Research on the impact of industry-finance cooperation on green total factor productivity from the perspective of indirect financing

Qiu, Wei et al. Article Research on the impact of industry-finance cooperation on green total factor productivity from the perspective of indirect financing Journal of Innovation & Knowledge (JIK) Provided in Cooperation with: Elsevier Suggested Citation: Qiu, Wei et al. (2024) : Research on the impact of industry-finance cooperation on green total factor productivity from the perspective of indirect financing, Journal of Innovation & Knowledge (JIK), ISSN 2444-569X, Elsevier, Amsterdam, Vol. 9, Iss. 1, pp. 1-11, https://doi.org/10.1016/j.jik.2024.100470 This Version is available at: https://hdl.handle.net/10419/327375 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Research on the impact of industry−finance cooperation on green total factor productivity from the perspective of indirect financing Wei Qiu a,b,1 , Yaojun Bian c,2 , Siyu Ren d,3 , Jingxia Chai e,f,4 , Shang Gao e,5 , Haitao Wu e,6, * a School of Economics, Xinjiang University of Finance and Economics, Urumqi 830012, Xinjiang, PR China b Postdoctoral Research Station of Applied Economics, Fudan University, Shanghai 200433, PR China c School of Economics and Management, Xinjiang University, Urumqi 830046, Xinjiang, PR China d School of Economics, Nankai University, Tianjin, 300071, PR China e School of Management and Economics, Beijing Institute of Technology, Beijing, 100081, PR China f School of Economics and Management, Yan’an University, Yan’an, 716000, PR China ARTICLE INFO Article History: Received 16 December 2022 Accepted 27 February 2024 Available online 11 March 2024 ABSTRACT The Chinese government aims to use environmental policy measures to improve society’s overall green total factor productivity (GTFP) and create a new development pattern in which the economy and the environment are equally important. Based on the perspective of indirect financing, this study divides 272 prefecture-level industry−finance cooperation cities in China from 2015 to 2021 into 109 resource-based cities and 163 nonresource-based cities; we use the PSM-difference-in-differences (DID) model to explore their impact on GTFP. The research results are as follows. First, industry−finance cooperation significantly promotes GTFP. Second, although the nonresource-based cities have insignificant industry−finance cooperation, resourcebased cities have significantly positive industry−finance cooperation. Third, from the perspective of indirect financing, industry−finance cooperation can effectively improve GTFP. Fourth, the reliability of the conclusions is verified by successively using the parallel trend test, the placebo test, the dynamic time window test, the counterfactual test, and the elimination of other policy interferences. Based on the above findings, the Chinese government must always adhere to the important “comprehensive regulation and measures according to local conditions”policy and fully use “information technology and green credit.” © 2024 The Authors. Published by Elsevier España, S.L.U. on behalf of Journal of Innovation & Knowledge. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Keywords: Green total factor productivity Industry−finance cooperation Indirect finance PSM-DID model JEL codes: O13 O15 P28 Introduction The central contradiction in Chinese society has recently been transformed into “the contradiction between the people’s ever-growing needs for a better life and unbalanced and insufficient development.”China’s pollutant emissions ranked first in the world in 2009 (Lin et al., 2019), and environmental pollution has become one of the essential causes of major social conflicts (Liu and Kalirajan, 2024;Qiu et al., 2022b). Therefore, the Chinese government clearly stated that promoting total factor productivity through supply-side structural reforms is necessary to ensure the coordination and unity of economic development and ecological environment construction; however, traditional total factor productivity does not reflect the impact of environmental pollution on social welfare and sustainable economic development. Therefore, scholars have proposed green total factor productivity (GTFP), a comprehensive indicator considering economic development and ecological environment construction. China’s industrial policy, implemented in the 1980s, has made remarkable contributions to the country’s rapid economic growth * Corresponding author. E-mail address: [email protected] (H. Wu). 1 Dr. Wei Qiu is currently a full associate professor at School of Economics, Xinjiang University of Finance and Economics 2 Mr. Yaojun Bian is currently a Ph.D. student on economics at School of Economics and Management, Xinjiang University. 3 Mr. Siyu Ren is currently a Ph.D. student on economics at School of Economics, Nankai University 4 Ms. Jingxia Chai is currently a Ph.D. student on economics at School of Management and Economics, Beijing Institute of Technology (BIT) 5 Mr. Shang Gao is currently a Ph.D. student on Business Administration at School of Management and Economics, Beijing Institute of Technology (BIT) 6 Dr. Haitao Wu is a highly regarded scholar at the School of Management and Economics, Beijing Institute of Technology (BIT). His-research focuses on macroeconomics, business economics, ecological economics, and energy economics. With an impressive track record, he has authored over 80 peer-reviewed papers in renowned SCI/SSCIindexed journals such as the Journal of Business Ethics (FT50), Environmental and Resource Economics (ABS/AJG 3), The Energy Journal (ABS/AJG 3), Energy Economics (ABS/AJG 3), Kyklos (ABS/AJG 3), Technological Forecasting and Social Change (ABS/ AJG 3), Business Strategy and the Environment (ABS/AJG 3), and Journal of Environmental Management (ABS/AJG 3). Notably, he was selected as top 2 % of scientists (Stanford and Elsevier) and highly cited young scientist (Scilit).Furthermore, he was honored with the Best Paper Award in Energy Economics in 2022. Dr. Wu’s contributions have garnered significant attention, as evidenced by his H-index exceeding 40 and his papers being cited over 7000 times according to Google Scholar. https://doi.org/10.1016/j.jik.2024.100470 2444-569X/© 2024 The Authors. Published by Elsevier España, S.L.U. on behalf of Journal of Innovation & Knowledge. