Sage Journals: Discover world-class research

Abstract

This study explores how the institution of the host countries affects the natural resource-seeking and strategic asset-seeking motives of the Chinese outward foreign direct investment. Using the system GMM estimation method and data of the Chinese outward foreign direct investment during the period from 2003 to 2020, the study shows that the search for strategic assets by the Chinese outward foreign direct investment is directed toward the host countries with well-established institutions, while the search for natural resources is driven by poor institutions of the host countries. Findings from this study provide empirical evidence for policymakers to formulate appropriate foreign investment policies given the dynamics of Chinese foreign direct investment worldwide.

Keywords

China FDI economic institution reform motive natural resources strategic assets

Introduction

The remarkable increase in the amount of the Chinese outward foreign direct investment (OFDI), which is defined as cross-border direct investment from China, has been observed worldwide (UNCTAD, 2018). In recent years, the Chinese OFDI has been shifting from developing countries to developed countries (Deng, 2009; Quer et al., 2012; Ramamurti & Hillemann, 2018; Ramasamy et al., 2012). A great number of studies have provided insights into this direction and the determinants of these changes (Buckley et al., 2018; Cui et al., 2014; Ramamurti & Hillemann, 2018). Nonetheless, there is no consensus in the literature about motives of the Chinese OFDI (Buckley et al., 2007, 2018; Deng, 2009).

A seminal paper of Buckley et al. (2007) explores these factors by using a special theory nested within the general theory for the Chinese OFDI based on three arguments about capital market imperfections, the ownership advantages of Chinese multinational enterprises (MNEs) with firm-specific advantages, and the institutions. The capital market imperfection supports the Chinese MNEs to invest abroad. Both Chinese state-owned enterprises (SOEs) and private firms have networks with the government officials (Wang, Hong, Kafouros, & Boateng, 2012; Wang, Hong, Kafouros, & Wright, 2012). These networks plus the networks of Chinese diaspora in the host countries help the Chinese MNEs to search for natural resources and other strategic assets. These networks also serve as a buffer against political risks in the host countries. Buckley et al. (2007) confirm that Chinese OFDI prefers the host countries which have high political risks. Kolstad and Wiig (2012) also argue that Chinese OFDI is attracted to the host countries that are rich in natural resources and have poor institutions.

Since 2003, when there was a change in the OFDI policy of China, private firms have been allowed to invest abroad. The Chinese MNEs have been increasingly seeking firm-specific advantages by acquiring global brands, human capital, managerial know-how and advanced technology. This motive is coined by the strategic asset-seeking internationalization (Christofi et al., 2022; Deng, 2009; Liang et al., 2022). Chinese MNEs have more rapid internationalization than their counterparts in developed countries (Cui et al., 2014; Peng, 2012; Vecchi & Brennan, 2022) and have a dominant Merger and Acquisition (M&A) mode of entry to acquire these assets (Meyer et al., 2014; Vecchi & Brennan, 2022).

Nonetheless, Buckley et al. (2018) confirm that the only two motives of the Chinese OFDI are market-seeking and resource-seeking. Chinese OFDI is found to be positively associated with host market characteristics and the endowments of natural resources of the host countries. Buckley et al. (2018) do not find any evidence to support the argument that a motive of Chinese OFDI is to seek strategic assets. This disagreement in the literature is highlighted by Buckley et al. (2018) that little is known about the determinants of the pattern of internationalization by Chinese MNEs. As a result, there is a research gap in understanding the determinants of Chinese OFDI.

Our study attempts to shed light on these contradictory arguments on Chinese OFDI in the literature. Utilizing the data on Chinese OFDI during the period from 2003 to 2020, this study re-examines the motives of Chinese OFDI. To correct for possible bias found in the previous studies of Buckley et al. (2007), Cheung and Qian (2009), Cheng and Ma (2010), Kolstad and Wiig (2012), we used the inverse hyperbolic sine (HIS) transformation method to calculate the Chinese OFDI values (Bahar & Rapoport, 2018; Bellemare & Wichman, 2020; McKenzie, 2017; Pence, 2006) and applied the dynamic generalized method of moments (GMM) for estimation.

Our study has several contributions to the literature. Firstly, the market-seeking motive for Chinese OFDI is confirmed. Secondly, we find that the host countries with rich natural resources and poor institutions are attracted to Chinese OFDI. Thirdly, we show that Chinese OFDI is seeking strategic assets in the host countries with well-established institutions, which are measured by the number of resident and non-resident patent applications. Our results are robust regardless of different proxies of institutions and various estimation methods. These findings close the gap in the literature by uncovering the effects of the host country institution on the direction of the Chinese OFDI. Thirdly, we show that the Chinese OFDI is searching for natural resources in the host countries with poor institutions.

The remainder of this paper is structured as follows: Section 2 provides stylized facts about the Chinese OFDI; Section 3 presents the theoretical background and hypothesis development; Section 4 describes methodology and data; Section 5 presents the empirical results; and Section 6 concludes the paper.

Chinese Outward Foreign Direct Investment: Stylized Facts

The initiation of the “going global” strategy in 2000 and China’s becoming a member of the World Trade Organization (WTO) in 2001 have stimulated the increase in Chinese OFDI more than ever before (Luo et al., 2010). As a result, China’s investment abroad has been growing, despite an overall decline in FDI worldwide after the global financial crisis (Liang et al., 2022; Salidjanova, 2011). As China is growing rapidly, the country faces shortages in all types of resources. Therefore, the Chinese MNEs have been investing increasingly more in a variety of sectors in host countries to seek a wide spectrum of resources ranging from natural resources to strategic assets (Liang et al., 2022; Salidjanova, 2011).

The data on Chinese OFDI, which are consistent with the international standards, are collected from 2003 to 2020. The previous studies on Chinese OFDI also used the data collected from 2003 (Cheung & Qian, 2009; Hao et al., 2020; Kolstad & Wiig, 2012; Liang et al., 2022; Yang, Wang et al., 2018). These data show that the total amount of Chinese OFDI was ranked as the third most all over the world, which was only behind the U.S. and Japan. By the end of 2020, approximately 28 thousand Chinese firms had invested in 189 countries and had established over 45 thousand overseas firms around the world, which are owned or have more than 10% of voting rights by Chinese firms (MOFCOM, 2021).

Table 1 presents the largest 15 countries that have hosted Chinese OFDI from 2003 to 2020. The total Chinese OFDI during the period from 2008 to 2020 was 23.3 times more than that during the pre-global financial crisis period from 2003 to 2007. Apart from tax havens such as Cayman Islands, Hong Kong, and British Virgin Islands, the countries belonging to the OECD (the Organization for Economic Co-operation and Development) have hosted a considerable proportion of the total of the Chinese OFDI. This observation is different from the trend during the period from 2003 to 2007, when the Chinese OFDI had been hosted by both OECD and non-OECD countries (Kolstad & Wiig, 2012; Liang et al., 2022). Figure 1 presents the shift of the Chinese OFDI among the main host regions in the world, including Europe, North America, ASEAN, and Africa. The Chinese OFDI to Africa has increased dramatically during the period before the global financial crisis in 2008, although it has gone down quickly afterward. By contrast, the Chinese OFDI to Europe, North America, and ASEAN remained stable before the crisis. However, since the crisis, it has increased substantially.

Table 1.

Top 15 Host Countries of Chinese OFDI (in Current Million USD).

