The Impact of Hong Kong, Macao and Taiwan Investment on Enterprise Innovation in the Pearl River Delta Region

Based on the micro-enterprise data of the nine cities in the PRD in the database of China’s industrial enterprises from 1998 to 2007, from two aspects of innovation input and innovation output, this article analyzes the impact of Hong Kong, Macao and Taiwan capital entering the Pearl River Delta region on the innovation capabilities of enterprises in the region. The empirical results show that at the enterprise level, the company’s R & D investment and new product output decrease with the increase of Hong Kong, Macao and Taiwan capital entry. At the industry level, the deeper the entry of Hong Kong, Macao and Taiwan-funded industries, the stronger the inhibitory effect on corporate R & D investment and new product output. Combining the results of these two aspects, the estimation results of this article show that the entry of Hong Kong, Macao and Taiwan capital into the Pearl River Delta region has a negative effect on the innovation ability of enterprises.


Introduction
Innovation ability is an important source and driving force for determining whether a country or region can achieve technological progress, industrial structure upgrade, and economic growth is an indispensable condition for achieving sustainable development. The theory of technology spillovers believes that FDI can directly or indirectly bring technological transfer to host country enterprises.
Will the entry of foreign capital have a positive impact on the innovation capa-article focuses on whether the entry of Hong Kong, Macao and Taiwan capital into the Pearl River Delta region will help strengthen the Pearl River Delta on the independent innovation capabilities of enterprises.
The rest of the paper is arranged as follows: the second part reviews and summarizes the relevant literature; the third part is data description and the definition of variables; the fourth part is model setting and empirical analysis; the fifth part summarizes the full text.

Literature Review
The research in this paper is mainly related to two literatures, one is the literature on the factors that affect the innovation ability of enterprises, and the other is the literature about how foreign direct investment affects the innovation of enterprises. Summarizing previous research, we can divide the factors affecting corporate innovation into two major categories: First, the company's own characteristic variables, including enterprise size, financing constraints, capital intensity, per capita wages, human capital, ownership factors, and the degree of foreign ownership etc.; Second, external environmental variables, including exports, environmental factors of external demand, market competition, government contact factors, etc. (Zhang et al., 2007;Nie et al., 2008;Kang, 2011;Zhang et al., 2012;Wu, 2012). Among them, the impact of FDI on corporate innovation has been measured from data from different perspectives and at different levels.
Generally, foreign capital entering has direct and indirect effects on the host country, among which indirect effects include competition effects, technology spillover effects, and related effects. The combined effect of these effects makes the following three views on the impact of foreign investment on domestic enterprise innovation.

Promotion Theory
Scholars who advocate that foreign capital entry will promote the independent innovation capabilities of national enterprises believe that foreign direct investment mainly promotes the improvement of technological innovation capabilities in host countries through the indirect effects of competition effects, technology spillover effects, and front-to-back industry connections. For example, Wang et al. (2006) verified that the driving force for the improvement of independent innovation capabilities of national enterprises in the host country came from the competitive effect of international capital's participation in the local market, using local panel data from 1998 to 2003. It is inevitable to increase research and development funds to improve the technological level of the enterprise. Jiang & Xia (2005), Tong & Yan (2005) demonstrated the conclusion that foreign capital will promote independent innovation of national enterprises from the perspective of technology spillover effects. The former takes the domestic high-tech industry as a research object, and believes that FDI will promote the R & D activities of domestic enterprises through demonstration and imitation effects, scien- uses relevant data at the inter-provincial level and takes the number of patent applications as a proxy for innovation capabilities, and concludes that foreign investment has a significant positive spillover effect on the number of patent applications in China. Zhang (2008) also verified that foreign capital has a significant spillover effect on the independent innovation output of Chinese enterprises from the perspective of innovation input and innovation output. On the other hand, Qiu et al. (2008) and Wang et al. (2010b) etc. have verified the independent innovation effect of foreign capital on enterprises in the host country based on the perspective of industrial association. Qiu et al. (2008)  The above literature is based on empirical analysis using macro-statistical caliber data, and most of them have obtained optimistic conclusions between foreign investment and corporate innovation. Turning the perspective to the micro level may lead to different conclusions. Du & Li (2016) use panel data at the enterprise level, and the research results show that foreign investment entered in general and through mergers and acquisitions will promote domestic companies' technological research and development and product innovation. The promotion of R & D innovation has been significantly reduced. Guadalupe et al. (2012) used data from Spanish manufacturing companies from 1990 to 2006 and found that domestic companies acquired by multinational companies can take advantage of the lower innovation costs of multinational companies to increase their innovation activities. However, some literatures have used microdata to conclude that FDI can inhibit the innovation capabilities of domestic enterprises. Fan et al. (2008) Wang & Zhang (2012) concluded that FDI has no obvious effect on the technological progress of Chinese domestic enterprises. Some scholars believe that the effect of FDI is not obvious because of the "threshold effect".

