Effects of Science, Technology, and Innovation Official Development Assistance on Foreign Direct Investment in Developing Countries

Abstract

This study investigated the effects of Science, Technology, and Innovation Official Development Assistance (STI ODA) on Foreign Direct Investment (FDI) in developing countries. The study sought to empirically analyze whether STI ODA has contributed to attracting FDI in recipient countries and whether specific sub-variables, such as absorptive capacity, human capital, infrastructure, and scientific and technical journal articles have a moderating effect. The analysis method was a panel analysis model that combined cross-sectional and time-series data as fixed-effect and random-effect models to reflect individual effects in the empirical analysis. This study highlights the importance of STI ODA in achieving sustainable development goals and fostering economic growth.

1. Introduction

Science, technology, and innovation (STI) is a key driver of economic and social development [1] and STI plays an essential role in accomplishing sustainable development goals in developing countries. In 2015, the United Nations addressed 17 sustainable development goals (SDGs) and launched a technology facilitation mechanism (TFM) to implement the SDGs. The TFM was established to promote multi-stakeholder cooperation for STI access through various examples and policy advice, including the sharing of information and experience among member states, to achieve the SDGs.

The TFM also aims to develop STI levels in developing countries to achieve the SDGs [2]. The OECD stated in 2015 that the 2030 Agenda for SDGs recognized the importance of STI as a key enabler in achieving several of the 17 goals. In 2002, the OECD encouraged foreign direct investment (FDI) as a major driver of development because it plays an essential role in an open and effective economic system that benefits investors and the countries receiving investments [3]. Recent studies have also demonstrated an empirical link between FDI and economic growth [3,4].

Some studies have found a relationship between FDI and economic growth and established the positive impact of FDI on economic performance [5,6,7]. However, little research was conducted on the contribution of STI ODA to the FDI of beneficiary countries. One reason for this is that statistical measurements were difficult because of the lack of a common definition owing to the cross-cutting nature of STI ODA [8]. Our study assessed whether STI ODA attracts FDI from developing countries. It is also critical to examine whether the absorptive capacity of developing countries has a moderating effect on FDI attraction. Therefore, it is important to identify and organize the sub-variables of absorptive capacity.

This study aims to examine empirically how STI ODA impacted the FDI in recipient countries. The initial analysis method was a panel analysis model that combined cross-sectional and time-series data as fixed- and random-effect models to reflect individual effects in the empirical analysis.

The reason why we want to analyze this impact is that recent research was conducted to empirically analyze the impact of STI ODA on innovation capabilities in developing countries [9], and the importance of evaluating STI ODA effectiveness is increasing [1]. The findings of the present study could highlight the importance of STI ODA in resolving global issues such as climate change, nuclear non-proliferation, and energy crises in developing countries, as well as the importance of international development cooperation in science and technology innovation [10].

Therefore, through this study, we will try to answer the question of whether STI ODA has contributed to receiving FDI. In addition, we will try to answer whether the sub-variables have a moderating effect. The results of this study may be helpful in determining which direction to implement when making a policy.

Moreover, data from the OECD or World Bank do not provide statistics (amount of funding received by country) by clearly distinguishing the STI ODA sector. It is difficult to find research that statistically collects STI ODA by country and empirically analyzes its impact on FDI in those countries.

Thus, research related to FDI, which can lead to improved economic performance in countries with STI ODA benefits (using statistical data), may be useful. In addition, this study classified STI ODA by country to determine the correlation between STI ODA and FDI, which could help establish a tool for better visualization and formulation of aid policies for STI ODA in advanced countries.

2. Literature Review

2.1. STI ODA and FDI

ODA is defined as government aid that promotes and specifically targets the economic development and welfare of developing countries, and the volume of ODA in 2022 (204 billion) has increased 13.6% from 2021 (185.9 billion) [11]. As the amount of ODA increases, the focus and transfers of science and technology increase, along with traditional areas such as health and education [12]. STI can be a necessary and sufficient condition to achieve the SDGs adopted by the United Nations General Assembly in 2015 [13].