Host country	2003	2005	2007	2009	2011	2013	2015	2017	2019	2020	Total	Share (%)
Total (all countries)	2,854.7	12,261.2	26,506.1	56,529.0	74,654.0	107,843.7	145,667.2	158,288.3	136,907.6	153,710.3	1,573,182.5	100.0
Hong Kong, China	1,149.0	3,419.7	13,732.4	35,600.6	35,654.8	62,823.8	89,789.8	91,152.8	90,550.1	89,145.9	922,931.1	58.7
Cayman Islands	806.6	5,162.8	2,601.6	5,366.3	4,936.5	9,253.4	10,213.0	−6606.0	−4356.7	8,562.2	74,093.8	4.7
British Virgin Islands	209.7	1,226.1	1,876.1	1,612.1	6,208.3	3,221.6	1,849.0	19,301.2	8,682.6	6,975.6	86,557.9	5.5
United States	65.1	231.8	195.7	908.7	1,811.4	3,873.4	8,028.7	6,425.5	3,806.7	6,018.7	69,556.3	4.4
Singapore	−3.2	20.3	397.7	1,414.3	3,269.0	2,032.7	10,452.5	6,319.9	4,825.7	5,923.4	51,417.2	3.3
Australia	30.4	193.1	531.6	2,436.4	3,165.3	3,458.0	3,401.3	4,242.0	2,086.7	1,198.6	36,944.6	2.3
Netherlands	4.5	3.8	106.8	101.5	167.9	238.4	13,462.8	−223.1	3,893.2	4,938.3	26,538.2	1.7
United Kingdom	2.1	24.8	566.5	192.2	1,419.7	1,419.6	1,848.2	2,066.3	1,103.5	922.2	16,757.2	1.1
Russian Federation	30.6	203.3	477.6	348.2	715.8	1,022.3	2,960.9	1,548.4	−379.2	570.2	12,427.0	0.8
Canada	−7.3	32.4	1,032.6	613.1	554.1	1,008.7	1,562.8	320.8	472.9	210.0	13,123.3	0.8
Germany	25.1	128.7	238.7	179.2	512.4	910.8	409.6	2,715.9	1,459.0	1,375.6	14,741.8	0.9
Indonesia	26.8	11.8	99.1	226.1	592.2	1,563.4	1,450.6	1,682.3	2,223.1	2,198.4	16,526.8	1.1
Switzerland	0.0	0.6	1.2	21.0	17.2	128.3	246.8	7,514.2	678.3	1,074.6	6,609.1	0.4
France	0.5	6.1	9.6	45.2	3,482.3	260.4	327.9	952.2	87.2	147.8	7,376.5	0.5
South Africa	8.9	47.5	454.4	41.6	−14.2	−89.2	233.2	317.4	338.9	400.4	7,728.9	0.5

Source. Authors’ compilation.

Figure 1.

Chinese OFDI by regions, excluding Cayman Islands and British Virgin Islands (%).

While it is clear that the Chinese firms search for natural resources by investing in African countries (Rjoub et al., 2017; Williams et al., 2022), they do not aim at obtaining natural resources in Europe and North America. With the framework provided by Buckley et al. (2018), it is likely that, after the global financial crisis, the Chinese OFDI has been shifting to the advanced countries for either market or strategic asset seeking objectives.

Theory and Hypothesis Development

Theoretical Background

The general theory of FDI suggests that there are four main motives of FDI including the natural resource-seeking, the market-seeking, the efficiency-seeking or cost reduction, and the strategic asset-seeking (Buckley et al., 2007, 2018; Dunning, 1998; Dunning & Lundan, 2008). These motives could be realized through both mergers and acquisitions (M&A) and greenfield investment (Dunning, 1998; Dunning & Lundan, 2008).

According to Buckley et al. (2007), as FDI firms look for alternatives for domestically scarce natural resources, they invest abroad with this being their first motive. Through FDI, firms could acquire natural resources at lower prices than those in their home country (Dunning & Lundan, 2008; Hao et al., 2020; Ramasamy et al., 2012). While the natural resource-seeking strategy may enhance a firm’s supply reliability, thus securing its present competitive position, it may neither increase the added value of the products, nor boost the firm’s capability (Cui et al., 2014). The natural resource-seeking strategy is often conducted by the MNEs from the emerging markets. This internationalization process is also used to overcome the latecomer disadvantages that have faced the MNEs (Luo & Tung, 2007; Vecchi & Brennan, 2022).

For the second motive, FDI firms search for a larger market to utilize resources and exploit economies of scale and scope efficiently (Loncan, 2021; UNCTAD, 1998). It is also widely agreed that larger markets present more opportunities for generating profits than smaller markets do (Lim, 1983). Buckley et al. (2007) confirm that the host country characteristics, which is measured by GDP, had a positive effect on Chinese OFDI during the period from 1984 to 2001. During this period, market-seeking was a key motive for the Chinese OFDI. In a recent study, Buckley et al. (2018) reconfirm the market-seeking motive of the Chinese OFDI.

Regarding the third motive, FDI firms may search for efficiency by choosing cheaper locations for their operations. For example, FDI firms may move to countries where the labor cost is low. Nonetheless, Buckley et al. (2007, 2018) argue that this efficiency-seeking strategy was not relevant for the Chinese OFDI before the turn of the millenium as the competitiveness of Chinese firms was still based on low labor costs from the domestic market.

The FDI firms, particularly from developing countries, may follow the fourth motive to search for strategic assets abroad to close the gap with the world leaders and increase their competitiveness (Cui et al., 2014; Liang et al., 2022). Strategic asset seekers normally acquire technological assets and human competences from firms in host countries to improve their global competitiveness. However, the empirical findings about the adoption of this strategy among the Chinese MNEs are mixed.

Hypothesis Development

One strand of studies confirms that this strategy has been carried out extensively by the MNEs from the emerging economies such China and India (Dunning & Lundan, 2008). For example, the need to survive increasing competition has forced Chinese firms to invest abroad through M&A to gain the firm specific assets to improve their competitive advantages in both their home and global markets (Deng, 2009; Liang et al., 2022; Luo et al., 2010; Rui & Yip, 2008). Other studies also support the argument that the Chinese OFDI’s strategic asset-seeking strategy has been carried out during their internationalization process in developed countries (Deng, 2009; Luo et al., 2010; Vecchi & Brennan, 2022). Furthermore, Meyer et al. (2014) argue that foreign subsidiaries of both listed state-owned and privately owned Chinese MNEs favor M&A over greenfield entries to acquire strategic assets.

Another strand of studies disagrees with the findings mentioned above. Buckley et al. (2007) was the first to use patents to proxy for the proprietary ownership advantage endowments to test the strategic asset-seeking motive of the Chinese OFDI. Nevertheless, this motive has not been found among the Chinese MNEs during the period from 1984 to 2001. Similarly, the study by Ramasamy et al. (2012) shows that Chinese OFDI is not attracted by countries with core research, which is proxied by the number of registered patents. In some estimations, they even find a significantly negative effect of the number of registered patents on the amount of Chinese OFDI.

The four motives of OFDI may be affected by institutional environments, particularly country-specific and government-created advantages, which are generated from government policies on the internationalization of the MNEs, both at home and at the host countries (Kolstad & Wiig, 2012; Ramamurti & Hillemann, 2018). It is argued that a high-quality institution is a crucial factor of FDI activity, because it lowers both the risk and the cost of doing business. A great number of empirical studies have also confirmed this relationship (Bénassy-Quéré et al., 2007; Christofi et al., 2022; Globerman & Shapiro, 2002; Liang et al., 2022). These findings are robust while controlling for the possible reverse causality between institutions and FDI flows (Bénassy-Quéré et al., 2007).

Similarly, a large number of previous studies have attempted to analyze the role of host institutions in the relationship between these motives on the Chinese OFDI (Yang, Wang et al., 2018). Findings from Cui and Jiang (2012) are in line with the literature in that the effects of host institutions on Chinese state-owned MNEs’ preference to establish joint ventures are higher to help them search for these motives.

