Foreign Direct Investment in Central Europe and Differences in Transition between post- communist Central European Economies
©2010
Masterarbeit
60 Seiten
Zusammenfassung
Inhaltsangabe:Introduction:
The transition process from a centrally planned to a market economy followed a very different path in East Germany compared to all other former communist countries. The German Democratic Republic acceded the Federal Republic of Germany in 1990, while other former socialist countries in Central and Eastern Europe (CEE) had to start from square one after becoming independent from the USSR. In contrast to other post-soviet countries, East Germany subsequently received massive transfers from the Western part of the country. A significant part of these transfers was invested into infrastructure improvement, while a larger share was spent for consumption, raising the purchasing power in the East of Germany, allowing it to sustain a higher wage level and living standard than would have been economically possible without aid from the West.
Twenty years after the breakdown of the iron curtain and the reunification of Germany, the infrastructure in the Eastern part of the country is en par with the West. The East German wage level remains only slightly lower than the Western level (as does productivity), but is significantly higher than in neighbouring post-communist CEE-countries. Because of these differences in economic transition, it can be expected that East Germany attracts a different kind of foreign direct investment compared to other CEE-countries. The objective of this dissertation is to empirically identify the factors affecting foreign direct investment into the region and to discuss the implications of the empirical findings for regional and national economic policy. The region is represented in this paper by East Germany and three of its Central-European neighbour-countries, the Czech Republic, Poland and Hungary.
The present study is organised as follows: A brief historical overview of the four economies transition processes will be given in chapter 2. The theoretical background, the methodology and the dataset used are being discussed in section 3. Hypotheses derived from the regions history, economic theory and previous literature on foreign direct investment will be presented in section 4. In section 5 they will be tested and the econometric results identifying the factors affecting foreign investors investment decisions in Central Europe will be discussed. Finally, the main empirical findings and their political implications will be summarised in the concluding chapter […]
The transition process from a centrally planned to a market economy followed a very different path in East Germany compared to all other former communist countries. The German Democratic Republic acceded the Federal Republic of Germany in 1990, while other former socialist countries in Central and Eastern Europe (CEE) had to start from square one after becoming independent from the USSR. In contrast to other post-soviet countries, East Germany subsequently received massive transfers from the Western part of the country. A significant part of these transfers was invested into infrastructure improvement, while a larger share was spent for consumption, raising the purchasing power in the East of Germany, allowing it to sustain a higher wage level and living standard than would have been economically possible without aid from the West.
Twenty years after the breakdown of the iron curtain and the reunification of Germany, the infrastructure in the Eastern part of the country is en par with the West. The East German wage level remains only slightly lower than the Western level (as does productivity), but is significantly higher than in neighbouring post-communist CEE-countries. Because of these differences in economic transition, it can be expected that East Germany attracts a different kind of foreign direct investment compared to other CEE-countries. The objective of this dissertation is to empirically identify the factors affecting foreign direct investment into the region and to discuss the implications of the empirical findings for regional and national economic policy. The region is represented in this paper by East Germany and three of its Central-European neighbour-countries, the Czech Republic, Poland and Hungary.
The present study is organised as follows: A brief historical overview of the four economies transition processes will be given in chapter 2. The theoretical background, the methodology and the dataset used are being discussed in section 3. Hypotheses derived from the regions history, economic theory and previous literature on foreign direct investment will be presented in section 4. In section 5 they will be tested and the econometric results identifying the factors affecting foreign investors investment decisions in Central Europe will be discussed. Finally, the main empirical findings and their political implications will be summarised in the concluding chapter […]
Leseprobe
Inhaltsverzeichnis
Jan Angenendt
Foreign Direct Investment in Central Europe and Differences in Transition between
post- communist Central European Economies
ISBN: 978-3-8428-0644-3
Herstellung: Diplomica® Verlag GmbH, Hamburg, 2011
Zugl. University of Warwick, Coventry, Großbritannien, MA-Thesis / Master, 2010
Dieses Werk ist urheberrechtlich geschützt. Die dadurch begründeten Rechte,
insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von
Abbildungen und Tabellen, der Funksendung, der Mikroverfilmung oder der
Vervielfältigung auf anderen Wegen und der Speicherung in Datenverarbeitungsanlagen,
bleiben, auch bei nur auszugsweiser Verwertung, vorbehalten. Eine Vervielfältigung
dieses Werkes oder von Teilen dieses Werkes ist auch im Einzelfall nur in den Grenzen
der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundesrepublik
Deutschland in der jeweils geltenden Fassung zulässig. Sie ist grundsätzlich
vergütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des
Urheberrechtes.
