1 The effects of the corporate tax, the personal income tax, the tax on goods and services and social security contribution tax on the level of inward FDI
stocks and the level of inward FDI stocks for eight major industries
Master thesis, MSc international Economics and Business University of Groningen, Faculty of Economics and Business
July 5, 2019
Wouter Crans Student number: s1991876
Supervisor: dr. R.K.J. Maseland
2 The effects of the corporate tax, the personal income tax, the tax on goods and services
and social security contribution tax on the level of inward FDI stocks and the level of inward FDI stocks for eight major industries
ABSTRACT
Although the relation between the corporate tax rate and the inflow of FDI has been studied by multiple authors, these studies have overlooked the role of other taxes, such as the personal income tax, the tax on goods and services and the social security contributions. This
study examines the relation between these taxes and the level of inward FDI stocks and find out which of these sorts of taxation is the best for attracting inward FDI stocks. The results show that level of inward FDI stocks is negatively related to the corporate income tax. The level of inward FDI stocks does not have a relation with the personal income tax and the social security tax, while it has a positive relation with the taxes on goods and services. The
corporate tax rate has a significant relation with two of the eight sectors. Therefore, a fiscal policy with a lower corporate tax rate and a high tax rate on goods and services should be the
best to attract inward FDI. Furthermore, this paper also examines the effects of taxation for the FDI inflow for different industries. The data reports FDI inflows of industries have
different relations with the different sorts of taxation.
Keywords
foreign direct investment, corporate income tax rate, personal income tax rate, tax on goods and services, social security contributions, industries
Introduction
3 Governments have many instruments at their disposal to steer economic activities. Taxes are one of these instruments and they can be used to attract or repel certain economic activities.
The OECD issues tax directives to harmonize fiscal policies and avoid a race to the bottom (OECD 2013). However, with the upcoming Brexit and the possibility of withdrawal of other countries from the European Union, things may chance. Withdrawing countries have more autonomy and can implement their own tax legislation without being bound by these tax directives. Therefore, they have the option to implement favourable tax legislation to attract investments from foreign countries. This makes tax competition a hot topic and an important aspect for policy makers and governments.
By setting fiscal policies, governments are influenced by the expectation of the behaviour of investors from foreign countries. Will they think it is profitable to invest in countries abroad or not? One method that investors use to invest abroad is foreign direct investment (FDI). FDI can be an important driver for economic growth for both origin and destination countries in terms of, for example, economic growth, productivity, wages and employment. This is because the internationalization and integration of economies help to better exploit the advantages of enterprises, to increase their competitiveness, to increase the competences of the local workforce, to increase employment and wages, to stimulate technology transfers and innovative activity. Furthermore, the expansion of multinational enterprises (MNEs) are often accompanied by the creation of complex cross-border production chains, which can also have important positive economic implications (Carril-Caccia and Pavlova, 2018). The relevance of these foreign direct investments has grown rapidly over the last two decades. Between 2005 and 2018, the share of FDI stock in Global GDP has increased from 24.7% to 35.4% (OECD, 2019).
Firms can have different reason to invest in foreign markets. Dunning (1993a) developed a popular framework named the Ownership, Location and Internalization (OLI) paradigm or
“eclectic” approach to determine if a firm should engage in foreign direct investment. Previous literature studied the factors that influence the location of firms to further explain where firms conduct FDI. This literature studied the relationship between FDI and several macroeconomic factors. These factors give benefits to firms and thereby attract FDI flows. Some of these factors are market size and market potential (Culem, 1988; Wheeler and Mody, 1992; Barrel and Pain, 1997; Bevan and Estrin, 2004; Bevan A. and K., 2004; Busse and Hefeker, 2007 and Trevino
4 et al., 2008), labour costs (Konings and Murphy, 2006; Bellak et al., 2008; Braconier et al., 2005), natural resources (Asiedu, 2006; Babatunde and Adepeju, 2012), the degree of openness (Campos and Kinoshita ,2010; Trevino et al.,2008; Sekkat and Veganzones-Varoudakis, 2007;
Babatunde and Adepeju, 2012; and Pereira 2011), infrastructure (Bénassy-Quéré, Gobalraja, and Trannoy , 2007; Bellak and Leibrecht , 2009; Globerman and Shapiro, 2003; Campos and Kinoshita, 2010; Moreiraa, 2009; Asiedu , 2006; Demekas et al., 2007) and macroeconomic stability (Campos and Kinoshita, 2008; Asiedu, 2006; Antonakakis and Tondl, 2010).
Taxes can also influence the location where firms conduct FDI. The literature on the effects of taxation on the investment of multinational firms has mostly focused on the taxation on corporations and capital (Hartman, 1984; Devereux and Freeman, 1995). Most of these studies found evidence of a negative relationship between the tax rates and the level of inward FDI.
However, few papers have researched the association with taxes other than the taxes on corporations and capital, even though these taxes may also be important for the attraction of FDI. This paper will include four different kinds of taxes to find out which tax is the best for the attraction of inward FDI stocks. Answering this question will help policy makers give more insight on the relationship between the different kinds of taxes and the level of inward FDI stocks. This will allow them to set a fiscal policy that generates the needed revenues for the government expenditures, but which will also attract inward FDI stocks.
The relation between the different taxes and inward FDI stocks will not only be investigated on the total level of inward FDI stocks, but also for the level of inward FDI on the industry level for eight major industries. It is likely that industries have different relationships with the different sorts of taxes because the characteristics of firms differ between industries. A low corporate tax rate might be attractive for certain industries, while other industries will look for other location factors for the location decision. Some industries are mobile and can quickly move to other countries to enjoy favourable tax regimes, while others are bound to certain places (e.g., require natural resources). Some industries have a lot of profit from capital gains, so they seek locations with a low tax rate on capital gains, while other industries are more labour intensive and seek locations with cheap labour and low taxes on labour. A better insight in the relationship between the different sorts of taxes and the level of inward FDI stocks for different industries is relevant and very important information for policy makers so they know which tax sorts attract or repel certain industries. For example, a policy maker wants to lower a tax rate to attract FDI in the manufacturing industry to try to reduce unemployment or to benefit from spill-over effects. Will a lower corporate tax rate attract FDI for this industry? Or maybe a lower tax rate on goods and services? This paper will try to answer these questions.
This paper seek to contribute to the literature in two ways. First, it will not only focus on the effects of the corporate tax rate, but also the effects of the personal income tax, the tax on goods and services, the social security contributions to find out which form of taxation is the best for the attraction of FDI inward FDI stocks. To the best of my knowledge, this has not been researched before.
