- Email: [email protected]
University R&D activities and firm innovations
Journal Pre-proof
University R&D Activities and Firm Innovations Xiaoying Li , Ying Tan PII: DOI: Reference:
S1544-6123(19)30816-5 https://doi.org/10.1016/j.frl.2019.101364 FRL 101364
To appear in:
Finance Research Letters
Received date: Revised date: Accepted date:
6 August 2019 10 November 2019 16 November 2019
Please cite this article as: Xiaoying Li , Ying Tan , University R&D Activities and Firm Innovations, Finance Research Letters (2019), doi: https://doi.org/10.1016/j.frl.2019.101364
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc.
Highlights A positive relation between total funding of universities and firm innovations in the same city. Funding from the government has larger spillovers than funding from other channels on firm innovation. High-tech firms and firms from the main board generally get larger spillovers from university R&D activities
University R&D Activities and Firm Innovations
1.Xiaoying Li (Corresponding author), Email: [email protected] Affiliation: Institute of Guangdong, Hong Kong and Macao Development Studies, Center for Studies of Hong Kong, Macao and Pearl River Delta, Sun Yat-sen University, Guangzhou, China, 510275. 2. Ying Tan [email protected] Affiliation: School of Economics and Trade, Guangdong University of Finance Acknowledge: This research is supported by National Social Science Foundation of China (NO. 19BJY011). Abstract This paper combines firm R&D activities with local university R&D activities to investigate the spillover effect of universities on firm innovations. First, we find a positive relationship between universities’ total funding and firm innovations in the same city, which is evidence of positive spillovers from university R&D activities. Funding from the government has larger spillovers than funding from other channels. Second, firms’ internal R&D expenditure does not affect the extent to which they obtain spillovers from local university R&D activities. High-tech firms and firms from the main board generally gain larger spillovers from university R&D activities. Keywords: Firm innovation, University spillovers, R&D funding
1 Introduction The main focus of this paper is on identifying the relationships among local university R&D activities, firm R&D expenditure, and firm innovations. Both firms’ internal attributes and regional factors affect firms’ innovation performance. Researchers find that corporate governance is an essential factor that affects firms’ innovation (Hirukawa and Ueda, 2011; Lee and Oh, 2018; Yang and Okada, 2019). In addition, firms’ R&D endowments and expenditures are of great importance to their innovation performance (Heimonen, 2012; He and Tian, 2013; Dang and Motohashi, 2015; Wang et al., 2016; Zhang et al., 2019). Firms interact with the surrounding economic environment to improve innovation (Brunow et al., 2017; Ferreira et al., 2017; Zhang et al., 2019). Academic institutions for education and research as well as firm innovation are critical elements of regional innovation systems (Berry and Glaeser, 2005; Edquist, 2010). Face-to-face contact between firms and organizations is important for knowledge exchanges, which in turn facilitates innovations (Storper and Venables, 2004;
Asheim et al., 2007; McCann, 2007). Universities are key players not only in creating knowledge through basic research but also in supplying well-educated workers who will participate in R&D (Audretsch and Feldman, 1996; Veugelers and Cassiman, 2005). Lin (2015) finds that firms locating near a research university innovate more. Policies play important roles in facilitating the transfer of scientific and technological achievements to firms which further promote firm innovation (Mowery et al., 2015; Gan and Xu, 2019; Tsai et al., 2019). This article combines firm R&D activities with local university R&D activities to find out how firm R&D investment interacts with regional R&D activities. It contributes to the literature in the following three dimensions. First, we measure university R&D spillovers by the actual R&D activities undertaken by universities. Second, we divide the total university R&D funds into sub-categories, which helps us to differentiate the effects of funds from different channels. Third, we investigate the interactions between firm R&D activities and university R&D activities on firm innovation.
2 Data and statistics Firm variables and city variables are matched to get an unbalanced panel with 1568 Chinese listed firms and 10829 observations by firm-year level. The firm variables come from CSMAR and RESSET from 2011 to 2015. CSMAR and RESSET are two data resources that cover stock market information, annual report information and accounting information of all the listed firms in mainland China. Here we collect the basic information of each firm, such as the number of employees, sales, profits, leverage (debt/asset ratio), R&D investment, average education level of employees, and ratio of employees with above college education. The city variables fall into two groups. The input and output of R&D activities in colleges and universities come from the “Compilation of Science and Technology Statistics of Colleges and Universities” from 2009 to 2017, which is publicly available from the website of the Science and Technology Department, Ministry of Education, in China.1 The information about the number of universities and colleges in a city is manually collected. Other city-level variables come from China’s City Statistical Yearbook. Patent applications are used as proxy for innovation outcomes in each firm. Detailed information on patents is available from the data sets of CSMAR. Patents can be divided into three categories by their content: invention patents, utility model patents and design patents. The shortcomings of using patent data as a measure of firm innovation performance are as follows. First, patent data reflect only technical innovation rather than all innovation, including management innovation and business model innovation. Second, there is an implicit assumption that the contributions of all kinds of patents are the same. However, from the viewpoint of the availability and comparability of data, patent data are still the most widely used indicator for firm innovation performance. The mean of patent applications for listed firms is 47 for all types of patents. For innovation patents, the average number of applications is 23 per year; for utility model patents, the average number of applications is 19 per year; and, for design patents, the mean number of applications is 5 per year. In the later regressions, these variables are transformed into the natural logarithm of the application number plus 1 to alleviate the right skewness. At the city level, the variable of interest is local university R&D activities. The total R&D 1
http://www.moe.gov.cn/s78/A16/A16_tjdc/
funds of universities as well as funds from the government, funds from firms and institutions, and funds from other channels are used in the later analysis. Generally speaking, university funds from the government are larger than university funds from firms, institutions, and other channels. Firm R&D expenditures are critical financial factors for firm innovations. In our data, we use the logarithm of R&D expenditure as an independent variable. We also control firm characteristics, such as log(R&D personnel), log(revenue), log(net profits), and dummies for different types of firms. Table 1 Summary statistics Obs.
Mean
Std
Min.
Max.