Die Wiedergabe von Gebrauchsnamen, Handelsnamen, Warenbezeichnungen usw. in
diesem Werk berechtigt auch ohne besondere Kennzeichnung nicht zu der Annahme,
dass solche Namen im Sinne der Warenzeichen- und Markenschutz-Gesetzgebung als frei
zu betrachten wären und daher von jedermann benutzt werden dürften.
Die Informationen in diesem Werk wurden mit Sorgfalt erarbeitet. Dennoch können
Fehler nicht vollständig ausgeschlossen werden und der Verlag, die Autoren oder
Übersetzer übernehmen keine juristische Verantwortung oder irgendeine Haftung für evtl.
verbliebene fehlerhafte Angaben und deren Folgen.
© Diplomica Verlag GmbH
http://www.diplomica.de, Hamburg 2011
i
Contents
1. Introduction
1
2. Historical Background
1
3. Methodology
8
3.1. Location Decisions and Economic Theory
8
3.2. Nested and Conditional Logit
10
3.3. Dataset
12
3.3.1. Primary Data
13
3.3.2. Secondary Data
15
4. Hypotheses
19
5. Empirical Results
24
5.1. Whole Sample
24
5.1.1. Nesting structures
24
5.1.2. Results
26
5.2. Firmsize
33
5.3. Industrial Sectors
34
5.4. Investing Countries
35
5.5. Change over time
37
6. Conclusion
40
6.1. General findings
40
6.2. Policy Implications
41
Appendix
v
Appendix A. Further results
v
Appendix B. Derivation of equations
ix
Appendix C. Derivation of the Nested Logit Model
ix
Appendix D. The IWH-FDI basic population
xiii
References
xvi
List of Figures
xviii
Wordcount: 11,982 words
1
1. Introduction
The transition process from a centrally planned to a market economy followed a very
different path in East Germany compared to all other former communist countries. The
German Democratic Republic acceded the Federal Republic of Germany in 1990, while
other former socialist countries in Central and Eastern Europe (CEE) had to start
from square one after becoming independent from the USSR. In contrast to other
post-soviet countries, East Germany subsequently received massive transfers from the
Western part of the country. A significant part of these transfers was invested into
infrastructure improvement, while a larger share was spent for consumption, raising
the purchasing power in the East of Germany, allowing it to sustain a higher wage level
and living standard than would have been economically possible without aid from the
West.
Twenty years after the breakdown of the iron curtain and the reunification of Ger-
many, the infrastructure in the Eastern part of the country is en par with the West.
The East German wage level remains only slightly lower than the Western level (as
does productivity), but is significantly higher than in neighbouring post-communist
CEE-countries. Because of these differences in economic transition, it can be expected
that East Germany attracts a different kind of foreign direct investment compared to
other CEE-countries. The objective of this dissertation is to empirically identify the
factors affecting foreign direct investment into the region and to discuss the implica-
tions of the empirical findings for regional and national economic policy. The "re-
gion" is represented in this paper by East Germany and three of its Central-European
neighbour-countries, the Czech Republic, Poland and Hungary.
The present study is organised as follows: A brief historical overview of the four
economies' transition processes will be given in chapter 2. The theoretical background,
the methodology and the dataset used are being discussed in section 3. Hypotheses
derived from the region's history, economic theory and previous literature on foreign
direct investment will be presented in section 4. In section 5 they will be tested and
the econometric results identifying the factors affecting foreign investors' investment-
decisions in Central Europe will be discussed. Finally, the main empirical findings and
their political implications will be summarised in the concluding chapter 6.
