How to improve the government-enterprise relationships to retain enterprise?

Abstract How can enterprises and governments establish and improve government-enterprise relationships (GER) to increase the efficiency of business operations and enterprise retention (ER), and ensure the sustainable development of enterprises and regional economies? Based on resource dependence theory and institutional theory, from the enterprise level and the institutional level, GER is composed of five elements: organisational political strategy, individual political strategy, government fairness, government service, government convenience. This article studies the multiple paths and the mechanism of the coupling of the five elements of GER to achieve high ER. Fuzzy-set qualitative comparative analysis (fsQCA) and the data of GER in China's 31 provinces from 2010 to 2019 were used to conduct configuration analysis and the causal relationship between multiple factors of GER and ER. Sixteen thousand eight hundred fifty-four observations of Chinese A-share listed companies are used to conduct an empirical analysis of corporate political strategy on enterprise retention willingness in government business environment, and to test the robustness of the configuration analysis results. Results indicate that a single factor of GER is not a necessary condition to produce high ER, four types of GER configurations (Government-driven government-enterprise separation, enterprise-driven government-enterprise separation, government-and-enterprise-driven government-enterprise cooperation, government-assisted government-enterprise separation) can produce high ER, and three kinds of GER that produce non-high ER are obtained. The empirical analysis of listed companies tested the configuration analysis results and robustness.


Introduction
The process of economic transformation is the process of reconstructing GER (Nie, 2020). In the transitional economy, how enterprises can improve their profitability and operating efficiency by establishing GER, and how the government can increase ER by improving GER is of vital importance to ensuring the sustainable development of enterprises and regions. The external environment for the survival and development of enterprises is composed of GER, which determines the strategic decisionmaking and business behavior of enterprises (Luo & Liu, 2009). ER refers to the number of enterprises at the end of the year after all new enterprises enter and existing enterprises exit in a region within a year. It is an important indicator of a region's business environment, economic development, industrial agglomeration and investment promotion (Yang & Zhang, 2009).
Before the 18th National Congress of the Communist Party of China, the central government delegated power to the local government, with GDP growth as the evaluation index for official promotion. The local government launched fierce competition around GDP growth, and the central and local governments also implemented various measures to protect state-owned enterprises, such as taxation, subsidies, prices and credit (Wang et al., 2014). The government has failed to create a level playing field for private enterprises, small and medium-sized enterprises and foreign enterprises. In many regions, rent-seeking enterprises and corrupt government officials have emerged, and the GER is relatively strained. After the 18th National Congress of the Communist Party of China in 2012, China's economic growth slowed down, from high-speed economic growth to medium speed and high-quality growth. Through the comprehensive deepening of economic system reform, the decisive role in resource allocation began to be transferred from local governments to the market, the functions of local governments were gradually reduced, most of the market activities were carried out by enterprises, and the transaction costs of enterprises participating in market activities were reduced. The Fifth Plenary Session of the 18th CPC Central Committee in 2015 pointed out that 'let the market play a decisive role in resource allocation, the government and the market form cooperation to ensure the maximum efficiency of resource allocation'. The reform of 'streamline administration and delegate power, improve regulation, and upgrade services' has greatly improved the fairness, serviceability, and convenience of the government business environment, stimulated the vitality of the market and enterprises, and improved the GER. Corresponding to different stages of economic development, government behavior has evolved from local government competition to the government giving way to market, to government and market cooperation and the relationship between government and enterprise is also improving continuously.
