogue.« The evaluation report confirms that the initiative for a more co-
ordinated NGO action was relevant and adequate, and the pro-active role of the Ministry (for public
services) was effective and positive. The projects carried out were strategic projects helping to form a
new infrastructure for the operations of the non-government sector, but a lack of a stable funding for
NGOs (after the project has ended) is defined as a major threat for activities which are to be performed by
the regional centres and horizontal NGO networks.
Mid-term evaluation of the OP HR
Oikos (2011) reports that Slovenia has failed to define the details of objectives and necessary instruments
for the human resource development strategy. Slovenia has not properly specified priority economic
sectors to be stimulated. Lack of coordination of instruments to fulfil the priority axis and the inability to
include monitoring of the long-term needs of government, employers and employees were also pointed
out.
Evaluation of labour market priority themes within OP HR, managed by the Ministry for labour, family, social
affairs and equal opportunities (Oikos, 2012)
The subject of evaluation were 37 instruments co-financed under priority theme “Training and education
for competitiveness and employability”, “Scholarship schemes”, “Enhancing the development of new
employment opportunities”, “ Encouraging the employability of job seekers and inactive”, “The reform of
job market institutions”.
The report finds the instruments suitable for a changing environment, and education and training of employees
have yielded positive effects on competitiveness of the companies. New job opportunities were not created,
however effects were present on sustaining existing jobs and improving competitive advantage of participants.
The self-employment support sufficiently facilitated self-employed. Graduates in the sample strengthened their
theoretical knowledge and added dynamics and mobility into organization. On the other side, the report points
out that 86% of all the funds targeted unemployed and failed to address the roots of the problem, and the process
of preparation of any instrument still does not include all target groups (cooperation of employers and
employees). Also, the implementation of instruments has to be linked institutionally and content-wise, and
procedures should be simplified.
4.1.1.2. Evaluations within the Operational Programme for Environmental and Transport Infrastructure
Development for the Period 2007 – 2013 (OP HRD),
Mid-term evaluation of OP ETID 2007 – 2013 (Oikos, 2010)
The evaluation of the projects within OP ETID 2007 – 2013 assessed the majority of them (66 out of 111)
as feasible. The potential issues were recognized in the implementation process itself (procurements,
permits, etc.).
The focus of OP ETID 2007 – 2013 is primarily on the Slovenian highways, which are seen as enhancing
the gravitational role of the larger urban areas while neglecting the regional development centres. They
have also failed to provide a permanent connection between regional and municipal centres. With the
exception of the second railway track between Koper and Divača, further improvements of railway
92
system were not predicted by the OP ETID 2007-2013, yet it needs further attention. The main airport,
Jože Pučnik, continues to be insufficiently connected by public transportation. Overall, the objectives of
improving Slovenian public transportation were not met, on contrary, Oikos reported of a step backwards.
The waste management system, planned in this OP is consistent with the needs of different municipalities. All
projects addressing the dangers of flooding are deemed unfeasible. Investments in buildings’ energy renovation
are meaningful. Actions on lowering the greenhouse gases need to be directed towards both, the consumers of
the energy, and the suppliers and energy producers (both small and larger investments in energy systems).
4.1.1.3. Evaluations within the Operational Programme for Strenghthening Regional and Development potentials
for the period 2007 – 2013 (OP RD)
Evaluation of the fourth priority axis »Regional development« within OP RD 2007 – 2013 (Pitija, 2009a).
The report targeted 487 of approved operations (in total of €311,763,271), where 35.5% of funds were
used for transport infrastructure, 31.63 % for environmental infrastructure, 16.44 % for tourism
development, while remaining was put in economic and social infrastructure and the development of
urban settlements. In general, operations were relatively successful, with two goals (number of people
with access to safer and higher quality water system, number of communally equipped agglomerations)
already been surpassed in 2009. Three objectives reached 50% and two reached 30% of their targets.
Many projects within Natura 2000 were assessed as unachievable, as majority of the instruments focused
on infrastructure and neglected biodiversity preservation and establishment of managerial structures.
Achieving the planned goal for large development projects, renovation and remedy of urban areas, which
was set above two million EUR, will be also difficult, where only 10 % was reached until 2009. The goal
of raising the number of people using public transportation was underachieved, with 0 % increase. The
number of newly created gross vacancies nearly met the targets (90 %). Business zones were on the rise
with the complementary infrastructure as well. The development of regional urban areas was sufficient
with the investments in the renovation of town centres and municipal infrastructure. The least efficient
efforts were made in the area of the development potential of regions. Operations that were financed were
assessed as beneficial for a sustainable development of the regions with opportunity to improve
economic, social and environmental elements. Operations did not have a negative effect on providing
equal opportunities to marginal groups and were beneficial for all inhabitants.
Midterm report on evaluation of key activities of innovation policy in Slovenia for the period 2007-2013 (MK
Projekt, 2012)
The objects of evaluation were the following priorities within two Operational programmes:
“Competitiveness and research excellence” (OP RD); “Experts and researchers for competitive
enterprises” (OP HR); “Scholarship schemes” (OP HR), “Quality, competitiveness and responsiveness of
higher-education” (OP HR).
The report suggests that innovation policy has achieved considerable improvements in innovation activity with
the particular contribution to strengthening cooperation between science, research and technological
development. Partnership is in particular improved between academic field and businesses. Indicators of effects
and results are exhibiting successful mid-term achievement of results and in some cases even realisation of goals
for the entire period (some crucial project results have been correlated with external statistical data which
confirms overall positive impact of the innovation policy). Institutional weaknesses of innovation policy are
seen as the main factor of lower efficiency, where administrative management is too often seen as more
important than substantive goals of innovation policy. Evaluation results also revealed there are still
considerable unused potentials for innovation which are not addressed by the current innovation policy. For their
93
activation they suggest intensifying inter-ministry coordination as well as exchange of knowledge between
responsible decision-makers and beneficiaries in preparation of future instruments.
4.1.2. Evaluating cohesion policy for firm-level R&D in Slovenia: counterfactual methods
Although case studies can provide an important and useful feedback about the workings of cohesion policy, they
tend to be highly qualitative (answering mainly the question “why does it work”) and focus on spending and
outputs (e.g. number of people trained) and/or results (number of people getting a job after training) rather than
impacts (the effects of a better trained workforce on firm performance) and answering the question “what
works”. But with growing importance of cohesion policy, the concerns about the effective use of funds have also
grown and the European Commission evaluation guidance documents have become more decisive, for example:
“In DG Regio we see a potential role for the use of counterfactual evaluation. Throughout 2007-13, together with
Member States and regions, DG Regional Policy will test the merits and limits of counterfactual evaluations
through pilot studies. We believe that this method can become a powerful additional tool that will need to be
complemented by others, including qualitative approaches.” (Stryczynski, 2008)
“In the programming period 2014 - 2020 performance and results will receive increased attention. This will
require a review of current monitoring and evaluation systems and capacities, including data collection
arrangements. Moreover, evaluation plans will become obligatory, and more emphasis is to be placed on impact
evaluation. This shift in focus towards a performance and results orientation is important. High-quality
evaluation strategies and techniques are essential for generating knowledge useful to all member states about
which interventions ‘work’ and which do not. Strengthening the quality of evaluations and developing reliable
evidence of value added is essential.” (European Commission, DG Employment, 2012, p.6)
Following this line of thought, Slovenia is one of a few cases where microeconometric counterfactual analysis
was used for cohesion policy evaluation (see Section 3.3 for a presentation of similar studies). It is based on
firm-level data and thoroughly presented in Šlander (2010). Below we outline the estimated model and discuss
the conclusions.
The study considered the effects of cohesion funds (subsidies, guarantees, credits) for R&D in Slovenia in the amount of €128 mio, given directly to 969 firms between 2004 and 2008, and analysed firm-level performance
of cohesion funds recipients relative to non-recipients.
The preliminary analysis10 found that the recipient firms of cohesion funds in Slovenia are on average more productive (by 25%), have higher capital-labour ratio, are larger, have above average investment- and export-intensity, higher labour costs per unit (which is also indicator of a higher level of human capital) and better
energy efficiency even before receiving cohesion funding. This suggests that the firms winning the cohesion
funds (hereafter CF) tenders have not been chosen randomly which is a causality issue and needs to be taken into
account to get unbiased estimates of the effect of cohesion funds on the performance of recipient firms11.
The effect of cohesion R&D funds on firm-level performance of the recipients was first estimated using the
Heckman two-step selection model (following the procedure for panel data as suggested in Wooldridge, 1995).
In the first step, the non-randomness property of the cohesion funds selection process was modelled where tender
conditions were used as covariates. In the second step, an augmented Cobb-Douglas production function in
logarithms was estimated:
Output
+β
it =α *β ilabourit +β k capitalit
m materialit +γ Cohesion policyit +µ Control_variablesit +ε it
10 The preliminary analysis compared the characteristics of cohesion funds’ recipients to both the entire population of Slovenian firms and to the average of their respective 2- and 3-digit NACE class.
11 The non-random selection of cohesion funds recipients was also confirmed with the binary probit selection model.
94
and the results suggest that the growth in total factor productivity (TFP) was on average higher by 1.8 percentage
points in firms receiving cohesion R&D funds during 2004 and 2008, compared to the population of Slovenian
firms. The impact is higher in the services sector than industry sector where the estimated difference in TFP
growth was 0.9 percentage points on average.
Next, the impact of cohesion R&D funds on firm-level performance was evaluated using the matching estimator
(e.g. Rubin, 1972-1979; Heckman, Ichimura, Smith, & Todd, 1998) which allows us to form a control group of
firms, similar to the cohesion R&D funds recipients (the treated firms) in all (observable) characteristics but the
fact that they did not receive cohesion R&D funds. Comparing the results of the treated group (CF recipients)
with their matched counterparts gives us unbiased estimates of the treatment effect (which is receiving CF for
R&D).
The matching of the CF recipients with their control group of similar firms was based on numerous firm
characteristics, which were reduced to a single-dimensional index (containing all the relevant information) with
the propensity score technique (i.e. Hahn, 1998, Rosenbaum & Rubin, 1983; Donald B., Rubin, 1992), where the
treated units are linked to their control units based on the propensity - probability to be selected into the cohesion
policy programmes. The selection propensity can therefore be defined as conditional probability that a firm will
be selected into programme (D), conditional on its characteristics before the programme (X):
(1) p(X) = PrD = 1|X = ED|X
This means that we can link each CF recipients to a firm or firms which have not received funds, but are the
most similar based on the probability of selection. When the control group is formed in this manner, the average
impact of cohesion funds (average treatment effect, ATT) is calculated as the difference between the average
results of both groups (CF recipients and control group of non-recipients):
(2) ATT = E(Y
|
|
|
1i – Y0i Di = 1) = EE Y1i Di = 1,p(Xi) - EE Y0i Di = 0,p(Xi)| Di = 1
where the external parentheses relates to the distribution (p(X |
i) Di = 1) and Y1i , Y0i to the potential results of
both groups – the treated firms (CF recipients) and the control group.
Following the procedure outlined above (equation 1) to estimate the effects of cohesion funds on firm-level
performance of recipients, a pooled probit selection model was estimated for the propensity of firms to be
selected into CP funding (propensity score). The period of estimation was 2004-2008 and one-year lags of
independent variables were used to avoid endogeneity.
The results of the model reveal that the selection was positively related to labour productivity, export intensity,
capital intensity and investment intensity, meaning that the probability to apply for cohesion policy tenders (and
the selection into funding) was higher in firms which were above-average investors even before being selected to
receive cohesion funds. Due to the specific tender conditions, the probability for selection was lower in firms
with a higher capital loss, while the size of firms was linked to selection in a quadratic function, where the
selection was positively affected by size only up to a certain size, followed by a negative correlation (this is the
direct effect of tender conditions, where a part of funds were dedicated only to micro, small and medium-sized
firms).
95
In general, the average effect of CF on the group of recipients can be defined for all matching algorithms
following Morgan, Winship (2009, p.106):
(3) ATTmatch =1/ n 1 Σ[( y |
( |
i di =1) − Σjωij yi di =1)]
where n 1 is the number of treated units, index i denotes the treated units, j denotes the control units, while ωij are weights measuring the distance of each control unit to its treated unit.
We report the results on the average treatment effect based on the nearest neighbour matching procedure. The
average effect of cohesion R&D funds on recipient firm labour productivity is estimated at 0.64 in the year of
receiving funds, which means that the productivity of firms, receiving cohesion R&D funds between 2004 and
2008 was higher by 6.4% in the year when funds were received (which was after the end of the
project/investment) compared to the control group of otherwise similar firms. A year later, the effect is
estimated at 7.2% (but significant only at 10% level due to loss of observations). Receiving cohesion R&D funds
has also positively (and significantly) affected firm export intensity (by 10% in the year of receiving funds and
23% a year after) and capital intensity (higher by 19% in the year of receiving funds, compared to the control
group). Employment wasn’t significantly affected in the year of receiving funds, but it was higher by 5.3% in
the year after (relative to the control firms). Energy intensity does not seem to affected by the cohesion funding
in the year of receiving funds (even though this was stated as one of the goals of cohesion R&D funds) while the
effect was wagely positive (not significant) a year after. Summa summarum, the empirical analysis suggests that
the cohesion policy for firm-level R&D in Slovenia was mostly successful in achieving the goals set out by the
National Strategic Reference Framework (Government Office for Local and Regional Development, 2007).
4.2. Macroeconomic evidence
The first models to estimate the macroeconomic impacts of cohesion funds in Slovenia date back to 1999, when
basic versions of the HERMIN models (Simončič et al., 1999) were estimated. The competitive model, a computable general equilibrium model SloMod (which was part of the EcoMod modelling framework) was used
first as part of ex-ante CP evaluation of the 2004-06 period (Kavaš et al., 2003) and later to measure the
macroeconomic impacts in the context of ex-ante Cohesion policy evaluations for the 2007-13 financial
framework (as part of the NSRF and operative programmes preparation). It was estimated that cohesion funds in
Slovenia would result in an average of 0.75 percentage points higher economic growth and 33,900 gross (27,500
net) additional jobs, leading to an increase in national employment of 1.7 percentage points and decrease in
registered unemployment of 2.2 percentage points.
Currently, macroeconomic impacts of cohesion policy in Slovenia are estimated by DG Regio, with the
Slovenian model within Cohesion System of HERMIN Models (CSHM, presented in Section 3.2.2). The model
works as an integrated system of cca 250 inter-dependent equations, 20 of which are stochastic and their
parameters are calibrated empirically based on economic theory. While the parameters of spillover elasticities
(parameters that allow for the long-run, supply-side impacts of cohesion funds in the model) within the CSHM
are chosen on the basis of a substantial international literature review (see Bradley, Untied, 2007, for an
exposition), we have modified slightly the Slovenian HERMIN model to use an empirically estimated parameter
of R&D factor productivity spillover (based on a microeconometric, firm-level model which is presented in
section 4.1 above).
The macroeconomic impacts of cohesion funds in Slovenia (for the total of 2004-06 period and 2007-13
financial perspective), estimated with the DG Regio’s Slovenian HERMIN model and presented below, are
based on the actual payment profiles until 2009 and planned annual allocations afterwards, for the period 2010-
15 (taking account of the ‘N+2’ rule).
96
Figure 2 presents the share of cohesion (public) funds12 in Slovenian GDP for the period 2004-15 as incorporated in the HERMIN model. Cohesion funds were relatively scarce during the 2004-06 period (€458
mio, which was below 0.4% GDP annually) but have increased significantly in the financial period 2007—13
(with the planned average allocations of cca. € 600 mio per annum). As in Figure 2, average yearly cohesion
funds allocations for the 2007-15 period is about 1.2% GDP.
Figure 2: Cohesion policy funds in Slovenia during the period 2004-15 as share of GDP (actual payments
during 2004-09, planned annual allocations in 2010-15)
Source: European Commission, authors’ calculations
Below we present the impact of cohesion policy in Slovenia on key macroeconomic variables. The total impact
comes from both short-run (Keynesian) effect of increased demand (with a multiplied impact due to inter-
sectoral linkages that spill over the entire economy) and the supply-side effect that accumulates over the years
into a long-run effects. The impact of cohesion funds is calculated as a percent-difference between the baseline
simulation (Slovenian economy without cohesion funds) and the cohesion policy simulation.
To sum up the macroeconomic impacts of cohesion policy, estimated with the Slovenian model within CSHM
and presented in Figure 3, cohesion funds are expected to cause a 1.18% increase in GDP on average for the
entire period 2004-2020, while the figure for the period 2007-2015 is 1.78% (compared to the baseline scenario
of no cohesion funds). The employment effect (the % increase relative to the no-cohesion funds scenario) is on
average 0.87% for the entire period while it was 0.17% during 2004 and 2006 and 1.15% between 2007 and
2015. The employment effect stays positive (on average 0.44% above the baseline of no cohesion funds) even
after 2015, when the Keynesian demand effect fades out. The national unemployment rate is on average lower by
0.87% due to cohesion policy programmes (2004-20 period), the difference is even -1.36% on average between
2007-15, but even after the cohesion policy induced inflated demand phase is over in 2015, the effect is still
evident (between 2016-20, the number of unemployed is on average lower by 0.42%, compared to the baseline
scenario of no cohesion funds).
