Adequacy of governance in the link between foreign direct investment and structural transformation

Abstract

The inflow of foreign direct investment has increased in developing nations. The study used panel data from sub-Saharan Africa and East Asia and the Pacific countries to observe the adequacy of governance in linking FDI and structural transformation. Pooled Mean Group estimation technique is utilized for analysis purposes using 31-year-old data. The econometric results suggest that FDI inflows boost structural transformation, while debt and inflation dis-improve structural transformation. Across countries, in the short run, FDI with governance had a negative effect for South Korea. Accordingly, the governments of these regions that attempt to succeed in structural transformation should focus on active utilization of FDI and minimize massive inflow of foreign public debts.

Keywords:

• Foreign direct investment
• Governance
• pooled mean group and structural transformation

PUBLIC INTEREST STATEMENT

Today, the least developed countries are characterized by a low level of savings and, thereby, a low level of domestic investment. Low levels of domestic investment forced these countries to attract FDI by providing different incentives. On the other hand, these countries are struggling to transform their economies. Hence, FDI inflows, particularly those augmented with technology, can catalyze the structural transformation of nations. Furthermore, the effect of FDI in improving the structural transformation of nations depends on the accommodation capacity of the receiving countries, like the quality of governance or institutions, the level of human capital, and so on. Hence, this study is intended to examine the adequacy of governance in the relationship between FDI and structural transformation, considering selected countries from the SSA and EAP regions.

1. Introduction

At the end of the 20^th century, owing to globalization, most developing countries have significantly reduced restrictions on the flow of FDI, and many have offered attractive tax incentives and subsidies in demand for foreign capital. The motive behind this policy implementation is the theory that pinpoints FDI by stimulating capital accumulation and through positive externalities in the form of knowledge and productivity spillovers to domestic infant firms, contributing to overall economic growth. However, the question remains whether this holds true. Scholars who doubt the theory indicate, among other facts, that FDI can reduce capital accumulation when investors from abroad claim scarce resources, such as skilled manpower, import licenses, credit facilities, and so on, and hence crowding out investment from domestic sources. Moreover, it is argued that knowledge spillovers are often hypothetical and more illusive because local firms use traditional production technology and unskilled laborers are typically unable to learn from foreign transnational corporations (TNCs). Finally, since TNCs generally have lower marginal costs due to firm-specific benefits, criticizers show that they can attract demand from local firms, forcing domestic firms to reduce their production. Thus, competition from TNCs can, paradoxically, lessen the productivity of local firms and the shutdown of incompetent local firms (Aitken & Harrison, 1999; Haddad & Harrison, 1993).

The SSA region is among the top FDI destinations in Africa and the leading recipient, with an average yearly inflow of 12.68 billion US dollars for the period from 1970 to 2019. In addition, the region accumulated a massive amount of FDI stock, which grew from around 295.19 billion US dollars in 2010 to 633.8 billion US dollars in 2019, showing a dramatic increment. While the FDI inflows for East Asia and the Pacific reached 8.81 trillion US dollars with an average annual inflow of 176.15 billion US dollars for the year 2019. However, currently the inflow of FDI for East Asian countries is declining from 609.27 billion US dollars in 2018 to 493.72 billion dollars in 2019 (IMF, 2020).

Recent empirical data show that both regions accommodate massive inflows of FDI stock, which accounts for approximately 9.44 trillion dollars (WB, 2020). The statistics also indicated that the GDP share of major economic sector composition for SSA countries is around 15.48 percent of the primary sector, 11.34 percent of the manufacturing sector, and 50.66 percent for the service sector, and for East Asian countries, agriculture, manufacturing, and service sectors accounted for 4.46 percent, 22.79 percent, and 60.57 percent, respectively. This shows that the two regions are at different angles in relation to the structural transformation theme but have the same economic history in the past decades of the 1950s and early 1960s.

The majority of empirical works were focused on linking structural transformation with economic growth and development, and little attention has been given to the role of FDI in the structural transformation of economies. Such investigations done by Bernard et al. (2023), Oduola et al. (2022), Muller (2021), Mamba et al. (2020), Mensah et al. (2016), and Samouel and Aram (2016) were conducted in an African context. While others, for instance, Thirion (2020), Muhlen and Escobar (2020), Cruz and Montes (2020), and Maroof et al. (2019), examined the Asian economy, and studies by Chenaf (2019) and Topcu (2016) considered the economies of developing countries but only targeted countries in the higher middle-income category. These few studies have also adopted labor mobility in highly productive sectors as a proxy for structural transformation. However, the labor mobility approach is more about structural change, which does not alone capture structural transformation.

At the cross-country level, various empirical findings indicate a robust and significant relationship between FDI and economic growth. However, some of them indicate that the positive impact depends on the economic characteristics of nations, and others indicate negative and mixed results. Empirical investigations by Alfaro et al. (2004) using cross-country data for 71 developing and developed countries concluded that FDI plays a significant role in supplementing economic growth, yet the development level of the financial markets of the countries is important to sustain these positive effects. Accordingly, scholars argue that local firms specifically need to reorganize their structure, such as purchasing new machinery, hiring skilled managers, and laborers, to enjoy the positive return from FDI that induces knowledge spillover effects, which is difficult to achieve with poor performance systems in financial markets. Blomström et al. (1994), using cross-country data for 78 developing countries, indicated that countries with lower income levels do not enjoy significant growth benefits from FDI, while countries with higher income levels do. Based on their findings, they conclude that a certain threshold level of development is essential to absorb new technology from investment in TNCs. Balasubramanyam et al. (1996) also supplement the above findings by examining a sample of 46 developing countries and found that FDI has a strong and significant effect on the growth of these countries, which are more open to international trade. They argue that more open economies are more likely to attract a higher volume of FDI and promote more efficient resource utilization than closed economies. Similarly, (Borensztein et al., 1998) examining a cross-country study of 69 developing countries and concluded that the positive effect of FDI on economic growth depends on the level of human capital (in both dimensions, education and health) in the host country. Specifically, they generalize the idea that FDI positively contributes to economic growth if and only if the education level is higher than a certain threshold level.

