外商直接投資

外商直接投資
Foreign direct investment

外商直接投資（以下簡稱外商直接投資）是當今全球經濟高度關注的重要組成部分之一。在過去幾十年中，外國直接投資對發展中國家的巨大增長已經引起了經濟學界和金融界的巨大興趣。已經進行了大量的研究，目的是實證研究什么促使企業參與跨國投資和激勵國家進行不同的政策改革和其他措施，追求吸引外國直接投資。外商直接投資的研究表明，外商直接投資可以從不同的方式提高受體國家的發展，即使外國企業不提供外部性。外國直接投資對接受國的好處不明確。外國直接投資被視為一個國家的經濟困境的解決方案，提供最需要的外國資本，促進東道國的經濟活動。根據Goldberg（2000），外商直接投資導致外國企業的技術和其他技能向當地企業轉移。這是通過外國直接投資，如資金，管理，技術和人才等，提供給東道國的補充資源。這些資源可以刺激東道國的現有的經濟活動，鼓勵內部競爭，提高國家產出水平。

外國生產商主要是認為有利于東道國公民通過引入各種優質的質量和更低的價格的新產品進入國內市場。最重要的是,外國直接投資是一個受援國的渠道進入國際金融市場,賺取外匯。作為外國實體在整個東道國在不同地點和相對較高的工資率建立業務單元，外國直接投資創造了大量的就業機會。向后聯系和溢出效應在外商直接投資接受國享有的好處中是次要的。

CHAPTER ONE: INTRODUCTION——第一章:介紹

1.1 INTRODUCTION——簡介

Foreign Direct Investment (henceforth FDI) is one of the most important integral parts of todays highly talked about global economy. The enormous growth of FDI towards developing countries over the past few decades has ignited a huge interest from researchers in both economics and finance fields. A number of studies have been undertaken with the aim to empirically examine what motivates firms to be involved in cross-border investments and what motivates countries to undertake different policy reforms and other measures in pursuit of attracting FDI. There is a consensus among FDI researchers that FDI can improve the recipient country's development in various ways even when foreign firms do not provide externalities. The benefits of FDI to recipient countries are not ambiguous. FDI is seen as a solution to a country's economic woes by providing the most needed foreign capital that boosts the economic activities of a host country. According to Goldberg (2000), FDI leads to transfer of technology and other skills from foreign firms to local firms. It is through FDI that supplementary resources such as capital, management, technology and personnel become available to host countries. These resources may stimulate existing economic activities in a host country, encourage internal competition, and raise the level of national output.#p#分頁標題#e#

The presence of foreign producers is primarily believed to benefit the host country's citizens by introducing a variety of new products into the domestic market, which are of superior quality and lower prices. Most importantly, FDI is a channel through which recipient countries gain access to international financial markets and earn foreign exchange. FDI creates a number of employment opportunities as foreign entities establish business units in various locations throughout the host country and relatively higher wage rates are offered. Backward linkages and spillovers are secondary benefits of FDI enjoyed by the recipient countries. Spillovers spur strong growth in industries into which FDI flows, especially when the competition between domestic and foreign firms is efficient. Foreign firms also go into joint ventures with domestic firms and a large percentage of profits generated through such collaborations are ploughed back into the domestic market, thereby contributing to the host country's macroeconomic growth and development without necessarily providing externalities. Externalities provide another form of benefits that FDI recipient countries enjoy. The existence of foreign direct investors spawns the seepage of managerial, personnel, and technological expertise from the foreign to domestic companies. For instance, Old Mutual plc training programme in South Africa may benefit the South African insurance and financial sectors as a whole.

It is because of the above-mentioned benefits that developing countries are actively embarking on measures involving macroeconomic as well as socio-political reforms with clear intentions of advancing their investment climate to attract FDI on a large scale and achieving sustained economic growth. Developing countries have formerly depended on loans and official development assistance as a source of foreign capital, principally provided by international agencies such as the World Bank and OECD countries. However, the flow of such funds from these institutions has been declining. For instance, Asiedu (2002) reports that financial assistance to Sub-Saharan Africa (SSA) fell from 6% in 1990 to just 3.8% in 1998 and foreign aid per capita fell from an average of $35 for the period 1989-92 to $2 for the period 1993-97. It is in situations like this, that FDI plays a pivotal role as an alternative source of foreign capital for the developing world.

For developing and the least developed countries that are making efforts to attract FDI as a way to enhance their economic growth and hence their sustained development, it is particularly essential to identify and comprehend the prime factors that shape FDI inflows as they apply to each country in particular. Since FDI plays a pivotal role in the growth dynamics of a country, a number of factors that are believed to influence FDI inflows towards developing countries have been intensely investigated. Among these factors the following have been most frequently considered: exchange rate volatility, market size, GDP growth, trade openness, infrastructure development, country size (also size of economy), per capita GDP, quality of the labour force, labour cost, inflation, return on capital, export orientation, political stability. Such analysis has immediate policy relevance as it identifies areas of comparative advantage that these countries should favour in terms of resource allocation.#p#分頁標題#e#

1.2 OBJECTIVES OF THE STUDY——研究的目標

The main aim of this study is to carry out an empirical investigation of the factors explaining FDI inflows to South Africa over the period 1980-2003. The review of previous theoretical research and the review of previous empirical evidence are means to this end.

1.3 THE STRUCTURE OF A PAPER——一篇dissertation的結構

The structure of this paper is as follows: Chapter 2 provides a review of both the theoretical and empirical literature on factors believed to be the major driving force behind inward FDI activities for host countries. In Chapter 3, the paper provides an overview of the South Africa's macroeconomic performance, FDI regulatory framework currently in place and incentives provided by the government to foreign investors. Chapter 4 discusses the data set and econometric methods used to carry out time series analysis for the study. The data and variable specifications are also described and clarified in this chapter. Chapter 5 reports and discusses the econometric results. A summary of findings, conclusions and policy implications are presented in Chapter 6.

CHAPTER TWO: REVIEW OF FDI LITERATURE——第二章:外國直接投資方面的文獻

2.1 INTRODUCTION——簡介

The main aim of this chapter is to present theories of FDI as developed by previous research and review the empirical evidence on the determinants of FDI. Before these are considered in more details, a brief overview of the definition of FDI and related concepts is provided. Given that FDI has direct effects on the economic growth of the host country, a specific section provides a brief exploration of the relationship between FDI and growth.

2.2 FDI: DEFINITION AND CONCEPTS——外國直接投資:定義和概念

In its archetypal form, FDI is conventionally defined as the physical investment made by acquiring foreign assets such as land and factory buildings with operational control residing with the parent company (Buckley, 2004). The definition can be extended to include such investments that seek to exercise considerable influence on the management of the foreign entity. A parent company is required to hold at least 10% of the ordinary shares or voting rights in order to exercise control over an incorporated foreign company. An ownership stake of less than the stated 10% is regarded as foreign portfolio investment and does not qualify as FDI.

FDI made by transnational companies (TNCs) is an indication of internal growth and it can be made in the form of greenfield investment or mergers and acquisitions. Greenfield FDI refers to an investment where a new entity is established in a foreign location. It entails formation of completely new production facilities in the recipient country (Eun and Resnick, 2007). Cross-border mergers and acquisitions are twofold. On one hand, they involve merging both domestic and foreign companies into one bigger company. On the other hand, they involve an acquisition of a domestic company by a foreign company. FDI is beneficial for the host country as it is a channel through which foreign capital and new technology are provided. FDI is regarded as a stimulus for economic activities and accelerated growth. The OECD (2002) describes FDI as a catalyst for speeding up the development process. Nevertheless, a recipient country must display a certain standard of development before it attracts FDI. Dunning (1977, 1979) identifies fundamental factors including firm specific and host country specific advantages that must be met before FDI occurs. These advantages are discussed in detail below.#p#分頁標題#e#

2.3 FDI AND ECONOMIC GROWTH——外商直接投資和經濟增長

A number of empirical studies have analysed the relationship between FDI and the economic growth of a host country. Lipsey (2000) finds that FDI and economic growth are positively correlated. Abwona (2001) urges that the effects of FDI on growth may vary between countries, as not all countries are at the same level of development. Lim (2001) points out that FDI positively affects economic growth of a host country by ''transferring advanced technology from the industrialized to developing economies'' (Lim, 2001, p.3). Conceptually, FDI increases GDP growth because it increases the amount of goods and services produced in a host country. Benefits of FDI are only evident in increased level of output because of the host country's ability to absorb technological spillovers from foreign firms. This, therefore, suggests that the positive contribution of FDI on growth is conditional upon the recipient country's absorptive capacity of all the benefits that FDI brings.

