The effect of Chinese foreign direct investment on the economic growth in Africa

(1)

The Effect of Chinese Foreign Direct Investment on the Economic

Growth in Africa

UNIVERSITY OF AMSTERDAM

AMSTERDAM SCHOOL OF ECONOMICS

Bachelor Economics & Business

Bachelor Specialisation Finance & Organisation

Author:

R.T.M. Stevens

Student number:

10675248

Thesis supervisor: Dr. J.J.G. Lemmen

Finish date:

31-01-2017

(2)

ABSTRACT

The aim of this paper is to investigate the effect of the Chinese Foreign Direct Investment (FDI)

on the economic growth in Africa. First, a literature review is conducted in order to generate a

clear view and understanding of FDI that includes FDIs consequences relevant for this paper, the

Chinese and African economy, and the history of the economic relationship between China and

Africa. Next, two linear OLS regression analyses will be performed on 50 African countries

within the period of 2005 until 2014. The dependent variable of the first model is the economic

growth of Africa, represented by the change of GDP in a country. The dependent of the second

model is the level of GDP in Africa. If the chosen variables continue to cause a higher level of

GDP, they indirectly result in economic growth. Based on existing literature six independent

variables are selected, assuming that they each will have a significant effect on the dependent

variables. The included variables are population, the amount of all other than Chinese FDI

inflow, Chinese FDI inflow, openness to trade, human development, an index for corruption and

natural resource endowment. Except for the data included for the amount of Chinese FDI inflow,

which is obtained from the Chinese Commerce Yearbook, all the data is collected from the

Worldbank database. The results show that in the first model only openness to trade and natural

resources have a significant impact on the change of GDP. This means that Chinese FDI does not

have an effect on the economic growth in Africa. The conclusion drawn from the second model

is that all variables, except for the level of corruption, have an impact on the GDP level in

Africa. Ultimately it is concluded that the Chinese FDI has a positive effect on the GDP level in

Africa, and thereby indirectly on the economic growth in Africa.

Keywords: Foreign Direct Investment, Africa, China, Economic Growth, GDP change

JEL Classification: F21

NON-PLAGIARISM STATEMENT

By submitting this thesis the author declares to have written this thesis completely by himself/herself, and not to have used sources or resources other than the ones mentioned. All sources used, quotes and citations that were literally taken from publications, or that were in close accordance with the meaning of those publications, are indicated as such.

COPYRIGHT STATEMENT

The author has copyright of this thesis, but also acknowledges the intellectual copyright of contributions made by the thesis supervisor, which may include important research ideas and data. Author and thesis supervisor will have made clear agreements about issues such as confidentiality.

(3)

LIST OF ABREVIATIONS

FDI

Foreign Direct Investment

GDP

Gross Domestic Product

MNC

Multination Corporations

OECD

Organisation for Economic Co-operation and Development

OLS

Ordinary Least Squares

UNCTAD

United Nations Conference on Trade and Development

(5)

LIST OF FIGURES

Figure 1 (I): Inflows of FDI (Billions of dollars and %) page 1 Figure 1 (II): Africa’s FDI inflow (Billions of dollars and %) page 2

Figure 2.2 (I) China’s FDI inflow (% of GDP) page 10

Figure 2.2 (II) China’s FDI outflow (% of GDP) page 10

Figure 2.4 (I) Africa; FDI inflows, top 5 host economies (2014) page 12 Figure 2.4 (II) Africa’s inward FDI stock by sector (2012) page 12 Figure 2.5 (I) China’s FDI outflow in Africa (2013) page 13

LIST OF TABLES

Table 1 A summary of other papers about FDI and economic growth in Africa page 2 Table 2 A summary of literature about determinants of FDI page 8 Table 3 Summary of the independent variables of the regression analysis

with GDP growth as dependent variable page 17

Table 4 Summary of the independent variables of the regression analysis

with GDP level as dependent variable page 18

Table 5 Summary of the regression analysis with GDP growth as

dependent variable page 21

Table 6 Summary of the regression analysis with GDP level as

dependent variable page 23

(6)

1. Introduction

One of the consequences of globalization on the world economy is the increase of Foreign Direct Investment (FDI) and the growth of Multinational Corporations (MNCs) that in part are responsible for the FDIs (Brink, 2004). The definition of Foreign Direct Investment as is determined by UNCTAD is as follows; “When an investment is made to gain an operation outside of the domestic economy of the investing enterprise”. The main forms of FDI are through mergers and acquisitions, these are called Brownfield Investments, and by Greenfield Investments. With a Greenfield Investment, a firm builds a new production facility in the host country. This kind of investment is mostly done in developing countries. Brownfield Investments are made by acquiring or merging with a firm that is already located in the target host country (Skovgaard Poulsen & Hufbauer, 2011).

The last 30 years, the amounts of FDI flow directed to developed and developing countries have significantly differed, as can be seen in figure 1 (Skovgaard Poulsen & Hufbauer, 2011). Historically, the largest FDI flows have been to developed countries. Since 1990, the FDI flows directed to developing countries has significantly increased (Brink, 2004). In 2012, developing countries received for the first time ever, more FDI inflows than developing countries. Whereas the FDI inflows to developed countries declined to their near lowest levels in history with an amount of $561 billion, compared to $700 billion in developing countries (UNCTAD, 2013). The main sources of FDI inflow are from developed countries, but China as a developing country is playing an increasingly significant role. Thereby, China’s segment of FDI outflow as developing country to other developing countries is the fastest growing of this new type of FDI (Skovgaard Poulsen & Hufbauer, 2011).

Figure 1 (I): Inflows of FDI (Billions of dollars and %)

Source: (Skovgaard Poulsen & Hufbauer, 2011)

The relationship between Chinese FDI and Africa is one of the most debated relations concerning this new type of FDI. Their financial relationship between them goes back many years, but has intensified in the recent years (Foster, Butterfield, Chen, & Pushak, 2009). Moreover, due to becoming a more open economy, China became one of the leading countries in FDI outflow. As a result, in 2012 China moved

(7)

from the sixth to the third world’s largest investing country in Africa. Also in terms of FDI stock, next to Malaysia, South Africa and India has China become one of the largest sources for Africa (UNCTAD, 2013). This increasing activity between China and Africa is measured through financial assistance, as well as a focused improvement of the infrastructure. Here, a better infrastructure will improve transportation and consequently stimulate trade (Foster, Butterfield, Chen, & Pushak, 2009).

Africa holds 49 developing countries and has seen their total FDI inflow grown every year since 2010. But not only their FDI inflow increased, also their FDI outflow increased from 5% in 2011 to 14% in 2012. Opposed to this growing number in African’s developing countries, the FDI inflow numbers from other developing countries like Asia and Latin-America remained at the same level as they had in 2011 (UNCTAD, 2013). In addition, after 2012 there was still an overall decline of FDI flows in developing countries, but in Africa the FDI inflow increased by 5%. Therefore, this growing development in Africa compared to other developing countries implies a relative economic growth (UNCTAD, 2016).

