“Polish investments at home and abroad: Differences
in firm & location characteristics”
Reinier de Jonge
Student number: 1481924
“Polish investments at home and abroad: Differences
in firm & location characteristics”
Reinier de Jonge
University of Groningen, Faculty of Economics and Business
Preface
I started writing this thesis in May 2008, one might be thinking; why the heck does it take such a long time to write a thesis? Well to be honest there were a lot of other ‘businesses’ that I had to care of, literately and figuratively.
Finally it is finished, which will most likely result in the award of a degree in International Business & Management from the University of Groningen which I am very pleased with.
There are a couple people I would like to thank; these are my parents, my sister and my friends. All these people have had their share in completion of this paper.
Secondly I like to thank my supervisor Rudi de Vries because he has taken the time to give comments to my work.
Thank you all.
Executive Summary
This report is the result from a research of Polish firms investing abroad and at home. We have looked at a total of 205 investments within and outside Poland for a total of 5 years (2003-2007). Within this time period Polish manufacturing firms made a total of 48 foreign direct investments commitments. We wanted to know if the firm and location characteristics of these Polish firms investing abroad are different from Polish firms investing at home.
We have seen that both the location and firm characteristics are significantly different from each other. In terms of firm characteristics we found that firm age and profitability are significantly different. Polish firms that invest abroad are younger than Polish firms that invest at home. Previous research suggests otherwise as Lall (1983) and Chen (1983) argues that older firms are more likely to have greater assets, which allow them to go more easily abroad. Additionally we find that Polish investors investing abroad are more profitable than Polish investors investing at home. This is in line with previous research such as Cantwell & Sanna-Fandaccio (1993) who found that profitable firms organize their activities more efficiently and thus create resources that are needed for expansion.
In terms of location advantages we looked at the institutional characteristics and found that both Voice and Accountability (measures perceptions of a country’s citizens to which they are able to participate in selecting their government) and Political Stability were significant. A remarkable finding was that Voice and Accountability had a negative sign indicating that Polish firms investing abroad invest in countries with a lower rating of freedom of expression and free media. This contradicts our expectations but could be explained by other theories such as Johanson & Vahlne (1977) who argue that investors invest in countries with similar cultural proximities. As to Political Stability we find a positive sign indicating that it is more likely that investments flow to countries with stable governments that are not likely to be overthrown. This finding is in line with our expectations and other theories.
Table of Contents
Executive Summary ... 3
1. Introduction ... 7
2. Theoretical framework: What explains foreign Direct Investment? ... 10
2.1 Neo-classical theories... 11
2.2 Resource-advantage theory ... 13
2.3 International trade theory ... 14
2.4 Transaction cost theory ... 14
2.5 Evolutionary theory... 15
2.6 The OLI framework ... 16
2.6.1 Ownership ... 18
2.6.2 Location advantages ... 19
3. Country environment of Poland ... 21
3.1 Poland and Foreign Direct Investment... 21
3.2 Poland in transition... 21
3.3 Poland and inward FDI ... 23
3.4 Poland and outward FDI ... 24
3.5 Development and government indicators... 25
4. Data Description, sample and methodology ... 27
4.1 Operationalisation of Ownership advantages... 27
4.2 Operationalisation of Location advantages ... 30
4.3 Control variables ... 34
4.4 Sample and methodology ... 35
5. Empirical analysis ... 41
6. Conclusion and Future Research... 48
6. Appendices ... 52
Figures and Tables
Figure 1: FDI flows Poland from 1990 to 2007 ... 9 Figure 2: Difference restricted and unrestricted model... 40
Table 1: EFFECT OF OWNERSHIP ADVANTAGES ON COMMITMENT TO FDI ... 30 Table 2: EFFECT OF GOVERNMENT INDICATORS ON INFLOWS OF FDI FOR ANY
GIVEN COUNTRY ... 33 Table 3: POLISH FIRMS THAT ENGAGED IN INVESTMENT ACTIVITY ABROAD
FROM 2003 TO 2007 ... 36 Table 4: ENGAGEMENT IN INVESTMENT ACTIVITY BY HOST COUNTRY AND
YEAR ... 37 Table 5: CHI-SQUARE TEST TO TEST SIGNIFICANT RELATIONSHIPS BETWEEN
HOST REGION AND TIME PERIOD. ... 37 Table 6: CHI-SQUARE TEST TO TEST SIGNIFICANT RELATIONSHIPS BETWEEN
1. Introduction
Foreign Direct Investment is a well-known and widely researched phenomenon. The definition of FDI can be described as follows: an investment involving a long- term relationship and reflecting a lasting interest and control of a resident entity in one economy (“parent enterprise”) in an enterprise resident other than that of the foreign direct investor (“foreign enterprise”) (source: Unctad.org). FDI has seen a strong increase after the Second World War. FDI inflows to developing countries (coming from developed countries) increased rapidly.
Nowadays we see also flows coming from developed nations and flowing to transition economies, an example are the flows to Poland. Where Poland received on average US$3.705 million between 1990 and 2000 it received US$17.580 million in 2008 (UNCTAD world investment report 2008). This phenomenon is described by various authors including Johnson (2006), McMillan (1993), and many others. These authors argue that financial flows from developed countries to less developed countries can be explained by changes in FDI regulations in the transition economies, transitional market demands and because of transition-specific determinants such as privatization. It is obvious that these capital flows were vital for the transition process in Eastern Europe, because developing economies (such as the Netherlands just after the second World War) also needed inflows of capital in order to start building an industry.
Outward FDI
Outward Foreign Direct Investment is defined as: “A domestic firm establishing a facility abroad.” (Deardorff International Economics). Outward FDI can be explained if it takes place from a developed country to a lesser-developed country, current literature however does not tell us a lot about the other way around. Theoretical models are therefore not fully able to explain why there are FDI flows from Poland to Western Europe. There has been a lot of research on FDI that can be classified in a number of theories. The oldest theory dates from 1776 and the newest from the second half of the twentieth century. All these theories have been modified and changed in order to account for changes in the world and economies. One of the leading theorists on outward FDI is Dunning. His works have contributed to how we explain the rationale behind FDI. Some people do not recognize it as a theory but rather an all-encompassing model, or as Dunning puts it himself as an “eclectic” model. The model looks as three important pillars: Ownership, Location and Internalisation advantages. Ownership advantages are a specific set of sills that distinguish a firm from another company and gives the firm a ‘competitve advantage’. Location advantages are things that make a certain country attractive, an example are low wages. The last is internalisation advantages (no to be mistaken with internationalisation). Internalisation advantages are a combination of Location and Ownership advantages and ultimately decide if it is wise to internalise a foreign investment. In reality this means that if a given firm has a specific set of ownership advantages and they wish to set up a new plant in a given country that has a specific set of location advantages they can decide to set up their own plant, buy another existing plant, export or franchise their concept. In the first two instances we speak of internalising where the last two are arm length agreements. If a firm eventually decides to internalise its production plant (e.g. set up a new production plant in a country) they will invest a certain amount of money. This sum of money is also called a foreign direct investment. If you would accumulate all foreign direct investments from all firms in a country you would get the outward foreign direct investment of that country.