STI ODA includes science, technology, and innovation, although the OECD Development Assistance Committee (DAC) does not make particular distinctions in this regard [11]. STI ODA was classified using the OECD creditor reporting system (CRS) purpose code. The OECD DAC CRS code is used to produce ODA statistics, which includes assistance between multilateral and bilateral countries [14]. In addition, the OECD CRS code was used as the primary data source, and World Bank analysis was performed to calculate the top donor’s share of development finance toward STI ODA disbursed in 2017 [15].

The amount of STI ODA fluctuated between USD 11–17 billion between 2010 and 2016, and one of the reasons for the large fluctuations was that the amount of aid differed annually [15]. Several studies have examined the impact of international aid on FDI. Bhavan et al., Carro, and Larrú [16,17] found that international aid attracts FDI in Bangladesh, Sri Lanka, Pakistan, and India. Wang and Bal Asubramanyam [18] showed that international aid positively affects the attraction of FDI and contributes to economic growth. Burnside and Dollar [19] showed that international aid has a positive impact on growth in recipient countries. Blaise [20] found a positive effect of international aid on infrastructure projects.

Asiedu [21] states that international aid chains positively impact FDI by reducing investment risk. Castellani and Zanfei [22] state that FDI can have an impact on recipient countries beyond the effect of the increase in aggregate productivity and growth Cipolina [4] showed the impact of FDI on the economic growth of investment countries using panel data on 14 manufacturing sectors in 22 developed and developing countries from 1992–2004. Iamsiraroj [23] found that FDI was positively correlated with economic growth. Graham [24] suggested that FDI has a significant positive effect on economic growth. Vukov [25] mentioned that effect of FDI on growth is beneficial for both foreign investors and host countries with the Romania’s automotive industry.

There is also skepticism about the impact of international aid and FDI. Jansky [26] found that aid flowing into production sectors has a lower impact on FDI, using data from 180 countries from 1971 to 2007. In addition, according to Amusa [27], with international aid, the production sector was more effective in terms of FDI than social infrastructure. Ono [3] used panel data for 2003–2020 to investigate the impact of ODA in major donor countries (France, Germany, Japan, the United Kingdom, and the United States) on FDI and concluded that ODA had not affected FDI since the 2000s. Furthermore, Liao et al. [28] confirmed the significantly negative effect of international development aid on FDI based on fixed-effect regression analysis.

2.2. Absorptive Capacity and FDI

Cohen and Levinthal [29] addressed the concept of absorptive capacity, which refers to the ability to recognize the value of new information, assimilate it, and apply it to commercial ends. Applying this definition to STI ODA, it can be defined as the capacity of developing countries to achieve economic growth by recognizing, absorbing, commercializing, and accepting advanced external technologies. In their research on technology and spillover effects through FDI, Cohen, and Levinthal [30] noted that the results (effects) vary depending on absorptive capacity.

Castellacci and Natera presented human capital and infrastructure as factors to identify the factors of absorptive capacity in the national innovation system [31]. In addition, Feeny and Silva also present human capital and physical capital (infrastructure) [32]. Through this, it is shown that it can be used as a basic indicator that constitutes absorptive capacity.

Farkas [33] mentioned that the magnitude of the spillover effect, that is, the impact of FDI on economic growth, depends on the absorptive capacity of the host economy. Girma [34] used company-level data from the United Kingdom and found that FDI effectiveness in increasing total factor productivity (TFP) depends on absorptive capacity. Girma and Gorg [35] showed that FDI interacts with absorptive capacity. They showed a U-shaped relationship between this interaction term and TFP growth, suggesting that improving absorptive capacity at the company level can increase the FDI spillover effect. De Mello and Blömstrom et al. [36,37] state that the positive spillovers from the presence of FDI are likely to depend on the level of human capital. Li and Liu [38] noted that human capital is an important catalyst for FDI to have a positive effect on the growth of economies.

