The Influence of Zombie Enterprises on Quantity and Quality of Entrepreneurship: Does Banking Competition Matter

Abstract

Entrepreneurship plays a pivotal role in driving economic development and social progress. In contrast to existing literature that primarily focuses on the impact of zombie enterprises on innovation, resource distortion, and environmental pollution, this study aims to investigate the influence of zombie enterprises on entrepreneurship by utilizing data from prefecture cities in China. Additionally, we explore how banking competition moderates this relationship. Employing the system GMM estimator, our findings reveal that zombie enterprises significantly hampers urban entrepreneurship quantity and diminishes entrepreneurship quality. Furthermore, the inhibitory effect of zombie enterprises on entrepreneurship quality is greater than that of entrepreneurship quantity. In addition, we observe that state-owned zombie enterprises have a more pronounced negative impact compared to private zombie enterprises. Notably, banking competition exhibits a nonlinear moderating effect on the association between zombie firms and entrepreneurship; joint-stock banks and city commercial banks play prominent roles in shaping this effect. Threshold model regression analysis confirms these nonlinear effects. Overall, our conclusions offer important policy implications for accelerating the disposal of zombie firms as well as enhancing entrepreneurship quantity /quality to promote high-quality urban development.

Plain language summary

Zombie firms, banking competition, and urban entrepreneurship

Entrepreneurship plays an important role in economic development and social progress. Different from the existing literature that concerns the impact of zombie enterprises on innovation, resource distortion, and environmental pollution, this paper attempts to study the effect of zombie enterprises on entrepreneurial behavior from the dimensions of quantity and quality. Based on the data from prefecture cities in China, this paper examines the impact of zombie enterprises on entrepreneurial activity and quality and focuses on the moderating effect of banking competition on this relation. The study finds that the degree of zombification has a significant inhibitory effect on urban entrepreneurial activity and entrepreneurial quality, the inhibitory effect of zombification on entrepreneurial quality is stronger than that on entrepreneurial activity, and the inhibitory effect of state-owned zombie enterprises on entrepreneurial behavior is greater than that of private zombie enterprises. Banking competition has a nonlinear moderating role in the inhibitory effect of zombie enterprises on entrepreneurship, and this moderating effect is mainly produced by joint-stock banks and city commercial banks. The threshold model regression further confirms the existence of nonlinear moderating effects. The conclusions of this paper provide policy implications for accelerating the disposal of zombie enterprises, enhancing entrepreneurial activity and quality, and promoting high-quality urban development.

Keywords

zombie enterprises entrepreneurship quantity entrepreneurship quality banking competition

Introduction

Entrepreneurship can boost employment and economic growth(Wennekers & Thurik, 1999; K. Zhang et al., 2018). In recent years, driven by governments at all levels, mass entrepreneurship in China has been extensively promoted with remarkable significance for the transformation of old and new drivers of economic growth, upgrading industrial structure, promoting employment, and enhancing people’s well-being. In the new era, conducting further research on the determinants of entrepreneurship and implementing targeted measures is conducive to enhancing entrepreneurial activity and quality, thereby facilitating high-quality economic development.

Concurrently, the phenomenon of zombie enterprises has garnered considerable attention in government and academics due to supply-side reforms. The term “zombie enterprises” refers to enterprises that exhibit low profitability and would typically be eliminated in market competition while these enterprises manage to survive by accessing a significant amount of credit resources at interest rates lower than the optimal market rate, essentially relying on external financial support (Hoshi & Kim, 2013). This type of enterprise holds vast social resources, but the efficiency is very low, which distorts the allocation of capital and destroys the market competition mechanism (Gouveia & Osterhold, 2018; Geng et al., 2021; Tan et al., 2016). Zombie enterprises, a persistent issue, continue to pose an inescapable reality in China’s current stage of development, constituting a significant impediment to the country’s economic progress. Entrepreneurship necessitates adequate financial support and a favorable business environment; hence, the prevalence of zombie enterprises within regions may exert inhibitory effects on entrepreneurial activities. However, it is worth noting that limited scholarly literature exists regarding the influence of zombie enterprises on entrepreneurship. The persistent presence of a substantial number of zombie enterprises necessitates the exploration of effective measures to mitigate their adverse impact, which holds significant implications for fostering urban entrepreneurship. The intensified competition in the banking sector is considered an effective mechanism for enhancing banking efficiency and reducing enterprise financing costs (Barth et al., 2003). Under competitive pressures, banks are apt to prioritize credit allocation toward high-quality and efficient enterprises, thereby rectifying resource misallocation. Banking competition enhances the market’s fundamental role in resource allocation and can alleviate financing constraints faced by small and medium-sized enterprises and private enterprises (X. Zhang et al., 2020). Consequently, banks are inclined to reduce credit provision to zombie firms while increasing credit rationing for efficient firms under competitive pressures. This optimization of credit resource allocation simultaneously amplifies survival risks for zombie enterprises. Therefore, banking competition may weaken the inhibitory effect of zombie enterprises on entrepreneurship; however, limited research is concerned about this aspect. Furthermore, the existing literature presents divergent views regarding the influence of banking competition on zombie firms, including negative perspectives (Chen et al., 2020; Y. Shen et al., 2023) alongside positive conclusions (X. Zhang & Huang, 2022).

The paper aims to address the following inquiries: (1) What is the impact of zombie enterprises on urban entrepreneurship? (2) How does banking competition moderate the relationship between zombie firms and urban entrepreneurship? (3) What are the distinct moderating roles of different types of banks? (4) If the moderating effect of banking competition is nonlinear, What are the thresholds for different regimes? So, this study employs a system GMM and a panel threshold model to examine how zombie firms influence both the quantity and quality of entrepreneurship at the prefecture level in China, with a particular focus on exploring how banking competition moderates these effects. The marginal contribution of this paper is mainly reflected in the following aspects: First of all, in terms of topic selection, the existing literature primarily focuses on examining the impact of zombie enterprises on innovation, industrial upgrading, and environmental pollution. However, we adopt a novel approach by investigating the adverse effects of zombie enterprises on high-quality economic development from an entrepreneurial perspective. This study expands the research scope concerning the negative economic influence of zombie enterprises. Secondly, the existing literature on entrepreneurship pays more attention to the “quantity” of entrepreneurship, namely the scale or activity of entrepreneurship, but pays less attention to the “quality” of entrepreneurship. This paper analyzes the impact of zombie enterprises on both entrepreneurial quantity and quality to provide a more comprehensive and in-depth analysis. Thirdly, concerning data collection platform design and measurement of entrepreneurship quantity, this paper utilizes the “Qi cha cha” as the data collection platform to collect the registration data of enterprises. Large-scale micro-enterprise data offer greater detail and accuracy in describing entrepreneurship quantity. Finally, building upon our examination of the relationship between zombie enterprises and entrepreneurship, this paper introduces banking competition as a moderating factor and employs nonlinear analysis methods like threshold panel models to investigate its nonlinear moderating effect of banking competition. We find that when the level of banking competition is low, it exacerbates the inhibitory effect of zombie firms on entrepreneurial behavior; However, when the degree of banking competition is high, it contributes positively toward mitigating the inhibitory effect caused by zombie enterprises. This paper can enrich existing literature about the relationship between banking competition and zombie firms.

The structure of the remainder is as follows: the second section is the literature review and theoretical hypothesis, which puts forward three theoretical hypotheses. The third section is the research design, which mainly introduces the model, variables, and data sources. The fourth section is the results, which report the regression results of the benchmark model, the robustness test results, the moderating effect, and the panel threshold regression results. The sixth section is the discussion and policy enlightenment.

Literature Review and Theoretical Hypothesis

Literature Review

The Detrimental Impact of Zombie Enterprises

The malicious motivation of banks to cover up the losses of non-performing loans is considered to be an important reason for the formation of “zombie enterprises” (Hoshi & Kashyap, 2004). Another reason is the government’s deregulation and support (Chang et al., 2021; Fang et al., 2018; Jaskowski, 2015; Nie et al., 2016). The existence of zombie enterprises poses a significant detriment to the functioning of the economy, which causes market congestion and destroys spontaneous “creative destruction” (Caballero et al., 2008). Zombie firms are thought to have caused productivity stagnation and hindered Japan’s economic recovery(Ahearne & Shinada, 2005; Griffin & Odaki, 2009; Kwon et al., 2015). zombie capital impedes the entry and exit of enterprises, (Carreira et al., 2022; Gouveia & Osterhold, 2018; McGowan et al., 2017). In recent years, more and more scholars have paid attention to the impact of zombie enterprises on China’s economy. Zombie firms have been shown to have a crowding-out effect on investment in non-zombie firms(Tan et al., 2016), causing and exacerbating China’s overcapacity(G. Shen & Chen, 2017), reducing patent applications and total factor productivity of normal enterprises in the same industry (Qiao et al., 2022), affecting the commercial credit of upstream and downstream enterprises through the supply chain relationship(Y. Dai et al., 2021), crowding out the pollution investment of normal enterprises and intensified sewage discharge (Chao et al., 2022; Wu et al., 2023). To sum up, although there are a lot of studies on the impact of zombie enterprises, there is little literature on the influence of zombie enterprises on entrepreneurship.

