180 Int. J. Economics and Business Research, Vol. 4, Nos.

1/2, 2012

FDI and intra-industry trade: theory and empirical

evidence from the Visegrad Countries

àukasz Ambroziak
Institute for Market, Consumption and
Business Cycles Research,
Al. Jerozolimskie 87,
02-001 Warsaw, Poland
Fax: +(22) 628 24 79
and
Jean Monnet Chair of European Integration,
Warsaw School of Economics,
ul. MadaliĔskiego 6/8,
02-513 Warsaw, Poland
Fax: +(22) 646 61 15
E-mail: lukasz.ambroziak@ibrkk.pl

Abstract: This paper analyses impact of the foreign direct investment (FDI) on
the intra-industry trade (IIT) patterns in the Visegrad Countries (VCs)
(the Czech Republic, Hungary, Poland and Slovakia) in the period 1995–2008.
The FDI has been a driving force of these countries’ foreign trade, especially
for the IIT. From the theory and previous empirical studies, it results that the
FDI has a positive impact on vertical intra-industry trade (VIIT), while the
influence of this variable on horizontal intra-industry trade (HIIT) is
ambiguous. Using a panel data approach, the determinants of HIIT and VIIT
were identified. The obtained results confirmed that the FDI in the VCs
stimulated not only VIIT, but also HIIT.

Keywords: IIT; intra-industry trade; HIIT; horizontal intra-industry trade;

VIIT; vertical intra-industry trade; FDI; foreign direct investment; VCs;
Visegrad Countries.

Reference to this paper should be made as follows: Ambroziak, à. (2012)

‘FDI and intra-industry trade: theory and empirical evidence from the Visegrad
Countries’, Int. J. Economics and Business Research, Vol. 4, Nos. 1/2,
pp.180–198.

Biographical notes: àukasz Ambroziak is an Assistant in the European

Integration Department of the Institute for Market, Consumption and Business
Cycles Research in Warsaw and PhD student in the Collegium of World
Economy at the Warsaw School of Economics. Professional interests include
foreign trade of the EU New Member States, intra-industry trade, trade
competitiveness indexes, product fragmentation in the EU New Member States.
In 2008, he obtained grant from the Prof M. Kucharski Scholarship Fund
(Foundation for Managerial Education Development of the Warsaw School of
Economics) on the research project: ‘Changes in trade specialisation of the
New Member States after their EU accession and internalisation of production
processes’.

Copyright © 2012 Inderscience Enterprises Ltd.

 FDI and IIT: theory and empirical evidence from the VCs 181

1 Introduction

An important part of foreign trade in the Central and Eastern European Countries
(CEECs) has been driven by the intra-industry trade (IIT). One of the crucial factors
which affected the IIT development in the CEECs was foreign direct investment (FDI). It
played a crucial role in the economic transformation of these countries. The need of
Central European companies to obtain access to know-how, capital and distribution
channels was one of the reasons for dynamic inflow of foreign capital in the 1990s. As a
result, the technological gap between CEECs and the European Union (EU) countries has
been decreasing, which has positively affected the development of the IIT between
trading partners. Recently, the FDI inflow to the CEECs has been connected with the
fragmentation of production process. Therefore, a special attention in this paper is paid to
the impact of FDI on the IIT.
The aim of this paper is to define the impact of the FDI on the IIT development in
four Visegrad Countries (VCs, i.e. the Czech Republic, Hungary, Poland and Slovakia)
during the period 1995–2008. The econometric analysis follows the review of theoretical
relationship between these above-mentioned variables and the presentation of the FDI
and IIT patterns in the analysed countries. A choice of the VCs resulted from the fact that
they were pioneers among CEECs in the integration process with the EU (the first half of
the 1990s). Furthermore, the presence of the multinational enterprises (MNEs) in VCs
was the most intensive in this region of Europe and those countries were the largest
recipients of FDI among CEECs.
The structure of this paper is as follows. Sections 2 and 3 give us the theoretical
framework about the impact of the FDI on the IIT and a brief survey of the empirical
studies on the IIT determinants. Section 4 discusses methodology of IIT measurement.
Sections 5 and 6 present the recent trends in the FDI and the IIT development in the
analysed VCs. Section 7 explains the nature of model, the used variables and advances
some research hypotheses about the horizontal intra-industry trade (HIIT) and vertical
intra-industry trade (VIIT). In Section 8, the results obtained from the model are
presented. The final Section 9 contains conclusions.

2 FDI and IIT – theoretical framework

The crucial step in the evolution of the IIT theory was the publication by Grubel and
Lloyd (1975). It stimulated an enormous interest in this type of foreign trade. The
evidence of the IIT was contrary to the classical trade theory associating trade between
two countries with their differences in factor endowments. The first works, which tried to
explain the nature of this phenomenon, were considerations of Krugman (1979, 1983),
Lancaster (1980) and Helpman (1981). Generally, the theories explaining the IIT are
classified into two groups, namely HIIT and VIIT theory. On the one hand, Finger
(1975), Lipsey (1976), Falvey (1981), Falvey and Kierzkowski (1987), Shaked and
Sutton (1984), Flam and Helpman (1987), Jones and Kierzkowski (1990), Arndt and
Kierzkowski (2001) and Cheng and Kierzkowski (2001) introduced the vertical
differentiation model. On the other hand, Lancaster (1980), Helpman (1981), Krugman
(1979, 1980) and Dixit and Norman (1980) developed the horizontal differentiation
model.
182 à. Ambroziak

The theoretical literature on determinants of the IIT is substantial. Originally,

