Driving factors for coordinating urbanization with conservation of the ecological environment in China

Abstract

Urbanization is a dominant component of social and economic development around the world, but this process creates tremendous pressure on the ecological environment. How to achieve coordination between urbanization and conservation of this environment has become a key issue, especially in developing countries. It is necessary to identify the driving factors that affect this coordination. To identify these factors, we chose 290 Chinese prefecture-level cities to analyze the driving factors behind urbanization using spatial regression analysis, and explored the spatial differences among regions in these factors. Our results show that industrial upgrading and technological progress were the main factors that promoted coordinated development, with industrialization having positive effects under government management, but with differences among regions in how the driving forces affected coordinated development. Using technological progress to promote industrial upgrading, creating new employment to absorb surplus rural labor, and providing workers with skills training so they can take advantage of new jobs can promote win–win solutions that coordinate urbanization with conservation of the ecological environment.

Introduction

Urbanization is one of the main driving forces behind socioeconomic development (Bai et al. 2014; Nagendra et al. 2018). However, the rapid rate of urbanization around the world has created serious problems for the ecological environment, and these problems may intensify in coming decades if urban planners cannot find ways to make the urbanization process more ecologically and environmentally sustainable (Azam and Khan 2016; Liu et al. 2019). This is essential, since the ecological environment is the basis for human survival and human activities, and damage to this environment will restrict urban development (Schewenius et al. 2014). How to coordinate urbanization with conservation of this environment has therefore attracted extensive attention from researchers and policy-makers around the world, especially in developing countries (Güneralp et al. 2015; Dame et al. 2019). The goal of this research is to provide scientific decision-making methods that will support the development of win–win solutions that both protect the ecological environment and sustain urbanization (Ahmed and Islam 2014; Wang et al. 2019b).

As the world’s largest developing country, China has undergone rapid urbanization since the “Reform and Opening Up” policy was implemented in 1978, with the urbanization rate (here, defined as the proportion of the population living in an urban area) increasing from 17.9% in 1978 to 58.5% in 2017 (NBS, National Bureau of Statistics 1979–2018). However, this rapid development has created problems for the ecological environment, such as air pollution, resource shortages (e.g., water), and food security issues (Wang et al. 2016; Liang et al. 2019; Wei et al. 2020). At the same time, large differences in economic foundations, development models, natural resources, and social culture among China’s diverse regions have led to large differences among regions in their driving forces of these problems (Juknys et al. 2016). This not only makes it more difficult for China to coordinate urbanization with conservation of the ecological environment, but also severely hinders sustainable regional development based on this conservation (Liao and Wei 2015). In addition, since China has a huge population and economic scale, the sustainability of its urbanization process will have impacts on the global ecological environment and economy (Liu and Liu 2019). Therefore, it is important to identify the driving factors that most strongly affect China’s coordination between urbanization and conservation of ecological environment (Hao et al. 2020).

Unfortunately, previous scholars have generally ignored the driving factors responsible for coordination of development between urbanization and conservation, and could not propose practical and effective development measures for policy-makers. Based on the environmental Kuznets curve hypothesis, researchers around the world have improved the index system for evaluating the degree of coordination and coupling between urbanization and the ecological environment (Grossman and Kreuger 1994; Li et al. 2012; Güneralp et al. 2015; Zhao et al. 2017). Combined with spatial analysis, many scholars have analyzed the spatial pattern and clustering of this coupling coordination (Liu et al. 2018; Yuan et al. 2018; Wang et al. 2019b). These studies have provided significant insights into the coordination between urbanization and conservation of the ecological environment (Grimm et al. 2008; Estoque and Murayama 2013; Zhang et al. 2020; Zhou et al. 2020). However, the driving factors responsible for this coordinated development remain poorly understood, especially in terms of regional differences among the main factors. As a result, researchers have been unable to provide effective guidance for policy development to coordinate urbanization with conservation of the ecological environment.

To identify the driving factors for this coordination and provide empirical evidence to support government urbanization planning, we studied Chinese prefecture-level cities using spatial regression analysis models to identify the dimensions that most strongly affected this coordination. We then used a geographically weighted regression model to explore the spatial differences among regions in the key driving factors, and propose appropriate development strategies that target the most important factors for each region. Two innovations of our study are the use of spatially explicit analysis of a large number of cities and the goal of identifying differences among cities in a large number of key factors to reveal how the key factors vary among regions. Our results provide guidance for Chinese policy development and a valuable reference to help other countries coordinate urbanization with conservation of the ecological environment.

