Foreign Aid–Human Capital–Foreign Direct Investment in Upper-Middle-Income Economies

Abstract

The study examined the influence of foreign aid on foreign direct investment (FDI) in upper-middle-income economies using panel data (2011–2021) analysis methods such as two-stage least squares (2SLS) and system GMM (generalized methods of moments). The study also explored if human capital development enhanced foreign aid’s influence on FDI in upper-middle-income economies during the same timeframe. The conflicting, divergent, and mixed results and views on the relationship between foreign aid, human capital development, and foreign direct investment (FDI) motivated the undertaking of this study to fill in the existing gaps. Apart from FDI enhanced by its own lag, foreign aid significantly improved FDI (under system GMM). FDI was also improved significantly by human capital development across all two panel methods. Under 2SLS and system GMM, foreign aid significantly improved FDI through the human capital development channel. To promote FDI inflows, upper-middle-income economies should develop and implement policies aimed at attracting foreign aid and enhancing the development of human capital. The study suggests that further research on threshold regression analysis on foreign aid–FDI nexus in upper-middle-income economies could better help develop an FDI policy that is beneficial toward economic growth.

1. Introduction and Background of the Study

Three sub-sections constituting this section include background discussion, contribution of the study (problem statement), and structure of the paper.

Background discussion: The efficacy of foreign aid in less developed nations is a perennial debate given more impetus with the publication of research by Moyo (2010) and Tandon (2008). Moyo (2010) argued that foreign aid to developing nations should come to an end given the not-so-good conditions imposed by donors aimed at dictating the political and economic frameworks of the receiving countries. The research pointed out that foreign aid to less developed nations should be replaced by FDI and trade as the two are durable, reliable, and friendly sources of capital for development, an argument that had earlier on been crushed by Thirlwall (1976), whose study described it as a naïve move. The reason behind the critique was that most less developed nations find it difficult to put in place the locational advantages that not only attract and retain FDI but also enable them to enjoy the spillover benefits associated with the inflow of such long-term external direct investments (C. Wang & Balasubramanyam, 2011). The proposition that instead of seeking to replace each other, foreign aid could complement and enhance the efficacy of FDI in the process of promoting development and growth in less developing nations emerged (C. Wang & Balasubramanyam, 2011, p. 722).

The theoretical literature on the influence of foreign aid on FDI exists but is mixed, conflicting, and divergent on the subject matter. Harms and Lutz (2006) explained that foreign aid improves FDI inflow by increasing foreign private capital productivity, an argument which was generally supported by Asiedu (2006) and R. Quazi et al. (2019). Harms and Lutz (2006) also noted that foreign aid promotes rent-seeking unproductive activities, thereby dissuading and impeding the flow of FDI into host countries. Kosack and Tobin (2006) argued that no relationship exists between foreign aid and FDI as the former is more directed toward augmenting budgetary needs for health and education, not at promoting private investments.

Contribution of the study/Problem statement: Empirical research work also delved into the subject matter, but the findings can be grouped into five categories such as the U-shaped relationship, the positive view, the negative perspective, the neutrality hypothesis, the feedback view, and the absorption capacities narrative. These differing, conflicting, and divergent results are an indication that the topic is not yet fully explored. In addition, these similar empirical studies are characterized by some methodological deficiencies. Firstly, the majority of them ignored the effects of endogeneity on the quality of results (D. Wang & Fillat-Castejon, 2024; Annageldy, 2011). Secondly, they did not address the influence of the omitted variable bias (C. Wang & Balasubramanyam, 2011; Karakaplan et al., 2005). Thirdly, the data they used is outdated and can no longer be suitable for current policy making purposes (Addison & Baliamoune-Lutz, 2020; Karakaplan et al., 2005; C. Wang & Balasubramanyam, 2011; Annageldy, 2011). Fourthly, the majority wrongly assumed that the impact of foreign aid on FDI is in a straight-line format (Tian, 2024; R. Quazi et al., 2019; Karakaplan et al., 2005; C. Wang & Balasubramanyam, 2011). Fifthly, none of these empirical researchers exclusively focused on upper-middle-income economies, an important economic bloc whose stature and importance in the global economic arena is fast growing (Karakaplan et al., 2005; Addison & Baliamoune-Lutz, 2020; C. Wang & Balasubramanyam, 2011; Annageldy, 2011; R. Quazi et al., 2019). Sixthly, the majority of them did not consider the fact that FDI is affected by its own lag (Tian, 2024; Addison & Baliamoune-Lutz, 2020; R. Quazi et al., 2019; C. Wang & Balasubramanyam, 2011; Karakaplan et al., 2005). The current study fills in these gaps.

