Impact of fiscal decentralization and local government competition on the supply of basic public services: Based on the empirical evidence of prefecture-level cities in China

Abstract

Promoting equal access to public services and improving people's well-being is a key link in building a modern national governance system in China. However, under the Chinese-style fiscal decentralization system, local governments face the “dilemma” of economic growth goals and the improvement of people's livelihoods. China's basic public services still have the problems of insufficient supply quantity, unbalanced structure and low efficiency. This paper aims to explore the impact of fiscal decentralization and local government competition on basic public services, and provide a theoretical and practical basis for deepening the reform of China's fiscal and taxation system, perfecting the transfer payment system, and improving the public service provision at the present stage. Under the premise of theoretical hypothesis, based on the panel data of 178 prefecture-level cities in China from 2008 to 2019, which is obtained from the Statistical Yearbook and the Work report of prefecture-level governments, the fixed effect model and threshold model are used for the empirical test. The results show that: first, the increase of fiscal decentralization has a negative impact on supply of basic public services in prefecture-level cities; second, the local government competition aiming at economic growth will intensify the negative impact, and make the relationship between fiscal decentralization and basic public services show nonlinear characteristics; Third, compared with developed areas, the strengthening regulatory effect of local government competition is more obvious in less-developed areas, such as western and northeastern regions, and third-tier, fourth-tier, and fifth-tier cities. Based on these findings, this paper draws the following policy implications: strengthen the reform of the fiscal and taxation system below the provincial level, promote the construction of a direct transfer payment mechanism, adjust the standards and methods of performance appraisal, and use modern information technology to improve the public service demand and interest expression mechanisms.

1. Introduction

Since reform and opening up, China has made remarkable achievements in economic growth and people's living standards have improved significantly. However, due to the mismatch between the financial and administrative powers of the central and local governments, China's basic public services still have the problems of insufficient supply quantity, unbalanced structure and low efficiency [1]. In particular, the quality of public services in non-economic areas such as basic education, medical and health care, and social security needs to be improved. With the reform of China's tax distribution system, local governments have obtained certain fiscal revenue rights, which provides financial support for the supply of public services. However, Chinese-style fiscal decentralization has made a new division of financial power and administrative power between the central and local governments [2]. The central government has gained more financial power while delegating more administrative power to the local governments, resulting in local governments facing the financial dilemma of “insufficient funds but excessive responsibilities ". In recent years, the structural tax reduction policies implemented in China have further increased the financial pressure on local governments, exacerbating the misalignment and asymmetry between local government financial resources and public services functions.

At the same time, under the performance evaluation mechanism that prioritizes economic development, the main competition among local governments is the management of regional economic growth targets. The local governments set economic growth targets to motivate lower-level governments and departments to strive to develop the economy and win the championship. Due to the characteristics of high investment and slow effect of public service, in order to promote the economic growth of the jurisdiction, local government officials tend to invest limited financial resources in the economic field, thus reducing the investment in the public service field [3]. China's fiscal decentralization gives local governments some autonomy in revenue and expenditure, allowing them to have the power to decide the size and structure of budget expenditures independently. Under the dual pressure of vertical fiscal imbalance and performance appraisal in China, sub-provincial governments are faced with the dilemma between economic development and providing public services: The limited financial funds should be invested in “economic” areas closely related to political achievements, or in “non-economic” public welfare areas such as basic education, medical care, and social security?

How does fiscal decentralization affect local governments' supply of basic public services? How does local government competition with economic growth as the main goal affect the relationship between fiscal decentralization and the supply of basic public services? This paper attempts to study these issues in light of China's actual situation. It has significant practical guiding value for improving the reform of the fiscal system, adjusting and optimizing the fiscal expenditure structure, and promoting the equalization of basic public services in the new era.

In view of this, this paper theoretically explores the influence mechanism of fiscal decentralization and local government competition on basic public services from the perspective of fiscal systems below the provincial level. At the empirical level, based on the panel data of prefecture-level cities in China from 2008 to 2019, this paper uses the fixed effect model to empirically test the impact of fiscal decentralization on the supply of public services of basic education, medical and health care, and social security. Moreover, the moderating effect model and threshold model are used to analyze the nonlinear impact of local government competition aiming at prefecture-level cities' economic growth on the relationship between fiscal decentralization and basic public services. The purpose of this paper is to establish a clear relationship between the financial and administrative power of the central and local governments, and improve the incentive system for officials, so as to alleviate the contradiction of vertical fiscal imbalance and improve the supply level of public services for people's livelihood in the new era. The possible marginal contributions of this paper are as follows: first, compared with most previous studies at the provincial level, this paper extends the research object of fiscal decentralization to the prefecture-level governments, which is the core providers of public services; Second, the selection of government competition indicators is more accurate, using the local economic growth target data manually collated from the work reports of China's prefecture-level municipal governments all over the country; third, in the setting of the model, considering both linear and nonlinear effects, the two-way fixed effect model is combined with the threshold model, and the instrumental variables are used to solve the endogeneity problem; fourth, unlike the existing literatures that analyze the local governments as a homogeneous whole, this paper considers the differences caused by resource endowments, and analyzes the regional heterogeneous effects of fiscal decentralization and government economic competition on public service provision, in order to make the research findings and policy recommendations more relevant.

The remainder of this paper is organized as follows: Section 2 summarizes the literatures, and consists of the theoretical analysis and research hypotheses; Section 3 introduces the model, variable, and data; After that, Section 4 presents results and discussion of empirical test; Section 5 is the heterogeneity analysis; And Section 6 elaborates the conclusions and policy recommendations of this paper.

