International Studies of Economics
Volume 17, Issue 4 pp. 459-483
RESEARCH ARTICLE
Open Access

The Belt and Road Initiative, political involvement, and China's OFDI

Haoyuan Ding, 

Haoyuan Ding

College of Business, Shanghai University of Finance and Economics, Shanghai, China

Contribution: Funding acquisition, Resources, Writing - original draft

Search for more papers by this author
Yi Li, 

Yi Li

College of Business, Shanghai University of Finance and Economics, Shanghai, China

Contribution: Conceptualization, Formal analysis, Supervision, Validation

Search for more papers by this author
Liang Wang, 

Corresponding Author

Liang Wang

College of Business, Shanghai University of Finance and Economics, Shanghai, China

Correspondence Liang Wang, College of Business, Shanghai University of Finance and Economics, 777 Guoding Road, Shanghai 200433, China.

Email: [email protected]

Contribution: Data curation, ​Investigation, Methodology, Software

Search for more papers by this author
Chang Xue, 

Chang Xue

Economics and Management School, Wuhan University, Wuhan, China

Contribution: Project administration, Supervision, Writing - review & editing

Search for more papers by this author
First published: 19 July 2022
https://doi.org/10.1002/ise3.15
Citations: 4

Abstract

The Belt and Road Initiative (BRI) is an important strategy for China. This study examines the effect of political involvement on firms' outward foreign direct investment (OFDI) in belt-road countries after the BRI. Using merged Chinese nonfinancial listed firm data, the fDi Markets database, and the Thomson One database (formerly known as SDC Platinum) for the period 2008–2018, we find that political involvement has positive effect on firms' OFDI in belt-road countries after the BRI. Furthermore, we find that the positive effect is heterogeneous across state-owned enterprises (SOEs) and non-SOEs. Political involvement has a positive effect on M&A for SOEs and Greenfield investment for non-SOEs in belt-road countries after the BRI. Our findings suggest that political involvement promotes firms' OFDI in belt-road countries after the BRI and is helpful to the macro-policy implementation.

1 INTRODUCTION

The Belt and Road Initiative (BRI), started in 2013, led by China, is a global influential strategy of international economic cooperation that draws plenty of attention from economic scholars (Du & Zhang, 2018; Foo et al., 2020). BRI is one crucial component of President Xi “Major Country Diplomacy.” Since the development of the strategy is not merely market-oriented, political factors can be seen as important factors in determining the following-up strategies and economic outcomes of firms.

China is a socialist country led by the Communist Party of China (CPC), with a lot of branches of the CPC pervasively existing in the society, such as firms, schools and other units. These branches are responsible for “political leadership,” including the implementation of the party's policy, and can be regarded as one kind of political factors. In particular, the effect of the firm's party organization on agency cost, earnings management, export, and corporate social responsibility (CSR) have been examined by existing researchers (Ding et al., 2018; Li & Zhou, 2005). Given that the BRI is an important strategy for the CPC, the party organizations of firms are indeed obligated to implement and enforce the BRI. Accordingly, the effect of party organizations' intervention on firms' outward foreign direct investment (OFDI) in belt-road countries after the launch of BRI deserves further investigation. Accordingly, the effect of party organizations' intervention on firms' OFDI in belt-road countries after the launch of BRI deserves further investigation.

Our study focuses on the intervention of firms' party organizations of the CPC and refers to this type of political intervention as political involvement. Political involvement is defined as the proportion of party organizations' members in firms as directors, supervisors, or senior managers. Using merged Chinese nonfinancial listed firm data, the fDi Markets database and the Thomson One database for the period 2008–2018, we conduct a generalized difference-in-differences (GDID) regression to examine the effect of political involvement on firms' OFDI in belt-road countries after the BRI. Our findings suggest that political involvement has a positive effect on firms' OFDI in belt-road countries.

In terms of investment modes, the positive effect is heterogeneous across SOEs (Stated owned enterprises) and non-SOEs. Political involvement has positive effects on M&A for SOEs and Greenfield investment for non-SOEs in belt-road countries after the BRI. Our results hold in a variety of robustness checks. We also pay close attention to the role of local institutions and corporate governance. It is found that the positive effects of political involvement on SOEs' M&A and non-SOEs' greenfield investments in belt-road countries after the BRI are significant in the group of firms located in areas with worse market institutions. Meanwhile, we find that the positive effect of political involvement on SOEs' M&A in belt-road countries after the BRI is only significant in the group of firms with better corporate governance, but in terms of non-SOEs, political involvement significantly promotes the greenfield investments of non-SOEs, in both sample groups, including better and worse corporate governance, in belt-road countries after the BRI.

Our study is related to several strands of literature. First, this study aims to contribute to the literature by exploring the effects of political involvement on firms' participation in national strategy. The notion that political involvement over firms' decision making is detrimental to firm performance is widespread in the literature on corporate governance and public choice (Chang & Wong, 2004). Most theoretical arguments rely on the assumption that politicians bargaining with managers to promote firms pursue political and social objectives, and that these are detrimental to the firm performance (Boycko et al., 1996; Shleifer & Robert, 1994; Shleifer & Vishny, 1998). As mentioned by various studies, this negative effect on the firm performance can be seen as an additional political cost (Li, 2000; Qian, 1994, 1996; Xu et al., 2005). Thus, a trade-off occurs between political objective and preservation and/or value creation of companies. Cheng (2022) examines the causal effects of having a CPC branch on the coverage of labor contracts and social insurance among employees in private enterprises and finds positive effects. However, our study focuses on the effect of political involvement specifically by calculating the overlapping ratio between the members of firms' party organizations and firms' management team. Through using the sample of Chinese listed firms, including SOEs and non-SOEs, we explore the effect of political involvement on firms' participation in BRI and estimate the heterogeneity across these two kinds of companies.