Structure of the paper: The rest of the paper is structured as follows: Section 2 extensively reviews the relevant literature. This includes the theoretical and empirical literature on the impact of foreign aid on FDI; the link between foreign aid, human capital development, and FDI; and other factors that influence FDI. Section 3 presents and explains the research methodological framework. Section 4 focuses on results presentation, discussion, and interpretation. Section 5 concludes.

2. Literature Review

There are four sections that discuss the literature review, and these are presented next.

Theoretical literature on the impact of foreign aid on FDI: Three dominant theoretical views explain the relationship between foreign aid and FDI. Firstly, foreign aid enhances FDI, an argument that was theoretically supported by Harms and Lutz (2006). They explained that foreign aid enhances FDI through its ability to increase private capital productivity and through financing infrastructural public investments at a low cost to the host country. Foreign aid is viewed by international foreign firms as positive outside help increasing their appetite to invest in countries characterized by high risk (Asiedu, 2006). According to Garriga and Phillips (2014), the flow of foreign into a country is a positive signal for foreign direct investors to begin investing in that country, especially in post-conflict nations. The reason is that investment information in these post-conflict nations is not only expensive to collect but also takes a long time to gather. R. Quazi et al. (2019) argued that because of its alignment with non-political factors, multilateral aid is more likely to enhance capital productivity, which in turn attracts FDI.

Secondly, foreign aid has adverse effects on FDI, a theoretical view that was supported by Harms and Lutz (2006). The theoretical argument is that foreign aid creates incentives for rent-seeking unproductive activities (more pronounced in countries characterized by volatile political environments and weak property rights) in the receiving country thereby discouraging the flow of FDI into the host country. R. Quazi et al. (2019) argued that bilateral aid is more likely to be channeled toward non-productive projects because it is aligned with selfish geopolitical interests, thereby dissuading FDI inflows. Meernik et al. (1998) explained that foreign aid (bilateral) flow determined by security concerns and geopolitical interests (the United States in most instances) is a warning sign to foreign direct investors against putting their long-term capital into that country.

Thirdly, a non-existent relationship between foreign aid and FDI was supported by Kosack and Tobin (2006). The argument is that foreign aid is generally directed toward supplementing the government budget and developing human capital development (health, education, and skills); therefore, it is not related to FDI. In support of that view, Jansky (2012) explained that foreign aid does not substitute inadequate FDI inflows; hence, there is no room for it to correct market potential failures.

Empirical literature on the impact of foreign aid on FDI: Empirical literature is discussed in the next few paragraphs. Addison and Baliamoune-Lutz (2020) examined whether FDI is influenced by foreign aid in Sub-Saharan Africa (SSA) and Latin American countries using panel data (1985–2008) analysis. The study noted that in countries characterized by high levels of human capital development, FDI is crowded out by foreign aid. In other regions, foreign aid positively enhanced FDI whilst a negative influence of foreign aid on FDI was observed in Latin American and SSA nations. Moreover, the interaction between social cohesion and foreign aid had a U-shaped influence on FDI in SSA and Latin American groups of countries. Using panel data analysis, Karakaplan et al. (2005) examined the interplay between foreign aid and FDI in developing nations. The study produced results that show that foreign aid enhanced FDI competitiveness in developing countries.

C. Wang and Balasubramanyam (2011) examined the impact of foreign aid on FDI in Vietnam using descriptive statistical analysis with data ranging from 1985 to 2010. The study noted that FDI was enhanced by foreign aid whilst the complementarity between FDI and foreign aid encouraged economic growth in Vietnam. Using generalized methods of moments, Ono and Sekiyama (2023) examined the FDI’s impact on foreign aid in five donor countries (Germany, United Kingdom, France, Japan, and the United States) versus sixty-three recipient countries using panel data ranging from 1996 to 2020. Foreign aid from Japan, Germany, and the United Kingdom promoted FDI through channels such as finance, energy, telecommunication, and transport infrastructural development enhancement. A foreign aid–FDI nexus was studied by Selaya and Sunesen (2012) in developing countries using panel data analysis (generalized methods of moments). The study noted that foreign aid increases the effectiveness of FDI if it is channeled toward human capital development and public infrastructural projects.

D. Wang and Fillat-Castejon (2024) examined the interplay between political power, foreign aid, and FDI in Africa using panel data (2002–2016) analysis. The results show that high-quality political institutions and foreign aid positively influenced FDI in Africa. Using panel data (1987–2013) analysis, Ahrsjo (2016) explored whether FDI is substituted or complemented by foreign development aid in thirty-five middle-income countries. The results show that foreign development aid complemented FDI, mainly because of foreign development aid’s ability to address market failures that spur investment shortages in developing countries. Foreign aid was generally found to be a catalyst for private foreign capital inflows into developing countries, therefore enhancing long-term economic growth. Tian (2024) examined the effect of foreign aid on FDI in developing nations using the instrumental variable approach with panel data spanning from 1987 to 2019. The study observed that foreign aid enhanced FDI in developing countries during the period under study.