Factors Affecting Entrepreneurship

The research primarily centers on three key dimensions when examining the determinants of entrepreneurship: individual characteristics, resource base, and external environment. Cooper pointed out in his analysis of high-tech entrepreneurship that three main factors affect the establishment of high-tech enterprises: first, entrepreneur characteristics, including various background characteristics that make individuals tend to start businesses; Second, the organizations you have worked in will become the incubation carrier; The third is external environmental factors, including the availability of venture capital and attitudes toward entrepreneurship (Cooper, 1973). The Global Entrepreneurship Monitor’s (GEM) conceptual model of entrepreneurship incorporates individual and ecological factors into the analytical framework (GEM, 2017), and its overall dimension is similar to Cooper’s division.

The external environment of entrepreneurship is commonly called the entrepreneurial ecological environment. Scholars mainly discussed the mechanism of the entrepreneurial ecological environment on entrepreneurial activities from the perspectives of financial support, government policies, government projects, education and training, R&D transfer, business/legal professional infrastructure, market openness, physical infrastructure, and cultural and social norms(Hechavarría & Ingram, 2019; Malecki, 2018).

In the study of the entrepreneurial ecological environment, there has been an increasing focus on examining the impact of the financial environment on entrepreneurship. Scholars have provided evidence to support the existence of financial constraints during the entrepreneurial period, which adversely affect entrepreneurship (Cagetti & De Nardi, 2006). Many potential entrepreneurs face challenges in successfully initiating a business due to difficulties in obtaining start-up capital or maintenance capital during the initial stages. Even in countries like the United States, where minority groups encounter liquidity constraints resulting from limited access to bank loans, entrepreneurs find it more challenging to meet their financing needs for starting new ventures (Prieger, 2023). Therefore, improving credit channels plays a crucial role in fostering entrepreneurship (Qin & Kong, 2022).In addition, Regional financial development also promotes entrepreneurial activities(Dutta & Meierrieks, 2021).

Impact of Banking Competition

Previous studies have found that the increase in banking competition will have two opposite effects on corporate financing constraints. On the one hand, according to the perspective of agency cost, the intensification of banking competition is conducive to improving the efficiency level of the whole industry and reducing financing costs(Barth et al., 2003; Pagano, 1993; Rajan, 1992). On the other hand, the view of information asymmetry holds that the intensification of banking competition will reduce the incentive for banks to establish close ties with enterprises, weaken the supervision function of banks, and thus aggravate corporate financing constraints(Petersen & Rajan, 1995; Zarutskie, 2006 ).

The impact of banking competition on innovation has attracted the attention of a large literature. Due to performance pressure, banking competition will encourage banks to actively collect and mine financial information and operating information of enterprises, which can indeed optimize the allocation of credit resources among enterprises and promote the inflow of bank credit funds into higher technological innovation industries(Amore et al., 2013; Cornaggia et al., 2015), which will not only increase the probability of high-efficiency enterprises entering the innovation sector but also encourage incumbent enterprises to carry out innovation activities(Huang et al., 2021). Additionally, empirical evidence has demonstrated that intensified competition in the banking sector fosters investment in innovation among privately owned enterprises, small-scale businesses, and technology-driven firms (Hou & Lu, 2022).

Limited research has been conducted on the impact of banking competition on zombie firms, and existing literature presents significant variations. Chen et al. (2020) and Y. Shen et al. (2023) have documented that banking competition facilitates the proliferation of zombie enterprises and amplifies their adverse externalities. However, X. Zhang and Huang (2022) found that banking competition is conducive to inhibiting the formation of zombie enterprises, which can reduce the negative impact of zombie enterprises to a certain extent.

Theoretical Hypothesis

The influence mechanism of zombie enterprises on entrepreneurship is reflected in the following aspects, and Figure 1 depicts our theoretical framework.

Figure 1.

Theoretical hypothesis framework.

First of all, zombie enterprises enhance the financing constraints of non-zombie enterprises, which damages entrepreneurship. Tan et al. (2016) argued that the government’s distorted investment behavior and favoritism of state-owned enterprises were the main reasons for the formation of zombie enterprises, which crowded out the investment of normal enterprises. There is an obvious misallocation of financial resources between zombie enterprises and non-zombie enterprises. Zombie enterprises have lower capital-output rates and labor productivity, lower profit rates, and return on assets, but they get a lot of financial resources at a lower cost (Qiao et al., 2022). However, some non-zombie enterprises with higher capital output and labor productivity have difficulty obtaining credit funds. However, credit resources have more important significance for entrepreneurial enterprises. Kerr and Nanda (2009) claimed that financing constraints had a prominent inhibitory effect on entrepreneurial behavior in the United States, and bank deregulation promoted entrepreneurial behavior by easing financing constraints. Gao and Jiang (2021) documented that easing financing constraints boost residents’ entrepreneurial choices and scale. Therefore, by crowding out credit resources, zombie enterprises enhance the financing constraints of start-up enterprises, thus weakening entrepreneurial behavior. The term “credit crowding effect” of zombie enterprises is used in this paper to refer to this mechanism channel.

Secondly, the presence of zombie enterprises hampers the innovation of non-zombie enterprises and consequently impedes entrepreneurial behavior. Innovation has a market expansion effect. Since innovation produces new products or processes, it may generate new market opportunities and bring higher profits, thus promoting the generation of entrepreneurial behavior. Simultaneously, it exerts a potent market usurpation effect due to the heightened impact of innovative products on consumers or their formidable price competitiveness. Consequently, innovative enterprises may seize market share from competitors and even compel less competitive rivals to exit the market, thereby fostering their advancement. However, zombie enterprises harm non-zombie enterprise innovation(Chen et al., 2020; Qiao et al., 2022; Wang et al., 2018). In addition, Y. Dai et al. (2021) argued that zombie enterprises will affect the commercial credit of upstream and downstream enterprises through the supply chain relationship. Commercial credit, serving as a crucial form of informal financing, can effectively compensate for the limitations of credit financing, thereby holding significant implications for enterprise innovation. Consequently, the presence of zombie enterprises may impede entrepreneurship by suppressing the innovative capabilities of non-zombie enterprises. This study designates this mechanism as the “innovation-inhibiting effect.”

Finally, zombie enterprises distort the optimal allocation of resources, thus inhibiting entrepreneurship. Factor price distortion will cause factor price to deviate from its opportunity cost, increase the cost of entrepreneurship, weaken the entrepreneurial motivation and vitality, and thus reduce the level of entrepreneurship(Ma & Wu, 2017). The distortion of the capital factor market and labor factor market significantly inhibited the development of household entrepreneurial activities. However, zombie enterprises destroy the optimal allocation of factor markets, resulting in distortion of factor markets, and gain the cost advantage of unfair competition ability, which undermines the principle of resource allocation in the market economy that market participants have fair access to factor opportunities(Yu & Wu, 2020). Caballero et al. (2008) found that zombie enterprises destroyed the spontaneous “creative destruction” of the market. Gouveia and Osterhold (2018) posit that zombie firms can disrupt resource allocation through both intensive and extensive margins, leading to factor market distortions that increase the costs of entrepreneurial enterprises and hinder entrepreneurship. This mechanism is referred to as the “factor distortion effect.”

Hypothesis 1: zombie enterprises weaken entrepreneurship.

According to the classic industrial organization theory, monopolistic and competitive market structures have vastly different economic implications. In terms of the banking industry, the monopolistic banking market structure will lead to insufficient loan supply and high loan interest rates, while the competitive banking market structure can reduce financing costs, and increase credit availability(Carlson et al., 2022; Love & Martínez Pería, 2015), and promote credit expansion. The pressure of competition can force banks to improve loan efficiency and promote the collection and screening of corporate information(Schaeck & Cihák, 2012), thus effectively reducing the information asymmetry between banks and enterprises. Under the pressure of competition, for business performance and risk control, banks will pay more attention to the credit of high-quality and efficient enterprises and reverse the phenomenon of resource misallocation, that is, the allocation of credit resources based on enterprise efficiency rather than ownership attributes.