Poland
also coupled with the disappearance of import protection (Kalotay, 2005). Poland first attracted a lot of FDI, which started in the 1990’s.
- 5 000 5 000 10 000 15 000 20 000 M il li o n o f U S d o ll a r s 90 92 94 96 98 00 02 04 06 Year
FDI flows Poland from 1990 to 2007
Outflows Inflows Figure 1: FDI flows Poland from 1990 to 2007
http://www.unctad.org
The above figure shows the flows of FDI to and from Poland since 1990. As one can see the inflows (flows to Poland) are indicated in orange and are still increasing, the outflows (capital leaving Poland) are indicated in purple.
Research question
Since 1990 a lot has changed in Poland, this gives us the possibility to research outward FDI in relation to firms that invest at home. We can then see if there are differences in firm specific characteristics and the location of firms that show a FDI commitment. The purpose of this thesis is to find quantitative results that will answer the following research question:
What is the impact of ownership and location advantages in determining the investment decision of Polish firms participating in the Warsaw stock exchange investing abroad or at home?
One can explain Foreign Direct Investment streams in many ways. It is possible that the change in the institutional environment of Poland since 1990 has contributed a lot. Maybe it is the change in strategy of firms that have changed the reason to invest abroad. Or maybe the receiving countries of FDI have become more attractive for Polish firms. In order to give answer to our research question we will utilize a number of theories that can explain our research question.
Set up of this thesis
To give an answer to the main research question we will have an elaborate review of the existing theories on FDI in Chapter 2. This theoretical framework will help us to develop a conceptual model that will identify the variables we need.
The third chapter will discuss the country environment of Poland. We will discuss the changes in Poland since 1990, as these changes have had an impact on the country. This chapter will mainly give us a background of the variables that were important contributors to the change.
The fourth chapter will illuminate the data we used and how we used it. Additionally we will also present the firms in the chapter that were selected and we also present the methods we used to come to our conclusions.
The fifth chapter will describe the results we got from our research, we empirically test our hypotheses in this chapter and we also present the logit regression model we used to test our model. We will also interpret the results that we got from data analysis.
2. Theoretical framework: What explains foreign Direct Investment?
2.1 (Neo)-classical theories
To explain the reasoning behind foreign direct investment (FDI) we have to go back to the explanation of why trade occurs. The first theory that describes why countries trade with each other dates back to 1815 when Robert Torrents concluded that it would be more advantageous for England to trade grain with Poland than to produce grain in England. This conclusion was partly based on the principle of absolute advantage where Adam Smith (1776) (in his book “The wealth of Nations”) showed that a country could benefit from trade if the country has the lowest absolute cost of production in a good (Country X can produce more output per unit than any other country).
Absolute advantage was refined by David Ricardo (1821) in his book “On the principles of political economy and taxation” where he argued that it is not absolute advantage but comparative advantage that explains trade. Comparative advantage explains how trade can benefit all parties (countries, regions, firms, individuals) involved. In a classical example he takes England and Portugal and shows that in Portugal it is possible to produce both wine and cloth with less work than it takes in England. However the relative costs of producing these two goods are different in the two countries. In England it is hard to produce wine, and only reasonably difficult to product cloth. Consequently while it is cheaper to produce both cloth and wine in Portugal than it is in England, it is still cheaper for Portugal to produce excess wine and trade this for English cloth. England would also benefit from this trade as its cost for producing cloth has not changed but it can now get wine at cheaper costs. The conclusion of Ricardo is that a country should specialise in products and services which it has a comparative advantage. The country should trade with other countries that also have a comparative advantage in another product/ service. This way both countries benefit from trade.
computers etc. They argue that a country specialises either in labour intensive production or capital intensive production. For example: a country that produces two goods (for example cloth & wine) an assumption must be made as to which industry has the larger capital- labour ratio. So if wine requires more labour per capital unit than clothing production, we would say that wine is labour-intensive relative to clothing production. The H-O model argues that if a country has abundant labour (many labour intensive industries) relative to capital then the country will specialise in labour intensive industries and also export labour intensive goods. If a country is capital abundant then this country will specialise in capital intensive industries and also export capital intensive goods.
The role of FDI and why it takes place has been explained through the H-O theorem until the 1970’s. There have been very many authors to explain why FDI takes place based on the H-O theorem but a key reference in the Hecksher Ohlin inspired literature on FDI was Robert Mundell (1957). Mundell elaborated on the factor endowments approach through trade flows. His work was in particular focused on sectors rather than firms. His conclusion was that FDI is mainly driven by international differences in rates of return. To our understanding the incapacity in this approach is that in the H-O theorem, the assumption is made that the factors of production (labour and capital) only move within countries. This implies that developing economies and the countries in transition should be the main recipients of FDI and not the developed economies. This for the reason that developing countries have a comparative disadvantage of capital but are labour abundant because of the relative cheap wages. In the traditional H-O model we would not expect to see the United States invest in Europe. As this is the case, the H-O model does not correctly represent the reasoning behind the occurrence of FDI. Furthermore the H-O model cannot explain the outward FDI (FDI from transition economies to developed economies) at all.
2.2 Resource-advantage theory
The resource advantage theory was first introduced by Stephen Hymer (1960). Hymer was the first economist to apply the theory of industrial organization to FDI. He believed that FDI takes place because of four factors; Lower cost in acquiring factors of production, control of special knowledge of the process of distribution, better marketing and distribution facilities and last a differentiated product. He was therefore the first to look specifically at the resources that are apparent within a firm and that should lead to FDI. In principle, in the conclusion of his doctoral thesis he states that it should be firms from the most advanced countries that enjoy these advantages and should exploit them through internationalisation. Unfortunately his empirical chapters do not support his ideas; there is no use of econometric work and his ideas rely heavily on the works of Dunning (1958).