Materials and Methods

Study area and data sources

To more comprehensively explore the driving factors behind the coordination of urbanization with conservation of the ecological environment in China, we selected 290 prefecture-level cities in 27 provinces based on the availability of data on the key driving factors (see below), and collected the most recent data available (from 2017). The data mainly come from the China City Statistical Yearbook, statistical yearbooks of the provinces and cities, and national economic development and statistical bulletins for each city (NBS 2018; http://www.stats.gov.cn/).

Evaluation of the coordination

Given the complexity of the coordination, we selected 12 urbanization indicators in the categories population urbanization, economic urbanization, social urbanization, and spatial urbanization, and 10 indicators of the ecological environment in the categories ecological and environmental level, ecological and environmental pressure, and protection of the ecological environment (Table 1). These indicators were selected based on their relevance for urbanization or ecological conservation reported in recent research (Li et al. 2012; Zhao et al. 2017; Dong et al. 2019). Our goal was to establish a comprehensive indicator system for assessing the degree of coordination between urbanization and conservation of the ecological environment.

Table 1.

Index system used to evaluate the relationship between urbanization and conservation of the ecological environment. The attribute column of the table indicates whether a given index increases (+) or decreases (−) urbanization (first part of the table) or conservation of the ecological environment (second part of the table)

System	Type	Index	Attribute
Urbanization	Population urbanization	Population urbanization rate	+
		Urban population density	+
	Economic urbanization	City’s per capita GDP	+
		Value added by tertiary industries as a percentage of the city’s GDP	+
		Value added by secondary industries as a percentage of the city’s GDP	+
		Per capita commercial trade	+
		Per capita fixed asset investment	+
	Social urbanization	Number of students with at least a primary school education per 10 000 people	+
		Number of medical beds per 10 000 people	+
		Number of Internet broadband subscribers per 10 000 people	+
	Spatial urbanization	Urban area as a percentage of the total land area	+
		Per capita urban road area (all categories of road combined)	+
Conservation of the ecological environment	Ecological and environmental level	Per capita grain planting area	+
		Water supply as a percentage of the total land area	+
		Proportion of urban area covered by vegetation	+
	Ecological and environmental pressure	Per capita industrial wastewater emission	−
		Per capita industrial dust emission	−
		Per capita industrial sulfur dioxide emission	−
		Natural growth rate of the population (birth rate minus death rate)	−
	Ecological and environmental protection	Proportion of wastewater that is treated centrally	+
		Proportion of consumption wastes that are treated	+
		Proportion of industrial solid waste that is recycled	+

First, because the indicators used different units of measurement, we standardized the indicators using Z-values (Li et al. 2012):

$$\text{Positive}\text{indicator:}{Z}_{\mathit{ij}}=\frac{X_{\mathit{ij}}-min\{X_j\}}{max\{X_j\}-min\{X_j\}},$$ 1

$$\text{Negative}\text{indicator:}{Z}_{\mathit{ij}}=\frac{max\{X_j\}-X_{\mathit{ij}}}{max\{X_j\}-min\{X_j\}},$$ 2

where positive indicators represent factors that promote urbanization or conservation of the ecological environment (Table 1), and negative indicators interfere with these processes. X_ij is the original value of the jth indicator for the ith city, and max{X_j} and min{X_j} represent the maximum and minimum values (respectively) of indicator j for all cities.

Second, we used the entropy method to determine the index weights for urbanization and conservation of the ecological environment. We chose this approach because it can reflect the original information that each index represents and enhance the objectivity of the evaluation (Liu et al. 2018):

$$Y_{\mathit{ij}}=\frac{Z_{\mathit{ij}}}{{\sum }_{i=1}^m{Z}_{\mathit{ij}}},$$ 3

$$e_j=-\frac{1}{lnm}\sum _{i=1}^m{Y}_{\mathit{ij}}ln{Y}_{\mathit{ij}},$$ 4

$$f_j=1-e_j,$$ 5

$$w_j=\frac{f_j}{{\sum }_{j=1}^n{f}_j},$$ 6

$$U_s=\sum _{j=1}^n{w}_j{Z}_{\mathit{ij}},$$ 7

where Y_ij is the information entropy of the jth index the ith city, Z_ij is the standardized value of the jth indicator for the ith city, m represents the number of cities (here, 290), e_j is the information entropy of indicator j, f_j is the redundancy of the information entropy, w_j is the weight of indicator j, n represents the number of indicators, and U_s is the comprehensive evaluation index of subsystem s (here, s = 1 represents the comprehensive value for all urbanization indicators and s = 2 represents the comprehensive value for all indicators for the ecological environment).

