ECN 360_ARIZONA STATE UNIVERSITY_ASSIGNMENT_FOREIGN DIRECT INVESTMENT (FDI) NATURAL RESOURCE ABUNDANCE AND ECONOMIC GROWTH

FOREIGN DIRECT INVESTMENT (FDI) NATURAL RESOURCE

ABUNDANCE AND ECONOMIC GROWTH

Introduction:

Indicators of the success of a country's economic development can generally be measured

through the level of economic growth in the country. A country's economic growth is influenced

by capital accumulation (investment in land, equipment, infrastructure, and facilities and human

resources), natural resources, human resources both in number and quality level of the

population, technological progress, access to information, the desire to innovate and develop

themselves and work culture (Todaro and Smith 2006).

Governments in developed and developing countries implement policies related to the

acceptance of foreign financing sources or Foreign Direct Investement (FDI) to stimulate

economic growth. The World Trade Organization (WTO) defines FDI as the activity of investors

based in one country (home country) acquiring assets in another country (host country) with the

intention of managing those assets. The enormous economic impact of FDI has been attributed to

its ability to benefit both the host, which is mostly developing countries, and the home country,

which is mostly developed countries (Kayani et al. 2021). Moreover, FDI has been considered

the most stable and prevalent component of foreign capital inflows in developing countries

(Hobbs et al. 2021).

FDI has a direct influence on the inflation rate and the level of government spending

which in turn affects the GDP growth rate in the recipient country (Kayani et al. 2021). FDI is

expected to stimulate domestic investment and complement the limited savings and investment

potential to achieve economic growth (Mehic et al. 2013). In addition, according to Miao et al.

(2021) the improvement of institutional quality indicators also plays a role in the effect of FDI

on economic growth. Institutional quality is the average of six different indicators of institutional

quality and governance, namely political stability and absence of violence, voice and

accountability, control of corruption, government effectiveness, rule of law and regulatory

quality published by Worldwide Governance Indicators (WGI). Trade openness factors can also

promote economic growth through various intermediaries including technology transfer,

attraction of FDI, source of foreign exchange, means of gaining access to capital equipment to

enhance development (Opoku et al. 2019).

Sasana and Fathoni (2019) stated that the entry of FDI into a country is influenced by the

conditions and strategies of foreign investors (push factors) and the conditions of the country

receiving FDI (pull factors). Push factors that affect FDI include the investor's investment

production strategy, as well as the risk perception of the recipient country, while pull factors

include competitiveness, industry/trade-related policies, FDI liberalization policies and the

availability of natural resources in the host country.

Global foreign direct investment (FDI) flows in 2021 reached $1.58 trillion, a 64%

increase from the previous year's global FDI level of just under $1 trillion according to the

United Nations Conference on Trade and Development (UNCTAD) in its World Investment

Report (WIR) 2022 publication (Figure 1). This substantial growth is mainly due to the policy of

recovery after the decline in global FDI levels due to Covid-19 in 2020. Global investment levels

in 2021 returned to similar levels achieved in 2019 before the pandemic was declared. However,

growth was uneven across the world. Investment flows to developed countries grew by 134% in

2021, while developing countries only grew by 34%. However, it is worth noting that developed

countries were significantly more (negatively) impacted in 2020 than developing countries. This

suggests that developed countries are much more sensitive to global shocks.

Global FDI flows generally improved in 2021 across all regions of the world. Figure 2

shows the developed country group experiencing a rapid increase, FDI flows to European and

North American countries grew by 171% and 145% respectively. Similarly, FDI flows to

developing Asian countries grew by 19% which is an all-time high with a total investment value

of $619 billion, this increase was dominated by FDI flows to East and Southeast Asian countries.

In addition, FDI flows also increased by 56% in Latin American and Caribbean countries and

more than doubled in African countries. This indicates that world countries have made FDI an

instrument of economic recovery after the Covid-19 pandemic.

Abundant natural resources also play an important role in increasing economic growth

(Ridzuan et al. 2021). Countries with abundant natural resources are believed to have the

capacity to achieve higher growth and development compared to countries that do not have

abundant natural resources (Rahim et al. 2021). Some countries, especially low- and middle-

income countries, tend to rely on natural resource wealth for economic growth and development

because it may be the only source of capital available (Barbier 2019). A growing population

further increases the demand for natural resource use. Finally, for many low- and middle-income

countries, natural resource-based development and primary commodity exports are the main

engines for long-term growth and development. Primary commodity exports are the percentage

of exports of agricultural raw materials, fuels, ores and metals to total goods exports.

Based on data processed from the World Bank in the World Development Indicators, over the

period 2001-2021 the share of primary product exports averaged 17.40% globally, 19.58% in

low- & middle-income countries and 16.20% in high-income countries. In Figure 3, long-term

changes in the concentration of primary commodity exports for low- and middle-income

countries are shown compared to global trends and high-income countries. Although in general

there has been no significant increase or decrease in the share of primary product exports across

all economic groups, the share of primary product exports of low- & middle-income countries is

still relatively higher than that of high-income countries.

Developing countries that are endowed with abundant natural resources may not be able

to exploit this natural wealth efficiently and generate productive investment. This background is

one of the attractions for multinational companies to enter these countries through FDI. FDI

inflows are more aggressive in countries with abundant natural resources (Zeeshan et al. 2020).

FDI is believed to help improve the technical skills of the host country's workforce and their work

efficiency. In addition, FDI also brings technology, assets, and capital that ultimately impact the

economic growth of the host country (Asghari et al. 2014).

1.1 Problem Formulation

Theoretically, the benefits of FDI are very important for the economic growth of the host

country. So far, the understanding taught by some world economists is that FDI can benefit both

parties, both for the investor's home country, which is dominated by developed countries, and for

the host country, which is dominated by developing countries that are relatively rich in natural

resources and cheap and abundant labor. However, FDI can also have a negative effect because

FDI can be used as a mechanism to exploit and gain control over developing countries by

western industrialized countries (Ali et al. 2019) shows that FDI will not support the economic

growth of developing countries because domestic industries will not compete with foreign

industries that provide products at low prices. Developed countries only utilize the assets of

developing countries for profit and source their production inputs. Developing countries are just

a place to obtain cheap raw materials, a place to produce with cheap labor, and become the

largest consumers of MNCs' products (Haryadi 2008). The existence of MNCs is only a tool that

is able to strengthen the hegemony of capitalists in developing countries, which causes the

increasing dependence of developing countries on MNCs (Damanhuri 2010).

Empirically, the effect of FDI on economic growth in both developed and developing

countries is still debatable. Some researchers such as Mehic et al. (2013); Alvarado et al. (2017);

Hayat (2017); Sultanuzzaman et al. (2018); Dinh et al. (2019); Shittu et al. (2020); Zeeshan et

al. (2020); Mohamed et al. (2021); Orji et al. (2021); Ahmad et al. (2022) found that FDI can

boost economic growth, Gochero and Boopen (2020) found that FDI in the mining sector can

increase economic growth both in the short and long term. Awunyo-Vitor and Sackey (2018)

also found that FDI in the agricultural sector can accelerate economic growth, which is supported

by the research of Chandio et al. (2019) who found a strong causal relationship between FDI in

agriculture and economic growth. However, some other studies found different results.

Bouchoucha and Bakari (2019) found that FDI has a negative effect on economic growth in both

the short and long run. Kolisi (2021) found that FDI in the manufacturing sector has a negative

impact on long-term economic growth, and Meivitawanli (2021) found that FDI has a negative

impact on overall economic growth. Research on the effect of FDI on economic growth has been

widely conducted. Some studies reveal that the relationship between FDI and host country

economic growth depends on many other relevant factors and variations in these factors are

thought to substantially alter the FDI-economic growth relationship. This study focuses on one of

these factors, namely the abundance of natural resources, which is still often overlooked.

The effect of natural resource abundance on economic growth has been widely

researched and studied. Sachs and Warner (1995) proved that countries with abundant natural

resources tend to grow slower than countries with scarce natural resources. This phenomenon is

often referred to in the literature as the "natural resource curse". There have also been many

studies on the effect of natural resources on FDI. Poelhekke and Ploeg (2012) found that the

abundance of natural resource assets has a positive impact on natural resource FDI but a negative

impact on non-natural resource FDI. In addition, Asiedu (2013) found that natural resources have

an adverse effect on FDI and the resource curse of FDI occurs even after controlling for

institutional quality and several other important determinants of FDI.

The effect of a country's abundance of natural resources is still often ignored in some

studies on the effect of FDI on economic growth. Therefore, this study tries to fill the gap by

exploring the effect of FDI on economic growth by considering the factor of natural resource

abundance in the host country. Research on the effect of natural resource abundance on economic

growth has been widely conducted using the approach of the share value of natural resource

exports to total exports of goods and services as an indicator of the size of natural resources. In

general, the data coverage used is the sum of exports of fuels, ores and metals. Bakari and

Mabrouki (2017), who examined the impact of agricultural exports on economic growth in

Southeast Europe, found that the exports of the agricultural sector have a significant impact on

economic growth agriculture is strongly correlated and has a positive impact on economic

growth. Zagozina (2014) also found similar results in the former Soviet Union member states.