However, when the intensity of banking competition is feeble, local governments may still play a crucial role in bank credit decisions, especially for state-owned commercial banks(Chen et al., 2020). Simultaneously, to seize market share, banks are likely to relax inspection conditions and lower credit standards(Y. Shen et al., 2023), and provide credit support to “bad” enterprises such as zombie enterprises, the credit crowding out effect, innovation inhibition effect, and factor distortion effect of zombie enterprises are thus exacerbated.

With the intensification of competition in the banking market, the number of business outlets of small and medium-sized banks has expanded rapidly, and foreign banks have also entered, breaking the monopoly position of state-owned banks, leading to the contraction of the number of business outlets of state-owned commercial banks and strengthening the competition among various banks (J. Zhang et al., 2017). Moreover, the vertical reform of the banking system and the upward movement of credit approval authority limit the direct influence of local governments on the credit decisions of state-owned banks. Therefore, the intensification of bank competition promotes commercial banks to have more autonomy, and the impact of government on them is gradually weakening. To reduce liquidity loss, banks must enhance the quality management of credit assets, elevate the level of business services, strive for more market share, and selectively extend credit to enterprises with higher operational efficiency. To reduce the occurrence of non-performing loans as much as possible, inefficient zombie enterprises may exit the market because they cannot obtain financial support. It helps play the “creative destruction” effect of the market(He et al., 2019). Moreover, banks are compelled to curtail credit provision to zombie enterprises further to mitigate operational risks while augmenting credit allocation toward efficient enterprises. It will further optimize the allocation of credit resources, thus alleviating the credit crowding out effect, innovation inhibition effect, and factor distortion effect of zombie enterprises.

Hypothesis 2: Banking competition has a nonlinear moderating effect on the inhibitory effect of zombie enterprises on entrepreneurship. When the degree of banking competition is intense, banking competition mitigates the inhibitory effect of zombie firms on entrepreneurship.

From the perspective of banking market structure, diverse categories of commercial banks have significant differences in the objects, capabilities, and efficiency of serving the real economy. Large state-owned commercial banks pay more attention to “hard” information such as corporate collateral value, financial statements, and credit rating when making loan decisions(Berger & Black, 2011). It is more in line with the loan requirements of large banks and easier to obtain loans from large banks. However, small and medium-sized banks such as joint-stock banks and city commercial banks have more advantages in organizational structure and access to information(Berger et al., 2005), relatively less government intervention, shorter information transmission chain, more flexibility in providing loans to small and medium-sized enterprises and new enterprises, and more rapid response to the market (Jayaratne & Wolken, 1999). At the same time, small and medium-sized banks are more likely to establish long-term relationships with small and medium-sized enterprises in adjacent areas and are better able to collect and process enterprise “soft” information and make credit decisions. From the perspective of bank competitiveness, joint-stock banks, and city commercial banks are relatively weak in competitiveness. To take the initiative in the competition, these small and medium-sized banks will extend more loans to promising start-ups. Therefore, the involvement of joint-stock commercial banks and city commercial banks in banking competition can effectively mitigate the credit crowding out effect, innovation inhibition effect, and factor distortion effect associated with zombie enterprises.

Hypothesis 3: joint-stock banks and city commercial banks play a greater role in mitigating the adverse impacts of zombie enterprises.

Research Design

Econometric Methodology

Benchmark Model

To address potential endogeneity between explanatory variablesCand residual errors in the model, this paper utilizes (1) generalized method of moments (GMM) for control, considers possible serial autocorrelation of variables and endogenous correlation between variables, and establishes the following dynamic panel data model:

E n_{ct} = α_{0} + β_{1} E n_{ct - 1} + β_{2} Zo m_{ct} + γ X + μ_{c} + φ_{t} + ε_{ct}

(1)

Where c is the prefecture city, t is the year, and En is the dependent variable, that is the entrepreneurship of each city. This paper considers two variables: entrepreneurship quantity and entrepreneurship quality; Zom is the zombification level of each city. X represents a set of control variables. μ and φ represent city and year effects, respectively, and ε is a random error term.

Moderating Effect Model

To examine the nonlinear moderating effect of banking competition on the relationship between zombie enterprises and entrepreneurship, and to validate hypothesis 2, this study incorporates two interaction terms into the benchmark model: one is the interaction term between the degree of urban zombification and the banking competition (Zom × BC), and the other is the interaction term between the degree of urban zombification and squared of banking competition (Zom × (BC)²).

\begin{matrix} E n_{ct} = α_{0} + β_{1} Zo m_{ct} + β_{2} B C_{ct} + β_{3} Zo m_{ct} \times B C_{ct} \\ + β_{4} Zo m_{ct} \times (B C_{ct})^{2} + γ X + μ_{c} + φ_{t} + ε_{ct} \end{matrix}

(2)

Where BC is urban banking competition As the ordinary least squares method may result in biased estimates for dynamic panel models (1) and (2), Arellano and Bover (1995) and Blundell and Bond (1998) suggest that System GMM( generalized method of moments) can address the issue of weak instrumental variables caused by differential GMM. Therefore, this paper employs a two-stage system GMM to estimate models (1) and (2).

Panel Threshold Regression Model

The interactive term method was employed to examine the nonlinear moderating effect of banking competition; however, this approach did not yield a specific threshold value, resulting in a relatively imprecise conclusion and limited policy implications. So, this paper adopts the panel threshold regression model proposed by Hansen (1999) and takes the urban banking competition degree as the threshold variable to estimate and test the threshold characteristics of banking competition. Assuming there are two threshold values θ₁ and θ₂, the panel threshold model follows:

\begin{matrix} E n_{ct} = α_{0} + β_{1} Zo m_{ct} + β_{2} B C_{ct} + β_{3} Zo m_{ct} \times B C_{ct} \\ \times IBC \leq θ_{1} + β_{4} Zo m_{ct} \times B C_{ct} \times I θ_{1} \leq BC \leq θ_{2} \\ + β_{5} Zo m_{ct} \times B C_{ct} \times IBC > θ_{2} + γ X + μ_{c} + φ_{t} + ε_{ct} \end{matrix}

(3)

In the Equation 3, I (*) is the indicative function, which takes the value of 1 when the conditions in the brackets are met, and 0 when they are not. β₃β₄β₅ represent the moderating effect coefficients of banking competition when BC ≤θ₁, θ₁ < BC ≤θ₂, and BC <span>> θ₂, respectively.

Independent Variables

The independent variable is the degree of urban zombification, which refers to the proportion of zombie enterprises in city C as defined by Shao et al. (2022). To calculate this variable, it is necessary to first identify zombie enterprises. The identification method for zombie enterprises was previously proposed by American and Japanese scholars. One of the most representative methods was introduced by Caballero et al. (2008), who used the presence of bank interest subsidies as a criterion for identifying zombie enterprises and developed the widely accepted CHK model, which served as a strong foundation for subsequent quantitative identification of such entities internationally. Building upon this framework, Fukuda and Nakamura (2011), Nakamura and Fukuda (2013), Imai (2016), Goto and Wilbur (2019) have made revisions to refine the methodology. Drawing on insights from Tan et al. (2016) and Fukuda and Nakamura (2011), we employ the following steps to identify zombie enterprises within industrial sectors.

First, estimate the minimum interest RA_i,t that enterprises need to pay under normal operating conditions:

R A_{i t} = r s_{t - 1} B S_{i t - 1} + (\frac{1}{5} Σ_{j = 1}^{5} r l_{t - j}) B L_{i t - 1}

(4)

BS_i,t and BL_i,t represent short-term (<1 year) bank loans and long-term (>1 year) bank loans, respectively. In the formula, rs_t and rl_t denote the average benchmark lending rates of banks for 1 and 5 years, respectively. Then, we compare the actual interest RB_it paid by the enterprise with the minimum interest RA_it that we assume the enterprise needs to pay at least. We then standardize this difference with the total amount of interest-bearing loans from the beginning of the current period to obtain the calculation formula for interest spread:

G A P_{i t} = (R B_{i t} - R A_{i t}) / (B S_{i t - 1} + B L_{i t - 1})

(5)

As with Caballero et al. (2008), if GAP_i,t<0, then enterprise i get subsidies, and its zombie index is 1; Otherwise, it has a zombie index of zero. Finally, according to Fukuda and Nakamura (2011), we further adjust the corporate zombie index by using Profitability Criteria and Evergreen Loan Criteria. First, companies whose Earnings Before Interest and Tax (EBIT) exceed the assumed risk-free interest payments are not zombie companies, according to the Profitability Criteria. Second, according to the Evergreen Lending Criteria, enterprises that are unprofitable, highly leveraged (leverage ratio above 0.5), and have increased external borrowing are zombie enterprises. Specifically, if an enterprise’s EBIT is lower than the assumed risk-free interest payment of period t, the total external debt of the enterprise exceeds half of its total assets, and the increased liabilities of period t are greater than that of period T-1, it is a zombie enterprise of period t.