Second, this study is related to the recently emerged literature on the BRI. The range of activities belonging to the BRI is very wide, including policy coordination, infrastructure, trade and investment and financial and people-to-people exchanges (De Soyres et al., 2019). De Soyres et al. (2019) focus on the consequences of transport infrastructures linked with the BRI and quantify the associated decrease in shipment times and trade costs. On the basis of this measurement of decrease in shipment times, Baniya et al. (2020) further quantify the potential trade effects of the BRI and find that it increases trade flows among participating countries. In term of investments, Du and Zhang (2018) find that the BRI promotes Chinese OFDI, especially M&A, in belt-road countries. They argue that the BRI strategy of massive investment in infrastructure would improve the quality and availability of logistics facilities in belt-road countries, which may boost foreign direct investment (FDI) inflows from China. Lu et al. (2020) find the implementation of the BRI significantly promote the growth of Chinese firms' greenfield investment, and the positive effect can be realized and explained through “five connectivities.” They explain the effects of BRI from the perspective of infrastructure or the five aspects of the initiative, which give high-level external support for economic activity. We contribute to these studies by examining the effect of political involvement, as a corporate governance factor, on firms' investment in belt-road countries.

Third, our study also contributes to the literature related to foreign market entry. Firms can choose among several modes of foreign market entry, including export, greenfield investment or M&A (Nocke & Yeaple, 2007). A large chunk of studies analyses the endogenous selection of firms' foreign market entry modes. Melitz (2003) analyses the decision of firms to enter foreign markets by becoming exporters. Helpman et al. (2004) consider only two modes of foreign market entry: exports and greenfield FDI. They find that only the most productive firms engage in foreign activities, and of those firms that serve foreign markets, only the most productive engage in FDI. Nocke and Yeaple (2007) introduce the “resource-based view of the firm” into the analysis and find that the decision of firms' foreign market access depends on the difference in their mobile or immobile capabilities. Researchers have also made substantial progress in understanding how current choices of foreign entry mode are influenced by past choices of the company and other companies (Ang et al., 2015; Chan & Makino, 2007; Lu, 2002; Shaver, 2013; Vermeulen & Barkema, 2001; Yiu & Makino, 2002). Basically, companies imitate their previous actions or those of their peers to achieve intracompany or external legitimacy (Hennart & Slangen, 2015). Thus far, however, few papers have considered the implication of corporate governance on the foreign market access of firms. Mariotti et al. (2021) claim that the interplay between the generation ruling the firm and the presence of nonfamily board members directs the choice between a greenfield investment versus the acquisition of a local unit. We contribute to this literature by analyzing the effect of political involvement on firms' decision for foreign market entry and considering the quality of the local institution and the level of corporate governance as moderating factors.

The rest of the study is structured as follows. Section 2 is the literature review and we develop a hypothesis, Section 3 describes the empirical strategy, sample coverage, and variable measurements; Section 4 presents the empirical results. The final section offers conclusions.

2 HYPOTHESIS DEVELOPMENT

In accordance with the CPC's basic line and basic strategy, the purpose of national macroeconomic policy is developing productivity and achieving social interests. Firms' behaviors, implementing the national strategies and policies, aim to increase social interests and fulfill the social responsibility beyond the goal of maximizing the profit of shareholders (McWilliams & Siegel, 2001). Specifically, the BRI is an important national strategy to actively optimize the allocation of resources, enhance the quality of economic developments, which obviously contributes to social interests, and is essentially consistent with typical CSR behaviors such as environmental protection and labor rights protection. The influence of the political involvement on firms' OFDI in belt-road countries after the BRI can be analyzed under the framework of CSR theory.

First, studies on determinants of CSR have shown that strong and effective government regulation requires companies to fulfill their social responsibility will lead to an increase in socially responsible behavior (Campbell, 2007). Thus, regulatory efforts exert an important effect on the fulfillment of social responsibility of enterprises. Political involvement is the core component of China's national governance system. According to the party constitution, the party organization is subject to the leadership of the corresponding higher organization and fulfills the responsibility of implementing the CPC's guidelines and policies. Political involvement is related to the strength and effectiveness of the supervision mechanism for the implementation of national macro policies in the corresponding enterprises. Enterprises with a higher degree of political involvement will fulfill their social responsibilities by trying to implement national macro policies due to they are subject to stronger supervision.

Second, spiritual and cultural factors such as beliefs and ideologies have an important influence on CSR behavior (Brammer et al., 2007), and the Party organization is the propagation ground of the Party's ideology, theory, and morality. According to the requirements of the party constitution and other regulations, the enterprise party organization has the obligation to carry out organizational activities for propaganda and education, so that party members can learn the party's theories, guidelines, and policies. This enables party members to learn more and carry forward the spirit of collectivism advocated by the party, such as “subordinate personal interests to the interests of the party and the people,” and then make the enterprise fully optimize its operation to safeguard the overall interests of society. At the same time, the Party organization's educational activities on specific policies can also deepen their internal recognition of macro policies, thus improving the enthusiasm of enterprises to implement policies. Moreover, the party organization is not isolated, but is embedded in each enterprise by the network of national party organization system, which has the effect of value introjection, and the activities of the party organizations embedded in enterprises will make enterprises internalize the party's values and ideas into their activities (Portes, 1998), and enterprises will have a stronger awareness to implement the CPC's policies. Therefore, the higher the degree of political involvement, the greater the impact of party organization's missionary activities on business operations will be, and the above analysis can be summarized in the following hypothesis.

Hypothesis: The political involvement has positive effect on firms’ OFDI in belt-road countries after the BRI.

3 EMPIRICAL STRATEGY AND DATA

3.1 Empirical specifications

To examine the effects of political involvement on firms' OFDI in belt-road countries after the announcement of BRI, we employ a GDID strategy to estimate the effect. This estimation strategy has also been employed in the literature on OFDI (Mariotti et al., 2021).

President Xi Jinping unveiled his vision of the Silk Road Economic Belt at Nazarbayev University on September 7, 2013, as part of his state visit to Kazakhstan. The concept of the New Maritime Silk Road was announced by President Xi before the Indonesian Parliament on October 3, 2013, as part of his state visit to Indonesia. These two concepts are combined as the BRI. Following some related literatures (Du & Zhang, 2018; Lyu et al., 2019), we set the year 2013 as the divider and treat the 6-year period from 2013 to 2018 as the post-shock or post-strategy period and the corresponding 5-year period from 2008 to 2012 as the pre-shock or pre-strategy period. The 6-year post-strategy period is sufficiently long to incorporate the changes in the OFDI of the Chinese firms in response to the announcement of the BRI national strategy.

