Political Risks to China’s Energy Infrastructure Investment in Countries along the Belt and Road

Abstract

Energy infrastructure connectivity is a key implementation area for the success of the Belt and Road Initiative. In the current complex international environment, China’s energy infrastructure investment in the countries along the Belt and Road is often restricted due to varying degrees of heated disputes with the host countries, thus facing the political risk of investment loss or even failure. This paper used the Heckman two-stage selection model to overcome the sample selection problem in traditional literature and systematically studied the political risks to China’s energy infrastructure investment in countries along the Belt and Road within a unified framework. The results showed that Chinese energy infrastructure investments exhibited market-seeking and energy resource-seeking features; while political risks had a significant impact on choosing the locations of energy infrastructure investments, socio-environmental factors were not given sufficient attention. At the national level, the implementation of the Belt and Road Initiative was beneficial to China’s foreign energy infrastructure investments under certain regulatory levels and environmental policy performance.

1. Introduction

The Belt and Road Initiative (BRI) is a major development initiative of China that integrates both domestic and international interests. The purpose of the BRI is to promote regional cooperation, and the key to achieving this establishing infrastructure connectivity. In March 2015, China’s National Development and Reform Commission, Ministry of Foreign Affairs, and Ministry of Commerce jointly released the “Vision and Actions on Jointly Building Silk Road Economic Belt and 21st-Century Maritime Silk Road”, which proposed that “facilities connectivity regarding energy, transportation and communications trunk line networks is a priority area for the construction of the BRI”. From 2013 to 2018, China’s direct investment in countries along the Belt and Road (B&R) exceeded USD 90 billion, the turnover of foreign contracted projects exceeded USD 400 billion, and 82 economic and trade cooperation zones were established in B&R countries, with a total investment of over USD 30 billion.

As the foundation of modern economy, logistics system, industrial system, and social service system, energy infrastructure investment is an important part of the implementation of the BRI. First of all, countries along the B&R have different resource endowments, and thus, improving energy infrastructure construction is conducive to strengthen energy cooperation between China and B&R countries and realize complementary production factors and resource advantages. Secondly, the current Russia–Ukraine war caused by the energy crisis further highlights the importance of energy security. The BRI opens up a road for onshore energy production, reduces the security risks in China’s energy imports, and is conducive to achieve national energy security under the condition of openness. Thirdly, in the context of global climate change, the B&R countries, as developing economies, are lagging in energy infrastructure, but are also undergoing the largest green and low-carbon transformation in history. On 22 September 2021, President Xi attended the general debate of the 76th session of the UN General Assembly, and proposed that China will strongly support the green and low-carbon development of energy in developing countries and no new coal power projects will be built outside China. This also has brought worldwide attention to China’s investment in energy infrastructure in B&R countries.

However, a crucial issue is that since the introduction of the BRI, there has been a lot of international skepticism about it. The U.S. Secretary of State Mike Pompeo said during a visit to the United Kingdom in 2019 that the BRI undermines the sovereignty of all countries, and the U.S. urged the British to remain vigilant and speak out against the initiative. B&R countries are also wary and suspicious of the real intentions behind this initiative and the possible adverse conditions it might engender. Some foreign scholars have suggested that the BRI will have a negative impact not only on interstate foreign policy, but also on geopolitical tensions. These controversies often lead to restrictions on or even resistance to China’s energy infrastructure investments in B&R countries to varying degrees or in different ways, resulting in investment losses or even failure. According to the State Ministry of Commerce, the failure rate of overseas investments by Chinese companies in energy and other infrastructure sectors has been as high as 70% in recent years. Overseas investments by Chinese firms may be driven by opaque non-economic factors and non-market-oriented behavior, and are thus different in nature from cross-border investments under market-oriented behavior in Western countries.

This research was motivated by the skepticism about the BRI and China’s current international energy cooperation problem. Without conducting any empirical research, are the international criticisms and even worries about China’s energy infrastructure investments in the countries along the B&R reasonable? Has the current state of international politics affected China’s investments in energy infrastructure?

In fact, political risk has become one of the main obstacles to China’s energy infrastructure investments in B&R countries. These countries have different levels of economic development, complex geopolitical relations, religious beliefs and ethnic conflicts, and the instability of the domestic and international situation brings great uncertainty to energy infrastructure investment. Moreover, energy infrastructure investment is a huge upfront, easy to operate after completion and can maintain stable income, which is easily coveted by the host government and nationalized companies. Therefore, ways to assess the political risks to China’s infrastructure investments in B&R countries and to propose risk-averse instruments are urgent issues at present.

Although academics have greatly contributed to the study of political risks, there are also certain drawbacks: First off, the majority of political risk studies based on B&R countries use the investment risk of unspecified fields as their research subject, and there are not many articles focusing on the field of energy investment risk. Qualitative research is the main method utilized. Furthermore, the majority of quantitative research is conducted from the perspective of management science, which limits research on the detection and mitigation of existing energy investment risks. Secondly, most of the previous quantitative research on the field of energy investment risks focuses on the political risk assessment perspective, and research on the effect is relatively lacking, so analyses of the effect of political risks on the field of energy infrastructure is insufficient.