Along with this line of research, a strand of recent studies provides empirical evidence to support the argument that Chinese MNEs prefer investing in host countries with high risks. This observation is explained as being due to the monopoly and dominance of SOEs among the Chinese MNEs and the imperfections in the capital market in China (Buckley et al., 2007; Meyer et al., 2014; Ramasamy et al., 2012). Buckley et al. (2007) was among the first studies to find that Chinese OFDI flows to host countries with higher political risks. The study by Kolstad and Wiig (2012) suggests that institutions and natural resources should particularly be analyzed together instead of in isolation. They also find that institutions have an effect on the relationship between Chinese OFDI and natural resources in developing countries. Added to that, Ramasamy et al. (2012) find no significant evidence to support the hypothesis that the publicly listed Chinese firms invest in host countries with good institutions in the search for strategic assets.

Nonetheless, another strand of research shows that it is still an open question as to whether Chinese OFDI prefers investing in the host countries with high political risks (Yang, Wang et al., 2018). Ramasamy et al. (2012) argue that only Chinese state-owned MNEs favor natural resources and political risks in the host countries. They could not find any evidence to support the argument that all Chinese MNEs are attracted to host countries with high political risks and rich natural resources. Quer et al. (2012) further confirm that the Chinese MNEs listed in the Fortune Global 500 during the period from 2005 to 2009 are not attracted by high political risks in host countries. Chinese OFDI is found to have different strategies depending on development levels of the host countries, for example between OECD and non-OECD host countries. During the period after the global financial crisis, while global FDI has declined, China’s investment abroad has increased dramatically (Salidjanova, 2011; Sattar et al., 2022), particularly in economies with well-established institutions in Europe and North America (Figure 1). Large foreign exchange reserves and China’s surplus of savings are indicated as crucial factors that promote Chinese OFDI while searching for strategic assets in these markets (Christofi et al., 2022; Morck et al., 2008; Salidjanova, 2011). Based on these arguments, we formulate the following testable hypotheses.

Hypothesis 1: Well-established institutions negatively moderate the effect of natural resource endowment of host countries on Chinese OFDI.

Hypothesis 2: Well-established institutions positively moderate the effect of strategic assets of host countries on Chinese OFDI.

Methodology and Data

Methodology

Previous studies on the determinants of Chinese OFDI used conventional estimation models such as Fixed Effects, Random Effects, pooled OLS or Poisson models to tackle the partially unobserved heterogeneity in individual effects such as country fixed effects (Buckley et al., 2007; Cheung & Qian, 2009; Kolstad & Wiig, 2012; Ramasamy et al., 2012; Yang, Wang et al., 2018). These estimation methods have a problem in using panel data with “small T, large N.” In our data, T is 18 years from 2003 to 2020. N is the number of host countries, which is more than 150. This small T is due to the use of lagged Chinese OFDI, which may be correlated with the fixed effects in the error term and autocorrelated within individuals leading to the Nickell bias (Nickell, 1981). In addition, the previous studies contain endogenous regressors, which are proxied by different variables at the macro level such as those which explain market-seeking OFDI, for example GDP per capita, trade openness, GDP growth rate, inflation, those which explain resource-seeking OFDI, for example natural resources, patents, and finally, those which reflect institutional environments, for example, the rule of law. In a dynamic panel data setting with “small T, large N,” a dynamic GMM estimation method should be used. This estimation method could handle modeling concerns such as unobserved heterogeneity, endogeneity of regressors, and heteroskedasticity and autocorrelation within individual units’ errors, which avoid the dynamic panel bias or the Nickell bias (Nickell, 1981). We thus use the dynamic GMM method to estimate equation (1) below. In fact, there are some approaches to estimate panel data models in time series settings. For example, Latif et al. (2018) use the fully modified least squares (FMOLS) technique to estimate the dynamic relationship between ICT, foreign direct investment (FDI), economic growth incorporating trade and globalization for BRICS economies over the period of 2000 to 2014. This technique only accounts for heterogeneity and cross-sectional dependence in panel data while the dynamic GMM can deal with the heterogeneity and endogeneity problems simultaneously.

Following Buckley et al. (2007), Cheng and Ma (2010), and Kolstad and Wiig (2012), to test the advanced Hypotheses H1 and H2, we apply the following estimation equation:

\begin{matrix} COFD I_{it} = ρ COFD I_{it - 1} + β_{1} NaturalRe s_{it - 1} + β_{2} Patent s_{it - 1} \\ + β_{3} Institution s_{it - 1} + β_{4} NaturalRe s_{it - 1} * Institution s_{it - 1} \\ + β_{5} Patent s_{it - 1} * Institution s_{it - 1} + Contro l_{it - 1} θ \\ + β_{6} EthicChines e_{it} + β_{7} Dis t_{it} + c_{i} + ε_{it} \end{matrix}

(1)

where $COFD I_{it}$ is the dependent variable for Chinese OFDI flows to host country i in year t during the period from 2003 to 2020. The values of Chinese OFDI used for regressions are calculated by the inverse hyperbolic sine (HIS) transformation of the values of Chinese OFDI in millions of current US dollars. The $COFD I_{it - 1}$ variable is the 1-year lag of the dependent variable, which is included to allow for the persistence of Chinese OFDI.

A number of empirical studies investigated the determinants of Chinese OFDI using the approved investment on FDI (Buckley et al., 2007; Cheung & Qian, 2009) and OFDI flows (Cheng & Ma, 2010; Kolstad & Wiig, 2012) as the dependent variable. The approved investment does not include reinvested earnings, which may lead to severe underestimations of Chinese OFDI (Kolstad & Wiig, 2012). The use of OFDI flows has two possible problems. First, when the OFDI flows are used without data transformation, the results may be biased due to heteroscedasticity and outliers in the values of OFDI flows. Second, when the logarithm of OFDI flows is taken, the zero-valued and negative-valued observations become undefined, leading to missing values. These missing values increase the biases in the estimation. Normally, we can add one to the variable prior to taking the logarithm if the number of zero values is small to avoid the reduction in the number of observations. Nevertheless, this is impossible if the original variable includes both zero values and negative values in FDI flows. One way to solve this problem is to apply the inverse hyperbolic sine (HIS) transformation of data (Bahar & Rapoport, 2018; Chen, 2013; Clemens & Tiongson, 2017; McKenzie, 2017; Pence, 2006). The HIS transformation of data is similar to taking logarithms while retaining zero-valued and negative-valued data (Bellemare & Wichman, 2020; Burbidge et al., 1988; MacKinnon & Magee, 1990; Pence, 2006). The HIS transformation can be performed with a random variable, x, such that $x_{HIS} = \ln (x + \sqrt{x^{2} + 1})$ . Bellemare and Wichman (2020) provide a practical guide to interpreting the meaning of the estimated coefficients. They note that the HIS transformation can generally be treated as a natural logarithm. In our dataset, the number of negative values of the Chinese OFDI accounts for approximately 10% of the total number of values. Hence, we apply the HIS transformation for the data in this study.

The main independent variables in equation (1) include the specific asset-seeking variable ( $Patent s_{it - 1}$ ), the natural resource-seeking variable ( $NaturalRe s_{it - 1}$ ), and the institution variable ( $Institution s_{it - 1}$ ) of the host country i at time t−1. Following Buckley et al. (2007) and Meyer et al. (2014), the specific asset-seeking variable, $Patent s_{it - 1},$ is measured by the HIS transformation of the number of patent applications by residents and nonresidents of host country i at time t−1. Following Buckley et al. (2007) and Kolstad and Wiig (2012), the $NaturalRe s_{it - 1}$ variable, which represents the natural resource-seeking motive, is proxied by the ratio of fuels, ores, and metals exports to the total GDP of the host countries (Kolstad & Wiig, 2012). Following Kolstad (2009) and Mehlum et al. (2006), the $Institution s_{it - 1}$ variable is measured by the rule of law. Kolstad (2009) and Mehlum et al. (2006) highlight that the rent-seeking problems in resource rich countries can be analyzed by the rule of law. The host countries with weak rules of law and rich natural resources are appealing to Chinese investors. We will also use the remaining indicators of Worldwide governance indicators (WGI) of the World Bank for the robustness checks.