Following Lin and Ye (2018), we specify the GDID regression model as follows:
()
where i and t denote firm and time, respectively; the dependent variable OFDI represents a firm's OFDI in belt-road countries; Par refers to a firm's political involvement; and Post is a dummy variable that takes the value of 1 if the time interval belongs to the post-shock period, and 0 otherwise. CVs refer to control variables, which aim to control other heterogeneous characteristics at the firm level. This study controls for state-owned enterprise (SOE), firm size (SIZE), leverage (LEV), firm age (AGE), return on asset (ROA), board size (BOARD), power concentration (BOTH), and the ratio of the largest shareholder control (TOP1). Furthermore, firm-specific fixed effects μi are added to the regression to account for time-invariant characteristics, and time-specific fixed effects θt are used to capture all time-variant macro-level factors that are common to firms. c refers to the constant term. As Post is in a time series at the national level and does not change across firms, it is absorbed by time-specific fixed effects. Specifically, to account for potential serial correlation and heteroskedasticity, we cluster standard errors at the firm level following the suggestion by Bertrand et al. (2004).

In Model (1), the roles of Post in a firm's OFDI in belt-road countries are associated with the political involvement of the firm. Here, the coefficient of interest is β1, which depicts the difference in the firm's OFDI in belt-road countries between low-level and high-level political involvement after the announcement of BRI (i.e., the GDID). β1 should be significantly positive because political involvement can promote firms' OFDI.

3.2 Sample coverage and data sources

We combine four data sources to examine the effects of political involvement on firms' OFDI in belt-road countries after the announcement of the BRI. The first data source is the China Stock Market Accounting Research (CSMAR) database, which contains detailed information about all Chinese listed firms' top management team (TMT) and board members as well as annual reports and firm financial information. The second data source is the fDi Markets database, which is the most comprehensive online database of cross-border greenfield investments in the market and provides detailed data on China's outbound greenfield investments, including the values of greenfield investments and the jobs created for host countries. The third data source is the Thomson One (formerly known as SDC Platinum) database, which is a widely used data source in cross-border M&A literature (Erel et al., 2012; Ferreira et al., 2010). The marketization index is obtained from Fan et al. (2019). More importantly, taking into account the missing information of corporate party organization membership in the curriculum vitae of TMT/board members from the “Profile of Directors and Senior Managers” obtained in the CSMAR database, we manually search such membership information for all listed companies through their related websites (e.g. related company websites, Sina Finance, Baidu, etc.).

We select all Chinese nonfinancial publicly listed companies listed on the Shanghai and Shenzhen exchanges from the period 2008–2018. In 2007, the split-share reform of listed companies was completed and the new accounting standard was implemented. Hence, choosing 2008 as the beginning avoids the need to account for statistical calibre and equity factors.

We obtain our regression sample for the empirical analysis through merging the above four databases by manually matching company names. The raw sample includes 28,051 firm-year observations. Following related previous research (Ding et al., 2018; Schmidt, 2015), we impose the following restrictions: (1) we delete firms from the financial and real estate industry; (2) we remove firms with missing TMT information, which leads to failure to calculate the measure of political involvement; (3) we exclude observations with insufficient data for calculating financial and corporate governance variables; and (4) we delete firms that were specially treated. Our final sample includes 24,247 firm years. This study winsorizes the continuous variables at the 1% and 99% levels to minimize outlier effects. Table A1 lists the definitions of all the variables, Table 1 reports the summary of statistics.

Table 1. Summary statistics
Variables N Mean SD Median Min Max
OFDI 24,247 0.011 0.104 0 0 1
GF_Times 24,247 0.014 0.152 0 0 4
GF_Value 24,247 4.270 137.758 0 0 17,800
MA_Times 24,247 0.002 0.040 0 0 1
MA_Amount 24,247 0.235 15.456 0 0 1890.937
Par 24,247 0.234 0.239 0.156 0 0.984
Par_Board 24,247 0.238 0.246 0.167 0 1
Par_Sup 24,247 0.243 0.300 0 0 1
Par_TMT 24,247 0.222 0.231 0.143 0 0.952
SOE 24,247 0.396 0.489 0 0 1
Size 24,247 21.933 1.308 21.775 13.763 28.509
Leverage 24,247 42.753 22.085 41.542 2.779 181.777
Age 24,247 2.658 0.446 2.708 0.000 3.932
ROA 24,247 0.034 0.371 0.038 −48.316 22.005
Board 24,247 2.146 0.202 2.197 1.099 2.890
Duality 24,247 0.263 0.440 0 0 1
Top1 24,247 0.353 0.150 0.334 0.022 0.900
  • Note: Variables are obtained from the firm-year-level data.

3.3 Political involvement

To construct a measure of political involvement at the firm level, we focus on all members of the TMT, board of directors, and board of supervisors, including chief executive officer (CEO), chief financial officer, and other top managers, as well as board members consisting of the chairman, executive board members, supervisors, and independent board members. By manually tracking their curriculum vitae, we can identify if a TMT or board member belongs to the Party Committee of the firm. If a TMT or a board member is currently a member of the firm's Party Committee, then they can be identified as an overlapping member. We then calculate separately the overlapping ratio of members between the firm's board of directors and Party Committee (Par_Board), the overlapping ratio of members between the firm's supervisory board (Par_Sup) and Party Committee and the overlapping ratio of members between firm's TMT and Party Committee (Par_TMT). Finally, we define Par, which equals the average of Par_Board, Par_Sup, and Par_TMT as the measurement of political involvement. This measurement method has also been employed in the literature on political involvement (Guo et al., 2019; Ma et al., 2012).

The summary statistics of Political involvement in Table 1 show that the mean of Par_Board is 0.238, which indicates that 23.8% of board members belong to the Party Committee. The mean of Par_Sup is 0.243, which indicates that 24.3% of supervisory board members belong to the Party Committee. The mean of Par_TMT is 0.222, which indicates that 22.2% of top management team members belong to the Party Committee.

