The impact of urban digital platforms on entrepreneurial activity: Evidence from China

Hu, Guosheng et al. Article The impact of urban digital platforms on entrepreneurial activity: Evidence from China Journal of Innovation & Knowledge (JIK) Provided in Cooperation with: Elsevier Suggested Citation: Hu, Guosheng et al. (2024) : The impact of urban digital platforms on entrepreneurial activity: Evidence from China, Journal of Innovation & Knowledge (JIK), ISSN 2444-569X, Elsevier, Amsterdam, Vol. 9, Iss. 1, pp. 1-15, https://doi.org/10.1016/j.jik.2024.100468 This Version is available at: https://hdl.handle.net/10419/327374 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/ The impact of urban digital platforms on entrepreneurial activity: Evidence from China Guosheng Hu a,b,1 ,SiHe c,1 , Xiaoqi Dong d,1 , Chengming Li c,1 , Zeyu Wang e,1 , Zengwen Wang f,1, *, Abbas Mardani g a School of Economics and Business Administration, Chongqing University, Chongqing 400030, China b School of Economics, South-Central Minzu University, Wuhan, Hubei 430074, China c School of Economics, Minzu University of China, Beijing 100081, China d Department of Applied Mathematics and Statistics, Stony Brook University, Stony Brook, NY 11794-3600, United States e School of Public Administration, Guangzhou University, Guangzhou 510006, China f Researching Center of Social Security, Wuhan University, Wuhan, Hubei 430072, China g Business School, Worcester Polytechnic Institute, Worcester, MA 01609-2280, United States ARTICLE INFO Article History: Received 16 January 2023 Accepted 19 February 2024 Available online 26 February 2024 ABSTRACT The development of urban digital platforms has changed its entrepreneurial environment and affected regional innovation vitality. We calculated the development index of digital platforms of 294 prefecture-level cities in China between 2013 and 2020 using the principal component analysis method. We used the microdata of enterprise registration information to describe the urban entrepreneurial activity. Digital platforms have promoted urban entrepreneurial activity significantly. Moreover, alleviating labour market distortions, optimizing the urban financial environment, and improving technological innovation are virtual channels for digital platforms to increase urban entrepreneurial activity. Furthermore, digital platforms play a more significant role in promoting urban entrepreneurial activity in the eastern region and cities with better industrial structures. This impact has a nonlinear increasing “marginal effect”in which the faster the development of digital platforms, the more significant the promoting effect on urban entrepreneurial activity. © 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: Digital platforms Urban entrepreneurial activity Labour market distortion Urban financial environment Technological innovation level JEL classification: E20 L26 O30 Introduction Cities are essential to economic expansion, and entrepreneurial activities are an important driving factor for urban economic development, so the entrepreneurial activity of a city reflects its economic vitality. With the wide application of digital technology, urban digital platforms have developed rapidly in recent years. So, will the development of digital platforms affect urban entrepreneurial activity? Existing studies mainly examine the influencing factors of entrepreneurial activity from the macro and micro levels. At the macro level, studies primarily focus on economic development (Asencio et al., 2022), institutions (Chowdhury et al., 2018), and innovation (Gregori & Holzmann, 2020); at the micro level, researchers focused on individual entrepreneurs (Dougherty et al., 2019) on entrepreneurial activity. Digital platforms are not only an essential component of the digital economy, but they also contribute significantly to economic growth. Digital platforms combine resources that enable external producers and consumers to create value and interact. Such digital platforms typically have three core characteristics: mediating effects, demand-side driving, and digital technology empowerment (Hu, Qi, et al., 2023;Song, 2019;Wang, Zhang, et al., 2023). However, the continuous expansion of digital platforms will also bring some disadvantages. For example, Srnicek (2017) believes that digital platforms will become a tool for capital to intercept profits by acquiring massive amounts of data and controlling and monitoring the existing rules, thus not conducive to the country’s development. Schor (2017) also emphasized that the platform economy’s development would weaken workers’bargaining power and further aggravate the income gap of the bottom 80 %. Thus, what role do digital platforms play in the advancement of society? As a solution to this issue, we compiled the body of research and discovered that the majority of the papers addressed the topic of digital platform governance (Chen et al., 2020; Gawer, 2022) and their ecosystem (Hein et al., 2020;Karhu et al., * Corresponding author. E-mail address: [email protected] (Z. Wang). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.jik.2024.100468 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) 100468 Journal of Innovation &Knowledge https://www.journals.elsevier.com/journal-of-innovation-and-knowledge 2018). In addition, a few studies focus on the impact of platforms on innovation and entrepreneurship (C. Li et al., 2023a,b; S. Li et al., 2023c;Schreieck et al., 2019;Srinivasan & Venkatraman, 2018;Srnicek, 2017), but these studies were almost exclusively qualitative rather than quantitative. There still needs to be more research on digital platforms and even less on their impact on urban entrepreneurial activity. Data released by the Shanghai Entrepreneurship Evaluation Center shows that the failure rate of start-ups in China is as high as 90 %. However, many start-ups’survival and growth rates based on digital platforms are significantly higher. Through reading and sorting relevant literature, we find that the development of digital platforms may affect urban entrepreneurial activity through the following three channels: one is to alleviate the distortion of the labour market and promote the free flow of human resources; The second is to enhance the efficiency of capital allocation and optimize the urban financial environment; Third, we will raise the level of technological innovation and give full play to the enabling role of technology. Specifically, in alleviating labour market distortions, digital platforms can promote workers’full employment and motivate enterprises’enthusiasm for independent innovation (Hu, Xu, et al., 2023;Laudien & Pesch, 2019). From the perspective of optimizing the financial environment, digital platforms, based on their characteristics, such as economies of scale and scope, can provide entrepreneurs and startups with low-cost funds from pre-loan to post-loan and other links (Li, Li, et al., 2022;Wang et al., 2021). In terms of improving the efficiency of technological innovation, digital platforms can effectively play the role of technological innovation from the three aspects of entrepreneurial costs, entrepreneurial opportunities, and entrepreneurial resources through technological empowerment. Based on this, this paper uses the empirical analysis method and the panel data of 294 prefecture-level cities in China from 2013 to 2020 to verify the impact of the development of digital platforms on urban entrepreneurial activity. First, this paper concludes that developing digital platforms can promote urban entrepreneurial activity through the Bidirectional fixed effects model. Second, considering the existence of endogenous problems such as autocorrelation, this paper uses the lag phase of the digital platforms development index to replace the current value for regression and uses the instrument variables for testing. In addition, after promulgating the “Broadband China”pilot policy, we also see that digital platforms are growing more rapidly. Therefore, we use the differential method to verify further the relationship between digital platforms and urban entrepreneurial activity, and the results remain unchanged. Possible contributions of this paper are as follows: First, from the city level, we research the impact of digital platforms on urban entrepreneurial activity through empirical analysis, enriching the effect of research on digital platforms. The existing literature mainly starts from the related concepts, categories, and applications of digital technology (Li, He, et al., 2022;Wang, Deng, et al., 2023), exploring digital artefacts (Porter & Heppelmann, 2014;Wang, Liang, et al., 2023) and digital infrastructure (Chalmers et al., 2020;Wang et al., 2022) on the process, the practice and outcome of new entrepreneurship (Hu, Xiong, et al., 2023;Nambisan, 2017). A few studies focus on platforms’impact on innovation and entrepreneurship (Schreieck et al., 2019). However, all the above studies only conducted qualitative analysis without quantitative analysis. To this end, we use quantitative studies to complement the existing literature. Second, this paper constructs the digital platform development level index through the principal component analysis. Referring to Szerb et al. (2022) and data available at the city level, we set up a multi-dimensional evaluation system consisting of three first-level indicators: digital technology base, multilateral digital platforms, and digital users. Regarding the selection of specific indicators, this paper takes Szerb et al. (2022) as the benchmark, selects 12 second-level indicators, calculates the development of digital platforms in 294 cities in China from 2013 to 2020, and finally obtains the development index of digital platforms. In addition, according to the structural differences, this paper also constructs corresponding development indexes from the three dimensions of digital technology foundation, multilateral digital platform, and digital users, which further enriches the research content and provides more research ideas. Third, this paper makes some innovations in the influencing factors of urban entrepreneurial activity, especially analyzing its impact on urban entrepreneurial activity from the perspective of digital platforms (Asencio et al., 2022). Specifically, we test the model by building a mediation effect to ease labour market distortions, optimize the financial environment, improve technological innovation in the three channels, and further depict the digital platforms of urban influence mechanism of entrepreneurial activity. In addition, it explores the heterogeneous impact of digital platforms on entrepreneurial activities in different cities. It provides more direct suggestions for developing entrepreneurial activities in the city. The structure is as follows: Part 2 is the theoretical analysis and research hypothesis. Section 3 is empirical design, including model construction, data, and variable definition. Part 4 is empirical results and analysis. The part 5 is the mechanism analysis. Section 6 provides further discussion. Section 7 provides conclusions and policy recommendations. The research framework is shown in Fig. 1. Theoretical analysis and research hypothesis Digital platforms are characterized by economies of scale and scope (Gawer, 2014), which show strong network effects, data, and algorithm advantages, optimize factor allocation, and promote technological innovation. At the same time, the development of a city is also inseparable from the role of human, material, and financial resources. According to the available research, We found that entrepreneurial activity in cities is influenced by environmental factors (Aslan & Kumar, 2021), individual entrepreneurs or individual firms (Zhang, et al., 2023), and Entrepreneurial attitudes (Bosma & Schutjens, 2009). Therefore, we explore the mechanism