Figure 1 (II): Africa’s FDI inflow (Billions of dollars and %)

Source: UNCTAD Report 2016

Studies about the potential relationship between China and domestic characteristics of Africa already exists. However, a clear relation between Chinese FDI and the economic growth in African nations is lacking, due to the limited number of case studies. The goal of this thesis is to investigate the effect of the growing Chinese FDI on the economic development in Africa. Since African countries are developing countries, it is assumed that this relationship has a positive effect on the economic growth and development. If this positive effect is tested, the results indicate that African countries should try to attract more FDI income flows and take advantage of this economic phenomenon. Thus, if a positive relationship is tested as is suggested above, FDI inflow will have a positive influence on the overall economic performance of Africa. However, when negative results of this relationship will be tested, it is favourable for African policy makers to decrease the FDI inflow. In attempt to clarify the relationship between Chinese FDI inflow on African countries, the following research question has been formulated;

(8)

What is the effect of the Chinese FDI on the economic growth in African countries?

First, there will be a literature review that consists of a theoretical part. The literature review will provide an in depth-insight in the background of FDI, the influence of FDI on the economic growth, the Chinese FDI in Africa, and a summary of the economic situation in China and Africa. Hereafter, a quantitative analysis will be performed through multiple regressions in complement of the theoretical questions as are described above. To study the effect stated in the research question, an OLS-regression will be performed using multiple independent variables. The dependent variable of this regression will be the economic growth measured by the changes of gross domestic product (GDP) in a country. The Gross Domestic Product of a country is the total value of all the goods and services that are produced within the economy (Krugman, Obstfeld, & Melitz, 2012). According to Nunnenkamp & Spatz (2004), is the change of GDP level the most used determination for economic change. Eventually, the results of the regression will be explained using the Gravity Model. This empirical model helps to determine the volume of trade between any pair of countries and helps to explain the restrictions that possibly limit international trade (Krugman, Obstfeld, & Melitz, 2012).

(9)

2. Literature Review

This section will give a review of the economic literature that explores FDI, a brief explanation of the economic status of Africa and China and their economic relationship, an overview of some relevant studies about the FDI in Africa, and lastly the Gravity Model will be explained.

2.1 FDI

Following the cross economic-border investment principle, a FDI creates a financial inflow for the host country, and a financial outflow for the investing firm. Subsequently, the purpose of the investing entity is to gain a long-lasting relationship and a fully or partly managing position in the invested entity, in this thesis perceived as the host firm. However, from a regular foreign portfolio investment perspective, it is difficult to determine the boundaries of the scope of the investing entities. FDI is not only on an organizational level, but as well on country level. At this geographical level, a country invests in another country. If a firm owns and thereby controls the assets of another firm in a foreign country, the controlling firm is called the parent enterprise. For example, in Transnational Corporations (TNCs) two or more entities located in different countries are linked by the means of a shared ownership, as with parent enterprises. Normally, these parent enterprises own a stake of equity within the invested foreign firm. This equity stake is gained through FDI and normally contains a required rate of owning 10% of the normal shares or voting power to own the control of the assets (UNCTAD, 2007). FDI is an important fundamental of GDP growth within the host country, as well as this GDP growth is an important part in the attraction of FDI. Essentially, FDI inflow does not only bring cash into the host country, but also transfers technology improvements (Borensztein, De Gregorio, & Lee, 1998).

There are two main forms of FDI, namely, horizontal FDI and vertical FDI. Horizontal FDI is when a foreign firm manufactures their products and/or goods in the local markets of the host country. In this manner, a foreign firm is able to avoid transportation costs and restrictions (Kinda, 2013). Horizontal FDI can be for example motivated by market-seeking FDI. Market-seeking FDI involves the exportation of firms to foreign countries in order to increase their global market size and their sales (Dunning, 1993). A benefit of horizontal FDI is the introduction of new products, services and production methods into the host-country. Subsequently, the introduction of these foreign investments methods lead to an increased level of competition on the local market. The increasing competition can be beneficial for some domestic producers as new assets will be introduced to their value chain. On the other hand, the entrance of foreign investments into local markets can move out other local players due to the increasing intense competition. In addition, when for example the parent firm’s home country has adopted export regulations which will negatively affect exportation activities, foreign economic activities such as horizontal FDI are limited (Nunnenkamp & Spatz, 2004).

In a vertical FDI, a firm set up a new plant or moves a part of their value chain to a foreign country with lower overall production costs. A vertical FDI is mostly done by firms that can separate their production process into different steps and thereby allocate these steps to different foreign locations (Kinda, 2013). Multiple motivations for a vertical FDI are studied. For instance, resource-seeking FDI is

(10)

aimed to obtain local resources which are not available in the home country (Dunning, 1993). This kind of FDI requires an up-front transfer of capital, technology and know-how. Here, issues can arise when a foreign enterprise has little connection with the domestic products and labour markets of the host country. Then, the foreign enterprise can for example easily be tricked by local corrupt elites. Therefore, resource-seeking FDI may not always automatically result in economic growth, in part limiting FDI activities (Nunnenkamp & Spatz, 2004).

Next, efficiency-seeking FDI occurs when firms can gain economies of scale in foreign markets. Firms use a few countries to serve multiple larger markets. Important motivation factors for the adoption of such strategies of FDI are location, possible resource endowment and governmental rules (Dunning, 1993). With this type of FDI, technology and know-how is transferred into the host-country with amounts that are compatible with the local level of development. This is beneficial for domestic suppliers and competitors due to the possibility of adaption and imitation (Nunnenkamp & Spatz, 2004). The transfer can help to develop human capital in the host county through labour training, skill acquisition and by the introduction of a different, more efficient, management organisation (Li & Lui, 2005).

Furthermore, amongst the other dissimilarities as earlier briefly were introduced, significant differences between horizontal and vertical FDI are that horizontal FDI is mostly determined by financing and human capital, while vertical FDI is largely determined by infrastructure and institutional constraints. Lastly, horizontal FDI is more attracted to larger markets, while vertical FDI is more attracted to cheaper production possibilities (Kinda, 2013).

In 2008 the global financial crisis began. According to Skovgaard Poulsen & Hufbauer (2011), the global FDI flows decreased together with the fall of real estate values, stock markets, consumer confidence and world trade. This decline in FDI was mainly caused by three reasons. The first reason was the global financial crisis. This crisis caused liquidity problems for Transnational Corporations around the world. Their access to credit decreased and their corporate balance sheets deteriorated. When in 2009 the sales began to recover, firms decided not to invest their profit anymore. This caused a shrink in credit disbursement and a decrease in FDI flows. Even when there was an opportunity, bigger firms would still not invest. Another reason was the slowdown of the economic growth. This was an effect of the decrease of FDI flows and the strong correlation between these two elements. This made it even less attractive for bigger companies to invest abroad. Finally, the financial crisis caused managers to be more cautious concerning risk-taking. Their investments moved from high-risking projects to investing in much safer assets.

