Research on the Investment Effect of Chinese Cultural “Going Out” —Empirical Evidence from the Countries along the “Belt and Road”

This paper is based on 2006-2017 relevant data of 56 countries along the “Belt and Road” to find out the investment effect of cultural “going out”. The empirical results show that: cultural “going out” has a significant promotion effect on China’s foreign direct investment; this effect has national heterogeneity, which is more effective for emerging markets and developing countries; for some traditional influence factors, such as traditional GDP per capita and time for diplomatic relations, cultural “going out” has a substitution effect. Finally, according to the empirical results, put forward targeted suggestions.

the information asymmetry in the international market [2], which is conducive to the local production and business activities of Chinese-funded enterprises. As shown in Figure 1.

Analysis of Influencing Factors of China OFDI
Dunning's [3] international production trade-off theory and monopoly advantage theory are paradigms for studying the factors or drivers of OFDI. Dai Zhongqiang [4] empirically analyzed that the above theories have sufficient explanatory power for China's OFDI. Factors such as economic development level, geographical distance, natural resources, infrastructure, and technological development level have been widely verified [5] [6] [7] [8] [9].
With the rise of new institutional economics (NIE), analyzing the influence factors of China's OFDI from institutional factors has become a new research perspective. Culture is an important part of the informal system, and the cultural difference between China and the host country is also an important factor affecting China's OFDI. Lankhuizen et al. [10] believe that cultural differences are conducive to market segmentation of multinational companies, thus promoting OFDI. Zhang Jipeng and Li Ning [11] use empirical analysis of enterprise panel data to verify that cultural distance has a negative impact on OFDI of Chinese  Jianhong et al. [12] found that there are threshold effects on the influence of cultural differences on OFDI.

Research on the Impact of Cultural "Going Out" on Trade and Investment
Some scholars regard the cross-border movement of talents as a bridge for cultural "going out". Among them, Wei Hao and Chen Kaijun [13] found that the cross-border movement of talents will reduce the transaction costs and opportunity costs of international trade; Gu Yuanyuan and Qiu Bin [14] studied the interactive relationship between overseas education and China's foreign direct investment, they believed that after returning to motherland, foreign students can penetrate Chinese culture into the host country's culture, thereby accelerating the promotion of China's foreign direct investment. Another scholar used the Confucius Institute as a proxy variable for cultural "going out". Lian Daxiang [13] found that compared with export trade, the Confucius Institute has a more significant positive effect on OFDI; Chen Yinmo et al. [14] from the perspective of home culture promotion, examine the influence mechanism of Confucius Institutes on OFDI of Chinese companies in countries along the "Belt and Road"; Xie Mengjun [1] used differential GMM and systematic GMM methods to verify that the investment promotion effect of Confucius Institutes has obvious Economic heterogeneity and intercontinental heterogeneity.
There are many literatures that studied the influencing factors of OFDI in

Sample Selection
According to the statistics of the "ydyl.  (Table 1).

Variable Selection
This article selects China's OFDI flow to the host country as the explained variable.   Table 2.

Model Setting
Among them, OFDI, CI and other variables are described in Table 2, β is the model parameter, i is the country, t is the year, λ is the time effect, and ε is the random disturbance term.

Estimation Method
Since the mean-variance inflation factor (Mean VIF) of the explanatory variables is 2.51, and the variance inflation factor (VIF) of each explanatory variable is less than 10, there is no serious multicollinearity problem. The Hausman test results indicate that a fixed-effect model is used. One of the explained variables, the geographic distance does not change with time in a country, so this article uses the Least Squares Dummy Variable Model (LSDV) for regression, introduces the time dummy variable, and uses the clustering robust standard error, which is consistent with the methods of Qi Jianhong et al. [12], Gu Yuanyuan and Qiu Bin [14]. This paper uses Stata14.0 software for data processing and analysis.

Regression Results and Analysis of the Full Sample
The results in Table 3

Regression Results and Analysis of Sub-Samples
According to the classification of countries by the International Monetary Fund (IMF), the countries along the B & R are divided into two samples for emerging markets and developing countries and developed countries, as shown in Table 1.
This article builds two models separately for two samples, as shown in Table 4.

Increase the Regression Results and Analysis of Interactive Items
Wang Yongqin et al. [17], Gu Yuanyuan and Qiu Bin [14] all use the interaction terms to examine the interaction between variables. This paper draws on its ideas and introduces interaction terms into the regression model. In Table 5, 10% significance level, which shows that there is a substitution effect between cultural "going out" and GDP per capita, that is cultural "going out" can reduce

Conclusions and Recommendations
In this paper, the number of Confucius Institutes (Classrooms) is used as a proxy variable for cultural "going out", using the relevant data of 56 countries