Since Schumpeter (1912) proposed innovation theory, most economists have taken as a given that innovation is particularly important for the economic development. From the macroscopic view, innovation is essential for the sustained growth of national economy, and is the key to determine international status and competitiveness. From the microcosmic view, enterprises want to survive in the fierce market competition; it is necessary to innovate. In the past 20 years, China’s R&D expenditure has been rising, according to the “2017 National Science and Technology Investments Statistics Bulletin”, in 2017, China’s investments in research and development of a total of 1.7606 trillion yuan, increase 12.3% over the previous year, investments intensity is 2.13%. Although since 2014, R&D Investments intensity of China has been more than 2%, it is also far from the OECD countries average R&D investments intensity of 2.4%, the government and enterprises need to increase investments in R&D continually.
R&D requires constant investments of funds, but, the internal funds of small and medium-sized enterprises, which we use SMEs to represent in the following chapters, are insufficient and external funds are required. On the one hand, with small size and a few assets, most SMEs cannot meet the loan conditions of banks, they are difficult to obtain bank loans, or have a high cost for debt financing. On the other hand, there is a high listing threshold and long IPO cycle in China, the cost of equity financing is high too. Data shows that by the end of October 2018, the average queue time for listing is about 15 months, and the average cost exceeds 70 million yuan, lack of funds has become a prominent problem hindering the research and development of SMEs.
Technology and knowledge have the spillover characteristics of public goods, R&D activities will inevitably encounter market failure and underinvestment (Tassey, 2004)  , therefore, government support is the key to influence the corporate R&D activities. Theoretically, there may be crowding out or motivating effects of government subsidies on corporate R&D investments. On the one hand, Government subsidies are classified as part of an enterprise’s income, and can be directly invested in R&D activities to compensate for the lack of marginal returns in the private sector, alleviate the externalities of R&D activity, then have a stimulating effect on R&D investments (Weiming Jie et al. 2009)  . Also, according to signaling model, government support can send positive signals to the market, reduce the information asymmetry between enterprises and investors, reduce the cost of debt financing, promote enterprises to obtain more external financing, increase R&D activities. On the other hand, government subsidizes specific enterprises, which will promote the hitchhiking behavior of other enterprises in the same industry, and increase the cost of R&D activities, then crowding out some of the private R&D investments.
This paper uses the 2013-2017 panel data of SMEs in China to study, focusing on whether government subsidies will promote the R&D investments of SMEs? How government subsidies affect the relationship of corporate financial structure and R&D investments? And considering that the cost of obtaining funds is different between state-owned enterprises and non-state-owned enterprises, we further regard the property rights as an important factor in this study. This paper contributes to the existing study in the following two respects. First, we supplement the literature on government subsidy, financial structure and corporate R&D investment. Second, our finding provides theoretical support for government to formulate relevant policies.
The rest of this paper is organized as follows. Section 2 presents hypothesis development. Section 3 discusses the data and empirical model. Section 4 presents a discussion of main results. Section 5 concludes.
2. Literature Review and Hypotheses Development
In order to study the relationship between government subsidy, financing structure and R&D Investment, this paper first review the relevant literature, then analyze theoretical relationship between financing structure and R&D activities, and the theoretical relationship between government subsidies and R&D activities, finally, further consider the interaction effect of government subsidy and financing structure.
2.1. Financial Structure and R&D Investments
Pecking order theory holds that, in a market with asymmetric information but no friction, the cost of internal financing is lower than that of external financing, and equity financing will convey the negative information of the company’s operation, therefore, when the company is financing, it will give priority to internal financing, followed by debt financing, and finally equity financing (Myers & Majluf, 1984)  . The investments behavior of an enterprise is limited by the cost of internal financing and external financing, therefore, internal financing is the main source of investments for R&D investments. Hall (2002) finds that small companies experience high costs of R&D capital, and there is a positive relationship between internal cash flow and R&D investments  . Himmelberg and Petersen (1994) find that internal financing is the main source of R&D investments of small high-tech enterprises  . Jie Zhang et al. (2012) find for China that the internal cash flow and registered capital are the main source of funding for SMEs’ R&D investments  . Most of studies indicate that internal funds will promote R&D investments, while financing constraints will deter R&D investments. Based on this, this paper puts forward the following assumption:
H1: Internal financing has the greatest positive effect on R&D investments, followed by debt financing and equity financing.
