1 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 Research on World Agricultural Economy https://journals.nasspublishing.com/index.php/rwae Copyright © 2023 by the author(s). Published by NanYang Academy of Sciences Pte. Ltd. This is an open access article under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License. (https://creativecommons.org/licenses/by-nc/4.0/). 1. Introduction Agriculture can contribute to ending severe impoverish- ment, encourage shared wealth, and feed a projected 9.7 billion people by 2050 [1]. Progress in the agriculture sec- tor is between two to four times more useful in growing incomes among the most impoverished relative to other sectors [1]. Further, agriculture is also essential to economic progress: representing 4% of global gross domestic product DOI: http://dx.doi.org/10.36956/rwae.v4i3.861 Received: 19 May 2023; Received in revised form: 15 June 2023; Accepted: 27 June 2023; Published: 5 July 2023 Citation: Badu-Prah, C., Agyeiwaa-Afrane, A., Gidiglo, F.K., et al., 2023. Trade, Foreign Direct Investment and Agriculture in Developing Countries. Research on World Agricultural Economy. 4(3), 861. http://dx.doi. org/10.36956/rwae.v4i3.861 *Corresponding Author: Justice Gameli Djokoto, Department of Agribusiness Management, Central University, Ghana, P. O. Box DS 3210, Dansoman, Accra, Ghana; Email: dgameli2002@gmail.com RESEARCH ARTICLE Trade, Foreign Direct Investment and Agriculture in Developing Countries Charlotte Badu-Prah Akua Agyeiwaa-Afrane Ferguson K. Gidiglo Francis Y. Srofenyoh Kofi Aaron A-O. Agyei-Henaku Justice Gameli Djokoto* Department of Agribusiness Management, Central University, Ghana, P. O. Box DS 3210, Dansoman, Accra, Ghana Abstract: Agriculture continues to make significant contributions to developing countries in the presence of globalisation. Thus, international trade and foreign capital flows are important to developing countries. The authors used data on 115 developing countries from 1995 to 2020 to investigate the effect of inward and outward foreign direct investment (FDI) on trade in the agricultural sector of developing countries. Inward FDI enhanced exports, imports, and trade openness. However, outward FDI did not affect exports, imports, and trade openness. To escalate international trade in agricultural products, developing countries must continue to promote the inflow of FDI into agriculture (AIFDI). This requires paying attention to appropriate management of the macroeconomy, keeping down the inflation rate, optimising the currency exchange rate, and keeping interest rates down to boost investment among others. Whilst these would enhance AIFDI that would promote trade, these would directly promote trade. As developing countries have often suffered foreign exchange pressures, they must enhance foreign exchange receipts through increased exports. Increasing human capital can increase exports. Unlike existing studies, the authors used more current data covering many developing countries and accounted for endogeneity. Keywords: Agricultural exports; Agricultural imports; Agricultural trade openness; Capital flow; Foreign capital http://dx.doi.org/10.36956/rwae.v4i3.861 http://dx.doi.org/10.36956/rwae.v4i3.861 http://dx.doi.org/10.36956/rwae.v4i3.861 mailto:dgameli2002@gmail.com https://orcid.org/0000-0001-6144-2142 https://orcid.org/0000-0001-6348-9737 https://orcid.org/0000-0002-6669-6181 https://orcid.org/0000-0003-1922-7582 https://orcid.org/0000-0002-2159-2944 2 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 (GDP) and exceeding 25% of GDP in some developing countries. As global investment needs are in the range of $5 trillion to $7 trillion per year, the estimates for investment needs in developing countries are between $3.3 trillion and $4.5 trillion per year [2,3]. The developing countries would require foreign investments to meet this need. Foreign direct investment (FDI) is an investment made by an occupant firm in one economy to find a long-term interest in a firm that is a denizen in another economy. The long-term interest infers the presence of a lasting relation- ship between the direct investor and the direct investment firm and a significant degree of influence on the leadership of the firm. The basis of the long-term relationship is the control of 10% or more of the firm by a direct investor [4-8]. Under the directional principle, direct investment is shown as either direct investment abroad (outward, OFDI) or direct investment in the reporting economy (inward, IFDI) [4,8]. Developing countries have pursued varied poli- cies to attract FDI into their respective agricultural sector to stimulate local investment and supply of funds, in- crease export capacity, increase employment, and enhance technology transfer [2,9-13]. Regarding exports, Aihu and Chedjou [14], Harding and Javorcik [15] and Kang [16] did find that IFDI promotes exports for the total economy and the manufacturing sector. For imports, whilst Aihu and Chedjou [14] reported a positive effect of IFDI, Djokoto [17] and Latif and Younis [18] found a neutral effect. The effects of IFDI on trade openness are most inconsistent; Aihu and Chedjou [14], Karaca, Güney, and Hopoğlu [19] and Yaox- ing [20] found a positive effect, Umar, Chaudhry, Faheem, and Farooq [21] found a negative effect for lower-income and lower-middle-income countries, but the neutral effect for upper middle-income countries. Although developing countries are generally net recipients of capital flows [22-27], Sun and Zhang [28] found trade openness enhances the ef- fect of OFDI from China. Considering these inconsisten- cies, what is the effect of FDI on trade in the agriculture sector in developing countries? Existing studies on FDI and trade nexus have focused on the total economy [19,23]. Harding and Javorcik [15] and Kang [16] addressed manufacturing, only Djokoto [17] and Latif and Younis [18] studied agriculture. Whilst Djokoto [17] studied a single country, Latif and Younis [18] studied four countries with data from 1995 to 2017. Some limitations emerge especially, regarding agricultural studies. First, the dependent variable in the agriculture studies has been exports and imports and not trade openness, a more in- clusive measure of trade. Second, the number of develop- ing countries covered is limited, thus, the results of the studies cannot be generalised for developing countries. Third, although, the data used were current at the time, these are not the most current now. Fourth, the studies did not account for endogeneity. This could have led to the correlation of the error term with some of the explanatory variables thereby violating an assumption of undergirding ordinary least squares. This could cause an inaccurate ef- fect of FDI on trade. Finally, the analyses ignored the role of OFDI, the counterpart of IFDI, which also affects trade. This could result in omitted variable bias. We make up for these limitations as follows. Firstly, in addition to exports and imports, we assessed the effect of FDI on trade open- ness. Secondly, we covered 115 developing countries in Africa, Latin America and the Caribbean, Asia, and the Pa- cific. Thirdly, we used data from 1995 to 2020. In the fourth place, we took account of endogeneity in macroeconomic variables and finally, included OFDI in our model. Inward foreign direct investment enhanced exports, im- ports, and trade openness. To escalate international trade in agricultural products, developing countries must contin- ue to promote the inflow of FDI into agriculture (AIFDI). This requires paying attention to appropriate management of the macro economy; keeping down the inflation rate, optimising the currency exchange rate, and keeping inter- est rates down to boost investment among others. Whilst these would enhance AIFDI that would promote trade, these would directly promote trade. As developing coun- tries have often suffered foreign exchange pressures, they must enhance foreign exchange receipts through increased exports. Increasing human capital can increase exports. In what follows, we present the theories of trade and cross-border capital flows. We examined the pertinent literature on developing countries to assess the scope of knowledge on the title of the study, assess the differences and similarities among them and provide the information needed for the discussion section. In Section 3, the model- ling is presented with a description of the data and estima- tion procedures. The results of the estimation are reported, and these are explained considering the relevant literature in Section 4. In the final section, we conclude the study with some policy recommendations. 