The reason why this paper will focus on the corporate tax, the personal income tax, the tax on goods and services and the social security contributions is because these taxes make up most of the total tax burden for the selected countries. I consider the tax rates on payroll and the tax
5 rates on property to be too small to be relevant and therefore they are not included in the models.
The results show that the level of inward FDI stocks is negatively related to the corporate income tax. The level of inward FDI stocks does not have a relation with the personal income tax and the social security tax and it has a positive relation with the taxes on goods and services.
Therefore, a lower corporate tax rate and a high tax on goods and services is the best for the attraction of inward FDI stocks.
The results for the effects on different industries show that a high corporate tax rate repels FDI stock for the finance and insurance industry, while a high personal tax rate repels FDI stocks in the water supply and waste management industry. A high tax rate for the tax on goods and services attract FDI in the services industry and high social security contribution tax rates repel FDI stock in the services industry.
This paper uses two panel data sets. To answer the first research question, a dataset ranging from 2005 until 2016 is used. To answer the second research question, a dataset containing industry FDI data is used that ranges from 2013 until 2017. A longer time period would give better estimates and more accurate data; however, this was not available or incomplete. The data extracted from the OECD Database and the World Bank database.
The remainder of the paper is structured as follows: the next section will examine previous literature, which will lead to two research questions. The section after that will discuss the research methodology, the data and the variables used the models. The section after that will show the results and discuss them. The last section will conclude this study.
Literature Review
To analyse which form of taxation is the best for attracting FDI, it is important to understand when a firm engages in FDI and what determines the investment location. This information can be used to create a model to analyse the effects of the different taxes on the level of inward FDI.
This section will first define FDI and indicate the benefits of FDI for the host country. Then the OLI framework will be used to understand when a firm engages in FDI and after that the factors that decide where a firm is going to invest. Finally, the research questions will be formulated.
Foreign Direct Investment
This paper will use the same definition for FDI that is used by the OECD. This is because most of the data used is extracted from the OECD database. According to the OECD (2008), the definition of foreign direct investment is: a category of investments that reflect the establishing of a lasting interest by a resident enterprise in one economy (direct investor) in an enterprise (direct investment enterprise) that is resident in an economy other than that of the direct investor.
6 A “lasting interest” implies the existence of a long-term relationship between the direct investor and the direct investment enterprise and a significant degree of influence on the management of the enterprise (OECD, 2008).
The “significant degree of influence on the management of the enterprise” exists if the direct or indirect owner owns 10% or more of the voting power of an enterprise resident in one economy by an investor resident in another economy (OECD, 2008).
The benefits of FDI for the host country
Foreign direct investment is one of the key drivers of international economic integration, and it is an important driver for economic growth (OECD, 2002a and 2002b). For the recipient country it can bring several benefits. The arrival of new multinational entities (MNEs) can foster efficiency through an increase in competition. FDI is often accompanied by positive spill-over effects through the integration of domestic firms in the production process of the MNE through forward and backward linkages(Ashraf, Herzer and Nunnenkamp, 2016; Bloom, Sadun and Van Reenen, 2012; Dachs and Peters, 2014; Girma and Görg, 2007). In addition, it can bring new technology, provide access to new markets, increase the competences of the local workforce and increase employment and wages (Blomström and Kokko 1998).
With the right policy framework, FDI can provide financial stability, promote economic development and enhance the wellbeing of societies (OECD 2008).
When firms engage in FDI
Multiple strategies exist to enter foreign markets, for instance by exporting, licensing, franchising or by investing in the ownership position in an organization located abroad (FDI).
To evaluate if a firm should do a large equity investment in the form of an FDI, different frameworks can be used. The most popular framework is the “OLI” or “eclectic” approach developed by John Dunning (1977). According to this approach a firm should have three sources of advantage when engaging in foreign direct investment.
The first advantage of ownership (O) refers to the competitive advantages of the enterprises who seek to engage in FDI. These advantages are specific to the ownership of the investing enterprises, i.e. their ownership (Dunning, 2000).
The second advantage of Location (L) are advantages related to the locational attractions of alternative countries or regions where the value adding activities of the MNE will be.
According to traditional location theories (Hoover, 1948; Hotelling, 1929; Isard, 1956; Losch, 1954; Lloyrd & Dicken, 1990; Weber, 1929) location advantages can refer to demand related variables, e.g. size, character and potential growth of local and adjacent markets. Supply oriented variables, e.g. availability, quality and price of natural resources, transportation costs, artificial barriers to trade, supply oriented variables, especially those related to comparative advantages of immobile assets, e.g. labour, land and infrastructure or the location and price of created assets, including those owned by firms likely to be acquired. The more these immobile, natural or created endowments, which firms need to use in combination with their own competitive advantages favour a presence in a foreign rather that a domestic location, the sooner firms will engage in FDI (Dunning 2000).
7 The third advantage of internalization (I) offers a framework to evaluate alternative ways in which firms may organize the creation and exploitation of their core competencies in certain countries or regions.
When all three advantages (Ownership, Location and Internalization) are met by the MNE, then the foreign direct investment should be beneficial.
Location factors
The OLI framework primarily aims to explain why a firm conducts FDI. However, the framework is limited towards explaining where firms should locate. According to the theory, companies should locate their FDI in locations where they can derive greater benefits from foreign market characteristics compared to the home market characteristics.
Previous literature on the factors that influence the location of firms can be used to further explain where firms conduct FDI. This literature studied the relationship between FDI and several macroeconomic factors. These factors give benefits to firms and thereby attract FDI flows. These factors are important because the aim of this paper is to know if the different tax rates are location factors that attract FDI and which sort of tax is the best at attracting FDI to a specific location. Therefore, other location factors must be controlled so the effect of taxes are clear. Some of these other factors that influence FDI flows are the market size and market growth, the openness of the economy, the exchange rate valuation, the cost of labour, the infrastructure, the abundancy of natural resources and the government effectiveness.
Market size and Market Potential
Larger markets are attractive for firms due to a larger potential demand and lower costs due to economies of scale (Amiti, 1998; Krugman, 1979). Indeed, Resmini (2000) found that countries in central and eastern Europe with larger population tend to attract more FDI.
Numerous researchers have studied the relation between the host countries market size and FDI and found a significant positive effect on the size of the market and the FDI inflows(Culem, 1988; Wheeler and Mody, 1992; Barrel and Pain, 1997, 1999; Bevan and Estrin, 2004; Bevan A. and K., 2004).
Different methods have been used to incorporate the market size index into the calculation of FDI inflows. Srinivasan (1985) and Reddy (2011) used the real GDP of countries as measure for market size and found a positive and significant relationship between market size and GDP.
