The Possible Responsiveness of FDI to Electric Power Consumption in Different Income Level

Modern energy in the form of electricity is vital for economic activities such as for getting clean water, and healthcare, for getting stable and effective lighting, heating, and cooking. However, in developing countries there is a huge shortage of electricity and big gaps in access, while access to electricity in developed countries almost reached a hundred percent. Hence, the purpose of this paper is to shows the possible responsiveness of FDI to electric power consumption in different income-level. Hence, this study aims to address questions, including how is the relationship between economic activities such as FDI and electricity consumption in different income-level. For motivating the research, 131 countries data have been collected from WDI and US-EIA from 1992 to 2016 and both quantitative and qualitative methods used. The cross-country regression result shows that there exists an inverse-U shaped relationship between EPC and FDI net inflow because most high-income countries have a high level of EPC and therefore EPC becomes less important for them to attract FDI. But when we separate the total sample into two, EPC can significantly increase net FDI inflow for middle & low-income countries because for these countries, especially for low-income countries, sufficient electricity supply is important for FDI inflow.


Introduction
Economic growth is the most powerful instrument for decreasing poverty and improving the quality of human wellbeing by creating the resources required for human development. It is motivated by several factors, including process, product, and organizational improvements established on technological transformation. Hence, the economic growth in the past several years have shown big progress. It is well known that for this progress different sectors have contributed a lot. Among those sectors, foreign direct investment and energy played an important role in each country. To know the importance of energy in economic growth, it is essential; to begin with, the role of energy (Apergis & Payne, 2011) in production because we can see that energy input is inducing production outputs very significantly. A more vibrant economy will have to use more energy to keep up with a higher level of production, as well as people's increasing demand for energy following the increased income level (Sadorsky, P., 2009). As income per capita of a country increases, demand for electric energy also increases to secure its people's well-being as well as to build a strong and productive economic foundation.
Reliance and demand for energy of countries have been growing due to increased innovations, industrialization, and globalization. The finding of Stern (2011) confirms that energy plays a matching role in labor and capital in the process of production. This means that with the use of energy in the production process, the efficiency of labor and capital grow together with the competitiveness of the nation's (Stern, 2011). Moreover, the economic development of a country depends upon its investment level. FDI is a catalyst for productivity enhancements and improved output levels in the host economy, permitting the local industry to reinvest its profits into the industry. On the other hand, investment, in turn, depends on the availability of infrastructure like electrical energy. Numerous researchers have tried to study the relationship between net inflows of FDI and energy consumption. The research has been concluded that as FDI allows for cheaper and easier access to capital, demand for energy increase with FDI inflows increase. This can, in turn, be used for expanding production, thus increasing energy demand and consumption Access to modern energy is important for the delivery of clean water, hygiene, and healthcare, for delivery of consistent and effective lighting, heating, cooking, mechanical power, transport, internet, television, and telecommunications services (IEA, 2014). When electricity is accessible, food and drug can be kept in the refrigerator for an extended period, it will help more people to read and raises the adult learning rate so that the living situation will improve.
On the other hand, the deficiency of access to electricity can have substantial consequences on public health. This was confirmed by Hosking, Fletcher, and WHO (2011) that exposed the emissions of CO2 and hydrocarbons due to the burning of biomass frequently leads to sicknesses and death in several developing nations (Hosking, J., Mudu P., Dora C., & WHO, 2011). In terms of number WHO (2018) report indicates that each year, due to household air pollution approximately 4 million people die using biomass fuels and kerosene that are cauterized as inefficient cooking practices (WHO, 2018). However, such an amount of premature death could be minimized by increasing access to electricity.
Electricity access refers to the percentage of individuals that have moderately basic, stable access to electricity (IEA, 2017) in a given region. The IEA electricity access explanation involves more than just supplying electricity to the household, which specifies the lowest threshold in the urban household per year is 500 kWh and in the rural household per year is 250 kWh (IEA, 2017). In general, at the global level, the population with access to electricity was growing over in the past twenty-five years that increased from 73.45% in 1993 to 88.85% in 2017. Hannah Ritchie and Max Roser (2018) have also found that upper-middle-income countries have access to electricity around 89-100 percent (Hannah R. & Max R., 2018). In this context, the global population in terms of their income level that represents access to electricity at the household level from 1990 to 2017 is shown in figure 9.
Data source: World Development Indicator    For realizing access to electricity the world would need to increase investments. This means considerable development effects can be attained by investing in electricity supply and that would have downstream economic special effects in return. Therefore, electric power access and utilization do not only make life comfortable but also conserve time for production and accelerate economic growth. Hence, it is important to give special consideration for electric power access as it could assist to boost human development.
Moreover, as discussed above the world energy scheme is by its nature is very much complex.  These are the top-ten electricity consuming nation's that account for more than two-thirds of worldwide electricity consumption.
Data source: World Development Indicator consumption growth was also negative 1.77%. Therefore, next, we found and discussed theoretically above about electric power consumption, and global economic activities such as FDI and GDP are also investigated empirically.
Therefore, the purpose of this paper is to investigate the possible responsiveness of FDI to electric power consumption in different income level at the global perspective. For motivating the research, 131 countries data have been collected from WDI and US-EIA from 1992 to 2016.
The paper is structured as follows; Section-1 comprises an introduction that includes background and purpose of the paper, Section-2 deals with related works, Section-3 discuss methodology, Section-4 covers findings and discussions, Section-5 includes conclusion and recommendation.