2. Literature review, theoretical analysis, and research hypothesis

2.1. Literature review

Traditional fiscal decentralization theory suggests that, compared with the central government, local governments are closer to voters and have information advantages such as easy identification of residents' needs and preferences and understanding of the cost of public service provision in their jurisdictions [4]. The Tiebout model introduces competition into government departments, proving that local competition under decentralization can motivate local governments to improve the efficiency of public goods supply based on voters' preferences, and residents can use “vote with their feet” to force local governments to allocate welfare resources such as education and healthcare effectively [5]. However, many of the assumptions of Tiebout's model cannot be met simultaneously in reality, and mainstream public economics after Tiebout has revised its problems [[6], [7], [8]]. Moreover, other scholars proved that only by reasonably matching tax and expenditure responsibilities and establishing a sound accountability mechanism, fiscal decentralization is conducive to efficiently allocating public resources and improving public welfare for residents [9]. A large number of relevant studies show that, inter-governmental “yardstick competition” under decentralized systems allows residents to evaluate government efficiency in their local government with reference to other regions, thus prompting local governments under decentralization to optimize government expenditure and provide public services with significant regional differences in pursuit of welfare maximization [10]. Luis and Enric (2019) found that fiscal and political decentralization positively affects the quality of public service delivery, using the micro-educational data from 22 countries [11]. Christian (2009) conducted his research from the perspective of regional disparities, based on cross-section and panel data of twenty-three OECD countries, the result is that poor regions appear to benefit from fiscal decentralization rather than suffer from it [12]. Similarly, Tiang boho (2019) finds that fiscal decentralization enhances citizens' access to public services and reduces poverty, mainly through enhancing access to education, by evidence from Côte d'Ivoire municipalities in a conflict setting [13]. However, other case-based studies have found that local governments under fiscal decentralization engage in fierce tax competition, which will increase infrastructure investment and reduce the proportion of livelihood-based public service spending, resulting in systematic distortions in the structure of fiscal expenditure [14]. Therefore, the negative impact of government actions on public service provision cannot be ignored, especially in developing countries [15,16].

Compared with fiscal decentralization based on federalism, Chinese-style fiscal decentralization is established on a framework of delegated responsibilities between the central and local governments, forming a politically centralized and economically decentralized institutional arrangement with the typical color of Chinese administrative decentralization. The government behavior influences the relationship between Chinese-style fiscal decentralization and public service provision. Chinese scholars conducted systematic research on the relationship between fiscal decentralization and public service supply. Currently, there are three main viewpoints: (1) fiscal decentralization promotes the supply of basic public services. Empirical studies by Sun and Andrews (2022) suggest that expenditure and revenue decentralization from provincial to sub-provincial governments positively impact healthcare efficiency, and it is stronger in wealthier provinces, using the panel data for 2006–2017 in China [17]. (2) However, fiscal decentralization has a restraining effect on the supply of basic public services to some extent. Chen and Li (2018) and Gong and Lu (2003) have empirically tested the adverse effects of fiscal spending decentralization on the efficiency of healthcare service allocation and the expenditure of minimum living guarantee. The intensified government investment competition brought about by decentralization aggravated the adverse effects [18,19]. (3) There is a complex relationship between the impact of fiscal decentralization on the provision of basic public services. Scholars have used different indicators to measure the degree of fiscal decentralization and obtained inconsistent results. zhan and Liu(2020) constructed a comprehensive evaluation index system of public service quality, and found that the relationship between fiscal decentralization and public service quality follows an inverted U-shaped curve [20].

Local government competition is also an important factor affecting the supply of public services. Local government competition is also an important factor affecting the provision of public services. Competition among local governments often lies in economic growth targets, but this preference for economic spending has led to the neglect of non-economic public services related to people's welfare, such as education, health care and social security. Moreover, the vertical delegated agency of “top-down” decentralization in China makes it difficult for residents and authorities in the jurisdictions to fully play the role of supervision over the government at the same level, resulting in a severe disconnect between the supply structure of local public goods and the demand structure of the residents in the jurisdictions [20]. The academia has not reached a consensus on the relationship among fiscal decentralization, local government competition and public services. some scholars tend to use linear and nonlinear models to study their relationship. As for the empirical methods, the fixed effect model is usually used to test linear relationships, and the threshold model, quantile regression and moderating effect model are used to study the nonlinear relationship [21,22]. The research fields involve many aspects, such as Trade openness, carbon emissions, environmental pollution [23,24]. In the field of fiscal decentralization and public service, local government competition can be used as a threshold variable to study the nonlinear relationship between fiscal decentralization and public service, and it can also be used as a moderating variable to test whether it has a strengthened or weakened regulatory effect. For example, Xu and Lin(2022) found that fiscal decentralization drives local governments to increase public health expenditure, and it has a nonlinear effect on public health, utilizing the fixed effect model and threshold model [25]. Scholars represented by Fu and Zhang (2007) studied the vertical delegation and agency of “top-down” decentralization and found that the degree of fiscal decentralization has exacerbated the distortion of the public expenditure structure characterized by “prioritizing basic construction over public services” by affecting the economic competition among regions [3].

Overall, academics have systematically studied the effects of fiscal decentralization and local competitive behaviors on public service supply structure, scale, and efficiency from different perspectives and specifically analyzed their impact mechanisms at a deeper level according to China's reality, providing many experiences and insights for this article. However, In terms of research objectives, the majority of studies on China's fiscal decentralization are based on the provincial level, and there are fewer studies on the sub-provincial level or county level. In the aspect of research methods, some scholars used the linear models to test the relationship between fiscal decentralization and public services, only considering linear models in the absence of a theoretical basis will lead to the bias of empirical results. Therefore, this paper attempts to further enrich the existing research in terms of research objects, research methods and data. From the perspective of fiscal decentralization below the provincial level, using the local economic growth target data manually collated from the work reports of China's prefecture-level municipal governments, this paper analyzes the effects of fiscal decentralization and government economic competition on public service provision.

2.2. Theoretical analysis and research hypothesis

China has a combination of fiscal decentralization and centralized political system. Local governments are subjected to the pressure of fiscal vertical imbalance brought by the upward shift of fiscal power and decentralization of affairs under the fiscal decentralization system，and the stress of performance assessment under the political centralization system. Under the influence of the dual pressure, the impact of fiscal decentralization and local government economic competition on public service provision presents a complex relationship.

From the aspect of fiscal decentralization. The increase of fiscal decentralization can bring more financial autonomy to local governments, and the “moral person” attribute of local governments requires them to perform the function of a “service-oriented” government and provide undertakings and guarantees for people's livelihood [26]. However, the tax-sharing reform has redistributed the local tax sources and the tax-sharing ratio between the central and local government, and the local tax revenue mainly comes from enterprise value-added tax and income tax. When the government faces the more significant vertical fiscal imbalance of “less money and more things”, in order to obtain more tax revenue and ease the fiscal deficit, the “rational economist” attribute of local governments tends to invest more financial resources in productive infrastructure and gradually weaken the supply of basic public services [27]. From the perspective of government economic competition. The most direct manifestation of local government competition is the “GDP race” [28]. The reasonable setting of local GDP targets can provide local governments with abundant tax revenues, which in turn provide sufficient funds to invest in public services. However, in the face of the pressure of political achievement assessment, coupled with the limited number of political promotion opportunities and the short term of office, the “political person” attribute of local governments tends to set higher regional economic growth targets, and officials in neighboring regions are often prone to vicious competition [29]. Since economic infrastructure such as electricity, transportation, and communication directly enter the production function of local officials during their term of officials, and they are highly efficient and contribute directly to economic growth in the short term, local governments tend to invest more in productive construction than non-economic public services such as basic education, health care, social security, and public culture, which have spillover and time lag. In order to pursue rapid GDP growth to improve political performance, it will even lead to local protectionism and duplicate construction [3]. At the same time, with the increase of fiscal autonomy, local governments have more rights and freedom to adjust the structure of fiscal expenditures, which further leads to the distortion of the fiscal expenditure structure that is characterized by “emphasizing economic construction” and “neglecting people's livelihood".