Loertscher and Wolter (1980) noted that IIT between countries is intense if gross
domestic product (GDP) per capita of the trading countries is high, if the difference in
this indicator is relatively small and if the average size of their aggregate outputs is high
and similar. According to the above-mentioned literature, IIT can be influenced also by
the geographical proximity (measured as a distance between capitals of the countries,
presence of the common border), membership in the free trade agreement and presence of
MNEs. Since 1980s, the presence of MNEs has significantly affected the international
trade, and, as a result, also the IIT. Dunning (1993) defined a multinational or
transnational enterprise as an enterprise that engages in FDI and owns or, in some way,
controls value-added activities in more than one country. This is the threshold definition
of MNE. Thus, the concept of ‘MNEs’ is intrinsically connected with the concept of
‘FDI’. In the further part of this section, the special attention is paid to the impact of the
FDI on IIT from the theoretical point of view. It is worth stressing that the theory of the
FDI impact on the IIT is only a part of the theory, which tries to describe the mutual
relations between FDI and trade flows. They depend on the character of capital flows.
Generally, horizontal foreign direct investment (HFDI) displaces trade and is positively
related with trade costs. Next, vertical foreign direct investment (VFDI) complements
trade and is eased by low trade costs.
The pioneering work, which tried to explain the impact of the FDI on the IIT, was the
publication of Helpman and Krugman (1985). The authors found that the emergence of
multinational corporations changes in an important way the link between differences in
relative factor endowments and the share of IIT. When the difference in factor
composition becomes large enough so as to bring about the emergence of multinational
corporations, this association turns positive, as long as the capital-rich country is a net
exporter of manufactures. When the difference in composition of factor endowments
becomes large enough so that the capital-rich country begins to be a net importer of
manufactures, the negative association between factor dispersion and the share of IIT is
restored. Therefore, the larger the involvement of multinational corporations among
analysed countries, the weaker the effect of changes in the degree of dispersion in per
capita income on the share of IIT. Furthermore, Helpman and Krugman (1985) found
that the volume of IIT will depend on how narrow one defines product categories. If in
the industrial classification, finished products are classified to be different from the
intermediate products, then only IIT in finished products contributes to the volume of
total IIT.
If Helpman and Krugman (1985) have focused in their last model on the vertically
integrated enterprises, then Markusen (1984, 2002) and Markusen and Venables (1998,
2000) have focused on the HFDI. The presence of multinationals unambiguously reduces
trade values. If multinationals are allowed to exist, then as countries converge in size, so
an increasing share of world production is undertaken by multinationals. Multinationals
substitute for IIT, so reduces the value of trade.
The next step in the evolution of the theory was the publication of Markusen and
Maskus (2002). They distinguished three main types of firms: national enterprises,
multinational horizontal and vertical enterprises. The existence of these three types of
enterprises depends only on two variables, i.e. on trade costs and investment costs. Trade
liberalisation can lead to the increase of the IIT shares between two trading countries.
However, investment liberalisation can contribute to decline the IIT intensity in case if
trade costs are too high or too low. Thus, the IIT index will be high, on the one hand,
 FDI and IIT: theory and empirical evidence from the VCs 183

when two countries are similar to each other or when the small country is skilled-labour
abundant, and on the other hand, when trade costs are low and investment costs are high.
Next, Fukao et al. (2003) made an attempt to answer the question, how trade patterns
are influenced by FDI costs, trade costs and the factor price gap between the two
countries. In particular, the authors studied the following three situations: firstly, zero
trade costs coupled with prohibitively high FDI costs; secondly, zero trade and FDI costs
and finally, substantial trade costs and zero FDI costs. The main results of the theoretical
analysis were summarised as follows. Firstly, VIIT occurs only when both FDI costs and
trade costs are small. If there exist substantial FDI costs, gains from the international
division of labour within firms will be surpassed by FDI costs. It means that firms in the
developed country will not conduct VFDI, which is indispensable for VIIT. If it is very
costly to trade products from the developed country to the developing country, then firms
in the developed country will replace their exports from their home country with local
production in the developing country. Because of this HFDI, VIIT becomes very small.
Secondly, if there exist substantial costs of FDI, the share of VIIT in total trade will
depend on the factor price gap between the two countries. If the factor price gap is small,
then firms will have limited incentive to engage in the international division of labour
through FDI, and VIIT will become small.
Okubo (2004) presumed that the currently increasing IIT is mainly determined by
trade between developed (Organisation for Economic Cooperation and
Development (OECD)) and developing (non-OECD) countries. This increase cannot be
explained by the Helpman–Krugman model. So, the author suggests that these changes
can be explained by the VIIT or fragmentation in the production process. Introducing
technology transfer through FDI into the model can explain the current IIT well.
Favourable circumstances for technology transfer in host countries, like a small
difference in educational level enhances FDI, which, in turn, increases reimports. In the
presence of wage and technology gaps, IIT increases when the degree of technology
transfer is sufficiently small, but a large degree of technology transfer decreases IIT.
Furthermore, if the technology transfer corresponds to production networks, we can
suggest that production networks promote IIT.
Recently, in the theoretical literature the special attention has been paid to the
modification of traditional model with multinationals in the form of 2 × 2 × 2 (two
countries, two factors of production and two goods). This modifications lay in the adding
of additional factor of production (Egger et al., 2007), additional country (Ekholm et al.,
2007) or these two variables simultaneously (Baltagi et al., 2007). These modifications
were induced by still changing form of the multinational activities. The division of MNEs
into horizontal (market-seeking production) and vertical (resource-seeking investments)
integrated FDI not fully reflects the investment strategies of the enterprises. More and
more frequently, the multinationals apply complex investment strategies, e.g. export-
platform FDI. According to Ekholm et al. (2007), export-platform FDI is defined as
investment and production in a host country where the output is largely sold in third
markets, not the parent or host country markets.
Figure 1 summarises the previous considerations. Some authors (i.e. Neary, 2009)
suggested that in the global economy, there do not exist any clear division into HFDI and
VFDI, respectively. However, from the point of view of this analysis, this division
enables us to better understand the mutual relationship between the FDI and the IIT.
HFDI, stimulated mainly by the high trade costs, substitutes trade flows and, as a result,
also the IIT. HFDI can also indirectly lead to the growth of HIIT, because it influences
184 à. Ambroziak

the economic development in the host country and diminishes differences in market
potential and level of development between the host and home country.
VFDI, also identified with the fragmentation of production processes,1 is of great
importance for the IIT development. It is very difficult to indicate which part of VFDI
has a typical vertical character and which part occurs in form of the export-platform FDI.
Generally, both types of the VFDI promote mainly VIIT. Typical VFDI complements
trade flows, so influences intensity of vertical specialisation. In case of export-platform
FDI, its impact on the IIT intensity depends on the market, where there are exported
goods produced in the plant. On the assumption that imports of parts and accessories to
host country exist, the growth of the IIT occurs, if final goods are exported to the home
country. However, if the final goods are exported to the third countries, the result is fall
of the IIT. Next, in case of the so-called global platform FDI (final goods are exported to
the third countries as well as to the home country), the scale of the IIT growth will
depend on what part of the export of final goods reaches the home market. The increase
of the IIT intensity will be resulted from the VIIT, because the differences between unit
value in export and import will be significant. It is worth to stress that the IIT with final
and the intermediate goods can only occur, if the final and intermediate goods are defined
as the same industry. VFDI can theoretically create the IIT with horizontally
differentiated goods. It can happen between the home and host countries where
simultaneous export and import of intermediate goods occur, which not significantly
differ in the unit value.