The idea of attracting investment from foreigners has received considerable attention from new growth theorists, given that overall productivity is demonstrated to be endogenous and pertinent to external inputs such as FDI (Berthélemy & Démurger, 2000; Bilgili et al., 2016; Borensztein et al., 1998; Romer, 1990; Su & Liu, 2016). With respect to economic growth in the host nation, FDI maintains extra benefits; for instance, it is less volatile, subject to minimum political collaterals, and difficult to reverse, as opposed to portfolio equity and debt flows (Lensink & Morrissey, 2006). Nonetheless, not everyone shares such an optimistic view, and argues that FDI crowds out domestic investment (DI), hurting the long-term growth potential of the host nation. In part, this pessimistic view stems from the assumption that due to FDI inflows, there exists intensive competition both in local factors and product markets that might reduce indigenous companies’ willingness to re-invest or drive incompetent companies out of the market altogether (Aitken and Harrison, 1999). This long-term effect of FDI on DI and thereby economic growth was witnessed during the 1997 Asian currency crisis and the global financial crisis in 2008, which showed an unexpected decline in FDI, resulting in a major obstacle impeding the recovery process in many emerging economies.

The significance of FDI inflow directly on economic growth as of the theories, and indirectly by affecting the accumulation of reserves and domestic saving on economic growth, has therefore triggered discussions among academic scholars, researchers, and policy planners, as well as analysts on which studies have been conducted immensely in developed countries and in some developing countries, focusing more on the economy of a single country (Gudaro et al., 2012; Semwanga, 2011; Prince & Vijay, 2019). Currently, there are very few studies that link FDI and economic growth in developing countries (Amensisa, 2018; Betelhem, 2016; Dejene, 2015) but these studies did not consider the interrelationship between FDI and growth across economic sectors,¹ and ignored the comparison of the impact of FDI on economies having the same initial point but differing in terms of their growth rate. These studies also lack the relationship between FDI inflows and structural transformation of a nation's economy, considering a panel of countries. Furthermore, irrespective of the intention of TNC, the World Bank and the IMF argued that the problem of Africa, particularly sub-Saharan Africa, is far beyond foreign sources of capital in the form of FDI inflows, and reiterated that other than linking low levels of African economies to huge FDI inflows—China’s FDI—indicators such as poor governance systems and weak institutions such as voice and accountability issues, political instability and presence of violence/terrorism acts, lack of government effectiveness, poorer regulatory quality, absence of rule of law, and lack of control of corruption should be addressed. According to Brautigam and Knack (2014), one of the reasons for poor governance in most African nations is due to the fact that colonialism did little or nothing to develop the quality of institutions and strong governance systems capable of addressing the demands of current states.

Similarly, the literature on the effect of FDI inflows on structural transformation is limited. They totally ignored the interactive effect of FDI with good governance in predicting variations in the structural transformation of nations. On the other hand, since countries may have differences in terms of governance or institutional quality, this study intends to cover these gaps by considering SSA and EAP countries. However, due to their respective shortcomings, this study proposes to analyze FDI and structural transformation of nations adopting a panel data analysis approach taking four countries from sub-Saharan Africa, namely Ethiopia, Kenya, South Africa, and Nigeria and three countries from East Asia, and the Pacific, namely Singapore, China, and South Korea for the period from 1989 to 2019. There exists heterogeneity among these selected countries based on their level of development, particularly focusing on the level of income based on World Bank (2022) classification of countries. Accordingly, low-income economies are defined as those with a GNI per capita of $1,135 or less; Lower middle-income economies are those with a GNI per capita between $1,136 and $4,465; Upper middle-income economies are those with a GNI per capita between $4,466 and $13,845; High-income economies are those with a GNI per capita of $13,846 or more. Thus, Kenya and Nigeria are under the category of lower middle-income, South Africa and China are in the upper middle-income category, Ethiopia and South Korea are in the low-income category, and Singapore is under the high-income category.

The panels were selected due to the following reasons:

• First, the development literature on sub-Saharan Africa suggests that these three countries might serve as more important sources of inspiration than other developed nations.

• Second, Singapore, China, and South Korea are more generalizable than other developed nations in the region as reference points for today’s developing countries (particularly SSA), and their development process starts from an agrarian base that is typically taken to be the starting point for industrialization. On the other hand, they transformed their economies faster than SSA countries despite having similar initial conditions, especially in the 1960s.

• Third, from an imitation and copying of technology point of view for today’s SSA countries, looking particularly in South Korea, Singapore, and China is useful (Waal, 2013). Hence, when we examine the historical trends of growth, no nations have grown as fast as the three EAP countries from low to high income levels, particularly for the period between 1960 and the 1990s. Thus, these nations might act as references for SSA countries in their structural transformation. Furthermore, this study aims to identify whether FDI is an accelerator or impediment across the selected countries in the context of structural transformation.

Furthermore, this study aims to identify whether FDI is an accelerator or impediment across these selected countries in the story of structural transformation. In particular, the study captures the following objectives:

• We examine the effect of FDI on structural transformation among sub-Saharan African and East Asian countries using panel data.

• We examine the adequacy of governance on the link between FDI and structural transformation of the economy among sub-Saharan African and East Asian countries using panel data.

2. Literature review

The empirical work of many researchers hangs on the theoretical foundation of the traditional neo-classical growth theories and the new growth theories, which are endogenous growth theories. The first one is based on Solow’s growth model (Solow & Solow Robert, 1956), which indicates that foreign direct investment boosts economic growth by increasing the volume of investment (Petrakos et al., 2007). Therefore, this paper theoretically adopts the endogenous growth models, which provided the theoretical foundation for most studies in the literature on the nexus between FDI and economic growth. Nevertheless, this study does not claim that the endogenous growth model by itself is free from some drawbacks, like the symmetrical assumptions involved in different investigations.