2. Historical Background
The fall of the iron curtain in 1989 opened up opportunities for the countries of Cen-
tral Europe to return to national sovereignty and to overcome the forced and decade-
lasting separation from the Western world. While the Western part of Europe was able
to benefit from economic integration with the rest of the free world during the cold war
2
and its modern market-economies allowed for rapid technological development, Central
and Eastern European countries had to (nearly) start economic development at square
one after the breakdown of communism.
But even before the Second World War, large parts of Central Europe were no part
of the Western world. They were "the lands between" (as Good puts it), between
a rapidly growing market-oriented Western world and a lagging-behind Russia, which
later became the Soviet Union [Good, 1992, p. 1]. While the Czech Republic and impe-
rial Germany (of which today's East Germany was a part of) followed a more Western
European path of development, Hungary and Poland were to be found somewhere in
the middle of the two systems [Good, 1992, p. 1]. This historic pattern continues
today: It is the Czech Republic that is closest to the East German level of prosperity,
despite not enjoying the vast subsidies received by East Germany, and some authors
predict the Czech Republic to overtake East Germany in 2013 [Hunt, 2006, p. 1f.].
Figure 2.1. GDP growth (market prices) of the four countries 1990-
2000. Source: [Burda and Hunt, 2001], own illustration.
Looking back, many authors agree with Hunt that the generous welfare benefits,
early retirement, public work jobs and the excessively high wage level are likely to have
lowered East Germany's attractiveness as an investment location, offsetting beneficial
pull-factors such as the comparably high quality of the East German infrastructure.
The most important factors leading to the very special development of the new German
states compared to their Eastern neighbours are related to either transfer payments
and/or political and economic institutions.
3
As the German Democratic Republic, from a legal perspective, joined the Federal
Republic of Germany on October 3rd 1990, it is widely believed that East Germany
should have been initially a more attractive region to invest than its Eastern neigh-
bour countries from an institutional perspective [Krkoska and Robeck, 2008, p. 569].
However, adapting the institutions of a developed economy may not be appropriate for
an emerging economy, as which East Germany can be described after the collapse of
the GDR. The difficulties associated with the adaption of West German institutions
are various: For example the currency union between East and West Germany led
to a massive appreciation of the East German Mark against the Deutschmark, as it
was exchanged on a 1:1 base (up to a certain limit) [Krkoska and Robeck, 2008, p.
569], letting the East German wage-level and the value of the debt of East German
companies skyrocket as a consequence of the exchange rate shock. To illustrate the
degree of overvaluation, the official exchange rate has to be compared with black mar-
ket rates, which were fluctuating between 4:1 and 5:1 [Niemietz, 2009, p. 72]. But
the circumstances under which such reforms had to be conducted should not be left
out of consideration: The adaption of the German Mark allowed the East Germans to
benefit from a very stable currency and helped to avoid the consequences of a very low
exchange rate of the East German against the West German Mark, which would have
led to a reduction in real wages and thus a decrease in living standards [Paqu´e, 2009, p.
31]. The consequence of a large gap in real wages between the Eastern and Western
part of Germany would have been mass-emigration from East to West. This is also
part of the explanation why the East German wage level rose within less than two years
from January 1990 to October 1991 from 31 to 49 percent of the West German level,
against the background of an economy in ruins [Paqu´e, 2009, p. 51] (also see figure
2.2). Another factor behind this unusual development were the West German trade
unions, which were expanding eastwards. High wage claims seemed like a sure method
to increase their support base in East Germany, while their West German members
were afraid of low-wage competition within their own country [Paqu´e, 2009, p. 52].
But facing 20 to 30 percent underemployment, German trade unions never had the
chance to become as powerful in the Eastern part of the country as they were in the
West [Paqu´e, 2009, p. 152]. The lower degree of organisation of East German employees
allowed East Germany to regain competitiveness lost in the past: One should note that,
despite the many comments of excessively high wages in East Germany, East German
labour costs are nowadays 14 percent lower than those in the old German states [Paqu´e,
2009, p. 148]. This has not always been the case: Unit labour costs had to shrink from
133 percent in 1995 to 100 percent in 2000 and to 85.7 percent of the Western level
in 2008. These days, in competition as an investment location with the West, East
Germany has an advantageously low wage-level compared to the Western level, possibly
4
also enabling it to make up for its disadvantages, such as its remote position within
Europe and its smaller market size.