2. Literature Review 2.1 Theoretical Review The workhorse theory about trade and capital flows is the Hecksher-Ohlin framework [29,30]. In this framework, trade and capital flows are perfect substitutes under a two- country, two-factor model and two-commodity. This con- dition is sufficient to ensure factor price equilibrium and this equilibrium is adequate to guarantee commodity price equilibrium. Mundell [26] states, ‘….the ability to engage in commodity trade can eliminate the need for capital to 3 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 flow from the capital-abundant countries to the capital- scarce countries since the rate of return differences can be eliminated through trade alone’. In acknowledging the factor substitutability of FDI and trade, Mundell [31] noted that increasing trade restrictions enhances factor move- ments, and an increasing restriction to factors enhances trade. Notwithstanding the significant role of the Heck- sher-Ohlin-Mundell framework in explaining trade and capital flows, it is constrained in its ability to provide a complete analysis of trade and capital flows and their col- laboration under a rich set of circumstances. Specifically, capital mobility in the static two-country, two-factor, two- commodity framework is restricted to the apportionment of capital across countries, for a fixed level of world capi- tal stock [26]. Despite the Hecksher-Ohlin-Mundell position of sub- stitutability between trade and capital flows, Antras and Caballero [32] have however, shown the complementarity between trade and capital flows when relative advantages across countries are not decided only by factor endow- ments, but also by financial conglomeration. 2.2 Empirical Review These theories have informed the developing country literature on the effects of FDI on trade that addressed agriculture [17,18], manufacturing [15,16], and the total econo- my [14,19-21,28,33]. The geographies included China [28], Cote d’Ivoire [20], Ghana [17], Jordan, Morocco, Egypt, and Thai- land [18], BRICS-T [19], Africa [14], and developing countries [15,21]. Djokoto [17], Karaca et al. [19] and Yaoxing [20] employed Granger causality, Sun and Zhang [28], and Umar et al. [28] employed fixed effects, random effects, and general method of moments. Harding and Javorcik [15] applied the difference-in-difference method. Inward FDI was positively related to trade openness [14,19,20]. However, Umar et al. [21] found a negative relationship for lower income (LIC) and lower-middle-income countries (LMIC) but a neutral effect for upper-middle-income countries (UMIC). Harding and Javorcik [15] reported a positive effect of FDI presence on exports of developing countries. The effect was stronger for developing countries than for developed countries. “A weaker and quantitative- ly smaller effect for developed countries is consistent with the view that foreign presence is closing a technology gap. For a developed economy, there is less of a technology gap to close, and the foreign presence has a minor effect on the unit values of exports.” [15]. Aihu and Chedjou [14] reported positive effects of inward FDI on exports and imports in the total economies of Africa. Kang [16] found a positive effect of FDI on Korean manufactured exports to developing but a negative effect on manufactured exports to developed countries. In the only study that investigated the role of outward FDI (OFDI) on trade, Sun and Zhang [28] found a positive effect of China’s OFDI on Belt and Road countries on trade in China. The effect of population growth on trade openness was positive [19,34,28] but Osei et al. [33] found a neutral effect for LIC and LMIC. The effects of GDP growth on trade openness have been mixed. A positive effect [19,15,28,21]. Osei et al. [33] reported a positive effect for lower-income countries and a negative effect for lower-middle-income countries. Mbogela [34] matched the negative effect with evidence on African countries. Aihu and Chedjou [14] however, reported a neural effect on exports, imports, and trade openness. As in the case of GDP growth, the effect of population growth is also mixed. Whilst Osei et al. [33] did not find a significant effect of population growth on trade openness, Harding and Javorcik [15] found a negative effect on exports whilst Karaca et al. [19], Mbogela [34] and Sun and Zhang [28] found a positive effect of population growth on trade openness. Mbogela [34] measured trade policy as the freedom to trade internationally and found that the variable did not significantly influence trade openness in Africa. However, Umar et al. [21] reported a positive effect on trade openness. Whilst the effect of inflation and domestic investment had a positive effect on trade openness, the effect of human capital was mixed; negative for lower-income countries [21], and neutral for lower-middle-income countries [21]. Djokoto [17] and Latif and Younis [18] are specific agri- cultural papers on FDI-trade nexus. In the only agricul- tural FDI-trade nexus paper, Djokoto [17] investigated the effect of FDI inflow on imports and exports in Ghana. Us- ing Granger’s instantaneous causality approach with data from 1961 to 2008, FDI substituted imports whilst FDI did not have a discernible effect on exports in the short- run. In the long run, imports and FDI complemented each other. Djokoto [17] explained that MNEs would need to import some capital items and raw from abroad including from parent companies. To some extent, employees of for- eign firms would generally prefer goods from their home country that could drive up imports of finished goods. Latif and Younis [18] studied Jordan, Morocco, Egypt, and Thailand collectively using data from 1995 to 2017. Whilst FDI promoted net exports, exports and imports were not significantly affected by FDI. It would be observed that the studies that investigated the effect of FDI on trade used FDI inflow, not FDI out- flow except Sun and Zhang [28]. Although the two studies focused on agriculture, attention was given to exports and imports and not trade openness. Moreover, the analysis did not consider other variables that explain exports and 4 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 imports. We fill these gaps by investigating the effect of inward and outward FDI on exports, imports, and trade openness in agriculture in developing countries. 3. Data and Methods 3.1 Models and Data Congruent to the literature on FDI and trade [14,33-35], we specify Equations (1)-(3). 6 by investigating the effect of inward and outward FDI on exports, imports, and trade openness in agriculture in developing countries. 3. Data and Methods 3.1 Models and Data Congruent to the literature on FDI and trade [14,33-35], we specify Equations (1)-(3).  