Much as there is general awareness that FDI influences economic growth, there is no general agreement on the causal relation between FDI and economic growth. De Mello (1997) argues that the relationship could run either way, as the prospects of economic growth make the host country more attractive to FDI. Once operational in a host country, FDI enhances growth by allowing the host country to integrate new inputs and technology to expand production.

2.4 THEORIES OF FDI——對外直接投資理論

In a study of TNCs, Hymer (1976) explains that because foreign TNCs have offsetting monopolistic advantages over domestic companies, they are able to compete with domestic companies that are in a better position with regard to knowledge and understanding of the domestic market. Kindleberger (1969) who suggests that these advantages should be adequate to tower over limitations and must be company specific reiterates this. According to Hymer (1976) and Kindleberger (1969), these advantages can be in the form of access to ownership patents, technological expertise, managerial expertise, marketing skills, etc. These skills should be scarce or completely unavailable to domestic companies. The basis of the argument here is on the theory of market imperfections in factor markets and product differentiation. In circumstances where market imperfections exist, firms find it rewarding to engage in cross-border direct investment instead of exporting to foreign markets or licensing. In this way, they can fully utilize their monopolistic market supremacy (Assefa and Haile, 2006).

In the same year as Hymer (1976), Buckley and Casson (1976) developed the internalization theory. This theory stipulates that in some instances it is desirable for TNCs to refrain from licensing and to choose cross-border direct engagement over exporting. FDI occurs when TNCs undertake an internal operation rather than a market operation. TNCs internalize their activities to circumvent impediments presented to them by the external market. These impediments arise because of market flaws such as lack of managerial expertise, human capital, etc. The benefits of internalization include both time and cost savings. According to Moosa (2002), the rationale behind internalization is the persistence of externalities in both goods and factor markets. However, Rugman (1980) disapprove of the internalization theory by contesting that it cannot be empirically analysed. Twelve years after introducing internalization theory, Buckley (1988) came back to warn that the theory cannot be directly analysed, claiming that the theory needs to be modified in order to allow rigorous analysis. He urges that there is still a room to develop advanced theories that can be empirically tested.#p#分頁標題#e#

Another line of study dealing with factors influencing FDI is based on Vernon's (1979) product life cycle hypothesis. Assefa and Haile (2006) and Udo and Obiora (2006) relate their studies to Vernon's product life cycle to explain that FDI is a stage in the life cycle of a new product from innovation to maturity. Home production is unique and strategic for some time, after which the new product reaches maturity and looses uniqueness. New similar products also enter the market and intensify competition. At this stage, the firms would then replicate the home production in lower cost foreign locations that offer cost benefits to the firm. The lower cost that can be achieved by producing in foreign locations can be due to cheaper factors of production and complementary government policies.

Dunning (1993) develops an eclectic theory which is referred to as OLI framework. OLI is the short name for ownership, location and internalization advantages. In his theory, Dunning identifies three sets of advantages that must be met for a firm to get involved in foreign direct investment. The first set is ownership advantages, which entail technological expertise, patents, marketing skills, managerial capabilities and the brand name. A TNC must have these firm specific advantages over its rivals in a foreign location. In the absence of these advantages, the firm will be exposed to fierce competition from its rivals in the market it serves. The second set of advantages that must be met is the location advantages, which Dunning (1993) explains as the degree to which the foreign location is more favourable to invest in. Examples of these advantages are an abundance of natural resources, exceptional infrastructure development, political and macroeconomic stability. These advantages should be adequate to validate investment in a preferred foreign location. The third set of advantages is internalization benefits. The ability of a TNC to internalize its operations is the manner in which it enters foreign location. According to Dunning (1993), this could be by a greenfield project, product licensing or acquisition of foreign assets as long as it fits the management strategy, the nature of the firm's business and the firm's long-term strategic plan.

Another rationale for FDI to occur is embodied in the in the industrial organization hypothesis (IOH) (Tirole, 1988). This hypothesis presupposes that there are various potential uncertainties that a TNC faces in foreign markets. The uncertainties may be political, religious, social, cultural and so on. If the firm decides to establish a subsidiary in a foreign location despite these uncertainties then the benefits accrued should be adequate to outweigh these obvious risks and restrictions. Lall and Streeten (1977) emphasize the importance of marketing and managerial skills, availability and ownership of capital, production technologies, scale economies and access to raw materials. They also put forward that FDI occurs because of the complexity of trading intangible assets abroad. Examples of these untradeable intangible assets include a TNC's organizational ability, executive skills, position in foreign financial markets and a well-established network with different government bureaucrats.#p#分頁標題#e#

2.5 MOTIVES FOR FOREIGN DIRECT INVESTMENT——為外國直接投資的動機

Assefa and Haile (2006) assert that the ownership and internalization advantages as developed in Dunning (1993) eclectic theory are firm specific advantages, while location advantages are regarded as host country qualities. Firms choose locations where all these advantages can be combined together to advance the firms' long-term profitability. Asiedu (2002) and Dunning (1993) distinguish the motives of FDI as either market seeking or non-market seeking (efficiency and resource seeking). According to Dunning (1993), a market seeking FDI is that which aims at serving the domestic and regional markets. This means that goods and services are produced in the host country, sold and distributed in the domestic or regional market (Asiedu, 2002). This kind of FDI is therefore, driven by host country characteristics such as market size, income levels and growth potential of the host market and so on. A non-market seeking FDI can either be classified as resource/asset seeking and/or efficiency seeking. Resource seeking FDI aims at acquiring resources that may not be available in the country of origin.

Such resources may comprise natural resources, availability and productivity of both skilled and unskilled labour forces as well as availability of raw materials. Efficiency seeking FDI aims at reducing the overall cost of factors of production especially when the firm's activities are geographically scattered (Dunning, 1993). This allows the firm to exploit scale and scope economies as well as diversify risks. Apart from the economic factors that are believed to be the major motivation for FDI, the host country's FDI policy also plays a major role in attracting or deterring FDI. This therefore, suggests a need for the host country to develop policies that provide a conducive environment for business if the authorities believe in the benefits of FDI. This necessitates a regular monitoring of the activities of TNCs and an acceptance by the host government that, if FDI is to make its best contribution, policies that were appropriate in the absence of FDI may require amendments in its presence. For example, macroeconomic policies may need to be altered in order to provide a favourable climate for FDI. Stronger competition as a result of FDI may also induce a host government to operate an effective and efficient competition policy.

