1. Introduction
China’ s urbanization rate had exceeded 60% by the end of 2020, which indicates the process of economic development and the transfer of industrial structure to non-agricultural industries. It is also accompanied by the continuous increasing in the urbanization rate [
1]. From the experience of developed countries, the core of urbanization is migration and agglomeration of labor and enterprises, while the expansion of urban space is the market-clearing equilibrium result of labor and land. The market mechanism is the main driving force for urban development in developed countries, and China’s rapid urbanization is characterized by government-led, large-scale and land-based public ownership [
2]. Rapid urbanization has not only contributed to the expansion of cities but also brought about serious energy consumption. It is estimated that China’s urban economy will account for 90% of the country’s gross domestic product (GDP) by 2025, and the urban energy demand will account for 85% of the country’s total energy demand. Every 1% increase in the urbanization rate will require an additional consumption of 60 million tons of standard coal [
3]. If energy consumption is not controlled, carbon emissions and air pollution (PM
2.5, nitrogen oxides, sulfur dioxide and other air pollutants) will continue to worsen, and the annual economic cost of air pollution in China is estimated at 1.2% of GDP, which is based on the cost of disease, and it will rise to 3.8% if it is based on willingness to pay [
4]. All of these factors mean that Chinese cities face the serious task of energy saving and consumption reduction, and it remains an open question whether government-led urban expansion can have a positive effect on reducing energy intensity.
China has implemented a series of administrative division adjustment policy after the reform and opening up in 1978, which aimed to promote industrialization and urbanization by adjusting interregional relations. In particular, the adjustment of administrative divisions has gradually become the main force in promoting the improvement of urbanization since the 1990s. Generally speaking, the increase in urbanization is mainly achieved through the growth of the number of cities and the expansion of the size of cities, both of which are closely related to the reorganization of administrative divisions in China: the creation of new cities (county-to-city upgrading) and the expansion of existing cities (city-county mergers). For the latter, city-county mergers are the abolition of counties (county-level cities) that were previously part of the prefecture-level city and the establishment of municipal districts in the administrative area of the former counties (county-level cities), thus achieve the purpose of expanding the size of the city. The small-town strategy was considered to be the most suitable urbanization path for China at the beginning of the reform and opening up in 1978 [
5]; this policy did not achieve the expected goal of promoting urbanization and industrialization, although a large number of small cities and towns were newly established through county-to-city upgrading [
6]. After the abolition of the county-to-city upgrading policy in 1997, the city-county merger policy began to be widely implemented. Are there also differences in the effects of these two types of policies in terms of reducing urban energy intensity? It has been a matter of debate which type of policy is more efficient between researchers and policy makers [
7]. Unlike county-to-city upgrading, city-county mergers essentially expand the boundaries of existing cities and allow the population size of cities to rise further. A total of 119 prefecture-level cities and 174 counties (county-level cities) participated in city-county mergers during the period of 2000–2017. From
Figure 1, we can see that there is a clear feature of uneven advancement of city-county mergers. Before 2000, the number of mergers was relatively small before 2000, and city-county mergers experienced two waves after 2000: the first wave was from 2000 to 2004, and there are 39 prefecture-level cities and 45 counties (county-level cities) which participated in the mergers. With the vigorous promotion of the new urbanization strategy, the second wave sprouted in 2011 and completely broke out in 2012–2017. A total of 78 cities have undergone 94 city-county mergers between 2011 and 2017. After this round of mergers, the number of Chinese cities has decreased, but the existing prefecture-level cities have realized a rapid expansion in population and land size, and the characteristics of population agglomeration in large cities have become more obvious. The city-county mergers will help optimize China’s urban space system and break down administrative boundaries between prefecture-level cities and the counties under their jurisdiction, which can achieve unified decision-making in urban energy planning.