Demirhan and Masca (2008) stated that the absolute value of GDP or GDP per capita does not affect FDI and the real growth of GDP per capita should be taken as proxy instead to measure market size. Pereira (2011) used GDP per capita as a measure of market size and found a positive and significant relationship between market size and FDI.
According to Morrissey and Rai (1995), investors take into consideration the market potential of the host country for the location of investment. The annual growth rate of GDP at market prices is used as a proxy for market potential. According to Lim (1983), there are better opportunities in rapidly growing economies than economies that have slowed down or are not growing. Not all studies find a positive relation between GDP growth and FDI. Schneider and Frey (1985), Culem (1988) and Billington (1999) find a positive relation between growth and
8 FDI, but not all studies find a positive relation. Nigh (1985) found a negative relation between the growth rate and FDI.
Degree of Openness
Another important factor influencing the location of firms is the openness of the economy. An open market benefits firms by giving producers access to bigger markets, which allows them to increase the scale of their production and encourage market competition and innovation.
Besides this, companies can also benefit from open markets because it makes new technologies move more freely (OECD 2019). An open economy can be achieved by having a liberal trade regime, experienced international trade relations, many bilateral trade or free trade agreements and are often seen as a positive setting for investors (Antonakakis and Tondl 2015).
Multiple researchers found that the openness of the host country is an important determinant for FDI inflows (Campos and Kinoshita 2010, Trevino et al. 2008, de Boyrie 2010 and Sekkat and Veganzones-Varoudakis 2007, Djankov et al. 2010, Babatunde and Adepeju 2012 and Pereira 2011). Waldkirch (2010) found evidence that FTAs of the United States with Central America have generated important FDI flows, while Baltagi et al. (2008) found that FTAs between EU and Eastern Europe also generated FDI flows. According to Júlio et al. (2013), a country must have a good degree of openness together with a good political stability, governance and human resources to attract FDI.
An indication of the openness of the economy can be the exports plus imports as share of GDP (Antonakakis and Tondl 2015).
Labour costs
Cheap labour prices is also an important factor that determines the location of firms. Firms are often attracted to locations with cheap labour, especially in labour intensive industries because it can minimize the cost of production (Eckel 2003). The FDI to developing countries is therefore encouraged by high wage gaps (Economou et al. 1996)
Konings and Murphy (2006) analysed the employment behaviour of MNEs in Europe using firm-level panel data of European multinational parent enterprises and their European affiliates.
They found for the parent firms in the manufacturing sector that the labour cost elasticity of parent employment with respect to North European affiliates’ labour costs is positive and significant. This implies that employment substitution between parent and their North European based affiliates takes place in response to wage costs differences between the parent company and its affiliates. However, they did not find evidence for this effect between parent employment and their affiliates that are in the low-wage regions of the European Union or in Central and Eastern Europe. FDI in these low-wage regions are discouraged in high labour cost countries. Bellak et al. (2008) analysed the determinants of FDI across Central and Eastern European Countries, focussing on labour costs. They conclude that higher unit labour costs and higher total labour costs affect FDI negatively, while higher labour productivity impacts FDI positively. Braconier et al. (2005) explored how wage costs for high-skilled and less-skilled labour in host countries affect the level of FDI. They found that more vertical FDI is conducted in countries where less skilled labour is relatively cheaper. For horizontal FDI they find that skilled-wage cost premia also affect the FDI.
9 Natural Resources
Natural resources is also an important factor that influence the location of firms. Natural resources attract firms that want to extract these resources. Firms that use these natural resources in their production process might also benefit from locating themselves close to the natural resources because of lower transportation costs.
Asiedu (2006) used panel data for 22 countries over the period of 1984-2000 to study the impact of natural resources, market size, government policies, political instability and the quality of the host country’s institutions on FDI. The paper concludes that natural resources are in some regions, like Africa, the major reason for FDI inflows. Babatunde and Adepeju (2012) studied the effect of natural resources on FDI in developing countries. They conclude that natural resources are a determinant of FDI in developing countries.
However, there are also papers that argue that the presence of natural resources might have a negative effect on the attraction of FDI in the long run (Robinson et all. 2002). Which is the so called “Resource Curse”, this refers to the paradox that countries with an abundance of natural resources have less economic growth and worse development outcomes than countries without these natural resources.
Infrastructures
According to the literature, well-established and quality infrastructure is an important factor that determines the location of firms and FDI. Bénassy-Quéré, Gobalraja, and Trannoy (2007) and Bellak and Leibrecht (2009) found that there is a positive relationship between a good infrastructure in Eastern European countries and FDI. Globerman and Shapiro (2003) that countries need a certain threshold level of infrastructure to attract FDI from US firms. Campos and Kinoshita (2010) found that a good telecommunication is important for the attraction of FDI. According to Moreiraa (2009), the quality of the electricity, water, transportation and telecommunications can be proxies for the quality of the infrastructure. According to Asiedu (2006), the quality of the infrastructure is one of the important determinants for the FDI inflow in Africa. Demekas et al. (2007) used, besides the statutory corporate tax rate, also a proxy for infrastructure in their study. Their panel analysis of aggregate FDI flows during the period 1995 and 2003 show that infrastructure determines the attractiveness of developing countries.
This is not valid for countries that are more economically developed. Also, they report that the tax effects are significant.
Macroeconomic stability and Government Credibility
Numerous studies have stressed the importance of macroeconomic stability as an important determinant of the location of firms and FDI (Campos and Kinoshita, 2008; Melanie Lansbury and Smidkova, 1996; Asiedu, 2006). A credible government and macroeconomic stability are associated with higher economic growth which attracts firms. Besides this, high corruption levels tend to add investment costs and reduce profits (James and Jiangyan, 2010).
The definitions and measures for macroeconomic stability vary between studies. According to Antonakakis and Tondl (2010), macroeconomic stability is important to avoid risk, and this involves low inflation rates, a stable currency and low external debt. Srinivasan (2002) argues
10 that investors prefer locations where market uncertainties are low, measuring it by using the inflation and exchange rates.
A weak exchange rate might increase vertical FDI because firms take advantage of relatively lower prices in the host country for the purchase of facilities or machinery. Serven (2002) argues that the exchange rate uncertainty discourages investors, while Busse and Hefeker (2007) argue that inflation levels are the important determinant for FDI inflows. Djankov et al (2010) use inflation and the exchange rate to measure the macroeconomic stability and found that the inflation has a significant negative relationship with FDI inflow, while exchange rates do not. However, according to Swenson (1993), the exchange rate does influence the inflow of FDI.
