Outward foreign direct investment (hereinafter referred to as OFDI) refers to Chinese domestic investors’ business activities such as establishing, purchasing foreign enterprises in Hong Kong, Macao and Taiwan or foreign areas in forms of cash and intangible assets, aiming to gain control over management and administrative rights of enterprises. According to 2015 World Investment Report released by UNCTAD, in 2014, the outflow of global OFDI was 1.35 trillion dollars. The year-end stock was 25.87 trillion dollars. Taking this number as the base, in 2014, OFDI of China accounted for 9.1% and 3.4% of global OFDI and stock respectively. The outflow of global OFDI ranked third in global nations (regions) for three consecutive years. The proportion increased by 1.5% compared with last year. The monetary stock ranked eighth, and advanced by three positions compared with last year.
In September and October 2013, while visiting countries in Central Asia and Southeast Asia, Chinese President Xi Jinping proposed strategies of “Economic Belt of Silk Road” and “21st Century Maritime Silk Road” (hereinafter referred to as The Belt and Road). In 2014, during the APEC Beijing Summit, China announced the establishment of the Asia Infrastructural Investment Bank and the Silk Road Fund, in order to support and enhance connectivity and construction among countries along “The Belt and Road”, which attracted a lot of international attention. In March 2015, China formally issued Vision And Action For Promoting Joint Development Of Economic Belt Of Silk Road” And “21st Century Maritime Silk Road”. It signified that the initiative of “The Belt and Road” began the stage of comprehensive development.
Under the initiative of “The Belt and Road”, how to develop and enhance Chinese direct investments on nations along “The Belt and Road” is an important and realistic issue. This issue involves important significance for properly promote the implementation of “The Belt and Road” initiative, as well as reference values for enhancing Chinese direct investments on nations along “The Belt and Road”. In recent years, research on the “The Belt and Road” initiative has aroused extensive attention in the academic community. However, literary studies and research based on the perspective of solid evidence are still inadequate. We can’t help wondering: are direct investments on nations along “The Belt and Road” dominated by the government, or corporate choices based upon profit maximization? As most developing countries do not indicate many institutional differences from China, will they pose unique influences on such direct investment? What are the roles and influences of ethnic Chinese in this initiative?
According to previous research on institution,  proposed that OFDI indicate preference for countries and regions with favorable environment and conditions. According to data research on developed nations’ FDI conducted by  , it demonstrated that absolute differences of bilateral institutions are negatively related to FDI. They proposed “theory of institutional proximity”, i.e. less absolute differences of bilateral institutions are more conducive for bilateral investments. Buckley’s research on Chinese OFDI demonstrated that Chinese enterprises indicate obvious motives in market quest. However, some other studies, such as those conducted by  indicated that Chinese OFDI did not reveal obvious “preference” over institutions in host countries. Although many scholars pay attention to influences of institutional quality on OFDI, they still have not developed unified conclusion concerning this issue. Some literary research was about influence of informal institutions on foreign direct investments. For instance, studies conducted by   found out that ethnic Chinese are conducive for the inflow of Chinese direct investment. But these research and studies mainly focus on primary effects of ethnic Chinese on corporate choices of outward investment in foreign locations. They paid less attention to regulatory effects of situational factors.
Compared with current research and studies, innovation in this paper involves the two following aspects: 1) In research and studies concerning influences of institutional differences on Chinese enterprises’ investment on countries in “The Belt and Road” strategy, the paper introduces the informal institutional factor of ethnic Chinese in host countries, in order to explore regulatory effects of institutional differences and Chinese enterprises’ choice on OFDI locations. 2) While studying how institutional differences and ethnic Chinese in host countries affect Chinese investments on countries in “The Belt and Road”, we conduct tests based on corporate classification. According to investment motivations, we divide companies into different types: commercial services, local production and sales, technology research and development, and resource development, in order to verify different influences of institutional disparity and ethnic Chinese on different types of enterprises.
2. Theoretical Basis
 proposed that institutions are a series of “rules of game” that include formal institutions and informal institutions, concerning different aspects such as economy, laws and social system, etc.  reduced institutions to the sum of social rules and regulations, including three basic elements: regulation, management and perception. Current research on OFDI location distribution mainly involves two theories. One is the eclectic theory of international production proposed by  . The other on is the theory of heterogeneity proposed by  . The eclectic theory of international production explains corporate OFDI from three aspects: ownership advantage, internalization advantage and regional advantages. Only when enterprises possess all of the three advantages will they choose for conduct OFDI. Institutional differences and ethnic Chinese around the global are the main regional advantage factors that impact Chinese OFDI. Helpman thought that OFDI should pay for more fixed costs overseas. It indicated that only the most efficient enterprises can conduct OFDI. Fixed costs directly impact the amount and distribution of OFDI.  believed that countries with less institutional differences are more likely to accept each other’s context and trading regulations, so as to reduce costs required by corporate OFDI.
