Foreign direct investment in the Czech Republic: the role of origin effects and government promotion abroad.

Deichmann, Joel I.

INTRODUCTION

After more than a decade and a half of political and economic reforms, the Czech Republic is widely considered to be among the most successful cases of transformation among the members of the former Soviet bloc. In 2006, foreign direct investment (FDI) in the Czech Republic, as both a driver of this success and a reflection thereof, reached more than $60 billion in cumulative stock, translating to $6,337 per capita, the highest in the region (Berglof, 2008). (1) In recognition of this success, the country has been placed in the category of 'frontrunner' (UNCTAD, 2007) in terms of inward FDI performance and potential, and the World Bank (2008) reports that annual inflows from 1993 through 2007 averaged 5.6% of gross domestic product, underscoring the importance of FDI to the economy of this landlocked Central European state.

In pursuit of convergence with Western Europe, Czech leadership has striven to distinguish the country as the leader among transition states. The first democratically elected governments assertively courted constructive investment from abroad as an engine for development while limiting potentially negative effects (Pavlinek, 2004). Governments recognize the importance of marketing their national advantages as FDI destinations through promotion agencies as part of broader policy initiatives (Breznitz, 2007), (2) and recent work by Capik (2007), Benacek (2008), and Drahokoupil (2008) has explored aspects of such promotion in Central and Eastern Europe (hereafter CEE).

As investment into CEE accelerated, a substantial literature also emerged from research into the determinants of the spatial distribution of these inflows. Location factors that are competently researched and empirically corroborated include trade links (Brenton et al., 1998), agglomeration (Head et al., 1999), trade openness (Kinoshita and Campos, 2003) and other relational factors (Bandelj, 2002), labor costs and gravity factors (Bevin and Estrin, 2004), policy consistency and economic stability (Ass and Beck, 2005; Grosse and Trevino, 2005), and political risks (Brada et al., 2006), among others (see Blonigen (2005) for a thorough review). Many of these factors simultaneously interact with industry (Eckert and Rossmeissl, 2005) and entry mode (Dikova and van Witteloostuijn, 2007). Another related literature critically assesses the impact of investment upon regional development in the host countries (Pavlinek, 2004) and broader convergence (Eckert and Rossmeissl, 2005).

In part because the EU accession process occurred over several years and attracted great international attention, economic integration in Europe had already been identified in several works as an important enabling factor of FDI when the 2004 candidates had only preliminary EU status (Brenton et al., 1998; Bandelj, 2002; Deichmann, 2004), and this was confirmed in the Czech case as well (Kinoshita and Campos, 2003; Eckert and Rossmeissl, 2005). Recognizing the success of EU accession states, they are now generally considered a cohort distinct the former Soviet Republics such as Belarus, Ukraine, Moldova, and the countries of Turkestan, which lag far behind Central Europe in nearly all socio-economic measures (Janicki and Wunnava, 2004; Brada et al., 2006; Berglof, 2008; World Bank, 2008).

The present paper's intent is to augment the existing literature by examining FDI from the angle of country-of-origin effects (hereafter 'origin effects'), an approach that is essential, yet less thoroughly understood, according to Dunning (1980, 2001) and echoed by Hunya (2000). It is important to understand why the Czech Republic has differential appeal to firms from different locations around the world. The largest sources of FDI with values of $30 million or more are presented in Table 1, and they are concentrated in Western Europe, North America, and East Asia. The top five countries are Germany, the Netherlands, Spain, Austria, and France, all of which are geographically close to the Czech Republic. Moreover, Germany and Austria also share long borders with the country, suggesting that effects of proximity might be at work.

Outside of Europe, most of the other leading home countries--notably the United States (#6), Japan (#11), and Canada (#18)--are relatively large economies, as measured by GDP. In addition, it is worth noting that several former eastern-bloc trade partners, including Slovakia (#10), Poland (#13), Hungary (#21), Russia (#22), and Slovenia (#30) appear in the list of leading origins. As part of Czechoslovakia until 1993, Slovakia fares prominently probably because of remaining business units in the Czech lands. Cursory observation of the other patterns hints that explanatory variables identified elsewhere are worth investigating in this context, particularly border effects (O'hUallachain and Reid, 1992), gravity model factors (Grosse and Trevino, 1996), and cultural factors (Zhao and Zhu, 2000; Bitzenis, 2004).

The ranking of home countries should be interpreted with caution, especially the cases of the Netherlands and Spain. A favorable bilateral investment treaty between the Netherlands and Czech Republic facilitates transactions, and the Dutch registration of foreign corporations from third countries holds special tax benefits for companies as well). (3) For example, the Prague office of Bank of Tokyo-Mitsubishi (the Netherlands) is registered with Dutch--rather than Japanese--origins (BusinessInfo.cz, 2006). The Seychelles, Saint Lucia, and the Bahamas all rank among the top 40 origins for similar reasons. Outliers provide another reason for caution in interpretation: Telefonica's 2005 acquisition of Cesky Telecom for nearly $5 billion is the largest single transaction to date and is largely responsible for Spain's ranking of third after being ranked twentieth as recently as 1999.

LITERATURE

John Dunning's eclectic (OLI) theory of international production (1980, 2001) has been the inspiration for a wide range of published work on FDI. The framework is simple: advantages relevant to ownership, location, and internalization are all essential in order to fully understand why FDI takes place. These dimensions represent three separate sets of questions in any given context, and have been widely researched as such. The most common approach is to examine location-specific advantages across countries (Bevin and Estrin, 2004; Janicki and Wunnava, 2004; Blonigen, 2005; Brada et al., 2006). Hunya (2000) explores the interaction of advantages in origin and CEE destination countries, while other scholars simultaneously look at ownership advantages and sub-national location choice (O'hUallachain and Reid, 1992; Zhao and Zhu, 2000). In general, the literature favors breadth of analysis over depth by examining large data sets rather than specific cases.

Notwithstanding the few recent exceptions noted above, Hunya (2000, p. 87) laments that the country-origin of FDI is 'a neglected issue'. He contributes a thorough descriptive assessment of the flows of FDI from home countries to CEE states and speculates that trade is the main explanation, but stops short of empirical tests. Citing evidence from other works (in particular, pan-European research by Brenton et al., 1998), he finds evidence of three types of trade relationships related to FDI: capital, exports, and imports with Germany, Austria, France, and the UK; capital links with the USA and Netherlands; and trade links with Russia, other CEE countries, and Italy. The author also recognizes that proximity plays a role. Finally, he reckons that policy can be instrumental when a 'foreigner friendly' privatization scheme is in place, because at the time of writing about half of the FDI in the countries in question had come through privatization (Hunya, 2000, p. 102).

In the tradition of Dunning's (1980) seminal work on origin effects, a handful of empirical inquiries have been published on such enabling factors, providing a template for the present research. These include analyses of FDI in the United States (O'hUallachain and Reid, 1992; Grosse and Trevino, 1996), China (Luo, 1998; Zhao and Zhu, 2000), Poland (Deichmann, 2004), Bulgaria (Bitzenis, 2004), and Europe (Brenton et al., 1998). From these county-specific inquiries a list of mainstream variables has evolved.