Based on the identification of zombie enterprises, referring to the methods of Tan et al. (2016) and Shao et al. (2022), the urban zombification degree is measured by the proportion of zombie enterprise assets in the total assets of each city in the current year.

{H Z o m}_{c t} = \sum_{i = 1} {A s s e t}_{c i t} / \sum_{i = 0, 1} {A s s e t}_{c i t}

(6)

In Equation 6, if i = 1, it indicates that the enterprise is a zombie enterprise; if i=0, it shows that the enterprise is a non-zombie enterprise; Asset_cit represents the total assets of enterprise i in city c in year t; HZom is the urban zombified degree.

Dependent Variables

Entrepreneurship quantity (En1): A commonly used indicator of entrepreneurship is the number of private businesses in a region; however, this data is only available at the provincial level. Building upon the concepts proposed by Bai et al. (2022) and utilizing the “Qi cha cha” database as a platform for data collection, this study gathers micro-data on start-ups with an investigation period of 42 months or less. To mitigate the influence of city size, we assess entrepreneurship quantity within a city based on the number of new enterprises per 1,000 individuals.

Entrepreneurship quality (En2): Referring to the methods of Mao and Lu (2020), the “China Regional Innovation and Entrepreneurship Index” developed by the Enterprise Big Data Research Center of Peking University is used to comprehensively measure the quality of urban entrepreneurship by integrating indicators such as the number, innovation level, and ability to accept venture capital. Similarly, we use the per capita form to overcome the impact of urban size and take the logarithm.

Moderating Variables

The financial license information published on the website of the China Banking and Insurance Regulatory Commission is utilized in this paper to extract the institution code, detailed address, approved establishment date, and exit date of each branch of commercial banks. Subsequently, the number of branches for each commercial bank in every prefecture city is calculated, followed by the construction of the Herfindahl index to reflect banking competition.

HH I_{ct} = \sum_{k = 1}^{K_{ct}} (Branc h_{ckt} / \sum_{k = 1}^{K_{ct}} Branc h_{ckt})^{2}

(7)

Where K_ct is the number of bank types owned by city c in the year t, Branch_ckt is the number of branches of k-type commercial banks in city c in the year t; HHI_ct represents the Herfindahl index of city c in year t, which takes the value between 0 and 1. The index is a negative indicator. To better explain the results, we refer to the practice of Li and Liu (2022) for transformation, and obtain the proxy indicator of banking competition after transformation:

HHI B_{ct} = 1 - HH I_{ct}

(8)

In addition, to ensure the robustness of the results, according to J. Dai et al. (2020) the market share of the top five banks is used to measure the degree of bank competition (CR), that is, the proportion of branches of the top five banks (cr5) including Industrial and Commercial Bank of China, Agricultural Bank of China, Bank of China, China Construction Bank and Bank of Communications to the total number of commercial bank branches in the city. The value range of the CR index is 0 to 1, which satisfies the formula:

C R_{ct} = 1 - cr 5_{ct}

(9)

Control Variables

According to the existing literature and research objectives, this paper controls the following variables:

The level of economic development (lnGDP): the higher the level of economic development, the greater the population agglomeration and economic capacity, which benefits increasing the returns of entrepreneurial activities and thus enhancing the residents’ entrepreneurial willingness. It is measured by GDP per capita.

Foreign direct investment (lnFDI): Foreign direct investment brings necessary capital to the host country and also generates positive spillover effects, including knowledge spillover and technological innovation. Domestic enterprises learn from both the successful and unsuccessful experiences of foreign enterprises, which significantly impact the entrepreneurial behavior of the host country. Following the method of K. Zhang et al. (2018), we measure it using actual amounts of foreign capital utilized in each city while taking a logarithm.

Development level of the digital economy (DE): It promotes mass entrepreneurship as digitalization greatly reduces various barriers to entry, leading to increased entrepreneurship opportunities. Zhao et al. (2020) argued that through influencing market scale, knowledge spillover, factor combination, accelerating information exchange, and idea dissemination; the digital economy cultivates more entrepreneurial opportunities and enriches resources for urban entrepreneurship. Due to data availability limitations in this study, this paper uses the number of Internet users per 100 people to measure the level of digital economy development in a city.

Industrial structure (IndS). The differences in urban industrial structure objectively reflect the differences in resource endowment, input factors, and other conditions at the urban level. This difference may lead to different degrees of difficulty and choice of space for entrepreneurship in different cities, affecting entrepreneurial activity and quality. This paper uses the proportion of urban tertiary industry to measure the industrial structure.

Marketization degree (Market). To a certain extent, the level of marketization impacts the transaction cost and market entry barriers for enterprises, thereby influencing urban entrepreneurship. This paper comprehensively considers the availability of city-level data and adopts the approach proposed by Bai et al. (2022) to utilize the GDP-to-government budget ratio as an approximate measure of marketization. Table 1 presents the descriptive statistics of main variables.

Data Sources

Given that the latest available data on China’s industrial enterprises is from 2015, this study covers the period from 2000 to 2015. The data about zombie enterprises is sourced from the Industrial Enterprise Database of the National Bureau of Statistics, while information on new enterprises is obtained from the “Qi cha cha” database – a pioneering mobile one-stop platform for enterprise credit information queries with over 300 million users in 2021. Entrepreneurship quality data originates from Peking University’s Enterprise Big Data Research Center and their China Regional Innovation and Entrepreneurship Index. Commercial bank branch data are extracted from CBRC’s official website, which has been disclosing financial license information since 1949, including codes, names, addresses, and establishment approval dates of all branches; city locations were derived using address details. Other sources include China City Statistical Yearbooks, provincial Statistical Yearbooks, and the China Economic Network database.

Table 1.

Descriptive Statistics of the Main Variables.

Variables	Observations	Mean	Max	Min	Std. dev
En1	4,352	5.791	3261.328	0.003	26.784
En2	4,352	1.715	2	0	1.438
Zom	4,352	0.227	0.956	0.000	0.142
HHIB	4,352	0.561	0.965	0.207	0.053
CR5	4,352	0.459	0.723	0.604	0.088

Results

Estimation Results of the Benchmark Model

Based on the system GMM method, this paper employs the Sargan test and AR test to address issues of over-identification and autocorrelation in the regression model. Table 2 presents the benchmark model testing the impact of urban zombification degree on entrepreneurship quantity and quality. The results indicate that for entrepreneurship quantity and quality, the coefficient of zombification degree (Zom) is significantly negative, confirming our theoretical hypothesis H1 regarding a significant inhibitory effect of zombie enterprises on entrepreneurship. Moreover, when comparing entrepreneurship quality with quantity, it is observed that the absolute value of the coefficient is higher for quality. It suggests that zombie enterprises have a more pronounced inhibitory effect on entrepreneurship quality due to their impact on core elements such as capital, technology, and the overall entrepreneurial environment. Consequently, these zombie enterprises not only impede new business formation but also potentially hinder foreign investment inflows, patent authorizations, and trademark registrations; thus exerting a more prominent negative influence on entrepreneurship quality.

Table 2.

Estimation Results of the Benchmark Model.

Variables	En1		En2
En(-1)	0.698^*** (9.652)	0.707^*** (5.446)	0.564^*** (8.372)	0.550^*** (6.241)
Zom	−0.364^*** (−6.308)	−0.319^*** (−6.472)	−0.453^*** (−8.064)	−0.487^*** (−7.051)
lnGDP		1.478^***(6.220)		1.502^***(6.531)
lnFDI		1.133^***(5.820)		0.988^***(6.416)
DE		0.695^***(5.184)		0.627^***(6.805)
IndS		0.081(1.065)		0.135(1.308)
Market		0.903^**(2.016)		0.929^***(4.134)
Sargan-P	0.286	0.349	0.457	0.294
AR(1)-P	0.005	0.005	0.035	0.039
AR(2)-P	0.629	0.578	0.741	0.706
Observations	4,352	4,352	4,352	4,352

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

Analysis of Endogeneity

Endogeneity in model estimation is a crucial factor that often undermines the reliability of estimation results, with missing variables and reverse causality being the two most prevalent factors leading to biased outcomes. While system GMM offers better handling of endogeneity issues compared to OLS regression, there may still be endogeneity problems arising from missing variables and reverse causality. Consistent with the concepts proposed by Tan et al. (2016) and Shao et al. (2022), we employ an instrumental variable approach by multiplying the asset proportion of state-owned enterprises during the early sample period (1999) for each city with the asset-liability ratio of national state-owned enterprises in the previous year, thereby capturing the degree of zombification. The instrumental variable design fully satisfies all necessary conditions for instrumental variables. Firstly, according to numerous research findings, the proportion of state-owned enterprises exhibits a close relationship with zombie enterprise prevalence. Secondly, incorporating both the proportion of state-owned enterprises and their asset-liability ratio interaction product enables a more accurate reflection of how state-owned enterprises absorb credit resources while meeting correlation and homogeneity requirements for instrumental variables. The results are presented in Table 3, demonstrating that zombie enterprises significantly impede both quantity and quality improvements in entrepreneurship, which aligns with prior conclusions.