The previous empirical work done by Mamba et al. (2020) in the West African Economic and Monetary Union, Topcu’s (2016) study on 19 upper middle-income developing countries, Mensah et al. (2016) in 21 countries, and Gui-Diby and Renard (2015) in 49 African countries found that FDI has no statistically significant impact on structural transformation. Contrary to Thirion (2020) from Mexico, Jie and Shamshedin (2019)’s Vector Error Correction Model (VECM) estimation in Ethiopia, and Samouel and Aram (2016) from Africa, FDI has a statistically significant effect on structural transformation.

In the same way, Oduola et al. (2022) using GMM’s estimation technique for 43 SSA economies for a period 1996–2018, Muller (2021) PCSE’s output in 47 SSA economies for a period 1996–2017, and Maroof et al. (2019) using ARDL methods of estimation for South Asian economies for a period 1996–2015 found that FDI has a negative and statistically significant effect on structural transformation, focusing on the industrialization dimension of the concern. They further justified that FDI engaged in oil exploration and mining, aggravated by weaker institutional quality, was identified as the major hindrance to achieving structural transformation in these countries.

Even though there is a mixture of results for studies at the cross-country level, many authors claim methodological problems when estimating cross-country growth equations, thus pinpointing serious doubts regarding the validity of the findings of these studies (Carkovic & Levine, 2005). The implicit assumption of the existence of a common economic structure and identical techniques of production across countries is one of the criticisms behind studies at the cross-country level. In the real world, however, techniques of production, institutional setups, and policies differ considerably among countries, so that country-specific omitted variables are more likely to lead to biased cross-country regression output, which is misleading the findings. Furthermore, a statistically significant relationship between FDI and economic growth does not necessarily mean the result of a causal impact of FDI on economic growth, but rather because of better profit opportunities created by the rapid level of economic growth. Thus, cross-country studies may suffer from serious endogeneity problems (Nair-Reichert et al., 2001).

Various scholars have adopted panel estimation techniques to mitigate the drawbacks of cross-country studies. The panel estimation technique is advantageous for considering unobserved country-specific characteristics, thereby avoiding problems related to omitted variable biases. In addition, by incorporating lagged predictors, the panel estimation technique allows us to control for potential endogeneity problems and, furthermore, enables us to explicitly test for Granger causality. Scholars such as Carkovic and Levine (2005) use the GMM panel estimator, which was previously used by Arellano and Bover (1995) and Blundel and Bond (1998), to control for the possibility of biased outcomes due to problems related to endogeneity and omitted variables. They examined a sample of 65 panels from both developing and developed countries and found that FDI has no robust effect on economic growth, even by making the field feasible by accommodating differences across countries in terms of trade openness, per capita income, human capital level measured by education, and the development level of domestic financial markets. However, the authors underline that FDI is not irrelevant for growth given that the FDI variable turns out to be positive and statistically significant in many specifications. In contrast to the above findings, Busse and Groizard (2008) adopt an Arellano and Bond (1991) style GMM difference estimator using data for 84 developing and developed countries and find that the impact of FDI on economic growth depends on the development level of financial markets in the domestic economy. Accordingly, their results suggest that the growth effect of FDI is negatively related to the level of financial regulation in the host country. They argued that restrictive regulations impede both the allocation of foreign capital to the most productive sectors and the creation of spillovers to local firms.

Carkovic and Levine (2005) and Busse and Groizard (2008) explicitly showed the common pitfalls of the panel estimation technique by incorporating the assumption of homogeneity imposed on slope parameters. Nevertheless, advances in the heterogeneous panel literature suggest that estimation and inference in standard dynamic panel models can be misleading when the slope coefficients differ across cross-sectional units. Nair-Reichert et al. (2001) used what they referred to as the mixed fixed and random coefficient (MFR) approach to test for causality between FDI and growth by curbing the problems due to the homogeneity of the slope term. The MFR approach allows for complete heterogeneity in the coefficients of the explanatory variables, thus avoiding biases induced by possibly incorrect homogeneity restrictions. Accordingly, they examined a sample of 24 developing countries using time-series data covering 25 years and found that, on average, FDI has a positive causal effect on economic growth, but this growth effect is highly heterogeneous.

Moreover, the other pitfalls related to FDI and economic growth at the cross-country and panel levels are the consideration treatment of the variables under study (Acemoglu & Ventura, 2002; Tiwari & Mutascu, 2011). Accordingly, most studies in both approaches used GDP growth rates as the outcome variable and either the nominal value of FDI or the growth rates of FDI (FDI-to-GDP ratio) as the predictor. However, due to their respective shortcomings, this study proposes to use the structural transformation of the economy of nations and the per capita FDI inflows as target variables.

The originality of this paper lies in the following points: Firstly, this paper is a stepping stone for linking the significance of governance to the relationship between FDI and the structural transformation of national economy. This stands on the belief that the problem of the least developed nations goes far beyond foreign sources of investment in the form of FDI, and hence the level of governance should be addressed. Secondly, theoretically, the study utilizes the augmented production function approach and has been extended to accommodate the endogenous growth model, introduced by Lucas (1988), and Rebelo (1991), in linking the variables. Thirdly, to mitigate the drawbacks of using cross-country investigations, it adopted a panel data approach, taking seven countries from the two regions. Last but not least, it adopted Pooled Mean Group model for estimation purpose. The model assumes the long-run coefficients to be identical for all panels. However, there is a difference in the short-run coefficients and error variance across the panels as an emerging approach to panel data analysis.