2.1.2 Does FDI lead to growth?

The results of earlier studies researching the relationship between FDI and economic growth are mixed. Some proponents assume that FDI has a positive effect on the economic growth as the result of increasing domestic savings and investments, the support of the transfer of technology between countries, the higher competition in the domestic markets of the receiving country, and the increase in exports and by attracting other foreign stimulating externalities (Ram & Zhang, 2002). Most of these studies investigate the importance of the presence of certain characteristics in a country to optimally utilize the FDI inflow. For

(11)

example, Borenzstein et al. (1996) studied the effect of the FDI on economic growth in developing countries. He compared the efficiency of transferring technology to the developing country between FDI and domestic investment, using data of FDI inflow into 69 developing countries. This study concluded that FDI is economically more efficient and as result has a positive contribution to economic growth. However, it is noted that higher productivity as the result of FDI falls under the condition that a significant level of human capital in the host country needs to be present. Moreover, given that the home country possesses a certain level of human capital, the invested country will be better equipped to take advantage of the technology transferred.

Li and Liu (2004) found that FDI has a significant influence on the economic growth from the mid-1980s onwards. Their study implies both a direct and indirect positive effect on FDI. The authors stated that human capital exerts a strong positive effect on the growth in developing countries.

Alfaro et al. (2004), did a cross-country regression analysis with data from 1975 until 1995. In this study a positive correlation was found between the development of the state of the local financial market in the host country and FDI activities. Ultimately, empirical evidences obtained by showing that FDI plays an important role in economic growth.

Immurana et al. (2015) investigated the effect of FDI on economic growth and service sector value in Ghana. They did a cross-country regression analysis with data from 1980 until 2013. This study concluded that FDI has a positive significant influence on economic growth in the long run and short run. However, the impact of FDI on the service sector has shown only short-term positive results.

The research by Nunnenkamp and Spatz (2003) has adopted a different approach. They studied the effect of different kinds of FDI on the economic growth. They tested separately the influences of resource-seeking, market-seeking and efficiency-seeking FDI. They concluded that efficiency-seeking FDI is most likely to lead to economic growth. The reason for this is the spill over of technology and know-how, that resource- and market-seeking FDI are concerned with. Moreover, on the one hand, the resulting modernizing may be beneficial for local markets due to the transfer of capital and technology, and by introducing new products and services for instance as is explained before. But on the other hand, they possibly exclude domestic smaller competitors. The interaction between the host-economy and industry characteristics mainly affect the possibility for the host-country to take advantage of the FDI inflow, due to the adaption capabilities. Therefore, because of the different effects of the FDIs, the authors conclude that the positive effect of FDI on economic growth is not guaranteed.

Other studies found opposing results and found a negative correlation between FDI and economic growth. For example, Ram & Zhang (2002) emphasise the possible negative impact of FDI in their research. They state that with FDI, the total outflow of the investments possibly unequal is to the total FDI inflow and therefore no economic gains will be the result for the host country. For example, when unnecessary technology into the host country is transferred. This has negative consequences for domestic enterprises since they do not benefit from the introduction of new production methods and assets and thereby are pushed out of the market due to the increased competition. Subsequently, firms focussing too much on the domestic market instead of on export possibilities, which will change domestic policies and

(12)

thereby creates benefits for only foreign investors, and ultimately will cause negative changes within the political and social networks within the host country.

Carkovic and Levine (2005) try to estimate the impact of FDI on the economic growth. They use two related samples and test these against the hypothesis that FDI inflows do affect economic growth. The data used is from 72 developing and developed countries during the period of 1960 till 1995. The conclusion made by the authors is that FDI does not even have an independent influence on economic growth in the host country.

Akinlo (2004) investigated the potential effect of FDI on economic growth in Nigeria. The data used is from the period 1970 until 2001. The results show that neither private capital nor foreign capital has any significant effect on the economic growth. The author also stated that export, labour force and human capital have a positive effect on economic growth. In addition, financial development has a negative effect on growth.

Beugelsdijk et al. (2008) conducted a comparable research approach as Nunnenkamp and Spatz (2003), but came to another conclusion. Using total FDI as a benchmark, they tested the different effects of vertical and horizontal FDI. The data used for this study is obtained from 44 host countries receiving US FDI over the period of 1983 until 2003. First, they concluded that the effect of horizontal FDI dominated the effect of vertical FDI. Secondly, they stated that there was no significant evidence of any effect on the economic growth in developing countries by either horizontal or vertical FDI.

In table 1 below, there are some relevant researches summarized concerning FDI and economic growth in Africa during the period between 2006 and 2015,

Table 1: A summary of other papers about FDI and economic growth in Africa

Title

and

author

Method

and

data

period

Conclusion

Botha (2006) Theoretical research (1990 -2006)

(i) Africa may become dependent on revenues from resource endowment, with the ‘resource curse’ for Africa as effect

(ii) FDI should be accompanied with domestic development programmes

Hennix & Beradovic (2009)

Linear OLS regression analysis (1999 – 2006)

There is a positive relationship between FDI and openness of trade, and their combined effect on the economic growth in Africa

Indopu & Talla (2010)

Market size and possible resource endowment are the most important determinants for Chinese FDI in Africa

Claassen (2011) Panel regression analysis (2003 – 2008)

(i) More Chinese FDI causes higher African GDP, and a higher African GDP attracts more Chinese FDI

(ii) Countries with a higher level of corruption attract more Chinese FDI

Kapoor (2014) Panel regression analysis (2000 – 2010)

(i) Higher developed countries rely less on aid

(ii) Institutional quality, represented by corruption, is not a determinant for Chinese aid allocations

(13)

Joutsen & Norling (2014)

(i) FDI is positively correlated to economic growth

Ahl & Fukino (2014)

Case studies (i) China’s motivation for FDI in Africa is mostly related to the resource-seeking motives

(ii) In most African countries the influence of the Chinese FDI is positive, due to the creation of jobs and a better infrastructure for instance

Breivik (2014) Econometric analysis (2003 – 2011)

(i) Chinese FDI is attracted to large markets and resource-rich African countries

(ii) Chinese FDI is different than other FDI, because institutions are not a significant determinant for China

Utesch (2015) Panel regression analysis (2002 – 2012)

(i) Especially market size attracts Chinese FDI

(ii) In resource-rich African countries, Chinese FDI is attracted by control of corruption and negatively correlated with better implementation of rule of law

Ter Braak Linear OLS regression analysis (1995 – 2009)

(i) Chinese FDI increase inequality and corruption in African countries

(ii) Chinese FDI helps the economic growth in Africa due to the presence of China decreases environmental damage

2.2 The determinants of FDI

The literature on FDI provides multiple variables that explain the effect of FDI. The empirical studies on FDI considered multiple combinations of these variables with varying results. This section will give a summary of some literature on the subject. In table 2 some relevant studies concerning the determinants of FDI from the period of 2002 until 2012 are summarized.

