Entrepreneurship promotion policies and urban carbon emission efficiency: does environmental concerns matter?

To foster economic and social development, many countries have adopted entrepreneurship promotion policies. However, existing research on entrepreneurship promotion policy primarily focuses on its employment incentive effect, while neglecting to examine the underlying impact mechanism and its effect on environmental sustainability. Based on the quasi-natural experiment of China’s National Entrepreneurial Pilot Cities, this paper employs the difference-in-differences model to examine the impact of the National Entrepreneurial Cities pilot policy on urban carbon emission efficiency. The study finds that China’s National Entrepreneurial Cities pilot policy has generally impaired urban carbon emission efficiency, and a series of robustness tests show that the results are robust. National Entrepreneurial Cities pilot policy decreases urban carbon emission efficiency by increasing the proportion of necessity entrepreneurship, inhibiting the upgrading of industrial structure, and increasing overcapacity. The government’s environmental concerns and the public’s environmental concerns play a prominent role in improving the adverse impact of the National Entrepreneurial Cities on carbon emission efficiency and mainly play a significant role in central and western cities, cities with low per capita GDP, cities with high economic growth pressure, and resource-dependent cities. The research of this paper provides a reference for the government to promote entrepreneurship while taking environmental effects into account.


Introduction
Entrepreneurship is a prominent engine of economic growth [1,2].In recent years, in response to the gradual deceleration of China's economic growth and the simultaneous increase in labor costs and unemployment rates, the issue of optimizing economic structure and transforming growth drivers has gained increasing prominence.The government has advocated for 'mass entrepreneurship and innovation' as a means to effectively harness the pivotal role played by entrepreneurship in enhancing market dynamism and improving development quality.With concerted efforts from the central government, governments at all levels implemented more proactive policies to foster entrepreneurial activities.Widespread entrepreneurship has produced an outstanding effect in boosting employment, improving people's livelihood, and driving economic growth.
However, the continuous and rapid development of entrepreneurship is accompanied by the increasingly salient issue of environmental pollution.According to CEADs data, China's total carbon emissions increased from 6.546 billion tons in 2007 to 9.794 billion tons in 2019, representing an increase of nearly fifty percent.China's total carbon emissions ranked first in the world in 2019, accounting for 28.8% of global carbon emissions [3].The significant amount of carbon emissions has brought about a series of environmental and social problems that restrict the sustainable development of the economy and society.Therefore, the government must explore the potential correlation between policies promoting entrepreneurship and carbon emissions to facilitate entrepreneurial growth while considering environmental implications, thereby achieving a sustainable economic transformation towards low-carbon development.
The government and academia have reached a consensus on the positive role of entrepreneurship in economic and social development among the triple bottom line of sustainable development [4][5][6].Entrepreneurship has been proven to promote economic growth and create wealth [7,8]; It is conducive to job creation [9,10], poverty reduction [11,12], improving well-being [13,14].However, there are few studies on the relationship between entrepreneurship and environmental pollution in the existing literature [15], and there are large differences.Existing research on entrepreneurship promotion policy primarily focuses on its employment incentive effect [16], while neglecting to examine the underlying impact mechanism and its effect on environmental sustainability.Since entrepreneurial activities aim at profits, the implementation of relatively lenient policies promoting entrepreneurship may result in the proliferation of numerous startups characterized by low environmental standards, consequently exacerbating environmental pollution.Accordingly, three urgent questions need to be answered: What is the impact of government policies promoting entrepreneurship on environmental pollution?What is the influence mechanism?What measures can be taken to mitigate the effect?The answers to these questions will bring new and crucial implications to governments that attempt to promote economic growth and employment through entrepreneurship promotion policies.However, to the best of our knowledge, there is little research on this topic in the existing literature.Therefore, this study aims to address this gap by utilizing China's National Entrepreneurial Cities pilot policy as a natural experiment and employing the difference-in-differences (DID) model to assess the impact of this significant entrepreneurship promotion policy on urban carbon emission efficiency.
The possible marginal contribution of this paper is as follows: (1) Currently, the evaluation of entrepreneurship promotion policies primarily focuses on employment outcomes, neglecting their environmental implications.This study expands the scope of assessing the effectiveness of such policies by considering their environmental impact.(2) The existing literature on the impact of entrepreneurship on environmental pollution generally has strong endogeneity.The quasi-natural experiment of the National Entrepreneurial Cities in China is adopted to conduct research with the difference-in-difference method and strengthen the causal identification.Propensity score matching is used for robustness tests, which makes the estimation results more reliable.(3) This paper deeply explores the mechanism by which entrepreneurship promotion policies inhibit carbon emission efficiency.We find that National Entrepreneurial Cities in China can reduce the carbon emission efficiency of cities by increasing the proportion of necessity entrepreneurship, inhibiting the upgrading of industrial structure, and increasing overcapacity.Consequently, this will contribute to an enhanced comprehension of the environmental implications of policies aimed at promoting entrepreneurship.(4) This article examines the moderating influence of environmental concerns, encompassing both government and public perspectives, on the impact of the National Entrepreneurship Pilot Cities' carbon emission efficiency.The findings offer valuable policy implications for governments seeking to promote entrepreneurship while simultaneously enhancing carbon emission efficiency.
The rest of the paper is organized as follows.Section 2 presents the review policy and theoretical hypotheses.Section 3 is the Model, variables, and data; it introduces the methodologies of empirical model building, variable screening, and data selection.Section 4 provides the empirical results.Section 5 discusses differences, and limitations.Section 6 presents the conclusions and policy implications.