Model
We use the following equation to estimate the impact of the entry of Hong Kong, Macao and Taiwan capital on the innovation of enterprises in the Pearl River Delta region. The model constructed in this paper is as follows: In the above formula, Y is an index reflecting the independent innovation of the enterprise; ε is a random error term. HMT_Firm it and HMT_Sector it is the natural logarithm of Hong Kong, Macao and Taiwanese investment in enterprise i and industry j, respectively. Because company-level FDI and industry-level FDI have a cross-complementarity to improve a company's R & D efficiency (Fan et al., 2008), this article adds HMT_Firm it *HMT_Sector it interaction terms to the model to test how the impact of Hong Kong, Macao and Taiwanese direct investment on enterprises at the industry level varies with the degree of participation of different enterprises in Hong Kong, Macao and Taiwan. X represents the set of control variables for the characteristics of the enterprise. Sector, city and year represent dummy variables of industry, city, and year, which are used to control the differences between different industries, cities, and years.

Data
The data in this article comes from the database of China's industrial enterprises above designated size from 1998 to 2007, which is widely used by economists. As in the existing literature, when processing data, refer to the data processing method of Brandt et al. (2012) to match the original data and delete error records that do not conform to the basic logical relationship. such as: 1) Key indicators are missing; 2) The main business income is less than 5 million RMB, the number of employees is less than 8 or the R & D expenditure is negative; 3) The total assets are less than the total of current assets, the total of fixed assets, and the average annual balance of fixed assets. In addition, according to the research purpose of this article, the cities are coded into the cities of Guangzhou, Shenzhen,Zhuhai,Foshan,Jiangmen,Zhaoqing,Huizhou,and Zhongshan,represented by 4401,4403,4404,4406,4407,4412,4413,4419,and 4420. The enterprise data is extracted to form the enterprise data set of the nine cities in the Pearl River Delta. Then it analyzes the impact of Hong Kong, Macao and Taiwan capital entering the Pearl River Delta on the innovation capabilities of enterprises in the

Variable Definitions
Most literatures mainly select the indicators of enterprise R & D investment, new product output, and the number of enterprise patents when selecting enterprise innovation indicators. Since the data of the number of patents of enterprises in the database of Chinese industrial enterprises is not available, this article uses the two indicators of "research and development expenditure" and "new product output value" in the database to measure corporate innovation. Drawing on the practices of Nie et al. (2008), Kang (2011) and others, in order to more fully reflect the innovation situation of enterprises, we adopt two different ways to reflect the independent innovation of enterprises. One is the innovation decision of the enterprise, which is represented by dummy variables. It takes a value of 1 when its new product output is greater than zero, a value of 0 in other cases, a value of 1 when its R & D expenditure is greater than zero, and a value of 0 in other cases; the other is innovation density, which takes the output value of new products/total industrial output value as a proxy variable for measuring innovation output, and uses R & D investment/sales as a proxy variable for measuring In the regression of this paper, the natural logarithm is adopted to reduce the sample loss caused by some variables being zero. Table 1 gives definitions and simple statistical descriptions of all variables.