Third, we developed a model for the degree of coupling to comprehensively evaluate the level of coordination between urbanization and conservation (He et al. 2017):

$$C=\frac{2\sqrt{U_1\times U_2}}{U_1+U_2},$$ 8

$$T=\alpha U_1+\beta U_2,$$ 9

$$D=\sqrt{C\times T},$$ 10

where C is the degree of coupling, U₁ and U₂ represent the comprehensive indicator values for urbanization and conservation of the ecological environment (respectively) in Eq. 7, T is a comprehensive index for the coordination between urbanization and conservation, α and β represent the coefficients for (respectively) the contributions of urbanization and conservation of the ecological environment (which we assumed had equal priority to the government, so that α = β = 0.5), and D is the degree of coordination between urbanization and conservation.

Finally, we used the spatially explicit Moran’s I index (using both the overall Moran’s I for China and the local Moran’s I) to analyze the spatial distribution characteristics (including spatial agglomeration trends and spatial heterogeneity) to reveal spatial correlations and differences among regions and cities in the relationship between urbanization and conservation of the ecological environment (Guo et al. 2020). That is, we wanted to determine whether cities located close to each other were more likely to be controlled by the same key factors. We employed the spatial autocorrelation analysis module of the GeoDA software (https://geodacenter.github.io/) to calculate I using the software’s default settings. The overall Moran’s I measures the relationship between spatial elements for China as a whole, and a statistically significant value demonstrates the existence of overall autocorrelation. The calculation formula is:

$$I=\left(\frac{n}{{\sum }_i{\sum }_j{W}_{\mathit{ij}}}\right)\left(\frac{{\sum }_i{\sum }_j{W}_{\mathit{ij}}(x_i-\overline{x})}{{\sum }_i{(x_i-\overline{x})}^2}\right),$$ 11

where n is the number of spatial units (here, 290 cities); x_i is the value of the tested variable i; $\stackrel{-}{x}$ is the average value of the tested variable; and W_ij is the corresponding value in the spatial distance weight matrix for variable i in city j. The local Moran’s I (the local indicator of spatial autocorrelation, LISA) represents the degree of spatial correlation between each city and its surrounding cities. The calculation formula is:

$$I_i=\left(\frac{n(x_i-\overline{x}){\sum }_i{W}_{\mathit{ij}}(x_j-\overline{x})}{{\sum }_i{(x_i-\overline{x})}^2}\right),$$ 12

where I_i represents the LISA value for variable i, and the other variables have the same meaning as in Eq. 11.

Selection of key indexes from the driving factors

We chose the level of coordination between urbanization and conservation of the ecological environment (i.e., D) as the dependent variable (here, high values of D represent good coordination of urbanization with conservation). For the independent variables, the level of coordination between urbanization and conservation results from a variety of factors, so we selected 13 measurement indicators that may affect D: industrialization, industrial upgrading, foreign trade, degree of land development, urban–rural income ratio, urban greening, urban construction, technological progress, government management, population pressure, resident education, resident living intensity, and climate. To avoid double calculation of indicators and determine the main factors, we first standardized the data using Z scores to make indicators with different units of measurement comparable. We then used stepwise regression to identify the variables with a significant effect on D, with P < 0.05 and > 0.10 as the criteria for retaining and eliminating variables, and then added the eliminated but economically meaningful variables. When this analysis was complete, we retained the eight variables in Table 2. Each of these variables had a variance inflation factor (VIF) < 3, which suggests that they lacked multicollinearity. This process selected independent variables that were both strong determinants of the degree of coordination and a logical choice (Wang and Mu 2018).

Table 2.

