This study aims to comprehensively evaluate the influence of FDI on economic growth, considering the diverse adaptive capabilities of host countries. Spanning a period of 33 years from 1989 to 2021, the research will delve into the non-linear relationship between FDI and economic growth, carefully examining how host countries adapt to the inflow of foreign investment. By integrating data from various sources and employing qualitative analysis, this study seeks to identify meaningful correlations between the two variables and gain deeper insights into their intricate interplay. The findings of this research hold significant potential for policymakers, providing them with valuable insights into harnessing the economic advantages of foreign investment while leveraging the unique strengths and abilities of each host country.
To accomplish our objective, we embarked on our study by employing the fundamental Cobb-Douglas Production Function Framework. This framework serves as a cornerstone for establishing a technological relationship between inputs, specifically Capital and Labour, and the anticipated output level. In our analysis, we consider Capital and Labour as the inputs, while focusing on GDP Growth, which represents the expansion of Gross Domestic Product, as the output variable. By utilizing this well-established production function framework, our study aims to shed light on the intricate interplay between inputs and output, thereby facilitating a deeper understanding of the factors driving economic growth. Using this, our production function framework is as follows:
In our analysis, we consider the variables Gross Domestic Product (GDP) Growth, denoted as Y, Gross Capital Formation (% of GDP), represented by K, and the labour force, referred to as L, which encompasses individuals who contribute their skills and efforts to the production of goods and services within a specified timeframe. With Eq. (1) as our foundation, we posit that a nation's GDP growth will experience an upward trajectory when accompanied by increased capital and labour inputs, assuming a constant level of technology. By acknowledging these relationships, our study seeks to elucidate the factors that contribute to the expansion of a country's GDP, providing insights into the underlying mechanisms driving economic growth.
To build upon prior research and further extend the scope of our study, we have incorporated additional variables into our equation. Notably, Adeniyi (2020) and Bibi et al. (2014), concluded that FDI is a robust predictor and significantly favours the host country's economic progress. Conversely, Inflation exerts an unfavourable impact and shares a significant negative association with economic progress. Additionally, research conducted by Bibi et al. (2014) depicted that Trade Openness shares an unfavourable connection with the country’s growth. A study by Mustafa (2019) found that high Inflation rates in a country affect the FDI inflow and slow down the economic growth and development of that country.
Additionally, the level of Human Capital, the amount of FDI, and the economic growth rate were related (Kar, 2013). As a result, a country's economic health would improve, and more job possibilities would be created with an increase in the inflow of FDI. This motivates the government and individuals to invest more in education and increase literacy. Furthermore, Kar (2013) found that Human Capital development is linked to better political and social stability by targeting more FDI inflow. Therefore, the increased inflow of FDI and Human Capital share a positive and favourable relationship. Similarly, empirical research conducted by Noorbakhsh et al. (2001) hypothesised that Human Capital is one of the most critical determinants of FDI inflows and affects economic growth statistically significantly, and its importance is growing over time. Additionally, Bekhet & Al-Smadi (2014) suggest that FDI, Inflation, and Broad Money are associated, sharing a casualty and long–run relationship.
After the contributions, we would extend our production function framework Eq. (1). We would also consider additional factors, such as net FDI inflows, Human Capital, Macroeconomic Stability, Financial Development, and Liquid Liabilities. This paper has abbreviated FDI as FDI, Inflation (a proxy for Macroeconomic Stability) as INF, Gross Capital Formation as GCF, Human Capital as HC, Broad Money (Liquid Liabilities) as LL, Labour as LAB, Trade Openness as TO, and Economic Growth as GDPEG.