According to Júlio et al., (2013), investors prefer a host country for their investment with a good legal and bureaucratic environment. The Government quality is a part of this environment as can be seen an important determinant of FDI. A high index of corruption is strong negative determinant of FDI.
Taxes
Taxes can also be an important location characteristic for the attraction of firms and FDI. In the past, many countries have lowered their taxes to attract FDI (Loretz, 2008). The corporate taxation and tax treatment affect the wedge between the pre and post-tax rates of return on FDI.
Locations with a lower corporate tax rate will have more profit compared to locations with a higher corporate tax rate while having the same amounts of revenue. Therefore, a high corporate tax rate might have a negative influence on the inflow of FDI because it reduces corporate profits. This has been examined by a bulk of literature.
Hartman (1984) used aggregate time series data to examine the correlation of the after-tax rates of return and the volume of inward FDI flows. He made the distinction between FDI financed with retained earnings and transfer of funds and concluded that the tax rate elasticity for retained earnings is significant, while the tax rate elasticity for transfers is not significant.
Cassou (1997) used bilateral FDI data on six investing countries and found a significant negative impact of the corporate tax rate on inward FDI. Devereux and Freeman (1995) estimated the impact of taxation on foreign direct investment flows by using a sample of seven OECD countries since 1984 through 1989 and introduced the cost of capital into their analysis which reflected the effective marginal tax burden on investments. They found that the choice between domestic investment and total outward FDI is not significantly affected by taxation, but it does affect the location of where outward FDI is invested. Buettner (2002) used a panel of intra-EU transfer of funds between 1991 and 1998. Besides the statutory corporate income tax rates, he uses the effective marginal tax burden and the effective average tax rate in the determination of FDI. Becker et al. (2006) estimated that a reform of tax policy, which decreases tax rates, tend to attract FDI. They emphasised that tax policy makers should consider the costs of lowering taxes, which could lead to lower tax revenues as well as a loss of well- being. Kinda (2014) did not find a significant effect from tax rates on the inflow of FDI.
Most of the studies that examined the effects of taxation on FDI focused primarily on the effects of corporate tax. However, other taxes can also influence FDI flows. When income taxes are
11 lower, people have more after-tax income. This income can be used to buy goods and services, increasing the market potential. Bénassy-Quéré et al. (2005) analysed FDI from US firms in 18 EU countries in the period from 1994 until 2003. They argued that there is a trade-off between a lower tax rate and public spending. They both increase FDI. The findings confirm this, and they emphasize that extreme tax competition is not desirable for countries because of negative spill overs, such as lower funding for public goods.
When the tax on goods and services is low, more of these goods and services can be bought, because prices can be cheaper, which increases the revenues of the firms selling the goods and services. These effects of taxation would suggest that there is a negative relationship between the tax rate and inflow of FDI. Radulescu and Egger (2008) analysed the labour tax impact on the location decisions made by multinational enterprises. According to the authors, high wage taxes borne by highly skilled employees or managers lead to reduced worker efforts, which lead to higher production costs and to less profits. The high employer born taxes on highly skilled labour have a direct negative impact on the firm’s profits, while the high employee borne taxes indirectly affect the profits by reducing the effort of managers. Therefore, it becomes more profitable to set up an MNE abroad where taxes are lower and serve consumers by using local subsidiaries. The authors argue that higher taxes on wages reduce not only a country’s attractiveness for the location of headquarters, but also for the location of outward FDI. As a result, they find that both the capital income tax and the rate of labour income tax have a negative relation with the use of foreign-owned subsidiaries and the employee borne part of the labour taxes determine bilateral FDI significantly. Desai et al. (2004) examined the impact of indirect taxes on FDI by American multinational firms. They found that high tax rates are associated with reduced FDI and that this is the same for all types of taxes. Indirect tax rates are, to the same degree as corporate taxes, negatively correlated with investment levels. They conclude that American affiliates located in countries with 10% higher indirect tax rates have 7.1% lower affiliate assets and that affiliates in countries with 10% higher corporate income tax rates have 6.6% fewer assets. Also, a 10% higher indirect tax rate is associated with 1.9% less output, while a 10% higher income tax rate is associated with 1.9%
less output.
Taxes can also have (indirect) positive effects that influence the profits and therefore make a market attractive for investment. As explained before, firms benefit from good working governments and markets with a good infrastructure. The government activities and the investment in infrastructure are paid by the government and therefore are funded by taxes and social contributions. According to the literature, a well-established government and quality infrastructure is an important determinant for attracting FDI inflows.
So, different taxes and social security contributions can have a different effect on the attraction of FDI. These effects have been studied in previous literature. However, to the best of my knowledge, no research has been devoted to the comparison between taxes and which tax is the best for attracting FDI. In this paper I will analyse the influence of different taxes and find out which form of taxation is the best in attracting FDI. Therefore, the research question is as follows:
Research question 1: Which form of taxation is the best for attracting foreign direct investment?
12 Firms in different industries often have different firm-characteristics. Therefore, it is likely that firms in different industries differ in their preference for a location. As Dunning (2000) argued, different industry sectors might generate different parts of the total FDI inflow because their location factors differ. Some industries favour the production in a different country than the home country due to the large amounts of natural resources. Other industries such as the pharmaceutical sector might prefer a highly skilled workforce. The amounts of natural resources and level of the labour force might have a big influence on the attraction of FDI for these sectors, while the amount of taxation might be less of an influence. There exists a large body of literature dealing with the location of industries. According to Huizinga and Laeven (2007), profit shifting might also be a big location factor for MNE’s FDI, and this can have a different impact on different industries. The multinational firm’s indebtedness in a country depends on a weighted average of national tax rates and differences between national foreign tax rates. Multinationals have an incentive to shift debts to countries with higher tax rates to reduce the amount of profits in that country. This might have a different effect on industries that have relatively higher debts compared to other industries. Grubert and Slemrod (1998) found that Puerto Rico, a country with a low corporate tax rate, is attractive for affiliates of U.S. corporations with highly mobile profits. Their results suggest that income shifting is a predominant reason for U.S. investments in Puerto Rico. A few years later, Grubert (2003) investigated the location choice of U.S. parent corporations and their manufacturing subsidiaries to better understand the income shifting process and its implications. He concluded that research & development (R&D) intensive subsidiaries from U.S. parent companies significantly invested more in countries with either a very high or very low corporate tax rate.