According to Barbier (2019), most low- and middle-income countries today rely heavily on the

exploitation of their natural resources for export-oriented commercial economic activities.

Exports of primary products such as agricultural raw materials, food, ores and metal

commodities tend to dominate especially in low- and middle-income countries.

In contrast to previous studies, this study adds agricultural raw material export data from

WDI as part of natural resource exports. This research is expected to add best practices on the

effect of FDI on economic growth, especially by considering the factor of natural resource

abundance.

2.1 Economic Growth Theory

Mankiw (2013) states that economic growth is the impact of economic activity in

generating additional income for society at a certain point in time. Economic activity is the

process of using inputs or factors of production to produce output. GDP is a good indicator to

measure economic growth because it reflects the total value added generated by production

activities. Indirectly, an increase in GDP shows the rewards for each factor of production used in

the production activity.

Classical growth theory explains the importance of natural resources as a factor in the

production process because the availability of natural resources has a maximum effort in

increasing economic growth. On the other hand, an increase in the number of workers

(population) will lead to a decrease in productivity "law of diminishing returns". Furthermore,

Solow's growth theory or often called neoclassical growth theory explains that growth is based

on factors of production and uses determinants such as labor growth, capital accumulation, and

technological progress (exogenous). There are several assumptions that need to be known in this

theory. The first assumption is constant return to scale, where Solow (1956) considers the

economy to be in a constant state.

There is a large capacity so that an increase in labor and capital will result in the same

change in output. Second, there is perfect substitution between labor and capital. Third, there

will be a decline in the marginal productivity of each factor of production.

The next endogenous growth theory or new growth theory was first proposed by

Romer (1986). According to this theory, economic growth can be generated by factors in the

production process that are influenced by the level of human capital through technological

development. This theory is supported by Lucas (1988) who argues that human capital has the

same role and is also needed in the production process in addition to physical capital. The

endogenous theory takes into account the elements of externalities that can create increasing

returns to scale, thus complementing the previous assumptions held by the neo-classical theory,

which only assumes constant returns to scale. In addition, the endogenous growth model

suggests an active role for public policy to drive economic development through foreign direct

and indirect investment.

Limited natural resources and land can cause output per worker to eventually fall. The

declining amount of natural resources and land per worker is a drag on economic growth. But

this can be overcome by technological progress, which is one of the drivers of economic growth.

If the driving force is greater than the drag, then there will be sustained growth in output per

worker. This is what has happened over the last few centuries.

Land stocks are fixed, and the use of natural resources must eventually decline. Therefore,

although technology has been able to overcome resource and land limitations over the past few

centuries, it may still appear that such limitations must eventually become binding constraints on

the ability to produce.

The gains from technological progress are thought to have outweighed the losses from

resource and land constraints (Romer 2019). But this does not tell us how large those losses are.

For example, the losses may be large enough that only a moderate slowdown of technological

progress would make overall income growth per worker negative.

2.2 Foreign Direct Investment (FDI)

According to OECD (2002), FDI is an integral part of an open, effective international

economic system and a major catalyst for development. FDI can trigger technology spillovers,

help human capital formation, contribute to international trade integration, help create a more

competitive business environment and enhance enterprise development which will further

contribute to economic growth. FDI is more than just the transfer of capital or the establishment

of a local factory in a developing nation. Multinational enterprises bring with them production

technologies, tastes and lifestyles, managerial philosophies, and diversified business practices

(Todaro and Smith 2006). The main challenge facing host countries to get the maximum benefit

from FDI is the need to establish an enabling policy environment for investment in a transparent,

broad and effective manner and to build the human and institutional capacity to implement it.

Multinational Enterprises (MNEs) or foreign investors in conducting production

processes in the host country may choose between the following strategies: (1) greenfield FDI,

i.e. setting up a new foreign affiliate or factory in the host country to produce goods locally; (2)

mergers and acquisitions (M&A), i.e. acquisition of local firms and their production capacity or

also called brownfield FDI; and (3) cooperation with local firms by establishing joint ventures.

According to Krugman et al. (2012), the reasons why a company or MNE chooses to operate an

affiliate in a foreign location can be classified into 2 types, namely: (1) Vertical FDI; i.e. FDI

that operates in a vertical manner involves a geographically decentralized process of the firm's

production flow. The company will carry out production activities in countries that have low

labor costs, then the production in these countries will be channeled back to the parent country.

For example, a good whose production requires a lot of capital will transfer its production

process to a capital intensive country. Conversely, if the good requires high-skilled labor, the

production process will be carried out in a country with an abundance of high-skilled labor

resources. (2) Horizontal FDI; that is, FDI that is carried out horizontally or carries out the

production process of similar goods in various countries. This type of FDI aims to create new

markets, and has a surplus in terms of shipping cost efficiency, as the production location of the

goods becomes more accessible to consumers. An example would be a car company that

completely replicates its production process in various foreign countries.

According to Cahyani (2019), FDI flows that occur in developed countries take into

account economic openness, size, and potential economic growth. Meanwhile, the reasons for

investors to invest in developing countries depend on economic factors, such as the availability

of skilled labor, production factors, labor wages, infrastructure, and the abundance of natural

resources.

2.3 Abundance of Natural Resources

Until the late 1980s, much conventional research assumed that the relationship

between abundant natural resources and a country's development was mutually beneficial.

Abundant natural resources were generally and unquestionably seen as an important pillar of

economic progress and prosperity. According to Badia-Miró and Ducoing (2015) natural

resource abundance is not a fixed situation, but a process of reaction to changes in the

commodity price structure and supporting factors. Economic development requires capital, labor,

technical change and appropriate institutional arrangements. History shows that institutional

quality is a key factor for managing abundant natural resources, especially the rewards derived

from their use and exploitation.

According to Willebald et al. (2015) economic development is no longer considered to

depend solely on the accumulation of physical and human capital. There is a third form of

economic capital or asset that is essential for the performance of production systems,

consumption, investment, savings and welfare. This distinct type of capital is the endowment of

natural resources and the environment available to an economy, and is commonly referred to as

"natural capital". Natural capital is essential for sustainable economic development, but the

economy's increasing dependence on natural resource exploitation is a barrier to growth and

development for most of the world's low- and middle-income people (Barbier 2019). Low- and

middle-income countries tend to depend on their natural resource endowments for economic

growth and development because in low-income countries natural capital may be the only source

of capital available to them. In accordance with the Hecksher-Ohlin (H-O) modern trade theory,

a country tends to export commodities with abundant and relatively cheap factors of production

and will import factors of production that are scarce or expensive in the country.

Raul Prebisch and Hans Singer, known as the Prebisch-Singer Hypothesis, in the late

1940s explained that the terms of trade between primary commodities and manufactured

commodities would decline over time in the long run. In this approach, countries in the world are

divided into two groups, developed countries that produce manufactured commodities and

developing countries that only produce primary commodities. The decline in the exchange rate

of primary commodities of developing countries against manufacturing commodities of

developed countries causes the production of developed countries to become more expensive

than the production of developing countries. As a result, the budget of primary commodity-

producing countries is used more to import manufactured commodities from developed

countries, which results in countries that produce primary goods experiencing a trade balance

deficit, which means that it will also reduce the portion of savings and investment, which

ultimately has an impact on the decline in economic growth rates.

According to Todaro and Smith (2006), the income elasticity of demand for primary

product commodity groups is relatively low. The percentage increase in the amount of primary

agricultural products and raw materials required by most of the primary agricultural products and

raw materials is relatively low. importers (the richest countries) will increase by less than the

percentage increase in their gross national income. In contrast, the income elasticities of fuel,

certain raw materials, and manufactured goods are relatively high. Thus, as incomes in rich

countries increase, demand for food, foodstuffs, and raw materials from developing countries

increases relatively slowly, while demand for manufactured goods increases relatively quickly.

The end result of such low elasticity of demand is that the relative prices of primary products

tend to fall over time. Furthermore, since the price elasticity of demand (and supply) of primary

commodities also tends to be quite low (inelastic), any shift in the demand or supply curve can

lead to large and volatile price fluctuations. The income elasticity of demand is the

responsiveness of the quantity of a commodity demanded to changes in consumer income,

measured by the proportional change in quantity divided by the proportional change in income.

Whereas the price elasticity of demand is the responsiveness of the quantity of a commodity

demanded to changes in price, expressed as the percentage change in quantity demanded divided

by the percentage change in price. These two elasticity phenomena contribute to what has come to

be known as export income volatility, which has been shown to lead to slower and less

predictable economic growth.

2.4 Previous Research

There have been many studies conducted on the impact of FDI on economic growth.