We add the $Contro l_{it - 1}$ variables to examine the determinants of Chinese OFDI as suggested in the literature. These variables include GDP per capita (GDPPerCapita), annual GDP growth (GDPGrowth), openness to trade (Trade) and inflation rates (Inflation) of host country i in year t−1. The two former variables can be proxied for the characteristics of the host country. We are interested in exploring whether the Chinese OFDI follow the market-seeking motive. In the literature, FDI will flow to the host countries with high economic growth rates because foreign investors have more opportunities to make a profit in these host countries (Buckley et al., 2007; Khan et al., 2020; Kong et al., 2020; Latif et al., 2018; Vu et al., 2022; Zhang & Daly, 2011). The Trade variable is trade openness measured by the ratio of merchandise trade (which is the sum of exports and imports) to the total GDP of the host country. This variable shows the economic integration of the host country into the global market (Chou et al., 2011; Kolstad & Wiig, 2012). The Inflation of the host country variable is included to control for the uncertainty in macroeconomic conditions, which may discourage market seekers (Buckley et al., 2007). Following Buckley et al. (2007), we control for cultural proximity by including the EthnicChinese dummy variable, which is equal to one if the percentage of the ethnic Chinese in the total population of the host country is more than 1%, and zero otherwise. Following Buckley et al. (2007), the Dist variable is included in equation (1). It is measured by the HIS transformation of distance between Beijing and the capital of the host country. This variable is proxied for investment costs of the Chinese OFDI. We also include year fixed effects in equation (1) to control the effects of any business cycle. Finally, $c_{i}$ is firm fixed effects and $ε_{it}$ is the error term.

There are two approaches in estimating the GMM, which are the difference GMM (Arellano & Bond, 1991; Holtz-Eakin et al., 1988) and the system GMM (Arellano & Bover, 1995; Blundell & Bond, 1998) for dynamic panel data. The difference GMM uses the first difference transformation to eliminate the country fixed effects and the instrumental variables, which are internal instruments based on their lags for the differenced equation. The system GMM simultaneously estimates the differenced equation and the level set equation. Regarding unbalanced panel data, the first-difference transformation is weak due to the gaps in this setting. Therefore, Arellano and Bover (1995) suggest that both estimators should apply the orthogonal deviations instead of the first difference.

Our study applies the orthogonal deviations for the estimation. To examine whether the regressors are strictly exogenous or endogenous, the estimation should be difference GMM or system GMM. To determine which instrumental variables should be selected (Bond, 2002; Roodman, 2009b), we estimate our regressions using the xtabond2 command of Roodman (2009b) in Stata, which will be elaborated more in the results section.

Table 2 presents basic statistics of all variables. Table 3 shows the variance inflation factor (VIF) to check the multicollinearity and correlation coefficients of variables in our model. All VIF values are less than five, suggesting that our models have no multicollinearity.

Table 2.

Descriptive Statistics of Variables.

	Descriptive	Source	Mean	SD	Min	Max
COFDI	Chinese OFDI	MOFCOM	3.28	3.26	−9.75	12.34
NaturalRes (%)	Ratio of fuels, ores, and metals exports to the total GDP	WDI	9.30	13.65	0.00	76.17
Patents	The number of resident and nonresident patent applications of host countries	WDI	5.34	4.06	0.00	14.03
Institutions	Rule of Law of host country	WGI	0.09	0.98	−1.89	2.13
NaturalRes*Institutions	Interaction between NaturalRes and Institutions		−1.07	14.69	−92.82	64.11
Patents*Institutions	Interaction between Patents and Institutions		2.21	7.32	−14.53	23.11
GDPPerCapita	GDP per Capita of host country	WDI	9.41	1.535	5.476	12.42
Trade (%)	Trade openness is measured by merchandise trade as share of GDP of host country	WDI	67.77	45.55	11.71	419.96
GDPGrowth (%)	GDP growth of host country	WDI	3.26	4.78	−54.01	43.48
Inflation (%)	Inflation rate of host country	WDI	6.03	17.51	−30.20	558.56
EthnicChinese (dummy)	1 if percentage of ethnic Chinese in total population is more than 1%	(Poston & Wong, 2016)	0.16	0.37	0.00	1.00
Dist	Distance between Beijing in China and capital of host country	CEPII	9.65	0.55	7.56	10.56
Observations			2,151

Note. COFDI, GDPPerCapita, Patents and Dist are taken the HIS transformation, which are computed as $x_HIS = \ln (x + \sqrt{x^{2} + 1})$ .

Table 3.

The Variance Inflation Factor (VIF) and Correlation Matrix.

		VIF	1	2	3	4	5	6	7	8	9
1	COFDI		1.00
2	NaturalRes	1.42	0.02	1.00
3	Patents	1.49	0.31	−0.11	1.00
4	Institutions	3.75	0.15	−0.14	0.43	1.00
5	GDPPerCapita	4.11	0.18	0.12	0.46	0.80	1.00
6	Trade	1.30	0.14	0.21	0.08	0.24	0.22	1.00
7	GDPGrowth	1.21	0.01	0.09	−0.11	−0.17	−0.22	0.03	1.00
8	Inflation	1.10	−0.03	0.05	−0.07	−0.19	−0.16	−0.05	−0.01	1.00
9	EthnicChinese	1.19	0.29	−0.01	0.20	0.16	0.13	0.29	0.05	−0.05	1.00
10	Dist	1.28	−0.21	−0.08	−0.29	−0.11	−0.09	−0.30	−0.08	0.05	−0.29

Data

The data for this study are extracted from various sources as presented in Table 2. The values of Chinese OFDI are taken from the annual Statistical Bulletins of China’s outward FDI, which were compiled by MOFCOM (2021) over 150 countries from 2003 to 2020. It is noted that the data before 2003 were not available.

Other data were collected from a variety of sources. Data on the number of patent applications by residents and nonresidents of the host countries (Patents) are collected from World Development Indicators (WDI) of the World Bank. We attempt to try some other ways of measuring this figure, and the results will be discussed in the robustness checks section. Data to calculate the ratio of fuels, ores, and metals exports to the total GDP of the host countries (NaturalRes) are taken from the WDI database of the World Bank. The rule of law indicator (Institutions) is extracted from the Worldwide Governance Indicators (WGI) of the World Bank. The WGI is used in many studies to assess the quality of governance over 200 countries, including six sub-indicators as the control of corruption, government effectiveness, political stability and the absence of violence, regulatory quality, rule of law, and voice and accountability. The range of index runs from −2.5 to 2.5, with the higher index indicating that a country has better institutions. GDP per capita (GDPPerCapita), annual GDP growth (GDPGrowth), trade openness (Trade) and inflation rates (Inflation) of the host countries are also taken from the WDI database of the World Bank. The percentage of ethnic Chinese (EthnicChinese) in the total population is adopted from Poston and Wong (2016). The HIS transformation of distance (Dist) between Beijing and the capital city of the host country is taken from the Centre d’Études Prospectives et d’Informations Internationales (CEPII).

Regression Results

Following the suggestions of Bond (2002) and Roodman (2009b) in the dynamic GMM estimator, we shall examine whether the regressors are strictly exogenous or endogenous. For example, time-invariant variables, which are considered as being strictly exogenous, include ethnic Chinese people in the host country and distance from Beijing to the capital city of the host country. Other time-variant variables need to be tested for strict exogeneity. Su et al. (2016) develop a practical test for strict exogeneity in panel data, which is consistent with the suggestion of Wooldridge. Wintoki et al. (2012) also use the suggestion of Wooldridge to test for strict exogeneity of the governance variable with current firm performance. Wooldridge (2010) suggests a simple test for strict exogeneity by using the Fixed Effects estimator in the following regression.