The estimated effect of cohesion policy on labour productivity comes not only from the demand side - increased
stock of capital (infrastructural, research and human capital) but also from technological spillovers, tied to some
of the cohesion expense categories (presented in Section 2.2). This effect is positive and increasing during the
entire 2004-20 period, when the average effect of cohesion funds on value added per employee, relative to the
12 Funds from the EU + Slovenian budget
97
baseline (no cohesion policy), is 0.28%. Over time, this effect is enhanced by the impact of cohesion policy on
increased competitiveness of the supply side of the economy and reaches the maximum in 2015, when labour
productivity is higher by 0.4% relative to the baseline scenario.
Figure 3: Macroeconomic impacts of cohesion policy in Slovenia (impact on GDP, employment,
unemployment and labour productivity) during 2004–2020 (%-increase relative to the no-cohesion policy
baseline simulation)
Source: authors’ calculations based on DG Regio’s Slovenian model within CSHM
To sum up the Slovenian CP evaluations presented above, we argue that the Slovenian model, by using both case
studies, microeconometric counterfactual methods as well as macroeconomic simulation model, not only
represents a successful case of triangulation in CP evaluation but modifies the basic triangulation scheme with an
addition of a feed-in mechanism where the results of the microeconometric CP evaluation model are fed directly
into the macroeconomic HERMIN model. In this, we support the European Commission’s view that a move
towards microeconometric, firm-level research based on counterfactual methods is needed to gain insight into
what works and what doesn’t in Cohesion Policy and argue further for an integration of micro- and
macroeconometric evaluation models for more accurate estimates of CP impact on macroeconomic variables.
5. CONCLUSIONS
Cohesion policy is financially the second strongest policy area in the EU budget, accounting for 36% of funds in
the period 2007-2013. Its objective is to contribute to reduced regional development disparities and at the same
time to promote growth across the European union. The impact of cohesion policy is being regularly evaluated
by the European Commission, responsibility for evaluation however, according to Council Regulation
1083/2006, also lies with the Member States. Cohesion policy has become one of the most widely reported and
evaluated policies in Europe (Barca & Bachtler, 2008) yet the European Commission concludes its last Cohesion
Report by noting that “the monitoring and evaluation systems need to be improved across the EU to track
performance and to help redirect efforts as necessary to ensure that objectives are attained.” (European
Commission, 2010, p. 257). It also emphasizes that this “requires a greater recourse to rigorous evaluation
methods, including counterfactual impact evaluation …” (ibid, p.257) striving to disentangle the effects of the
intervention/policy measure from the contribution of other factors.
98
A range of methods and approaches is available but they all have their strengths and weaknesses, leading to the
European Commission’s (2014) recommendation “Whenever possible, evaluation questions should be looked at
from different viewpoints and by different methods”. Following Fay (1996) and Alecke, Blien, Frieg, Otto and
Untiedt (2010) a comprehensive, triangulation-based evaluation process consists of: (i) estimation of the impacts
of the measures on the individual firm (microeconometric evaluation); (ii) examination whether this is the best
outcome that could have been achieved for the money spent (efficiency, e.g. case study based cost-benefit
analysis); (iii) examination whether the impacts are large enough to yield net social gains if all spillover effects
and side-effects are taken into account (macroeconomic evaluation).
Even though the European Commission lately emprises the need for counterfactual evaluations, studies using
microeconometric models are still relatively scarce, since they tend to be data- and methodological knowledge-
demanding. Historically, quantitative, case- studies are the preferred mode of CP evaluation. The European
Commission, DG Regio, also uses two macroeconomic models (QUEST, HERMIN) in its regular Report on
economic and social cohesion to report on longer-term, economy wide impacts of cohesion funds.
We present Slovenia as a successful case of triangulation in CP evaluation, with an additional improvement in
the form of a feed-in mechanism. In the 2007-2013 financial perspective, Slovenia has 4.2 billion EUR of
available commitment appropriation at its disposal, which, given its size, puts the question of the effects of
resources at the forefront of our research interest. Although case studies of programmes, priorities and
instruments have been well established and preferred in CP evaluation, efforts have also been put into the other
two dimensions. For the microeconomic evaluation, matching and difference-in-difference were used to evaluate
the effects of CP direct aid for R&D to Slovenian firms based on firm-level data for 2004-2008 period. The fact
that the recipient firms were on average more productive, had higher capital-labour ratio, were larger, had above
average investment- and export-intensity, higher labour costs per unit and better energy efficiency even before
receiving cohesion funding was a motivation for the selection of the counterfactual approach. The results
confirm a positive and significant impact of cohesion policy R&D funds on the selected firm-level performance
indicators, compared to the (matched) control group of otherwise similar firms: labour productivity in recipient
firms is estimated to be higher (by 6.4 % on average) in the year of receiving funds (after the project is
completed) and the year after, and the same for export- and capital intensity. Employment was estimated to be
higher a year after receiving funds, while no significant effect of cohesion funds on energy efficiency was found.
Because microeconomic results do not include effects such as spillovers, feedbacks and externalities, and
therefore lack the ability to represent the economy-wide effects on macroeconomic variables, they must be
complemented by a fully-specified macroeconomic model. The DG Regio’s Slovenian model within CSHM has
therefore been used as the third model to complete the comprehensive CP evaluation in Slovenia, but it has been
improved slightely to include the empirically estimated parameter of R&D factor productivity spillover (rather
than use the choice made on literature review which is the mode of action in other countries’s models) obtained
from the model explained above. In this way, we have modified the original triangulation process in CP
evaluation with a feed-in mechanism from the results of the microeconometric model directly into the
macroeconomic HERMIN model to get more accurate estimates of CP impact on macroeconomic variables. So
not only we confer with the European Commission’s view that a move towards microeconometric, firm-level
research is needed to gain insight into what works and what doesn’t in Cohesion Policy, we argue further for an
integrated micro- and macroeconometric approach as well as a modification of the basic triangulation procedure
where microeconometric results could be fed into a macroeconomic model as much as possible to improve the
accuracy of the estimates of CP economy-wide effects.
99
REFERENCES
Alecke, Blien, Frieg, Otto and Untiedt (2010). Ex post evaluation of Cohesion Policy programmes 2000-2006, Final report.
[http://ec.europa.eu/regional_policy/sources/docgener/evaluation/pdf/expost2006/wp6c_final_report_en.pdf]
ASVAPP (2012). Counterfactual impact evaluation of Cohesion policy: impact and cost-effectiveness of investment subsidies in Italy. Final report. [http://ec.europa.eu/regional_policy/information/evaluations/impact_evaluation_en.cfm#1]
Bähr, Cornelius (2008). How does Sub-National Autonomy Affect the Effectiveness of Structural Funds? Kyklos. 61: 3-18.
Barca, F. (2009). An Agenda For A Reformed Cohesion Policy. A Place- Based Approach to Meeting European Union Challenges and Expectations, Independent Report prepared at the request of Danuta Hübner, Commissioner for Regional Policy, 244 pp.
Basile, R., Castellani, D., & Zanfei, A. (2008). Location choices of multinational firms in Europe: The role of EU cohesion policy. Journal of International Economics, 74(2), 328-340.
Batterbury, S. C. (2006). Principles and purposes of European Union cohesion policy evaluation. Regional Studies, 40(02), 179-188.
Bayar, A. (2007), Study on the Impact of Convergence Interventions2007-2013, Working paper, ULB/EcoMOD, May.
Bayar, A. (2007), Study on the Impact of Convergence Interventions 2007-2013, Working paper, ULB/EcoMOD, May.
Bechara, J., & Van de Ven, A. H. (2011). Triangulating philosophies of science to understand complex organizational and managerial problems. Research in the Sociology of Organizations, 32, 343-364.
Becker, S. O., Egger, P. H., & 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), 648-668.
Beugelsdijk, M. and S. C. W. Eijffinger (2005). The Effectiveness of Structural Policy in the European Union: An Empirical Analysis for the EU-15 in 1995-2001, Journal of Common Market Studies. 43: 37-51.
Beutel, J. (1993) ‘The Economic Impacts of the Community Support Frameworks for the Objective 1 Regions 1989–93’, Report prepared for DG XVI, Brussels: Commission of the European Communities, April.
Bondonio, D., & Greenbaum, R. T. (2006). Do business investment incentives promote employment in declining areas? Evidence from EU
Objective-2 regions. European Urban and Regional Studies, 13(3), 225-244.
Bradley, J. & Untiedt, G. (2007). The COHESION System of HERMIN country and regional models: Version 2. Muenster: report for European Commission, 1–258.
Bradley, J. & Untiedt, G. (2009). Analysis of EU Cohesion Policy 2000–2006 using the CSHM: Aggregate impacts and inter-country comparisons. [http:// ec.europa.eu/regional_policy/sources/docgener/evaluation/pdf/expost2006/wp3_hermin_aggregate.pdf]
Bradley, J., Fitzgerald, J. and Kearney, I. (1992). The Role of the Structural Funds: Analysis of Consequences for Ireland in the Context of 1992, Policy Research Series No. 13, The Economic and Social Research Institute, Dublin.
Bradley, J., Modesto, L. & Sosvilla-Rivero, S. (1995). HERMIN A macroeconometric modelling framework for the EU periphery. Economic Modelling, 12, 221–247.
Bradley, J., Untiedt, G. (2008). Do economic models tell us anything useful about Cohesion Policy impacts?, in Regional Economic Policy in Europe: New Challenges for Theory, Empirics and Normative Interventions, (eds.) U. Stierle-von Schutz, M. Stierle, F. Jennings and A.
Kuah, Edward Elgar, Cheltenham, UK, pp. 159-180.
Czarnitzki D., Lopes Bento C. and T. Doherr (2011). Counterfactual impact evaluation of cohesion policy. Final report.
[http://ec.europa.eu/regional_policy/information/evaluations/pdf/impact/ciewp2_final.pdf]
Dall’erba Sandy and Julie Le Gallo J. (2007). The Impact of EU Regional Support on Growth and Employment, Czech Journal of Economics and Finance. 57(7): 325-340.
Ederveen, Sjef, Henri L.F. de Groot and Richard Nahuis (2006). Fertile Soil for Structural Funds? A Panel Data Analysis of the Conditional Effectiveness of European Cohesion Policy. Kyklos. 59: 17-42.
100
European Commission (2007). Growing Regions, Growing Europe. Forth Report on the Economic and Social Cohesion. Brussels: European Commission. 226 pp.
European Commission (2010 ). Investing in Europe’s future. Fifth Report on the Economic andSocial Cohesion. Brussels: European Commission. 266 pp.
European Commission (2014). Guidance Document on Monitoring and Evaluation − European Regional Development Fund and Cohesion Fund, Concepts and Recommendations. The Programming Period 2014-2020.
European Commission, DG Employment (2012): The Design and Commissioning of Counterfactual Impact Evaluations. A Practical
Guidance for ESF Managing Authorities. Brussels: European Commission.
Fay, R. (1996). Enhancing the Effectiveness of Active Labor Market Policies: Evidence from Programme Evaluations in OECD Countries, Labour Market and Social Policy Occasional Papers, OECD.
Gaspar V., and Pereira A. (1995). The Impact of Financial Integration and Unilateral Public Transfers on Investment and Growth in EC
capital-importing countries. Journal of Development Economics, 48 (1), pp.43-66.
Government Office for Local and Regional Development (2007): Nacionalni strateški referenčni okvir 2007–2013. Ljubljana: SVLR.
Goybet, P. and M. Bertold, 1994, “The Efficiency of the Structural Funds,” in Improving Economic and Social Cohesion in the European Community, J. Mortensen, editor, New York: St Martin’s Press, pp. 229–40
Hagen, T., & Mohl, P. (2009). How does EU Cohesion Policy Work? Evaluating its Effects on Various Outcome Variables.
[http://www.socialpolitik.ovgu.de/sozialpolitik_media/papers/Hagen_Tobias_uid645_pid580.pdf]
Hahn, J. (1998). On the role of the propensity score in efficient semiparametric estimation of average treatment effects. Review of Economic Studies, 64, 605–654.
Hart M. and K. Bonner (2011). Data-Linking and Impact Evaluation in Northern Ireland. Final report.
[http://ec.europa.eu/regional_policy/information/evaluations/impact_evaluation_en.cfm#3]
Heckman, J., Ichimura, H. & Todd, P. (1998). Matching as an econometric evaluation estimator. The Review of Economic Studies, 65(2), 261–294.
Jick, T. D. (1979). Mixing qualitative and quantitative methods: Triangulation in action. Administrative science quarterly, 602-611.
Kavaš, D., Majcen, B., Stanovnik, P., Stropnik, N., Bevc, M., Koren, M., et al. (2003). Predhodno vrednotenje Enotnega programskega dokumenta RS 2004-2006. Ljubljana: Institute for economic research, Research voor Beleid International, Institute Jožef Stefan.
Lolos, Suwa-Eisenmann, Zonzilo, Bourgignon (1995) “Evaluating the CSF With an Extended Computable General Equilibrium Model: The Case of Greece (1988–1995)” in Journal of Policy Modeling 17(2) pp. 177–197.
MK Projekt (2012): Vrednotenje ukrepov za spodbujenje raziskovalno razvojnih aktivnosti v gospodarstvu in institucijah znanja. Ljubljana, oktober 2012.
Mohl, P., & Hagen, T. (2008). Does EU cohesion policy promote growth? Evidence from regional data and alternative econometric approaches. Evidence from Regional Data and Alternative Econometric Approaches, 08-086.
Mohl, P., & Hagen, T. (2010). Do EU structural funds promote regional growth? New evidence from various panel data approaches.
Regional Science and Urban Economics, 40(5), 353-365.
Morgan, S. L. & Winship, C. (2009). Counterfactuals and Causal Inference: Methods and Principles for Social Research. New York: Cambridge University Press.
Oikos (2010): Vmesno vrednotenje Operativnega programa razvoja okoljske in prometne infrastrukture 2007-2013. Domžale, 19.oktober 2010.
Oikos (2011): Končno poročilo vmesnega vrednotenja 1. in 2. Prednostne usmeritve izboljšanja kakovosti in učinkovitosti sistemov izobraževanja in usposabljanja ter izboljšanje posameznika za delo in življenje v družbi, temelječi na znanju 3. Razvojne prioritete OP RČV
Razvoj človeških virov in vseživljenjskega učenja. Kamnik, avgust 2011.
101
Oikos (2012): Vrednotenje prednostnih usmeritev s področja trga dela, ki jih v okviru OP rČV izvaja Ministrstvo za delo, družino in socialne zadeve. Kamnik, julij 2012.
Pellegrini, G., Terribile, F., Tarola, O., Muccigrosso, T., & Busillo, F. (2013). Measuring the effects of European Regional Policy on economic growth: A regression discontinuity approach. Papers in Regional Science, 92(1), 217-233.
Pitija (2009a): Vrednotenje instrumenta Javni razpis za izbor razvojnih projektov za dvig zaposljivosti ranljivih družbenih skupin na področju kulture in podporo njihovi socialni vključ enosti v letih 2008–2009 s pripravo priporoč il za nadaljnje izvajanje aktivnosti za doseganje ciljev prednostne usmeritve 4.3 v OP RCV. Ljubljana, 14. december 2009.
Pitija (2009b): Vrednotenje javnega razpisa za sofinanciranje projektov horizontalnih NVO mrež in regionalnih stičišč za leto 2008.
Ljubljana, 6. Januar 2009 .
Pitija (2009c): Vrednotenje četrte razvojne prioritete »Razvoj regij« Operativnega programa krepitve regionalnih razvojnih potencialnov za obdobje 2007-2013. Ljubljana, 23. April 2009.
Polverari, L., Mendez, C., Gross, F., & Bachtler, J. (2007). Making Sense of European Cohesion Policy 2007–13: On-Going Evaluation And Monitoring Arrangements.
Romer, P. M. (1994). The origins of endogenous growth. The journal of economic perspectives, 3-22.
Rosenbaum, P. & Rubin, D. B. (1983). The Central Role of the Propensity Score in Observational Studies for Causal Effects. Biometrica, 70(1), 41–55.
Rubin, D. B. (1972). Assignment to treatment group on the basis of a covariate. Journal of Educational Statistics, 2, 1–26. 146.
Rubin, D. B. (1973a). Matching to Remove Bias in Observational Studies. Biometrics, 29, 159–83.
Rubin, D. B. (1974). Estimating Causal Effects of Treatments in Randomized and Non-Randomized Studies. Journal of Educational Psychology, 66, 688–701.
Rubin, D. B. (1976). Multivariate matching methods that are equal pecent bias reducing, I: some examples. Biometrics, 32, 109–120.
Rubin, D. B. (1992). Characterizing the Effect of Matching Using Linear Propensity Score Methods with Normal Covariates. Biometrica, 79, 797–809.
Simončič. M., Kuzmin, F., Pfajfar, L. and J. Potočnik (1999). The HERMIN macromodel of Slovenia, Working Paper 3, Ace-Phare project P96-6242-R, IER, Ljubljana.
Šlander, S. (2010): Ocena ekonomskega vpliva kohezijske politika v Sloveniji z uporabo kombiniranega mikro- in makroekonometričnega modela. Ljubljana: Ekonomska fakulteta.
Stryczynski, Kai (2008): Rigorous impact evaluation: the magic bullet for evaluation of Cohesion Policy?
[http://www.alt.degeval.de/calimero/tools/proxy.php?id=20232]
Varga, J., in ’t Veld, J. (2010), The Potential Impact of EU Cohesion Policy Spending in the 2007–2013 Programming Period: A Model-Based Analysis. ECFIN European Economy Economic Paper, no. 422.