O'hUallachain and Reid (1992) contend that border effects may enable FDI. Border effects arise from home country familiarity with adjacent territory or from cultural similarities between the home country and host region. The authors identify patterns using location quotients for Canadian, British, Latin American, and Japanese investment in the USA, and analyze regression results for each home country. Border effects can be traced to subtle linkages of communications and transportation infrastructure that facilitate international migration and tourist flows, which the authors conclude are important for Japanese, Latin American, and Dutch manufacturing investment.

Grosse and Trevino (1996) test an extensive list of determinants using multivariate regression. They find that FDI flows to the USA are governed by trade, geographical and cultural distance, and political risk at home. Among the authors' most interesting findings are the divergent and significant impacts of each direction of trade flow. FDI is found to be positively related to home countries' exports to the USA, while imports to the home country have a negative effect on FDI in the USA.

Brenton et al. (1998) report upon their simple but revealing application of Linnemann's (1966) gravity model to the EU and Central Europe. The authors compare the coefficients of FDI and trade (imports and exports handled separately) in a straightforward equation featuring income, population, geographical distance, and dummies for preferential trade agreements. They find evidence that FDI into CEE countries follows the same determinants as elsewhere, conforming to the gravity model and exhibiting a complementary relationship with trade (in other words, they find that FDI does not serve as a substitute for trade).

Zhao and Zhu's (2000) work attempts to identify origin-specific responses to location factors within China. The authors cluster home countries according to cultural similarities (Hong Kong, Singapore, Japan, USA, and Europe). From Ordinary Least Squares (OLS) regression, they conclude that management style, business logic, and cultural backgrounds shared within these groups are indeed important in selecting locations, as are complementary factors of production between the host region and locations within China.

In the context of Poland, Deichmann (2004) upholds the applicability of the gravity model by confirming the importance of GDP and geographical distance as determinants of FDI origins. The role of early 'Europe Agreements' is found to be strong, even prior to Polish accession to the EU. The author also finds evidence that some FDI can be attributed to Polonia, or the relocation of poles primarily to North America and Western Europe.

Bitzenis (2004) assesses origin-specific factors in Bulgaria, a country where FDI has thus far been relatively scarce, appropriately framing the research question around 'explanatory variables for low (FDI) interest'. Bulgaria's legal system, bureaucracy, and crime are all perceived differentially by investors according to their home country. According to UNCTAD (2007), by the end of 2004, Bulgaria had attracted only 13.4% as much investment as the Czech Republic (7.6 billion versus 56.4 billion). The survey approach adopted by Bitzenis unveils barriers, obstacles, and disincentives to investing in Bulgaria, yet observes 'significant' Russian FDI there during years when Russia's situation was even more volatile.

Finally, Brada et al. (2006) note the dramatically different levels of FDI in Central European states vis-a-vis the Balkan region, and examine the role of political risk as a factor. Dissatisfied with the literature's measurement of political instability with reference to strikes, riots, civil unrest, the authors apply the same equation to the Balkans that they use in seven countries of the Visegrad and Baltic regions, then examine the differences between expected and actual outcomes. Using time-series data, they find clear evidence that FDI to countries involved in conflict (notably Croatia, Bosnia, and Slovenia) and their neighbors (Bulgaria and Romania) suffered, especially prior to the Dayton Peace Accords of 1995. The return of relative stability since that time has begun to attract better-than-expected inflows.

Table 2 summarizes the empirical contributions that follow the 'ownership' (country-of-origin effects) dimension of Dunning's theoretical flamework as described in this section.

Together, these contributions represent a small but valuable literature on home country motivating or enabling factors for FDI. Based upon these examples, it can be concluded that an appropriate and manageable scope of analysis is to focus upon a single destination country in order to pay adequate attention to the variables themselves.

Two intuitively plausible enabling factors that are absent from the specific literature discussed above are borrowed from elsewhere as appropriate additions to the models. The first is agglomeration, pioneered by Knickerbocker's (1973) work on oligopolistic reaction and later empirically corroborated as a FDI location determinant by Head et al. (1999). In short, when expanding to unfamiliar territory, firms show a tendency to cluster in an effort to replicate success, obtain a portion of a proven market, and/or share ancillary services. The second additional variable also draws from research by Head et al. (1999), who examine the role of investment promotion offices abroad without empirical confirmation, leading to the inclusion of this variable in the present study.

CzechInvest abroad: a context-specific origin-effect?

The agency CzechInvest was established in 1992 within the Czech Republic's Ministry of Industry and Trade. Its staff handles investment promotion and support in three phases: (1) International Marketing, (2) Site Selection and Development, and (3) Aftercare. The available services can be summarized as follows: information provision, incentive handling, properties and supplier identification, infrastructure development, and facilitation of structural funds (CzechInvest, 2009). Examples include the 2009 completion of the Holesov Strategic Industrial Zone, and the identification of partners for expanding small and medium-sized investing firms.

Because Czech policy explicitly treats foreign business entities identically to one another and to domestic entrepreneurs, the expectation that policy in practice favors investors from certain locations warrants some clarification. In short, following expectations by Head et al. (1999) that US states' investment promotion offices in Japan might play a role in investment to those states, it is similarly anticipated that firms from countries with CzechInvest offices will be more likely to invest in the Czech Republic. In Phase I of its charge (International Marketing), the agency's presence abroad captures enormous potential to create awareness of opportunities and promote the Czech brand directly in targeted countries.

Breznitz (2007) examines government investment policy in Taiwan, Israel, and Ireland. Ireland's operation of 17 investment offices abroad is comparable to the Czech promotion strategy. Breznitz also overviews tools at the disposal of government to guide industrial change with the help of FDI, especially in the information technology sector.

Other aspects of government policy have already been analyzed in the literature, although Beyer (2002) concludes that policy in general has had only minimal impact on FDI in the region. Moreover, Mallya et al. (2004) as well as Drahokoupil (2008) argue that any success in attracting investment may not be worth the expense. Survey research by Mallya et al. (2004) investigates the Czech National Incentive Scheme in 'crowding-in' investment, finding that, at best, it has increased FDI in manufacturing by 10%. They argue that by using tax holidays as the main incentive lever, a considerable net cost of US $200,000-$400,000 is incurred by the host society for each new job. Capik (2007) finds cause for optimism, but acknowledges that the benefits to regions in Poland and the Czech and Slovak Republics are undermined by fierce competition between investment promotion agencies.

Focusing exclusively on Central Europe, Drahokoupil (2008) assesses critically what he calls the investment-promotion 'machines' as ad hoc amalgamations of self-interested parties led by agencies such as CzechInvest. He argues that the unfortunate Visegrad Four result is 'competition states' that lobby too hard for investment, often at the expense of their own citizens and the integrity of their own laws (2008, p. 206). Although his case studies confirm their effectiveness in luring well-known foreign firms such as Nemak, LG Philips, BMW, PSA Peugeot Citroen, Kia Motors, and Hyundai, he laments that they may also lead to 'suffering and conflict within communities' (2008, p. 218).