City-county mergers have become an institutional choice for urban spatial expansion during a period of rapid urban development, with the aim of making effective use of economic scale and improving the scope and quality of public service provision [
8]. City-county mergers essentially reflected in the state power based on a specific geographical space and the country’s political organization structure, which is used to carry out regional division or element allocation. The large-scale city-county mergers have also shaped space and governance characteristics in the long-term development of Chinese cities [
9,
10]. The reorganization of administrative districts to meet the needs of urban expansion is not unique to China; developed countries such as the UK, Germany and France have optimized urban management in the context of rapid urbanization by reorganizing their administrative districts since the 1960s. The success of city-county mergers in developed countries is subject to referendum by the local people and therefore reflects the needs of the people and the market. In contrast, the process of administrative redistricting in China is government-led that may run counter to the laws of the market and easily lead to pseudo-urbanization and urban sprawl, which will result in county resources being seized by prefecture-level cities. In a positive sense, merger policies may have promoted economic linkages and market integration between counties and city districts, which helps to reduce administrative barriers and enhance urban agglomeration economies, and also can accelerate population urbanization and economic development [
11]. Therefore, whether the city-county mergers can promote the reduction in urban energy intensity remains to be tested empirically, which is the core motivation of this paper.
On the other hand, the state has gradually decomposed the binding targets for reducing energy intensity. starting with China’s 11th Five-Year Plan (FYP): the energy-saving indicators that reach the provinces are further decomposed into cities, and the cities further decompose the tasks into urban districts, so that urban districts become important units to implement the task of energy saving and consumption reduction. They must take a series of measures to achieve the energy intensity reduction targets, which are set in the Five-Year Plan. City-county mergers are almost irreversible policy and require the government to make scientific decisions and comprehensively evaluate various potential impacts on the urban energy intensity. The policy provides good target for examining whether government-led urban expansion through city-county mergers can reduce urban energy intensity, and the answers can help to assess the policy effects of the government-led green urbanization with city-county mergers. To our knowledge, there is a lack of in-depth research on this topic in the existing literature, and we attempt to fill this research gap by presenting this paper as the first empirical study about the impact of city-county mergers on urban energy intensity in the process of China’s rapid urbanization.
In this paper, we use city-level data from 2000–2017 to empirically analyze the urban energy intensity reduction effect of city-county mergers in Chinese cities. In line with most studies, we focus on city-county mergers after 2000 [
7]. This is due to the fact that China’s urban development strategy has shifted from quantitative expansion to scale expansion with the abolition of county-to-city upgrading, and city-county mergers have been given a more important status and policy implications, which are more valuable to examine. The possible contributions of this paper include two main aspects. Firstly, the paper adopts the DID method to circumvent the possible endogenous problem in the empirical analysis, while at the same time systematically investigating the time-lagged effects of the impact of city-county mergers on urban energy intensity, which have practical significance for the proper understanding and appreciation of the policy effects of city-county mergers. Secondly, this paper focuses the impact of city-county mergers on urban energy intensity, while previous studies have mostly focused on economic growth and the efficiency of local public goods provision, and it identifies the mechanisms of merger effects on urban energy intensity from the perspective of local decentralization, agglomeration effect and regional integration. In general, we provide empirical evidence for evaluating the energy-saving performance of city-county mergers, which have led to a reduction in urban energy intensity to some extent. It comes from the formation of “active government” in reducing energy intensity, the energy-saving effect produced by spatial agglomeration, the free flow of factors and the optimal resources spatial allocation, thereby bringing about long-term performance of reducing energy intensity.
The remainder of the paper is organized as follows:
Section 2 introduces the background and theoretical mechanisms of the urban administrative division system;
Section 3 presents the econometric model and description of the relevant data;
Section 4 gives the empirical results;
Section 5 gives the impact mechanisms, and these are discussed in
Section 6;
Section 7 draws conclusions and policy implications.
5. Analysis of Impact Mechanisms
Why city-county mergers show a more pronounced effect in reducing urban energy intensity is the interesting question considered in this section. As can be seen from the previous section, city-county mergers may affect urban energy intensity in three ways: the decentralization effect, agglomeration effect and regional integration effect. This section identifies and tests the above transmission pathways with the help of the mediating effects method. For fiscal decentralization (FD), the ratio of local general per capita budgeted fiscal expenditures to national general per capita budgeted fiscal expenditures is used for measurement [
54]. For the agglomeration effect (POP), we used population density as its proxy variable, considering that rapid population agglomeration within urban space is a distinctive feature of urbanization [
31]. As for regional integration (MS), we used market segmentation (total retail sales of consumer goods/GDP) as its reverse proxy variable [
55]. The data on national general per capita budgetary expenditure were obtained from the China Financial Statistics Yearbook (2001–2018), respectively, and all other data were obtained from the China Urban Statistics Yearbook (2001–2018). We used the mediating effects to further analyze the possible paths of city-county mergers affecting urban energy intensity, which were proposed by Baron and Kenny (1986) [
56]. The recursive model is constructed as follows:
where
represents the mediating effect, and the coefficient
β0 of
treatedc ×
postct in model (1) represents the total effect of city-county mergers on urban energy intensity.