These subsidiaries also undertake a significantly larger volume of intercompany transactions, responding to the opportunities for income shifting. Firms that have the comparative advantage of cross-border profit shifting prefer to make high costs in host companies with high corporate tax rates, while they try to shift their profits countries with low corporate tax rates. Stöwhase (2005) used bilateral FDI data from 8 EU countries to the UK, Germany and the Netherlands between the years 1995 and 1999. He showed that the primary sector was not sensitive to taxation, while the secondary and the tertiary sectors are.
So, firms in different industries have different preferences for location factors, and they may also have different preferences for the different sorts of taxation. Therefore, the second research question is as follows:
Research question 2: Does the preference for the different taxes for the attraction of foreign direct investment differ between industries?
Methodology
This section will describe the methods used to answer the two research questions. First, the model to answer the first research question will be discussed. Second, the model to answer the second research question will be discussed. After that the different variables used in the models will be further discussed and finally the data used in the analysis will be discussed.
13 The first goal of this paper is to analyse which form of taxation is the best for attracting foreign direct investment. Theoretically, there are multiple variables that might influence the level of inward FDI. These variables are discussed in the literature section.
Cross-sectional data only allows the analysis of the variables affecting the inward FDI stocks at one point in time. Time series data would only allow the analysis of the relationship between the taxes and the level of inward FDI stock over time, but only for one country. The main feature of panel data is that it has a time series and a cross-section dimension, and therefore panel data allows the analysis of the relation between the taxes and the level of inward FDI stocks for different countries over different years. Besides this, the variables that influence the level of inward FDI stocks may chance over time and may be different for different countries.
Panel data can uncover these dynamic relationships.
Panel data can be categorized into balanced panel data and unbalanced panel data. When observations are missing, the panel is unbalanced. When there are no missing values, the panel is balanced. The dataset used in this paper is strongly balanced. This means that some observations for certain variables are missing. This has the consequence for regression analysis (used in Stata) that all observations with missing values for any of the variables used in the model will be dropped. This is known as listwise deletion.
To estimate the effects of the different tax sorts on the level of inward FDI stocks a regression analysis will be conducted. The simplest method to estimate the coefficients of the various variables that relate to the level of inward FDI is by using the Ordinary Least Squared Model.
This method does not deal with country or temporal effects, it pools all the data and runs an ordinary least squares regression model. There is a chance of specific country or temporal effects since countries have individual characteristic which may influence the inward FDI.
Therefore, I do not consider this method to be appropriate for the analysis in this paper.
Regression models that recognize the panel data structure and deal with country effects are the fixed effects model and the random effects model. These two models also have a way of dealing with omitted variables. It is possible that there are other variables influencing the inflow of FDI stocks which are not discussed in the literature section such as certain economic variables as industry agglomerations, workforce education or cultural differences, etc. This would result in omitted variable bias if these are not controlled for. The fixed effects and the random effects model can eliminate the omitted variable bias.
The fixed effect model identifies the relationship between de independent and dependent variables within the countries, while it controls for the differences between countries. Each country may have individual characteristic which may influence the predictor variables. The fixed effects model removes the effect of time-invariant country level unobserved variables that may bias the model so the net effect of the predictors on the outcome can be assessed (Hsiao 2003). In theory, it is very likely that there are unobserved time invariant country effects that influence the FDI inflow. Variables such as culture or geographical variables might change over time in the very long term, but in the period used in this model I will consider them as time invariant country effects and therefore, the fixed effects model is appropriate. One limitation of the fixed effects model is that the time-invariant variables will be excluded from the model. For example, when there are country specific characteristics such as the education level of the workforce (considering these are fixed over time), then these time-invariant
14 coefficients cannot be estimated in the fixed effects model. However, the changes in the effects of the time-invariant variables can be observed. I do not consider the taxes to be time-invariant, therefore the fixed effects model should be appropriate. Another limitation of the fixed effects model is that it may have to many dummy variables which may plague the model with multicollinearity, which increases the standard errors and lowers the model’s statistical power to detect significance.
When the unobserved factors are not fixed but random, the random effects model can be used to control for these random effects. The rationale behind the random effects model is that the variation across entities is random and uncorrelated with the independent variables in the model (Green 2008). In the case of the model in this paper, it assumes that the country effect (heterogeneity) is not correlated. The assumption of the random effects model is that the covariance of the between-entity error and the independent variables are zero. When this assumption is true, the random effects model gives a better estimation compared to the fixed effects model. However, the random effects model reduces the number of parameters to be estimates but will produce inconsistent estimates when the individual specific random effect, in this case the country effect, is correlated with regressors.
The Hausman test can be used to determine whether the random effect panel data model or the fixed effect panel data model is appropriate. This test is further discussed in the appendix III.
Certain challenges arise when testing which taxes are the best for the attraction of FDI. These are multicollinearity, outliers and the possibility of spurious regression.
Multicollinearity
Multicollinearity means that there is a problematic amount of correlation between the independent variables (explanatory variables). The issue is that these explanatory variables all give the same information in explaining the dependant variable and it will be hard to pinpoint which of the explanatory variables have how much effect on the independent variable. The consequences with a high degree of multicollinearity is that the accuracy of the estimates which is indicated by the variances of the coefficient estimates is compromised. In the case of this paper, the estimates of the different variables that influence the level of inward FDI are biased.
The variances are larger, the standard errors are increasing, and this means that the T statistics will be lower than it would be without the high degree of multicollinearity. The coefficients themselves will still be consistent, however the significance test results will not be trustworthy.
To deal with multicollinearity Variance Inflation Factors (VIF) will be used to detect multicollinearity. The VIF measures how much the variance of the estimated regression coefficient is inflated by the existence of correlation among the predicting variables in the model. A VIF of 1 means that there is no correlation between a predictor and the other predictor variables. A commonly used rule of thumb is that a VIF of 4 or more indicates an excessive or serious collinearity which should be addressed (O'brien, 2007).
To remove the issue of multicollinearity, a correlation table together with a Variance Inflation Factors (VIF) table will determine which independent variables have high correlation (and a high VIF score). The independent variables that have high VIF scores and correlate will be not be included in the model. Therefore, the estimates of the coefficients should be more accurate.
15 Another correlation problem arises when different taxes are included in the same regression, since the level of different taxes should theoretically correlate with each other. However, dropping the taxes that correlate to remove this problem is not a possibility since the aim of this paper is to analyse the different effects of taxes on FDI and find out which form of taxation is the best for attracting FDI. Therefore, separate regressions will be made to estimate the coefficients for each different form of the four types of taxation. However, when each form of taxation is separately calculated in the regressions, knowing that these types of taxation correlate, then it will be hard to pinpoint which part of the estimated coefficient is due to the correlation of taxes and which part is tax specific. Therefore, the correlation between taxes will be controlled by making separate regression for each form of taxation but by including the total tax burden as an independent variable together with the specific tax form as percentage of the total tax burden. The estimates of the coefficients will be more accurate when the correlation between taxes is controlled.