The second important contribution to the resource-advantage theory was brought by Raymond Vernon (1966). He puts less emphasis on the comparative cost doctrine and more upon the timing of innovation, the effects of scale economies and the roles of ignorance and uncertainty in influencing trade patterns. He, just like Hymer, also focused on advanced countries by taking the United States as an example. He argues that U.S. firms will be more likely to participate in FDI activities because; U.S. average income is higher than anywhere in the world, and U.S. is characterised by high unit labour costs which means that production facilities would likely be situated outside the U.S. He continues however that the rationale behind putting up production facilities outside the U.S. should be based on the lowest costs possible. A firm which desires to produce a new product will therefore first produce in the U.S. because of transportation costs and import duties; he calls this the early stages of introduction. After the product matures, a number of other stages are reached until production takes place in another country. All these stages combined are called the “product cycle” (See appendix 1). Vernon describes that there are 4 stages for new products on a macro-economic level:
• Stage 1:U.S. exports products • Stage 2: Foreign production starts
• Stage 3: Foreign production competitive in export markets • Stage 4: Import competition begins.
perspective and are focused on the production of industrial goods in manufacturing sectors and ignore trade in intangibles such as services and brand names.
2.3 International trade theory
After the H-O model seemed to be incapable to explain North-North flows Paul Krugman (1978) came up with a new extension to international trade, which is now called the “new trade theory”. He argued that comparative advantage was not the whole story but that the role of increasing returns was more important. The new trade theory looks at a macro-economic level and thus only compares or explains why countries trade with each other. It does not specifically describes the rational behind FDI, however it is important to mention because new trade theory combines the idea of free trade with something called the network effect. New trade theory argues that for nations it might be effective to shelter infant industries until they have grown to sufficient size to compete nationally. Sheltering would mean that nations impose subsidies or quotas to either support the firm to export or to restrict competition to import. When the industry would have sufficient economies of scale it would dominate the world market. The new trade theory therefore specifically addresses the political conditions of countries that impact the reasoning behind exporting or FDI. The rationale behind outward FDI therefore lies also in the political policies of countries that influence FDI, trade or international production.
2.4 Transaction cost theory
because of government measures (sales taxes, rationing, price controls etc) it is cheaper to keep transactions internal rather than using the market system. The advantage of internalisation therefore increases the size of the firm. Thus internalising additional transactions continues until the cost of making that transaction internally is equal to making that transaction in the market. If the firm’s costs exceed the market’s costs the firm will not internalise the transaction. According to Coase the size of the firm is thus dependent on how the entrepreneur organises the transactions he does and on the costs of using the market mechanism. These two factors combined determine how many products a firm produces and how large a firm is. Transaction costs theory was long regarded of why FDI occurs because if a firm decides to internalise across national borders a firm becomes multinational. Some economists however suggested that internalisation does not adequately describe the reasons of why FDI occurs. They believe that it is not a sufficient explanation as transaction cost theory is unable to give answers to questions such as: Under what circumstances would a firm replace an open market and instead use an internal transaction? Is this because of product quality or unstable supply of raw materials or market knowledge? It seems that there are other reasons besides minimizing the transaction costs that could lead to internalising transactions. 2.5 Evolutionary theory
internationalisation patterns of firms and distinguishes itself by focusing on the individual firm rather than sectors (such as the transaction theory) or countries (such as the trade theory). However it lacks the explanatory power to describe firms that start up a foreign production plant directly or firms that are global in their first 3 years of existence.
2.6 The OLI framework
Albeit the inconstancy in the various theories to explain the rationale behind FDI, one framework emerged that tries to explain FDI by combining theories as discussed above. The framework which was developed by John Dunning (1977) has proved a successful way of thinking about multinational enterprises and has found its way into the applied work of economics and international business. We have to note that the framework in itself is not a theory but rather categorises and combines empirical research. The origin of the framework lies in an empirical investigation of Dunning (1958) where he noticed that the labour productivity in the US manufacturing industry was, on average 2 to 5 times higher than that in the UK industry. He posed two questions: Is the difference in productivity a result of superior indigenous resources of the US (vs UK) or because of the ability of managers of US firms (vs UK firms) to organise these resources in a better way? These two questions resulted in two hypotheses: If the superior productivity was entirely managerially related, US manufacturing affiliates in the UK should perform at least as well as their parent companies, and fair considerably better than their indigenous competitors. This is known as ownership advantages. If US affiliates in the UK recorded no better performances than their UK competitors, and hence, much poorer than that of their parents companies then this would be due to the non-transferable characteristics of the US economy, this is called location advantages. The outcome of his research was that US affiliates were not as productive as their parent companies, but were more productive than their local competitors, Dunning explained that this was partly explainable by location (L) and partly explainable by ownership (O) specific characteristics. In a later paper Dunning (1977) extended the O and L characteristics with internalisation (I) characteristics. The reason to do so was because he argued that one also had to explain why such firms opted to generate add/or exploit their O specific advantages internally, rather than to acquire and/or sell these, or their rights, through the open market, this he identified as the I characteristics.
Enterprises” (page 176). As such he argues that this will be determined by the configuration of three sets of forces:
1. The (net) competitive advantages (O) which firms of one nationality possess over those of another nationality in supplying any particular market or set of markets. 2. The extent to which firms determine to internalise (I) the markets for the generation
and use of these assets
3. The extent to which firms choose to locate (L) these value-adding activities outside their national boundaries.
Furthermore he also mentions a number of restrictions: First that each of these advantages and the configuration between them is likely to be context specific and will vary across industries, regions or countries. Furthermore Dunning argues that the variables necessary to explain import-substituting FDI are likely to be different from the variables that are needed to explain resource-orientated FDI. Thus the rationale behind FDI also depends on the type of investment a company is seeking.