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (1)  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (2)  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (3) There are i developing countries and t years. The α, β and γ are parameters to be estimated. The ω, φ and τ are idiosyncratic error terms. Agricultural export (AEX) is the ratio of agricultural exports to agricultural value added. Agricultural import (AIM) is the ratio of agricultural imports to agricultural value added. The sum of AEX and AIM is agricultural trade openness (ATO). AEX, AIM, and ATO constitute measures of TRADE. Anderson [36], de Azevedo et al. [37], Djokoto [2,10,38,39], Kastratović [40], Narteh-Yoe, Djokoto and Pomeyie [41] and Osei, et al. [33] measured trade similarly. The inflow of FDI into agriculture in developing countries is AIFDI, measured as the ratio of FDI to agricultural value added. We measured AOFDI = 1 for observation of the outflow of FDI into agriculture and 0 otherwise. This is outward FDI out of agriculture in developing countries. The use of the dummy variable was necessitated by limited non-zero values reported for agricultural OFDI at the source. AINV is agricultural domestic investment measured as the ratio of agricultural gross fixed capital formation to agricultural value added [2,10,39,42]. We defined AGDPG as the annual growth rate of agricultural value at 2015 prices. Growth of the agricultural sector can absorb agricultural imports through the consumption of agricultural inputs and agricultural products as raw and intermediate goods for processing. Agricultural exports would be acquired from domestic agricultural production resulting from increased AGDPG. The rest of the variables are not specific to the agricultural sector. The official exchange rate EXRATE is measured as the annual average of the number of the developing country’s currency per US$ 1. A high EXRATE would raise the prices of agricultural imports and could dampen agricultural imports whilst promoting agricultural exports. Agricultural produce exporters would expect more revenue denominated in the domestic currency. Umar et al. [21] reported the effect of the exchange rate on agricultural trade. We define FTTRADE (1) 6 by investigating the effect of inward and outward FDI on exports, imports, and trade openness in agriculture in developing countries. 3. Data and Methods 3.1 Models and Data Congruent to the literature on FDI and trade [14,33-35], we specify Equations (1)-(3).  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (1)  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (2)  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (3) There are i developing countries and t years. The α, β and γ are parameters to be estimated. The ω, φ and τ are idiosyncratic error terms. Agricultural export (AEX) is the ratio of agricultural exports to agricultural value added. Agricultural import (AIM) is the ratio of agricultural imports to agricultural value added. The sum of AEX and AIM is agricultural trade openness (ATO). AEX, AIM, and ATO constitute measures of TRADE. Anderson [36], de Azevedo et al. [37], Djokoto [2,10,38,39], Kastratović [40], Narteh-Yoe, Djokoto and Pomeyie [41] and Osei, et al. [33] measured trade similarly. The inflow of FDI into agriculture in developing countries is AIFDI, measured as the ratio of FDI to agricultural value added. We measured AOFDI = 1 for observation of the outflow of FDI into agriculture and 0 otherwise. This is outward FDI out of agriculture in developing countries. The use of the dummy variable was necessitated by limited non-zero values reported for agricultural OFDI at the source. AINV is agricultural domestic investment measured as the ratio of agricultural gross fixed capital formation to agricultural value added [2,10,39,42]. We defined AGDPG as the annual growth rate of agricultural value at 2015 prices. Growth of the agricultural sector can absorb agricultural imports through the consumption of agricultural inputs and agricultural products as raw and intermediate goods for processing. Agricultural exports would be acquired from domestic agricultural production resulting from increased AGDPG. The rest of the variables are not specific to the agricultural sector. The official exchange rate EXRATE is measured as the annual average of the number of the developing country’s currency per US$ 1. A high EXRATE would raise the prices of agricultural imports and could dampen agricultural imports whilst promoting agricultural exports. Agricultural produce exporters would expect more revenue denominated in the domestic currency. Umar et al. [21] reported the effect of the exchange rate on agricultural trade. We define FTTRADE (2) 6 by investigating the effect of inward and outward FDI on exports, imports, and trade openness in agriculture in developing countries. 3. Data and Methods 3.1 Models and Data Congruent to the literature on FDI and trade [14,33-35], we specify Equations (1)-(3).  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (1)  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (2)  = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +  (3) There are i developing countries and t years. The α, β and γ are parameters to be estimated. The ω, φ and τ are idiosyncratic error terms. Agricultural export (AEX) is the ratio of agricultural exports to agricultural value added. Agricultural import (AIM) is the ratio of agricultural imports to agricultural value added. The sum of AEX and AIM is agricultural trade openness (ATO). AEX, AIM, and ATO constitute measures of TRADE. Anderson [36], de Azevedo et al. [37], Djokoto [2,10,38,39], Kastratović [40], Narteh-Yoe, Djokoto and Pomeyie [41] and Osei, et al. [33] measured trade similarly. The inflow of FDI into agriculture in developing countries is AIFDI, measured as the ratio of FDI to agricultural value added. We measured AOFDI = 1 for observation of the outflow of FDI into agriculture and 0 otherwise. This is outward FDI out of agriculture in developing countries. The use of the dummy variable was necessitated by limited non-zero values reported for agricultural OFDI at the source. AINV is agricultural domestic investment measured as the ratio of agricultural gross fixed capital formation to agricultural value added [2,10,39,42]. We defined AGDPG as the annual growth rate of agricultural value at 2015 prices. Growth of the agricultural sector can absorb agricultural imports through the consumption of agricultural inputs and agricultural products as raw and intermediate goods for processing. Agricultural exports would be acquired from domestic agricultural production resulting from increased AGDPG. The rest of the variables are not specific to the agricultural sector. The official exchange rate EXRATE is measured as the annual average of the number of the developing country’s currency per US$ 1. A high EXRATE would raise the prices of agricultural imports and could dampen agricultural imports whilst promoting agricultural exports. Agricultural produce exporters would expect more revenue denominated in the domestic currency. Umar et al. [21] reported the effect of the exchange rate on agricultural trade. We define FTTRADE (3) There are i developing countries and t years. The α, β and γ are parameters to be estimated. The ω, φ and τ are idiosyncratic error terms. Agricultural export (AEX) is the ratio of agricultural exports to agricultural value added. Agricultural import (AIM) is the ratio of agricultural im- ports to agricultural value added. The sum of AEX and AIM is agricultural trade openness (ATO). AEX, AIM, and ATO constitute measures of TRADE. Anderson [36], de Azevedo et al. [37], Djokoto [2,10,38,39], Kastratović [40], Narteh-Yoe, Djokoto and Pomeyie [41] and Osei, et al. [33] measured trade similarly. The inflow of FDI into agriculture in developing countries is AIFDI, measured as the ratio of FDI to agricultural value added. We measured AOFDI = 1 for observation of the outflow of FDI into agriculture and 0 otherwise. This is outward FDI out of agriculture in developing countries. The use of the dummy variable was necessitated by limited non-zero values reported for agri- cultural OFDI at the source. AINV is agricultural domestic investment measured as the ratio of agricultural gross fixed capital formation to agricultural value added [2,10,39,42]. We defined AGDPG as the annual growth rate of agricultural value at 2015 prices. Growth of the agricultural sector can absorb agricultural imports through the consumption of ag- ricultural inputs and agricultural products as raw and inter- mediate goods for processing. Agricultural exports would be acquired from domestic agricultural production resulting from increased AGDPG. The rest of the variables are not specific to the agricul- tural sector. The official exchange rate EXRATE is meas- ured as the annual average of the number of the