Figure 1 The impact of the FDI on the IIT – summary of the theoretical considerations

Source: Own studies.

 FDI and IIT: theory and empirical evidence from the VCs 185

3 Review of empirical studies on the IIT determinants

The empirical literature on the determinants of IIT is extremely extensive.2 Already in the
1990s, analyses were conducted to investigate both changes in IIT of the CEECs and
determinants of IIT growth in these countries. The special interest in such analyses has
been usually paid to these countries’ trade with the EU Member States, mainly due to the
fact that EU countries had a dominant role in trade of CEECs.
The changes of the IIT intensity in the CEECs were analysed in detail by Caetano and
Galego (2006), Fidrmuc and Djablik (2003), Gabrisch and Segnana (2002, 2003),
Gabrisch (2006), Kandogan (2003), Hoekman and Djankov (1996), Ferto and Soos
(2006), while only in case of the Central European Free Trade Agreement – ýernoša
(2007) and only in case of Poland – CieĞlik (2003), Czarny and ĝledziewska (2009),
Molendowski (2007), Klimek (2006) and Kawecka-Wyrzykowska (2009).
Next, the determinants of the IIT intensity in the CEECs were investigated by
Caetano and Galego (2006), Fidrmuc and Djablik (2003), Gabrisch and Segnana (2002,
2003), Aturpane et al. (1997), Kandogan (2003) and CieĞlik (2008).

4 Measurement of the IIT

The IIT in this paper is measured by the Grubel–Lloyd (GL) index (Grubel and Lloyd,
1975), which is based on the concept of ‘trade overlap’ and represents the share of the
absolute value of IIT in trade turnover in a particular industry i, i.e.:

GLkk ' X kk '

i ,t 
 M ikk,t '  X ikk,t '  M ikk,t '
1
X ikk,t '  M ikk,t '
i ,t
X kk '
i ,t  M ikk,t '  X kk '
i ,t  M ikk,t ' 
where X ikk,t ' and M ikk '
,t refer to country k’s exports and imports, respectively, to/from
country k’ over one particular year t, in a particular industry i (here: four-digit
harmonized system (HS)). This measure takes values between 0 (exclusive inter-industry
trade) and 1 (exclusive IIT).
GL indices can be aggregated across N industries, as a trade-weighted average of the
industry indices:
N kk '
N
¦ X i ,t  M ikk,t ' X ikk,t '  M ikk,t '
GLkk
t
'
¦ wikk,t ' GLkk
i ,t
'
1 i 1
N
and wikk,t ' N
i 1 ¦ X i 1
kk '
i ,t  M ikk,t '  ¦ X
i 1
kk '
i ,t  M ikk,t ' 
or can be aggregated across K’ partner countries and across N industries:
K' N kk '
¦ ¦ X i ,t  M ikk,t '
GLkt 1 k' 1
K'
i 1
N
¦ ¦ X k' 1 i 1
kk '
i ,t  M ikk,t ' 
186 à. Ambroziak

or can be aggregated across K’ partner countries, across K reporter countries and across N
industries:
K K' N kk '
¦ ¦ ¦ X i ,t  M ikk,t '
k 1 k' 1 i 1
GLt 1 K K' N
¦ ¦ ¦ X k 1 k' 1 i 1
kk '
i ,t  M ikk,t ' 
where wikk,t ' denotes the share of trade in product i in the total trade between k and k’
trading partners, N could be denoted a number of industries in the total trade (or in the
trade section) between k and k’ trading partners, K’ denotes a total number of trading
partners or group of trading partners (here: EU-15, New Member States (NMS)-123 and
third countries) and K denotes a number of reported (analysed) countries (here: four
Central European Countries (CEC-4)).
According to the Greenaway et al. (1994) methodology, the total IIT has been divided
into its two types – HIIT and VIIT – using the so-called ‘product similarity criterion’.
This criterion is based on the ratio between the unit value in exports and the unit value in
imports in trade between two trading partners. HIIT is defined to exist for trade in
industry i between two trading partners that satisfy the criterion:

UVikk
,t
', x
1 D d d 1 D
UVikk
,t
', m

and VIIT when such criteria are satisfied:

UVikk
,t
', x
UVikk
,t
', x
d 1  D or t 1 D
UVikk
,t
', m
UVikk
,t
', m

where x ((UVikk
,t
', x
) /(UVikk
,t
', m
)) is relative unit value of exports and imports (the ratio
between exports and imports unit value); Į is the dispersion factor. The parameter Į is
arbitrarily fixed. In general, Į assumes the value of 0.15. Some authors, e.g. Fontagné
et al. (2006), assume the value of 0.25. However, to distinguish HIIT and VIIT, Yoshida
et al. (2009) use some different relative unit values of exports and imports on the level of
0.25, 0.5, 0.67, 0.8, 1.25, 1.50, 2 and 4.
The above-presented methodology is based on the assumption that differences in
quality are reflected in price differences (with prices as unit values of analysed goods).
The relation price quality is supported by the idea that in a perfect information context, a
certain variety of a product can only be sold at a higher price if its quality is superior.
Stiglitz (1987) claimed that even in the context of imperfect information, the quality will
be always reflected in the prices.
In this paper the IIT will be divided into four types of trade:
1 VIIT low quality: Low-quality vertical intra-industry trade when x < 0.85; it means
that the unit value of exports is relatively lower than unit value of imports (country k
exports low-quality goods and imports high-quality ones).
2 HIIT: Horizontal intra-industry trade when 0.85 ” x ” 1.15; it means that country k
exports and imports goods within the same industry, which are the same price
(quality) but differ in some other features, such as colours, country of origin, etc.
 FDI and IIT: theory and empirical evidence from the VCs 187

3 VIIT high quality: High-quality vertical intra-industry trade when x > 1.15; it means
that the unit value of exports is relatively high in comparison with the unit value of
imports (country k exports high-quality goods and imports low-quality ones).
4 IIT not allocated: It means that the relative unit value of exports and imports is
impossible to be computed. There can be some reasons of it, i.e. lack of data on
export quantity, lack of data on import quantity, lack of data on both export and
import quantity.
All trade data used in this paper are taken from the World Integrated Trade Solution
(WITS) database, jointly developed by the World Bank and the United Nations
Conference on Trade and Development. The basic information source is the United
Nation Statistical Division’s Commodity Trade database (COMTRADE). I retained all
bilateral imports and exports in value terms (current US dollars).
The IIT indices were computed at the four-digit HS for four EU countries: the Czech
Republic, Hungary, Poland and Slovakia in 1995–2008. Then the indices were
aggregated across industries (to the level of total trade) and across trading partners (to the
EU-15, NMS-12, third countries or to the trade with all partners) and sometimes also
across reporter countries (to the level of CEC-4).