of the impact of the growth of digital platforms on urban entrepreneurial activities through three channels: easing labour market distortions, optimizing the financial environment, and improving the efficiency of technological innovation. The development of digital platforms can alleviate labour market distortions, thereby increasing urban entrepreneurial activity Digital platforms can not only promote the free flow of human resources but also mobilize the enthusiasm of enterprises to innovate independently, thus easing the distortion of the labour market and promoting the activities of urban entrepreneurship. From the perspective of workers, according to the “pre-market discrimination”theory, Discrimination in the labour market occurs when a worker or a group of workers get paid less than other workers for doing the same job. The four main types of discrimination in labour markets are wage, employment, employee, and customer. Wage discrimination is the most common type of labour discrimination. Specifically, labour employment is based on employment opportunities and expected earnings. When the labour market distortion is severe, the probability of work and the expected wage will be very low; the expected return of labour can hardly make up for the human capital workers invest. On the one hand, it will reduce the enthusiasm of potential employees and the investment of human capital. On the other hand, it will weaken the motivation of those who have already entered the labour market to learn and train the human capital level of future generations. From the perspective of enterprises, labour market G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 2 distortions will provide cost advantages and increase profit margins for enterprises, so enterprises will be more inclined to use undervalued factors to reduce production costs rather than invest in technological research and development, which will reduce the enthusiasm of enterprises to carry out technological innovation. However, relevant studies show that digital platforms are vital in reducing market friction, especially in effectively alleviating the information asymmetry between customers and suppliers (Galperin & Greppi, 2017;Liu et al., 2023;Pallais, 2014;Tadelis, 2016). Specifically, the “ABCD”technology builds an interconnected network among various subjects. On the one hand, digital platforms are conducive to acquiring shared and user information feedback so employees can better understand market information and promote employment equity. In turn, labour market distortions are alleviated, and entrepreneurial activity in the region is boosted (Gaglio et al., 2022). On the other hand, it is helpful to break the market access barriers of enterprises, limit the government’s intervention in the market, promote the free flow of labour factors on cross-regional digital platforms, and improve the innovation motivation of enterprises (Sandberg et al., 2020;Simsek et al., 2019). Based on this, each party can capitalize on their unique advantages, improve collaboration and the division of labour, achieve complementary advantages, achieve an efficient flow of production factors, lessen labour market distortion, and boost the city’s entrepreneurial vibrancy. Therefore, we propose hypothesis 1: Developing digital platforms can alleviate labour market distortions and promote urban entrepreneurial activity. The development of digital platforms can optimize the financial environment and improve urban entrepreneurial activity According to its characteristics, such as economies of scale and scope, the digital platforms can alleviate financing constraints for entrepreneurs and start-ups from the links of pre-loan and post-loan, optimize the financial environment, and thus promote urban entrepreneurial activity. In the pre-loan stage, it is difficult for entrepreneurs to obtain enough start-up capital through traditional financing methods due to asymmetric information in the market, so seeking funds to alleviate financing constraints has become a fundamental goal for entrepreneurs. Digital platforms can broaden the financing channels of entrepreneurs and reduce their financing costs. Specifically, the digital platforms through a digital credit evaluation system, relying on the new digital financial business such as the network, microfinance lending, and digital payments, reduce the threshold of the financial services for entrepreneurs to provide affordable, diversified, comparable, and comprehensive financial services and products, meet the demand of entrepreneurs in the different start-up period of financial services, enhance the willingness of entrepreneurs to make entrepreneurial decisions and carry out entrepreneurial activities. Wu and Mao (2020) also drew a similar research conclusion: entrepreneurial motivation is significantly influenced by socioeconomic conditions and the availability of financial and non-financial support. Barbara Bernhofer and Li (2014) also agree with this view. He believes the entrepreneurial environment includes cultural, economic, and Fig. 1. Research framework. G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 3 political factors, and individuals have different motivations for entrepreneurship in different environmental backgrounds. In the middle stage of lending, digital platforms can use the mature Internet and extensive data analysis technology to change the risk assessment mode and reduce credit assessment costs, improve the allocation efficiency of entrepreneurial credit resources, and alleviate the credit mismatch probability of entrepreneurs. Specifically, for individual entrepreneurs, digital platforms can help them lower the threshold of entrepreneurship (von Briel et al., 2017), promote the equalization of entrepreneurial opportunities, and stimulate individual entrepreneurial vitality. For entrepreneurial enterprises, digital platforms can provide a better atmosphere for enterprise