COFD I_{it} = γ X_{it} + ω W_{it + 1} + α_{0 i} + ϵ_{it}

(2)

where $X_{it}$ includes patents, natural resources, institutions, and other control variables, which are GDP per capita, trade, GDP growth, inflation, and year dummy variables. These variables affect Chinese OFDI ( $COFD I_{it}$ ). is a different subset of $X_{it}$ at the first-order lead time. Under the null hypothesis of strict exogeneity, $ω$ = 0 and is uncorrelated with current Chinese OFDI. The results of the strict exogeneity test are two subsets of $W_{it + 1}$ . The first test in equation (2) is with patents, natural resources, and institutions present evidence of strict exogeneity, where F-test (3, 164) = 0.73 and p-value = .537. The second test in equation (2) is with GDP per capita, trade, GDP growth, and inflation provides evidence of endogeneity, where the F-test (4, 167) = 4.50 and p-value = .002.

In terms of selecting the difference or the system GMM, Bond (2002) and Roodman (2009b) suggest to obtain the coefficient $ρ$ of the $COFD I_{it - 1}$ variable from the pooled OLS model, which is upward biased, and from the Fixed Effects (FE) model, which is downward biased. The coefficients on lagged Chinese OFDI ( $ρ$ ) of equation (1), from using the pooled OLS and the FE models, fell from 0.333 (OLS) to 0.096 (FE). The $ρ$ coefficient from equation (1), which is obtained from the difference or system GMM estimator, should be within this interval. If the value of the difference GMM is close to or below the coefficient of the FE estimator, it may be downward biased due to weak instruments. We estimate equation (1) by using the one-step difference and two-step difference GMM. The coefficients on lagged Chinese OFDI are 0.091 and 0.071, respectively. The results are available upon request from the authors. Therefore, the system GMM estimator should be preferred. Our main results report both a one-step system GMM with robust standard errors and a two-step system GMM with robust standard errors, as indicated by Windmeijer (2005).

The selection of instrumental variables in the dynamic GMM model is crucial in determining the joint validity of the instruments by the Hansen test. Arellano and Bond (1991) develop an autocorrelation test to be applied to the residual in differences to determine whether some lags are valid or invalid instruments. This test examines the autocorrelation in the idiosyncratic disturbance term. If the Arellano–Bond test for serial correlation in the first-differenced disturbance term is negative at the first order and there is no serial correlation at the second order, the linear GMM regression on panel data is appropriate and lags can be used as instruments (Roodman, 2009b). Moreover, the dynamic GMM model normally generates too many instruments, which weakens the Hansen test’s joint validity of the full set of instruments or instrument proliferation (Roodman, 2009a). Roodman (2009a, 2009b) note that three main techniques to limit the number of instruments are: to apply specific lags of variables instead of all available lags for instruments; to apply the “collapse” option to limit the smaller subset of instruments; to use a principal components analysis (PCA) to reduce the number of instruments by combining the correlation instruments subsets into common factors with eigenvalues of at least one (Bai & Ng, 2010; Mehrhoff, 2009). We report the first technique as suggested by Roodman (2009a) by selecting the specific lags of the instruments. We also provide the regression results of the “collapse” and the PCA techniques to show a robustness check.

The results in Table 4 are obtained from estimating equation (1) by using the system GMM estimator. Columns from (1) to (3) report the one-step system GMM estimation with robust standard errors. Columns from (4) to (6) present the two-step system GMM estimation with corrected standard errors’Windmeijer (2005). In Columns (1) and (4), the interaction term between natural resources and institutions is added. In Columns (2) and (5), the interaction term between patents and institutions is included. Both interaction terms are included in Columns (3) and (6).

Table 4.

Determinants of Chinese OFDI Results Using the Dynamic GMM Estimator with One-step System and Two-step System.

	(1)	(2)	(3)	(4)	(5)	(6)
	OneSYS	OneSYS	OneSYS	TwoSYS	TwoSYS	TwoSYS
COFDI (lag)	0.188***	0.187***	0.185***	0.197***	0.191***	0.188***
	(0.048)	(0.049)	(0.049)	(0.052)	(0.054)	(0.052)
Patents (lag)	0.185***	0.185***	0.182***	0.184***	0.196***	0.187***
	(0.039)	(0.040)	(0.041)	(0.045)	(0.042)	(0.047)
NaturalRes (lag)	−0.002	0.000	−0.002	−0.004	0.002	−0.003
	(0.012)	(0.012)	(0.012)	(0.011)	(0.011)	(0.011)
Institutions (lag)	0.615	−0.263	0.077	0.620	−0.333	0.019
	(0.454)	(0.479)	(0.547)	(0.448)	(0.510)	(0.583)
NaturalRes*Institutions (lag)	−0.027***		−0.022**	−0.031***		−0.026***
	(0.009)		(0.010)	(0.009)		(0.010)
Patents*Institutions (lag)		0.085***	0.071**		0.100***	0.085**
		(0.029)	(0.030)		(0.037)	(0.036)
GDPPerCapita (lag)	−0.344	−0.243	−0.296	−0.312	−0.238	−0.279
	(0.319)	(0.318)	(0.334)	(0.326)	(0.320)	(0.350)
Trade (lag)	0.015***	0.013***	0.014***	0.014***	0.012***	0.013***
	(0.003)	(0.004)	(0.003)	(0.004)	(0.004)	(0.004)
GDPGrowth (lag)	0.037	0.048	0.045	0.042	0.053*	0.052*
	(0.030)	(0.029)	(0.030)	(0.028)	(0.029)	(0.029)
Inflation (lag)	0.006**	0.005*	0.005*	0.006**	0.005**	0.005**
	(0.003)	(0.003)	(0.003)	(0.003)	(0.002)	(0.002)
EthnicChinese	1.409***	1.238***	1.339***	1.348***	1.246***	1.334***
	(0.342)	(0.303)	(0.309)	(0.400)	(0.362)	(0.348)
Dist	0.014	0.023	0.011	0.029	−0.034	0.000
	(0.208)	(0.190)	(0.192)	(0.236)	(0.217)	(0.214)
	(0.402)	(0.401)	(0.400)	(0.402)	(0.384)	(0.385)
Constant	3.720	2.670	3.231	3.293	3.200	3.181
	(2.945)	(2.890)	(2.965)	(2.821)	(3.140)	(2.927)
Year dummies	YES	YES	YES	YES	YES	YES
Observations	1,912	1,912	1,912	1,912	1,912	1,912
No. of instruments	187	187	188	187	187	188
Arellano-Bond test for AR (1)	−7.935***	−7.922***	−7.937***	−5.976***	−5.936***	−5.939***
Arellano-Bond test for AR (2)	0.030	0.049	0.008	0.066	0.091	0.043
Hansen-J (p-value)	.374	.312	.342	.374	.312	.342
Hanen-J’s Degree of freedom	159	159	159	159	159	159

Note. Using the second-order lags as instruments in the system GMM estimator; Robust standard errors in parentheses clustered by country in Columns from (1) to (3) and corrected by Windmeijer (2005) in Columns from (4) to (6). *p < .10. **p < .05. ***p < .01.

