China’s Belt and Road Initiative and large-scale outbound investment

Abstract

Purpose

The purpose of this paper is to comprehensively examine the influence of formal and informal institutional differences on enterprise investment margin, mode and result.

Design/methodology/approach

This paper is based on 2,440 micro samples of large-scale outbound investment from 609 Chinese enterprises from the years 2005 to 2016.

Findings

The study has found that formal institutional differences have little impact on investment scale, but significantly affect investment diversification. In order to avoid the management risks brought by formal institutional differences, enterprises tend to a full ownership structure. However, the choice between greenfield investment and cross-border mergers and acquisitions is not affected by formal institutional differences. In contrast, the impact of informal institutional differences is more extensive. Both formal and informal institutional differences significantly increase the probability of investment failure. Further research found that the Belt and Road Initiative (BRI) bridges the formal institutional differences.

Originality/value

The study concludes that developing the BRI, especially cultural exchanges with countries alongside the Belt and Road, will help enterprises to “go global” faster and better.

1. Introduction

Despite the late start of Chinese enterprises’ outbound investment, in recent years, under the initiative and guidance of the Chinese Government, the pace of “going global” of Chinese enterprises has accelerated rapidly. According to the 2016 World Investment Report of the United Nations Conference on Trade and Development (UNCTAD), China’s outbound investment ranked second in the world for the first time in 2015, which exceeded the amount of inward foreign investment of the same year for the first time, thus becoming a net exporter of capital. Better enterprises outbound investment will not only meet the investment needs of other countries in the world, especially developing countries, but is also conducive to China’s overcapacity reduction and industrial upgrading. However, at present, international protectionism characterized by anti-globalization prevails, and Chinese enterprises are facing challenges relating to host country’s legal regulations and social culture in further deepening outbound investment cooperation. Therefore, it is of great significance to examine the institutional factors behind the decision making and result of enterprise outbound investment from the micro perspective of enterprise large-scale outbound investment.

Unlike global outbound investment, which mainly takes place between developed countries, China’s outbound investment not only goes into developed countries, but also flows to many developing economies (Jiang and Jiang, 2012a). The diversification of investment targets determines that in outbound investment, Chinese enterprises not only face the challenges of formal institutional differences such as property rights protection, legal norms and judicial systems, they are also challenged by differences in culture, psychology, customs and other informal institutional differences. Moreover, Chinese outbound investments are mainly carried out by state-owned enterprises, which often enjoy ownership advantages such as soft budgetary constraints. They also have certain non-market-driven motives, so a single investment often amounts to hundreds of millions of dollars or even billions of dollars. Compared with small-scale outbound investment, these large-scale investments are more likely to induce discriminatory policies in the host country. Consequently, their investment scale, scope, model and result are special. The existing literature mostly studies from the perspective of Chinese enterprises’ total investment and lacks a focused discussion on this issue. Moreover, after China announced the Belt and Road Initiative (BRI), whether the impact of institutional differences between the countries alongside the Belt and Road and China on the decision making and result of large-scale investment will change is vital to the implementation of the BRI in the future. Since few papers have discussed this issue, this paper tries to fill the gap. Specifically, this paper uses 2,440 micro samples of 609 Chinese enterprises outbound investment from 2005 to 2016 to discuss the effects of formal and informal institutional differences on the decision making and result of enterprise investment. The paper shall also discuss whether the impact varies due to the host country’s participation in the BRI.

This study has discovered that, first, formal institutional differences significantly hinder the scope of large-scale outbound investment by enterprises but have no significant impact on the scale of investment. At the same time, formal institutional differences drive enterprises to adopt a full ownership structure but have no impact on the investment establishment model. Second, in contrast, the informal institutional differences have more extensive influence on large-scale outbound investment decision-making: expanding the scale of enterprise investment, reducing the scope of enterprise investment and driving the enterprise to adopt a full ownership structure and merger and acquisition mode. Third, both formal and informal institutional differences significantly increase the probability of failure for enterprise large-scale outward investment. Fourth, the complementarities created by the BRI on institutional differences are mainly reflected in formal institutions, while the Initiative’s function of bridging the gap between different cultures is relatively inadequate.

The contributions of this paper are as follows: first, this paper provides a comprehensive framework to study the outbound investment by enterprises. Most of the existing literature that discussed the decision making of enterprises’ outbound investment focuses on the location choice of enterprises’ investment (Jiang and Jiang, 2012b; Wang et al., 2014; Qi and Yang, 2012). The author believes that the outbound investment by enterprises is a systematic process, which includes the decisions of investment scale, investment scope, ownership structure and establishment mode besides location selection. It is of great significance to analyze these aspects of decision-making into a unified research framework for a comprehensive understanding of enterprises’ outbound investment behavior.

Second, this paper expands the study of the BRI. Most of the existing literature uses qualitative methods to analyze the effect of the BRI on the countries alongside the Belt and Road (Zhang and Liu, 2015). Only few literatures use a quantitative method, but they fail to analyze the complementation of the BRI and institutional differences (Sun et al., 2017). This study provides reliable empirical evidence to understand the policy effects of the BRI and its relevant conclusions provide guidance for further implementation of the BRI.

Finally, this paper analyzes detailed micro data of large-scale investment. The existing studies mostly use the data of the total amount of Chinese enterprises’ outbound investment, and it is difficult to reveal the structural characteristics of the outbound investment (Chen, 2014). The use of microdata on large enterprises’ outbound investment (over US$100m) helps to analyze the structural characteristics of the enterprises’ outbound investment. In addition, compares with small scale investment, large outbound investments are more susceptible to institutional differences between home and host countries, then the use of micro data allow for a more accurate estimation of the relationship between the two.

The rest of this paper is arranged as follows: the second part is literature review and research hypothesis, the third part is research design, the fourth part is the results analysis, the fifth part is further discussion and the sixth part is the conclusions and policy implications.

2. Literature review and research hypothesis

2.1 Literature review

The research on Chinese enterprises’ outbound investment is mainly carried out from three aspects. First, some literature has demonstrated the motivation of Chinese enterprises’ outbound investment from the perspective of enterprises’ individual characteristics. These studies generally believe that companies with stronger competitive advantages and higher productivity are more likely to invest abroad and are more likely to make larger investments (Tian and Yu, 2012; Ge and Luo, 2013). Second, some literature has studied the impact of enterprise outbound investment. Li and Jin (2011) found that Chinese enterprises’ outbound investment creates reverse technology spillover in the regions where the enterprises are located. Yu and Yang (2014) and Jiang and Jiang (2014) studied the export effects of outbound investment, but due to the differences in research perspectives, the conclusions were different. Furthermore, some literature studies the influencing factors of Chinese enterprises’ outbound investment from the institutional environment of the host country and the institutional differences between the home country and the host country (Habib and Zurawicki, 2002; Zong et al., 2012). Among them, how the institutional differences affect the location choice of enterprises’ outbound investment has received the most attention (Jiang and Jiang, 2012b; Wang et al., 2014). And, both formal and informal institutional environment and their differences are seen as important factors influencing enterprises’ outbound investment (Jiang and Jiang, 2012a; Qi and Yang, 2012).