innovation and entrepreneurship, improving the matching efficiency between supply and demand (Autio, 2017), the entrepreneurial opportunities, and the success rate. In the post-loan stage, developing digital platforms can constrain lenders’behaviour, reducing loan-related risks (Sutherland, 2018). Digital platforms also use technologies to promptly detect fund use violations or other potential default risks, use machine learning to deal with threats, ensure the security of borrowed funds, and further alleviate the financing constraints of all parties. Based on this, this paper puts forward hypothesis 2: The expansion of digital platforms helps to optimize the financial environment, reduce the financial barriers faced by business owners, and boost urban entrepreneurship. The development of digital platforms has the potential to spur technical innovation and, in turn, boost urban entrepreneurship With the help of technological advantages, the network effect formed by digital platforms can effectively promote the opportunity interaction and resource integration of all parties and reduce the cost of entrepreneurship, thus providing many entrepreneurial opportunities and promoting urban entrepreneurial activity. From a start-up cost perspective, digital platforms’core value is pooling information and reducing information asymmetry. On the one hand, it can reduce the start-up cost of digital technology (Song, 2019). Specifically, virtual network subjects and digital platforms are thriving in the digital economy era. As an intermediary, digital platforms can form an entrepreneurial network centred on entrepreneurs through network effects, connect all entrepreneurial subjects, and form a digital entrepreneurial team, thereby reducing the cost of resource acquisition (Ferreira et al., 2019;Huang et al., 2017;Matarazzo et al., 2021) and team communication (Nambisan et al., 2017; Siachou et al., 2021), thus improving innovation efficiency (Hung et al., 2023). On the other hand, digital platforms use technologies represented by cloud computing so that enterprises can obtain the required computing, storage, and network resources at a lower cost, lowering the technological threshold. According to the Ali Research Institute, using cloud computing can reduce IT costs by 70 % and improve innovation efficiency by three times. From the perspective of entrepreneurial resources, digital platforms use digital technologies to empower traditional industries, which can promote the upgrading of industrial structure and then release many production factors to improve more available resources for regional entrepreneurial subjects. Furthermore, with the power of efficient digital information technology, digital platforms can connect all parties, improve the penetration and synergy of resources, promote the reorganization and integration of entrepreneurial resources (Sousa-Zomer et al., 2020;Yoo et al., 2012), and then promote entrepreneurship and improve the entrepreneurial vitality of the region. From the angle of entrepreneurial opportunities, digital platforms use digital technology to increase the effectiveness of information and thus create new industries and fields, stimulating the main market body of entrepreneurial enthusiasm and promoting regional entrepreneurial energy (Khan & Tao, 2022). Digital platforms can not only overcome the “resource constraint”and “cost constraint”faced by new enterprises but also provide more entrepreneurial opportunities for entrepreneurs and significantly improve regional entrepreneurial activity. Based on this, this paper puts forward hypothesis 3: The development of digital platforms is conducive to improving technological innovation efficiency and thus promoting urban entrepreneurship’s activeness. Research design Model construction UEAit ¼a0þa1DPDIi;tþa2Controli;tþmiþdtþeit ð1Þ Where istands for the city, tstands for the year, UEA is the entrepreneurial activity of the city, and DPDI represents the developmzent level of the digital platforms. Control is a series of Control variables, including city scale, urban economic development level, urban public expenditure, urban financial environment, the average annual wage of urban residents, and the proportion of urban tertiary industry employees. miand dtare the city and time fixed effect, respectively, and eit is the random error term. It should be noted that the coefficient mainly concerned in this paper is a1. If the coefficient of a1is greater than 0, x positively promotes y, and vice versa. Variable measurement and explanation Measurement of urban entrepreneurial activity In the existing research, the measurement of entrepreneurial activity mainly includes the labour market and ecological research approaches. The basic logic of both methods is to measure the regional entrepreneurial activity by observing the number of new ventures in the period (Reynolds et al., 2005). Therefore, we select the number of newly registered enterprises in a city and take the logarithm to measure city entrepreneurial activity. Measurement of digital platforms At present, many scholars have measured the digital economy and platform economy index, but there needs to be a measurement system for the index of digital platforms. Referring to Szerb et al. (2022) and combining with the data available at the city level, this paper sets up a multi-dimensional evaluation system consisting of three first-level indicators: digital technology foundation, multilateral digital platform, and digital user. In the selection of specific indicators, this paper mainly refers to Szerb et al. (2022), selects 12 second-level indicators, calculates the development of digital platforms in 294 cities in China from 2013 to 2020, and finally obtains the development index of digital platforms, which is recorded as DPDI. a. Digital technology foundation The foundation of digital technology is people’s access to digital infrastructure. With access to the Internet, people