β1μ2 is the product of the coefficient
β1 of
treatedc ×
postct in model (3) and the coefficient
μ2 of
Mct in model (4), which represents the indirect effect of mergers on urban energy intensity, and the coefficient
β2 of
treatedc ×
postct in model (4) represents the direct effect of mergers on urban energy intensity. The total effect is equal to the sum of the indirect and direct effects, that is
β0 =
β2 +
β1μ2.
Combining the sequential and bootstrap tests, the baseline regression model already reports the first step results of the recursive model (coefficient of the explanatory variable is −0.170), and
Table 7 reports the second and third step results of the recursive model estimation, which confirmed the initial judgment of the second part of the mechanism analysis that mergers significantly reduces fiscal decentralization, market segmentation and increases population density, thus enabling mergers to reduce urban energy intensity through the channels of urban fiscal centralization, agglomeration effects and regional integration.
First, the fiscal centralization effect of city-county mergers is analyzed. The coefficient of the
treatedc ×
postct in column (1) is significantly negative, suggesting that mergers can significantly reduce fiscal decentralization. Combined with column (4), we find that city-county mergers reduce urban energy intensity by reducing fiscal decentralization. This is consistent with the findings of Zhang et al. (2011) in that fiscal decentralization may be an important institutional source of declining regulatory standards and increasing energy consumption [
57]; therefore, we argue that the focus of improving the current situation of high urban energy intensity is not to abandon the decentralization system but to strengthen its rationality. On the one hand, the fiscal centralization brought about by city-county mergers reduces local government competition, and on the other hand it may undermine incentives for local officials in counties merged, thereby reducing the efficiency gains from intergovernmental competition. Competition between district governments around the goal of reducing energy intensity in their own jurisdictions has to some extent contributed well to regional energy saving and consumption reduction, and excessive competition can also lead to ineffective allocation of resources and intensification of social conflicts. Improving the current performance appraisal system and introducing a multi-objective incentive mechanism will certainly make local governments more willing to pay attention to the effect of energy saving and consumption reduction while developing the economy, and establish a public service-oriented local government on the basis of maintaining the moderate enthusiasm of local governments, thus promoting the fundamental method for coordinating the tasks of energy saving and consumption reduction among urban districts.
Second, we focus on the agglomeration effect of city-county mergers. The coefficient of the
treatedc ×
postct in column (2) is significantly positive, indicating that mergers significantly increase urban population density, and the results in column (5) indicate that city-county mergers reduce urban energy intensity by increasing population density. It is not just a simple combination of regions, but city-county mergers increase the effective size of the city and improve the level of public services in the counties merged, and although it has a certain dilution effect on the public services of other urban districts, it promotes the agglomeration of elements throughout the city [
58]. The externalities of energy saving and consumption reduction are considered as a “black box” of agglomeration, which can result in “self-cleaning” of negative environmental externalities [
59]. Therefore, the positive externalities brought about by agglomeration, such as the reduction in transportation and information costs and the promotion of technological spillover effect among enterprises, can lead to the optimization of resource allocation and the improvement of utilization efficiency by directing the flow of resources from low-tech industries to high-tech industries within the region, thus promoting the development of high-tech industries and the elimination of low-tech industries, and reducing the energy intensity by correcting the distortion of enterprise factors and promoting the transformation and upgrading of the original industries.
Third, we continue to explain the impact of the regional integration effect of city-county mergers on urban energy intensity. The estimation results in column (3) show that mergers significantly reduce the market segmentation of prefecture-level cities, indicating that mergers increase the level of regional integration, and the results in column (6) suggest that city-county mergers can reduce urban energy intensity by increasing regional integration. The municipal governments combine urban districts and the merged counties into one unit for urban planning, transportation system development, construction project approval and land supply, which provides an institutional basis for reducing administrative barriers and improving the transport spatial linkages between urban districts and the merged counties [
60]. City-county mergers can not only improve market integration, but it can also eliminate the serious district government competitions and market fragmentation caused by the “administrative district economy”. Under the unified urban planning of municipal governments, expanding the radiation range of a city center while reducing district governmental frictions will improve the efficiency of resource allocation among the districts and county merged, including unified industrial layout, transport, communication networks and other infrastructure development, thus contributing to the reduction in the overall urban energy intensity.