Accordingly, this paper puts forward research hypothesis (1): the supply level of local basic public services shows a downward trend with the increase of fiscal decentralization, and the competition of local governments aiming at economic growth will intensify the negative impact of fiscal decentralization on the supply of public services.

When local governments set economic growth targets, they are motivated or constrained by national macroeconomic policies and the fiscal decentralization policies of higher-level governments. They then combine the region's actual resource endowments and economic development stages to formulate annual economic growth targets and publish investment priorities, industry policies, and fiscal support policies [30,31]. Since local economic growth targets are set differently, local governments may exhibit differentiated competitive behaviors and degrees of competition, which are manifested in different degrees of public service provision. Therefore, the relationship between fiscal decentralization and regional basic public service provision may show nonlinear characteristics.

Thereby, research hypothesis (2) is proposed: due to the different degrees of economic competition among local governments, fiscal decentralization has a nonlinear effect on local basic public service supply.

3. Model, variable, and data

3.1. Model

The two-way fixed effect model can take into account the interaction between time and individuals while controlling the influencing factors, providing a more comprehensive and accurate interpretation of data analysis. According to the purpose of this paper, Referring to Li and Qi's method [32], the benchmark regression model of fiscal decentralization and basic public service supply is constructed as follows:

$$Y_{i,t}={\alpha }_0+{\alpha }_{1}fi{d}_{i,t}+{\alpha }_2{X}_{i,t}+{\mu }_i+{\nu }_t+{\epsilon }_{i,t}$$ (1)

$Y_{i,t}$ represents the expenditure level of basic public services of city i in year t, including edu_it, hos_it, ins_it, respectively representing the expenditure level of education, medical and health services and social security. In order to eliminate the influence brought by the difference of population in different cities, the basic public service supply level is expressed by the per capita amount and taken as the natural logarithm. $fi{d}_{i,t}$ denotes the degree of fiscal decentralization of the prefecture, $X_{i,t}$ are the control variables, including $in{d}_{i,t}$、 $fi{n}_{i,t}$、 $ta{x}_{i,t}$、 $te{c}_{i,t}$、 $h{r}_{i,t}$, respectively representing the level of industrial structure, financial development, regional macro tax rate, strength of science and technology investment, and human capital. ${\mu }_i$ refers to individual fixed effect, ${\nu }_t$ refers to time fixed effect, and ${\epsilon }_{i,t}$ refers to random disturbance term.

According to research hypothesis (1), each prefecture-level city's public service supply is directly affected by fiscal decentralization and local government competition, which caused by fiscal decentralization may exacerbate the negative relationship between the degree of fiscal decentralization and basic public services. Therefore, it is necessary to add prefecture-level city government competition ($ec{u}_{i,t}$) to the benchmark model to further analyze the comprehensive effects of fiscal decentralization and government competition on the level of local government basic public service supply. Model (2) adds local government competition $ec{u}_{i,t}$ to the baseline model based on model (1). Model (3) introduces the interaction term of fiscal decentralization and government competition based on model (2) to further explore the moderating effect of local government economic competition on the relationship between fiscal decentralization and public service provision. The specific model is shown in the following equation:

$$Y_{i,t}={\alpha }_0+{\alpha }_{1}fi{d}_{i,t}+{\alpha }_{2}ec{u}_{i,t}+{\alpha }_3{X}_{i,t}+{\mu }_i+{\nu }_t+{\epsilon }_{i,t}$$ (2)

$$Y_{i,t}={\alpha }_0+{\alpha }_{1}fi{d}_{i,t}+{\alpha }_{2}ec{u}_{i,t}+{\alpha }_{3}fi{d}_{i,t}*ec{u}_{i,t}+{\alpha }_4{X}_{i,t}+{\mu }_i+{\nu }_t+{\epsilon }_{i,t}$$ (3)

$ec{u}_{i,t}$ refers to the competitive behavior of local governments of city i in year t, specifically expressed as the difference between the economic growth target in the government work report of the current year and the actual economic growth rate of the previous year. $fi{d}_{i,t}*ec{u}_{i,t}$ refers to the interaction term between fiscal decentralization and the competitive behavior of local governments.

The threshold model is proposed for non-dynamic panel data analysis with specific individual characteristics, and it can classify the sample according to the values of the observed variables. The threshold model is mainly to investigate the nonlinear relationship between explained variables and explanatory variables in the panel model by estimating the appropriate threshold value to observe whether the estimated coefficients of important explanatory variables will change significantly. To further test research hypothesis (2), Hansen's (1999) panel threshold model was selected based on equation (1), using the difference between the current year's local economic growth target and the previous year's growth target as the threshold variable to estimate the heterogeneous effect of fiscal decentralization on the supply of basic public service, with the dynamic change of the threshold variable. The model for the nonlinear effect of fiscal decentralization on the supply level of basic public services is constructed as follows:

$$Y_{i,t}={\alpha }_{1}fi{d}_{i,t}I\left(Z_{i,t}\le {\lambda }_1\right)+{\alpha }_{2}fi{d}_{i,t}I\left({\lambda }_1<Z_{i,t}\le {\lambda }_2\right)+\cdots +{\alpha }_{n}fi{d}_{i,t}I\left(Z_{i,t}>{\lambda }_n\right)+\beta X_{i,t}+u_i+v_t+{\epsilon }_{i,t}$$ (4)

I (·) denotes the threshold indicator function, and the variable Z in parentheses is the specific threshold variable, $\lambda$ is the threshold value. I (·) takes 1 when the condition in parentheses holds. Otherwise, it takes 0.