Since 2017, the central government and local governments of China have made it one of their key tasks to build a new type of political business relationship and optimise the business environment (Nie, 2020), preferential policies to promote investment promotion have been introduced continuously to attract enterprises to enter. At the same time, a large number of enterprises exit every year. According to the development of national market entities released by the comprehensive planning department of the State Administration of market supervision and administration, from 2016 to 2020, the number of actual market entities increased from 98.148 million to 138.407 million, of which the number of existing enterprises increased from 25.961 million to 43.314 million at the end of the year. The number of new enterprises is 5.528 million, 6.074 million, 6.7 million, 7.391 million and 8.035 million each year. It can be seen that, with the rapid development of market players, more than 1 million enterprises have exited each year since 2017. From 2017 to 2020, the number of enterprises exited is respectively 1.152 million, 1.669 million, 2.859 million, 3.304 million and the number rises sharply.
Therefore, enterprises are actively exploring what kind of political strategy to establish GER to reduce the uncertainty of the institutional environment and enhance the acquisition of resources and opportunities, thereby improving organisational performance and sustainable development. At the same time, local governments are accelerating the reform of the economic system, optimising GER, and creating a good business environment to attract and retain more enterprises. From the perspective of enterprises and the government, what strategy should be adopted to build GER to increase ER, and then ensure the stable development of enterprises and the regional economy has become an important issue facing China.
On the one hand, the previous studies start from the perspective of enterprises, take institutional environment and policy uncertainty as environmental variables in the context of a transitional economy, and establish a formal or informal institutional relationship with the government through the implementation of political strategy. The relationships enable enterprises to influence public policy or government decision-making, which has an impact on enterprise performance and retention intentions (Guo et al., 2014). Some researchers believe that the establishment of GER through political strategies by enterprises can cope with the uncertainty of the institutional environment, strengthen legitimacy and resource possession (Li & Huang, 2017), create a relaxed regulatory environment and reduce fraud penalties, thereby enhancing enterprises performance, stable income and promote the survival and development of enterprises (Chen & Liu, 2019). Other researchers have found that the implementation of political strategies, especially the establishment of strong ties with the government, has caused enterprises to be affected by different institutional logics, which reduces operating efficiency, squeezes out the development space of enterprises, and reduces enterprise retention willingness (Chen & Liu, 2019). For example, enterprises need to help the government achieve social policy goals such as expanding employment (Yuan et al., 2015) and promoting social stability (Dang et al., 2015), and spending a lot of money to maintain GER to inhibit enterprise creative investment (Xia et al., 2019), changes in local government officials have led to fluctuations in the performance of government-enterprise-related enterprises (Shi et al., 2014). On the other hand, previous studies focus on the perspective of the government's institutional supply, and the research results are keen to evaluate and optimise the government business environment from indicators, rankings or plans, mainly from the aspects of government efficiency, administrative approval, and market supervision (Lou & Zhang, 2018), use the ranking of the government business environment to judge the GER and the impact on ER.
The lack of previous studies is that it ignores the heterogeneity of corporate political strategy, as well as its different effects on the path difference in the acquisition of resources and legitimacy, and its different effects on GER and ER. The institutional environment is regarded as the external variable of the rational choice of enterprises. Meanwhile, the government's institutional supply lacks the research on the demand side of enterprises as the main body of GER. This article is mainly reflected in: First, based on the overall perspective, the elements of GER are integrated from the enterprise level and institutional level. Second, starting from important phenomena (GER and ER), construct a theoretical model of configuration analysis. Third, explore the multiple paths and mechanism of the coupling of the five elements of the GER to achieve high or non-high ER. An empirical analysis of the data of Chinese listed companies to verify the configuration analysis results. The results are helpful to provide theoretical reference for enterprises and the government to build and improve the GER to increase ER, and to ensure the stable development of enterprises and the regional economy in the context of China's transitional economy.