Albeit the criticisms, the OLI frame work is to our understanding one of the most compelling paradigms to describe FDI, as it is solidly grounded in transaction, evolutionary, trade theory and resource advantage theory. We will therefore partly use this framework in order to answer our main research question: What is the impact of ownership and location advantages in determining the investment decision of Polish firms participating in the Warsaw stock exchange investing abroad or at home? We will not use the traditional model that was introduced by Dunning (1977) but the more extensive version of the eclectic paradigm as suggested by Dunning (2001) as a response to the criticism. The most notable difference in the extended version is the inclusion of institutional theory. The extension was proposed by Dunning (2006) himself and has an influence on the choice of the location advantage variables. Dunning (2006) argues that location advantage variables can both be firm level as well as country specific. Country specific variables were further developed by Guisinger (2001), who calls them “geovalent elements”. Country specific variables (and thus location advantage variables) include political risk, government restrictions, culture, and legal systems. As the applicability of the OLI framework is context specific and can vary across firms, regions or even countries, our research will be limited to Poland only. Furthermore we will only use manufacturing firms to account for a specific industry. We will therefore look at Polish manufacturing firms that are expanding abroad through means of FDI. Other means of entrance such as arm length’s agreements, export and licensing will be neglected, as these entrance methods require a different set of variables and theories to explain their behaviour. As explained in the introduction this study compares two groups of firms; the ones that invest abroad and the ones that invest at home. Although the theory of Dunning tries to explain why firms decide to invest abroad rather than export (and thus tries to explain the different entry modes) we use this framework to explain our research question as we believe it incorporates most of the important theories.
2.6.1 Ownership
control of special knowledge of the process of distribution, better marketing and distribution facilities and last a differentiated product. As setting up distribution facilities or production facilities abroad face high costs and these costs are not faced by domestic producers the foreign firm must have a certain set of characteristics that allow the firm to go abroad and invest heavily with increased risks. Additionally Dunning (1988a, 1988b) points out that the specific set of characteristics can stem from the nationality of the firm and Cantwell & Narula (2001) indicate that ownership advantages can be both firm specific and country specific. All three authors therefore agree that ownership advantages are specific for a number of firms in a specific context in a specific country; our research will therefore be focused on Polish firms. As ownership advantages have to compensate for the additional costs associated with setting up and operating abroad that are not faced by domestic producers we have to indicate a number of quantitative variables that allow us to measure the compensation to invest abroad and that can be used as an indication to either engage in FDI or not engage in FDI. There have been a number of studies that go into depth on the variable specification that could explain the ownership part of the OLI framework. Johnson (2006) has tried to summarize the variable specifications of previous studies (See appendix 2).
2.6.2 Location advantages
3. Country environment of Poland
3.1 Poland and Foreign Direct Investment
Albeit the criticism on the OLI framework, we decided to use it because in our opinion it is the most valuable and comprehensive framework to explain Foreign Direct Investment. The various factors that make up the OLI framework are discussed above, as one knows the OLI framework looks at firm specific factors, industry factors, but also institutional factors of the country that a firm is likely to invest in. We believe however that the data that supports the OLI framework leaves out one important factor; institutional factors of the home country. This chapter will describe the institutional environment of Poland as this has been an important contributor for both inward and outward FDI streams. We will use a certain set of variables as a proxy for the change in the institutional environment of Poland. Besides including the institutional environment of Poland, we will also describe the changes in Poland since 1989 as this was an important milestone for Poland. Looking at home country environments was also carried out by a number of other authors. Pradhan (2004) was one of these authors who looked at Indian manufacturing firms and their reason to go abroad. She also included the home environment of India, she argued that the policies of the government changed over the years which allowed firms to go abroad.
3.2 Poland in transition
Poland has also gone through a transition process, which was the result of the fall of the communist regime. Immediately after the transition started, trade liberalization became the first method of reintegrating Poland into the world economy (EBRD, 1999). The result of the collapse of the communist regime was that the trade liberalization was not only radical it was also coupled with the disappearance of import protection (Kalotay, 2005) Poland first attracted a lot of FDI, which started in the 1990’s. In the context of the development of Poland, 1990 was a year of radical, institutional change that was the start of evolutionary adjustments in the Polish economy. The economy of Poland was opening after the end of the communist system and the following factors had the most powerful influence on the opening of the country to the inflow of foreign capital in the form of FDI (Kubielas et al. 1996, p 428):
• Liberalization of legal regulations concerning FDI inflow
Before the privatization or liberalization, Poland was regulated by the enclave model, which treated FDI in a special way compared to the rest of the economy (Samonis 1992). The enclave model had the following characteristics:
• FDI was allowed only in so-called small foreign business operations, or in joint-venture companies with a minor share of foreign capital.
• Foreign investors had to obtain permits through a complicated procedure.
• A wide range of sectors were out of bounds for FDI, either forbidden or rationed. • There was an obligation to resell foreign currency revenues from exports to domestic
banks.
• There were restrictions on the transfer of profits abroad and on the purchase of real estate.
• FDI could benefit from tax holidays on corporate income tax.
After 1990 when the Polish economy went through a transition, certain legal and institutional changes were implemented that had a direct effect on the inflow of foreign capital. For the first time FDI was treated the same way as domestic investments, which meant that FDI contributed significantly to the national treatment of FDI, the most important features included the following:
• There were no restrictions on the transfer abroad of profits and initial capital
• Foreign investors only needed permits that were issued by the government if foreign investors bought equity or leased equity of state-owned firms.
• Corporate income tax holidays were abandoned.
• Full guarantee of compensation in the unlikely case of expropriation
• Ability for foreign firms to start their activities in two forms: limited liability companies and joint-stock companies.
3.3 Poland and inward FDI
The inward FDI became probably the most important engine of successful reintegration into the world economy (Kalotay, 2001). Between 1995 and 2001 the inward FDI in the CEE (Central Eastern Europe) countries increased rapidly from $40 billion to $160 billion (UNCTAD, 2002: 113). After 5 years (in 2000) the CEE countries had almost completely caught up with the rest of the world (Kalotay, 2005). Surprisingly there were a number of CEE countries (including Poland) that had a very high intake of FDI, additionally inward FDI increasingly determined the trade performance of Poland. Foreign investors accounted for 48% of Poland’s export, consequently trade performance was less decided by government actions but rather by corporate decisions. Poland moved from central planning to decentralization where state-owned firms where transformed to privatized firms. It is important to notice that almost 80% of Poland’s GDP was produced by these former nationalized firms. Additionally almost 88% of the employment in Poland came from these nationalized firms and came from non-agricultural sectors (Rondinelli & Yurkiewicz. 1996).