Table 2: A summary of literature about determinants of FDI Study Method and data period Conclusion

Asiedu (2002) Panel study (2002), tested in 22 African countries

FDI determinants are different than for other developing countries:

(i) A higher return on investment and a better infrastructure is a determinant for developing countries, but not for Africa (ii) Openness to trade is a determinant for Africa, but not for the other countries

Agiomirgianakis et al (2006)

Panel regression analysis (1975 – 1997), tested in 20 OECD countries

Human capital, liberal trade regimes, infrastructure are significant determinants for FDI

Naude and Krugell (2007)

Dynamic GMM estimator (1970 – 1990), tested in 43 African countries

Important determinants for FDI inflow to Africa are good government (e.g. political stability, accountability & rule of law), inflation rate, investment, government consumption and initial literacy

Asiedu (2009) Panel study (1985 – 2000), tested in 22 African countries

(i) Natural resources, large market size, lower inflation, good infrastructure, educated population, openness to FDI, less corruption, political stability and reliable legal system increase FDI

(ii) African countries with small resource endowments can increase FDI inflow by improving their institutions and

(14)

institutions

Reiter and

Steensma (2010)

Linear regression analysis (1980 – 2005), tested in 49 developing countries

FDI inflows are positively correlated to improvement in human development, especially when corruption is low

Anyanwu (2012) Cross-country regression (1996 – 2008), tested in 53 African countries

(i) Market size, openness to trade, stricter rule of law, past FDI inflows, foreign aid, natural resource endowment all attract FDI

(ii) Higher financial development decreases FDI inflow Kolstad and Wiig

(2012)

Econometric analysis (2003 – 2006), tested in 142 host countries

Chinese FDI is attracted to large markets, and to countries having a combination of poor institutions and high natural resources endowments

Based on the literature summary above, the determinants used in this thesis are openness to trade, human capital, the level of corruption and natural resource endowment. Using these determinants, the effect of Chinese FDI on the economic growth in Africa will be tested.

2.3 The Chinese Economy

According to Si (2014) the globalization of China is separable into four stages using the Investment Development Path (IDP). This model illustrates the relation between FDI inflow and outflow, and the economic development of the country based on data from developed countries. The first stage to be drawn is from 1978 until 1991. Before this stage, China was an economically closed country with little

involvement in the global economy. During this stage, the Chinese economy began to open. The Chinese government preferred FDI inflow above FDI outflow, which resulted in a very little outflow of below 1 billion dollars.

The second stage was from 1992 until 1998. Chinese FDI inflow stayed more valued than FDI outflow. FDI inflow amounted 61 billions of dollars a year, while FDI outflow stayed at accounted for just 10% of the inward FDI. At the beginning of this period, FDI outflow increased at similar speed as FDI inflow. However, FDI outflow slowed down due to the Asian Crisis since the government became even more suspicious against outgoing FDI. At the end of the period some companies made overseas investments. Even though this was on a small scale, they were becoming the most successful inside their field in China and made themselves an international brand (Si, 2014).

The third stage from 1999 until 2005 was the official beginning of the ‘Go Global’ policy. The previously closed economy of China opened and the FDI outflow became larger than the FDI inflow. The government even encouraged companies to internationalize. In 2001 the 10th Five Year Plan of China, which was the guideline for Chinese development, included the ‘Go Global’ policy. China’s

manufacturing industries grew the most during this period and joined the overseas. Since this step to go overseas, more FDI outflow went into developing countries than to developed countries. This was the starting point of the geographical distribution changes (Si, 2014).

The final period is noted from after 2006. Following the World Investment Report (2016) of UCTAD, China grew in this period on multiple aspects the last years. To start, the FDI outflow of China grew more than the FDI inflow. This year, the FDI outflow of China grew by 3.6 percent to a record-high

(15)

of $12.6 billions dollars a year since 2015. This rapid flow of Chinese FDI is likely to keep increasing. This increase in FDI inflow is mainly the result of the investments in the retail, transport and finance sector. Investments in these infrastructure-related industries supported the investment climate and increased the attractiveness by foreign regions for manufacturing FDI inflow. But China not only started to invest abroad. Also within their own region Chinese companies are responsible for a minimum of one fifth of the total number of projects. The graphs in figure 2.3 (I) and (II) present the changes of FDI inflow and outflow as percentage of GDP.

Figure 2.3 (I): China’s FDI inflow (% of GDP)

Figure 2.3 (II): China’s FDI outflow (% of GDP)

Source both graphs: Worldbank

2.3.1 Is Chinese FDI different?

Multiple studies investigated the difference between Chinese and Western FDI. The results from these studies are that the FDIs of China are different compared to FDIs made by Western countries. For

(16)

example, Kolstad and Wiig (2012) concluded that Chinese FDI is different than other FDI with respect to developing and developed countries. Opposed to the pre-assumption that there is a positive correlation between political stability, resource endowment and the attractiveness of FDI, these authors had different conclusions after their empirical study. First, they concluded that China’s FDI is, just like the rest of the world, positively correlated to market size. Secondly, they discovered that the Chinese are attracted to markets with large natural resources combined with poor institutions. The greater the possibility of resource endowment, the more they are attracted by high political risk. This is different since FDI is normally attracted to a more stable political situation. Regarding Africa, this preference of China implies China to invest in countries whereas Western countries are not attracted by to invest in (Buckley, et al., 2007). This explains the larger share of China in the total investments made in countries with a weak government.

Buckley et al. (2007) investigate the determinants of Chinese FDI outflow. They used a regression analysis with data from the period of 1984 until 2001. The results show that Chinese FDI is attracted to high levels of political risk, host market size, geographic proximity and host natural resource endowment possibilities. These results suggest that Chinese FDI is not significantly affected by political environment (i.e. stability) as China perceives political risk differently compared to Western firms.

2.4 FDI and economic growth in Africa

According to the World Investment Report of 2015 of UNCTAD, Africa’s share of global FDI flows increased from 3.6 percent to 4.4 percent in 2013. Drivers that initiate this FDI trend are the rising FDI between African countries, expansion of emerging-market firms, growing of less traditional investors like the private equity sector, and a growing consumer market in Africa. The countries receiving the most FDI inflow are South-Africa, Congo, Mozambique, Egypt and Nigeria, as seen in figure 2.4 (I). However, not in all the African countries has been a growth of FDI noted. For example, in many countries located in West Africa a decrease of FDI inflow is measured. This decline is the result of the amongst other things the Ebola outbreak, regional conflicts and falling commodity prices, affecting the countries in this region.

(17)

Figure 2.4 (I): Africa; FDI inflows, top 5 host economies (2014)

The largest sector in Africa receiving FDI inflow is the service sector, as seen in figure 2.4 (II). Trends in 2014 show the continued importance of service investments, but as well the manufacturing sector (UNCTAD, 2015). During 2009 till 2012 32.4% of all the jobs in Africa were found in the service sector. In this period the African service sector grew more than twice as fast compared to the world average growth of this sector. More specific, the transport, storage and communications sector grew the most, which are significant important sectors for the development of Africa. This growth rate matches the characteristics of developing countries, meaning that it has a great chance of becoming a constant growth in the future (UNCTAD, 2015).