3.4 Corporate governance

Following Zhou et al. (2020), we construct a comprehensive index of corporate governance using principal component analysis, which is conducted from various aspects, such as supervision, incentives, and so on. Particularly, we select the ratio of executive compensation to executive shareholding to indicate the incentive factor, the ratio of independent directors to the size of the board of directors to indicate the supervisory role of the board of directors, the ratio of institution-investors' shareholding and the cumulative ratio of shareholding from the second to the fifth largest shareholders to indicate the supervisory role of shareholders and a dummy variable equal to 1 if the CEO is also the chairman of the board to indicate the decision-making power of the CEO. We construct a comprehensive index of corporate governance using the principal component analysis based on these indicators. Then, we define the first principal component score as the corporate governance index. Finally, we adopt the sorting method to classify all firms into two groups, using the median value of the corporate governance index.

3.5 Marketization index

Consistent with Ang et al. (2015) and Zhang et al. (2016), we use the marketization indices constructed by Fan et al. (2019), who use the arithmetical average to generate an overall indicator and five secondary indicators to measure the degree of marketization in the province level. These indicators include the relationship between government and market, development of the non-state economy, marketization of product market, marketization of factor market and order of intermediary organizations and law. Note that Fan et al. (2019) only provide the marketization indices for each province over 2008–2016, the data over 2017–2018 cannot be directly obtained for our empirical research. To address this, we select the overall indicator as the measurement of marketization degree and adopt the historical average growth rate of the overall indicator to forecast the indices over 2017–2018 in the province level. We then adopt the sorting method to classify all provinces into two groups, using the median value of the marketization indices. Provinces whose marketization is above the median are defined to exhibit high-level marketization. Finally, we manually match marketization indices to the listed firm according to the registered address of the firms in the province level and the year.

3.6 Other variables

Following the previous related literature (An et al., 2016; Gulen & Ion, 2016; Kim & Kung, 2017), we define the measurement of other variables at the firm level as follows: SOE is a dummy variable that is equal to 1 for an SOE, and 0 otherwise; Size measured by the natural logarithm of total assets; Leverage is defined as the ratio of total liabilities divided by total assets; Age is measured by the interval starting from the year when a firm established until the current year; ROA measured by the ratio of net profits relative to total assets; Board measured by the natural logarithm of the number of board members and Duality is a dummy variable that equal to 1 if CEO is the chair, and 0 otherwise. Top1 is defined as the largest shareholder' control for the firm and is measured by percentage of shares held by the largest shareholder.

4 EMPIRICAL RESULTS

4.1 Benchmark regression results

We conduct GDID regression to examine the effects of political involvement on firms' OFDI in belt-road countries after the BRI. Table 2 reports the results of the benchmark analysis. Column (1) is the main specification based on the full sample. We also differentiate the effect of political involvement depending on whether the companies are state-owned. Columns (2)–(5) report the results for the SOE subsamples while Columns (6)–(9) report the results for non-SOE subsamples.

Table 2. Benchmark results
All SOEs Non-SOEs
(1) (2) (3) (4) (5) (6) (7) (8) (9)
OFDI GF_Times GF_Value MA_Times MA_Value GF_Times GF_Value MA_Times MA_Value
Par × Post 0.012 0.002 −0.893 0.007 0.629 0.008 3.780 −0.003 −2.847
(0.005) (0.018) (6.251) (0.004) (0.303) (0.003) (1.314) (0.004) (2.357)
Par −0.011 0.003 −19.779 −0.002 1.045 −0.022 −3.176 0.004 2.930
(0.006) (0.022) (12.123) (0.001) (0.575) (0.005) (0.533) (0.003) (1.795)
SOE × Post −0.004
(0.003)
SOE −0.002
(0.004)
Size −0.004 0.005 8.780 0.001 −0.463 0.004 −0.100 −0.000 −0.207
(0.003) (0.004) (8.800) (0.001) (0.265) (0.001) (0.372) (0.000) (0.199)
Leverage −0.002 −0.000 −0.163 −0.000 −0.024 0.000 0.010 0.000 0.023
(0.004) (0.000) (0.192) (0.000) (0.013) (0.000) (0.004) (0.000) (0.017)
Age 0.004 0.050 99.067 −0.003 0.234 −0.004 −1.280 −0.004 −0.993
(0.001) (0.030) (92.640) (0.002) (0.184) (0.001) (0.499) (0.002) (0.305)
ROA 0.000 0.011 −0.891 −0.000 −0.789 −0.000 0.017 0.000 0.017
(0.000) (0.012) (4.608) (0.000) (0.795) (0.000) (0.018) (0.000) (0.014)
Board 0.001 −0.019 −2.588 −0.004 −0.571 0.003 −7.040 0.004 0.805
(0.009) (0.020) (4.805) (0.002) (0.296) (0.003) (1.824) (0.001) (0.991)
Duality 0.000 −0.015 −2.969 0.001 −1.234 0.002 0.593 −0.001 −0.158
(0.000) (0.007) (1.608) (0.001) (0.532) (0.001) (0.242) (0.000) (0.251)
Top1 −0.003 −0.031 −38.240 −0.002 −0.555 −0.015 −2.596 −0.004 −3.245
(0.005) (0.020) (55.646) (0.001) (0.643) (0.009) (4.350) (0.002) (1.157)
Constant −0.057 −0.129 −377.586 0.001 12.065 −0.058 19.719 0.003 4.474
(0.025) (0.105) (360.469) (0.002) (6.727) (0.029) (10.944) (0.005) (1.764)
Obs 24,247 9598 9598 9598 9598 14649 14,649 14,649 14,649
R2 0.003 0.005 0.004 0.002 0.002 0.002 0.002 0.002 0.002
Time FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
  • Note: Robust standard errors are given in parentheses.
  • *** p < 0.01
  • ** p < 0.05
  • * p < 0.1.

In Column (1) of Table 2, the estimated coefficient of Par × Post is positive and significant at the 5% level, meaning political involvement has a positive effect on firms' OFDI in belt-road countries after the BRI. For SOEs, the estimated coefficient of Par × Post is positive and significant in Columns (4)–(5), which indicates political involvement has a positive impact on SOEs' M&A in belt-road countries after the BRI. For non-SOEs, the estimated coefficient of Par × Post is positive and significant in Columns (6)–(7), indicating that political involvement has positive effect on non-SOEs' Greenfield investment in belt-road countries after the BRI.