can effectively use relevant services, including digital platforms. Therefore, the foundation of digital technology is the basis for developing digital platforms. In this paper, the development of the digital technology foundation is measured by Internet access, Internet resources, and the number of Internet sites, and then measured by the number of broadband access ports per 100 people, the number of CN domain names per person, the number of outlets per 10,000 people and other indicators. b. Multilateral digital platforms Multilateral digital platforms are intermediaries for users and enterprises to connect and knowledge media to realize and promote innovation, entrepreneurship, and value creation (Song, 2019). This paper measures the development of multilateral digital platforms from the perspective of the construction and application of digital G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 4 platforms. Specifically, in constructing digital platforms, this paper adopts the index of the proportion of computer service and software employees to measure. This paper uses e-commerce sales, e-commerce purchase amount, and the number of e-commerce platforms enterprises to measure in applying digital platforms. c. Digital users Digital users are residents who can participate in digital activities and enjoy digital services. The rapid development of digital platforms is also inseparable from the massive growth of digital users, so this paper measures the development of digital users from the two dimensions of digital literacy and digital openness. Precisely, this paper measures the number of educational practitioners in terms of digital literacy. Regarding digital openness, we measure four secondary indicators: Internet users per 100 people, mobile phone users per 100 people, total telecom services per capita, and Internet broadband access users. Table 1 is the measurement system of the digital platform development index in this paper: Next, we first studied the relationship between the above 12 second-level indicators and obtained the correlation coefficient matrix of the 12 indicators. In addition, this paper first standardised the original data to eliminate the possible adverse effects caused by different dimensions. Then, Bartlett’s Test of Sphericity and KMO test shows that the data in this paper are suitable for PCA. Furthermore, this paper uses the method to determine whether the cumulative variance contribution rate of the first K principal components reaches more than 85 %. Finally, as shown in (2), the development index of the digital platforms obtained in this paper is standardized to the interval [0,1] to measure the development level of the digital platforms. DPDI ¼½Si=Max Si ðÞMin Si ðÞ0:4þ0:6ð2Þ Mediating variables a. Financial environment With the promotion and reference of digital technology, digital platforms have nurtured more digital-inclusive financial services. The development degree of urban digital inclusive finance reflects its economic environment to some extent, including the higher the degree of digital finance development, the greater the vitality of urban financial innovation and the better the economic environment. Therefore, this paper uses China’s Peking University Digital Financial Inclusion Index (PKU-DFII) as a proxy variable of the financial environment. If the PKU-DFII is more extensive, finance is more inclusive, the financial environment is better, and vice versa. b. Labor factor market distortion Concerning Yang et al. (2018), we use the production function method to measure the distortion degree of factor markets and adopt the trans-log form of the production function, which is expressed as follows: lnYit ¼λ0þλ1lnLit þλ2lnKit þ1=2λ3ln2Lit þ1=2λ4ln2Kit þλ5lnKitlnLit þeit ð3Þ Where Yis the regional output, measured by the gross regional product and balanced to 2013 constant prices using the GDP deflator, L is the regional labour force, which adopts the number of employed persons in urban units at the end of each region. λ0is a constant term, and λ1,λ2,λ3,λ4and λ5represent the regression coefficients of the explained variables, respectively. eit denotes the random disturbance term. According to Eq. (3), the marginal output of the labour force can be obtained by taking the derivative of L: MPL¼ðλ1þλ3lnL þλ5lnKÞY=Lð4Þ In Eq. (4),MPLis the marginal output of the labour force. According to the definition of factor market distortions, labour factor market distortions can be expressed as the marginal output of labour divided by its price, i.e., DIST_L¼MPL=vð5Þ In Eq. (5),DIST_Lrepresents the degree of labour market distortion. vis the labour price, that is, the wage level. In this paper, the average index of urban unit employment in each region is used to characterize it, and it is calculated as the constant price of 2013 based on the consumer price index of urban residents. c. Technological innovation This paper refers to Riaz et al. (2018) and uses the total number of patent applications to measure the level of technological innovation. The more patent applications a city has, the higher technological innovation. Control variables Refer to Capozza et al. (2018),Goel and Saunoris (2020), and we select city scale, urban economic development level, urban public expenditure, urban financial environment, the average annual wage of urban residents, and the proportion of employees in the tertiary industry as control variables. The above variables are calculated as shown in Table 2. Samples and data sources Since this paper studies the impact of digital platforms on entrepreneurial activity in cities, the research samples focus on the city level. The primary data are from the China City Statistical Yearbook, the State Statistical Bureau, the annual statistical Table 1 Measurement index of digital platforms. One-level indicators Two-level indicators Definition of indicators The foundation of digital technology Internet Access Number of