Research conducted by  indicated that overseas Chinese could share information and promote completion of bilateral trading through informal network connections.  found out that inherent punishment mechanism set up by overseas Chinese could prevent potential opportunistic behaviors from occurred in trading.  thought that cultural differences and the lack of “overseas networks” lead to “outsiders’ disadvantages” in transnational business operations. With the rapid and fast development of information and communication technology, global economic integration accelerates its pace, global economic integration and communication also become increasingly frequent, thus promoting the development of transnational business operations. As a result, “outsiders’ disadvantages” pose more negative influences than “strangers’ disadvantages” on transnational businesses. Overseas network is conducive to overcoming such disadvantages. Therefore, through formation and informal connections, overseas Chinese network is conducive for Chinese enterprises to integrate into local businesses in host countries, obtain knowledge and intangible assets needed in local business operations, set up commercial relationship with intangible assets, in order to gain legitimacy in host countries.
Hence, based on above theories, the paper proposes two assumptions:
Assumption 1: China and host countries indicate increasingly obvious institutional differences. Chinese enterprises are less likely to invest in such countries.
Assumption 2: in host countries, over Chinese network indicates more vitality, it is more like to reduce negative impacts of institutional differences on Chinese enterprises’ OFDI.
3. Data and Models
1) Data samples
This paper considers the number of Chinese OFDI projects in countries along “The Belt and Road” as the proxy variable in Chinese enterprises’ OFDI. Data source is Statistical Bulletin Of Chinese Foreign Direct Investment. It selects data concerning investments in 65 countries along “The Belt and Road” from 2003 to 2014. “The silk road economic belt” and “Marine silk road economic belt” has a long history, “The Belt and Road” involved 65 countries and regions (Table 1).
2) Variable declarations (Table 2)
Linearization of several variables, there will be heteroscedasticity, affect the
Table 1. Sample data.
Table 2. Variable and data sources and descriptions.
Table 3. Descriptive statistics of different variables.
stability of statistical results. The natural logarithm of these variables can eliminate the phenomenon of heteroscedasticity. Since the logarithm is strictly monotone increasing function, it will not change the causality between the data, and can avoid the violent fluctuation between the variables.
Descriptive statistics of different variables are shown in Table 3.
3) Model construction
This paper constructs basic models based on following formulas:
Chinese enterprises’ investments in countries along “The Belt and Road” = f (institution difference, overseas Chinese network, market demands, wage levels, technical levels, resource level, export, geographical distance, economic development, cultural distance).
Considering that explained variable selected in this paper is Chinese enterprises’ investment projects in countries along “The Belt and Road” in 2013 and 2014, which is nonnegative integer of discrete variable. Poisson’s regression model is used in this enumeration data. After calculation, sample variance is greater than expected value, indicating excessive decentralization. If Poisson’s regression is still used at this point, it may generate significant errors. So it adopts the negative binomial regression model:
is unobserved individual effects. Assume that conforms to Gam- ma distribution with as parameters, and is independent identically distributed. Then, conforms to the negative binomial distribution:
Mean value and variance are and
. When is any nonzero constant, conditional variance is greater than conditional mean. When is close to infinity, negative binomial distribution converges in Poisson distribution model. As a result, Poisson distribution model is a unique example of negative binomial distribution. Estimated values of and can be obtained through the following log-likelihood function of negative binomial distribution:
As zero value accounts for big proportions in explained variables of samples, which may impact estimation results. So zero-expansion negative binomial regression is adopted. The mixed distribution is as follows:
Zip option is added in regression, so as to provide a likelihood-ratio test to determine whether parameter of excessive decentralization is 0. The result shows that parameter is not 0. So it proves that zero-expansion negative binomial regression should be adopted.
To sum up, the model should be set as:
i and t represents year and country, ninvest represents Chinese enterprises’ investments in countries along “The Belt and Road”, is a constant of national differences, cgzcfl represents institution distance, llaborc represents the delay stock of labor dispatching in host countries. cgzcf *llaborc is introduced to verify Assumption 2, lndistance represents geographical distance, lngdp represents market demands, lnhgdp represents labor costs, lnex represents export, lnhighex represents technological levels of host countries, lnre represents resource of host countries, lncgdp represents economic development level of China, cd represents cultural distance between China and host countries. As for (6), through the stata 12 panel data, obtain the estimated parameters based on negative regression estimation method.