Figure 2.2. East German consumption, GDP per capita and nominal
wages as percentage of the West German levels. Source: [Burda, 2006, p.
369], own illustration.
But let us go back in time again from the present situation to the events that
occurred after the breakdown of socialism. The magnitude of the economic downturn
after 1989 was enormous. East German GDP contracted by 15.6% in 1990 and 22.7% in
1991 [Burda and Hunt, 2001, p. 6], before returning to growth with 7.3% in 1992. The
recovery continued with growth figures in 1993 and 1994 remaining above 8 percent, but
GDP began to stagnate from 1995 onwards. Unemployment rose from 10.3% in 1991
(officially zero before 1991) to more than 19% in 1997 and has been slowly retreating
in recent years. How can these figures go along with a rapidly rising wage level? The
answer to this question lies in massive West-to-East transfers.
The total net amount of the transfers from the West adds up to between 1.2 and 1.4
billion Euros [Paqu´e, 2009, p. 184]. Figure 2.2 illustrates the gap between East German
GDP and consumption per capita. A large share of the transfers were spent for con-
sumption, i.e. by paying generous benefits to the unemployed or pensioners. Massive
active labour market programs to fight unemployment were started in the early years
after the reunification, which partially explains an 18-percentage-point gap between
the official unemployment rate and actual underemployment in 1992 [Burda and Hunt,
5
2001, p. 5]. By adapting the generous West German welfare system and its high wage
replacement rate, which East Germans could not have afforded on their own, the cost
of doing business in East Germany was further increased [Niemietz, 2009, p. 72]. High
social security contributions on top of the standard wages coexisted with economically
unreasonable high reservation wages. This combination seems toxic for economic de-
velopment. Niemitz correctly notes that East Germany was the first underdeveloped
economy that had to develop under the influence of a German-Scandinavian welfare
state significantly distorting economic forces. To illustrate this point, one should note
that West Germany spent 22% on welfare in 1950 and welfare spending only reached
26% in 1970 the time of the "Wirtschaftswunder". Modern day Germany's spending
on welfare amounts to approximately one third of GDP and this figure illustrates the
size of the welfare state that influenced the East German recovery process [Niemietz,
2009, p. 72]. Consequently, initial expectations of a repetition of the rapid develop-
ment of West Germany after the Second World War might have to be judged as rather
unrealistic.
Though inner-German competition for foreign investment, as addressed above, is
likely to be important for the future development of East Germany, it can be argued
that East Germany's natural competitors for foreign direct investment (FDI) in the
transition process were not the West German regions, but those in the other transition
economies of Central and Eastern Europe. Those economies received larger FDI-inflows
than East Germany (see figure 2.4), in absolute and per-capita terms, which can only
be partially explained by the fact that Germany is a major investor in Central Europe,
but West Germany is not being treated as a foreign investor in East Germany in the
data. As a location for foreign investment, East Germany has performed less well than
its neighbour countries.
Compared to the East German economy shocked by reunification, the economies
of the Czech Republic, Poland and Hungary fared significantly better in the years
directly following the 1990 revolution. The three countries' per capita GDP declined
in three consequent years by a lesser extent than the East German in the first years
(see figure 2.1), but their recovery started at an initially lower pace, as they did not
receive a boost from transfer payments following the post-communist decline. However,
growth in the three countries subsequently gained momentum. As can be seen from
figure 2.3, the gap in GDP per capita between the East German regions and the Central
European countries is nowadays closing, but they are not equally successful. The Czech
Republic is the wealthiest East Central European country in the present study, while
Hungary is being closely followed by Poland. If the catching-up-process continues, it
can indeed be expected that the Czech Republic may overtake the poorest East German
region, Brandenburg-Nordost, in the medium run. All countries benefit from foreign
6
Figure 2.3. GDP per capita of the three Central European countries,
Germany and Eastern Germany's richest (Berlin) and poorest NUTS-2-
region (Brandenburg-Nordost), 1995-2008. Source: Eurostat, own illus-
tration.
investment: Hungary was a major target for foreign investors from the start, while
Poland and the Czech Republic followed in 1995 [Dauderst¨adt, 2004, p. 4]. Foreign
investment flows into the other EU-accession countries (except Slovakia) were rather
limited, due to their smaller market sizes.