3. Methodology of the study

This study adopts a quantitative research design using a panel data analysis approach. This study used secondary data from the WB and IMF datasets. Descriptive and econometric analysis methods were used. Under econometric analysis, the study adopted the Pooled Mean Group (PMG) estimation technique for the full panel, across the two groups, and among countries.

Using a standard augmented Cobb-Douglas production function framework, the level of FDI inflows is augmented with capital and labor, and modifications are developed to accommodate structural transformation issues via growth and development. According to economic theories, as growth takes place, we expect a change in the structure of the economy, which is a shift from a primary sector dependency to manufacturing and service sectors; thus, our outcome variable of interest is the structural transformation of the economy of nations as proxied by the relative share of the manufacturing sub-sector from GDP.

$\mathrm{S}\mathrm{T}{\mathrm{I}}_{\mathrm{t}}=\mathrm{f}\left(\mathrm{L},\mathrm{K},\mathrm{F}\mathrm{D}\mathrm{I}\right)\dots \dots \dots \dots \dots \left(1\right)$

where STI is the index for structural transformation of nations’ economy at period t, L is labor, K is capital, and FDI is foreign direct investment inflow. Following Grossman and Elhanan (1991); Barro and Sala-I-Martin (1995) and Balasubramanyam et al. (1996), this production’s function has been extended as of the new growth theory, the endogenous growth model, introduced by Lucas (1988), and Rebelo (1991). Endogenous growth theories show that the long-run growth of a country is not only inclined by the volume of physical investment but also depends on the efficiency of utilizing investment. In addition, McMillan and Rodrik’s (2011) measure structural transformation of the economy using a proxy variable, which is the reallocation of labor between sectors for a representative country at time t. This hangs on the economic theory that propagates as the economy develops resources, especially labor, which moves from the traditional sector to the industrial sector. However, this might not happen only because of development; rather, it is better to consider the parallel movement of aggregate output change with respect to the share of each sector of the economy in the national output (GDP). Hence, it is better to examine the value added by the major sectors of the economy.

According to the theoretical perceptive structural transformation of the economy is measured from either or both the production side and the consumption side, in this study, we used the structural transformation index as proxied by the relative share manufacturing sub-sector from the GDP, which indicates how transformation is related to productivity. We adopted a macro-panel data analysis because our target groups are seven countries from SSA and EAP and have a time dimension of more than 20 years. The current heterogeneous nature of the economic level and status of the selected countries is important for minimizing the cross-country dependency problem of the macro-panel model approach. Based on the extensions of endogenous growth models (Equation 1) above log transformation and extended by including other important variables as represented by C as a vector for other control variables in equation (2):

$\mathrm{S}\mathrm{T}{\mathrm{I}}_{\mathrm{t}}=\mathrm{f}\left(\mathrm{L},\mathrm{K},\mathrm{F}\mathrm{D}\mathrm{I},\mathrm{C}\right)\dots \dots \dots \dots \dots \left(2\right)$

where C is a control variable for the other exogenous factors. The purpose of this study is to examine the impact of FDI on the structural transformation of nations economy considering SSA and EAP regions for the period covering 1989–2019. Moreover, since countries vary in terms of quality of governance, we examine the interaction of FDI with the governance index to examine the role of institutional quality on the link between FDI and structural transformation. According to the World Bank (2021), the Worldwide Governance Indicators (WGI) measure the quality of governance, which has six components: political stability and absence of violence/terrorism, voice and accountability, government effectiveness, rule of law, regulatory quality, and control of corruption. Each indicator had a magnitude running from approximately −2.5 − 2.5, a movement from weak to strong governance performance. This study used a composite index based on these components that measures the quality of governance. Hence, the impact of FDI on the structural transformation of countries is analyzed using the following macro-panel econometric equation:

$ST{I}_{it}={\alpha }_{0i}+{\alpha }_{1i}FD{I}_{it}+{\alpha }_{2i}FD{I}^{\ast }Go{v}_{it}+{\alpha }_{3i}FP{D}_{it}+{\alpha }_{4i}SD{I}_{it}+{\alpha }_{5i}F{D}_{it}+{\alpha }_{6i}{I}_{it}+{\alpha }_{7i}{H}_{it}+{\epsilon }_{it}\dots ..(3)$where α₀ stands for the unknown intercept term for each country in the panel; ɑ₁, ɑ₂, ɑ₃, ɑ4, ɑ₅, and ɑ₆ are partial slope parameters to be estimated; FDI is foreign direct investment in current USD, included to examine the autonomous effect of FDI on the structural transformation of countries; FDI*Gov stands for the interaction of FDI with the governance index; SDI stands for the share of domestic investment from GDP in percent; FPD stands for the total amount of foreign public debt in current USD; FD stands for the level of financial sector development as measured by money supply over GDP; I stands for the inflation rate; and H stands for human capital variable; i = 1,2, … ,7 stands for the number of countries in the study; t = 1,2, …,31 stands for the time periods, and ε_it stands for the residual term (see Table 1).

Table 1. Symbol, definition and measurement, and data sources of variables included in the study

Symbol	Definition and Measurement	Sources
StrTransn	Stands for structural transformation of the economy of nations. McMillan and Rodrik (2011) indicated that value added productivity in the modern sector enhances structural transformation, and accordingly this study used the value added in the manufacturing sub-sector divided by GDP as a proxy for structural transformation. Value added is the net output of a sector after adding up all outputs & subtracting intermediate inputs. It is measured in current U.S. dollars.	WB
FDIndicator	Stands for financial development indicator and measured by the ratio of money supply, M2, to GDP.	WB
PcFPD	Stands for foreign public debt. It is the sum of public, publicly guaranteed, and private non-guaranteed long-term debt, use of IMF credit, and short-term debt, & measured as the ratio of FPD to population.	WB
SDI	Stands for share of domestic investment from GDP.	WB
Inflation	Stands for Inflation rate, average consumer prices.	IMF
PcFDI	Stands for FDI inflows. It is the sum of equity capital, reinvestment of earnings, other long-term capital, & short-term capital as shown in the balance of payments. It is measured by the ratio of FDI to population.
FDIGov	Stands for interaction of FDI with governance indicators. Governance indicator is a composite index that includes the following components: Voice & Accountability, Political Stability & Absence of Violence, Government Effectiveness, Regulatory Quality, Rule of Law, and Control of Corruption. Estimate of governance indicators ranges from −2.5 to 2.5 (from weak to strong) governance performance.	WB
H	Stands for the level of human capital as measured by literacy rate. Literacy rate is the percentage of people ages 15 & above who can both read & write with understanding a short simple statement about their everyday life.	WB