3.2. Variable

The explained variable is the expenditure of basic public services (ps), which is the sum of three key indicators, namely per capita local government expenditure on basic education (edu), medical and health care (hos), and social security and employment (ins). It should be noted this paper chooses input indicators, the fiscal expenditures related to the livelihood area of prefecture-level cities, to measure the level of basic public service provision rather than the output index (such as the enrollment rate and the per capita number of beds in welfare homes). This approach not only avoids the impact caused by the difference in the efficiency of transforming people's livelihood expenditure into people's livelihood public goods in prefecture-level cities but also eliminates the deviation caused by the difference in the supply quantity and supply capacity of public goods by private enterprises or non-profit organizations in prefecture-level cities. This approach not only avoids the effects of the variability in the efficiency of livelihood-related expenditures into livelihood-related public goods across prefectures but also eliminates the bias caused by the differences in the quantity and supply capacity of public goods supplied by private enterprises or non-profit organizations across prefectures. In addition, since the supply of basic public services involves many elements, only some output indicators may not represent the real level of basic public services in each prefecture-level city.

Fiscal decentralization refers to giving local governments certain taxation power and scope of expenditure responsibility, allowing local governments to independently decide the scale and structure of budget expenditure and entitle them to relative autonomy to manage local affairs. Referring to Fu and Zhang's research [3], this paper chooses the degree of fiscal autonomy (fid), the specific algorithm is the ratio of per capita general budget revenue to per capita general budget expenditure of prefecture-level cities, to represents the degree of fiscal decentralization, and the larger the value the greater the degree of fiscal decentralization of local governments. In order to eliminate the influence caused by the difference in population in different cities, fiscal revenue and expenditure are expressed in per capita amount. Since the current pattern of fiscal funds allocation in China is still mainly linked to the number of household registration population in the location, the measurement of fiscal decentralization degree is based on the number of household registration population.

Referring to Zhan and Liu(2020) [22], this paper chooses the economic growth target of prefecture-level cities to represent the degree of local economic competition (ecu). Specifically, it is measured by the difference between the economic growth target of the current year and the actual economic growth rate of the previous year in prefecture-level cities. Compared with specific governmental behavior such as fixed asset investment and foreign direct investment, the economic growth target of prefecture-level municipalities is a more comprehensive, accurate, and representative measure of the competitive behavior among local governments.

In addition to the degree of fiscal decentralization and government economic competition, many other factors affect the supply of regional basic public service. Referring to Chen and Li (2018) [19], to make the obtained results more robust and reliable, this paper introduces the level of industrial structure (ind), the level of financial development (fin), the regional macro tax rate (tax), science and technology investment (tec) and human resources (hr) as control variables.

3.3. Data

This paper uses panel data of 178 prefecture-level cities from 2008 to 2019. Since the statistical caliber of fiscal expenditures in China changed significantly in 2007, in order to avoid the impact on the research results due to the different statistical caliber, the sample time period selected for this paper is 2008–2019. Based on the accessibility of public finance expenditure data at the prefecture and city level, the study sample includes only 178 prefecture-level cities in Inner Mongolia, Liaoning, Jilin, Heilongjiang, Jiangsu, Anhui, Shandong, Henan, Hunan, Guangxi, Sichuan, Guizhou, and Gansu. The data were mainly obtained from the China Regional Economic Statistical Yearbook (https://data.cnki.net/yearBook/single?id=N2015070200), China City Statistical Yearbook (https://data.cnki.net/yearBook/single?id=N2020050229), EPS databases (http://olap.epsnet.com.cn/auth/platform.html?sid=195238B7ECA4A6A59B9FFE42BDECA5A0_ipv409845072), and the work reports of prefecture-level municipal governments. The data on basic public service expenditure and fiscal decentralization were obtained from EPS database and provincial and municipal statistical yearbooks. Meanwhile, each prefecture-level city's economic growth target data were obtained by manually collating 2136 government work reports of prefecture-level city governments from 2008 to 2019.

The missing data of individual regions and years were supplemented by the interpolation method. In order to exclude the influence of demographic factors on the correlation, the variables are processed per capita. Meanwhile, in order to eliminate possible heteroskedasticity in the regression analysis, the indicator data of the explained variables were taken as natural logarithms. The descriptive statistics of the raw data of each variable in the model are shown in Table 1.

Table 1.

Descriptive statistics of variables.

	Variables	Obs	Mean	Std. Dev.	Min	Max
Explained variables	Public service (ps)	2136	906.3514	464.8051	158.7301	3348.427
	Public basic education (edu)	2136	1147.2510	588.3717	153.253	4829.740
	Medical and health care (hos)	2136	586.8396	311.2988	73.3214	2453.430
	Social security and employment services (ins)	2136	985.8595	675.5836	36.6445	4631.570
Core explanatory variables	the degree of fiscal autonomy (fid)	2136	0.4426	0.2177	0.0932	1.0168
Moderating variables	the degree of local economic competition (ecu)	2136	−0.0691	2.4626	−6.73	10.94
Control variables	Level of industrial structure (ind)	2136	86.2445	8.0851	60.2300	98.6700
	the level of financial development (fin)	2136	1.2448	0.4799	0 0.5019	2.9406
	the regional macro tax rate (tax)	2136	0.0711	0.0238	0.0292	0.1351
	science and technology investment (tec)	2136	1.3879	1.2521	0.1276	6.9302
	human resources (hr)	2136	1.5736	1.8596	0.000	9.7638

4. Empirical analysis

4.1. Empirical test of the comprehensive impact of fiscal decentralization and government competition on the supply of basic public services

This paper uses a panel two-way fixed-effects model for regression estimation. Table 2 reports the results of the baseline regression. It can be seen that the coefficients of fiscal decentralization on the overall supply level of basic public services and the supply levels of various types of public services such as basic education, medical and health care, and social security are all significantly negative at the level of 1%, indicating that the supply level of basic public services shows a decreasing trend with the increase of fiscal decentralization, which is consistent with hypothesis (1) above. This result is similar to the research of Li et al., （2020） [32] Comparing the regression coefficients of various types of public services, the degree of fiscal decentralization has the largest negative impact on social security, followed by the level of health care and the level of education. Among the control variables, industrial structure, regional macro tax rate, the level of scientific and technological importance, and the level of human capital positively influence the level of basic public service provision, and the level of financial development has a negative influence on it.

Table 2.

Baseline regression results.