While the existing literature mostly focused on location selection, the process of enterprises’ outbound investment involves not only location selection, but also the selection on investment scale and scope, ownership structure and investment modes. Although some papers have analyzed the investment decisions of enterprises from different perspectives, they lack a unified framework in discussing investment scale and scope, as well as the ownership structure and investment mode (Liu and Yang, 2016; Zhou and Zhang, 2014). Furthermore, the ultimate goal of enterprises’ outbound investment decision making is to improve the investment performance, so it is necessary to examine the enterprises’ investment decision making and performance at the same time (Liu and Yang, 2016; Yang et al., 2016). In view of this, this paper attempts to construct a complete research framework in order to comprehensively examine the impact of institutional differences on enterprises’ outbound investment decisions and results.

2.2 Theoretical analysis and research hypotheses

Generally speaking, to avoid risks, enterprises tend to choose a host country with a formal institutional environment that is similar to that of its home country. Therefore, the smaller the formal institutional differences between the home country and the host country, the large the scale of outbound investment (Habib and Zurawicki, 2002). A number of studies have pointed out that this theory is only applicable to international investment between developed countries, while Chinese enterprises are special in their outbound investment. Specifically, since China’s outbound investment is mainly carried out by state-owned enterprises, they have “two motives” and ownership advantages in outbound investment and may prefer to invest in host countries whose formal institutions are very different from China (such as African countries) (Jiang and Jiang, 2012a; Morck et al., 2008). However, when investing in host countries with similar formal institutions, Chinese enterprises can replace formal institutions with “connections” to enjoy many conveniences brought by “non-market skills”, thus greatly reducing the costs of institutional adaptation and organizational coordination (Kolstad and Wiig, 2012). For multinational enterprises, the external environment is far more complex than that of domestic operations, and the similarity of formal institutions makes it easier for Chinese enterprises to understand the environments in both countries, which is crucial for both market and non-market investment incentives. Therefore, when the formal institutions of the host country and China differ greatly, the possibility of Chinese enterprises’ outbound investment is reduced. Enterprises’ foreign investment involves not only the scope of investment, but also the scale of investment. For large enterprises, once they decide to enter a certain host country, then the decision is a dominant strategy, and the impact of formal institutional differences on investment scale may be extremely limited.

Similar to formal institutional differences, informal institutional differences such as culture differences between home country and host country are also important factors affecting enterprises’ outbound investment. Peretiatko and D’Souza (2005) pointed out that cultural similarity can promote enterprises’ outbound investment. For example, similar cultural backgrounds allow Australia to receive more outbound investment from US companies than any other countries in the Asia-Pacific region. Chang and Rosenzweig (2001) found that cultural differences can prevent firms from entering a host market. The influence of cultural differences stems from the fact that different cultures contain different business philosophy. Enterprises that have experienced the cultural influence of their home country may face cultural conflicts when investing abroad, and their original internal management model and organization strategy may not adapt to the cultural background of the host country. Even large enterprises with a good management system and strong adaptability cannot fully avoid such influence. Therefore, when the cultural and other informal institutions between the host country and China differ greatly, the possibility of large-scale outbound investment by Chinese enterprises will also decrease. However, informal institutional differences such as cultural difference are different from formal institutional differences. Compared with formal institutions, which is displayed through explicit knowledge, the informal institutional differences such as cultural difference are more hidden. Even when the home country and the host country share great similarities in the formal institutions, the social culture of the host country may still hide resistance to transnational corporations. Therefore, when enterprises decide to invest in a host country, in order to avoid the cultural divergence, they tend to make a large investment in order to strengthen the cultural attributes of the home country (Cui and Jiang, 2009).

Based on the above analysis, this paper proposes the following hypothesis:

H1. Formal institutional differences significantly reduce large-scale outbound investment scope but have no significant impact on investment scale. Informal institutional differences such as cultural differences significantly reduce the scope of investment and help to expand the scale of outbound investment.

As for outbound investment, enterprises not only need to decide the scope and scale of investment, but also need to choose an appropriate ownership structure and establishment mode. According to the transaction cost theory, both formal institutional difference, such as law, and informal institutional difference, such as cultural difference, will significantly increase the cost of information exchange and technology transfer between the home country and the host country. Therefore, when there is a large institutional difference between the home country and the host country, the enterprise tends to adopt a prudent investment strategy to reduce the intensity and control of investment in the host country, resulting in the joint venture model in the ownership structure (Pan and Lu, 2006). However, the institutional differences between the home country and the host country may also make it difficult to effectively transplant internal regulations and corporate culture into the joint venture, which could incur risk for outbound investment. Therefore, enterprises may choose to gain absolute control of the host country subsidiaries and to reduce the potential risk of internal management when faced with institutional differences (Brouthers and Brouthers, 2000). For enterprises making large outbound investments, risk aversion is clearly a higher priority than reducing the cost through share transfer. Moreover, factors such as sufficient funding and bargaining power often drive enterprises to favor the wholly-owned equity mode.

Enterprises choose the mode of cross-border merger and acquisition can also reduce the intensity of outbound investment. In cross-border merger and acquisition, the parent company directly acquires or merges foreign companies. While another important mode, greenfield investment, means that the parent company establishes a brand-new enterprise in the host country (Zhang et al., 2012). Through cross-border mergers and acquisitions, the parent company can obtain the resources of the acquired company. Through greenfield investment, the parent company can gain absolute control over the new company. But, of course, it also needs to bear more fixed investment costs (Jiang and Jiang, 2017). For enterprises that make large-scale outbound investments, fixed costs are not the main constraint, and companies tend to grasp control in order to maximize the company’s unique advantages. Due to the existence of control claims, the positive relationship between institutional differences and corporate bias toward cross-border M&A entry patterns may be greatly weakened. However, when multinational companies invest in other countries’ markets, compared with formal institutional differences, cultural and other informal institutional differences are more prominent sources of risk at the national level and cannot be eliminated in a short period of time. For large multinational corporations, operating risks caused by cultural differences are especially serious. Therefore, when faced with huge cultural differences, these corporations are more inclined to choose the entry mode of cross-border mergers and acquisitions. Therefore, this paper proposes the following hypothesis:

H2. Formal institutional differences drive companies to adopt a full ownership structure, but doesn’t induce them choose cross-border M&A mode more. Informal institutional differences such as culture not only increase the probability of companies adopting a full ownership structure, but also make them choose cross-border mergers and acquisitions rather than greenfield investment.