Patent application
10,830
47.1
258.6
0
7073
Patent application: innovation
10,830
23.5
181.6
0
5855
Patent application: utility model
10,830
18.5
93.04
0
3979
Patent application: design
10,830
5.2
24.64
0
682
lnR&D expenditure
10,830
15.7
1.4
0
23.7
Revenue
10,829 10,004 10,830 10,830 10,830 10,830
21.3 18.6 0.7 0.4 0.2 0.4
1.5 1.6 0.4 0.5 0.4 0.5
11.60 11.66 0 0 0 0
28.69 26.35 1 1 1 1
University R&D funding in city
10,830
3.610e+06
5.960e+06
0
2.310e+07
University R&D funding from gov. in city
10,830
2.410e+06
4.066e+06
0
1.540e+07
University R&D funding from firm in city
10,830
1.033e+06
1.729e+06
0
7.056e+06
University R&D funding from other in city
10,830
166560
224460
0
1.067e+06
Population density of city
9,096
843.2
539.2
5.100
2648
GDP per capita of city
10,829
83706
55055
4396
467749
Dependent variables:
Firm characteristics:
Net profits Dummy for high-tech firm Dummy for SME board Dummy for growth enterprise market Dummy for main board City characteristics:
3 Empirical Framework and Results 3.1 Model specifications Pakes and Griliches (1984) and Hall, Griliches, and Hausman (1986) are the first to study the relationship between firm expenditure and firm innovations. Following the model specification, we obtain model (1), that is: 𝑝𝑎𝑡𝑒𝑛𝑡𝑖,𝑐,𝑡+1 = 𝛼 + 𝛽1 ∗ 𝑓𝑟𝑑𝑖,𝑡 + 𝛽2 ∗ 𝑋𝑖𝑡 + 𝛽3 ∗ 𝑋𝑐,𝑡 + 𝜇𝑐 + 𝜃𝑡 + 𝜀𝑖,𝑐,𝑡 (1) where 𝑝𝑎𝑡𝑒𝑛𝑡𝑖,𝑐,𝑡+1 is the patent number of firm i in city c and year t+1. 𝑓𝑟𝑑𝑖,𝑡 is the firm R&D investment, and 𝑋𝑖,𝑡 and 𝑋𝑐,𝑡 represent firm and city covariates. The firm-level characteristics of firm i year t include employees’ education structure, annual sales, profits, leverage (ratio of debt to total assets), and R&D investment. 𝑋𝑐,𝑡 stands for the city-level characteristics other than university R&D activities in city c and year t, including the income per capita, which stands for the economic status of the city, and the total population of the city. 𝜇𝑐 represents the time-invariant city fixed effect, and 𝜃𝑡 is the year fixed effect. 𝜀𝑖𝑡 is the error term. All the
continuous variables are log transformed. The coefficient of interest is 𝛽1 . University spillovers could be defined by externalities towards firms, for which a university is the source of the spillover but is not fully compensated (Harris, 2001). We use the regional knowledge production function to relate the technological performance of firms to the activities of universities located in that region (Jaffe, 1989; Griliches, 1991). The following model specification is used to measure the spillover effects from university R&D activities: 𝑝𝑎𝑡𝑒𝑛𝑡𝑖,𝑐,𝑡+1 = 𝛼 + 𝛾1 ∗ 𝑢𝑟𝑑𝑐,𝑡 + 𝛽2 ∗ 𝑋𝑖𝑡 + 𝛽3 ∗ 𝑋𝑐,𝑡 + 𝜇𝑐 + 𝜃𝑡 + 𝜀𝑖,𝑐,𝑡 (2) where 𝑢𝑟𝑑𝑐,𝑡 is the university R&D activities in city c and year t. Different proxies are used for university R&D activities, including university R&D funding from the government, R&D funding from firms and institutions, and R&D funding from other channels. The coefficient of interest is 𝛾1. The covariates in model (2) follow those in model (1). To further identify interactions between firm R&D expenditure and university R&D funding, we extend model (1) and model (2) by adding the interaction term of 𝑓𝑟𝑑𝑖𝑡 ∗ 𝑢𝑟𝑑𝑐𝑡 to obtain model (3) as follows: 𝑝𝑎𝑡𝑒𝑛𝑡𝑖𝑐𝑡+1 = 𝛼 + 𝛽1 ∗ 𝑓𝑟𝑑𝑖𝑡 + 𝛾1 ∗ 𝑢𝑟𝑑𝑐𝑡 + 𝛿1 ∗ 𝑓𝑟𝑑𝑖𝑡 ∗ 𝑢𝑟𝑑𝑐𝑡 + 𝛽2 ∗ 𝑋𝑖𝑡 + 𝛽3 ∗ 𝑋𝑐𝑡 + 𝜇𝑐 + 𝜃𝑡 + 𝜀𝑖𝑐𝑡 (3) where the interaction term between external university R&D activities and internal financial indicators is 𝑓𝑟𝑑𝑖𝑡 ∗ 𝑢𝑟𝑑𝑐𝑡 and coefficient 𝛿1 shows the relationship between firm financial investment in R&D and university R&D activities. As patent numbers follow the non-negative binomial distribution, the log-log OLS estimation methodology is the primary strategy for estimation in this article. The results are as follows. 3.2 Estimation results (1) Firm R&D expenditure and firm innovation First, we consider the relationship between firm R&D activities and firm patent applications. Table 2 presents the estimation results of Equation (1). In columns (1), (3), (5), and (7), we run the total patent applications, innovation patent applications, utility model patent applications, and design patent applications on firm and city characteristics, with year and industry fixed effects. In columns (2), (4), (6), and (8), we extend the above regressions with city fixed effects. All the standard errors are clustered at the firm level. The estimation results show that firm patents are positively related to firm R&D expenditure except design patents. Meanwhile, we find that firms with more revenues on average have larger numbers of patents and firms’ net profits are positively related to their patents. Table 2 Firm R&D expenditure and firm patents (1) VARIABLES Log(firm R&D)
Log(revenue)
Log(netprofits)
Log(density)
(2)
Log(total patents)
(3)
(4)
(5)
(6)
Log(innovation
Log(utility model
patents)
patents)
(7)
(8)
Log(design patents)
0.008***
0.008***
0.012***
0.012***
0.009***
0.009***
0.002
0.002
(2.88)
(2.88)
(4.05)
(4.05)
(3.02)
(3.02)
(0.87)
(0.87)
0.205***
0.205***
0.230***
0.230***
0.136***
0.136***
0.040
0.040
(3.38)
(3.38)
(4.01)
(4.01)
(2.76)
(2.76)
(0.95)
(0.95)
0.064***
0.064***
0.054***
0.054***
0.071***
0.071***
0.028**
0.028**
(3.41)
(3.41)
(3.03)
(3.03)
(3.90)
(3.90)
(2.18)
(2.18)
-0.174
-0.174
-0.302
-0.302
-0.162
-0.162
-0.028
-0.028
(-0.53)
(-0.53)
(-0.96)
(-0.96)
(-0.50)
(-0.50)
(-0.12)
(-0.12)
Log(per GDP)
0.249***
0.249***
0.350***
0.350***
0.151***
0.151***
0.043
0.043
(4.87)
(4.87)
(6.83)
(6.83)
(3.24)
(3.24)
(1.08)
(1.08)
-5.465***
-5.465***
-5.798***
-5.798***
-3.120
-3.120
-1.108
-1.108
(-2.66)
(-2.66)
(-3.03)
(-3.03)
(-1.46)
(-1.46)
(-0.73)
(-0.73)
Year FE
yes
yes
yes
yes
yes
yes
yes
yes
Industry FE
yes
yes
yes
yes
yes
yes
yes
yes
City FE
no
yes
no
yes
no
yes
no
yes
R
0.076
0.076
0.103
0.103
0.060
0.060
0.012
0.012
Observations
7456
7456
7456
7456
7456
7456
7456
7456
Constant
2