2.6 EMPIRICAL LITERATURE——實證文獻

There is an extensive empirical literature on the determinants of FDI. A large share of this literature focuses on the pull factors or, equivalently, on the host country's location advantages. Given the increasing flows of FDI towards developing countries, especially in Asia, most academic researchers have been poised to investigate what factors influence flows of FDI into those countries. The African continent instead remains under researched, especially a country such as South Africa, which is regarded by many as one of the economic giants of the continent. The following studies have attempted to examine the link between FDI and predictor variables such as market size, GDP growth, inflation, exchange rate, political stability and many more.#p#分頁標題#e#

Goldar and Ishigami (1999) use panel data techniques to empirically analyse the determinants of FDI for 11 developing countries of East, Southeast and South Asia for the years 1985-1994. The authors estimate two separate models, one for Japanese FDI and the other for total FDI flowing towards these countries. They report a strong positive relationship between the size of the economy and FDI inflows for the two models. They also find a positive relationship between FDI inflows and the exchange rate. Intuitively, this suggests that lower exchange rates should make the host country's exports more competitive in foreign markets and therefore act as an important factor for attracting FDI. Interestingly, Goldar and Ishigami (1999) report contrasting results for the relationship between FDI inflows, domestic investment and trade openness for the two models. They find both domestic investment and trade openness to be significant determinants for the Japanese FDI model but the authors fail to find support for the two variables for the total FDI model.

Market size has generally been accepted as an important factor influencing FDI by many empirical studies. In a cross-country empirical study, Chakrabarti (2001) finds strong support for market size as an important determinant of FDI. He further reports that the relationship between FDI and other explanatory variables, such as trade openness, tax, wages, the exchange rate, the GDP growth rate and the trade balance is highly sensitive to small changes in these variables. That is, a small change in these variables is likely to deter or increase FDI inflows with a large magnitude.

Campos and Kinoshita (2003) use two trade related variables to examine the extent to which trade openness influences FDI inflows for 25 countries in transition between 1990 and1998. The authors employ an external liberalization index and trade dependence as proxies for trade openness. They argue that, the greater degree of trade openness does not only increase international trade, but also increases FDI inflows. Chang et al. (2009) also study the importance of trade openness in the economic growth of a country. They use a sample of 82 countries from all over the world for the period 1960-2000. They conclude that trade openness, used in conjunction with complementary trade policies, enhances economic growth through increased FDI inflows into a country. Their finding further bolsters the position of trade openness as an important determinant of FDI.

Root and Ahmed (1979) seek to analyse the determinants of manufacturing FDI in 58 developing countries for the period 1966-1970. They classify these determinants under three categories: economic, social and political. Among the four important economic variables they study, infrastructure, per capita GDP and GDP growth rate appear to be important predictor variables while the absolute of size of GDP does not help to predict FDI. They also put forward that foreign investors view long-term political stability and the extent of urbanization as important factors when choosing the location for their investments.#p#分頁標題#e#

Nunnenkamp (2002) adopts Spearman correlations to study the relationship between FDI and its determinants for 28 developing countries between 1987 and 2000. His findings show that variables such as GDP growth rate, entry restrictions, post-entry restrictions, market size, infrastructure and quality of the labour force, as measured by the years of schooling, have no effect on FDI inflows. However, he also reports positive effects on FDI of non-traditional factors such as factor costs. That is, the cheaper costs of factors of production, like lower costs of raw materials and lower costs of labour, are essential for attracting efficiency-seeking FDI. Tsai (1991) presents an opposing view by arguing that in Taiwan FDI inflows have increased with increasing labour costs. His findings suggest that in Taiwan, there are far more important determinants of FDI than cheaper costs of factors. Indeed Tsai (1991) observes that for the years 1965-1985, Taiwan's economic performance was spectacular, with an expanding domestic market and purchasing power of the economy as measured by rising GDP per capita.

Singh and Jun (1995) look into three determinants of FDI in developing countries. They pay particular attention to socio-political stability, favourable business operating conditions and export orientation. They employ other macroeconomic variables as control factors. The authors use a political risk index (PRI) developed by Business Environment Risk Intelligence, S.A (BERI, 2009), as a measure of socio-political instability. When used in conjunction with traditional determinants of FDI such as GDP growth rate and per capita GDP, PRI appears to be highly associated with FDI for 31 developing countries for the period 1970-1993. Schneider and Frey (1995) also find political instability to be a significant deterrent of FDI for 54 less developed countries, considered over three different years 1976, 1979 and 1980. Furthermore, Singh and Jun (1995) employ an operation risk index (ORI) also developed by BERI, S.A, and taxes on international trade and transactions (ITAX). They further find that ORI is highly associated with FDI flows and has a positive sign. ITAX is also reported as an important determinant of FDI. They also find export orientation to be highly related to FDI flows, especially manufacturing exports.

Asiedu's (2002) investigation on the determinants of FDI in 71 developing countries provides further evidence of the importance of these extensively studied FDI determinants. Asiedu goes a little further to assess whether the same determinants of FDI have a similar impact on FDI flowing towards Sub-Saharan Africa (SSA) between 1980 and 1998. After a detailed analysis, she concludes that factors influencing FDI in other developing countries do not necessarily have a similar impact on FDI flows to SSA. She notes that infrastructure development and a high return on capital influence FDI flows towards developing countries in other continents, but the same factors have no influence on FDI flows towards SSA.#p#分頁標題#e#

Bende-Nabende (2002) also undertakes a study probing into the determinants of FDI flows towards 19 SSA countries for the period 1970-2000. In this study, market growth, trade liberalization and export orientation turn out to be important factors explaining why countries are able to attract foreign investors. Obwona (2001) argues that the Ugandan market size has been one of the driving forces behind FDI flows into the Ugandan economy between 1975 and 1991. Yasin (2005) uses panel data from a sample of 11 SSA countries for the period 1990-2003. His findings lead to the conclusion that official development assistance (ODA), trade openness, growth of the labour force and the exchange rate are important determinants of FDI, while his findings do not suggest that the growth rate of per capita GDP, a political repression index and a composite risk index have explanatory power.

Ahmed et al. (2005) examine the composition of capital flows between 1975 and 2002, to assess if South Africa is different from 81 other emerging markets. They classify relevant factors into macroeconomic performance, investment environment, infrastructure and resources, quality of institutions, financial development and global factors. The authors argue that some factors influencing capital flows are limited to particular forms of capital, while other factors have comparable effects on FDI, bond and equity flows. They further assert that the more open the economy is to international trade, the richer it is in natural resources, and the fewer barriers to profit repatriation it has, the more likely it is to attract FDI on a large scale. The implication of this study is that South Africa should ease its capital controls in order to avoid deterring FDI. South Africa is currently allowing repatriation of profits only within six months (Ahmed et al., 2005).

The agglomeration effect is another important factor that can be used to explain why countries that have attracted FDI in the past still receive a large share of it. Notable magnitudes of existing FDI stocks in recipient countries tend to attract more FDI inflows (Lim, 2001). Foreign investors incorporate the size of the existing FDI stocks in their decisions as they seek locations where to expand their operations. Campos and Kinoshita (2003) also confirm that for 25 countries in transition between 1990 and1998, foreign investors tend to locate where others are investing, as the decision by other firms implies a favourable atmosphere for investment offered by counties in which they locate.

Asiedu (2003), Loree and Guisinger (1995) and Lee (2005) consider effective government policies, especially the monetary and fiscal policies, as important factors that can be used to effectively attract or deter FDI inflows. Loree and Guisinger (1995) use effective tax rates for 48 countries (classified into developed and developing countries) in 1977 and 1982. After finding the 'effective tax policy variable' to be significant, they conclude that government policies are important and that policy reforms are likely to attract more FDI. Lee (2005) investigates some evidence on the effectiveness of policy barriers to FDI for 153 developing and developed countries between 1995 and 2001. He finds strong support for the proposition that restrictive public policies deter FDI inflows. Asiedu (2003) studies the effects of investment policy on FDI for 22 SSA countries over the period 1984-2000. Her findings show that governments can increase FDI flows by developing and implementing policies that provide investor friendly environment.s#p#分頁標題#e#

CHAPTER 3: SOUTH AFRICA'S MACROECONOMIC OVERVIEW, FDI REGULATORY FRAMEWORK AND INCENTIVES.——第三章:南非的宏觀經濟概況,外國直接投資的監管框架和激勵

INTRODUCTION——簡介

The purpose of this chapter is to provide an overview of the South African macroeconomic performance over the period of study (1980-2003). This has been based on two regimes – the apartheid regime and the democratic regime. Other sections of the chapter are dedicated to the regulatory framework for FDI and the incentives offered to foreign investors by the South African government.