Simon and Feeny [32] found that international aid effectiveness varied depending on the recipient country’s absorptive capacity. Narula and Marin [39] confirmed that firms with high-absorptive capacity are able to have positive spillover generated by FDI. Feeny et al. [40] suggested that the effectiveness of aid in promoting growth and reducing poverty (from national or international donor organizations) was not compromised by excessive aid compared with the absorptive capacity level of recipient countries. Barrios and Strobl [41] found that Spain could apply positive externalities related to FDI in spin-only domestic firms with adequate absorptive capacity.

Borensztein [42] established that FDI has a positive growth effect only if the host country has minimum human capital. Samir and Mefteh [43] addressed that the economic environment plays important role in attracting FDI using the cases of transports and ICT. Kumari and Sharma [44] mentioned that market factors, such as market size and trade openness, are still dominant factors in FDI. Dang [45] suggested that developing countries must have flexible policies in each stage of the economic cycle to attach FDI effectively.

3. Research Methodology

3.1. Methodology

This study aims to verify by empirically analyzing the effect of STI ODA on FDI in recipient countries. This study sought analyze the effects of human capital, infrastructure (high-speed internet accessibility), scientific and technical journal articles on moderating effects. To this end, STI ODA data from 140 countries (see Appendix A) from 2002–2020 (OECD and World Bank) were collected and analyzed [10,46,47]. This study used panel and random effect models for empirical analysis to control for temporal and nationally differentiated characteristics. Based on the literature review, it can be assumed that STI ODA positively affects FDI in recipient countries. Therefore, the present study proposes and verifies the following hypothesis:

Hypothesis 1 (H1). 

STI ODA has a positive impact on FDI in recipient countries.

Based on the above discussion in the literature review, it can be assumed that the AC of developing countries positively moderates the impact of STI ODA. Therefore, the present study proposes the following hypotheses:

Hypothesis 2 (H2). 

Absorptive capacity has a moderating effect on STI ODA and FDI in recipient countries.

Hypothesis 2-1. 

Human capital (male/female) has a significant positive effect on FDI in recipient countries.

Hypothesis 2-2. 

Infrastructure (fixed broadband subscriptions and high-speed internet accessibility) has a significant positive effect on FDI in recipient countries.

Hypothesis 2-3. 

Scientific and technical journal articles have significant positive effects on FDI in recipient countries.

The Figure 1 is presented to facilitate the understanding of the research hypotheses. H1 indicates that the independent variable of STI-ODA affects FDI. Secondly, human capital (H2-1), infrastructure (H2-2), and scientific and technical journal articles (H2-3) can be considered sub-variables within absorptive capacity, which aim to determine whether the sub-variables within absorption capacity have a moderating effect on FDI in H1.

The following model was used in the present study based on Mitra and Abedin [45,47].

$$\begin{array}{ll}\mathrm{F}\mathrm{D}\mathrm{I}=\alpha & +\left({\beta }_{1}STIODA+{\beta }_2{HumanCapital}_{male}+{\beta }_{3}Human{Capical}_{female}+{\beta }_{4}InFra+{\beta }_{5}STPublications\right)\\ & +({\gamma }_{1}STIODA\times H{C}_m+{\gamma }_{2}STIODA\times H{C}_f+{\gamma }_{3}STIODA\times InFra\\ & +{\gamma }_{4}STIODA\times STPublications)\\ & +\left({\gamma }_{5}H{C}_m\times H{C}_f+{\gamma }_{6}H{C}_m\times InFra+{\gamma }_{7}H{C}_m\times STPublications\right)\\ & +\left({\gamma }_{8}H{C}_f\times InFra+{\gamma }_{10}H{C}_f\times STPublications\right)+\left({\gamma }_{11}InFra\times STpulications\right)\end{array}$$

3.2. Variables and Data Collection

(1)

STI ODA (independent variable)

The independent variable represents the net STI ODA expenditure received by the recipient country (in US dollars). These data were collected based on the 2002–2020 period and were provided by the OECD. The volume of ODA support by country can be confirmed through the OECD, and the OECD and World Bank established a date in 1967 to understand the flow of aid in the international community. It has a CRS purpose code based on criteria for classifying aid projects. It is necessary to identify the contents of aid among ODA and separate them into STI ODA [48]. The OECD provides data that can be classified according to the CRS codes, and only data from 2002 to 2020 are available (accessed on 5 May 2023).