Results of the indicator selection (variables retained at P < 0.05 in the stepwise linear regression) and of the multicollinearity test (to avoid multicollinearity) to select the key factors that have influenced the coordination between China’s urbanization and conservation of the ecological environment. TOL, tolerance factor; VIF, variance inflation factor

Variables	Calculation method	Multicollinearity test
		TOL	VIF
Industrialization	Value added by secondary industries as a proportion of the city’s GDP	0.393	2.546
Industrial upgrading	Value added by tertiary industries as a proportion of the city’s GDP	0.397	2.517
Urban–rural income ratio	Urban residents income/rural residents income	0.810	1.234
Urban greening	Proportion of urban area covered by vegetation	0.820	1.219
Technological progress	Technology expenditure per capita	0.720	1.389
Government management	Proportion of industrial solid waste that is recycled	0.912	1.096
Population pressure	Birth rate minus death rate	0.712	1.405
Resident education	Number of students with at least a primary school education per capita	0.708	1.413

Spatial analysis

To determine whether the impacts of the variables exhibited spatial variation, we performed three forms of regression: ordinary least-squares (OLS), spatial error model (SEM), and spatial lag model (SLM) regression (Chaurasia et al. 2020). We performed this analysis using the regression analysis module of GeoDA (https://geodacenter.github.io/) to calculate the regression coefficients. The calculation formulas were:

$$\text{OLS:}\mathbf{Y}=\mathbf{X}\beta +\mathbf{\epsilon },$$ 13

$$\text{SLM:}\mathbf{Y}=p\mathbf{WY}+\mathbf{X}\beta +\mathbf{\epsilon },$$ 14

$$\text{SEM:}\mathbf{Y}=\lambda \mathbf{W}\epsilon +\mathbf{X}\beta +\mathbf{\mu },$$ 15

where Y is an N × 1 vector for the degree of coordination between urbanization and conservation of the ecological environment (N = 290 cities); X is an N × K matrix of K = 8 exogenous explanatory variables (Table 2); β is a K × 1 vector of the corresponding regression coefficients (β_k); ε is a vector of normally distributed error terms; þ represents the spatial autoregressive coefficient, which reflects the spatial effect of the observed value in the nearby cities (WY) on local observation value Y; WY represents a vector for a spatial lag-dependent variable; λ is the spatial autocorrelation coefficient of error; W is an N × N spatial distance weight matrix containing contiguity relationships; and µ is an N × 1 vector of residuals. We defined the best model as the model that had the highest R² and Lagrangian multiplier (LM) value (Anselin 1995). We also used the absolute value of the coefficients to quantify the degree of influence of different factors, which we defined as their contribution (Cao et al. 2014). The calculation formula is:

$${\text{Con}}_k=\frac{{\text{AC}}_k}{{\sum }_{k=1}^k{\text{AC}}_k}\times 100\%,$$ 16

where Con_k is the contribution of variable k to the dependent variable (Y), and AC_k is the absolute value of the coefficient β_k.

To further explore the effects of the driving factors in different regions, we developed a geographically weighted regression model to comprehensively describe the spatial effects of the coordination between urbanization and conservation of the ecological environment. We performed this analysis using version 10.2 of the ArcGIS software (www.arcgis.com) for the eight selected variables (Fotheringham et al. 1998; Wang et al. 2019a):

$$Y_i={\beta }_0(m_i,n_i)+\sum _{j=1}^n{\beta }_j(m_i,n_i)x_{\mathit{ij}}+{\epsilon }_i.$$ 17

where Y_i represents the dependent variable (i.e., D), β₀(m_i, n_i) is a regression constant (the intercept), β_j is the regression coefficient for variable j in city i, (m_i, n_i) is the coordinates of the geographic center of gravity for city i, and ε_i is a residual term that follows a normal distribution, which we calculated using the geographically weighted regression function provided by ArcGIS. We then mapped the values of the key driving factors to display their spatial distribution intuitively.