In Table 4, the Arellano-Bond tests for the first-order serial correlation, that is, AR(1), are significantly negative (z = −7.937 and p-value = .000 for the one-system GMM in Column (3); z = − 5.939 and p-value = .000 for the two-system GMM in Column (6)). The second-order serial correlations, that is, AR(2), are not significant (z = 0.008 and p-value = .993 for the one-system GMM in Column (3); z = 0.043 and p-value = .966 the two-system GMM in Column (6)). These findings indicate that it is appropriate to use the linear GMM estimation with lags as instruments. The Hansen J test of overidentification restrictions is not significant in all columns, suggesting that our instruments are valid. All of the results are from specifications in reducing the instrument counts that use the second-lag instruments only for all the endogenous variables instead of all lags in the system GMM estimator (Table 4). We also provide the estimation results using the second-order and the third-order lags as the instruments, the estimation results using the second-order, the third-order, and the fourth-order lags as instruments and the estimation results using all lags as instruments. The estimation results are shown in Table A1 in the Appendix. The results show a perfect Hansen J of 1.00 (p-value = 1), which is a potential problem (Roodman, 2009a, 2009b), even though these results are still consistent with the results in Table 4. We also provide robustness checks by using the “collapse” and PCA estimation methods to reduce the number of instruments. Table A2 in the Appendix shows that the results are similar to those reported in Table 4, indicating that our results are robust regardless of the estimation techniques.

In Table 4, the results using the one-step GMM estimator are largely similar to those using the two-step GMM estimator, suggesting that the bias in the estimation is negligible (Roodman, 2009a, 2009b). In Columns (1) and (4), the coefficients of the interaction term between natural resources and institutions are negative (β = −.027 and p-value < .01; β = −.031 and p-value < .01, respectively). This finding supports Hypothesis H1. This finding indicates that the Chinese OFDI is attracted to the host countries that are rich in natural resources and have poor institutions (Kolstad & Wiig, 2012). Our estimation results are consistent with our observation in Figure 1 that, before the global financial crisis, the Chinese OFDI had been increasing in Africa, where abundant natural resources are available and poor institutions exist. Researchers posit two perspectives to explain this finding. First, the dominance of SOEs in Chinese multinational firms brings them with institution proximity to adapt to an imperfect market and deal with legal difficulties in the host country. Moreover, the experience of the Chinese SOEs enhances their ability to deal with risks from the poor institutions in the host country (Buckley et al., 2007; Kolstad & Wiig, 2012; Morck et al., 2008). Second, Chinese multinational enterprises are late comers. They had challenges to enter the markets in developed countries. Thus, they were located in countries with poor institutions to seek for cheaper natural resources (Buckley et al., 2008; Guo et al., 2014; Morck et al., 2008).

To interpret the interaction terms intuitively, we draw a contour graph to predict the Chinese OFDI based on the results of the moderation of institutions with natural resources reported in Column (4). Figure 2 shows that the host countries having poor institutional indices, which range from −1.2 to −2.5, and having from 40% to 60% of the total GDP contributed by the export values of the natural resources, have attracted the amount of Chinese OFDI ranging of from 45.0 to 272.3 million USD on average. We calculate $y = lnln (x + \sqrt{x^{2} + 1}) \approx \ln (2 x)$ , then $x \approx (e^{y - \ln (2)})$ . Chinese OFDI is equal to $(e^{4.5 - \ln (2)}) =$ 45.0 and $(e^{6.3 - \ln (2)}) =$ 272.3.

Figure 2.

Predicted Chinese OFDI by moderating natural resources with rule of law.

Results regarding the strategic asset-seeking motive and institutions are presented in Columns (2) and (5) in Table 4. The coefficients of the interaction term between these two factors are positive and significant (β = .085 and p-value < .01; β = .100 and p-value < .01, respectively), thus supporting Hypothesis H2.

This finding suggests that well-established institutions positively moderate the relationship between the strategic asset-seeking motive and the Chinese OFDI. In other words, the Chinese MNEs, which are seeking strategic assets, are attracted to the host countries which have well-established institutions. Similarly, this finding is consistent with our observation in Figure 1 that, after the global financial crisis, there has been a remarkable increase in the Chinese OFDI to EU and North America, where the conditions of good institutions and the availability of strategic assets hold. We provide the first evidence of the strategic asset-seeking motive of Chinese OFDI at macro level as assumed by Buckley et al. (2007 and 2018). The previous studies suggested that the need to survive increasing competition and cheap capital in China (Morck et al., 2008; Salidjanova, 2011) has forced Chinese firms to invest abroad through M&A to gain firm specific assets to improve their competitive advantages in both their home and global markets (Deng, 2009; Luo et al., 2010; Meyer et al., 2014; Rui & Yip, 2008).

To provide a more intuitive interpretation of the estimation results, we also prepare a contour graph, which is presented in Figure 3. This graph is drawn from the regression results in Column (5). It shows that, on average, a host country having an institutional index ranging from 1.5 to 2.5 and having 49,357 more patent applications attracts an amount of Chinese investment ranging from 300.9 to 999.1 million USD. The number of registered patents is computed as $(e^{11.5 - \ln (2)}) =$ 49,357. Chinese investment is $(e^{6.4 - \ln (2)}) =$ 300.9 to $(e^{7.6 - \ln (2)}) =$ 999.1 million USD.

Figure 3.

Predicted Chinese OFDI by moderating patents with rule of law.

Columns (3) and (6) in Table 4 report the estimation results with both interaction terms, that is, one with natural resources and institutions, and one with the number of patent applications and institutions. The coefficients of the interaction term between the number of patent applications and institutions are statistically significant and positive (β = .071 and p-value < .05 in Column (3) and β = .085 and p-value < .05 in Column (6)). The coefficients of the interaction term between natural resources and institutions are, however, statistically significant and negative (β = −.022 and p-value < .05 in Column (3) and β = −.026 and p-value < .01 in Column (6)). These findings suggest that the Chinese OFDI seeks for strategic assets in the host countries with well-established institutions, while it searches for natural resources in the host countries with poor institutions.

Robustness Analysis

We conduct additional analyses to check for the robustness of our estimation results. Firstly, we use an alternative measure to proxy for the variable of strategic asset-seeking motive. This alternative variable is the HIS transformation of the number of resident patent applications (PatentsResi variable). We then re-estimate equation (1) with this variable by using the two-step GMM estimation method. The results are reported in Table A3 in the Appendix. These results are consistent with those reported in Table 4.

Secondly, we re-estimate equation (1) with alternative indicators of institutions adopted from the WGI including voice and accountability; political stability and absence of violence; government effectiveness; regulatory quality; rule of law; and control of corruption. The results are reported in Table A4 in the Appendix. These results show that the coefficients of the interaction terms of the number of patent applications and these indicators are positive and significant. Therefore, these results are similar to those reported in Table 4, which use the rule of law to proxy for institutions.

Finally, the previous studies on the Chinese OFDI have focused on tax haven host countries (Cheng & Ma, 2010; Morck et al., 2008) and tax evasion (Morck et al., 2008). The tax haven host countries are normally excluded from the estimation (Cheng & Ma, 2010; Kolstad & Wiig, 2012). We thus re-estimate equation (1) with a new sample of host countries excluding Hong Kong. We cannot include both Cayman Islands and British Virgin Islands in our analysis due to missing data in the independent variables. The estimation results are reported in Table A5 in the Appendix. These results are also consistent with those reported in Table 4.

Concluding Remarks

This study attempts to revisit the motives for the Chinese OFDI, which have been the subject for a great number of previous studies in the literature. The study contributes to the literature by taking institutions from the host countries into consideration to re-examine the effects of the natural resource-seeking motive and the strategic asset-seeking motive on the Chinese OFDI. Using a large dataset collected from various sources for the period from 2003 to 2020 and applying the system GMM estimation method, the study finds that the Chinese OFDI is attracted to the host countries that are rich in natural resources with poor institutions. This finding is in line with other previous studies in the literature (Buckley et al., 2007, 2018; Kolstad & Wiig, 2012). Additionally, it is revealed that the Chinese OFDI is directed to the host countries with well-established institutions to search for strategic assets, which are measured by the number of patent applications. These findings are robust to a wide spectrum of indicators for institutions and different estimation techniques.