Political involvement plays a critical role in corporate decision. Party organizations' members in firms must fulfil duties of monitoring enterprises to comply with policies and regulation rules, helping companies better understand national policies (Ang & Boyer, 2007) and following the call of the government. After the implementation of the “two-way entry, cross-appointment” scheme, members of the Party Committee could directly enter the board of directors, board of supervisors and management team (Guo et al., 2019). They can directly participate in business decisions. Thus, the members of firms' party organizations directly involved in management have duties and capacities to promote their firms' OFDI in belt-road countries after the BRI.

Many political parties also exert tremendous effort imposing their ideological propositions and governance principles on their members and supporters to improve their identification through cultural education and propaganda (He & Jiang, 2020). Thus, party members carry forward the spirit of collectivism advocated and learn more about international political cooperation, policy coordination, and government support embedded in BRI. In this sense, party members directly involved in company management have an intrinsic motivation to promote their company's OFDI in belt-road countries after the BRI.

Compared to greenfield investments, M&A investment transactions can be more speedily executed (Du & Zhang, 2018) and allow a firm to get access to the country-specific capabilities of the acquired firm (Nocke & Yeaple, 2007). In addition, the BRI is an infrastructure-led paramount national strategy announced by the Chinese government. Given that SOEs play a pivotal role in the commanding heights of the national economy and are reliable forces to achieve government goals (Nocke & Yeaple, 2007), they are expected to play a leading part in investments in infrastructure sectors. By engaging in greenfield investments, a firm will need to construct new production facilities, which are costly and risky for infrastructure sectors. Consequently, political involvement has a positive impact on M&A of SOEs in belt-road countries after the BRI.

Infrastructure development, such as the quality of and access to transportations and telecommunication networks, plays a crucial role in attracting FDI (Bellak et al., 2009; Cheng & Kwan, 2000; Coughlin et al., 1991). The massive state-led OFDI in infrastructure sectors can generate related investment opportunities for non-SOEs to play a complementary role in non-infrastructure-related industries, such as agricultural, finance, technology and business services. Compared to infrastructure sectors, these industries generally suffer less costly construction of new production facilities. Furthermore, high-level international political cooperation, policy coordination, and government support embedded in the BRI can considerably reduce policy uncertainty of the host country and political risks for Chinese firms investing in belt-road countries (Du & Zhang, 2018), which further reduce the cost of searching information and adapting to the local environment for greenfield investors. However, M&A suffers from internal conformity costs and difficulties of the transfer and integration of the competitive advantage (Vaara et al., 2012). Consequently, political involvement has positive effect on Greenfield investment of non-SOEs in belt-road countries after the BRI.

4.2 Robustness checks

In this subsection, we conduct a variety of sensitivity analyses to check the robustness of our benchmark regression results. First, we check whether our results are robust to alternative measures of political involvement. We employ Par_Board, Par_Sup and Par_TMT as alternative measures respectively. The estimation results obtained from using these measures of involvement are shown in Table 3. Using alternative measures of political involvement do not alter our benchmark regression findings.

Table 3. Robustness checks: Alternative measures of political involvement
Non-SOEs SOEs
(1) (2) (3) (4) (5) (6)
GF_Value GF_Value GF_Value MA_Value MA_Value MA_Value
Par_Board × Post 7.534 0.721
(2.462) (0.295)
Par_Board −5.440 1.497
(1.414) (0.756)
Par_Sup × Post 0.298 0.286
(0.384) (0.185)
Par_Sup 0.134 −0.016
(0.507) (0.058)
Par_TMT × Post 2.111 0.738
(0.839) (0.340)
Par_TMT −3.424 1.303
(0.736) (0.709)
Constant 20.691 19.219 19.941 12.189 11.843 12.158
(11.319) (10.873) (11.121) (6.748) (6.595) (6.750)
Obs 14,649 14,649 14,649 9598 9598 9598
R2 0.002 0.002 0.002 0.002 0.002 0.002
Controls Yes Yes Yes Yes Yes Yes
Time FE Yes Yes Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes Yes Yes
  • Note: Robust standard errors are given in parentheses.
  • *** p < 0.01
  • ** p < 0.05
  • * p < 0.1.

Second, some of the sample firms may also be listed on other stock markets, for example, Hong Kong and United States, which give them external sources of finance. To make sure that the positive effects of political involvement on firm's OFDI in belt-road countries after the BRI are not driven by those external sources of finance. We introduce a dummy variable, denoted Crosslist, to indicate whether a listed firm also listed on Hong Kong or United States. Then, we include this dummy variable as additional control. The results shown in Table 4 suggest that our benchmark regression findings are not driven by external sources of finance.

Table 4. Robustness checks: Additional control of cross-listing
All SOEs Non-SOEs
(1) (2) (3) (4) (5) (6) (7) (8) (9)
OFDI GF_Times GF_Value MA_Times MA_Value GF_Times GF_Value MA_Times MA_Value
Par × Post 0.012 0.002 −0.840 0.007 0.629 0.008 3.779 −0.003 −2.743
(0.005) (0.018) (6.252) (0.004) (0.304) (0.003) (1.313) (0.005) (2.381)
Par −0.011 0.003 −19.802 −0.002 1.045 −0.022 −3.175 0.003 2.804
(0.006) (0.022) (12.136) (0.001) (0.575) (0.005) (0.532) (0.003) (1.823)
SOE × Post −0.004
(0.003)
SOE −0.002
(0.004)
Size 0.004 0.005 8.792 0.001 −0.463 0.004 −0.100 −0.000 −0.214
(0.001) (0.004) (8.818) (0.001) (0.265) (0.001) (0.372) (0.000) (0.186)
Leverage 0.000 −0.000 −0.164 −0.000 −0.024 0.000 0.010 0.000 0.023
(0.000) (0.000) (0.192) (0.000) (0.013) (0.000) (0.004) (0.000) (0.017)
Age 0.001 0.050 99.388 −0.003 0.234 −0.003 −1.280 −0.004 −0.856
(0.009) (0.030) (92.762) (0.002) (0.189) (0.001) (0.499) (0.002) (0.318)
ROA 0.000 0.011 −0.901 −0.000 −0.789 −0.000 0.017 0.000 0.021
(0.000) (0.012) (4.608) (0.000) (0.795) (0.000) (0.018) (0.000) (0.018)
Board −0.003 −0.019 −2.604 −0.004 −0.571 0.002 −7.039 0.004 0.229
(0.005) (0.020) (4.826) (0.002) (0.296) (0.003) (1.823) (0.001) (1.060)
Duality 0.001 −0.016 −2.949 0.001 −1.234 0.002 0.593 −0.001 −0.159
(0.000) (0.007) (1.603) (0.001) (0.532) (0.001) (0.242) (0.000) (0.252)
Top1 −0.000 −0.031 −38.398 −0.002 −0.555 −0.015 −2.596 −0.004 −3.232
(0.008) (0.020) (55.768) (0.001) (0.645) (0.009) (4.351) (0.002) (1.170)
Crosslist 0.009 0.012 −10.264 −0.000 −0.016 0.056 −0.115 0.033 37.160
(0.003) (0.006) (9.201) (0.000) (0.221) (0.017) (0.417) (0.010) (11.147)
Constant −0.057 −0.129 −377.322 0.001 12.066 −0.057 19.718 0.003 4.996
(0.025) (0.105) (360.255) (0.002) (6.731) (0.028) (10.939) (0.004) (1.922)
Obs 24,247 9598 9598 9598 9598 14,649 14,649 14,649 14,649
R2 0.003 0.005 0.004 0.002 0.002 0.002 0.002 0.002 0.010
Time FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
  • Note: Robust standard errors are given in parentheses.
  • *** p < 0.01
  • ** p < 0.05
  • * p < 0.1.