Poland went through an economic stagnation from 1990 to 1992 and the national income fell by 13% while inflation reached a staggering 585% (Rondinelli & Yurkiewicz, 1996). As one can imagine, this had also an influence on the foreign direct investment. FDI inflows still increased from 88 million dollars in 1990 to 678 million US dollars in 1992 (UNCTAD, 2007). To decrease the inflation Mr Balcerowicz (prime minister of Poland at that time) took some drastic measures. He implemented price controls, he devaluated the zloty, increased taxes, decreased trade restrictions, decreased wage increases and lowered government spending. All these drastic measures led to a more stabilized Poland and a stronger Polish currency.
decrease in FDI inflows can be explained by the world economic situation in 2000 / 2001 when the “internet bubble” collapsed. From 2002 to 2007 we can see an increase in FDI going to Poland, from 4.1 billion US dollars in 2002 to 13 billion US dollars in 2004 when Poland joined the European Union. Additionally Poland has received almost one third of all FDI inflows to Central and Eastern Europe if one would accumulate all inflows from 1990 to 1996, which is remarkable given the number of countries in Central and Eastern Europe. Poland follows after the Czech Republic and Hungary in terms of the quantity of outward FDI measured by the total amount of dollars in any given year. This can be explained by the population of Poland which is smaller than Czech Republic and Hungary.
3.4 Poland and outward FDI
We have tried to explain the phenomenon behind inward FDI to Poland and have graphically shown by a bar-chart, this study focuses however on outward FDI. As outward FDI is primarily explained as a micro-economic occurrence as research is focused on a firm level, we also have to look at the macro-economic variables and in this case Poland. As indicated above Poland has gone through a dramatic reform, which has attracted FDI but has also allowed for outward FDI to take place. In order to explain the outward FDI from a macro-economic perspective we have to take the Investment Development Path (IDP) into mind. The IDP concept was (again) introduced by Dunning (1981; 1986) and was refined by Dunning and Narula (1996) to incorporate trade. The basic idea behind the IDP theory is that FDI (both inward and outward) can be explained through the economic situation / economic development of a country. The NET outward FDI of a country can be calculated by subtracting the inward FDI from the outward FDI in any given year. The result of the calculation is called the Net Outward Investment (NOI) and is likely to change throughout the years as it is tied to the economic development of a country. Dunning and Narula argue that the NOI has five (5) stages and is intrinsically related to related to country’s economic development.
Countries in stage 3 of the IDP have an increasing NOI as inward FDI decreases and outward FDI increases which result in an almost positive NOI. In stage 4 of the IDP outward FDI increases and rises more rapidly than the inward FDI. The NOI position of a country becomes for the first time positive and usually fluctuates around the 0 level. The country has transformed itself and country attractive factors are usually because of the assets that were created the years before. This stage (stage 4) and stage 5 is the stage of most developed countries (including the Netherlands). Stage 5 is not much different from stage 4 (as NOI still fluctuates around 0) but distinguishes itself from the increased volume. In stage 5 both the inward and outward FDI increases. According to Dunning (2001) the IDP has a parallel connection to the trade development path (TDP) and has its roots in the traditional trade theory as discussed above (Krugman 1978). An example of a country’s life cycle can be found in appendix 2.
3.5 Development and government indicators
4. Data Description, sample and methodology
To give an answer to our research question we make use of Dunning’s ecletic framework as discussed in chapter 3. This framework consists of three important pillars namely ownership, location and internalisation advantages. If both the location and the ownership advantages are constructive a firm can choose to internalise these advantages by making an investment abroad. This chapter will operationalise these advantages by quantifying variables that constitute the framework of Dunning.
4.1 Operationalisation of Ownership advantages
We will measure ownership advantages by using four variables: firm size, firm age, profitability and leverage. The reason to choose these variables in particular is because of data availability. There are also a number of other variables such as R&D expenditures that can have a significant influence on ownership advantages, however these expenditures are not mentioned in the annual reports of the firms we investigated. We will discuss the variables we chose for this study individually and also explain how they are related to ownership advantages.
Firm size
and Caves (1996) also acknowledge that the larger the firm grows relative to the domestic market, the less profitable it would be to even further increase the share of the domestic market, hence it would be a better choice to go abroad. Pradhan (2008) however noticed that there could also be a threshold effect to firm size, as he argues that firm size may have a positive effect up to a critical level of size, after the firm has grown to a certain size the added value might not be that contributing anymore.
We will measure firm size looking at the total assets of a firm in a given year. These assets are mentioned on the balance sheet of a firm, which can be retrieved from annual reports.
Firm age
A second determinant that could influence the decision to invest abroad might be the age of the firm. The amount of intangible assets, which accumulates over time can be expect to grow with age. Hence older firms are most likely to be more established and have accumulated valuable business and production experience, which younger firms do not have. The firm’s age and the contribution to the relationship of overseas production have been tested in a number of studies in the past. Important contributors are Lall (1983) and Chen (1983) which both found that age has a positive effect on outward investment. Pradhan (2008) also noticed that age (just as firm’s size) might have threshold and could even have a negative effect.
We will measure firm age by subtracting the year under investigation from the year of incorporation of the firm. As our research has a time element and looks at a number of years the age of a single firm changes throughout the years of investigation.
Profitability
found that more profitable firms would like to face less risk when investing abroad and thus might opt to invest in non-equity participations.
We will measure profitability by looking at the profit margin of a firm. We measure the profit margin by dividing the profit (loss) before taxation by operating revenue/turnover. This number is then multiplied by 100 to get a percentage.
As indicated by Cantwell & Narula (2001) internalisation advantages can be both firm
specific and country specific. In this study we will therefore differentiate between the two and assume that Polish firms that are investing abroad are also looking at the institutional
environment of the host country. Leverage
Leverage indicates the degree to which a business is utilizing borrowed Money.
Businesses that are highly leveraged can be limited in the benefits of FDI (Krugman 1998). We try to control for the degree of financial leverage since we expect that a firm’s debt to equity ratio is positively correlated with FDI. This as we believe that a higher leverage may lead to an increased desire for diversifaction, and thus risk reducing investment. One study by Tan and Vertinsky (1996) measured liquidity as cash flow as a percent of total assets but failed to find any type of relationship between cash flow and FDI. One could also argue the other way that firms with a lower financial leverage would be in a better position to undertake FDI, however empirically this has not been proved so we will follow the first logic.
Table 1: EFFECT OF OWNERSHIP ADVANTAGES ON COMMITMENT TO FDI
Given our research question we can now come to our first hypothesis given the theoretical framework we have developed above. Our first hypothesis is therefore:
Hypothesis 1: Polish firms that invest abroad have greater ownership advantages than Polish firms that invest at home.
4.2 Operationalisation of Location advantages
As indicated in chapter three we will use government indicators as a proxy for location advantages. We thus measure the institutional environment of a certain country by using six different dimensions. These six different dimensions are constructed by Kaufmann, Kraay and Zoido-Lobaton. The indices they employ are aggregated from 352 individual variables and taken from 31 different organizations. We have to make a distinction between aggregated indices and individual indices. We have chosen to use aggregated governance indicators as they have higher chance of following a normal distribution and thus can provide more reliable and statistically significant measures than individual variables.