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Journal of Innovation & Knowledge 9 (2024) 100470 Journal of Innovation &Knowledge https://www.journals.elsevier.com/journal-of-innovation-and-knowledge (Cao et al., 2023,2024;Han et al., 2014;Huang et al., 2023;Zhang & Feng, 2024;Zhang & Zhang, 2023), but it has also caused increasingly serious environmental pollution problems. For example, in 2015, China’s industrial sulfur dioxide emissions were 27.5571 million tons (see Fig. 1). With the slogan "clear waters and green mountains are golden mountains and silver mountains," the emission of industrial sulfur dioxide respectively decreased to 1.7518 million tons; however, GTFP decreased from 1.011 in 2015 to 1.009 in 2021 (see Fig. 1). This change may be due to the failure of existing industrial policies to effectively promote technological progress. Furthermore, improving the ecological environment will be at the expense of rapid economic growth in the case of low-production technology. Given this, China’s Ministry of Industry and Information Technology and four other departments launched the pilot project of industry−finance cooperation in 2016, aiming to improve the financing environment of real enterprises and GTFP and complete the upgrading of production technology. The research on industry−finance cooperation mainly focuses on enterprise innovation (Chen and Xia, 2019;Dzhambova, 2024;Stice et al., 2023;Xiong and Gui, 2019;Yang et al., 2024), enterprise growth (Lin, 2010;Shen et al., 2019), asset appreciation (Wu et al., 2018), investment efficiency (Hu et al., 2019;Li et al., 2014), and financing constraints (Wan et al., 2015;Yang et al., 2023). At the same time, the GTFP literature mainly focuses on environmental regulation (Ambec et al., 2013;Wu et al., 2024;Zhang, 2021;Zhuo et al., 2022), Internet development (Amin et al., 2019;Hao et al., 2024;Popkova et al., 2022;Zhang et al., 2022b), and artificial intelligence (Zhao et al., 2022); however, no unified opinions exists on whether industry−finance cooperation can effectively improve GTFP. One view is that the industrial policy using industry convergence as an example is effective (Barkhordar et al., 2018;Criscuolo et al., 2012). Some scholars suggest that because different types of industrial policies will produce different effects (Schleich et al., 2017;Mao et al., 2023), different industrial policies need to be combined to effectively promote the coordinated development of the economy (Reichardt et al., 2016;Weber et al., 2012;Wang et al., 2019,2020). Moreover, from the perspective of enterprise ownership change, industry−finance cooperation can significantly improve total factor production efficiency (Ongena et al., 2009;Yang, 2017;Zhu et al., 2014), upgrade production technology, and achieve the goal of GTFP improvement. Conversely, another view is that industrial policy is ineffective (Blonigen, 2016), and problems such as government corruption will make the industrial policy with industry−finance cooperation, for example, unable to effectively achieve its purpose of promoting GTFP (Chang, 2006). Furthermore, industry−finance cooperation will lead to the financialization of real enterprises and reduce their investment in productive innovation activities, thereby reducing the overall level of innovation in society (Wang et al., 2017), inhibiting GTFP improvement. The large profit gap between the real and virtual economies and the high risk of productive innovation activities may be the critical reasons for the inhibitory effect of industry−finance cooperation on GTFP. Therefore, to avoid the negative impact of industry−finance cooperation on GTFP, the market and government means should be effectively combined. Based on the perspective of the combination of market and government, this paper elaborates on the influence logic of industry−finance cooperation on GTFP. At the same time, in the empirical aspect, more scholars use the difference-in-differences (DID) model to study GTFP (Li et al., 2022;Zhang et al., 2020;Zhang et al., 2022a). Moreover, because the propensity score matching-DID (PSM-DID) model can better solve the defect, the traditional DID model must meet the “random grouping”requirement. Therefore, this paper uses the PSM-DID model to verify the relationship between industry−finance cooperation and GTFP. Furthermore, the financing constraints faced by smalland medium-sized enterprises (SMEs) are also one of the reasons why the innovation level is low and GTFP cannot be significantly improved. The cooperation between industry and finance can solve the problem of financing constraints (Ma and Wang, 2017;Wang et al., 2016;Zhou et al., 2018). Simultaneously, with the continuous development of digital technology, indirect financing can break the limitation of time and space and provide diverse financial services to people in remote areas (Chen and Chen, 2018;Fu et al., 2018;Ren et al., 2019;Wang and Guo, 2022). Indirect financing can also alleviate the asymmetry between information, increase the allocation efficiency of financial resources (Wu et al., 2022), improve the ecological environment (Adams and Klobodu, 2018;Tamazian et al., 2009;Salahuddin et al., 2015), and significantly promote the improvement of GTFP (Cheng et al., 2023). Therefore, based on the perspective of indirect financing, this study starts from the market and government levels to analyze the impact of industry−finance cooperation on GTFP. The contributions of this study include the following points. First, in terms of topic selection, although many empirical studies examine