In recent years, the BRI has been considered as a significant strategic deployment to promote China’s opening-up policy. And there is a large amount of specialized literature on the risk of China’s investment in the countries along the B&R, but there are relatively few research studies on the risk analysis of energy project investment. This paper overcomes the problem of sample selection in traditional literature, and systematically investigates the location choice and investment scale of China’s energy infrastructure investments in B&R countries from four aspects of market search, resource access, social environment and political risk within a unified framework. Meanwhile, considering the special strategic background of the BRI, we further test whether the political risks to China’s energy infrastructure investment have changed after the implementation of the BRI, and whether the implementation of the BRI is a driving force for China’s energy infrastructure investment. This research improves and develops the existing international investment theory and new institutional economics theory in the field of energy investment, while also providing empirical evidence for China’s current adjustment to energy investment strategies for countries along the B&R and the formation of enterprise international competitiveness.

Compared with previous studies, the main contributions of this paper are as follows. Firstly, in terms of research content, due to the special nature of the energy infrastructure industry, and the fact that energy security is related to the lifeblood of a country and carries high risks at the political and public interest levels, a systematic and specialized study on energy infrastructure investment is primarily conducted in this paper. Secondly, in terms of measurement methods, based on the Heckman two-stage model, this paper distinguishes between location choice and investment scale, effectively solves the problem of biased estimation caused by sample selection in previous literature, and systematically studies the location choice decision and investment scale decision of China’s foreign energy infrastructure investments within a unified framework. Thirdly, in terms of data source, this paper draws on the China Global Investment Tracker, which is a globally authoritative dataset created by the Heritage Foundation, thus avoiding the peculiarities of previous incomplete samples or incomplete conclusions.

2. Theoretical Foundations and Micro Evidence: A Brief Summary

The international monopoly dominance theory, the internalization theory and the product cycle theory have been adopted to study the motives, evolutionary laws and trends of international investment from different perspectives. With regard to developed countries, issues such as incomplete monopoly advantage of market competition, internalized market knowledge, cost transfer, and national gradient evolution combining monopoly advantage and location choice have been investigated based on these theories. For theoretical studies related to overseas investment in non-developed countries, Lewis Wells’ “small-scale technology theory” of 1977 and Cantwell’s “technological innovation and industrial upgrading theory” of 1990 have been utilized to analyze the phenomenon of overseas investment by developing enterprises from diverse angles. By studying Japanese overseas investment cases, Terutomo Ozawa’s “investment-development cycle theory” advocates that overseas investment patterns should be adapted to the orderly flow of capital in response to structural changes in the world economy. All these representative theories have served as guidance for overseas investment activities of enterprises in non-developed countries. The above theories provide a regular summary of the characteristics and trends of overseas investment practices of multinational corporations at a specific stage from a macro level. Nevertheless, they fail to provide guidance for Chinese overseas investment in a general sense and lack applicability in line with industry characteristics.

For overseas investing firms, both investment location choice and investment scale are critical to their performance and even survival [1]. With the increasing damage caused by non-economic risks to Chinese firms abroad, political risks have raised high concerns in the study of cross-border investments. As for the definition of political risk, the academia has not yet reached a unified definition. In the 1950s, U.S. scholars introduced the concept of “political risk” to reflect the nationalization of companies in host countries; in the 1970s, a new perspective on political risk research emerged, with increasing emphasis on the impact of governmental or social actions that occur both inside and outside the host country. In the 21st century, with the rise of economic globalization, the economic development of countries and the improvement in people’s welfare, a series of extreme political risks have evolved, which have become a research focus because of the changes in host country policies, protection of resources, cultural differences, nationalism and religious conflicts. Overall, there is a distinction between broad and narrow political risk. Narrowly defined political risk is the possibility of loss to investors due to political instability and changes in the political environment, policies and regulations in the host country. In contrast, political risk in a broad sense takes into account the interaction between political, economic, social and ecological environments. The current scholarly concept of “political risk” is an umbrella term for discontinuous adverse effects related to politics, economics, society and resources, and is a complex multifaceted phenomenon that depends on the institutional and economic development stages of countries and regions [2]. In general, a stable political environment can provide a stable institutional environment for domestic and foreign investors, offer good property rights protection and social environment, and avoid damage to multinational enterprises due to political unrest and economic fluctuations [3,4,5]. According to the new institutional economics, political risk is in fact partly a reflection of the institutional environment and changes in the institutional system, which can directly affect the advantages of the firms and thus the return on investment [6]. In terms of the influence of host country’s political risk on China’s transnational investment, scholars’ research conclusions also differ. Refs. [6,7] believed that the level of governance of host government is closely linked to Chinese outbound investment. Some scholars, however, hold a different view. For instance, Ref. [8] showed that China’s multinational enterprises were not sensitive to political risk in the host country and even had immunity against political risk; Ref. [9] analysis based on industry data revealed that host country political risk had no significant effect on Chinese outward manufacturing investment. Ref. [9] also argued that the impact of different dimensions of political risk in host countries on the location choice of Chinese outward investment is different and cannot be generalized. Some scholars have also contended that the “institutional risk preference” based on the perspective of international capital flows is largely due to the fact that China’s outward foreign direct investment (OFDI) flows more to regions with lower levels of economic development and more abundant natural resources [10,11]. The introduction of new institutionalism theory into the study of overseas investment provides a valuable perspective for a deeper exploration of overseas investment risks. Ref. [12] introduced institutional variables into the study of economic history and discovered the impact of institutions and institutional change on economic growth, producing a set of new institutional economics analytical frameworks and methods. The core premise is that “economic man” creates new demand for institutions in order to minimize losses, and that this demand is constantly changing in order to develop a more appropriate institutional system. Due to the large number of overseas investment subjects and stakeholders, more and more scholars regard the national system, government behavior, and social culture as key factors for the outward investment of emerging-country enterprises, and opine that the disparity of institutional differences between them will become key for the investment’s success or failure [13].