Outliers
The dataset consists of the different variables of 26 different countries, most developed countries and three developing countries. These variables might differ a lot and this might generate issues. The residuals of the fixed effects model are assumed to be normally distributed and homogeneous, however it might be possible that the countries in the dataset have country- specific heteroskedasticity or autocorrelation over time or outliers that might bias the estimations. This might be especially concerning because a special event is presence in the period of the dataset: the great recession (2007 until 2009). To address these issues, a robust regression will be performed. One particular country that might have be considered as a outlier might be Ireland, especially since it also has extreme values in the level of inward FDI stocks (independent variable). The different values between the non-robust regression and the robust regression are shown in the appendix V.
Spurious correlation
Spurious correlation refers to a mathematical relationship in which variables are associated with each other, but are not causally related, due to coincidence or third (unobserved) variables (Pearl 2000). However, this should not be an issue in the models of this paper, since the variables used in this paper have been researched in previous literature and should theoretically influence the level of inward FDI stocks. Besides this, the literature did not suggest alternative variables which affects both the level of inward FDI stocks and one of the variables.
Taking the above into consideration and assuming the fixed effects model will be appropriate, the following model will be used to estimate the coefficients of the different taxes on FDI:
𝒀𝒊,𝒕 = 𝒂𝒊+ 𝜷𝟏 𝑿𝒊,𝒕+ µ𝒊,𝒕 Where
- 𝑎. (i=1…n) is the unknown intercept for each entity.
16 - 𝑌.,0 is the dependent variable: The total level of inward FDI stocks, where I = the entity
and t = the time.
- 𝑋.,0 represent an independent variable.
- 𝛽3 is the coefficient for the independent variable.
- µ.,0 is the error term.
The value and the significance coefficients will determine which form of taxation is the best for attracting FDI.
The second goal of this paper is to analyse if the preferences for the different tax sorts differ for the attraction of FDI for different industries. To do this, four fixed effects analyses will be conducted to test the relationship between the preferences of different taxes and the total level of inward FDI stocks. This will be done in the same way as the model that will answer the first research question, so the total tax burden will be included in the regression and the tax rates for the different tax forms are measured as a share of the total tax burden. The results will be displayed in a total of four tables, one for each different tax, and these tables will contain a fixed effects model for each of the 8 major industries.
The data used to answer the second research question is less balanced compared to the data used to answer the first research question. Besides this, values of the variables are collected for the years 2013 until 2017, which is a shorter time period compared to the period 2005-2016. A smaller sample size increases the margin of error which makes it harder to detect significant effects (decreases the statistical power). As a consequence, the results of the models that answer the second research question are less conclusive.
The variables
This section provides the definitions of the dependent and independent variables used in the models. The independent variables are selected according to the literature.
Dependent Variable
The models that will answer the first research question will use the level of Inward Foreign Direct investment stocks as share of GDP as dependent variable. The model that answer the second research question will use the inward FDI stock per industry sector as dependent variable.
The Inward Foreign Direct Investment Stocks (as share of GDP) measures the total level of direct investments in the host country per year. This is the value of foreign investor’s equity in and net loans to enterprises in the host country. They are measured in USD as a share of the total GDP of that country (OECD 2019).
The Inward Foreign Direct Investment stocks by industry measures the total level of direct investments in the host country per year by industry sector as a percentage of total inward FDI stocks in the host country. It is the value of the total equity and net loans received by the enterprise of the host country in a given sector from foreign investors at the end of each year.
17 The level of inward foreign direct investment stocks per industry are recorded for eight major ISIC4 industries
Independent variables
There are eight independent variables which represent the host country characteristics (market size, labour cost, natural resources, degree of openness, the degree of macroeconomic stability which consists of the inflation and exchange rates and the government effectiveness). Five independent variables are used to represent the host countries taxation policy. These are the total tax revenue, the corporate tax, the personal tax, the tax on goods and services and the social security contributions.
Market size (SIZE)
The proxy for market size is the Gross domestic product per capita measured in US dollars at current prices. The Gross domestic product is the most commonly used measure for the country’s overall economic activity. It represents the total value at constant prices of final goods and services produced within the country for one year (World Economic Outlook April 2019).
Labour cost (LABOUR)
Unit labour costs are often viewed as a measure of international price competitiveness.
Following the papers from Konings and Murphy (2006), Bellak et al. (2008) and Braconier et al. (2005), the labour costs will be used as independent variable as a measure of international price competitiveness. It is defined as the average cost of labour per unit of output produced and it is expressed as the ratio of total labour compensation per hour worked to output per hour worked (labour productivity) and it is measured in percentage change from the previous year.
Natural resources (RESOURCES)
According to Asiedu (2006) and Babatunde and Adepeju (2012), resources are an important factor why investors select a location. In some countries, the earnings from natural resources account for a large share of the GDP and these consist mostly from extracting natural resources such as fossil fuels and minerals. The independent variable for the natural resources is the Total natural resources rents as percentage of GDP. The total natural resources rents are the total revenue that can be generated from the extraction of the natural resource, minus the cost of the extraction. The natural resources rents are calculated as the difference between the average world price of a commodity and the average cost of producing it. The total revenue is the sum of oil rents, natural gas rents, coal rents, mineral rents and forest rents (World Bank, 2011) degree of openness (OPEN)
As independent variable for the degree of openness, the model will follow the papers of Antonakakis and Tondl (2015) and this variable is the sum of exports and imports of goods and services measured as a share of gross domestic product.
degree of macroeconomic stability (INFL and EXCHA)
The model will use two variables as proxy for the macroeconomic stability, the inflation rate and the exchange rate. In the literature these are commonly used as to measure for the
18 macroeconomic stability (Antonakakis and Tondl, 2010; Srinivasan, 2002; Serven, 2002;
Busse and Hefeker, 2007). The inflation is measured by the consumer price index (CPI) and this is defined as the change in the prices of a basket of goods and services that are typically purchased by specific groups of households (OECD 2019). The inflation is measured in terms of the annual growth rate per year.
The exchange rates are defined as the price of a country’s currency in relation to another. They are measured as the national currency unit per US dollar.
Infrastructure (INFRA)
According to multiple previous literature (Bénassy-Quéré, Gobalraja, and Trannoy, 2007;
Bellak and Leibrecht, 2009; Globerman and Shapiro, 2003; Asiedu 2006), there is a positive relationship between a good infrastructure and attracting FDI. These can be measured as the quality of electricity, water, transportation or telecommunication.