industry−finance cooperation, most are in the market context; however, little literature discusses the relationship between industry −finance cooperation and GTFP by combining the market and government. Second, for research methods, this study uses the PSM-DID Fig. 1. Line chart of industrial so 2 emissions and GTFP in China from 2015 to 2021. W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 2 model to empirically analyze the relationship between industry −finance cooperation and GTFP. Our sample is based on the panel data of 272 prefecture-level cities in China from 2015 to 2021, which avoids selectivity bias. Third, in terms of research perspective, although many studies have examined the pairwise relationship between industry−finance cooperation, indirect financing, and GTFP, the extant literature does not place the three into the same framework. The remainder of this paper is structured as follows. Section ’Mechanisms and hypotheses’discusses the mechanism between industry−finance cooperation, GTFP, and indirect financing and presents the corresponding research hypotheses. Section ’Methods and data’briefly expounds the basic form of the PSM-DID model and introduces the calculation method of the relevant variables. The empirical results are discussed mainly in Section ’Results and discussion’. Section ’Conclusions and policy recommendations’briefly summarizes this study’s conclusions and makes relevant recommendations. Mechanisms and hypotheses The influence mechanism of industry − finance cooperation on GTFP The cooperation between industry and finance mainly completes upgrading production technology through benefit synergy, management synergy, and financial synergy to improve GTFP. First, from the perspective of benefit synergy, in the case of entity enterprises participating in financial enterprises, the cooperation between industry and finance can promote the gradual convergence of interests between entity enterprises and financial enterprises, ease the disputes between the two, and thereby reduce transaction costs. The reduction of transaction costs between real enterprises and financial enterprises increases the investment desire of financial enterprises, enabling real enterprises to complete their financing goals and promoting the improvement of GTFP. When a real enterprise holds more shares of a financial enterprise, it can influence the decision-making of the board of directors of the financial enterprise to a greater extent, making it easier for the financial enterprise to pass the risk assessment of innovative projects. In this case, the interests of the entity enterprise and the financial enterprise are more consistent, and the entity enterprise can disclose the technological innovation that reflects its comparative advantage to a certain extent. Disclosing this information can alleviate the information asymmetry between enterprises, thus alleviating their financing constraints (Myers et al., 1984). Furthermore, even if a real company only holds a small share of financial companies, it can more easily gain the trust of financial companies, reduce transaction costs between companies, and obtain loan support from financial institutions. Second, from the perspective of management synergy, the cooperation between industry and finance makes the relevant functions of some departments of real enterprises overlap with those of financial enterprises, which prompts real enterprises to reduce such costs. Reducing production costs can benefit enterprises by allowing them to concentrate funds to upgrade production technology and maintain their competitive advantages while improving GTFP (Sirower et al., 1997). Finally, from the perspective of financial synergy, industry−finance cooperation will further reduce production costs through reasonable tax avoidance, ensure the smooth development of innovation activities, and improve GTFP. In the case of entity enterprises participating in nonlisted financial institutions, each subsidiary has different capabilities, and its marginal profit and marginal cost also differ. Therefore, entity enterprises can make tax payments under the optimal profit scale by stabilizing the profit scale of each subsidiary, thereby achieving the goal of reasonable tax avoidance. Given this, the cooperation between industry and finance will reduce the difficulty of loans for enterprises, reduce the production costs of enterprises, and reasonably avoid the tax costs of enterprises through benefit synergy, management synergy, and financial synergy, thus completing the technological upgrading and achieving the goal of GTFP improvement. Nonetheless, relying solely on the market’s industry−finance cooperation model can easily lead to problems such as over-investment and under-investment. In the early stage of China’s economic development, real enterprises tended to complete productive expansion through extensive development methods to obtain excess profits. Financial companies have also relaxed financing constraints on real companies because of their optimistic economic forecasts, which has led to problems such as excessive investment. Moreover, as China’s economy gradually enters the “new normal”stage, the profit gap between the real and virtual economies continues to expand. To improve the existing profit level, state-owned enterprises use their good relationship with financial enterprises to obtain loans and lend to SMEs through asset securitization and other methods (Demir, 2009). This behavior crowds out the productive investment of stateowned entity enterprises and leads to the continuous expansion of the credit chain. When financial institutions tend to issue loans to large enterprises, the credit chain expansion further increases SMEs’ financing costs, resulting in insufficient investment in society (Orhangazi, 2008). Moreover, the industry−finance cooperation model under the market background can only solve the problem of financing constraints of large enterprises. At the same