Institutional differences not only affect the enterprises outbound investment decision making in many aspects, they ultimately will also increase the risk of investment failure. Formal institutional differences between the home country and the host country can easily trigger political conflicts between the two countries. Although the risks of direct confiscation, expropriation and nationalization of enterprises’ outbound investment assets by the host country have been significantly reduced, the host country government may transfer the cost of political conflict caused by formal institutional differences to the enterprises through discriminatory policies and other means and increase the probability of failure of outbound investment (Yang et al., 2016; Yang, 2012). Moreover, when the cultural differences between the home country and the host country are relatively large, in the context of counter-globalization and increasing national sentiment, multinational corporations with the cultural imprint of the home country may become the target of public criticism and bear high operational risks. In case of an outbreak of national sentiment, enterprise outbound investment projects are likely to fall victim to the cultural gap. Especially in cases of large-scale outbound investment, the impact on the host country’s economy, employment and culture will be enormous, so these enterprises often bear the brunt in the conflict. Therefore, this paper proposes the following hypothesis:

H3. Both formal and informal institutional differences will significantly increase the probability of failure of large-scale enterprise outbound investment.

The impact of institutional differences on enterprises outbound investment is not static. It may depend on the bilateral investment agreements and political relations between the home country and the host country (Zong et al., 2012; Liu and Yang, 2016; Yang et al., 2016). China launched the BRI at the end of 2013, which is conducive to achieving connectivity and expanding exchanges and cooperation among the governments of countries alongside the Belt and Road. As of August 2016, as many as 34 countries and international organizations have signed intergovernmental cooperation agreements with China. One of the priorities of the BRI is to speed up the implementation of the bilateral national free trade zone strategy, which will promote the improvement of investment rules between China and the countries alongside the Belt and Road, overcome the investment risks brought about by formal institutional differences and create a formal institutional environment conducive to outbound investments. However, in addition to the formal institutional environment, cultural factors such as social customs, religious beliefs and social credit in countries alongside the Belt and Road are also important factors that determine whether foreign capital can be recognized by the host country. At present, the BRI still mainly focus on intergovernmental cooperation and exchanges and lacks functional arrangements of informal institutions. Therefore, this paper proposes the following hypothesis:

H4. The complementarities created by the BRI are mainly shown in formal institutional differences. However, the Initiative still lacks functional arrangements to bridge cultural differences.

3. Research design

3.1 Regression method

First, in order to analyze the influence of institutional difference between China and host countries on the dual margin of large-scale outbound investment, the following model is proposed:(1) $$ofd{i}_{ijt}={\alpha }_0+{\alpha }_{1}dif{f}_{jt}+\theta X_{jt}+{\lambda }_i+{\mu }_t+{\epsilon }_{ijt}.$$

In this model, i, j and t represent the enterprise, the host country and year, respectively, and ofdi represents the dual margin of outbound investment, which is decomposed into intensive (investment scale) and extensive margin (investment scope). diff denotes the institutional difference between China and the host country, including both formal and informal institutional difference. Control variables on the national level are expressed as X, industry fixed effect of investing enterprises as λ_i, time fixed effects as μ_t and the residual as ε_ijt.

Second, the following model is created to demonstrate the influence of institutional difference between China and the host country on enterprises’ investment mode:(2) $$\mathrm{Pr}\mathrm{ob}\left(\mathrm{mod}{e}_{ijt}=1\right)={\alpha }_0+{\alpha }_{1}dif{f}_{jt}+\theta X_{jt}+{\lambda }_i+{\mu }_t+{\epsilon }_{ijt}$$

Outbound investment mode by enterprises is investigated from two perspectives, ownership structure and mode of entry. In Equation (2), mode represents, on the one hand, different ownership structures. The value of this variable is 1 for wholly owned subsidiaries and 0 for joint ventures. On the other hand, it also represents different modes of entry. The value of mode is 1 for greenfield investment and 0 for mergers and acquisitions.

Third, the following model is constructed to analyze the influence of institutional difference between China and the host country on investment result:(3) $$\mathrm{Prob}\left(troubl{e}_{ijt}=1\right)={\alpha }_0+{\alpha }_{1}dif{f}_{jt}+\theta X_{jt}+{\lambda }_i+{\mu }_t+{\epsilon }_{ijt}$$

In this equation, trouble represents the investment result. The value is 1 when the outbound investment failed and 0 when succeeded.

Finally, to explore whether the BRI has changed the influence of institutional difference between China and countries along the Belt and Road on outbound investment by enterprises, the following model is constructed:(4) $$ofd{i}_{ijt}={\alpha }_0+{\alpha }_{1}dif{f}_{jt}+{\alpha }_{2}dif{f}_{jt}\times ba{r}_{jt}+{\alpha }_{3}ba{r}_{jt}+\theta X_{jt}+{\lambda }_i+{\mu }_t+{\epsilon }_{ijt},$$ (5) $$\text{Prob}\left({\mathrm{mode}}_{ijt}=1\right)={\alpha }_0+{\alpha }_{1}dif{f}_{jt}+{\alpha }_{2}dif{f}_{jt}\times ba{r}_{jt}+{\alpha }_{3}ba{r}_{jt}+\theta X_{jt}+{\lambda }_i+{\mu }_t+{\epsilon }_{ijt},$$ (6) $$\mathrm{Prob}\left(troubl{e}_{ijt}=1\right)={\alpha }_0+{\alpha }_{1}dif{f}_{jt}+{\alpha }_{2}dif{f}_{jt}\times ba{r}_{jt}+{\alpha }_{3}ba{r}_{jt}+\theta X_{jt}+{\lambda }_i+{\mu }_t+{\epsilon }_{ijt}.$$

In this model, bar_jt represents whether the host country j has been influenced by BRI in year t. As President Xi Jinping proposed the BRI at the end of 2013, for countries along the Belt and Road, when t⩾2014, bar_jt is 1. Otherwise, the value is 0. diff_jt and diff_jt×bar_jt are two core explanatory variables to pay attention to. If α₂ and α₁ are both positive or negative, it indicates that BRI has reinforced institutional difference’s influence on outbound investment. If the signs of α₂ and α₁ are opposite, it indicates that BRI has weakened institutional difference’s influence on outbound investment.

