How Does the Improvement of Ecological Compensation Efficiency Affect Urban Economic Resilience? Evidence from the Yangtze River Economic Belt in China

Abstract

This study examines whether and through what channels ecological compensation efficiency affects urban economic resilience from a watershed-scale perspective. Using panel data for 108 prefecture-level cities in China’s Yangtze River Economic Belt from 2011 to 2023, ecological compensation efficiency is first measured with a super-efficiency SBM model incorporating undesirable outputs. A two-way fixed effects model, a mechanism-testing framework, robustness checks, and a spatial Durbin model are then employed to investigate its direct effect, transmission mechanisms, and spatial spillovers. The results show that (1) ecological compensation efficiency significantly enhances urban economic resilience, and this finding remains robust under alternative indicator measurements and model specifications; (2) mechanism analysis indicates that ecological compensation efficiency strengthens urban economic resilience by promoting green technological innovation and facilitating digital–real economy integration; and (3) spatial analysis further reveals significant positive spillover effects on neighboring cities. These findings suggest that improving ecological compensation efficiency can enhance both local and regional economic resilience. This study enriches the literature on ecological compensation and resilient urban development and provides policy implications for efficiency improvement, green and digital transformation, and cross-regional collaborative governance.

1. Introduction

Against the backdrop of shrinking demand, supply shocks, weakening expectations, geopolitical shifts, more frequent extreme climate events, and the restructuring of global value chains, cities are increasingly required to sustain stable growth and strengthen governance capacity under conditions of heightened uncertainty and multiple constraints [1]. The 14th Five-Year Plan (2021–2025) for Economic and Social Development of the People’s Republic of China and the Outline of the Long-Range Objectives Through 2035 explicitly identifies the development of “livable, innovative, smart, green, culturally rich, and resilient cities” as a major strategic task. Strengthening urban economic resilience has therefore become essential to high-quality economic development, regional sustainability, and the capacity of cities to transform risks into development opportunities [2]. Shaker and Mackay (2021) further reveal hidden patterns of sustainable development in Asia and their correlations with global change, highlighting the need to understand regional resilience within an integrated sustainability framework [3].

The Yangtze River Economic Belt is both a major national economic corridor and a key region for watershed ecological protection and cross-regional ecological compensation. It connects urban agglomerations and industrial chains across the upper, middle, and lower reaches of the Yangtze River, while also facing the public-good nature of watershed-scale environmental governance and cross-regional ecological externalities. In recent years, China has continuously improved the legal, policy, and institutional framework for ecological protection compensation, especially in key watershed regions. Under fiscal constraints and the requirements of high-quality development, the policy focus has gradually shifted from expanding the scale and coverage of ecological compensation to improving its performance and efficiency. This shift raises an important question: whether more efficient ecological compensation can contribute not only to environmental improvement, but also to the economic resilience of cities.

Ecological compensation efficiency reflects the input–output performance of ecological compensation and ecological governance under given resource constraints. It is related not only to the cost-effectiveness and coordination of environmental governance, but also to cities’ ability to reallocate resources, improve ecological quality, stabilize expectations, and maintain economic functions under external shocks [4]. Studies on ecological compensation and payments for ecosystem services (PESs) have examined efficiency from perspectives such as spatiotemporal variation, budget constraints [5,6], additionality [7], ecological compensation baselines [8], and cost–benefit comparisons [9]. In terms of measurement methods, DEA, SFA, and SBM-based DEA models have been widely used to measure environmental or ecological compensation efficiency, especially when undesirable outputs and input–output slacks need to be considered [10,11].

Economic resilience was initially understood as the ability of an economic system to cope with shocks [12], and has gradually been extended to include resistance, recovery, adjustment, and adaptation to new development conditions [13,14]. Existing studies measure economic resilience using either single indicators, such as employment rates [15] or trade-related indicators [16], or multi-indicator systems based on resistance, adjustment, response, and transformation capacities [17,18]. Its determinants have been examined from the perspectives of knowledge base, trade openness [19], digital economy [20], industrial structure [21], agglomeration effects [22,23], institutional and policy environments [24,25], and technological innovation [26,27]. Recent studies also show that economic resilience may generate spatial spillovers across regions, suggesting that urban resilience is shaped not only by local factors but also by interregional linkages.

Taken together, international studies on PES, ecological fiscal transfers (EFTs), and regional resilience provide an important analytical basis for this study while also revealing several limitations. PES research shows that ecological compensation can help align the incentives of ecosystem service providers and beneficiaries, but its effectiveness depends on conditionality, additionality, monitoring capacity, and the link between compensation standards and ecological outcomes [28]. EFT studies further demonstrate that intergovernmental fiscal arrangements can internalize spatial ecological externalities and encourage local governments to protect ecosystems that generate cross-regional benefits [29]. However, these studies mainly focus on policy design, fiscal incentives, conservation outcomes, and environmental performance, while paying less attention to whether ecological compensation resources are used efficiently and whether such efficiency can strengthen urban economic resilience. Meanwhile, regional resilience studies have mostly explained resilience through industrial, technological, financial, and institutional factors, with limited attention to ecological governance instruments. Therefore, this study shifts the focus from compensation scale and policy participation to ecological compensation efficiency, and examines its direct effects, transmission mechanisms, and spatial spillovers on urban economic resilience in the Yangtze River Economic Belt.

Building on the above discussion, this study shifts attention from the amount of ecological compensation funds and the strength of policy implementation to the efficiency with which ecological compensation resources are used. The study first evaluates whether higher ecological compensation efficiency contributes to the resilience of urban economies. It then explores whether this effect operates through green technological innovation and digital–real economy integration, and further investigates whether the influence extends across neighboring cities within the watershed-based regional system.

Compared with existing studies, this paper makes three main contributions. First, it shifts the research focus from the scale and implementation of ecological compensation to ecological compensation efficiency, thereby emphasizing how ecological governance inputs are transformed into ecological, economic, and social outcomes. Second, it links ecological compensation efficiency with urban economic resilience, extending the discussion of ecological compensation from environmental policy performance to the resilience of urban economic systems. Third, it examines the mechanisms, heterogeneity, and spatial spillover effects of this relationship at the watershed city scale, providing empirical evidence on how efficiency-oriented ecological compensation contributes to resilient urban development in the Yangtze River Economic Belt.

The rest of the paper is arranged as follows. Section 2 presents the theoretical analysis and hypotheses. Section 3 introduces the research design, including data sources, variable definitions, and model specifications. Section 4 and Section 5 provide the empirical findings and further analysis. Section 6 summarizes the main conclusions, policy implications, limitations, and directions for future research.

2. Theoretical Analysis and Research Hypothesis

This study develops its theoretical framework by integrating ecological economics, institutional theory, the resource-based view, and resilience theory. Ecological economics explains how ecological compensation internalizes environmental externalities and improves the allocation of ecological governance resources. Institutional theory highlights the role of formal governance arrangements in reducing coordination costs and strengthening cross-regional cooperation. The resource-based view suggests that green technological innovation and digital–real economy integration can serve as strategic resources and dynamic capabilities that support industrial upgrading and adaptive transformation. Resilience theory provides the basis for understanding urban economic resilience as the capacity of an urban economic system to resist shocks, recover from disturbances, adjust its structure, and achieve long-term transformation. Based on these perspectives, this study explains how ecological compensation efficiency affects urban economic resilience through direct, mechanism, and spatial spillover effects.

2.1. The Direct Impacts of Ecological Compensation Efficiency on Urban Economic Resilience

Improving ecological compensation efficiency may enhance urban economic resilience through three related mechanisms: externality internalization, resource allocation optimization, and governance coordination. Together, these mechanisms can strengthen cities’ resistance, recovery, adaptive, and transformative capacities when facing environmental risks, market fluctuations, or policy shocks.