Review policy and theoretical hypothesis
This section first briefly reviews China's entrepreneurship promotion policies' background, development, policy objectives, and main points.On this basis, three hypotheses are proposed to address the research questions.

Brief review policy
The Chinese government's stance towards entrepreneurship has transitioned from stringent control to relaxation, and subsequently to encouragement since the initiation of reform and opening up.However, in light of evolving domestic and international circumstances, particularly the gradual decline of demographic dividends coupled with the simultaneous rise in labor costs and unemployment rates, it has become increasingly evident that urgent adjustments to the economic structure are necessary for resolving labor employment issues.Consequently, over time, the Chinese government's attitude towards entrepreneurship has grown more positive.During China's 17th National Congress of the Central Government, a significant policy was proposed:'implementing a development strategy focused on expanding employment through entrepreneurial endeavors.'This initiative sparked a wave of government-led entrepreneurship as the primary means for fostering employment opportunities.In addition, in 2008, China's Ministry of Human Resources and Social Security issued a notice promoting the establishment of entrepreneurial cities aimed at driving employment through entrepreneurial activities; this marked the commencement of national-level efforts to construct such cities.The policy is to foster a conducive business environment that facilitates entrepreneurship and harnesses the employment multiplier effect associated with it.The policy mandates pilot cities to actively explore and enhance policies and measures addressing prominent challenges, such as market entry, administrative management, financing channels, and other pertinent aspects.Additionally, efforts will be made to establish dedicated support funds for entrepreneurship while bolstering proactive employment policies through implementing supportive measures such as tax exemptions, fee waivers, small-scale guaranteed loans, and site arrangements.Finally, a policy framework will be established to promote entrepreneurship and cultivate an enabling entrepreneurial ecosystem.
In 2009, the General Office of the Ministry of Human Resources and Social Security announced the list of the first batch of National Entrepreneurial Cities, totaling 78.These cities are dispersed in all provinces except municipalities directly under the central government, and there are 1-4 pilot cities in each province (figure 1).The National Entrepreneurial City pilot program has been proven to have significantly promoted urban entrepreneurship [17], but the impact of this policy on the environmental pollution has not been paid attention by scholars.

Theoretical hypothesis 2.2.1. Effect of entrepreneurship promotion policies on urban carbon emission efficiency
What is the impact of government policies promoting entrepreneurship on environmental pollution?What is the influence mechanism?The answer to these questions will contribute to bridging the research gap in entrepreneurship promotion policies, which has predominantly focused on employment impacts while neglecting environmental consequences.This section will address this query from a theoretical standpoint.
A large amount of literature has confirmed the positive effect of entrepreneurship on economic growth, and the government also regards entrepreneurship as a sharp tool to boost local economic growth and promote employment.Entrepreneurship has been proven to promote urban economic growth and social development through channels such as the job creation effect, and wealth creation effect [10,12,14].
However, scholars have provided conflicting and uncertain results regarding the impact of entrepreneurship on environmental pollution.For example, Khezri and Muhamad (2023) believed that the improvement of entrepreneurship promotes the alleviation of environmental degradation by improving technical conditions, but the development of entrepreneurship worsens the negative impact of energy intensity rising on environmental pollution [18].
Entrepreneurship promotion policies have a positive role in carbon emission efficiency, mainly because environment-related market failures may generate opportunities for entrepreneurs to reduce behaviors that damage the environmental economy while achieving profits [19].Entrepreneurial behavior can address environmental challenges by introducing new technologies and resilience measures to mitigate and adapt to climate change, and through promoting environmentally sustainable practices and consumption patterns [20,21].York and Venkataraman (2010) claim that the positive effect of entrepreneurship on the environment was far more significant than the environmental damage [22].Based on the research at the enterprise level, Gu et al (2021) found that both the innovation spirit and risk-taking spirit of entrepreneurship promoted the pollution emission control of enterprises [23].He et al (2020) show that entrepreneurship can protect the ecosystem, reducing the disorderly deforestation of forests, improving the green function of agricultural activities, and promoting the development of ecological diversity [24].Nakamura and Managi (2020) used data from 62 countries to find that countries with high entrepreneurship rates have lower CO 2 emissions per capita [25].
However, unsustainable entrepreneurial enterprises triggered by profit expectations would abuse resources and hurt the environment [26].Riti et al (2015) employed the fully modified least squares technology to examine the impact of entrepreneurship on environmental quality and showed that there was a negative correlation between the two, making sustainable development impossible [27].Dhahri and Omri's (2018) study in 28 developing countries also confirmed the negative effect of entrepreneurship on environmental quality [28].Youssef et al (2018) found that informal entrepreneurship would increase pollution, especially in low-income countries.Moreover, entrepreneurship and pollution have an inverted U-shaped relationship in high-income countries [4].Youssef et al (2018), with Africa data, and Philip et al (2020), with Turkey data, both found that entrepreneurship aggravated environmental degradation, but higher innovation and institutional quality could improve this effect [4,29].
Omri (2018) confirmed that entrepreneurial activities rapidly increased the total CO 2 emissions of some countries, and his research further showed an inverted U-shaped relationship between entrepreneurship and CO 2 emissions [30].Omri and Afi (2020) claimed that these four entrepreneurial behaviors all resulted in environmental degradation by studying the impact of four different forms of entrepreneurship on the environment in 32 developing countries, among which informal and necessity entrepreneurship played a more prominent role in air pollution [31].Venancio and Pinto (2020) used the sample data from 67 countries and also found that entrepreneurial activities hurt sustainable development, which was caused by necessity entrepreneurship and non-innovative entrepreneurship [5].
To sum up, by stimulating the entrepreneurial vitality of entrepreneurial subjects, entrepreneurship promotion policies affect their industrial choice, resource consumption, product production, and investment in pollution control, and affect urban carbon emission efficiency.However,China's National Entrepreneurial Citiespilot policy is employment-oriented, which can be seen from the guiding ideology of the policy and relevant implementation departments.Under this guidance, the thresholds of enterprise registration and environmental impact assessment have lowered, giving rise to a large number of necessity entrepreneurship, informal entrepreneurship, and labor-intensive entrepreneurship, inhibiting the urban industrial structure upgrading and aggravating overcapacity, thus having a significant negative impact on urban carbon emission efficiency.
From the perspective of necessity entrepreneurship, the motivation for necessity entrepreneurship is survival, and entrepreneurs have significant passivity [32].The primary purpose of opportunity entrepreneurship is to take advantage of business opportunities and realize personal value.Compared with opportunity entrepreneurship, necessity entrepreneurs lack the initiative and passion to pursue entrepreneurial opportunities, and it is difficult to transform entrepreneurial opportunities into commercial value.Therefore, necessity entrepreneurship may neglect the impact on the environment to obtain immediate benefits due to myopia and utilitarianism.Both Venancio and Pinto (2020) and Omri and Afi (2020) found that necessity entrepreneurship increased pollution emissions, which was not beneficial to environmentally sustainable development [5,31].
From the perspective of industrial structure, the National Entrepreneurial Cities pilot policy with employment as their primary focus have resulted in a proliferation of labor-intensive enterprises.While these enterprises have effectively addressed a substantial portion of employment challenges, their relatively low labor productivity contributed to stagnation in urban industrial coupling.This stagnation in an industrial structure constitutes a significant factor contributing to inefficient carbon emissions [33].From the perspective of overcapacity, under the incentive of the National Entrepreneurial Cities pilot policy, entrepreneurial enterprises will choose more industries with low capital threshold and technology threshold, which leads to the fast growth of supply in these industries, far exceeding the existing demand, and the insufficient effective demand leads to the excess of production capacity and products.However, in some high-end industries, production capacity is inadequate.Overcapacity not only harms economic growth but also intensifies pollution emissions, causing air pollution and water pollution [34].
Hypothesis 1: National Entrepreneurial Cities pilot policy hurt urban carbon emission efficiency.
Hypothesis 2: National Entrepreneurial Cities pilot policy hinders the enhancement of urban carbon emission efficiency by promoting necessity entrepreneurship, impeding industrial structure upgrading, and exacerbating overcapacity.