3.2 THE APARTHEID SOUTH AFRICA——南非種族隔離

According to Lester et al.(2000), the pre-1994 period was one during which South Africa was politically unstable as a result of the racial discrimination philosophy (apartheid) that was adopted by the Nationalist government then in power. This created internal resistance by 'black' South Africans, who protested against the racial discrimination practices. In 1960, the Nationalist government banned organizations such as the African National Congress (ANC) that represented 'black' South Africans and this intensified demonstrations against the government. In response to the protests, the Nationalist government using armed forces, killed and arrested many 'black' South Africans activists. These incidents drew attention from the international community, who condemned the repression against 'black' South Africans by a Nationalist government that predominantly looked after the interests of 'white' South Africans. These events also caused distress among foreign investors, who became sceptical of the prospects of South Africa's economic stability and the protection of their investments. This led to capital outflow, a mounting pressure on the South African rand in foreign exchange markets, effective economic sanctions and the isolation of South Africa from the rest of the world.

Despite being blessed with an abundance of natural resources, especially gold, diamonds, platinum and other minerals, South Africa was not a favoured location for investment during this period. Political instability, investment insecurity and violation of human rights severely inhibited FDI. Realizing the challenges that came with economic isolation, the Nationalist government, in an attempt to uphold domestic capital growth, introduced incentives for 'import substitution industries such as car manufacturers and military equipment' (Lester et al., 2000, p.187).

Despite these attempts, this period saw South Africa undergoing falling investment, diminishing international reserves, sinking economic growth, soaring rates of inflation and lofty interest rates (Nowak, 2005). During this period, the economic performance was notably disappointing as investors left South Africa in search of other locations that offered better conditions for their investments. As shown in Table 3.1, the GDP growth rate was declining in some years with an average annual growth estimated at 1.44% per annum for the years 1980-1990. In addition, inflation and interest rates were continuously increasing with no signs of stabilising. All these were signals of loss of effective control of macroeconomic policies.#p#分頁標題#e#

Table 3.1: GDP growth, inflation and interest rates in South Africa for the period 1980-1990

Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
GDP annual % change
5.4
-0.4
-1.9
5.1
-2.2
0.0
2.1
4.2
2.4
-0.3
-1.0
Inflation, annual % change
15.6
15.7
14.4
12.6
11.2
16.2
18.8
16.2
12.9
14.5
14.3
Interest rate annual %
4
10
17
14
20
18
11
10
14
19
19

3.3 DEMOCRATIC SOUTH AFRICA——民主的南非

This period marked the birth of a democratic system in South Africa. The general elections held in 1994 resulted in the replacement of the Nationalist government by an ANC-led government. On assuming power, the ANC government had effectively inherited a disheartening bequest of economic and social disparities left by the apartheid regime that called for immediate attention (Lester et al., 2000). This meant that the ANC had to commit itself to the redistribution of the economic benefits that were concentrated in the hands of few 'white' citizens. For this to be achieved, the government chose to develop and implement policies that would assist in attracting FDI on a large scale to stimulate economic growth, with an emphasis on empowering those who were previously disadvantaged.

In 1994 when the ANC came to power, it dumped the inward looking policies that had been implemented by the apartheid government to protect businesses owned by 'white' South Africans. The ANC-led government instigated a number of reform processes in order to alter the direction of the economic approach of the period before 1994. These reforms entailed a significant reduction of trade barriers, which aided the liberalization of South African markets, accelerated the integration of the South African economy with the rest of the world and promoted an expansive participation of the private sector in the development of a non-racial economy (Nowak, 2005). Stabilizing the South African economy was particularly important for the new government because of its keen interest in attracting FDI inflows, which were regarded as strategic for the promotion of economic growth. In its bid to achieve wealth redistribution and the elimination of socio-economic inequalities, in 1994 the ANC government launched the Reconstruction and Development Programme (RDP), a people-oriented programme that was aimed at redeveloping the self-confidence of the 'black' and other races that had been previously disadvantaged (Lester et al., 2000). According to Lester et al. (2000), goals of the RDP included provision of basic services (housing, electricity, water, health care), job creation through public works, land reform, social security and welfare, development of human capital through education and training, democratization of state and society, thus building an economy that would address economic inequalities and promote urban and rural development.#p#分頁標題#e#

Despite successes in certain areas, the RDP drew criticism from a wide spectrum of analysts and was terminated in 1996, when the government adopted a more comprehensive macroeconomic strategy that came to be known as Growth, Employment and Redistribution (GEAR) strategy. The GEAR strategy entailed liberal policies that were internationally accepted (Lester et al., 2000). The GEAR also played a pivotal role in integrating the South African economy with the rest of the global markets, as increasing the participation of the private sector was one of its pillars.

Ever since 1994, South Africa has seen positive changes in its real GDP. As shown in Table 3.2, economic growth during the period 1994-2003 improved substantially, with an average estimate of 3% per annum, which is twice the growth rate registered between 1980 and the early 1990s (Nowak, 2005). However, growth in South Africa has been slow due to low levels of public sector investment and limited private sector investment. While it is acknowledged that there have been fluctuations in the growth rate, it is worth noting that, growth has been positive ever since 1994. The rate of inflation dropped from 14% in 1990 (see Table 3.1) to just below 6% in 2003. Per capita GDP has also been growing, implying that on average South Africans were getting richer, although the benefits from economic growth were still concentrated in the hands of few South Africans.

Table 3.2: GDP growth, per capita GDP and inflation in South Africa for the period 1994-2003

Year
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
GDP per capita, (in Rands)
11272.86
12480.13
13884.03
15317.88
16633.03
17538.77
18599.18
20337.27
22070.24
24659.99
GDP annual % change
1.2
3.2
3.1
4.3
2.6
0.8
2
3.5
2.7
3.6
Inflation rate annual %
9.9
8.8
8.7
7.3
8.6
6.9
5.2
5.4
5.7
5.8
A decline in the rate of inflation can be attributed to a more disciplined monetary policy. In 2002, the South African Reserve Bank (SARB) formally moved from money supply targeting to an inflation targeting policy (Ricci, 2005). Since then, inflation has been contained within the range of 3-6%. According to Fisher's effect, lower levels of inflation imply lower interest rates (Buckley, 2004). Nowak (2005) also notes that lowering inflation in South Africa has been successful mainly because fiscal policy has been developed in a manner that supports the anti-inflation drive by the SARB and the government budget deficit has been reduced significantly, without resorting to borrowing from the SARB. The exchange rate was also under intense pressure due to low levels of international reserves and this forced the SARB to interfere with the currency market in order to influence the value of the rand (Nowak, 2005). The main concern here was the belief that a weakening currency could affect the value of GDP. However, the SARB stopped intervening in the foreign exchange market in 1998 and the rand was set free to float unreservedly against other major currencies. Both the inflation targeting and floating exchange regimes have kept inflation and exchange rates under control without retarding economic growth.#p#分頁標題#e#

3.3 THE FDI REGULATORY FRAMEWORK IN SOUTH AFRICA——在南非的外國直接投資的監管框架

As was discussed earlier, the ANC led government made a remarkable progress in breaking down old economic policies that were based on anti-competitive practices and broad government intervention in the economy. The ANC government welcomes FDI into South Africa. It has taken major steps towards reducing the role of government in the economy because it has a keen interest in promoting and ensuring the competitiveness of private sector investment and hence attracting investment from foreign investors. The current FDI regulatory framework in South Africa provides foreign investors with a choice of mode of entry into South African markets. TNCs are allowed to enter the market through greenfield investments or through mergers and acquisitions. The latter have to be in accordance with the South African Company Act of 1973 that regulates the formation, conduct of affairs and liquidation of all companies. The Act makes no distinction between domestic and foreign companies. Foreign investment is welcome in almost all business sectors of the economy, with few having set limits to allow a certain degree of foreign participation. One example of this is the Financial Institution (Investment Funds) Act No. 39 of 20 March 1984, which allows a limited foreign equity investment in financial institutions other than banks (Department of Trade and Industry, 2009). Despite these restrictions, currently all investors (domestic and foreign) receive similar treatment. There is no discrimination made as to which investors have access to investment incentives such as tax breaks, export initiatives and trade regulations.