Considering BRI is a massive infrastructure-led economic integration plan (Du & Zhang, 2018), firm's OFDI in belt-road countries after the initiative is likely to be industry specific. Thus we define industry dummy including Manufacture (i.e., equal to 1 if firm is belongs to the manufacture industry, and equal 0 otherwise), Business (i.e., equal to 1 if firm is belongs to the business industry, and equal 0 otherwise), Public (i.e., equal to 1 if firm is belongs to the public utility industry, and equal 0 otherwise) and Composite (i.e., equal to 1 if firm is belongs to the composite industry, and equal 0 otherwise). Then, we add the triple-interaction terms including Par × Post × Manufacture, Par × Post × Business, Par × Post × Public, Par × Post × Composite, respectively into the benchmark regression. The results shown in Table 5 indicate that the positive effects of political involvement on firm's OFDI in belt-road countries after the BRI are not driven by specific industry.

Table 5. Robustness checks: The role of specific industry
(1) (2) (3) (4)
Manufacture Business Public Composite
Par × Post 0.012 0.012 0.012 0.013
(0.006) (0.005) (0.005) (0.005)
Par × Post × Manufacture −0.000
(0.009)
Par × Post × Business −0.001
(0.010)
Par × Post × Public 0.003
(0.011)
Par × Post × Composite −0.014
(0.009)
Par −0.011 −0.011 −0.011 −0.011
(0.006) (0.006) (0.006) (0.006)
SOE × Post −0.004 −0.004 −0.004 −0.004
(0.003) (0.003) (0.003) (0.003)
SOE −0.002 −0.002 −0.002 −0.002
(0.004) (0.004) (0.004) (0.004)
Size 0.004 0.004 0.004 0.004
(0.001) (0.001) (0.001) (0.001)
Leverage 0.000 0.000 0.000 0.000
(0.000) (0.000) (0.000) (0.000)
Age 0.001 0.001 0.001 0.001
(0.009) (0.009) (0.009) (0.009)
ROA 0.000 0.000 0.000 0.000
(0.000) (0.000) (0.000) (0.000)
Board −0.003 −0.003 −0.003 −0.003
(0.005) (0.005) (0.005) (0.005)
Duality 0.001 0.001 0.001 0.001
(0.000) (0.000) (0.000) (0.000)
Top1 −0.000 −0.000 −0.000 −0.000
(0.008) (0.008) (0.008) (0.008)
Constant −0.057 −0.057 −0.057 −0.057
(0.025) (0.025) (0.025) (0.025)
Obs 24,247 24,247 24,247 24,247
R2 0.003 0.003 0.003 0.003
Time FE Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes
  • Note: Robust standard errors are given in parentheses.
  • *** p < 0.01
  • ** p < 0.05
  • * p < 0.1.

4.3 The role of the local institution

We further explore the potential heterogeneity across local institutions. To do so, we split each subsample (SOEs and Non-SOEs) into two groups, a better market institution (MKT Institution better) group and a worse market institution (MKT Institution worse) group, which includes firms located in a province with a lower quality of market institution. Columns (1) and (2) of Table 6 show the estimation results for SOEs. The positive effect of political involvement on SOEs' M&A in belt-road countries after the BRI is only significant in the sample group of firms located in areas with worse market institutions. Columns (3) and (4) show the estimation results for non-SOEs. Similarly, the positive effect of political involvement on non-SOEs' greenfield investments in belt-road countries after the BRI is only significant in the sample group of firms located in areas with worse market institutions. It can be inferred that the positive effects of the political involvement are significant in areas with worse institutions.

Table 6. The role of the local institution
SOEs Non-SOEs
(1) (2) (3) (4)
Better MKT institution Worse MKT institution Better MKT institution Worse MKT institution
MA_Value MA_Times GF_Value GF_Times
Par × Post 0.141 1.372 0.316 6.954
(0.083) (0.595) (0.326) (1.760)
Par −0.146 1.503 −3.416 −3.507
(0.087) (0.817) (0.468) (0.640)
Obs 3192 6406 7539 7110
R2 0.005 0.003 0.005 0.002
Controls Yes Yes Yes Yes
Time FE Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes
  • Note: Robust standard errors are given in parentheses.
  • *** p < 0.01
  • ** p < 0.05
  • * p < 0.1.

As to the causes, the lower the local marketization level, the less mature the factor-market and product-market and the less improved the legal and financial mechanisms (Zheng & Sheng, 2017). Therefore, by relying on their political capital, the members of firms' party organizations play a more important role in helping businesses reduce regulatory burdens, lower fees and taxes and gain easier access to loans, licenses and permits (Chen et al., 2017; Guo et al., 2014; Li et al., 2008; Liu et al., 2013), which further improve firm's investments in belt-road countries. The institutional environment also directly influences the quality of information available in the public domain (Piotroski & Wong, 2012), which influences the investment decision of the firm. The identity advantages of party organization members allow them get high-quality information about the BRI easily. Thus, for those firms located in worse institution provinces, their investments to belt-road countries are significantly promoted by political involvement.