broadband access ports per 100 people Information on Internet Resources CN domain name per person Number of Internet sites The number of outlets per 10,000 people is measured. Digital multilateral platforms Application of digital multilateral platforms E-commerce sales volume E-commerce purchase amount Number of e-commerce platform enterprises The establishment of digital multilateral platforms The percentage of computer service and software employees Digital users Digital Open Internet users per 100 people Mobile phone users per 100 people Total telecom service per capita Internet broadband access users Digital literacy Number of educational practitioners G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 5 bulletins of each city, and the CSMAR database. Besides the mediate variable, the Digital Financial Inclusion Development Index (DFIDI) comes from the Peking University Digital Finance Center. This paper conducts the following processing to make the sample data representative: first, eliminate the cities with apparent missing data; Second, the linear interpolation and smooth index methods are used to fill in some missing data reasonably. After the above processing, this paper takes 2013-2020 as the sample period and determines 294 prefecture-level cities as the whole sample for the empirical study. Descriptive analysis Descriptive statistical analysis is to describe the relevant data of all variables, mainly including the frequency analysis of data, the central trend analysis of data, the degree of data dispersion analysis, the distribution of data, and some basic statistical graphs. A descriptive statistical analysis describes the data distribution in this paper. Table 3 shows the results of the primary descriptive analysis in this paper. The mean value of the urban entrepreneurial activity is 10.529, the minimum value is 7.411, and the maximum is 13.767, which indicates that the entrepreneurial activity level of different cities has apparent differences, which is consistent with the conclusion of existing literature (Song & Winkler, 2014). The average value of the development index of digital platforms is 0.652, and the development of digital platforms has apparent differences. Individual differences in the other control variables are also evident, meaning that the sample is well-represented. Empirical analysis Baseline regression In Table 4, columns (1) and (2) show the results of mixed regression (OLS), and columns (3) and (4) show the results of fixed effects regression (FE). The regression coefficients of these four columns are significantly positive, indicating that digital platforms can promote urban entrepreneurial activity. In control variables, the regression coefficients of the city’s economic development level (GDP), size (SIZE), and average annual wage of residents (AWAGE) are all significantly positive, indicating that the faster the city’s economic growth, the more population, the more average yearly salary of residents, and the higher the urban entrepreneurial activity. Table 2 Definition of main variables. Variables name Variables definition Explained variable Urban entrepreneurial activity (UEA) Ln (Number of newly registered enterprises in the city in that year +1) Explanatory variables Digital Platform Development Index (DPDI) See above Digital Infrastructure Technology Index (DITI) See above Digital Multilateral Platforms Index (DMPI) See above Digital Users Index (DUI) See above Mediating variables Index of factor market distortions (DIST_L) The production function method is used to measure the distortion degree of the factor market. Digital Financial Inclusion Development Index (DFIDI) Peking University Digital Financial Inclusion Development Index Technology Innovation Index (TII) Total number of patent applications Control variables Urban size (SIZE) Ln (At the end of the population of ten thousand people +1) The level of urban economic development (GDP) Ln (Annual price of GDP billion yuan +1) Urban government Expenses (GS) Revenue and expenditure of local general public budget Citywide/ (10000* Annual price of regional GDP billion yuan) Urban Financial Environment (FIN) Financial institutions at the end of the year RMB - financial institutions at the end of the year RMB loans)/ (10000* Annual price of GDP billion yuan) Average wage of urban residents (AWAGE) Ln (The total wage of employees is 10,000 yuan/the average number of employees is 10,000 yuan +1) The proportion of employees in the tertiary industry in the city (TIND) Persons employed in the tertiary sector/all persons employed Table 3 Descriptive analysis. Variable Obs Mean Std. Dev. Min Max UEA 2360 10.529 0.917 7.411 13.767 DPDI 2360 0.652 0.046 0.600 1.000 GDP 2360 7.397 1.013 0.000 16.312 SIZE 2360 5.851 0.753 2.996 8.140 GS 2360 0.080 0.029 0.000 0.262 FIN 2360 1.545 1.246 0.377 38.690 AWAGE 2360 11.023 0.450 2.090 16.077 TIND 2360 0.554 0.146 0.166 1.059 DIST_L 2360 0.724 3.569 -8.796 63.814 Table 4 Baseline regression results. OLS FE (1) (2) (3) (4) DPDI 13.71*** 1.955*** 2.576*** 1.704*** (0.304) (0.249) (0.414) (0.417) GDP 0.500*** 0.0469*** (0.0148) (0.0149) SIZE 0.389*** 0.544*** (0.0164) (0.0666) GS 0.938*** -0.155 (0.295) (0.323) FIN 0.0423*** -0.00358 (0.00672) (0.00427) AWAGE 0.247*** 0.0553*** (0.0198) (0.0138) TIND 0.387*** -0.0425 (0.0618) (0.0978) _cons 1.591*** 0.197 8.849*** 5.318*** City fixed effect Time fixed effect (0.198) No No (0.219) No No (0.270) Yes Yes (0.450) Yes Yes N 2360 2360 2360 2360 R 2 0.464 0.825 0.957 0.959 Note: *** ** , and * indicate significance at the 1 %, 5 %, and 10 % statistical levels, respectively. G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 6 Nonlinear effect analysis The development of digital platforms is closely related to the Internet. In other words, the development of digital platforms will also exist in the “Metcalfe Law”of the Internet. The network value equals the square of the number of nodes, and the marginal