There has been an ongoing review of FDI regimes since 1994 when the ANC came to power in South Africa. Effective from 1st January 2001, South Africa's source–based tax system was changed to a residence–based system. As opposed to the period before 2001, when companies were required to pay local taxes on revenues earned in South Africa, the residence-based system requires South African-based TNCs to pay taxes on their worldwide income as well. Non-South African TNCs hosted in South Africa pay tax on earnings arising from South African sources, subject to double taxation agreements with other countries, especially with investors' countries of origin. Other reforms affecting foreign investment include an adjustments of tax brackets from R6 million to R14 million for qualifying small enterprises, a 100% tax exemption on research and development expenditure and many other tax allowances. The adjustment of exchange controls is another milestone achieved by the ANC government. In 2008, the government raised dividend (profit) repatriation from R0.75 million to R2million per shareholder. The government has also reduced its initial 50% requirement of shareholding in a foreign company investing in South Africa to 25%. This further bolsters the commitment of the government in promoting private investment and reducing its active role in the economy.#p#分頁標題#e#

Institutionally, the Department of Trade and Industry (DTI) has the responsibility of promoting, coordinating and facilitating overall direct investment activities in the country. The DTI offers a wide range of services to those interested in doing business in South Africa. Investment facilitation services offered by DTI include the provision of all necessary information on South Africa's sectors, consultation on South Africa's regulatory environment, facilitation of investment missions, links to joint ventures in South Africa, provision of information on incentive packages for investors, assistance with work permits and logistical support for relocation to South Africa (DTI, 2009).

The DTI is a mother body that consists of different agencies working together to enhance investment in South Africa. Trade and Investment South Africa (TISA) is one of the agencies that operate under the umbrella of the DTI to provide Industrial Development Zones (IDZs). These are purpose–built industrial estates that provide facilities and services customized for export-oriented industries. IDZs are linked to coastal or inland ports for exporters. TISA also provide special incentives for both domestic and foreign investors and facilitates policy development intended to establish an investor friendly climate.

3.4 INVESTMENT INCENTIVES OFFERED TO INVESTORS IN SOUTH AFRICA——在南非投資激勵措施提供給投資者

After years of isolation from international trade, the South African government has a sharp interest in promoting foreign direct investment in South Africa. The government offers a wide range of incentives to foreign investors who are keen to invest in South Africa. Some of the key incentives provided are discussed below.

INDUSTRIAL DEVELOPMENT ZONES (IDZs)——工業開發區(IDZs)

Apart from promoting the South African manufactured exports in the global markets, IDZs also provide outstanding industrial infrastructure to foreign investors. Other benefits of this programme include duty-free importation of intermediate goods and inputs, a VAT exemption on capital goods and other supplies purchased in a domestic market, links to international ports of entry and easy access to South African Revenue Services (SARS) for information and support.

DUTY CREDIT CERTIFICATE SCHEME (DCCS)——責任信用證書計劃(dcc)

The DCCS is an incentive programme available to textile and clothing manufacturers. The scheme was launched in April 1993 to induce global competitiveness of South African textile and clothing manufactures. Under this scheme, manufacturers receive no subsidies from the government but secure duty credits based on the value of their exports of certain prescribed textile and clothing products. A duty credit certificate allows the exporter an exemption from duty obligations of the value stated in a certificate.

MANUFACTURING INVESTMENT PROGRAMME (MIP)——制造業投資計劃(MIP)#p#分頁標題#e#

The MIP scheme launched in July 2008 aims at stimulating growth, employment and broadening participation in the South African manufacturing sector. The scheme is designed to provide a 15-30% capital allowance on the costs of plant and machinery for investments projects in excess of R5million. The scheme is available to both domestic and foreign investors who are keen to locate their operations in South Africa. The MIP incentive is for investment projects within the priority sectors as listed in South Africa's national industrial policy framework and for expansion or upgrading projects in the textile and clothing sector. The scheme is offered along with other incentives as determined by the relevant income tax legislation being implemented by the government to stimulate investment. Tax incentives include a 100% tax allowance applicable to those investing in the information technology sector, research and development activities on natural resources and engineering.

Other incentives include a higher customs duty charged on imported goods, in order to facilitate competition between goods produced within the Southern African Customs Union (SACU) and imported goods; a rebate of customs duty on intermediate goods that are not found within the SACU region and a tourism support programme, which carries similar principles and benefits as the MIP scheme, discussed above but customized for the tourism sector.

CHAPTER 4: DATA AND RESEARCH METHODOLOGY——第四章:數據和研究方法

4.1 INTRODUCTION——簡介

The objective of this paper is to carry out an empirical investigation of factors influencing FDI inflows to South Africa over the period 1980-2003. Therefore, this chapter discusses the econometric methodology adopted for this piece of work. Variables included in the study are specified and discussed in some detail, with particular attention being paid to their measurement and relative importance as postulated by previous studies. Data description and sources are briefly discussed. The model specification and the estimation methods are also covered in this chapter.

4.2 VARIABLE SPECIFICATIONS——變量的規范

DEPENDENT VARIABLE——因變量

In line with the reviewed literature, the dependent variable for this study is the ratio of FDI inflows to GDP. Buckley (2004) defines FDI as the physical investment made by acquiring foreign assets such as land and factory buildings, with operational control residing with the parent company. The definition is extended to include such investments that seek to exercise considerable influence on the management of the foreign entity. A parent company is required to hold at least 10% of the ordinary shares or voting rights in order to exercise control over an incorporated foreign company.

EXPLANATORY VARIABLES——解釋變量

The selection of variables for this paper is motivated by theoretical considerations, but unlike other empirical studies reviewed above, it is constrained by data availability. For instance, variables such as those representing political risk, operational risk, entry and post entry restrictions (trade barriers) are omitted in this study, as there is no data available on them for South Africa. Notwithstanding this drawback, the study makes use of numerous variables that are frequently selected in the literature as forces behind the attraction of FDI inflows.#p#分頁標題#e#

Macroeconomic Stability:
A stable economy acts as an assurance to foreign investors that their businesses will operate profitably in a foreign location. Exchange rate and inflation rate are frequently used as proxies in the literature to capture the effects of macroeconomic stability on FDI inflows. However, there has been a lack of general agreement across studies in terms of the nature of the relationship between the exchange rate and FDI inflows. For example, Chakrabarti (2001) argues that a weaker exchange rate is likely to deter inward FDI, while Goldar and Ishigami (1999) contend that a lower exchange rate improves the competitiveness of a country's exports in international markets, which would yield profits for foreign firms and thus increase FDI inflows in countries with weaker currencies. In turn, the effect of a lower inflation rate on FDI is always expected to be positive. High inflation rates may act as a hint to foreign investors that a host country is unable to manage its monetary policy efficiently. Asiedu (2002) makes the case that FDI will not occur in high inflation countries, as it creates uncertainty on investments. Inflation, based on the consumer price index (CPI), is included in this study as a proxy for macroeconomic stability. A negative association between inflation and inward FDI is expected.