3.2 Variables and data

3.2.1 Dependent variables

There are three types of variables: investment margin, investment mode and investment result.

3.2.1.1 Investment margin

Referring to Liu and Yang (2016) and Yang et al. (2016), this paper divides large-scale outbound investment margin into intensive and extensive margins. Intensive margin refers to the average outbound investment by enterprises to country–industry pairs in one year, and extensive margin refers to the quantity of country-industry pairs of outbound investment by enterprises in one year. The equation is as follows:(7) $$ofd{i}_{it}=\sum _{j=1,h=1}ofd{i}_{ijht}=\underset{\mathrm{int}ensive}{\underbrace{\left(ofdivalu{e}_{it}\right)}}\times \underset{extensive}{\underbrace{\left(\sum _{j=1}ofdinu{m}_{ijt}\right)}},$$ where ofdi_it stands for the total outbound investment by enterprise i in year t, and ofdi_ijht stands for the investment by enterprise i in year t to industry h of country j. ofdivalue_it represents the average investment made by enterprise i in year t to the country–industry pair, i.e. the intensive margin. ofdinum_ijt represents the number of industries in country j that enterprise i has invested in year t, and the total sum on national level is the extensive margin.

3.2.1.2 Investment mode

The outbound investment mode is investigated from two perspectives: the ownership structure and the mode of entry. First, in terms of ownership structure, enterprises have two options: fully owned subsidiaries and joint ventures. According to Zhou and Zhang (2014), a subsidiary is considered wholly owned by its parent company if over 95 percent of its share is held by the parent company. Otherwise, it is a joint venture. Thus a dummy variable sharemode is introduced to represent the ownership structure of the outbound investment. The value of this variable is 1 for wholly owned subsidiaries and 0 for joint ventures. Second, in terms of mode of entry, two options are available: greenfield investment and mergers and acquisitions. Thus, a dummy variable greenmode is introduced to represent the mode of entry. The value of this variable is 1 for greenfield investment and 0 for mergers and acquisitions.

3.2.1.3 Investment result

A dummy variable trouble is introduced to denote the result of the investment made to host country j by enterprise i in year t. The value is 1 if the investment failed and 0 if succeeded.

3.2.2 Independent variables

Institutional difference variables are classified into formal and informal institutions variables.

3.2.2.1 Formal institutional difference (Idiff)

Referring to Jiang and Jiang (2012a), Wang et al. (2014) and Kolstad and Wiig (2012), this paper intends to evaluate the quality of formal institutions of different countries based on World Bank’s Worldwide Governance Indicators The indicators includes six dimensions, i.e. Political Stability, Government Effectiveness, Regulatory Quality, Rule of Law, Voice and Accountability and Control of Corruption. The average of the six indicators is taken as each country’s final score of formal institutions. Formal institutional difference is calculated by the score difference between China and the other countries. When the direction is not considered, then the absolute difference of the two is taken. When the direction is considered, if the host country’s formal institution is better than China, then there is a positive difference; if the host country’s formal institution is worse than China, there is a negative difference.

3.2.2.2 Informal institutional difference (Cdiff)

The most important aspect of the informal institution of a country is the culture; therefore, this paper adopts culture difference as the indicator to measure informal institutional difference between countries. According to Hofstede (2005), cultural difference index incorporates four dimensions, namely, power distance, uncertainty avoidance, individualism vs collectivism and masculinity vs femininity. Considering states visits and cultural exchange activities after the establishment of diplomatic relations will help narrow down the culture difference, this study includes the history of diplomatic relations to the cultural difference index. The equation for evaluating the informal institutional difference between countries is as follows:(8) $$Cdif{f}_j=\frac{{\sum }_{i=1}^4\left[\frac{{\left(I_{ij}-I_{ic}\right)}^2}{V_i}\right]}{4}+\frac{1}{T_j},$$ where Cdiff_j denotes the informal institutional difference between China and country i. I_ij is the value of the ith dimension of cultural indicator of country j, and I_ic is the value of the ith dimension of cultural indicator of China. V_i denotes the variance of the ith dimension of cultural indicator. T_j is the years of diplomatic relation between China and country j. 1/T_j shows that the longer the diplomatic relation, the smaller the informal institutional difference.

3.2.3 Control variables

Referring to previous studies, this paper adds a series of control variables in the regression model (Buckley et al., 2007; Qi and Wang, 2012): labor standard of the host country (Labor), measured by natural logarithm of tuberculosis incidence (per 100,000 persons); natural resource of the host country (Resource), measured by the percentage of ores and metals export to the total export; signing of Avoidance of Double Taxation Agreement (sign): if agreement was signed between China and the host country in that year, the value is 1, otherwise the value is 0; host country being a member of GATT or WTO (Gatt): if so, the value is 1, otherwise the value is 0; economic development level of the host country (InGdp), measured by natural logarithm of total GDP of the country or region; years of diplomatic relationship (time), calculated based on the establishment year of diplomatic relationship between China and the host country; geographic distance (Distance), the natural logarithm of the geographic distance between China’s capital city and that of the host country.

3.2.4 Sample and data

The data of outbound investment by enterprises are from China Global Investment Tracker 2005–1016 created by American Enterprise Institute and the Heritage Foundation. The data set documented all outbound investment over US$100m (including verifiable investment and construction contracts) made by Chinese enterprises, which can be used to study large-scale investment decisions by Chinese enterprises. The data set provides detailed information about the volume, industry, country and result of each investment, based on which the dual margin, investment mode and investment result can be obtained. Data of formal institution quality are retrieved from Worldwide Governance Indicator of World Bank; data of culture distance is from Hofstede’s website; and the establishment year of diplomatic relationship is inquired from website of Ministry of Foreign Affairs, the People’s Republic of China. Countries along the Belt and Road are listed in the Statistical Communiqué on China’s Foreign Investment. Development indicators including host country’s total GDP, GDP deflator, tuberculosis incidence, ratio of ores and metals export to the total export are from the database of World Bank. Information of Avoidance of Double Taxation Agreement between China and the host country is from the website of State Administration of Taxation. The list of GATT and WTO members and geographic distance between China and the host country is inquired from CEPII. Partially missing values are filled with linear interpolation. Table I summarized descriptive statistics of major variables.