Orji et al. (2021) found that FDI can increase economic growth so that the government must

attract FDI in all sectors, especially the industrial and service sectors, supported by the research

findings of Ahmad et al. (2022) that FDI can spur economic growth. Gochero and Boopen

(2020) found that mining sector FDI can increase economic growth both in the short and long

term. The same results were also found by Awunyo-Vitor and Sackey (2018) that FDI in the

agricultural sector can accelerate economic growth, supported by Chandio et al. (2019) who

found that there is a strong causal relationship between FDI in the agricultural sector and

economic growth.

Different results were found in the research of Bouchoucha and Bakari (2019) which

stated that FDI has a negative effect on economic growth both in the short and long term. Kolisi

(2021) found that manufacturing sector FDI has a negative impact on long-term economic

growth, this is supported by the findings of Meivitawanli (2021) that FDI has a negative impact

on economic growth.

Another study was conducted by Hayat (2018) on the relationship between FDI and

economic growth and the impact of natural resource abundance in the host country on the

relationship between FDI and economic growth using data from 104 countries in 1996-2015. The

results found that FDI has a positive and significant effect on the economic growth of the host

country. However, the effect of FDI inflows on economic growth changes as the size of the

natural resource sector changes. The estimated positive effect of FDI inflows on economic

growth decreases as the size of natural resources expands. An increase of more further

increase in the size of the natural resource sector will lead to a negative effect of FDI on

economic growth.

Adika (2022) examined the complementarity role of economic integration, natural

resources and FDI in explaining economic growth in the Southern African Development

Community (SADC) region using OLS and IV estimators to control for potential endogeneity.

Empirical results reveal that regional economic integration significantly increases economic

growth in the region. Natural resources and FDI jointly and significantly impact economic

growth in resource-rich countries in the region. In addition, domestic resources such as gross

domestic savings and human capital significantly impact economic growth in the region.

The novelty of this research from several previous studies is that the coverage of

countries that become research units is more, namely 124 countries and the time period of data

used is longer, namely 25 years 1996-2020. This study adds agricultural raw material export data

derived from the World Development Index (WDI) World Bank Database as one part of natural

resource exports that have not been used in previous studies in addition to exports of fuel, ores

and metals as a proxy for the abundance of natural resources of the host country. In the

descriptive statistical analysis section of this research, a quadrant analysis was carried out to see

how the data distribution pattern and the general description of the relationship between each

dependent and independent variable of this research in countries in the world based on their per

capita income group which has not been done by many other researchers. In addition, in the

inferential analysis section of this study, we simulated the effect of total FDI on economic

growth under four different conditions of natural resource abundance: 1) no increase in natural

resource abundance, 2) natural resource abundance increases by 1%, 3) natural resource

abundance increases by its average value, and 4) natural resource abundance doubles or increases

by 100%. This makes it possible to analyze how the effect of FDI on economic growth changes

with changes in the abundance of natural resources in the host country in each of the per capita

income groups of the countries of the world under study. Framework of Thought

Foreign Direct Investment (FDI) is one of the main determinants of economic

development because it brings tremendous growth opportunities to the recipient country (Kayani

et al. 2021). In some countries, especially developing countries, FDI plays a very important role,

even considered as an engine of economic growth and development. Foreign capital can help

reduce the gap between capital needs and national savings, increase skill levels in the host

economy, improve market access and contribute to technology transfer and good governance

(Abbes et al. 2015).

Global FDI flows began to improve and increase in 2021 after the global economic

recovery due to the Covid-19 pandemic, reaching $1.58 trillion, an increase of 64 percent from

the previous year's global FDI level of less than $1 trillion (UNCTAD in World Investment

Report 2022).

Although theoretically the benefits of FDI are crucial for the economic growth of host countries,

the empirical evidence on the relationship between FDI and economic growth of developing

countries is still debatable. On the one hand many studies have reported that FDI promotes

economic growth of developing countries, while some other studies have found no evidence or

weak evidence that FDI promotes economic growth of developing countries. Some literature has

shown that the relationship between FDI and economic growth (FDI-led growth) depends on

many other factors, such as inflation, trade openness, government spending, population growth,

domestic investment, and institutional quality. However, one factor that is often overlooked is

the host country's natural resource abundance. Hayat (2018) found that the positive effect of FDI

inflows on economic growth may decrease as the size of natural resources expands. A further

increase in the size of the natural resource sector will potentially lead to a negative effect of FDI

on economic growth. Sachs and Warner (1995) proved that countries with abundant natural

resources tend to grow slower than countries with scarce natural resources. This phenomenon is

often referred to in the literature as the "resource curse".

Using descriptive and regression analysis methods, this study analyzes the effect of

FDI on the economic growth of several countries in the world and takes into account the

abundance of natural resources of the host country. Based on the results of the analysis, policy

recommendations will be formulated in order to increase the benefits that FDI can provide for

the host country.

3.1 Data Analysis and Processing Methods Descriptive Analysis

This analytical method is used to answer the first research objective, which is to

analyze the development of FDI inflows to various countries based on their per capita income

level groups. The analytical tools used are series graph analysis for the observation period 1996-

2020 and quadrant analysis using 2019 data.

Panel Data Analysis:

This method of analysis is used to answer the second research objective, which is to

analyze the effect of FDI on economic growth by considering the host country's natural resource

abundance factor and several other control variables based on its per capita income level group

in the observation period 1996-2020.

Panel data is data that has space (individual) and time dimensions. In panel data, the same cross

section data is observed over time. If each cross section unit has the same number of time series

observations, it is called a balanced panel, otherwise if the number of observations is different

for each cross section unit, it is called an unbalanced panel.

According to Baltagi (2021), the use of panel data in regression has the following advantages:

1. By combining time series and cross section data, panel data provides more data and more

complete and varied information. Thus, it will result in a larger degree of freedom and

increase the precision of the estimation.

2. Panel data is able to accommodate the level of heterogeneity of unobserved individuals that

can affect the results of modeling (individual heterogeneity). This cannot be done by time

series or cross section studies so that the results obtained through these two studies will be

biased.

3. Panel data can be used to study the dynamism of data. This means that it can be used to

obtain information on how the condition of individuals at a certain time compares to their

condition at other times.

4. Panel data can identify and measure effects that cannot be captured by pure cross section

data or pure time series data.

5. Panel data allows for the construction and testing of more complex models than pure cross

section data or pure time series data.

6. Panel data can minimize the bias produced by individual aggregation due to too many

observation units.

The procedure in this method will begin by determining the most appropriate

panel data estimation method to get the best results. There are three estimation methods that can

be used, namely the pooled least square method (PLS) or common effect model-CEM, fixed

effect model (FEM), and random effect model (REM).

Common Effects Model (CEM)/ Pooled Least Square (PLS)

The first simplest approach in panel data regression is to use the ordinary least squares

method or often called Pooled Least Square (PLS). In this method, it is assumed that there are no

differences in intercept and slope values in the regression results either on the basis of

differences between individuals or between time. The parameter estimation method in the CEM

model uses the Ordinary Least Square (OLS) method.

3.2 Definition of Operational Variables

The variable definitions that will be used in this study are as follows:

1. Economic growth is proxied by the growth rate of real GDP per capita (in percent).

2. Foreign Direct Investment (FDI) inflow is the ratio of FDI inflow to GDP (in percent).

3. Natural resource abundance is proxied by the share of natural resource exports to total

exports of goods and services. Natural resource exports include exports of fuels, ores,

metals and agricultural raw materials (in percent).

4. Inflation is an upward trend in the prices of goods and services in general that takes place

continuously. Inflation can also be interpreted as a decline in the value of money against

the value of goods and services in general (in units of percent).

5. Trade openness is the ratio of exports and imports to GDP (in percent).

6. Government expenditure is the ratio of final government consumption expenditure to GDP

(in percent).

7. Population growth is the annual rate of population growth (in percent).

8. Domestic investment is the gross domestic savings rate (in percent).

9. Institutional quality is the level of institutional quality obtained from the average of six

different indicators of institutional quality and governance: political stability and absence

of violence; voice and accountability; control of corruption; government effectiveness; rule

of law and regulatory quality. (in unit scores ranging from approximately -2.5 (weak) to

2.5 (strong)).

4.1 Development of FDI Inflows and Per Capita Income by Country Income Group

The development of FDI and GDP per capita by income group can be seen in Figure 5

and Figure 6. Both FDI inflows and per capita income in each country's income group have the

same pattern or trend, which is getting higher and higher. Although it cannot be directly

concluded that the higher the flow of FDI in a country, the higher the per capita income of the

population in that country. There is a wide gap in FDI inflows between groups of countries based

on their per capita income levels. Figure 5 shows that FDI inflows dominated to high income

countries (HIC) over the past 25 years with an average of US$ 1,204.21 billion per year with the

highest total realization in 2007 of US$ 2,599.07 billion and the lowest in 2018 of US$ 319.95

billion. Meanwhile, in upper-middle income countries (UMIC), the development of FDI inflows

was relatively stagnant with an average of US$ 324.59 billion. FDI flows seem to have gained

significant momentum due to a booming M&A market and rapid growth in international project

finance, which is likely due to favorable financing conditions and a large infrastructure stimulus

package.