However, the data provided by the OECD DAC does not organize the STI ODA field separately. Accordingly, STI ODA was reproduced and analyzed to support infrastructure, economic infrastructure, the production sector, and multiple parts by referring to the method of calculating science and technology ODA in the CSR code of Kang and Yim [14]. For example, each project relating to science and technology that occurred in each country was classified separately and collected again according to the method suggested by Kang and Yim [14].

(2)

Human capital (moderator variable)

The first moderator variable represents the percentage of tertiary education (degrees above university level) by country [48]. According to the World Bank, tertiary education refer to all formal post-secondary education, including public and private universities, colleges, technical training institutes, and vocational schools [49]. These data were compiled from data classified as male or female from 2002–2020 and were provided by the Work Bank [49]. To better understand the proportion of higher education in each country, we collected data classified into male or female categories. The data also shows the percentage of tertiary education attendance by age in each country.

(3)

Infrastructure (second moderator variable)

The second moderator variable represents data collected on the number of fixed broadband subscriptions (high-speed Internet accessibility) by country. A fixed broadband subscription provides access to the public Internet with a downstream speed of 256 kbit/s or more [50]. These data were collected from 2002–2020 and provided by the World Bank [49].

(4)

Scientific and technical journal articles (third moderator variable)

The third moderating variable represents the number of science and technology journals and articles. These data are based on the number of publications from 2002 to 2018 and were provided by the World Bank (National Science Foundation) [49]. The collected fields include physics, biology, chemistry, mathematics, clinical medicine, biomedicine, engineering, technology, and space science. Scientific and technical publications generally promote scientific development by reporting new research.

(5)

FDI (dependent variable)

The dependent variable represents the amount of FDI (in US dollars) annually among countries that received STI ODA. These data show the amount of FDI for each country from 2002 to 2020 and were provided by the World Bank [49].

(6)

Data collection

This study examined the impact of STI ODA on FDI in recipient countries. The selection of variables was based on the absorptive capacity of the national innovation system [32,33]. To this end, 140 countries, included as ODA recipients of OECD data, were selected; technical statistics and correlation analyses were conducted from 2002–2020, and panel data were constructed and analyzed by combining cross-sectional and time-series data.

4. Results

The present study analyzed whether STI ODA and absorptive capacity in recipient countries influence the amount of FDI. STATA software (version 15.0) was used for all analyses. The initial analysis method was a panel analysis model that combined cross-sectional and time-series data as fixed- and random-effect models to reflect individual effects in the empirical analysis. The fixed-effects model cannot estimate the coefficient of a constant (time-invariant) variable regardless of time. In addition, because the random-effects model considers individual effects as part of the error term, there is a possibility of convenience owing to the correlation between observable and unobservable variables [51,52]. The fixed-effects model assumes that the unobserved individual effect is correlated with the explanatory variable of the analysis model and is reflected in the constant term, whereas the random-effects model assumes that the unobserved individual effect is not correlated with the explanatory variable and is reflected in the error term [51,52]. Because the two-panel analysis models have both advantages and disadvantages, the Hausman test was conducted to select the random effects model if the statistics were less than the threshold and the fixed effects model if they were large.

(1)

Descriptive statistical analysis

The

Table 1. Descriptive statistical analysis.