4. Empirical Results
1) Analysis of regression results
The paper adopts the method of hierarchical regression. Model 1 conducts regression on control variables. Model 2 adds independent variables and regulated variables to Model 1. Model 3 adds interactive terms to Model 2. Results are shown in Table 4.
From regression results, we can see that for control variables, geographical
Table 4. Regression results.
t statistics in parentheses *p < 0.05, **p < 0.01, ***p < 0.001.
distance coefficient in Model 1 are obviously negative, indicating that geographical distance is still an important factor prohibiting Chinese enterprises’ investments in countries along “The Belt and Road”. Wage levels of host countries (lnhgdp) and export to host countries (lnex) coefficient symbols are identical to expectations, indicating that labor costs of host countries is still an important factor for Chinese enterprises’ outward investments. They may reduce investments in countries with higher levels of average per capita income in order to save costs. Export can promote foreign investments to some extent. Host countries’ GDP coefficient (lngdp) is negative. The possible reason is that in primary stages, investments in countries along “The Belt and Road” are guided by state-owned enterprises, which are mainly controlled by national government. As a result, Chinese OFDI in “The Belt and Road” tend to reveal diplomatic policies and government expectation. Natural resources of host nations (lnre) are negative, which is in conflict with expectations. The possible reason is that we merely regard mineral, metal and fuel export proportion in GDP as proxy variables of resources. In practical investments on countries along “The Belt and Road”, many enterprises do not include resources such as forest and rubber as proxy variables in host countries. Technological level coefficient (lnhighex) is negative. Through comparing lists, we can find that many foreign investment projects provide services in technical support and technology consulting for host countries. It also indicates that Chinese enterprises do not prefer to make investments on nations with high technological levels. Cultural difference coefficient (cd) between host countries and China is not obvious. The proxy variable(lncgdp)of China’s economic development level is obviously positive. It means that the enhancement of China’s economy also promotes increase in corporate outward and foreign investments.
Based on Model 2 and Model 3, it can be seen that institutional difference coefficient is obviously negative. It means that institutional differences between countries along “The Belt and Road” and China can reduce Chinese enterprises’ direct investments. As a result, Chinese enterprises are more likely to make investments in countries with less institutional differences compared with China. It is identical to research results of  . Similar institutions have similar trading rules. So it is conducive for transnational corporations to apply strategies of parent companies in subsidiary corporations in host countries. It is also conducive for Chinese enterprises to make full use of their own advantages in local conditions in host countries, so as to reduce related cost risks and increase outward investments. In Model 3, delayed stock of labor dispatch and interactive item coefficients are obvious. Meanwhile, compared with Model 2, institution difference coefficients decrease, while overseas Chinese coefficients increase. It means that activeness of overseas Chinese network poses positive influences on Chinese enterprises’ OFDI, and reduce negative impacts of institutional differences on “going out” of Chinese enterprises. In host countries, the overseas Chinese network can help enterprises reduce time on familiarizing themselves with local markets, searching for useful interpersonal connections and market information. Beyond doubt, it can facilitate and reduce institutional differences between host countries and host countries. Kolstad and Wiig’s research indicated that Chinese enterprises tend to make good use of non-market skills, set up relational network so as to replace formal institutions for their own interests.
2) Tests on institutional preference
According to research conducted by  and  , when making investments in host countries with better institutional contexts than home countries, transnational corporations choose countries with bigger institutional differences. When making investments in host countries with worse institutional contexts, transnational corporations tend to choose countries with smaller institutional differences. Can this study, conducted in developed countries, be applied in developing nations? Does it produce asymmetric effects on location selections in China’s OFDI in countries along “The Belt and Road”? In order to verify this issue, the paper discusses concepts of positive institution differences and negative institution differences. Regression results are shown in Table 5. It can be seen
Table 5. Tests on institutional preference.
t statistics in parentheses *p < 0.05, **p < 0.01, ***p < 0.001.
that as for countries along “The Belt and Road” that have better institutions than China, Chinese enterprises indicate preference of institution differences in outward investments. As for countries along “The Belt and Road” that have worse institutions than China, Chinese enterprises indicate preference in host countries with less institutional differences. Positive institution difference coefficients are obviously negative. Such results are identical to results of studies conducted by  , which indicate “institutional proximity”.
3) Test on investment motivations
According to main purposes of Chinese enterprises’ OFDI, the paper divide them into different types: commercial services, local production and sales, resource development, technology research and development. Commercial service companies means that investments mainly provide business services to enterprises, so as to facilitate domestic and foreign production and investment. Local production and sales OFDI focuses on production based on cheap local labor and other special elements, so as to reduce production costs and enhance competition. Resource development OFDI aims to development unique local resources, so as to ensure domestic production and gain materials. Technology research and development OFDI aims to make use of advanced production factors, as well as research and development contexts, so as to enhance R&D efficiency for subsidiary corporations in host countries and parent companies in home countries.