From the perspective of externality internalization, ecological compensation mechanism converts the positive externalities generated by watershed ecological protection into economic benefits for compensation recipients. This helps correct market failures caused by the mismatch between ecological protection costs and the spatial scope of benefits [30]. In watershed regions, upstream areas usually bear the costs of pollution control and ecological restoration, while downstream areas benefit from improved ecosystem services. Existing studies suggest that ecosystems can reduce social costs by buffering adverse events, reflecting “insurance value” of ecosystem services [31]. Therefore, higher ecological compensation efficiency allows compensation inputs to be more effectively transformed into ecosystem service improvements, thereby reducing the exposure of urban economic activities to environmental risks such as floods, droughts, and ecological degradation [32].

From the perspective of public finance, ecological compensation efficiency reflects whether limited fiscal resources can be effectively converted into ecological protection performance and environmental improvement [33]. Higher efficiency implies that public funds can be allocated more precisely to pollution prevention, ecological restoration, environmental infrastructure, and public service improvement, thereby reducing the misallocation of ecological governance resources. Since ecological restoration and natural capital improvement are important for disaster risk reduction and climate adaptation [34], efficient ecological compensation can help cities maintain basic economic functions under external shocks by improving environmental quality and reducing the negative effects of ecological degradation on production, public services, and residents’ livelihoods.

From the perspective of institutional economics and regional governance, improving ecological compensation efficiency can reduce the transaction and coordination costs of cross-regional ecological governance. Watershed ecological environments are characterized by strong cross-regional externalities, meaning that a single city can hardly bear the full cost of ecological protection while fully capturing the corresponding benefits. Ecological compensation mechanisms can therefore reallocate ecological protection responsibilities, governance costs, and environmental benefits across regions [35]. An efficient mechanism can clarify upstream and downstream responsibility, optimize compensation standards, and improve benefit-sharing arrangements. By reducing coordination barriers caused by unclear responsibilities and unequal benefit distribution [36], such institutional arrangements can improve policy responsiveness and resource mobilization in crisis situations, thereby supporting the recovery and adjustment of urban economic systems [37].

However, improving ecological compensation efficiency may involve short-term adjustment costs. A more performance-oriented compensation mechanism may strengthen environmental constraints, increase compliance costs for some firms, and accelerate the transformation or exit of pollution-intensive activities, thereby placing temporary pressure on traditional industries, employment, and local fiscal revenue. Nevertheless, from a dynamic perspective, these short-term costs are likely to be outweighed by the long-term benefits of externality internalization, more efficient fiscal resource allocation, improved ecological governance performance, and stronger cross-regional coordination. By guiding production factors toward greener and more productive sectors, ecological compensation efficiency can ultimately strengthen the adaptive and transformative capacity of urban economic systems under uncertainty.

Accordingly, this paper proposes Hypothesis 1:

Hypothesis 1 (H1).

Ecological compensation efficiency has a positive overall effect on urban economic resilience.

2.2. The Indirect Impacts of Ecological Compensation Efficiency on Urban Economic Resilience

In this study, ecological compensation efficiency is regarded as a policy-oriented efficiency concept rather than a general environmental efficiency indicator. It emphasizes the performance of ecological compensation and ecological governance under resource constraints, thereby providing the basis for examining how efficiency-oriented ecological governance may influence urban economic resilience through green technological innovation and digital–real economy integration.

Beyond its direct effect, ecological compensation efficiency may influence urban economic resilience by strengthening cities’ technological and digital capabilities. From the perspective of the resource-based view, green technological innovation and digital–real economy integration can be regarded as strategic resources and dynamic capabilities that support resource use efficiency, environmental compliance, information processing, resource matching, organizational flexibility, and industrial transformation. Therefore, this study identifies these two variables as key channels through which ecological compensation efficiency may enhance the resistance, recovery, adaptive, and transformative capacities of urban economic systems.

Green technological innovation can improve resource use efficiency, reduce environmental compliance costs, and promote green industrial upgrading. Because green technological innovation generates both knowledge spillovers and environmental benefits, innovators may find it difficult to fully internalize its returns. This can lead to underinvestment under market mechanisms [38], making external incentives from ecological compensation mechanisms necessary. As ecological compensation efficiency improves, compensation funds can be more effectively transformed into incentives for green development, encouraging local governments and enterprises to invest in the research, development, and application of energy-saving and environmental protection technologies.

Furthermore, green technological innovation can enhance urban economic resilience through several pathways. First, it can reduce firms’ pollution control costs and environmental compliance pressures, thereby improving cost efficiency and environmental performance [39]. Stronger innovation capacity also helps economic systems maintain greater resistance and recovery capacity under external shocks, which in turn strengthens urban economic resilience [40]. Second, green technological innovation improves resource use efficiency and production organization, enabling firms to resume operations more rapidly after shocks and thereby strengthening the recovery capacity of urban economic systems [41]. Finally, it helps traditional industries move toward low-carbon development and creates favorable conditions for new sectors such as clean energy, environmental services, and energy-saving industries. By reducing reliance on pollution- and energy-intensive growth models, green technological innovation strengthens the adaptive and transformative capacity of urban economic systems.

Accordingly, this paper proposes Hypothesis 2:

Hypothesis 2 (H2).

Ecological compensation efficiency enhances economic resilience through green technology innovation.

Digital–real economy integration refers to the incorporation of digital technologies into production, governance, and organizational activities within the real economy. Through data flows and intelligent decision-making, it can enhance operational efficiency and improve the ability of economic systems to respond to risks. Ecological compensation efficiency may promote digital–real economy integration through the governance demand generated by performance-oriented ecological compensation. Improving efficiency requires more precise allocation of ecological governance resources and the dynamic monitoring of governance performance, thereby increasing the demand for digital technologies. Local governments can use these technologies to improve environmental supervision, performance evaluation, and cross-regional coordination, while enterprises can apply them to cleaner production, resource management, and green supply-chain coordination. In this sense, ecological compensation efficiency does not promote digital–real economy integration mechanically, but rather through the practical need for more precise, transparent, and performance-based ecological governance. Existing studies also show that digital technology development can strengthen environmental governance capacity [42], while digital–real economy integration can facilitate firms’ green innovation and support industrial transformation and upgrading [43].

Furthermore, digital–real economy integration can strengthen urban economic resilience through three channels: information response, resource allocation, and industrial adjustment. First, by embedding digital technologies into the real economy, it can improve information transmission efficiency and enable governments and enterprises to identify risks and respond more quickly to environmental risks or market fluctuations [44]. Second, it can enhance the allocation efficiency of urban resources through algorithm-based decision support, platform-mediated matching, and information sharing, thereby strengthening the capacity of urban economic systems to mobilize resources and recover from shocks. Moreover, the embedding of digital technologies into real-sector activities can accelerate the intelligent transformation of traditional industries and facilitate industrial upgrading. This process helps weaken cities’ reliance on conventional growth paths and improves the adaptive and transformative capacity of urban economic systems [45].

Accordingly, this paper proposes Hypothesis 3:

Hypothesis 3 (H3).

Ecological compensation efficiency enhances economic resilience through digital–real economy integration.

2.3. Spatial Spillover Effects of Ecological Compensation Efficiency on Economic Resilience

Ecological compensation efficiency not only affects local urban economic resilience but may also generate spatial spillover effects on neighboring cities through the cross-boundary flow of ecosystem services, the diffusion of governance experience, and the coordinated allocation of production factors.

First, water bodies, atmospheric processes, and ecosystem services in watershed regions often extend across administrative boundaries. When ecological compensation efficiency improves in one area, ecological governance inputs can more effectively enhance ecosystem services and benefit neighboring areas through upstream–downstream linkages. This can reduce cities’ exposure to environmental risks and strengthen the resistance capacity of their economic systems [46].