Category	Minimum	Maximum	Mean	Standard Deviation	Skewness	Kurtosis
STI ODA	1.04	8.87	6.16	1.08	−0.29	0.23
FDI	4.74	11.46	9.03	0.88	−0.35	1.51
Human capital (male)	−0.48	2.10	1.27	0.50	−0.80	0.13
Human capital (female)	−0.21	2.08	1.28	0.37	−0.82	0.97
Infrastructure	1.40	8.68	5.05	1.31	−0.11	−0.36
Scientific and technical journal articles	−0.43	5.72	2.59	1.10	0.28	−0.27

Log.

below, univariate and multivariate normalities were reviewed for the related variables, and skewness and kurtosis were identified among the normality items. Unlike the symmetrical normal distribution of the mean, the positively and negatively skewed data were concentrated on the left and right sides of the mean, respectively, forming an asymmetrical distribution of the mean [53]. Kurtosis refers to the relative degree of the ratio of scores in the middle of the normal distribution curve, or the tail. Negative skewness refers to too many cases concentrated in the tail or a few cases in the middle, whereas positive kurtosis refers to a large concentration of data near the average [53]. If skewness is greater than two absolute values and kurtosis is greater than seven absolute values, there is a problem with data normality. Based on the normality review of the response data in this study, it was found that there was no problem with the univariate normality assumption of −0.80–0.28 for skewness and −0.36–1.51 for kurtosis.

(2)

Correlation analysis

Panel data regression analysis was conducted on the correlation results of the independent variable STI ODA, dependent variable recipient countries’ FDI, and regulatory variable absorptive capacity to examine this relationship in detail. In the case of the independent variable STI ODA and the dependent variable, recipient countries’ FDI, r = 0.265, showed a significantly positive correlation (p < 0.01), and there was no significant correlation between the independent variable STI ODA and the regulatory variable absorptive capacity infrastructure (fixed broadband subscriptions). The correlation between the independent variable STI ODA and the moderator variable (an absorptive capacity sub-variable) of science and technology journal articles (r = 0.25, was significant (p < 0.01). For the regulatory variable, there was a significantly positive correlation with the dependent variable, FDI (p < 0.01). The moderator variable, human capital (Male and Female) and the dependent variable, recipient country FDI (r = 0.120 [M) r = 0.146[F]), exhibited significantly positive correlations (p < 0.01), while the second moderator variable, infrastructure (fixed broadband subscriptions), and the dependent variable, recipient country FDI (r = 0.814), exhibited significantly positive correlation (p < 0.01). The third moderator variable, scientific and technical journal articles, and the dependent variable and recipient country FDI (r = 0.920), exhibited significantly positive correlation (p < 0.01). A panel data regression analysis was conducted to examine this correlation in detail. The

Table 2. STI ODA correlation analysis, absorptive capacity (sub-variables), FDI in recipient countries, and related variables.

Category	STI ODA	Human Capital (Male)	Human Capital (Female)	Infrastructure	Scientific and Technical Journal Articles	FDI
STI ODA	1
Human capital (male)	−0.027	1
Human capital (female)	0.010	0.940 **	1
Infrastructure	0.111 **	0.096 **	0.113 **	1
Scientific and technical journal articles	0.252 **	0.089 **	0.121 **	0.936 **	1
FDI	0.265 **	0.120 **	0.146 **	0.814 **	0.920 **	1

** p < 0.01.

was shown STI ODA correlation analysis, absorptive capacity (sub-variables), FDI in recipient countries.

(3)

Panel regression analysis

Looking at the likelihood ratio test results for the panel model, it is necessary to consider the individual characteristics of the panel by rejecting the null hypothesis at the 0.05 level. Therefore, it is more appropriate to use a panel model than an Ordinary Least Square (OLS model). As a result of the Hausman test, whether it is a fixed effect or a probability effect, which is an individual characteristic of the panel, chi²(2) is Prob > chi² = 0.0400 and less than p < 0.05, indicating that the random effect is desirable. Therefore, although a fixed-effect model diagram was presented, it was interpreted by focusing on the probability effect. The

Table 3. Panel data analysis of STI ODA impact on FDI in recipient countries.