develop- ing country’s currency per US$ 1. A high EXRATE would raise the prices of agricultural imports and could dampen agricultural imports whilst promoting agricultural exports. Agricultural produce exporters would expect more rev- enue denominated in the domestic currency. Umar et al. [21] reported the effect of the exchange rate on agricultural trade. We define FTTRADE as the freedom to trade inter- nationally [34]. FTTRADE is a composite measure of the absence of tariff and non-tariff barriers that affect imports and exports of goods and services. This is composed of the trade-weighted average tariff rate and non-tariff bar- riers. The weighted average tariff uses weights for each tariff based on the share of imports for each good. A low FTTRADE means a low prospect to trade than a high FT- TRADE. Whilst the former would discourage TRADE [34] the latter would enhance international trade (TRADE). HC is human capital, defined as secondary school enrol- ment percent of gross enrolment. High HC contributes to high employment in the production of goods and services that can be exported. HC can be combined with imported goods to produce for domestic and the export market. HC has a relationship with trade [21,43,44]. INFLA, inflation, is measured as the annual growth rate of the consumer price index. High INFLA reduces the value of the developing country’s currency. This could discourage imports as well as exports. However, Osei et al. [33] found that INFLA does not depress trade. POPG is the annual growth rate of the population of males and females. A high population in- creases the market for the consumption of imports as well as increased labour for production for exports. Therefore, POPG could influence TRADE [19,34]. Data for the study comprised 115 developing countries (Appendix) from 1995 to 2020. Aside from the availabil- ity of data, the period also covers increased foreign direct investment activity in developing countries. Data on AEM, AIM, AGDPG, and AINV were obtained from FAOSTAT [45] whilst World Development Indicators of the World Bank [46] was the source of EXRATE, HC, INFLA and POPG, The Heritage Foundation [47] is the source for FTTRADE. 3.2 Estimation Procedure The panel structure of the data (large cross-section than time series) necessitated the application of the estima- tion of fixed and random effects estimators. However, as macroeconomic variables could be plagued with endo- geneity, we employed the general method of moments (GMM) to take care of the possible endogeneity. We used xtdpdgmm [50] to reduce the number of instruments.a a We employed the Sargan test [51,52], to explore the overidentifying restrictions and the Arellano and Bond [48] test to test for the presence of second-order serial correlation. 5 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 4. Results and Discussion 4.1 Summary of the Data The standard deviation of ATO is about two times that of AIM and close to three times that of AEX (Table 1). This suggests a larger spread of ATO than AEX and AIM. The mean of AIFDI is lower than its standard deviation suggesting over-dispersion of AIFDI. As AOFDI was defined as a dummy variable, the mean represents the percentage of observations with AOFDI as 1. Specifically, only 8% of the 2,462 observations recorded AOFDI. This small proportion is in line with the fact that outward FDI tended to originate more from developed countries than from developing countries and is underscored in the litera- ture [22-27]. 4.2 Results We estimated Equations (1)-(3) and performed robust- ness checks on the estimates of the key coefficients (Table 2). The sign of the coefficients of AIFDI are positive and the magnitudes are similar across models 1-9. Similarly, the coefficients of AOFDI are similar in magnitude across models 1-9. These suggest the estimates of AIFDI and AOFDI are robust to the control variables. In the case of agricultural imports (AIM) as the depend- ent variable, the coefficients of AIFDI and AOFDI are similar across models 10-18 suggesting the robustness of the estimates of AIFDI and AOFDI (Table 3). For agricul- tural trade openness, ATO, as the dependent variable, the coefficients of AIFDI and AOFDI are also similar across models 19-27 suggesting the robustness of the key esti- mates (Table 4). It would be observed that the estimates of AIFDI in Table 4 are about two times the magnitude of those in Table 2 and more than those in Table 3. Also, across Tables 2-4, the coefficients of the lag of the de- pendent variable, are positive, statistically significant, and similar in magnitude. Whilst the statistical significance confirms that the endogeneity has been cared for, the simi- larity across models suggests the robustness of the esti- mates to control variables. The complete models in Tables 2-4 are assembled in Table 5. The probability of the second-order serial cor- relations tests is invalidated signifying no second-order correlation in the errors of models. The probability of the Sargan-Hansen test also shows values above 10%. This implies that the over-identifying restrictions imposed in the estimation are valid. Following these impressive model properties, the panel model estimated is appropri- ate. Whilst the estimates of the coefficients in Table 5 are similar, across the models, the estimates in model 27 ap- pear to be larger than those in models 9 and 18. This is not surprising as the dependent variable in model 27 (ATO) is the sum of the dependent variables in models 9 and 18 (AEX and AIM). The increased value of ATO resulted in higher coefficients than those in models 9 and 27. 4.3 Discussion of the Effects of Foreign Direct In- vestment on Trade The coefficient of AIFDI of 0.6882 suggests a US$ 1 rise in agricultural inward FDI will raise exports by 69 cents (Table 5). Although this is inelastic, nevertheless, it shows that FDI in the agricultural sector of develop- ing countries enhances trade. This can be attributable to multinational enterprises (MNEs) engaging in exports of their products to the parent company and other affiliates as well as non-affiliate customers outside the country. As many developing countries produce primary agricultural products, the exports to parent firms and other affiliates fit into the vertical integration of the MNEs. The export-en- hancing role of AIFDI, ceteris paribus should improve the foreign exchange receipts of developing countries. Whilst the finding is contrary to the Hecksher-Ohlin-Mundell po- sition of substitutability between trade and capital flows, it is consistent with the Antras and Caballero [32] position of complementarity of trade and capital flows. In the empiri- cal space, our results conform to that of the manufacturing sector in developing countries [16] and the total economies of Africa [14]. But Djokoto [17] and Latif and Younis [18] reported a neutral effect of AIFDI on trade in Ghanaian agriculture and the agriculture of Jordan, Morocco, Egypt, and Thailand, respectively. A US$ 1 increase in AIFDI will induce an 87 cents increase in imports. The investment codes of developing countries contain concessions on imports of raw materials Table 1. Summary statistics. Variable Observation Mean Standard deviation Minimum Maximum AEX 2,462 0.8113 5.2582 0 96.7905 AIM 2,462 1.4886 7.3732 0.0208 118.4649 ATO 2,462 2.2999 12.5249 0.0594 214.5246 AIFDI 2,462 0.0052 0.0327 –0.1076 0.8139 AOFDI 2,462 0.0804 0.2720 0 1 AINV 2,462 0.1010 0.0642 0.0089 0.4896 AGDPG 2,462 0.0299 0.0882 –0.7022 1.2342 EXRATE 2,454 1.26e+07 2.22e+08 0.0028 5.60e+09 FTTRADE 2,462 64.6789 15.0331 0 94.8000 HC 2,347 63.4170 29.9712 5.2834 212.5903 INFLA 2,460 11.2421 102.4682 –16.1173 4145.106 POPG 2,462 1.9142 1.3267 –16.8806 17.3991 6 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 Table 2. Estimations and robustness checks for the effect of foreign direct investment on exports. (1) (2) (3) (4) (5) (6) (7) (8) (9) VARIABLES AEX AEX AEX AEX AEX AEX AEX AEX AEX L.AEX 1.0833*** (0.0004) 1.0518*** (0.0233) 1.0831*** (0.0004) 1.0832*** (0.0004) 1.0826*** (0.0010) 0.3494* (0.2103) 1.0836*** (0.0007) 1.0877*** (0.0037) 0.3633* (0.1950) AIFDI 0.8699 (0.5739) 0.7516 (0.5086) 0.8467 (0.5978) 0.8647 (0.5741) 0.8596 (0.5645) 0.6825** (0.2847) 0.8699 (0.5753) 0.8488 (0.6238) 0.6882** (0.2855) AOFDI 0.0576 (0.0370) 0.0216 (0.0724) 0.0591 (0.0387) 0.0482 (0.0345) 0.0543 (0.0358) –0.0020 (0.0328) 0.0555 (0.0361) 0.0282 (0.0528) –0.0047 (0.0372) AINV 7.2822 (5.2882) 0.6663*** (0.2268) AGDPG –0.4083 (0.3494) –0.1026*** (0.0348) EXRATE –0.0000*** (0.0000) –2.41e-11 (1.65e-11) FTTRADE 0.0018 (0.0019) 0.0005 (0.0006) HC 0.0052*** (0.0016) 0.0046*** (0.0017) INFLA –0.0000 (0.0000) 1.42e-07 (5.59e-06) POPG 0.1547 (0.1527) 0.0009 (0.0075) CONSTANT –0.0274*** (0.0056) –0.6810 (0.4604) –0.0157 (0.0152) –0.0261*** (0.0055) –0.1459 (0.1195) –0.1105 (0.1099) –0.0275*** (0.0055) –0.3267 (0.2971) –0.1823 (0.1237) Model diagnostics Observations 2,347 2,347 2,347 2,340 2,347 2,239 2,346 2,347 2,235 Countries 114 114 114 114 114 113 114 114 113 1. Values in parenthesis are Windmeijer’s (2005) finite-sample correction as the default two-step standard errors are biased in finite samples due to the neglected sampling error in the weighting matrix. 