Table 3.

Instrumental Variables Tests.

Variables	En1		En2
Variables	(1)	(2)	(3)	(4)
IV	−0.171^***(−4.965)	−0.204^***(−5.329)	−0.225^***(−6.691)	−0.286^***(−6.271)
Control variables	N	Y	N	Y
City fixed effect	N	Y	N	Y
Year fixed effect	N	Y	N	Y
Observations	4,352	4,352	4,352	4,352
R ²	0.223	0.283	0.205	0.219

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

Robustness Test

Replace Independent Variables

In this section, we employ a robustness test by replacing the independent variables. Firstly, for the identification of zombie enterprises, we substitute the total profit volume with the total business volume as the re-identification standard, resulting in a new indicator of the zombification degree called Zom1. Secondly, we adopt Zom2 as the proportion of employees in zombie enterprises and Zom3 as the proportion of output value from these enterprises. Based on these modifications, benchmark model (1) is employed for further testing, and Table 4 presents the results obtained. The findings demonstrate that after replacing the measurement indicators for assessing urban zombification degree, except for certain variables, there is a significant negative impact of zombification degree on both entrepreneurship quantity and quality. These outcomes indicate the strong robustness of our model.

Table 4.

Robustness Test of Replacing Independent Variables.

Variables	En1			En2
Variables	Zom1	Zom2	Zom3	Zom1	Zom2	Zom3
En(-1)	0.680^***(8.926)	0.641^***(6.307)	0.720^***(7.518)	0.523 (5.367)	0.579^***(6.843)	0.508^***(4.614)
Zom	−0.211^***(−10.148)	−0.368^**(−2.027)	−0.293^***(−6.483)	−0.249 (−1.684)	−0.389^***(−6.035)	−0.337^***(−2.876)
Control variables	Y	Y	Y	Y	Y	Y
Sargan-P	0.529	0.710	0.484	0.565	0.733	0.504
AR(1)-P	0.026	0.010	0.032	0.012	0.006	0.020
AR(2)-P	0.371	0.508	0.517	0.429	0.483	0.590
Observations	4,352	4,352	4,352	4,352	4,352	4,352

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

Eliminate Some Samples

In cities with a higher level of entrepreneurial activity, local governments may implement specific supportive policies for entrepreneurial enterprises to effectively steer social entrepreneurial behaviors, thereby mitigating the adverse impact of zombie enterprises on entrepreneurship. To ensure that the regression results are not influenced by these particular samples, this study retests after excluding cities that rank within the top 10% of the number of startups (Table 5). Overall, the coefficients of core variables in the regression remain consistent with those obtained from the benchmark regression.

Table 5.

Robustness Test of Eliminate Some Samples.

Variables	En1	En2
En(-1)	0.764^***(9.280)	0.601^***(7.632)
Zom	−0.406^***(−4.713)	−0.455^***(−6.324)
Control variables	Y	Y
Sargan-P	0.620	0.588
AR(1)-P	0.014	0.009
AR(2)-P	0.683	0.427
Observations	3,904	3,904

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

Analysis of Heterogeneity

Fang et al. (2018) argue that small and medium-sized private enterprises constitute the main body of zombie enterprises, and this phenomenon is commonly referred to as the “enigma of small and medium-sized private enterprises.” We categorize zombie enterprises into state-owned and private based on their ownership attributes, subsequently quantifying the extent of state-owned and private zombie enterprises. The findings from our report indicate that the degree of state-owned zombie enterprises has a more pronounced inhibitory impact on urban entrepreneurship quantity and quality compared to that of private zombie enterprises. Our conclusion suggests that although small and medium-sized private enterprises form the majority of zombie enterprises, the detrimental effect caused by state-owned zombies on entrepreneurship quantity and quality is more significant than that posed by their counterparts. This discrepancy primarily stems from China’s financial market bias toward State-Owned Enterprises (SOEs), enabling them to access larger-scale and more stable credit support, thereby crowding out funds from other businesses, ultimately exerting direct and indirect negative repercussions on entrepreneurship quantity and quality. In contrast, for small- to medium-sized private enterprises, their zombification mainly results from a contagion effect driven by their lack of credit support—a predicament contributing to a lesser extent when compared with the comprehensive negative impact generated by state-owned zombies—thus accentuating the inhibitory influence imposed by state-owned zombie enterprises on entrepreneurial behavior (Table 6).

Table 6.

Results of Heterogeneity Analysis of Ownership Attributes of Zombie Enterprises.

Variables	En1		En2
Variables	state-owned zombie enterprises	private zombie enterprises	state-owned zombie enterprises	private zombie enterprises
En(-1)	0.730^***(6.632)	0.718^***(7.267)	0.580^***(10.114)	0.527^***(9.456)
Zom	−0.430^***(−7.206)	−0.113^**(−2.145)	−0.461^***(−6.324)	−0.138^*(−1.905)
Control variables	Y	Y	Y	Y
Sargan-P	0.548	0.553	0.792	0.746
AR(1)-P	0.026	0.021	0.008	0.013
AR(2)-P	0.631	0.657	0.597	0.608
Observations	4,352	4,352	4,352	4,352

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

The Moderating Effect Analysis

Based on Model (2) framework for examining moderating effects, we aim to validate Hypothesis 2 by testing the nonlinear moderating effect of banking competition on the inhibitory impact of zombie enterprises on entrepreneurship. The test results are shown in Table 7.

Table 7.

Moderating Effect of Banking Competition.

Variables	En1		En2
Variables	HHIB	CR5	HHIB	CR5
En(-1)	0.734^***(8.580)	0.710^***(7.796)	0.597^***(3.809)	0.566^***(4.891)
Zom	−0.286^***(−5.771)	−0.304^***(−6.287)	−0.372^***(−8.425)	−0.364^***(−7.264)
BC	0.104^***(4.530)	0.087^***(5.647)	0.076^***(5.439)	0.067^***(5.114)
Zom × BC	−0.063^***(−6.553)	−0.047^***(−6.580)	−0.052^***(−4.643)	−0.046^***(−11.534)
Zom × (BC)²	0.083^**(2.159)	0.063^***(8.726)	0.076^***(6.455)	0.050^***(6.743)
Control variables	Y	Y	Y	Y
Sargan-P	0.373	0.301	0.524	0.448
AR(1)-P	0.015	0.002	0.031	0.011
AR(2)-P	0.856	0.736	0.802	0.793
Observations	4,352	4,352	4,352	4,352

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

The coefficient of the first-order interaction term (β₃) is significantly negative, but the coefficient of the quadratic interaction term (β₄) is significantly positive, which shows that in cities with weak banking competition, banking competition intensifies the inhibitory effect of zombie enterprises on entrepreneurship. When the competition in the urban banking industry is more intense, the banking competition can mitigate the inhibitory effect of zombie enterprises on entrepreneurship.

To examine the heterogeneous moderating role of different types of banks on the inhibitory effect of zombie enterprises on entrepreneurship and verify Hypothesis 3, we calculate the contribution index for banking competition among state-owned commercial banks (HHIB1), joint-stock commercial banks (HHIB2) and city commercial banks (HHIB3). The calculation method involves replacing the numerator in Equation 7 with the number of corresponding types of commercial banks. The findings presented in Table 6 indicate that both the one-time interaction term (Zom × BC) and the quadratic interaction term (Zom × (BC)²) coefficients are not statistically significant for state-owned banks. However, significant negative coefficients are observed for first-order interaction terms (Zom × BC) with joint-stock banks and urban commercial banks, while significantly positive coefficients are found for quadratic interaction terms (Zom × (BC)²). It suggests that non-linear moderation by banking competition primarily occurs through joint-stock banks and city commercial banks, whereas state-owned banks do not play a significant role. Therefore, the conclusion provides empirical support for Hypothesis 3. The results also indicate that, in comparison to city commercial banks, joint-stock commercial banks play a more prominent role in enhancing the inhibitory effect of zombie enterprises on entrepreneurship. It is primarily due to the broader business scope, diversified operational methods, increased capital sources, stronger risk diversification capabilities, and more support from the government enjoyed by joint-stock commercial banks. Moreover, they possess the operational flexibility of urban commercial banks and the stable advantages of state-owned commercial banks. Consequently, the credit capital allocation efficiency of joint-stock commercial banks is higher and plays a pivotal role in mitigating the inhibitory effect of zombie enterprises on entrepreneurship (Table 8).