Theoretical analysis and research framework
GER is a kind of interactive mode formed in the game by the government through subsidies, taxation, regulation and other policies, and enterprises through political strategies such as senior management's political status and contacting government officials (Nie, 2020). The research on GER at present mainly focuses on political connection, government intervention, industrial policy, and business environment. Nie proposed that GER can be divided into four types: government-enterprise cooperation, government-enterprise collusion, government-enterprise separation and government-enterprise sabotage by two dimensions of government intervention mode and degree of intervention (Nie, 2020). This paper refers to Nie's research ideas. From a holistic perspective, based on resource dependence theory and institutional theory, five factors influencing GER constituted by the enterprise level and the institutional level are derived. They are the organisational political strategy and individual political strategy of the corporate political strategy, and the fairness, serviceability and convenience of the government business environment.

Corporate political strategy of GER
The resource dependence theory believes that the entry and survival of enterprises depend on market natural resources and government basic resources, and the government is a vital resource provider (Guo et al., 2014). Through effective political strategies, enterprises expect to obtain scarce resources to gain a competitive advantage, including government subsidies, government orders, tax incentives, bank loans, internal information, etc. At the same time, enterprises also try to obtain policy support and protection, or guide policy changes in the interests of enterprises, thereby reducing the uncertainty in the environment (Wei et al., 2021).
In the context of China's institutional transformation, corporate political strategy is divided into organisational political strategy and individual political strategy (Wei et al., 2021). Organisational political strategy refers to the fact that an enterprise takes the government and other authoritarian departments as its strategic implementation targets, and strives to embed in the formal institutional environment to achieve its economic and policy goals. Institutional political strategy refers to the fact that enterprises take government officials as their strategy implementation targets, and realise their own economic and policy goals by embedding in an informal institutional environment and establishing a network of interpersonal relationships between entrepreneurs and officials. The two have different effects on the ER in terms of access to resources and legitimacy.
The organisational political strategy ensures that the enterprise obtains stable, continuous, and comprehensive resources and legitimacy, and is subject to formal regulations that can effectively suppress the rent-seeking tendencies and opportunistic motives of both parties, and promote ER. Through organisational political strategies, enterprises have more opportunities to contact government departments at different levels, to obtain comprehensive and objective information (Yuan et al., 2015). Enterprises are cooperating with the government for economic interests and compliance needs, and there is no obvious lock-in effect. While improving the associated stability, it can also ensure the flexibility of the enterprise. The openness of formal institutional participation can enhance enterprise reputation, increase government and stakeholder support and its legitimacy in the institutional environment (Li & Huang, 2017). And help enterprises win more resources, deal with uncertainties in the transition period, and effectively suppress rent-seeking and opportunistic motivations (Li et al., 2013).
The individual political strategy brings resources to the enterprise. But it is limited to stability, continuity, comprehensiveness, and the maintenance cost is high. The exchange of interests between the government and the enterprise is concealed, subject to fewer supervises and constraints, increasing rent-seeking and opportunism motivation. The tendency is not conducive to the long-term development of enterprises and ER. Relying on individual officials to obtain resources has a lock-in effect, and the resources are limited (Li & Huang, 2017). Sudden changes in the business environment and the removal of officials will exacerbate the unsustainability of resource supply (Xia et al., 2019). Enterprises need continuous investment to effectively maintain relationships, and they are even 'grabbing hand' by officials to help them complete political tasks and affect normal business operations (Luo, 2000). Entrepreneurs and officials usually maintain relationships in private and informal ways, which are hidden and difficult to track, and imply non-compliance (Flaherty, 2019). Due to the difficulty of supervision, the rent-seeking costs and risks of the government and enterprises are weakened, and the rent-seeking and opportunistic motives of both parties are strengthened (Dong et al., 2016).

Government business environment and GER
The institutional theory believes that a good external institutional environment provides an institutional guarantee for the long-term stable development of enterprises . New institutional economics incorporates the institutional environment as an internal variable into the analysis of individual actions. By strengthening supervision, increasing financial support, and reducing government intervention, the government improves the fairness, serviceability and convenience of the government business environment, and helps attract and retain enterprises. Government fairness is reflected in the intensity of legal supervision. Sound supervision will increase the cost of interest exchange between the government and enterprises in violation of laws and regulations, reduce rent-seeking and corruption, help form a fair market competition environment and government fairness, and attract enterprises to enter (Flaherty, 2019). Government fairness is conducive to the protection of intellectual property rights, and reduces the 'crowding out effect' and 'substitution effect' of enterprise rent-seeking corruption on the R&D investment, increasing the enthusiasm of enterprise innovation, and promoting the sustainable development of enterprises (Du et al., 2020). Government fairness creates more transparent financing channels for enterprises, reduces information asymmetry, and provides strong financial support for the development of enterprises. It has a positive impact on enterprises' capital acquisition and R&D confidence, and has a positive effect on ER (Judge et al., 2020).
Government service is reflected in the degree of unpaid resource supply. The institutional theory believes that compliance with the institutional environment or consistency with institutional requirements can endow enterprises with legitimacy and social support (Li & Huang, 2017). Government service produces strong support or resource constraints on the production and operation activities of enterprises, and will have an impact on enterprises' location choice. In the stage of economic transformation, to implement national strategic objectives, the government usually intervenes in enterprises through industrial policies. Supporting policies are a compliant intervention method, which is conducive to attracting enterprises to enter and ER (He et al., 2019). However, to achieve economic targets, local governments with insufficient capabilities, while providing policy support, also force enterprises to undertake government goals or excessive investment, which interferes with business decision-making and increases additional costs for enterprises. Sometimes enterprises are even allowed to choose 'bad' production technologies (such as low-cost, high-pollution and unsafe) to achieve rapid economic growth in the short term, causing a large number of accidents and hindering the long-term development of the company (Nei, 2020).
Government convenience is reflected in the degree of government intervention. A highly open and sound market environment can reduce market entry barriers and facilitate enterprise entry (Nystr€ om, 2007). The extreme form of government intervention in enterprises is state-owned enterprises. The proportion of state-owned enterprises has an obstacle effect on enterprise entry and a positive effect on enterprise exit (Ren et al., 2015). Recessive budget constraints of state-owned enterprises lead to the occupation of social resources and hinder the entry and exit of marketoriented enterprises (Wang et al., 2014). Government convenience reduces ineffective intervention and can reduce the institutional transaction costs incurred by the time and cost of dealing with the government to start and operate the enterprise. It can restrain non-production or destructive behaviors caused by institutional cost pressure, stimulate market vitality, and attract enterprises to enter . Government convenience reduces the 'crowding-out effect' of innovation investment caused by the squeeze of the production costs of enterprises by institutional transaction costs, etc., providing enterprises with opportunities and space for sustainable development (Yu & Liang, 2019).
GER is composed of five elements, which are coupled to form different GER to affect ER. This paper focuses on two causal relationships: (1) whether some elements of GER are necessary for a region to achieve high ER; (2) how these elements of GER are coupled to achieve high ER. The complex mechanism of the interaction between multiple elements of GER, by using configuration analysis is explored. The theoretical model is shown in Figure 1.