The measure of infrastructure in this paper is the Total Inland Infrastructure Investment as share of the GDP. This covers the spending on new transport construction and the improvement of the existing infrastructure. The inland infrastructure includes roads, rails, inland waterways, maritime ports and airports and it takes account all sources of financing.
Government Effectiveness (GOVERN)
To capture the effect of quality of public services, the quality of the civil services and the degree of political pressures and the credibility of the government’s commitment to its policies, the Government Effectiveness Estimate is used. This is an index developed by the World Bank which estimates the country’s score, in units of a standard normal distribution, ranging from - 2.5 to 2.5.
Tax variables
In line with Pereira (2011), the total yearly revenue from taxes in national currency divided by the GDP will be the independent variable to estimate the influence of taxes on FDI. The reason for this is that the statutory tax rates often do not reflect the real tax burden of the specific tax.
This is because the tax legislation between countries differ and the tax bases differ between countries but may also differ between different types of firms.
The variables for the different taxes will be the total revenue per tax form as share of the total tax revenue for the respective country. This is done to control for multicollinearity, as discussed in the methodology. The total tax revenue will also be included in the model to control for the total amount of taxes raised.
The total tax revenue (TOTTAXREV) is defined as the revenues collected from taxes on income and profits, social security contributions, taxes levied on goods and services, payroll taxes, taxes on the ownership and transfer of property, and other taxes (OECD 2019). It indicates the share of a country’s output that is collected by the government through taxes and contributions. It can be seen as a measurement of how a government controls the country’s economic resources.
The Corporate Tax Rates (CORPTAX) is defined as the taxes levied on the net profits of enterprises based in the host country. It also includes the taxes on capital gains (OECD 2019).
19 The tax on personal income (PERSTAX) is defined as the taxes levied on the net income and capital gains of residents of the host country (OECD 2019).
The tax on goods and services (GOODSSERVTAX) is defined as the taxes levied on the production, extraction, sale, transfer, leasing or delivery of goods, and the rendering of services, or on the use of goods or permission to use goods or to perform activities. They consist mainly of value added and sales taxes (OECD 2019).
The Social security contributions (SOCIAL) are defined as compulsory payments paid to general government that confer entitlement to receive a (contingent) future social benefit. They include: unemployment insurance benefits and supplements, accident, injury and sickness benefits, old-age, disability and survivors' pensions, family allowances, reimbursements for medical and hospital expenses or provision of hospital or medical services (OECD 2019). They can be levied on both employers and employees.
Data
The data used in this paper is extracted from the for Economic Co-operation and Development and the World Bank databases.
The Organisation for Economic Co-operation and Development (OECD) is an intergovernmental economic organisation that act as a forum for the 36 member countries to develop economic and social policies. The OECD collects and disseminates economic data for a wide range of economic indicators. The data from the OECD is widely used in academic papers.
From the OECD I have extracted the data of the Market Size, Unit labour costs, the inflation and exchange rates, the Total Inland Infrastructure Investments, The corporate tax rates, the personal income tax rates, the tax rates for the tax on goods and services, the social security contributions, the total tax revenue, the annual Inward Foreign Direct Investment flows as share of GDP, the inward foreign direct investment stocks as share of GDP.
The dependent variable used in the models to answer the first research question is the inward foreign direct investment stock as share of GDP. This measures the total level of net inward direct investment stocks at the end of the year for each country. The net inward foreign direct investment stock is the value of foreign investors of the net loans to enterprises resident in the host country and the equity in these enterprises.
The net inward FDI can be calculated by taking the investments by foreign parents in resident affiliates and investments by foreign fellow enterprises in resident fellows which ultimately controlled by a non-resident parent, and then subtracting the by foreign affiliates in their resident parents and by foreign fellow enterprises in resident fellows which are ultimately controlled by a resident parent. Fellow enterprises are enterprises that are not in a direct investment relationship themselves but have a direct investor in common. They are included because, even though there is no direct investment, any transactions between them likely resulted from the influence that their common direct investor has. The change of the FDI stocks is equal to the value of financial transactions recorded during the year plus other changes in prices, exchange rates and volume. The main financial instruments components of FDI are debts and equity. Debts include marketable securities such as bonds, debentures, commercial
20 paper, promissory notes, non-participating preference shares and other tradable non-equity securities as well as loans, deposits, trade credit and other accounts payable or receivable. The interest returns on these instruments are included in these debts. The equity includes common and preferred shares, reserves, capital contributions and reinvestment of earnings. The Foreign direct investment stocks are measured in USD and as a share of the GDP. This data is extracted from the OECD International Direct Investment Statistics database. This database gives a comprehensive set of statistics on FDI into and out OECD countries from 1982 and onwards.
The FDI stock data is reported monthly, quarterly and yearly. The data for this variable is complete for all 28 countries for the period 2005 until 2016.
The variable market size is measured by the Gross domestic product per capita measured in US dollars at current prices. This data is extracted from the OECD National Accounts Statistics, which include annual and quarterly data ranging from 1955 until 2019 (OECD, 2019). The data is internationally comparable by the system of national accounts 2008 (SNA 2008), for all countries except for Japan and Turkey which report under SNA 1933. The data for this variable is complete for all 28 countries for the period 2005 until 2016.
The variable labour productivity is measured by the average cost of labour per unit of output produced. This is extracted from the OECD Productivity Statistics database (OECD, 2019).
The data is presented as annual datapoints from 1970 onwards. According to the OECD, this data is internationally comparable. The data for this variable is complete except for Israel, Mexico and Turkey for the period 2005 until 2016.
The inflation rate is extracted from the OECD Main Economic Indicators database, which include a wide range of areas from 1961 onwards, such as quarterly national accounts, business surveys, retail sales, industrial production, construction, consumer prices, total employment, unemployment rates, interest rates, money and domestic finance, foreign finance, foreign trade, and balance of payments for OECD countries and non-member economies. The data for this variable is complete for all 28 countries for the period 2005 until 2016.
The exchange rates are defined as the price of a country’s currency in relation to one US dollar.
The data is extracted from the OECD National Accounts Statistics database and includes annual and quarterly data ranging from 1955 until 2019 (OECD, 2019). The data for this variable is complete for all 28 countries for the period 2005 until 2016.
The measure of infrastructure in this paper is the Total Inland Infrastructure Investment as share of the GDP. The data for this variable comprises data collected on an annual basis from Transport Ministries, statistical offices and other institutions designated as an official data source (OECD, 2019). The data is available from 1970 and onwards for 59 member countries of the International Transport Forum. The variable has missing data for the countries Ireland, Netherlands, New Zealand, Portugal and Israel.