The indicators are estimated using to methods: Survey’s and reports from experts on a certain country and working for an international organization. We can therefore conclude that these indicators are based on human perceptions of governance situation in a certain country. We find this method useful as the indicators are not only determined by regulations but rather by the environment where they are applied. A potential problem could be however that perceptions can be culturally different from person to person or from organization to organization which means that it potentially could not completely reflect the true situation.
The governance estimates are normally distributed with a mean of zero and a standard deviation of one in each period. The numbers originally lie between -2.5 and 2.5, but because
Government indicators (location advantages)
Description: Expected outcome
1 Firm size Total assets (logaritm) of a firm in a given year + 2 Firm age Year under investigation subtracted from the
year of incorporation.
+ 3 Profitability Profit (loss) before taxation divided by
operating revenue/turnover times 100.
of potential problems with regression indicators (being 0) we have decided to add 2.5 which means that our numbers lie between 0 and 5. Furthermore it is important to mention that the numbers have been transformed so that higher numbers indicate a positive result (.e.g 5 for corruption means that the a country is not corrupt).
We will now discuss the governance indicators. Voice and Accountability (VA)
Voice and accountability measures perceptions of a country’s citizens to which they are able to participate in selecting their government, as well as freedom of expression, freedom of association and free media. These variables focuses in particular on those aspects that are related to they way authorities are selected and replaced. It combines the different indicators such as the political process, civil rights and institutions that facility citizen control of government actions. Economic theory does not (yet) support the significant relation to FDI. Arguments are made that firms ignore political rights because they might support suppressive regimes so that a firm can exploit country’s scarce resources. However papers that have done research into the relationship between democracy and FDI show that counties with a high degree of political and civil liberties. Examples of this conclusion can be found in the works of Busse (2003) but also Addison and Hesmati (2003) that found a positive effect for a sample of 110 countries.
Political Stability and Absence of Violence (PV)
This variable measures the perceptions of a country’s citizen of the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means, including politically motivated violence and terrorism. There are many studies on FDI that take political instability and violence as a variable in their research. Political instability has a direct impact on the continuity of policies in general which leave investments uncertain. Violence has the same effect and can pose threats to personnel of multinational firms.
talked earlier about the importance to distinguish firms on a an industry level to generalize findings. Kolstad and Villanger (2004) have researched political instability in the service sector in relation to FDI. They find that that stability is irrelevant for FDI in most sub-sectors. This is an important finding because it would mean that in general there is a negative causal relationship but for certain industry sectors there could be no or possibly even a positive relationship. Lastly Desborder and Vicard (2005) show that conflicts (specifically with the involvement of armoured weapons) discourage or have a negative impact on FDI.
Government effectiveness (GE)
Government Effectiveness measures the perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. The World Business Environment Survey (2004) shows that policy and regulatory uncertainty obstruct business operations. It is said that even when regulations are business forthcoming they have to be applied in an efficient manner to have a decent effect. Two authors have focused on the relationship between Government effectiveness and FDI; Busse and Hefeker (2005) and Bénassy-Queré et al (2005). They both find a positive relationship between Government Effectiveness and FDI.
Regulatory Quality (RQ)
This measures the perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development. Regulatory quality is closely related to government effectiveness as the first focuses on the input that is necessary to implement sound policies the latter focuses on policies in specific. Regulatory quality captures market-unfriendly policies such as price controls or unnecessary trade regulations. Price controls and trade regulations directly effect firms and therefore FDI as a firm is likely not to invest in countries, which have a low regulatory quality. Buch et al (2003) researched the presence of capital controls and exchange rate as a proxy for regulatory quality and they found that the presence of trade restriction or capital controls will lower FDI, while multiple exchange rates surprisingly increases FDI.
Rule of Law (RL)
police, and the courts, as well as the likelihood of crime and violence. The effect of rule of law on FDI is easily explainable: firms and investors need certain stable rules so that contracts can be enforced and established and so that transactions can be established without fraud and infringements of the law. A number of authors have done research on this relationship and Campos and Kinoshita (2003), Busse and Hefeker (2005) find that the extend of rule of law and the effect on FDI is positive. This implies that if a government has a sound set of rules it is more likely to attract FDI.
Corruption (CR)
Corruption is defined as the exercise of public power for private gain. It can affect FDI in multiple ways. First of all it adds costs of doing business, as it makes things more expensive and adds to investments and transactions. Secondly it can also hamper economic growth as it has an influence on competition because it diverts capital towards non-productive activities and negatively affects the provision of public services (Kennedy 2003). Other researchers claim that corruption may facilitate business by constituting an instrument for circumventing inefficient regulations Kaufmann and Wei, 2000). Wei (2006 b) argues that investors are more likely to not invest in countries that deal with corruption as FDI usually involves significant sunk costs that are hard to get back when corruption arises. Moreover it also has influences the bargaining position of the firm in relation to the corrupt bureaucrats as the investment is already made.
Table 2: EFFECT OF GOVERNMENT INDICATORS ON INFLOWS OF FDI FOR ANY GIVEN COUNTRY
Government indicators (location advantages)
Description: Expected outcome
1 Voice and Accountability
Perceptions of country’s citizens, freedom of expression, association and free media
+ 2 Political Stability and
absence of Violence
Perceptions of a country citizen of the likelihood that the government will be destabilized or overthrown
+
3 Government effectiveness
Perceptions of the quality of public services, quality of civil service and degree of independence
+
4 Regulatory Quality Perceptions of the ability of the government to formulate and implement sound policies and regulations.
+
5 Rule of Law Quality of contract enforcement, property rights, the police and the courts, as well as the likelihood and crime and violence.
+
Given the location characteristics as discussed above we can now define our second hypothesis, which is:
Hypothesis 2: The location characteristics of the countries that receive Polish investments have higher government indicators than the location characteristics of Poland.
4.3 Control variables
There are a number of different studies that use various control measures for FDI. Because there is such a vast amount of literature on FDI it is very difficult to find one set or one type of control variables, this is mainly due to differences in samples, methods, theory and perspectives. We have however found an important variable that is used throughout a number of different empirical studies.