4. Results and discussions

4.1. Descriptive data analysis

According to Table 2, the mean values of the structural transformation index, financial development indicator, human capital as measured by literacy rate, foreign public debt, SDI, FDI, inflation, and governance indicator index are displayed for the total sample as well as for the income categories. Except for inflation in all remaining variables, higher-income countries are, on average, performing better than lower- and middle-income countries. Accordingly, the data show that the average structural transformation index for higher-income countries is 26.3 percent which is the highest as compared to the average value of low income (4.81 percent) and middle income (13 percent) countries. The data also show that the FDI inflow was on average about US dollar 51.9 Billion for high-income countries, US dollar 2.2 Billion for middle-income countries and US dollar 848 million for low-income countries for the study periods.

Table 2. Summary statistics of variable in the panel data

Variable	Total sample	Low Income Sample	Middle Income	High Income
	Mean	Mean	Mean	Mean
StrTransn	17.52	4.81	13.0	26.3
FDIndicator	70.2	27.4	39.3	115.5
FPD	109.2 Billion	10.7 Billion	37.1 Billion	214.1 Billion
SDI	23.29474	0.6822138	28.88573	25.24127
FDI	23.3Billion	848 Million	2.2 Billion	51.9 Billion
Literacy rate	77.50015	37.98421	74.18536	93.98691
Inflation	8.38	10.6	12.7	3.3
Governance	−.0959908	−0.99	−0.53	0.64

As shown in Table 3, the governance index was computed as the composite index of the six indicators employed in the study for conciseness, and countries with positive values seem to have a viable government or better institutions. On the other hand, as the magnitude of the governance indicators runs from −2.5 to 2.5 pinpointing a movement from weak to strong governance performance. In particular, the results show that countries like Singapore and South Korea, followed by South Africa, have relatively better institutions than other countries in the panel, while countries like Nigeria and Ethiopia have the weakest governance. This shows that across regions, the four countries from sub-Saharan Africa have a weak level of governance (institutional quality) compared to the three countries from East Asia and the Pacific.

Table 3. Mean value of governance indicators and human capital across countries (1989–2019)

Country	Ethiopia	Kenya	South Africa	Nigeria	Singapore	China	South Korea
Governance index	−0.99	−0.7	0.29	−1.17	1.69	−0.49	0.70
Literacy rate	37.98	78.69	86.99	56.86	93.54	90.13	98.29

Source: Authors’ computation using Stata 14, 2022

4.2. Estimation and Econometrics Results

4.2.1. Stationarity test

According to the Im-Pesaran-Shin unit-root test result, the p-values for the variables GDP, inflation, FDI, and FDI with governance are less than 5 percent in level form; hence, we fail to reject the null hypothesis of stationarity. The variable structural transformation index, FPD, financial development indicator, and FDI with literacy are stationary in the first difference. Thus, there might be a long-run linear relationship between the variables (Table 4).

Table 4. Im-Pesaran-Shin unit-root test

Economic variables	At Level		At First Difference
	Test Statistics	P-value	Test Statistics	P-value
Structural transformation	2.743	0.997	−8.125	0.000
Financial development indicator	−0.478	0.316	−5.992	0.000
Inflation	−3.716	0.000
SDI	−2.2326	0.0128
PcFDI	−1.2653	0.1029	−10.5418	0.0000
H	−2.5243	0.0058
PcFPD	4.8511	1.0000	−6.0632	0.0000
FDI*Gov	−0.8343	0.2021	−10.1616	0.0000

Source: Own computation based on World Bank datasets using Stata 14, 2021.

4.2.2. Co-integration test

In Table 5, Pedroni’s co-integration test found the presence of co-integration among our variables, which means the existence of a long-run relationship. On the other hand, since the absolute value of the rho statistics is greater than the absolute value of the test statistics v, −0.7521, for both panel and groups, confirming the null hypothesis of no cointegration is rejected at the 1 percent level of significance level.

Table 5. Pedroni’s cointegration tests

No. of Panel units: 7 Regressors: 7 No. of obs.: 217 Avg obs. per unit: 31
Test Statistics	Panel	Group
v	−.7521	.
rho	2.233	2.596
t	.409	−.2367
adf	.3622	.04185
NB: All test statistics are distributed N(0,1), under a null of no cointegration, and diverge to negative infinity (save for panel v)

Source: Own computation based on World Bank datasets using Stata 14, 2022.

4.2.3. Identifying technique of estimation

After detecting the existence of cointegration, the next procedure should be selecting the ARDL with error correction method to adopt either the pooled mean group (PMG) or the mean group (MG) technique. The Hausman test results recommend the use of the PMG estimation technique (Table 6). However, since the Prob>chi2 value is greater than 5 percent, the null hypothesis of homogeneity cannot be rejected.