4.4 The role of corporate governance

In this subsection, we explore the potential heterogeneity across corporate governance. Similarly, we split each subsample (SOEs and Non-SOEs) into two groups, a better corporate governance group and a worse corporate governance group.

Columns (1) and (2) of Table 7 show the estimation results for SOE subsamples. The positive effect of political involvement on SOEs' M&A in belt-road countries after the BRI is only significant in the sample group of firms with worse corporate governance. The result could be explained by the fact that the main shareholder of SOEs is the state, and it is difficult for the state per se to monitor the managers of SOEs, a phenomenon also known as ‘owner absence' (Guo et al., 2019). Compared with state shareholders, managers share fewer of the political objectives of local party committees (Chang & Wong, 2004) and have less motivation to follow the call of the government in fulfilling various national or local goals. As a consequence, the backdrop of worse corporate governance causes managers to invest less in belt-road countries than state shareholders would like them to. The involvement of local party committees leads SOEs to focus more on social or political goals instead of economic profit (Guo et al., 2019). Therefore, political involvement promotes firms' investments in belt-road countries.

Table 7. The role of corporate governance
SOEs Non-SOEs
(1) (2) (3) (4)
Better corporate governance Worse corporate governance Better corporate governance Worse corporate governance
MA_Value MA_Times GF_Value GF_Times
Par × Post 0.025 0.536 1.288 9.762
(0.014) (0.298) (0.596) (2.239)
Par −0.000 1.982 −2.675 −2.407
(0.003) (1.122) (0.533) (0.836)
Obs 2678 6920 10,299 4350
R2 0.006 0.003 0.002 0.004
Controls Yes Yes Yes Yes
Time FE Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes
  • Note: Robust standard errors are given in parentheses.
  • *** p < 0.01
  • ** p < 0.05
  • * p < 0.1.

Columns (3) and (4) show the estimation results for non-SOEs subsamples. the positive effect of political involvement on non-SOEs' greenfield investments in belt-road countries after the BRI is simultaneously significant in the sample group of better corporate governance and the sample group of worse corporate governance. It can be inferred that the positive effects of political involvement on firms' investment are not affected by corporate governance. As to the causes, in the early stage of the BRI, some of motivating benefits for OFDI in belt-road countries were still largely expected ones instead of realized ones (Du and Zhang, 2018). Non-SOEs have low confidence in the success of the BRI. Since CPC members have more information about the BRI, they have more incentives to invest in belt-road countries. Thus, political involvement significantly promotes the greenfield investments of non-SOEs, in both sample groups, including better and worse corporate governance, in belt-road countries after the BRI.

5 CONCLUSIONS

Previous studies have consistently found that as an important means for the governance of firms, political involvement can promote firms' donations and reduce agency cost. However, whether political involvement can promote firms' OFDI in belt-road countries after the BRI has not been determined. We use data for Chinese nonfinancial private listed companies from 2008 to 2018 to examine the effect of political involvement on firms' OFDI in belt-road countries after the BRI.

Estimation results show that political involvement has a positive effect on firm's OFDI in belt-road countries after the BRI. In terms of investment modes, the positive effect is heterogeneous across SOEs and non-SOEs. Political involvement has a positive effect on M&A for SOEs and Greenfield investment for non-SOEs in belt-road countries after the BRI. We also pay close attention to the role of local institutions and corporate governance. It is found that the positive effects of political involvement on SOEs' M&A and non-SOEs' greenfield investments in belt-road countries after the BRI are significant in the group of firms located in areas with worse market institutions. Meanwhile, we find that the positive effect of political involvement on SOEs' M&A in belt-road countries after the BRI is only significant in the group of firms with better corporate governance, but in terms of non-SOEs, political involvement significantly promotes the greenfield investments of non-SOEs, in both sample groups, including better and worse corporate governance, in belt-road countries after the BRI.

This study argues that beyond the traditionally acknowledged effects of reducing agency cost and improving operation performance, political involvement also can promote firms' OFDI in belt-road countries after the BRI, which helps firms implement the macro-policy and fulfil CSR. Thus, policymakers should be care of the function of political involvement and consider that how make full use of the political involvement channel and achieve the maximum of social interest.

AUTHOR CONTRIBUTIONS

Haoyuan Ding: funding acquisition; resources; writing – original draft. Yi Li: conceptualization; formal analysis; supervision; validation. Liang Wang: data curation; investigation; methodology; software. Chang Xue: project administration; supervision; writing – review & editing.

ACKNOWLEDGMENTS

This study was supported by the National Natural Science Foundation of China (Nos. 72173082 and 71703086] and 2021 Postgraduate Innovation Funds for the Shanghai University of Finance and Economics (No. CXJJ-2021-404).

    CONFLICT OF INTEREST

    The authors declare no conflict of interest.

    ETHICS STATEMENT

    None declared.