Moderating effect of environmental concern on the relationship between entrepreneurship promotion policies and carbon emission efficiency
If the entrepreneurship promotion policy exerts a suppressive impact on urban carbon emission efficiency, what measures can be taken to mitigate it?This section aims to address the issue by examining it through the lenses of both governmental environmental concerns and public environmental concerns, thereby proposing an effective approach to mitigate the adverse impacts of entrepreneurship promotion policies.
The improvement of the public's environmental concern will have a positive impact on the willingness to green consumption [35].People with higher environmental concerns are more willing to buy green products, and People who are more concerned about the environment are more want to take action for environmental protection [36].In addition, consumers' environmental concerns have a crucial impact on their willingness to purchase environmentally friendly products.Positive market feedback will encourage entrepreneurs to invest in R&D of environmentally friendly production.Moreover, the cognitive factors of consumers provide opportunities for entrepreneurs and encourage them to transform environment-related technologies into market products, thus stimulating the enthusiasm of entrepreneurs and improving the success rate of entrepreneurship, thus affecting the environmental quality.
The government concern degree is a behavioral process in which the government decision-making body selectively pays attention to significant information and ignores other information.The higher the government's environmental concern is, the more importance local governments attach to environmental protection and the greater the guiding effect on environmental governance may be [37].Dong and Wang(2021) claim that both the public and the government's environmental attention significantly improved the level of green technology innovation and promoted the transformation of technology in the green direction [38].Farooq(2021) believed that the government's environmental concern has a significant effect on the green investment of enterprises [39].Wu et al (2022) shows that public environmental concern promotes local governments to improve the intensity of environmental regulation and investment in environment protection infrastructure construction by reducing the degree of information asymmetry between the central government, local governments, and enterprises, inhibiting the collusion between local governments and enterprises, and ultimately encourages enterprises to invest in environmental protection governance [40].On the other hand, public environmental attention increases the market's investment demand for green products and environmental protection enterprises, promotes the scale expansion of environmentally friendly enterprises, and ultimately improves the environmental quality.Yi et al (2022) found that public environmental concern can drive green innovation behavior of enterprises by strengthening government environmental regulation and improving enterprises' ecological concept [41].Reflected in the capital market, public environmental concerns can significantly improve the return rate of environmental protection stocks, affect individual investment behavior, and thus reallocation industrial capital [42].To sum up, the impact of entrepreneurial activities on the environment depends on the behaviors and motivations of entrepreneurs [43].The environmental quality of entrepreneurial enterprises is often closely related to the social reputation of the enterprises and affects the purchase intention of environmental protection consumers.Therefore, with the increasing government environmental concern and public environmental concern, driven by the baton of enterprise performance, entrepreneurial enterprises will inevitably improve the environmental quality, and thus enhance urban carbon emission efficiency.
Hypothesis 3: Environmental concern is conducive to mitigating the negative impact of the National Entrepreneurial Cities pilot policy on urban carbon emission efficiency.

Research design
This section is about research design, which includes three parts: model specification, variable selection, and data sources.