effect of network spillover increases. The impact of digital platforms on urban entrepreneurial activity may also have nonlinear characteristics. As digital platforms evolve, the marginal cost of interdepartmental interaction continues to decrease, and the benefits to participants will increase geometrically. This effect will become more and more evident as digital platforms evolve. Therefore, this paper takes the development index of digital platforms as the threshold variable to carry out threshold effect regression (Fig. 2). Before estimating the threshold model, this paper conducts the existence test of panel thresholds based on Hansen (1999) ’s method. After repeated sampling by the bootstrap method 1000 times, the results show that the threshold variable of the digital platforms development index has passed the double threshold significantly but failed the triple threshold test. It can be seen from Table 5 that with the development of digital platforms, the coefficient of urban entrepreneurial activity will increase correspondingly. Endogeneity analysis Control for fixed effects To avoid the endogeneity problems, such as causality between the development of digital platforms and urban entrepreneurial activity, we set the provincial fixed effect and the interaction effect between provinces and years. In Table 6, column (1) is the result after the province effect is fixed based on the baseline regression, and column (2) is the result after the interaction term of province and year is added based on column (1). However, the results of columns (1) and (2) are consistent with the basic regression. System GMM method A systematic GMM approach examines the baseline results to reduce potential endogeneity problems. Since urban entrepreneurial activity may persist over time, the system GMM can dynamically model urban entrepreneurial activity. Considering that the urban entrepreneurial activity of the current period may be affected by the entrepreneurial activity of the previous period, this paper introduces the lagged one-stage L.UEA of the explained variable. Then, the baseline results were tested using the system GMM estimation method. As shown in Table 7, the DPDI is significantly positive, consistent with the regression results above. Fig. 2. Regression results of the threshold effect. Table 5 Regression results of the threshold effect. The threshold model (1) GDP 0.122*** (0.0179) SIZE 0.398*** (0.0810) GS -1.074*** (0.391) FIN 0.0119** (0.00517) AWAGE 0.225*** (0.0155) TIND 1.129*** (0.0994) 0.cat#c.DPDI 6.449*** (0.485) 1. cat#c.DPDI 6.752*** (0.480) 2. cat#c.DPDI 6.976*** (0.468) _cons -0.132 City fixed effect Time fixed effect (0.491) Yes Yes N 2360 R 2 0.599 Note: *** ** , and * indicate significance at the 1 %, 5 %, and 10 % statistical levels, respectively. G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 7 Instrumental variable method The development of digital platforms will affect the entrepreneurial activity of cities; likewise, the development of cities will also promote the development of digital platforms. Therefore, digital platforms and urban entrepreneurial activity may have a reciprocal cause-effect relationship. To solve the potential endogeneity problem, this paper adopts the following two methods to avoid the influence of endogeneity. (1) The regression estimation is re-performed by using the digital platforms’lag period to replace the current value. In column (1) of Table 8, the coefficient of DPDI is still significantly positive. (2) Referring to relevant literature, this paper adopts the historical data of posts and telecommunications 1984 as the instrumental variable. The development of telecom technology facilities in history will affect the subsequent application of the Internet and other technologies and then affect the development of digital platforms. This paper introduces the time-changing variable of the number of Internet users in China in the previous year and multiplicates it with the historical data of posts and telecommunications of provinces in 1984 to construct instrumental panel variables and conduct regression analysis. In column (2) of Table 8, the regression coefficients of instrumental variables are still significantly positive, indicating that the results of this paper are still robust. Exogenous shock test − the impact of the pilot policy of “Broadband China” a. Difference-in-differences method We use the “Broadband China”pilot policy shock to analyze the causal relationship between digital platforms and urban entrepreneurial activity. In August 2013, the Ministry of Industry and Information Technology and the National Development and Reform Commission selected 120 cities in three batches as “Broadband China”demonstration sites in 2014, 2015, and 2016. This policy significantly promotes the local broadband user scale, broadband network speed, broadband coverage, and so on, which is conducive to developing digital platforms. Therefore, this paper uses the difference-in-differences model to test whether the pilot policy of “Broadband China”promotes the development of urban entrepreneurial activity and the mechanism behind it. Specifically, we constructed the following difference-in-differences model: UEAit ¼Constant þaPostit Treatit þgControlit þmtþdi þe:Model ð6Þ Where the coefficient of Postit Treatit reflects the treatment effect of this policy on urban entrepreneurial activity, subscript I represents city individuals, subscript t represents time, and UEA represents city entrepreneurial activity. Treatment is a dummy variable, 1 Table 6 Fixed effect regression. Control for fixed effects (1) (2) DPDI 1.951*** 1.951*** (0.422) (0.423) GDP 0.0455*** 0.0452*** (0.0150) (0.0151) SIZE 0.334*** 0.334*** (0.0983) (0.0983) FIN -0.00421 -0.00412 (0.00428) (0.00432) GS -0.197 -0.188 (0.326) (0.331) AWAGE 0.0633*** 0.0632*** (0.0160) (0.0161) TIND -0.00956 -0.00756 (0.0998) (0.101) dum 2.16e-09 (1.35e-08) _cons 6.294*** 4.620 City