Figure 2.4 (II): Africa’s inward FDI stock by sector (2012)

According to the fiscal year overview of Africa made by the World Bank, Africa has made improvements in terms of economic growth and poverty reduction in the last 10 years. But nowadays Africa faces significant challenges mainly due to the global decline of commodity prices and region-specific risks. The overall growth in Africa decreased from 4.5% in 2014 to 3.0% in 2015 and it is projected to decrease even more in 2016 to 2.5%. This is the slowest pace since 2009. The exact decrease varies between countries and depends on the resource richness of the country.

(18)

2.5 Chinese FDI in Africa

According to UNCTAD’s yearly report on Asian FDI in Africa (2007), Africa is one of China’s most important partners for trade and cooperation. The trade between China and Africa grew from 11 billion dollars in 2000, to nearly 40 billion dollars in 2005 within 48 African countries. This amount continued to grow during the following years, see figure 2.5 (I). Ever since, China became the third biggest investor in Africa, after the United States and France (UNCTAD, 2013). The motivations of Chinese firms to directly invest in the continent of Africa varies. For instance, a Chinese firm wants to obtain direct assets in the market, secure access to natural resources or increased market penetration (UNCTAD, 2007)

Figure 2.5 (I): China’s FDI outflow to Africa

The Chinese government motivated Chinese firms to invest in different types of projects in Africa. The first one are industrial processing projects. Due to a more generic experience, Chinese firms have a relative better understanding and know-how technology in multiple fields compared to African businesses. For example, in the field of electronics, building materials and machinery building, China has made significant developments. The second type is agriculture. China and Africa have been cooperating to increase the productivity and quality of agriculture. The motive behind this interest, is China’s food supply problem. Investing in Africa’s underdeveloped agriculture by well-established Chinese firms, may add value to Africa’s export and help China with their food supply problem. The third category is natural resources. China needs to secure their access to Africa’s natural resources to support its economic growth. Almost the entire continent of Africa is rich in natural resources, especially in petroleum and other high-value minerals. The last type of project China invests in is infrastructure and real estate development. So far, in many African countries China has started to build roads, houses, hotels, schools and other public buildings. They also set up oil drilling companies and drilled water holes. These infrastructure projects are becoming the priority of China’s investments (UNCTAD, 2007).

0 1000 2000 3000 4000 5000 6000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 C h in ese FDI f lo w in m illi o n o f d o llar s Year Source data: China Africa Research Initiative

(19)

2.6 The Gravity Model

The Gravity model is an economic empirical model that for a long time has been the most successful for understanding the distributions of goods and labour (Anderson, 2011). The model is initially based on Newton’s law of gravity. Just as the attraction between objects due to gravity is proportional to their masses minus the distance between them, the trade between countries also depends on their GDP and their distance (Krugman, Obstfeld, & Melitz, 2012). The fit of this original idea behind the Gravity Model increases due the recording of trade frictions. Such as the effect of political borders and different languages (Anderson, 2011). One of the main uses of the Gravity Model is to help identify these anomalies in the trade between countries. When the actual trade between countries is more of less than predicted by the model, economists will search for an explanation. An important role to consider in this situation while determining the amount of trade, is the role of transport costs and geography. The Gravity Model shows a significant negative correlation between distance and international trade (Krugman, Obstfeld, & Melitz, 2012).

The Gravity Model was previously unconnected to the rich and extended family of economies theories. The good fit and concentrated clustering of coefficient estimates, the empirical literature suggested that it must be built on some economic laws. But due to the absence of an accepted connection to any economic theory, the Gravity Model had mostly been ignored. However, opposed to other empirical models, the theoretical foundation of the Gravity Model has made significant progress the recent years (Anderson, 2011).

An important factor that differentiates this model from other economic models, is the possibility to investigate the economic interaction between multiple countries. Most other economic models focus on just two, or maybe three, countries. To possibility allocate the different levels, gravity is in this many-country case the result from its modularity. The modularity of a system is the degree in which its components are possible to separate and recombine again. Here, the distribution of factors and goods across space is determined by the different gravity forces resulting from the size of the economic activity of each country (Anderson, 2011).

It is applicable to the empirically modelling of factor movements. The model is usually used to build a structural model concerning migration, FDI and an international portfolio. In this thesis, the use of the Gravity Model for explaining FDI is of interest. A key element in the explanation of the location of linked production and demand, is the trade-off between lower production costs or lower distribution costs. It is possible for a firm to reduce unit costs by concentrating their more production at one location and thereby minimizing the separation of the production process. However, a firm can at the other hand choose to decrease distribution costs by allocating production to the different markets where it is locally active. With other words, a firm chooses between producing at home and export their products, or investing abroad by setting up a new plant (Anderson, 2011).

The Gravity Model is applicable to the situation of China and Africa due to this trade-off. For China, the set-up of a new plant seemed to be the priority in this trade-off, since so many Chinese firms and workers had moved to Africa.

(20)

3. Empirical Analysis

This section will explain the method and data used to perform an empirical study and

create an answer to the research question. Using an Ordinary Least Squares (OLS) there will be

two regression done using panel data from the period 2005 until 2014. This data will be

obtained through data of the World Bank and the Natural Bureau of Statistics of China. The first

regression will have GDP growth as dependent variable and the second one the total amount of

GDP in Africa.

3.1 Method

To test the effect of Chinese FDI on the economic growth in Africa, two regression tests will be conducted. The first regression will test the effect of FDI on the economic growth by adopting GDP growth as dependent variable. The second test will include the yearly amount of GDP in Africa as dependent variable. In both regression tests, almost the same independent variables will be used. However, the difference here is that the first model uses the annual growth rate of each independent variable, instead of the requirement to log most variables because of their size. The second model uses the real amounts of these variables.

3.2 Data

To capture the development of economic growth in Africa over time, data from 2005 until 2014 of 50 African countries is used. Using a panel data analysis will provide the possibility to capture the temporal effects of all the used independent variables. Panel data is available if there are repeated observations of the same factors, collected over a period of time. The availability of panel data allows us to specify and estimate more complicated and realistic models, than is possible with a single cross-section analysis. The obtained panel data are pooled together in time series of the several African countries. This cross-sectional comparison of countries provides significant information on the interaction between the included countries (Verbeek, 2004). Of each variable, all data from each country is summed up to obtain a level of this variable for each year.

For the analysis there will be data used from 50 African countries, that includes: Algeria, Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo Democratic Republic, Congo Republic, Cote d’Ivoire, Djioubti, Egypt, Equatorial Guinea, Eritrea, Ethiopia, Gabon, The Gambia, Ghana, Guinea, Kenya, Lesotho, Liberia, Libya, Madagascar, Malawi, Mali, Mauritania, Mauritius, Morocco, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tome and Principe, Senegal, Seychelles, Sierra Leone, South Africa, Sudan, Tanzania, Togo, Tunisia, Uganda, Zambia, and Zimbabwe.