Results

More than 70% of the cities had a degree of coordination between urbanization and conservation of the ecological environment that was less than 0.5, which is the threshold between balance and imbalance; that is, a large majority of the cities were unbalanced. Only 4 cities (Xiamen, Guangzhou, Chengdu, and Zhuhai) showed a good level of coordination (D > 0.6), indicating a balance between urbanization and conservation. The degree of coordination in southeastern and parts of northeastern China was higher than that in the western region (Fig. 1). The overall Moran’s I for China was 0.4 (P < 0.05), which means that the degree of coordination exhibited a significant spatial correlation (that is, cities with a similar level of coordination were located close to each other). The local Moran’s I showed that 231 cities (80% of the total selected cities) showed significant spatial correlation (Fig. 2). The statistically significant associations (P < 0.05) were mainly H–H and L–L, in which H represents high correlation, L represents low correlation, and the first and second letters represent the focal city and the adjacent cities, respectively. The cities with significant results were concentrated in eastern and western China for H–H and L–L, respectively. This means that the degree of coordination was related to a city’s geographic location, and that there was obvious spatial heterogeneity in this degree. Therefore, it was necessary to account for spatial factors in the model, and it will be more effective to analyze the driving factors both for China as a whole and for individual regions.

Fig. 1.

The spatial distribution of the degree of coordination (D) between urbanization and conservation of the ecological environment for the 290 selected Chinese cities in 2017. See the “Materials and Methods” section for details. Moderately unbalanced, 0.3 ≤ D < 0.4; Slightly unbalanced, 0.4 ≤ D < 0.5; Slightly balanced, 0.5 ≤ D < 0.6; Moderately balanced, 0.6 ≤ D < 0.7

Fig. 2.

The spatial distribution of the local Moran’s I (LISA) for the degree of coordination between urbanization and conservation of the ecological environment for the 290 Chinese cities in 2017. In the legend, H represents high correlation, L represents low correlation, and the first and second letters represent the focal city and the adjacent cities, respectively. See the “Materials and Methods” section for details

For China as a whole, the SEM regression appeared to be most suitable, as it had the highest R² and LM values (Table 3; Anselin 1995). Therefore, we have focused on the SEM regression results in the rest of our description. The contribution to the coordination was greatest for industrial upgrading (22.1%), followed by industrialization (20.8%) and technological progress (19.1%). The remaining five indicators were also significant (P < 0.05), but had much smaller contributions: resident education (9.9%), population pressure (8.6%), government management (7.7%), urban greening (6.4%), and the urban–rural income ratio (5.4%). A large population pressure and a large income ratio both had strong negative impacts on the degree of coordination.

Table 3.

Statistical significance and contributions to the coordination between urbanization and conservation of the ecological environment for the 8 socioeconomic and ecological dimensions selected by stepwise linear regression for the 290 Chinese cities. Data are from 2017, and the Lagrange multiplier (LM) diagnosis results were used to select the optimal model. The regression coefficients (β_k) were calculated using Eqs. 13–15. Results are for ordinary least-squares (OLS), spatial lag model (SLM), and spatial error model (SEM) regressions. The contribution was calculated using Eq. 16. See the “Materials and Methods” section for details. Significance levels: **0.01, *0.05, ns is not significant. MI, Moran’s I; DF, degrees of freedom; SARMA, spatial autoregressive model with autoregressive errors; R-LM, robust-Lagrange multiplier

Variables	OLS (R2 = 0.58)		SLM (R2 = 0.68)		SEM (R2 = 0.71)
	βk	Contribution (%)	βk	Contribution (%)	βk	Contribution (%)
Industrialization	0.170**	10.47	0.196**	13.50	0.354**	20.82
Industrial upgrading	0.227**	13.99	0.257**	17.70	0.375**	22.06
Urban–rural income ratio	− 0.222**	13.68	− 0.125**	8.61	− 0.091*	5.35
Urban greening	0.216**	13.31	0.139**	9.57	0.109**	6.41
Technological progress	0.364**	22.43	0.320**	22.04	0.325**	19.12
Government management	0.227**	13.99	0.128**	8.82	0.131**	7.71
Population pressure	− 0.052ns	3.20	− 0.130**	8.95	− 0.147**	8.65
Resident education	0.145**	8.93	0.157**	10.81	0.168**	9.88

Test	MI/DF	Value		MI/DF	Value
Moran’s I (error)	0.240	13.497**	LM (SARMA)	2	148.072**
LM (lag)	1	100.067**	R-LM (lag)	1	19.154**
LM (error)	1	128.918**	R-LM (error)	1	48.005**

From the local perspective, the geographically weighted regression (GWR) model (R² = 0.78, P < 0.05, VIF < 3) did a good job of explaining the impact of the driving factors on the differences between regions in the degree of coordination (Fig. 3). Specifically, industrial upgrading, industrialization, and resident education showed obvious differences between eastern and western China and between northern and southern China, with the regression coefficient decreasing from southeast to northwest (Fig. 3a, b, d). In contrast, technological progress showed the opposite trend (Fig. 3c). The regression coefficient for population pressure increased from western to eastern China (Fig. 3e), and was highest in northeastern China. The regression coefficient for government management decreased from southern to northern China (Fig. 3f), and was smallest in northeastern China. Urban greening and the urban–rural ratio showed a concentric circular pattern, but with the highest values at the center of the circle for urban greening (Fig. 3g) and the lowest values at the center of the circle for the income ratio (Fig. 3h).