Annageldy (2011) studied the interlinkages between FDI, foreign aid, and domestic investment in Central Asia (emerging and landlocked countries) using panel data analysis. Foreign aid enhanced FDI in Central Asia both at country and regional levels. The complementary relationship between FDI and foreign aid was also confirmed especially in countries characterized by low economic growth. The effect of foreign aid on FDI in post-conflict economies was studied by Garriga and Phillips (2014) using panel data (1973–2008) analysis. The results show that FDI was enhanced by foreign aid inflows depending on whether the foreign aid came from other nations or the United States. Foreign aid that came from other countries outside the United States enhanced FDI in post-conflict nations (positive signal). United States foreign aid’s positive influence on FDI in post-conflict nations came as a warning sign because of the United States’ tendency to provide aid for geostrategic motives.

A study on the influence of foreign aid on FDI in Africa was also examined by R. Quazi et al. (2019) using a feasible generalized least squares approach, with panel data spanning between 1996 and 2017. Foreign aid was found to have had a significant enhancing impact on FDI inflows into Africa. When disaggregated data were employed, bilateral aid had no impact on FDI whereas multilateral aid’s enhancing influence on FDI was well pronounced in Africa. Using pooled mean group and panel autoregressive distributive lag with data spanning from 1990 to 2019, Dash et al. (2024) studied foreign aid’s influence on capital formation in South Asian nations. Foreign aid crowded out FDI inflow in the long run but complemented FDI in the short run. A feedback relationship between FDI and foreign aid was also observed in both the short and long run.

Amusa et al. (2016) examined the interrelationship between FDI and foreign aid in Sub-Saharan Africa using panel data (1995–2012) analysis. Foreign aid related to socio-economic infrastructure was found to have no significant influence on FDI in the SSA group of nations. On the other hand, foreign aid linked to productive infrastructure complemented FDI in SSA countries. Socio-economic and productive infrastructure aid to oil-producing countries attracted less FDI in comparison to non-oil-producing nations. Employing system generalized method of moments (GMM) with panel data (1995–2015), Michael (2018) examined foreign aid’s influence on FDI in Africa. The foreign aid-led FDI hypothesis was supported by the results of this study. The results also confirmed that foreign aid channeled through multilateral institutions attracted more FDI into Africa.

Ndambendia and Njoupouognigni (2010) studied the interplay between FDI, foreign aid, and economic growth in Sub-Saharan Africa (SSA) using panel data (1980–2007) analysis. Foreign aid and FDI complemented each other in enhancing economic growth in SSA. Pham (2015) examined the influence of foreign aid on FDI in Vietnam using two-stage least squares with time series data (1998–2012) analysis. Foreign aid was found to have had a positive influence on FDI. Using panel data (1995–2012) analysis, R. M. Quazi et al. (2014) explored the relationship between FDI and foreign aid in East Asia and South Asia. A significant positive influence of foreign aid on FDI in East Asia and South Asia was noted. Apart from foreign aid, FDI was also affected by the rate of return, corruption, human capital development, infrastructure, and political stability.

Freyer et al. (2024) examined the interrelationship between FDI and foreign aid in Ukraine using multiregression analysis with time series data spanning from 1992 to 2022. Foreign aid spurred FDI according to the results whilst economic growth was also enhanced by the interaction between FDI and foreign aid in Ukraine. Time series data (1990–2018) analysis was also employed by Ibukun et al. (2024) to investigate the interrelationship between human capital development, foreign aid, and FDI in Nigeria. A long-run relationship using Johansen co-integration analysis was observed during the period under study. Dayaratna-Banda (2004) examined the role played by foreign aid in stimulating FDI inflows into Sri Lanka using the dynamic time series analysis. The results agree with the foreign aid-led FDI inflow perspective.

Ono and Sekiyama (2024) examined the nexus between FDI and foreign aid in the case of Japanese companies in India using multiregression analysis with primary data. Foreign aid’s positive influence on FDI was lesser in comparison to other variables which affected FDI into India. The study also noted that economic infrastructure development coming through foreign aid had a more significant enhancing effect on FDI inflows into India during the period under study. Furthermore, the study noted that FDI in India was enhanced by Japanese foreign aid because the latter enhanced the availability of locational and ownership-specific advantages in India.

Ifeakachukwu and Oladiran (2019) examined the complementarity between foreign aid and FDI in Nigeria using a vector error correction model (VECM) with time series data ranging between 1970 and 2016. The study noted that foreign aid and FDI are substitute investments in the process of enhancing economic growth in Nigeria. Using autoregressive distributive lag (ARDL) with time series data from 1992 to 2018, Slesman (2023) examined whether FDI is influenced by foreign aid in post-conflict Cambodia. In the long run, donor-specific aid and aggregate development aid from the United Nations and Australia significantly enhanced FDI inflows into post-conflict Cambodia. Foreign aid from the European Union enhanced FDI inflows only in the short run in post-conflict Cambodia.