Basic Regression Analysis
Due   Notes: The data in parentheses are t statistics; Y and N represents the dummy variables of the added and un-added, respectively; ***, **, and * indicate that the coefficient estimates are significant at 1%, 5% and 10% confidence levels, respectively; The following table is the same.
shares on innovation decisions in the PRD. First, we take the output value of the new product as the dependent variable without adding any control variables (column (1) in Table 2  Macao and Taiwan capital in the enterprise and the share of Hong Kong, Macao and Taiwan capital in the industry increase, the possibility of independent innovation by the enterprise will be reduced. In column (2), we controlled the factors such as export intensity, capital intensity, enterprise size, per capita wages, human capital, financing constraints, corporate profitability, and corporate age. The coefficient is still significantly negative. However, compared to the results in column (1), the coefficients of the share of Hong Kong, Macao and Taiwan capital of enterprises are not much different, and the estimated coefficient of the share of Hong Kong, Macao and Taiwan capital of the industry has increased from −0.958 to −0.321, indicating that it does not control some of the companies that affect corporate innovation Under the influence of factors, the inhibitory effect of the share of Hong Kong, Macao and Taiwan capital on the innovation output of the industry has been overestimated. By adding dummy variables of industry, city, and time, column (3) further controls the fixed differences between industries or regions and time, and adds the interaction between the Hong Kong, Macao, and Taiwan capital shares of enterprises and the Hong Kong, Macao, and Taiwan capital shares of industries. The results show that the interaction term coefficient is greater than 0, but it is not significant. The share of Hong Kong, Macao and Taiwan capital in enterprises and the share of Hong Kong, Macao and Taiwan capital in industries are significantly less than 0. Moreover, the coefficients of Hong Kong, Macao, and Taiwanese capital of enterprises and industries are smaller than the coefficients in column (2), which indicates that after further control, industry, city, time and other factors, the inhibitory effect of Hong Kong, Macao and Taiwanese capital on corporate innovation has increased, Product output tends to decline. When the explanatory variable is replaced by a binary selection variable of whether the company has independent R & D investment, the estimation results are listed in columns (4) to (6). The results show that without any control varia-ble (column (4)), the estimated coefficients of the shares of Hong Kong, Macao and Taiwan capital of enterprises and the shares of Hong Kong, Macao and Taiwan capital of industries are significantly negative. When controlling export intensity, capital intensity, enterprise size, per capita wages, human capital, financing constraints, corporate profitability, and corporate age (column (5)), the estimated coefficients of the Hong Kong, Macao and Taiwan capital shares of enterprises and the Hong Kong, Macao and Taiwan capital shares of industries It is still negative and not much different from the estimated coefficient in column (4). After further controlling the factors such as industry, city, time, etc., the estimated coefficient of foreign investment in enterprises is significantly less than 0, but the coefficient of Hong Kong, Macao and Taiwan investment in the industry has changed from negative to positive, but not significantly. The coefficients of the interaction terms are negative but not significant. Table 3 repeats the analysis of  Table 2 and Table 3, the impact of the entry of two types of Hong Kong, Macao, and Taiwanese capital on the innovation density of PRD enterprises is basically the same as that of their impact on corporate innovation decisions. However, as can be seen from the result in column (3), after controlling the characteristics of the enterprise and the industry, city, and time variables, the coefficient of the interaction term is significantly positive. It shows that the increase in the total amount of Hong Kong, Macao and Taiwan capital in the industry will increase enterprises' investment in technology research and development. Table 2 and Table 3 have two important results. First, Hong Kong, Macao, and Taiwan capital at the enterprise level have a significant negative impact on the innovation of an enterprise (∂ 1 < 0), and this impact is reflected in both the Probit model and the Tobit model. In other words, the more Hong Kong, Macao, and Taiwanese capital an enterprise participates in, the less input and output it has for innovation. This is most likely because Hong Kong, Macao and Taiwanese investors have better technology than the Pearl River Delta enterprises. The greater the Hong Kong, Macao and Taiwan investment participation of an enterprise, the more technology transfers from its Hong Kong, Macao, and Taiwanese business partners may reduce the need for its own technological innovation. Secondly, we also found that the correlation coefficient of the Hong Kong, Macao and Taiwanese investment share of the industry is negative and significant. Except when controlling corporate characteristics and virtual variables such as industry, city, time, etc., the impact of industry Hong Kong, Macao and Taiwan investment on corporate R & D investment and R & D investment decisions not obvious. In other words, the more Hong Kong, Macao, and Taiwanese capital participation in the industry, the greater the negative impact of Hong Kong, Macao, and Taiwanese capital on industry R & D.

Robustness Test
The results of corporate R & D activities are often difficult to achieve immediately. There may be a long lag from the time when enterprises invest in research and development to the launch of innovative products. Therefore, many companies currently engaged in research and development activities may not be able to make breakthroughs in product innovation. Considering the lag of the influence of these factors on innovation activities and reducing Endogenous problem, all explanatory variables are lagging one period. The regression results are shown in Table 4

Conclusion
This article establishes a simple model, based on the micro-enterprise data of the Pearl River Delta, and analyzes empirically the role of Hong Kong, Macao and Taiwan capital entering the Pearl River Delta on the independent innovation input and output of enterprises in the region. Studies consistently show that the estimated coefficients of the Hong Kong, Macao and Taiwanese shares of enterprises are significantly negative. This means that if the proportion of Hong Kong, Macao and Taiwanese capital in the company increases, the company will reduce R & D investment and new product output. The conclusion of this study has important implications for the adjustment of the investment promotion policy in the Pearl River Delta region. In response to the above conclusions, this article proposes the following policy recommendations: Firstly, Although innovation is one of the key factors in the new round of industrial structure transformation and growth mode transformation, we cannot ignore the role of Hong Kong, Macao and Taiwan capital in rapidly upgrading the technological level of the Pearl River Delta because we are worried that the entry of Hong Kong, Macao and Taiwan capital may inhibit independent innovation of enterprises. A more rational approach is to balance the impact on technology level and independent innovation when assessing the contribution of Hong Kong, Macao, and Taiwan capital to the Pearl River Delta, and consider adopting targeted methods to promote independent innovation of enterprises in order to avoid Hong Kong, Macao, and Taiwan capital's autonomy to enterprises Inhibition of innovation. Improving the financial environment to increase the availability of funds and reduce corporate financing costs, thereby optimizing corporate liquidity, will be one of the effective means to achieve this goal.
Secondly, the input of innovative manpower can significantly improve the scale and level of innovation activities of enterprises in the Pearl River Delta region. The government should further increase investment in education and increase the accumulation of knowledge in the Pearl River Delta region, lay a solid foundation for the introduction of high-level FDI enterprises. On this basis, encourage domestic-funded enterprises to attach importance to employee training, and attract talents with important skills to enter local enterprises from foreign-funded enterprises, in order to promote the improvement of local enterprises' independent innovation capabilities.

Conflicts of Interest
The author declares no conflicts of interest regarding the publication of this paper.