Fig. 3.

Spatial distributions of the regression coefficients (β_j) of the variables in the geographically weighted regression model for the degree of coordination between urbanization and conservation of the ecological environment for the 290 Chinese cities in 2017. The coefficients were calculated using Eq. 17

Discussion

From the overall perspective for China as a whole, industrial upgrading, industrialization, and technological progress were the main driving factors that affected the coordination between urbanization and conservation of the ecological environment (Table 3). However, increasing population pressure decreased the degree of coordination, especially in the relatively undeveloped western region (Fig. 3e). Although a growing population can create more demand for products and promote technological progress, it will also cause more environmental pollution and increase pressure on the ecosystems that sustain a city (Chen et al. 2020). Unfortunately, it appears to be infeasible to decrease the population to reduce these pressures, as China may be experiencing a demographic crisis in which there are insufficient young workers to support the growing population of retired citizens (Li and Lin 2016). The keys to relieving the population pressure will be technological progress and education, as these processes will help residents find better jobs with a higher salary in secondary and tertiary industries. This will combine improved economic development (greater income) with lower environmental impacts (because secondary and tertiary industries tend to produce less pollution), thereby promoting conservation despite the population growth (Hutcheson et al. 2018). From a local perspective, the regression coefficients for the eight key factors showed large spatial heterogeneity; that is, the effects of the factors differed among the regions (Fig. 3). In short, the driving factors that influence the degree of coordination show complex regional differences.

Industrial upgrading is a key driving factor to improve the degree of coordination (Hao et al. 2020). This is because industrial upgrading can attract more rural people to cities by providing more job opportunities, thereby promoting urbanization (Liang and Yang 2019). However, improved education will be necessary to let most of these workers work in these new jobs, which require different skills from what rural migrants have learned. Simultaneously, industrial upgrading can reduce the dependence on resources and energy as the industrial structure evolves from one dominated by primary industries to one with a greater proportion of secondary and tertiary industries, and this will also improve the environmental quality (Kang and Feng 2016). Since China’s “Reform and Opening Up” policy was implemented first in eastern China, the eastern coastal areas took the lead in reforming their economic institutions, and also benefited from preferential economic development policies that promoted industrial upgrading (Kuang et al. 2014). Therefore, industrial upgrading had the greatest effect in eastern China, with the effect decreasing toward northwestern China (Fig. 3a). China should encourage a transformation from low-value-added primary industries to high-value-added secondary and tertiary industries by developing policies that encourage this form of economic development, and can modify the policy details to account for the key constraints and opportunities in each region (Cheong and Wu 2014). This will both improve the degree of coordination and promote coordinated regional development by reducing regional differences.

As a core element of industrial upgrading, technological progress plays an important role in increasing the degree of coordination (Zhao et al. 2016). First, technological progress is an essential part of industrial upgrading, since improved technology increases the efficiency and decreases the environmental impacts of an industry (Hao et al. 2020). Second, technological progress can create new jobs and promote the transfer of surplus rural labor to cities, which is a decisive force in promoting urbanization (Wang and Wei 2020). Technological progress requires a large amount of capital and new talents, which are both closely related to the level of economic development; as a result, the degree of technological development decreases from the economically developed southeast region to the economically backward western and northeastern regions (Fraumeni et al. 2019). This gives technological progress in the western and northeastern regions more potential to promote coordination between urbanization and conservation. Therefore, it is necessary for the Chinese government to further increase investment in technology, continuously improve the level of innovation-driven development, and develop high-tech industries and green industries through technological innovation, especially in western and northeastern China (Fig. 3c).