Unfortunately, neither Mallya et al. (2004), Capik (2007), nor Drahokoupil (2008) specifically question the role of CzechInvest offices abroad, which have taken investment promotion to a new level with foreign offices in Cologne, Chicago, Silicon Valley, London, Paris, Yokohama, Hong Kong, and Brussels. It is no coincidence that Germany, the United States, the United Kingdom, France, and Belgium represent five of the top nine origins, with Japan following closely. Each CzechInvest office itself represents a strategic location decision by the Czech government, targeting industries in which Czechs consider themselves competitive. Frelich (2006) points out that the Chicago CzechInvest office courts American manufacturing companies, while the Silicon Valley location was selected to reach the largest US cluster of high technology firms. In addition to promoting the Czech Republic, publicizing its partners and a highly skilled Czech workforce, the office facilitates ongoing 'Aftercare' (Frelich, 2006).

METHODS AND DATA

This paper employs OLS regression to measure the role of CzechInvest and other independent variables on cumulative inflows from 2000-2005 ('FDI0005'), as well as inflows during the peak year of 2006 alone ('FDI06'). The dependent variables are reported by the Czech National Bank (2007). FDI0005 is calculated by subtracting the cumulative of FDI through 1999 from the value through the end of 2005 (in other words, resulting in the total FDI from 2000-2005). The rationale for this simple calculation is to retain cumulative FDI through 1999 as an independent variable for examining 'follow-the-leader' strategies without double counting. Following the specification of models for cumulative flows, FDI during 2006--the largest annual inflow thus far--becomes the dependent variable in order to compare the initial findings with those using more recently published data.

Table 3 defines the independent variables, for which mean values over 5 years (2000-2004) were used. The calculation of means is deemed to be more realistic than using a time lag in this case because it is capricious to estimate the length of time required for enabling factors to influence decision-making. Moreover, mean values keep the handful of observed outliers in check. Most of the independent variables are remarkably stable over time, although the dependent variable spiked in 2005 because of major investment activity from specific origins. (4) In the case of variables with skewed distributions (FDI values, GDP, AGG, and TRADE), logarithms were calculated to product normal distributions. Missing values are replaced by means, resulting in a maximum population of 191 countries in the models.

In the initial model, all variables are entered as follows:

log[(FDI0005).sub.i] = [alpha] + [B.sub.1] log([GDP.sub.i]) + [b.sub.2][GNI.sub.i] + [b.sub.3] log([AGG.sub.i]) + [b.sub.4][EXRATE.sub.i] + [b.sub.5] log([TRADE.sub.i]) + [b.sub.6][EU.sub.i] + [b.sub.5][CULT.sub.i] + [b.sub.8][TECH.sub.i] - [b.sub.9][DIST.sub.i] + [b.sub.10][CZI.sub.i] + [b.sub.11][CPI.sub.i]

where log([FDI0005.sub.ij]) =logarithmic transformation of the value of investments from home country i to the Czech Republic. Subsequent models use the dependent variable log([FDI06.sub.i]) to examine change in the role of the determinants, [alpha] = constant for fitting the equation, [b.sub.1], ..., [b.sub.11] = coefficients for each independent variable.

As summarized in Table 3, FDI in the Czech Republic is expected to be positively related to the size of the origin economy, its market strength, existing investment from that economy, exchange rates, trade flows, and R&D personnel. Following Bandelj's (2002) work on relational factors, FDI expected to be facilitated through EU membership and cultural proximity, and discouraged by geographic distance. With regard to government policy, in line with Benacek's (2008) affirmation of CzechInvest as a model of government competence, the expectation is that this agency is effective at attracting FDI. High corruption levels at home (low corruption perception index (CPI)) are expected to also encourage FDI in the Czech Republic as a relatively transparent environment (ranked forty-seventh), following Cuervo-Cazurra's (2008) assertion that additional costs and uncertainty are associated with operating in corrupt countries.

Moderate collinearity is symptomatic of OLS regression of many variables, and is anticipated especially in the initial model, which simply represents a means to a more parsimonious solution. (5) In addition to the large number of determinants called for by the literature, home countries of investors are generally high-income, and therefore share similar values on economic measures, which are highly correlated with spatial characteristics such as distance and cultural similarities. In response, variables are removed from successive models as flagged by problematic simple correlation coefficients, tolerance, and variance inflation factors (VIFs), keeping in mind the intention to retain for examination of the variables that are poorly represented in the literature and of greatest interest here: government policy (CZI), and agglomeration (AGG).

ANALYSIS

This section reports upon five models that were specified to identify the strongest origin effects, paying special attention to the role of CzechInvest. To begin with, Model 1 includes all of the variables as found to be justified by existing evidence from other contexts. Next, a more parsimonious model (#2) is presented, specified by Statistical Package for the Social Sciences (SPSS)'s forward selection algorithm, and including the three best-performing determinants. Third, a model (#3) is deliberately specified to include agglomeration, EU membership, and CzechInvest presence abroad as these variables are of greatest interest here. Model 4 replicates Model 3's independent variables, but considers FDI in 2006 as the dependent variable. Finally, Model 5 is specified by SPSS's forward selection algorithm to optimize for the year 2006 and compare mainly with Model 2 (optimized through 2005). The overarching goal of this approach is to triangulate through a series of model specifications and identify continuity in variables that perform well across the models.

As an initial model, Model 1 includes all 11 variables and performs quite close to expectations with most of the valence signs as predicted. The variables that perform best are the logarithmic transformations of agglomeration (AGG) and total trade (TRADE), both positive determinants of FDI that are significant at the 0.001 level of confidence. Evidence of follow-the-leader strategies in the Czech Republic corroborates findings by Head et al. (1995, 1999) from the US context. The observation that firms practice herd behavior by national origin is discussed further in this section. Trade emerges as the second most important origin-effect with regard to FDI in the Czech Republic, perhaps as a precursor to FDI as originally observed in Sweden by Johanson and Wiedersheim-Paul (1975). This evidence of complementarity refutes earlier theory by Vernon (1966) that trade and FDI are necessarily competitive modes of internationalization. This observation also supports findings elsewhere in CEE by Brenton et al. (1998) and Hunya (2000) that the two modes of internalization can reinforce one another and are not mutually exclusive.