5 Brief review of FDI in the VCs

First investment activities by a foreign corporation in the analysed VCs date back to the
early 1960s. A large-scale inflow of FDI to the VCs was observed in the 1990s. The
transition to market economies was accompanied by a growing demand for foreign
capital, in connection with changes in the type of ownership of industrial enterprises. In
the early 1990s, FDI inflows to these countries mostly concerned takeovers of existing
plants by foreign investors with the view of restructuring and joint venture investment.
Simultaneously, the analysed countries joined the process of production fragmentation.
Central and Eastern Europe as an investment destination offered multinational
corporations a number of advantages, such as workforce skills, relatively low labour costs
and favourable tax system for new investors. Investment inflow to the VCs was also
driven by economic cooperation of those countries with the EU, initiated in the 1990s,
and prospects for future EU membership.
A vital role in FDI inflow to the analysed countries was played – apart from
brownfield investment – by greenfield investment. Those mainly consisted in creating a
business manufacturing in the host country from scratch. Their advantage over the former
was that the location was not determined by the existing production plant, there was full
freedom of choice. Investors took account of factors such as the availability of labour, the
proximity of suppliers and outlets as well as transport infrastructure.
As a result, at of the end of 2008, the stock of FDI in the four VCs was circa EUR
288 billion (Table 1). It constituted nearly 71% of foreign capital, which had inflow to
the ten EU NMS (excluding Cyprus and Malta). The largest recipients of FDI among the
analysed countries were Poland, the Czech Republic and Hungary.
188 à. Ambroziak

The comparison of the stock of inward FDI in the new EU Member States fails to
reflect its full role in the economies of those countries. A more reliable measure is FDI
per capita, as well as the ratio of the FDI stock to GDP. On average, at the end of 2008,
the ratio of the FDI stock in the VCs to their GDP amounted to 42%, and the index of the
FDI stock per capita exceeded the level of EUR 4.4 thousand. Figure 2 shows that in
terms of the FDI stock per capita the Czech Republic and Hungary are the leaders in the
VCs (with EUR 7.8 thousand and EUR 6.2 thousand per capita at of the end of 2008,
respectively). The worst performer was Poland, with EUR 2.9 thousand per capita. The
above-expressed conclusions confirm the analysis based on the ratio of the FDI stock to
GDP. The unfavourable for Poland comparison with the other analysed countries could
be explained by a relative size of its economy. Generally, the foreign capital is of greater
importance in the small economies than in the large economies.

Table 1 Inward FDI stock in the VCs (in EUR billion)

2000 2001 2002 2003 2004 2005 2006 2007 2008

VCs 89.8 114.9 126.2 132.7 167.3 199.0 239.0 288.4 286.8
Czech Republic 23.3 30.7 36.9 35.9 42.0 51.4 60.6 76.3 81.3
Hungary 24.6 31.0 34.6 38.3 45.9 52.4 61.8 67.0 62.7
Poland 36.8 46.7 46.1 45.9 63.3 75.2 91.1 116.0 110.2
Slovakia 5.1 6.5 8.6 12.6 16.1 20.0 25.5 29.1 32.6
NMS-10 109.9 138.9 151.6 164.5 210.7 259.4 321.9 395.9 406.7
Source: Own study based on data of the Vienna Institute for International Economic
Studies (WIIW).

Figure 2 The stock of inward FDI in the CEC-4 per capita (EUR, left scale) and relative to gross
domestic product (% of GDP, right scale)

Source: Own study based on WIIW data.

 FDI and IIT: theory and empirical evidence from the VCs 189

6 Recent trends in the IIT of the VCs

The intensity of the IIT in the VCs has been generally increasing in the analysed period
(see Figure 3). Faster growth of the IIT indices was noted especially in the pre-accession
period, from 1996 to 2004. After the EU accession (in 2005), a decline of IIT shares in
total trade occurred, and then it was reverted. As a result, in 2008, the IIT share in total
trade of the VCs was higher by 8.2 p.p. than in 1995 and higher by nearly 1 p.p. than in
the accession year and amounted to 34.3% in 2008.
Changes of the IIT intensity in the individual VCs were slightly different. The largest
growth of IIT shares in the total trade was noticed in Poland (nearly 16 p.p. in 1995–
2008). Upward trend of the IIT intensity characterised the Hungarian and partly Slovak
foreign trade. The level of the IIT intensity in the Czech Republic was (except for the
second half of the 1990s) relatively stable and oscillated about 40%. As a result, the
highest shares of the IIT were recorded in the Czech Republic and Hungary (39% and
36% in 2008, respectively) and the lowest level in Slovakia and Poland (28% and 32%,
respectively).
Taking into account the geographical patterns of the IIT development in the VCs,
some remarks can be formulated (Figure 4). Firstly, the upward trends of the IIT indices
in relation with the EU-15 were almost identical like those in total trade, due to the fact
that trade with the EU-15 accounted for about 60% of the analysed countries’ trade.
Secondly, the level of the IIT in relations with the EU-10 was relatively stable in
1995–2000, and then increased much faster, especially after the EU accession. Finally,
the share of the IIT in trade with the third countries increased in 1996–2004, and
especially rapid growth was noticed after 2000. Since the accession, the downward trends
in the IIT indices have been noticed.

Figure 3 The IIT indices in trade of the VCs in 1995–2008, in percentage

Source: Own calculations based on WITS-Comtrade data.

190 à. Ambroziak

Figure 4 The IIT indices in trade of the VCs with the major trading group partners in 1995–2008,
in percentage

Source: Own calculations based on WITS-Comtrade data.