(21)

3.3 The Models

The first regression model will test the effect of Chinese FDI on economic growth in Africa by performing a cross-country analysis in which the growth rate of GDP is the dependent variable. As discussed in chapter 2.1.2., the studies on determinants of FDI have shown mixed results about which would have a significant effect. Considering the research question of this thesis and the available data, the determinants used will be the yearly growth rate of the amount of FDI other than Chinese, of the amount of Chinese FDI, of the openness to trade, of human development, of the lack of corruption level and the yearly growth of the natural resource endowment. The regression model used for this analysis is given below.

𝒚

𝒈𝒓𝒐𝒘𝒕𝒉

= 𝜶 + 𝜷

𝟏𝑭𝑫𝑰𝒈𝒓𝒐𝒘𝒕𝒉

+ 𝜷

𝟐𝑪𝑭𝑫𝑰𝒈𝒓𝒐𝒘𝒕𝒉

+ 𝜷

𝟑𝑶𝑷𝑬𝑵𝒈𝒓𝒐𝒘𝒕𝒉

+ 𝜷

𝟒𝑮𝑬𝑺𝒈𝒓𝒐𝒘𝒕𝒉

+ 𝜷

𝟓

𝑪𝑷𝑰𝑨

𝒈𝒓𝒐𝒘𝒕𝒉

+ 𝜷

𝟔𝑵𝑨𝑻𝑹𝑬𝑺𝒈𝒓𝒐𝒘𝒕𝒉

𝒚𝒈𝒓𝒐𝒘𝒕𝒉 Nominal percentage of change of GDP compared to the year before

𝜶 Intercept

𝜷_𝒏 Correlation Coefficient

𝑭𝑫𝑰_{𝒈𝒓𝒐𝒘𝒕𝒉} Nominal yearly growth rate of all other than Chinese FDI

𝑪𝑭𝑫𝑰𝒈𝒓𝒐𝒘𝒕𝒉 Nominal yearly growth rate of the Chinese FDI

𝑶𝑷𝑬𝑵_{𝒈𝒓𝒐𝒘𝒕𝒉} Nominal yearly growth rate of the Openness to Trade

𝑮𝑬𝑺_{𝒈𝒓𝒐𝒘𝒕𝒉} Nominal yearly growth rate of the Gross Human Development

𝑪𝑷𝑰𝑨𝒈𝒓𝒐𝒘𝒕𝒉 Absolute yearly growth rate of the lack of Corruption

𝑵𝑨𝑻𝑹𝑬𝑺_{𝒈𝒓𝒐𝒘𝒕𝒉}Nominal yearly growth rate of the Level of Natural Resources Endowment

In table 3 the dependent variables used to test the GDP growth in the first regression are summarized.

Table 3: Summary of the independent variables of the regression with GDP growth as dependent variable

Variable Description Units Expected

sign

Source

𝒚𝒈𝒓𝒐𝒘𝒕𝒉 Yearly growth of GDP of Africa

% World Bank

𝑭𝑫𝑰_{𝒈𝒓𝒐𝒘𝒕𝒉} Yearly growth of other than Chinese FDI

inflow into Africa

% + World Bank

𝑪𝑭𝑫𝑰_{𝒈𝒓𝒐𝒘𝒕𝒉} Yearly growth rate of the Chinese FDI

inflow into Africa

% + Natural Bureau

of Statics of China

(22)

𝑶𝑷𝑬𝑵𝒈𝒓𝒐𝒘𝒕𝒉 Yearly growth of the sum of Export and

Import of Goods and Services

% + World Bank

𝑮𝑬𝑺_{𝒈𝒓𝒐𝒘𝒕𝒉} Yearly growth of the Gross Enrolment in

Secondary Education (Both Sexes)

% + World Bank

𝑪𝑷𝑰𝑨_{𝒈𝒓𝒐𝒘𝒕𝒉} Yearly growth of the CPIA index:

Transparency, Accountability, and Corruption in the Public-Sector Rating

% - World Bank

𝑵𝑨𝑻𝑹𝑬𝑺_{𝒈𝒓𝒐𝒘𝒕𝒉} The yearly growth rate of the level of Natural Resource Rent

% + World Bank

The second regression will have the real GDP as dependent variable opposed to the growth of the GDP. The independent variables of this regression will be the same as is used in the first regression, but will use the real value of each year and each country. Namely, the amount of FDI other than Chinese FDI, the amount of Chinese FDI, the openness to trade, human development, the level of the lack of corruption and the possibility for natural resource endowment. If the chosen dependent variables will result in a higher level of GDP each year, they indirectly result in GDP growth. Since the level of change in GDP is an index for the economic growth, this regression will also show the effect of the chosen dependent variables on the economic growth in Africa. The model used in this regression with its independent and dependent variables is given below;

𝑮𝑫𝑷 = 𝜶 + 𝜷

_𝟏𝑭𝑫𝑰

+ 𝜷

_𝟐𝑪𝑭𝑫𝑰

+ 𝜷

_𝟑𝑶𝑷𝑬𝑵

+ 𝜷

_𝟒𝑮𝑬𝑺

+ 𝜷

_𝟓

𝑪𝑷𝑰𝑨

+ 𝜷

_𝟔𝑵𝑨𝑻𝑹𝑬𝑺 𝑮𝑷𝑫 Level of real GDP in Africa

𝜶 Intercept

𝜷_𝒏 Correlation Coefficient

𝑭𝑫𝑰 The total amount of all other than Chinese FDI 𝑪𝑭𝑫𝑰 The total amount of the Chinese FDI

𝑶𝑷𝑬𝑵 The level of Openness to Trade in the country

𝑮𝑬𝑺 Level of the Gross Enrolment Ratio in Secondary Education 𝑪𝑷𝑰𝑨 Level of lack of Corruption in the country

𝑵𝑨𝑻𝑹𝑬𝑺 Level of Natural Resources Rent in the country

In table 4 the dependent variables used to test the GDP growth in the first regression are summarized.

Table 4: Summary of the independent variables of the regression with GDP level as dependent variable

Variable Description Units Expected

sign

(23)

𝑮𝑫𝑷 Level of GDP in Africa

Millions of US dollars

World Bank

𝑭𝑫𝑰 Amount of other than Chinese FDI inflow into Africa

Millions of US dollars

+ World Bank

𝑪𝑭𝑫𝑰 Amount of Chinese FDI inflow into Africa Millions of US dollars + Natural Bureau of Statics of China

𝑶𝑷𝑬𝑵 Yearly growth of the sum of Export and Import of Goods and Services

% of GDP + World Bank

𝑮𝑬𝑺 Level of the Gross Enrolment in Secondary Education (Both Sexes)

Thousands of people

+ World Bank

𝑪𝑷𝑰𝑨 Level of the CPIA index: Transparency, Accountability, and Corruption in the Public-Sector Rating

Scale 1 (=low ) to 6 (=high)

- World Bank

𝑵𝑨𝑻𝑹𝑬𝑺 Level of Natural Resource Rent % of GDP + World Bank

3.2.1 Economic Growth & GDP level

The GDP level of a country shows the amount of Gross Domestic Product. This is an indication of the economic wealth of a country. According to Nunnenkamp and Spatz (2004), the GDP level gives the best view on the economic wellbeing of a country, rather than just the scale of an economy (Nunnenkamp & Spatz, 2004). Economic growth will be measured as the change of the real GDP in millions of US dollars between this year and previous year. Measuring GDP is the most used method to determine economic change (Nunnenkamp & Spatz, 2004).