In terms of the impact of the host country’s political system on China’s cross-border investment, scholars also arrived at different research conclusions, using different theories such as “institutional facilitation theories”, “institutional hindrance theories”, “institutional irrelevance theories”, etc. For instance, Ref. [14] believed that the impact of different dimensions of political risk in host countries on China’s outbound investment location choice is different and cannot be generalized. Based on the perspective of international capital flows, Ref. [15] argued that “institutional risk preference” is largely caused by the fact that China’s OFDI flows more to regions with a lower level of economic development and more abundant natural resources. Ref. [16] proved the point of view of “institutional obstructionism” in their research. After that, Ref. [17] analyzed the influencing variables of China’s direct investment in the countries along the B&R based on the spatial perspective. Ref. [18] argued that the imperfections of the host countries’ system of developing economies are motivating factors for China’s outward FDI.

With the implementation of the BRI, scholars began to conduct research on the political risk factors of China’s infrastructure investment. Ref. [19] argued that the host country’s payment capacity, project financing capacity, government interference risk, and the risk of policy and legal changes all have an impact on infrastructure investment. Refs. [14,15] analyzed the influence of political risks that China may face at home and abroad in the B&R energy infrastructure investment. Refs. [20,21,22] revealed that institutional quality, the participation of international financial institutions that can impact energy infrastructure investment’s risk-taking commitment, and financing constraints are the key factors determining the success or failure of the B&R infrastructure projects. Ref. [23] found that political system differences and institutional distance have a significant negative impact on China’s infrastructure investment, and the signing of bilateral investment agreements can reduce the negative effect of political system differences to a certain extent. Ref. [16] suggested that under the BRI, China’s investment in infrastructure in countries along the B&R has increased, while institutional differences can significantly increase the probability of problematic projects. Overall, current research on the political risk of infrastructure investment mainly focuses on qualitative analysis, while relevant quantitative analysis and empirical studies to promote energy infrastructure connectivity are relatively rare.

Chinese outbound energy investments have been active since the 1990s. The 2008 financial crisis brought about unprecedented opportunities for Chinese companies to invest in outbound energy, and data on outbound investment in the mining sector, including oil and gas resources, found that the flow, stock and share of overseas investment in the energy sector have continued to grow in tandem, and are firmly in the top three of all sectors of Chinese overseas investment. From 2000 to 2010, Chinese companies completed more than 369 mergers and acquisitions (M&A) concerning energy, mineral and other resources, amounting to more than USD 50 billion. China invested USD 102.2 billion in oil and power from 2006 to 2010, accounting for 47% of its total overseas investment. Data from Petroleum Intelligence Weekly show that from 2009 to 2013, China’s central oil companies invested a total of USD 109.7 billion overseas, with a net increase of USD 104.1 billion in transactions, which is 11.5 times more than the investment made by the three largest U.S. oil companies in the same period. Since the inception of the BRI, China’s overseas investment in mineral resources and energy has been growing steadily. With the transformation of China’s economic structure, its outbound energy investment has shifted from the initial global search for cheap energy resources to secure domestic supply to regional optimization of energy supply through the acquisition of overseas high-quality assets to enhance channel control of energy security. On one hand, China has very limited resources per capita, and the rapid industrialization process since the reform and opening-up has led to a rapid expansion of demand for resource products and energy and raw materials. On the other hand, the pricing power of international energy products is often in the hands of multinational monopolies, mainly in the developed countries. Only through large-scale overseas investments can Chinese energy-based enterprises secure energy support for economic development and enhance their voice in the international energy market [19].

In terms of the overall M&A situation, the development of cross-border M&A by Chinese enterprises can be roughly divided into three stages: the budding stage before the financial crisis, the rapid development stage after the financial crisis until 2016, and the adjustment stage after 2016. In terms of the number of M&A, it continued to experience rapid growth after 2016, with 107 cases in 2019, occupying 56% of the total number of overseas M&A in that year. Compared with 2008, there were only 10 cases of cross-border M&A, accounting for only 14.9%. In terms of M&A amount, China’s cross-border M&A to countries along the B&R reached a new record total of USD 36.62 billion in 2017. Since then, although there has been a significant decline due to the impact of the general environment of outward investment, the overall share of cross-border M&A has basically remained above 20%, increasing from 14.9% in 2008.

The volume and amount of cross-border M&A by Chinese firms along the B&R reveal an imbalanced distribution, with M&A destinations mostly focusing on West Asia and the ASEAN region. From 2008 to 2018, there were 487 successful Chinese M&A deals in nations along the B&R, with 237 cases in 10 ASEAN countries, and 70 and 62 cases in 18 West Asian countries and 8 South Asian countries, respectively. In terms of M&A value, the 18 countries in West Asia accounted for 31.57% of the total M&A value of the countries along the B&R, while the 10 ASEAN countries accounted for 29.52%. Concerning different industries in different countries, Figure 1 shows that the top five industries in China’s investment in energy infrastructure along the Belt and Road were energy, manufacturing, finance, transportation and real estate, with energy investment accounting for 72% of the total. Figure 2 shows that there were two characteristics of China’s overseas energy M&A in B&R countries. Firstly, both the number and amount of energy M&A increased and varied greatly between regions, with obvious differentiation between two levels. For example, China’s energy infrastructure investments in the Middle East, Central Asia, Russia and other countries and regions rich in energy endowments such as oil and natural gas were more concentrated. In terms of investment amount, Kazakhstan, Russia, Singapore, Indonesia and Iraq occupied a large proportion. Moreover, the amount of China’s M&A in “One Belt” countries was concentrated in some years, while that in “One Road” countries was more scattered. Secondly, the single amount of energy investment was high. Given the significant advantages of the B&R countries in energy resource endowments, China’s overseas M&A transactions in these destinations were concentrated on large energy projects, and the degree of volatility would change to a large extent with changes in the international energy supply and demand pattern. Therefore, the energy cooperation faces many kinds of risks in the process of China’s energy infrastructure investment in B&R countries with radically different national conditions.