As a result of these changes, the highest level of the IIT intensity was noticed in the VCs
trade with the EU-15 (44.0% in 2008) and it was by 11.4 p.p. higher than at the beginning
of the analysed period (Figure 4). Next, in trade with the NMS-12, the IIT index grew in
1995–2008 by 10.1 p.p. and amounted to 41.4% in 2007. Although the share of IIT in
total trade with the third countries has almost doubled in the analysed period, still it was
nearly four times lower than in relation with EU-27 countries. In 2008, only 11.0% of
trade with non-EU countries was determined by IIT specialisation. Such differences in
the level of IIT intensity between EU-27 and non-EU countries seem to be consistent
with the theory. They can be explained by different economic potential of trading
partners, EU-27 and non-EU, differences in their income level per capita, geographical
distance and level of the trade barriers.
Of great importance is also separation of the IIT into HIIT and VIIT. According to the
methodology by Greenaway et al. (1994), the basis for it is an assumption that price
changes reflect quality changes of trading goods. The criterion for this separation is based
on the ratio between the unit value in export and the unit value in import in trade between
two trading partners. In situation when it is impossible to compute this ratio, there is not
any possibility to make separation between HIIT and VIIT. This empirical analysis shows
that sometimes in the recent years, the defects of trade data were noticed.
In the VCs, the IIT was dominated all the time by trade in vertically differentiated
products (Figure 5). Its share in the total trade in the recent years was relatively stable.
However, big changes took place in the structure of VIIT in the analysed period.
A positive trend has been a declining significance of low-quality VIIT (countries
exported mainly low-quality products and imported high-quality ones) to the benefit of
high-quality VIIT’s share (countries exported mainly high-quality products and imported
low-quality ones). At the beginning of the analysed period, the latter accounted for about
one-fourth of total VIIT, while in 2008 it was circa a half. These changes reflect
structural transformation of the analysed economies. Another positive trend has been a
slow but gradual rise of HIIT: on average, its share in total trade of the analysed countries
grew from 4% to nearly 10% in the period 1995–2008.
 FDI and IIT: theory and empirical evidence from the VCs 191

Figure 5 The pattern of the IIT in the VCs (as a whole) in 1995–2008, in percentage

Source: Own calculations based on WITS-Comtrade data.

Some observations result from the above analysis. Firstly, the share of the HIIT was all
the time higher in trade with the NMS-12 countries than with the EU-15 countries. It
could be explained by the fact that on average the VCs are more similar to each other and
with regard to the whole group of NMS-12 (e.g. as measured by pattern of production
and the level of GDP per capita) than as compared with the EU-15. Secondly, the pattern
of VIIT with two groups of the EU countries was quite different in the analysed period.
The shares of high- and low-quality VIIT in trade with the NMS-12 were stable all the
time and accounted for about 10–13%, while in trade with the EU-15 more clear changes
of VIIT structure can be observed: in the analysed period, the share of low-quality VIIT
substantially declined and the share of high-quality VIIT substantially increased. It
reflected positive changes in the VCs trade pattern in relation with the EU-15: the
exported goods were of better and better quality, which means the progressive
specialisation of the VCs in more advanced products.

7 Model specification, explanatory variables and research hypothesis

This paper also analyses the determinants of the IIT (both types of IIT) in four Central
European Countries (Czech Republic, Hungary, Poland and Slovakia) in the period
1995–2008. Many previous papers have not considered the difference between VIIT and
HIIT. Several authors (e.g. Greenaway et al., 1994) have pointed out that this fact may
produce unbiased estimates. According to those authors, both types of the IIT have a
different nature. Following this approach, in this paper, different regressions for VIIT and
HIIT were estimated. A question was asked, whether there were differences between the
determinants of the two types of the IIT: HIIT and VIIT. Bilateral HIIT and VIIT indices
were computed in trade of each of the four analysed VCs for each year from the period
1995–2008. However, the size of the sample was determined by the availability of FDI
data. As a result, an unbalanced panel data with 1,769 observations were used. The
choice of the VCs trading partners and the years was determined by the availability of
192 à. Ambroziak

data, especially the availability of FDI data. Trade partners considered in the econometric
analysis accounted in most cases for more than 90% of individual VCs’ total trade.
Consistent with the literature on the determinants of IIT, the following regression
model was estimated:


HIITij VIITij   
E0  E1ln  GDPi   E2 ln GDP j  E3ln  GDPi _ pc 

   
 E4 ln GDP j _ pc  E5 EDij  E6 ln FDIij  E7 FTA ij

 
 E8 UEij  E9 borderij +E10 ln dist ij  const

HIITij (VIITij) means the horizontal (vertical) IIT index in trade between country i and j.
The independent variables were presented in Table 2.

Table 2 Variables and data sources

Expected change
a
Variables Description Data source HIIT VIIT
GDPi Gross domestic product of country i (reporter) – WEO database + +
purchasing power parity (PPP), in billion USD
GDPj Gross domestic product of country j (partner) – WEO database + +
PPP, in billion USD
GDPi_pc Gross domestic product of country i (reporter) WEO database + +
per capita – PPP, in USD per capita
GDPj_pc Gross domestic product of country j (partner) WEO database + +
per capita – PPP, in USD per capita
EDij ‘Economic distance’ computed by WEO database – +
max(GDPi _ pc,GDP j _ pc) / min(GDPi _ pc,
GDP j _ pc), (GDP_per capita in USD per capita,
PPP)
FDIij Foreign direct investments stock from country j WIIW database –/+ +
in the country i, in billion euro on FDI
FTAij Dummy variable with value 1 if country i and RTA-IS + +
country j have preferential trade agreement
(otherwise 0)
UEij Dummy variable with value 1 if country i and RTA–IS + +
country j are the EU Member States (otherwise 0)
borderij Dummy variable with value 1 if country i and CEPII database + +
country j have the common contig (otherwise 0)
distij Distance between capitals of country i and country j, CEPII database – –
in km
a
WEO database, World Economic Outlook database by the International Monetary Fund;
WIIW database on FDI, the FDI database of the Vienna Institute of International
Economic Studies; CEPII database, the French Research Centre in International
Economics database; RTA–IS, Regional Trade Agreement–Information System by the
World Trade Organisation.
 FDI and IIT: theory and empirical evidence from the VCs 193