Dependent Variable: FDI		Coef.	Std. Err.	t	p > t	[95% Conf. Interval]
Fixed effect	STI ODA	0.2749027	0.0238601	11.52 ***	0.000	0.2280839	0.3217214
	_cons	7.336131	0.1495448	49.06	0.000	7.042692	7.62957
sigma_e: 0.82975893 Prob > F = 0.0000, R-sq: 0.4041
Random effect	STI ODA	0.2650704	0.0233486	11.35 ***	0.000	0.2193079	0.3108329
	_cons	7.396865	0.1464272	50.52	0.000	07.109873	7.683857
sigma_e: 0.82975893, Wald chi2(1) = 0.000 R-sq: 0.1119

*** p < 0.001.

was shown the panel data analysis of STI ODA impact on FDI in recipient countries.

In the random-effects model, the independent variable STI ODA had a significantly positive effect on the dependent variable FDI (t = 11.35, p < 0.001). Therefore, it is shown that when the independent variable STI ODA is high, the dependent variable FDI increases. Additionally, it is shown that the coefficient of determination is 11%. These results show that the hypothesis adopted in the case of H1 STI ODA will positively affect FDI in recipient countries. The

Table 4. Panel data analysis on the moderating effect of human capital (male) as an absorptive capacity sub-variable of STI ODA impact on FDI in recipient countries.

Category		Coef.	Std. Err.	t	p > t	VIF	Prob > F/R-sq
Model.1	(Constant)	6.156	0.146	42.209	0.000		Prob > F = 0.0000, R-sq = 0.313
	STI ODA	0.299	0.021	14.542 ***	0.000	1.009
	Human capital (M)*	0.809	0.045	17.954 ***	0.000	1.009
Model.2	(Constant)	6.493	0.163	39.839	0.000		Prob > F = 0.0000, R-sq = 0.325
	STI ODA	0.248	0.023	10.636 ***	0.000	1.322
	Human capital (M)*	0.766	0.046	16.772 ***	0.000	1.055
	STI ODA*Human capital (M)*	7.022 × 10−11	0.000	4.478 ***	0.000	1.337

*** p < 0.001 (M)* = Male.

was shown moderating effect of human capital (male) as an absorptive capacity sub-variable of STI ODA impact on FDI in recipient countries.

Considering the moderating effect of human capital (male), an absorptive capacity sub-variable of the STI ODA impact on FDI in recipient countries, the determination coefficient was 31% because of the input of independent and regulatory variables in the model 1 panel data analysis. Considering the moderating effect in model 2, the combined effect of the independent variables STI ODA and human capital (male) had a significant positive effect (t = 4.478, p < 0.001). The determination coefficient was 32%. The

Table 5. Panel data analysis on the moderating effect of human capital (female) as a sub-variable of absorptive capacity of STI ODA impact on FDI in recipient countries.

Category		Coef.	Std. Err.	t	p > t	VIF	Prob > F/R-sq
Model 1	(Constant)	5.834	0.151	38.536	0.000		Prob > F = 0.0000, R-sq = 0.330
	STI ODA	0.285	0.020	14.054 ***	0.000	1.003
	Human capital (F)*	1.123	0.059	18.909 ***	0.000	1.003
Model 2	(Constant)	6.146	0.167	36.801	0.000		Prob > F = 0.0000, R-sq = 0.340
	STI ODA	0.239	0.023	10.549 ***	0.000	1.280
	Human capital (F)*	1.075	0.060	17.938 ***	0.000	1.038
	STI ODA*Human capital (F)*	6.444 × 10−11	0.000	4.269 ***	0.000	1.303

*** p < 0.001 (F)* = Female.

was shown moderating effect of human capital (female) as an absorptive capacity sub-variable of STI ODA impact on FDI in recipient countries.

