Effectiveness of R&D subsidies depending on firm financial risks: evidence from Chinese listed firms

ABSTRACT

In this study, we investigate firm heterogeneity with respect to the financial risk that determines the effectiveness of public R&D subsidies. Using panel data of Chinese listed firms from 2007 to 2018, we find that firms with high financial risk are more responsive to government innovation funding. This positive impact is significant with respect to both private R&D spending and patent applications, but only for risky firms. Moreover, simultaneously considering another conventional criterion, firm size, we find that financially risky small-sized firms are most sensitive to public funding. The generalised propensity score and text mining methods are used to control for endogeneity in terms of the quantity and quality of innovation subsidies for robustness checks. This study suggests that priority should be given to small risky firms, as this criterion can screen out enterprises that are innovative and financially constrained.

KEYWORDS:

• Financial risk
• innovation subsidy
• R&D spending
• patent
• text mining

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The risk level is an indicator that combines the Z score and leverage ratio of enterprises using principal component analysis (PCA). Detailed information concerning this point is introduced in Section 2.

2 These authors found that government R&D funding promotes innovation performance but that EU-funded grants for physical and human capital upgrading in Poland were inefficient.

3 Generally, private R&D input is measured via R&D spending, R&D intensity, and innovation-related employment (Aerts & Schmidt, 2008; Czarnitzki & Lopes-Bento, 2013; González & Pazó, 2008; Özçelik & Taymaz, 2008). Measures of R&D output include the number of patent applications, new products and services, and citations (Aerts & Czarnitzki, 2004; Bérubé & Mohnen, 2009; Hewitt-Dundas & Roper, 2010; Li et al., 2022).

4 In this study, the impacts on firm R&D activities are categorised into three cases: positive (including additionality, neutrality, and partial crowding out), insignificant (full crowding out), and negative (over-full crowding out). In some studies, the impacts on R&D inputs have been classified in further detail by being divided into additionality, neutrality (no effect), partial crowding out, full crowding out, and over-full crowding out, but the positive impact on R&D outputs cannot be distinguished in more specific terms (Boeing, 2016; Dimos & Pugh, 2016). The main concern of this study is the effectiveness of R&D subsidies in terms of both input and output, and increases in gross R&D spending and innovation outcomes are regarded as positive impacts.

5 A significant number of previous studies have been unpublished working papers or discussion papers (Aschhoff, 2009; Czarnitzki & Hottenrott, 2011; Schneider & Veugelers, 2010), but we mainly review published papers.

6 It remains controversial whether other characteristics, such as age or industry, can improve the effectiveness of public subsidies. Some of the literature has argued that public subsidies for young firms are ineffective (e.g., Koga, 2005; Gimenez-Fernandez et al., 2020; Koga, 2003). Some studies focusing on firm industry types have claimed that government funding is more effective in high-tech industries (Colombo et al., 2011; Czarnitzki & Delanote, 2015), while others have found evidence suggesting that supporting low-tech industries is more effective (González & Pazó, 2008; Hall et al., 2009; Becker & Hall, 2013). Moreover, while public subsidies are reliably more effective for small businesses, such effects involve greater financial difficulties for small firms rather than a greater potential for innovation (Hall, 2002; Hall & Lerner, 2010). Therefore, an alternative measure is needed to determine whether innovation subsidies are sufficiently effective to reflect a firm’s enthusiasm and capability that are necessary to innovate.

7 In China, firms listed as distressed are classified into two categories, ST and PT, which indicate firms with two years and three years of consecutive losses, respectively.

8 We use the number of patent applications instead of the number of patents granted because the patent granting procedure lasts some time, and not all patent applications are granted (Bronzini & Piselli, 2016). This variable has been widely used by previous studies as a measure of private innovation activity outputs (e.g., Ascani et al., 2020; Gilding et al., 2020; Gao et al., 2021; Wang et al., 2022).

9 A Z score lower than 1.8 indicates that the firm is more likely to go bankrupt due to financial difficulties. On the other hand, a score of 3 or higher indicates that the firm is in a safe zone and is unlikely to file for bankruptcy. If the score is between these two points, the firm is in a grey zone and has a moderate likelihood of filing for bankruptcy.

10 The observations of patent applications are less numerous than other observations because the data are available only for the period 2007–2017.

11 The data concerning firm R&D spending covers the period 2007–2018, while the data pertaining to patent applications cover the period 2007–2017. Thus, low- and high-risk groups are divided based on the entire sample, which caused the imbalance between the two groups in terms of R&D output results. We also divided the sample related to R&D outputs into halves in accordance with median risk levels and found the results to be almost the same as those currently reported.

12 We also conducted an analysis of the related words and found that the word ‘transfer’ appears mostly alongside the word ‘technology’.

13 ROA (Cucculelli & Ermini, 2013; Hirshleifer et al., 2012; Jia et al., 2016; Sunder et al., 2017), cash holding ratio (Hirshleifer et al., 2012; Sunder et al., 2017), PPE ratio (Hirshleifer et al., 2012; Tian & Wang, 2014), and ownerships (Tian & Wang, 2014).

14 The percentage of high-school students in the population is closely correlated to the per capita GDP of the province, so the proportion of college students is used as an alternative.

Additional information

Funding

This research was supported by the National Research Foundation of Korea (NRF) funded by the Ministry of Education [2021R1A6A1A14045741].

Notes on contributors

Haizhen Jin

Haizhen Jin is a research fellow in the Department of Industrial Systems Engineering and Management at the National University of Singapore. She holds a Ph.D. in Economics from Seoul National University. Her research interests include innovation strategy and policy, technology management, institutional economics, and the Chinese economy.

Hyerim Lee

Hyerim Lee is a specialist at the Center for Regulatory Studies at Korea Development Institute (KDI), South Korea. She got her Ph.D. in Economics from Seoul National University. Her field of research focuses on corporate finance, industrial organization, and innovation policy.