Our findings have lifted the curtain covering the nature of the mechanisms through which contextual factors affect the Chinese OFDI. A clear understanding about the determinants of the pattern of internationalization by Chinese MNEs has been achieved. These findings close the gap in the literature reviewing the contradictory findings about the strategic asset-seeking and natural resource-seeking motives for the Chinese OFDI. This study can be extended to the OFDI from other countries. As a result, the study provides practical evidence for policymakers to design appropriate foreign investment policies in response to the dynamics of the Chinese OFDI worldwide.

Our study has some limitations. First, we are not able to distinguish between Chinese OFDI by M&A and greenfield investment due to a lack of data. Therefore, we are not able to dig deep into the mechanism through which the strategic asset-seeking motive has been carried out by the Chinese MNEs. Second, we do not have data on the industries where the Chinese MNEs invested in the host countries. Hence, we cannot provide insights into the motives of the Chinese OFDI by industry. Third, other factors such as infrastructure, which may be proxied by utility consumption, or human resources of the host countries may affect the Chinese OFDI. In this study, due to limited data we are not able to examine these factors. Future studies should address these limitations to have better insights into the Chinese OFDI.

Footnotes

Appendix

Table A5.

Estimation Results for a Sample Without Hong Kong Using the Two-step GMM Estimator.

	(1)	(2)	(3)
COFDI (lag)	0.193***	0.189***	0.187***
	(0.052)	(0.053)	(0.007)
Patents (lag)	0.182***	0.191***	0.186***
	(0.044)	(0.044)	(0.008)
NaturalRes (lag)	−0.002	0.004	−0.000
	(0.012)	(0.012)	(0.002)
Institutions (lag)	0.624	−0.338	0.060
	(0.461)	(0.520)	(0.090)
NaturalRes*Institutions (lag)	−0.029***		−0.025***
	(0.009)		(0.002)
Patents*Institutions (lag)		0.098***	0.078***
		(0.035)	(0.008)
GDPPerCapita (lag)	−0.300	−0.202	−0.268
	(0.340)	(0.324)	(0.250)
Trade (lag)	0.010*	0.007	0.009***
	(0.005)	(0.005)	(0.001)
GDPGrowth (lag)	0.045	0.056**	0.054***
	(0.027)	(0.027)	(0.005)
Inflation (lag)	0.006**	0.005**	0.005***
	(0.003)	(0.002)	(0.000)
EthnicChinese	1.396***	1.319***	1.378***
	(0.392)	(0.330)	(0.103)
Dist	0.012	−0.038	−0.047
	(0.226)	(0.211)	(0.051)
Constant	3.602	3.239	3.784***
	(2.842)	(2.834)	(0.532)
Year dummies	YES	YES	YES
Observations	1895	1895	1895
No. of instruments	187	187	188
Arellano-Bond test—AR (1)	−5.953***	−5.948***	−6.896***
Arellano-Bond test—AR (2)	0.032	0.076	0.031
Hansen-J (p-value)	.491	.421	.420
Hanen-J’s Degree of freedom	159	159	159

Note. Robust standard errors corrected by Windmeijer (2005) in parentheses. * p < .10. **p < .05. ***p < .01.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was sponsored by Foreign Trade University under research grant number FTURP01-2020-01.

ORCID iD

Nam Hoang Vu

References

Arellano

Bond

(1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58(2), 277–297.

Arellano

Bover

(1995). Another look at the instrumental variable estimation of error-components models. Econometrics Journal, 68(1), 29–51.

Bahar

Rapoport

(2018). Migration, knowledge diffusion and the comparative advantage of nations. The Economic Journal, 128(612), F273–F305.

Bai

(2010). Instrumental variable estimation in a data rich environment. Economic Theory, 26(6), 1577–1606.

Bellemare

M. F.

Wichman

C. J.

(2020). Elasticities and the inverse hyperbolic sine transformation. Oxford Bulletin of Economics and Statistics, 82(1), 50–61.

Bénassy-Quéré

Coupet

Mayer

(2007). Institutional determinants of foreign direct investment. World Economy, 30(5), 764–782.

Blundell

Bond

(1998). Initial conditions and moment restrictions in dynamic panel data models. Econometrics Journal, 87(1), 115–143.

Bond

S. R.

(2002). Dynamic panel data models: A guide to micro data methods and practice. Portuguese Economic Journal, 1(2), 141–162.

Buckley

P. J.

Clegg

L. J.

Cross

A. R.

Liu

Voss

Zheng

(2007). The determinants of Chinese outward foreign direct investment. Journal of International Business Studies, 38(4), 499–518.

10.

Buckley

P. J.

Clegg

L. J.

Voss

Cross

A. R.

Liu

Zheng

(2018). A retrospective and agenda for future research on Chinese outward foreign direct investment. Journal of International Business Studies, 49(1), 4–23.

11.

Buckley

P. J.

Cross

A. R.

Tan

Xin

Voss

(2008). Historic and emergent trends in Chinese outward direct investment. Management International Review, 48(6), 715–748.

12.

Burbidge

J. B.

Magee

Robb

A. L.

(1988). Alternative transformations to handle extreme values of the dependent variable. Journal of the American Statistical Association, 83(401), 123–127.

13.

Cheng

L. K.

(2010). China’s outward foreign direct investment. In C. F. Robert & W. Shang-Jin (Eds.), China’s growing role in world trade (pp. 545–578). University of Chicago Press.

14.

Chen

M. K.

(2013). The effect of language on economic behavior: Evidence from savings rates, health behaviors, and retirement assets. American Economic Review, 103(2), 690–731.

15.

Cheung

Y. W.

Qian

(2009). Empirics of China’s outward direct investment. Pacific Economic Review, 14(3), 312–341.

16.

Chou

K.-H.

Chen

C.-H.

Mai

C.-C.

(2011). The impact of third-country effects and economic integration on China’s outward FDI. Economic Modelling, 28(5), 2154–2163.

17.

Christofi

Vrontis

Makrides

(2022). Exploring the role of institutions in Chinese OFDI: A systematic review and integrative framework. Asia Pacific Business Review, 28(2), 187–213.

18.

Clemens

M. A.

Tiongson

E. R.

(2017). Split decisions: Household finance when a policy discontinuity allocates overseas work. The Review of Economics and Statistics, 99(3), 531–543.

19.

Cui

Jiang

(2012). State ownership effect on firms’ FDI ownership decisions under institutional pressure: A study of Chinese outward-investing firms. Journal of International Business Studies, 43(3), 264–284.

20.

Cui

Meyer

K. E.

H. W.

(2014). What drives firms’ intent to seek strategic assets by foreign direct investment? A study of emerging economy firms. Journal of World Business, 49(4), 488–501.

21.

Deng

(2009). Why do Chinese firms tend to acquire strategic assets in international expansion? Journal of World Business, 44(1), 74–84.

22.

Dunning

J. H.

(1998). Location and the multinational enterprise: A neglected factor? Journal of International Business Studies, 29(1), 45–66.

23.

Dunning

J. H.

Lundan

S. M.

(2008). Multinational enterprises and the global economy. Edward Elgar Publishing.

24.

Globerman

Shapiro

(2002). Global foreign direct investment flows: The role of governance infrastructure. World Development, 30(11), 1899–1919.

25.

Guo

J. J.

Wang

G. C.

Tung

C. H.

(2014). Do China’s outward direct investors prefer countries with high political risk? An international and empirical comparison. China & World Economy, 22(6), 22–43.

26.

Hao

Guo

Ren

(2020). Does outward foreign direct investment (OFDI) affect the home country’s environmental quality? The case of China. Structural Change and Economic Dynamics, 52, 109–119.

27.

Holtz-Eakin

Newey

Rosen

H. S.

(1988). Estimating vector autoregressions with panel data. Econometrica: Journal of the Econometric Society, 56, 1371–1395.