Market size
The market size of the host country is a very important characteristic to determine the extend of FDI rather than arms lengths agreements such as export or licensing. This means that a larger market is more probable receiving larger amounts of FDI than smaller markets. For firms it is thus more interesting to invest in larger markets because the sunk costs of setting up a subsidiary is lower as there is a larger economy thus larger economies of scale. Some authors that find market size an important contributor of commitment of FDI are Kolstad and Villanger (2004), Nunnenkamp (2002), Addison and Heshmati (2003) and Chakrabarti (2001).
There are two proxies that are used to measure market size; these are GDP and GDP per capita. Where GDP measures the absolute size of the economy, GDP per capita measures the average income level. In some cases one can argue that GDP is a poor indicator to reflect the purchasing power of the inhabitants of a certain country as it rather looks as size of the population. The downside of GDP per capita is that it does not reflect the number of people in a country but rather looks that the purchasing power of a single average customer.
4.4 Sample and methodology
Our full sample consists of 65 Polish manufacturing firms enlisted in the Warsaw Stock exchange. We have selected manufacturing firms in particular as Dunning has pointed out that the eclectic framework is industry dependent. Furthermore we have selected firms from one country because the eclectic framework is not only industry dependent but also country dependent, as the motives for international expansion might vary from country to country because of differences in cultures, location characteristics or differences in economic growth.
The firms where selected using the Amadeus and Zephyr database, selecting the industry “UK-SIC P15 to 36”. Furthermore we restricted the activity search to “acquirer & majority stake”. To get a sufficient sample size we looked at “rumour date” instead of “finished deal” as a rumour already gives the intention of commitment of FDI. The firms that were selected are the most active ones in both domestic and international investment activity and symbolize a considerable part of the Polish manufacturing industry. We have tried to get a large sample size by taking the largest of number of years that were available in both databases. Because there was a lot of data missing before 2003, we have selected the years 2003 to 2007 (inclusive). Unfortunately there was not enough data available for 2008, which would have made the study more recent. We have used balance sheets and Profit and loss accounts for each firm and each year under investigation. The data was partly available in the Amadeus databases but the majority was retrieved from annual reports, which made this study a lengthy one as 205 individual annual reports were looked at.
For the location characteristics we looked at every host country a firm showed a commitment for FDI. In total we looked at 15 countries and their location characteristics for 5 years. In total we get 31 location dependent observations. As we looked at 6 different proxies to measure location advantages this gave us a sample size of 186 location characteristics. All these location characteristics were gathered from the IRIS-3 file of International Country Risk Guide (ICRG) database.
year by the same firm, these were treated as different events. An overview of the firms and their investment activities can be seen in the following table:
Table 3: POLISH FIRMS THAT ENGAGED IN INVESTMENT ACTIVITY ABROAD FROM 2003 TO 2007
The host countries of those 48 international investments from Poland cover a total of 15 different countries. These international investments took place in Western Europe (11 times), the Baltic states (10 times), Eastern Europe (19 times) and Central Europe (10 times). An overview of the countries that were invested in and in which year can be found in the following table:
Firm Acquisition/minority stake
(number) 2003 2004 2005 2006 2007
Bioton SA 5 4 1
Fabryka Farb i Lakierow
Sniezka SA 4 1 3
Polski Koncern Naftowy Orlen
SA 11 4 4 1 2
Ambra SA 4 1 3
Lubelskie Zaklady Przemyslu
Skórzanego PROTEKTOR SA 2 2 Koelner SA 3 1 2 Zaklady Urzadzen Komputerowych ELZAB SA 1 1 Decora SA 1 1 Relpol SA 1 1 Grupa Lotos SA 1 1 Barlinek SA 1 1
Polski Koncern Miesny DUDA
SA 3 2 1
Polcolorit SA 1 1
Cersanit SA 2 2
Hoop SA Zaklad Pracy
Table 4: ENGAGEMENT IN INVESTMENT ACTIVITY BY HOST COUNTRY AND YEAR
We see a clear increase in number of investments in the last two years of investigations. This can be explained by commitment of Poland to the European Union that have opened economic borders and has allowed Poland to implement political, economic and institutional reforms and liberalization of trade. Furthermore we have utilized a Chi-square test to test for statistical significant relationships between the host region and the time period. These results can be seen in the following table:
.
Table 5: CHI-SQUARE TEST TO TEST SIGNIFICANT RELATIONSHIPS BETWEEN HOST REGION AND TIME PERIOD.
With a Chi-square of 18.90 and a critical value of 21.026, our result is not significant, indicating that the country of the investment is not statistically associated with the actual timing of the investment. As the chi-square method requires values greater than 0 and that a
maximum of 20% of the data can be smaller than 4 we have created another table by adding regions and years.
Table 6: CHI-SQUARE TEST TO TEST SIGNIFICANT RELATIONSHIPS BETWEEN HOST REGION AND TIME PERIOD WITH GROUPED DATA
Regions 2003-2004 2005-2006 2007 TOTAL
Western/Central Europe 4 9 7 20
Eastern/Baltic 7 13 8 28
TOTAL: 11 22 15
Chi-square 0.28
With a Chi-square of 0.28 this result is far from significant which means that the region is not statistically associated with the actual timing of the investment.
There have been previous studies (Demos, 2004) that looked at the effect of an announcement of foreign direct investment of a firm and it’s stock evaluation. The results were promising as they found that these announcements have a positive effect on stock market evaluation of a firm. This study is focused on the underlying drivers that make firms expand abroad. We therefore look for an explanation why a firm undertakes an investment abroad. The dependent variable is therefore measuring the intention to invest abroad in a specific time and not the actual (or volume) of the investment. The dependent variable can therefore take a 0 when there is an investment at home or a 1 when there is an investment abroad. This approach was also used by Duran and Ubeda (2001), the difference with their study and ours is that they have used questionnaires to measure the intention to invest abroad while we have used actual announcements that were published in the news.
A dependent variable (=y) that can only take two values (either there is an investment abroad =1 or there is an investment within Poland= 0) is called a dichotomous (two-way) variable. To test dichotomous variables as a dependent factor with multiple independent variables the Logit regression model is the most appropriate model.
and P(y = 0) = 1 ¡ P(y = 1). This model has a convenient representation in terms of the odds of the event y = 1 as,
This means the log odds is simply the linear function β0 + β1x. This model has two parameters, but the parameter β1 will be of primary interest. It controls the degree of association between the response and predictor variables.