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. (2) University R&D activities and firm innovation This section investigates how local university spillovers affect firm innovations by estimating specification (2) as presented in table 3. Here we use university funding as a proxy for university R&D activities. Both the total funding and the funding from different resources in a city are used. In column (1), we explore the effects of the total R&D funding of universities in a city on firm innovations, and the coefficients show that there are significantly positive relations between the total funding of universities in the city and the firm innovations in the city. In columns (2), (3), and (4), we investigate the effects of funding from the government, firms or institutions, and others, and the results show general positive spillovers of university R&D activities. Funding from the government has particularly larger and significantly positive spillovers than funding from firms, institutions, or other channels. R&D funds from the government are particularly important for carrying out basic research, which usually has more externalities, which explains a larger positive effect on firm innovation from the government. Table 3 University R&D funding and firm innovation (1)
(2)
VARIABLES Log(university R&D funding in city)
(3)
(4)
(5)
Log(total patents)
0.085*** (4.49)
Log(university R&D funding from gov. in city)
0.087***
0.054**
(4.54)
(2.16)
Log(university R&D funding from firm in city)
0.049***
0.031**
(4.21)
(2.35)
Log(university R&D funding from other in city)
Log(revenue)
0.053***
0.014
(3.51)
(0.76)
0.203***
0.199***
0.212***
0.210***
0.198***
(6.84)
(6.69)
(7.22)
(7.10)
(6.65)
Log(net profits)
0.067***
0.067***
0.065***
0.066***
0.067***
(4.62)
(4.66)
(4.52)
(4.54)
(4.66)
Log(density)
-0.325
-0.327
-0.238
-0.261
-0.340
(-1.53)
(-1.53)
(-1.13)
(-1.23)
(-1.60)
0.254***
0.245***
0.271***
0.285***
0.249***
(6.32)
(6.05)
(6.85)
(7.24)
(6.09)
-5.540***
-5.348***
-5.993***
-5.913***
-5.368***
Log(per GDP)
Constant
(-3.50)
(-3.36)
(-3.80)
(-3.74)
(-3.38)
Year FE
yes
yes
yes
yes
yes
Firm FE
yes
yes
yes
yes
yes
City FE
yes
yes
yes
yes
yes
R2
0.077
0.077
0.077
0.076
0.078
Observations
7456
7456
7456
7456
7456
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. (3) Interaction between firm R&D expenditure and university R&D activities This section investigates the interaction between firm R&D expenditure and university R&D activities by estimating specification (3). In column (1) of table 4, the coefficients for both firm R&D expenditure and university R&D funding are positive, but the coefficient for the interaction term of firm R&D expenditure and university R&D funding is insignificant, which means that firms with higher R&D expenditures are not more likely to gain larger spillovers from local university R&D activities. In columns (2), (3), and (4) of table 4, R&D funding from the government, firms or institutions, and other channels are used to replace the university’s total R&D funding. The coefficients of firm R&D expenditure and university R&D funding from different resources are positive, but the interactions are insignificant, which means that firms with higher expenditures are not more likely to obtain more spillovers from different university R&D activities. Table 4 Interactions between firm R&D expenditures and university R&D funding (1)
(2)
VARIABLES Log(firm R&D)
(3)
(4)
(5)
Log(total patents) 0.021**
0.019**
0.016***
0.020**
0.015
(2.22)
(2.04)
(2.70)
(2.32)
(1.41)
Log(university R&D funding)
0.096***
Log(firm R&D)*
-0.001
(4.28)
Log(university R&D funding) (-1.51) Log(university R&D funding from gov.)
0.093***
0.012
(4.21)
(0.29)
-0.001
0.002
(-1.30)
(1.03)
Log(firm R&D)* Log(university R&D funding from gov.)
Log(university R&D funding from firm)
0.057***
0.043**
(4.18)
(2.06)
-0.001
-0.001
(-1.55)
(-0.70)
Log(firm R&D)* Log(university R&D funding from firm)
Log(university R&D funding from other)
0.067***
0.055
(3.41)
(1.38)
-0.001
-0.003
Log(firm R&D)* Log(university R&D funding from other)
(-1.51)
(-1.21)
Control variables
yes
yes
yes
yes
yes
Year FE
yes
yes
yes
yes
yes
Firm FE
yes
yes
yes
yes
yes
City FE
yes
yes
yes
yes
yes
R
0.079
0.079
0.079
0.078
0.081
Observations
7456
7456
7456
7456
7456
2
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. (4) R&D spillovers by different types of firms The relations between university R&D activities, firm R&D expenditures, and firm innovations might differ among different types of firms. In this section, we divide the firms into two perspectives. First, we divide firms according to whether they are recognized as national high-technology firms by the Ministry of Science and Technology of China. High-tech firms operate in the field of high technology, and they continue to carry out research and development and the transformation of technological achievements, hold core intellectual property rights, and implement production activities on this basis. In columns (1) and (2) of table 5, we find that high-tech firms generally gain larger spillovers from university R&D activities compared with non-high-tech firms. Second, we divide the firms according to the market through which they are listed, including main board market, growth enterprise market (GEM) and small and medium-sized enterprise board (SEM). The first is the main board market, which has higher requirements for the firms in terms of the size of equity, profitability level, minimum market value, and so on. Most listed companies on the main board market are large mature enterprises with a large capital scale and stable profitability. The second is the growth enterprise market, and most of the companies listed here are engaged in high-tech business and have high growth, but established in a relatively short time and small in scale. The third is the small and medium-sized enterprise board, which is a market for firms with less than 100 million yuan. In columns (3), (4), and (5) of table 5, we can see that the firms from the main board generally have higher spillovers from university R&D activities in the same city. Table 5 Interactions between firm R&D expenditures and university R&D funding by subgroups (1)
(2)
(3)
(4)
(5)
Log(total patents) By high-tech firms
By listed markets
VARIABLES Yes
No
Main board
SEM board
GEM board
0.006
0.049*
0.021**
-0.014
-0.290
(0.35)
(1.80)
(1.97)
(-0.59)
(-0.98)
0.089**
0.133**
0.102***
0.032
-0.064
(2.42)
(2.42)
(3.63)
(0.74)
(-0.21)
Log(firm R&D)*
0.000
-0.004*
-0.001
0.001
0.023
Log(university R&D funding)
(0.21)
(-1.76)
(-1.57)
(0.36)
(1.22)
Control variables
yes
yes
yes
yes
yes
Year FE
yes
yes
yes
yes
yes
Log(firm R&D)
Log(university R&D funding)
Firm FE
yes
yes
yes
yes
yes
City FE
yes
yes
yes
yes
yes
R2
0.071
0.120
0.133
0.073
0.078
Observations
5658
1798
3158
2882
1416
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