Using multiregression analysis with time series data (1990–2018), Kim and Cho (2023) studied the foreign aid–FDI–human capital nexus in Korea. Foreign aid was found to have had a moderating influence on the relationship between FDI and human capital development in Korea. The study also observed that human capital development was enhanced significantly by FDI in Korea. Bhavan (2014) also examined the FDI–foreign aid in South-East Asia using a fixed effects approach with panel data spanning from 1995 to 2012. The study noted that although foreign aid was found to be a catalyst for FDI inflows into South-East Asia, social infrastructure aid was observed to be more effective than physical infrastructure aid in terms of attracting FDI into South-East Asia. The study also observed that foreign aid from the United Kingdom had a more significant influence on FDI inflows whereas foreign aid from Germany, the United States and Netherlands’ impact on FDI inflows into South-East Asia was marginal.

Employing the error correction model (ECM) with time series data ranging from 1970 to 2020, Musakwa and Odhiambo (2023) studied the relationship between economic growth, FDI, and foreign aid in Kenya. The results showed a feedback relationship between foreign aid and FDI in the short run. In the long run, a bi-directional causality was observed running from foreign aid toward FDI in Kenya.

The theoretical literature on the foreign aid–FDI nexus is divided into three categories: the foreign aid-led positive FDI, foreign aid-led negative FDI, and the neutral hypothesis. The empirical literature produced findings that can be categorized into six sections: the results that support the positive influence of foreign aid on FDI and another group of empirical results that says that foreign aid negatively or crowds out FDI; views that support that there is no relationship between foreign aid and FDI and a group of findings that argues that certain locational advantages must be present in host countries before FDI can be enhanced by foreign aid; and another group of findings states that foreign aid and FDI influence each other whilst the last and final category of results observes that foreign aid and FDI’s relationship is U-shape in nature. It is, therefore, evident that there is no consensus on the influence of foreign aid on FDI. In fact, the findings are conflicting, mixed, and divergent and clearly show that the subject matter is still far from being resolved hence still requires some further empirical interrogation.

Link between foreign aid, human capital, and FDI: According to Kosack and Tobin (2006), foreign aid is in most instances channeled toward augmenting government budgets and development of human capital (health, education, and skills). The same study argued that only developing nations characterized by high levels of human capital development can effectively benefit from local knowledge spillovers associated with FDI because the resultant jobs created are mainly for skilled people.

Ono and Sekiyama (2023) noted that foreign aid from Japan, Germany, and the United Kingdom promoted FDI through channels such as finance, human capital, energy, telecommunication, and transport infrastructural development enhancement. Selaya and Sunesen (2012) also revealed that foreign aid increased the effectiveness of FDI if channeled toward human capital development and public infrastructural projects. Ono and Sekiyama (2024) also noted that FDI in India was improved by Japanese foreign aid because the latter enhanced the availability of locational and ownership-specific advantages in India.

According to Dunning (1980), FDI into the host country is lured by high levels of human capital quality because the multinational companies involved then find it cheaper to employ locally available educated and skilled workforce than hiring from other countries. Since skilled and educated people adapt to new technology at a faster rate, Freckleton et al. (2012) argued FDI is lured into countries characterized by high-quality human capital availability. Assimilation of FDI spillovers is easier and quicker in countries with high levels of human capital development.

Addison and Baliamoune-Lutz (2020) observed that in Sub-Saharan Africa (SSA) and Latin American countries characterized by high levels of human capital development, FDI was crowded out by foreign aid. Moreover, the interaction between social cohesion and foreign aid had a U-shaped influence on FDI in SSA and Latin American groups of countries. These contrasting views and results make the nexus between foreign aid, human capital development, and FDI complicated, unresolved, and far from being settled in the field of development economics.

Other factors that influence FDI: 

Table 1. Control variables’ theoretical influence on FDI.