6. Discussion
The current empirical literature on the policy effects of city-county mergers mainly focuses on the local public goods provision and economic development, and the results found that city-county mergers can increase the efficiency of local public goods provision and economic development [
8,
21,
23]. It is rare within the literature to explore how such mergers reduce urban energy intensity in developing countries. This study further found that the city-county mergers have a positive impact on promoting urban energy intensity in China. This research not only provides a new understanding of the city-county mergers, but also provides valuable enlightenment for urban energy planning.
In addition, the DID method is one of the most popular methods in the field of policy evaluation, and is also used to evaluate the intertemporal effects of policy implementation [
40,
61]. Because it can better overcome the endogenous problem in the evaluation and its model settings are simple, many researchers are like to use the method to conduct policy evaluation. The previously mentioned scholars used the DID method to evaluate the merger policy on the supply of public goods and economic efficiency, such as Tang and Hewings (2017), Liang et al. (2019) and Zhang et al. (2018) [
8,
21,
23]. The traditional DID method sets unified treatment group dummy variables and time dummy variables for the samples. The time for the city-county mergers to be studied in article is inconsistent in various places, which makes it unsatisfactory to meet the conditions of the traditional DID method. This requires the use of the multi-period DID method, which empirical strategy does not implement uniform treatment group dummy variables and time dummy variables; only the cross term of the two is included in the regression model and focuses on testing the sign and significance of the cross term.
Once the researchers need to empirically analyze the impact of a certain policy on changes in energy intensity, such as city development policies, they can use the DID method for analysis. We take the impact of city-county mergers on urban energy intensity as an example. First, we need to divide the sample into treatment group and control group. The treatment group is the city where merger occurred, and the control group is the city where the merger did not occur. Then, the regression method is used to examine the influence of the treatment group on the dependent variable, that is, the urban energy intensity. The results in this article reflect the effect of the merger policy, determined by the average gains of the treatment group after the policy change subtracted from average gains of the control group during the same period.
This research explores the influence of independent variables on dependent variables, obtaining the actual or theoretical relationship between them, and also attempts to further try to explore the internal mechanism or principle of the relationship; the mediating effect analysis provides the possibility to answer this question. Regarding the mechanism analysis of city-county merger effects, the use of mediation effects is the choice of many scholars in this field [
39]. It can be seen from the analysis of the mechanism that the policy design for the city-county mergers does not completely directly affect the urban energy intensity. In addition to the direct merger effect on urban energy intensity, it also indirectly has a positive effect on reducing urban energy intensity through the fiscal centralization, agglomeration effects and the regional integration, and the total effect of the baseline regression result is the sum of the direct effect and indirect effect.
In the context of China’s decentralization system, the city-county mergers can bring about a small-scale recentralization process under the overall decentralization framework [
15,
21,
22]. Consistent with the research by Zhang C Y et al., (2018), Song et al., (2018), Luo et al. (2017) and Zhang P et al. (2018) [
24,
25,
26,
27], we also found that moderate fiscal centralization, reduction in the competition intensity among governments and scientific incentives for local officials can have positive effects on reducing urban energy intensity. In addition, an important feature of the mergers is population agglomeration and expansion of the urban scale [
8,
29]. Consistent with the research by Morikawa (2012), Otsuka et al. (2018) and Lin et al., (2019) [
30,
31,
33], we also found population agglomeration after merger can reduce urban energy intensity. In addition, city-county mergers can potentially increase the efficiency of resources allocation with strengthening the coordination and economic integration between governments [
34,
35,
39]. The regional integration brought about by the mergers is conducive to the improvement of urban energy intensity, which is consistent with related research results by Qin et al., (2020), Li and Lin, (2017) and Zhang et al. (2017) [
36,
37,
38].