𝑦

_{𝑔𝑟𝑜𝑤𝑡ℎ}

=

𝑦

𝑡

− 𝑦

𝑡−1

𝑦

_𝑡−1

3.2.2 Other than Chinese FDI

The variable FDI shows the amount of foreign direct investment inflow into Africa from all countries except from China in millions of US dollars. The variable 𝐹𝐷𝐼𝑔𝑟𝑜𝑤𝑡ℎ shows the yearly growth rate of this

amount, as a nominal percentage. The effect of this growth variable is considered as positive, due to technology transfer, increased competition and increased money supply as is explained earlier. In chapter 2.1.1. a literature overview of the potential negative or positive effect of FDI on economic growth is presented.

3.2.3 Chinese FDI

The variable CFDI represents the total amount of Chinese foreign direct investment inflow into Africa in millions of US dollars. CFD𝐼𝑔𝑟𝑜𝑤𝑡ℎis an index for the yearly nominal growth of the CFDI variable. It is

(24)

assumed that Chinese FDI is positively related to the growth of a country. Since Chinese FDI will have a similar effect as FDI and FDI has a suspected positive correlation with economic growth.

3.2.4 Openness to Trade

The variable used for the yearly growth rate of the openness to trade of a country is calculated by adding the total amount of the export of goods and services to the amount of import of goods and services. Originally the amount of import of goods and services and the amount of export of goods and services are expressed in percentage of the GDP level, resulting that the openness of trade index is also expressed in percentage of the GDP level. The growth variable of this index shows the yearly nominal change rate in percentages. Based on the literature presented in chapter 2.1.1., it is assumed that a more open host market will have a positive influence on the change of GDP level.

3.2.5 Human Development

The variable used to describe the human development will be the yearly change of the Gross Enrolment Ratio in Secondary Education of both sexes measured in thousands of people. Due to missing data of multiple African counties of other more accurate variables, the yearly change of the Gross Enrolment Ration is the only reliable variable. According to the literature in chapter 2.2.1, human capital is expected to have a large positive effect. The growth variable shows the nominal growth rate of this variable on a yearly basis.

3.2.6 Corruption

Corruption will be measured using the absolute growth rate of the CPIA level. This index stands for Transparency, Accountability and Corruption in the public-sector rating. The scale of this index goes from 1 (= low) to 6 (= high). Due to expression as an absolute scale, the yearly growth rate of this variable will be absolute as well. For example, a change of level 1 to 5 in one year means a growth rate of 4. A higher level of corruption in a country will have a negative effect on the attraction of FDI, since all other than Chinese FDI will be attracted to a low level of corruption. However, from the literature review presented in chapter 2.2.1 it is possible to conclude that the Chinese FDI is attracted to a higher level of corruption combined with a high level or natural resources, but other FDI is attracted to a lower level of corruption. Since the index measures the lack of corruption, a positive effect of this index on the economic growth and GDP level is expected.

3.2.7 Natural Resource Endowment

The natural resources endowment will be tested by using the nominal yearly growth rate of the Natural Resource Rent as percentage of GDP. This index shows the possibility of resource endowment in a country. The yearly growth variable will be nominal. The effect of this variables will be positive, since both Chinese and other FDI are attracted to a higher level of natural resource endowment as seen in chapter 2.2.1.

(25)

4. Results & Analysis

This section will give the regression results. The aim of this thesis is to investigate the

effect of Chinese FDI on the economic growth in Africa. The previous described model will be

run with the obtained data. Based on the results a second regression analysis will be done.

4.1 Analysing the results

Besides the chosen regression tests in this paper, various other ways exist to measure the effect on economic growth. The indexes indicating the amount Chinese FDI and non-Chinese FDI, have different ways to be measured, such as the net outflow or the net total, opposed to the net inflow as is done with the current regression included in this paper. The GDP level can also be measured per capita, next to the total amount of GDP in a country. The gross enrolment ratio in secondary education as an index for the human development may not be the best variable, but due the lack of data I was forced to use this proxy. Scale values used might not be the most accurate way to get an estimation of the impact of that independent variable, since the dependent variable is not estimated in scale values. The obtained results may therefore not be completely reliable, but will still create a clear understanding and accurate indication of the researched relationship and ultimately answering the research question.

4.2 Analysing GDP growth

To test the effect of the Chinese FDI on the economic growth in Africa, first a regression analysis with the dependent variable GDP growth is conducted. The resulting coefficients and their t-value of this regression analysis are summarized in table 4 and the actual outcome of the regression can be found in the Appendix number 1 and 2. Appendix 3 shows a descriptive statistics table of the chosen independent variables.

Table 5: Summary of regression model 1 with GDP growth as dependent variable

Dependent variable: GPD growth

Independent

Variables

Description of

Independent

Variables

Estimated

Coefficient

Estimated t-value

(Based on OLS

Standard Errors)

Significant

(Significance level

of 1%, two-sided

test)

Constant

Constant variable 0.127 6.261

YES

𝑭𝑫𝑰

𝒈𝒓𝒐𝒘𝒕𝒉 Yearly growth of

other than Chinese FDI inflow into

Africa -0.001

-0.454

NO

𝑪𝑭𝑫𝑰

𝒈𝒓𝒐𝒘𝒕𝒉 Yearly growth rate of

the Chinese FDI

inflow into Africa -1.97E-05 -0.453

NO

𝑶𝑷𝑬𝑵

_{𝒈𝒓𝒐𝒘𝒕𝒉} Yearly growth of the

sum of Export and -0.386 -3.033

(26)

In order to make the variable significant in a two-sided test with a significance level of 1%, the t-value should be higher than 1.96 or below -1.96. In accordance, following this model the indexes for the amount of FDI other than Chinese, the amount of Chinese FDI, the human development and natural resources have no significant effect on the economic growth in Africa. Furthermore, the results show that only the variables openness of trade and the index for the lack of corruption are significant.

In addition, the variable openness of trade indicates that a one percent increase will result in a decrease of 38.6 percent of GDP growth, holding all other variables constant. This is contradicting with findings in the literature, as the literature assumes that openness of trade would stimulate the economic growth.