time, SMEs with inherently insufficient capital cannot obtain sufficient funds through equity participation or the establishment of financial departments. Therefore, as the “visible hand,”the government must supplement and improve the market. Establishing a pilot project for industry-fusion cooperation can accurately reflect this perspective. It can better inherit the advantages of industry−finance cooperation under the market background and compensate for the market’s shortcomings. First, industry−finance cooperation will give full play to its supervisory effect. Industry−finance cooperation can curb the extensive development model of real enterprises through environmental regulations and other means and solve problems such as excessive investment and environmental pollution. Second, due to the difficulty of reviewing small and micro enterprises and the low amount of a single loan, financial institutions are reluctant to lend to small and micro enterprises, thus hindering their long-term development; however, industry−finance cooperation can effectively exert its long tail effect. With the support of digital information technology, financial enterprises can more easily complete basic tasks such as loan enterprise reviews. Decreasing the review difficulty will reduce the business cost of financial enterprises, increase their desire to invest in small and micro enterprises, resolve problems such as insufficient investment, and improve GTFP. Finally, industry−finance cooperation can also form a market signal. Because industry−finance cooperation aims to create an environmentally friendly development model, investors are more inclined to invest in green innovation projects based on this market signal. The adjustment of investment tendency will broaden the direct financing channels for SMEs engaged in production and innovation activities, thereby improving the overall GTFP level of society. Based on the above analysis, this study puts forward the research hypothesis H1: From the market and the government perspective, industry−finance cooperation significantly promotes GTFP. The influence mechanism of industry − finance cooperation on indirect financing First, from the perspective of enterprises, financial institutions tend to set higher risk levels for SMEs due to their lack of asset collateral, increasing the severity of their financing constraints. However, the cooperation between industry and finance will reduce the loan risk of SMEs and transaction costs and encourage financial institutions to issue loans to them (Lu et al., 2012). Specifically, with the W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 3 support of digital technology, industry−finance cooperation will make full use of the convenience of government departments to collect and disclose information on the productive activities of SMEs. Improving information disclosure will reduce the risk level of SMEs and the transaction costs between SMEs and financial institutions. Afterward, industry−finance cooperation will provide small and medium enterprises with suitable loans through financing platforms to solve their financing constraints (Ghisetti et al., 2015). Compared with large companies, industry integration can improve the ability of indirect financing of SMEs. Second, from an industry perspective, industry−finance cooperation can better promote the improvement of indirect financing capabilities in highly competitive industries. In the market context, monopolistic industries with lower competition have a higher market share and more substantial bargaining power for commodities. Therefore, monopoly industries can control and adjust the quantity and price of circulating products, obtain higher profits, and avoid falling into a "cash flow predicament"; however, industries with higher levels of competition are more vulnerable to threats from competitors. Such industries need sufficient funds for productive expansion to maintain their leading position; therefore, the promotion effect of the industry−finance cooperation on indirect financing can be more reflected in industries with increased competition. Finally, from a regional perspective, industry−finance cooperation can improve the indirect financing level of the society as a whole by optimizing the redistribution of capital. Under the influence of the “siphon effect,”enterprises in central and western China face more serious financing constraints. However, with the assistance of local governments, industry−finance cooperation will guide the transfer of capital from the eastern coastal areas to the central and western regions by building a financing platform, thereby optimizing the redistribution of capital and improving the level of indirect financing for enterprises. Based on the above analysis, this study puts forward the research hypothesis H2: Whether at the enterprise, industry, or regional level, industry−finance cooperation has a significant role in promoting indirect financing. The influence mechanism of indirect financing on GTFP First, in terms of factor allocation efficiency, indirect financing mainly adopts green credit to improve capital utilization efficiency, thereby improving society’s overall GTFP. Specifically, financial institutions set higher loan interest rates for “three high”companies, which reduces their financing scale. Reducing the financing scale will inhibit the productive expansion of “three high”enterprises, forcing them to invest funds in green innovation activities (Zhang et al., 2011). Conversely, for environmentally friendly enterprises, financial institutions set lower loan interest rates and longer repayment periods to ensure the smooth development of their green innovation activities (Qiu et al., 2023). Therefore, under the guidance of green credit, indirect financing can improve society’s overall factor allocation efficiency by adjusting the loan interest rates of “three high” enterprises and environment-friendly enterprises, thereby improving GTFP (Li et