Second, compensation standards, governance models, and performance evaluation methods formed under an efficient ecological compensation mechanism may generate policy diffusion and demonstration effects. Through intergovernmental learning and regional cooperation, these practices can spread to surrounding areas, reduce policy trial-and-error costs and coordination costs, and improve neighboring cities’ capacity for environmental risk response and resource mobilization. This, in turn, enhances the recovery capacity of their economic systems [47].

Finally, by improving regional environmental quality, ecological compensation efficiency can promote the orderly flow of capital, technology, and green industrial activities within the region. It can also facilitate industrial structure optimization, thereby strengthening the adaptive and transformative capacities of neighboring urban economies [48].

Accordingly, this paper proposes Hypothesis 4:

Hypothesis 4 (H4).

Ecological compensation efficiency generates positive spatial spillover effects on urban economic resilience.

3. Methodology and Materials

This study adopts a quantitative research design based on city-level panel data. To match the panel structure of the data, this study employs econometric models that exploit both cross-city and intertemporal variation. Such models help distinguish the effect of ecological compensation efficiency from unobservable city attributes and common annual disturbances. Moreover, the Yangtze River Economic Belt is characterized by strong intercity economic connections and shared watershed-based ecological processes. Spatial econometric methods are therefore applied to examine whether the effect of ecological compensation efficiency extends beyond local cities and influences neighboring areas. The super-efficiency SBM model was estimated using MaxDEA 9 (Realworld Software Company, Beijing, China). All econometric analyses were performed using Stata 17.0 (StataCorp LLC, College Station, TX, USA).

3.1. Model Construction

To examine the direct effect of ecological compensation efficiency on economic resilience in the Yangtze River Economic Belt, this study specifies the following baseline regression model based on the preceding theoretical analysis:

$$RE{S}_{it}={\alpha }_0+{\alpha }_{1}EC{E}_{it}+{\alpha }_2{\displaystyle \sum _{i=1}^{5}Contro{l}_{it}}+u_i+v_t+{\epsilon }_{it}$$

(1)

In this specification, ${RES}_{it}$ denotes the resilience level of the economy in city during year $t$, while ${ECE}_{it}$ is the corresponding ecological compensation efficiency. ${Control}_{it}$ includes the set of covariates used in the regression. City-specific and year-specific effects are captured by $u_i$ and $v_t$, respectively, and ${\epsilon }_{it}$ is the disturbance term. The parameters ${\alpha }_0$, ${\alpha }_1$, ${\alpha }_2$ are the coefficients to be estimated, respectively.

Model (1) is designed to assess whether improvements in ecological compensation efficiency are linked to greater resilience in urban economies, conditional on differences in economic development, industrial upgrading, financial conditions, government intervention, and urbanization. The estimated coefficient of ${ECE}_{it}$ reflects the average marginal association between ecological compensation efficiency and economic resilience. A significantly positive estimate would imply that cities using ecological compensation resources more efficiently tend to display stronger capacity to withstand, recover from, and adapt to economic shocks.

To further investigate the potential mechanisms through which ecological compensation efficiency affects urban economic resilience, this study follows Jiang (2022) [49] and adopts a two-step mechanism-testing approach. The first step estimates the total effect of ecological compensation efficiency on urban economic resilience using the baseline model. The second step examines whether ecological compensation efficiency significantly affects the proposed mechanism variables. This approach does not aim to provide a strict decomposition of mediation effects, but rather to identify possible channels through which ecological compensation efficiency influences urban economic resilience. Compared with the traditional mediation-effect test, it helps avoid potential over-control bias caused by directly including post-treatment mechanism variables in the baseline regression. Based on this logic, the following model is specified:

$$M_{it}={\gamma }_0+{\gamma }_{1}EC{E}_{it}+{\gamma }_2{\displaystyle \sum _{i=1}^{5}Contro{l}_{it}}+u_i+v_t+{\epsilon }_{it}$$

(2)

In Equation (2), ${\mathrm{M}}_{it}$ refers to the proposed mechanism variable for city $i$ in year $t$, including green technological innovation and the degree of digital–real economy integration. The key parameter is ${\gamma }_1$, which measures whether changes in ecological compensation efficiency are associated with variations in the corresponding mechanism variable. Other terms are defined in the same way as in the baseline model.

In practical terms, Model (2) is used to identify possible transmission channels. Specifically, it tests whether ecological compensation efficiency significantly promotes green technological innovation or digital–real economy integration. If ecological compensation efficiency has a significantly positive effect on these mechanism variables, it suggests that these channels may help explain how ecological compensation efficiency contributes to urban economic resilience.

To assess whether ecological compensation efficiency generates spatial spillovers on economic resilience, this study further employs a spatial Durbin model (SDM), specified as follows:

$$RE{S}_{it}={\alpha }_0+\rho WRE{S}_{it}+{\beta }_{1}EC{E}_{it}+{\theta }_{1}WEC{E}_{it}+{\beta }_2{\displaystyle \sum _{i=1}^{5}Contro{l}_{it}}+{\theta }_{2}WContro{l}_{it}+u_i+v_t+{\epsilon }_{it}$$

(3)

In the SDM, $\rho$ characterizes the spatial dependence of urban economic resilience, reflecting how the resilience level of nearby cities is related to that of the focal city. ${\beta }_1$ and ${\theta }_1$ are the coefficients of local ecological compensation efficiency and its spatially lagged term, respectively. ${\beta }_2$ and ${\theta }_2$ are the coefficients of the local and spatially lagged control variables, respectively. $W$ denotes the spatial linkage matrix among cities. The benchmark spatial regression is estimated using the economic distance matrix.

In practical terms, Model (3) considers spatial linkages among cities. It examines not only whether ecological compensation efficiency affects the economic resilience of the local city, but also whether improvements in ecological compensation efficiency in one city generate spillover effects on neighboring cities. In this model, the spatially lagged terms capture the influence transmitted through inter-city economic and geographic connections. For a more intuitive explanation of the direct and indirect effects in the spatial Durbin model (An intuitive way to understand the spatial Durbin model is to view cities as interconnected ponds in a broader watershed system. When ecological compensation efficiency improves in one pond, the benefits are not limited to that pond alone; cleaner water and improved governance may also spread to nearby ponds. In this sense, the direct effect refers to the change occurring within the city itself, whereas the indirect effect reflects the influence transmitted to surrounding cities through spatial linkages).

As additional checks for spatial robustness, this study also re-estimates the spatial model using an adjacency matrix and an inverse geographic distance matrix. The economic distance matrix used in the benchmark specification is defined as follows:

$$W_{ij}=\left\{\begin{array}{l}{\displaystyle \frac{1}{\left|\overline{Y_i}-\overline{Y_j}\right|}}\\ 0,i=j\end{array}\right.,i\ne j$$

(4)

where $\overline{Y_i}$ and $\overline{Y_j}$ represent the sample-period average of real GDP per capita for city $i$ and city $\mathrm{j}$, respectively. After construction, the matrix is normalized by rows and then used in the spatial econometric estimation.