Variables	(1)	(2)	(3)	(4)
	lnps	lnedu	lnhos	lnins
fid	−0.466*** (−9.043)	−0.232*** (−3.391)	−0.239*** (−3.448)	−0.835*** (−9.234)
ind	0.011*** (10.963)	0.017*** (12.293)	0.018*** (12.794)	0.002 (1.207)
fin	−0.073*** (−5.673)	−0.167*** (−9.750)	−0.086*** (−4.959)	−0.072*** (−3.191)
tax	3.026*** (11.202)	2.908*** (8.111)	1.775*** (4.877)	3.983*** (8.407)
tec	0.011*** (3.225)	0.015*** (3.323)	0.009* (1.867)	−0.001 (−0.140)
hr	0.017*** (2.821)	0.023*** (2.894)	0.024*** (2.953)	0.028*** (2.602)
Constant	4.939*** (55.044)	4.773*** (40.073)	3.591*** (29.705)	5.909*** (37.551)
City FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
Obs	2136	2136	2136	2136

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

Table 3 reports the results of the effects of fiscal decentralization and government competition on the overall supply level of basic public services and the supply level of each classification of public services. Among them, models (1), (3), (5), and (7) represent the regression results after adding government economic competition to the original model. Models (2), (4), (6), and (8) represent the regression results after adding government competition and the interaction term of fiscal decentralization and government competition to the original model.

Table 3.

Moderating effect test results.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	lnps	lnps	lnedu	lnedu	lnhos	lnhos	lnins	lnins
fid	−0.461*** (−8.964)	−0.472*** (−9.173)	−0.225*** (−3.291)	−0.233*** (−3.402)	−0.233*** (−3.366)	−0.249*** (−3.589)	−0.834*** (−9.222)	−0.839*** (−9.257)
ecu	−0.003*** (−2.959)	0.002 (1.043)	−0.004*** (−3.403)	−0.001 (−0.391)	−0.004*** (−2.720)	0.003 (1.380)	−0.000 (−0.166)	0.002 (0.625)
fid*ecu		−0.011*** (−3.292)		−0.008* (−1.854)		−0.016*** (−3.538)		−0.005 (−0.871)
ind	0.011*** (10.538)	0.011*** (10.492)	0.016*** (11.815)	0.016*** (11.780)	0.018*** (12.383)	0.017*** (12.342)	0.002 (1.177)	0.002 (1.158)
fin	−0.066*** (−5.033)	−0.069*** (−5.249)	−0.156*** (−8.971)	−0.158*** (−9.078)	−0.077*** (−4.372)	−0.081*** (−4.607)	−0.071*** (−3.103)	−0.073*** (−3.153)
tax	2.974*** (11.011)	3.001*** (11.134)	2.829*** (7.897)	2.850*** (7.955)	1.711*** (4.700)	1.750*** (4.819)	3.978*** (8.377)	3.990*** (8.399)
tec	0.011*** (3.193)	0.012*** (3.307)	0.015*** (3.288)	0.015*** (3.349)	0.009* (1.835)	0.009* (1.955)	−0.001 (−0.142)	−0.001 (−0.114)
hr	0.017*** (2.786)	0.017*** (2.873)	0.023*** (2.855)	0.023*** (2.902)	0.024*** (2.921)	0.025*** (3.016)	0.028*** (2.599)	0.028*** (2.620)
Cons	4.961*** (55.212)	4.970*** (55.430)	4.806*** (40.326)	4.813*** (40.390)	3.618*** (29.877)	3.632*** (30.065)	5.912*** (37.430)	5.916*** (37.436)
City FE	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes
Obs	2136	2136	2136	2136	2136	2136	2136	2136

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

From the overall regression results, the coefficient of ecu in the model (1) is significantly negative at the 1% level, indicating that the supply of basic public services tends to decrease with the increase of local economic competition. The coefficient of the interaction term (fid*ecu) between government competition and fiscal decentralization in the model (2) is −0.011, which passes the significance test at the 1% level, and the coefficient of fid is −0.472, which decreases compared to the coefficient of −0.466 in Table 2, indicating that the negative effect of fiscal decentralization on the supply of basic public services will be exacerbated by the increase of government competition, which justifies the research hypothesis (1). Similar conclusions have also been reached in Fu and Zhang's study, which found that under the background of Chinese-style fiscal decentralization, the economic competition of local governments will lead to the structural distortion of fiscal expenditure “emphasizing economy over people's livelihood [3]. A specific comparison of the interaction term coefficients of various types of public services shows that the interaction between government economic competition and fiscal decentralization has the greatest impact on the level of medical and health expenditure, followed by the level of education expenditure, but the impact on social security expenditure is not obvious. The possible reason is that social security is more basic than education and health care.

4.2. Robustness test

Considering the endogeneity problem, this paper selects the instrumental variables that are more correlated with fiscal decentralization but less correlated with the level of per capita expenditure on basic public services and uses the method of two-stage least squares (2SLS) to estimate the correlation between fiscal decentralization and basic public services, choosing the indicator of population density as the specific instrumental variables. The corresponding two-stage least squares (2SLS) estimation results are given in Table 4. The results show that the Kleibergen-Paap rk LM and Kleibergen-Paap Wald rk F tests reject the original hypothesis of insufficient and weak identification of instrumental variables. It shows that there is a strong correlation between the instrumental variables and the potential endogenous variables, and the instrumental variables selected in this paper are more reasonable. As shown by the estimation results in Table 4, the degree of fiscal decentralization still shows a significant negative effect on the supply of basic public services after taking into account the possible endogeneity problems. Therefore, the core conclusions of this paper are robust.

Table 4.

Endogeneity test results.

Variables	(1)	(2)	(3)	(4)
	lnps	lnedu	lnhos	lnins
fid	−3.211***	−2.069***	−2.281***	−5.080***
	(-4.034)	(-3.008)	(-4.148)	(-4.532)
ind	0.050***	0.038***	0.034***	0.072***
	(4.671)	(4.330)	(4.487)	(4.576)
fin	−0.155	−0.175*	−0.154**	−0.196
	(-1.391)	(-1.920)	(-2.024)	(-1.182)
tax	12.280***	8.673***	8.355***	18.811***
	(3.877)	(3.218)	(3.882)	(4.097)
tec	0.123***	0.117***	0.095***	0.146**
	(2.779)	(3.016)	(3.342)	(2.338)
hr	0.083***	0.070***	0.068***	0.117**
	(2.607)	(2.664)	(3.145)	(2.508)
cons	2.291***	3.334***	2.672***	0.959
	(3.180)	(5.753)	(5.251)	(0.878)
Kleibergen-PaaprkLM	30.706***	30.706***	30.706***	30.706***
Kleibergen-PaapWaldrkF	36.272 [16.38]	36.272 [16.38]	36.272 [16.38]	36.272 [16.38]
City FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
Obs	2136	2136	2136	2136

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

By reducing the sample size, this paper tests the robustness of the model and result. Table 5 shows the estimated results after the provincial capital cities are excluded, and Table 6 shows the estimated results after the first and last years are excluded, the results show that the signs, significance levels, and numerical trends of the estimated coefficients of the core explanatory variables and control variables are basically consistent with the regression results of the benchmark sample, which further indicates that the regression results of the original model are relatively robust.