al., 2018). Second, in terms of investment in innovation, indirect financing, with the support of digital information technology, expands the scale of financial institutions’funds and resolves the predicament of insufficient funds for the promotion of GTFP. With the support of digital information technology, financial institutions can absorb idle funds in remote areas more efficiently and satisfy the preferences of various investors by enriching the types of financial products they offer. Consequently, financial institutions expand the scale of funds with the support of digital information technology (Gomber et al., 2017; Li et al., 2020). Further analysis shows that because financial institutions have more sufficient funds, enterprises that were initially facing financing constraints can obtain more loans, which is conducive to solving the problem of insufficient funds for enterprises’productive innovation activities and improving the overall GTFP of society. Finally, regarding the frequent exchange of information, financial institutions will tend to issue loans to green investment projects and promote the improvement of GTFP through technological upgrading. With the continuous improvement of economic development, the public gradually realizes the importance of ecological environment construction, so they are more inclined to buy environmentally friendly products. At the same time, in the contemporary era of increasingly frequent information exchanges, it is easier for financial institutions to accurately analyze the purchasing behavior of the public through information technology and to judge that green and sustainable projects have larger profit margins. Therefore, the investment willingness of profit-seeking financial institutions will shift from extensive scale expansion to green innovation to improve the utilization efficiency of indirect financing and then improve GTFP through upgrading production technology. Based on the above analysis, this study proposes research hypothesis H3: Indirect financing has a significant role in promoting GTFP in three aspects: factor allocation efficiency, investment in innovation, and frequent exchange of information. The mechanism analysis diagram is as follows (see Fig. 2). Methods and data Econometric methodology Benchmark regression model The traditional DID model requires the samples to meet the "random grouping" assumption, while the pilot cities for industry −finance cooperation need a solid industrial foundation, a high level of industrial chain, and rich financial resources. Therefore, the relationship between industry−finance cooperation and GTFP cannot be empirically analyzed using the traditional DID model; however, the PSM-DID model can effectively resolve the above problems (Heckman et al., 1997). Specifically, the PSM-DID model first uses the PSM method to find the control group samples similar to the experimental group; thus, the samples meet the “random grouping”assumption and balance requirements. After that, the PSM-DID model uses the basic DID model to empirically analyze the relationship between industry−finance cooperation and GTFP (Rosenbaum et al., 1983). The specific measurement model is as follows: lnGTFPi;t¼b0þb1timei;tcityi;t  þb2X 6 k¼1 lnXk;i;tþai þmtþei;t:ð1Þ Here, irepresents the prefecture-level city, and trepresents the year. time i,t is the time dummy variable—if tis 2015 and 2016, then time i,t =0;iftis 2017−2021, then time i,t =1.city i,t is a grouping dummy variable. When iis the control group, city i,t = 0; when iis the experimental group, city i,t =1.X k,i,t represents the control variable. a i and m t denote fixed effects for region and year, respectively. e i,t denote random disturbance terms. Lastly, Fig. 3 illustrates the overall research framework. The design of the mediation model This study explores the impact of industry−finance cooperation on GTFP from factor allocation efficiency of indirect financing, investment in innovation, and frequent exchange of information. First, based on factor allocation efficiency, industry−finance cooperation will give full play to its supervisory effect and examine the pollutant emissions of enterprises by implementing follow-up supervision and other measures. Such inspections can optimize the allocation of the W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 4 whole society’s factors and promote the improvement of GTFP. Second, from the perspective of investment in innovation, industry −finance cooperation will fully absorb idle funds from society and expand the scale of financial institutions’funds with the support of digital information technology. Afterward, the industry−finance cooperation association will issue loans to enterprises engaged in innovation activities in the form of “point-to-point”to increase the capital investment in productive innovation activities and then achieve the goal of improving GTFP. Finally, from the perspective of frequent exchange of information, with the assistance of local governments, industry−finance cooperation will promote green and environment-friendly lifestyles, increase the public’s desire to buy green products, and expand the profit margins of environmentfriendly products. The expansion of profit margins of environmentally friendly commodities will encourage financial institutions to invest in them, thus promoting the promotion of GTFP in society. This study proposes the following mediation model to test the above theories: lnGTFPi;t¼b0þb1timei;tcityi;tlnIFi;t  þb2X 6 k¼1 lnXk;i;tþaiþmtþei;t:ð2Þ Here, IF i,t is the indirect financing of region iin year t. Data Explained variable As a comprehensive indicator for economic development and ecological environment construction, GTFP is mainly calculated through input and output indicators (Hu and Yang, 2011;Li, Peng, & Ouyang, 2013). Fig. 2. Mechanistic diagram between industry-finance cooperation, indirect financing and GTFP. Fig. 3. Research framework. W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 