4. Results

4.1 Baseline regression

4.1.1 Influence of institutional difference on investment margin

Table II presents the regression result of the impact of institutional difference and the BRI on outbound investment dual margins. Based on columns (1)–(4), formal institutional difference exerts a positive influence on the intensive margin of outbound investment, but the influence does not prove a statistical significance. While informal institutional difference has a significant positive influence on the intensive margin of outbound investment, that is, the greater the informal institutional difference, the larger the outbound investment scale. The results show that cultural difference could bring about “Benefit of Foreignness.” Neither the estimated cross-term between BRI and the formal institutional difference nor that between BRI and informal institutional difference is significant, showing that in terms of an outbound investment scale, BRI fails to be complementary to institutional difference.

Based on columns (5)–(8), both formal and informal institutional differences exert significant negative influence on the extensive margin of outbound investment. This means that similar institution is significantly preferred by large-scale outbound investment enterprises. The more institutional similarities China and the host country share, the lower the cost of policy adaptation and organizational coordination, so that the more likely Chinese enterprises are to invest in the host country. Furthermore, compared with formal institutional difference, informal institutional difference exerts greater negative influence on the extensive margin, which means it is more difficult to bridge informal institutional difference including cultural and psychological gaps than to adapt to policies, laws and other formal institutional difference. It takes time to better understand the condition of host countries and effectively replicate the recessive knowledge of enterprises when faced with wide cultural gaps. The estimated cross-term between BRI and formal institutional difference is significantly positive and that between BRI and informal institutional difference failed the test of significance, showing that BRI helps to overcome formal institutional difference and expand investment scope, yet lacks the function of bridging cultural and psychological gaps. The abovementioned results proved H1 and H4.

4.1.2 Influence of institutional difference on investment mode

Table III shows the regression result of the impact of institutional difference and the BRI on outbound investment mode. From columns (1) to (4), estimated coefficients of formal institutional difference are positive; the value in column (1) is not significant, and the value in column (2) is significant at the 5 percent level. This means that the greater the institutional difference between China and the host country, the more likely the enterprises are to invest through wholly-owned subsidiaries. Estimated coefficients of informal institutional difference are also positive, and the value in columns (3) and (4) are significant. This also indicates that the greater the cultural difference, the more likely the enterprises invest through wholly owned subsidies. When faced with great institutional difference, for the purpose of risk aversion, having absolute control of overseas subsidiaries to reduce potential risks of internal management overweighs the consideration of reducing information transaction costs through equity transfer (Brouthers and Brouthers, 2000). When carrying out a large-scale outbound investment, well-funded parent enterprises pursuing discourse power are inclined to invest through wholly owned subsidiaries. The estimated coefficient between BRI and formal institutional difference is significantly negative and that between BRI and informal institutional difference failed the test of significance. It shows that BRI can be a complement to the formal institution and reduce risks of joint ventures, yet still lacks the function of bridging cultural gaps.

Based on columns (5)–(8), in terms of investment mode, estimated coefficients of formal institutional difference in columns (5)–(6) failed the test of significance, while the estimated coefficients of informal institutional difference in columns (7) and (8) are significantly negative. This means that the formal institutional difference does not promote cross-border M&A, while the larger the cultural difference is, the more likely the company is to adopt cross-border M&A. The parent company obtains the technology, brand, marketing network and other resources of the acquired enterprise through cross-border M&A; through greenfield investment, the parent company obtains more control but has to bear more fixed costs. For a parent company carrying out a large-scale outbound investment, its investment decision is less likely to be constrained by fixed costs, but it is more difficult to manage operational risk brought up by cultural difference than institutional difference. Therefore, enterprises tend to acquire cultural identity through cross-border M&A. The estimated cross terms between BRI and formal and informal institutional difference both failed the test of significance, showing that the general tendency of going global at a rapid pace is not influenced by BRI. Chinese enterprises tend to invest through cross-border M&A rather than greenfield investment in the pursuit of high-speed internalization. The abovementioned results proved H2 and H4.

4.1.3 Influence of institutional difference on investment results

Table IV shows the regression result of the impact of institutional difference and the BRI on investment performance. Based on columns (1)–(4), estimated coefficients of formal and informal institutional difference are significantly positive, meaning that the possibility of outbound investment failure grows with institutional difference. Compared with the research of Liu and Yang (2016), this result further proves that informal institutional difference can also enlarge risk of investment failure. The estimated coefficient between BRI and formal institutional difference is significantly negative, while that between BRI and informal institutional difference failed the test of significance. Thus, BRI can effectively reduce risk of investment failure due to BRI as a complement to formal institutional difference, yet it is not complementary to informal institutional difference. The abovementioned results proved H3 and H4.

4.2 Robustness checks

Table V shows the results of robustness checks carried out with different variables and different estimation methods. First, the formal and informal institutional difference used in the baseline regression was the average value of various sub-parameters; thus, it is necessary to verify the robustness with the sub-parameters. The parameter of “judicial effectiveness” is used in columns (1) and (3) to measure the formal institutional difference, and the results are basically the same as that of the baseline regression. The parameter of “individualism and collectivism” is used in columns (2) and (4) to measure informal institutional difference, and the results are basically the same as that of the baseline regression. Second, when explaining variables are dummy variables, the Probit model was adopted in this paper; thus, it is necessary to verify the robustness by Logit model. Results in column (5)–(10) demonstrate that the conclusions are still valid under a different estimation method.

4.3 Endogeneity

The core explaining variables of the paper are institutional difference and BRI, which are less threatened by endogeneity. Because, on the one hand, the institution of a country, which is a result of long-term evolution, is unlikely to change in a short term; on the other hand, shock from BRI comes up randomly, since the time when China proposes the initiative depends on if consensus has been reached with other countries. Hardly can any country predict when China will reach the equilibrium with other countries during the dynamic process of gaming (Sun et al., 2017). Besides, nations are included into the BRI span because of their geographical locations and historical backgrounds, which are exogenous in nature. However, errors may exist in the measurement of institutional difference, and in order to ensure a certain level of freedom, the author did not control fixed effect of country. Therefore, it is still necessary to test the endogeneity that may exist in institutional difference. The lagged variable of institutional difference after one year is adopted as the instrumental variable for TSLS regression. F-value is larger than 10 in the first stage, indicating little problems exist in weak instruments. A detailed report is omitted on account of space limitation. Table VI shows the results in the second stage, demonstrating that the main conclusions of the paper are still valid considering endogeneity.