Considering the moderating effect of human capital (female), a sub-variable in absorptive capacity on the STI ODA effect on FDI in recipient countries, the coefficient of determination was 33% as a result of the input of independent and regulatory variables in panel data analysis model 1. Considering the moderating effect in Model 2, the combined effect of the independent variables STI ODA and human capital (F) was significantly positive (t = 4.269, p < 0.001). The coefficient of determination was 34%. The

Table 6. Panel data analysis on the STI ODA moderating effects on FDI in recipient countries on the infrastructure (fixed broadband subscriptions, high-speed Internet accessibility) of absorptive capacity.

Category		Coef.	Std. Err.	t	p > t	VIF	Prob > F/R-sq
Model 1	(Constant)	0.106	0.017	6.192	0.000		Prob > F = 0.0000, R-sq = 0.560
	STI ODA	0.471	0.014	33.240 ***	0.000	1.086
	Infrastructure	6.101	0.115	53.025 ***	0.000	1.086
Model 2	(Constant)	0.095	0.017	5.546	0.000		Prob > F = 0.0000, R-sq = 0.565
	STI ODA	0.460	0.014	31.759 ***	0.000	1.116
	Infrastructure	0.000	0.000	3.596 ***	0.000	1.143
	STI ODA*Infrastructure	7.022 × 10−11	0.000	4.478 ***	0.000	1.112

*** p < 0.001.

was shown moderating effect on FDI in recipient countries on the infrastructure of absorptive capacity.

Considering the moderating effect of infrastructure (fixed broadband subscriptions- high-speed Internet accessibility) on FDI in recipient countries, the coefficient of determination was 56% because of the use of independent and regulatory variables in model 1 panel data analysis. Considering the moderating effect in Model 2, the combined effect of the independent variable STI ODA and infrastructure (fixed broadband subscription number, high-speed Internet accessibility) was significantly positive (t = 4.478, p < 0.001). The coefficient of determination was 56%. The

Table 7. Panel data analysis on the moderating effect of scientific and technical journal articles, which are sub-variables of absorptive capacity, on the STI ODA impact on FDI in recipient countries.

Category		Coef.	Std. Err.	T	p > t	VIF	Prob > F/R-sq
Model 1	(Constant)	6.929	0.102	68.142	0.000		Prob > F = 0.0000, R-sq = 0.602
	STI ODA	0.097	0.017	5.777 ***	0.000	1.087
	Scientific and technical	0.581	0.017	34.624 ***	0.000	1.087
Model 2	(Constant)	6.967	0.105	66.046	0.000		Prob > F = 0.0000, R-sq = 0.602
	STI ODA	0.092	0.017	5.421 ***	0.000	1.128
	Scientific and technical	0.576	0.017	33.353 ***	0.000	1.151
	STI ODA*Scientific and technical	3.611 × 10−15	0.000	1.334	0.183	1.133

*** p < 0.001.

was shown moderating effect of scientific and technical journal articles as an absorptive capacity sub-variable of STI ODA impact on FDI in recipient countries.

Considering the moderating effect of scientific and technical journal articles, which are absorptive capacity sub-variables on the STI ODA impact on FDI in developing countries (source countries), the coefficient of determination was 60.2% because of the input of independent and regulatory variables in the model 1 panel data analysis. Considering the moderating effect in model 2, there was no significant combined effect of STI ODA, an independent variable, and scientific and technical journal articles (t = 1.334, p > 0.05). The coefficient of determination was 60.2%, indicating that it was explained in a manner similar to that of model 1. These results show that (H2)’s absorptive capacity will have a moderating effect on STI ODA and FDI in recipient countries. It was adopted in the case of human capital (H2-1) and infrastructure (H2-2) and rejected in the case of scientific and technical journal articles (H2-3), indicating that the hypothesis was partially adopted.