Model specification 3.1.1. Difference-in-difference method
The primary objective of this study is to assess the influence of the National Entrepreneurial Cities pilot policy on urban carbon emission efficiency.China's National Entrepreneurial Cities pilot policy can be regarded as a quasi-natural experiment.Therefore, this paper employs the DID model to examine the impact of the National Entrepreneurial Cities on urban carbon emission efficiency.DID models are widely used in the field of economics and sociology for evaluating the effects of policy implementation [44].The accuracy of assessing these effects has been validated through numerous studies.The model is as follows: Where i is the city, t is the year, and CEE it is the carbon emission efficiency of the city.EnPC it is the dummy variable of the National Entrepreneurial Cities; if its value is 1, it means city i has become a national entrepreneurial pilot city in year t; otherwise, it is 0. Specifically, EnPC it = treatment i × year t , where treatment i = 1 if this city is a pilot city, otherwise, it equals 0; year t equals 1 for every year after this city was approved as a pilot city; otherwise, it equals 0. The treatment group is the pilot cities belonging to the National Entrepreneurial Cities, and the control group used in this paper is the cities not belonging to the National Entrepreneurial Cities.δ i and μ t are city and year fixed effects, respectively, and ε it is a random error.The coefficient β 1 can capture the effect of the national entrepreneurial city pilot program on urban carbon emission efficiency.In addition, to obtain an unbiased estimator β 1 for the core explanatory variable EnPC it coefficient in the model (1), it is necessary to include city feature vector variable Z so that EnPC it is independent of ε it .That is, after incorporating potential control variables into the model, it simultaneously satisfies the randomness of the year when the list of National Entrepreneurial Pilot Cities was formulated and the randomness in selecting cities for the National Entrepreneurial Cities list.Clustering at the city level is employed to alleviate potential intergroup correlation issues and ensure robust standard errors.

Mediating models of influence mechanisms
From hypothesis 2, it can be seen that the National Entrepreneurial Cities pilot policy may damage urban carbon emission efficiency by increasing the proportion of necessity entrepreneurship, inhibiting the upgrading of industrial structure, and increasing industrial overcapacity.This paper examines these three channels by building a mediating effect model.
Where M is the mediating variable, it includes the proportion of necessity entrepreneurship (NeE), upgrading of industrial structure (SR) and overcapacity(CU).if the coefficients of the mediating variable in the equation (3) should be significant, which indicates a mediation effect exists.

Moderating effect model
To consider the moderating effect of environmental concern, we analyze the moderating effect based on Model (1) by constructing the interaction term.
Where EC represents urban environmental concern, it includes government environmental concern (GEC) and public environmental concern (PEC).

Variable screening 3.2.1. Dependent variable
The dependent variable of this paper is urban carbon emission efficiency (CEE).Referring to the method of Liu and Sun(2023) [3], this paper uses the logarithm of the ratio of GDP to the total carbon emissions of a city (100 million yuan/thousand tons) to represent the carbon emission efficiency.As for the calculation of total urban carbon emissions, due to no detailed energy consumption and classification data at the city level [45], the carbon emission data calculated by the Particle Swarm Optimization-Back Propagation (PSO-BP) algorithm in the study of Chen et al (2020) [46] are used to sum the data to the prefecture-level city level according to administrative divisions.This method is based on DMSP-OLS and NPP-VIIRS night light data to simulate carbon emissions.The logic behind this simulation method is that the higher the night light intensity is, the higher the economic development level of the city is, and the corresponding energy consumption.Because of its strong scientific nature, this method has been widely used in recent years [47,48].

Independent variable
The selected independent variable is dummy variable of National Entrepreneurial Cities (EnPC).According to the list of cities published by the Circular of the General Office of the Ministry of Human Resources and Social Security on the Announcement of the First Batch of National Entrepreneurial Cities, the core explanatory variables reflecting the implementation effect of the pilot policy are built in combination with the construction time of the National Entrepreneurial Cities.

Other variables
The control variables selected are as follows: (1) Foreign direct investment (FDI) .Trade openness, particularly foreign direct investment, has the potential to transfer pollution to the host country.However, it also facilitates the exchange and dissemination of advanced green technologies, thereby contributing to a reduction in pollution emissions [49].It is represented by the proportion of FDI in GDP in each city.(2) Population density (lnPop) .To some extent, an increase in population density can lead to economic agglomeration effects that facilitate the sharing of infrastructure and optimize resource allocation, thereby improving urban carbon emission efficiency.However, excessive population density may result in a crowding effect [50], which could hinder carbon emission efficiency.This is reflected by the number of people per square kilometer.(3) Energy consumption structure (energy).The carbon emission intensity of different energy sources is significantly different, and the carbon emission intensity of renewable energy is very low, but that of coal is very high.The difference in urban energy consumption structure leads to the heterogeneity of carbon emission efficiency [51].This variable is measured by the ratio of urban coal consumption to total energy consumption.(4) Science and education expenditure (scedu).Increasing investment in science and education yields multiple benefits, including the promotion of green innovation, improvement of economic structure, enhancement of energy efficiency, and consequent influence on carbon emission efficiency [52].It is expressed by the ratio of science and education expenditure to GDP. (5) The level of economic development (lnGDP).Low-income regions tend to drive economic development by increasing production activities with higher carbon dioxide emission intensity, as measured by per capita GDP.(6) Urban infrastructure (Inf), Improved road infrastructure will stimulate transportation demand, promote industrial development, and may lead to an increase in carbon emissions.It is measured by the ratio of urban highway mileage to urban land area.Mediating variables (NeE, SR, CU) and Moderating variables (GEC, PEC) will be reported in the following sections.