3.2. Variables

3.2.1. Dependent Variable

Urban economic resilience reflects the ability of an urban economic system to perform better or worse than an expected benchmark growth path when facing macroeconomic fluctuations or external shocks [13,50]. Following the sensitivity-index logic proposed by Martin et al. (2019) [51], this study uses national real GDP growth as the benchmark and measures urban economic resilience by comparing each city’s actual GDP change with its expected change under the national growth trend. A positive deviation indicates stronger-than-benchmark performance and therefore higher economic resilience, whereas a negative deviation indicates weaker resilience. The calculation is as follows:

$$R_{it}=(\Delta Y_i^t-\Delta E)/\left|\Delta E\right|$$

(5)

In Equation (5), ${ R}_{it}$ is the relative resilience measure for city $i$ in year $t$. $\Delta Y_i^t$ indicates the actual variation in the city’s real GDP, while $\Delta E$ refers to the benchmark change implied by the national macroeconomic growth trend. The two terms are further specified in Equations (6) and (7).

$$\Delta Y_i^t=Y_i^t-Y_i^{t-k}$$

(6)

$$\Delta E=\left[\left(Y_E^t-Y_E^{t-k}\right)/Y_E^{t-k}\right]Y_i^{t-k}$$

(7)

where $Y_i^t$ and $Y_i^{t-k}$ represent the real GDP of city $i$ in year $t$ and year $t-k$, respectively.${ Y}_E^t$ and ${ Y}_E^{t-k}$ denote the national real GDP in year $t$ and year $t-k$, respectively. Based on this, the complete expression for relative economic resilience can be derived as follows:

$$R_{it}={\displaystyle \frac{(Y_i^t-Y_i^{t-k})/Y_i^{t-k}-(Y_E^t-Y_E^{t-k})/Y_E^{t-k}}{\left|(Y_E^t-Y_E^{t-k})/Y_E^{t-k}\right|}}$$

(8)

To facilitate subsequent comparative analysis, the results of Equation (8) are centralized.

$$RE{S}_{it}=R_{it}-{\displaystyle \sum _{i=1}^n{R}_{it}/n}$$

(9)

where $n$ is the number of cities, and $RES$ is the standardized urban economic resilience.

As an alternative measurement, this study constructs an alternative urban economic resilience index following Wang and Chen [52]. The index covers three dimensions: resistance and recovery, adjustment and regulation, and transformation and development. The entropy-weight method is used to determine indicator weights, and the final index is obtained by weighted aggregation. The indicator system and weights are shown in Table 1.

Table 1. Indicator system for the alternative urban economic resilience index.

Dimension	Tertiary Indicator	Weight
Resistance and recovery capacity	Per capita GDP	0.0467
	Per capita disposable income of urban residents	0.0208
	Savings balance of urban residents	0.0010
	Number of registered urban unemployed persons	0.1096
	Ratio of total imports and exports to GDP	0.0022
Adjustment and regulation capacity	Local fiscal revenue–expenditure ratio	0.0237
	Total retail sales of consumer goods	0.0959
	Share of the tertiary industry in GDP	0.0072
	Deposit–loan ratio of financial institutions at year-end	0.0232
	Fixed asset investment	0.0839
Transformation and development capacity	Number of patents granted	0.1826
	Number of students enrolled in regular higher education institutions per 10,000 persons	0.0864
	Fiscal expenditure on science and technology	0.0812
	Fiscal expenditure on education	0.2357

3.2.2. Core Independent Variable

Ecological compensation efficiency is measured using the super-efficiency SBM model with undesirable outputs. Compared with traditional DEA models, the SBM model incorporates input and output slacks, making it more suitable for efficiency evaluation when undesirable outputs are present [53]. The super-efficiency extension further allows efficiency scores to exceed 1, enabling the ranking of efficient decision-making units [54]. Since ecological compensation involves multiple inputs, desirable outputs, and undesirable outputs, this model is appropriate for evaluating whether compensation-related resources are effectively transformed into ecological, economic, and social outcomes while reducing pollution and social costs. The model is specified as follows:

$$\min \rho ={\displaystyle \frac{1+\frac{1}{m}{\displaystyle \sum _{i=1}^m\frac{s_i^{-}}{x_{ik}}}}{1-\frac{1}{q_1+q_2}\left({\displaystyle \sum _{r=1}^{q_1}\frac{s_r^{+}}{y_{rk}}+{\displaystyle \sum _{t=1}^{q_2}\frac{s_t^{b-}}{b_{rk}}}}\right)}}$$

(10)

$$s.t.\left\{\begin{array}{l}{\displaystyle \sum _{j=1,j\ne k}^n{x}_{ij}{\lambda }_j-s_i^{-}\le x_{ik}}\\ {\displaystyle \sum _{j=1,j\ne k}^n{y}_{rj}{\lambda }_j+s_i^{+}\ge y_{rk}}\\ {\displaystyle \sum _{j=1,j\ne k}^n{b}_{tj}{\lambda }_j-s_t^{b-}\le b_{tk}}\\ 1-{\displaystyle \frac{1}{q_1+q_2}}\left({\displaystyle \sum _{r=1}^{q_1}{s}_r^{+}/y_{rk}+{\displaystyle \sum _{t=1}^{q_2}{s}_t^{b-}/b_{rk}}}\right)>0\\ \lambda ,s^{-},s^{+}\ge 0\end{array}\right.$$

(11)

In this model, $\rho$ is the ecological compensation efficiency score derived from the objective function, and its value is allowed to be greater than 1. $x_i\left(i=1,2,\dots ,m\right)$, $y_r\left(r=1,2,\dots ,q_1\right)$ and $b_t\left(t=1,2,\dots ,q_2\right)$ denote the input variables, desirable output variables, and undesirable output variables, respectively. $s_i^{-}$, $s_i^{+}$ and $s_t^{b-}$ are the corresponding slack variables for inputs, desirable outputs, and undesirable outputs, while $\lambda$ denotes the weight vector.

Ecological compensation efficiency reflects each city’s capacity to transform compensation-related fiscal, technological, human, and resource inputs into environmental improvement, economic development, and social welfare enhancement while reducing pollution emissions and social costs. It differs from general environmental efficiency, which mainly evaluates the relationship between production inputs, economic outputs, and pollution emissions in economic production. By contrast, ecological compensation efficiency is more policy-oriented, emphasizing the performance of ecological compensation and ecological governance under resource constraints.

In terms of inputs, transfers from the central and provincial governments to cities reflect fiscal support from higher-level governments for local ecological governance and public service provision, and constitute an important funding source for ecological compensation. The share of environmental protection expenditure captures local fiscal commitment to pollution control and ecological restoration, while the share of science and technology expenditure reflects technological and innovation support in ecological governance. Employment in the water conservancy, environment, and public facilities management sector presents labor input in ecological governance, environmental regulation, and public facilities maintenance. Total water consumption is included to reflect the resource input associated with urban water use.

For desirable outputs, this study uses the proportion of days with good air quality, per capita park green space, the proportion of villages with centralized waste treatment, the effective irrigation rate, and the comprehensive utilization rate of industrial solid waste as indicators. These indicators capture ecological and environmental outcomes in terms of air quality improvement, urban ecological space, rural living environment governance, agricultural water conservancy support, and resource recycling. Per capita net income of rural residents and per capita disposable income of urban residents are included to reflect possible welfare gains associated with ecological compensation. Accordingly, the desirable outputs capture the ecological, economic, and livelihood benefits generated by ecological compensation.

For undesirable outputs, industrial wastewater discharge, sulfur dioxide emissions from industrial waste gas, and industrial smoke and dust emissions are selected to capture water pollution and air pollution pressures. The use of agricultural fertilizers and pesticides reflects agricultural non-point source pollution, while the registered urban unemployment rate captures potential social costs associated with ecological governance and industrial adjustment. Incorporating these outputs helps avoid evaluating ecological compensation efficiency solely from the perspective of economic growth or ecological improvement, while ignoring pollution emissions and social costs.

In summary, the indicator system developed in this study covers fiscal, technological, human, and resource inputs in ecological compensation, as well as environmental improvement, economic development, welfare enhancement, pollution emissions and social costs. It therefore provides a relatively comprehensive measure of ecological compensation efficiency for cities in the Yangtze River Economic Belt. Drawing on Dong [33], Long [55], Hou [56], and related studies, this study constructs the input and output indicator system reported in Table 2.

Table 2. Selected indicators of ecological compensation efficiency.