This study has collected data from the World Bank Indicators annually related to GDP growth, FDI inflow, Trade Openness, Macroeconomic Stability (Inflation), Gross Capital Formation, Human Capital (measured by secondary school enrolment percentage), Liquid Liabilities, and Labour. The collected data is used to observe the correlation between FDI and economic growth and how the different adaptive capabilities of host countries can influence this relationship. To measure economic growth, GDP growth (annual %) has been taken as a proxy, calculated as the Annual percentage growth GDP rate at market prices based on the constant local currency, which has been used in earlier studies (GDP Growth (Annual %) | Data, n.d.). The Human Capital School enrolment is proxied using the secondary (% gross) as a reference. It is calculated by dividing the total enrolment by the population of the age group corresponding to the level of education being considered, regardless of the age of the individuals enrolled (School Enrollment, Secondary (% Gross) | Data, n.d.). Macroeconomic stability is proxied by Inflation, calculated by the consumer prices index (annual %) (Inflation, Consumer Prices (Annual %) | Data, n.d.). Broad money, or Liquid Liabilities, a proxy of Financial Development, is calculated by adding up different types of readily available funds, such as cash not held in banks, certain types of checking accounts, savings accounts, foreign currency deposits, and certain types of negotiable instruments (Broad Money (% of GDP) | Data, n.d.). As indicated by the Trade (% of GDP) metric, Trade Openness is determined by the total of a country's imports and exports of goods and services as a GDP Percentage (Trade (% of GDP) | Data, n.d.). This measurement is used to determine the level of trade liberalisation country. Labour is taken as the total Labour force, which is made up of individuals engaged in the creation or manufacture of goods and services during the period (Labor Force, Total | Data, n.d.). FDI net inflows denoted as a percentage of total inflows are calculated by adding together short-term capital, the reinvestment of earnings, equity capital, and other forms of long-term capital as reported in a country's BOP (FDI, Net Inflows (% of GDP) | Data, n.d.). At last, Gross Capital Formation (GCF), measured as a % of GDP, includes spending on expanding the economy's permanent assets, as well as any changes in the amount of inventory held (Gross Capital Formation (% of GDP) | Data, n.d.).
Apart from all these variables, we have used various other interaction terms to scrutinise the association of Economic Growth and FDI in the availability of such adaptive characteristics and to check if FDI influences economic growth in the presence of specific host nation characteristics. Such variables are the interaction between FDI and INF as FDIINF, FDI and LL as FDILL, FDI and TO as FDITO, and FDI and HC as FDIHC. After the inclusion of all these variables, our production function framework is as follows:
Y = f (K, L, GCF, LL, HC, INF, FDI, TO, FDILL, FDIHC, FDITO, FDINF) (2)
We started with a descriptive statistic to analyse the relationship as mentioned above. Descriptive statistics give information about the mean and standard deviation of the data. It depicts the general behaviour of the data. The next section is a presentation of the correlation analysis, which explains the relationship that exists between the independent and dependent variables. Cross-sectional dependence is a common phenomenon in panel data. This could arise due to the spillover effects or unobserved common factors (Baltagi & Hashem Pesaran, 2007). The cross-sectional dependency of our data sets must be confirmed after analysing the relationship between these variables. Here, the Pesaran Cross-sectional Dependence test checks for the same. It is crucial to determine whether cross-sectional dependence exists; if so, it must be eliminated. The following is the cross-sectional dependence's null hypothesis:
Null Hypothesis (H0): No cross-sectional dependence (correlation) in residual.
After that, it's crucial to do a stationary test to confirm the stationarity of the variables. The cross-sectional dependency of the data determines the use of the First-Generation or Second-Generation unit root tests. Following the unit root test, different regression models examine how independent factors affect the dependent variable. There are three types of regression models. They are the Ordinary Least Square (OLS) Method, Panel Data with Random Effects, and Panel Data with Fixed Effects.
To decide on the best regression method, different tests have been conducted. An F test has been used to choose between OLS and Panel Data with the Fixed Effects Method. The Breusch-Pagan Lagrange multiplier (LM) test has been used to distinguish between panel data with fixed effects and panel data with random effects. Subsequently, the Hausman test was been used to determine the final choice between fixed and random effects.
Finally, to check the general underlying assumptions of the regression, various tests have been applied. The Wooldridge autocorrelation test checks the serial correlation; the Wald test is for heteroscedasticity. In contrast, the Collinearity Test (VIF), which determines whether multicollinearity is present, is undertaken before making signs.