Subsequent variables that fall slightly short of significance in the model include geographic distance (to which FDI is positively related), CzechInvest offices abroad (negatively related), and the CPI, which suggests that firms from countries with more transparent corporate cultures are more likely to do business in the Czech Republic. The unexpected valence relationship of FDI to geographic distance is understandable because the top FDI origins include the United States (#6), Japan (#11), Canada (#18), China (#27), and Mexico (#28)--all of which are located far from Central Europe. Reasons underlying the negative valence sign of government promotion in the first model are unclear and support the need to examine the question more thoroughly in subsequent specifications. However, especially given the fact that no CzechInvest offices are located within the borders of three leading origin countries--Netherlands, Austria, and Spain--it is plausible that the agency's principle strategy is to make investors in non-leading countries aware of opportunities in the Czech Republic. Certainly, this explanation is plausible when looking at Japan and China (including Hong Kong), both of which host CzechInvest offices but lie outside the list of the top 10 investors. With regard to transparency, it is worth noting that the Czech Republic also scores in the CPI's mid-range (4.3 out of possible 10), tied with Greece and Slovakia at forty-seventh out of 158. This mid-range score and arbitrary result suggests that foreign firms do not necessarily consider this detail relevant.

While the goal of Model 1 was to be inclusive of a wide range of variables found in the literature, it makes little sense to dwell on those that performed poorly in context of the Czech Republic. These variables are worth mentioning but not inflating. While market size (GDP) and strength (GNI), do predict FDI flow as expected, they are not statistically significant. Reflecting on the data set, FDI from countries with the world's largest markets (ie China and India) remains dwarfed by that from smaller (mainly European) markets. Similarly, a substantial amount of FDI from countries with relatively low GNIs (Slovakia, Poland, Russia, and Mexico) confounds the statistical significance for firms from high-income countries to invest (GNI leaders like Luxembourg, Norway, Switzerland, USA, and Japan, and much of Western Europe). As is common in OLS regression with many variables, the collinearity statistics for Model 1 leave room for improvement, with agglomeration having the highest VIF among the variables at 4.218. This magnitude of collinearity is interpreted as being unacceptably high, but such is to be expected with an inclusive initial model in this exercise.

A tendency exists, albeit not statistically significant (p=0.196), for cultural proximity to facilitate FDI in the Czech Republic, in harmony with Zhao and Zhu's (2000) insights from the Chinese context and Bandelj's (2002) examination of CEE. However, in opposition to Bandelj's findings on bilateral relations with EU members, actual EU membership since 2004 now seems to play a negative role, corroborating earlier observations by Bevin and Estrin (2004) that most importantly firms from non-EU countries see FDI as a strategy to gain a foothold in the EU. Annual change in the real exchange rate (REER) and the number of research and development personnel per 1000 (TECH) show no evidence of explaining the origins of FDI in the Czech context. This final inconclusive observation with regard to these variables reflects widely varying conditions in countries from which firms originate.

The subsequent models in this exercise are intended to clear up uncertainties about the results of Model 1. Model 2 is specified by the SPSS forward selection algorithm as the optimal explanation of FDI through 2005 and yields the variables agglomeration (AGG), trade (TRADE), and transparency (CPI) as the critical origin effects with an [R.sup.2] of 0.839 (Table 4), and a high level of significance: agglomeration (p=0.000), trade (p=0.000), and corruption perceptions index (p = 0.033). The results of collinearity tests in Model 2 were satisfactory, with the minimum tolerance of 0.398 and maximum VIF of 2.512 (both AGG), meaning that the standard error for the coefficient of AGG is 1.585 times as large as it would be if were uncorrelated with the other independent variables. In light of these results, this variable warrants specific discussion in the context of the Czech Republic.

Following Head et al. (1995), agglomeration is the realization of positive externalities associated with the proximate location of related firms, in this case related by a common national origin. Plausibly, this factor is particularly important for firms that come from countries that are culturally dissimilar to the Czech Republic (such as Japan and Korea), and for this reason tend to follow--the leader, either to provide ancillary services such as banking or insurance or in an attempt to emulate the success of competitors. Reflection upon this result and inspection of the data set suggest that agglomeration may be a key variable for a small number of large firms (high investment value) from Asia and North America, and not as important for firms from European origins, which already share substantial cultural proximity with the Czech Republic and therefore benefit from existing 'local' knowledge (O'hUallachain and Reid, 1992).

The trade variable again performs well in Model 2 as a positive and significant predictor of FDI, in harmony with findings by Brenton et al. (1998) and Hunya (2000) elsewhere in CEE. As a form of spatial interaction, FDI is itself not unlike trade, which Frankel (1997) recently confirms to remain governed by gravity rules. Both forms of interaction are directly related to the size of two objects--for example, population size--and inversely related to the (geographic) distance between them. What is interesting about the intuitive commonality of these flows is that FDI in this context is not governed by hypothesized similar predictors as originally postulated by Linnemann (1966). This is probably because of overarching intense linkages between the Czech Republic and major distant global traders such as the United States, Japan, and Korea. In fact, recognizing the new potential of FDI from Asia, the CzechInvest web site has recently expanded its language menu, adding Japanese and Korean to the four European languages (CzechInvest, 2009).

Finally, because the Czech Republic is not listed as a particularly transparent country (4.3 on a scale of 1-10), it is difficult to speculate upon the reason behind the tendency for firms from more transparent countries to invest there. The findings refute the expectation that corruption at home, as a form of home country risk, serves as a catalyst for investing in the Czech Republic, as was found in the more transparent environment of the USA (Grosse and Trevino, 1996). Because the Czech Republic is listed as only moderately transparent, the ambiguity of this variable suggests that other factors are at work in drawing FDI.

While Models 1 and 2 confirm the importance of agglomeration and trade origin effects that are widely recognized from research on other contexts, they offer little by way of new insights in the Czech context. For this reason, it is worthwhile to re-run the model specifying only the variables of greatest interest in this inquiry. Therefore, agglomeration, EU membership, and CzechInvest are entered into Model 3. Agglomeration is retained as it has been by far the best performing variable so far. EU membership is of interest because the Czech Republic joined the organization during the time period under investigation, and because expectations of integration long preceded the actual event on 1 May 2004 (Bandelj, 2002; Bevin and Estrin, 2004). The presence of CzechInvest offices abroad, of course, is the variable of main interest in this study. Although only agglomeration is statistically significant, together these three variables provide a very strong model with an [R.sup.2] of 0.825, slightly lower than that of Model 2. It is worth noting that very little explanatory power was lost by simplifying the models from 11 to three variables.

Given the strong performance of Model 3, Model 4 features the same specification for the dependent variable of FDI in 2006, (6) but with agglomeration defined as cumulative FDI through the end of 2005, rather than only 1999. The model performs quite well ([R.sup.2] = 0.479, p = 0.000), with all three expected determinants showing significance. As oligopolistic reaction, EU membership, and CzechInvest have all been thoroughly explained in preceding pages, there is no need to reiterate the interpretations.

It is, however, worth underscoring the remarkable consistency among determinants over time. The data for 2006 clearly show that EU membership (negative) and CzechInvest promotion abroad (positive) have grown in importance in 2006 vis-a-vis previously, as both the coefficients of these variables and their significance levels have increased using the new dependent variable; EU's p = 0.078; CZI's p = 0.021, and the collinearity statistics show that Model 4 is the most in-check of all five models with a peak VIF of 1.795.