The above-mentioned equation was estimated considering a fixed effect model or a

random effect model (shortly called as re model). Hausman tests led us to conclude that
in all cases the re effect model is more appropriate. To control for the time effects, a time
dummy variable was included into the re model. With regard to the fact that a large
portion of the IIT indices concentrated around zero in the sample, the Tobit model was
also adopted. For example, this method of estimation was recently used by Byun and Lee
(2005). All of the equations and diagnostic tests were estimated with the statistical
package STATA.
According to the majority of the previous studies, the following research hypotheses
seem warranted:
1 Both types of the IIT among countries are intense if their market sizes are large. In
this analysis, the market sizes are measured as the GDPs (at PPP, in billion USD) of
two trading partners (i and j).
2 IIT (both types) among countries is intense if their levels of development are high.
The GDPs per capita of country i and j (at PPP, in USD per capita) are taken to
indicate the development stages of these countries.
3 The smaller the difference in the level of development between two countries, the
higher the degree of HIIT and conversely for VIIT. In this empirical analysis, the
differences in development stage of the countries are measured as the so-called
‘economic distance’ computed as a ratio of the higher value of GDP per capita of
these countries to the lower value (GDPi,j_pc in USD per capita, PPP).
4 The larger the stocks of the FDI in the host country, the higher the share of VIIT,
while the influence of this variable on HIIT is ambiguous. The stocks of inward FDI
(from the major investor countries) in analysed countries (FDIij, in billion USD) are
used to test this hypothesis.4
5 The share of both types of the IIT will be higher when trading partners are
geographically close. The distance between capitals of two trading partners i and j
(distij, in km) and dummy variable borderij with value 1 if a reporter country i
borders with a partner country j (otherwise 0) are used to test this hypothesis.
6 The lower the barriers to trade, the higher the share of both types of the IIT. In this
analysis, the level of trade liberalisation is measured as EUij dummy variable with
value 1 if two trading partners are EU Member States and free trade area dummy
variable FTAij with value 1 if two trading partners are members of the same free
trade area.

8 The estimation results

As discussed earlier, three models have been estimated: the re model, the re model with
time effects and the Tobit model. The results of applied estimation methods are not the
same because not only value of coefficients and level of statistical significance differ, but
also sometimes the sign of a coefficient. In case of re models, the fit of the estimation for
HIIT and VIIT variables was relatively high, and the R2 exceeded 0.56 (HIIT) and 0.61
(VIIT).
Firstly, the estimated model was tested for the HIIT. The obtained results are
generally consistent with above-mentioned expectations (but economic distance).
194 à. Ambroziak

For several variables, the estimates are very stable for the different specifications, such as
GDPi, GDPj, GDPj per capita, distij and borderij. As expected, all variables (except for
distij) have positive and significant effects on the HIIT. On the contrary, GDPi per capita,
FTAij and UEij had positive impact on the HIIT and the coefficients of these variables
were not significant in all specifications. Although, according to the theory, the FDI
impact on the intensity of the HIIT is ambiguous, in all adopted specifications, the
estimates on FDIij were positively correlated with the HIIT and highly significant.
Finally, in case of economic distance, the signs of the estimated coefficients were
different and statistically not significant (Table 3).

Table 3 Determinants of the IIT in the VCs

HIIT VIIT
Random Random
Random effect, time Random effect, time
effect effects Tobit effect effects Tobit
Coefficients Coefficient Coefficient Coefficient Coefficient Coefficient
(SE) (SE) (SE) (SE) (SE) (SE)
GDPi 1.039* 1.021* 1.102* í0.035 1.348** 0.002
(0.263) (0.319) (0.166) (0.595) (0.65) (0.275)
GDPj 0.585* 0.612* 0.882* 2.278* 2.394* 2.262*
(0.12) (0.125) (0.082) (0.27) (0.256) (0.13)
GDPi_pc 2.092* 1.854 1.82* 2.211* 10.715* 4.123*
(0.52) (1.344) (0.444) (0.95) (2.612) (0.734)
GDPj_pc 0.794* 0.803** 0.731* 2.827* 3.213* 1.054*
(0.327) (0.348) (0.225) (0.664) (0.667) (0.37)
EDij 0.103 0.058 í0.004 0.017 í0.062 í0.516*
(0.206) (0.216) (0.128) (0.426) (0.433) (0.212)
FDIij 0.147** 0.133** 0.152** 0.417* 0.363* 0.924*
(0.069) (0.074) (0.056) (0.109) (0.114) (0.092)
FTAij 0.247 0.398 0.922* 1.066*** 1.070 5.641*
(0.399) (0.437) (0.342) (0.609) (0.675) (0.56)
EUij 0.554 0.972** 1.704* 0.441 1.38*** 4.511*
(0.459) (0.537) (0.4) (0.692) (0.819) (0.663)
distij í1.501* í1.433* í1.639* í3.922* í3.875* í2.803*
(0.235) (0.246) (0.168) (0.502) (0.483) (0.273)
borderij 3.137* 3.203* 3.317* 3.542* 3.896* 4.324*
(0.638) (0.657) (0.377) (1.488) (1.397) (0.629)
Const í23.304* í21.967 í21.681* í22.239* í112.831* í35.346*
(5.325) (14.884) (4.778) (8.886) (26.526) (7.897)
R2 0.569 0.575 – 0.616 0.646 –
No. of 1,769 1,769 1,769 1,769 1,769 1,769
observations
Note: (*) (**) and (***) denote values significant at 1%, 5% and 10%, respectively;
Time dummies were included in one specification but not reported.
Source: Own calculations.
 FDI and IIT: theory and empirical evidence from the VCs 195

The model was also reestimated using VIIT as a dependent variable. The most variables
which were stable for the HIIT were also stable for the VIIT model, such as GDPj, GDPj
per capita, distij and borderij. The estimates on these variables (but distij) were positive
and statistically significant. The signs on all mentioned variables confirmed the previous
expectations. As expected, the FDI variable was positively correlated with the VIIT and
the coefficients on this variable were significant at 1% significance level in all adopted
specifications. As for variables FTAij and EUij, the signs of their coefficients were
positive but only in the Tobit model were highly significant. Contrary to the HIIT model,
in one specification (also in the Tobit model) the coefficient on variable economic
distance was significant at 1% level. However, finally, the sign of this coefficient was
negative, which was not consistent with the theoretical framework. According to the
theory, differences in the economic distance between two trading partners may stimulate
fragmentation processes, and, as a result, inflows of the FDI, usually from the country on
the higher level of development to the country on the lower level.
It is important to notice that the estimated impact of the FDI on both HIIT and VIIT
was positive and highly significant in all applied specifications. The signs on this variable
for the VIIT were compatible with previous expectations. At the same time, the revealed
relationship between the FDI and the HIIT reflected not exactly a hypothetical
relationship between the variables resulting from the theoretical framework. It can mean
that the FDI inflows to the analysed Central European Countries have contributed to the
growth of the HIIT.