Based on GDP per capita, Figure 5 shows that there is a considerable gap in the

average value of per capita income levels between high income countries (HIC) and upper-

middle income countries (UMIC). In terms of data patterns, it can be seen that in both high

income countries (HIC) and upper-middle income countries (UMIC), the average level of per

capita income during the 1996-2020 period has a trend that tends to increase every year.

Hlaváček and Bal-Domańska (2016) found that foreign direct investment in Central and Eastern

European countries has become an important indicator of economic development and a measure

of external confidence in their economic stability and growth.

In Figure 6, FDI inflows to lower-middle income countries (LMIC) have experienced a

significant upward trend over the past 25 years compared to the other three income groups as

seen from the trendline shown with an average of US$ 82.11 billion with the highest total

realization in 2019 of US$ 147.23 billion and the lowest in 2000 of US$ 12.53 billion.

Meanwhile, in Low income countries (LIC) the development of FDI inflows is relatively

stagnant with an average of US$ 9.47 billion.

4.2 The Relationship between FDI Inflows, Natural Resource Abundance and Some

Other Control Variables on Economic Growth of Countries by Income Group

The relationship between economic growth (proxied by GDP per capita growth) with

FDI inflows, natural resource abundance (proxied by natural resource exports), inflation, trade

openness, government spending, population growth, domestic investment levels (proxied by

domestic savings) and institutional quality based on income groups of countries in the world is

simply depicted in a quadrant analysis graph. The data used for this part of the quadrant analysis

is 2019 data by considering the last condition factor of the world economy before the Covid-19

pandemic, which factually and empirically is thought to have affected economic growth in

almost all countries in the world since 2020 until now, but this study does not focus on the

impact of the pandemic. Because the data described in this quadrant analysis section is only

limited to the conditions of 2019, the conditions and characteristics of the variables observed for

each country may be different from a few years before or after the year observed.

The relationship between the share of FDI inflows to GDP and economic growth of

countries in the world based on per capita income groups is illustrated by quadrant analysis in

Figure 7. The average value of the share of FDI inflows to GDP in world countries is 3.94

percent, while the average value of economic growth of countries in the world in 2019 is 4.21

percent. In general, Figure 7 shows that the share of FDI inflows to GDP in world countries tends

to cluster around the average for all groups of per capita income levels with their respective

economic growth values. Furthermore, economic growth rates are higher in countries with higher

per capita income with FDI inflows centered around the average.

In quadrant I (the share of FDI inflows to high GDP and high economic growth) is

seen to be dominated by a group of HIC countries namely Hungary, Singapore, Malta,

Luxembourg, Hong Kong, Seychelles, Macao SAR, China, Estonia and several other HIC

countries and 3 UMIC countries. In quadrant IV (the share of FDI inflows to GDP is high and

economic growth is low) it tends to be dominated by a group of LMIC countries, namely

Mongolia, Cambodia, LIC country Mozambique and several other LMIC and LIC countries and

there are 5 UMIC countries.

Indonesia and most countries belonging to the LMIC group are in quadrant III (low

share of FDI inflows to GDP and low economic growth). Thus, countries in this quadrant need to

pay more attention to efforts to increase economic growth, especially those related to FDI inflow

acceptance policies so that they can be classified into quadrant I like HIC countries.

Based on the quadrant analysis in Figure 7, we can see the pattern of data distribution

that in the group of countries with higher GDP per capita, the level of economic growth also

tends to be higher than the group of countries with lower GDP per capita at any level of the share

of FDI inflows to GDP. Although empirically it cannot be concluded that there is a strong or

weak relationship between the share of FDI inflows to GDP and economic growth, with this

quadrant analysis it can be seen how the pattern of data distribution and the general picture of the

relationship between the two research variables, namely the share of FDI inflows to GDP and

economic growth in countries in the world based on their per capita income groups.

The relationship between natural resources exports (a proxy for natural resource

abundance variables) and the economic growth of countries in the world based on their per capita

income groups is illustrated by quadrant analysis in Figure 8. The average value of the share of

natural resources exports to total exports in world countries in 2019 was 25.53 percent, while

the average value of economic growth as mentioned earlier was 4.21 percent.

If detailed based on the components that make up the proxy for the abundance of natural

resources used in this study, based on data obtained from the World Development Indicators

(WDI) displayed in Table 2, it can be seen that in both the High Income and Low & Middle

income countries, the proportion of fuel exports is greater, namely 12.34 percent and 11.51

percent respectively, followed by the proportion of ore and metal exports of 3.57 percent and

4.03 percent respectively, the proportion of exports of agricultural raw materials looks relatively

lower than the other two components, namely 1.33 percent and 1.38 percent respectively in

2019.

In general, Figure 8 shows that the value of natural resources exports in the world

countries that are the unit of this study tends to be spread across all groups of per capita income

levels with their respective economic growth values. Furthermore, economic growth rates are

higher in countries with higher per capita income regardless of the value of natural resource

exports.

In quadrant I (high share of natural resources exports and high economic growth), it

appears to be dominated by a group of HIC countries, namely Kuwait, Brunei Darussalam,

United Arab Emirates, Saudi Arabia, Bahrain, Norway, Chile, Australia, and several other HIC

countries and 2 UMIC countries, namely Kazakhstan and Russian Federation. Quadrant IV (high

share of natural resources exports and low economic growth) is dominated by other than the

HIC group. In this quadrant, UMIC countries are Azerbaijan, Jamaica, Colombia, Peru, LMIC

countries are Mongolia, Nigeria, Zambia, Benin, Bolivia, Mauritania, Indonesia, LIC countries

Mozambique, Nigeria and several other UMIC, LMIC and LIC countries. If observed in more

detail, the share of natural resources exports in HIC countries in quadrant I tends to be

dominated by fuel exports (Table 4). It can be concluded that in general, developed countries

with a high share of natural resources exports and high economic growth are endowed with

abundant natural resources in fuel commodities. So that excess stocks can be exported to help

increase the economic growth of these countries.

Based on the quadrant analysis in Figure 8, we can see the pattern of data distribution

that in the group of countries with higher GDP per capita, the economic growth rate also tends to

be higher than the group of countries with lower GDP per capita at any level of natural

resources exports share. Although empirically, it cannot be concluded that there is a strong or

weak relationship between the share of natural resource exports and GDP per capita.

However, with this quadrant analysis we can see how the pattern of data distribution

and the general picture of the relationship between the two research variables, namely the share

of natural resources exports and economic growth in countries in the world based on their per

capita income groups.

The relationship between inflation and economic growth of countries in the world

based on their per capita income groups is illustrated by quadrant analysis in Figure 9. The

average value of inflation in world countries in 2019 was 6.72 percent, while the average value

of economic growth as mentioned earlier was 4.21 percent.

In general, Figure 8 shows that the inflation rate in the world countries that become the

unit of this study tends to be centered around the average value for all income per capita level

groups with their respective economic growth rates, except for one LMIC country, Zimbabwe,

where the inflation rate is very high at 440.83 percent with an economic growth rate of 3.36

percent and one UMIC country, Argentina, where the inflation rate is quite high at 50.92 percent

with an economic growth rate of 4.34 percent. Furthermore, economic growth rates are higher in

countries with higher per capita income with inflation values centered on the average value.

GDP per capita growth, PPP (%)

Indonesia as one of the UMIC countries is included in quadrant III (low inflation rate

and low economic growth) in 2019. If analyzed over the past twenty-four years, Indonesia's

annual inflation still tends to be high, for the period 1996-2019 the average annual inflation rate

in Indonesia reached an average of 11.85% per year. The government and Bank Indonesia agreed

on five strategic steps to strengthen inflation control in the High Level Meeting of the Central

Inflation Control Team (HLM TPIP) in 2022. The strategic measures are aimed at consistently

keeping inflation within the target range of 3.0%±1% and continuing to maintain the momentum

of national economic recovery. The strategic measures include:

1) Strengthen policy coordination to maintain macroeconomic stability and boost the

momentum of national economic recovery;

2) Mitigate the impact of upside risks, including the impact of normalization of global

liquidity policies and the increase in world commodity prices on inflation and people's

purchasing power;

3) Maintain volatile food inflation within the range of 3.0-5.0%. This is done by maintaining

supply availability and smooth distribution, especially ahead of the National Religious

Holidays (HBKN). Implementation of the strategy is focused, among others, on optimizing

the use of technology and digitalization of upstream-downstream agriculture, developing

connectivity, and strengthening inter-regional cooperation;

4) Strengthen the synergy of policy communication to support the management of public

inflation expectations;

5) Strengthening the coordination of the Central and Regional Governments in controlling

inflation through the organization of the National Coordination Meeting (Rakornas) on

Inflation Control 2022 with the theme: "Digitalization of Food MSMEs for Access and

Price Stabilization".