Compared with the optimal size, it should be noted that previous studies believe that the size of Chinese cities is generally small [
62]. However, Yao et al. (2017) believe that the relationship between city size and electricity intensity appeared as an inverted U shape [
32], meaning that our results indicate that many Chinese cities may reach or exceed the optimal scale. The results do not mean that urban energy intensity can be improved only through the adjustment of administrative divisions such as city-county mergers. Although we have confirmed the positive effects of mergers that occurred in 2000–2017, it depends on two conditions: scientific design and improvement of a diversified official appraisal system, and effective reduction in jurisdictional barriers and improvement of government efficiency. Facing both economic and political incentives for local governments, we found that the merger effects have a significant positive effect on reducing urban energy intensity from the third and subsequent three years after the policy implementation. On the one hand, it shows that the assessment which mainly consists of the improvement of energy efficiency already has some positive effect in reducing urban energy intensity. On the other hand, there is not an inverse U-shaped relationship between city officials’ tenure and non-visible expenditures, such as scale of gas and central heating (Wu et al., 2018), which indicates that a “non-visibility bias” will lead to a strategic response from officials [
28]. The central government should redesign performance evaluation systems to rebalance the incentives of local officials; it is important to establish an accountability mechanism that spans the tenure of officials, and officials should be held accountable even after they step down. In addition, the spatial reorganization is difficult to leap into an organic market community and may weaken the effective role of market in city-county mergers [
16]; whether the coordination costs can be reduced and government efficiency can be truly improved depends on the intensity and effect of the merger reform. Under these backgrounds, it is still remains to be seen whether the implemented policy of city-county mergers can achieve the expected goals and effectively promote the decline in urban energy intensity.
7. Conclusions and Policy Implications
Based on the above analysis, how city-county mergers reduce urban energy intensity in China has rarely been studied. We examined the impact of city-county merger policy on urban energy intensity in China, which not only provides new insights about the effect of mergers, but also provides valuable insights about urban energy planning. China is experiencing the largest rural-to-urban migration in the history of the world, and the Chinese government has set a clear binding target of reducing energy consumption per unit of GDP from the 11th to 13th Five-Year Plan. With the promotion of new-type urbanization and energy saving and consumption reduction becoming the two main themes in the green transformation of China’s economy, it is an urgent problem facing Chinese policymakers with establishment of a green development urban space system. In this context, the Chinese experience of promoting merger policy can be used as reference for many developing countries facing rapid urbanization.
City-county mergers are one of the key tools of China’s urbanization strategy, and such policy experiments provide a good research object for testing whether government-led urban expansion promotes the decline in urban energy intensity. While urban district governments are required to carry out special inspections and report to the municipal government on energy saving and consumption reduction every year due to the decomposition of energy-saving constraint indicators, it has rarely been discussed whether the situation of city-county mergers has an impact on urban energy intensity, which is an important theoretical and practical issue in China’s new-type urbanization strategy. We find that city-county mergers have a positive effect on the reduction in urban energy intensity, which starts to have a significant effect in the third year after implementation. The results have passed a series of robustness tests and provided a reasonable explanation from regional decentralization, agglomeration effect and regional integration, which are combined with the intermediary effect. City-county mergers are a spatial reorganization of the transformation of county economies into urban economies; the task of energy saving and consumption reduction can only be accomplished through the comprehensive and unified coordination of the prefecture-level city government from top to bottom. The positive effect of the mergers also relies on the scientific design and improvement of the diversified official appraisal system, effectively reducing jurisdictional barriers and improving government efficiency.
The above findings bring a lot of important enlightenments. They provide an important reference for other prefecture-level cities that are planning to achieve “win-win” of economic growth and energy saving through city-county mergers. It is worth noting that although the city-county mergers can help achieve energy saving and consumption reduction, it will inevitably take time to achieve the intended goals.
Although city-county mergers are conducive to the coordination of energy saving and consumption reduction by city governments, there is still a need to provide effective incentives to local officials of counties merged. The reform of city-county mergers is not “free lunch”, and it must face the large coordination costs between cities and counties merged. Municipal governments must decentralize power appropriately to ensure the weight of energy-saving component of the performance evaluation indicators while centralizing fiscal power; it is important to hold the counties merged government to participate in the energy-saving and consumption-reduction tasks as the top priority. The setting of performance assessment indicators should not follow the old path of sacrificing energy consumption for rapid economic growth, and it also cannot only pursue energy conservation at any economic cost. As a “hard constraint”, energy-saving performance indicators should be accompanied by more financial and administrative powers to encourage the “healthy competition” of energy-saving governance among urban district governments, so that the implementation of merger policy can achieve the twice the result with half the effort effect. In addition, the assessment of energy conservation and consumption reduction should highlight normalization and differentiated assessments, which can establish the leading goose effect in similar assessments and form active government in reducing energy intensity.