5. Discussion

There are few studies and analyses on STI ODA effectiveness, and the exact statistical definition of STI ODA, including that of the OECD, is yet been established. This study sought to identify the factors that affect foreign direct investment, especially STI ODA effectiveness. We examined the moderating effect of absorptive capacity on FDI. Various factors affecting FDI were identified using variables such as human capital, tertiary education enrollment rate (male and female), infrastructure (fixed broadband subscriptions- high-speed Internet accessibility), and science and technology journal articles collected by country from the OECD and World Bank data. A panel model was used for empirical analysis, and the Hausman test confirmed that the random effect was more appropriate. The results of this study can be summarized as follows:

(1) STI ODA was shown to affect FDI in developing countries. Moreover, it was shown that STI ODA also had a positive effect on attracting FDI in recipient countries. (2) Human capital (male/female), an absorptive capacity sub-variable, was shown to have a positive influence on the STI ODA effect on FDI. Thus, it was shown that human capital has a moderating influence on the STI ODA effect on FDI. (3) It was found that the number of fixed broadband subscriptions (high-speed Internet accessibility), an absorptive capacity sub-variable, had a positive effect. Fixed broadband subscriptions (high-speed Internet access) have also been shown to moderate the impact of STI ODA on FDI. (4) Regarding the effect of STI ODA on FDI, scientific and technical journal articles (studies), which are absorptive capacity sub-variables, were shown to have no moderating effect. Previous research on absorptive capacity has mainly used firm-level data; however, research using national-level data is difficult to find. This study is meaningful because it expands from a corporate-level scope to a national-level scope and empirically analyzes the moderating effect of absorptive capacity.

6. Conclusions

This study investigated the effect of STI ODA on attracting FDI from developing countries. The direct effect of STI ODA on FDI, which has an economic effect, and the control effect of absorptive capacity were investigated using a panel random-effect model.

This study is similar to the results of previous studies; for example, Mesghena Yasin analyzed the relationship between ODA and FDI based on panel data from 11 African countries (1990–2003), and as a result, aid between countries has a statistically positive effect on FDI [54].

In addition, Phillip Harms and Matthias Lutz applied the OLS, 2SLS, and GMM models to 92 countries based on panel data up to (1988–1999), and the results show that aid is positive for FDI if regulations are removed by country [10]. Donaubauer et al. also hypothesized that aid to infrastructure (e.g., telecommunications, transportation, energy) helps attract FDI inflow (using the economic infrastructure index of 81 countries (1990–2010), and studies showed that infrastructure support has a direct impact on FDI [55]. In contrast, Ono and Sekiyama used panel data for 2003–2020 to examine the impact of ODA in major donor countries (France, Germany, Japan, the United Kingdom, and the United States) on FDI, finding no clear evidence that ODA of any effect of FDI since the 2000s. In particular, the leading effect of Japan’s ODA was found to be insignificant, as Japan’s FDI and ODA have decreased since 2000 because of the long-term fiscal deficit and economic recession caused by the collapse of Japan’s bubble economy [3,56]. Chang and Park reviewed the effect of FDI through Korean ODA in ASEAN countries on economic development from 2014 to 2020, confirming that ODA had a positive effect on economic growth, and FDI also had a moderating effect on ODA [57]. Wehncke et al. verified the causal relationship between FDI and ODA using various econometric models for 20 African countries (2000–2018) [58]. The study found that countries that promote policies for continuous economic growth create an environment favorable for ODA and FDI, which inflows ODA and attracts FDI. Furthermore, ODA can be used to measure the risk of investment in beneficiary countries [28].

For STI ODA to be more economically effective, investments in human capital and infrastructure (e.g., the Internet) must be considered. Thus, when STI ODA is accompanied by efforts to expand human capital and infrastructure (high-speed Internet accessibility) in developing countries, it is expected to play a favorable role in attracting FDI, a major driving force of economic growth. Because this study conducted a statistical analysis by country, some areas were not reflected in the specificity or obstacles that can only be found on actual field trips. In addition, because data from 2002–2020 were collected and analyzed, there is a limit to generalizing the results of this study. Therefore, it is expected that further research on actual countries and surveys in the future will be more meaningful if various approaches are studied, including factors that are difficult to observe through statistical analysis.