28.

Khan

N. H.

Danish

Y. J.

Khan

Z. L. F.

(2020). Nexus between carbon emission, financial development, and access to electricity: Incorporating the role of natural sources and population growth. Journal of Public Affairs, 21(1), 21–31.

29.

Kolstad

(2009). The resource curse: Which institutions matter? Applied Economics Letters, 16(4), 439–442.

30.

Kolstad

Wiig

(2012). What determines Chinese outward FDI? Journal of World Business, 47(1), 26–34.

31.

Kong

Guo

Wang

Sui

Zhou

(2020). Home-country environment and firms’ outward foreign direct investment decision: Evidence from Chinese firms. Economic Modelling, 85, 390–399.

32.

Latif

Mengke

Latif

Ximei

Pathan

Z. H.

Salam

Jianqiu

(2018). The dynamics of ICT, foreign direct investment, globalization and economic growth: Panel estimation robust to heterogeneity and cross-sectional dependence. Telematics and Informatics, 35(2), 318–328.

33.

Liang

Giroud

Rygh

(2022). Strategic asset-seeking acquisitions, technological gaps, and innovation performance of Chinese multinationals. Journal of World Business, 57(4), 101325.

34.

Lim

L. Y. C.

(1983). Singapore’s success: The myth of the free market economy. Asian Survey, 23, 752–764.

35.

Loncan

(2021). The effects of project scale on FDI location choices: Evidence from emerging economies. Management International Review, 61, 157–205.

36.

Luo

Tung

R. L.

(2007). International expansion of emerging market enterprises: A springboard perspective. Journal of International Business Studies, 38(4), 481–498. https://doi.org/10.1057/palgrave.jibs.8400275

37.

Luo

Xue

Han

(2010). How emerging market governments promote outward FDI: Experience from China. Journal of World Business, 45(1), 68–79.

38.

MacKinnon

J. G.

Magee

(1990). Transforming the dependent variable in regression models. International Economic Review, 31, 315–339.

39.

McKenzie

(2017). Identifying and spurring high-growth entrepreneurship: Experimental evidence from a business plan competition. American Economic Review, 107(8), 2278–2307. https://doi.org/10.1257/aer.20151404

40.

Mehlum

Moene

Torvik

(2006). Institutions and the resource curse. The Economic Journal, 116(508), 1–20.

41.

Mehrhoff

(2009). A solution to the problem of too many instruments in dynamic panel data GMM. (Discussion Paper Series 1. No 31/2009).

42.

Meyer

K. E.

Ding

Zhang

(2014). Overcoming distrust: How state-owned enterprises adapt their foreign entries to institutional pressures abroad. Journal of International Business Studies, 45(8), 1005–1028. https://doi.org/10.1057/jibs.2014.15

43.

MOFCOM. (2021). 2010 statistical bulletin of China’s outward foreign direct investment. The Ministry of Commerce of the People’s Republic of China.

44.

Morck

Yeung

Zhao

(2008). Perspectives on China’s outward foreign direct investment. Journal of International Business Studies, 39(3), 337–350.

45.

Nickell

(1981). Biases in dynamic models with fixed effects. Econometrica, 49, 1417–1426.

46.

Pence

K. M.

(2006). The role of wealth transformations: An application to estimating the effect of tax incentives on saving. The BE Journal of Economic Analysis & Policy, 5(1), 1–24.

47.

Peng

M. W.

(2012). The global strategy of emerging multinationals from China. Global Strategy Journal, 2(2), 97–107.

48.

Poston

D. L.

Wong

J. H.

(2016). The Chinese diaspora: The current distribution of the overseas Chinese population. Chinese Journal of Sociology, 2(3), 348–373.

49.

Quer

Claver

Rienda

(2012). Political risk, cultural distance, and outward foreign direct investment: Empirical evidence from large Chinese firms. Asia Pacific Journal of Management, 29(4), 1089–1104.

50.

Ramamurti

Hillemann

(2018). What is “Chinese” about Chinese multinationals? Journal of International Business Studies, 49(1), 34–48.

51.

Ramasamy

Yeung

Laforet

(2012). China’s outward foreign direct investment: Location choice and firm ownership. Journal of World Business, 47(1), 17–25.

52.

Rjoub

Aga

Abu Alrub

Bein

(2017). Financial reforms and determinants of FDI: Evidence from landlocked countries in Sub-Saharan Africa. Economies, 5(1), 1–12.

53.

Roodman

(2009a). A note on the theme of too many instruments. Oxford Bulletin of Economics and Statistics, 71(1), 135–158.

54.

Roodman

(2009b). How to do xtabond2: An introduction to difference and system GMM in Stata. The Stata Journal, 9, 86–136.

55.

Rui

Yip

G. S.

(2008). Foreign acquisitions by Chinese firms: A strategic intent perspective. Journal of World Business, 43(2), 213–226.

56.

Salidjanova

(2011). Going out: An overview of China’s outward foreign direct investment. US-China Economic and Security Review Commission.

57.

Sattar

Tolassa

T. H.

Hussain

M. N.

Ilyas

(2022). Environmental effects of China’s overseas direct investment in South Asia. Sage Open, 12(1), 1–21.

58.

Zhang

Wei

(2016). A practical test for strict exogeneity in linear panel data models with fixed effects. Economics Letters, 147, 27–31.

59.

UNCTAD. (1998). World investment report 1998: Trends and determinants. United Nations.

60.

UNCTAD. (2018). World investment report 2018: Investment and new industrial policies. UN.

61.

Vecchi

Brennan

(2022). Two tales of internationalization: Chinese internet firms’ expansion into the European market. Journal of Business Research, 152, 106–127.

62.

N. H.

Bui

T. A.

Hoang

T. B.

Pham

H. M.

(2022). Information technology adoption and integration into global value chains: Evidence from small- and medium-sized enterprises in Vietnam. Journal of International Development, 34(2), 259–286.

63.

Wang

Hong

Kafouros

Boateng

(2012). What drives outward FDI of Chinese firms? Testing the explanatory power of three theoretical frameworks. International Business Review, 21(3), 425–438.

64.

Wang

Hong

Kafouros

Wright

(2012). Exploring the role of government involvement in outward FDI from emerging economies. Journal of International Business Studies, 43(7), 655–676.

65.

Williams

Abu Alrub

Aga

(2022). Ecological footprint, economic uncertainty and foreign direct investment in South Africa: Evidence from asymmetric cointegration and dynamic multipliers in a nonlinear ARDL approach. Sage Open, 12(2), 1–15.

66.

Windmeijer

(2005). A finite sample correction for the variance of linear efficient two-step GMM estimators. Econometrics Journal, 126(1), 25–51.

67.

Wintoki

M. B.

Linck

J. S.

Netter

J. M.

(2012). Endogeneity and the dynamics of internal corporate governance. Journal of Financial Economics, 105(3), 581–606.

68.

Wooldridge

J. M.

(2010). Econometric analysis of cross section and panel data. MIT Press.

69.

Yang

J. H.

Wang

K. L.

Yeh

C. Y.

(2018). Capital intensity, natural resources, and institutional risk preferences in Chinese outward foreign direct investment. International Review of Economics & Finance, 55, 259–272.

70.

Zhang

Daly

(2011). The determinants of China’s outward foreign direct investment. Emerging Markets Review, 12(4), 389–398.

A Revisit of Motives for Chinese Outward Foreign Direct Investment: The Role of the Institution in Host Countries

Abstract

Keywords

Introduction

Chinese Outward Foreign Direct Investment: Stylized Facts

Theory and Hypothesis Development

Theoretical Background

Hypothesis Development

Methodology and Data

Methodology

Data

Regression Results

Robustness Analysis

Concluding Remarks

Footnotes

Appendix

Declaration of Conflicting Interests

Funding

ORCID iD

References