Table 5.

Robustness test: regression results after excluding provincial capital.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	lnps	lnps	lnedu	lnedu	lnhos	lnhos	lnins	lnins
fid	−0.493***	−0.498***	−0.251***	−0.253***	−0.215***	−0.222***	−0.911***	−0.914***
	(-8.984)	(-9.094)	(-3.519)	(-3.541)	(-3.177)	(-3.281)	(-9.514)	(-9.519)
ecu		0.002		−0.000		0.002		0.001
		(1.016)		(-0.096)		(1.081)		(0.422)
fid*ecu		−0.011***		−0.010**		−0.013***		−0.004
		(-3.058)		(-2.116)		(-2.998)		(-0.610)
cons	4.905***	4.936***	4.711***	4.754***	3.655***	3.692***	5.853***	5.859***
	(54.592)	(54.911)	(40.275)	(40.598)	(32.925)	(33.225)	(37.324)	(37.174)
Control	yes	yes	yes	yes	yes	yes	yes	yes
City FE	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes
Obs	1968	1968	1968	1968	1968	1968	1968	1968

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

Table 6.

Robustness test: Regression results after excluding the first and last years.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	lnps	lnps	lnedu	lnedu	lnhos	lnhos	lnins	lnins
fid	−0.452***	−0.456***	−0.302***	−0.300***	−0.257***	−0.258***	−0.703***	−0.712***
	(-8.886)	(-9.002)	(-4.066)	(-4.050)	(-3.256)	(-3.272)	(-7.954)	(-8.047)
ecu		0.002		−0.001		0.000		0.005*
		(1.448)		(-0.613)		(0.128)		(1.897)
fid*ecu		−0.011***		−0.008*		−0.010**		−0.010*
		(-3.603)		(-1.808)		(-2.131)		(-1.864)
cons	5.069***	5.097***	4.814***	4.865***	3.820***	3.863***	5.978***	5.971***
	(49.414)	(49.752)	(32.093)	(32.448)	(23.994)	(24.244)	(33.512)	(33.355)
Control	yes	yes	yes	yes	yes	yes	yes	yes
City FE	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes
Obs	1780	1780	1780	1780	1780	1780	1780	1780

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

4.3. An empirical test of the nonlinear effect of fiscal decentralization on the supply of basic public services

It has been confirmed above that the competition behavior of local governments has a moderating effect on the impact of fiscal decentralization on the supply of basic public services. Does local government competition represented by local economic growth targets have a threshold characteristic when the moderating effect occurs? To test hypothesis (2), this paper further uses a panel threshold model to examine the nonlinear effect of fiscal autonomy on the supply of basic public services when the difference between the regional economic growth target and the previous year's economic growth rate changes. According to the F-statistic and the p-value obtained by the “self-sampling method” given in Table 7, it can be seen that when the regional economic competition is taken as the threshold variable, its single threshold effect on the overall public service, education, and medical and health expenditure level has passed the significance test, and there is no threshold effect on social security, which again proves that social security is more livelihood.

Table 7.

Threshold significance test and threshold estimation under different threshold variables.

Variables	Number of thresholds	F	P	BS	threshold estimate	95% confidence interval
lnps	A single threshold	17.65	0.0267	300	M1 (1) = 1.200	[0.725，1.330]
	Double threshold	9.61	0.1033	300
	The triple threshold	2.63	0.8467	300
lnedu	A single threshold	21.45	0.0000	300	M1 (1) = 1.200	[0.960，1.330]
	Double threshold	4.05	0.5633	300
	The triple threshold	3.55	0.5333	300
lnhos	A single threshold	24.21	0.0067	300	M1 (1) = −2.600	[-2.900，-2.540]
	Double threshold	4.05	0.5233	300
	The triple threshold	2.56	0.8600	300
lnins	A single threshold	10.35	0.1867	300	–	–
	Double threshold	8.64	0.1400	300
	The triple threshold	3.15	0.7100	300

Note: The BS column indicates 300 bootstrap sampling times.

After determining the single threshold effect of the economic competition of prefecture-level municipal governments, the regression results of the threshold model are shown in Table 8. From the regression results, it can be seen from the regression results that the degree of fiscal decentralization has a significant nonlinear effect on the level of basic education, medical and health care, and the overall public service supply, which shows that the regression coefficients of fiscal decentralization on the expenditure levels of basic education, health care, and overall public services become smaller as the degree of economic competition increases. Taking the overall supply level of basic public services as an example, when the difference between the current year's economic growth target of prefecture-level cities and the last year's actual growth rate is less than 1.2%, the coefficient of fiscal decentralization on the supply of basic public services is −0.457; when the difference is greater than 1.2%, the coefficient decreases to −0.514. This shows that the larger the difference between the current year's regional economic growth target and the last year's actual growth rate, the more fiscal decentralization's negative impact on basic public services. This finding is consistent with Zhan and Liu's research [20]. In the context of China's reality, when local governments face greater pressure of economic competition, the target of regional economic growth will be set too high, and local governments cannot afford the dual tasks of promoting economic growth and improving public service provision. With the expansion of fiscal autonomy, a structural distortion of “heavy infrastructure construction and light people's livelihood” will occur.

Table 8.

Regression results of the panel threshold model with different threshold variables.

Threshold variable	lnps	Threshold variable	lnedu	Threshold variable	lnhos
ecu≤1.200	−0.457***	ecu≤1.200	−0.219**	ecu ≤ −2.600	−0.155
	(-4.391)		(-2.096)		(-1.546)
ecu>1.200	−0.514***	ecu>1.200	−0.302***	ecu > −2.600	−0.254**
	(-4.901)		(-2.897)		(-2.442)

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

5. Heterogeneity analysis

Given the vast area of China and the differences in Geographical location and resource endowment, there may be regional differences in the effects of the degree of fiscal decentralization and government economic competition behavior on basic public services. Thus, further heterogeneity analysis is necessary. In this paper, all prefecture-level cities are divided into four categories based on their geographical locations: western, central, eastern, and northeastern, to conduct Regional heterogeneity analysis. Moreover, according to the National Urban Comprehensive Strength Ranking 2021 published by the Urban Development Institute, prefecture-level cities are divided into new first-tier, first-tier, second-tier, second-tier, third-tier, fourth-tier, and fifth-tier cities. Due to the small number of new first-tier and first-tier cities in the sample, considering the balance of sample size among different groups, this paper combine new first-tier and first-tier cities into one large category (first-tier cities) and explore the heterogeneity in different cities.