2.2.FDI and Energy Consumption
Numerous researchers have tried to study the relationship between net inflows of FDI and energy consumption. The research has been concluded that as FDI allows for cheaper and easier access to capital, demand for energy increase with FDI inflows increase. This can, in turn, be

METHODOLOGY
Regression is a common methodology for evaluating the statistical relationship between Since the data is a panel data, we also control country fixed effect and year fixed effect, which is a common time shock for all countries. Where: FDI it , GDP it , Electricity it , and Labor it , respectively representtotal foreign direct investment net inflow, per capita gross domestic product, per capita electricity power consumption, and total labor force, "i" denotes country i and "t" denotes year "t", "αᵢ", is the fixed effect of country i that controls all characteristics of the country that is constant over the sample period, year tcontrol the yearly random shocks that are common to all countries in a year, ₜ the error.
Since our regression equation includes two endogenous variables, GDP and EPC, making it necessary to test the endogeneity if E [Ɛ| X₁, X₂, Xₙ] = 0 ν X j , then we say that we have explanatory exogenous variables. If, for some reason such as the omission of relevant variables, measurement errors, simultaneity, etc., X j is correlated with ", we say that X j is an endogenous explanatory variable (Burcu Eke, 2020). Hence, we will test endogeneity in the regression equation by predicting "X j " residual and lets Xj be "e". If p-value of the residual "e" small it indicates that there is an endogeneity problem that also shows the estimation of the model is not consistent. We also use Instrument Variable (IV) method to estimate the equation. Therefore, we will use the lag of GDP, EPC, and EPC2 as an instrument.
Lagged explanatory variables remain commonly used as instrumental variables (IVs) to address endogeneity concerns in empirical studies with observational data (Davidson & Mackinnon, 1993). Finally to test the exogeneity we used the STATA command "dmexogxt" that gives the Davidson-Mackinnon test of exogeneity. Hence, as will be shown below, the regression analysis results confirm that the electricity supply is very important for countries, especially in low and middle-income countries, to attract FDI and drive economic growth.

3.1.Data
To choose the best suitable and appropriate materials as resources for the research topic numerous data had been collected from different sources. Most importantly world development indicators of World Bank, US-EIA, and IRENA data were utilized for the analysis. Hence, annual data from 1992 to 2016 for the 131 countries including "gross domestic product per capita (US$)", "foreign direct investment inflow per capita (million US$)", and "total labor force (thousands)" had been collected from WDI, which available online at http://www.worldbank.org.
While electricity consumption in kWh data is found from US-EIA available online at https://www.eis.gov/beta/international/data/browse, then per capita EPC in kWh and renewable electricity consumption share was calculated.  Table 2 All country's data to test the possible responsiveness of FDI to electric power consumption (EPC)