The tax variables (total tax revenue, the corporate tax rates, the tax rate on personal income, the tax rate on goods and services and the social security contributions) are measured as the total tax revenue for the respective tax form as share of GDP. The calculation of the total tax revenues and the revenues for the different tax sorts and the social security contributions are extracted from the country’s National Accounts.
The term “taxes” is confined to compulsory unrequited payments to general governments.
These do not include fines, penalties and compulsory loans paid to governments. The definition
21 of the “general governments” are according to the “2008 System of National Accounts” and consists of the central administration, agencies whose operations are under its effective control, state and local governments and their administrations, social security schemes and autonomous governmental entities, excluding public. The taxes paid by governments, such as social security contributions and payroll taxes paid by the government as employer, are also included as taxes.
The data of the tax revenues are recorded on an accrual basis at the time that the tax liability was created.
The tax on personal income is levied on the net personal income of individuals, i.e. gross income minus the allowable tax reliefs. These allowable tax reliefs differ between countries according to the tax laws applicable.
The corporate tax is levied on the profits of enterprises, i.e. the gross income minus the allowable tax reliefs. If a country taxes the gains on capital, then the corporate taxes also include the taxes on these capital gains.
The payments of social security contributions are classified as all the payments that confer an entitlement to receive a future social benefit that are paid to institutions of general governments that provide these benefits. These benefits include the types of social security such as:
unemployment insurance benefits and supplements, accident, injury and sickness benefits, old- age, disability and survivors’ pensions, family allowances, reimbursements for medical and hospital expenses or provision of hospital or medical services. These contributions can be levied on both employees and employers.
The tax on goods and services include all the taxes and duties which are levied on the production, extraction, sale, transfer, leasing or delivery of goods, and the rendering of services.
These include multi-stage cumulative taxes, general sales taxes (levied at manufacture, production, wholesale or retail level), value-added taxes, excises, taxes levied on the import and export of goods, taxes levied in respect of the use of goods and taxes on permission to use goods, or perform certain activities or the taxes on the extraction, processing or production of minerals and other products.
The tax revenues are extracted from the OECD Tax Statistics database (OECD, 2019). The data is provided for the period 1975 onward for all OECD countries. To compare the data, the taxation and social security contributions as share of GDP is used. The data for the tax variables is complete for all 28 countries over the period 2005 until 2016.
These variables are collected for the period 2005-2016 for the following 28 countries:
Table 1
COUNTRY COUNTRY CODE DEVELOPMENT (ACCORDING TO THE UN COUNTRY CLASSIFCIATION) AUSTRALIA AUS Developed
AUSTRIA AUT Developed
CANADA CAN Developed
CZECH REPUBLIC CZE Developed
22
DENMARK DNK Developed
FINLAND FIN Developed
FRANCE FRA Developed
DEUTSCHLAND DEU Developed
GREECE GRC Developed
HUNGARY HUN Developed
ICELAND ISL Developed
IRELAND IRL Developed
ITALY ITA Developed
JAPAN JPN Developed
MEXICO MEX Developing
NETHERLANDS NLD Developed NEW ZEALAND NZL Developed
POLAND POL Developed
PORTUGAL PRT Developed
SLOVAK REPUBLIC SVK Developed
TURKEY TUR Developing
UNITED KINGDOM GBR Developed UNITED STATES USA Developed
ESTONIA EST Developed
ISRAEL ISR Developing
SLOVENIA SVN Developed
LATVIA LVA Developed
LITHUANIA LTU Developed
From the World Bank database, the data for the Natural Resources Rents, the degree of openness and the government effectiveness is extracted for the period 2005-2016 and for the same 28 countries for the period 2005 until 2016.
The World bank is an international organization that fights poverty by offering developmental assistance to middle-income and low-income countries. They also offer advice for both private and public sectors (World Bank, 2019). The data provided by the Worldbank is of high quality by using standards, methodologies, sources, definitions and classifications that are internationally accepted. However, the data used in this paper comes from the statistical systems of the member countries and the quality depends on how well these national systems perform. The Worldwide Governance Indicators score the governments and institutions of countries on six criteria, which include the
23 The natural resources rents consist of oil rents, natural gas rents, hard and soft coal rents, mineral rents and forest rents and are reported annually. They are calculated as the difference between the price of a commodity and the average cost of producing it. The data for this variable is complete for all 28 countries for the period 2005 until 2016.
The degree of openness is measured as the sum of exports and imports of goods and services measured as a share of gross domestic product. This data is annually reported. The data for this variable is complete for all 28 countries for the period 2005 until 2016.
To capture the government quality, Government Effectiveness Estimate is used. This is an index developed by the World Bank which estimates the country’s score, in units of a standard normal distribution, ranging from -2.5 to 2.5. The Government Effectiveness Estimate consists of six broad dimensions of governance for over 200 countries over the period 1996 and onward.
The six broad dimensions are: Voice and Accountability, Political Stability and Absence of Violence, Government Effectiveness, Regulatory Quality, Rule of Law and Control of Corruption. The data for this variable is complete for all 28 countries for the period 2005 until 2016.
The dependent variable for the second research, the Inward Foreign Investment Stocks by industry, is extracted from the OECD International Direct Investment Statistics database for the period 2013-2016. This measures the total level of direct investment in the host country, at the end of the year, by industry sector. This is measured by market value and this is the value of the total equity and net loans received by the enterprise of the host country in a given sector from foreign investors at the end of each year. The level of inward foreign direct investment stocks per industry are recorded for eight major ISIC4 industries. The International Standard Industrial Classification of All Economic Activities (ISIC) is a united nations industry classification system, which classifies data according to economic activity. The appendix II shows the full list of the different ISIC industry classification. The eight different industries used in this paper are the manufacturing industry, the agriculture, forestry and fishing industry, mining and quarrying industry, electricity, gas etc industry, water supply and waste management industry, the construction industry and the services industry. Resident Special Purpose Entities are excluded. The data for the level of inward FDI stocks is for 88 percent complete.
Due to the incompleteness and limited availability of data, only eight major ISIC4 industries are part of the dataset. These eight major industries make up the majority of inward FDI stocks.
The inward foreign direct investment stocks per industry are collected for the years 2013 until 2017 for the following 25 countries:
TABLE 2
COUNTRY COUNTRY CODE AUSTRALIA AUS
AUSTRIA AUT
CANADA CAN
CZECH REPUBLIC CZE