Wooldridge, J. (1995). Selection corrections for panel data models under conditional mean independence assumptions. Journal of Econometrics, 68, 115–132.
102
Ernest Dautović
Université de Lausanne, Switzerland
Lucia Orszaghova
Národná banka Slovenska, Slovakia
Willem Schudel
De Nederlandsche Bank, The Netherlands
CONVERGING IN DIVERGENT WAYS:
EXPLAINING TRADE INTEGRATION BETWEEN CESEE COUNTRIES AND THE
EU-151
ABSTRACT
The period since the end of the Cold War has seen a rapid increase of two-way trade in similar products, or
intra-industry trade (IIT), between the EU15 and Central, Eastern and South-Eastern European (CESEE)
countries. IIT is an important mean to achieve real convergence towards the EU through trade
specialisation, and reduce the costs of an economic and monetary union. This paper assesses real,
nominal and institutional determinants of intra-industry trade between new EU member states, EU
candidates and potential candidates and the EU15 by using a product-level trade flow database and
employing linear and non-linear panel data specifications. Although the determinants of IIT for new EU member
states deviate considerably from those of candidate and potential candidate countries, the evidence suggests that
there exist common factors promoting IIT across the CESEE region, such as the corporate tax competitiveness,
the flexibility of exchange rate regimes and lower levels of corruption.
Keywords: economic integration, European Monetary Union, institutions, intra-industry trade, real
convergence, fractional response panel data.
JEL Classification: F14 (empirical studies of trade), F15 (trade-Economic Integration), F10 (Trade-general)
1. INTRODUCTION
One of the most profound economic developments of the past quarter of a century in Europe has been the
transformation of formerly centrally planned economies in Central, Eastern and South-Eastern Europe
(henceforth CESEE) towards open market-based economies and their closer economic integration and
convergence with the European Union. Trade has been an important aspect of this process: due to their relatively
closed economies, CESEE countries had only limited trade relations with the EU152 at the beginning of the transition process, whereas by 2010, the EU15 had become the destination of over 50 percent of these countries’
exports.
1 We would like to thank Letizia Montinari, Gilles Noblet, Leslie Papke, Mara Pirovano, Livio Stracca, Michael Sturm and Benjamin Vonessen as well as an anonymous referee for their valuable inputs on previous drafts of the paper. Mistakes and errors remain the sole responsibility of the authors.
2 The EU15 includes all countries which joined the EU before 1 May 2004, namely Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom.
103
One possible way to assess trade integration with the EU15 is by focusing on intra- industry trade between
individual CESEE countries and the EU15. Intra industry trade between countries entails a bilateral exchange of
the same type of goods and thus is regarded as a necessary pre-condition for real economic convergence,
external balance sustainability and trade competitiveness. Furthermore, the optimal currency area theory
suggests that it leads to more synchronised business cycles (Mundell, 1961, Frankel and Rose, 1998), a
necessary condition for the stability of a monetary union. In order to take stock of how far CESEE countries
have come with regard to trade integration with the EU, this paper focuses on factors which help determining IIT
between these countries and the EU15. Given the considerable differences between individual CESEE countries
in terms of convergence and integration with the EU15, this study distinguishes two groups of countries eight
EU candidate countries and potential candidates (CCPC)3 and eleven 'new' EU member states (NMS). 4
Analysing a period from 1998 to 2010 and employing variants of ordinary least square (OLS) models,
generalised method of moments (GMM) and fractional response models (FRM), we quantify the effects of
various macroeconomic and institutional variables on an aggregated country-level IIT measure.
The paper seeks to contribute to the literature in the following ways. First, it is the first study aimed at analysing
variables related to the quality of political institutions in combination with economic determinants of IIT.
Second, the analysis sheds a new light on efficient integration policies for CESEE countries, in view of the
importance of intra-industry trade in terms of achieving external competitiveness and real convergence which in
turn reduce the costs of a monetary union. As emphasised by Fidrmuc (2004) and the optimal currency area
literature, an increase in the share of IIT strengthens the synchronisation of business cycles within a monetary
union, reducing the costs of forsaking an autonomous monetary and exchange rate policy. Third, our results
show that the flexibility of the exchange rate regime is a significant factor that is conducive to higher IIT with
the EU15. This implies that on the convergence path toward a monetary integration a flexible exchange rate
seems to promote faster trade integration, a conclusion that is seemingly at odds with the recommendations of
the EU convergence criteria.
From a more technical standpoint, the main contribution of this study is the adoption of IIT indicators based on
highly disaggregated product-level bilateral trade data (at the 6-digit level) and the application of a
comprehensive econometric modelling approach which takes account of the structure and truncation of the
dependent variable (IIT). As such, this is the first study that tries to estimate the quantitative impact of CESEE
economic policies on IIT with EU15, not only by controlling for the sector and aggregation biases arising from
low levels of disaggregation, but also by taking the non-linear feature of our dependent variable appropriately
into account.
The structure of the paper is as follows: Section 2 provides a brief summary of the related literature on IIT and
Section 3 describes the data and the measurement of intra industry trade. Section 4 illustrates recent trends of
intra industry trade between the CESEE countries and the EU15. Section 5 describes our explanatory variables
and hypothesis, followed by the econometric analysis and results in section 6. Section 7 concludes.
2. RELATED LITERATURE
Intra-industry trade (IIT) first received scholarly attention in the 1960s, when it appeared to contradict prevailing
theories of international trade, which were based on the concept of comparative advantage and specialisation of
economies in particular types of goods. 5 The first vintages of IIT trade theory models predict that IIT would 3 The CCPC includes the following countries: Croatia (now a member of the EU, but a candidate country during the time span of the empirical analysis), Albania, FYR Macedonia, Montenegro, Serbia and Turkey (currently all EU candidate countries) and Bosnia-Herzegovina (EU potential candidate). Iceland, an EU candidate country, is also included, although it is not strictly referred to as CESEE
country, whereas Kosovo*, an EU potential candidate, was excluded from the sample due to data constraints.
4 The NMS includes the following countries that acceded to the EU on 1 May 2004 or 1 January 2007 respectively: the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, Slovenia, Bulgaria and Romania. Cyprus was also covered by this paper, although is not strictly referred to as the CESEE country, but Malta was excluded due to data availability.
5 See Verdoorn (1960); Balassa (1966) or Grubel (1967).
104
develop between countries which have a similar level of economic development and where specialisation would
be transferred at the firm level, whereas inter-industry trade (one-way trade) would prevail between partners with
differences in relative factor endowments. 6
One of the important contributions of this strand of theory is the distinction between horizontally and vertically
differentiated goods in intra-industry trade (HIIT and VIIT, henceforth), where the division is based on the
notion of product unit values (or 'quality').
Horizontal IIT (HIIT) is defined as a two-way trade in products of homogeneous quality, costs and technology
employed, but with different characteristics of attributes. The theoretical basis for this type of trade was
developed by Dixit and Stiglitz (1977), Lancaster (1980), Krugman (1979, 1981). It is associated with imperfect
competition or consumer preferences, but also with market structure (Brander and Krugman, 1983). It leads to
efficiency through economies of scale in production and welfare gains through a greater variety for consumers,
including producers’ gains in a variety of intermediate goods. The standard theoretical models suggest that the
share of horizontal IIT increases with a higher level of country similarity in terms of capital endowments.
Vertical intra-industry trade (VIIT) involves simultaneous imports and exports of same type of goods that are
differentiated by heterogeneous quality, technology or costs. The theoretical basis for this type of trade was
proposed by Falvey (1981), Shaked and Sutton (1984), Falvey and Kierzkowski (1987) and Flam and Helpman
(1987). These models entail a positive relationship between the level of vertical IIT and differences in capital
factor endowments. Countries specialise along the quality spectrum of a specific product, based on the
assumption that development of human or physical capital intensities are associated with higher product
qualities. Within this literature, economic distance between the countries, that is the distance in the accumulation
of physical or human capital, is hence a crucial determinant of VIIT and is not associated with inter-industry
trade as in the case of the pioneering contributions of Verdoon (1960) and Balassa (1966). 7
The link between regional integration and IIT has been under particular scrutiny since the beginning of the
European integration process and the abundant literature on trade patterns among EU15 countries provided the
base for the theoretical understanding of this phenomenon. There is a relatively ample literature on IIT in the
context of EU enlargement, in particular in the period around the accession of Central and Eastern European
countries in 200, 8 while a similar analysis for Western Balkan countries is almost non-existent. 9
This study emphasises the importance of studying IIT in the European context for two main reasons. First, the
evolution of trade patterns is an indicator of real convergence across countries: a higher degree of intra-industry
trade corresponds to an advanced level of economic integration, diversification of the economy and convergence
to EU15 industrial development levels. Second, as authors of the optimal currency areas suggest, a higher share
of intra-industry trade leads to a synchronisation of business cycles and a lower frequency of asymmetric shocks
between trading partners, the latter being a pivotal characteristic for the macroeconomic stability of a monetary
union. 10
6 See notably Helpman and Krugman (1985).
7 Fontagné and Freudenberg (1997) point out that intra-industry trade is not exclusively based on perfect competition and constant returns to scale and can occur without product differentiation, for instance in highly concentrated market structures and inter-industry trade can occur without comparative advantages (due to agglomeration effects or country size), . See also Balassa (1986) or Flam and Helpman (1987).
8 See among others Aturupane et al. (1997); Hoekman and Djankov (1996); Caetano and Galego (2007); Jensen and Lüthje (2009); Fidrmuc et al. (1999); Gabrisch (2006).
9 There are a few exceptions, namely Botrić (2012, 2013) and Mardas and Nikas (2008a and 2008b), however these authors use a lower disaggregation of data, which makes it rather difficult to compare their findings with ours.
10 For more details on the correlation between IIT and business cycles synchronisation see Frankel et al. (1998), Fidrmuc (2004), or Shin et al. (2003)
105
3. DATA AND MEASUREMENT OF INTRA-INDUSTRY TRADE
In order to quantify intra-industry trade we use a bilateral trade database at a 6-digit HS product level of
disaggregation (HS1996), which allows us to address any geographic or sector aggregation bias. 11 The data for this study were obtained from BACI, a detailed international trade database constructed by CEPII. Besides using
one of the finest product classifications available for international trade, BACI database also removes
discrepancies between import and export values providing comparable harmonised quantities. In addition, it
contains also data on the quantity of traded goods, allowing for the computation of goods’ unit product values.
Our initial product-level dataset spans the period 1998-2010, containing bilateral annual data for quantities,
imports and exports trade values for each traded product between each of the 15 EU members and every of 19
CESEE countries. The dataset provides an initial panel of approximately 18 million observations of bilateral
trade flows.
The dependent variable is intra-industry trade which is obtained by computing the Grubel-Lloyd Index (GLI) and
is based on the intensity (degree) of trade overlap for each individual product and partner. For each traded
product between two countries, a GLI is calculated based on the following formula, representing the intensity of
trade overlap:
𝐺𝐺𝑀𝑖𝑖′,𝑘,𝑖 = 100 ∗ �1 − �𝑋𝑖𝑖′,𝑘,𝑖−𝑀𝑖𝑖′,𝑘,𝑖��
(1)
𝑋𝑖𝑖′,𝑘,𝑖+𝑀𝑖𝑖′,𝑘,𝑖
where k represents a specific traded product, i the country in question and i’ the partner country, X represents exports, M imports and t stands for the year. Calculated in this way, the GLI takes a value between 0 and 100, where 100 indicates that all trade is intra-industry trade (two-way trade) and 0 that all trade is inter-industry (one-way trade). As a result, higher values of the index correspond to a larger involvement of a country in intra-industry trade with the EU15. The indices have been calculated according to equation (1) for each pair of trading
partners and for each product class.
Subsequently, the bilateral product-level GLI is aggregated to a country-level GLI, computing IIT between
each country in the sample of 19 CESEE countries and each partner country in the EU15. The GLI for a single
CESEE country vis-à-vis a country in the EU15 is a weighted average of the products GLIs, with weights given
by the share of product k in total trade with the partner country in the EU15.
≈5000
≈5000
𝐺𝐺𝑀
𝑘=1
𝑖𝑖′,𝑖 = �
𝑤𝑘,𝑖 ∗ 𝐺𝐺𝑀𝑖𝑖′,𝑘,𝑖 = 1 − ∑
�𝑒𝑖𝑖′,𝑘,𝑖−𝑀𝑖𝑖′,𝑘,𝑖�
≈5000
(2)
𝑘=1
∑𝑘=1 �𝑋𝑖𝑖′,𝑘,𝑖+𝑀𝑖𝑖′,𝑘,𝑖�
in which weights are given by:
𝑤𝑘,𝑖 =
𝑒𝑖𝑖′,𝑘,𝑖−𝑀𝑖𝑖′,𝑘,𝑖
(3)
∑≈
�
5000 𝑋
𝑘=1
𝑖𝑖′,𝑘,𝑖+𝑀𝑖𝑖′,𝑘,𝑖�
Following a similar weighting procedure, in which the weights correspond to the trade shares of partner
countries in the EU15, the data are grouped across partner countries in order to obtain a country-level GLI
defining IIT between every country in the CESEE region and the EU15 as one trading partner. Therefore, the
GLI index used for the following analysis explains a share of IIT in total trade with respect to the EU15.
11 A caveat in the calculation of the GLI is the geographic and the sector aggregation bias arising from a low level of data disaggregation.
The BACI database helps us to deal with this bias by providing a detailed product-partner trade database. On the BACI database see Gaulier and Zignago (2012). On the aggregation effect, see Grubel and Lloyd (1975), Greenaway and Milner (1986) or Fontagné and Freudenberg (1997). The HS stands for ‘Harmonised System’, which distinguishes about 5,000 product items.
106
In addition, we follow the procedure suggested by Fontagné and Freudenberg (1997), where only trade overlap
above a threshold of 10 percent is considered to be structural and hence of intra-industry nature, below this
threshold bilateral trade is considered to be one-way trade. Formally:
𝑀𝑖𝑛 �𝑋𝑖𝑖′,𝑘,𝑖𝑀𝑖𝑖′,𝑘,𝑖� > 0.1
(4)
𝑀𝑀𝑒 �𝑋𝑖𝑖′,𝑘,𝑖𝑀𝑖𝑖′,𝑘,𝑖�
Finally, all product classes have been further divided into horizontally and vertically differentiated products
(HIIT and VIIT, respectively) using the unit values, which are understood as proxies for quality. IIT is
considered of a horizontal nature if unit values satisfy the following equation:
𝑒
1
𝑈𝑈𝑘
1+𝑑 ≤ 𝑈𝑈𝑚 ≤ 1 + 𝑇
(5)
𝑘
in which 𝑈𝑈𝑒
𝑚
𝑘 and 𝑈𝑈𝑘 represent unit values of exports and imports of product k, and d is a chosen dispersion factor. If this condition is not fulfilled, IIT is considered to be vertically differentiated. Following most of the
studies, a dispersion factor of 15 percentage points is applied here. 12 Finally, lower and higher-quality vertical IIT are distinguished based on this dispersion factor, where lower quality VIIT is defined as the share of
bilaterally traded product classes of which the unit export value of CESEE countries is at least 15 percent lower
than that of the EU15, the opposite (greater than 15 percent) holds for high-quality VIIT.
4. RECENT TRENDS IN TRADE PATTERNS BETWEEN CESEE AND EU15
Figure 1a illustrates the evolution in IIT in Europe from 1998 to 2010 and Table 1 reports the average GLI index
over this period. Figure 1a is a triangular chart depicting the ‘ideal’ convergence path in terms of the overall
trade structure. At the beginning of the development trajectory, a country starts from the bottom left angle of the
triangle characterised by exclusively one-way trade. As it begins to integrate in a trade area, it increases its IIT
share. It reaches convergence when it is located at the centroid of the isosceles triangle. This point consists of
balanced proportions of one-way trade (25 percent) and horizontal IIT (25 percent) but a competitive edge in
vertical IIT (50 percent).
Figure 1a - Evolution of IIT in EU15 and CESEE countries with respect to the EU15.
Note: Blue dots indicate candidate countries and potential candidates, red dots belong to the new EU member states and the green dots represent the EU15 countries. The horizontal lines intersecting the triangles indicate the level of vertical IIT (for example France (FR) in 1998 had around one-third of its trade in vertical IIT with the EU15). The one-way trade is revealed drawing a negatively sloping parallel line from the base of the triangles (hence, France in 1998 had half of its trade with the EU15 of this nature). Similarly, for the horizontal IIT, parallel lines have to be drawn from the right hand side to the dot representing a country (almost 20 percent of total trade of France in 1998
was two-way trade of similar quality).
12 For a discussion about the different dispersion factors, see Fontagné and Freudenberg (1997) and Aturupane et al. (1997). We perform our analysis with the dispersion factor of 25 percent, our results do not change qualitatively.
107
As indicated by an increase over time in the IIT share in total trade, the transition and integration process of
many CESEE countries has been accompanied with profound changes in the composition of trade patterns with
the EU15. The figures for IIT differ substantially among individual countries, spanning from less than 2
percentage points in the case of Montenegro to almost 40 percentage points in the case of the Czech Republic.
Most new member states have higher GLI levels than candidate and potential candidate countries, indicating
their higher integration and convergence with trade patterns in EU15. As Figure 1a illustrates, some new
member states have reached IIT levels of EU15 countries by 2010. This is in particular true for the Czech
Republic, Hungary, Poland and Slovenia.