As can be seen from Table 9, in the eastern, central, and western regions, the influence coefficients of fiscal autonomy on the supply of basic public services are significantly negative at the level of 1%, indicating that the improvement of fiscal decentralization has an inhibitory effect on the level of basic public service expenditure. Comparing the coefficients reveals that the Western region is the largest, followed by central and eastern regions. After government economic competition and its interaction term with fiscal decentralization are added, the coefficients of the interaction term in the eastern, central, western, and northeastern regions are all negative. The western and northeastern regions pass the significance test of 10% and 1%, respectively. However, the eastern and central regions fail to pass the significance test, indicating that the interaction effect between economic growth target and fiscal decentralization is more obvious in the western and northeastern regions.

Table 9.

Regional heterogeneity analysis.

	East	East	Middle	Middle	West	West	Northeast	Northeast
	lnps	lnps	lnps	lnps	lnps	lnps	lnps	lnps
fid	−1.079***	−1.096***	−0.268***	−0.277***	−0.550***	−0.559***	−0.126	−0.105
	(-7.913)	(-8.081)	(-2.926)	(-3.024)	(-5.714)	(-5.808)	(-1.291)	(-1.110)
ecu		0.014		0.011		0.003		0.006**
		(0.822)		(1.578)		(1.346)		(2.339)
fid*ecu		−0.033		−0.021		−0.008*		−0.024***
		(-1.440)		(-1.620)		(-1.662)		(-4.514)
cons	3.333***	3.587***	4.140***	4.179***	5.256***	5.264***	6.539***	6.452***
	(9.115)	(9.485)	(25.464)	(24.813)	(34.515)	(34.570)	(35.258)	(35.340)
Control	yes	yes	yes	yes	yes	yes	yes	yes
City FE	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes
Obs	360	360	684	684	684	684	408	408

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

Table 10 shows that the influence coefficients of fiscal autonomy on basic public service provision in all classes of cities are negative, and all pass the significance test. After adding local economic competition and its interaction term with fiscal decentralization, the coefficients of the interaction terms for third-tier, fourth-tier, and fifth-tier cities are all significantly negative at the 5% level, indicating that economic competition in third-tier, fourth-tier, and fifth-tier cities exacerbates the negative impact of fiscal decentralization on basic public service provision in prefecture-level cities. On the contrary, the coefficients of the interaction terms of first-tier and second-tier cities are positive, indicating that the economic growth target set in first-tier and second-tier cities has a weakening effect on the negative relationship of fiscal decentralization and basic public services.

Table 10.

Prefecture-level cities' heterogeneity analysis.

	first--tier	first--tier	second-tier	second-tier	third-tier	third-tier	fourth-tier	fourth-tier	fifth-tier	fifth-tier
	lnps	lnps	lnps	lnps	lnps	lnps	lnps	lnps	lnps	lnps
fid	−0.709***	−0.559**	−0.494**	−0.525**	−0.560***	−0.575***	−0.630***	−0.685***	−0.345***	−0.306***
	(-3.053)	(-2.245)	(-2.336)	(-2.483)	(-5.373)	(-5.540)	(-6.680)	(-7.073)	(-3.817)	(-3.377)
ecu		−0.092*		−0.013		0.008		0.008*		0.000
		(-1.837)		(-0.537)		(1.095)		(1.664)		(0.063)
fid*ecu		0.116*		0.004		−0.026**		−0.020**		−0.009**
		(1.968)		(0.112)		(-2.015)		(-2.261)		(-2.087)
Cons	0.822	−1.271	3.268***	3.126***	4.872***	5.012***	3.356***	3.452***	5.513***	5.534***
	(0.317)	(-0.463)	(2.833)	(2.661)	(25.952)	(25.870)	(16.206)	(16.388)	(45.819)	(46.170)
Control	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
City FE	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Obs	96	96	168	168	516	516	576	576	780	780

Note: The numbers in parentheses are the corresponding t values; *, ** and *** indicate significance at 10%,5% and 1%, respectively.

Combined with the actual situation in China, the reason for the above situation is that the less developed regions are under greater pressure to develop their economy, which can be confirmed by the difference between the regional economic growth targets of the current year and the actual economic growth rate of the previous year (the descriptive statistics of the regions are not listed in the text due to the limitation of space). Specifically, the difference in the eastern, central, western and northeastern regions is −0.4885, −0.5415, −0.3060, and 1.5304, respectively, suggesting that the northeastern region is the largest, followed by Western and eastern and central regions. Similarly, the difference in first-tier, second-tier, third-tier, fourth-tier, and fifth-tier cities are −0.5274, −0.4787, −0.1708, and 0.4320 separately, revealed the fifth-tier city is the largest, followed by fourth-tier, third-tier, first--tier, and second-tier city. Generally, competition among local governments mainly includes four forms: investment, management rights, promotion, and public product supply. When regional livelihood construction reaches the minimum target, it is no longer subject to hard constraints of laws and policies, so public services are in a relatively unimportant position in regional competition. Governments tend to set higher economic growth targets for underdeveloped regions with weak foundations and high economic growth potential. Economic infrastructure such as electricity, transportation, and communication directly enter the production function of underdeveloped regions' officials during their term of office, which is highly efficient in the short term and makes direct contributions to economic growth. Meanwhile, it has strong additional benefits for attracting investment. However, non-economic public services such as basic education, health care, and social security serve local residents, and their input-output effects are asymmetric, time-lag, and spillover, and have a little external effect on attracting capital flows, which do not help much in the performance assessment of officials in the short term. Therefore, government competition in less developed regions has a significant inhibitory effect on the supply of livelihood-based public services and can exacerbate the negative impact of the degree of fiscal decentralization on basic public services. By contrast, the growth potential of GDP is small for developed regions, which have a better economic foundation and are affected by diminishing marginal utility. Compared with the pure pursuit of GDP, local governments are more inclined to take into account the areas of people's livelihood while developing the economy and promoting high-quality economic development to obtain political promotion. Therefore, the strengthening moderating effect of local government competition represented by the setting of economic growth target on the relationship between fiscal decentralization and public service supply is not significant in developed regions and may even weaken the negative relationship between fiscal decentralization and public service.