Variable	Theoretical Explanation and Proxy Used	Direction of Influence
Savings (SV)	Savings stimulates domestic and foreign investment projects thereby improving economic growth (Lucas, 1988; Romer, 1986). Domestic savings as a ratio of GDP was used as a measure of savings. Domestic savings add liquidity to the economy, thereby reducing the need for FDI-related capital injections (Selaya & Sunesen, 2012).	+
Financial development (FD)	Domestic credit to the private sector (as a ratio of GDP) is a proxy for financial development used in this study. Ezeoha and Cattaneo (2012) argued that developed financial markets improved foreign capital productivity.	+
Infrastructure development (INF)	Richaud et al. (1999) explained that for the host country to enjoy the spillover benefits of FDI, infrastructural development levels need to be very high. Developed energy, transport, road, and communication infrastructure promotes participation in international commerce and production networks (Limao & Venables, 2001).	+
Trade openness (TO)	Denisia (2010) noted that trade openness is a locational advantage of FDI in the host country thereby attracting FDI, consistent with the eclectic paradigm hypothesis. Denisia (2010) argued that access to international product markets is enhanced by high levels of trade openness hence international firms have less need to establish production plants in foreign lands. This means that trade openness substitutes FDI. Total trade (% of GDP) is a trade openness proxy used in this study.	+
Personal remittance (PR)	Personal remittance received (% of GDP) is a measure of remittances used in this study. Personal remittances promote domestic investment by availing liquidity in the economy from foreign countries. In other words, personal remittance inflow reduces the demand for FDI inflows required to meet the liquidity needs of the host country.	+

Source: Author.

below explains the control variables and how they each theoretically affect FDI.

3. Methodology

This study used panel data (2011–2021) which was collected from reputable, reliable, and publicly available databases such as the International Monetary Fund, World Development Indicators, and United Nations Development Programme. The period chosen (2011–2021) was the most ideal in this study because during this period, most upper-middle-income countries experienced very rapid foreign direct investment, economic growth, and financial market development. Data availability for the countries and variables chosen was also considered in selecting the sample period. The paper used upper-middle-income economies as a focal point of analysis and these include South Africa, Turkey, Republic of Korea, Peru, Mexico, Indonesia, Colombia, Brazil, Philippines, India, and Argentina. According to Mazzi (2013), upper-middle-income countries are nations that are developed but not to the level of achieving a developed nation’s status. They are developed to at least have or be characterized by rudimentary financial markets (Mazzi, 2013, p. 101). Upper-middle-income countries have the characteristics of a developed country without attaining the standards (highly developed financial markets and high income per capita) of a developed market (MSCI Index Research, 2014).

The selection of upper-middle-income economies was informed by the availability of data for the variables in question. Regional balance was also a factor because Africa, the Middle East, East Asia, and North America are all represented in this study. This study used upper-middle-income countries because they are an international economic and financial grouping that is characterized by rapid market liberalization, economic growth, political reform changes, financial market development, and foreign direct investment inflows (Cavusgil et al., 2013, p. 7).

The general model specification is represented in Equation (1), whose dependent variable is foreign direct investment (FDI) whilst foreign aid (FAID) is the independent variable. The control variables in this model include human capital development (HCAP), savings (SV), financial development (FD), infrastructure development (INF), trade openness (TO), and personal remittance (PR). The selection of control variables used in this study is in line with empirical studies such as Addison and Baliamoune-Lutz (2020), Ono and Sekiyama (2023), D. Wang and Fillat-Castejon (2024), Tian (2024), Garriga and Phillips (2014), R. Quazi et al. (2019), Dash et al. (2024), Amusa et al. (2016), Michael (2018), Freyer et al. (2024), Ibukun et al. (2024), Ifeakachukwu and Oladiran (2019), Kim and Cho (2023), Musakwa and Odhiambo (2023), and Slesman (2023).

FDI = f (FAID, HCAP, SV, FD, INF, TO, PR)

(1)

Net FDI inflow as a ratio of GDP (gross domestic product) is the proxy of FDI used in this study. Human capital development (HCAP) was measured using the human capital development index. The study also employed net official development assistance received as a ratio of GDP as a proxy of foreign aid (FAID).

The FDI econometric function is represented by Equation (2). FAID × HCAP is a variable that shows a complementarity between foreign aid and human capital development, consistent with existing empirical studies (Ono & Sekiyama, 2023; Selaya & Sunesen, 2012). Ono and Sekiyama (2023) noted that foreign aid from Japan, Germany, and the United Kingdom promoted FDI through channels such as finance, human capital, energy, telecommunication, and transport infrastructural development enhancement. Selaya and Sunesen (2012) also revealed that foreign aid increased the effectiveness of FDI if channeled toward human capital development and public infrastructural projects. The interaction term is, therefore, expected to have a positive effect on FDI, in the context of upper-middle-income countries.

$${\mathrm{FDI}}_{\mathrm{it}}={\beta }_0+{\beta }_1{\mathrm{FAID}}_{\mathrm{it}}+{\beta }_2{\mathrm{HCAP}}_{\mathrm{it}}+{\beta }_3 \left({\mathrm{FAID}}_{\mathrm{it}}·{\mathrm{HCAP}}_{\mathrm{it}}\right)+{\beta }_4{\mathrm{SV}}_{\mathrm{it}}+{\beta }_5{\mathrm{FD}}_{\mathrm{it}}+{\beta }_6{\mathrm{INF}}_{\mathrm{it}}+{\beta }_7{\mathrm{TO}}_{\mathrm{it}}+{\beta }_8{\mathrm{PR}}_{\mathrm{it}}+\mu +\mathsf{\epsilon }$$