Table 1. IIT average 1998-2010 with EU15
Mean
S.E.
Minimum
Maximum
EU12
Bulgaria
13.5
1.9
11.4
17.6
Cyprus
11.8
2.6
8.3
16.5
Czech Republic
39.2
0.6
38.4
40.5
Estonia
18.4
2.2
14.4
21.8
Hungary
27.8
1.4
25.7
29.6
Latvia
8.2
2.9
5.3
14.4
Lithuania
9.7
2.3
6.4
13.5
Malta
19.3
5.3
11.2
26.8
Poland
27.3
4.8
18.7
34.2
Romania
15.7
4.7
9.1
24.1
Slovakia
22.0
2.4
17.8
26.2
Slovenia
26.0
1.7
22.6
28.2
CCPC
Albania
20.9
2.4
16.2
24.2
Bosnia-Herzegovina
9.8
2.9
5.9
13.6
Croatia
19.8
1.4
17.3
21.4
FYR Macedonia
5.2
2.6
2.8
11.1
Montenegro
1.8
0.4
1.3
2.4
Serbia
10.4
2.4
7.4
14.6
Turkey
15.5
3.4
9.6
20.2
The divergence between NMS and CCPC can also be observed with respect to development of IIT shares over
the period of 12 years. Some CCPC have been losing their positions in IIT while increasing the share of one-way
trade with EU15. Most prominently, Albania and Croatia have lost IIT shares despite still enjoying relatively
high GLI, but also Montenegro and FYR Macedonia, where IIT levels in 2010 constituted less than 5 percent of
total trade, have witnessed falling shares (Figure 1b).
108
Figure 1b. Horizontal and Vertical IIT levels in CESEE countries in 2010
When looking at different components of IIT in Figure 1b, it shows that vertical IIT continues to dominate two-
way trade, pointing to a specialisation along quality range between CESEE and EU15 countries. Between 1998
and 2010 most of CESEE have increased their share of higher-quality vertical IIT as well as the share of
horizontal IIT in total IIT, indicating a continued convergence toward the EU15 industrial structures. This could
point to a relative improvement in the quality of goods produced by CESEE countries. NMS from Central
Europe, but also Romania, Turkey and Serbia are worth highlighting in this respect as countries with the highest
improvements in the quality of their products (in both relative and absolute terms).
The downward trend of IIT for these countries is persistent over the whole period, also when analysed separately
for the pre-crisis and crisis period, (Figure 1c). This is rather surprising, since the closer integration appears to
have led to IIT divergence between them and EU15. Most of the other countries have recorded convergence of
IIT patterns, which does not seem to be weakened even during 2008-2010 crisis period.
Figure 1c. Overall changes in IIT levels in CESEE countries
Note: Simple differences between first and last year of the two periods are showed. The difference between these two periods could be used as a proxy for the change since the crisis.
109
5. EXPLANATORY VARIABLES AND HYPOTHESIS
5.1. Unit Labour Cost
The cost of labour is one of the most frequently tested determinants of a country’s external competitiveness, but
to our knowledge no other paper has included labour cost differences as a determinant for IIT development. 13
ULC differences can promote IIT by increasing competitiveness through lower wage dynamics. 14 For instance, in the recent debate on competitiveness within the euro area, adjustment is considered to pertain exclusively to
the wage side of this relationship (ECB, 2012).
ULC with the EU15 is computed as the difference between the average unit labour costs in EU15 and the ULC in
a CESEE country. In addition, following Felipe and Kumar (2011), ULC is disentangled into the wage share of
labour in total production and a price deflator. From an estimation point of view, the split of ULC into two components allows relaxing the common parameter restriction, so as to assess the relative effects of wage and
general price dynamics on IIT.
The data for ULC come from the IMF’s International Financial Statistics and Eurostat, the ULC is specifically
measured as follows:
𝑤
w
𝑤𝐺
ULC = 𝐴𝐺𝑃 = (𝐺𝐺𝑃 𝑃
⁄ ) 𝐺
⁄ = 𝐺𝐺𝑃 ∗ 𝑃 = 𝐺𝑇𝑢𝑜𝑢𝑇 𝑆ℎ𝑇𝑇𝑇 ∗ 𝑃
(6)
Where w is the average money wage rate, ALP represents average labour productivity, P represents the price deflator index, GDP/P is the real GDP, L is the number of employed persons in a country. This distinction is primarily aimed at disentangling the specific impact on IIT of wage share vis-à-vis general price inflation in the
economy.
5.2. Capital Endowments
Capital endowment differences play an important role in international trade theory, both for the pattern and the
volume of trade. 15 IIT is viewed as a consequence of vertical product differentiation based on quality, driven by initial differences in endowments, labour productivity and technological possibilities. Ceteris paribus, a larger
stock of physical capital is assumed to increase productivity as well as the comparative advantage in
endowments and, thereby, country competitiveness. Two measures of capital in the economy are included: the
domestic stock of capital (stock of physical capital) and foreign direct investment inflow (investment capital,
FDI).
The computation of the stock of capital is performed via the perpetual inventory method, assuming an annual
depreciation rate of 15 percent and an annual output growth rate of 3 percent. 16 For the analysis, the natural logarithm of the difference in capital stock between the EU15 average and the country in question is used. We
expect a negative relationship between the domestic stock of capital and IIT.
FDI has been a major source of capital investment and technology transfer in CESEE countries. These long-term
investments have accelerated productivity convergence as well as convergence of trade patterns towards the
trade structures of advanced EU15 countries. 17 Due to countries’ proximity to the EU15 and a relatively lower wage costs, foreign companies have found it attractive to delocalise their production processes into CESEE
13 The cost of labour has been considered in the IIT literature only implicitly, where differences in labour endowment were assumed to include different labour costs.
14 See Marin (2006).
15 See Helpman et al. (1985), Falvey et al. (1987) and Falvey (1981).
16 For details on the methodology, see Dhareshwar and Nehru (1993). Data come from the World Bank Development Index.
17 See Bijsterbosch and Kolasa (2009).
110
countries. A measure of net FDI inflows as a percentage of GDP is included in the analysis so as to account for
the foreign long-term investment channel. Following the standard new-trade theory argument of capacity
building and product differentiation, we expect a positive impact of FDI on IIT for CESEE countries.
Moreover, FDI is linked to the fragmentation of production processes in Europe, with parent companies
specialising in capital intensive activities, whereas labour-intensive activities have been entrusted to their foreign
affiliates (efficiency-seeking FDI). The interaction of FDI and labour-intensive technologies can have
inflationary effects on the share of wages in GDP and promotes economies of scale; the latter can in turn
increase IIT (Helpman and Krugman 1985).
At the same time, Markusen (1984, 2002) shows how the story can be exactly opposite as FDI can substitute for
trade (domestic market-oriented FDI) on a global production scale, which would have a negative impact on IIT.
Similarly, Gaulier et al. (2012) explain how FDI inflows can have a direct demand effect on both tradable goods
and non-tradable goods in the domestic economy and tend to increase price levels in the tradable sector,
appreciating the real exchange rate and making the tradable goods less competitive. In our specifications we
control for this relation, by including an interaction terms between FDI and the deflator component of ULC.
5.3. Trade Agreements
The pioneering theories on IIT were developed in relation to the signing of the first regional trade agreements, in
particular between countries of the European Economic Community. Most of the early empirical studies found
some evidence that regional trade agreements stimulate intra-industry trade. 18 However, there appears to be some disagreement in the literature when it comes to the effect of trade agreements on trade patterns between
economically and geographically diverse countries. Some empirical studies suggest that the elimination of trade
barriers contributes to an increase in IIT, linked to the re-export to a richer country of goods assembled in the
lower-income country (Foster et al., 2010). Other empirical studies (Rodas-Martini, 1998) show that the impact
of trade agreements on IIT is statistically not significant, suggesting that the increased competition among local
and foreign firms due to the removal of trade barriers can imply that a relatively less developed country will not
be capable of exploiting the benefits of the opening towards a new market. In other terms, the opening of
markets can induce specialisation based on revealed comparative advantages and hence promote one-way trade.
It follows that the overall impact of trade agreements can either be positive or negative, depending in particular
on the quality of goods the two countries are able to supply. 19
Since CESEE countries have been partners in different trade agreements with the EU15, these agreements are
controlled for separately by means of including dummy variables for preferential trade agreements (PTA), free
trade agreements (FTA) and EU membership.
5.4. Exchange Rate Regime
Most of the empirical literature investigates the effects of the exchange rate regime and exchange rate volatility on the volume of bilateral trade, testing the underlying assumption that uncertainty about the final prices of
traded goods reduces the value of bilateral trade flows. 20 This is in line with the notion that a monetary union eliminates any exchange risk from transactions and thus promotes trade. The effect of exchange rate regime on
IIT has been studied in connection with the introduction of the Euro in 1999 and creation of the European
Monetary Union. 21
18 See Grubel and Lloyd (1975) and Balassa and Bauwens (1987).
19 Related to the increased competition argument see Herderschee and Qiao (2007) on the importance of sequencing in opening up domestic markets to foreign trade.
20 See Baldwin et al. (2005).
21 See Fontagné et al. (1999).
111
It has been argued by Fontagné et al. (2005) that different trade types are not affected in the same way by
exchange rate volatility. The authors argue that if the perceived elasticity of demand is very high, small
variations in exchange rates may have a large impact on IIT, with particular influence on horizontal IIT where
the products are differentiated by some minor attributes. Despite not directly addressing the issue of exchange
rate regimes, following this argument the elimination of exchange rate volatility would benefit IIT by reducing
trade transaction costs and related financial uncertainties.
However, a floating regime can serve as an absorber of external shocks, nominal depreciation vis-à-vis trading
partners can make the tradable sector more competitive and thus increase export volumes. In particular of the
low quality and horizontal type of traded goods since high quality vertical goods are intrinsically competitive. A
nominal depreciation can provide a cost efficiency mechanism to firms in developing countries trying to enter
and export their goods in more developed markets.
Therefore, the overall effect of the exchange rate regime on IIT for the group of CESEE countries is an empirical
issue. To estimate this effect, we included the Reinhart and Rogoff (2004) index on the exchange rate regime in
each country in a particular year. The index varies from 1 to 14 where number 1 represents a country with no
separate legal tender (for instance Montenegro) and 14 represents a country with a freely floating currency. 22
5.5. Institutional variables
In addition, three novel institutional variables are included in the analysis and used as differences from the EU15
average.
First, we account for the extent to which corruption is perceived in a country. The main rationale is that
corruption serves as an invisible tax on business and has been shown to reduce investment and growth. 23 We use data from Transparency International’s Corruption Perception Index, which run from 1 (very corrupt) to 10 (free
from corruption).
Furthermore, a discrete variable measuring the level of democracy is included in order to control for the broad
political environment. The data are sourced from the Polity IV database and they range between minus 10 (low
level of democracy) and 10 (high level of democracy).
Third, we include a variable on corporate taxation rates, collected from KPMG Global Corporate Tax Data, and
test whether differences in corporate tax rates can explain IIT. The corporate tax policy can be used by
governments to increase price competitiveness among domestic exporters which can facilitate trade integration
of CESEE countries with the EU15.
6. IDENTIFICATION STRATEGY AND ESTIMATION RESULTS
As indicated above, the database contains an annual panel dataset covering the period from 1998 up to 2010 for
19 CESEE countries, namely eleven NMS and eight CCPC. 24 The identification strategy is based on a set of dynamic panel regressions, 25 they account for endogeneity by exploiting IV-GMM estimators.
22 In line with the literature, our choice to study the exchange rate regime (as opposed to exchange rate variability) is motivated by an interest in monetary policy analysis and policy choices of monetary authorities. We included in separate regressions the exchange rate variability with respect to the Euro without finding statistically significant effects on IIT shares. Results are available from authors.
23 See for instance Barro (1996) and Shleifer and Vishny (1993).
24 For Montenegro and Serbia, the sample period is limited to 2006-2010, due to data availability.
25 In order to account for unobserved heterogeneity, we use fixed-effect panel estimations and employ robust standard errors to account for heteroscedasticity and serial correlation in the pooled residuals. We perform a Hausman test to see if a random effect model is more appropriate to use here. Yet, the test rejects this hypothesis. Nevertheless, we run several random effect specifications in order to allow for the effects of geographic and cultural time invariant variables. The results of these models are available upon request.
112
As noted previously, the GLI index explains a percentage share of IIT in total trade with respect to the EU15,
and is truncated at 0 and 100 percentage values by construction. A caveat is that the linear dynamic panel
regression cannot provide meaningful quantitative effects for the covariates due to the truncated nature of the
dependent variable: linear panel estimation would predict values outside of the specified boundaries, an outcome
that would be hard to reconcile with a meaningful economic interpretation. Therefore, a non-linear model able to
account for the continuous (yet bounded) nature of the dependent variable is needed. As such, two
methodologies are applied in this paper: a logistic transformation on the response variable and a fractional
response model methodology developed by Papke and Wooldridge (2008).
Given the longitudinal structure of the data, a series of pooled ordinary least squares (OLS) models is estimated.
The panel baseline regression specification takes the following functional form:
𝑔
LN � 𝐼𝐼𝐼𝑖𝑖 � = 𝛼𝑔 + 𝛾𝑔 + 𝛽
� + 𝛽
� + 𝛽
𝑔 + 𝛽
𝑔 + +𝛽
𝑔 + 𝛽
𝑔 +
1−𝐼𝐼𝐼
𝑖
1𝐺𝐿 � 𝐼𝐼𝐼𝑖𝑖−1
2𝐺𝑛 � 𝐾
3𝑈𝐺𝑈
4𝐹𝐺𝑀
5𝑇𝐴
6𝐸𝑋𝐸
𝑖𝑖
𝑖
1−𝐼𝐼𝐼𝑖𝑖−1
𝐺𝐺𝑃 𝑖𝑖
𝑖𝑖
𝑖𝑖
𝑖𝑖
𝑖𝑖
𝛽
𝑔
𝑔
7𝑀𝐿𝑆𝑇𝑖𝑖 + 𝜀𝑖𝑖
where i and t are the usual subscripts indicating respectively an i-th country and a t-th year. The superscript g is an index for a group of countries and represents either NMS or CCPC, αi and γt represent respectively the cross-section fixed effect.
ULC is the unit labour cost variable, FDI represents the share of foreign direct investments in the economy over the GDP. The row vector Ln(K/GDP) captures the effect of the natural logarithm of the capital stock (K) on IIT.
The variable TA is a row vector of dummy indicators aimed at capturing the effect of trade agreements between
the EU15 and the i- th country. The variable EXR is a discrete variable explaining the exchange rate arrangements for monetary policy. The indicator INST is a row vector capturing the effect of institutional variables on IIT, in particular three discrete variables, namely the Corruption Perception Index (CPI), an aggregate indicator of the
level of democracy process in a country and the differences of the corporate tax rate from the EU15 average. It is
important to recall that most explanatory variables, namely ULC, capital stock, wage share, corruption and
democracy indexes, enter the equation as the difference with respect to the EU15 average, to better capture the
(economic) distance from this ‘benchmark’ region. Intuitively, the difference with respect to the EU15 represents
the distance from the desired level of economic integration.
Table 2 illustrates the main descriptive statistical properties of our group of explanatory variables as well as of
IIT. Except for the inflow of FDI scaled by the GDP, the means between the NMS and the CCPC deviate
substantially. These differences render it likely that the effects of explanatory variables are heterogeneous in
these two groups. Therefore, in what follows two sets of specifications are presented: one for new member states
and one for the candidate and potential candidates. In addition, the motivation for a separate set of regressions is
also policy-driven: NMS and CCPC are subject to different trade agreements and institutional arrangements with
the EU15 and require different dummy variables capturing the effect on IIT.
113
Table 2. Descriptive statistics for CCPC and NMS countries
Candidates & Potential Candidates
Obs.
Mean
Std. Dev.
Min
Max
IIT with EU15 (%)
96
11.84
6.73
1.30
24.17
ULC (diff with EU15)
104
0.15
0.12
-0.27
0.33
Wage Share (diff EU15)
104
0.13
0.45
-3.73
0.33
Inflation (%)
104
12.69
22.46
-1.58
137.96
Net FDI inflow (% GDP)
88
6.21
7.28
0.31
36.88
Ln(Capital stock/ GDP) (diff EU15)
102
0.29
0.32
-0.41
1.53
Corporate taxation (diff EU15)
104
11.28
6.55
-6.05
21.41
PTA
104
0.63
0.48
0
1
FTA
104
0.42
0.50
0
1
FX regime
104
8.07
4.29
1
14
Corruption perception (diff EU15)
104
3.70
2.16
-2
6.30
Democracy (diff EU15)
83
3.15
2.93
0.92
15.92
New Member States
Obs.
Mean
Std. Dev.