Table 6. Hausman test for selecting technique of estimation

---- Coefficients ----
	(b)	(B)	(b-B)	sqrt(diag(V_b-V_B))
	MG	PMG	Difference	S.E.
PcFDI	−2.739479	.9531879	−3.692667	2.487685

b = consistent under Ho and Ha; obtained from xtpmg

B = inconsistent under Ha, efficient under Ho; obtained from xtpmg

Test: Ho: difference in coefficients not systematic

chi2(1) = (b-B)’[(V_b-V_B)^(−1)](b-B) = 2.20

Prob>chi2 = 0.1377

Pooled Mean Group Model estimation results across countries in the panel

Table 7 shows the results of the PMG model, taking into account the variations among the panels, indicate that in the long run, three variables, namely, inflation, foreign public debt, and FDI, significantly predict structural transformation. The interaction term, FDI and governance, has a negative sign and yet it is statistically insignificant. The sign of the interaction term indicates that countries without good governance results in misutilization of FDI other than the intended goal like enhancing structural transformation of nations’ economy. This judgement is drawn due to the fact that all selected countries in this study are on average performing lower in terms of their governance index as shown in Table 3, except Singapore, relative to the maximum index amount of governance, 2.5.

Table 7. Pooled Mean group estimation result across countries in the panel

Panel Variable (i): ID	Number of obs = 216			Number of groups =7		Avg =30.9
Time Variable (t): Year	Obs per group: min=30			Max =31		Log Likelihood=−179.3517
dStrTransn	Coef.	Std. Err.		z	P>z	[95% Conf.	Interval]
Long-run Dynamics
FDIndicator	.0049732	.0039685		1.25	0.210	−.0028049	.0127513
Inflation	−.0365415	.0109408		−3.34	0.001	−.057985	−.0150979
SDI	.0016318	.0063803		0.26	0.798	−.0108733	.0141369
PcFDI	.0137432	.0033881		4.06	0.000	.0071027	.0203836
PcFPD	−.0058168	.0004261		−13.65	0.000	−.0066519	−.0049817
FDIGov	−.0042943	.0043084		−1.00	0.319	−.0127385	.00415
H	.0000311	.0001642		0.19	0.850	−.0002908	.000353
Ethiopia
ec	−.3849955	.146755		−2.62	0.009	−.6726299	−.0973611
dFDIndicator	.060079	.03248		1.85	0.064	−.0035807	.1237387
dInflation	.0065623	.0090855		0.72	0.470	−.0112449	.0243695
dSDI	.133265	.4106843		0.32	0.746	−.6716615	.9381915
dPcFDI	−.0178125	.0869929		−0.20	0.838	−.1883154	.1526905
dPcFPD	−.0025334	.0051103		−0.50	0.620	−.0125494	.0074825
dFDIGov	−.0538945	.1043747		−0.52	0.606	−.258465	.1506761
dH	.0686078	.1168295		0.59	0.557	−.1603738	.2975895
_cons	2.214164	.8617044		2.57	0.010	.5252541	3.903073
Kenya
ec	−.2071475	.1107658	−1.87		0.061	−.4242445	.0099495
dFDIndicator	−.2274638	.0493189	−4.61		0.000	−.3241271	−.1308005
dInflation	.0018816	.0130951	0.14		0.886	−.0237843	.0275474
dSDI	−.0010071	.0010763	−0.94		0.349	−.0031166	.0011024
dPcFDI	−.1094742	.1032437	−1.06		0.289	−.3118282	.0928798
dPcFPD	−.0127373	.0050037	−2.55		0.011	−.0225444	−.0029302
dFDIGov	−.1824141	.1462805	−1.25		0.212	−.4691187	.1042905
dH	.0515222	.0999234	0.52		0.606	−.1443241	.2473685
_cons	2.554078	1.327741	1.92		0.054	−.0482466	5.156402
South Africa
ec	.0317382	.031995	0.99		0.321	−.0309709	.0944474
dFDIndicator	−.0440925	.0381271	−1.16		0.247	−.1188203	.0306353
dInflation	.046116	.0426097	1.08		0.279	−.0373974	.1296294
dSDI	5.829556	6.620948	0.88		0.379	−7.147264	18.80638
dPcFDI	−.0038958	.0041359	−0.94		0.346	−.012002	.0042105
dPcFPD	−.0001524	.0005118	−0.30		0.766	−.0011556	.0008507
dFDIGov	.0180659	.0138761	1.30		0.193	−.0091308	.0452625
dH	−.0052554	.0043346	−1.21		0.225	−.0137511	.0032404
_cons	−.9701993	.7432545	−1.31		0.192	−2.426951	.4865528
Nigeria
ec	−.0315815	.0359685	−0.88		0.380	−.1020784	.0389155
dFDIndicator	−.0160781	.0911386	−0.18		0.860	−.1947065	.1625502
dInflation	.0028786	.0148518	0.19		0.846	−.0262303	.0319875
dSDI	−.0087219	.0150134	−0.58		0.561	−.0381476	.0207039
dPcFDI	.169026	.1377927	1.23		0.220	−.1010427	.4390946
dPcFPD	.010293	.0057322	1.80		0.073	−.0009419	.0215279
dFDIGov	.1326863	.1152908	1.15		0.250	−.0932795	.3586521
dH	.0924553	.1086133	0.85		0.395	−.1204229	.3053336
_cons	.1771333	.5645667	0.31		0.754	−.9293972	1.283664
Singapore
ec	−.0043188	.0174855	−0.25		0.805	−.0385899	.0299522
dFDIndicator	−.0839568	.0305249	−2.75		0.006	−.1437845	−.024129
dInflation	−.1665111	.1249623	−1.33		0.183	−.4114327	.0784106
dSDI	−.0268196	.1633161	−0.16		0.870	−.3469133	.2932741
dPcFDI	.0004672	.0006135	0.76		0.446	−.0007353	.0016696
dPcFPD	−.0000248	.0001844	−0.13		0.893	−.0003862	.0003365
dFDIGov	−.0002938	.0003628	−0.81		0.418	−.0010049	.0004172
dH	−.0013671	.0031986	−0.43		0.669	−.0076364	.0049021
_cons	−.0406525	.4267079	−0.10		0.924	−.8769845	.7956796
China
ec	−.7526586	.0759384	−9.91		0.000	−.9014951	−.603822
dFDIndicator	−.0064114	.0096697	−0.66		0.507	−.0253637	.0125408
dInflation	−.0086458	.0144653	−0.60		0.550	−.0369972	.0197056
dSDI	.0228498	.0185701	1.23		0.219	−.0135469	.0592465
dPcFDI	.0007897	.0062478	0.13		0.899	−.0114558	.0130351
dPcFPD	.0031357	.0006999	4.48		0.000	.0017639	.0045075
dFDIGov	.0012981	.0096117	0.14		0.893	−.0175405	.0201367
dH	−.0040804	.0092241	−0.44		0.658	−.0221592	.0139984
_cons	23.47263	2.289301	10.25		0.000	18.98568	27.95958
South Korea
ec	−.3182814	.106719	−2.98		0.003	−.5274468	−.109116
dFDIndicator	−.0271466	.0138098	−1.97		0.049	−.0542132	−.0000799
dInflation	.0197008	.0511499	0.39		0.700	−.0805511	.1199527
dSDI	−.005044	.004477	−1.13		0.260	−.0138188	.0037308
dPcFDI	.0128582	.0057619	2.23		0.026	.0015652	.0241513
dPcFPD	.0072034	.0062532	1.15		0.249	−.0050527	.0194595
dFDIGov	−.0146657	.0063321	−2.32		0.021	−.0270764	−.0022549
dH	.0000454	.0021719	0.02		0.983	−.0042113	.0043022
_cons	7.931789	2.725876	2.91		0.004	2.58917	13.27441