QCA
QCA method was developed by the sociologist Ragin, and it has also been more and more applied in management in recent years. In this paper, QCA method is selected because its advantages help to reveal the causal complexity of regional GER and ER (Meyer et al., 1993). Firstly, QCA method focuses on the comprehensive explanatory power of variable combinations on the outcome variables, which helps to explore the independent effect of a single antecedent condition in GER that cannot perfectly explain the formation logic of ER (Dul et al., 2020). Secondly, the different causal paths leading to high ER can be identified by QCA. Multiple configurations will be formed among multiple antecedents of the GER, and these configurations may lead to high ER. Thirdly, QCA method infers the causal relationship between the condition and the result through the set relationship (Fiss, 2011). From the perspective of set theory, QCA method can identify which conditions (configurations) are sufficient or necessary for the results. The practical enlightenment based on this collective relationship is simple and clear. QCA uses Boolean algebra, which will not lead to missing variable bias. Therefore, there is no control variable requirement in QCA method .
QCA can be divided into three branches according to data types: csQCA (Crisp set qualitative comparative analysis, dichotomous data), mvQCA (multi-value qualitative comparative analysis, multivalued data), and fsQCA (Fuzzy set qualitative comparative analysis, continuous data) (Zhang & Du, 2019). Considering the advantages that fsQCA can handle both category and degree issues, and the conditional data type of this article is continuous, fsQCA is chosen as a research method to explore the causal and complex mechanism of GER affecting ER. The fsQCA is applied to make an overall analysis of cross-time and cross-case. To identify whether a single antecedent condition is a necessary condition to generate ER, and explores various condition configurations of GER to generate high or non-high ER.

Data source and measurement
The data covers the GER and ER in 31 provinces in China from 2010 to 2019. The data on organisational political strategy and individual political strategy comes from the CSMAR database, and government fairness, government service, government convenience, and ER comes from the database of the National Bureau of Statistics of China.
3.2.1. Outcome variables ER is measured by the number of enterprises in each province at the end of each year, and the data is dimensionless processed to produce a standardised value of ER between [0,1]. The closer the value is to 1, the higher ER in the province (Du et al., 2020).