2. *** p < 0.01, ** p < 0.05, * p < 0.1. 3. All models estimated with GMM (xtdpdgmm in Stata) using the collapse option to control for instrument proliferation. Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 7 Table 3. Estimations and robustness checks for the effect of foreign direct investment on imports. (10) (11) (12) (13) (14) (15) (16) (17) (18) VARIABLES AIM AIM AIM AIM AIM AIM AIM AIM AIM L.AIM 1.0212*** (0.0028) 0.9879*** (0.0247) 1.0210*** (0.0026) 1.0212*** (0.0028) 1.0211*** (0.0033) 1.0745*** (0.0103) 1.0212*** (0.0028) 1.0257*** (0.0026) 1.0736*** (0.0094) AIFDI 0.3811 (0.6453) 0.3617 (0.5270) 0.3066 (0.6224) 0.3805 (0.6450) 0.3936 (0.6776) 1.2703** (0.6404) 0.3719 (0.6480) 0.1613 (0.6152) 0.8694* (0.4866) AOFDI 0.0367 (0.0810) –0.0502 (0.1681) –0.0519 (0.1074) 0.0409 (0.0795) 0.0331 (0.1029) –0.0673 (0.0679) 0.0347 (0.0856) –0.0316 (0.0998) –0.0215 (0.0732) AINV 10.7769* (6.2746) 3.3783 (2.3380) AGDPG –1.1521*** (0.3969) –0.7748*** (0.2191) EXRATE –0.0000 (0.0000) –8.49e-12 (2.43e-11) FTTRADE 0.0004 (0.0034) 0.0044 (0.0030) HC 0.0062 (0.0042) 0.0053 (0.0050) INFLA 0.0000 (0.0001) 4.92e-05 (4.92e- 05) –0.0056 (0.0184) POPG 0.1739 (0.1859) –0.9352** (0.3918) CONSTANT 0.0248 (0.0418) –0.9496* (0.5194) 0.0638 (0.0471) 0.0246 (0.0414) 0.0017 (0.2254) –0.4263* (0.2590) 0.0252 (0.0427) –0.3057 (0.3289) –8.49e-12 (2.43e-11) Model diagnostics Observations 2,347 2,347 2,347 2,340 2,347 2,239 2,346 2,347 2,235 Countries 114 114 114 114 114 113 114 114 113 1. Values in parenthesis are Windmeijer’s (2005) finite-sample correction as the default two-step standard errors are biased in finite samples due to the neglected sampling error in the weighting matrix. 2. *** p < 0.01, ** p < 0.05, * p < 0.1. 3. All models estimated with GMM (xtdpdgmm in Stata) using the collapse option to control for instrument proliferation. Table 4. Estimations and robustness checks for the effect of foreign direct investment on trade openness. (19) (20) (21) (22) (23) (24) (24) (26) (27) VARIABLES ATO ATO ATO ATO ATO ATO ATO ATO ATO L.ATO 1.0482*** (0.0011) 1.0174*** (0.0188) 1.0480*** (0.0010) 1.0482*** (0.0011) 1.0479*** (0.0019) 1.1107*** (0.0181) 1.0483*** (0.0011) 1.0528*** (0.0033) 1.1116*** (0.0181) AIFDI 1.5770 (1.0206) 1.3055 (1.0604) 1.4877 (1.0466) 1.5777 (1.0235) 1.5557 (1.0434) 2.5702** (1.0243) 1.5339 (1.0205) 1.4261 (0.9688) 2.0284** (0.9029) AOFDI –0.0052 (0.1194) –0.2510 (0.4114) –0.0952 (0.1177) –0.0093 (0.1223) –0.0237 (0.1323) –0.1339 (0.1423) –0.0219 (0.1235) –0.0905 (0.1701) –0.0792 (0.1202) AINV 16.7347* (9.1374) 4.4784 (2.8000) AGDPG –1.5769** (0.7245) –0.9604*** (0.2828) EXRATE –0.0000*** (0.0000) –8.04e-11** (4.00e-11) FTTRADE 0.0018 (0.0062) 0.0064* (0.0036) HC 0.0068 (0.0046) 0.0058 (0.0049) INFLA 0.0000 (0.0000) 3.35e-05 (2.58e-05) POPG 0.3380 (0.3292) 0.0169 (0.0267) CONSTANT –0.0078 (0.0355) –1.5124* (0.7926) 0.0532 (0.0604) –0.0067 (0.0361) –0.1196 (0.3792) –0.5035* (0.2719) –0.0079 (0.0350) –0.6356 (0.6162) –1.2819*** (0.4048) Model diagnostics Observations 2,347 2,347 2,347 2,340 2,347 2,239 2,346 2,347 2,235 Countries 114 114 114 114 114 113 114 114 113 1. Values in parenthesis are Windmeijer’s (2005) finite-sample correction as the default two-step standard errors are biased in finite samples due to the neglected sampling error in the weighting matrix. 2. *** p < 0.01, ** p < 0.05, * p < 0.1. 3. All models estimated with GMM (xtdpdgmm in Stata) using the collapse option to control for instrument proliferation. 8 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 9 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 and other resources. The abuse of the system can cause an increase in imports of agricultural resources. Also, the expatriates as well as the growing middle class of developing countries’ populations tend to develop a taste for foreign foods. These also contribute to an increase in agricultural exports. Our finding is consistent with the theoretical position of Antras and Caballero [32] of comple- mentarity of trade and capital flows but contrary to those of the Hecksher-Ohlin-Mundell position. Our finding is also inconsistent with the findings of Djokoto [17] and Latif and Younis [18] on the agricultural sector of Ghana, Jordan, Morocco, Egypt, and Thailand, respectively. These re- ported negative and neutral effects, respectively. The effect of AIFDI on trade openness is also positive. The elastic magnitude of 2.0284 implies that a US$ 1 in- crease in AIFDI would induce a more than US$ 1 increase in trade openness. The estimate turns out to be the highest among the statistically significant estimates. The elasticity can be attributable to the complementarity of AIFDI and imports and exports on one hand and the synergistic effect of imports and exports on the other. Theoretically, foreign direct investment and trade are related [29-31]. Whilst Heck- sher [29], and Mundell [30,31], posited a substitution effect, Antras and Caballero [32] noted a complementary effect in line with our findings. Our findings are consistent with empirical evidence from developing countries [14,15,19,20]. Whilst Umar et al. [21] found a negative effect for lower- income countries and lower-middle-income countries, a neutral effect was reported for upper-middle-income countries. This result points not only to the presence of capital flows and trade in developing countries but also to a significant trade-enhancing role of AIFDI in developing country agriculture. As foreign capital and trade are prox- ies of globalisation [54-56], these pointers are evidence of the globalisation of agriculture in developing countries. The coefficients of AOFDI on exports, imports and trade openness are negative and statistically insignificant. Recalling that the AOFDI was measured as a dummy, the negative sign suggests fewer observations of AOFDI than non-observation of AOFDI. This is not surprising as de- veloping countries are not generally the source of foreign capital, rather they are recipients [22-27]. Dunning [57] and Dun- ning and Narula [58] theorised that developing countries are in stages I and II of development in which the inflow of FDI outstrips the outflow of FDI. Although empirical evidence shows some developing countries have moved to stage III [59-64], many developing countries are still far from becoming a net exporter of capital. Our finding is incon- sistent with the theory of substitution [29,26,30] and comple- mentarity [32] between capital flows and trade. Our findings also departed from the statistically significant positive ef- fect of OFDI, and trade found by Sun and Zhang [28]. 