Based on the quadrant analysis in Figure 9, it can be seen that in the group of countries

with higher GDP per capita, the level of economic growth also tends to be higher than the group

of countries with lower GDP per capita at any level of inflation. Although empirically it cannot

be concluded that there is a strong or weak relationship between the inflation rate and economic

growth, but with this quadrant analysis it can be seen how the pattern of data distribution and the

general picture of the relationship between the two research variables, namely the inflation rate

and economic growth in countries in the world based on its per capita income group.

The relationship between trade openness and economic growth of countries in the

world based on per capita income groups is illustrated by quadrant analysis in Figure 10. The

average value of the share of trade openness to GDP in world countries is 92.66 percent, while

the average value of economic growth as mentioned earlier is 4.21 percent. In general, Figure 10

shows that the level of trade openness in the world countries that are the unit of this study tends

to be spread across all income per capita level groups with their respective economic growth

values. Furthermore, economic growth rates are higher in countries with higher per capita

income regardless of the level of trade openness.

Quadrant I (high level of trade openness and high economic growth) is dominated by

HIC countries, namely Luxemburg Hong Kong SAR China, Singapore, Malta, Ireland,

Seychelles United Arab States and several other HIC countries, as well as UMIC countries,

namely Malaysia, Bulgaria and four other UMIC countries. Quadrant IV (high level of trade

openness and low economic growth) is dominated by other than the HIC group. In this quadrant,

UMIC countries are Georgia and two other UMIC countries, LMIC countries are Vietnam,

Cambodia, Belize and five other LMIC countries, and one LIC country Mozambique.

Indonesia, which falls into quadrant III (low trade openness and low economic growth),

experienced a decline in the world trade openness index in 2019, ranking 68th in the Global

Index for Economic Openness. Indonesia's trade to Gross Domestic Product (GDP) ratio of 43

percent is the lowest in ASEAN. This indicates a low level of economic openness that is not in

line with its economic growth. Therefore, Increasing trade cooperation agreements with other

countries needs to be carried out by Indonesia and other countries other than in quadrant I to be

able to encourage trade performance (exports and imports) to grow and expand Indonesia's trade

market so that later it will have an impact on the economy (Fitriani 2021).

Based on the quadrant analysis in Figure 10, we can see the pattern of data distribution

that in the group of countries with higher GDP per capita, the level of economic growth also

tends to be higher than the group of countries with lower GDP per capita at any level of

economic openness. Although empirically it cannot be concluded that there is a strong or weak

relationship between the level of economic openness and economic growth, with this quadrant

analysis we can see how the pattern of data distribution and the general picture of the

relationship between the two research variables, namely the level of economic openness and

economic growth in countries in the world based on their per capita income groups.

The relationship between government expenditure and economic growth of countries

in the world based on their per capita income groups is illustrated by quadrant analysis in Figure

11. The average value of the share of government expenditure to GDP in world countries in 2019

was 16.14 percent, while the average value of economic growth as mentioned earlier was 4.21

percent. In general, Figure 11 shows that the value of the share of government expenditure to

GDP in the world countries that are the unit of this study tends to be spread across all groups of

per capita income levels with their respective economic growth values. Furthermore, economic

growth rates are higher in countries with higher per capita income regardless of the level of

government expenditure.

In quadrant I (high share of government spending and high economic growth), it

appears to be dominated by the HIC group of countries, namely Kuwait, Brunei Darussalam,

Saudi Arabia, Austria, Qatar, and several other HIC countries as well as UMIC countries,

namely Botswana, Russian Federation and five other UMIC countries. Quadrant IV (high share

of government spending and low economic growth) is dominated by countries other than HICs.

In this quadrant, UMIC countries are Namibia, Brazil, Jordan, LMIC countries are Morocco,

Bolivia, Zambia, LIC countries Burundi, Mozambique, Burkina Faso and several other UMIC

and LMIC countries. Indonesia as one of the UMIC countries is included in quadrant III (low

government spending share and low economic growth) in 2019.

Based on the quadrant analysis in Figure 11, we can see the pattern of data distribution

that in the group of countries with higher GDP per capita, the level of economic growth also

tends to be higher than the group of countries with lower GDP per capita at any level of

government expenditure share. Although empirically it cannot be concluded that there is a strong

or weak relationship between the share of government spending and economic growth, with this

quadrant analysis we can see how the pattern of data distribution and the general picture of the

relationship between the two research variables, namely the share of government spending and

economic growth in countries in the world based on their per capita income groups.

The relationship between population growth and and economic growth of countries in

the world based on their per capita income groups is illustrated by quadrant analysis in Figure

12. The average value of population growth rates in world countries in 2019 was 1.17 percent,

while the average value of economic growth as mentioned earlier was 4.21 percent.

In general, Figure 12 shows that the values of population growth rates in the world

countries that are the units of this study tend to be spread out for all groups of per capita income

levels with their respective economic growth values. Furthermore, it can also be seen that

economic growth rates are higher in countries with higher per capita income, but conversely

population growth rates are higher in countries with lower per capita income.

Quadrant I (high population growth rate and high economic growth) is dominated by HIC

countries, namely Bahrain, Malta, Luxemburg, Qatar, Saudi Arabia and several other HIC

countries and UMIC countries, namely Malaysia, Turkey and three other UMIC countries.

GDP per capita growth, PPP (%)

Quadrant IV (high population growth rate and low economic growth) is dominated by other than

the HIC group. In this quadrant, UMIC countries are Guatemala, Jordan, South Africa, LMIC

countries are Senegal, Ghana, Pakistan, Zimbabwe, Morocco, as well as several other UMICs

and LMICs and all LIC countries are included in this quadrant.

Based on the quadrant analysis in Figure 12, we can see the pattern of data distribution

that in the group of countries with higher GDP per capita, the level of economic growth also

tends to be higher than the group of countries with lower GDP per capita at any level of

population growth. Although empirically it cannot be concluded that there is a strong or weak

relationship between population growth rate and economic growth, with this quadrant analysis

we can see how the pattern of data distribution and the general picture of the relationship

between the two research variables in countries in the world based on their per capita income

groups.

The relationship between domestic investment and economic growth of countries in

the world based on their per capita income groups is illustrated by quadrant analysis in Figure

13. The average value of domestic investment levels in world countries in 2019 was 22.09

percent, while the average value of their economic growth as mentioned earlier was 4.21 percent.

In quadrant I (high level of domestic investment and high economic growth), it can be

seen that it is dominated by a group of HIC countries, namely Singapore, United Arab States,

Brunei Darussalam, Bahrain, Korea Rep, Norway, Spain, Austria, and several other HIC

countries and UMIC countries, namely China, Kazakhstan, Thailand, Turkiye and five other

UMIC countries. Quadrant IV (high level of domestic investment and low economic growth) is

dominated by other than the HIC and LIC groups. In this quadrant, UMIC countries are Ecuador,

LMIC countries are Indonesia, Tanzania, Mauritania, Morocco, Zambia, and several other

UMICs and LMICs.

Based on the quadrant analysis in Figure 13, we can see the pattern of data distribution

that in the group of countries with higher GDP per capita, the level of economic growth also

tends to be higher than the group of countries with lower GDP per capita at any level of domestic

investment. Although empirically it cannot be concluded that there is a strong or weak

relationship between the level of domestic investment and economic growth, with this quadrant

analysis we can see how the pattern of data distribution and the general picture of the

relationship between the two research variables, namely the level of domestic investment and

economic growth in countries in the world based on their per capita income groups.

4.3 Panel Data Regression Results of the Effect of FDI and Natural Resource

Abundance on Economic Growth with Four Separate Models

In the inferential analysis section, to answer the second research objective, panel data

regression analysis was conducted. This study produces four outputs of panel data regression

results consisting of an overall group of 124 countries and grouping based on the country's per

capita income level, namely High Income Countries (HIC), Middle Income Countries (MIC) and

Low Income Countries (LIC). Each of these models is analyzed separately, so that each stage of

the modeling is also carried out separately. The modeling stages carried out are the selection of

the best model, non-heteroscedasticity assumption test, residual cross-section dependence test

and non-multicollinearity test.

A. Overall Group of 124 Countries

1) Selection of the best model

The best model selection stage initially conducted a Chow test to choose the best

model between the common effect model (CEM) or the fixed effect model (FEM) for the total

data of 124 countries, with the research hypothesis:

H0 = prob. Chi-square > 0.05 then the common effect model (CEM) is selected.

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then reject H0 , the fixed effect model

(FEM) is selected. Furthermore, the Hausman test is conducted to select the best model between the fixed

effect model (FEM) or the random effect model (REM). The research hypothesis is as follows:

H0 = prob. Chi-square > 0.05 then the random effect model (REM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then

reject H0 , the fixed effect model (FEM) is selected. Thus, from the results of the Hausman test

that has been carried out, the fixed effect model (FEM) is more appropriate among the other two

models, namely the common effect model (CEM) and the random effect model (REM).