5 Among them, the input indicators are divided into labor input, capital input, and energy input. The labor input is the average number of employees in the prefecture-level cities, the capital input is the capital stock obtained using the perpetual inventory method, and the citywide electricity consumption data measure the energy input. In addition, output indicators are mainly divided into expected output and undesired output. The expected output is each city’sgross domestic product (GDP), and the environmental pollution index represents the undesired output. Specifically, this study uses the entropy method to calculate the comprehensive environmental pollution index according to pollutant discharge indicators such as industrial wastewater discharge, industrial waste gas discharge, and industrial solid waste discharge. Core explanatory variable In terms of the year before and after the implementation of industry−finance cooperation, four departments, including the Ministry of Industry and Information Technology, launched the application for the pilot application of industry−finance cooperation in 2016, and at the end of 2016, announced the list of 37 “China’s industry−finance cooperation pilot cities”including Beijing. Therefore, this study takes 2017−2021 as the implementation year of industry−finance cooperation and sets 2015−2016 as the period before the policy was introduced. Additionally, in the division of the experimental group and the control group, this study is based on the list of “China’s Industry −finance cooperation pilot cities”announced at the end of 2016. After excluding the county-level city samples, 31 pilot cities, such as Harbin and Xiamen, were selected as the experimental group. The rest of the prefecture-level city samples were the control group. Mediator Indirect financing mainly refers to the process of capital-shortage units and capital-sufficient units through financial institutions to achieve financial integration. Specifically, entities with sufficient funds deposit idle funds in financial institutions using deposits, purchase of securities of financial institutions, etc. After that, financial institutions can lend funds to capital-strapped units through loans, discounts, etc. Furthermore, the measurement method of indirect financing adopts the ratio of the balance of deposits and loans provided by financial institutions to GDP. Control variables First, the secondary production to GDP ratio refers to industrial upgrading. Second, human capital (HC), which is measured by the number of college students per 10,000 people. Third, per capita GDP (PGDP) is the ratio between the regional gross domestic product and the total population at the end of the year. Fourth, the marketization level (ML) is represented by the proportion of fiscal expenditure in GDP. Fifth, the urbanization rate is calculated by the ratio between the urban population and the total population at the end of the year. Sixth, the trade openness (TO) is the percentage of each prefecturelevel city’s total import and export to GDP. Data sources Due to the lack of data in Tibet, Hong Kong, and other regions, this study selected the panel data of 272 prefecture-level cities in China from 2015 to 2021. Moreover, the data used in this study are from the EPS database and “China Urban Statistical Yearbook.”Table 1 provides the names of the variables in this study, variable descriptions, and descriptive statistics. This study divided the data from 2015 to 2021 into two parts to examine the mean, standard deviation (SD), minimum (Min), and maximum (Max) for each variable. Although the mean for GTFP in 2015−2017 was not significantly different from the mean in 2018−2021, the SD for GTFP decreased significantly in 2018−2021, indicating a gradual decrease in the gap between GTFP in each city. For the control variables, the mean of ML decreased significantly, indicating that the national marketization level increased significantly. At the same time, with the gradual development of China’s economy, the average PGDP has increased significantly. Results and discussion Baseline estimate Whether the results of PSM are reliable determines whether the samples meet the requirements of balance, affecting the stability of the empirical results. Given this, this study carries out the following verifications. First, the PSM balance test is used to determine whether the results of the PSM satisfy the "conditional independence assumption" (see Table 2). Most matched variables (Matched) had standard deviations less than 20 in absolute value, indicating that the results of PSM were valid (Rosenbaum and Rubin, 1983). Second, since the T values of most of the matching variables are not significant after matching, the results of PSM conform to the null hypothesis that the means of matching variables after matching are equal; that is, the results of PSM are valid. Finally, to test the quality of the PSM results, this study compared whether the overlap of the nuclear density maps between the control group and the experimental group before and after matching increased (see Fig. 4). The test results show that the overlapping part of the nuclear density maps of the control and experimental groups increased significantly after matching; that is, the matching quality improved. Therefore, the results of PSM are Table 1 Descriptive statistics. Variables Total 2015−2017 2018−2021 GTFP:Green total factor productivity Mean 1.009 1.009 1.009 S.D. 0.025 0.029 0.023 Min 0.960 0.960 0.961 Max 1.060 1.060 1.060 IF:Indirect financing (%) Mean 2.762 2.554 2.917 S.D. 1.281 1.188 1.326 Min 0.910 0.910 1.201 Max 21.301 12.567 21.301 IU:Industrial upgrading (%) Mean 0.426 0.455 0.404 S.D. 0.105 0.105 0.100 Min 0.127 0.139 0.127 Max 0.745 0.745 0.729 HC:Human capital (person/10,000 people) Mean 183.343 187.745 180.042 S.D. 205.475 221.162 192.913 Min 6.000 6.000 8.000 Max 1294.000 1294.000 1003.000 PGDP:GDP per capita (RMB 10,000/person) Mean 6.130 5.494 