5. Further discussion

5.1 Estimates of different industries

The above analysis shows that the institutional differences between China and the host country have an impact on the large-scale outbound investment by enterprises, and the BRI also bridges institution difference to some extent. The next question is whether the conclusion still holds for different industries. To answer this question, the author analyzes the heterogeneity between manufacturing and service industries, and the results are showed in Table VII[1].

Panel A in Table VII shows estimates based on samples of the manufacturing sector. It shows that the formal institutional differences between the host country and China have significantly hindered the diversification of enterprises’ outbound investment in the manufacturing industry, increased the probability of enterprises adopting a full ownership structure and raised the possibility of enterprises’ failure in outbound investment in the manufacturing industry. Similarly, informal institutional differences also significantly hinder the diversification of enterprises’ outbound investment in the manufacturing industry and increase the probability of failure of enterprises’ outbound investment in the manufacturing industry. What is different is that informal institutional differences also significantly reduce the probability that enterprises will adopt the greenfield investment when investing in the manufacturing industry. In addition, the complementarities created by the BRI are more conspicuous in formal institutional differences than in informal institutional differences.

Panel B in Table VII shows the estimates based on the service industry samples. It can be seen that both formal and informal institutional differences have significantly expanded the scale of enterprises’ outbound investment in the service industry. However, formal institutional differences have significantly reduced the scope of enterprises’ outbound investment in service industries. This supports the fact that China’s outbound investment in service industry is mainly concentrated in a few developed countries. In addition, the bigger the difference between the formal and the informal institutions, the more likely for enterprises to adopt a full ownership structure when investing in the service industry. It is worth mentioning that the differences between formal and informal institutions have not significantly affected the results of enterprises’ outbound investment in service industry. This shows that the receiving countries of Chinese enterprises’ service industry investment have effective market mechanisms and strict institutional norms, which, to some extent, alleviates the risks brought about by institutional differences.

5.2 Estimates of positive and negative institutional differences.

The above analysis does not distinguish positive and negative institutional differences when examining formal institutional differences. Therefore, the author further studies whether the above conclusion still holds true in countries with superior and inferior institutions compared with China, and the results are shown in Table VIII.

Panel A in Table VIII reports the estimates based on positive institutional difference samples. It showed that the enterprises’ outbound investment scale did not increase as the institutional differences for the host country whose formal system is superior to China. This shows that the escape response theory cannot effectively explain the outbound investment behavior of Chinese enterprises (Zong et al., 2012). Formal institutional differences have not significantly hindered the diversification of enterprises’ outbound investment, indicating that a strong institutional environment has a moderating effect on formal institutional differences. In sharp contrast to this, informal institutional differences still significantly hinder the diversification of enterprises’ outbound investment, indicating that the strong institutional environment has failed to effectively bridge cultural differences.

Panel B in Table VIII reports the estimates based on samples of negative institutional differences. It can be found that for the host country whose formal system is inferior to China, the formal system differences have not significantly increased the scale of enterprises’ outbound investment, but have significantly reduced the scope of enterprises’ outbound investment. This shows that Chinese enterprises prefer a good market economy system for outbound investment. Although a host country whose formal institution is inferior to China may have a large market and rich natural resources, Chinese enterprises still favor a good institutional environment (Wang et al., 2014). It is worth noting that neither formal nor informal institutional differences have significantly affected the ownership structure and establishment of enterprises’ outbound investment. The possible reason is that Chinese enterprises have a strong unique advantage when investing in a host country whose formal system is inferior to China; therefore, the host country’s environmental factors have limited influence on the decision-making of investment.

6. Main conclusions and policy implications

The report of the 19th National Congress of CPC clearly pointed out that “We should pursue the Belt and Road Initiative as a priority, give equal emphasis to ‘bringing in’ and ‘going global’; We will develop new ways of making outbound investments, promote international cooperation on production capacity.” Optimizing outbound investment concerns not only location decisions, but also the investment structure, equity mode, establishment pattern and investment performance. However, most of the existing studies are focused on one aspect of outbound investment; they lack a global vision and a unified framework. Moreover, among quantitative analyses of the BRI, few studies have been carried out on the relationship between the BRI and institutional differences. Based on the micro data of 2,440 large-scale outbound investments of 609 enterprises in China from 2005 to 2016, this paper examines the institutional factors affecting large-scale outbound investment decision-making and result, as well as the functional arrangement of the BRI from the perspective of formal and informal institutional differences, thus filling the gap in the existing research field.

The study shows that formal and informal institutional differences have different impacts on large-scale outbound investment decisions of enterprises. Formal institutional differences have significantly hindered the scope of large-scale outbound investment by enterprises, but have no significant impact on the scale of investment. At the same time, with formal system differences in place, enterprises are more likely to adopt a full ownership structure, yet their choose greenfield investment or not remain unaffected. In contrast, informal institutional differences have a wider impact on large-scale outbound investment decisions of enterprises: they expand the scale of enterprise investment, hinder the scope of enterprise investment and are more likely to give rise to a full ownership structure and a merger and acquisition mode. As for the performance of large-scale outbound investment by enterprises, both formal and informal institutional differences have significantly increased the probability of failure. The complementarities created by the BRI is mainly reflected in formal institutions, and its function of bridging cultural differences is still insufficient.

There is heterogeneity among different industries and institutions about the above conclusions: on the one hand, unlike the manufacturing industry, formal institutional differences have significantly expanded the scale of enterprises’ outbound investment in service industry, and none of the institutional differences has significantly increased the probability of enterprises’ investment failure in service industry. On the other hand, the negative impact of formal institutional differences on the diversification of investment by Chinese enterprises is only reflected in cases where the formal institution is inferior to that of China, reflecting that Chinese enterprises tend not to invest in countries with inferior institutions. Compared with the host country whose institution is superior to China, the outbound investment mode and performance of Chinese enterprises are less related to the institution difference when investing in a country with an inferior institution, which may be due to the ownership advantage of state-owned enterprises.

The policy implication of this paper is that in helping enterprises to make better and faster outbound investment, the government should accurately assess the risks that may be brought about by institutional differences, and put in place relevant preventive measures and risk guarantees. Based on the institutional environment of the host country and the specific attributes of industry of the enterprise’s investment, government can guide enterprises in decision-making regarding a reasonable investment scale, an appropriate investment scope, a suitable percentage of ownership and an establishment mode. In addition, deepening the BRI and building closer regional cooperation is one of the key strategies to be considered in the top-level design to help Chinese enterprises’ large scale outbound investment. We should not only strengthen the interconnection with the official institutions of the countries along the Belt and Road through exchange of visits between political leaders and building sister cities, but also carry out functional arrangements to bridge cultural differences such as establishing Confucius Institute and holding cultural festivals.