Figure 2.4. FDI stocks in East Germany (excluding Berlin), Poland,
Hungary and the Czech Republic. Source: Eurostat and Deutsche Bun-
desbank, own illustration.
7
Though the transition of East Germany is unique, it should be clear that the tran-
sition process in Central Europe also differed from country to country. Of the eight
countries that joined the European Union in 2004, only two existed in their present
borders (Poland and Hungary). The challenges they had to face were ranging from es-
tablishing democracy to transforming the economies and integration in world markets.
They had to privatise their companies and other state-owned property, which went
quickly in the Czech Republic and Hungary, but took longer in Poland [Dauderst¨adt,
2004, p. 1]. The governments had to introduce new currencies and start financing
their needs on international capital markets. Though not being able to simply adapt
an existing legal system, European law gave an orientation in the lawmaking process
in many cases. In contrast to East Germany, the other countries had to cope with
balance of payments deficits and inflation. This led to a pressure to devalue their cur-
rencies and devaluation then helped them to increase their competitiveness on world
markets [Dauderst¨adt, 2004, p. 4f.].
However, as the East Central European countries had to build their institutions from
scratch, their quality is not always perceived as equivalent to those in East Germany
[Krkoska and Robeck, 2008, p. 573]. Krkoska and Robeck evaluate data from an
enterprise survey and find that the quality of the legal system is likely to be perceived
as poorer by the enterprises' management in the East Central European countries
compared to Germany. On the other hand, the quality of the public administration is
likely to be perceived as nearly equivalent [Krkoska and Robeck, 2008, p. 574], though
corruption seems to be a bigger problem in East Central Europe compared to East
Germany [Krkoska and Robeck, 2008, p. 575] and access to finance is worse than in
East Germany.
In contrast, the share of part-time employees on the labour market tends to be lower
in the CEE-countries and East Central European firms invest more into the training of
their workforce [Krkoska and Robeck, 2008, p. 580]. As access to a high-wage labour
market was limited for the citizens of the East Central European countries before their
accession to the EU, none of them did go through a phase of mass emigration. The
Czech population fell by one percentage point since 1990 compared to 7 percent for
East Germany since 1991 [Hunt, 2006, p. 3]. Nonetheless, Hunt remarks that in the
19952005 period, Czech GDP per capita grew on average at 3.3% per year, while East
Germany only grew at an average rate of one percent, indicating a significant increase
in Czech productivity or employment.
The influence of trade unions in the transition process of the CEE-economies was
limited, lowering pressure on wage levels. Even nowadays the degree of unionisation
remains low of the three CEE-countries it is lowest in Poland (around 15%) and
highest in the Czech Republic (around 20%) [Kohl, 2008, p. 3]. Labour costs compared
8
to Germany remain low for example in the manufacturing industry they are at 20% of
the German level in Poland, at 26% in the Czech Republic and 25% in Hungary [Paqu´e,
2010, p. 14]. But the competitiveness of the Central European economies is still
hampered by rather low levels of productivity, ranging between 30 and 36% of the
German level in the manufacturing industry.
To summarise the above discussion, it is obvious that East Germany is wealthier
than its neighbours today, but the Central European countries are catching up. It
remains unclear whether the economic shocks and distortions caused by the integration
of East Germany into the Federal Republic were comparably harmful for its economic
development and future prospects, relative to those of its more "naturally" developing
neighbours. In the next chapter, the empirical models used to gain insights on this
issue will be presented.
3. Methodology
3.1. Location Decisions and Economic Theory. The analysis of investment deci-
sions in this dissertation is motivated by the New Economic Geography (NEG) litera-
ture founded by Paul Krugman [Krugman, 1991]. NEG models emphasise the impor-
tance of cross-border regional characteristics in the decision-making process of firms.