fixed effect Time fixed effect Province fixed effect (0.601) Yes Yes Yes (10.43) Yes Yes Yes N 2331 2331 R 2 0.959 0.959 Note: *** ** , and * indicate significance at the 1 %, 5 %, and 10 % statistical levels, respectively. Table 7 System GMM regression results. GMM (1) L.UEA 0.817*** (24.31) DPDI 0.418 (1.16) SIZE 0.107*** (5.07) GDP 0.066 (1.50) GS 0.375 (1.34) FIN 0.003 (0.56) AWAGE 0.058*** (3.66) TIND -0.051 City fixed effect Time fixed effect (-0.61) Yes Yes Observations 2065 Number of code 295 Hansen 0 ar1 0 ar2 0.514 Note: ***** , and * indicate significance at the 1 %, 5 %, and 10 % statistical levels, respectively. Table 8 Instrumental variable method. IV1 IV2 (1) (2) L.DPDI 1.836 *** (0.474) SIZE 0.586 *** 0.303 *** (0.0668) (0.0976) GDP 0.0391 *** 0.146 *** (0.0146) (0.0321) GS -0.0375 0.326 (0.346) (0.393) FIN -0.00275 -0.00696 (0.00420) (0.00780) AWAGE 0.0566 *** -0.0173 (0.0154) (0.0304) TIND 0.0227 -0.111 IV (0.102) (0.123) 0.184 ** (0.0745) _cons 5.054 *** 5.639 *** City fixed effect Time fixed effect (0.507) Yes Yes (1.108) Yes Yes N 2065 1983 R 2 0.963 0.965 Note: ***, **, and * indicate significance at the 1 %, 5 %, and 10 % statistical levels, respectively. G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 8 Riaz, M. F., Cherkas, N., & Leit~ ao, J. (2018). Corruption and innovation: Mixed evidences on bidirectional causality. Journal of Applied Economic Sciences, 13(2), 56. Sandberg, J., Holmstr€ om, J., & Lyytinen, K. J. (2020). Digitization and phase transitions in platform organizing logics: Evidence from the process automation industry. Management Information Systems Quarterly, 44(1), 129–153. doi:10.25300/misq/2020/ 14520. Schor, J. B. (2017). Does the sharing economy increase inequality within the eighty percent?: Findings from a qualitative study of platform providers. Cambridge Journal of Regions, Economy and Society, 10(2), 263–279. doi:10.1093/cjres/rsw047. Schreieck, M., Wiesche, M., Kude, T., & Krcmar, H. (2019). Shifting to the cloud - How SAP’s partners cope with the change. In Proceedings of the Hawaii international conference on system sciences. Siachou, E., Vrontis, D., & Trichina, E. (2021). The moderating role of absorptive capacity and strategic interdependence. Journal of Business Research, 124, 408–421. doi:10.1016/j.jbusres.2020.11.011. Simsek, Z., Vaara, E., Paruchuri, S., Nadkarni, S., & Shaw, J. D. (2019). New ways of seeing big data. Academy of Management Journal, 62(4), 971–978. doi:10.5465/ amj.2019.4004. Song, A. K. (2019). The digital entrepreneurial ecosystem—a critique and reconfiguration. Small Business Economics, 53(3), 569–590. doi:10.1007/s11187-019-00232-y. Song, L., & Winkler, C. (2014). China’s trans-regional entrepreneurship. Journal of Entrepreneurship in Emerging Economies, 6(3), 202–222. doi:10.1108/JEEE-06-20140020. Sousa-Zomer, T. T., Neely, A., & Martinez, V. (2020). Digital transforming capability and performance: A microfoundational perspective. International Journal of Operations & Production Management, 40(7/8), 1095–1128. doi:10.1108/IJOPM-06-2019-0444. Srinivasan, A., & Venkatraman, N. (2018). Entrepreneurship in digital platforms: A network-centric view. Strategic Entrepreneurship Journal, 12(1), 54–71. doi:10.1002/ sej.1272. Srnicek, N. (2017). Platform capitalism. John Wiley & Sons. Sutherland, A. (2018). Does credit reporting lead to a decline in relationship lending? Evidence from information sharing technology. Journal of Accounting and Economics, 66(1), 123–141. doi:10.1016/j.jacceco.2018.03.002. Szerb, L., Komlosi, E. S., Acs, Z. J., Lafuente, E., & Song, A. K. (2022). The digital platform economy index 2020. Springer. Tadelis, S. (2016). Reputation and feedback systems in online platform markets. Annual Review of Economics, 8(1), 321–340. doi:10.1146/annurev-economics-080315015325. von Briel, F., Davidsson, P., & Recker, J. (2017). Digital technologies as external enablers of new venture creation in the IT hardware sector. Entrepreneurship Theory and Practice, 42(1), 47–69. doi:10.1177/1042258717732779. Wang, Y., Hu, H., Dai, W., & Burns, K. (2021). Evaluation of industrial green development and industrial green competitiveness: Evidence from Chinese urban agglomerations. Ecological Indicators, 124, 107371. doi:10.1016/j.ecolind.2021.107371. Wang, Z., Deng, Y., Zhou, S., & Wu, Z. (2023). Achieving sustainable development goal 9: A study of enterprise resource optimization based on artificial intelligence algorithms. Resources Policy, 80, 103212. doi:10.1016/j.resourpol.2022.103212. Wang, Z., Liang, F., Li, C., Xiong, W., Chen, Y., & Xie, F. (2023). Does China’s lowcarbon city pilot policy promote green development? Evidence from the digital industry. Journal of Innovation & Knowledge, 8,(2) 100339. doi:10.1016/j. jik.2023.100339. Wang, Z., Lu, J., Li, M., Yang, S., Wang, Y., & Cheng, X. (2022). Edge computing and blockchain in enterprise performance and venture capital management. Computational Intelligence and Neuroscience. doi:10.1155/2022/2914936 20222914936. Wang, Z., Zhang, S., Zhao, Y., Chen, C., & Dong, X. (2023). Risk prediction and credibility detection of network public opinion using blockchain technology. Technological Forecasting and Social Change, 187, 122177. doi:10.1016/j.techfore.2022.122177. Wu, F., & Mao, C. (2020). Business environment and entrepreneurial motivations of urban students. [Original Research] Frontiers in Psychology, 11. doi:10.3389/ fpsyg.2020.01483. Yang, M., Yang, F., & Sun, C. (2018). Factor market distortion correction, resource reallocation and potential productivity gains: An empirical study on China’s heavy industry sector. Energy Economics, 69, 270–279. doi:10.1016/j.eneco.2017.11.021. Yoo, Y., Boland, R. J., Lyytinen, K., & Majchrzak, A. (2012). Organizing for innovation in the digitized world. Organization Science, 23(5), 1398–1408. doi:10.1287/ orsc.1120.0771. G. Hu, S. He, X. Dong et al. Journal of Innovation & Knowledge 9 (2024) 100468 15