Category	Indicator
Input	Transfers from the central and provincial governments to cities
	The share of environmental protection expenditure
	The share of science and technology expenditure
	Employees in water conservancy, environment, and public facilities management sector
	Total water consumption
Desirable Output	Proportion of days with good air quality
	Per capita park green space
	Proportion of villages with centralized waste treatment
	Effective irrigated rate
	Comprehensive utilization rate of industrial solid waste
	Per capita net income of rural households
	Per capita disposable income of urban households
Undesirable output	Industrial wastewater discharge
	Sulphur dioxide emissions from industrial exhaust
	Industrial soot and dust emissions
	Agricultural fertilizer and pesticide use
	Registered urban unemployment rate

3.2.3. Control Variables

To mitigate potential omitted-variable bias, this study refers to the relevant literature [57,58] and incorporates several city-level control variables. Economic development level (ECO) is measured using the average intensity of DMSP-OLS nighttime light data for China. Industrial structure upgrading (IND) is defined as the proportion of the combined value added of the secondary and tertiary industries in regional GDP. Financial development (FIN) is calculated as the year-end balance of deposits and loans of financial institutions divided by regional GDP. Government intervention (GOV) is measured by the share of local general public budget expenditure in regional GDP. The urbanization rate (UR) is represented by the proportion of urban permanent residents in the total regional population.

3.2.4. Mechanism Variables

Green technological innovation (GTI) is an important mechanism for reconciling economic growth with environmental protection and provides key support for strengthening urban economic resilience. Following Hong et al. [59], this study measures green technological innovation using the share of green invention patents in total invention patents at the city level.

Digital–real economy integration (DRE) describes the technological connections and industrial linkages between the digital economy and the real economy. It reflects the extent to which digital technologies are incorporated into traditional industries and production activities. Following Kwon et al. (2020) [60], this study adopts a technology-driven perspective on industrial integration and uses patent co-classification analysis to convert technological relatedness into industrial linkage information. Specifically, patents jointly classified into digital technology fields and real-economy industrial fields indicate the technological connections between digital capabilities and traditional production activities. This measure captures the extent to which digital technologies are integrated into real economy sectors, thereby reflecting the level of digital–real economy integration.

3.3. Sample Selection and Data Sources

This study uses balanced panel data for 108 prefecture-level cities in the Yangtze River Economic Belt from 2011 to 2023. Prefecture-level cities are selected because they are key units for implementing ecological compensation, environmental governance, and urban development policies, while also allowing the study to capture intra-provincial differences that provincial-level data cannot fully reflect. This study focuses on the Yangtze River Economic Belt because it combines a strategic national economic corridor with intensive watershed ecological protection and cross-regional ecological compensation practices. Tongren and Bijie are omitted from the final balanced panel owing to incomplete data. The economic development variable is constructed using DMSP data, the urbanization rate is compiled from municipal statistical bureaus, and green technological innovation is measured using data from the Chinese Research Data Services Platform (CNRDS). The remaining variables are obtained from the China City Statistical Yearbook, China Environmental Yearbook, China Energy Statistical Yearbook, municipal Statistical Bulletins on National Economic and Social Development, and the China Economic and Social Big Data Research Platform. Missing observations were supplemented, where possible, using municipal statistical yearbooks, Statistical Bulletins on National Economic and Social Development, and relevant official databases. Most missing values result from incomplete disclosure or changes in city-level reporting, especially for environmental indicators such as industrial wastewater discharge after 2019. For isolated missing values, linear interpolation was applied within each city to maintain the continuity and comparability of the balanced panel. Table 3 reports the descriptive statistics.

Table 3. Descriptive statistical analysis.

Variables	N	Mean	Std. Dev.	Min	Max
RES	1404	1.1567	0.7546	−4.3143	8.4068
ECE	1404	0.8798	0.1644	0.4206	1.1493
GTI	1404	0.0893	0.0392	0.0143	0.5890
DRE	1404	4.2504	11.7408	0.0001	116.7996
ECO	1404	8.9817	10.0899	0.2376	53.9171
IND	1404	0.8875	0.0638	0.6777	0.9979
FIN	1404	2.5431	1.0089	0.7642	7.1745
GOV	1404	0.1757	0.0741	0.0666	0.6211
UR	1404	0.5626	0.1306	0.1815	0.8960

It is worth noting that urban economic resilience and digital–real economy integration exhibit relatively high dispersion. This is mainly because the two variables capture relative economic performance and the linkage intensity between the digital economy and the real economy, both of which may vary substantially across cities.

4. Results

4.1. Baseline Regression Analysis

Table 4 shows the benchmark regression results. The model is estimated under four specifications, ranging from a simple regression to specifications that gradually include control variables, city fixed effects, and year fixed effects. Across these specifications, the coefficient of ecological compensation efficiency is consistently positive and significant at the 1% level. This indicates that ecological compensation efficiency has a stable positive association with urban economic resilience. Therefore, Hypothesis 1 is supported.

Table 4. Results of the baseline regression analysis.

Variables	(1)	(2)	(3)	(4)
	RES	RES	RES	RES
ECE	0.4679 ***	0.4759 ***	0.4775 ***	0.4969 ***
	(4.2445)	(3.6111)	(4.2832)	(3.5683)
ECO			0.0018	0.0159 **
			(0.7237)	(2.3357)
IND			0.0146	5.5985 ***
			(0.0208)	(4.4445)
FIN			0.0459 *	−0.4538 ***
			(1.9285)	(−4.3333)
GOV			−0.8382 **	−1.8190 *
			(−2.0860)	(−1.8802)
UR			−0.9733 ***	−2.3153 ***
			(−2.6585)	(−2.6896)
City FE	NO	YES	NO	YES
Year FE	NO	YES	NO	YES
Cons_	0.7450 ***	0.7380 ***	1.2855 **	−1.6161
	(8.7784)	(6.3639)	(2.3682)	(−1.4197)
N	1404	1404	1404	1404
R2	0.0097	0.1481	0.0163	0.2035

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

In terms of economic significance, the coefficient in Column (4) indicates that a one-unit increase in ecological compensation efficiency is associated with a 0.4969-unit increase in urban economic resilience. Given that the standard deviation of ecological compensation efficiency is 0.1644, a one-standard-deviation increase corresponds to an increase of approximately 0.0817 units in urban economic resilience, accounting for about 10.8% of its standard deviation. This suggests that the estimated effect is not only statistically significant but also economically meaningful. In the context of the Yangtze River Economic Belt, higher ecological compensation efficiency means that ecological governance inputs can be more effectively transformed into environmental improvement, welfare enhancement, and pollution reduction. This process helps convert the positive externalities of environmental protection into economic incentives, encourages resources to shift away from highly polluting and energy-intensive activities, and strengthens the adaptive and transformative capacity of urban economic systems under external shocks. Although the R² is relatively modest, this is not unusual for city-level panel models of economic resilience, which is shaped by multiple structural factors and external shocks.

For the control variables, industrial structure upgrading has a significantly positive effect on urban economic resilience, with a relatively large coefficient. This may be because industrial upgrading reduces dependence on resource-intensive activities and strengthens the diversity and innovation capacity of the urban economy. By contrast, financial development shows a significantly negative association with economic resilience. One possible explanation is that excessive financial development may draw resources away from the real economy and increase leverage risk and asset price volatility, thereby weakening the shock resistance and recovery capacity of urban economic systems [61]. The urbanization rate also has a negative effect, which may be related to the resource overconsumption, environmental degradation, rising inequalities, and governance difficulties associated with rapid and extensive urbanization [62].

4.2. Endogeneity Test

Given the possible two-way relationship between ecological compensation efficiency and urban economic resilience, this study further adopts the two-stage least squares (2SLS) method to alleviate endogeneity concerns related to reverse causality and omitted-variable bias. Drawing on previous studies that construct instrumental variables by interacting exogenous geographical distance with time-varying factors [63], this study constructs an instrumental variable using the interaction between each city’s spherical distance from Hangzhou and the one-period lag of ecological compensation efficiency [64]. This variable is denoted as $IV$.