This is why in this study, firms are assumed to make their location-decisions on the
level of the European Union's regional statistical units (NUTS-2 regions). It is being
assumed that the decision-making firm has already decided to invest in Central and
Eastern Europe, for example because it wants to serve the local market by producing
locally instead of exporting goods to the region or to take advantage of lower production
costs.
Following and slightly modifying the approach taken in [Spies, 2009] I assume that
the investing firm faces the following profit function:
(3.1.1)
ik
=
J
j
(1 - t
i
)(P
ij
(x
ij
) - c
ik
ij
)x
ij
- f
ik
The expected profit for a company operating in sector k and investing in region i is
determined by the profit tax t
i
, the price p
ij
and the quantity x
ij
the company can sell
in all markets j J and region i itself, as well as on marginal costs c
ik
and fixed costs
f
ik
. Marginal costs c
ik
are given by c
ik
= c
ik
(w
ik
,
i
, L
ik
,
), where w
ik
is the wage level,
L
ik
the number of employees working in industry k in region i and
i
is the tax-wedge
on labour, while is a parameter influencing the productivity of the workforce. Fixed
costs are given by f
ik
= f
ik
(z
ik
, s
ik
,
), where , < 0, z
ik
is a variable measuring
the diversity of region i's economy accounting for possible backward-linkages and s
ik
9
measures the specialisation of region i in industry k, representing possible competitive
advantages on regional level, while captures all other factors, such as land prices.
The total quantity the firm located in region i sells, x
i
, depends on the size of region
i
's own market GDP
i
and the sizes of all other markets j accessible from this region
J
j
GDP
j
as well as on the price P
ij
for which the company sells its product in the
respective market
(3.1.2)
x
i
= x
i
(GDP
i
,
J
j
GDP
j
, P
ij
)
Assuming monopolistic competition and maximising equation (3.1.1), we get the
company's price depending on marginal production costs in region i and transport
costs from region i to regions j
(3.1.3)
P
ij
=
-
1
c
ik
ij
where
ij
are iceberg-type transportation costs, which will influence the company's
competitiveness on foreign markets and are likely to influence its location with regard
to market access, and is the price-elasticity of demand. Using the result in equation
(3.1.3), we can rewrite the profit equation as
1
(3.1.4)
ik
=
J
j
(1 - t
i
)
1
1 -
c
ik
ij
x
ij
- f
ik
Taking logs, the following log-linear specification can be found
ln
ikt
=
0
+
1
ln t
it
+
2
ln
it
+
1
ln w
ikt
+
2
ln L
ikt
+
3
ln
it
(3.1.5)
+
4
ln Y
it
+
5
ln
C
c
Y
ct
D
ic
+
6
ln z
it
+
7
s
ikt
+
8
ln
it
+
9
ln
it
where now denotes coefficients on country-level variables and coefficients on
region-specific variables. Taking the variables in logs and using a dataset with a large
number of location decisions also has the advantage that, in the econometric analysis,
the estimated coefficients are close to elasticities of the probability of choosing a region
for the average investor [Mayer et al., 2009, p. 16]. Firms choose from the set of 40
NUTS-2-regions of East Germany, the Czech Republic, Hungary and Poland and locate
their affiliate where they can expect the maximum profit from the investment. Region
i
is being chosen if
1
See appendix B for the (relatively simple) derivation.
10
(3.1.6)
ci
>
cj
j C
holds for all alternative regions j in the set of regions J in all countries C. The
most appropriate econometric strategies to model the individual firms' choices are
(additive) random utility models the conditional and the nested logit model, which
will be discussed below in subsection 3.2. Afterwards the dataset used will be presented
in sections 3.3.1 and 3.3.2, the latter section including a further discussion of the
explanatory variables.
3.2. Nested and Conditional Logit. The dataset used in this study includes 40 re-
gions in four different countries. As regions within a country (or a group of countries)
share several characteristics e.g. from a common culture and language to similarities
in taxation it seems appropriate to consider these similarities in the econometric
modelling of firm-level location choices. The nested logit model proposed by [McFad-
den, 1978] takes the heterogeneity of the regions within the four countries into account
and the modelled decision-making process can be interpreted as a two-level process:
Investors first choose the country (supranational region) they want to invest in, and
a region within the country (supranational region) afterwards. The interpretation of
the model as a model describing sequential decision-making processes is not necessary
though investors are simply assumed to be aware of the similarities shared by re-
gions in a country when choosing one of the 40 regions. The nested logit model allows
correlation of the error terms for regions within the country (supranational region)
level.