2.2. Government Subsidies and R&D Investments
The impact of government subsidies on R&D has been the focus of academic debate, based on the theory of State intervention, most scholars argue that government should properly intervene the market, improve the effective demand for R&D activities in the private sector. However, some scholars believe that government subsidies have substitution effect and crowding out effect on R&D investments. Also there are some scholars believe that the motivation effect and crowding out effect coexist.
Because of the positive externality of R&D activities, marginal returns from private R&D investments are lower than the marginal benefits from social R&D investments, therefore, it will appears to market failures if relying solely on business investments (Arrow, 1962)  . So we need government departments to give redress and remedy by financial subsidies, and government subsidies have a significant stimulating effect on corporate R&D investments (Weimin Jie, 2009). There is a large literature that establishes that government subsidies can motivate enterprises to invest in R&D activities (Hamberg, 1966; Scott, 1984). The higher the technical level of the industry, or the larger the scale of enterprises, the worse the incentive effect of government subsidies on corporate R&D activities (Lach, 2002)  . There is also some literatures believe that, government subsidies for specific businesses will create free rider problems, and raise the price of relative factors, then crowding out parts of the private R&D investments (Shrieves, 1978; Lichtenberg, 1984)   . Zheng (2009) based on a study of China’s high-tech industry, find that government R&D activities will replace or crowd out some of the private R&D activities, and will weaken the incentive effect of fiscal policy  . A few studies show that, in China, the motivation effect and the crowding out effect of government subsidies coexist, and within a certain range, the motivation effect of government subsidies exceeds the crowding out effect (Gorg and Strobl, 2006)  . Based on this, this paper makes the following assumptions:
H2a: Government subsidies have stimulating effect on R&D investments of SMEs;
H2b: Government subsidies have crowding out effect on R&D investments of SMEs.
2.3. Government Subsidies, Financial Structure and R&D Investment
Due to the high uncertainty of innovation, information asymmetry exists between R&D activities and external investors. Government subsidies make efforts on complementing internal cash flows, alternatively they can also delivery positive signals to the market, eliminate information asymmetry in some way, then more external capitals invest in enterprises’ R&D activities (Zhiyong Kang, 2013)  . Czarnitzki (2006) finds that, because of the high amount of government subsidies, the R&D investments of East German firms is not sensitive to the external financing constraints, by using Tobit model  . Xiaofang Bi (2017) suggests that government subsidy and financial redundancy can significantly stimulate the investments of firms’ R&D, and they a complementary effect  . Based on this, this paper makes the following assumption:
H3: The more the government subsidies companies received, the more willing they are to invest the money from other financing channels into R&D activities.
To summarize, researches about the influence of government subsidy on enterprise innovation have produced abundant theoretical and empirical results, but because of the different methods or samples used, there is not yet a unified conclusion, and researches about the impact of financial structure on R&D investments are rich. However, there are only few studies combine these two factors, and deeply explore the specific links between government subsidies, financial structure and R&D investments. In view of this, this paper will combine two factors of government subsidy and financial structure, discuss how government subsidies affect the financial structure, thus further affecting R&D investments.
3. Data and Methodology
3.1. Sample Construction
In this paper, we use the annual financial data of firms listed in SME board from 2013 to 2017 as the initial sample, on the basis of the initial sample, the following processing is done: 1) We removed ST or *ST companies; 2) Eliminated observations with incomplete data; 3) Winsorized main variables. The final sample is 3570 observations of 714 companies listed in SME board, the data mainly from the wind database, processing with Stata.
3.2. Variable Construction
This paper mainly studies the influence of government subsidy and financing structure on the R&D investment of SMEs, therefore, the dependent variable of this paper is the R&D investments, which is measured by R&D investments divided by total assets. The main independent variables are government subsidy and internal and external financing, in this paper we measure government subsidy (Sub) as current government subsidies of SMEs in financial statement. We uses the net cash flow of business activities to measure internal financing, and debt financing (DF) is measured by short term loan plus long term loan and payable bonds, equity financing is measured by issued capital plus capital reserve.