Therefore, for the overall group of countries, the econometric output results with the FEM model

will be analyzed and interpreted.

2) Testing the assumption of non-heteroscedasticity

With the selection of the fixed effect model (FEM), the Glejser test needs to be done to

see whether there is a violation of the non-heteroscedasticity assumption in the regression model

or not. The Glejser test is carried out by generating a series of reabs = abs(resid) in the Eviews-

12 data processing software. The next step is to regress all independent variables with the reabs

variable as the dependent variable. Then the output of the Glejser test will be generated. From

the output results, what needs to be considered is the probability value on each independent

variable. If the prob.<0.05 value on the independent variable then there is heteroscedasticity.

Conversely, if the prob.>0.05 value, it is free from violations of heteroscedastic assumptions.

From the results of the Glejser test, it can be seen that there is a violation of heteroscedasticity,

namely in the NR variable (prob. 0.0000), INF (prob. 0.0000), TRADE (prob. 0.0000), GE (prob.

0.0000), and IQUAL (prob. 0.0011).

3) Testing residual cross-section dependence

Residual cross-section dependence test is conducted to check whether there is serial

correlation between cross-sections. If the prob.value <0.05 then there is cross-section serial

correlation. Conversely, if prob.>0.05 then free from cross-section serial correlation. The test

results show that in all tests the value of prob.<0.05 there is cross-section serial correlation. So,

to overcome these two problems, a fixed effect regression model with the dependent variable

LOG(GDP3P) and weighted white/robust standard error regression was re-run to overcome the

heteroscedasticity and autocorrelation problems. According to Baltagi (2021), the application of

white/robust standard errors can overcome the problem of heteroscedasticity and/or serial

correlation. The output of these regression results will then be used for analysis and

interpretation.

4) Multicollinearity test

Next, from the multicollinearity test results, it can be seen that there is no correlation

between variables worth more than 0.8 (Spearman's Rho Correlation). It can be concluded that

the independent variables in the model do not have a linear relationship with other independent

variables.

5) Regression results of research data

The effect of FDI on economic growth by considering the host country's natural

resource abundance factor and several other control variables in the overall group of 124

countries is shown in Table 6. The results of determining the best regression model that has been

done previously show that the fixed effect model (FEM) with white/robust standard error

regression weighting is selected as a better model for analysis and interpretation. High-Income

Countries (HIC) group

6) Selection of the best model

The best model selection stage initially conducted a Chow test to select the best model between

the common effect model (CEM) or the fixed effect model (FEM) for the data category of the 47

High Income Countries (HIC) countries with the research hypothesis:

H0 = prob. Chi-square > 0.05 then the common effect model (CEM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then reject H0 , the fixed effect model

(FEM) is selected. Furthermore, the Hausman test is conducted to select the best model between the fixed

effect model (FEM) or the random effect model (REM). The research hypothesis is as follows:

H0 = prob. Chi-square > 0.05 then random effect model (REM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then reject H0 , the fixed effect model

(FEM) is selected. Thus, from the results of the Hausman test that has been carried out, the FEM model is

more appropriate among the other two models, namely the CEM model and the REM model. Therefore,

for the overall group of countries, the econometric output results with the FEM model will be analyzed

and interpreted.

7) Testing the assumption of non-heteroscedasticity

With the selection of the fixed effect model (FEM), the Glejser test needs to be done to

see if there is a violation of the assumption of non-heteroscedasticity in the regression model or

not. The stages of conducting the Glejser test are the same as those previously described. From

the Glejser test results, it can be seen that there is a violation of heteroscedasticity in the NR

(prob. 0.0000) and TRADE (prob. 0.0011) variables.

8) Testing residual cross-section dependence

Furthermore, the residual cross-section dependence test is conducted to check whether

there is serial correlation between cross-sections. The test results show that in all tests the

prob.value <0.05 there is cross-section serial correlation. So, to overcome these two problems, a

fixed effect regression model with the dependent variable LOG(GDP3P) and weighted

white/robust standard error regression is performed again to overcome the heteroscedasticity

and autocorrelation problems. The output of these regression results will then be used for

analysis and interpretation.

9) Multicollinearity test

Next, from the multicollinearity test results, it can be seen that there is no correlation

between variables worth more than 0.8 (Spearman's Rho Correlation). It can be concluded that

the independent variables in the model do not have a linear relationship with other independent

variables.

B. Middle Income Countries (MIC) group

1) Selection of the best model

The best model selection stage initially conducted a Chow test to select the best model between

the common effect model (CEM) or the fixed effect model (FEM) for the Middle Income

Countries (MIC) data category of 67 countries with the research hypothesis:

H0 = prob. Chi-square > 0.05 then the common effect model (CEM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then

reject H0 , the fixed effect model (FEM) is selected. Furthermore, the Hausman test is conducted

to select the best model between the fixed effect model (FEM) or the random effect model

(REM). The research hypothesis is as follows:

H0 = prob. Chi-square > 0.05 then random effect model (REM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then

reject H0 , the fixed effect model (FEM) is selected. Thus, from the results of the Hausman test

that has been carried out, the fixed effect model (FEM) is more appropriate among the other two

models, namely the common effect model (CEM) and the random effect model (REM).

Therefore, for the overall group of countries, the econometric output results with the FEM model

will be analyzed and interpreted.

2) Testing the assumption of non-heteroscedasticity

With the selection of the fixed effect model (FEM), the Glejser test needs to be done to see

whether there is a violation of the non-heteroscedasticity assumption in the regression model or

not. The stages of conducting the Glejser test are the same as those previously described. The

results of the Glejser test show that there is a violation of heteroscedasticity except for the GE

variable, namely the variables FDI (prob. 0.0000), NR (prob. 0.0001), FDI * NR (prob. 0.0008),

INF (prob. 0.0000), TRADE (prob. 0.0000), POP (prob. 0.0209), DS (prob. 0.0001) and IQUAL

(prob. 0.0000).

3) Testing residual cross-section dependence

Furthermore, the residual cross-section dependence test is conducted to check whether there is

serial correlation between cross-sections. The test results show that in all tests the prob.value

<0.05 there is cross-section serial correlation. So, to overcome these two problems, a fixed effect

regression model with the dependent variable LOG(GDP3P) and weighted white/robust standard

error regression is performed again to overcome the heteroscedasticity and autocorrelation

problems. The output of these regression results will then be used for analysis and interpretation.

4) Multicollinearity test

Next, from the multicollinearity test results, it can be seen that there is no correlation between

variables worth more than 0.8 (Spearman's Rho Correlation). It can be concluded that the

independent variables in the model do not have a linear relationship with other independent

variables.

5) Regression results of research data

The effect of FDI on economic growth by considering the host country's natural resource

abundance factor and several other control variables in the middle income countries (MIC) group

is shown in Table 8. The results of determining the best regression model that has been done

previously show that the fixed effect model (FEM) with white/robust standard error regression

weighting is selected as the better model for analysis and interpretation. Then, for partial testing

using the t-test statistic which is then summarized in Table 8 as follows:

The Effect of Foreign Direct Investment (FDI) on Economic Growth

Table 10 shows that based on the total country model at the 1% significance level, it is

found that FDI has a positive significant effect on economic growth as seen from the coefficient

value of 0.000632. This means that if there is an increase in FDI inflow by 1 percent, then

economic growth is expected to increase by 0.000632 percent, ceteris paribus.

When viewed based on the per capita income group, at a significance level of 1%, it

was found that in High Income Countries (HIC) and Middle Income Countries (MIC), FDI has a

significant positive effect on economic growth with a coefficient value of 0.000539 and

0.015543 respectively. This means that in the HIC and MIC country groups, if there is an

increase in FDI inflow by 1 percent, then economic growth is expected to increase by 1 percent

increased by 0.000539 percent and 0.015543 percent, respectively, ceteris paribus. Capital

inflows can serve as a strong stimulus to a country's economy and can also lead to the transfer of

new technologies that can help increase productivity, thus leading to overall economic growth.

According to De Gregorio (2003), FDI allows a home country to bring in new technology and

knowledge that is not available to domestic investors, and in this way increases productivity

growth throughout the host country's economy. FDI can also bring in expertise that the country

does not have, and foreign investors may have better access to global markets. This result is in

line with the research findings of Mehic et al. (2013); Alvarado et al. (2017); Hayat (2017);

Sultanuzzaman et al. (2018); Dinh et al. (2019); Shittu et al. (2020); Zeeshan et al. (2020);

Mohamed et al. (2021); Orji et al. (2021); Ahmad et al. (2022) who found that an increase in

FDI inflows is one of the crucial aspects that will stimulate economic growth. Asghari et al.