Data sources and description
Given the limitation of data availability, the panel data of prefecture-level cities in China from 2005 to 2015 is employed.Since province-level municipalities have special policy support, which is quite different from other cities, this paper excludes the data of Beijing, Shanghai, Tianjin, and Chongqing.
There are 78 National Entrepreneurial Cities in the treatment group and 203 in the control group, totaling 281 cities.The carbon emission data come from China Carbon Emission Accounts &Datasets (CEADs, https:// www.ceads.net/),which use night light and statistical carbon emission data to obtain spatial carbon emission data, and then extract the total carbon emission at the prefecture-level city scale.The enterprise-level data are from the Statistical Survey Database of Chinese Industrial Enterprises, and the data on the proportion of necessity start-ups was calculated after the query of Qi chacha, an industrial and commercial enterprise query platform.The city-level data come from the China Statistical Yearbook for Regional Economy, China City Statistical Yearbook, Statistical Yearbooks of provinces, Statistical Yearbooks of cities, and the China Economic

Estimation results of the benchmark model
Table 2 reports the regression results of the benchmark model.Columns (1) and (2) are the DID estimation results obtained under two-way fixed effects (FE), and columns (3) and (4) are the DID estimation results obtained under feasible generalized least squares (FGLS).It can be seen from the results that the coefficients of EnPC are all significantly negative, indicating that the entrepreneurship promotion policy represented by the National Entrepreneurial Cities has a significant negative effect on urban carbon emission efficiency.This result confirms Hypothesis 1.According to the findings from the feasible generalized least squares (FGLS) analysis, while holding other variables constant, National Entrepreneurial Cities exhibit an average reduction in carbon emission efficiency of 2.8% compared to non-National Entrepreneurial Cities.
The regression results of the control variables show that both the degree of openness to the outside world (FDI) and the expenditure on science and education (scedu) can significantly improve urban carbon emission efficiency.There is a significant U-shaped relationship between the level of economic development (lnGDP) and carbon emission efficiency.However, the increase in population density (lnPop) significantly reduces urban carbon emission efficiency, which may be because, in the current process of urban expansion in China, environmental pollution control facilities and green technology innovation capacity fail to match the increase of population density, which is not conducive to the improvement of carbon emission efficiency.The energy consumption structure (engry) shows a significant negative impact on urban carbon emission efficiency, indicating that the high-carbon energy structure brings prominent carbon emissions to the city.None of the coefficients on infrastructure(Inf) is significant.

Parallel trend analysis
Satisfying the hypothesis of parallel trend is a fundamental condition for the reliability of the different-indifferences test.Referring to the practice of Li et al (2016) [53], this paper adopts the event analysis method to analyze parallel trends and estimates the regression equation as follows: Where EnPC iK is a series of dummy variables, it represents the policy variables of the years before and after the implementation of the National Entrepreneurial Cities pilot policy.Since the sample period is from 2005 to 2015, and the National Entrepreneurial Cities pilot policy implemented in 2009, we choose four periods ahead and six periods behind for analysis.From the parallel trend test results reported in table 3, it can be seen that the dummy variable regression coefficients of the first four years of the National Entrepreneurial Cities are not significant, indicating that there is no significant difference between the treatment group and the control group before the city becomes a pilot city, that is, the hypothesis of the parallel trend established.

Robustness test 4.3.1. Propensity score matching test
To overcome the bias of benchmark estimation results caused by systematic differences in the changing trends of National Entrepreneurial Cities and non-National Entrepreneurial Cities, this paper uses the PSM-DID method to control the endogeneity problems caused by selection bias.The PSM-DID method can effectively reduce the differences in measurable variables between the treatment group and the control group.According to the existing literature, the control variable in Model (1) is employed as the matching variable, and three methods of radius matching, kernel matching, and nearest neighbor matching are applied to match the samples, respectively.It is easy to see that the estimated coefficients of EnPC in table 4 are all significantly negative, that is, further supporting the benchmark regression results.

Other robustness tests
This paper also conducts the following robustness tests.First, the dependent variable is changed, and the SBM-DEA method is employed to measure the urban carbon emission efficiency.Referring to Shao et al (2022) [54], this paper takes labor and capital as input factors, gross regional product as the expected output, and carbon dioxide emissions as unexpected output and adopts a non-radial and non-angular SBM-DEA model to measure urban carbon emission efficiency.Capital and labor inputs are measured by the capital stock and the annual average number of urban employment, respectively.Carbon dioxide emissions are consistent with the previous measurement.Second, perform a placebo test.Because the effects of pilot policies identified by DID may also be disturbed by missing variables and other unobservable urban characteristics, the problem of 'implicit bias' will be brought up.Therefore, to identify the estimation errors that may be produced by the unobservable characteristics of cities and omitted variables, this paper further conducts a placebo test on the empirical results.The identification strategy of the placebo test is to construct the counterfactual that the neighboring cities of the National Entrepreneurial Cities are also the National Entrepreneurial Cities [17].In this paper, we search for placebo cities for National Entrepreneurial Cities within the same provincial administrative region for matching.The regression results of the matched samples are shown in table 5.The coefficient of National Entrepreneurial Cities is no longer significant, indicating that the reduction of carbon emission efficiency after implementing the National Entrepreneurial Cities pilot policy is not caused by missing variables or other unobservable urban characteristics.
Third, eliminate the samples of high-level cities.Since the policy environment of cities separately listed in the state plan and cities at sub-provincial level is quite different from that of prefecture-level cities, to exclude the possible impact of some particular policies on the samples of high-level cities, the regression is conducted after excluding cities separately listed in the state plan and cities at sub-provincial level.
Fourth, the time interval of the benchmark regression sample is from 2005 to 2015, and the period after the implementation of the pilot policy is long, which increases the possibility that the estimation results of this paper are affected by other factors.Therefore, this paper selects the samples from 2005 to 2012 for regression to examine the short-term effects of implementing the pilot policy.