4.4 Discussion of Control Variables The coefficients of AINV are positive but statistically significant for exports. Thus, a US$ 1 increase in AINV Table 5. Complete models for exports, imports, and trade openness. (9) (18) (27) VARIABLES AEX AIM ATO L.AEX 0.3633* (0.1950) L.AIM 1.0736*** (0.0094) L.ATO 1.1116*** (0.0181) AIFDI 0.6882** (0.2855) 0.8694* (0.4866) 2.0284** (0.9029) AOFDI –0.0047 (0.0372) –0.0215 (0.0732) –0.0792 (0.1202) AINV 0.6663*** (0.2268) 3.3783 (2.3380) 4.4784 (2.8000) AGDPG –0.1026*** (0.0348) –0.7748*** (0.2191) –0.9604*** (0.2828) EXRATE –2.41e-11 (1.65e-11) –8.49e-12 (2.43e-11) –8.04e-11** (4.00e-11) FTTRADE 0.0005 (0.0006) 0.0044 (0.0030) 0.0064* (0.0036) HC 0.0046*** (0.0017) 0.0053 (0.0050) 0.0058 (0.0049) INFLA 1.42e-07 (5.59e-06) 4.92e-05 (4.92e-05) 3.35e-05 (2.58e-05) POPG 0.0009 (0.0075) –0.0056 (0.0184) 0.0169 (0.0267) CONSTANT –0.1823 (0.1237) –0.9352** (0.3918) –1.2819*** (0.4048) Model diagnostics Observations 2,235 2,235 2,235 Countries 113 113 113 Probability of 2nd order serials 0.7288 0.4430 0.9352 Probability of the Sargan- Hansen test 0.3024 0.1282 0.1002 1. Values in parenthesis are Windmeijer’s (2005) finite-sample correction as the default two-step standard errors are biased in finite samples due to the neglected sampling error in the weighting matrix. 2. *** p < 0.01, ** p < 0.05, * p < 0.1. 3. All models estimated with GMM (xtdpdgmm in Stata) using the collapse option to control for instrument proliferation. 10 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 will induce less than a US$ 1 (66 cents) increase in ex- ports. Exports originate from the supply which also arises from production. AINV, therefore, contributes to agricul- tural exports. This is consistent with the findings of Osei et al. [33] and Tahir et al. [43]. The coefficients of AGDPG are negative and statisti- cally significant for exports, imports, and trade openness. It was expected that an increase in output would contrib- ute to the production, supply, and export of commodities, hence a positive effect. However, this did not turn out to be the case. Regarding imports, the negative sign sug- gests a substitution effect of agricultural growth and imports. An increase in agricultural output would lead to increased provision of agricultural goods that would otherwise have been imported. Whilst this is consistent with Mbogela [34] for African countries and Osei et al. [33] for lower-middle-income countries, others have reported a positive effect [15,19,21,28,43]. The negative coefficient of EXRATE suggests increas- ing currency value per US$ would discourage trade openness. Although increasing EXRATE would provide increased local currency sales revenue from exports, the cost of production for export would go up and ultimately discourage exports. For imports, it is a truism that in- creased EXRATE means imports become more expensive, discouraging imports. The combination of these explains the negative relationship between EXRATE and ATO, albeit a minuscule value. Umar et al. [21] found a positive sign for EXRATE for lower-middle-income countries but a neutral effect for lower-middle-income and upper-middle- income countries. The coefficient for FTTRADE is positive for all three models in Table 5 but weakly significant for model 27. Thus, freedom to trade internationally enhances trade openness. This result is expected because the freedom to trade reduces the constraints to trade, thus, encouraging trade. The neutral effect of FTTRADE found by Mbogela [34] disagrees with our findings. The coefficient of HC is positive for exports, imports, and trade openness. However, the magnitude is statistical- ly indistinguishable from zero for the export model. HC contributes to labour. Recalling that the marginal produc- tivity of labour is positive, HC would enhance production, the source of export supplies. Tahir et al. [43] and Umar et al. [21] also found a human capital-enhancing role in trade, albeit for trade openness. The positive finding of Umar et al. [21] was about upper-middle-income countries. For lower-income countries, however, Umar et al. [21] re- ported a neutral effect. Aihu and Chedjou [14] reported a neutral effect of HC for all the trade measures. The coefficients of INFLA and POPG are statistically indistinguishable from zero regarding exports, imports, and trade openness. The results for inflation are contrary to the negative effects reported by Osei et al. [33]. Our results for the population are also consistent with those of Osei et al. [33] for trade openness. Whilst Harding and Javorcik [15] reported a negative effect on exports, Karaca et al. [19], Mbogela [34], and Sun and Zhang [28] found posi- tive effects of population on trade openness. It must be noted that some results are inconsistent with the previous literature, such as the effect of inflation or population. This may be because all countries are considered for the analysis at the same time, and no differentiation is made at all. Consequently, the effect of certain variables on the data in specific types of countries remains obscured. 5. Conclusions and Recommendations Following gaps in the trade and capital flow literature regarding agriculture, we estimated the effect of FDI on exports and imports and trade openness, using 115 de- veloping countries from 1995 to 2020 taking account of endogeneity in macroeconomic variables. Whilst AIFDI has a positive effect on AEX and AIM, the effect of the latter is higher than that of the former. The larger effect of the latter over the former would impose foreign ex- change pressure on developing countries. The estimate of the coefficient of AIFDI on trade openness turns out to be the highest among the statistically significant estimates. Freedom to trade internationally enhanced trade openness. Agricultural output growth and exchange rate did not en- hance trade, however, measured. Human capital enhanced exports. AOFDI, INFLA and POPG had no effect on trade however measured. To escalate international trade in ag- ricultural products, developing countries must continue to promote AIFDI. This requires paying attention to ap- propriate management of the macro economy; keeping down the inflation rate, optimising the currency exchange rate, and keeping interest rates down to boost investment among others. Whilst these would enhance AIFDI that would promote trade, these would directly promote trade. As developing countries have often suffered foreign ex- change pressures, they must enhance foreign exchange receipts through increased exports. Increasing human cap- ital can increase exports. This would provide the needed labour for production and increase supplies that lead to increased exports. Developing countries must continue to support measures that promote freedom to trade. As many developing countries have acceded to the World Trade Organisation agreement, it provides a regimen that will compel developing countries to follow policies that make for more free trade among members. A limitation of this study lies in the absence of partial 11 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 analysis by country groups, which would have provided a better understanding of the phenomenon under study. This study is also limited to developing countries that are net recipients of AIFDI. Further research can consider devel- oped countries and transition economies. Author Contributions Charlotte Badu-Prah: Contributed data and analysis tools; Wrote the paper; Reviewed the paper. Akua A. Afrane-Arthur: Contributed data and analysis tools; Wrote the paper; Reviewed the paper. Ferguson K. Gidiglo: Contributed data and analysis tools; Wrote the paper, Re- viewed the paper. Francis Y. Srofenyoh: Contributed data and analysis tools; Wrote the paper, Reviewed the paper. Kofi Aaron A-O. Agyei-Henaku: Contributed materials, and analysis tools; Wrote the paper, Reviewed the paper. Justice G. Djokoto: Conceived and designed the experi- ments, Analysed and interpreted the data Wrote the paper; Reviewed the paper. Funding This research received no external funding. Data Availability Data used in the study were extracted from publicly available international sources. Conflict of Interest All authors disclosed no conflict of interest. References [1] Agriculture and Food [Internet]. World Bank; 2023 [cited 2023 Apr 7]. Available from: https://www. worldbank.org/en/topic/agriculture/overview [2] Djokoto, J.G., 2021. Level of development, foreign direct investment and domestic investment in food manufacturing. F1000Research. 