Antecedent variables
3.2.2.1. Organisational political strategy. Use the political status of enterprise executives to measure the strength of the organisational political strategy (Wei et al., 2021). Senior executives with political identities can have a certain impact in the formulation and implementation of policies, and can use their political identities to fight for the benefits of the enterprise. When there are current NPC deputies or members of the CPPCC among the enterprise executives, the enterprise is deemed to implement the organisational political strategy, otherwise, it is deemed that the organisational political strategy has not been implemented. The degree of organisational political strategy is measured by the proportion of the number of enterprises implementing organisational strategy in the number of regional enterprises. The higher the proportion, the higher the organisational political strategy.
3.2.2.2. Individual political strategy. Use enterprise hospitality and travel expenses as a percentage of operating income to measure the strength of an individual political strategy. Using Luo's management expenses such as hospitality and travel expenses to measure the company's personal relations with government officials (Luo, 2000), the higher the proportion of management expenses in operating income, the higher the individual political strategy.
3.2.2.3. Government fairness. Use legal supervision force to measure government fairness. It is the binding force to supervise the behavior of the government and enterprises (Nie, 2020). Measured by the number of enterprises under provincial supervision and random inspection of product quality, it can directly reflect the strength of legal supervision and indirectly reflect the strength of social supervision. The larger the number, the stronger the supervision, indicating that the government fairness is higher.
3.2.2.4. Government service. Use government intervention to measure government service. It refers to the compliance of how the government intervenes in the production or business activities of enterprises (Nie, 2020). It is measured by the death toll of work safety accidents in the GDP of 100 million yuan. The lower the number is, the more compliant the government intervention methods are, indicating the higher government service. Therefore, the reciprocal processing of the indicator is performed, that is, the greater the value of the indicator after processing, the higher government service.
3.2.2.5. Government convenience. Use the degree of government intervention to measure government convenience. It refers to the degree of government intervention in the production or business activities of enterprises. The extreme form of government intervention is state-owned enterprises. The proportion of state-owned enterprises in the number of regional enterprises is used to measure the degree of government intervention. The lower the proportion, the lower the degree of intervention (Nie, 2020), indicating the higher government convenience. Therefore, the reciprocal of the index is processed, that is, the greater the value of the index after processing, the higher government convenience.

Variable calibration
QCA uses the objective quantile value to determine the location of three qualitative anchors (Furnari et al., 2021). The three calibration points(which are full-subordinate, intersection and no-subordinate) of five conditional variables and one outcome variable are set as the upper quartile (75%), the median and, the lower quartile (25%) of the descriptive statistics of the case samples (Du et al., 2020). The calibration of nonhigh ER is realised by taking the non-set of high ER. The calibration anchor points and descriptive statistics of each variable are shown in Table 1.

Necessity analysis of the single condition
Use fsQCA 3.0 software for necessity analysis. Use the consensus threshold of 0.9 widely recognised by existing research as the criterion for judging its necessity (Misangyi & Acharya, 2014). It can be seen from Table 2 that the consistency values of all conditions are lower than 0.9, indicating that there is no single condition that can constitute the necessary condition for high or non-high ER. Therefore, it is necessary to investigate the influence of conditional configuration on ER.

Sufficient analysis of conditional configuration
FsQCA 3.0 software is used to analyse the configuration of GER that leads to high ER and non-high ER. These different configurations represent different GER that achieves the same result. The original consistency threshold is set to 0.8, the PRI consistency threshold is set to 0.7, and the case frequency threshold is set to 1. Due to the lack of evidence and theory that environmental conditions affect the exact direction of the results, in the counterfactual analysis, this study assumes that whether a single GER condition exists or not can contribute to high (or nonhigh) ER (Du et al., 2020).  (Du et al., 2021). According to the symbolic expression of QCA, in the results of configuration analysis in Table 3, the solid circle (᭹) indicates that the condition exists, the cross circle () indicates that the condition is absent, and the space indicates that the condition may or may not appear (Du et al., 2020). The big circle represents the core condition. The small circle represents the edge condition (There are no edge conditions in the results in Table 3.). Table 3 shows that four configurations (S1, S2, S3, S4) produce high ER, while three configurations produce non-high ER (NS1, NS2, NS3).
The consistency of all configurations is greater than 0.90, which is higher than the generally accepted consistency standard of 0.80. The unique coverage and original coverage of configuration S1 and NS1 were the highest in high ER and non-high ER, respectively. It indicates that S1 and NS1 are the most empirically relevant configurations. According to the theoretical process of configuration, the configuration found in this paper is named.