Mechanism test result 4.4.1. Increase the proportion of necessity entrepreneurship
We collected the micro data of new ventures less than or equal to 42 months during the investigation period through the 'Qichacha' platform and defined the entrepreneurial enterprises whose registration information is The numbers in parentheses are Z-statistics.
'individual business' as necessity entrepreneurship [55], and those with other industrial and commercial attributes as opportunity entrepreneurship.Based on the above groups, we recalculate the number of newly registered enterprises at the city-year level and calculate their respective proportions to obtain the proportion of necessity entrepreneurship (NeE) and the proportion of opportunity entrepreneurship (OpE), which are then substituted into Model (1) as explained variables for regression.From the regression results reported in table 6, it can be seen that the National Entrepreneurial Cities pilot policy of has a significantly positive effect on the proportion of necessity entrepreneurship, but has no significant impact on the proportion of opportunity entrepreneurship.This shows that China's employment-oriented pilot policy of National Entrepreneurial Cities has increased the proportion of necessity entrepreneurship, but has no significant impact on the proportion of opportunity entrepreneurship.In the results of Column (3), the coefficients of NeE is significantly negative, indicating that the National Entrepreneurial Cities pilot program reduces the carbon emission efficiency of the city by increasing the proportion of necessity entrepreneurship.

Restrain the upgrading of industrial structure
The upgrading of urban industrial structures is an outstanding reason for reducing carbon emissions and improving urban carbon emission efficiency [56].Referring to the method of Sun et al (2020) [57], industrial structure upgrading is measured according to the coupling degree of urban factor input structure and output structure, and the mechanism of entrepreneurship promotion policy inhibiting carbon emission efficiency by destroying industrial structure upgrading is examined.
Where SR it is the degree of industrial coupling in the city i in year t, Y imt represents the output of industry m in the city i in year t, Y it denotes the output of the city i in year t, L imt is the employment of industry m in the city i in year t, L it represents the employment in the city i in year t, the larger the SR it is, the more reasonable the industrial structure is.SR is substituted into the mediating effect model and the regression results are shown in table 7. The results indicate that the National Entrepreneurial Cities pilot policy has significantly inhibited the upgrading of urban industrial structure.The results in Column (2) show that the coefficient of SR is positive and passes the significance test, indicating that the National Entrepreneurial Cities pilot policy reduces the urban carbon emission efficiency by suppressing the industrial structure upgrading.

Exacerbating overcapacity
Drawing on the idea of Guo et al (2022) [58], this paper uses DEA to measure urban capacity utilization rate (CU) and regards capacity utilization rate as a reverse indicator to measure the degree of overcapacity.From the results in Column (3) of table 7, it can be seen that the national-level entrepreneurial city pilot program has a significantly negative effect on the urban capacity utilization rate, which indicates that the national-level entrepreneurial city pilot program is not conducive to the improvement of urban capacity utilization rate, that is, it aggravates overcapacity.The coefficient of CU in Column (4) is significantly positive, which confirms that The numbers in parentheses are Z-statistics.
the national entrepreneurial pilot city program inhibits the improvement of urban carbon emission efficiency through the channel of worsening overcapacity.

Empirical test results of the moderating effect of environmental concern
Government environmental concern (GEC) uses the word frequency statistics method to calculate the relevant word frequency in the work reports of prefecture cities' governments [59] and logarithmically.At the present stage, China's high-quality economic development continues to advance, residents' pursuit of a better life is also increasing, and the government, as the leader of institutional and mechanism construction, is also paying more attention to environmental protection.Statistical keywords include environmental protection, ecological environment, treatment plant, ecological construction, comprehensive treatment, coverage rate, returning farmland to forest, pollutants, sewage treatment, clear water and green mountains, air, PM2.5, garbage, energy saving, and clean.
Referring to the practice of Gu et al (2022) [60], this paper selects the Baidu search index of environmental protection keywords to describe public environmental concerns (PEC).On the one hand, Baidu, the largest Chinese search engine in China, has a greater coverage area and high data availability.On the other hand, keywords such as 'low-carbon,' 'green,' 'new energy,' and 'haze' can effectively represent the public's preferences and behavioral intentions in terms of environment and energy.Therefore, this paper chooses the Baidu index of the above keywords as the proxy variable of public environmental concern and performs logarithmic processing on the variable.
Table 8 considers the moderating effect of the urban government's environmental concern and the public environmental concern.In the results, the coefficient of the interaction term between the National Entrepreneurial Cities and government/public environmental concerns have coefficients of 0.038 and 0.023, respectively.These are positive and significant, indicating that both the government environmental concern and the public environmental concern are conducive to improving the negative effect of entrepreneurship promotion policies on carbon emission efficiency.The empirical results confirm Hypothesis 3. Comparing, the coefficient of the interaction term between government environmental attention and National Entrepreneurial Cities is greater than the coefficient of the interaction term between public environmental concern and entrepreneurship promotion policies, which indicates that the government's environmental attention can more significantly improve the negative effect of National Entrepreneurial Cities to the carbon emission efficiency.4.6.Heterogeneity analysis 4.6.1.Regional heterogeneity To check the differential impact of regional heterogeneity, we first divide the sample into the eastern, central, and western regions.And then use sub-sample data to analyze the influence of entrepreneurship promotion policies on urban carbon emission efficiency.According to the results in table 8, the National Entrepreneurial The numbers in parentheses are Z-statistics.
Cities pilot program has a significantly negative effect on carbon emission efficiency in the central and western regions but has not in the eastern cities.The reason is that compared with the eastern cities, both the central and western cities are geographically remote, with weak infrastructure and scarce talents, and most low-end industries with high energy consumption concentrating here, a large proportion of the entrepreneurial enterprises brought about by the National Entrepreneurial Cities are high-carbon industries.Therefore, in both regions, the National Entrepreneurial Cities significantly weakened carbon emission efficiency.
In the central and western cities, the upgrading of government environmental attention and public environmental attention has significantly improved the inhibitory effect of the National Entrepreneurial Cities on carbon emission efficiency.In Western cities, the improvement effect is more prominent.