Taking into account common method in this area, this article also adds variables that may have a significant impact on R&D activities as control variables: enterprise size (Size), enterprise age (Age), equity concentration (Con), profitability (Roe), leverage ratio (LEV), year virtual variable (years) and industry virtual variable (Ind). Details are provided in Table 1.
Because the dependent variable is the R&D investments, but some R&D investments of SMEs are 0, there is a problem of 0 points deletion, therefore, this paper uses Tobit model to study the impact of government subsidy and financing structure on enterprises’ R&D investment, model 1 and model 2 used to test hypothesis 1 and hypothesis 2, then, in order to study the interaction between government subsidy and enterprise financing structure, so we add interaction items into model 2, and we build model 3 to test hypothesis 3, we the model as follows:
Table 1. Variable definitions and calculations.
3.4. Descriptive Statistics of Variables
Following analysis is based on the balanced sample, which tracks the same firms over 5 years. Table 2 reports the means and standard deviations of main variables. The mean of R&D investments relative to total assets is 2.28%, and minimum value is 0, maximum value is 16.21%, which Indicates that the R&D investments of most SMEs are at a low level. The mean of debt financing is lower than internal financing and equity financing, but standard deviation of debt financing is the highest, this may be due to the fact that, the SMEs have many obstacles to obtain the credit support, such as lacking collateral. The values of the main variables are within a reasonable range, so the regression results are less affected by outliers.
4. Empirical Results
4.1. Government Subsidies, Financial Structure and R&D Investments of SMEs
This paper first studies the impact of government subsidies on R&D Investments,
Table 2. The description of the main variables.
The result in the first column of Table 3 shows that, the regression coefficients of internal financing, debt financing and equity financing are significantly positive at 1% confidence level without taking into account of government subsidies, indicating that internal financing, debt financing and equity financing all have positive effects on innovation investments of SMEs, and The regression coefficient of internal financing is greater than that of debt financing and equity financing, which means that R&D investment is more sensitive to internal financing than debt financing and equity financing, China’s SMEs tend to give priority to the use of internal funds for R&D activities, followed by debt financing, and finally equity financing, hypothesis 1 is verified. Then we studies the impact of government subsidies on R&D investments under the condition that financing structure remains unchanged, in the second column of Table 3, we add the variable of government subsidy, the result shows that both the government subsidy and the enterprise’s internal and external financing variable regression coefficient are significantly positive at 1% confidence level, therefore, government subsidies can significantly stimulate the R&D investments of enterprises, and confirmed hypothesis 2a.
Through the above analysis, this paper has found that different financing methods have a positive impact on the innovation activities of enterprises, and then, this paper further studies the interaction between government subsidies and different financing methods. In columns 3, 4 and 5 of Table 3, the regression coefficients of the interaction items are significantly positive, indicating that the government subsidy has a significant positive moderating effect to the relationship of financing and R&D investment, hypothesis3 is established.
4.2. Regression Results of State-Owned and Non-State-Owned Enterprises
Considering that there are obvious differences in access to financial resources and government supports between enterprises of different ownership, it’s necessary
Table 3. The effect of government subsidy and financial structure on R&D investment.
Notes: Standard errors in parentheses and aren’t robust standard errors; *10% significance level; **5% significance level; ***1% significance level.
to test the influence of government subsidy and financing structure on the R&D investment of different ownership enterprises. In order to test the hypothsis3, Table 4 divides the samples into two groups, state and non-state. From the return results of state-owned enterprises and non-state-owned enterprises, for non-state-owned enterprises, the regression coefficients of Sub*IF, Sub*Debt and Sub*Equity, are significantly positive, which shows that government subsidy has obvious stimulating effect on the innovation financing of non-state-owned enterprises, that is, the greater the government subsides, the more the SMEs are willing to invest in R&D activities. For state-owned enterprises, the regression coefficient of interaction between government subsidy and internal financing is not significantly negative, indicating that the more the government subsidies,
Table 4. Regression results of state-owned and non-state-owned enterprises.