4 Conclusions This paper analyses the impact of local university R&D activities on innovation in Chinese listed firms in the same city. We combine firm R&D activities with local university R&D activities to investigate how firm R&D investment interacts with regional R&D activities to affect firm innovations. First, we investigate the relationship between firm R&D activities and firm patent applications using different measurements of patents. Consistent with the existing literature, we find that the total patent applications, innovation patent applications, and utility model patent applications are positively related to firm R&D expenditures. Firms with more revenues on average have larger amounts of patents, and firms’ net profits are positively related to their patents. Second, we investigate how local university spillovers affect firm innovations in the same city. Both university total funding and funding from different resources are used to estimate university spillovers on firm innovation. Significantly positive relations exist between the total funding of universities in the city and the firm innovations in the city, which is evidence of positive spillovers of university R&D activities. Meanwhile, we find that funding from the government has particularly larger spillovers compared to the effects of funding from firms, institutions, or other channels. Moreover, we do not find that firms with higher R&D expenditures obtain more substantial spillovers from local university R&D activities. We further explore the relations by different types of firms. We divide the firms according to whether they are recognized as high-tech firms, and we find that high-tech firms generally gain more spillovers from university R&D activities than non-high-tech firms. Moreover, we divide the firms according to the market through which they are listed, and we see that firms from the main board generally have higher spillovers from university R&D activities. Some issues are left for future work due to lack of proper datasets. First, future research should study the mechanism through which university spillovers work, i.e. the cooperation between firms and universities on innovation. Second, it is worthwhile to investigate how academic fields of universities affect firm from different industries in different ways. References Asheim, B., Coenen, L., Vang, J., 2007. Face-to-face, buzz, and knowledge bases: sociospatial implications for learning, innovation, and innovation policy. Environ. Plan. C: Gov. Policy 25(5), 655–670. Audretsch, D.B., Feldman, M.P., 1996. R&D spillovers and the geography of innovation and production. Am. Econ. Rev. 86(3), 630–640. Berry, C.R., Glaeser, E.L., 2005. The divergence of human capital levels across cities. Pap. in Reg. Sci. 84(3), 407–444.
Brunow, S., Hammer-Langrock, A., McCann, P., 2017. Innovation and Location in German Knowledge Intensive Business Service Firms (No. 22/2017). IAB Discussion Paper. Dang, J., & Motohashi, K., 2015. Patent statistics: A good indicator for innovation in China? Patent subsidy program impacts on patent quality. China Economic Review, 35, 137-155 Edquist, C., 2010. Systems of innovation perspectives and challenges. African Journal of Science, Technology, Innovation and Development, 2(3), 14-45. Ferreira, J.J., Fernandes, C.I., Raposo, M.L. (2017). The effects of location on firm innovation capacity. J. Knowl. Econ. 8(1), 77–96. Gan, W., & Xu, X., 2019. Does anti-corruption campaign promote corporate R&D investment? Evidence from China. Finance Research Letters, 30, 292-296 Griliches, Z., 1991. The search for R&D spillovers, No. w3768. National Bureau of Economic Research, https://www.nber.org/papers/w3768.pdf. Hall, B. H., Griliches, Z., & Hausman, J. A., 1986. Is there a second, technological opportunity) factor. International Economic Review, 27, 265-283 Harris, R.G., 2001. The knowledge-based economy: intellectual origins and new economic perspectives. Int. J. Manag. Rev. 3(1), 21–40. He, J. J., & Tian, X., 2013. The dark side of analyst coverage: The case of innovation. Journal of Financial Economics, 109(3), 856-878. Heimonen, T., 2012. What are the factors that affect innovation in growing SMEs?. European Journal of Innovation Management, 15(1), 122-144 Hirukawa, M., & Ueda, M., 2011. Venture capital and innovation: which is first?. Pacific Economic Review, 16(4), 421-465. Jaffe, A.B., 1989. Real effects of academic research. Am. Econ. Rev. 957–970. Lee, H. H., & Oh, F. D., 2018. Corporate innovation and credit default swap spreads. Finance Research Letters, https://doi.org/10.1016/j.frl.2018.12.030 Lin, S., 2015. Are ivory towers truly ivory? Knowledge spillovers and firm innovation. J. Econ. Bus. 80, 21–36. McCann, P., 2007. Sketching out a model of innovation, face-to-face interaction and economic geography. Spat. Econ. Anal. 2(2), 117–134. Mowery, D.C., Nelson, R.R., Sampat, B.N., Ziedonis, A.A., 2015. Ivory Tower and Industrial Innovation: University–Industry Technology Transfer before and after the Bayh-Dole Act. California: Stanford University Press Pakes, A., Griliches, Z., 1984. Estimating distributed lags in short panels with an application to the specification of depreciation patterns and capital stock constructs. Rev. Econ. Stud. 51(2), 243–262. Storper, M., Venables, A.J., 2004. Buzz: face-to-face contact and the urban economy. J. Econ. Geogr. 4(4), 351–370. Tsai, L. C., Zhang, R., & Zhao, C., 2019. Political connections, network centrality and firm innovation. Finance Research Letters, 28, 180-184 Veugelers, R., Cassiman, B., 2005. R&D cooperation between firms and universities. Some empirical evidence from Belgian manufacturing. Int. J. Ind. Organ. 23(5–6), 355–379. Wang, Z., Zhang, M., Sun, H., & Zhu, G., 2016. Effects of standardization and innovation on mass customization: An empirical investigation. Technovation, 48, 79-86. Yang, A. S., & Okada, H., 2019. Corporate innovations as institutional anomie: Patent activities
and financial performance of the international aerospace industry. Finance Research Letters, 28, 328-336. Zhang, D., Guo, Y., Wang, Z., & Chen, Y., 2019. The impact of US monetary policy on Chinese enterprises’ R&D investment. Finance Research Letters. https://doi.org/10.1016/j.frl.2019.09.016 Zhang, D., Tong, Z., & Li, Y. (2019). The role of cash holding towards cleaner production in China’s manufacturing sectors: A financial constraint perspective. Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.118875.