3. Data and Model Setting

3.1. Data Source

The data used in this paper were retrieved from the China Global Investment Tracker, a database created by the Heritage Foundation which records detailed information about Chinese outbound investments. The database is currently the only comprehensive set of data covering China’s global investment and construction, recording 3100 large transactions including energy, transportation and real estate, as well as 280 problematic transactions. In this study, the data on China’s investment in energy infrastructure in B&R countries were extracted, and due to issues with data availability, 31 target B&R countries were selected, with the sample spanning from the period of 2005 to 2019.

3.2. Econometric Model

China’s investment in energy infrastructure in countries along the B&R is discontinuous, in that the flow of investment in some countries was zero in different years. The previous literature examining cross-country investment excluded these samples, thus inevitably leading to biased estimation results. The reason is that ignoring or excluding these samples may not cause bias only if China’s investment in B&R countries is random, which, however, is not the case. That is to say, zero Chinese investment in B&R countries is neither accidental nor inevitable, but the result of optimal decision-making. To address similar problems of sample selection bias, we adopted Heckman two-stage selection model to divide China’s investment decisions in energy infrastructure in B&R countries into two stages. The first stage was the Probit investment choice model, which looks at the factors influencing China’s investment choices, and the second stage was the modified investment quantity model, which further observes the factors influencing the quantity of China’s investment in B&R countries.

Thus, we estimated the investment location choice equation and calculated the inverse Mills ratio in the first stage, and then introduced the inverse Mills ratio into the second stage to estimate the investment scale equation. The specific model is

$$\mathrm{Pr}\left(energy=1\right)=\mathsf{\Phi }({\alpha }_0+\alpha Controls+{\sum }_n{\theta }_k{X}_{ijt}^k+\beta {\gamma }_{ijt}+{\lambda }_j+{\nu }_t+{\epsilon }_{ijt})$$

(1)

$${energy}_{ijt}={\alpha }_0+\alpha Controls+{\sum }_n{\theta }_k{X}_{ijt}^k+\beta {\gamma }_{ijt}+{\lambda }_j+{\nu }_t+{\epsilon }_{ijt}$$

(2)

where ${energy}_{ijt}$ indicates the actual amount of energy infrastructure investment. If ${energy}_{ijt}$ > 0, then ${energy}_{ijt}$ = 1; otherwise ${energy}_{ijt}$ = 0. Equation (1) is the Heckman first-stage investment selection model, where $\mathrm{Pr}\left(energy=1\right)$ indicates that China’s investment in the host country’s energy infrastructure is greater than zero; $\mathsf{\Phi }$ is the probability distribution function for the standard plus distribution; $X_{ijt}^k$ is the core explanatory variable; ${\nu }_t$ is the fixed effect; ${\epsilon }_{ijt}$ is the random disturbance term. We introduced an extra ${\gamma }_{ijt}$ term which represents inverse Mills ratio in Equation (2) to overcome sample selection bias. This term was calculated according to the estimated results of the first stage, and the specific calculation formula is

$${\gamma }_{ijt}=\frac{\varphi ({\alpha }_0+\alpha Controls+{\sum }_n{\theta }_k{X}_{ijt}^k+\beta {\gamma }_{ijt}+{\lambda }_j+{\nu }_t+{\epsilon }_{ijt})}{\mathsf{\Phi }({\alpha }_0+\alpha Controls+{\sum }_n{\theta }_k{X}_{ijt}^k+\beta {\gamma }_{ijt}+{\lambda }_j+{\nu }_t+{\epsilon }_{ijt})}$$

(3)

where $\varphi (\cdot )$ and $\mathsf{\Phi }\left(\cdot \right)$ represent the probability density function and probability distribution function of standard plus distribution, respectively. If ${\gamma }_{ijt}$ is significantly non-zero, it indicates the existence of sample self-selection problem, and that the Heckman selection model is effective.

In addition, previous studies have shown that the political and economic environment of the host country and energy infrastructure investments might have a two-way causal relationship and lead to endogeneity problems. In view of this, we lagged the host country’s economic variables by one period to address the endogeneity issue; see model setup for details. It should be noted that we did not lag the institutional variables by one period for the following reasons: firstly, Chinese investments in energy infrastructure in B&R countries are relatively small in scale and do not yet affect the host country’s institutions; secondly, institutional evolution is a long-term process and does not change much in the short term. Therefore, the host country’s system is exogenous.