Min
Max
IIT with EU15 (%)
156
19.91
9.18
5.28
40.47
ULC (diff with EU15)
143
0.04
0.06
-0.12
0.13
Wage Share (diff EU15)
143
0.05
0.05
-0.09
0.15
Inflation (%)
156
5.96
8.11
-3.71
55.22
Net FDI inflow (% GDP)
142
6.20
7.65
-32.88
52.05
Ln(Capital stock /GDP) (diff EU15)
156
0.07
0.16
-0.55
0.48
Corporate taxation (diff EU15)
156
7.79
6.06
-8.46
19.70
EU Member
156
0.5
0.50
0
1
FTA
156
0.42
0.50
0
1
FX regime
156
6.83
3.55
1
14
Corruption perception (diff EU15)
143
2.85
1.09
0.69
4.93
Democracy (diff EU15)
143
0.63
0.84
-0.08
3.92
6.1. Logistic Transformation Dynamic Panel Estimates
The results of the dynamic panel regressions are presented in Table 3, with first three columns covering CCPC
and the second part covering NMS. Standard diagnostic tests and regression statistics are shown at the bottom of
the table. Following the results of the Monte-Carlo experiment by Judson and Owen (2001) on macro dynamic
panels, the estimates are run with the one-step Arellano-Bond-GMM estimator. Four lags are implemented for all
GMM-type instruments, namely the lagged dependent variable and all other our covariates except the time
dummies and the unobserved cross-country heterogeneity. 26 All specifications have exogenous instruments, as confirmed by the Sargan test, and there is no sign of second or higher order correlation between the lagged
dependent variable and the error term.
26 The choice of four lags warrants an appropriate degree of balance between the bias-efficiency trade-off, see by Judson and Owen (2001).
We adopt the one-step GMM estimator which performs better than the two-step GMM estimator as reported in Arellano-Bond (1991) and Kiviet (1995). We include time-variant fixed effects in all our reported specifications.
114
Table 3. Determinants of IIT Central Europe and South Eastern European Countries
CCPC
New Member States
(1)
(2)
(3)
(4)
(5) (6)
(7)
LT(IIT) Lag(-1)
0.461***
0.516***
0.435***
0.411**
0.411**
0.414**
0.494**
(0.125)
(0.133)
(0.056)
(0.203)
(0.206)
(0.202)
(0.202)
ULC, diff EU15
0.951
-0.114
(0.641)
(0.698)
FDI/GDP
-0.013**
0.226***
0.294***
0.002
-0.099*
-0.072
-0.084
(0.005)
(0.085)
(0.076)
(0.006)
(0.058)
(0.060)
(0.056)
FDI*ULC, diff EU15
-0.005
-0.007
(0.119)
(0.047)
Ln(K.Stk/GDP)diffEU15 -0.373*
-0.569***
0.007
0.169
(0.222)
(0.158)
(0.273)
(0.265)
FDI*Ln(K.Stk/GDP)
0.018*
0.053**
-0.005
-0.024
(0.010)
(0.025)
(0.034)
(0.033)
Corp.Tax, diff EU15
0.013**
0.013**
0.023**
0.004
0.005*
0.008**
0.009**
(0.007)
(0.007)
(0.011)
(0.003)
(0.003)
(0.004)
(0.004)
W/GDP, diff EU15
1.467
0.708
-0.339
-0.843
-0.896
(1.173)
(0.566)
(0.930)
(0.807)
(0.835)
Deflator
-0.327***
0.947***
-1.021**
-1.155*
-1.22**
(0.119)
(0.308)
(0.513)
(0.622)
(0.591)
FDI*W/GDP,diff EU15
0.046
0.053*
0.017
0.003
0.009
(0.060)
(0.032)
(0.025)
(0.015)
(0.018)
FDI*Deflator
-0.196**
-0.252***
0.109
0.071
0.087
(0.077)
(0.061)
(0.068)
(0.061)
(0.058)
FTA (EU member)
-0.199***
-0.128**
0.060*
(0.073)
(0.062)
(0.033)
XR Regime
0.115***
0.015***
0.015***
(0.034)
(0.004)
(0.004)
Corruption, diff EU15
-0.208***
-0.006
-0.009
(0.038)
(0.027)
(0.022)
Democracy, diff EU15
0.014
-0.024
-0.024
(0.019)
(0.039)
(0.039)
Intercept
-1.244***
-0.792***
-2.523***
-0.812**
0.122
0.352
0.375
(0.339)
(0.301)
(0.714)
(0.320)
(0.482)
(0.611)
(0.588)
Time Fixed Effects
Yes Yes Yes
Yes Yes Yes
Yes
F-Test .000*** .000***
.002*** .000*** .000*** .000*** 0.000***
SW -Test
.000*** .000***
.000*** .000*** .000*** .000*** 0.000***
Sargan Test
0.206 0.336
0.183
0.623 0.679 0.594
0.587
BIC
203.82 210.39
148.97
235.00 242.32 207.70
210.26
AB Test - 1st order
0.029** 0.024**
0.024** 0.117 0.120 0.105
0.104
AB Test - 2nd order
0.089* 0.835
0.855
0.323 0.300 0.352
0.334
AB Test - 3rd order
0.370 0.837
0.367
0.264 0.246 0.235
0.237
Observations
73 73 62
110 110 106
118
N. of countries 8 8 7
11
11 11
11
Note: p<0.1*, p<0.05**, p<0.01***.Coefficients: Std. Errors in parentheses robust with respect to serial correlation and heteroscedasticity.
F-Test, p-values for joint significance of time fixed effects under Ho: no joint effect of time fixed effects. SW (Shapiro-Wilk) test for normality of residuals, p-values reported under Ho: residuals are normally distributed. Sargan Test for over-identifying restrictions, p-values reported under Ho: the instruments as a group are exogenous. AB (Arellano-Bond) test for autoregressive residuals of 1st, 2nd or 3rd order, reported p-values for Ho: no serial correlation.
115
For both groups results show the importance of the dynamic lagged effect of IIT, confirming that IIT is an
intrinsically dynamic concept. 27 In columns 1 and 2 the marginal effect of the differential in the stock of fixed capital formation in CCPC is significant only after controlling for the interaction with the inflow of FDI. The
interaction between the FDI and the capital stock is statistically significant and positive, suggesting that FDI
contributes to the accumulation of the stock of capital in CCPC and, as standard theory predicts, the marginal
effect of capital stock distance on IIT is negative. The variables for economic distance, and in particular physical
capital stock, are not significant for the NMS countries (columns 4 and 5). 28
At first inspection, no direct evidence of significant effects of ULC differentials on IIT is found for both groups.
In other words, the generally lower levels of ULC in CESEE do not present a significant impact on IIT. In order
to further understand the effect of ULC distances on IIT, we decompose them in two components: the share of
labour compensation in GDP and the general level of prices in the domestic economy, both measured as the
distance from the average EU15 values. The increasing price level, as measured by the deflator, appears to be a
stronger and more significant determinant than the wage share of GDP in deterring IIT with the EU15. The
marginal effect of price levels is statistically significant and negative for both NMS and CCPC whereas the
effect of labour compensation share is not statistically significant. However, and only in the case of CCPC
(column 3), it is interesting to notice that the interaction between FDI and the deflator is significant and negative.
This suggests that the inflow of FDI, which CCPC countries witnessed during the period, has brought
generalised upward pressures on price levels. Importantly, the effect of the interaction of FDI and price deflator
indicates that local monetary authorities could have played a greater role in containing inflationary pressures
promoting IIT with the EU15.
The marginal effect of FDI is slightly positive and significant. In our preferred specification for CCPC (column
3), the marginal effect for CCPC, evaluated at mean deflator value (1.1269) and mean wage share of GDP
difference (0.13), the impact of FDI on IIT is significant and positive. 29 FDI thus appears to contribute to the capacity building in CCPC region and mitigates the lag in the convergence process. A caveat is necessary, due to
the previously positive interaction of FDI and the general level of prices, the positive effect of FDI is supported
only if the inflationary pressures on the tradable sector are well managed. 30
Policy rather than structural variables seem to have a higher weight in determining IIT for the NMS. With this
respect, the impact of FDI on IIT does not show the same patterns as for the CCPC. The inflow of FDI into the
NMS does not have a strong significant effect on IIT even after interacting FDI with the stock of physical
capital. Intuitively, given a lower NMS gap of the physical stock of capital with respect to the EU15, the inflow
of FDI did not have a significant marginal contribution to the increase of IIT.
A very important result of these estimates, which is robust to different specifications and across the two sets of
countries, is the one of the corporate tax distances. In all specifications, the corporate tax difference with respect
to the EU15 is statistically significant and positive. Corporate tax differentials can have a strong effect on trade
patterns by making one country’s exports relatively cheaper than the similar goods produced in the EU15. The
transmission channel allows a lower corporate tax rate to provide room for a more competitive pricing of
tradable goods and hence promotes IIT with the EU15. This result underlines how the tax structure can be a very
important policy lever to increase trade integration with the EU and emphasises further the importance of the
fiscal discipline which could allow for some fiscal space to lower (or maintain a low level of) corporate taxes to
promote trade integration with the EU15 block. 31
27 See the dynamic marginal IIT contribution by Brülhart (1991).
28 We run separate regressions where we include also GDP per capita distance from the EU15 average as a measure of economic distance: the estimates are insignificant. The results are available from the authors.
29 Using the coefficients of column 3 we perform the following computation: 0.294+0.053*0.13-0.252*1.1269=0.017.
30 As already discussed by Gaulier et al (2012), a positive shock of foreign capital helps to build economic capacity but the influx of long-term investment creates an internal demand shock, that is the demand (and thus price level) for tradable and non-tradable products in the domestic economy increases, making the export sector relatively less competitive.
31 A caveat is due, corporate tax rates can also influence the impact of FDI on IIT (see OECD 2007). Most of CCPC have a sizably lower corporate tax rate than EU15 countries (on average 11.3 percentage points, Table 2), acting as a catalyst in attracting FDI inflows. To control 116
As for the effect of free trade arrangements, they have a significant and negative effect on IIT among CCPC
(column 3). Yet, this result should not be surprising due to the symmetric nature of such agreements coupled
with the lower exporting capabilities of the CCPC compared to the EU15. 32 Similarly to candidate countries, the symmetric FTAs that were in place prior to the EU accession in most of the NMS have a markedly negative and
statistically significant coefficient (column 6). 33 When splitting the regressions including the EU Membership binary variable in column 7 instead of the FTA dummy, a positive and significant effect for the EU membership
is found for the NMS. 34 The estimate for the EU membership dummy indicates that only after a period of real and nominal convergence to EU rules and a period of transition from socialist industrial structures, NMS
benefited from the EU partnership in terms of IIT convergence.
The variable for the exchange regime has a positive and highly significant impact on IIT for both groups of
countries, suggesting that a less restricted exchange rate mechanism allows less competitive countries to enter
EU15 markets via standard competitive devaluation argument. The result reconciles with the notion of
devaluation as an adjustment tool to gain external competitiveness relative to trade partners. Is important to
emphasize then that the positive effect of flexible exchange rate regime on IIT is not in line with the assertion of
fixed exchange rate being a mean to eliminate exchange rate risk, anchor inflation expectations and thus promote
trade, which is the standard prescription a country receives when applying for the single currency.
Another interesting result in Table 3 is the heterogeneous effects of corruption on CCPC and NMS, pointing to
the importance of the institutional convergence of CESEE to the EU. In particular, the corruption perception
index has a very strong and significant effect for CCPC, indicating that higher relative levels of corruption
reduce IIT. 35 At the same time corruption distance from EU15 average does not have any significant effect on IIT for NMS. The absence of any impact of corruption for the NMS is not surprising since these countries have
undergone recently a comprehensive legislative confluence path toward the EU’s acquis communautaire, which
reduced overall investment risk and increased the trust of trading partners. The overall impact of the institutional
convergence process, despite still incomplete given relatively higher levels of corruption in some of the NMS
than in the EU15, has been successful and did not harm the IIT flows with the EU15.
6.2. Fractional response Model
In this section we address the bounded nature of our dependent variable by exploiting the non-linear fractional
response model. Our pooled fractional probit model36 has the form:
𝔼 �𝑀𝑀𝑇
𝑔
𝑔
𝑔
𝑖𝑖�𝒙𝒊𝒊,𝒙𝒊𝒊, … 𝒙𝒊𝒊 � = 𝛷�𝛾𝑀𝑖 + 𝒙𝑖𝑖𝛽𝑀 + 𝒙
�𝑖 𝜗𝑀�
It is estimated using the one-step pooled Bernoulli quasi-MLE (QMLE) derived by maximising the pooled
probit log-likelihood. To correct for arbitrary serial dependence and misspecified conditional variance37 robust standard errors are used. We then compute the partial effects averaged across the population, the average partial
𝑔
effects (APE), to have an estimate of the relative importance of the various determinants. The variable 𝛾𝑀𝑖
represents the intercept and the subscript t indicates that the average IIT is allowed to differ across years. As for this interaction, we include interaction terms of FDI with corporate tax differences in other specifications. However, the interaction is not significant. The estimates are available from the authors.
32 In a different set of estimations we find some evidence that a more asymmetric trade agreement favouring the relatively weaker countries such as the Preferential Trade Agreement (PTA) has a positive effect on IIT. Nevertheless, the PTA effect is not robust across different specifications.
33 Similar results have been found by Herderschee and Qiao (2007).
34 From a trade integration perspective, the fundamental differences between the EU Membership and the FTA is that the former encompasses also free movement of factors of production and a common external trade policy whereas the latter is a mere removal of trade tariffs and quotas with no common trade policy.
35 Similar results have been documented in the trade and corruption literature. See for example de Jong and Bogmans (2011) with respect to corruption at border.
36 We use the probit model as it was shown to be superior to the conditional logit estimation, the latter is not consistent when the response variable is not binary and serial dependence is an issue. For more details see Wooldridge (2002), section 15.8.3.
37 In an alternative estimation method we allow for misspecifications in the conditional variance and adopt the generalised estimating equation approach (GEE) with an exchangeable working correlation matrix. The results, available from the authors, are very similar to the Bernoulli QMLE.
117
before, g represents either the CCPC or NMS countries. The subscript a represents the scaling factor: in fact, all of the QMLE estimated coefficients depend on the scaling factor a, without it the QMLE coefficients would not
𝑔
be identifiable. 38 The explanatory variables are represented by the matrix 𝒙𝑖𝑖. Importantly, the inclusion of the
𝑔
time averages of the covariates ( 𝒙
�𝑖 ) controls for correlation between country unobserved fixed effects and the
covariates and helps in estimating, with relative ease, the coefficients of interest up to a scaling factor. 39
Table 4 illustrates the results of the pooled Bernoulli quasi-MLE estimator for the two groups of countries.
Although the coefficients of the pooled fractional response model (that is columns 1, 3, 5 and 7) can be used to
evaluate qualitative effects, they do not have meaningful quantitative economic interpretation. To gauge the
quantitative effect of the covariates we refer to the average partial effects (columns 2, 4, 6, and 8) where we use
the scaling factor to obtain the APE coefficients and bootstrapped standard errors.
The non-linear estimates show that the APEs have the same qualitative signs as the dynamic panel regressions
although there are some fundamental differences in terms of quantitative effects and statistical significance of
some variables. For both groups of countries, it is safe to confirm that the dynamic inertial effects are an
important feature of intra-industry trade: the coefficients on the lagged dependent variable are close to 0.5,
suggesting that about half of the intra-industry share in one year is carried over to the next. For the NMS and
mirroring the linear models, most of the explanatory variables are not significant except for the floating
exchange regime which has a positive effect. In the following we focus our attention on the results for the
candidate and potential candidates, i.e. columns 2 and 4.
Table 4. Pooled Fractional Response Model and the APEs for CCPC and NMS
Candidates and Potential Candidates
New Member States
Dep. Var.:
Pooled
APE
Pooled
APE
Pooled
APE
Pooled
APE
IIT with EU15
QMLE
QMLE
QMLE
QMLE
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
IIT Lag(-1)
2.694***
0.741***
1.898***
0.555***
1.601**
0.475**
1.600**
0.481**
(0.609)
(0.065)
(0.625)
(0.079)
(0.688)
(0.211)
(0.622)
(0.189)
FDI /GDP
0.092**
0.025***
0.088***
0.025***
-0.054
-0.016
-0.031
-0.009
(0.045)
(0.007)
(0.033)
(0.005)
(0.040)
(0.014)
(0.038)
(0.012)
Ln (K.Stk/GDP)
-0.319*** -0.088***
-0.042
-0.012
(0.089)
(0.025)
(0.175)
(0.046)
FDI*Ln (K.Stk/GDP)
0.023**
0.006
0.005
0.001
(0.011)
(0.004)
(0.021)
(0.006)
Corp. Tax, diff EU15
0.006*
0.002**
0.002*
0.001
(0.004)
(0.001)
(0.001)
(0.001)
W/GDP, diff EU15
0.467
0.129
-0.566*** -0.166***
-0.216
-0.064
-0.341
-0.102
(0.527)
(0.091)
(0.195)
(0.031)
(0.548)
(0.155)
(0.487)
(0.138)
Deflator
-0.218**
-0.060
0.007
0.002
-0.282
-0.084
-0.356
-0.107
(0.102)
(0.056)
(0.097)
(0.038)
(0.298)
(0.087)
(0.331)
(0.114)
FDI* W/GDP
0.007
-0.002
0.002
0.000
-0.008
-0.002
-0.016*
-0.005
(0.022)
(0.003)
(0.010)
(0.002)
(0.016)
(0.006)
(0.009)
(0.003)
FDI*Deflator
-0.089**
-0.025**
-0.093*** -0.027***
0.046
0.014
0.179
0.005
(0.044)
(0.009)
(0.029)
(0.004)
(0.040)
(0.013)
(0.035)
(0.012)
FTA
-0.119*** -0.035***
-0.060**
-0.018
(0.032)
(0.005)
(0.030)
(0.013)
Float XR Regime
0.038***
0.011***
0.012***
0.004**
(0.012)
(0.002)
(0.003)
(0.002)
Corruption, diff EU15
-0.091**
-0.027***
0.006
0.002
(0.037)
(0.008)
(0.017)
(0.006)
38 See Papke and Wooldridge (2008) for further details on QMLE.
39 See Chamberlain (1980).
118
Log pseudolikel.