(2)

where $\mu$ is time-invariant and unobserved country-specific effects, and ε is the error term. ${\beta }_0$ is the intercept term. Subscript $t$ stands for time whilst subscript $i$ represents country. The dynamic characteristic of FDI data as represented by FDI_it−1 (which is the lag of FDI), consistent with Walsh and Yu (2010, p. 5) is captured in Equation (3). It also resonates with Selaya and Sunesen (2012), whose study pointed out the persistent characteristic of FDI data.

$${\mathrm{FDI}}_{\mathrm{it}}={\beta }_0+{\beta }_1{\mathrm{FDI}}_{\mathrm{it}-1}+{\beta }_2{\mathrm{FAID}}_{\mathrm{it}}+{\beta }_3{\mathrm{HCAP}}_{\mathrm{it}}+{\beta }_4\left({\mathrm{FAID}}_{\mathrm{it}}·{\mathrm{HCAP}}_{\mathrm{it}}\right)+{\beta }_5{\mathrm{SV}}_{\mathrm{it}}+{\beta }_6{\mathrm{FD}}_{\mathrm{it}}+{\beta }_7{\mathrm{INF}}_{\mathrm{it}}+{\beta }_8{\mathrm{TO}}_{\mathrm{it}}+{\beta }_9{\mathrm{PR}}_{\mathrm{it}}+\mu +\mathsf{\epsilon }$$

(3)

When co-efficient ${\beta }_4$ is positive and significant, it means that FDI is significantly improved by the complementarity between foreign aid and human capital development in upper-middle-income economies. To address the dynamic nature of FDI, the FDI influence of the complementarity variable, and the endogeneity problem associated with the relationship between FDI and foreign aid, this study employed the system GMM econometric estimation approach to estimate econometric Equation (3). The superiority of the system GMM approach is as follows: it can eliminate country-fixed effects using the forward orthogonal transformation approach, and it approximates the endogenous regressor and combines both cross-section and time series data into a panel threshold data analysis. Other alternative methods are incapable of matching this standard.

4. Results Presentation and Discussion

This section presented and discussed results on the correlation of the variables, descriptive statistics, panel stationarity tests, panel co-integration, and main data analysis.

Table 2. Analysis of correlation results.

	FDI	FAID	HCD	SV	FD	INF	TO	PR
FDI	1.00
FAID	0.17 *	1.00
HCD	−0.16 *	−0.24 ***	1.00
SV	−0.35 ***	−0.30 ***	−0.07	1.00
FD	−0.34 ***	0.10	0.30 ***	0.18 **	1.00
INF	−0.11	−0.07	0.87 ***	−0.22 **	0.44 ***	1.00
TO	−0.26 ***	−0.14	0.16 *	0.46 ***	0.39 ***	−0.003	1.00
PR	0.02	0.08	−0.40 ***	0.42 ***	−0.26 ***	−0.35 ***	0.17 *	1.00

Source: author. ***, **, and * denote 1%, 5%, and 10% levels of significance, respectively.

presents correlation analysis results whilst

Table 3. Statistics of descriptive in nature.

	FDI	FAID	HCD	SV	FD	INF	TO	PR
Mean	2.32	0.14	0.75	23.67	59.40	54.27	46.54	1.88
Median	2.02	0.09	0.76	21.50	49.06	53.76	45.39	0.91
Maximum	9.66	0.69	0.92	37.35	170.82	97.57	105.57	9.99
Minimum	0.28	0.00	0.58	11.39	13.67	10.07	13.52	0.07
Standard deviation	1.41	0.14	0.08	8.11	36.96	21.65	19.53	2.68
Skewness	1.60	1.21	0.23	0.27	1.26	−0.08	0.64	2.13
Kurtosis	8.29	4.06	2.94	1.61	3.64	2.47	3.25	6.56
Jarque–Bera	193.01	34.91	1.13	11.27	33.86	1.55	8.46	155.80
Probability	0.00	0.00	0.57	0.00	0.00	0.46	0.01	0.00
Observations	121	121	121	121	121	121	121	121

Source: Author.

shows descriptive statistics.

In Table 2, a significant positive correlation between FDI and foreign aid was observed. On the other hand, a negative but significant correlation was also noted between (1) FDI and human capital development, (2) FDI and domestic savings, (3) FDI and financial development, and (4) FDI and trade openness. The relationship between infrastructure development and FDI was observed to be negative and insignificant. Personal remittances were also positively and non-significantly related to FDI. A multi-collinearity problem was noted between human capital development and infrastructure development (87%), in line with Stead (2007)’s argument.