The investment gravity model [24] was used as the benchmark model. Also, the variables to be studied were added to the model, which is expressed as

$$\begin{array}{cc}\hfill \mathrm{Pr}\left(energy\right)=& \mathsf{\Phi }[{\alpha }_0+{\alpha }_1\ln \left({GDP}_{j\left(t-1\right)}\right)+{\alpha }_2\ln \left({GDPP}_{j\left(t-1\right)}\right)+{\alpha }_3\ln \left({ENDOW}_{j\left(t-1\right)}\right)\hfill \\ & +{\alpha }_4\ln \left({MUR}_{j\left(t-1\right)}\right)+{\alpha }_5\ln \left({DEPEND}_{j\left(t-1\right)}\right)+{\alpha }_6\ln \left({DEPEND}_{j\left(t-1\right)}\right)\hfill \\ & +{\alpha }_7\ln \left({DTAX}_{j\left(t-1\right)}\right)+{\alpha }_8\ln \left({LABOR}_{j\left(t-1\right)}\right)+{\alpha }_9\ln \left({DIST}_{j\left(t-1\right)}\right)+{\sum }_n{\theta }_k{X}_{ijt}^k\hfill \\ & +\beta {\gamma }_{ijt}+{\lambda }_j+{\nu }_t+{\epsilon }_{ijt}]\hfill \end{array}$$

(4)

$$\begin{array}{cc}\hfill \ln \left({energy}_{ijt}\right)=& {\alpha }_0+{\alpha }_1\ln \left({GDP}_{j\left(t-1\right)}\right)+{\alpha }_2\ln \left({GDPP}_{j\left(t-1\right)}\right)+{\alpha }_3\ln \left({ENDOW}_{j\left(t-1\right)}\right)\hfill \\ & +{\alpha }_4\ln \left({MUR}_{j\left(t-1\right)}\right)+{\alpha }_5\ln \left({DEPEND}_{j\left(t-1\right)}\right)+{\alpha }_6\ln \left({DEPEND}_{j\left(t-1\right)}\right)\hfill \\ & +{\alpha }_7\ln \left({DTAX}_{j\left(t-1\right)}\right)+{\alpha }_8\ln \left({LABOR}_{j\left(t-1\right)}\right)+{\alpha }_9\ln \left({DIST}_{j\left(t-1\right)}\right)+{\sum }_n{\theta }_k{X}_{ijt}^k\hfill \\ & +\beta {\gamma }_{ijt}+{\lambda }_j+{\nu }_t+{\epsilon }_{ijt}\hfill \end{array}$$

(5)

where energy indicates the amount of China’s energy infrastructure investment in B&R countries. If ${energy}_{ijt}$ > 0, then ${energy}_{ijt}$ = 1; otherwise ${energy}_{ijt}$ = 0 (the data were from the China Global Investment Tracker). Gross National Product (GDP) for each country indicates the size of the host country market (constant 2010 USD); Gross National Product per capita (GDPP) measures the market purchasing power of a country’s residents (constant 2010 dollars); ENDOW indicates energy endowment which is measured using crude oil reserves, natural gas reserves and coal reserves (the data were from the EIA database); MUR indicates social security which is measured by the number of murders per 100,000 people per year; DEPEND represents energy dependence and is measured using the country’s coal, oil and gas consumption as a percentage of merchandise exports and imports. The above data were collected from the Worldwide Governance Indicators (WGI) database of the World Bank. DTAX means double taxation avoidance, which is set to 1 if China has signed a double taxation avoidance agreement with the host country, otherwise it is 0 (the data were from the State Administration of Taxation); LABOR indicates labor force abundance, and the total labor force of each country is used to reflect the size of the labor force (the data were derived from the World Development Indicators (WDI) database of the World Bank); DIST represents the geographical distance from China to the host country (the data were obtained from the CEPII database). $X_{ijt}^k$ is the key variable to be examined, and three indicators were selected to measure the political risk of the host country, namely, the quality of government regulation, the integrity of government, and the performance of environmental policies. According to the definition of political risk in the “Political Risk Assessment Report on Energy Resources Investment in the Belt and Road” released by Renmin University of China, political risk in a broad sense involves the interaction between politics, economy, society and ecological environment, and changes in any party will enhance uncertainty on energy infrastructure investment, increase social and ecological problems such as terrorism, corruption and climate change, and trigger political instability and change. In this paper, three different dimensions of political risk were selected as proxy variables. Data on the quality of government regulation were obtained from the WGI, those on government integrity from the Global Corruption Perception Index published by Transparency International, an international anti-corruption NGO, and those on environmental policy performance from the Environmental Performance Index provided by Yale University. The descriptive statistical results of variables are shown in

Table 1. Descriptive statistics.

Variable	Description	Mean	Standard Deviation	Minimum	Maximum
LENERGY	Amount of investment in energy infrastructure	6.72	1.11	4.61	9.13
LGDP	GDP	25.57	1.59	22.07	28.71
LGDPP	GDP per capita	8.36	1.21	6.39	10.98
ENDOW	Energy resource endowment	0.32	2.11	−5.44	4.05
MUR	Social environment	1.05	0.96	−1.85	3.21
DEPEND	Energy dependence	4.34	0.31	3.20	4.60
RQ	Government regulation quality	−0.41	0.83	−2.26	2.26
CPI	Government integrity	2.35	1.28	0.26	4.46
EPI	Environmental policy performance	4.00	0.26	3.22	4.46
DTAX	Avoidance of double taxation	0.82	0.37	0	1
LABOR	Labor endowment	16.37	1.36	13.87	20.01
DIST	Geographical distance	8.42	0.40	7.06	9.06
ENTCOST	Cost of entry	2.09	0.43	0.69	2.94

Note: obtained by the authors from the original data.

.

4. Empirical Analysis

4.1. Basic Regression

The results of the Heckman two-stage model estimation are reported in

Table 2. Basic regression results.