10, 72. [3] World Investment Report 2014 [Internet]. United Nations; 2014. Available from: https://unctad.org/ system/files/official-document/wir2014_en.pdf [4] Balance of Payments and International Investment Position Manual, Sixth Edition [Internet]. IMF; 2009 [cited 2023 Apr 26]. Available from: https://www. imf.org/external/pubs/ft/bop/2007/pdf/bpm6.pdf [5] OECD, 1999. OECD benchmark definition of foreign direct investment, 3rd Ed. OECD Publishing: Paris. [6] Punthakey, J., 2020. Foreign direct investment and trade in agro-food global value chains. OECD Food, Agriculture and Fisheries Papers, No. 142. OECD Publishing: Paris. [7] World Investment Report 2009—Transnational Cor- porations, Agricultural Production and Development [Internet]. UNCTAD; 2009. Available from: https:// reliefweb.int/report/world/world-investment-re- port-2009-transnational-corporations-agricultural-pro- duction-and?gclid=EAIaIQobChMI8_7P2OTk_wIVo- AutBh2ZaABDEAAYASAAEgLhWvD_BwE [8] UNCTAD 2023. Foreign direct investment: Inward and outward flows and stock, annual. [cited 2023 Apr 26]. Available from: https://unctadstat.unctad.org/ wds/TableViewer/tableView.aspx?ReportId=96740 [9] De Mello Jr, L.R., 1997. Foreign direct investment in developing countries and growth: A selective survey. The Journal of Development Studies. 34(1), 1-34. [10] Djokoto, J.G., 2021. Foreign direct investment into agriculture: Does it crowd-out domestic investment? Agrekon. 60(2), 176-191. [11] Farla, K., De Crombrugghe, D., Verspagen, B., 2016. Institutions, foreign direct investment, and domestic investment: Crowding out or crowding in? World Development. 88, 1-9. [12] Gallova, Z., 2011. Foreign direct investment in selected countries of Balkan, does it crowd out or crowd in domestic investment? Czech Financial and Accounting Journal. 6(4), 68-78. [13] Oualy, J.M.R., 2019. Do Foreign Direct Investments (FDI) Crowd in or Crowd Out Domestic Investment in Cote d’Ivoire? [Internet]. Available from: https:// ssrn.com/abstract=3505572 [14] Aihu, W., Chedjou, A.B., 2018. Effect of Chinese outward foreign direct investment on international trade of Africa. African Journal of Business Manage- ment. 12(7), 188-211. [15] Harding, T., Javorcik, B.S., 2012. Foreign direct in- vestment and export upgrading. Review of Econom- ics and Statistics. 94(4), 964-980. [16] Kang, K., 2012. Is the relationship between foreign direct investment and trade different across developed and developing countries? Evidence from Korea. Asian-Pacific Economic Literature. 26(2), 144-154. [17] Djokoto, J.G., 2012. Does causal relationships exist between external trade and foreign direct investment flow to agriculture in Ghana? International Journal of Business and Management. 7(2), 179. [18] Latif, B.F., Younis, A.A.A., 2020. The role of for- eign direct investment in external agricultural trade of chosen developing countries to the period 1995- 2017. Tikrit Journal of Administration and Econom- ics Sciences. 16(49), 252-273. [19] Karaca, T., Güney, N. & Hopoğlu, S. 2022. Ticari https://www.worldbank.org/en/topic/agriculture/overview https://www.worldbank.org/en/topic/agriculture/overview https://unctad.org/system/files/official-document/wir2014_en.pdf https://unctad.org/system/files/official-document/wir2014_en.pdf https://www.imf.org/external/pubs/ft/bop/2007/pdf/bpm6.pdf https://www.imf.org/external/pubs/ft/bop/2007/pdf/bpm6.pdf https://reliefweb.int/report/world/world-investment-report-2009-transnational-corporations-agricultural-production-and?gclid=EAIaIQobChMI8_7P2OTk_wIVoAutBh2ZaABDEAAYASAAEgLhWvD_BwE https://reliefweb.int/report/world/world-investment-report-2009-transnational-corporations-agricultural-production-and?gclid=EAIaIQobChMI8_7P2OTk_wIVoAutBh2ZaABDEAAYASAAEgLhWvD_BwE https://reliefweb.int/report/world/world-investment-report-2009-transnational-corporations-agricultural-production-and?gclid=EAIaIQobChMI8_7P2OTk_wIVoAutBh2ZaABDEAAYASAAEgLhWvD_BwE https://reliefweb.int/report/world/world-investment-report-2009-transnational-corporations-agricultural-production-and?gclid=EAIaIQobChMI8_7P2OTk_wIVoAutBh2ZaABDEAAYASAAEgLhWvD_BwE https://reliefweb.int/report/world/world-investment-report-2009-transnational-corporations-agricultural-production-and?gclid=EAIaIQobChMI8_7P2OTk_wIVoAutBh2ZaABDEAAYASAAEgLhWvD_BwE https://unctadstat.unctad.org/wds/TableViewer/tableView.aspx?ReportId=96740 https://unctadstat.unctad.org/wds/TableViewer/tableView.aspx?ReportId=96740 https://ssrn.com/abstract=3505572 https://ssrn.com/abstract=3505572 12 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 Dışa Açıklık-Doğrudan Yabancı Yatırımlar-Ekonomik Büyüme İlişkisi: BRICS-T Ülkeleri Örneği (Turkish) [The relationship between trade openness-foreign direct investment and economic growth: The case of BRICS-T countries]. Bingöl Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi. 6(1), 109-146. DOI: https://doi.org/10.33399/biibfad.937591 [20] Yaoxing, Y., 2010. The relationship between foreign direct investment, trade openness and growth in Cote d’Ivoire. International Journal of Business and Man- agement. 5(7), 99-107. [21] Umar, M., Chaudhry, I.S., Faheem, M., et al., 2021. Do governance, foreign direct investment and human capital matter to bolster trade liberalization? Fresh insight from developing countries. Review of Eco- nomics and Development Studies. 7(3), 325-341. [22] Dunning, J.H., 1977. Trade, location of economic activity and MNE: A search for an eclectic approach. The international allocation of economic activity. Macmillan: London. pp. 395-418. [23] Dunning, J.H., 1988. The eclectic paradigm of inter- national production: A restatement and some possible extensions. Journal of International Business Studies. 19(1), 1-31. [24] Dunning, J.H., 2001. The eclectic (OLI) paradigm of international production: Past, present and future. International Journal of the Economics of Business. 8(2), 173-190. [25] Hymer, S.H., 1976. The international operation of national firms, a study of direct foreign investment. MIT Press: Cambridge, MA. [26] Mundell, R.A., 2010. International Trade and Inter- national Capital Flows: A Theoretical Perspective [Internet] [Accessed April 23rd, 2023]. Available at: https://personal.lse.ac.uk/jink/pdf/globalization_ chapter.pdf [27] Vernon, R., 1966. International investment and inter- national trade in the product cycle. Quarterly Journal of Economics. 80(2), 190-207. [28] Sun, K., Zhang, X., 2022. Study on the trade effects of China’s outward foreign direct investment—data from the countries along the Belt and Road. Eco- nomic and Business Management. CRC Press: Boca Raton. pp. 49-54. [29] Heckscher, E., 1919. The effect of foreign trade on the distribution of income. Ekonomisk Tidskriff. 21, 497-512. [30] Ohlin, B., 1933. Interregional and international trade. Harvard University Press: Cambridge, MA. [31] Mundell, R.A., 1957. International trade and factor mobility. The American Economic Review. 47(3), 321-335. [32] Antras, P., Caballero, R.J., 2009. Trade and capital flows: A financial frictions perspective. Journal of Political Economy. 117(4), 701-744. [33] Osei, D.B., Sare, Y.A., Ibrahim, M., 2019. On the de- terminants of trade openness in low-and lower-mid- dle-income countries in Africa: How important is economic growth? Future Business Journal. 5(1), 2. [34] Mbogela, C.S., 2019. An empirical study on the determinants of trade openness in the African econ- omies. Advances in Management and Applied Eco- nomics. 9(3), 9-42. [35] Nga, N.T.V., 2020. Analysis of the determinants of trade balance: A case study of Vietnam. Journal of Applied Finance and Banking. 10(3), 21-35. [36] Anderson, K., 2022. Agriculture in a more uncertain global trade environment. Agricultural Economics. 53(4), 563-579. [37] de Azevedo, D.S., Arruda, E.F., de Brito, A.C., et al., 2023. Dynamic elasticities of the trade balance of Brazilian agriculture in relation to the exchange rate. The International Trade Journal. 1-18. [38] Djokoto, J.G., 2013. Openness and agricultural per- formance in Ghana. Journal of Science and Technol- ogy (Ghana). 