Table 8.

Moderating Effects of Different Types of Banks.

Variables	En1	HHIB2	HHIB3	En2	HHIB2	HHIB3
Variables	HHIB1	HHIB2	HHIB3	HHIB1	HHIB2	HHIB3
En(-1)	0.688^***(12.703)	0.694^***(11.165)	0.702^***(9.378)	0.540^***(8.377)	0.576^***(8.941)	0.553^***(7.915)
Zom	−0.310^***(−8.425)	−0.277^***(−6.325)	−0.291^***(−7.432)	−0.357^***(−6.036)	−0.360^***(−6.544)	−0.349^***(−6.458)
BC	0.094(1.28)	0.088^***(4.714)	0.090^***(5.625)	0.086 (1.492)	0.101^***(6.538)	0.097^***(6.008)
Zom × BC	0.035(0.974)	−0.051^*(−1.985)	−0.041^***(−4.763)	−0.009(−1.375)	−0.054^***(−6.430)	−0.060^***(−5.881)
Zom × (BC)²	0.124(1.066)	0.058^***(6.532)	0.043^***(5.446)	0.067(1.135)	0.049^***(5.680)	0.042^***(8.719)
Control variables	Y	Y	Y	Y	Y	Y
Sargan-P	0.903	0.894	0.912	0.783	0.676	0.650
AR(1)-P	0.005	0.029	0.016	0.001	0.040	0.031
AR(2)-P	0.388	0.524	0.509	0.417	0.583	0.474
Observations	4,352	4,352	4,352	4,352	4,352	4,352

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

Further Examinations Utilizing Panel Threshold Regression Models

According to Model (3), this paper initially examines the presence of a threshold effect in banking competition. Given the similarity in test results for both measures of banking competition, we solely present the findings when HHIB serves as the threshold variable, as depicted in Table 9.

Table 9.

Threshold Effect Test.

threshold Variable	Dependent variable	Number of thresholds	F statistic value	p value
HHIB	En1	Single thresholdDouble thresholdThree thresholds	16.98311.4479.526	.039.126.137
	En2	Single thresholdDouble thresholdThree thresholds	12.08310.6259.430	.052.109.115

The results shown in Table 9 demonstrate a significant single threshold effect on the moderating impact of banking competition for entrepreneurship quantity (En1), as evidenced by an F statistic value of 16.983, surpassing the 5% significance level test. Similarly, for entrepreneurship quality (En2), there exists a single threshold effect with an F value of 12.083, passing the test at a significance level of 5% (Table 10).

Table 10.

Results of Threshold Estimates.

Dependent variable	Number of thresholds	Threshold value	Interval of confidence(α = 95%)
En1En2	Single thresholdSingle threshold	0.9040.922	(0.871, 0.910)(0.908, 0.929)

The results from Table 11 reveal that in terms of entrepreneurship quantity (En1) when the banking competition is less than 0.904, the coefficient of the interaction term reflecting its moderating effect is −0.059, which passes the significance test of 1%. When the banking competition degree exceeds 0.904, the coefficient of the interaction term is 0.033, which is significant at the level of 5%. By entrepreneurship quality (En2), when the competition degree of the banking industry is less than 0.922, the coefficient of the interaction term reflecting its moderating effect is −0.071, which passes the significance at the 10% level. When the support strength exceeds 0.922, the coefficient of the interaction term is 0.028, which is significant at the level of 1%. When the banking competition degree is lower than the threshold value, the banking competition degree is in the low regime, and the moderating effect of banking competition on the inhibitory effect of zombie enterprises on entrepreneurship is significantly negative. However, when the banking competition degree is higher than the threshold value, that is, the banking competition degree is in the high regime, the moderating role of banking competition on the inhibitory effect of zombie enterprises on entrepreneurship is significantly positive. This result can, to a certain extent, coordinate the “positive theory” (X. Zhang & Huang, 2022) and the “negative theory” (Chen et al., 2020; Y. Shen et al., 2023) on the effect of banking competition on zombie enterprises. That is, banking competition has a nonlinear threshold effect on the inhibitory effect of zombie enterprises on entrepreneurship. In the low regime, banking competition intensifies the inhibitory effect of zombie firms on entrepreneurship. However, in the high regime, banking competition can mitigate the inhibitory effect of zombie firms on entrepreneurship. In addition, the results also show that for entrepreneurship quality, the threshold value of banking competition is higher.

Table 11.

Panel Threshold Model Test Results.

Variables	En1	Variables	En2
Zom	−0.331^***(−4.667)	Zom	−0.409^***(−5.726)
BC	0.128^***(7.630)	BC	0.131^***(4.292)
Zom × BC(BC <0.904)	−0.059^***(−6.268)	Zom × BC (BC <0.922)	−0.071^*(−1.948)
Zom × BC(BC ≥0.904)	0.033^** (2.144)	Zom × BC (BC ≥0.922)	0.028^***(3.711)
Control variables	Y	Control variables	Y
City fixed effect	Y	City fixed effect	Y
Year fixed effect	Y	Year fixed effect	Y
Observations	4,352	Observations	4,352
R ²	0.344	R ²	0.359

Note.*, **, and *** indicate significance at the levels of 10%, 5%, and 1%.

Discussion and Policy Implications

Discussion

The paper employs dynamic panel models based on system GMM estimation and the panel threshold model to investigate the relationship between zombie firms and the “quantity” and “quality” of entrepreneurship, based on data from 272 prefecture cities in China. It specifically examines the moderating effect and threshold effect of banking competition. This study not only expands the research scope regarding the negative impacts caused by zombie enterprises (Chao et al., 2022; Chen et al., 2020; Kwon et al., 2015; Qiao et al., 2022; Wu et al., 2023) but also enriches existing literature concerning the association between banking competition and zombie enterprises (Chen et al., 2020; Y. Shen et al., 2023; X. Zhang & Huang, 2022). The study finds:

(1) The zombie enterprises have a significant inhibitory impact on both urban entrepreneurship quantity and quality. Importantly, the degree of zombification has a more pronounced inhibitory effect on entrepreneurship quality compared to entrepreneurship quantity. State-owned zombie enterprises have a more noticeable dampening effect on entrepreneurship than private zombie enterprises. In contrast to the research conducted by Qin and Kong (2022) and Bai et al. (2022), which solely focused on entrepreneurship quantity, this study encompasses an evaluation of the quality dimension of entrepreneurship, thereby expanding the conceptual framework of entrepreneurship and providing a broader perspective for investigating the influence of zombie enterprises on entrepreneurship.

(2) Banking competition plays a nonlinear moderating role in mitigating the inhibitory impact of zombie enterprises on entrepreneurship. This moderating effect is primarily driven by joint-stock banks and city commercial banks. The panel threshold model test reveals that when the degree of banking competition falls below the threshold value, indicating a low level of competition, the moderating effect of banking competition on the inhibitory influence of zombie enterprises on entrepreneurship is significantly negative. However, when the degree of banking competition surpasses the threshold value, signifying a high level of competition, the moderating effect becomes significantly positive. These findings reconcile both “positive theory” (X. Zhang & Huang, 2022) and “negative theory” (Chen et al., 2020; Y. Shen et al., 2023) regarding the impact of banking competition on zombie enterprises.

Policy Implications

The policy implications of the research conclusions are as follows:

Firstly, zombie enterprises have a significant inhibitory effect on both the quantity and quality of entrepreneurship, especially on the quality. Therefore, while actively adopting policies to promote entrepreneurial activity and improve the quality of entrepreneurship, the government must further promote supply-side structural reform and accelerate the establishment and improvement of an effective exit mechanism for zombie enterprises, to create a healthier urban entrepreneurial environment. It is crucial to fully utilize market-oriented policies and the “creative destruction” mechanism, gradually reducing zombie enterprises in the long run to enhance the quantity and quality of entrepreneurship.

Secondly, distinct measures should be explored to dispose of state-owned and private zombie enterprises. Given the more pronounced inhibitory impact of state-owned zombie enterprises on entrepreneurial activity and quality, it is particularly imperative to strengthen their disposal by reducing government intervention policies and excessive protection measures. Regarding private zombie enterprises, appropriate support may be provided to those with development prospects while expanding financing channels and enhancing supervision of interconnection and mutual insurance among private enterprises - especially in regions where private lending is prevalent - to prevent financial risks and contagion effects.