Variable	Political Risk: The Quality of Government Regulation		Political Risk: Government Integrity		Political Risk: Environmental Policy Performance
	Location Choice	Investment Scale	Location Choice	Investment Scale	Location Choice	Investment Scale
LGDP	1.887 ** (0.666)	0.289 (0.709)	1.586 ** (0.619)	0.141 (0.704)	1.341 ** (0.639)	0.405 (0.692)
LGDPP	−1.016 * (0.595)	−1.156 * (0.667)	−1.345 ** (0.583)	−1.028 (0.680)	−0.748 (0.614)	−1.338 * (0.697)
Energy and resource endowment	0.014 (0.130)	0.667 *** (0.155)	0.070 (0.123)	0.589 *** (0.147)	0.065 (0.130)	0.574 *** (0.152)
Social environment	0.159 (0.189)	0.546 ** (0.183)	0.191 (0.172)	0.499 ** (0.184)	0.042 (0.176)	0.514 ** (0.177)
Energy dependence	−0.016 (0.014)	−0.028 (0.015)	−0.005 (0.013)	−0.022 (0.015)	−0.003 (0.013)	−0.015 (0.014)
Institutional variable	−0.522 * (0.274)	−0.209 (0.299)	0.016 ** (0.007)	−0.003 (0.007)	−0.032 ** (0.009)	0.073 (0.574)
Avoidance of double taxation	−0.624 (0.571)	1.121 (0.677)	−0.598 (0.532)	1.026 (0.688)	−0.553 (0.548)	1.125 * (0.641)
Labor force abundance	−1.241 * (0.649)	−0.904 (0.774)	−1.161 * (0.626)	−0.684 (0.783)	−0.706 (0.654)	−0.860 (0.769)
Geographical distance	−1.092 ** (0.431)	1.304 ** (0.529)	−0.553 (0.373)	1.222 ** (0.503)	−0.778 ** (0.373)	1.124 ** (0.462)
Entry cost	−1.647 *** (0.464)		−0.998 ** (0.447)		−1.151 ** (0.409)
Lambda		4.259 ** (1.620)		3.272 ** (1.573)		3.384 * (1.839)
Constant	−4.903 (3.116)	12.045 ** (3.533)	−3.031 (2.891)	11.975 ** (3.528)	−4.797 (3.004)	9.401 ** (3.735)

Note: “***”, “**” and “*” show the results significant at the level of 1%, 5%, and 10%, respectively.

, where the three sections measure the political risk using the quality of government regulation, government integrity (the inverse measure of corruption) and environmental policy performance, respectively. The first column of each section reports the estimates of China’s investment location decisions for energy infrastructure in B&R countries, while the second column reports the estimates of investment scale decisions.

As can be seen from Table 2, the entry costs (ENTCOST) were all significantly negative, which is logical, and indicates a significantly lower probability of making energy infrastructure investments in countries (regions) with higher entry costs. Meanwhile, the inverse Mills ratios (lambda) were all significant at the 5% or 10% level, making it important to use the Heckman two-stage method.

We focused on the impact of core explanatory variables on energy infrastructure investments. Firstly, from the perspective of the market, the GDP of host countries was significantly positive in the first regression stage, but was not significant in the second stage, indicating that the market size affected whether Chinese enterprises enter the market, but did not affect the investment scale. This reveals that the characteristics of China’s pursuit of energy infrastructure investment market along the Belt and Road are reflected in location selection. GDP per capita significance results for host countries were not robust. This demonstrates that market opportunity and market potential of the target country had no obvious influence on China’s energy infrastructure investment, and hence were not important predictors. Secondly, in terms of energy endowment, the coefficients of energy dependence were not significant, and the energy endowment coefficient was only significantly positive in the scale of investment. Therefore, energy endowment affected the investment scale of Chinese enterprises rather than their location choice. Thirdly, regarding the labor force and social environment, the insignificant labor coefficient indicates China’s lack of significant labor-seeking motivation for energy infrastructure investment; however, the MUR coefficient was not significant in the phase of location selection, but was significantly positive in the phase of investment scale, indicating that China did not consider the local social environment in the location selection of energy infrastructure investment in B&R countries, and even invested more in countries with a higher degree of social risk. Fourthly, from the perspective of political risk, the quality of government regulation and environmental policy performance were significantly negative in the location choice stage, while the degree of government integrity was significantly positive in the stage. All three, however, were not significant in the investment scale stage. This suggests that local political risks were fully considered by Chinese enterprises when choosing the location of energy infrastructure investment in B&R countries. It can therefore be summarized that China tended to make energy infrastructure investment in countries with loose government regulation, poor environmental performance and high government integrity. The increase in political risk will influence the accuracy of enterprise forecast, which will directly affect the management, thereby raising the difficulty in the judgment of investment management on future investment location choice and policy situation. As a result, the unpredictable legal environment and the quality of government supervision will become important political risks to investment inflow, thus influencing the location and scale decision of enterprises [25]. Nonetheless, China has invested a huge amount in the energy infrastructure of B&R countries. According to the industry M&A data collated by Thomson Financial SDC Platinum Database, energy infrastructure investment ranks first in all industries. Countries with high-quality government regulation, strong awareness of environmental protection and high government integrity generally have more stringent access to and control of the energy industry, so companies may be forced to invest in countries with lower political risk.