-20.12
-
-20.11
-40.42
-
-39.39
-
AIC
0.670
-
0.669
0.826
-
0.837
-
BIC
-324.99
-
-325.01
-537.76
-
-514.98
-
Observations
81
81
81
81
118
118
118
118
N. of clusters
8
8
8
8
11
11
11
11
Note: ***p<0.01, **p<0.05, *p<0.10, Standard errors in parentheses, robust to general second moment misspecification, conditional variance and serial correlation. All models have time dummies from 1999 to 2010. All models are estimated with pooled Bernoulli QMLE
and have time averages of the explanatory variables except the interaction terms and the dummies for trades agreements and EU
membership. The standard errors for the APE are obtained with 500 bootstrap replications.
The average partial effect of the difference in the stock of capital is negative: a ten percentage point reduction of
the gap from the average EU15 capital stock could have contributed to an increase of 0.8 percentage points in the
fraction of intra-industry trade. 40
One of the most interesting results of this group of estimates is the negative APE of FDI after taking into account
the partial effects of the interactions with capital (not significant), wage share (not significant) and the deflator
(significant). 41 The result is crucial to understand the policy implications for FDI impact on IIT and the convergence process with the EU15. 42
As shown in Table 4, the result is exclusively driven by the negative effect of the FDI interaction with the
deflator: this illustrates the danger of the possible crowding out effect of FDI on IIT due to the induced
inflationary pressures in the tradable sector after a surge in FDI. Domestic monetary authorities can try to
cushion the rise in prices when witnessing a surge in FDI inflow into their country in order to maintain the
competitiveness of their export sector. With this respect, a comparison with the NMS results is worth
considering: between 1998 and 2010, the intensity of FDI inflow in proportion to GDP levels was similar in the
two regions, 6.21 for CCPC and 6.20 for the NMS (see Table 2). However, for the NMS we do not find any
negative interaction effect of the inflow of FDI and the deflator. We conclude that NMS had more success in
controlling inflationary pressures induced by the inflow of FDI, keeping all else constant.
There is some evidence that the effect of unit labour costs, decomposed in wage share and price deflator,
decreases IIT. However the effect is significant only when we run the estimates unconditional on the capital
stock distance (column 4). In particular, this time the APE of the wage share has a negative and significant effect
on IIT, this is an indication that wage share can have negative effects on IIT at extreme distributional values,
namely for countries with very high levels of wages. Quantitatively, a 1 percent decrease in the wage share
increased IIT with the EU15 by 16 basis points. Furthermore, the interaction with FDI and deflator has a
negative sign, making the average partial effect of the increase in prices negative: a 1 percentage point increase
of the deflator, evaluated at average 1998-2010 FDI ratio in CCPC, decreases IIT by 16 basis points.
Free trade agreements have a negative effect on trade integration, although this effect is quantitatively marginal,
(column 4). Similarly, the degree of flexibility of the exchange rate can have a positive effect on IIT, for instance
a drastic paradigm shift in the exchange rate policy from the value of 1 (euroisation) in the Reinhart-Rogoff
(2004) scale to the value of 14 (fully floating exchange rate), ceteris paribus, could increase IIT with the EU15
by 14.3 basis points.
40 The estimated coefficient is considerably lower with respect to the linear model estimate. This implies that the linear model prediction is not performing well due to the non-linear nature of the response variable.
41 Looking at column 4 and evaluating at the average deflator level of CCPC we calculate the following: 0.025-0.027*1.1269=-0.005: a 10
percentage point rise of the FDI/GDP ratio decreased intra-industry exchange between CCPC and the EU15 block by 5 basis points. For the sake of illustration and extrapolating further, we note that in the period 1998-2010 the average growth rate of the share of FDI in GDP for CCPC was 27 percent per year. This translates into an average dampening effect on IIT of approximately 13.5 basis points per year.
42 As a caveat recall that the linear model may have a good approximation of the effect at the average of the FDI distribution, however at extreme values of FDI inflow the linear model performs poorly.
119
The effect of the difference in corruption perception for the CCPC is also highly statistically significant. The
estimate shows how a reduction of the index by one unit with respect to the EU15 average leads to a 2.7 basis
points rise in IIT. 43
6.3. Vertical and Horizontal Intra-Industry Trade
To conclude the empirical analysis, we present evidence for the determinants of vertical and horizontal IIT
which helps to shed new lights on the previous results. Table 5 illustrates the determinants of horizontally and
vertically differentiated goods, in addition it shows the estimates also for low and high quality differentiated
goods.
Table 5. Determinants of Vertical and Horizontal IIT for CCPC and NMS
Candidates and Potential Candidates
New Member States
Horizontal
Vertical
V-Low
V-High
Horizontal Vertical
V-Low
V-High
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Lag HIIT
-0.028
-0.134
(0.035)
(0.164)
Lag VIIT
0.247***
0.519***
(0.061)
(0.145)
Lag VIIT L
0.166
0.441**
(0.139)
(0.186)
Lag VIIT H
-0.133
0.329***
(0.129)
(0.052)
FDI/GDP
-0.488**
0.588***
0.432**
0.867***
-0.117
-0.065
0.019
-0.118
(0.193)
(0.115)
(0.177)
(0.262)
(0.157)
(0.067)
(0.068)
(0.171)
Ln (K.Stock)
0.559
-1.415***
-0.844
-1.183
-0.299
-0.062
-0.179
-0.167
(0.926)
(0.533)
(0.956)
(0.799)
(0.531)
(0.250)
(0.402)
(0.337)
FDI*lnKStk
-0.058
0.131***
0.012
0.369***
-0.035
0.004
0.034
-0.007
(0.120)
(0.045)
(0.093)
(0.095)
(0.067)
(0.027)
(0.052)
(0.046)
Corp. Tax
0.015
0.028**
0.023
0.021
0.046***
0.001
0.006
-0.001
(0.014)
(0.012)
(0.015)
(0.021)
(0.015)
(0.003)
(0.006)
(0.006)
W/GDP
-1.630
0.539
1.385
-3.452***
0.916
-1.061
-2.189***
0.455
(1.581)
(0.735)
(1.221)
(1.212)
(1.530)
(0.647)
(0.828)
(0.819)
Deflator
0.276
0.936**
1.811
-0.596
-3.451*
-0.456
-0.165
-0.359
(0.564)
(0.408)
(1.144)
(1.567)
(1.886)
(0.507)
(0.874)
(1.869)
FDI*W/GDP
-0.152
0.037*
0.032
-0.088**
0.062
-0.014
-0.053**
0.040
(0.093)
(0.022)
(0.025)
(0.041)
(0.050)
(0.018)
(0.027)
(0.037)
FDI*Deflator
0.362
-0.592***
-0.400**
-1.030***
0.158
0.051
-0.057
0.141
(0.238)
(0.123)
(0.199)
(0.295)
(0.180)
(0.070)
(0.075)
(0.190)
FTA
-0.584***
-0.083
-0.168
0.222
-0.339*
-0.131
-0.231
0.048
(0.113)
(0.099)
(0.210)
(0.253)
(0.178)
(0.092)
(0.231)
(0.220)
Float XR
0.019
0.104***
0.100*
0.058
0.035***
0.014***
0.020**
0.008
(0.044)
(0.035)
(0.057)
(0.067)
(0.011)
(0.005)
(0.009)
(0.007)
Corruption
0.144
-0.372***
-0.145
-0.313*
0.058
-0.030
-0.114**
0.096*
(0.164)
(0.100)
(0.176)
(0.181)
(0.092)
(0.023)
(0.049)
(0.050)
Democracy
-0.041
0.110***
0.139
0.084
-0.004
-0.004
-0.057
0.049
(0.073)
(0.035)
(0.089)
(0.096)
(0.043)
(0.043)
(0.046)
(0.050)
Intercept
-4.508**
-1.710***
-4.670**
-1.729
-1.011
-0.152
-0.549
-1.979
(1.995)
(0.560)
(1.824)
(1.849)
(1.913)
(0.652)
(1.232)
(1.861)
Time Fix Eff
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
F-Test
.000***
.000***
.001***
.000***
.000***
.000***
.000***
.000***
43 The corruption perception index ranges from 0 to 10 (the 10 meaning no corruption perception). As an example, in 2010 the average EU15
value was 7.25, the Croatian was 4.1 (3.14 points distant from the EU15 average): ceteris paribus, if Croatia had reduced this distance completely over the past decade, its IIT with the EU15 block could have benefited by approximately 9 basis points.
120
SW -Test
.000***
.086*
.168
.000***
.000***
.000***
.001***
.000***
Sargan Test
0.540
0.010**
0.650
0.218
0.427
0.219
0.130
0.156
BIC
264.52
151.50
254.94
279.76
381.95
187.31
267.34
283.23
AB Test - 1st
0.034**
0.191
0.032**
0.043**
0.024**
0.013**
0.011**
0.040**
AB Test - 2nd
0.293
0.567
0.036**
0.104
0.188
0.913
0.396
0.024**
AB Test - 3rd
0.185
0.394
0.078*
0.264
0.134
0.474
0.350
0.090**
Observations
62
62
62
62
106
106
106
106
N. countries
7
7
7
7
11
11
11
11
Note: p<0.1*, p<0.05**, p<0.01***Std. Errors in parentheses robust with respect to serial correlation and heteroscedasticity. F-Test, p-values for joint significance of time fixed effects under Ho: no joint effect of time fixed effects. SW (Shapiro- Wilk) test for normality of residuals, p-values reported under Ho: residuals are normally distributed. Sargan Test for over-identifying restrictions, p-values reported under Ho: the instruments as a group are exogenous. AB (Arellano-Bond) test for autoregressive residuals of 1st, 2nd or 3rd order, reported p-values for Ho: no serial correlation.
For NMS the dynamic effect are particularly important for vertically differentiated goods, suggesting a learning
and cumulative effect of IIT over the years. In contrast, the dynamic nature of IIT is a weaker feature for CCPC.
One interpretation can be that inertial effects of established intra-industry relationships and trade patterns persist
over time when industrial structures are more similar and there is some already established trade relationship.
The negative effect of the ULC is mostly driven by their effect on the high quality IIT for CCPC while for NMS
the wage inflation hinders trade in low quality goods. For NMS, horizontal IIT is penalised by increasing general
price dynamics, whereas the low quality range of IIT is reduced by increasing wage dynamics. In the regressions
for CCPC, the wage share of GDP has a remarkable negative effect for the high range of the quality spectrum:
over the period under study, and evaluating at average FDI inflow, a 1 percentage point increase in wage share
decreased the IIT in high quality goods by 4 percentage points. Similarly, an increase in the general level of
prices has also a considerable negative effect on high quality IIT, driven exclusively by the interaction with FDI
inflow.
In the CCPC group one of the most important aspects of the quality-partitioned estimates is that the negative
effect of FDI on IIT is driven mainly by the effect FDI has on the similar quality range of exports. At the same
time, FDI has also a negative effect on the high quality goods after accounting for the partial effects and a
slightly weaker effect on the low quality goods. 44
The floating exchange rate regime has some competitive benefits for the lower end of the quality spectrum as
well as for IIT in similar goods. This indicates that high quality goods are not affected by the competitive
devaluation argument. An interpretation can be that they are able to compete in the foreign market solely through
their intrinsic quality.
Interestingly, in the estimates for NMS corruption perception distance has a statistically significant and positive
coefficient on the highest quality range. With regards to the positive impact of corruption on the high quality
range, the literature shows45 that it is likely that the highest quality range of producers correspond to companies having a greater disposal of financial resources whereby invest in lobbying activities in order to improve their
market access into the EU.
44 For the high quality range we perform the following calculation: 0.867-0.088*0.13-1.030*1.1269+0.369*0.29 = -0.198
For the low quality range: 0.432-0.4*1.1269 = -0.019
45 See Meunier and Nicolaidis (2006).
121
7. CONCLUDING REMARKS
Over the past quarter of a century, CESEE countries have opened up to trade as part of a process of economic
transformation and integration into the European Union. This paper has looked into trade integration from the
point of view of intra-industry trade (IIT) between CESEE countries and the EU15. The focus on IIT is of first
order importance for the EU integration process for two main reasons: firstly, IIT is an important determinant of
competitiveness and sustainable current account balances; secondly, IIT is a tool for achieving more
synchronised business cycles and, as a consequence, reduce the effects of asymmetric shocks in an integrated
trade area.
By describing and analysing the factors behind these in a panel data set up, using the most disaggregated level of
bilateral trade data available and applying various statistical modelling techniques, this paper finds that EU
candidate countries and potential candidates are lagging behind in terms of IIT integration with respect to the
new EU member states.
The study identifies common factors behind IIT between the agglomerate CESEE region and EU15 countries,
such as fiscal incentives (corporate tax rate) and the exchange rate regimes. In particular, the strong significance
of the corporate tax differential indicates that tax policy could play an important role in promoting a faster
convergence process toward the EU trade structure. Furthermore, we find evidence that for both groups of
countries unit labour costs and their interplay with the influx of FDI are negative drivers of IIT. The findings in
the baseline specifications are echoed by results from a fractional response model, which underlines the
considerable quantitative effects of the variables.
Common determinants notwithstanding, there is considerable variation between CCPC on the one hand and
NMS on the other hand. While the trade competitiveness of CCPC with the EU15 is affected by institutional
quality and the distance in stock of physical capital, none of these factors appear to play a salient role in
explaining IIT between NMS and EU15. Our analysis shows that corruption perception plays a critical role in
hampering trade integration of CCPC into the EU. The disaggregated analysis of vertical versus horizontal IIT
reveals more important distinctions between the explanatory variables, and in particular the importance of highly
innovative and qualitative goods for intra-industry trade development.
REFERENCES
Arellano, M., Bond, S. (1991). ‘Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations’, Review of Economic Studies, 58, pp. 277–297.
Aturupane, C., Djankov, S. and Hoekman, B. (1997). ‘Determinants of intra-industry trade between East and West Europe’, World Bank policy research working paper No. 1850.
Balassa, B. (1966). ‘Tariff reduction and trade in manufactures among the industrial countries’, American Economic Review, 56 (3), pp.
466-473.
Balassa, B. (1986). ‘The Determinants of Intra-industry Specialisation in United States Trade’, Oxford Economic Paper, 38, pp. 220-233.
Balassa, B. and L. Bauwens (1987). ‘Intra-Industry Specialisation in a Multi-Country and Multi-Industry Framework’, The Economic Journal, 97, Dec., pp. 923-939.
Baldwin, R., Skudelny, F. and Taglioni, D. (2005). ‘Trade effects of the Euro. Evidence from sectoral data’, ECB Working Paper No. 446.
Barro, R.J. (1996). ‘Determinants of Economic Growth: A Cross-Country Empirical Study’, NBER Working Papers 5698, National Bureau of Economic Research, Inc.
Bijsterbosch M. and Kolasa M. (2009). ‘FDI and Productivity Convergence in Central and Eastern Europe: an Industry-level Investigation’, ECB Working Paper Series No. 992.
122
Botrić, V. (2012). ‘Intra-industry trade between the European Union and Western Balkans: A close-up’, EIZ Working Paper No. EIZ-WP-1202.
Botrić, V. (2013). ‘Determinants of Intra-industry trade between Western Balkans and EU15: Evidence from bilateral data’, International Journal of Economic Sciences and Applied Research, 6, pp. 7-23.
Brander, J.A. and Krugman, P.R. (1983). ‘A reciprocal dumping model of international trade’, Journal of International Economics, 15, pp.
313–321.
Brülhart, M. (1991) ‘Marginal Intra-Industry Trade: Measurement and Relevance for the Pattern of Industrial Adjustment’,
Weltwirtschaftliches Archiv, 130, pp. 600-613.
Caetano, J. and Galego, A. (2007). ‘In search for the determinants of intra-industry trade within an enlarged Europe’, South-Eastern Europe Journal of Economics, pp. 163-183.
Chamberlain, G., (1980). ‘Analysis of Variance with Qualitative Data’, Review of Economic Studies, 47, pp. 225-238.
Dhareshwar A. and Nehru V., (1993). ‘A New Database of Physical Capital Stock: Sources, Methodology and Results’, Revista de Analisis Economico, 8(1).
Dixit, A. K., & Stiglitz, J. E. (1977). ‘Monopolistic competition and optimum product diversity’, The American Economic Review, 297-308.
ECB, (2012). ‘Competitiveness and External Imbalances within the Euro Area’, ECB Occasional Paper No.139.
Falvey, R. E. (1981). Commercial policy and intra-industry trade. Journal of international economics, 11(4), 495-511.
Falvey R., Kierzkowski H. (1987). ‘Product Quality, Intra-Industry Trade and (Im)perfect Competition, in Kierzkowski H. (ed.) (1987).
Protection and Competition in International Trade, Blackwell, Oxford
Felipe J. And Kumar U. (2011). ‘Unit Labour Costs in the Eurozone: The Competitiveness Debate Again’, Levy Economics Institute Working Paper No. 651.