The validity of this instrumental variable is mainly reflected in two aspects. First, regarding the relevance, Hangzhou is an important core city in the Yangtze River Economic Belt, especially in the Yangtze River Delta region, and has demonstration and spillover effects in ecological governance, institutional innovation, and green development. Cities closer to Hangzhou are more likely to be influenced by its policy diffusion, governance experience, and regional coordination effects. Therefore, geographical distance to Hangzhou, interacted with the lagged ecological compensation efficiency, can provide variation for current ecological compensation efficiency. Second, regarding exogeneity, the spherical distance between a city and Hangzhou is determined by historical geographical location and is unlikely to be affected by current urban economic resilience in the short run. Meanwhile, using lagged ecological compensation efficiency helps reduce the possibility that the instrumental variable is directly correlated with the current error term.

Table 5 reports the endogeneity test results. The first-stage estimates confirm a significant positive association between the instrumental variable and ecological compensation efficiency at the 1% level. Meanwhile, the Kleibergen–Paap rk LM and Wald F statistics indicate that the instrumental variable passes the under–identification and weak–instrument tests. In the second-stage regression, ecological compensation efficiency continues to exert a positive effect on urban economic resilience at the 5% significance level. These results are consistent with the benchmark estimates and reinforce the reliability of the main conclusion.

Table 5. Results of endogeneity tests.

Variables	2SLS-Stage (1)	2SLS-Stage (2)
	ECE	RES
IV	0.0698 *** (15.6707)
ECE		0.7093 ** (2.3144)
ECO	0.0019 * (1.9299)	0.0028 (1.0067)
IND	0.0843 (0.4367)	0.1040 (0.1308)
FIN	−0.0058 (−0.6402)	0.0528 * (1.8231)
GOV	0.4344 *** (3.9959)	−1.0314 ** (−2.0043)
UR	0.2298 *** (2.6340)	−1.2014 *** (−2.6834)
Cons_	0.2049 (1.2907)	1.1396 * (1.8049)
N	1284	1284
Kleibergen-Paap rk LM	52.029 ***
Wald F	245.57

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

4.3. Robustness Test

The reliability of the baseline findings is further examined through several robustness checks. Specifically, this study re-estimates the model by adopting an alternative measure of urban economic resilience, removing observations from the COVID-19 and post-pandemic recovery period, and applying 1% winsorization to all variables. The corresponding results are presented in Table 6.

Table 6. Results of the robustness test.

Variables	(1)	(2)	(3)
	Replacement of the Dependent Variable	Exclusion of Special Year Effects	Exclusion of Extreme Outliers
ECE	0.0261 ***	0.2766 **	0.4440 ***
	(3.1894)	(1.9826)	(3.3877)
ECO	0.0038 ***	0.0145 *	0.0115 *
	(4.3308)	(1.8277)	(1.6674)
IND	−0.0808	9.2314 ***	4.4497 ***
	(−1.3547)	(5.8010)	(4.0816)
FIN	0.0047	−0.5207 ***	−0.3456 ***
	(0.5862)	(−4.7464)	(−3.5517)
GOV	−0.0396	−1.8821 *	−2.1715 **
	(−0.9895)	(−1.7030)	(−2.4702)
UR	0.0126	−3.0903 ***	−2.3791 ***
	(0.1592)	(−2.9886)	(−3.0325)
City FE	YES	YES	YES
Year FE	YES	YES	YES
Cons_	0.0989	−4.2633 ***	−0.6932
	(1.2836)	(−3.2689)	(−0.7535)
N	1404	1188	1404
R2	0.9084	0.2608	0.2318

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

For the first test, the dependent variable is replaced with a composite resilience index constructed in Section 3.2.1. Compared with the GDP deviation-based measure used in the benchmark analysis, this alternative indicator provides a broader characterization of resilience. As reported in Column (1), ecological compensation efficiency remains significantly positive, suggesting that the main conclusion is not sensitive to the way urban economic resilience is measured.

For the second test, observations from 2020 to 2021 are removed to account for the possible disturbance caused by the COVID-19 pandemic and the subsequent recovery period. These two years were characterized by substantial disruptions to urban production, employment, industrial operations, and local fiscal conditions. After excluding this period, ecological compensation efficiency still shows a positive and statistically significant coefficient, indicating that the baseline conclusion is not driven by the direct impact of the pandemic shock.

For the third test, this study applies 1% winsorization to all variables to reduce the disturbance caused by extreme values. Column (3) shows that the coefficient of ecological compensation efficiency is still significantly positive. Overall, the robustness checks demonstrate that the main finding remains valid after using an alternative resilience measure, excluding the special period, and treating outliers.

4.4. Quantile Regression Analysis

To examine whether the effect of ecological compensation efficiency varies across the conditional distribution of urban economic resilience, this study estimates quantile regressions at the 0.2, 0.4, 0.6, and 0.8 quantiles. The results are reported in Table 7. At the 0.2 quantile, the coefficient of ecological compensation efficiency is positive but statistically insignificant. From the 0.4 quantile onward, the coefficient becomes significantly positive and increases gradually at higher quantiles. This suggests that the resilience-enhancing effect of ecological compensation efficiency becomes more pronounced among cities with higher levels of economic resilience.

Table 7. Quantile regression results.

Variables	(1)	(2)	(3)	(4)
	0.2	0.4	0.6	0.8
ECE	0.118 (0.237)	0.190 * (0.106)	0.301 *** (0.114)	0.420 ** (0.208)
ECO	0.021 (0.013)	0.014 ** (0.006)	0.015 ** (0.006)	0.011 (0.012)
IND	0.637 (1.739)	1.732 ** (0.780)	4.167 *** (0.835)	3.190 ** (1.531)
FIN	−0.152 (0.132)	−0.280 *** (0.059)	−0.277 *** (0.063)	−0.305 *** (0.116)
GOV	−0.929 (1.181)	−0.633 (0.530)	−0.402 (0.567)	−0.263 (1.039)
UR	−0.598 (1.012)	−1.221 *** (0.454)	−1.746 *** (0.486)	−1.816 ** (0.890)
City FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
Cons_	0.652 (2.032)	1.039 (0.911)	−1.005 (0.976)	0.249 (1.789)
N	1404	1404	1404	1404

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

High-resilience cities generally have stronger economic foundations, more diversified industrial structures, greater fiscal and governance capacity, and higher levels of innovation. These conditions enable them to absorb ecological compensation resources more effectively and convert improvements in ecological compensation efficiency into environmental quality, green technological progress, industrial upgrading, and risk-response capacity. In contrast, low-resilience cities often face weaker industrial bases, greater dependence on traditional or resource-intensive sectors, limited fiscal space, and insufficient innovation capacity. As a result, the transformation of ecological governance performance into economic resilience may be constrained, and the effect of ecological compensation efficiency may be delayed or weakened by short-term adjustment costs such as industrial restructuring pressure and environmental compliance costs.

4.5. Heterogeneity Analyses

To examine whether the effect of ecological compensation efficiency differs across city characteristics, this study conducts heterogeneity analyses from five dimensions: urban location, city size, administrative hierarchy, natural resource endowment, and industrial structure. The results are reported in Table 8 and Table 9.

Table 8. Regression results for heterogeneity in urban location and size.