While Models 3 and 4 are specified based upon intuitive rationale, Model 5 is run in order to observe the determinants that are statistically most parsimonious. SPSS's algorithm selects in a forward direction the most important independent variables for explaining the 2006 FDI inflow, and accordingly its [R.sup.2] is 0.586, somewhat higher than Model 4. Consistent with Models 1-4, the most important explanatory variable is agglomeration.

Agglomeration is the only common explanatory variable among the five models, and in each case it is significant at the 0.001 level. What is new in the present study is that agglomeration is tested as an origin-effect, in contrast to its use as a location determinant elsewhere (Knickerbocker, 1973; Head et al., 1999). Aside from the role of ancillary services linking firms from the same origin, a company's demonstration of success likely attracts the attention of other decision-makers at home, enticing them to follow suit. This behavior has been documented by Head et al. (1999), among others, and has also been anecdotally observed in the case of Japanese auto manufacturers in the Czech Republic (Frelich, 2006).

It is important to note that cultural proximity (CULT) also appears in Model 5, and that TRADE does not. Explained elsewhere as the desire by firms to avoid problems related to cultural uncertainty (Grosse and Trevino, 1996; Zhao and Zhu, 2000), it can also been seen as a form of local knowledge (O' hUallachain and Reid, 1992). Why would it appear first as a significant variable in 2006? A glance at the Czech National Bank's investment figures yields the observation that many Asian firms first arrived in the Czech Republic only in 2006. Japan and South Korea each accounted for more than $118 million in FDI, led by Hyundai, IPS Alpha, and Hitachi. A conspicuous absence of TRADE in Model 5 reveals a changing relationship with FDI that appears only in the unprecedented 2006 inflow of FDI. This observation merits further discussion in the forthcoming section on 'Implications and Conclusions'.

The extended process of the Czech accession to the European Union in May of 2004 has been instrumental in facilitating inward investment (Bevin and Estrin, 2004). In addition, it is shown here that home country EU membership plays little role in facilitating FDI, but also that FDI from non-EU countries including European countries such as Switzerland (#8) and Norway (#20) continues to thrive. This hints that the Czech Republic may be seen from outside the EU as a favorable export platform to the EU. Because this finding goes against Deichmann's (2004) observations in neighboring Poland, another interpretation is that the nature of inflows has changed over time and EU-based firms are expanding farther eastward. Indeed, Czech National Bank (2007) data confirm that hundreds of millions of dollars worth of FDI from EU members, Italy, Spain, Belgium, Sweden, Denmark, and Portugal, were withdrawn from the country.

The third major finding, based upon three of the five models, is that CzechInvest's foreign promotion indeed represents an effective government policy tool, lending credence to Benacek's (2008) assertion of the agency's competence. While Mallya et al. (2004) attempted elsewhere to interrogate the expense of government policy vis-a-vis its benefits, it remains impossible to speculate on how much investment would have taken place without CzechInvest offices. Moreover, one can only estimate the intangible value for investing firms to have Czech investment support available near their home country headquarters.

The strong support found here for the three dominant variables should not entirely discount other variables. Among the conspicuously weak variables from the perspective of economic geography is geographic distance, which grosse and Trevino (1996) and Deichmann (2004) found to be statistically significant in earlier work. The insignificance of the variable in this case plausibly suggests that in the midst of ever-increasing transportation and communications sophistication, distance is becoming less of a barrier to economic interaction. Given the clear importance of the variables discussed above, it is conceivable that this traditionally prominent variable--one of two considerations in Linnemann's (1966) traditional gravity model--could become displaced by other considerations or the notion of distance, especially when reflecting upon the substantial level of investment from smaller neighboring countries as well as distant places such as Canada with economies relatively smaller than traditional gravity-model champions such as the USA and Japan.

IMPLICATIONS, LIMITATIONS, AND CONCLUSIONS

The empirical evidence presented here for the role of agglomeration, trade flows, government policy, and the European Union generates several suggestions for policy and further research. These findings are particularly noteworthy in the context of the Czech Republic because of its unique situation as a new EU member with a prominent investment-promotion machine.

Although agglomeration strategies are well documented as a determinant of location choice (Head et al., 1995, 1999), it is puzzling that oligopolistic reaction has for the most part escaped scholarly attention as an origin effect, or a strategy of follow-the-leader (from one's own country). Moreover, oligopolistic reaction may become more common as the complexity of international production and distribution chains continues to increase. Further research should at least take notice of a tendency for firms from a given origin to 'travel in packs'. Moreover, it would be most beneficial if this complex question is approached through qualitative analysis, by asking direct questions of decision-makers regarding inter-firm linkages and reporting responses that might either support or refute the present findings.

Trade flows are found to be a significant predictor of FDI through 2005, but not so in 2006. As FDI patterns develop over time, perhaps they do so at the expense of trade flows, corroborating Vernon's (1966) explanation of FDI as a more intimate level of internalization that supplants arms-length trade. Evidence from an unprecedented inflow of FDI in 2006 therefore supports Vernon's suggestion that the two modes of serving an international market can be competitive and redundant. As only modest evidence is offered here, this issue remains to be settled through further inquiry in a broader context.

The European Union is the global frontrunner among economic integration efforts. Its customs union, shared currency among an ever-increasing number of member states, and near elimination of travel and trade restrictions were expected to facilitate cross-border FDI. Kinoshita and Campos (2003) and Eckert and Rossmeissl (2005) identified membership prospects as an advantage among transition states in attracting FDI, and the present data indicate that non-EU firms are more interested than EU firms in investing in the Czech Republic, plausibly as a means to access the EU. Brenton et al's early (1998) expectation that accession plans would be unlikely to generate a surge appear to have underestimated the role of integration. With reference to the transient nature of firms and change in FDI patterns over time, Eckert and Rossmeissl (2005) argue that FDI might yet move eastward, into (new EU or non-EU) states with lower production costs. As time progresses, further research examining Bulgaria, Romania, and other places may shed light on this issue.

Empirical evidence for the variable CZI in 2006 suggests that Czech policymakers should be encouraged with the results of Phase I of the investment process: reaching out to transnational firms in their home countries. Not only does CzechInvest's presence abroad provide exposure to investment opportunities, it demonstrates a commitment to support investors both in their home countries and in the Czech Republic (Phases II and III), providing a compelling advantage for firms from those seven countries. However, following Mallya et al. (2004) and Drahokoupil (2008), policymakers should also reflect upon the expense of this presence abroad, yet another direction for future inquiry.

Implications for government policy can also be extended beyond the Czech Republic. This empirical evidence supports Benacek's (2008) portrayal of CzechInvest as a model of public administration, so additional credence is provided for his suggestion of general rules to emulate this model, including the importance of policy continuity at the national level and the need for such agencies to behave like a private consultancy without charging their clients, rather than as a typical government bureaucracy.