University R&D Activities and Firm Innovations Xiaoying Li , Ying Tan PII: DOI: Reference:
S1544-6123(19)30816-5 https://doi.org/10.1016/j.frl.2019.101364 FRL 101364
To appear in:
Finance Research Letters
Received date: Revised date: Accepted date:
6 August 2019 10 November 2019 16 November 2019
Please cite this article as: Xiaoying Li , Ying Tan , University R&D Activities and Firm Innovations, Finance Research Letters (2019), doi: https://doi.org/10.1016/j.frl.2019.101364
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc.
Highlights A positive relation between total funding of universities and firm innovations in the same city. Funding from the government has larger spillovers than funding from other channels on firm innovation. High-tech firms and firms from the main board generally get larger spillovers from university R&D activities
University R&D Activities and Firm Innovations
1.Xiaoying Li (Corresponding author), Email: [email protected] Affiliation: Institute of Guangdong, Hong Kong and Macao Development Studies, Center for Studies of Hong Kong, Macao and Pearl River Delta, Sun Yat-sen University, Guangzhou, China, 510275. 2. Ying Tan [email protected] Affiliation: School of Economics and Trade, Guangdong University of Finance Acknowledge: This research is supported by National Social Science Foundation of China (NO. 19BJY011). Abstract This paper combines firm R&D activities with local university R&D activities to investigate the spillover effect of universities on firm innovations. First, we find a positive relationship between universities’ total funding and firm innovations in the same city, which is evidence of positive spillovers from university R&D activities. Funding from the government has larger spillovers than funding from other channels. Second, firms’ internal R&D expenditure does not affect the extent to which they obtain spillovers from local university R&D activities. High-tech firms and firms from the main board generally gain larger spillovers from university R&D activities. Keywords: Firm innovation, University spillovers, R&D funding
1 Introduction The main focus of this paper is on identifying the relationships among local university R&D activities, firm R&D expenditure, and firm innovations. Both firms’ internal attributes and regional factors affect firms’ innovation performance. Researchers find that corporate governance is an essential factor that affects firms’ innovation (Hirukawa and Ueda, 2011; Lee and Oh, 2018; Yang and Okada, 2019). In addition, firms’ R&D endowments and expenditures are of great importance to their innovation performance (Heimonen, 2012; He and Tian, 2013; Dang and Motohashi, 2015; Wang et al., 2016; Zhang et al., 2019). Firms interact with the surrounding economic environment to improve innovation (Brunow et al., 2017; Ferreira et al., 2017; Zhang et al., 2019). Academic institutions for education and research as well as firm innovation are critical elements of regional innovation systems (Berry and Glaeser, 2005; Edquist, 2010). Face-to-face contact between firms and organizations is important for knowledge exchanges, which in turn facilitates innovations (Storper and Venables, 2004;
Asheim et al., 2007; McCann, 2007). Universities are key players not only in creating knowledge through basic research but also in supplying well-educated workers who will participate in R&D (Audretsch and Feldman, 1996; Veugelers and Cassiman, 2005). Lin (2015) finds that firms locating near a research university innovate more. Policies play important roles in facilitating the transfer of scientific and technological achievements to firms which further promote firm innovation (Mowery et al., 2015; Gan and Xu, 2019; Tsai et al., 2019). This article combines firm R&D activities with local university R&D activities to find out how firm R&D investment interacts with regional R&D activities. It contributes to the literature in the following three dimensions. First, we measure university R&D spillovers by the actual R&D activities undertaken by universities. Second, we divide the total university R&D funds into sub-categories, which helps us to differentiate the effects of funds from different channels. Third, we investigate the interactions between firm R&D activities and university R&D activities on firm innovation.
2 Data and statistics Firm variables and city variables are matched to get an unbalanced panel with 1568 Chinese listed firms and 10829 observations by firm-year level. The firm variables come from CSMAR and RESSET from 2011 to 2015. CSMAR and RESSET are two data resources that cover stock market information, annual report information and accounting information of all the listed firms in mainland China. Here we collect the basic information of each firm, such as the number of employees, sales, profits, leverage (debt/asset ratio), R&D investment, average education level of employees, and ratio of employees with above college education. The city variables fall into two groups. The input and output of R&D activities in colleges and universities come from the “Compilation of Science and Technology Statistics of Colleges and Universities” from 2009 to 2017, which is publicly available from the website of the Science and Technology Department, Ministry of Education, in China.1 The information about the number of universities and colleges in a city is manually collected. Other city-level variables come from China’s City Statistical Yearbook. Patent applications are used as proxy for innovation outcomes in each firm. Detailed information on patents is available from the data sets of CSMAR. Patents can be divided into three categories by their content: invention patents, utility model patents and design patents. The shortcomings of using patent data as a measure of firm innovation performance are as follows. First, patent data reflect only technical innovation rather than all innovation, including management innovation and business model innovation. Second, there is an implicit assumption that the contributions of all kinds of patents are the same. However, from the viewpoint of the availability and comparability of data, patent data are still the most widely used indicator for firm innovation performance. The mean of patent applications for listed firms is 47 for all types of patents. For innovation patents, the average number of applications is 23 per year; for utility model patents, the average number of applications is 19 per year; and, for design patents, the mean number of applications is 5 per year. In the later regressions, these variables are transformed into the natural logarithm of the application number plus 1 to alleviate the right skewness. At the city level, the variable of interest is local university R&D activities. The total R&D 1
http://www.moe.gov.cn/s78/A16/A16_tjdc/
funds of universities as well as funds from the government, funds from firms and institutions, and funds from other channels are used in the later analysis. Generally speaking, university funds from the government are larger than university funds from firms, institutions, and other channels. Firm R&D expenditures are critical financial factors for firm innovations. In our data, we use the logarithm of R&D expenditure as an independent variable. We also control firm characteristics, such as log(R&D personnel), log(revenue), log(net profits), and dummies for different types of firms. Table 1 Summary statistics Obs.
Mean
Std
Min.
Max.