The results on the impact of FDI on structural transformation in the full panel, as shown in the eight columns of Table 8, reveal that FDI has a positive and significant impact on structural transformation. This implies that a percentage increase in FDI inflows leads to an approximately 1.37 percent increase in the structural transformation of economies, ceteris paribus. This finding suggests that FDI inflow on its own adds to the structural transformation of nations. This finding supplements the neoclassical economic growth theory and overhang theory, which argue that external sources of financing in the form of FDI encourage economic growth, thereby productivity, and enhance structural transformation. This is also in line with the findings of Pineli et al. (2019) and Muhlen and Escobar (2020) in developing countries.

Table 8. LM test for serial correlation and Jarque–Bera test for normality

I. Durbin-Watson test for serial correlation H0: no autocorrelation at lag order
Durbin-Watson d-statistic(12, 37) = 2.117218
II. Breusch-Godfrey LM test for autocorrelation H0: no autocorrelation at lag order
Lag(p)	chi2	df	Prob > chi2
1	0.797	1	0.3721
III. White’s test for Heteroskedasticity H0:: homoskedasticity
chi2(36)=37.00	Prob > chi2=0.4226
IV. Cameron & Trivedi’s decomposition of IM-test for Heteroskedasticity
chi2	df	p
36	35	0.4215
Jarque-Bera test for normality: H0: Residuals are normally distributed
Equation	chi2	Prob > chi2
D_FER	.7306	.6277

Similarly, the results in Table 8 reveal that in the long run, foreign public debt and inflation indicators negatively and significantly predict a variation in structural transformation across all panels. In the long run, a one percent increase in foreign debt decreases structural transformation by approximately 0.58 percent, ceteris paribus. This shows that the accumulation of debt retards nations’ objectives of achieving structural transformation. This result might be due to the fact that debt is provided in the form of foreign currency in which countries are expected to pay back the principal and interest amount in the long run, which causes capital out-flight and thereby decreases productivity. On the other hand, in the long run, a percentage increase in inflation rate causes the level of structural transformation to decrease by approximately 3.65 percent, which is statistically significant at less than the 1 percent level of significance. This might be because inflation causes production costs to increase, thereby reducing production and productivity.

The short-run (SR) result depends on the heterogeneous nature of economies across countries in the panel. In general, the following findings are summarized as country-specific endogenous variables:

4.2.4. Ethiopia

Only financial development indicators positively and significantly predict structural transformation, and the error correction coefficient converges towards its long-run equilibrium by around 38.5 percent. Accordingly, a percentage increase in the financial development indicator increases structural transformation by around 6 percent, which is statistically significant at less than the 10 percent level of significance, ceteris paribus. This might be due to the fact that improvements in the financial development indicator can strengthen the competitiveness of the economy and thereby enhance productivity and structural transformation.

4.2.5. Kenya

Financial development indicators and foreign public debt adversely affect structural transformation, and the error correction coefficient converges towards its long-run equilibrium. A percentage increase in the financial development indicator decreases structural transformation by approximately 22.74 percent, which is statistically significant at the 1 percent level of significance, ceteris paribus. Similarly, a percentage increase in foreign public debt causes the structural transformation to decrease by approximately 1.27 percent, which is statistically significant at less than the 5 percent level of significance, ceteris paribus. This might be due to the misutilization of public debt in the Kenyan economy.

4.2.6. Nigeria

Foreign public debt positively and significantly predicts structural transformation. All the remaining variables have the expected signs and are insignificant. A percentage increase in foreign public debt leads to an increase of approximately 1.03 percent increment in structural transformation, ceteris paribus. This shows that Nigeria might effectively utilize foreign debts on these activities, which were promising and productive for the economic growth of the country.

4.2.7. South Africa

All variables are statistically insignificant in predicting the variation in structural transformation of the economy of South Africa. According to the data for the study period, the relationship between these included explanatory variables and structural transformation was statistically insignificant, which suggests their Linkage is by random chance. This calls on other scholars to identify other measurement approaches to reconcile these differences.

4.2.8. Singapore

The financial development indicator has an adverse effect on structural transformation. When the financial development indicator increases by 1 percent we expect that structural transformation will increase by around 8.39 percent, which is statistically significant. China: Foreign public debt positively and significantly predicts structural transformation, whereas inflation negatively predicts structural transformation. The coefficient of the error correction term converges towards its long-run equilibrium by around 75 percent. Keeping others constant, as foreign public debt increases by 1 percent we expect that structural transformation will increase by around 0.31 percent, which is statistically significant. This result also supplements the need to effectively utilize debt to enhance the structural transformation of countries.