9 Conclusions

The importance of the IIT in the VCs has been gradually increasing. In 2008, more than
one-third of VCs trade turnover was of an IIT character. The growth of this type of trade
has been driven by both HIIT and VIIT. In comparison with global changes of the IIT
and its pattern, the scale of the VIIT in the VCs is relatively not high. There are also
different factors affecting the both types of the IIT. HIIT is mainly an effect of the
increasing economic convergence between the VCs and their main trading partners, it is
the other EU countries. Next, VIIT is mainly influenced by the FDI inflows, resulting
from joining of the VCs into international fragmentation of production processes. These
above-listed relationships were confirmed in the econometric analysis. Thus, one can
conclude that the FDI has a positive impact, both on HIIT and VIIT. In case of the HIIT,
the result seems to be a little bit surprising. If the impact of the FDI (VFDI) on the VIIT
is beyond all doubt, then the impact of FDI on the HIIT is unambiguous. The growth of
HIIT can be, in general, only directly an effect of the FDI inflows, because they promote
the economic development in the host country and diminish differences in market
potential and level of development between the host and home countries.

Acknowledgement

This research project was funded from a grant obtained from the Prof M. Kucharski
Scholarship Fund (Foundation for Managerial Education Development of the Warsaw
School of Economics).
196 à. Ambroziak

References
Arndt, S. and Kierzkowski, H. (Eds.) (2001) Fragmentation: New Production Patterns in the World
Economy. Oxford: Oxford University Press.
Aturpane, C., Djankov, S. and Hoekman, B. (1997) ‘Determinants of intra-industry trade between
east and west Europe’, The World Bank: Policy Research Working Paper, No. 1850.
Baltagi, B., Egger, P. and Pfaffermayr, M. (2007) ‘Estimating models of complex FDI: are there
third-country effects?’, Journal of Econometrics, Vol. 140, No. 1, pp.260–281.
Byun, J.J. and Lee, S.H. (2005) ‘Horizontal and vertical intra-industry trade: new evidence from
Korea, 1991–1999’, Global Economy Journal, Vol. 5, No. 1, pp.1–29.
Caetano, J. and Galego, A. (2006) ‘In search for determinants of intra-industry trade within an
enlarged Europe’, Economics Working Papers, Vol. 2, University of Évora.
ýernoša, S. (2007) ‘Horizontal and vertical intra-industry trade between the former CEFTA
Countries and the European Union’, Managing Global Transitions, Vol. 5, No. 2,
pp.157–178.
Cheng, L. and Kierzkowski, H. (Eds.) (2001) Global Production Networks and Trade in East Asia.
Boston: Kluwer Academic Publishers.
CieĞlik, A. (2003) ‘Handel wewnątrzgaáĊziowy Polski z Unią Europejską: Stan obecny i
perspektywy rozwoju’, in J.J. Michaáek, W. SiwiĔski and M. Socha (Eds.), Od liberalizacji do
integracji Polski z Unią Europejską: Mechanizmy i skutki gospodarcze. Warsaw: PWN,
pp.62–76.
CieĞlik, A. (2008) ‘Wpáyw przedsiĊbiorstw miĊdzynarodowych na fragmentaryzacjĊ produkcji
i handel wewnątrzgaáĊziowy Polski z krajami OECD’, Gospodarka Narodowa, Vol. 10,
pp.1–21.
Czarny, E. and ĝledziewska, K. (2008) ‘Poland’s intra-industry trade with the European Union at
the beginning of the 21th century’, in M.A. Weresa (Ed.), Poland Competitiveness Report
2008 Focus on Services. Warsaw: World Economy Research Institute, Warsaw School of
Economics.
Dixit, A.K. and Norman, V. (1980) Theory of International Trade. Cambridge: Cambridge
University Press.
Dunning, J.H. (1993) Multinational Enterprises and the Global Economy. Wokingham: Addison-
Wesley Company.
Egger, H., Egger, P. and Greenaway, D. (2007) ‘Intra-industry trade with multinational economics’,
European Economic Review, Vol. 51, No. 8, pp.1959–1984.
Ekholm, K., Forslid, R. and Markusen, J. (2007) ‘Export platform foreign direct investment’,
Journal of the European Economic Association, Vol. 5, No. 4, pp.776–795.
Falvey, R. (1981) ‘Commercial policy and intra-industry trade’, Journal of International
Economics, Vol. 11, pp.495–511.
Falvey, R. and Kierzkowski, H. (1987) ‘Product quality, intra-industry trade and (im)perfect
competition, in H. Kierzkowski (Ed.), Protection and Competition in International Trade.
Oxford: Blackwell, pp.143–161.
Ferto, I. and Soos, K.A. (2006) ‘The development of intra-industry trade between the European
Union and European Former Communist Countries before the 2004 enlargement’, INDEUNIS
Papers, Hungarian Academy of Sciences, Institute of Economics, February.
Fidrmuc, J. and Djablik, M. (2003) ‘Intraindustry trade between the EU and the CEECs – the
importance of FDI in trade structure’, Paper Presented at the East-West Conference, Vienna,
Austria, 2–4 November 2003.
Finger, J.M. (1975) ‘Trade overlap and intra-industry trade’, Economic Inquiry, Vol. 13, No. 4,
pp.581–589.
Flam, H. and Helpman, E. (1987) ‘Vertical product differentiation and North–South trade’,
American Economic Review, Vol. 77, pp.810–822.
 FDI and IIT: theory and empirical evidence from the VCs 197