Thirdly, in disposing of zombie enterprises, it is imperative to enhance banking competition. Therefore, in the supply-side structural reform of finance, we should accelerate the opening of the banking market, lower the market access threshold, give full play to the decisive role of the market mechanism in the banking reform, and mobilize the vitality and competition degree of the banking market under the condition of improving the quality of supervision. More small and medium-sized banks, private banks, and foreign banks should participate in the market competition, which will help improve the allocation efficiency of credit resources in the whole society, eliminate inefficient and non-viable zombie enterprises, and improve the overall efficiency of the banking industry in serving the real economy.

Fourthly, when enhancing banking competition, it should be targeted toward different cities to prevent a “one size fits all” approach. For cities with low levels of banking competition, efforts should be made to open the banking market, promote financial marketization reforms, encourage diversification of financial products, and reduce barriers to bank entry. In cities where banking competition exceeds the threshold value, it becomes necessary to establish an effective information identification mechanism and a scientifically reasonable loan pricing mechanism that can mitigate information asymmetry between banks and enterprises while fostering a conducive external financing environment for businesses. Additionally, there is a need to enhance marketization levels and legal frameworks while strengthening supervisory cooperation among cities. It includes promoting comprehensive exchange and sharing of supervision information to minimize bank credit risks and foster financial stability. With effective risk prevention measures in place, financial regulators should empower local banks with appropriate credit business authority while further stimulating efficiency within the credit market for enterprise identification and screening.

Footnotes

Acknowledgements

No.

Author Contributions

Conceptualization, Xiangmin He, Shenrun Yan; Methodology, Xiangmin He, Shenrun Yan; Jingyun Miao; formal analysis, Xiangmin He, Shenrun Yan; Jingyun Miao; Investigation, Xiangmin He, Shenrun Yan; writing—original draft preparation, Xiangmin He, Shenrun Yan; Jingyun Miao. All authors have read and agreed to the published version of the manuscript

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project was provided by The National Social Science Fund of China (20BJL033), National Natural Science Foundation of China (71863024), Humanities and Social Science Foundation of Ministry of Education (18YJC790046).

Ethics Statement

Not applicable.

ORCID iD

Xiangmin He

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Ahearne

A. G.

Shinada

(2005). Zombie firms and economic stagnation in Japan. International Economics and Economic Policy, 2(4), 363–381. https://doi.org/10.1007/s10368-005-0041-1

Amore

M. D.

Schneider

Žaldokas

(2013). Credit supply and corporate innovation. Journal of Financial Economics, 109(3), 835–855. https://doi.org/10.1016/j.jfineco.2013.04.006

Arellano

Bover

(1995). Another look at the instrumental variable estimation of error-components models. Journal of Econometrics, 68(1), 29–51. https://doi.org/10.1016/0304-4076(94)01642-D

Bai

Zhang

Bian

(2022). Whether innovation-driven policies promote urban entrepreneurial activity: empirical evidence from the national innovative city pilot policy. China’s Industrial Economy, 6, 61–78. https://doi.org/10.19581/j.cnki.ciejournal.2022.06.016

Barth

J. R.

Nolle

D. E.

Phumiwasana

Yago

(2003). A Cross-Country Analysis of the Bank Supervisory Framework and Bank Performance. Financial Markets Institutions and Instruments, 12(2), 67–120. https://doi.org/10.1111/1468-0416.t01-2-00001

Berger

A. N.

Black

L. K.

(2011). Bank size, lending technologies, and Small Business Finance. Journal of Banking & Finance, 35(3), 724–735. https://doi.org/10.1016/j.jbankfin.2010.09.004

Berger

A. N.

Miller

N. H.

Petersen

M. A.

Rajan

R. G.

Stein

J. C.

(2005). Does function follow organizational form? Evidence from the lending practices of large and small banks. Journal of Financial Economics, 76(2), 237–269. https://doi.org/10.1016/j.jfineco.2004.06.003

Blundell

Bond

(1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143. https://doi.org/10.1016/s0304-4076(98)00009-8

Caballero

R. J.

Hoshi

Kashyap

A. K.

(2008). Zombie lending and depressed restructuring in Japan. American Economic Review, 98(5), 1943–1977. https://doi.org/10.1257/aer.98.5.1943

10.

Cagetti

De Nardi

(2006). Entrepreneurship, frictions, and wealth. Journal of Political Economy, 114(5), 835–870. https://doi.org/10.1086/508032

11.

Carlson

Correia

Luck

(2022). The effects of banking competition on growth and financial stability: Evidence from the National Banking Era. Journal of Political Economy, 130(2), 462–520. https://doi.org/10.1086/717453

12.

Carreira

Teixeira

Nieto-Carrillo

(2022). Recovery and exit of zombie firms in Portugal. Small Business Economics, 59, 491–519. https://doi.org/10.1007/s11187-021-00483-8

13.

Chang

Zhou

Liu

Wang

Zhang

(2021). How does government intervention affect the formation of zombie firms?Economic Modelling, 94(1), 768–779. https://doi.org/10.1016/j.econmod.2020.02.017

14.

Chao

Guo

Sun

(2022). Zombie problem: Normal firms’ wastewater pollution. Journal of Cleaner Production, 330,129893. https://doi.org/10.1016/j.jclepro.2021.129893

15.

Chen

Zhou

Liu

(2020). Zombie firms and firm Innovation: The perspective of Bank Competition. Economics and Management, 42(12), 5–22. https://doi.org/10.19616/j.cnki.bmj.2020.12.001

16.

Cooper

A. C.

(1973). Technical entrepreneurship: what do we know?R&D Management, 3(2), 59–64. https://doi.org/10.1111/j.1467-9310.1973.tb01002.x

17.

Cornaggia

Mao

Tian

Wolfe

(2015). Does banking competition affect innovation?Journal of Financial Economics, 115(1), 189–209. https://doi.org/10.1016/j.jfineco.2014.09.001

18.

Dai

Yang

Liu

(2020). Banking competition, innovation resource allocation and enterprise innovation output: based on the empirical evidence of Chinese industrial enterprises. Journal of Financial Research, 2, 51–70.

19.

Dai

Liu

(2021). Throwing good money after bad: Zombie lending and the supply chain contagion of firm exit. Journal of Economic Behavior & Organization, 189, 379–402. https://doi.org/10.1016/j.jebo.2021.06.042

20.

Dutta

Meierrieks

(2021). Financial development and entrepreneurship. International Review of Economics & Finance, 73, 114–126. https://doi.org/10.1016/j.iref.2021.01.002

21.

Fang

Zhang

Nie

(2018). The puzzle of small and medium-sized private enterprises becoming zombie Enterprises. Academic Monthly, 50(03), 75–86. https://doi.org/10.19862/j.cnki.xsyk.2018.03.009

22.

Fukuda

Nakamura

(2011). Why did ‘Zombie’ firms recover in Japan?World Economy, 34(7), 1124–1137. https://doi.org/10.1111/j.1467-9701.2011.01368.x

23.

Gao

Jiang

(2021). Small and medium-sized banks, financial structure and residents’ Entrepreneurship. Nankai Economic Studies, 3, 17–32. https://doi.org/10.14116/j.nkes.2021.03.002

24.

Geng

Liu

(2021). How do zombie firms affect China’s industrial upgrading?Economic Modelling, 97, 79–94. https://doi.org/10.1016/j.econmod.2021.01.010

25.

Global Entrepreneurship Monitor. (2017) Global Entrepreneurship Monitor 2016/2017 Global Report [EB/OL]. https://www.gemconsortium.org/file/open?fileId=49812

26.

Goto

Wilbur

(2019). Unfinished business: Zombie firms among SME in Japan’s lost decades. Japan and the World Economy, 49(3), 105–112. https://doi.org/10.1016/j.japwor.2018.09.007

27.

Gouveia

Osterhold

(2018). Fear the Walking Dead: Zombie firms, spillovers and exit barriers. OECD Productivity Working Papers<xyz/>, 3(25), 13. https://doi.org/10.1787/24139424

28.

Griffin

N. N.

Odaki

(2009). Reallocation and productivity growth in Japan: Revisiting the lost decade of the 1990s. Journal of Productivity Analysis, 31, 125–136. https://doi.org/10.1007/s11123-008-0123-5

29.

Hansen

B. E.

(1999). Threshold effects in non-dynamic panels: Estimation, testing, and inference. Journal of Econometrics, 93(2), 345–368. https://doi.org/10.1016/S0304-4076(99)00025-1

30.