6.606 S.D. 3.390 3.129 3.499 Min 1.099 1.099 1.266 Max 21.812 21.549 21.812 ML:Marketization level (%) Mean 4.025 5.633 2.819 S.D. 3.017 2.141 3.017 Min 0.059 1.448 0.059 Max 15.111 12.584 15.111 UR:Urbanization rate (%) Mean 0.565 0.549 0.577 S.D. 0.162 0.157 0.165 Min 0.001 0.001 0.001 Max 1.000 1.000 1.000 TO:Trade openness (%) Mean 0.211 0.203 0.217 S.D. 0.496 0.330 0.591 Min 0.001 0.001 0.001 Max 11.836 2.773 11.836 W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 6 more reliable. Through the PSM-DID model, this study can further verify the relationship between industry−finance cooperation and GTFP. To accurately verify research hypothesis H1, this study adopts model (1) (a static fixed-effect model) and uses a dynamic fixedeffect model to empirically analyze the relationship between industry−finance cooperation and GTFP (see Table 3). The specific steps of the dynamic fixed-effect model are as follows. Since the start year of the industry−finance cooperation is 2017, we set 2017, 2018, 2019, 2020, and 2021 as year dummy variables that take the value of 1; the other years are 0. After that, the above dummy variables and grouping dummy variables (city i,t ) are used as interaction terms. If the coefficients are significant, industry−finance cooperation has a dynamic effect (Xuan et al., 2020). On the one hand, the static fixed effect shows that the industry−finance cooperation coefficient is significantly positive at the 10 % level, verifying research hypothesis H1. The promoting effect of industry−finance cooperation on shortening the credit chain may be an important reason for the above results. Differences in the attitudes of financial institutions toward risk-heterogeneous enterprises and the profit gap between real enterprises and the virtual economy are essential reasons for shadow banking, which crowded out the productive investment of large enterprises and enlarged the credit chain within the financial system. The expansion of the credit chain increases the financing cost of high-risk SMEs, reduces the productive investment of SMEs, and thus significantly reduces the productive innovation activities of the whole society. However, the cooperation between industry and finance will promote direct communication between financial and real enterprises by building a financing platform to shorten the credit chain. Further analysis shows that shortening the credit chain reduces SMEs’financing costs, promotes the expansion of their production scale, indirectly narrows the profit gap between the virtual economy and the real economy, and reduces the participation of large enterprises in shadow banking. Moreover, the shortening of the credit chain can effectively alleviate systemic risks, increase the willingness of financial institutions to invest in innovative projects and improve society’s overall GTFP. Therefore, industry−finance cooperation has a significant promoting effect on the GTFP. On the other hand, from the perspective of dynamic fixed effects, although industry−finance cooperation was formally implemented in 2017, its coefficient was not significantly positive at the 10 % level until 2018 (see Table 3). The time lag of the fusion work is one of the main reasons for this result. At the government level, the approval of administrative procedures needs a certain amount of time. At the same time, the deployment of personnel and materials related to industry−finance cooperation also needs some time to complete. Furthermore, at the enterprise level, due to the information asymmetry, the enterprise needs to spend a certain amount of time to grasp the information related to the industry−finance cooperation accurately and in detail. Fig. 4. Nuclear density map before and after PSM. Table 2 PSM balance test results. Match variables Before matching (U) Mean Standard deviation (%) T test After matching (M) Test group Control group Deviation Reduce P>|t| lnIU U 0.89133 0.88777 1.3 281.2 0.859 M0.81599 0.82957 4.8 0.719 lnHC U 5.4946 4.6484 95.1 99.0 0.000 M 4.9356 4.9444 1.0 0.941 lnPGDP U 2.0521 1.6332 86.1 98.2 0.000 M 1.7821 1.7745 1.6 0.903 lnML U 0.76799 0.68472 5.1 21.4 0.455 M 0.89362 0.95905 4.0 0.774 lnUR U 0.40873 0.75706 47.5 97.0 0.000 M0.5451 0.53475 1.4 0.741 lnTO U 1.946 2.6695 50.5 65.1 0.000 M2.2186 2.471 17.6 0.179 Table 3 Benchmark estimation results. Variables Static effect Dynamic effect time £city 0.010* year17 £city 0.014 year18 £city 0.018* year19 £city 0.009 year20 £city 0.009* year21 £city 0.003 Control variables Control Control Time fixation Control Control Individual fixation Control Control N 1904 1904 R 2 0.0084 0.0100 Note: Significance of the coefficients at 10 %, 5 %, and 1 % are indicated as *, * *, and * * *, respectively. W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 7 Given this, industry−finance cooperation did not effectively promote GTFP until 2018. In addition, the coefficient of industry−finance cooperation in 2020 is also significantly positive at 10 % (see Table 3), which may result from increased government support. Due to the outbreak of COVID-19 in 2020, most SMEs are at risk of capital chain rupture. In response, local governments have increased their support for SMEs. Therefore, the 2020 industry−finance cooperation can significantly improve GTFP. Heterogeneous effects This study divides the sample of prefecture-level cities into resource-based and nonresource-based cities 7 and draws the following conclusions (see Table 4): The coefficient of resource-based cities is significantly positive at the 1 % level. This result may occur because the resource-based industry is the pillar industry of a resource-based city (Qiu et al., 2022a), and the financial development level of resource-based cities is relatively low. Resource-based cities with resource-based industries as their pillar industries tend to take advantage of the low-cost advantage of fossil energy, such as coal, to take the road of extensive economic development, which makes it easier to