Human capital:
Human capital is a location advantage that is expected to have positive impact on FDI inflows. A host country with an educated labour force that is skilled in the operation of recent production technologies should be able to attract a larger share of FDI. Foreign investors view labour market skills as an important factor when making location decisions, as this reduces the costs of training employees. Schneider and Frey (1985) use the ratio of an age group with secondary education, school enrolment and availability of technical and professional workers as proxies for human capital but find none of these variables to be significantly associated with FDI inflows. Noorbakhsh et al. (2001) find human capital to be a significant determinant of FDI inflows, and warn that developing countries that do not develop labour market skills may find it difficult to attract FDI over time. A variety of measures of human capital has been employed across empirical studies. For example, Root and Ahmed (1979) use the ratio of literacy and school enrolment, Nunnenkamp (2002) uses the number of years in school, while Noorbakhsh et al. (2001) use accumulated years in secondary and tertiary education. These measures vary across the studies, because it is hard to find a good measure of the level of education of a whole country. This study follows the existing literature and makes use of the literacy rate as a measure of human capital. The hypothesis that the literacy rate has a positive effect on inward FDI is maintained.

Infrastructure:
Host countries that display high quality infrastructure development should be able to attract FDI inflows on a large scale. Quality infrastructure can take a variety of forms, including sophisticated telecommunications connections, railway and road networks, to mention but a few. Asiedu (2002) and Ahmed et al. (2005) observe a positive relationship between infrastructure and inward FDI. These authors support the assertion that a good infrastructure development increases inward FDI flows. Therefore, a positive relationship between FDI and infrastructure, as measured by the number of main telephone connections per 1000 inhabitants is expected in this study too.#p#分頁標題#e#

Labour cost:
Compensation of employees is usually viewed as a critical factor that foreign investors take into account when deciding where to locate their investments. Existing flaws in labour markets are motivational for efficiency-seeking FDI and foreign investors will more often direct their investments to those locations where not only skilled but also cheap labour is available. Nunnenkamp (2002) finds a positive relationship between inward FDI flows and a lower industrial wage rate. In contrast, the results obtained by Tsai (1991) demonstrate that a lower labour costs may not always be a positive determinant of FDI, as in Taiwan inward FDI increased with increasing labour costs over the period 1965-85. These conflicting results demonstrate that the effect of labour cost on FDI inflows depends on whether FDI finances high-tech or low-tech industries. If FDI finances high-tech industries, the labour cost should probably be the last factor foreign investors worry about. This is because high-tech industries require a skilled labour force and high wages must be offered in order to attract skilled workers. However, if the investment goes into low-tech industries, then high labour cost is expected to deter FDI inflows. Despite this controversy, this study supports the hypothesis that a lower labour cost is attractive to FDI. Therefore, an inverse relationship between wage rates and inward FDI is expected.

Openness:
The more open a country is to international trade, the more it is likely to attract export oriented foreign investments. Asiedu (2002) empirically supports the role of trade openness by reporting that FDI increases with an increasing degree of trade openness. Singh and Jun (1995) also find strong support for trade openness as a significant force driving FDI inflows. Foreign investors consider the ability to move capital freely across borders while conducting their business transactions as an important factor when making location decisions. The ratio of exports to GDP has served as a measure of openness in some studies (see for example Asiedu, 2002), while other studies have used net exports as a measure of the export orientation of the economy (see for example Chakrabarti, 2001). This study also makes use of the ratio of exports to GDP and proposes a positive relationship between openness and inward FDI.

Market size:
The market size of a host country is an important factor that influences decisions of market-seeking foreign firms. Market size is a significant determinant of FDI that features in almost all empirical studies dealing with the determinants of FDI. Obwona (2001) supports the assertion that market size is an important factor influencing inward FDI by reporting that it has been a formidable force behind Uganda's inward FDI between 1975 and 1991. A large market allows foreign firms to use resources more efficiently and exploit economies of scale (Chakrabarti, 2001). Different proxies have been employed in the literature as measures of market size. Some studies employ the level of GNP and/or per capita GNP (for example. see Root and Ahmed, 1979 and Schneider and Frey, 1985) while others utilize GDP level and/or per capita GDP (see for example Yasin, 2005; Chakrabarti, 2001 and Tsai, 1991). For the purpose of this work, the absolute level of GDP is included to capture the effect of the market size. Thus, the study advocates a positive association between market size and FDI inflows.#p#分頁標題#e#

Natural Resource Endowment:
Availability of natural resources in the host country can increase the country's location advantage for resource-seeking FDI. The natural resource endowment hypothesis states that countries that are rich in natural resources should be able to attract a large share of resource-seeking FDI. Campos and Kinoshita (2003) support the potential of abundance of natural resources to play a crucial role in attracting FDI inflows. The authors report that abundance of natural resources has a significant positive effect on inward FDI. This study makes use of net gold exports as a measure of availability of natural resources in South Africa. Hence, a positive association between natural resources and FDI inflows is expected.

Political regime:
A dummy variable is included to capture the possibility of structural break due to a change in government regime. As previously discussed in Chapter 3, the South African economy was effectively isolated from that of the rest of the world during the apartheid regime. During this time, inward FDI declined. However, in 1994 when the ANC came to power and introduced reforms, the country started witnessing increasing flows of inward FDI. The dummy variable assumes the value of zero for the period before 1994 (the apartheid era) and one for the period after 1994 (the democratic era).

4.3 DATA DESCRIPTION——數據描述

This study utilizes secondary time series annual data covering the period 1980-2003. The determination of the sample size and the sampling frequency for this study is influenced by the availability of data in some of the variables included in the model. The data is obtained from three sources – the IMF International Financial Statistics, the UN database (ILO Labour Statistics Yearbook database and UNESCO UIS Literacy Statistics database) and the South African Reserve Bank. Specifically, data on adult literacy rate, labour cost, and number of main telephone line per 1000 people were obtained from the UN database; data on inward FDI flows and inflation rate were obtained from the IMF International Financial Statistics, while GDP and net gold exports came from South African Reserve Bank.

4.4 MODEL SPECIFICATION——模型設定

The existing literature presumes a linear relationship between the dependent and the explanatory variables such as those described in section 4.2 (see for example Chakrabarti, 2001). Thus, FDI is expressed as linear function of the inflation rate, natural resource endowment, market size, labour cost, human capital, trade openness, infrastructure and a dummy for the political regime. The model takes the following general form: ........... (1)

where;
FDI is inward foreign direct investment in constant prices measured as a ratio of real GDP;
INFR is the annual inflation rate based on consumer price index (CPI);
RES is net gold exports;
GDP is the real gross domestic product;#p#分頁標題#e#
WAGE is the labour cost based on manufacturing wages;
LIT is the adult literacy rate;
OPEN is a measure of openness to international trade;
TELE is the number of main telephone lines per 1000 people;
DUMMY is a dummy variable assuming value zero for the apartheid era (1980-1993) and one for the democratic era (1994-2003).

4.5 ESTIMATION METHOD——估算方法

Most of the empirical studies on the determinants of FDI, including those reviewed in Chapter 2 above, use panel data methods for the estimation of the regression coefficients. The reason for this, among others, is that many countries are usually included in the analysis and such estimation methods lay a solid foundation for making comparisons across countries (see for example Asiedu, 2002; Chakrabarti, 2001; Goldar and Ishigami, 1999 and Root and Ahmed, 1979). The approach in this study will be different from the approaches adopted in these earlier studies. This study uses time series analysis and formally takes on board an autoregressive-distributed lags (ARDL) model specification to explore how the selected explanatory variables influence inward FDI flows to South Africa over time.