Industrialization is a key driving force for urbanization, but will inevitably have some negative effects on the ecological environment for primary industries and many secondary industries (Kurucu and Chiristina 2008; Brahmasrene and Lee 2017). In China, many cities have not completed their industrialization process, so they cannot abandon industrialization to protect the ecological environment that sustains them; doing so would lead to an infeasible balance between urbanization and conservation, resulting in a low degree of coordination (Zhang and Su 2016). More importantly, a higher level of sustainable urbanization is an important foundation for conservation, as increases in technology and income provide technical and financial support for conservation (Liang and Yang 2019). We found that industrialization increased the degree of coordination, but that the effect weakened from eastern to western China (Fig. 3b). Possible solutions include accepting a certain degree of pollution as the price for economic growth, strengthening government management of the discharge of industrial pollution (e.g., setting emission reduction targets and increasing waste treatment capacity), and increasing technical support to develop a circular (recycling) economy (Fig. 3f; Li et al. 2012; Zhou et al. 2020). In addition, increasing the proportion of the urban area covered by vegetation is a good plan, especially in the central region of northern China, as this approach can both improve the quality of life of residents and improve the ecological environment (Fig. 3g; Imam and Banerjee 2016; Chan and Vu 2017).

Education of urban residents and of migrants to cities, including improvements in their work skills and willingness to protect the ecological environment, will also be important (O’Neill et al. 2020). Due to the higher income that is available in cities, large numbers of rural residents have been flocking to China’s rapidly growing cities, especially in southeastern China, which has a high level of urbanization (Fig. 3d; Zhang 2016). However, if they lack the work skills to adapt to urban employment, they may be forced to return to the rural area, thereby slowing or reversing urbanization (Wang et al. 2011). In addition, research on the environmental Kuznets curve suggests that residents with lower incomes have a lower desire to protect the environment, and in some cases, the need to earn a satisfactory living may force them to carry out economic behaviors that damage their ecological environment because they have no alternatives (Grossman and Kreuger 1994; Liu 2011). Therefore, it is necessary to improve their education by providing job skills training and environmental protection education, to help them both achieve income growth and protect the environment. China has successfully tested approaches such as a model based on natural ecological restoration (i.e., the environment is protected against further degradation to allow natural recovery); for example, in Changting County, the regional government subsidized the construction of barns to raise livestock (Cao et al. 2017, 2020). This method promotes ecological restoration by removing grazing pressure from the region’s grasslands, but also provides a sustainable livelihood for residents of the project area (Fu et al. 2020).

Conclusion

Urbanization is a seemingly inevitable consequence of socioeconomic development, and the resulting ecological and environmental problems have begun to pose serious challenges to global sustainable development. Therefore, governments and researchers must cooperate to identify the main driving factors that promote coordination between urbanization and conservation of the ecological environment, and must implement specific policies to target these factors. Our research demonstrates one way to identify the main driving factors. Using this approach, we found that technological progress can promote industrial upgrading and transformation of a city’s industrial structure, and increasing work skills and willingness to protect the environment will both promote conservation and create new employment opportunities to absorb rural surplus labor.

Technological progress is the key factor to promote coordination between urbanization and conservation of the ecological environment. Therefore, investments in science and technology must be increased to make full use of technological progress to optimize the industrial structure, to improve livelihoods of residents to achieve income growth and increase their willingness to protect the environment (Li et al. 2018; Hao et al. 2020). In addition, given the strong regional differences in the main driving factors for urbanization, governments should formulate development measures that focus on the key challenges for their region and strengthen regional cooperation to take advantage of each region’s strengths and compensate for its weaknesses (Liu et al. 2018; Tian et al. 2020). These approaches can increase the likelihood of achieving a win–win solution that coordinates urbanization with conservation of the ecological environment. As the world’s largest developing country, China’s diverse regions contain a range of stages of economic development that can provide lessons for other countries and regions at similar stages and that face similar problems.

Biographies

Zhaoyang Cai

is a Student at Minzu University of China. Her interests include ecological economics, institutional economics, and politics.

Weiming Li

is a Student at Minzu University of China. His interests include ecological economics, institutional economics, and politics.

Shixiong Cao

is a Professor at Minzu University of China. His interests include ecological economics, institutional economics, and politics.

Author contributions

SC designed the research; ZC and WL performed the data analysis and wrote the main manuscript text. ZC and WL contributed equally to this work.

Conflict of interest

The authors declare no conflict of interest. The opinions expressed here are those of the authors and do not necessarily reflect the position of the government of China or of any other organization.