In order to achieve the energy saving and consumption reduction targets as early as possible, we need to appropriately relax administrative intervention policies that excessively restrict economic and population clustering. In the process of city-county mergers, municipal governments should encourage to introduce effective policies to actively promote the rational clustering of various economic activities within cities, and promote the urban development in the direction of specialization, industrialization and agglomeration. Taking into account the balance of polarization and diffusion effects, city-county mergers should achieve positive mutual promotion with urbanization; on the one hand, energy intensity can be reduced through industrial structure upgrading with intensive urbanization layout, strengthening infrastructure construction and green knowledge spillover. On the other hand, the agglomeration effects can reduce the cost of information transmission, improve enterprise production efficiency and promote the formation of economies of scale, making it play an important role in reducing energy intensity. While expanding the scale of the cities through city-county mergers, attention should be paid not only to preventing excessive agglomeration, but also to reduce the negative effects caused by the inefficiency of “urban sprawl”. The control of the urban size and density should take into account the differences in the environmental carrying capacity of different cities, avoiding setting uniform standards based on the existing classification of Chinese cities’ tiers, and impose a limit on the cities with existing population sizes exceeding the optimal sizes based on energy efficiency. For cities whose population size is below the optimal population size, they should give full play to their population absorption potential.
Finally, regional integration can improve energy saving and consumption reduction by promoting the free flow of factors and optimizing the efficiency of resource space allocation, which is an important tool to promote coordinated regional economic development. However, energy-intensive enterprises within cities tend to migrate to urban boundaries and areas with low environmental barriers to entry [
63]. In order to avoid the threat posed by the transfer of high energy consumption in the integration process to the completion of energy-saving and consumption reduction tasks in counties merged as much as possible, the formulation of criteria for city-county mergers that match the stage of economic development and strict merger procedure will not only help the city-county mergers to break down administrative barriers to drive regional market integration, but also help strengthen the cooperation of municipal district governments in energy saving and consumption reduction during the integration process. By further strengthening the institutional arrangements for integrated regional development strategies between municipal districts, it will promote a synergistic mechanism for energy saving and consumption reduction policies, which can further break market segmentation and create an “effective market”, that is to say, accelerating the flow and diffusion of economic factors among neighboring districts and promoting a mutual reduction in energy intensity. By striving to form an effective interregional connection in related energy policies such as enterprise energy-saving subsidies, new energy industry support and incentives for the application of new energy technologies, this not only can provide basic guarantees for the realization of overall energy-saving goals, but also could be an important way to achieve the goal of reducing energy intensity.
The shortcomings of this paper lie in several points. It is still necessary to compare and analyze the economic strength of merged counties. Strong counties can still maintain their relative independence after merger policy is implemented, and this will have adverse effect for coordinating the task of energy conservation and consumption reduction. In addition, there are also differences between the active adaption and passive adjustment of city-county mergers; the former is actively implemented based on the needs of urban development and overall planning, while the latter is accompanied by undoing prefecture and setting up prefecture-level city, and what impact they will have on energy intensity is a topic worthy of further study. There are counties merged that are further away from the city center, which tends to make the support and supervision of energy conservation governance from the municipal government relatively small. Therefore, it will help to examine the local government’s energy conservation and consumption reduction strategies with analysis of the formation and mechanism of the boundary effect on energy conservation and consumption reduction, which is based on the spatial and geographical characteristics of cities under the pressure of energy conservation assessment. We noted that the effect of mergers occurs in the third year, so the energy-saving effects of regional decentralization, agglomeration effects and regional integration are all delayed, and it will also be an important part of our future work with systematic quantification of the impact mechanisms as well as heterogeneity. Due to the above reasons, the impact of city-county mergers on urban energy intensity is still complex and the effect of mergers should be evaluated carefully.