Fontagné, L., Freudenberg, M. and Gaulier, G. (2006) ‘A systematic decomposition of world trade
into horizontal and vertical IIT’, Review of World Economics, Vol. 142, No. 3, pp.459–475.
Fukao, K., Ishido, H. and Ito, K. (2003) ’Vertical intra-industry trade and foreign direct investment
in East Asia’, Journal of the Japanese and International Economies, Vol. 17, pp.469–506.
Gabrisch, H. (2006) ‘Vertical intra-industry trade between EU and accession countries’,
IHW-Discussion Paper, No. 12, July.
Gabrisch, H. and Segnana, M.L. (2002) ‘Why is trade between the European Union and the
transition economies vertical?’, Discussion Paper, No. 7, University of Trento.
Gabrisch, H. and Segnana, M.L. (2003) ‘Vertical and horizontal patterns of intra-industry’, Institut
für Wirtschaftsforschung Halle, April.
Greenaway, D., Hine, R.C. and Milner, C. (1994) ‘Country-specific factors and pattern of
horizontal and vertical intra-industry trade in the UK’, Weltwirtschaftliches Archiv, Vol. 130,
pp.77–100.
Grubel, H.G. and Lloyd, P.J. (1975) Intra-Industry Trade: The Theory and Measurement of
Intra-Industry Trade in Differentiated Products. London: Macmillan.
Helpman, E. (1981) ‘International trade in the presence of product differentiation, economies
of scale and monopolistic competition: a Chamberlin-Hecksher-Ohlin approach’, Journal of
International Economics, Vol. 11, pp.305–340.
Helpman, E. and Krugman, P.R. (1985) Trade Policy and Market Structure. Cambridge, MA:
M.I.T. Press.
Hoekman, B. and Djankov, S. (1996) ‘Intra-industry trade, foreign direct investment, and the
reorientation of Eastern European exports’, The World Bank: Policy Research Working
Papers, No. 1652.
Jones, R.W. and Kierzkowski, H. (1990) ‘The role of services in production and international trade:
a theoretical framework’, in R.W. Jones and A.O. Krueger (Eds.), The Political Economy of
International Trade. Oxford: Blackwell.
Kandogan, Y. (2003) ‘Intra-industry trade of transition countries: trends and determinants’,
William Davidson Working Paper, No. 566, May.
Kawecka-Wyrzykowska, E. (2009) ‘Evolving pattern of intra-industry trade specialization of the
new Member States (NMS) of the EU: the case of automotive industry’, European Economy.
Economic Papers, No. 364, March.
Klimek, A. (2006) ‘Influence of European integration on intensity of intra-industry trade of
Poland’, Paper presented at the ETSG Conference, Vienna, Austria, 7–9 September 2006.
Krugman, P.R. (1979) ‘Increasing returns, monopolistic competition, and international trade’,
Journal of International Economics, Vol. 9, pp.469–479.
Krugman, P.R. (1980) ‘Scale economies, product differentiation, and the pattern of trade’,
American Economic Review, Vol. 70, No. 5, pp.950–959.
Krugman, P.R. (1983) ‘New theories of trade among industrial countries’, American Economic
Review, Vol. 73, No. 2, pp.343–348.
Lancaster, K. (1980) ‘Intra-industry trade under perfect monopolistic competition’, Journal of
International Economics, Vol. 10, No. 2, pp.151–175.
Leitão, N.C., Faustino, H.C. and Yoshida, Y. (2010) ‘Fragmentation, vertical intra-industry trade,
and automobile components’, Economics Bulletin, Vol. 30, No. 2, pp.1006–1015.
Lipsey, R. (1976) ‘Review of Grubel and Lloyd (1975)’, Journal of International Economics,
Vol. 6, pp.312–314.
Loertscher, R. and Wolter, F. (1980) ‘Determinants of intra-industry trade: among countries and
across industries’, Weltwirtschaftliches Archiv, Vol. 116, pp.280–293.
Markusen, J.R. (1984) ‘Multinationals, multi-plant economies, and the gains from trade’, Journal
of International Economics, Vol. 16, Nos. 3–4, pp.205–226.
Markusen, J.R. (2002) Multinational Firms and the Theory of International Trade. Cambridge, MA:
M.I.T. Press.
198 à. Ambroziak

Markusen, J.R. and Maskus, K.E. (2002) ‘A unified approach to intra-industry trade and
direct foreign investment’, in P.J. Lloyd and H-H. Lee (Eds.), Frontiers of Research in
Intra-Industry Trade. New York: Palgrave Macmillan.
Markusen, J.R. and Venables, A.J. (1998) ‘Multinational firms and the new trade theory’, Journal
of International Economics, Vol. 46, pp.183–203.
Markusen, J.R. and Venables, A.J. (2000) ‘The theory of endowment, intra-industry and
multinational trade’, Journal of International Economics, Vol. 53, No. 2, pp.209–234.
Molendowski, E. (2007) Liberalizacja wymiany handlowej krajów Europy ĝrodkowowschodniej w
okresie transformacji ze szczególnym uwzglĊdnieniem doĞwiadczeĔ krajów CEFTA. Cracow:
Wydawnictwo Uniwersytetu Ekonomicznego w Krakowie.
Neary, J.P. (2009) ‘Trade costs and foreign direct investment’, International Review of Economics
and Finance, Vol. 18, No. 2, pp.207–218.
Okubo, T. (2004) ‘Intra-industry trade and production network’, HEI Working Paper, Vol. 2004,
No. 13.
Shaked, A. and Sutton, J. (1984) ‘Natural oligopolies and international trade’, in H. Kierzkowski
(Ed.), Monopolistic Competition and International Trade. Oxford: Oxford University Press,
pp.34–50.
Stiglitz, J.E. (1987) ‘The causes and consequences of the dependence of quality on price’, Journal
of Economic Literature, Vol. 25, No. 1, pp.1–48.
Yoshida, Y., Leitão, N.C. and Faustino, H.C. (2009) ‘Vertical intra-industry trade and foreign
direct investment between Japan and European Countries’, Atlantic Economic Journal,
Vol. 37, No. 4, pp.351–365.

Notes
1
Leitão et al. (2010) stressed that VIIT capture two-way trades with different prices, which can
reflect the quality differentials or can be an effect of the fragmentation of production processes.
2
Generally speaking, we could classify these empirical analyses into studies that consider only
country-specific variables, only industry-specific variables and, finally, studies that consider both
types of variables.
3
NMS-12 means EU New Member States: ten countries, which joined the EU on 1 May 2004
(Cyprus, the Czech Republic, Hungary, Estonia, Latvia, Lithuania, Malta, Poland, Slovakia and
Slovenia) and two countries, which became the EU members on 1 January 2007 (Bulgaria and
Romania).
4
The choice of the inward FDI stock as a variable reflecting inflow of foreign capital to the home
economies was determined only by the availability of FDI data. A measurement of FDI impact on
the intra-industry trade seems to be quite complicated. There doesn’t exist a perfect FDI variable,
which could be without problems adopted into the model. On the one hand, the intra-industry trade
intensity has been influenced by FDI inflows in the analysed year, on the other hand, also by FDI
cumulated to this year, which has come in the analysed economy. The division of the two types of
impact is practically impossible. The usage of the FDI inward stock could entail some problems.
The FDI inward stocks adopted in the next years reflect every time a whole cumulated value of
FDI. Thus, it could influence estimation results (in plus). These above-mentioned problems will be
considered in the further research.