(2014) argue that FDI can contribute to economic growth in several ways. One is by increasing

domestic savings, which can help overcome capital accumulation limitations. Another is by

facilitating technology transfer, which can lead to more efficient use of resources and increased

productivity. In addition, FDI can help boost exports by increasing the capacity and

competitiveness of domestic production.

In contrast to the results in other groups, it was found that FDI has a significant negative effect

on economic growth in Low Income Countries (IC) seen from the coefficient value of -0.011104.

This means that in the LIC group of countries if there is an increase in FDI inflow by 1 percent,

then economic growth is expected to decrease by -0.011104 percent, ceteris paribus. In

underdeveloped countries, the effect of the presence of FDI will be very small on economic

growth due to the existence of threshold externalities at the level of development, such as

inadequate levels of education, technology, infrastructure (OECD 2002). In addition, the level of

institutional quality also tends to be low. As evidenced by the results of the quadrant analysis

shown in Figure 13, only HIC countries are in quadrant I, which is a high level of institutional

quality and high economic growth, while other countries are still outside quadrant I with a

relatively low level of institutional quality. Institutional heterogeneity and differences in

government efficiency and political freedom are responsible for differences in capital

accumulation and labor productivity. In addition, according to structural dependency theory, the

underdevelopment that occurs in third world countries is a result of the exploitative structure of

the world economy so that the economic surplus of these countries shifts to advanced industrial

capitalist countries, one of which is through FDI, so that it will hamper economic growth in these

countries. So it can be concluded that the presence of FDI in the LIC group of countries tends to

be detrimental to the host country. According to De Gregorio (2003), only countries that have

reached a certain level of income can absorb new technologies and benefit from technological

diffusion, because most of the population has a high level of education which is relatively easier

to absorb the benefits of new technology, so as to reap the benefits offered by FDI to the

economy. This result is in line with the research findings of Alvarado et al. (2017); Bouchoucha

and Bakari (2019); Kolisi (2021); Meivitawanli (2021).

Based on these results, it is suspected that the effect of FDI inflow on economic

growth is different in each group of countries based on the level of income per capita. The effect

of FDI inflow is greater in the Middle Income Countries (MIC) group followed by the High

Income Countries (HIC) group, while in the Low Income Countries (LIC) group the effect of

FDI inflow is lower and even negative on economic growth. So it is concluded that FDI inflow

will not have the same effect on economic growth for countries based on their per capita income

groups.

The Effect of Natural Resource Abundance on Economic Growth

Based on the total country model at the 1% significance level, it is found that the

abundance of natural resources proxied by the share of natural resources exports has a positive

significant effect on economic growth as seen from the coefficient value of 0.002177. This

means that if there is an increase in the abundance of natural resources by 1 percent, then

economic growth is expected to increase by 0.002177 percent, ceteris paribus.

When viewed based on the per capita income group, at the 1% significance level, it is

found that the abundance of natural resources in High Income Countries (HIC) and Middle

Income Countries (MIC) has a significant positive effect on economic growth with a coefficient

value of 0.005096 and 0.003674, respectively. This means that in the HIC and MIC country

groups, if there is an increase in the abundance of natural resources by 1 percent, economic

growth is expected to increase by 0.005096 percent and 0.003674 percent, respectively. percent,

ceteris paribus. According to De Gregorio (2003), the high starting level of human capital in

Scandinavian countries and in other high-income countries with extensive natural resource bases

helps explain why natural resource exploitation is not detrimental to their economic growth. In

addition, industrialized countries got richer because they protected their natural resources, the

need for raw materials was reduced due to new discoveries. Natural resource abundance tends to

increase per capita income in countries with less government intervention, more sound money,

better protection of property rights, which are less open to international markets, and/or exhibit

lower government corruption (Kim and Lin 2017). Furthermore, Erum and Hussain (2019) claim

that proper management and governance of natural resources can promote economic growth.

This result is in line with the research findings of Redmond and Nasir (2020); Haseeb et al.

(2021); Adika (2022).

In contrast to the results in other groups, it was found that the abundance of natural

resources has a significant negative effect on economic growth in Low Income Countries (LIC)

as seen from the coefficient value of -0.001823. This means that in the LIC country group, if

there is an increase in natural resource abundance by 1 percent, then economic growth is

expected to decrease by -0.001823 percent, ceteris paribus.

Some of the reasons why natural resources can negatively affect economic growth

according to Satti et al. (2014) are exchange rate volatility, inefficient allocation of natural

resource proceeds due to corruption, a false sense of economic capability with an abundance of

natural resources resulting in poorly designed governance, and a lack of prioritization and

development of human resources to manage these natural resources. There is strong evidence

that having an educated workforce is key to economic growth, especially to fully utilize FDI and

natural resources (De Gregorio 2003). In addition, according to Barbier (2019), increasing

economic dependence on natural resources is associated with poor economic performance. The

implication for Low Income Country (LIC) countries is that the "take-off" stage towards

sustainable and structurally balanced economic growth and development is still a long time away,

and thus their overall economic dependence on natural resources will persist in the long run as

well. This result is in line with the research findings of Rahim et al. (2021); Tabash et al. (2022)

who found that natural resources can negatively affect economic growth because natural

resources in these countries are not used in the right way to obtain their economic benefits.

Based on these results, it is suspected that the effect of natural resource abundance on

economic growth is different in each group of countries based on the level of income per capita.

The effect of natural resource abundance is greater in the High Income Countries (HIC) group

followed by the Middle Income Countries (MIC) group, while in the Low Income Countries

(LIC) group the effect of natural resource abundance is lower and even negative on economic

growth. So it is concluded that the abundance of natural resources will not have the same effect

on economic growth for countries based on their per capita income groups.

The Effect of Inflation on Economic Growth

Based on the total country model in Table 10, it can be seen that with a significance

level of 1%, it is found that the inflation rate has a significant negative effect on economic

growth as seen from the coefficient value of -0.001010. This means that if the inflation rate

increases by 1 percent, then economic growth is expected to decrease by 0.001010 percent,

ceteris paribus.

When viewed based on the per capita income group, it is found that the inflation rate

also has a significant and negative effect on economic growth in High Income Countries (HIC)

with a significance level of 1% and in Middle Income Countries (MIC) with a significance level

of 5% with a coefficient value of -0.003758 and -0.000906 respectively. This means that in the

HIC and MIC country groups, if the inflation rate increases by 1 percent, then economic growth

is expected to decrease by 0.003758 percent and 0.000906 percent, respectively, ceteris paribus.

While in Low Income Countries (LIC) the inflation rate also has a negative but insignificant

effect on economic growth with a coefficient value of -0.000762. This means that in the LIC

country group if the inflation rate increases by 1 percent, then economic growth is expected to

decrease by 0.000762 percent. These results are in line with the research findings of Dammak

and Helali (2017); Ndoricimpa (2017); Sequeira (2021) which states that a low inflation rate can

encourage economic growth.

Effect of Trade Openness on Economic Growth

Based on the total country model in Table 10, it can be seen that with a significance

level of 1%, it is found that the level of trade openness has a positive significant effect on

economic growth as seen from the coefficient value of 0.002104. This means that if the level of

trade openness increases by 1 percent, then economic growth is expected to increase by

0.002104 percent, ceteris paribus.

When viewed based on the income per capita group, it is found that the level of trade

openness also has a significant and positive effect on economic growth in High Income

Countries (HIC) and Low Income Countries (LIC) at the 1% significance level with a coefficient

value of 0.002985 and 0.005450 respectively. This means that in the HIC and LIC country

groups, if the level of trade openness increases by 1 percent, then economic growth is expected

to increase by 0.002985 percent and 0.005450 percent, respectively, ceteris paribus. While in

Middle Income Countries (MIC), the level of trade openness also has a positive but insignificant

effect on economic growth.

Effect of Government Spending on Economic Growth

Based on the total country model in Table 10, it can be seen that with a significance

level of 1%, it is found that government spending has a significant positive effect on economic

growth as seen from the coefficient value of 0.012571. This means that if the inflation rate

increases by 1 percent, then economic growth is expected to increase by 0.012571 percent,

ceteris paribus. When viewed based on the per capita income group, it is found that government

spending also has a significant and positive effect on economic growth in Middle Income

Countries (MIC) at the 1% significance level with a coefficient value of 0.018172. This means

that in the MIC group, if government spending increases by 1 percent, then economic growth is

expected to increase by 0.018172 percent, ceteris paribus. While in the High Income Countries

(HIC) government spending also has a positive but insignificant effect on economic growth with

a coefficient value of 0.002648. This means that in the HIC group of countries if government

spending increases by 1 percent, then economic growth is expected to increase by 0.002648

percent. In contrast to the results in other groups, it was found that government spending has a

negative but insignificant effect on economic growth in Low Income Countries (HIC).