Nelson and Plosser (1982) have shown that most macroeconomic variables are linearly related, and tend to move closely together over time. Variables such as GDP, GDP growth, infrastructure development, interest rates, foreign investments, inflation, wages and many more have trended together over time. This kind of behaviour displayed by macroeconomic variables generates integrated processes that affect the stationarity requirement of the underlying time series. Integrated processes are usually denoted by I(1) or I(2) in the literature, reflecting the order of integration. I(1) processes become stationary after differencing once, while I(2) processes become stationary after differencing twice. Two I(1) random processes are said to be cointegrated if there exists a linear combination between them that is stationary, denoted by I(0) ( Engel and Granger, 1987). If the linear combination between two I(1) processes is not I(0), then estimation by OLS will not yield consistent and efficient results, because the resulting errors are autocorrelated.

As argued by Granger and Newbold (1974), the use of non-stationary variables in regression analysis may produce spurious regressions that are characterised by high measures of goodness of fit (R2 and ) but with an extremely low Durbin-Watson statistic diagnosing the problem of autocorrelated errors. According to these authors, statistical estimations and inferences of models involving interrelated variables necessitate adoption of techniques that confront the problem of non-stationarity in the original series for the phenomenon under study. This is where the ARDL model comes in handy. The main advantage from using an ARDL model is the ability of such an approach to produce consistent and efficient results regardless of whether the underlying explanatory variables are integrated processes, stationary or mutually cointegrated (Pesaran and Shin, 1998). Another advantage is that ARDL makes it possible to explore both the long-run and the short-run dynamics of the phenomenon being studied. The empirical counterpart of the general model given by equation (1) is expressed as a linear regression model of the following form:#p#分頁標題#e#

where all variables are as explained in equation (1), is an intercept term, t represents a time trend, , i and ßj, i, j = 1......7 are unknown regression parameters, is a white noise error term with mean zero and finite variance, pi are respective specific optimum lag orders and is the first difference operator.

Since the number of observations used in this study is admittedly too small, mainly due to lack of data in some of variables, the study imposes one lag and then uses Schwarz's Bayesian Criteria (SBC) to choose the optimal lag length that enters the ARDL model given in equation (2). Imposing lags of higher order require sufficient observations, which is a drawback of this study. The argument put forward by Nelson and Plosser (1982) that economic variables are integrated and tend to trend together over time suggests that a deterministic trend should be included in equation (2).

Equation (2) allows for the 'bounds test' proposed by Pesaran et al. (2001) to be applied to test for the existence of long run cointegrating relations between FDI and its determinants. The authors argue that their test is more reliable than other standard tests for unit roots, as it does not require the explanatory variables to follow a certain order of integration. In this test, the null hypothesis of no long-run cointegrating relationship is tested against the alternative that there is at least one cointegrating relationship. Testing for no long run cointegration among the variables is equivalent to testing whether the long-run coefficients ( i's) in equation (2) are equal to zero. The null and associated alternative hypotheses are therefore formulated as follows:

.........................

Pesaran et al. (2001) have provided two sets of critical values for the test, one set assuming that all explanatory variables are I(0) and the other set assuming that the explanatory variables are I(1). The set assuming I(0) variables provides lower bounds while the other set assuming I(1) variables provides upper bounds for the test. The decision rule is as follows: if the value of the F-statistic is larger than the value of the upper bound, the null hypothesis given in (3) is rejected. If it is smaller than the value of the lower bound, the null hypothesis is not rejected and if it lies between the lower and the upper bound, the test is inconclusive.

CHAPTER 5: DISCUSSION OF ECONOMETRIC RESULTS——第五章:計量經濟學的討論結果

5.1 INTRODUCTION——簡介

This Chapter discusses the econometric results presented by tables 1 to 4 in the Appendix. After discussing stationarity, cointegration and structural stability tests, the Chapter goes further to discuss the meaning of the estimated parameters of the explanatory variables used in equation (2), what kind of association exists between them and FDI inflows - their significance and direction of relationship.

5.2 STATIONARITY TESTS——穩定性測試#p#分頁標題#e#

The first and perhaps the most important step to consider when dealing with time series analysis is to verify whether the variables used in a model satisfy the stationarity requirements or not. Verbeek (2008) asserts that a stationary time series alternates around its mean and has a finite variance that does not depend on changes in time horizon, while a non-stationary time series tends to drift away from its mean and has an infinite variance. He further explains that a stationary time series is inclined to return to its mean in the long run and has a dumpy memory of its past behaviour, whereas a non-stationary time series has an unlimited memory of its past with permanent effects on the process. Therefore, it is very important to investigate the properties of time series to ensure that the stationarity requirement is satisfied before carrying out any estimation. Failure to adhere to this principle may result in a spurious regression as discussed in Chapter 4 above.

To determine whether the variables are stationary or not, this study employs Dickey-Fuller (DF) and Augmented Dickey-Fuller (ADF) tests. In these tests, the null hypothesis of a unit root is tested against the alternative that the series is stationary. If the null hypothesis cannot be rejected, then the conclusion is drawn that the series is not stationary. Based on the results in Table 1, it is apparent that the null hypothesis cannot be rejected hence all variables, except trade openness, have a unit root in their levels. This is shown by the DF and ADF statistics that are greater than the critical values (shown in brackets) at the 5% significance level. The degree of trade openness variable (OPEN) is the only exception, which marginally rejects the null hypothesis by the ADF test, but not by the DF test. However, the null hypothesis is decisively rejected after differencing once, suggesting that all variables become stationary after first differencing. It should be noted that all the DF and ADF statistics (with and without trend) become insignificant for differenced variables. This intuitively makes sense, because differencing should eradicate the trend and probably induce stationarity in the original series.

5.3 COINTEGRATION TESTS——協整檢驗

To tests the existence of long-run cointegrating relations among the variables, this study follows the Pesaran et al. (2001)'s 'bounds test'. The null hypothesis of no long run cointegrating relationship is tested against the alternative of at least one long-run cointegrating relationship. This is tantamount to testing stationarity of the cointegrating relationship between two variables that are not stationary. The 95% critical value bounds from Pesaran et al. (2001) Case V and the F-statistic for the test are reported in Table 2 in the Appendix. As discussed in Chapter 4, the null hypothesis is rejected when the estimated F-statistic is greater than the upper bound; on the other hand, the null hypothesis is not rejected if the F-statistic is less than the lower bound. Based on the test results shown in Table 2, the null hypothesis is certainly rejected, because the F-statistic is much larger than the upper bound of the test. This suggests that FDI and the other variables used in the model have a long-run cointegrating relationship. As argued by Pesaran et al. (2001), this kind of a conclusive decision can be drawn without knowledge of the cointegration status of the variables.#p#分頁標題#e#

5.4 STRUCTURAL STABILITY TESTS——結構穩定性測試

This study anticipates a structural break due to government regime changes for the periods before and after 1994. A dummy variable is included in the model to account for the possibility of structural instability in the estimated FDI model. Brown et al. (1975) developed different methods for testing the structural stability of regression relationships in time series analysis. This study follows the cumulative sum (CUSUM) and cumulative sum of squares (CUSUMSQ) tests based on recursive residuals advanced by the authors to investigate the stability of the estimated regression of FDI on its determinants. The null hypothesis in these tests is that the underlying structure of the regression is stable/constant over time. The test is carried out by plotting CUSUM and CUSUMSQ of recursive residuals against time. If in both the CUSUM and CUSUMSQ plots the observed sample path lies within two critical bounds a decision is made that the underlying structure is stable; otherwise, the null hypothesis is rejected. Figures 1 and 2 in the Appendix provide a diagrammatical illustration of CUSUM and CUSUMSQ tests respectively. The critical bounds are shown as straight lines and are obtained from the functions provided by Brown et al. (1975). It is evident from both figures that the null hypothesis cannot be rejected, indicating structural stability in the regression of FDI on its determinants. Therefore, this does not support the inclusion of a dummy variable in equation (2).