This paper is supported by the National Social Science Fund Major Project (14ZDA023). Thanks to the amendments proposed by the anonymous reviewers, the author has made changes accordingly and shall take sole responsibility for this paper.

Note

1.In this paper, the number of outbound investment in agriculture is only 26; therefore, the sample size is too small to be taken in to consideration.

Table I

Descriptive statistics of main variables

Variables	Connotation	Sample size	Mean value	SD	Minimum	Maximum
Intensive	Natural logarithm of intensive margin of outbound investment	2,440	6.213	0.841	4.605	9.680
Extensive	Extensive margin of outbound investment	2,440	6.784	7.494	1	28
Sharemode	Ownership structure (wholly-owned=1, joint venture=0)	709	0.230	0.421	0	1
Greenmode	Mode of entry (Greenfield=1, M&A=0)	2,440	0.134	0.340	0	1
trouble	Investment result (fail=1, success=0)	2,440	0.083	0.276	0	1
Idiff	Natural logarithm of formal institutional difference	2,440	−0.607	1.248	−7.711	0.866
Cdiff	Informal institutional difference	1,806	2.211	1.382	0.322	5.626
bar	Influenced by the Belt and Road Initiative or not	2,440	0.134	0.341	0	1
Labor	Natural logarithm of tuberculosis incidence (per 100,000 persons)	2,440	3.288	2.073	0.470	6.884
Resource	Natural logarithm of percentage of ores and metals export to total export	2,440	0.949	1.589	−8.756	4.459
Sign	Signed Avoidance of Double Taxation Agreements with China or not	2,440	0.792	0.406	0	1
Gatt	Whether the host country is a member of GATT or WTO	2,440	0.716	0.451	0	1
Gdp	Natural logarithm of host country’s economic development level	2,440	25.980	4.057	19.223	30.460
Time	Years of diplomatic relation	2,440	42.680	13.230	0	68
Distance	Natural logarithm of geographic distance between the host country and China	2,440	8.825	0.997	6.696	9.868

Table II

Institutional difference, BRI and investment margins

	Intensive				Extensive
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Idiff	0.014 (0.015)	0.008 (0.017)	–	–	−0.071*** (0.020)	−0.092*** (0.022)	–	–
Idifff×bar	–	0.020 (0.033)	–	–	–	0.205*** (0.043)	–	–
Cdiff	–	–	0.063*** (0.022)	0.059** (0.025)	–	–	−0.112*** (0.028)	−0.119*** (0.030)
Cdifff×bar	–	–	–	0.003 (0.047)	–	–	–	0.050 (0.063)
bar	–	−0.052 (0.063)	–	−0.073 (0.100)	–	0.374*** (0.079)	–	0.016 (0.136)
Constant term	5.092*** (0.261)	5.928*** (0.260)	5.901*** (0.357)	5.960*** (0.356)	1.249*** (0.243)	1.146*** (0.241)	1.823*** (0.359)	1.751*** (0.366)
lnalpha	–	–	–	–	−0.344*** (0.027)	−0.356*** (0.027)	−0.331*** (0.031)	−0.332*** (0.031)
Control variable	Y	Y	Y	Y	Y	Y	Y	Y
Industry fixed effect	Y	Y	Y	Y	Y	Y	Y	Y
Time fixed effect	Y	Y	Y	Y	Y	Y	Y	Y
Sample size	2,440	2,440	1,806	1,806	2,440	2,440	1,806	1,806

Notes: Values in brackets are heteroscedasticity-robust standard errors; the same with the tables below. The generalized least squares approach is adopted from column (1) to (4), and negative binomial regression is used from column (5) to (8). **,***Significant at the 5 and 1 percent significant levels, respectively

Table III

Institutional difference, BRI and investment mode

	Sharemode				Greenmode
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Idiff	0.150 (0.101)	0.312** (0.125)	–	–	−0.005 (0.030)	0.004 (0.035)	–	–
Idifff×bar	–	−0.513** (0.207)	–	–	–	0.001 (0.071)	–	–
Cdiff	–	–	0.137* (0.079)	0.159* (0.089)	–	–	−0.186*** (0.051)	−0.171*** (0.057)
Cdifff×bar	–	–	–	−0.281 (0.229)	–	–	–	−0.033 (0.107)
bar	–	−0.610* (0.356)	–	0.001 (0.509)	–	0.189 (0.138)	–	0.139 (0.211)
Constant term	−4.340*** (1.436)	−4.831*** (1.485)	−2.997 (1.958)	−2.608 (1.864)	−2.096*** (0.528)	−2.227*** (0.542)	−3.883*** (0.965)	−3.946*** (0.979)
Control variable	Y	Y	Y	Y	Y	Y	Y	Y
Industry fixed effect	Y	Y	Y	Y	Y	Y	Y	Y
Time fixed effect	Y	Y	Y	Y	Y	Y	Y	Y
Sample size	709	709	593	593	2,391	2,391	1,775	1,775

Notes: *,**,***Significant at the 10, 5 and 1 percent significant levels, respectively

Table IV

Institutional difference, BRI and investment result

	trouble
Variables	(1)	(2)	(3)	(4)
Idiff	0.112** (0.047)	0.174*** (0.051)	–	–
Idifff×bar	–	−0.248** (0.099)	–	–
Cdiff	–	–	0.134*** (0.050)	0.136** (0.058)
Cdifff×bar	–	–	–	0.011 (0.108)
bar	–	−0.105 (0.179)	–	0.216 (0.283)
Constant term	−0.924 (0.602)	−0.973 (0.610)	−1.325* (0.776)	−1.496* (0.803)
Control variable	Y	Y	Y	Y
Industry fixed effect	Y	Y	Y	Y
Time fixed effect	Y	Y	Y	Y
Sample size	2,391	2,391	1,753	1,806

Notes: *,**,***Significant at the 10, 5 and 1 percent significant levels, respectively