Patent application
10,830
47.1
258.6
0
7073
Patent application: innovation
10,830
23.5
181.6
0
5855
Patent application: utility model
10,830
18.5
93.04
0
3979
Patent application: design
10,830
5.2
24.64
0
682
lnR&D expenditure
10,830
15.7
1.4
0
23.7
Revenue
10,829 10,004 10,830 10,830 10,830 10,830
21.3 18.6 0.7 0.4 0.2 0.4
1.5 1.6 0.4 0.5 0.4 0.5
11.60 11.66 0 0 0 0
28.69 26.35 1 1 1 1
University R&D funding in city
10,830
3.610e+06
5.960e+06
0
2.310e+07
University R&D funding from gov. in city
10,830
2.410e+06
4.066e+06
0
1.540e+07
University R&D funding from firm in city
10,830
1.033e+06
1.729e+06
0
7.056e+06
University R&D funding from other in city
10,830
166560
224460
0
1.067e+06
Population density of city
9,096
843.2
539.2
5.100
2648
GDP per capita of city
10,829
83706
55055
4396
467749
Dependent variables:
Firm characteristics:
Net profits Dummy for high-tech firm Dummy for SME board Dummy for growth enterprise market Dummy for main board City characteristics:
3 Empirical Framework and Results 3.1 Model specifications Pakes and Griliches (1984) and Hall, Griliches, and Hausman (1986) are the first to study the relationship between firm expenditure and firm innovations. Following the model specification, we obtain model (1), that is: 𝑝𝑎𝑡𝑒𝑛𝑡𝑖,𝑐,𝑡+1 = 𝛼 + 𝛽1 ∗ 𝑓𝑟𝑑𝑖,𝑡 + 𝛽2 ∗ 𝑋𝑖𝑡 + 𝛽3 ∗ 𝑋𝑐,𝑡 + 𝜇𝑐 + 𝜃𝑡 + 𝜀𝑖,𝑐,𝑡 (1) where 𝑝𝑎𝑡𝑒𝑛𝑡𝑖,𝑐,𝑡+1 is the patent number of firm i in city c and year t+1. 𝑓𝑟𝑑𝑖,𝑡 is the firm R&D investment, and 𝑋𝑖,𝑡 and 𝑋𝑐,𝑡 represent firm and city covariates. The firm-level characteristics of firm i year t include employees’ education structure, annual sales, profits, leverage (ratio of debt to total assets), and R&D investment. 𝑋𝑐,𝑡 stands for the city-level characteristics other than university R&D activities in city c and year t, including the income per capita, which stands for the economic status of the city, and the total population of the city. 𝜇𝑐 represents the time-invariant city fixed effect, and 𝜃𝑡 is the year fixed effect. 𝜀𝑖𝑡 is the error term. All the
continuous variables are log transformed. The coefficient of interest is 𝛽1 . University spillovers could be defined by externalities towards firms, for which a university is the source of the spillover but is not fully compensated (Harris, 2001). We use the regional knowledge production function to relate the technological performance of firms to the activities of universities located in that region (Jaffe, 1989; Griliches, 1991). The following model specification is used to measure the spillover effects from university R&D activities: 𝑝𝑎𝑡𝑒𝑛𝑡𝑖,𝑐,𝑡+1 = 𝛼 + 𝛾1 ∗ 𝑢𝑟𝑑𝑐,𝑡 + 𝛽2 ∗ 𝑋𝑖𝑡 + 𝛽3 ∗ 𝑋𝑐,𝑡 + 𝜇𝑐 + 𝜃𝑡 + 𝜀𝑖,𝑐,𝑡 (2) where 𝑢𝑟𝑑𝑐,𝑡 is the university R&D activities in city c and year t. Different proxies are used for university R&D activities, including university R&D funding from the government, R&D funding from firms and institutions, and R&D funding from other channels. The coefficient of interest is 𝛾1. The covariates in model (2) follow those in model (1). To further identify interactions between firm R&D expenditure and university R&D funding, we extend model (1) and model (2) by adding the interaction term of 𝑓𝑟𝑑𝑖𝑡 ∗ 𝑢𝑟𝑑𝑐𝑡 to obtain model (3) as follows: 𝑝𝑎𝑡𝑒𝑛𝑡𝑖𝑐𝑡+1 = 𝛼 + 𝛽1 ∗ 𝑓𝑟𝑑𝑖𝑡 + 𝛾1 ∗ 𝑢𝑟𝑑𝑐𝑡 + 𝛿1 ∗ 𝑓𝑟𝑑𝑖𝑡 ∗ 𝑢𝑟𝑑𝑐𝑡 + 𝛽2 ∗ 𝑋𝑖𝑡 + 𝛽3 ∗ 𝑋𝑐𝑡 + 𝜇𝑐 + 𝜃𝑡 + 𝜀𝑖𝑐𝑡 (3) where the interaction term between external university R&D activities and internal financial indicators is 𝑓𝑟𝑑𝑖𝑡 ∗ 𝑢𝑟𝑑𝑐𝑡 and coefficient 𝛿1 shows the relationship between firm financial investment in R&D and university R&D activities. As patent numbers follow the non-negative binomial distribution, the log-log OLS estimation methodology is the primary strategy for estimation in this article. The results are as follows. 3.2 Estimation results (1) Firm R&D expenditure and firm innovation First, we consider the relationship between firm R&D activities and firm patent applications. Table 2 presents the estimation results of Equation (1). In columns (1), (3), (5), and (7), we run the total patent applications, innovation patent applications, utility model patent applications, and design patent applications on firm and city characteristics, with year and industry fixed effects. In columns (2), (4), (6), and (8), we extend the above regressions with city fixed effects. All the standard errors are clustered at the firm level. The estimation results show that firm patents are positively related to firm R&D expenditure except design patents. Meanwhile, we find that firms with more revenues on average have larger numbers of patents and firms’ net profits are positively related to their patents. Table 2 Firm R&D expenditure and firm patents (1) VARIABLES Log(firm R&D)
Log(revenue)
Log(netprofits)
Log(density)
(2)
Log(total patents)
(3)
(4)
(5)
(6)
Log(innovation
Log(utility model
patents)
patents)
(7)
(8)
Log(design patents)
0.008***
0.008***
0.012***
0.012***
0.009***
0.009***
0.002
0.002
(2.88)
(2.88)
(4.05)
(4.05)
(3.02)
(3.02)
(0.87)
(0.87)
0.205***
0.205***
0.230***
0.230***
0.136***
0.136***
0.040
0.040
(3.38)
(3.38)
(4.01)
(4.01)
(2.76)
(2.76)
(0.95)
(0.95)
0.064***
0.064***
0.054***
0.054***
0.071***
0.071***
0.028**
0.028**
(3.41)
(3.41)
(3.03)
(3.03)
(3.90)
(3.90)
(2.18)
(2.18)
-0.174
-0.174
-0.302
-0.302
-0.162
-0.162
-0.028
-0.028
(-0.53)
(-0.53)
(-0.96)
(-0.96)
(-0.50)
(-0.50)
(-0.12)
(-0.12)
Log(per GDP)
0.249***
0.249***
0.350***
0.350***
0.151***
0.151***
0.043
0.043
(4.87)
(4.87)
(6.83)
(6.83)
(3.24)
(3.24)
(1.08)
(1.08)
-5.465***
-5.465***
-5.798***
-5.798***
-3.120
-3.120
-1.108
-1.108
(-2.66)
(-2.66)
(-3.03)
(-3.03)
(-1.46)
(-1.46)
(-0.73)
(-0.73)
Year FE
yes
yes
yes
yes
yes
yes
yes
yes
Industry FE
yes
yes
yes
yes
yes
yes
yes
yes
City FE
no
yes
no
yes
no
yes
no
yes
R
0.076
0.076
0.103
0.103
0.060
0.060
0.012
0.012
Observations
7456
7456
7456
7456
7456
7456
7456
7456
Constant
2