The configuration of the GER with a high ER
Four configurations (S1, S2, S3, S4) of GER produce high ER.
4.2.1.1. Government-driven government-enterprise separation. Configuration S1 points out that GER with high government fairness and high government convenience as the core conditions can produce high ER. It shows that no matter whether the enterprise implements or implements any political strategy, the government only needs to strictly legal supervision, increase the protection of property rights, less interfere with the market and enterprise behavior, and play its 'invisible hand' role.
In a market-oriented environment with high government fairness and convenience, enterprises can develop well and orderly. The GER is government-enterprise separation, resulting in high ER. S1 has the highest unique coverage and original coverage in all configurations, indicating that the Government-driven government-enterprise separation has the strongest correlation with high ER. The largest proportion of the configuration S2 cases in Guangdong Province. Guangdong's business environment reform focuses on fair management and convenient services. By improving the standardisation and transparency of regulatory enforcement, improving the regulatory standard system in various fields, reducing the time for real estate registration and processing, and promoting tax facilitation and other measures to form a fair and convenient business environment has stimulated the vitality of the market and enterprises. According to the 'Report on Business Environment Evaluation by Ten Thousands of Private Enterprises in 2020', Guangdong has been rated as 'the province with the best reputation for the business environment'. In 2019, the ER in Guangdong was 3.038 million, accounting for 7.9% of China, ranking first. 4.2.1.2. Enterprise-driven government-enterprise separation. Configuration S2 points out that the GER with high organisational political strategy, high government convenience, and non-high individual political strategy as the core conditions can produce high ER. It shows that by implementing high organisational political strategies, enterprises can improve their social status, obtain policy support and protection, and make it easier to obtain scarce resources. Usually, the profitability of the enterprise is better. In addition, the degree of government intervention is low, providing high government convenience. Enterprises rely on their capabilities and a free market environment to develop healthily, without excessive spending on hospitality to implement individual political strategies. The GER is government-enterprise separation, resulting in high ER. The largest proportion of the configuration S2 cases in Jiangsu Province. From 2010 to 2019, the management expenses of listed companies in Jiangsu accounted for an average of 7% of operating income, and their individual political strategies were low. The Jiangsu government transforms government functions, simplifies everything that can be simplified, and speeds up everything that can be accelerated, making government services more convenient, and addressing the needs of enterprises with precise and effective measures. From 2017 to 2019, the ER in Jiangsu was more than 2 million, ranking second in China, second only to Guangdong.

Government-and-enterprise-driven government-enterprise cooperation.
Configuration S3 points out that the GER with high organisational political strategy, high government fairness, and high government service as the core conditions can produce high ER. It shows that the company adopts a high organisational political strategy to lay a good foundation for its development. The government provides industrial policies and financial support, and provides good government services for enterprises. Under strict supervision, high government fairness is guaranteed, and the opportunity cost of government-enterprise collusion is increased. Both government and enterprise tend to cooperate. The GER is government-enterprise cooperation, resulting in high ER.
The largest proportion of configuration S3 cases in Hunan Province. About 30% of the listed companies in Hunan have political status, and the company and the government maintain an open and transparent cooperative relationship at the formal institutional level. Hunan has strict legal supervision and market supervision. The 'ten Bans' on optimising the business environment of Hunan Province's market supervision system stipulates strict regulatory red lines 4.2.1.4. Government-assisted government-enterprise separation. Configuration S4 points out that the GER with high organisational political strategy, high government service, and high government convenience as the core conditions can produce high ER. It shows that the enterprises adopt a high organisational political strategy and maintain an open and transparent relationship with the government at the formal institutional level. The government has played the role of 'helping hand' and adopted highly compliant intervention measures, such as policy benefits, pollution disclosure and work safety, to promote economic development and improve government service. The degree of government intervention is low, and enterprises enjoy the benefits of the 'helping hand' in a convenient market environment. The GER tends to government-enterprise separation, resulting in a high ER.
The largest proportion of configuration S4 cases in Shanghai. While providing highly convenient marketisation for enterprises, Shanghai also pays great attention to financial support to enterprises. Since 2016, the proportion of commercial service industry expenditures in total fiscal expenditures has reached 2%, the highest in China. Since 2012, the ER in Shanghai has reached more than 400,000, and in 2017 it has reached more than 450,000, ranking second among cities, second only to Beijing.

The configuration of the GER that produces the non-high ER
Three configurations (NS1, NS2, NS3) of GER produce non-high ER.
NS1 shows that in the GER that lacks high government fairness, lacks high government service, and lacks high government convenience, no matter whether the company implements or chooses which political strategy to implement, ER is not high. It shows that the institutional environment is bad, and the government is likely to use public power to grab the benefits of enterprises, causing government-enterprise sabotage, leading to non-high ER.
NS2 shows that in the GER that lacks high organisational political strategy, high individual political strategy, high government service, and high government convenience, no matter how government fairness is, ER is not high. It shows that when government services and convenience are absent, enterprise performance is limited, and organisational political strategies and individual political strategies are not high. To survive, enterprises may resort to illegal methods to deceive the government, causing government-enterprise sabotage, leading to non-high ER.
NS3 shows that in the lack of high organisational political strategy, high government fairness, and high government convenience, enterprises adopt high individual political strategies in the GER, regardless of the government service, ER is not high. It shows that the supervision is weak, and the government has taken certain actions to intervene in the enterprise highly, making the business environment lacking fairness and convenience. In order to survive, enterprises adopt high individual political strategy without the support of organisational political strategy, and form government-enterprise collusion. Enterprises may use 'bad' economic growth methods, leading to high accidents, high pollution, cutting corners, and tax evasion. This kind of GER can only bring short-term conspiracy benefits to enterprises, and cannot support their long-term development, leading to non-high ER.
Through comprehensive analysis, it is found that the three configurations of high ER include high organisational political strategy and high government convenience. All the three configurations of non-high ER lack of high government convenience. This shows that in China's transition period, the government's reduction of intervention and acceleration of market-oriented reform are very effective in promoting ER.