Table V

Robustness testing: alternative indicator and estimation methods

	Intensive		Extensive		Sharemode		Greenmode		trouble
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Idiff	0.016 (0.018)	–	−0.090*** (0.021)	–	0.688** (0.271)	–	0.009 (0.063)	–	0.343*** (0.112)	–
Idifff×bar	−0.046 (0.035)	–	(0.047)	–	−1.112*** (0.417)	–	−0.006 (0.122)	–	−0.477** (0.219)	–
Cdiff	–	0.005 (0.009)	–	−0.039*** (0.010)	–	(0.163)	–	−0.279*** (0.081)	–	0.257** (0.113)
Cdifff×bar	–	−0.029 (0.068)	–	−0.191 (0.118)	–	−0.496 (0.528)	–	−0.078 (0.200)	–	0.056 (0.213)
bar	−0.110 (0.061)	−0.065 (0.075)	0.249*** (0.077)	0.154 (0.099)	−1.187 (0.760)	0.043 (1.071)	0.320 (0.252)	0.277 (0.379)	−0.186 (0.397)	0.357 (0.614)
Constant term	5.925*** (0.261)	6.127*** (0.350)	1.189*** (0.243)	1.368*** (0.352)	−8.797** (2.702)	−4.968 (3.827)	−3.820*** (1.059)	−7.341*** (1.764)	−1.698 (1.275)	−2.497 (1.615)
Control variable	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Industry fixed effect	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Time fixed effect	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Sample size	2,440	1,806	2,440	1,806	709	593	2,391	1,775	2,391	1,753

Notes: **,***Significant at the 5 and 1 percent significant levels, respectively

Table VI

Endogeneity testing

	Intensive		Extensive		Sharemode		Greenmode		trouble
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Idiff	0.024 (0.022)	–	−0.253 (0.179)	–	0.049** (0.023)	–	−0.005 (0.009)	–	0.023*** (0.007)	–
Idifff×bar	0.008 (0.062)	–	1.551*** (0.414)	–	−0.169** (0.073)	–	−0.025 (0.029)	–	−0.016 (0.013)	–
Cdiff	–	0.058** (0.028)	–	−0.572*** (0.215)	–	0.031 (0.024)	–	−0.026** (0.012)	–	0.026*** (0.010)
Cdifff×bar	–	0.022 (0.050)	–	0.222 (0.415)	–	−0.034 (0.040)	–	−0.007 (0.021)	–	0.004 (0.020)
bar	−0.054 (0.084)	−0.082 (0.105)	2.553*** (0.771)	−0.114 (1.035)	−0.254** (0.115)	−0.046 (0.103)	0.010 (0.037)	0.047 (0.051)	0.002 (0.024)	0.035 (0.037)
Constant term	6.088*** (0.218)	5.798*** (0.479)	11.888*** (2.547)	12.723*** (5.585)	−0.536 (0.349)	−0.334 (0.513)	0.032 (0.081)	−0.081 (0.231)	0.304** (0.123)	0.062 (0.142)
Control variable	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Industry fixed effect	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Time fixed effect	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Sample size	1,905	1,505	1,905	1,505	580	492	1,905	1,505	1,905	1,505

Notes: **,***Significant at the 5 and 1 percent significant levels, respectively

Table VII

Estimates for different industries

	Intensive		Extensive		Sharemode		Greenmode		trouble
Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Panel A: manufacturing industry
Idiff	−0.027 (0.020)	–	−0.074*** (0.023)	–	0.214* (0.126)	–	0.013 (0.040)	–	0.171*** (0.056)	–
Idifff×bar	0.019 (0.039)	–	0.194*** (0.047)	–	−0.784*** (0.276)	–	0.045 (0.085)	–	−0.284*** (0.107)	–
Cdiff	–	0.004 (0.033)	–	−0.118*** (0.039)	–	0.109 (0.106)	–	−0.135* (0.069)	–	0.189** (0.076)
Cdifff×bar	–	0.058 (0.057)	–	0.093 (0.072)	–	−2.896 (1.897)	–	0.028 (0.127)	–	−0.222 (0.139)
Sample size	1,515	1,100	1,515	1,100	415	345	1,515	1,100	1,515	1,078
Panel B: service industry
Idiff	0.150*** (0.036)	–	−0.092* (0.054)	–	0.925** (0.421)	–	−0.016 (0.094)	–	0.242 (0.219)	–
Idifff×bar	0.051 (0.090)	–	0.066 (0.128)	–	0.752 (1.331)	–	−0.158 (0.209)	–	0.229 (0.306)	–
Cdiff	–	0.191*** (0.045)	–	−0.045 (0.057)	–	0.512** (0.212)	–	−0.175 (0.152)	–	−0.039 (0.113)
Cdifff×bar	–	−0.191** (0.085)	–	−0.109 (0.116)	–	−0.419 (0.315)	–	−0.282 (0.295)	–	0.699*** (0.213)
Sample size	743	559	743	559	193	161	690	524	644	491

Notes: Regression includes dummy variable (bar), control variables, the fixed effect of industry and year. *,**,***Significant at the 10, 5 and 1 percent significant levels, respectively

Table VIII

Estimates of positive and negative institutional differences

	Intensive		Extensive		Sharemode		Greenmode		trouble
Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Panel A: positive institutional differences
Idiff	0.007 (0.025)	–	−0.041 (0.032)	–	0.378** (0.176)	–	0.040 (0.048)	–	0.155** (0.070)	–
Idifff×bar	0.031 (0.045)	–	0.136** (0.060)	–	−0.481 (0.305)	–	−0.095 (0.096)	–	−0.276** (0.129)	–
Cdiff	–	0.032 (0.033)	–	−0.121*** (0.038)	–	0.185* (0.097)	–	−0.232*** (0.073)	–	0.296*** (0.075)
Cdifff×bar	–	0.029 (0.058)	–	0.011 (0.071)	–	−0.240 (0.225)	–	−0.227 (0.220)	–	0.022 (0.111)
Sample size	1,563	1,332	1,563	1,332	597	526	1,537	1,313	1,516	1,294
Panel B: negative institutional differences
Idiff	0.027 (0.029)	–	−0.061* (0.032)	–	−0.028 (0.035)	–	−0.002 (0.012)	–	0.017* (0.010)	–
Idifff×bar	−0.035 (0.053)	–	0.231*** (0.063)	–	−0.051 (0.091)	–	0.029 (0.020)	–	−0.016 (0.025)	–
Cdiff	–	0.143*** (0.048)	–	0.021 (0.059)	–	0.006 (0.024)	–	0.027 (0.022)	–	−0.024 (0.017)
Cdifff×bar	–	−0.184* (0.108)	–	−0.190 (0.156)	–	−0.013 (0.033)	–	0.009 (0.069)	–	0.061* (0.034)
Sample size	877	474	877	474	112	67	877	474	877	474

Notes: Regression includes dummy variable (bar), control variables, the fixed effect of industry and year. *,**,***Significant at the 10, 5 and 1 percent significant levels, respectively