Heterogeneity of economic development
Dong and Wang (2021) point out that urban wealth has a selective preference for green technology [38].That is to say, cities with more developed economies have an enormous demand for environmental protection.According to the average per capita GDP, the samples were divided into two sample groups: high per capita GDP and low per capita GDP.The empirical results (table 9) show that the National Entrepreneurial Cities have a significant inhibitory effect on urban carbon emission efficiency in the group of low per capita GDP, but no significant influence on the group of high per capita GDP, and in the low per capita GDP group, both the government and the public environmental concern help to mitigate the inhibition effect of the National Entrepreneurial Cities pilot policy to the carbon emission efficiency.

Economic growth targets heterogeneity
Under the pressure of economic growth, government departments at all levels often take various 'extraordinary actions' to achieve the established growth target [61].Higher economic growth targets may lead the government to lower environmental standards and acquiescence in more start-ups to invest in high-carbon industries with quick short-term results, thus affecting the carbon emission impact of entrepreneurship promotion policies.This paper chooses the degree of deviation between the economic growth target and actual economic growth as the proxy for pressure on economic growth.Considering the impact of economic fluctuations and the better measurement of regional economy growth endowment, referring to the practice of Wang et al (2021) [62], we choose the ratio of the economic growth target of prefecture-level cities in the current year to the average actual growth rate of the previous three years (taking the natural logarithm) to measure the pressure on economic growth faced by local governments.Based on the median pressure on the economic growth of each city in the current year, the higher economy growth pressure group and the lower pressure on economic growth are divided.The results show that National Entrepreneurial Cities has a significant adverse impact on urban carbon emission efficiency in the high economy growth pressure group, but has no significant effect in the low economic growth pressure group.The government environmental concern and the public environmental concern in the higher economy growth pressure group are both to improve the inhibitory effect of the National Entrepreneurial Cities pilot policy on carbon emission efficiency.

Resource endowment heterogeneity
Resource-dependent cities are particular industrial cities that rely on natural resources for development.Such cities rely too heavily on resource endowments, and their industrial development has formed a strong path dependence, which makes it easy to fall into the trap of recession.Therefore, entrepreneurship in resourcedependent cities may also be more resource-dependent.According to the National Sustainable Development Plan for Resource-based Cities(2013-2020) issued by The State Council of China, the cities are divided into resource-dependent cities and non-resource-dependent cities, and the test is carried out again.From the results in columns (3) and (4) of table 10, it can be seen that the National Entrepreneurial Cities have a significantly negative effect on urban carbon emission efficiency in resource-dependent cities.However, the influence is not significant in non-resource-dependent cities.The government's environmental concern and the public's environmental concern are also conducive to decreasing the inhibition effect of the National Entrepreneurial Cities pilot policy on carbon emission efficiency in resource-dependent cities.

Discussion
To foster economic development and enhance employment opportunities, many countries have implemented policies aimed at promoting entrepreneurship.This study seeks to investigate the impact of these policies on carbon emission efficiency and proposes a novel approach to mitigate their adverse environmental effects.By addressing a research gap in the existing literature that predominantly focuses on the influence of entrepreneurship promotion policies on entrepreneurial vitality and employment [16,63], with limited exploration of their implications for environmental pollution, this research offers a fresh perspective for comprehensively understanding the effects of entrepreneurship promotion policies on economic and social development while underscoring the significance of deepening our awareness regarding environmental concerns.
Concerning the influence of the National Entrepreneurial Cities pilot policies on urban carbon emission efficiency, this study's findings partially corroborate the conclusions drawn by Omri and Afi (2020) [31], Philip et al (2022) [29], and Gu et al (2020) [34].Specifically, it suggests that entrepreneurship contributes to environmental degradation.However, unlike these studies, our research primarily focuses on evaluating the impact of policies promoting entrepreneurship on the environment.Methodologically, we employ a difference- in-differences model to examine the causal effect of the National Entrepreneurial Cities pilot policies on urban carbon emission efficiency, effectively addressing endogeneity concerns between variables.The study in this paper contributes a Chinese case to the study of the relationship between entrepreneurship and environmental pollution.This paper finds that a crucial channel for the National Entrepreneurial Cities pilot policy to inhibit urban carbon emission efficiency is to expand the proportion of necessity entrepreneurship, which also supports the findings of Venancio and Pinto (2020) [5] and Omri and Afi (2020) [31].This paper argues that because the policy of China's the National Entrepreneurial Cities pilot policy is employments oriented, under this guidance, the government may lower the standards of enterprise registration and the threshold of environmental impact assessment, giving rise to a large number of necessity entrepreneurship, informal entrepreneurship, and laborintensive entrepreneurship, inhibit the upgrading of urban industrial structure, and exacerbate overcapacity.Thus, it has a significant adverse impact on urban carbon emission efficiency.These findings offer valuable insights for countries grappling with employment challenges by incentivizing entrepreneurship, highlighting the potential drawbacks of employment-focused entrepreneurial promotion policies that may foster an excessive proliferation of survival-based and informal businesses, ultimately leading to environmental degradation.Moreover, this finding aligns with the perspective of Shane (2009) [64] and Ajide et al (2024) [65], suggesting caution in promoting widespread business creation as a policy approach, particularly in developing nations.The finding of the significant moderating role of government and public environmental concern on the carbon emission efficiency inhibitory effect of entrepreneurship promotion policies, as discussed in this article, echoes the conclusion of Ouadghiri et al (2021) [66] and Aliano et al (2023) [67].Government and public environmental concerns are conducive to guiding businesses towards green investments and reducing environmental pollution.It provides a reliable approach for governments to mitigate the environmental pollution caused by entrepreneurship promotion policies.