    APPENDIX A

    Table A1. Variable definitions
    Variables Definitions
    Dependent variables
    OFDI A dummy takes a value of 1 if a firm has OFDI in belt-road countries, and 0 otherwise.
    MA_Times Number of times a firm carries out M&A in belt-road countries
    MA_Value Value of a firm carrying out M&A in belt-road countries
    GF_Times Number of times a firm carries out greenfield investment in belt-road countries
    GF_Value Value of a firm carrying out greenfield investment in belt-road countries
    Explanatory variables
    Par_Board Overlapping ratio of members between a firm's board of directors and Party Committee
    Par_Sup Overlapping ratio of members between a firm's board of supervisors and Party Committee
    Par_TMT Overlapping ratio of members between a firm's TMT and Party Committee
    Par (Par_Board + Par_Sup + Par_TMT)/3
    Post A dummy variable that equals 1 if the year is after 2013, and 0 otherwise.
    Control variables
    SOE A dummy variable that equal to 1 for an SOE, and 0 otherwise
    Size Natural logarithm of total assets
    Leverage Ratio of total liabilities divided by total assets
    Age Number of years after the firm's establishment
    ROA Ratio of net profits relative to total assets
    Board Natural logarithm of the number of board of directors of the firm
    Duality A dummy variable that equal to 1 if CEO is the chair, and 0 otherwise
    Top1 Number of shares held by the company's largest shareholder/total number of shares; %
    Table A2. A comparison of political involvement between SOEs and non-SOEs
    SOEs non-SOEs t-Test
    Observations Mean Observations Mean Difference T-value
    Par 9598 0.388 14,649 0.134 −0.255 −95.05
    Par_Board 9598 0.378 14,649 0.146 −0.232 −80.8
    Par_Sup 9598 0.424 14,649 0.124 −0.299 −87
    Par_TMT 9598 0.362 14,649 0.130 −0.234 −88.5
    • *** p < 0.01.
    Table A3. Results based on “2014-Divider”
    All SOEs Non-SOEs
    (1) (2) (3) (4) (5) (6) (7) (8) (9)
    OFDI GF_Times GF_Value MA_Times MA_Value GF_Times GF_Value MA_Times MA_Value
    Par × Post 0.010 0.009 1.459 0.006 0.584 −0.002 3.295 −0.004 −2.473
    (0.004) (0.018) (5.851) (0.003) (0.264) (0.002) (1.323) (0.003) (1.666)
    Par −0.009 0.000 −21.041 −0.001 1.120 −0.016 −2.593 0.004 2.486
    (0.006) (0.020) (13.276) (0.001) (0.620) (0.004) (0.406) (0.002) (1.234)
    Constant −0.060 −0.128 −377.332 0.001 12.014 −0.059 19.535 0.003 4.613
    (0.024) (0.105) (359.742) (0.002) (6.714) (0.029) (10.868) (0.005) (1.873)
    Obs 24,247 9598 9598 9598 9598 14,649 14,649 14,649 14,649
    R2 0.003 0.005 0.004 0.002 0.002 0.002 0.002 0.002 0.002
    Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes
    Time FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
    Firm FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
    • Note: Robust standard errors are given in parentheses.
    • *** p < 0.01
    • ** p < 0.05
    • * p < 0.1.
    Table A4. Results based on “2015-Divider”
    All SOEs Non-SOEs
    (1) (2) (3) (4) (5) (6) (7) (8) (9)
    OFDI GF_Times GF_Value MA_Times MA_Value GF_Times GF_Value MA_Times MA_Value
    Par × Post 0.006 0.011 −7.218 0.005 0.685 −0.008 2.733 −0.003 −2.141
    (0.003) (0.017) (5.214) (0.002) (0.325) (0.003) (1.307) (0.002) (1.553)
    Par −0.007 −0.000 −17.255 −0.001 1.126 −0.014 −2.068 0.003 2.131
    (0.005) (0.018) (10.818) (0.000) (0.608) (0.002) (0.256) (0.002) (1.390)
    Constant −0.069 −0.128 −377.911 0.001 12.012 −0.059 19.467 0.003 4.663
    (0.023) (0.105) (361.451) (0.002) (6.712) (0.029) (10.855) (0.005) (4.946)
    Obs 24,247 9598 9598 9598 9598 14,649 14,649 14,649 14,649
    R2 0.003 0.004 0.002 0.001 0.001 0.001 0.001 0.001 0.001
    Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes
    Time FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
    Firm FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
    • Note: Robust standard errors are given in parentheses.
    • *** p < 0.01
    • ** p < 0.05
    • * p < 0.1.
    Table A5. Test for nonlinear relationship
    (1) (2)
    OFDI OFDI
    Par −0.004 −0.017
    (0.005) (0.010)
    Par × Par 0.017
    (0.017)
    SOE −0.004 −0.004
    (0.003) (0.003)
    Size 0.004 0.004
    (0.001) (0.001)
    Leverage 0.000 0.000
    (0.000) (0.000)
    Age 0.002 0.002
    (0.008) (0.008)
    ROA 0.000 0.000
    (0.000) (0.000)
    Board −0.003 −0.003
    (0.005) (0.005)
    Duality 0.001 0.001
    (0.000) (0.000)
    Top1 −0.000 −0.000
    (0.009) (0.009)
    Crosslist 0.009 0.009
    (0.003) (0.003)
    Constant −0.059 −0.058
    (0.022) (0.023)
    Obs 24,247 24,247
    R2 0.003 0.003
    Time FE Yes Yes
    Firm FE Yes Yes
    • Note: Robust standard errors are given in parentheses.
    • *** p < 0.01
    • ** p < 0.05
    • * p < 0.1.

    • 1 President Xi Jinping unveiled his vision of the Silk Road Economic Belt at Nazarbayev University on September 7, 2013 as part of his state visit to Kazakhstan. The concept of New Maritime Silk Road was announced by President Xi before the Indonesian Parliament on October 3, 2013, as part of his state visit to Indonesia. These two concepts are combined as the Belt and Road Initiative (BRI).
    • 2 By the end of 2013, countries covered by the BRI (i.e. belt-road countries) included Mongolia, Singapore, Malaysia, Indonesia, Burma, Myanmar, Thailand, Lao, Cambodia, Vietnamese, Brunei, Philippines, Iran, Iraq, Turkey, Syria, Jordan, Lebanon, Israel, Palestine, Saudi Arabia, Yemen, Oman, Arab Emirates, Qatar, Kuwait, Bahrain, Greece, Cyprus, Sinai, India, Pakistan, Bangladesh, Afghanistan, Sri Lanka, Maldives, Nepal, and Bhutan. Source: BELT AND ROAD PORTAL.
    • 3 We also report the difference of political involvement between SOEs and Non-SOE in Table A2. The degree of the political involvement of SOEs is significantly higher than non-SOEs. Specifically, the difference between the degree of the political involvement of SOEs and non-SOEs is 25.5% in terms of Par.
    • 4 Considering that the response time of Chinese firms to the announcement of the BRI is long, we reset “Divider” as 2014 and 2015 respectively and conducted the benchmark regression, which has proved the robustness of our benchmark regression results (Tables A3 and A4). We also examine the nonlinear relationship between Par and OFDI and find no evidence to support it (Table A5).

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.