Holding all other variables constant, the results also show an increase of 23.1 percent in the economic growth, under the condition that the index of the lack of corruption would increase by one level. It is noted that since this variable has a relative small scale of 1 to 6, an increase with1 level has relatively a more significant impact. That being said, a similar effect on the GDP growth level by a one level increase of CPIA is reasonable. However, this does not confirm the findings in the literature on the estimated effect of the level of CPIA. Here, the results for difference of Chinese FDI compared to all the other FDI is that Chinese FDI would be more attracted to economies with higher levels of corruption. An increase of one level of the CPIA index indicates that the lack of corruption increases, and therefore the level of corruption decreases. Based on the literature it was assumed that an increase of CPIA would result in a less attraction of Chinese FDI and in consequence a decrease of GDP growth.

The correlation between the independent variables is summarized in Appendix 3. Here, in case of a correlation lower than 0,6 it is concluded that there is no multicollinearity between the two variables. On the other hand, when the correlation between two variables is higher than 0.6 these variables are highly correlated, meaning they have multicollinearity. This weakens the individual effect on the independent variable and therefore decreases the reliability. As can be seen in appendix 3, there is no multicollinearity between the chosen independent variables of this regression model.

Import of Goods and Services

𝑮𝑬𝑺

Gross Enrolment in Secondary Education

(Both Sexes) 0.032 0.144

NO

𝑪𝑷𝑰𝑨

CPIA index: Transparency, Accountability, and Corruption in the Public-Sector Rating 0.231 2.567

YES

𝑵𝑨𝑻𝑹𝑬𝑺

𝒈𝒓𝒐𝒘𝒕𝒉 The yearly growth

rate of the level of Natural Resource

Rent 0.022 1.197

(27)

The R-square of a model represents the ability of the model to explain the dependent variable by using the independent variable. A higher R-square value indicates a better and more accurate model. The R-square of this model is 0.125 meaning that this model explains the effect of the independent on the dependent variables for 12.5 percent. The adjusted R-square has a value of 0.078 and describes that the model is adjusted to its sample size, and thereby explains the model for 7.8 percent.

4.3 Analysing GDP

Following a two-sided test with a significance level of 1%, the second model show that the variables chosen for the amount Chinese FDI, the total amount of all FDI other than Chinese, openness of trade, the level of human development and for the level of natural resources are significant. Only the index for the lack of corruption does not have a significant impact on the GDP level. The obtained estimated coefficients of the dependent variables and their t-value are presented in figure 6 below. See Appendix 4-6 for more regression results of this model.

Table 6: Summary of regression model 2 with GDP level as dependent variable

Dependent variable: GPD

Independent

Variables

Description of

Independent

Variables

Estimated

Coefficient

Estimated t-value

(Based on OLS

Standard Errors)

Significant

(Significance level

of 1%, two-sided

test)

Constant

Constant variable 23.116 0.002

NO

𝑭𝑫𝑰

Amount of FDI

inflow into Africa

other than Chinese 13.66 9.592

YES

𝑪𝑭𝑫𝑰

Amount of Chinese FDI inflow into

Africa 82.174 2.863

YES

𝑶𝑷𝑬𝑵

Yearly growth of the sum of Export and Import of Goods and

Services -489.029 -5.653

YES

𝑮𝑬𝑺

Level of the Gross

Enrolment in

Secondary Education

(Both Sexes) 447.273 3.335

YES

𝑪𝑷𝑰𝑨

Level of the CPIA index: Transparency, Accountability, and Corruption in the Public-Sector Rating 2859.239 0.85

NO

𝑵𝑨𝑻𝑹𝑬𝑺

Level of Natural Resource Rent 994.869 5.424

YES

(28)

Following the results, it is concluded that this model supports the assumption that Chinese FDI has a significant impact on the GDP level in Africa and therefore on the economic growth in Africa. More specific, should the amount of Chinese FDI increase with one million US dollars, holding all other variables constant, the amount of GDP will increase with 82.174 US million dollars.

In addition, next to the Chinese FDI have all other FDI inflows as well a positive impact on the GDP level of Africa. Should this amount of FDI increase with one US million dollars, the GDP level will increase with 13.66 US million dollars. Therefore is concluded that the effect of Chinese FDI is stronger compared to the effect of the other FDI included in the tests. Since it was predicted that FDI would have a positive effect on the level of GDP, the obtained positive effect of FDI in the regression test, confirm the statements that were found in literature and presented in Chapter 2.

Next, concluded from the literature was a positive effect of the openness of trade to the level of GDP, since this would stimulate possible import and export. However, the index for the openness of trade shows that the GDP level will decrease by 489.029 million US dollars if this index increases by one percent. Thus, these results are contradicted to the literature about the estimated effect of the openness of an economy relative to economic growth.

The variable indicating the level of natural resources and the GDP level have a positive correlation. If the natural resource rent increases by one percent of GDP, the GDP level will increase with 994.869 million US dollars. This is the same effect as was predicted based on the literature. Moreover, since FDI would be attracted by the level of naturel resources present in the host country, an increase of the level of natural resources would have a positive effect on the level of GDP.

The correlation between the independent variables is summarized in Appendix 6. With a correlation lower than 0,6 it is concluded that no multicollinearity between these variables exists. As can be seen in appendix 6, there is no multicollinearity between any of the independent variables of this regression model.

The R-square of this model is 0.523. This means that this model 2 has an ability of describing the dependent variable with an accuracy of 52,3 percent. The adjusted R-square has a value of 0.508. Adjusted to its sample size, this model describes the dependent variable for 50,8 percent.

4.4 Comparing models

The most important difference between the two models is the insignificance of the CFDI and FDI variables in model 1, and their significant effect on the dependent variable in model 2. This states that according to the first model the economic growth in Africa is neither effected by the amount of FDI inflow from China or from the rest of the world. However, in model 2 both FDI and Chinese FDI has a positive effect. Here, a one percentage increase in the amount of FDI inflow by countries other than Chinese, shows an increase of 13.66 US million dollars of the GDP level in Africa. The effect of a one percent increase of the amount of Chinese FDI is even stronger. A one percent increase of this variable shows an increase of 82.174 US million dollars of the GDP level.

(29)

In both models the variables that represent the ‘openness of trade’ are significant. Moreover, in model 1 the effect on the growth level of GDP caused by a one percent increase in the growth of ‘openness of trade’ results in a total decrease of 38.6 percent. In addition, model 2 show for the same variable a negative effect. Here, if the openness of trade increases by a fraction of one percent of GDP, the level of GDP will decrease with 489.029 million US dollars.

The variable for the level of natural resources is significant in model 1 but not significant in model 2. The variables for level of human development and for the level of natural resources are contrary significant in model 2, but not in model 1.

All the significant variables of the two models can be interpreted as variables that the Gravity Model is interested in. With the increase of some variables and decrease of others, the GDP level grew. This indicates that this change of this variable eased up the trade between China and Africa. One of the main reasons why the Gravity Model is used is to identify the possible anomalies between countries (Krugman, Obstfeld, & Melitz, 2012). The factor that made the Gravity Model unique is its ability to investigate the economic interaction between more than two countries (Krugman, Obstfeld, & Melitz, 2012). In the models used in this thesis the Gravity Model investigated the interaction between China and 50 African countries, making it a total of 51 countries.

The effect of Chinese foreign direct investment on the economic growth in Africa