5.5 LONG RUN COEFFICIENTS——長期系數

The long-run coefficients for the ARDL model are reported in Table 3 in the Appendix. As it can be seen from the table, inflation has a positive sign, implying a positive long-run relationship between inward FDI and inflation rate. This result is in contradiction with what the study had initially hypothesised. A negative association between these two variables was expected. The reported coefficient is 0.31895 and it is significant at the 10% level. This means a unit increase in inflation rate will increase FDI inflows by a small magnitude of 0.31895 units. Against this backdrop, one possible implication of this result is that, despite failing to control inflation, South Africa's macroeconomic policies may be doing well in other monetary items, promising a high degree of stability. Thus, the overall stability more than compensates the detrimental effect of inflation on inward FDI, so that the net effect of macroeconomic policies is to attract FDI and hence we observe simultaneously increasing inflation and increasing inward FDI.

The natural resource endowment variable has an unexpected sign and it is not significant. The result clashes with empirical evidence provided by previous studies, where natural resources have been found to act as one of the key factors attracting inward FDI (see Chapter 2). This finding is discouraging because South Africa is known as one of the world's biggest gold producers. The country is not only rich in gold deposits but also in other minerals, such as diamonds and coal, which it uses to produce fuel and electricity. The effect of the natural resource endowment variable on inward FDI was not only expected to be positive but also to be statistically significant. One possible implication of this finding is that South Africa's inward FDI may not be resource seeking, but may be motivated by other reasons. Nevertheless, this finding does not rule out the possibility that gold may not be the best proxy for general resource endowment of different kinds in South Africa. It may also be irrelevant for FDI going to South Africa or completely uncorrelated with other natural resources, which are instrumental in attracting FDI to the country. It should also be remembered that FDI is fundamentally a substitute for exports, so net gold exports may not work well as a proxy for natural resources.#p#分頁標題#e#

The sign of market size variable (GDP) is contrary to what was expected and it is significant at the 5% level. The negative coefficient of GDP implies inward FDI declines by 0.80834 units for every unit increase in GDP level. This finding is inconsistent with empirical evidence provided by previous studies on the determinants of FDI, but it may make sense in the case of South Africa, because its GDP growth had been declining and recording negative rates in some years over the period of study (see Chapter 3). A declining GDP growth means that GDP level is increasing at a declining rate, and this suggests that the market size is shrinking instead of expanding. Foreign investors prefer to locate their operations in a market that is forever growing as it provides better opportunities for profit generation. Since GDP does not only represent the market size, but also represents the economic wellbeing of a country, a declining GDP implies declining income for households, which eventually means no purchasing power for foreign investors' products. This would have direct effect on the profitability of foreign firms' operations and hence no incentive for investing in South Africa. In sum, it should not be surprising that this measure of market size has a negative long-run effect on inward FDI flowing to South Africa.

The results reveal that labour cost variable (WAGE) has no significant long-run impact on FDI inflows. This variable is one of the most controversial determinants of FDI, with some studies supporting the notion that lower labour costs have a positive effect on inward FDI (see for example Nunnenkamp, 2002) and others arguing that labour costs do not have any effect on FDI inflows (see for example Tsai,1991). As argued in Chapter 4, the effect of labour cost depends on the type of industry FDI goes to; hence, this finding suggests that FDI in South Africa goes into high-tech industries, where labour cost is deemed less important. When FDI goes into high-tech industries, foreign investors are willing to offer highly competitive wages in order to attract a skilled and productive labour force to their operations. As expected, the human capital variable, as measured by the adult literacy rate, has a positive sign and is significant at the 1% level. This variable does not only contribute positively towards FDI inflows, but has a larger effect on inward FDI than any other significant variable used in the model. The empirical result indicates that FDI inflows to South Africa would increase by 46.1192 units for every unit increase in the literacy rate. The result is consistent with findings of Noorbakhsh et al. (2001), who report human capital as a significant determinant of inward FDI flows for all developing countries they study.

The estimated long-run coefficient of the degree of trade openness (OPEN) is 0.58496, which has the expected positive sign and is significant at the 5% level. It indicates that in the long run, a unit change in degree of openness is followed by an increase of 0.58496 units of FDI inflows. This finding is also in line with empirical evidence provided by some earlier studies (Obwona, 2001; Asiedu, 2002 and Yasin, 2005). The contribution of infrastructure development, as measured by telephone lines per 1000 population, towards FDI inflows is found to be positive and significant at the 1% level. In comparison with most African nations and other new emerging markets, South Africa is at a more advanced level in terms of infrastructure development. Infrastructure development in South Africa includes a well-developed transportation network (road, air and railway) that facilitates easy movement of goods and services between coastal and inland ports of entry and different places within the country.#p#分頁標題#e#

5.6 SHORT RUN DYNAMICS AND ERROR CORRECTION REPRESENTATION——短期動力學和糾錯表示

Table 4 in the Appendix presents the short run terms of the model and its error correction mechanism towards long run equilibrium. The inflation rate, a greater degree of trade openness and advanced infrastructure development contribute positively towards increasing inward FDI flows in the short run and are significant at 1% level. On the other hand, market size and labour cost have negative short run effects on FDI inflows and are significant at conventional 5% and 1% levels respectively. The natural resource endowment and the human capital variables do not play any significant role in influencing inward FDI flows to South Africa in the short run.

The estimated coefficients imply that a unit increase in inflation, trade openness and infrastructure development will increase FDI inflows by 0.9136, 1.0095 and 8.4822 units respectively, whilst a unit increase in market size and labour cost will decrease FDI inflows by 1.3950 and 0.0839 units respectively. The error correction term (ECM) is significant and has the expected sign. This finding verifies the cointegration test explained in section 5.2 above and validates the conclusion drawn there from. Hence, this provides strong evidence of the existence of a long-run cointegrating relationship between FDI and its determinants. The ECM coefficient is greater than one, which implies that short-run dynamic adjustments towards long-run equilibrium take place instantaneously.

CHAPTER 6: SUMMARY AND CONCLUSIONS——第六章:總結和結論

The objective of this study was to investigate the effects of macroeconomic stability, natural resource endowment, market size, labour cost, human capital, trade openness, and infrastructural development on inward FDI flows to South Africa. A review of theoretical and empirical literature, South Africa's economic performance and FDI regulatory framework were means to this end. The study employs time series data covering the period 1980-2003 to test the significance of the aforementioned variables in explaining FDI inflows. The empirical results indicate that inflation (a measure of macroeconomic stability) has a significant positive impact on inward FDI flows to South Africa. Trade openness and infrastructure development and human capital (measured by adult literacy rate) also have significant positive effects on FDI inflows, while GDP level (a measure of market size) carries a wrong sign and appears to be significant at the 5% level. The natural resource endowment and labour cost (as measured by manufacturing wage) variables are statistically insignificant.

Therefore, the implication of these results is that inflation rate, trade openness, market size, infrastructure development and human capital significantly justify FDI activities by TNCs in South Africa. The empirical results of the inflation rate and the GDP are different from what was expected and contrary to what standard literature on determinants of FDI postulates. One possible cause of this may be lack of quality data. The significance of infrastructure development variable also suggests that TNCs may be taking advantage of well-developed infrastructure in South Africa since their businesses are more export oriented. The empirical results indicate that natural resources in South Africa do not make it an attractive destination for FDI. Development of policies that improve the country's attractiveness to resource-seeking FDI would be beneficial as South Africa is rich in natural resources. Drawing from the less important role played by labour cost in attracting FDI inflows, this study observes that wage demands in South Africa may be upward rigid and this does not help in increasing FDI inflows. Curbing trade unions powers and adjusting the minimum wage legislation would be helpful in luring efficiency-seeking FDI into the country. Finally, it should be noted that due to data limitations in this study, the empirical results of some variables look suspicious and unreliable. The study recommends that, if we are to understand FDI activities in South Africa, then good quality data must be continuously collected. Therefore, FDI remains a subject of interest for future research in South Africa.#p#分頁標題#e#