Table 3 indicates that data for all the variables is not normally distributed, as shown by the probability values of the Jarque–Bera criterion which is equal or close to zero.

Table 4. System GMM results.

	2SLS	GMM-SYS
$FD{I}_{i,t-1}$	-	0.05 ***
FAID	0.01	0.63 **
HCD	0.34 *	0.39
FAID.HCD	0.03 ***	0.02 **
SV	0.88	0.16
FD	0.92	0.04
INF	0.07	0.43
TO	0.19	0.26
PR	−0.11	−0.17
AR1	-	−1.8
Pr > z	-	0.004
AR2	-	0.87
Pr > z	-	0.001
Sargan test	0.54	14.38
Prob test > chi2	0.85	0.99
Hansen test	33.92	38.01
Prob test > chi2	0.43	0.72
Observations	121	121
Number of countries	11	11

Source: author. ***, **, and * denote 1%, 5%, and 10% levels of significance, respectively.

presents the results of panel stationarity tests. The data for all the variables was stationary at first difference hence integrated of the first order. The natural logarithm format of the data allowed the study to address the problems of multi-collinearity, outliers, and abnormally distributed data sets, in line with Aye and Edoja (2017)’s argument.

The results of the main data analysis are included in Table 4,

Table A1. System GMM results—institutional quality.

	Political Stability	Regulatory Frameworks	Control Level	Perceived Risks	Institutional Weaknesses
$FD{I}_{i,t-1}$	0.04 ***	0.78 ***	0.03 ***	0.09 ***	0.17 ***
FAID	0.32	0.18 *	0.47	0.26	0.42 *
HCD	0.43 **	0.48	0.23 *	0.39 *	0.27
FAID.HCD	0.02 ***	0.18 *	0.26	0.33 **	0.18 ***
SV	0.12	0.32	−0.48	0.38	−0.43
FD	0.94	0.91	0.18	0.43	0.37
INF	0.38	0.32	0.28	0.37	0.27
TO	0.21	0.34	0.87	0.38	0.28
PR	0.83	0.22	0.38	0.82	0.77
AR1	−1.03	−1.34	−1.32	−1.43	−1.89
Pr > z	0.003	0.008	0.01	0.004	0.009
AR2	0.89	0.93	0.43	0.17	0.29
Pr > z	0.38	0.28	0.43	0.19	0.63
Sargan test	0.35	0.43	0.39	0.28	0.89
Prob test > chi2	1.000	1.004	0.89	0.78	0.93
Hansen test	43.09	28.92	54.86	34.83	33.87
Prob test > chi2	0.62	0.74	0.94	0.48	0.32
Observations	121	121	121	121	121
Number of countries	11	11	11	11	11

Source: author. ***, **, and * denote 1%, 5%, and 10% levels of significance, respectively.

and

Table A2. System GMM results—country-specific factors.

	Unique Economic Structures	Governance Levels	Policy Environments
$FD{I}_{i,t-1}$	0.44 ***	0.32 ***	0.05 ***
FAID	0.84	0.31	0.43 **
HCD	0.03 *	0.34	0.83 ***
FAID.HCD	0.43 ***	0.94 *	0.17 *
SV	0.27 *	−0.74	0.28 *
FD	0.03 *	0.48 *	0.09 *
INF	0.33 **	0.28	0.17 ***
TO	0.43 *	−0.27	0.22 *
PR	−0.19	−0.43	−0.39
AR1	0.11	0.54	0.75
Pr > z	0.35	0.53	0.39
AR2	0.84	0.92	0.98
Pr > z	1.03	1.32	0.91
Sargan test	46.83	28.03	77.29
Prob test > chi2	0.43	0.65	0.76
Hansen test	43.09	27.00	39.20
Prob test > chi2	0.62	0.26	0.84
Observations	121	121	121
Number of countries	11	11	11

Source: author. ***, **, and * denote 1%, 5%, and 10% levels of significance, respectively.

.

There are two methodological limitations of this study worth noting. The data used only covers the period from 2011 to 2021 and only includes 11 countries. It, therefore, means that these results cannot be generalized to other income groups or to longer time periods.

This paper then used system GMM, an approach that is more efficient in addressing endogeneity problems in comparison to different GMMs employed by Kremer et al. (2013). Two-stage least squares was also used as part of the robustness checks.

Table 4 also shows that FDI was affected by its own lag in a significant manner, in agreement with Selaya and Sunesen (2012), whose study revealed the persistent characteristic of FDI data. The results support Walsh and Yu (2010)’s argument that the flow of FDI acts as a positive signal for a favorable investment climate, hence attracting more foreign investment inflow into the host country.