step into the “resource curse”trap, and it is difficult to complete the GTFP upgrade alone. Although the local governments of resource-based cities try to force the “three high”enterprises to flow funds into productive innovation activities through environmental regulation, due to the shortage of internal funds, it is difficult for enterprises to complete the transformation of production technology and improve GTFP by themselves. However, industry−finance cooperation will actively use its long tail effect to absorb idle funds from society effectively and expand the scale of funds of financial institutions. Afterward, it will issue loans in a “peer-to-peer”form, alleviating resource-based enterprises’“cash flow dilemma”and increasing capital investment in their productive innovation activities. At the same time, industry−finance cooperation will also form a market signal and increase investors’desire to invest in green and innovative projects, thereby expanding the channels for direct financing of resource-based enterprises. Further analysis shows that expanding financing channels for green innovation projects can ensure that resource-based enterprises can obtain sufficient funds. Moreover, resource-based enterprises with sufficient funds can improve their marginal production efficiency by purchasing advanced equipment and taking independent research and development measures to reduce pollution emissions. Furthermore, nonresource-based cities in the eastern coastal area have a relatively high level of financial development, which can reduce the cost of productive innovation activities through existing financial tools and share their risks to achieve the goal of upgrading production technology. Conversely, for resourcebased cities in the central and western regions with low financial development, in a purely market context, the high-risk characteristics of innovation activities make local financial institutions reluctant to issue loans to green investment projects; however, industry −finance cooperation will give full play to its role in government regulation. This action will encourage financial institutions in resourcebased cities to invest in green and innovative projects, thereby reducing the difficulty of innovative corporate loans. Therefore, industry −finance cooperation has a more significant impact on GTFP in resource-based cities. Mediation effect To verify the correctness of the research hypotheses H2 and H3, we adopt model (2) and draw the following conclusions (see Table 5). The time £city £lnIF coefficient is significantly positive at the 5 % level. An important reason is that industry−finance cooperation has enriched the types of financial products and has a supervisory effect. By enriching the types of financial products, industry−finance cooperation meets the financing needs of different types of enterprises, thereby expanding the scale of indirect financing for society. “Small and micro easy loans”can ease the financing constraints of small and micro enterprises by calculating interest daily and repaying loans as needed; they can also accelerate their financing speed by taking the form of online processing. At the same time, for green and innovative entity enterprises, bank-insurance linkage products such as “cloud computing service insurance”reduce the risk of their productive innovation activities and expand their financing channels. Further analysis shows that industry−finance cooperation will give full play to its supervisory effect and ensure that funds are used to upgrade production technology, thereby improving GTFP. Specifically, before issuing loans, financial institutions will actively use digital information technology to review the business status of enterprises thoroughly and determine whether they can carry out productive innovation activities. Furthermore, after issuing loans, financial institutions will cooperate with local government departments to check whether the enterprise conforms to the green development mode of production. Therefore, in terms of indirect financing, industry −finance cooperation significantly promotes GTFP. Robustness test Parallel trend test This study conducted the following parallel trend test (see Fig. 5) to ensure that the above conclusions are unbiased. Before the implementation of industrial−financial cooperation, the trend of GTFP change was not apparent. After the implementation of the policy, GTFP increased significantly; that is, the trend of GTFP change before and after the implementation of the policy. Therefore, this study believes the above conclusions align with the parallel trend assumption. Placebo test This study conducted the following placebo test (see Fig. 6)to exclude interference from other unknown factors with the above conclusions. The test results show that the absolute value of the tvalue of most of the sampling estimated coefficients is within 2, indicating that most of the estimated coefficients of generatrix cooperation are insignificant in the placebo test. Therefore, the above results were considered relatively robust. Table 4 Results of heterogeneous effects. Variables Resource-based cities Non-resource-based cities time £city 0.024*** 0.006 Control variables Control Control Time fixation Control Control Individual fixation Control Control N 763 1141 R 2 0.0211 0.0213 Note: Significance of the coefficients at 10 %, 5 %, and 1 % are indicated as *, * *, and * * *, respectively. Table 5 Mediating effect results. time £city £lnIF Control variables Time fixation Individual fixation NR 2 0.010 ** Control Control Control 1904 0.0083 Note: Significance of the coefficients at 10 %, 5 %, and 1 % are indicated as * * *, and * * *, respectively. 7 Nonresource-based cities mainly include 163 prefecture-level cities including Beijing, Chengdu, and Shanghai; resource-based cities mainly include 109 prefecturelevel cities, such as Ordos and Panzhihua. W. Qiu, Y. Bian, S. Ren et al. Journal of Innovation & Knowledge 9 (2024) 100470 8