Notes: Standard errors in parentheses; *10% significance level; **5% significance level; ***1% significance level.
the less willing the enterprises are to invest internal funds in R&D activities, the coefficient of interaction between government subsidy and debt financing is significantly positive, indicating that government subsidies can promote more debt financing for R&D activities. The difference may be that, the Chinese government, in constructing the financial system, gives more care to state-owned enterprises on the basis of ideological considerations, forming a resource allocation of discrimination against non-state-owned enterprises. The biased financial support of the financial system to state-owned enterprises leads to the more serious shortage of funds for the R&D investments of non-state-owned enterprises.
4.3. Robustness Tests
In order to test the robustness of research conclusions, firstly, we uses the fixed effect model and the random effect model to estimate the impact of government subsidy and financial structure on R&D investment, Table 5 shows the empirical results of fixed effects model. Table 6 shows the empirical results of random effects model. As we can see from the following tables, the regression coefficients of internal financing, debt financing and equity financing are still significantly positive, and government subsidies have a significant positive moderating effect to the relationship of financing and R&D investments, the conclusions has not been greatly affected, so our main conclusions remain valid.
Secondly, there may be endogenous problems between government subsidies and R&D investments. On the one hand, government subsidies can affect R&D investments, and on the other hand, corporate innovation will in turn influence government subsidies (Lichtenberg, 1988). Therefore, we use different variables
Table 5. Results of fixed effects model.
Table 6. Results of random effects model.
to further test the impact of government subsidies and financing structure on enterprise R&D investments, using the same method as Huidong Li et al. (2013)  , we use the registered city and ownership as agent variables of government subsidies, if the company is registered in prosperous cities, the government has more ability to provide subsidies to enterprises, at the same time, enterprises in these places have easier access to external resources, in addition, there is no direct relationship between the registered place of enterprises and the R&D investments. Then we use the Heckman two stage regression, Table 7 shows the results, the conclusions are still hold.
In recent years, the intensity of China’s R&D investments is rising, but there is still a gap with the world level, because the capital market is not perfect, China’s small and medium-sized enterprises innovation financing channels blocked,
Table 7. Results of Heckman two step regression.
government subsidies as an important source of enterprise innovation Investments, can alleviate the financing difficulty of enterprise innovation activities, and stimulate enterprises to invest in innovation. Based on the theory of government intervention and the theory of optimal order financing, this paper uses the empirical data of Chinese small listed companies from 2013 to 2017, discusses the relationship between government subsidy, financing structure and SMEs’ R&D investments. The study finds that: 1) Both internal financing and external financing have positive effect on the R&D investments of SMEs, among which the effect of internal financing is the greatest, followed by debt financing and equity financing. 2) Government subsidies and corporate financing all can significantly influence the R&D investments of enterprises. 3) Government subsidies have a significant positive moderating effect to the relationship of financing and R&D investments, and government subsidies can release positive signals, then enable enterprises to obtain more external financing for their R&D activities.
R&D activities of SMEs in China are faced with the problem of insufficient capital and external financing channels. Government support and financial system reform are effective ways to alleviate the financing difficulties and promote investments in innovation of small company. The research in this paper shows that government subsidy has a significant positive moderating and, it can effectively alleviate the problem of insufficient innovation funds, and then promote the development of small and medium-sized enterprises. The government should strengthen financial support for R&D activities of SMEs, deepen financial system reform, eliminate discrimination in financial markets, and realize the effective allocation of financial resources, improve financing environment for SMEs innovation, and encourage SMEs to play a leading role in technological innovation.
 Myers, S. and Majluf, N. (1984) Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not. Journal of Financial Economics, 13, 187-221.
 Himmelberg, C.P. and Petersen, B.C. (1994) R&D and Internal Finance: A Panel Study of Small Firms in High-Tech Industries. The Review of Economics and Statistics, 76, 38-51.
 Zheng, X.T. (2009) The Influence of Public Policy to Private R&D Activities: An Empirical Analysis Based on 1995-2006 Years of High-Tech Industry in China. Forum on Science and Technology in China, 3, 29-33.
 Czarnitzki, D. (2006) Research and Development in Small- and Medium-Sized Enterprises: The Role of Financial Constraints and Public Funding. Scottish Journal of Political Economy, 53, 335-357.
 Li, H.D., Tang, Y.J. and Zuo, J.J. (2013) Innovate with Corporate Internal Funds or External Funds? Research on Financing Structure and Innovation of Chinese Listed Companies. Journal of Financial Research, 2, 170-183.