4.2. Impact of the Belt and Road Initiative

We set the virtual variable BAR (the Belt and Road) to be 1 in the year after the BRI took effect; otherwise it was set to 0. In view of the previous finding that the BRI and political risk had a combined effect, we added the interaction term of BAR and political risk of the host country into the regression. The regression results are shown in

Table 3. Impact of the BRI on China’s energy infrastructure investment.

Variable	Political Risk: The Quality of Government Regulation		Political Risk: Government Integrity		Political Risk: Environmental Policy Performance
	Location Choice	Investment Scale	Location Choice	Investment Scale	Location Choice	Investment Scale
LGDP	2.396 ** (0.697)	0.077 (0.797)	1.580 ** (0.628)	0.160 (0.701)	1.152 * (0.652)	0.303 (0.757)
GDP per capita	−0.022 (0.014)	−0.021 (0.016)	−0.544 (0.376)	−0.751 (0.793)	−0.687 (0.623)	−1.236 (0.759)
Energy and resource endowment	0.022 (0.127)	0.599 ** (0.162)	0.071 (0.124)	0.465 ** (0.158)	0.103 (0.131)	0.581 ** (0.165)
Social environment	0.149 (0.128)	0.540 ** (0.197)	0.195 (0.176)	0.313 (0.201)	0.003 (0.179)	0.509 (0.180)
Energy dependence	−0.22 (0.014)	1.174 ** (0.547)	−0.544 (0.376)	−0.017 (0.015)	−0.004 (0.013)	−0.202 (0.017)
Institutional variable	−1.015 ** (0.331)	−0.032 (0.394)	0.015 * (0.008)	−0.008 (0.008)	−2.531 ** (0.731)	0.189 (0.578)
Avoidance of double taxation	−0.623 (0.571)	1.121 (0.676)	−0.588 (0.532)	1.026 (0.687)	−0.553 (0.548)	1.125 * (0.641)
Labor force abundance	−1.241 * (0.649)	−0.904 (0.774)	−1.162 * (0.636)	−0.684 (0.783)	−0.707 (0.654)	−0.860 (0.769)
Geographical distance	−1.092 ** (0.421)	1.314 ** (0.528)	−0.553 (0.363)	1.222 ** (0.503)	−0.778 ** (0.372)	1.124 ** (0.462)
Initiative	0.364 (0.331)	0.428 (0.337)	0.085 (0.833)	1.296 (0.805)	7.715 * (4.178)	−0.053 (0.448)
Interaction term	1.123 ** (0.416)	0.319 (0.457)	0.004 (0.021)	−0.022 (0.019)	2.007 * (1.033)	0.046 (0.104)
Entry cost	−1.877 *** (0.519)		−0.985 ** (0.450)		−0.689 * (0.464)
Lambda		3.236 * (1.684)		3.037 ** (1.565)		1.742 (1.694)
Constant	−5.592 * (3.047)	12.770 ** (3.703)	−3.032 (2.892)	11.216 ** (3.578)	2.460 (3.315)	12.503 ** (4.057)

Note: “***”, “**” and “*” show the results significant at the level of 1%, 5%, and 10%, respectively.

.

It can be seen from Table 3 that the interaction item of BAR and political risk of the host country was added into the regression. The coefficient significance of the BRI was not robust, but the interaction between BAR and government regulatory quality as well as that between BAR and environmental policy performance was significantly positive. This indicates that at the national level, with a certain degree of regulation and environmental policy performance, energy infrastructure investment would be more favorable after the implementation of the BRI compared with before.

4.3. Discussion of Empirical Results

In general, the “behavioral pattern” of China’s energy infrastructure investments in B&R countries was generally consistent with the outbound investment patterns summarized in the mainstream literature, with market, energy endowment, social environment and political risk all having an impact on energy infrastructure, although not in the same direction. Energy infrastructure differs in the following ways. Firstly, energy infrastructure investment is a matter of national lifelines, and investment agents are sensitive to political risk control, but lack certain long-term considerations. For example, China had a significant negative reflection on the quality of government regulation affecting current costs, but showed a clear risk appetite for the long-term social environment. Secondly, Chinese energy infrastructure investments, which were inferior to political risks in host countries, showed risk aversion to different dimensions of political risks. This suggests that China did not differentiate between different dimensions of political risk while making energy infrastructure investments, which contradicts the existing literature emphasizing that China’s outbound investments have heterogeneous preferences for political risk. This highlights the need to break down political risk to specific investment areas, and to differentiate between investment sectors when studying outbound investment behavior, rather than simply making generalizations. At the same time, “the Belt and Road” initiative, as an important starting point for China to expand its opening-up, helps to reduce the political risk of China’s investment in energy infrastructure in countries along the “the Belt and Road”, effectively safeguards the rights and interests of China and countries along the “Belt and Road”, enhances the breadth and depth of energy cooperation, and strengthens the implementation of high-quality development of energy cooperation between China and countries along the “Belt and Road”.

5. Conclusions and Policy Recommendations

Based on the Heckman two-stage selection model, this paper has investigated the impact of political risk on China’s energy infrastructure investment in B&R countries from two levels, location choice and scale decision, using the data on China’s energy infrastructure investment in B&R countries during 2005–2019. The basic findings are as follows. Firstly, energy endowment is a primary predictor of China’s energy infrastructure investment scale decision. As the energy industry is linked to the lifeline of a country, China is more sensitive to the political risks in the B&R host countries when choosing the location for energy infrastructure investment. It shows significant risk aversion to political risks such as government regulation, environmental rules and corruption which affect the immediate costs, but exhibits a risk preference for the social environment, which has long-term impacts. The latter implies that the BRI is more sensitive to political risks in the host country. Secondly, the interaction of the BRI with market regulations and environmental regulations can offset the negative impact of political risks and promote Chinese energy infrastructure investment.