Based on regression results in Table 6, they still indicate differences from basic tests on samples. Institution difference (cgzcfl) coefficient are both obviously negative in results of production and sales OFDI enterprises, and resource development OFDI enterprises. Meanwhile, it can be seen that resource development enterprises are mostly impacted by institutional differences. Ramasamy’s research on China’s outward investment found out that seeking for natural resources is an important reason for Chinese enterprises to make outward investments. Buckley also found out that in host countries with worse contexts, Chinese enterprises are inclined to seek for more natural resources. This paper’s research results are different previous ones. They indicated that resource-orien- tated Chinese enterprises were more inclined to make investments in host countries with less institution differences from China’s. In contrast, institution difference coefficient does not pose obvious impacts on commercial service and technology research and development OFDI enterprises sample. Overseas Chinese network (llaborc) and interactive items (cgzcfl *llaborc) do not pose obvious impacts on technology research and development OFDI enterprises sample. It means that overseas Chinese network in host countries do not impact outward foreign direct investment of such enterprises. Although overseas Chinese network promoted development of outward foreign direct investment for commercial service enterprises, it does not obviously impact closing gaps in regulatory effects.
5. Robustness Tests
This paper conducts robustness tests on labor dispatch (laborl) and empirical results. Test results are shown in Table 7. Based on regression results, regardless of core explanatory variables or regulated variables, coefficients and significance of multiplied items do not indicate essential changes. As a result, test conclusions are still robust.
Table 6. Test on investment motivations.
t statistics in parentheses *p < 0.05, **p < 0.01, ***p < 0.001.
Table 7. Robustness tests.
t statistics in parentheses *p < 0.05, **p < 0.01, ***p < 0.001.
6. Conclusions and Inspiration
Based on zero-expansion negative binomial regression method for corporate OFDI project data from 2003 to 2014, this paper conducts empirical analysis on influence of institutional differences on Chinese enterprises’ outward foreign direct investment on countries in “The Belt and Road”, as well as regulatory effects between labor dispatching of these countries and regional choices for Chinese enterprises’ outward foreign direct investment. Moreover, according to four main purposes of Chinese enterprises’ outward foreign direct investment, the paper conducts related classification, tests and studies on commercial services, local production and sales, resource development, as well as technology research and development OFDI enterprise, as well as basic samples. Studies reveal: firstly, institutional differences and Chinese enterprises’ OFDI are negatively related. Even if for investments countries in “The Belt and Road”, Chinese enterprises’ outward foreign direct investment still involves “institutional proximity”. Secondly, overseas Chinese network in host countries reduces impacts of institutional differences on regional choices for Chinese enterprises’ outward foreign direct investment. In host countries with more active network of overseas Chinese, despite bigger institutional differences, Chinese enterprises are still more inclined to make investments. Overseas Chinese network in host countries can help Chinese enterprises overcome business barriers caused by institutional differences, so as to expand investments. Lastly, impacts on institutional differences, overseas Chinese network, commercial service, technology research and development enterprises’ outward foreign direct investment are not significantly obvious. It means that these enterprises tend to focus on market volume, technical context, as well as research and development conditions, instead of labor dispatching in host countries.
Based on the above conclusion, we should respond to the national initiative of opening to the outside world, comply with trends of world multipolarization, economic globalization, cultural diversification and social informationalization, adhere to the spirit of regional cooperation, and dedicate to maintaining the global free trade system. For investments on countries in “The Belt and Road”, Chinese enterprises also takes the regional element of institutional differences into consideration, so as to reduce investment costs, and risks of legal disputes. Due to economic globalization, as well as political and economic integration, underdeveloped countries learn from advanced countries. As a result, over the long run, reducing the institutional differences can promote China’ outward foreign direct investment on the whole. In the meantime, as for institutional differences, we can attempt to reduce them from non-economic perspectives. For instance, increase labor dispatching to countries in “The Belt and Road” and enhance labor flow across countries. Moreover, setting up more convenient and fast convenient and fast, as well as improving investment context can also provide more resources and information for Chinese enterprises investments on countries in “The Belt and Road”. In addition, for enterprises that focus on the motivation of technology research and development, they are also less likely to be negatively influenced by institutional differences and related economic losses. Instead, they can focus more attention on selecting the most optimal location for product research and development among regions and countries in “The Belt and Road”. In this way, it can better promote and stimulate innovation, technical enhancement, as well as research and development efficiency for parent enterprises in China.