The limitations of the approach adopted here should also be acknowledged. While many important enabling factors of home countries have been identified, these findings cannot be generalized to the cases of individual firms. Especially with regard to motivations of individual decision-makers, a full understanding calls for parallel inquiry at the firm level. For this reason, the case study approach would enhance the depth of understanding presented here. Neither inquiry restricted to the macroeconomic level nor that at the individual level is alone sufficient; therefore more work remains to be done at the corporate level with reference to firm characteristics and local conditions. Moreover, to inspire substantiated suggestions for firms engaging in FDI, it would be helpful to reflect upon the performance of firms that have made decisions to invest, especially in the context of the current global economic crisis.

In the interest of guiding further research, two additional trends that were observed in the data warrant mentioning. The first is the enormous and sustained outward flow of FDI from countries that were until recently themselves to be considered 'transition countries' and researched only as recipients of FDI. The second is an enormous withdrawal of FDI by Western European and North American firms from some leading transition countries (Wilson, 2009), presumably in many cases to be relocated farther eastward as anticipated by Eckert and Rossmeissl (2005), but also because of corporate hardship during global economic downturn of the new millennium's first decade. It has been noted in the present data set that disinvestment in the hundreds of millions of dollars has taken place by firms from Italy, Spain, Sweden, Denmark, Portugal, Taiwan, and Australia, all of which have been large past suppliers of FDI to CEE. In addition to measuring these trends, scholars should consider whether they can be linked to conditions in the home countries, in the Czech Republic, perceived better opportunities elsewhere, or some combination of these. Finally, it will be worthwhile to examine the impact on investment flows to the Czech Republic as data emerge from the present period of global recession.

In conclusion, this research generates several models for assessing the importance of specific origin effects and their ability to explain the home countries of firms conducting FDI in the Czech Republic. Supporting Knickerbocker's (1973) oligopolistic reaction theory, agglomeration is found to be the most important variable across all models, indicating that investors engage in 'follow-the-leader' behavior in an attempt to mimic successful ventures from their own countries, as well as extending existing inter-firm trade relationships to the Czech Republic. Evidence is also presented that economic integration facilitates FDI from outside the EU by firms seeking a foothold inside this growing market of 500 million, the world's largest supranational bloc of accelerating human, capital, and material mobility. The complementarity of trade and FDI is confirmed through 2005, following Brenton et al. (1998), while evidence from 2006 adds credence in the Czech context to Vernon's (1966) theory that local production through FDI can take the place of trade. Finally, this research demonstrates the importance of maintaining CzechInvest office as tools for FDI promotion and support, as well as reinforcing crucial role of cultural proximity (Bandelj, 2002) between home countries and the Czech Republic.

In sum, this paper begins to fill a gap in the literature on origin effects that enable and facilitate FDI, responding to supplications by Hunya (2000, p. 87) to address this 'neglected issue'. The findings provide insights into the unique Czech context during and following its accession to the European Union and in response to its investment promotion efforts abroad. The evidence does not categorically refute mainstream origin effects, but rather examines them in a new context, thereby adding to them the value of insights from the unique Czech case. The paper also calls for further examination of under-researched determinants, in particular government investment promotion and economic integration. As the European Union expands eastward, the 12 new additions from 2004 and 2007 are integrating and changing rapidly, offering rich laboratories for further research on FDI as a facet of globalization. Subsequent projects should therefore reflect upon this changing context, interrogate the findings presented here, and expand our collective understanding of origin effects.

APPENDIX See Table A1.

Table A1: Countries in the data set

Afghanistan
Albania
Algeria
Andorra
Angola
Antigua and Barbuda
Argentina
Armenia
Australia
Austria
Azerbaijan
Bahamas
Bahrain
Bangladesh
Barbados
Belarus
Belgium
Belize
Benin
Bhutan
Bolivia
Bosnia
Botswana
Brazil
Brunei
Bulgaria
Burkina Faso
Burma (Myanmar)
Burundi
Cambodia
Cameroon
Canada
Cape Verde
Central African Rep
Chad
Chile
China
Colombia
Comoros
Congo (Brazzaville)
Congo (Kinshasa)
Costa Rica
Cote d'Ivoire
Croatia
Cuba
Cyprus
Denmark
Djibouti
Dominica
Dominican Republic
East Timor
Ecuador
Egypt
Et Salvador
Equatorial Guinea
Eritrea
Estonia
Ethiopia
Fiji
Finland
France
Gabon
Gambia
Georgia
Germany
Ghana
Greece
Grenada
Guatemala
Guinea-Bissau
Guinea
Guyana
Haiti
Honduras
Hungary
Iceland
India
Indonesia
Iran
Iraq
Ireland
Israel
Italy
Jamaica
Japan
Jordan
Kazakhstan
Kenya
Kiribati
Korea (North)
Korea (South)
Kuwait
Kyrgyzstan
Laos
Latvia
Lebanon
Lesotho
Liberia
Libya
Liechtenstein
Lithuania
Luxembourg
Macedonia
Madagascar
Malawi
Malaysia
Maldives
Mali
Malta
Marshall Islands
Mauritania
Mauritius
Mexico
Micronesia
Moldova
Monaco
Mongolia
Morocco
Mozambique
Namibia
Nauru
Nepal
Netherlands
New Zealand
Nicaragua
Niger
Nigeria
Norway
Oman
Pakistan
Palau
Panama
Papua New Guinea
Paraguay
Peru
Philippines
Poland
Portugal
Puerto Rico
Qatar
Romania
Russian Federation
Rwanda
Saint Kitts and Nevis
Saint Lucia
Saint Vincent
Samoa
San Marino
Saudi Arabia
Senegal
Serbia/Montenegro
Seychelles
Sierra Leone
Singapore
Slovakia
Slovenia
Solomon Islands
Somalia
South Africa
Spain
Sri Lanka
Sudan
Suriname
Swaziland
Sweden
Switzerland
Syria
Taiwan
Tajikistan
Tanzania
Thailand
Togo
Tonga
Trinidad and Tobago
Tunisia
Turkey
Turkmenistan
Tuvalu
Uganda
Ukraine
United Arab Emirates
United Kingdom
United States
Uruguay
Uzbekistan
Venezuela
Vietnam
Yemen
Zambia
Zimbabwe

Acknowledgements

The author thanks the editor and three anonymous reviewers for their helpful suggestions, and Bayar Tumennasan for his methodological assistance.

REFERENCES

Ass, N and Beck, M. 2005: Foreign direct investment in post-Soviet and Eastern European transition economies. The Business Review 3(2): 13-20.

Bandelj, N. 2002: Embedded economies: Social relations as determinants of foreign direct investment in Central and Eastern Europe. Social Forces 81 (2): 411-444.

Benacek, V. 2008: Management of the policies of development in the Czech Republic: Czechlnvest as an example of public administration competence. Proceedings of the 16th NISP Acee Annual Conference. Bratislava, Slovakia, http://wwwl.ceses.cuni.cz/benacek/Benacek-NISPA_Bratislava.pdf, accessed 10 November 2009.

Berglof, E (Chief Economist, Editor). 2008: Transition report: Reforms, growth, and the macroeconomy. London: European Bank for Reconstruction and Development.