33(2), 24-36. [39] Djokoto, J.G., 2021. Drivers of agricultural foreign divestment. Studies in Agricultural Economics. 123(1), 43-51. [40] Kastratović, R., 2023. The impact of foreign direct investment on agricultural exports: The evidence from developing countries. The Journal of Interna- tional Trade & Economic Development. 1-18. [41] Narteh-Yoe, S.B., Djokoto, J.G., Pomeyie, P., 2022. Aid, domestic and foreign direct investment in small states. Economic Research-Ekonomska Istraživanja. 1-18. [42] Bekoe, B.B., Chen, J., Tougem, T.O., et al., 2023. Does domestic investment matter? A multivari- ate time series analysis of the energy-CO2 emis- sion-growth nexus in Ghana. Environmental Science and Pollution Research. 30(17), 49536-49550. [43] Tahir, M., Hasnu, S.A.F., Ruiz Estrada, M., 2018. Macroeconomic determinants of trade openness: Empirical investigation of SAARC region. Journal of Asia Business Studies. 12(2), 151-161. [44] Tsaurai, K., 2021. Determinants of trade openness in transitional economies: Does the complementarity between foreign direct investment and human capital development matter? International Journal of Eco- nomics & Business Administration (IJEBA). 9(1), 318-330. https://personal.lse.ac.uk/jink/pdf/globalization_chapter.pdf https://personal.lse.ac.uk/jink/pdf/globalization_chapter.pdf 13 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 [45] FAOSTAT, Data [Internet]. Food and Agricultural Organisation of the United Nations; 2023 [cited 2023 Mar 31]. Available from: https://www.fao.org/faostat/ en/#data [46] World Development Indicators [Internet]. World Bank; 2023 [cited 2023 Jan 30]. Available from: https://databank.worldbank.org/source/world-devel- opment-indicators [47] 2023 Index of Economic Freedom. Explore the data [Internet]. The Heritage Foundation; 2023 [cited 2023 Apr 12]. Available from: https://www.heritage. org/index/explore?view=by-region-country-year [48] Arellano, M., Bond, S., 1991. Some tests of specifi- cation for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies. 58(2), 277-297. [49] Roodman, D., 2009. How to do xtabond2: An intro- duction to difference and system GMM in Stata. The Stata Journal. 9(1), 86-136. [50] Kripfganz, S. (editor), 2019. Generalized method of moments estimation of linear dynamic panel data models. Proceedings of the 2019 London Stata Con- ference; 2019 Sep 5-6; London. [51] Sargan, J.D., 1958. The estimation of economic rela- tionships using instrumental variables. Econometrica: Journal of the Econometric Society. 26(3), 393-415. [52] Sargan, J.D., 1988. Testing for misspecification after estimating using instrumental variables. Contributions to Econometrics: John Denis Sargan. 1, 213-235. [53] Rosewarne, S. 2019. The structural transformation of Australian agriculture: Globalisation, corporatisation and the devalorisation of labour. Journal of Austra- lian Political Economy. 84, 175-218. [54] Fujie, S., Bevan, P., Frankowska, A., 2019. The im- pact of globalisation on agriculture in a coffee-pro- ducing rural community in Southern Ethiopia since 2011. International Journal of Ethiopian Studies. 13(1), 91-106. [55] Nugroho, A.D., Lakner, Z., 2022. Impact of economic globalisation on agriculture in developing countries: A review. Agricultural Economics. 68(5), 180-188. [56] Scoppola, M., 2021. Globalisation in agriculture and food: The role of multinational enterprises. European Review of Agricultural Economics. 48(4), 741-784. [57] Dunning, J.H., 1981. Explaining the international direct investment position of countries: towards a dynamic or developmental approach. Weltwirtschaft- liches Archiv. 117, 30-64. [58] Dunning, J.H., Narula, R., 1996. The investment de- velopment path revisited. Foreign Direct Investment and Governments: Catalysts for Economic Restruc- turing. 1-41. [59] Djokoto, J.G., 2021. The investment development path theory and small states. Research in Globaliza- tion. 3, 100048. [60] Djokoto, J.G., Pomeyie, P., 2021. Level of in- come and the investment development path the- ory: Evidence from Africa. SAGE Open. 11(4), 21582440211061334. [61] Dunning, J.H., 1986. The investment development cycle revisited. Weltwirtschaftliches Archiv. 122, 667-676. [62] Duran, J.J., Ubeda, F., 2005. The investment devel- opment path of newly developed countries. Interna- tional Journal of the Economics of Business. 12(1), 123-137. [63] Frenken, J.H., Mbuvi, D., 2017. Country risk, FDI flows and convergence trends in the context of the Investment Development Path. UNU-MERIT Work- ing Papers. 5, 1-17. [64] Satoglu, E.B., 2017. Emerging through foreign in- vestment: Investment development path estimation of MINT economies. Advances in Economics and Business. 5(5), 256-264. Appendix 1. List of acronyms. AEX Agricultura exports AGDPG Agricultura GDP growth AIFDI Agricultural inward foreign direct investment AIM Agricultural imports AINV Agricultural investments ATO Agricultural trade openness EXRATE Exchange rate FDI Foreign Direct Investment FTTRADE Absence of tariff and non-tariff barriers on trade GDP Gross Domestic Product HC Human capital IFDI Inward foreign direct investment INFLA Inflation LIC Lower income LMIC Lower-middle-income countries OFDI Outward foreign direct investment POPG Population growth rate TRADE Trade UMIC Upper-middle-income countries https://www.fao.org/faostat/en/#data https://www.fao.org/faostat/en/#data https://databank.worldbank.org/source/world-development-indicators https://databank.worldbank.org/source/world-development-indicators https://www.heritage.org/index/explore?view=by-region-country-year https://www.heritage.org/index/explore?view=by-region-country-year 14 Research on World Agricultural Economy | Volume 04 | Issue 03 | September 2023 25 Appendix 2. List of developing countries in the data. Afghanistan Comoros India Morocco Singapore Algeria Congo Indonesia Mozambique Solomon Islands Angola Congo, DR Iran Namibia South Africa Bahamas Costa Rica Iraq Nepal Sri Lanka Bahrain Côte d’Ivoire Israel Nicaragua Suriname Bangladesh Djibouti Jamaica Niger Syria Barbados Dominica Jordan Nigeria Tanzania Belize Dominican Rep. Kenya Oman Thailand Benin Ecuador Kiribati Pakistan Timor-Leste Bolivia Egypt Kuwait Panama Togo Botswana El Salvador Laos Papua New Guinea Tonga Brazil Equatorial Guinea Lesotho Paraguay Trinidad and Tobago Brunei Darussalam Eswatini Liberia Peru Tunisia Burkina Faso Ethiopia Libya Philippines Türkiye Burundi Fiji Madagascar Republic of Korea Uganda Cabo Verde Gabon Malawi Rwanda UAE Cambodia Gambia Malaysia Saint Lucia Uruguay Cameroon Ghana Maldives Saint Vincent and the Grenadines Vanuatu Central African Republic Guatemala Mali Sao Tome and Principe Venezuela Chad Guinea Mauritania Saudi Arabia Viet Nam Chile Guinea-Bissau Mauritius Senegal Yemen China, mainland Guyana Mexico Seychelles Zimbabwe Colombia Honduras Mongolia Sierra Leone Appendix 3. Data analyses strategy. General method of moments 2 step 1 step Instrument proliferation Accepted model Appendix 3. Data analyses strategy. Appendix 2. List of developing countries in the data. Afghanistan Comoros India Morocco Singapore Algeria Congo Indonesia Mozambique Solomon Islands Angola Congo, DR Iran Namibia South Africa Bahamas Costa Rica Iraq Nepal Sri Lanka Bahrain Côte d’Ivoire Israel Nicaragua Suriname Bangladesh Djibouti Jamaica Niger Syria Barbados Dominica Jordan Nigeria Tanzania Belize Dominican Rep. Kenya Oman Thailand Benin Ecuador Kiribati Pakistan Timor-Leste Bolivia Egypt Kuwait Panama Togo Botswana El Salvador Laos Papua New Guinea Tonga Brazil Equatorial Guinea Lesotho Paraguay Trinidad and Tobago Brunei Darussalam Eswatini Liberia Peru Tunisia Burkina Faso Ethiopia Libya Philippines Türkiye Burundi Fiji Madagascar Republic of Korea Uganda Cabo Verde Gabon Malawi Rwanda UAE Cambodia Gambia Malaysia Saint Lucia Uruguay Cameroon Ghana Maldives Saint Vincent and the Grenadines Vanuatu Central African Republic Guatemala Mali Sao Tome and Principe Venezuela Chad Guinea Mauritania Saudi Arabia Viet Nam Chile Guinea-Bissau Mauritius Senegal Yemen China, mainland Guyana Mexico Seychelles Zimbabwe Colombia Honduras Mongolia Sierra Leone