Flam, H. and E. Helpman (1987). ‘Vertical Product Differentiation and North-South Trade’, American Economic Review, 77, pp. 810-22.
Fidrmuc, J., Grozea-Helmenstein, D. and Wörgötter, A. (1999). ‘East-West intra-industry trade dynamics’, Weltwirtschaftliches Archiv, 135(2), pp. 332-346.
Fidrmuc, J. (2004). ‘The endogeneity of the optimum currency area criteria, intra-industry trade and EMU enlargement’, Contemporary Economic Policy, 22 (1), pp. 1-12.
Fontagné, L. and Freudenberg, M. (1997). ‘Intra-industry trade: Methodological issues reconsidered’, CEPII Working Paper No 97-01.
Fontagné, L. and Freudenberg, M. (1999). ’Endogenous Symmetry of Shocks in a Monetary Union’, Open Economies Review, 10, pp. 263-287.
Fontagné, L., Freudenberg, M. and Gaulier, G. (2005). ‘Disentangling horizontal and vertical intra-industry trade’, CEPII Working Paper No.
2005-10.
Foster, N. and Stehler, R. (2010). ‘Preferential Trade Agreements and the Structure of International Trade’, WIIW Working Paper No. 61.
Frankel, J.A., and Rose A. K. (1998). ‘The endogeneity of the optimum currency area criteria.’ The Economic Journal 108.449 1009-1025.
Gabrisch, H. (2006). ‘Vertical intra-industry trade between EU and Accession countries’, IHW Discussion paper No. 12, Institut für Witschaftsforschung Halle.
Gaulier G., Taglioni D. and Vicard V. (2012). ‘Tradable Sectors in the Eurozone periphery, Countries did not underperform in the 2000s’, http://www.voxeu.org/article/tradable-sectors-eurozone-periphery
Gaulier, G. and Zignago, S. (2012). ‘BACI: International trade database at the product level. The 1994-2007 Version’, CEPII Working Paper No 2010-23.
Greenaway, D. and Milner, C. (1986). ‘The Economics of Intra-Industry Trade’, Oxford: Blackwell.
123
Grubel, H.G. (1967). ‘Intra-industry specialisation and the pattern of trade’, Canadian Journal of Economics and Political Science, 33(3), pp.
374-388.
Grubel, H.G. and Lloyd, P.J. (1975). ‘Intra-Industry Trade: The theory and measurement of international trade with differentiated products’, London: Macmillan.
Helpman E. and Krugman P.R. (1985). ‘Market Structure and Foreign Trade’, Wheatsheef Boods, Havester Press, MIT.
Herderschee J., Qiao Z., (2007). ‘Impact of Intra-European Trade Agreements, 1900-2005: Policy Implications for Western Balkans and Ukraine’, IMF Working Paper No. WP/07/126.
Hoekman, B. and Djankov, S. (1996). ‘Intra-industry trade, foreign direct investment and the reorientation of Eastern European exports’, World Bank policy research working paper No. 1652.
Jensen, L. and Lüthje, T. (2009). ‘Driving forces of vertical intra-industry trade in Europe 1996-2005’, Review of World Economics, 145(3), pp. 469-488.
de Jong, E., Bogmans, C., (2011). ‘Does corruption discourage international trade?’, European Journal of Political Economy, 27, pp. 385–
398.
Judson, R.A. and Owen, A.L. (2001). ‘Estimating Dynamic Panel Data Models: a Guide for Macroeconomists’, Economic Letters, 65, pp. 9-15.
Kiviet, J.F., (1995). ‘On bias, inconsistency and efficiency of various estimators in dynamic panel data models’, Journal of Econometrics, 68, pp. 53–78.
Krugman, P. R. (1979). ‘Increasing returns, monopolistic competition, and international trade’, Journal of international Economics, 9(4), 469-479.
Krugman, P. R. (1981). ‘Intra-industry specialization and the gains from trade’, The Journal of Political Economy, 959-973.
Lancaster, K. (1980). ‘Intra-industry trade under perfect monopolistic competition’, Journal of International Economics, 89 (5), pp. 151-175.
Mardas, D. and Nikas, C. (2008a). ‘Economic Integration and Intra-industry trade between the European community and the Western Balkan countries’, Transition studies review, 15(3), pp. 511-523.
Mardas, D. and Nikas, C. (2008b). ‘European Integration, Intra-Industry Trade in Vertically Differentiated Products and the Balkan Countries’, International advances in Economic Research, 14(4), pp. 355-368.
Marin, Dalia. 2006. ‘A New International Division of Labour in Eastern Europe: Outsourcing and Offshoring to Eastern Europe.’ Journal of the European Economic Association, 4(2–3), 612–622.
Markusen, J.R. (1984). ‘Multinationals, multi-plant economies and the gains from trade’, Journal of International Economics, 16(3–4), pp.205–226.
Markusen, J.R. (2002). ‘Multinational Firms and the Theory of International Trade’. Cambridge, MA:M.I.T. Press
Meunier, S. and Nicolaidis, K. (2006). ‘The European Union as a conflicted trade power’, Journal of European Public Policy, 13(6), pp. 902–
921.
Mundell, R. A. (1961). ‘A theory of optimum currency areas.’ The American Economic Review,657-665.
Papke, L. E., & Wooldridge, J. M. (2008). ‘Panel data methods for fractional response variables with an application to test pass rates’, Journal of Econometrics, 145(1), 121-133.
OECD (2007). ‘Tax Effects on Foreign Direct Investment’ Tax Effects on Foreign Direct Investment – No. 17, Recent Evidence and Policy Analysis, 2007.
Reinhart, C.M. and Rogoff, K.S. (2004). ‘The Modern History of Exchange rate Arrangements: a Reinterpretation’, The Quarterly Journal of Economics, CXIX (1)
Rodas-Martini, P. (1998). ‘Intra-industry trade and revealed comparative advantage in the Central American Common Market’, World Development, 26(2), pp. 337-344.
124
Shaked A. and Sutton J. (1984). ‘Natural Oligopolies and International Trade’, in Kierzkowski H. (ed.) (1987). Protection and Competition in International Trade, Blackwell, Oxford
Shin, K. and Wang, Y. (2003). ‘Trade integration and business cycle synchronisation in East Asia’, ISER Discussion Paper No. 574.
Shleifer A. and Vishny, R.W. (1993). ‘Corruption‘, The Quarterly Journal of Economics, MIT Press, 108(3), pp. 599-617.
Verdoorn, P.J. (1960). ‘The Intra Block Trade of Benelux’, In: Robinson, E.A.G (ed). Economic consequences of the size of nations, London: Macmillan.
Wooldridge, J.M. (2002). ‘Econometric Analysis of Cross Section and Panel Data’, MIT Press, Cambridge, MA.
125
SECTION III
THE CHALLENGES OF OTHER WBCs IN THE PROCESS OF EUROPEAN
INTEGRATION
126
Branka Topić-Pavković
Faculty of Economics, University of Banja Luka, Bosnia and Herzegovina
FISCAL AND MONETARY ASPECTS OF ACCESSION OF
BOSNIA AND HERZEGOVINA TO THE MONETARY UNION
ABSTRACT
Bosnia and Herzegovina needs to become a member of the EU and to achieve the criteria for membership,
before gaining the position of candidate for EMU. Since the criteria for accessing the EMU are quantitatively
more precise than a wide range of other criteria, in this paper we focus on the fiscal and monetary specificity of
joining BiH the monetary union. Due to the heterogeneity of members of the monetary union, the main question
remains whether the loss of monetary sovereignty and unique monetary policy can be optimal for all its
members? Considering the theoretical and empirical knowledge about the benefits and disadvantages of
monetary integration, the aim goal of this paper is to analyze possible fiscal and monetary implications on BiH
in asseccion to monetary union. The results suggest that a rational solution for BiH, after joining the EU, is
based on gradual process of monetary integration, with stable monetary policy, effective management of public
finances and careful management of public debt. Central Bank of Bosnia and Herzegovina, which functions on
the principle of the currency board succeeded in maintaining monetary and financial stability even in times of
crisis. Since the institutional arrangement of the currency board does not allow budget deficit financing, the
monetary system of BiH will have certain advantages in terms of the Treaty of Maastricht. On the other hand, in
the next period monetary authorities should be devoted to assure the development of money market in BiH, and
then adjusting the structure of euro area monetary aggregates. Analysis of fiscal convergence criteria related to
the budget deficit and public debt, currently shows acceptable results for BiH, because the deviation from the
reference value is minimal. However, keeping in mind that the dynamics of the public debt directly depends on
the level of increase in GDP, exports and disposable income for debt servicing, the decision on further
borrowing will have to be associated with production projects or funding projects that would contribute to
further economic growth. This emphasizes that in the future, BiH needs to manage fiscal policy more efficiently,
especially due to negative effects of the recent debt crisis. Аccording to the optimum currency area theory,
member state must maintain a certain degree of flexibility and autonomy, and manage fiscal policy with clear
rules and budgetary principles. Fiscal aspect of monetary integration is significant because fiscal policy in EU is
based on coordination of single member states fiscal (budgetary) policies through Maastricht convergence
criteria and the Stability and Growth Pact. The long-term goal of BiH lies in achieving real convergence
through increased productivity and competitiveness. Consequently, our main purpose is to highlight the question
of conducting effective economic policy and necessary reforms before entering the E(M)U, because if
implemented quality as it should be, asseccing the monetary union will have more benefits than costs to the
economy of Bosnia and Herzegovina.
Key words: monetary integration, fiscal policy, public debt, monetary policy, debt crisis
JEL classification: E5, H6, F02
127
1. INTRODUCTION
Bosnia and Herzegovina's capacity for the accession to the European Union has been determined by
characteristics influencing monetary and fiscal system of BiH, from political, economic and social aspect.
For a country that is in a transition to a market economy, it is important to ensure macroeconomic stability which
is a key condition for stable national currency with solid basis for the investments in economic development. If
we observe the achieved level of macroeconomic stability and confidence in domestic currency, we can say that
Currency Board arrangement in Bosnia and Herzegovina has achieved its main goal. However, past experience
and theoretical aspects of functioning of Currency Board also show limitations in the results of growth of real
investments, GDP and employment. In terms of accession to the Monetary Union, the currency board as a form
of monetary arrangement has positive implications.
In conditions of passive monetary policy, country's economic policy and real GDP growth is significantly
dependant on fiscal and budgetary policy. High budget deficit from previous period and slow implementation of
political and economic reforms have resulted in increase of BiH borrowing from international financial
institutions and in significant increase in borrowing from commercial banks in the domestic capital market.
Combined with projected slow economic growth and high budget deficit, the issue of public debt sustainability
comes into the focus. Main treats to public debt management are: credit rating, reduced capacity for borrowing
from international financial institutions, political (in)stability, impact of global financial and economic crisis,
decrease in inflow of foreign direct investments, negative balance of payments, high unemployment rate,
reduced transfers from abroad and etc. In addition to mentioned characteristics, macroeconomic environment in
BiH was also largely influenced by the last economic crisis and high risk and uncertainty which challenged
economic policy in maintaining financial stability and in selecting economic policy as a response to crisis.
The crisis in majority EU countries has affected financial system, and afterwards the real sector causing slow
economic growth, increase in unemployment and increase of fiscal pressures. As a consequence, the slowdown
of global economy indicates fall in overall consumption and investment activities which also caused significant
deterioration in trade conditions in BiH. Main effects of economic stagnation in the EU (and eurozone) to
domestic economy were reflected in reduction of foreign demand for our export and insufficient capital inflows.
Weak domestic demand, deterioration in fiscal position of country and pressure on foreign currency reserves are
the primary generators of negative economic growth in BiH.
Considering the advantages and disadvantages, it is evident that classic currency board represents appropriate
tool for the fiscal stabilization but not the mechanism for a more dynamic economic growth. At the same time,
tendency of growth of public debt of Bosnia and Herzegovina in conditions of current global economic crisis,
stagnation of GDP and budget revenues, as well as the fact that settlement of long-term liabilities as a priority
requires higher percentage of budget funds, indicates the need for caution in future borrowing in terms of
contracting new borrowings under the more favorable terms, adapting repayment schedule to expected revenues,
and selection of priority projects. One of the main treats is a reduced credit rating of BiH influencing reduction
in inflow of foreign direct investments, as well as the deterioration in borrowing conditions of BiH from
international financial institutions, as well as limited opportunities for borrowing in the domestic market due to a
limited domestic accumulation.
The analysis presented is aimed at acquiring information on the current monetary and fiscal parameters in BiH,
and their aspects in the assessment of level of achieved convergence of BiH to the Maastricht criteria, that
candidate countries and potential candidates must meet on their way to the accession to E(M)U.
128
2. THE CRITERIA FOR MONETARY UNION ACCESSION
One of the main characteristics of the European integration, in all phases through which European history has
passed was heterogeneity of countries making the Union. This heterogeneity is reflected in socio-cultural
characteristics, history and level of economic development. As a primary goal of integration of Europe was a
need for prevention of new wars and accomplishment of deeper economic integration in order to defend its
economic interests and create strong economic force that will be equally competitive in the global market.
According to the theory of optimum currency area, which critically evaluates costs and benefits of Monetary
Union, similarity between the Member States, especially the achieved level of economic growth, is considered as
a prerequisite of successful functioning of Monetary Union.
Crucial moment in the history of European integration was the Maastricht Treaty, signed in December 1991 by
the EU Member States. Accession to Monetary Union was conditioned by meeting of convergence criteria. Why
did Member States have to meet convergence criteria for the creation of Monetary Union? First, we will look at
the theoretical framework and definitions of convergence.
2.1. The Criteria of Nominal Convergence
The concept of economic convergence indicates an accelerated process of social development resulting in
convergence of the values of economic variables among Member States, and primarily referring to the nominal
and real convergence. In order to adopt common currency Member States, according to the Maastricht Treaty,
are required to comply with nominal convergence (five convergence criteria). Nominal convergence indicates
meeting quantitatively determined criteria prescribed by Maastricht Treaty on readiness of country to join
eurozone. Since mentioned criteria are quantitative and more precisely defined than Copenhagen criteria, in the
economic literature they have become synonym of readiness of candidate countries for joining EMU. Country
may accede Monetary Union if it meets determined criteria, i.e. nominal convergence criteria (De Grauwe,
2003), such as:
1. inflation rate not more than 1,5% higher than the average inflation rates of the three Member States
with the lowest inflation,
2. Long-term interest rates should be no more than 2% higher than in the three Member States with the
lowest inflation rate,
3. Applicant country must accept exchange rate mechanism (ERM II) of the European Monetary System,
and must not devaluate its currency during the 2 years before the accession to EU,
4. The ratio of budget deficit to GDP must not exceed 3% (or if the deficit exceeds reference value, deficit
must decline until reaching the level of 3%) or, on the other hand, if the excess has a temporary nature
and, is close to the reference value, i.e. 3%,
5. The ratio of Government debt to GDP must not exceed 60% (or if debt exceeds reference value, debt
must be diminished and must be approaching reference value at a satisfactory pace).
2.2. Real Convergence
Concept of real convergence indicates decrease in differences in the levels of development of Member States. It
is defined as similarity in GDP per capita, level of nominal wages, equilibrium of real exchange rate and
similarity of price levels and ratio of foreign trade and local goods (Gaspar, 2005). Human capital is also quoted
as crucial criteria of convergence.
Bjorksten (Björksten, 2000) defines real convergence as reduction of differences in productivity and price level between the States. Real convergence requires sustainable economic growth in potential candidate countries, and
this requires appropriate micro and macro-economic policies, and effective mechanism for transition to a market
economy. According to Kowalski (Kowalski, 2003), real convergence refers to similarities of real structures and
129
business cycles in countries introducing or that have introduced common currency, in terms of productivity
convergence and higher standards of living measured by reduction of differences in GDP per capita.
3. CHARACTERISTICS OF MONETARY ARRANGEMENT OF BOSNIA AND HERZEGOVINA
3.1.Scope and limitations of currency boards in terms of monetary integration
Currency Board was the only adequate form of monetary policy for the stabilization of financial sector in the
political and economic environment characteristic for Bosnia and Herzegovina after the war. For a country that is
in a transition to a market economy, it is important to ensure macroeconomic stability which is a key condition
for stable national currency with solid basis for the investments in economic development. The primary task of
currency board in countries undergoing transition and reform is to secure currency stability, i.e. to keep inflation
at the lowest possible level.
For developing countries, tight fixing of exchange rate to the foreign currency of any leading monetary authority
may represent good strategy for the economic stabilization. Lack of exchange rate risks makes market
participants unaware of economic differences of country that pegged its currency compared to the country with
anchor currency, so the borrowing conditions of these countries converge. Appearance of external shocks lead to
exponential growth of dispersion of financing conditions of developed countries compared to the less developed,
especially in countries which have tightly fixed its exchange rates to foreign currencies. Reason for this is a lack
of flexibility of exchange rate, which in time of crisis leads to situation in which negative effects of the crisis are
fully reflected to the real sector.
In terms of the achieved level of macroeconomic stability, we can say that Currency Board arrangement in
Bosnia and Herzegovina has achieved its main goal. On the other hand, creating favorable investment
environment and strengthening of competitive position should represent main goal of BiH economy, and
accelerate the implementation of criteria of the real convergence. In other words, meeting the macroeconomic
stability is a good base for successful economic development in the long run. In conditions of passive monetary
policy, the essential question for BiH economy has been aimed at raising international competitiveness of
country in order to reduce the current account deficit.