Configuration analysis test
The robustness of the GER configuration of high and non-high ER is verified. The threshold of the number of cases is increased from 1 to 2 (Zhang & Du, 2019). The PRI consistency is increased from 0.7 to 0.75 (Du et al., 2020). Deleting 20 cases of Beijing and Shanghai (political and economic resources are obviously superior) from 2010 to 2019. The configurations obtained by the three analyses are basically the same, and the configuration analysis results are robust.

Empirical analysis test
To further ensure the robustness of the configuration model and results, the ER is replaced by enterprise retention willingness (ERW). Use Stata 16.0 to analyse the interaction effects of corporate political strategy and government business environment. The model takes ERW as the dependent variable, organisational political strategy and individual political strategy as independent variables, government fairness, government service and government convenience as moderator variables. Taking the 2010-2019 A-share listed companies in China's Shanghai and Shenzhen stock markets as the initial sample (data source CSMAR database), we exclude enterprises with ST, PT, delisting, negative net assets, missing major variables and financial enterprises. All continuous variables were processed with winsorize at the 1% level. In the end, a total of 16,854 observations were obtained, and a panel data with a time range of 2010-2019 was established. The ERW is measured by the return on total assets of the enterprise. The higher profitability, the higher ERW and the higher ER (Wei et al., 2021). The measurement of organisational political strategy is that the senior management is the current NPC or CPPCC member, taking 1; otherwise, taking 0. The remaining variables are consistent with the measurements of configuration analysis. After standardising all continuous variables, Table 4 is the final interaction effect analysis result.
Model 1 is the regression result of independent variable OrgCps and IndCps. The coefficient of OrgCps is positive but not significant (b ¼ 0.013), and the coefficient of IndCps is negative and significant (b ¼ À0.171, p < 0.01). It shows that the organisational political strategy is not directly related to ERW. The higher the individual political strategy, the lower the ERW. Therefore, it is verified that the single condition of organisational political strategy in the configuration analysis is not a necessary condition for ER. High individual political strategy has a significant inhibitory effect on ER (NS3).
Model 2 adds the moderating variable GovFairness and its interaction terms with independent variables. The results show that the primary term of the moderating variable is not significant (b ¼ À0.004), the coefficient of OrgCps Ã GovFairness is positive but not significant (b ¼ 0.014), and IndCps Ã GovFairness is positive and significant (b ¼ 0.020, p < 0.01). It shows that the government fairness weakens the inhibitory effect of individual political strategy on ERW. Therefore, it is verified that the single condition of the government fairness in the configuration analysis is not a necessary condition for ER. The high individual political strategy intensifies non-high ER, which is more obvious with the addition of non-high government fairness (NS3). Model 3 adds the moderating variable GovService and its interaction terms with independent variables. The results show that the primary term of the moderating variable is not significant (b ¼ À0.010), the coefficient of OrgCps Ã GovService is positive and significant (b ¼ 0.044, p < 0.01), and IndCps Ã GovService is positive and significant (b ¼ 0.026, p < 0.01). It shows that the interactive effect of government service and organisational political strategy plays a significant role in promoting ERW. The government service weakens the inhibitory effect of individual political strategy on ERW. Therefore, it is verified that the single condition of government service in the configuration analysis is not a necessary condition for ER. The combination of high organisational political strategy and high government service contributes to the generation of high ER (S3, S4), and the combination of non-high organisational political strategy and non-high government service produces non-high ER (NS2).
Model 4 adds the moderating variable GovConvenience and its interaction terms with independent variables. The results show that the coefficient of the primary term of the moderating variable is positive and significant (b ¼ 0.032, p < 0.1), the coefficient of OrgCps Ã GovConvenience is positive and significant (b ¼ 0.045, p < 0.01), and IndCps Ã GovConvenience is negative and significant (b ¼ À0.022, p < 0.01). It shows that the higher the government convenience, the higher the ERW. The interactive effect of the government convenience and the organisational political strategy plays a significant role in promoting ERW. The government convenience strengthens the inhibitory effect of individual political strategies on ERW. Therefore, it is verified that the high government convenience in the configuration analysis has a universal effect on the production of high ER (S1, S2, S4), and the non-high government convenience has a universal effect on the production of non-high ER (NS1, NS2, NS3). The combined effect of high organisational political strategy and high government convenience helps to produce high ER (S2, S4), and the combined effect of non-high organisational political strategy and non-high government convenience produces nonhigh ER (NS2, NS3).