limitations
Although the research presented in this paper holds significant theoretical value and practical significance, it is not without limitations.Firstly, regarding the selection of the dependent variable, we have exclusively focused on urban carbon emission efficiency and have not conducted a comprehensive analysis of the impact of entrepreneurship promotion policies.In future studies, our objective is to gather a large sample of corporate carbon emission data to thoroughly examine the effects and mechanisms of the National Entrepreneurial Cities pilot policies on corporate carbon emission efficiency.Secondly, about pollutant emissions, this study solely focuses on carbon emissions.Are there differential effects of the National Entrepreneurial Cities pilot policies on other pollutants such as SO 2 , water pollution, and haze?Would these findings alter the overall conclusion?Future research endeavors will encompass a broader range of pollutants to enhance the reliability of our conclusions.

Conclusions and policy implications 6.1. Conclusions
This study attempts to evaluate entrepreneurship promotion policies from the perspective of environmental effects, aiming to fill the gap in existing literature regarding the evaluation of entrepreneurial policies from an environmental standpoint.Using China's the National Entrepreneurial Cities as a natural experiment, this research employs a DID model to examine the impact of this policy on urban carbon emission efficiency.Additionally, it investigates the moderating role of government and public environmental concerns on these effects.The findings include: firstly, the National Entrepreneurial Cities pilot policy hinders improvements in carbon emission efficiency, with pilot cities exhibiting an average of 2.8% lower carbon emission efficiency than non-pilot cities.This conclusion suggests that the National Entrepreneurial Cities pilot policies lead to environmental degradation.Secondly, mechanism analysis reveals that the National Entrepreneurial Cities pilot policy negatively affects urban carbon emission efficiency through three channels: increasing the proportion of necessity entrepreneurship, inhibiting industrial structure upgrading, and exacerbating overcapacity issues.Thirdly, interaction terms between the National Entrepreneurial Cities and government/public environmental concerns have coefficients of 0.038 and 0.023, respectively, for urban carbon emission efficiency, indicating that both government and public environmental concerns mitigate the adverse effects caused by this policy.Lastly, heterogeneity analysis shows that the National Entrepreneurial Cities pilot policy primarily suppresses carbon emission efficiency in central-western cities, low per capita GDP levels, high economic growth pressure, and resource-dependent cities; furthermore, the moderating effect of environmental concerns is observed in these cities.

Policy implications
The policy implications are as follows: First of all, in the process of encouraging widespread entrepreneurship, the government should also pay attention to the possible environmental damage caused by start-ups and pay attention to the screening and management of entrepreneurial enterprises.Especially for necessity entrepreneurship, it is necessary to strengthen publicity, help them overcome short-sighted behaviors, guide them to carry out clean production and reduce their damage to the ecological environment to weaken the negative effect of entrepreneurial behavior on urban carbon emission efficiency.Secondly, in promoting entrepreneurship, city governments need to pay more attention to the environment, formulate appropriate environmental protection policies, strengthen environmental regulation, and avoid excessive reduction of environmental standards to promote employment.At the same time, attach importance to the public environmental attention, optimize the way of public environmental participation, and reasonably guide the public environmental attention, establish and promote the green mark system in production to meet the public demand for green products.And give full play to its role in guiding and constraining entrepreneurial enterprises.Finally, in promoting entrepreneurship, the government needs to increase the intensity of environmental attention according to local conditions and formulate different environmental regulation policies.It sufficiently takes the differences in the region and economic development level of diverse regions into account.Regions under tremendous pressure of economic growth or resource-dependent cities should not relax the attention of the government and the public to the environment while promoting economic development.

Figure 1 .
Figure 1.Spatial distribution of the National Entrepreneurial Cities.

Table 1 .
Descriptive statistics of variables.Network database.The government's environmental concern is assessed through text analysis of the government work report, while the public's environmental concern is gauged based on Baidu Index data.The descriptive statistics of the data are presented in table 1.

Table 2 .
Benchmark regression results for the impact of the national entrepreneurial cities on carbon emission efficiency.* and *** indicate significance at the level of 10%, 5% and 1%, respectively.
* , *The numbers in parentheses are Z-statistics.

Table 3 .
Result of parallel trend test.

Table 7 .
Mechanism test: restrain the upgrading of industrial structure and exacerbating overcapacity.

Table 9 .
Results of regional heterogeneity and economic development heterogeneity analysis.

Table 10 .
Results of economic growth targets heterogeneity and resource endowment heterogeneity analysis.* and *** indicate significance at the level of 10%, 5% and 1%, respectively.The numbers in parentheses are Z-statistics.