Hechavarría

D. M.

Ingram

A. E.

(2019). Entrepreneurial ecosystem conditions and gendered national-level entrepreneurial activity: A 14-year panel study of GEM. Small Business Economics, 53(2), 431–458. https://doi.org/10.1007/s11187-018-9994-7

31.

Tie

Liu

(2019). Does bank deregulation in China promote optimal resource allocation. Financial and Economic research, 45(4), 83–95. https://doi.org/10.16538/j.cnki.jfe.2019.04.007

32.

Hoshi

Kashyap

A. K.

(2004). Japan’s financial crisis and economic stagnation. Journal of Economic Perspectives, 18(1), 3–26. https://doi.org/10.1257/089533004773563412

33.

Hoshi

Kim

(2013). Macroprudential Policy and Zombie Lending in Korea. ABFER Working Paper, No. 2251. https://www.abfer.org/media/abfer-events-2013/annual-conference/international-macroeconomics/track-1-macroprudential-policy-and-zombie-lending-in-korea.pdf

34.

Hou

(2022). Bank competition, financing constraints and Enterprise Innovation. Journal of Applied Finance and Banking, 12, 43–56. https://doi.org/10.47260/jafb/1263

35.

Huang

Deng

(2021). Does banking competition stimulate regional innovation? Evidence from China. Pacific-Basin Finance Journal, 70, 101674. https://doi.org/10.1016/j.pacfin.2021.101674

36.

Imai

(2016). A panel study of zombie SMEs in Japan: Identification, borrowing and investment behavior. Journal of the Japanese and International Economies, 39, 91–107. https://doi.org/10.1016/j.jjie.2015.12.001

37.

Jaskowski

(2015). Should zombie lending always be prevented?International Review of Economics & Finance, 40, 191–203. https://doi.org/10.1016/j.iref.2015.02.023

38.

Jayaratne

Wolken

(1999). How important are small banks to small business lending?: New Evidence from a survey of small firms. Journal of Banking & Finance, 23(2-4), 427–458. https://doi.org/10.1016/s0378-4266(98)00085-5

39.

Kerr

W. R.

Nanda

(2009). Democratizing entry: Banking deregulations, financing constraints, and entrepreneurship★. Journal of Financial Economics, 94(1), 124–149. https://doi.org/10.1016/j.jfineco.2008.12.003

40.

Kwon

H. U.

Narita

(2015). Resource Reallocation and zombie lending in Japan in the 1990s. Review of Economic Dynamics, 18, 709–732. https://doi.org/10.1016/j.red.2015.07.001

41.

Liu

(2022). Research on the impact of banking competition on corporate investment and financing maturity mismatch. International Finance Research, 7, 68–76. https://doi.org/10.16475/j.cnki.1006-1029.2022.07.007

42.

Love

Martínez Pería

M. S.

(2015). How Bank competition affects firms’ access to finance. World Bank Economic Review, 29(3), 413–448. https://doi.org/10.1093/wber/lhu003

43.

Malecki

E. J.

(2018). Entrepreneurship and entrepreneurial ecosystems. Geography Compass, 12(3), 12359. https://doi.org/10.1111/gec3.12359

44.

Mao

(2020). How land mismatch affects the quality of urban innovation and entrepreneurship in China: empirical evidence from prefecture-level cities. Industrial Economic Research, 3, 17–29. https://doi.org/10.13269/j.cnki.ier.2020.03.002

45.

(2017). Empirical analysis on the influence of factor price distortion on entrepreneurship. Statistics and Decision Making, 12, 175–178. https://doi.org/10.13546/j.cnki.tjyjc.2017.12.044

46.

McGowan

Andrews

Millot

(2017). The Walking Dead? Zombie Firms and Productivity Performance in OECD Countries, OECD Economics Department Working Papers, 25,45. https://doi.org/10.1787/18151973

47.

Nakamura

J. I.

Fukuda

S. I.

(2013). What happened to” zombie” firms in Japan?: Reexamination for the lost two decades. Global Journal of Economics, 02(02), 1350007. https://doi.org/10.1142/s2251361213500079

48.

Nie

Jiang

Zhang

Fang

(2016). The current situation, Reason and countermeasure of China zombie Enterprise. Macroeconomics Management. 31(9), 63–88. https://doi.org/10.19709/j.cnki.11-3199/f.2016.09.017

49.

Pagano

(1993). Financial Markets and growth. European Economic Review, 37(2-3), 613–622. https://doi.org/10.1016/0014-2921(93)90051-b

50.

Petersen

M. A.

Rajan

R. G.

(1995). The effect of credit market competition on lending relationships. Quarterly Journal of Economics, 110(2), 407–443. https://doi.org/10.2307/2118445

51.

Prieger

J. E.

(2023). Local banking markets and barriers to entrepreneurship in minority and other areas. Journal of Economics and Business, 124(4), 106108. https://doi.org/10.1016/j.jeconbus.2023.106108

52.

Qiao

Song

Fan

(2022). How do zombie firms affect innovation: From the perspective of credit resources distortion. Asian-Pacific Economic Literature, 36(1), 67–87. https://doi.org/10.1111/apel.12349

53.

Qin

Kong

(2022). Access to credit and Entrepreneurship: Evidence from China. Economic Development and Cultural Change, 71(1), 295–331. https://doi.org/10.1086/714440

54.

Rajan

R. G.

(1992). Insiders and outsiders: The choice between informed and Arm’s-Length Debt. Journal of Finance, 47(4), 1367–1400. https://doi.org/10.2307/2328944

55.

Schaeck

Cihák

(2012). Banking competition and capital ratios. European Financial Management, 18(5), 836–866. https://doi.org/10.1111/j.1468-036x.2010.00551.x

56.

Shao

Yin

Fan

Yang

(2022). Zombie enterprises and low-carbon transformation development: Based on the perspective of carbon emission performance. Journal of Quantitative and Technological Economics, 10, 89–108. https://doi.org/10.13653/j.cnki.jqte.20220920.004

57.

Shen

Chen

(2017). Zombie firms and over-capacity in Chinese manufacturing. China Economic Review, 44(3), 327–342. https://doi.org/10.1016/j.chieco.2017.05.008

58.

Shen

Ren

Zhao

(2023). Bank competition and zombie company: Empirical evidence from China. Economic Analysis and Policy, 80, 297–318. https://doi.org/10.1016/j.eap.2023.08.011

59.

Tan

Huang

Woo

W. T.

(2016). Zombie firms and the Crowding-Out of private investment in China. Asian Economic Papers, 15(3), 32–55. https://doi.org/10.1162/asep_a_00474

60.

Wennekers

Thurik

(1999). Linking entrepreneurship and economic growth. Small Business Economics, 13, 27–56. https://doi.org/10.1023/A:1008063200484

61.

Chang

Bai

Wei

(2023). How do zombie enterprises hinder climate change action plans in China?Energy Economics, 124,106854. https://doi.org/10.1016/j.eneco.2023.106854

62.

(2020). Factor market distortion and family entrepreneurship: An empirical analysis based on China Family Tracking Survey (CFPS) data. Journal of Yunnan University of Finance and Economics, 36(1), 78–88. https://doi.org/10.16537/j.cnki.jynufe.000558

63.

Zarutskie

(2006). Evidence on the effects of bank competition on firm borrowing and investment. Journal of Financial Economics, 81(3), 503–537. https://doi.org/10.1016/j.jfineco.2005.07.008

64.

Zhang

Zheng

Xin

(2017). Banking deregulation, structural competition and corporate innovation in China. China Industrial Economics, 10, 118–136. https://doi.org/10.19581/j.cnki.ciejournal.2017.10.008

65.

Zhang

Gao

(2018). The impact of foreign direct investment on Mass Entrepreneurship. China Industrial Economics, 23(12), 79–96. https://doi.org/10.19581/j.cnki.ciejournal.2018.12.005

66.

Zhang

Gao

(2020). Banking competition and resource misallocation: Evidence from Chinese industrial firms. Studies of International Finance, 6, 54–63. https://doi.org/10.16475/j.cnki.1006-1029.2020.06.006

67.

Zhang

Huang

(2022). Does bank competition inhibit the formation of zombie firms?International Review of Economics & Finance, 80, 1045–1060. https://doi.org/10.1016/j.iref.2022.03.011

68.

Zhao

Zhang

Liang

(2020). Digital economy, entrepreneurial activity and high-quality development: empirical evidence from Chinese cities. Managing the World, 10,65-76. https://doi.org/10.19744/j.cnki.11-1235/f.2020.0154