Theoretical enlightenment
Based on resource dependence theory and institutional theory, a holistic perspective provides new ideas for the study of GER and ER. First, build a configuration framework for GER, and determine the five conditions of GER: organisational political strategy, individual political strategy, government fairness, government service, government convenience. Second, a single factor of the GER does not constitute a necessary condition for high ER, but high government convenience has a universal effect on high ER. Third, the GER that exhibits a limited diversity of various element combinations can produce high or non-high ER. Fourth, for China during the economic transition period, the marketisation of high government convenience has played a key role in high ER, and the poor government business environment is an important reason for non-high ER.

Management enlightenment
There are three management enlightenments for the government and enterprises to build GER and increase ER in the context of China's transformation. First, although each enterprise has different choices and implementation levels of organisational political strategy and individual political strategy, there are differences in government fairness, government service and government convenience that each province can provide. It does not necessarily hinder enterprises from achieving high ER by improving organisational political strategies and governments through multiple methods such as improving the government convenience. Second, for enterprises, implementing a political strategy is more conducive to increasing the ER than not implementing a political strategy, such as S3, S4 and NS2. The implementation of organisational political strategy is more conducive to ER than the implementation of individual political strategy, such as S2 and NS3. Third, for the government, vigorously advancing the process of marketisation, strict supervision, reducing direct intervention, and providing enterprises with high government fairness and convenience are important measures to produce high ER, such as S1 and NS1.

Limitations and directions for future research
The measurement of government fairness, government service and government convenience used in this article focuses on the perspective of government institutional supply, and lacks of demand side consideration of entrepreneurs as the main body of the GER. Therefore, interviews and questionnaires should be carried out to study the relationship between entrepreneurs' subjective cognition of GER and ER from the perspective of entrepreneurs' perception, and to explore mild and low-cost cognitive intervention to improve ERW with the help of entrepreneurs' subjective perception of GER, which is worthy of further research. First, the ER is affected by enterprise retention willingness, and the main body that determines enterprise retention willingness and operating decisions is the entrepreneur (Egfjord & Sund, 2020). Therefore, it is possible to study the effectiveness of the optimisation of the GER and its relationship with enterprise retention willingness from the perspective of entrepreneurs' perception, and put forward more targeted suggestions for adjusting the government's institutional supply. Second, the judgment of entrepreneurs on GER does not completely follow the evaluation of objective indicators. According to the expectation perception theory, the judgment of entrepreneurs on GER is the result of comparing expectations and perceptions (Lou & Zhang, 2018). The expectation of GER comes from the direct and indirect experience of the entrepreneur. The comparison of expectations and perceptions can reflect the characteristics of the GER and the opportunities and risks it provides to the enterprise, and thereby determine the entrepreneur's understanding of GER and ERW (Xu et al., 2021). Therefore, we can further study the mediating effects of GER satisfaction, opportunity, and risk on the perception of GER and ERW. Finally, since the perception of GER is a psychological mechanism, it is more importantly affected by factors such as experience, attitude preferences, and cognitive processes (Hoogendoorn et al., 2019). When decision makers are in a decision-making situation with limited or vague information, they tend to take 'psychological shortcuts' in the process of information processing, and stereotypes help them make quick decisions (Fiske et al., 2002). Therefore, we can further study the moderating effect of stereotypes on GER preference and ERW. At the same time, it can reveal the causes of the perception differences and cognitive biases in the regional GER, and provide a scientific basis for the governance of the stereotype of the regional GER.

Disclosure statement
No potential conflict of interest was reported by the authors.