System GMM noted that foreign aid significantly improved FDI whilst 2SLS shows that FDI was non-significantly enhanced by foreign aid. These results agree with authors such as Harms and Lutz (2006), whose study explained that FDI is enhanced by foreign aid through the latter’s ability to improve the productivity of private capital by ensuring low-cost financing of infrastructural public investments. The results were also supported by Garriga and Phillips (2014) and Asiedu (2006), whose studies agreed that foreign aid provides a positive signal for international foreign investors to take an interest in that country.

System GMM indicates that human capital development insignificantly improved FDI whilst 2SLS had a significant influence impact on FDI. Both 2SLS and system GMM enhanced FDI in a significant manner. The results are consistent with Freckleton et al. (2012), whose study argued that host countries characterized by developed human capital assimilate FDI spillovers quicker and easier. The results also confirm the assertion that FDI is attracted to host countries characterized by an educated, healthy, and skilled workforce. The correlation results in Table 2 indicated a significant negative relationship between FDI and human capital development whilst all three econometric methods observed that human capital had a significant enhancing effect on FDI. Such a contradiction shows that the relationship between these two variables is non-linear. It is influenced by the existence of other variables that affect FDI.

The complementarity between foreign aid and human capital development significantly enhanced FDI in upper-middle-income economies. The results are consistent with Selaya and Sunesen (2012), whose empirical study observed that if foreign aid is channeled through human capital and public infrastructure development projects, its effectiveness on FDI promotion and retention improves. They are also in sync with those produced by Ono and Sekiyama (2023) regarding foreign aid–human capital–FDI nexus.

A non-significant relationship running from domestic savings toward FDI was observed under both models. These findings resonate with Lucas (1988), whose study explained that foreign and domestic investment is stimulated by savings. Financial development’s influence on FDI was positive and non-significant across the two models. The results resonate with Ezeoha and Cattaneo (2012), whose study noted that foreign capital productivity was spurred by a developed financial system.

FDI was non-significantly improved by infrastructural development under both models. The results generally mean that infrastructure development improved FDI in upper-middle-income economies, in line with Richaud et al. (1999), whose study noted that a developed infrastructure promotes beneficiation from FDI spill overs and enhances participation in international commerce and production networks.

Trade openness’s impact on FDI was found to be positive and non-significant under both models in general agreement with Denisia (2010)’s theoretical view which singled out that trade openness is a locational advantage of FDI within the eclectic paradigm hypothesis. Personal remittances’ influence on FDI was non-significantly negative across both models, in line with the existing theoretical literature which explains that personal remittances provide liquidity into the economy thereby potentially substituting FDI inflows.

For robustness purposes, system GMM considered the role of institutional quality (such as political stability, regulatory frameworks, and control of corruption) which significantly influences FDI decisions (see Table A1 in the Appendix A section). System GMM considered the country-specific factors such as unique economic structures, governance levels, and policy environments, which can lead to varying impacts of foreign aid on FDI (see Table A2 in the Appendix A section).

Under 2SLS, the chi-square statistic is 0.54 (Sargan test) and 0.43 (Hansen test) and their associated probability values are greater than 0.05. The study, therefore, accepted the alternative hypothesis that the independent variables are not endogenous under both 2SLS and system GMM. This means that evidence that any of the independent variables in both models are correlated with the error term is not available. The results can, therefore, be confidently interpreted without the worry of the endogeneity problem.

5. Conclusions

The study examined the FDI influence of foreign aid in upper-middle-income economies using panel data (2011–2021) analysis methods such as 2SLS and system GMM (generalized methods of moments) The same study also explored if human capital development enhanced foreign aid’s influence on FDI in upper-middle-income economies during the same period. Apart from FDI enhanced by its own lag, foreign aid significantly improved FDI (under system GMM). FDI was also significantly enhanced by human capital development across all three panel methods. Under 2SLS and system GMM, foreign aid significantly improved FDI through the human capital development channel in upper-middle-income countries. To promote FDI inflows, upper-middle-income economies should develop and implement policies aimed at attracting foreign aid and enhancing the development of human capital. Upper-middle-income countries should also implement policies that enhance savings, financial development, infrastructure development, and trade openness to facilitate the inflow and retention of FDI.

The study had two main limitations: The study could have included more upper-middle-income countries if the data were available. Moreover, more variables could also have been included in the econometric model if data were available. This study does not consider that the effectiveness of foreign aid can vary based on how it is administered and utilized. Factors such as tied aid—which requires recipients to spend aid on donor-country products—can limit the positive impact of aid on development outcomes, including FDI. Further research is recommended to capture this very important aspect of the literature on FDI. The study suggests that further research on threshold regression analysis regarding the relationship between foreign aid and FDI in upper-middle-income economies could better help develop an FDI policy that is beneficial toward economic growth.