Bevin, AA and Estrin, S. 2004: The determinants of foreign direct investment into European transition economies. Journal of Comparative Economics 32: 775-787.

Beyer, J. 2002: 'Please invest in our country'--how successful were the tax incentives for foreign investment in transition countries? Communist and Post-Communist Studies 35(2): 191-211.

Bitzenis, A. 2004: Explanatory variables for low western investment interest in Bulgaria. Eastern European Economics 42(6): 5-38.

Blonigen, B. 2005: A review of the empirical literature on FDI determinants. Atlantic Economic Journal 33(December): 383-403.

Brada, JC, Kutan, AM and Yigit, TM. 2006: The effects of transition and political instability on foreign direct investment flows: Central Europe and the Balkans. Economics of Transition 14(4): 649-680.

Brenton, P, Di Munro, F and Lucke, M. 1998: Economic integration and FDI: An empirical analysis of foreign investment in the EU and in Central and Eastern Europe. Kiel Working Paper #890. Available at http://papers.ssrn.com/downloaded 12 October 2006.

Breznitz, D. 2007: Innovation and the state: Political choice and strategies for growth in Israel. New Haven and London: Yale University Press.

BusinessInfo.cz. 2006: (Czech Business Web Portal). Bank of Tokyo-Mitsubishi UFJ opens a branch in Prague, http://businessinfo.cz/en/articles/czech-republic-business-news, last accessed 17 October 2006.

Capik, P. 2007: Organising FDI promotion in Central-Eastern European regions. Place Branding and Public Diplomacy 3(2): 152-163.

Cuervo-Cazurra, A. 2008: Better the devil you don't know: Types of corruption and FDI in transition economies. Journal of International Management 14: 12-27.

Czech National Bank. 2007: Balance of payment statistics--Foreign Direct Investment. Available at http://www.cnb.cz/en/statistics/bop_stat/fdi/, last accessed 12 December 2009.

CzechInvest. 2009: CzechInvest homepage, http://www.czechinvest.org/en, last accessed 10 November 2009.

Deichmann, JI. 2004: The origins of foreign direct investment in Poland, 1989-2001. Journal of Business and Economic Studies 10(1): 12-28.

Dikova, D and van Witteloostuijn, A. 2007: Foreign direct investment mode choice: Entry and establishment modes in transition economies. Journal of International Business Studies 38: 1013-1033.

Drahokoupil, J. 2008: The investment-promotion machines: The politics of foreign direct investment in Central and Eastern Europe. Europe-Asia Studies 60(2): 197-225.

Dunning, J. 1980: Toward an eclectic theory of international production: Some empirical tests. Journal of International Business Studies 11: 9-31.

Dunning, J. 2001: The eclectic (OLI) paradigm of international production: Past, present, and future. International Journal of the Economics of Business 8(2): 173-190.

Eckert, S and Rossmeissl, E 2005: Consequences of convergence--Western firms' FDI activities in Central and Eastern Europe at the dawning of EU-enlargement. Journal for East European Management Studies 10(1): 55-74.

Frankel, J. 1997: Regional trading blocs in the world economic system. Washington DC: Institute for International Economics.

Frelich, B. 2006: Personal interviews with CzechInvest's Chicago director. Chicago, IL, and telephone discussions, March and April.

Froot, K and Stein, J. 1991: Exchange rates and foreign direct investment: An imperfect capital markets approach. Quarterly Journal of Economics 106(4): 1191-1218.

Grosse, R and Trevino, L. 1996: Foreign direct investment in the United States: An analysis by country of origin. Journal of International Business Studies 27(1): 139-155.

Grosse, R and Trevino, L. 2005: New institutional economics and FDI location in Central and Eastern Europe. Management International Review 45(2): 123-145.

Head, CK, Ries, JC and Swenson, DL. 1999: Attracting foreign manufacturing: Investment promotion and agglomeration. Regional Science and Urban Economics 29: 197-218.

Head, K, Ries, J and Swenson, D. 1995: Agglomeration benefits and location choice: Evidence from Japanese manufacturing investment in the United States. Journal of International Economics 38(3-4): 223-247.

Hunya, G. 2000: Home country patterns of foreign direct investment in Central and Eastern European countries. Russian and East European Finance and Trade 36(2): 87-104.

Janicki, HP and Wunnava, PV. 2004: Determinants of foreign direct investment: Empirical evidence from EU accession candidates. Applied Economics 36: 505-509.

Johanson, J and Wiedersheim-Paul, R 1975: The internalization of the firm--Four Swedish cases. Journal of Management Studies 12 (3): 305-322.

Kinoshita, Yand Campos, NF. 2003: Why does FDI go where it does? New evidence from the transition economies. William Davidson Institute Working Paper Number 573, June.

Knickerbocker, F. 1973: Oligopolistic reaction and multinational enterprise. Boston: Harvard University.

Linnemann, H. 1966: An econometric study of international trade flows. North-Holland: Amsterdam.

Luo, Y. 1998: Strategic traits of foreign direct investment in China: A country of origin perspective. Management International Review 38(2): 109-132.

Mallya, TJS, Kukulka, Z and Jensen, C. 2004: Are incentives a good investment for the host country? An empirical evaluation of the Czech National Incentive Scheme. Transnational Corporations 13(1): 110-148.

O'hUallachain, B and Reid, N. 1992: Source country differences in the spatial distribution of foreign direct investment in the United States. The Professional Geographer 44: 272-285.

Pavlinek, P. 2004: Regional development implications of foreign direct investment in Central Europe. European Urban and Regional Studies 11(1): 47-70.

Transparency International. 2007: Corruption perceptions index, http://www.transparency.org/policy_research/surveys_indices/cpi, accessed 12 August 2007.

UNCTAD. 2007: World investment report 2007 New York and Geneva: United Nations. Also available at http://www.unctad.org/en/docs/wir2008_en.pdf, accessed 15 December 2008.

UNCTAD. 2008: FDI inward stock, by host region and economy, 1980-2006, http://stats.unctad.org/FDI/, accessed 15 December.

Vernon, R. 1966: International investment and international trade in the product cycle. Quarterly Journal of Economics 81: 190-207.

WAIPA. 2008: World association of investment promotion agencies. Members List. Available at http://www.waipa.org/members.htm, accessed 15 December.

Wilson, A. 2009: Czech republic sees FDI drop-off. Eastern European Times. Available at www.eetinres.eu/eastern_Europe, accessed 22 July.

World Bank. 2008: Key development data and statistics, http://web.worldbank.org/, data obtained 20 July.

Zhao, H and Zhu, G. 2000: Location factors and country-of-origin differences: An empirical analysis of FDI in China. Multinational Business Review 8(1): 60-73.

(1) UNCTAD defines FDI as direct investment made to acquire lasting interest of at least 10% ownership in enterprises operating outside of the economy of the investor, and reports annual and cumulative national inflows at http://stats.unctad.org/FDI/.

(2) Poland's PAIiIZ, Hungary's ITDH, Russia's NAPI, for example; a complete list is available at http://www.waipa.org/members.htm.