Effect of Population Growth on Economic Growth

Based on the total country model in Table 10, it is found that population growth has a

positive but insignificant effect on economic growth as seen from the coefficient value of

0.007194. This means that if population growth increases by 1 percent, then economic growth is

expected to increase by 0.007194 percent, ceteris paribus. When viewed based on the per capita

income group, it is found that population growth also has a positive and significant effect on

economic growth in High Income Countries (HIC) at the 1% significance level with a coefficient

value of 0.013366. This means that in the HIC group of countries, if the population growth rate

increases by 1 percent, then economic growth is expected to increase by 0.013366 percent

respectively, ceteris paribus.

In contrast to the results in the High Income Countries (HIC) group, in the Low

Income Countries (LIC) population growth has a negative and significant effect on economic

growth with a significance level of 1% with a coefficient value of -0.028585. This means that in

the LIC group, if the population growth rate increases by 1 percent, then economic growth is

expected to decrease by 0.028585 percent, ceteris paribus.

While in Middle Income Countries (MIC) population growth also has a negative but

insignificant effect on economic growth. According to the United Nation - Department of

Economic and Social Affairs (DESA) (2022), rapid population growth makes it difficult for low-

and lower-middle-income countries to pay for the increased per capita public spending needed to

eradicate poverty, end hunger and malnutrition, and ensure universal access to health care,

education and other essential services, thus resulting in slowing economic growth.

The Effect of Domestic Savings on Economic Growth

Based on the total country model in Table 10, it can be seen that with a significance

level of 1%, it is found that the domestic savings rate has a positive significant effect on

economic growth as seen from the coefficient value of 0.012191. This means that if the domestic

savings rate increases by 1 percent, then economic growth is expected to increase by 0.012191

percent, ceteris paribus.

When viewed based on the income per capita group, it is found that the domestic

savings rate also has a significant and positive effect on economic growth in High Income

Countries (HIC), Middle Income Countries (MIC) and Low Income Countries (LIC) with a

significance level of 1% with a coefficient value of 0.007559, 0.013717 and 0.012776

respectively. This means, if the domestic savings rate increases by 1 percent, then economic

growth is expected to increase in HIC, MIC and LIC respectively by 0.007559 percent, 0.013717

percent and 0.012776 percent, ceteris paribus.

Effect of Institutional Quality on Economic Growth

Based on the total country model in Table 10, it can be seen that with a significance

level of 1%, it is found that institutional quality has a positive significant effect on economic

growth as seen from the coefficient value of 0.445932. This means that if institutional quality

increases by 1 percent, then economic growth is expected to increase by 0.445932 percent,

ceteris paribus. When viewed based on the per capita income group, it is found that institutional

quality also has a significant and positive effect on economic growth in High Income Countries

(HIC), Middle Income Countries (MIC) and Low Income Countries (LIC) with a significance

level of 1% with a coefficient value of 0.935819, 0.438520 and 0.263311 respectively. This

means, if the institutional quality increases by 1 percent, then economic growth is expected to

increase in HIC, MIC and LIC respectively by 0.935819 percent, 0.438520 percent and 0.263311

percent, ceteris paribus.

4.4 Panel Data Regression Results of the Effect of FDI and Natural Resource

Abundance on Economic Growth with Dummy Slope Model

Panel data regression in this study was conducted with two approaches as a

comparison to see the effect of FDI and the abundance of natural resources on economic growth,

namely first by modeling each for each country income group with the assumption that the effect

of each independent variable is different for each country income group. Second, modeling is

done using a dummy slope model with the assumption that the influence of the independent

variables FDI and the abundance of natural resources is different for each country income group,

but the influence of the control variables is the same for each country income group.

Both approaches aim to analyze the effect of FDI, natural resource abundance and

several control variables on economic growth. Similar to the panel data regression results with

the previous approach, the panel data regression results with the dummy slope model also found

that FDI and natural resource abundance have a positive effect on economic growth in the HIC

and MIC country groups, while in the LIC country group, FDI and natural resource abundance

have a negative effect on economic growth. The results of the analysis of the natural resource

abundance factor in changes in the effect of foreign direct investment (FDI) on economic growth

are also found to be the same as the previous results, namely changes in the value of the NR

variable are expected to change the effect of FDI on economic growth in a negative direction in

the High Income Countries (HIC) and Middle Income Countries (MIC) groups, while in the

Income Countries (LIC) changes in the value of the NR variable are expected to change the effect

of FDI on economic growth in a positive direction.

In more detail, the modeling stages carried out are the same as the previous modeling and can be

explained as follows, namely the selection of the best model, non-heteroscedasticity assumption

test, residual cross-section dependence test and non-multicollinearity test.

1) Selection of the best model

The stage of selecting the best model was initially carried out the Chow test to choose the best

model between the common effect model (CEM) or the fixed effect model (FEM), with the

research hypothesis:

H0 = prob. Chi-square > 0.05 then the common effect model (CEM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , fixed effect model (FEM) selected

The test results obtained that the prob value. Chi-square = 0.0000 then reject H0 , the

fixed effect model (FEM) is selected. Furthermore, the Hausman test was conducted to select the

best model between the fixed effect model (FEM) or the random effect model (REM). The

research hypothesis is as follows:

H0 = prob. Chi-square > 0.05 then random effect model (REM) is selected

H1 = prob. Chi-square < 0.05 then reject H0 , the fixed effect model (FEM) was selected The test

results obtained that the prob value. Chi-square = 0.0000 then

reject H0 , the fixed effect model (FEM) is selected. Thus, from the results of the Hausman test

that has been carried out, the FEM model is more appropriate among the other two models,

namely the CEM model and the REM model. Therefore, for the overall group of countries, the

econometric output results with the FEM model will be analyzed and interpreted.

2) Testing the assumption of non-heteroscedasticity

With the selection of the fixed effect model (FEM), the Glejser test needs to be done to

see whether there is a violation of the non-heteroscedasticity assumption in the regression model

or not. The Glejser test is performed by generating series reabs = abs(resid) in the Eviews-12

data processing software. The next step is to regress all independent variables with the reabs

variable as the dependent variable. Then the output of the Glejser test will be generated. From

the output results, what needs to be considered is the probability value on each independent

variable. If the prob.<0.05 value on the independent variable then there is heteroscedasticity.

Conversely, if the prob.>0.05 value, it is free from violations of heteroscedastic assumptions.

From the results of the Glejser test, it can be seen that there is a violation of heteroscedasticity,

namely in the variables FDI (prob. 0.0024), NR (prob. 0.0000), FDI * NR (prob. 0.0054), FDI *

D1 (prob. 0.0038), NR * D1 (prob. 0.0000), INF (prob. 0.0000), TRADE (prob. 0.0000), and GE

(prob. 0.0000).

3) Testing residual cross-section dependence

Residual cross-section dependence test is conducted to check whether there is serial correlation

between cross-sections. If the prob.value <0.05 then there is cross-section serial correlation.

Conversely, if prob.>0.05 then free from cross-section serial correlation. The test results show

that in all tests the value of prob.<0.05 there is cross-section serial correlation. So, to overcome

these two problems, a fixed effect regression model with the dependent variable LOG(GDP3P)

and weighted white/robust standard error regression was re-run to overcome the

heteroscedasticity and autocorrelation problems. According to Baltagi (2021), the application of

white/robust standard errors can overcome the problem of heteroscedasticity and/or serial

correlation. The output of these regression results will then be used for analysis and

interpretation.

Conclusion:

Based on the results of the analysis and discussion in the previous section, it can be concluded as

follows:

1. The results of the descriptive analysis show that the development of FDI inflows to

various countries has increased rapidly over the past 25 years with considerable

variation/inequality between groups of countries based on the level of per capita income.

FDI inflows are higher in countries that are large in economic size as measured by high

levels of per capita income. The results of the quadrant analysis show Economic growth

rates are higher in the group of countries with higher levels of per capita income regardless of the

distribution of FDI inflows and the abundance of natural resources.

2. The estimation of the economic growth model using panel data regression analysis was

conducted using both approaches. The results, whether done with separate regression

models with each model for three country income groups or panel data regression with

slope dummy models, both show that FDI inflows and natural resource abundance have a

positive significant effect on host country economic growth. However, with a higher

increase in natural resource abundance, it is expected that the total effect of FDI influence

on economic growth will decrease and will even turn out to be negative. In the High

Income Countries (HIC) and Middle Income Countries (MIC) groups, FDI inflows and

abundance of natural resources also have a significant positive effect on economic

growth, an increase in the abundance of higher natural resources is also expected to make

the total effect of FDI on economic growth will decrease and will even turn into a

negative value. Different results in the Low Income Countries (LIC) group, FDI inflows

and abundance of natural resources have a negative significant effect on economic

growth, but with a higher increase in the abundance of natural resources, it is suspected

that the total effect of FDI on economic growth will increase and will even turn into a

positive value.

ECN 360_ARIZONA STATE UNIVERSITY_ASSIGNMENT_FOREIGN DIRECT INVESTMENT (FDI) NATURAL RESOURCE ABUNDANCE AND ECONOMIC GROWTH

Students also viewed