4.2.Discussion
In table 4.3, column (1) reports the regression results for all countries that show the GDP size and total labor force of a country can significantly increase FDI inflow, while EPC has a negative but insignificant effect on FDI inflow. Since EPC is vital for production, it supposedly to increases FDI. We suspect that most high-income countries have a high level of EPC and therefore EPC becomes less important for them to attract FDI. We hence include the square term of EPC in the second column, and the result shows that there is an inverse-U shaped relationship between EPC and FDI. That means as EPC increases, FDI increases first and then becomes flat or falls with a threshold value of 31,272 = (3.415/ (2*0.0000546)) kWh. However, we also suspect an endogenous problem in variables (GDP and EPC), we test endogeneity in the regression equation by predicting the "e" residual. The small p-value of the residual "e" indicates that there is an endogeneity problem that also shows the estimation of the model is not consistent. So it can be concluded that there is an endogeneity bias in the OLS fixed estimates and it needs to instrument the variables (GDP, EPC, and EPC 2 ) to make them exogenous see appendix C. Hence, we use the lag of GDP, EPC, and EPC 2 as an instrument. Lagged explanatory variables remain commonly used as instrumental variables (IVs) to address endogeneity concerns in empirical studies with observational data (Davidson & MacKinnon, 1993). To test the exogeneity we used the STATA command "dmexogxt" that gives the Davidson-Mackinnon test of exogeneity higher p-value 0.5166 so that the model gives good results without endogeneity problem.
Therefore, according to column (3), the coefficient of lagged GDP 1.064 implies that one dollar increase in GDP per capita, the FDI inflow per capita will be increased by around 1.064 dollars one year later, ceteris paribus and statistically significant at 1%. Similarly, the coefficient of lagged EPC 3.135, which suggests that one unit increase in EPC, FDI inflow also increased by 3.135 dollars one year later, ceteris paribus and statistically significant at 5%. Finally, the coefficient of LF 0.974 implies that one unit increase in the total labor force, FDI net inflow per capita increases by about 0.974 dollars, ceteris paribus, and statistically significant at 1%.
After doing the whole country panel data analysis to test the possible different responsiveness of FDI to EPC in different countries, we separate the total sample into two subsamples: highincome countries, and middle & low-income countries as shown in table 4.4 above in column (1) the regression result of EPC for only high-income countries is not significant that implies EPC has no impacts on net FDI inflow in high-income countries. As electricity supply in high-income countries is abundant and is not a concern for consumption and production in their countries, EPC will not be a factor to be considered for foreign investors. Since most high-income countries have a high level of EPC and therefore EPC becomes less important for them to attract FDI.
Then we include the square term of EPC in column (2), and the result shows that there is an inverse-U shaped relationship between EPC and FDI. That is, as EPC increases, FDI increases first and then becomes flat or falls. The threshold value is 37,947 = (3.180/ (2*0.0000419)) kWh.
However, since we already identified an endogenous problem in variables (GDP and EPC), it needs to instrument the variables (GDP, EPC, and EPC 2 ) to make them exogenous and we use the lag of GDP, EPC, and EPC 2 as an instrument and tested the exogeneity the result shows a higher p-value 0.2888 so that the model gives good results without endogeneity problem.
Therefore, according to column (3), the coefficient of lagged GDP 0.877 implies that one dollar increase in GDP per capita, the FDI inflow per capita will be increased by around 0.877 dollars one year later, ceteris paribus and statistically significant at 1%. Similarly, the coefficient of LF 9.043 implies that one unit increase in the total labor force, FDI net inflow per capita increases by about 9.043 dollars, ceteris paribus, and statistically significant at 1%.
But for EPC, even after the lagged effect incorporated the result show statistically insignificant, because EPC is not a factor for high-income countries to attract investors. On the other hand, considering only middle and low-income countries the results are reversed. For these countries, especially for low-income countries, and abundant electricity supply is important for FDI inflow as shown in the table below. In table 4.4 column (4) & (5) above we report the regressions to result for the middle and low-income countries that show EPC can significantly increase net FDI inflow. As we already identified an endogenous problem in variables (GDP and EPC), it needs to instrument the variables (GDP and EPC) to make them exogenous and we use the lag of GDP and EPC as an instrument and tested the exogeneity; The result shows a higher p-value 0.9295 so that the model gives good results without endogeneity problem. Therefore, according to column (5), the coefficient of lagged GDP 0.890 implies that one dollar increase in GDP per capita, the FDI inflow per capita will be increased by around 0.890 dollars one year later, ceteris paribus and statistically significant at 1%. Similarly, the coefficient of lagged EPC per capita is 11.65 and statistically significant at 1%, which implies that one unit increase in EPC the net inflow FDI would increase by 11.65 units. Moreover, the coefficient of LF 0.788 implies that one unit increase in the total labor force, FDI net inflow per capita increases by about 0.788 dollars, ceteris paribus, and statistically significant at 1%.

Conclusion
The finding of this research shows that as an essential input for production and indispensable good for consumption, electric power consumption is the driving force for economic activities.
Especially, a stable and abundant supply of electric power is crucial to attracting foreign direct investment (FDI). The cross-country regression also shows that there exists an inverse-U shaped relationship between EPC and economic growth such as FDI net inflow in the whole 131 countries panel data. That is, as EPC in an economy increases, the net FDI increases first and then keeps rather stable or falls. Furthermore, by separating the sample into two, high income, and middle & low-income countries, it is shown that the impact of EPC on FDI mainly comes from the middle & low-income countries. The regression results show that EPC can significantly increase net FDI inflow for middle & low-income countries but has no impact on net FDI inflow in high-income countries. As electricity supply in high-income countries is abundant and is not a concern for consumption and production in their countries, EPC will not be a factor to be considered for foreign investors. However, for middle & low-income countries, the results are reversed. For these countries, especially for low-income countries, sufficient electricity supply is important for FDI inflow. Therefore, we conclude that to improve economic development it is essential to raise electricity supply.