(3) See, for example http://www.tax-consultants-international.com/.

(4) For example, Spain is the leading origin for 2005 only because of Telefonica's $4.8 billion investment. All other Spanish investment totals about $500 million.

(5) A collinearity matrix has been generated as is available from the author. In the interest of space, it has been omitted from this paper.

(6) The correlation between the two dependent variables (logFDI0005 and logFDI06) is 0.720, significant at the 0.01 level.

JOEL I. DEICHMANN

Global Studies Department, Bentley University, 175 Forest Street, Waltham, Massachusetts 02452, USA. E-mail: jdeichmann@bentley.edu

Table 1: Leading cumulative origins of
FDI in the Czech Republic through 2005

Rank   Country           Value ($mil)

1      Germany              14,099.64
2      Netherlands          10,583.38
3      Spain                  5100.23
4      Austria                4473.49
5      France                 3759.42
6      United States          2459.52
7      United Kingdom         2232.54
8      Switzerland            1598.99
9      Belgium                1491.94
10     Slovakia               1089.89
11     Japan                   981.57
12     Sweden                  926.68
13     Poland                  715.84
14     Luxembourg              671.06
15     Italy                   529.98
16     Denmark                 511.20
17     Cyprus                  391.48
18     Canada                  318.91
19     Liechtenstein           166.06
20     Norway                  140.44
21     Hungary                 135.06
22     Russia                  128.02
23     Ireland                 122.99
24     Finland                  81.37
25     Malta                    75.93
26     Portugal                 69.29
27     China                    51.17
28     Mexico                   48.94
29     Seychelles               45.29
30     Slovenia                 37.88

Source: Czech National Bank (2007)

Table 2: Summary of empirical research on origin effects

Authors (Year)                  Context    Variables

Froot and Stein (1991)          USA        Exchange Rates (-)

O'hUallach6in and Reid (1992)   USA        Geographic distance Local
                                           knowledge (+)

Grosse and Trevino (1996)       USA        Imports (-), exports (+),
                                           home GDP (+), geographic
                                           distance
cultural distance
                                           (-), home risk (+),
                                           exchange rates (-)

Brenton et al. (1998)           CEE        Geographic distance (-),
                                           population (+), trade (+)

Head et at. (1999)              USA        Agglomeration by origin
                                           (+), offices abroad (+)

Zhao and Zhu (2000)             China      Cultural similarities (+)

Hunya (2000)                    CEE        Geographic distance (-),
                                           population (+)

Bandelj (2002)                  CEE        Cultural relations between
                                           investors and hosts (+),
                                           including trade (+)

Deichmann (2004)                Poland     GNP (+), trade (+), EU
                                           membership (+), migration
                                           (+), distance (-)

Bitzenis (2004)                 Bulgaria   Cultural similarities (+),
                                           geographic distance (-),
                                           historical links (+)

Table 3: Explanation of variables

Variable    Definition (units)                    Valence source

FD70005     FDI value 2000-2005 inclusive         Czech National
            (thousand US$)                        Bank (2007)

FD106       FDI inflow during the year 2006       Czech National
            (thousand US$)                        Bank (2007)

GDP         Gross domestic product, mean          + World Bank (2008)
            2000-2004 (million US$)

GNI         Gross National Income per capita      + World Bank (2008)
            (mean 2000-2004)

AGG         Cumulative Value of FDI at the        + Czech National
            end of 1999                           Bank (2007)

REER        Average annual change in exchange     + World Bank (2008)
            rate index, 2000-2004

TRADE       Imports+Exports with partner i,       + Czech National
            mean 2000-2004 (in Kc) (a)            Bank (2007)

EU          Dummy for 15 European Union           + www.europa.eu
            members in 2004 (1, 0)

CULT        Cultural proximity between origin     + Deichmann (2004)
            i and destination j (most=5,
            least=1) (b)

TECH        Researchers in science and            + World Bank (2007)
            engineering (per million)

DIST        Distance between leading cities       - http://
            of country i (as defined by           www.mapcrow.info/
            Frankel, 1997) and Prague (in km)

CZI         CzechInvest offices in origin         + CzechInvest (2009)
            country i (number: 0, 1, 2)

CPI         Corruption Perceptions Index          - Transparency
            (high value=low corruption)           International
                                                  (2007)

(a) Initially, the modest were run with imports and exports as
separate variables, in part to investigate Frankel's (1997) finding
that in the USA they behave differently as origin effects. However,
in the Czech context this resulted in collinearity, so a new variable
TRADE' (imports + exports) was calculated. The simple correlation
between imports and exports in the data set was 0.981, and that
between their log transformations was 0.988, both significant at the
0.01 level.

(b) Home countries are assigned scores of 1-5 with reference to
cultural similarities to the Czech Republic, using the proxy of
language and alphabet. '5' is for countries using Slavonic languages
and Latin alphabet lie Poland); '4' is for Slavonic languages and
Cyrillic alphabet (ie, Russia), or Germanic languages and Latin
alphabet (Austria, USA--noting that most Czechs speak English); '3'
is for Romance languages (France); '2' is for all others using the
Latin alphabet as Czechs do (Albania); and '1' is for minimal
similarities (China, Japan).

Table 4: Coefficients and significance
levels of variables in the models

                         Model 1        Model 2        Model 3
                       coefficients   coefficients   coefficients

Independent Variable   LogFDI0005     LogFDI0005     logFDI0005
Constant               -1.046         -0.222          0.645 (a)
logGDP                  0.045
GNI                     0.00001266
AGG (99/05)             0.941 (a)      0.896 (a)      0.943 (a)
REER                    0.000
logTRADE                0.163 (c)      0.175 (a)
EU                     -0.845                        -0.023
CU LT                   0.245
TECH                   -0.000068
DIST                    0.000071
CZI                    -1.214                        -0.042
CPI                     0.177          0.204 (a)
[R.sup.2]               0.847 (a)      0.839 (a)      0.825 (a)

                         Model 4        Model 5
                       coefficients   coefficients

Independent Variable   LogFDI06       logFDI06
Constant                0.577 (b)     -0.856
logGDP
GNI
AGG (99/05)             0.555 (a)      0.426 (a)
REER
logTRADE
EU                     -1.976 (b)     -1.096
CU LT                                  1.081 (a)
TECH
DIST
CZI                     2.437 (b)      2.662 (b)
CPI
[R.sup.2]               0.479 (a)      0.528 (a)

(a) Statistically significant at the 0.001 level (two-tailed).

(b) Statistically significant at the 0.05 level (two-tailed).

(c) Statistically significant at the 0.005 level (two-tailed).

Model 1: All variables entered (through 2005).

Model 2: Stepwise (optimal model, through 2005).

Model 3: Enter (variables of greatest interest, through 2005).

Model 4: Enter (variables of greatest interest, 2006).

Model 5: Stepwise (optimal model for 2006).

Note: AGG is defined as cumulative FDI value through 1999 for
Models 1, 2, and 3 and cumulative value through 2005 for Models 4
and 5