Variables	Urban Location			Urban Scale
	Upstream Cities	Midstream Cities	Downstream Cities	Large Cities	Medium-Sized Cities	Small Cities
ECE	−0.0729	0.1144	0.8957 ***	0.2754	0.9605 ***	0.4811
	(−0.2595)	(0.4772)	(3.5830)	(1.2932)	(3.6237)	(1.0998)
ECO	−0.0005	0.0121	0.0051	0.0154	0.0058	0.0246
	(−0.0133)	(0.6189)	(0.4818)	(1.0353)	(0.4334)	(1.5436)
IND	7.2207 ***	1.6058	6.4415 **	4.6206 **	5.2641 **	6.4111 **
	(3.2180)	(0.5980)	(2.4156)	(2.4056)	(2.4057)	(2.5125)
FIN	−0.5699 ***	−0.0225	−0.3618 **	−0.3884 ***	−0.4562 *	−0.4456 **
	(−3.8599)	(−0.1039)	(−2.1593)	(−2.7368)	(−1.8989)	(−2.7220)
GOV	−0.4756	−4.8373 **	−2.6398	−4.9139 **	−5.9379 ***	−0.6108
	(−0.5010)	(−2.3684)	(−1.5785)	(−2.3882)	(−2.8330)	(−0.7471)
UR	−4.2228 **	0.1167	−2.1535 *	0.4248	−2.5628	−3.2659 **
	(−2.3538)	(0.0709)	(−1.9141)	(0.2781)	(−1.3684)	(−2.4470)
City FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
Cons_	−1.1941	0.3411	−2.8690	−1.7613	−1.0276	−2.1511
	(−0.5123)	(0.1369)	(−1.2020)	(−0.7463)	(−0.5325)	(−0.9661)
N	403	468	533	531	469	403
R2	0.3665	0.2651	0.1917	0.2674	0.2011	0.1797

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

Table 9. Regression results for heterogeneity in urban administrative levels, natural endowments, and industrial structure.

Variables	Administrative Hierarchy		Natural Resource Endowment		Industrial Structure
	Central Cities	Ordinary Prefecture-Level Cities	Non-Resource-Based Cities	Resource-Based Cities	Service-Oriented Stage	Industrialization Stage
ECE	0.0642	0.6088 ***	0.5597 ***	0.2481	0.7988 ***	0.3127
	(0.2889)	(3.8645)	(3.4332)	(0.9259)	(2.9808)	(1.6569)
ECO	0.0303	0.0153 **	0.0113	0.0342	0.0384 *	0.0237 **
	(1.4133)	(2.1242)	(1.3472)	(1.6178)	(1.8046)	(2.1938)
IND	11.8481	5.4248 ***	3.7594 *	7.0137 ***	3.8891	6.0501 **
	(1.5265)	(4.1944)	(1.8032)	(4.3714)	(1.5631)	(2.3731)
FIN	−0.4519 **	−0.4764 ***	−0.2254 *	−0.8576 ***	−0.4507 ***	−0.5454 ***
	(−2.9636)	(−3.7153)	(−1.9577)	(−4.0089)	(−3.2135)	(−2.6712)
GOV	1.5988	−1.7008	−4.0146 **	0.1153	−2.6245 **	−1.7549
	(0.3993)	(−1.6406)	(−2.1353)	(0.1563)	(−1.9970)	(−1.0283)
UR	−3.3624 ***	−2.2742 **	−1.1035	−3.3442 **	−1.9280	−3.6567 **
	(−3.9897)	(−2.3011)	(−1.1701)	(−2.6253)	(−1.2813)	(−2.4985)
City FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
Cons_	−6.3665	−1.6393	−0.9524	−1.6926	−0.3870	−1.2722
	(−0.9034)	(−1.4163)	(−0.4085)	(−1.2426)	(−0.1374)	(−0.6875)
N	143	1261	897	507	623	777
R2	0.2504	0.2037	0.1633	0.2512	0.2231	0.2018

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

First, according to the geographic location of cities in the Yangtze River Economic Belt, the sample is divided into upstream, midstream, and downstream cities. The results show that ecological compensation efficiency significantly improves urban economic resilience only in downstream cities, with a coefficient of 0.8957. This suggests that downstream cities, with stronger economic foundations, more mature market systems, and greater demand for green transformation, are better able to convert improvements in ecological compensation efficiency into resilience enhancement. By contrast, the effects in upstream and midstream cities are insignificant, possibly because these cities face greater industrial transformation pressure and have weaker short-term capacity to transform ecological governance inputs into green productivity.

Second, following the State Council’s 2014 adjustment to the criteria for classifying city sizes [65], cities are grouped by permanent urban population: large cities have more than one million residents, medium-sized cities have between 500,000 and one million residents, and small cities have fewer than 500,000 residents. The results indicate that the effect of ecological compensation efficiency is most pronounced in medium-sized cities, with a coefficient of 0.9605. Medium-sized cities generally have relatively complete infrastructure and industrial foundations while retaining greater flexibility for structural adjustment, making them more responsive to improvements in ecological compensation efficiency. The insignificant effects in large and small cities may be related to the diversified resilience drivers of large cities and the weaker resource absorption capacity of small cities.

Third, based on China’s urban administrative hierarchy, the sample is divided into central cities, including provincial capitals and municipalities directly under the central government, and ordinary prefecture-level cities, including non-capital and non-municipality cities. The coefficient of ecological compensation efficiency is significantly positive for ordinary prefecture-level cities but insignificant for central cities. This may be because ordinary prefecture-level cities rely more directly on local ecological resources and environmental governance performance, whereas central cities have more diversified economic drivers, making the marginal contribution of ecological compensation efficiency less visible.

Fourth, based on the National Sustainable Development Plan for Resource-Based Cities (2013–2020) [66] issued by the State Council, cities are classified into resource-based and non-resource-based groups. The coefficient of ecological compensation efficiency is significant only for non-resource-based cities, while the estimate for resource-based cities is insignificant. This difference may be related to industrial diversification. Non-resource-based cities usually have broader industrial bases and stronger capacity to absorb ecological governance gains, whereas resource-based cities are more likely to rely on resource extraction and heavy industries. Such path dependence may limit the transformation of ecological compensation efficiency into economic resilience.

Finally, cities with a ratio of tertiary industry value added to secondary industry value added greater than 1 are classified as service-oriented cities, whereas those with a ratio below 1 are classified as industrializing cities. The results show that ecological compensation efficiency significantly improves economic resilience in service-oriented cities, but not in industrializing cities. Service-oriented cities are generally more compatible with green transformation, innovation-driven development, and lower-pollution growth models, making them more capable of transforming ecological governance performance into stronger economic resilience.

Overall, the heterogeneity results suggest that the resilience-enhancing effect of ecological compensation efficiency depends on cities’ development stage, industrial structure, resource endowment, and absorptive capacity. These findings indicate that ecological compensation policies should avoid a uniform design and instead be adjusted according to local conditions.

4.6. Mechanism Analysis

To examine the transmission mechanisms through which ecological compensation efficiency affects urban economic resilience, this study tests whether green technological innovation and digital–real economy integration serve as possible channels. Table 10 reports the mechanism test results.

Table 10. Results of the mechanism analysis.

Variables	(1)	(2)
	GTI	DRE
ECE	0.0195 **	4.6131 **
	(2.1658)	(2.1592)
ECO	−0.0008	1.0502 ***
	(−1.0494)	(3.9318)
IND	0.2192 *	−37.6638 **
	(1.6684)	(−2.2962)
FIN	−0.0044	1.7710
	(−0.7013)	(0.6880)
GOV	−0.1367	−8.3363
	(−1.5889)	(−0.7311)
UR	−0.1463 *	−0.3968
	(−1.7556)	(−0.0173)
City FE	YES	YES
Year FE	YES	YES
Cons_	−0.0030	21.3567
	(−0.0209)	(1.0119)
N	1404	1404
R2	0.3276	0.6841

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

As shown in Column (1), ecological compensation efficiency has a significantly positive effect on green technological innovation at the 5% level. This suggests that higher ecological compensation efficiency can encourage local economies to engage in green innovation activities. The result is consistent with theoretical expectation and supports the Porter Hypothesis, which argues that well-designed environmental regulations can stimulate innovation, offset compliance costs, and create new sources of competitive advantage rather than merely constraining economic development. By providing stable policy signals and a clear green orientation, efficient ecological compensation mechanisms reduce the uncertainty associated with technological innovation and encourage enterprises to transform environmental pressure into momentum for green transition. This process helps strengthen their capacity to cope with economic fluctuations and external shocks.