4.2.9. China

Foreign public debt is positively and significantly predicting structural transformation, and inflation is negatively predicting structural transformation. The coefficient of the error correction term shows convergence towards its long-run equilibrium by around 75 percent. Keeping others remain constant, as foreign public debt increases by 1 percent we expect that structural transformation to increase by around 0.31 percent, which is statistically significant. It pinpoints that foreign public debt has an important role in improving the structural transformation of the country’s economy. This result supplements the need to have effective utilization of debts to enhance structural transformation move of countries. This finding is also inline with the New Structural Economics’ point of view, that debts will be sustainable, if the debts are used to assist the transformation of the manufacturing and industrial sector, and thereby promote comparative advantages in the country.

4.2.10. South Korea

Financial development indicators and FDI with governance negatively and significantly predict structural transformation, and FDI positively predicts structural transformation; the coefficient of the error correction term shows convergence towards its long-run equilibrium by approximately 31.8 percent. Keeping others constant, as financial development indicators and FDI with governance increasing by 1 percent, we expect the structural transformation to decrease by around 2.71 percent and 1.47 percent, respectively, which are statistically significant. As FDI increases by 1 percent, we expect the structural transformation to increase by approximately 1.29 percent, ceteris paribus. These results indicate that FDI without good governance or viable institutions plays no role in enhancing the structural transformation of countries.

4.3. Diagnostic checks for the estimated model

To check the validity of our models, we conducted different diagnostic checks, including the Durbin–Watson test and Breusch–Godfrey test for serial correlation of the error term, Jarque–Bera test for normality, and White’s test for Heteroskedasticity. The results in Table 8 indicate that we accept the null hypothesis that states there is no autocorrelation or heteroskedasticity, and the normality problem due to the p-value is greater than the 5 percent level of significance in bias of accepting the null hypotheses (H_os).

5. Conclusions and policy recommendations

Over decades, the objective of attaining structural transformation has been pivotal, but it is still difficult to achieve in the least developed countries due to lack of technology, low level of literacy, poor governance and institutional setups, lack of managerial expertise, and low development of the manufacturing sub-sector to lead it. FDI is a critical factor in catalyzing the structural transformation of countries. Even though there is a massive inflow of FDI in developing countries, especially in the SSA region, most countries have failed to effectively utilize FDI inflows to improve the structural transformation of their economies, with some even becoming fragile. In addition, previous empirical studies have ignored the interactive effect of FDI and good governance on structural transformation. Furthermore, in the last 60 years, these selected countries from SSA and EAP had almost similar economic backgrounds, yet those from EAP had already transformed their economies. Thus, this study examined the effect of FDI supplemented with good governance on structural transformation, considering four countries from SSA and three countries from EAP for the period 1989 to 2019.

According to the descriptive analysis of this study, on average, structural transformation was lower in these four SSA countries than in the three EAP countries. The results of the PMG model estimation suggest that FDI inflows promote structural transformation in the long-run. Moreover, foreign sources of finance in the form of public debt and inflation rates acted as a discouraging factor in achieving structural transformation. Finally, the above outcomes may indicate that even the existing minimal amount of structural transformation in these selected countries is due to a massive inflow of FDI, among other factors.

Based on the findings of this study, the governments of SSA and EAP countries should better utilize the FDI-accommodative capacity of their economies. In addition, the government should focus on minimizing foreign public debt and control the existing inflationary pressure. This research has few limitations like inclusion of large number of countries from different regions of the world and other important macroeconomic variables. Finally, since this study uses the composite index of governance, it calls on others to identify the relative adequacy of six governance indicators one by one on the link between FDI and structural transformation.

Acknowledgments

I would like to extend my thanks to Wolaita Sodo University and Arba Minch University for giving me the opportunity to pursue my PhD program. Furthermore, thanks go to the World Bank for the free provision of datasets. We also appreciate the journal editorial team and anonymous reviewers who provided valuable comments. Last but not least, I appreciate my family for their direct and indirect support on the journey of the PhD. Thanks to the almighty God.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data are available and can be provided over the email querying directly to the author at the corresponding author ([email protected]) or found in the data depository doi: 10.4121/b26b4989-f903-4b69-84b8-99ac7cc67b04 in Stata Software version 14.

Additional information

Funding

No funding was used in this study.

Notes on contributors

Solomon Kebede Menza

Solomon Kebede Menza is a scholar in Development Economics at Arba Minch university, Ethiopia. He participated as a presenter in more than five national and international research conferences. His teaching and research interests include macroeconomic analysis, financial economics, poverty and livelihood analysis, impact assessment and analysis, value chain analysis and institutional economics.

Zerihun Getachew Kelbore

Zerihun Getachew Kelbore is Country Economist, Macroeconomics Trade and Investment Global Practice, World Bank. He has published many articles in internationally reputable journals. His teaching and research experience includes macroeconomics, time series, and panel data analysis.

Tora Abebe Duka

Tora Abebe Duka Assistant Professor in Economics at Arba Minch University, Ethiopia. His research interests include impact analysis, poverty analysis, and efficiency analysis. He has published several articles in internationally reputable journals.

Berihanu Kuma Shano

Berhanu Kuma Shano is an Associate Professor in Agricultural Economics at Wolaita Sodo University, Ethiopia. He has published more than 30 articles in internationally reputable journals. His research interests include agricultural economics, value chain analysis, dairy farming and livelihood, and poverty analysis.

Notes

1. Economic sectors in this study mean the aggregate classification of the economy as agricultural, manufacturing, and service sectors. And here-onwards we will use the word sectoral synonymously with sector.