The conditional logit in contrast assumes independence of all errors and is a special
case of the family of nested logit models [Cameron and Trivedi, 2009, p. 512]. In
a simple conditional logit setup, all 40 regions are treated as if they were perfect
substitutes and thus the nested logit model can be expected to be the more appropriate
approach of the two models in this study. To illustrate the econometrics used to derive
the results in chapter 5, assume that a company's expected profit from investing in
regon i is given by
2
(3.2.1)
ci
= V
ci
+
ci
where c is the country-subscript, i denotes the region within the country. V is the
deterministic part of the profit function, while represents the stochastic part. The
countries (or alternatively supranational regions) c are on the upper level of the decision
making tree, while the region i within the respective country is being chosen in a second
2
The time-subscript has been omitted in equation (3.2.1).
11
step. The probability of choosing region i in country c is given by (as in [Cameron and
Trivedi, 2005, pp. 508ff.])
3
(3.2.2)
p
ci
= p
c
× p
i|c
p
i|c
is the conditional probability of choosing region i, conditional on country c being
chosen by the investor, and p
c
is the probability of the investor choosing country c.
The deterministic part of the profit function is in our case given by
(3.2.3)
V
cik
=
c
+ z
c
+ x
ci
where are the country-specific and the region-specific regression-parameters,
while z
c
are explanatory variables on country- and x
ci
on region-level respectively.
4
The nested logit model is now given by
(3.2.4)
p
ci
=
exp(
c
+ z
c
+
c
I
c
)
C
m=1
exp(a
m
+ z
m
+
m
I
m
)
×
exp(x
ci
/
c
)
L
m
l=1
exp(x
ml
/
c
)
where L
m
the number of regions within a nest, C is the number of nests and I is
the inclusive value the average profit a firm can expect from investing in a region of
a nest (see [Hausman and McFadden, 1984, p. 1227]), defined by
(3.2.5)
I
c
= ln(
Ic
i=1
exp
(x
ci
/
c
))
If the coefficient on the inclusive value equals one (
c
= 1), the conditional logit
model arises [Cameron and Trivedi, 2005, p. 511].
5
The conditional logit is specified
for the 40 regions in our model as
(3.2.6)
p
ci
=
exp(
c
+ x
ci
+ z
c
)
C
m
L
m
l=1
exp(
m
+ z
m
+ x
ml
)
For a nested logit model to be consistent with random utility maximisation, the
coefficient on the inclusive value should lie in the interval [0, 1] [Cameron and Trivedi,
2005, p. 511]. However, [Hausman and McFadden, 1984, p. 1228] state that it is
still possible to use the results of a nested logit model if the coefficient
c
lies outside
3
See appendix C for a detailed derivation of the model.
4
As can be seen from equation (3.2.4), it is not relevant for a random utility maximisation nested
logit estimation whether a nest-specific independent variable is specified for the nest or all alternatives
within the nest (also see [Heiss, 2002, p. 12] or Stata 10's nlogit command.)
5
In the conditional logit model, country-specific factors can be taken into account by including
country-dummies. However, similar reasons apply to also include nest-specific-dummies in the nested
logit specification, in the present case to account for the component of the profit-function that is
constant across the alternatives in a nest (e.g., see [Lubowski et al., 2005, p. 7] and equation (C.0.22)).
Details
- Seiten
- Erscheinungsform
- Originalausgabe
- Jahr
- 2010
- ISBN (eBook)
- 9783842806443
- DOI
- 10.3239/9783842806443
- Dateigröße
- 1.8 MB
- Sprache
- Deutsch
- Institution / Hochschule
- University of Warwick – Economics
- Erscheinungsdatum
- 2010 (November)
- Note
- 75%
- Schlagworte
- central europe foreign direct investment transition economic geography