6. Main conclusions and policy recommendations

6.1. Main conclusions and discussion

This study explores the Impact of fiscal decentralization and local government competition on the supply of basic public services from the theoretical and empirical level, based on the panel data of 178 prefecture-level cities in China from 2008 to 2019, using the fixed effect model and threshold model. We found that fiscal decentralization has a negative impact on the supply of basic public services, and local government competition aiming at economic growth plays an strengthen regulatory role. Moreover, with the increase of economic growth target, the negative effect of fiscal decentralization on basic public services has nonlinear characteristics. In addition, the impact of fiscal decentralization and local government competition on public service supply has regional heterogeneity. In the less developed regions represented by west and northeast cities, third-tier, fourth-tier, and fifth-tier cities, local economic growth competition strengthens the negative relationship between fiscal decentralization and public services. However, the strengthen regulatory effect is not obvious in developed regions.

It can be seen from the above results that in the case of vertical fiscal imbalance and horizontal economic competition, the improvement of fiscal decentralization will cause the structure distortion of public expenditure, and reduce the investment in public services related to people's livelihood. The local governments competition triggered by regional fiscal pressure and performance assessment pressure exacerbates the negative impact. Therefore, it is very important to design a scientific and reasonable fiscal decentralization system and performance appraisal mechanism to ensure the ability of public finance to support basic public services.

6.2. Policy recommendations

The following policy recommendations are proposed based on the above findings.

Firstly, accelerate the reform of the fiscal and tax systems, build a relationship of financial power and authority between central and local governments with clear rights and responsibilities, coordinated financial resources, and introduce incentive and restraint mechanisms into the framework fiscal and tax system reform. On the one hand, it is necessary to reshape the vertical fiscal relationship of the government and scientifically divide the financial and administrative power of the central, provincial, and prefecture-level cities. The expenditure responsibilities of the central and provincial-level governments can be moderately increased, and the central government can supervise public service projects with strong externality, whereas those with weak externality can be supervised by local governments, reducing the financial pressure on the basic-level government. Narrow the fiscal revenue and expenditure gap faced by municipal and county-level governments and reduce the degree of vertical fiscal imbalance. On the other hand, considering the horizontal game and competition relationship between regions, a long-term mechanism for horizontal tax coordination and incentives should be established. Give local governments the authority to set taxes (such as resource tax and mineral tax) and tax rates, according to their own resource endowments, and promote a tax revenue sharing mechanism that matches the contribution of regional production factors. For example, the property tax being piloted can be taken as the main direction of local tax cultivation to broaden the source channels of local fiscal revenue. Moreover, improve the benefit sharing and risk compensation mechanism in the field of “public land” to promote healthy competition among regions with the goal of high-quality economic development.

Secondly, increase transfer payments to less developed areas and improve the mechanism for direct transfer payments. The reform of transfer payments should balance fairness and efficiency, increase fund coordination and tilt efforts, and focus on increasing the scale and proportion of transfer payments to fourth-tier and fifth-tier cities and impoverished areas in the western region with a large elderly and juvenile population. The provincial level should strictly implement the requirement of direct transfer funds to the grassroots and refine the specific plan of distributing all funds to cities and counties. At the same time, the city and county governments should coordinate the direct funds from higher levels and other funds, scientifically arrange public finance expenditures. The allocation of funds should closely focus on the central government's requirements for equalizing basic public services in urban and rural areas and strengthening people's shortcomings, highlight key areas, Furthermore, it is necessary to improve the specific implementation steps for direct funds: in the allocation stage, ensure that the time of issuing direct funds from the central government is consistent with the time of preparing local financial budgets; in the allocation stage, ensure that the time of the central direct fund plan is in line with the local financial budget preparation time; in the liquidation stage, establish a long-term management mechanism that can cross years for liquidation, enhancing the flexibility of grassroots finance; in the supervision and assessment stage, conduct accurate supervision and effective constraints according to the specific property and use of direct transfer funds, conduct performance assessments based on the budget year and project process, and establish corresponding reward and punishment mechanisms.

Thirdly, improve the political performance appraisal mechanism, formulate differentiated governmental performance evaluation criteria, establish dynamic tracking and assessment mechanism for officials, and promote the “bottom-up” assessment method. This paper finds that some regions (especially the less developed regions) overly pursue economic growth rate, which aggravates the negative impact of fiscal decentralization on the supply of basic public services. Although various livelihood and public services and ecological environment indicators have been included in the performance appraisal in recent years, the complex indicators may make the appraisal ineffective. Therefore, differentiated indicators and weights should be set according to the actual situation in different regions (such as ecological conservation areas and impoverished areas with a large elderly and juvenile population), and the minimum evaluation standards for compulsory education, medical care, and social security in underdeveloped areas should be raised to guide local governments to shift their competition from the GDP race to public service-oriented competition. What is more effective is to combine the “top-down” and “bottom-up” evaluation approaches because for governments below the provincial level, no one knows more about the performance of local officials than the residents in their jurisdictions, who rate the public opinion of the relevant departments based on their satisfaction and sense of gain from the public services. Besides, supervisory efforts should be strengthened, and channels of supervision should be expanded to fully leverage local residents' supervisory and restrictive role in implementing public services by local departments. For the public service fields such as basic education, which have a long input-output cycle and are difficult to achieve obvious results during the term of office of officials, a long-term tracking assessment system should be established to conduct all-round and multi-channel assessment of officials.

Finally, improve the mechanism for expressing public service needs and interests, broaden the expression mode, and perfect democratic decision-making and supervision mechanisms. Make full use of modern information technology such as Internet+, big data, and artificial intelligence to improve the information transparency of public service projects. A network information system for democratic decision-making on public services should be established, and a combination of online and offline methods should be used to solicit public opinion. This allows grassroots communities to participate in the decision-making process of basic public service supply through direct or indirect channels. Furthermore, an interactive feedback mechanism for public demands and suggestions should be constructed to enhance the precision of public service supply.

Funding statement

This study was supported by Humanities and Social Sciences Research Program of Chongqing Municipal Education Commission (22SKGH196); Postgraduate Research and Innovation Project of Chongqing (CYB22270); and the Research Projects of Macao Polytechnic University (grant number RP/ESCHS-03/2020 and RP/ESCHS-04/2020).

Ethical approval

This article does not contain any studies with human participants performed by any of the authors.

Data availability statement

Data will be made available on request.

CRediT authorship contribution statement

Yamei Wang: Writing – original draft. Xiuquan Huang: Writing – review & editing, Software, Data curation. Tao Zhang: Investigation. Bo Jiang: Methodology. Xi Wang: Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.