With China’s investment in energy infrastructure in B&R countries continuining in full swing, the findings of this paper are helpful for understanding and responding to the challenges and obstacles faced in the process of energy infrastructure investment. For example, what are the main considerations for Chinese companies investing in B&R countries? This paper finds that, as summarized in the mainstream literature, energy infrastructure investments should take into account factors such as market, resource endowment, social environment and political risk. Although the direction of the impact of different political risk indicators on energy infrastructure investment is not entirely consistent, which contradicts previous empirical discoveries, this is related to the particular context of the BRI and is in line with economic intuition.

Current systematic theoretical study on the political risk of energy investment focuses primarily on qualitative descriptions or the development of indicators, with quantitative research still being absent. In the context of geopolitics of becoming a new risk for global energy cooperation, this paper aims to investigate the risk of China’s energy infrastructure investment in countries along the B&R from the standpoint of new institutional economics, and aims to incorporate political risk and the BRI into a comprehensive framework of research on the risk of energy investment, which is a useful addition to the existing literature, and also has an important practical application value.

Meanwhile, the conclusions of this paper have significant practical and policy implications. The global economy is currently under greater downward pressure, the global cross-border investment momentum is insufficient, the general trend of strengthening investment regulation in some countries continues, and Chinese enterprises’ participation in the global industrial division of labor and cooperation is limited. Furthermore, the pandemic, global inflation, the Ukraine crisis, and other hazards have made it more difficult for Chinese firms to secure their interests abroad, and Chinese enterprises face new hurdles in the areas of “soft power”, such as compliance management. China, on the other hand, will continue to engage in high-quality and long-term outbound investment cooperation. Regionally, the role of the countries’ along the B&R in China’s outward investment structure will be strengthened, particularly in ASEAN countries, such as Singapore, Indonesia, and Malaysia, which will be the focus of China’s future investment. From an industry standpoint, the new round of technological revolution and industrial change characterized by digitalization, intelligence and greening is accelerating, and China will increase its investment in 5G communications, wind energy, solar energy, energy-saving buildings, transportation facilities, and other infrastructure areas.

However, the imbalance between the global energy strategy, energy transformation and regional economy results in high energy investment, production and operation costs. The imperfect investment environment and frequent political risks require all parties of interest to rationalize their thinking in the energy cooperation between China and B&R countries. Resolving the frequent political risks and establishing a long-term political risk prevention mechanism are most critical at present.

First of all, risk assessment and screening of energy infrastructure investment in B&R countries should be performed well to reduce the probability of political risk emanating from the source. Due to the long cycle and large scale of cross-border energy projects, China and B&R countries need to further improve the cooperation framework to deal with security issues such as terrorism, separatism and extremism, and maintain a high level of attention to political issues including political democracy and anti-corruption in the host countries. Chinese energy companies should also effectively assess the economic situation, legal environment, environmental policies and social environment of the host countries and develop risk warning programs. At the same time, to cope with the complex international and national situations, Chinese energy companies should establish independent compliance management departments to deal with political risks such as national security reviews and environmental protection disputes within host countries, and aim for effective risk prevention, constant follow-up and fruitful learning experiences.

Secondly, strengthening the coordination of communication and collaboration mechanisms with B&R countries is the main means to prevent political risks. With economic cooperation as the core, the government will continue to highlight the pioneering role of energy infrastructure in the construction of the BRI. At the enterprise level, it is of vital significance to strengthen technology research and development, cultivate the core competitiveness of energy enterprises, and realize the transformation and improvement in enterprises in their participation in the construction of the initiative. Owing to the special nature of energy infrastructure investment, cooperation among state-owned enterprises can effectively improve the probability of successful investment. Therefore, we should actively guide the cooperation among state-owned enterprises, effectively enhance the enterprises’ risk prevention, comprehensively consider the political forces, ethnicity and religion of the host countries as well as the influence of NGOs, and establish extensive contact and regular communication mechanisms with them. In the meanwhile, it is important to respect the religious beliefs and customs of the host countries, consciously protect the legitimate rights and interests of laborers and the local environment, and reduce the risks of enterprise operation. At the social level, with the progress of the BRI, there is a growing need to cultivate management talents who are familiar with the language and culture, laws and regulations, and investment policies of B&R countries, as well as involve consultants capable of assessing risks and formulating strategies.

Thirdly, the energy transition can change the global distribution of power, interstate relations and the risk of conflict, thereby rebalancing countries and effectively resolving energy conflicts. With the goal of “carbon peaking and carbon neutrality”, China is supposed to accelerate the transformation of energy cooperation while promoting the construction of a low-carbon, clean, safe and efficient modern energy system; the bilateral cooperation should insist on shifting from traditional fossil energy exploitation to a low-carbon, clean-oriented energy economy industrial chain. In addition, the country should strive to seize the golden opportunity in the global energy competition, transforming the clean energy technology advantages into policy advantages, the technical and practical advantages in solar and wind energy into market advantages, and energy competition into energy cooperation to break the geopolitical influence. In so doing, its clean energy technology cooperation with B&R countries will be strengthened, while providing them with technical support and financial support; this will aid in effectively avoiding bilateral cooperation conflicts, promoting the safety and sustainability of bilateral energy cooperation, and achieving mutual benefit and a win-win situation.