Statistical inference for one model parameter typically involves a hypothesis test, or a confidence interval for the estimated odds ratio. A confidence interval for β is:
where Z* is the appropriate multiplier from the standard normal distribution. Confidence intervals whose endpoints do not contain zero indicate a relationship between the predictor x and the response after adjusting for any other predictor variables in the model. Confidence bounds containing zero do not show significant evidence of a relationship between the predictor and response. A hypothesis test of H : β = δ vs Ha : β= δ uses the standard normal test statistic,
When δ = 0, this test statistic and p-value are given in eviews, and should correspond to the inference that would be made if a confidence interval was computed. When this test gives a small p-value, it will correspond to a confidence interval for β that does not contain zero (there was an investment). A confidence interval for the odds ratio exp (β ) is obtained by simply exponentiating the confidence limits for the parameter. Suppose the confidence bounds for β are (a, b) then the confidence limits for OR are (exp(a), exp(b)).
Goodness of Fit
Likelihood Ratio
The Likelihood ration test
With the likelihood ratio test we test the hypothesis that all parameters of the model are equal to zero, against the alternative hypothesis that all parameters do not all equal to zero:
0 1 : 0 : 0 H H
θ
θ
= ≠For this test we maximize the log-likelihood function for the unrestricted model θ ≠0 and the restricted model θ =0 . Then we calculate the difference between the restricted and unrestricted model, which result in the likelihood ratio:
ˆ ˆ R U L L
λ
=Figure 2 shows this ratio graphically. The parameter θ is on the x-axis and the value of the
log likelihood function on the y-axis. Furthermore,
θ
ˆMLEis the parameter value at which theunrestricted model is maximized and θ =0the parameter value if our restriction is valid. The hypothesis will be rejected if the difference between the restricted and unrestricted model is large. When testing whether the unrestricted model is significantly different form the restricted we use the large sample distribution of −2 lnλ , which has a chi-squared
distribution, with degrees of freedom equal to the number of restrictions imposed. In the regression output table we report the p-values of this test. If the p-value is lower then 0.05% we reject the null-hypothesis that all parameters are equal to zero.
Figure 2: Difference restricted and unrestricted model
ln ( )Lθ
The Pseudo R-squared
The pseudo R-squared or McFadden R-squared compares the value of the likelihood function of the restricted and unrestricted model:
2 ln ˆ 1 ˆ ln u r L Pseudo R L = −
The Pseudo R-squared explains the promotion of the total variability of the outcome that is accounted for by the model. As we looking at location and firm specific variables we include these all in one model. If the R-squared value from this model would result in .72 this would mean that the variables in our model predict 72% of the variability in the variable characteristics. This would thus mean that most of the variability is accounted for, however you may want to increase the number of variables or decrease the number of variables to increase the explanatory power of the complete model.
Higher pseudo R-squareds report that this model better fits the outcome data than the previous model. While pseudo R-squares cannot be interpreted independently or compared across datasets, they are valid and useful in evaluating multiple models predicting the same outcome on the same dataset. In other words, a pseudo R-squared statistic without context has little meaning. A pseudo squared only has meaning when compared to another pseudo R-squared of the same type, on the same data, predicting the same outcome. In this situation, the higher pseudo R-squared indicates which model better predicts the outcome.
5. Empirical analysis
We have constructed the following equation that will be tested using the Logit regression. FDIi = β 0 + β 1Firm agei + β 2Firm sizei + β 3Profitabilityi+β4 Leveragei + β 5Regulatory
Table 7: VARIABLES, MEASUREMENT, DESCRIPTION AND SOURCE
Table 7 (see above) presents the various variables and how it is measured, furthermore you see a summary of the description of each variable and the source. The results of this research can be viewed in Table 8
Variable Measurement Abbreviation Description Source
Dependent Variable
Foreign Direct Investment
Decision
FDI_DUMMY Value can take 1 if there is a foreign investment and 0 if there is a domestic investment
Author’s own Design
Firm
Variables Firm Age FIRM_AGE
Year of investment (home or abroad) minus the year of incorporation Annual Reports of firms enlisted in Warsaw Stock exchange (2003 – 2007), Amadeus, Zephyr database
Firm Size SIZE Logaritm of total assets As Above
Profitability PROFITABILITY Profit (loss) before taxation divided by operating revenue/turnover times 100. As Above
Leverage LEVERAGE (total liabilities / total assets) As Above
Location institutional Variables Regulatory quality RQ IRIS-3 File of International Country Risk Guide (ICRG) data Voice and Accountability VA
Perceptions of country’s citizens, freedom of
expression, association and free media As Above Rule of Law RL
Quality of contract enforcement, property rights, the police and the courts, as well as the likelihood
and crime and violence.
As Above Political
Stability PS
Perceptions of a country citizen of the likelihood that the government will be destabilized or
overthrown
As Above Corruption CR Exercise of public power for private gain As Above Governance
Effectiveness GE
Perceptions of the quality of public services, quality of civil service and degree of
independence
As Above Location
Economic Variables
Market Size MARKET_SIZE in millions of 1990 US$ (converted at Geary
Table 8: RESULTS RESEARCH, RUNNING THREE DIFFERENT MODELS
Model 1 Model 2 Model 3
Dependent Variable FDI Dummy (0/1) Independent Variables
Coefficient z-statistic Coefficient z-statistic Coefficient z-statistic
FIRM_AGE -2.629122 -0.950277 -2.086021 -2.962044*** -1.823554 -2.193479** SIZE -0.165605 -0.14446 0.707478 2.342565** 0.311453 0.885383 PROFITABILITY 0.09807 1.357979 0.018566 1.084292 0.035631 1.696555* LEVERAGE -0.303288 -0.252659 -0.442426 -1.723568* -0.270813 -0.885997 RQ 3.407498 0.433955 VA -38.81377 -1.630555 -5.900506 -5.0466*** RL 11.31144 0.640143 PS 4.595425 0.609313 6.719765 4.500369*** CR 25.74553 1.01533 GE -8.195561 -0.695821 MARKET_SIZE -11.24336 -2.627478** -4.318697 -5.296451*** -2.329813 -2.941772*** C 90.57879 2.726062 21.00348 4.453401 11.59158 2.114101 No. Observations 205 205 205 Pseudo R-squared (McFadden) 0.889312 0.286625 0.462505 Probability (LR stat) 0 1.84E-12 0 Log-likelihood -12.34931 -79.59039 -59.96764
The estimates this table lists are based on probit models. The dependent variable is indicated in the table. The definition of the other variables can be found in Table 7. Coefficients are listed in the first column, with z-statistics reported in italics in the second column. *** Significant at 1%, ** significant at 5%, * significant at 10%.