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. (2) University R&D activities and firm innovation This section investigates how local university spillovers affect firm innovations by estimating specification (2) as presented in table 3. Here we use university funding as a proxy for university R&D activities. Both the total funding and the funding from different resources in a city are used. In column (1), we explore the effects of the total R&D funding of universities in a city on firm innovations, and the coefficients show that there are significantly positive relations between the total funding of universities in the city and the firm innovations in the city. In columns (2), (3), and (4), we investigate the effects of funding from the government, firms or institutions, and others, and the results show general positive spillovers of university R&D activities. Funding from the government has particularly larger and significantly positive spillovers than funding from firms, institutions, or other channels. R&D funds from the government are particularly important for carrying out basic research, which usually has more externalities, which explains a larger positive effect on firm innovation from the government. Table 3 University R&D funding and firm innovation (1)
(2)
VARIABLES Log(university R&D funding in city)
(3)
(4)
(5)
Log(total patents)
0.085*** (4.49)
Log(university R&D funding from gov. in city)
0.087***
0.054**
(4.54)
(2.16)
Log(university R&D funding from firm in city)
0.049***
0.031**
(4.21)
(2.35)
Log(university R&D funding from other in city)
Log(revenue)
0.053***
0.014
(3.51)
(0.76)
0.203***
0.199***
0.212***
0.210***
0.198***
(6.84)
(6.69)
(7.22)
(7.10)
(6.65)
Log(net profits)
0.067***
0.067***
0.065***
0.066***
0.067***
(4.62)
(4.66)
(4.52)
(4.54)
(4.66)
Log(density)
-0.325
-0.327
-0.238
-0.261
-0.340
(-1.53)
(-1.53)
(-1.13)
(-1.23)
(-1.60)
0.254***
0.245***
0.271***
0.285***
0.249***
(6.32)
(6.05)
(6.85)
(7.24)
(6.09)
-5.540***
-5.348***
-5.993***
-5.913***
-5.368***
Log(per GDP)
Constant
(-3.50)
(-3.36)
(-3.80)
(-3.74)
(-3.38)
Year FE
yes
yes
yes
yes
yes
Firm FE
yes
yes
yes
yes
yes
City FE
yes
yes
yes
yes
yes
R2
0.077
0.077
0.077
0.076
0.078
Observations
7456
7456
7456
7456
7456
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. (3) Interaction between firm R&D expenditure and university R&D activities This section investigates the interaction between firm R&D expenditure and university R&D activities by estimating specification (3). In column (1) of table 4, the coefficients for both firm R&D expenditure and university R&D funding are positive, but the coefficient for the interaction term of firm R&D expenditure and university R&D funding is insignificant, which means that firms with higher R&D expenditures are not more likely to gain larger spillovers from local university R&D activities. In columns (2), (3), and (4) of table 4, R&D funding from the government, firms or institutions, and other channels are used to replace the university’s total R&D funding. The coefficients of firm R&D expenditure and university R&D funding from different resources are positive, but the interactions are insignificant, which means that firms with higher expenditures are not more likely to obtain more spillovers from different university R&D activities. Table 4 Interactions between firm R&D expenditures and university R&D funding (1)
(2)
VARIABLES Log(firm R&D)
(3)
(4)
(5)
Log(total patents) 0.021**
0.019**
0.016***
0.020**
0.015
(2.22)
(2.04)
(2.70)
(2.32)
(1.41)
Log(university R&D funding)
0.096***
Log(firm R&D)*
-0.001
(4.28)
Log(university R&D funding) (-1.51) Log(university R&D funding from gov.)
0.093***
0.012
(4.21)
(0.29)
-0.001
0.002
(-1.30)
(1.03)
Log(firm R&D)* Log(university R&D funding from gov.)
Log(university R&D funding from firm)
0.057***
0.043**
(4.18)
(2.06)
-0.001
-0.001
(-1.55)
(-0.70)
Log(firm R&D)* Log(university R&D funding from firm)
Log(university R&D funding from other)
0.067***
0.055
(3.41)
(1.38)
-0.001
-0.003
Log(firm R&D)* Log(university R&D funding from other)
(-1.51)
(-1.21)
Control variables
yes
yes
yes
yes
yes
Year FE
yes
yes
yes
yes
yes
Firm FE
yes
yes
yes
yes
yes
City FE
yes
yes
yes
yes
yes
R
0.079
0.079
0.079
0.078
0.081
Observations
7456
7456
7456
7456
7456
2
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. (4) R&D spillovers by different types of firms The relations between university R&D activities, firm R&D expenditures, and firm innovations might differ among different types of firms. In this section, we divide the firms into two perspectives. First, we divide firms according to whether they are recognized as national high-technology firms by the Ministry of Science and Technology of China. High-tech firms operate in the field of high technology, and they continue to carry out research and development and the transformation of technological achievements, hold core intellectual property rights, and implement production activities on this basis. In columns (1) and (2) of table 5, we find that high-tech firms generally gain larger spillovers from university R&D activities compared with non-high-tech firms. Second, we divide the firms according to the market through which they are listed, including main board market, growth enterprise market (GEM) and small and medium-sized enterprise board (SEM). The first is the main board market, which has higher requirements for the firms in terms of the size of equity, profitability level, minimum market value, and so on. Most listed companies on the main board market are large mature enterprises with a large capital scale and stable profitability. The second is the growth enterprise market, and most of the companies listed here are engaged in high-tech business and have high growth, but established in a relatively short time and small in scale. The third is the small and medium-sized enterprise board, which is a market for firms with less than 100 million yuan. In columns (3), (4), and (5) of table 5, we can see that the firms from the main board generally have higher spillovers from university R&D activities in the same city. Table 5 Interactions between firm R&D expenditures and university R&D funding by subgroups (1)
(2)
(3)
(4)
(5)
Log(total patents) By high-tech firms
By listed markets
VARIABLES Yes
No
Main board
SEM board
GEM board
0.006
0.049*
0.021**
-0.014
-0.290
(0.35)
(1.80)
(1.97)
(-0.59)
(-0.98)
0.089**
0.133**
0.102***
0.032
-0.064
(2.42)
(2.42)
(3.63)
(0.74)
(-0.21)
Log(firm R&D)*
0.000
-0.004*
-0.001
0.001
0.023
Log(university R&D funding)
(0.21)
(-1.76)
(-1.57)
(0.36)
(1.22)
Control variables
yes
yes
yes
yes
yes
Year FE
yes
yes
yes
yes
yes
Log(firm R&D)
Log(university R&D funding)
Firm FE
yes
yes
yes
yes
yes
City FE
yes
yes
yes
yes
yes
R2
0.071
0.120
0.133
0.073
0.078
Observations
5658
1798
3158
2882
1416
Notes: The t-values are reported in the parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