Required prerequisite for this is acceptable ratio of productivity growth and wage adjustment. If the gross wages
in major sectors are growing faster than productivity in these sectors, this could increase inflationary pressures
and destimulate export on the one hand, boosting consumption and import on the other hand, which could at
certain point lead to unsustainable deficit of the current account and put into the question existence of currency
board arrangement and parity between the EURO and Convertible Mark (Kristic, 2007).
3.2. Criteria of Inflation Convergence
The primary task of currency board in countries undergoing transition and reform is to secure currency stability,
i.e. to keep inflation at the lowest possible level. One of these requirements implies convergence of inflation rate
to the inflation rate of anchor currency, i.e. currency to domestic currency is pegged. If we look at the inflation in
BiH (Chart 1), we can see that inflation declined in period from 2000 to 2004 and ranged below the value of the
inflation in the eurozone. A slight increase in 2005 was caused by exogenous pressures caused by the increase in
oil prices on the world markets. In 2006, level of inflation increased in prices due to the introduction of Value
Added Tax (VAT). Prices of raw oil had significant increase in the first part of 2008 which significantly affected
global inflation trends. Inflation pressures were more pronounced, and annual inflation almost reached double
figures in the middle of the year. Start of inflation was caused by the increase in prices of oil and food on the
world market, but inflationary spiral accelerated due to growth of local wages and utility services. In addition,
130
inflation was also characterized by emphasized fiscal expansion, mostly through the growth of social transfers
and current spending. International position was further weakened by increased foreign trade deficit.
Chart 1 Inflation in BiH
Source: World Economic Outlook Database, interpretation of the author
Despite the increase of merchandize export, net export (foreign trade deficit) sustained deterioration and
practically had negative contribution to the economic growth. Domestic spending was stimulated by the growth
of wages (particularly in public sector), large amount of new loans to the population, and continuous inflow of
remittances from abroad (information from Central Bank BiH).
We can see that inflation converges to the reference value of inflation rate in the eurozone, except in 2006 and
2008, which was a result of mentioned exogenous factors caused by the increase in oil prices on the world
market. The downward trend in inflation was present since the beginning of 2011 and continued in 2013, with
the deflationary pressures showing in the second part of the year. Annual inflation measured by consumption
prices index (CPI) in 2013 was -0,1%. In the end of 2013, inflation rate was -1,2%. Deflation in 2013 was the
result of continued downward trend in food and oil prices on the world market. The only significant divergences
in primary inflation were in 2010 as a result of simultaneous increase of excise duties to alcohol and tobacco.
Week domestic demand despite deflation in addition to deferred consumption due to expectations of further price
reductions indicates weak purchasing power of population.
4. FISCAL ASPECT OF INTEGRATION OF BOSNIA AND HERZEGOVINA
Fiscal aspect of integration of Bosnia and Herzegovina will be observed through the prism of fiscal criteria of
convergence. Please note that with the accession of a new member to the fiscal system of EMU, Member States
experience changes within their public finances, both in public revenues and also in public expenditures. Also, in
conditions of passive monetary policy, country's economic policy and real GDP growth is significantly
dependant on fiscal and budgetary policy.
Fiscal aspects of joining E(M)U are important due to several significant reasons: (Shaw, 1996)
1. it seems that accession regularly leads to fiscal pressures in the new Members, regardless of the principle that
new Members State should not immediately contribute to the EU budget, there were even talks of possibility of
fiscal crisis caused by enlargement;
2. After accession, new Member States must conduct fiscal policy in accordance with the rules of Stability and
Growth Pact, which could also cause certain fiscal consequences.
131
4.1. The Budget Deficit
From the analysis and studies on the importance of convergence, among the basic criteria is the criteria of budget
convergence, that requires: budget deficit of member country must not exceed 3% of GDP (and in case of higher
deficit, deficit must decline continuously and substantially before reaching rate of 3%), or on the other hand, if
this divergence from the referent value is caused by exceptional circumstances and has a temporary nature and is
close to the referent value, i.e. 3%.
The bottom line of the fiscal sustainability criteria is reflected in stabilization of debt-to-GDP ratio („stabilization
of primary budget“) (Bajo and Pezer, 2011). Issue of sustainability may be formulated in following manner:
budget deficit leads to the increase in government debt which will have to be serviced in the future. If interest
rates on government debt exceed growth rate of the economy, debt is set dynamically, which leads to the
increase in government debt in relation to the GDP. Government must ensure that primary budget has a surplus.
If there is no surplus, debt/GDP ratio will increase which will certainly lead to default in government debt (De
Grauwe, 2003).
Development of budget surplus/deficit in Bosnia and Herzegovina presented in the Chart no. 2.
Chart 2 Budget surplus/deficit in BiH
Source: Central Bank of Bosnia and Herzegovina, interpretation of the author
Budget of BiH had deficit over 753 million BAM or 2,5% of GDP in 2010, which is lower by 29,3% compared
to the previous fiscal year. Although we see the improvement of fiscal balance of BiH in the next year, 2012 and
2013 were characterized by further growth of deficit. According to these parameters, Bosnia and Herzegovina
still meets the Maastricht criteria in connection to the budget deficit.
4.2. Public debt
With the increase of country's indebtedness and expansion of its financial activities problem of defining debt
limit emerges. Last debt crisis has produced growth of public debt which has increased in previous years at rate
higher than growth of GDP in the majority of European States. Country with growth of public debt creates effect
of negative spillover to the other countries. Size and structure of public debt influences all trends in the
economy, and managing public debt is becoming more and more important segment of the overall economic
policy of the country. The growth of public debt in the long run must be lower than the economic growth rate, if
we want to avoid problems with liquidity. Therefore, a primary criterion for accession to the European Monetary
Union sets the limit for the public debt-to-GDP ratio to 60%.
132
Basic indicators of public debt of Bosnia and Herzegovina according to the above criteria, classify BiH as a
medium indebted country. Bosnia and Herzegovina in 2013 had debt-to-GDP ratio of 39,69%. However,
analysis of situation shows constant growth of public debt, and attention should be focused on how we spend
borrowed money and level of public debt sustainability in BiH.
4.2.1. Public Debt Trends in Bosnia and Herzegovina
Since 2008, increase in fiscal deficit, as well as the escalation of the economic crisis in 2009, has influenced
growth of debt of Bosnia and Herzegovina. In the period from 2008-2012 Bosnian economy had real fall of 2,2%
which led to the decline of public revenues, and country failed to adjust public expenditures which led to the
fiscal deficit that is present throughout the observed period. These developments were the main cause of the
sudden increase in public debt that has increased significantly over the period of 4 years.
Table1 Percentage increase/decrease of foreign debt of BiH compared to the previous year
Year
2004.
2005.
2006.
2007.
2008.
2009.
2010.
2011.
2012.
2013.
%
0,45
7,59
-6,15
-2,70
7,04
23,44
20,16
5,92
7,42
3,48
Source: Author
Increase of foreign debt on 31.12.2013 compared to 31.12.2012 was 249.209.054 BAM, i.e. 3,48%. Mentioned
increase in 2013, was a result of use of granted loans in amount of 1.009 million BAM (EIB 247 million BAM,
IMF 240 million BAM, EBRD 200 million BAM, etc.), minus the amount of paid principals (approx. 600
million BAM), with correction of part referring to the foreign exchange movements (approx. 160 million BAM)
for the observed period. The main risk in projecting internal debt is: potential changes in legislation regulating
the obligation of payment of internal debt in different manner from the existing legal arrangements, and thus
preventing planning and control of repayment, and potentially new obligations.
Level of debt sustainability is significantly influenced by currency structure of foreign debt. In the end of 2013,
foreign debt of BiH included 4 major currencies: EUR, SDR, USD and CPU1. Since the Central Bank of BiH
maintains monetary stability in accordance with the currency board arrangement, we can say that EUR holds a
second place in currency structure of debt (two currencies have majority share, EUR and SDR with 85%). Such a
high share of EUR provides high degree of predictability of future liabilities and BiH is exposed to a lower
currency risk.
When it comes to the currency structure of foreign debt servicing in the period 01.01.- 31.12.2013, share of paid
debt to the IMF created currency structure of payments dominated by the SDR. SDR2 is exposed to currency risk, although effective payments are realized in EUR. Taking into account already said, if all payments realized
in EUR were presented, share of this currency in total currency structure would be 84,01%. Focus should be on
the loans in USD because rise of USD could affect increase of foreign debt, which would at the same time
1CPU- Current pool unit of World Bank for the liabilities under the consolidation loan - IBRD no. 40390, which were in 2012 serviced in USD and EUR, and in 2013 in JPY and USD and are included in currency structure of serviced debt .
2SDR - Special Drawing Rights are supplementary foreign exchange reserve assets defined and maintained by IMF, World Banka and some other international financial institutions. The value is based on the basket of key international currencies (USD 41,9%, EUR 37,4%, Japanese yen 9,4% and British pound 11,3%). In total currency structure of serviced liabilities for the period 01.01.-31.12.2013, settled liabilities to IMF were presented separately, while other liabilities were in SDR (to World Bank and IFAD-u) and paid in EUR and USD, and these were included in currency structure of serviced debt in these currencies.
133
require more domestic currency for the servicing of foreign debt. Having in mind aforementioned, we can
conclude that future loans should be in EUR.
Servicing of debt refers to payment of funds each fiscal year for the principal, interest, discounts, other
obligations originating from debt, including all other associated costs. Since majority of loans granted to Bosnia
and Herzegovina have due, i.e. grace period has expired, share of paid principals in the structure of totally
serviced liabilities, i.e. compared to the collected interest, servicing and other costs, and has an growth tendency.
5. FINAL CONCLUSIONS
Results of research have shown that indicators determining the economic variable of criteria for the accession to
the Monetary Union from the monetary and fiscal aspect have been at a satisfactory level in Bosnia and
Herzegovina. Owing to the currency board as a monetary arrangement present in Bosnia and Herzegovina,
monetary criteria would be met relatively fast in the process of the accession to the ERM 2. Inability of
monetarisation of budget deficit, elimination of the exchange rate risk and low inflation have provided strong
basis for the required macroeconomic stability of Country.
Analysis of fiscal convergence criteria have shown that Bosnia and Herzegovina is currently classified as a
medium indebted country. Criteria relating to the budget deficit and public debt to GDP ratio, currently show
acceptable results for BiH, because divergence from referent values is minimal. However, considering that
development of public debt and servicing of the same is directly dependent on the degree of increase/decrease of
GDP, export and available income for servicing of debt, decisions on future borrowing will have to be associated
with production projects, or financing of projects which would help future economy growth. Characteristics
related to the public debt in BiH are reflected through inability to pursue active monetary policy and exchange
rate policy. Having in mind aforesaid, dominant segment of public debt management belongs to the fiscal policy
and expenditure control policy.
It should be underlined that BiH will have to manage efficiently fiscal policy in the future, and particularly when
BIH accedes E(M)U, because then monetary policy will be under the European Central Bank whose member
will also become Central Bank BiH. In accordance with the OCA Theory, it is recommended that BiH, i.e.
institutions managing its fiscal policy, must keep certain level of flexibility and autonomy and manage fiscal
policy (with clear rules and principle of budget equilibrium in terms of managing budget debt to GDP ratio).
Namely, this will be mandatory because we will have to meet the requirements defined under the Stability and
Growth Pact.
Within the passive monetary policy, significant efforts for the economy of BiH are aimed at boosting
international competitiveness of country in order to reduce deficit of current account. Creation of favorable
investment environment and strengthening of competitive position should represent main goal of BiH economy
and accelerate meeting of criteria of real convergence. The purpose of the entire process of convergence is
achievement of real convergence, gradual move towards equalizing the level of per capita income of regional
countries to the average income of EU Member States.
134
REFERENCES
Bajo, A. and Pezer, I. (2011): Strategije i ciljevi upravljanja javnim dugom, Ekonomski fakultet Zagreb
Björksten, N. (2000). Real convergence in the enlarged euro area: a coming challenge for monetary policy, Bank of Finland, Economics department, Working papers. 1/2000. URL:
http://www.researchgate.net/publication/5059202_Real_Convergence_in_the_Enlarged_Euro_Area_a_Coming_
Challenge_for_Monetary_Policy
Central Bank of Bosnia and Herzegovina, URL: www.cbbh.ba
De Grauwe, P. (2012 ). Economics of Monetary Union, Ninth Edition, Oxford University Press.
Gaspar, P. (2005). Real and Nominal Convergence of Pre-Accession Economies and the Choice of Exchange Rate Regime, International Centre for Economic Growth (ICEG) and Budapest University of Economics (BUE), Pаper presented on the conference Alternatives for Exchange Rate RegimeinPre-Accession Economies. Vienna: September 20-21, 2005. URL:
http://www.ecb.int/pub/pdf/other/neweumemberstatesen2005en.pdf
Kowalski, P. (2003). Nominal and Real Convergence in Alternative Exchange Rate Regimes in Transition Countries: Implications for the EMU Accession, Center for Social and Economic Research,Warsaw. URL: http://www.case-research.eu/upload/ publikacja _plik/1708281_270.pdf
Kristić, I. (2007). Održivost aranžmana valutnog odbora u BiH, Direkcija za ekonomsko planiranje Savjeta ministara BiH-DEP. URL: http://www.dep.gov.ba /dep_publikacije/doc/?id=102
Shaw, J. (1996). Law of the European Union, Macmillan, London.
135
LIST OF AUTHORS
Botrić, Valerija
Orszaghova, Lucia
The Institute of Economics, Zagreb
Národná banka Slovenska, Slovakia
Trg J.F. Kennedy 7
lucia_orszaghova@nbs.sk
10000 Zagreb, Croatia
vbotric@eizg.hr
Pána, Lubomír
Col ege of European and Regional Studies
Dautović, Ernest
Žižkova 6, 370 01 České Budějovice, Czech Republic
Université de Lausanne, Switzerland
pana@vsers.cz
ernest.dautovic@unil.ch
Schudel, Willem
Dušek, Jiří
De Nederlandsche Bank, The Netherlands
Col ege of European and Regional Studies
c.j.w.schudel@dnb.nl
Žižkova 6, 370 01 České Budějovice, Czech Republic
jiridusek@centrum.cz
Šlander, Sonja
University of Ljubljana, Faculty of Economics,
Golob, Marino
Kardeljeva pl. 17, 1000 Ljubljana, Slovenia
Colegium Fluminense Polytechnic of Rijeka, Rijeka,
katja.zajc@ef.uni-lj.si
Croatia
Topić-Pavković, Branka
Golob, Martin
University of Banja Luka, Faculty of Economics, Bosnia
Mara Mara d.o.o., Pazin, Croatia
and Herzegovina
branka.topic-pavkovic@efbl.org
Gonza, Tej
Erasmus University of Rotterdam
Zajc Kejžar, Katja
gonza.tej@gmail.com
University of Ljubljana, Faculty of Economics,
Kardeljeva pl. 17, 1000 Ljubljana, Slovenia
Kandžija, Tomislav
katja.zajc@ef.uni-lj.si
Primorsko−goranska County, Rijeka, Croatia
Zaninović, Vinko
Kumar, Andrej
University of Rijeka, Faculty of Economics, Croatia
ECSA Slovenija
vinko.zaninovic@efri.hr
andrej.kumar@siol.net
136
Conference Proceedings
COOPERATION CHALLENGES AFTER THE EU
ACCESSION OF CROATIA
Edited by
Andrej Kumar, Katja Zajc Kejžar
Document Outline
Preface
OPENING LECTURE
EU TRADE STRATEGY AND THE BALKANS
EU TRADE STRATEGY AND THE BALKANS
SECTION I
IMPLICATIONS OF EUROPEAN INTEGRATION FOR CROATIA:
PRE-ACCESSION EVIDENCE AND EARLY MEMBERSHIP EXPERIENCES
CROATIAN INSURANCE MARKET OVERVIEW AFTER EU ACCESSION
INDUSTRY WAGE PREMIUM AND EU TRADE EFFECTS IN CROATIAN MANUFACTURING SECTOR
INTENSIVE AND EXTENSIVE MARGINS OF CROATIAN MANUFACTURING EXPORTS: EVIDENCE FROM 2000-2012 PERIOD
SECTION II
EU MEMBER STATES’ EXPERIENCES IN DIFFERENT POLICY AREAS
THE USE OF PPP PROJECTS ON THE LEVEL OF STATES, REGIONS AND MUNICIPALITIES UNDER THE CONDITIONS OF THE CENTRAL EUROPEAN REGION
THE PROBLEM AREA OF CZECH REPUBLIC'S USE OF EU STRUCTURAL FUNDS IN THE PROGRAMME PERIOD OF 2007/2013
EU COHESION POLICY AND ABSORPTION IN SLOVENIA
EVALUATION OF EU COHESION POLICY: LESSONS FROM SLOVENIAN CASE
CONVERGING IN DIVERGENT WAYS:
EXPLAINING TRADE INTEGRATION BETWEEN CESEE COUNTRIES AND THE EU-1513F
SECTION III
THE CHALLENGES OF OTHER WBCs IN THE PROCESS OF EUROPEAN INTEGRATION
FISCAL AND MONETARY ASPECTS OF ACCESSION OF
BOSNIA AND HERZEGOVINA TO THE MONETARY UNION