4 Conclusions This paper analyses the impact of local university R&D activities on innovation in Chinese listed firms in the same city. We combine firm R&D activities with local university R&D activities to investigate how firm R&D investment interacts with regional R&D activities to affect firm innovations. First, we investigate the relationship between firm R&D activities and firm patent applications using different measurements of patents. Consistent with the existing literature, we find that the total patent applications, innovation patent applications, and utility model patent applications are positively related to firm R&D expenditures. Firms with more revenues on average have larger amounts of patents, and firms’ net profits are positively related to their patents. Second, we investigate how local university spillovers affect firm innovations in the same city. Both university total funding and funding from different resources are used to estimate university spillovers on firm innovation. Significantly positive relations exist between the total funding of universities in the city and the firm innovations in the city, which is evidence of positive spillovers of university R&D activities. Meanwhile, we find that funding from the government has particularly larger spillovers compared to the effects of funding from firms, institutions, or other channels. Moreover, we do not find that firms with higher R&D expenditures obtain more substantial spillovers from local university R&D activities. We further explore the relations by different types of firms. We divide the firms according to whether they are recognized as high-tech firms, and we find that high-tech firms generally gain more spillovers from university R&D activities than non-high-tech firms. Moreover, we divide the firms according to the market through which they are listed, and we see that firms from the main board generally have higher spillovers from university R&D activities. Some issues are left for future work due to lack of proper datasets. First, future research should study the mechanism through which university spillovers work, i.e. the cooperation between firms and universities on innovation. Second, it is worthwhile to investigate how academic fields of universities affect firm from different industries in different ways. References Asheim, B., Coenen, L., Vang, J., 2007. Face-to-face, buzz, and knowledge bases: sociospatial implications for learning, innovation, and innovation policy. Environ. Plan. C: Gov. Policy 25(5), 655–670. Audretsch, D.B., Feldman, M.P., 1996. R&D spillovers and the geography of innovation and production. Am. Econ. Rev. 86(3), 630–640. Berry, C.R., Glaeser, E.L., 2005. The divergence of human capital levels across cities. Pap. in Reg. Sci. 84(3), 407–444.
Brunow, S., Hammer-Langrock, A., McCann, P., 2017. Innovation and Location in German Knowledge Intensive Business Service Firms (No. 22/2017). IAB Discussion Paper. Dang, J., & Motohashi, K., 2015. Patent statistics: A good indicator for innovation in China? Patent subsidy program impacts on patent quality. China Economic Review, 35, 137-155 Edquist, C., 2010. Systems of innovation perspectives and challenges. African Journal of Science, Technology, Innovation and Development, 2(3), 14-45. Ferreira, J.J., Fernandes, C.I., Raposo, M.L. (2017). The effects of location on firm innovation capacity. J. Knowl. Econ. 8(1), 77–96. Gan, W., & Xu, X., 2019. Does anti-corruption campaign promote corporate R&D investment? Evidence from China. Finance Research Letters, 30, 292-296 Griliches, Z., 1991. The search for R&D spillovers, No. w3768. National Bureau of Economic Research, https://www.nber.org/papers/w3768.pdf. Hall, B. H., Griliches, Z., & Hausman, J. A., 1986. Is there a second, technological opportunity) factor. International Economic Review, 27, 265-283 Harris, R.G., 2001. The knowledge-based economy: intellectual origins and new economic perspectives. Int. J. Manag. Rev. 3(1), 21–40. He, J. J., & Tian, X., 2013. The dark side of analyst coverage: The case of innovation. Journal of Financial Economics, 109(3), 856-878. Heimonen, T., 2012. What are the factors that affect innovation in growing SMEs?. European Journal of Innovation Management, 15(1), 122-144 Hirukawa, M., & Ueda, M., 2011. Venture capital and innovation: which is first?. Pacific Economic Review, 16(4), 421-465. Jaffe, A.B., 1989. Real effects of academic research. Am. Econ. Rev. 957–970. Lee, H. H., & Oh, F. D., 2018. Corporate innovation and credit default swap spreads. Finance Research Letters, https://doi.org/10.1016/j.frl.2018.12.030 Lin, S., 2015. Are ivory towers truly ivory? Knowledge spillovers and firm innovation. J. Econ. Bus. 80, 21–36. McCann, P., 2007. Sketching out a model of innovation, face-to-face interaction and economic geography. Spat. Econ. Anal. 2(2), 117–134. Mowery, D.C., Nelson, R.R., Sampat, B.N., Ziedonis, A.A., 2015. Ivory Tower and Industrial Innovation: University–Industry Technology Transfer before and after the Bayh-Dole Act. California: Stanford University Press Pakes, A., Griliches, Z., 1984. Estimating distributed lags in short panels with an application to the specification of depreciation patterns and capital stock constructs. Rev. Econ. Stud. 51(2), 243–262. Storper, M., Venables, A.J., 2004. Buzz: face-to-face contact and the urban economy. J. Econ. Geogr. 4(4), 351–370. Tsai, L. C., Zhang, R., & Zhao, C., 2019. Political connections, network centrality and firm innovation. Finance Research Letters, 28, 180-184 Veugelers, R., Cassiman, B., 2005. R&D cooperation between firms and universities. Some empirical evidence from Belgian manufacturing. Int. J. Ind. Organ. 23(5–6), 355–379. Wang, Z., Zhang, M., Sun, H., & Zhu, G., 2016. Effects of standardization and innovation on mass customization: An empirical investigation. Technovation, 48, 79-86. Yang, A. S., & Okada, H., 2019. Corporate innovations as institutional anomie: Patent activities
and financial performance of the international aerospace industry. Finance Research Letters, 28, 328-336. Zhang, D., Guo, Y., Wang, Z., & Chen, Y., 2019. The impact of US monetary policy on Chinese enterprises’ R&D investment. Finance Research Letters. https://doi.org/10.1016/j.frl.2019.09.016 Zhang, D., Tong, Z., & Li, Y. (2019). The role of cash holding towards cleaner production in China’s manufacturing sectors: A financial constraint perspective. Journal of Cleaner Production, https://doi.org/10.1016/j.jclepro.2019.118875.