Column (2) shows that ecological compensation efficiency also has a significantly positive effect on digital–real economy integration at the 5% level. This indicates that higher ecological compensation efficiency helps strengthen the connection between digital technologies and real economic activities. By improving the digital allocation of ecological governance resources and supporting the green transformation of the real economy, ecological compensation mechanisms may further contribute to urban economic resilience through digital–real economy integration.

5. Further Analysis: Spatial Spillover Analysis

To investigate the spatial spillover effects of ecological compensation efficiency on urban economic resilience, this study conducts spatial dependence tests before estimating the spatial Durbin model. The Moran’s I statistics and spatial model selection tests are provided in Appendix A. The results indicate that urban economic resilience shows significant spatial dependence in most years, and the LM, LR, Wald, and Hausman tests support the use of the spatial Durbin model. Accordingly, a spatial Durbin model with city and year fixed effects is estimated. Table 11 reports the spatial Durbin model results under three spatial weight matrices: the economic distance matrix, the adjacency matrix, and the inverse geographic distance matrix. The coefficient of ecological compensation efficiency is significantly positive under all three matrices, indicating that improvements in local ecological compensation efficiency can enhance local urban economic resilience. This finding is consistent with the baseline regression results and further confirms the positive effect of ecological compensation efficiency from a spatial perspective. It is also consistent with the international literature on ecological fiscal transfers, which suggests that ecological services often generate benefits beyond local administrative boundaries and that intergovernmental fiscal transfer mechanisms can help internalize such spatial externalities [67].

Table 11. Regression results for the spatial Durbin model.

	(1)	(2)	(3)
	Economic Distance Matrix	Adjacency Matrix	Inverse Geographic Distance Matrix
ECE	0.3835 ** (2.5627)	0.3599 ** (2.4994)	0.3752 ** (2.5629)
ECO	0.0019 (0.1647)	0.0081 (0.7957)	0.0004 (0.0335)
IND	6.5262 *** (5.6577)	6.0094 *** (5.3741)	6.5548 *** (5.7612)
FIN	−0.4362 *** (−4.9801)	−0.4647 *** (−5.5616)	−0.4639 *** (−5.3494)
GOV	−1.8385 ** (−2.4285)	−1.8031 ** (−2.4889)	−1.8547 ** (−2.5166)
UR	−2.4082 *** (−3.7048)	−2.6900 *** (−4.2227)	−2.4209 *** (−3.8340)
City FE	YES	YES	YES
Year FE	YES	YES	YES
N	1404	1404	1404
R2	0.0120	0.0415	0.0054

Note: ***, ** indicate significance at the 1%, 5% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

Since the coefficients of the spatial Durbin model include both local feedback effects and cross-regional spillovers, this study further decomposes the estimated effects into direct, indirect, and total effects. The results are reported in Table 12. The direct effects of ecological compensation efficiency are significantly positive under all three spatial weight matrices, suggesting that higher local ecological compensation efficiency directly strengthens local economic resilience. More importantly, the indirect effects are also significantly positive, indicating that improvements in ecological compensation efficiency in one city can generate positive spillover effects on neighboring cities.

Table 12. Spatial effect decomposition results of the SDM under different spatial weight matrices.

Variables	Matrix Types	Direct Effect	Indirect Effect	Total Effect
ECE	Economic distance matrix	0.4792 *** (2.8696)	10.3065 * (1.8081)	10.7857 * (1.8694)
	Adjacency matrix	0.4336 *** (2.7984)	0.7998 ** (2.2833)	1.2334 *** (2.8894)
	Inverse geographic distance matrix	0.5652 *** (3.0279)	19.8974 ** (1.9696)	20.4626 ** (2.0031)
ECO	Economic distance matrix	0.0018 (0.1717)	0.0451 (0.2831)	0.0469 (0.3000)
	Adjacency matrix	0.0080 (0.8508)	0.0037 (0.2401)	0.0117 (0.7329)
	Inverse geographic distance matrix	0.0023 (0.2234)	0.2596 (0.9168)	0.2619 (0.9293)
IND	Economic distance matrix	5.6902 *** (4.7614)	−108.3005 ** (−2.0178)	−102.6103 * (−1.8955)
	Adjacency matrix	6.0076 *** (5.5735)	−1.3072 (−0.5822)	4.7004 * (1.7972)
	Inverse geographic distance matrix	6.0726 *** (5.2191)	−63.9091 (−1.3105)	−57.8365 (−1.1758)
FIN	Economic distance matrix	−0.4362 *** (−4.9801)	−0.4086 *** (−4.6995)	3.1200 (1.0609)
	Adjacency matrix	−0.4618 *** (−5.5634)	0.0325 (0.2044)	−0.4293 ** (−2.2501)
	Inverse geographic distance matrix	−0.4451 *** (−5.0923)	1.9970 (0.4867)	1.5519 (0.3759)
GOV	Economic distance matrix	−1.8385 ** (−2.4285)	−1.9983 *** (−2.6183)	−19.3685 (−0.6008)
	Adjacency matrix	−1.8727 *** (−2.6192)	−0.9285 (−0.5607)	−2.8012 (−1.4366)
	Inverse geographic distance matrix	−1.9527 ** (−2.4889)	−11.4953 (−0.2821)	−13.4480 (−0.3273)
UR	Economic distance matrix	−2.4082 *** (−3.7048)	−2.1573 *** (−2.9589)	25.5869 (0.9567)
	Adjacency matrix	−2.7078 *** (−4.2501)	−0.4238 (−0.3226)	−3.1315 * (−1.9569)
	Inverse geographic distance matrix	−1.7319 ** (−2.0462)	71.4873 (1.4380)	69.7554 (1.3866)
City FE		YES	YES	YES
Year FE		YES	YES	YES
N		1404	1404	1404

Note: ***, ** and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-statistics based on robust standard errors clustered at the city level.

These findings can be further interpreted in relation to the international literature on PES and EFTs. Existing studies suggest that ecosystem services and ecological governance benefits often extend beyond administrative boundaries, thereby providing a theoretical basis for interregional compensation and ecological fiscal transfers [28,29]. This evidence supports the above argument and shows that ecological compensation efficiency can generate cross-city resilience benefits. Beyond confirming spatial spillovers, this study further demonstrates that the influence of ecological compensation is not limited to environmental improvement or conservation incentives, but also extends to urban economic resilience. This transmission may occur through several channels. First, improvements in ecological governance efficiency can enhance regional environmental quality and ecosystem services, thereby benefiting neighboring cities through watershed and atmospheric linkages. Second, efficient ecological compensation practices may diffuse across regions through policy learning, intergovernmental cooperation, and demonstration effects. Third, better ecological governance can improve the regional allocation of green capital, technology, and industrial activities, thereby supporting the resilience of surrounding cities. However, the magnitude of the indirect effects varies across spatial weight matrices, indicating that such spillovers depend on the strength of intercity economic linkages and geographic proximity. Therefore, ecological compensation policies in the Yangtze River Economic Belt should not only focus on local governance performance, but also strengthen cross-regional coordination and benefit-sharing mechanisms.

6. Conclusions, Implications, and Future Research

6.1. Conclusions

This study uses panel data from 108 prefecture-level cities in the Yangtze River Economic Belt during 2011–2023 and combines the super-efficiency SBM model, two-way fixed effects model, mechanism tests, and spatial Durbin model. It examines the effect of ecological compensation efficiency on urban economic resilience and further analyzes heterogeneity and spatial spillovers. The main findings are as follows.

First, ecological compensation efficiency significantly improves urban economic resilience, and this conclusion remains valid after a series of robustness checks and endogeneity tests. Second, the positive effect becomes more pronounced at higher quantiles of urban economic resilience, suggesting that the resilience-enhancing effect depends on cities’ existing development conditions and absorptive capacity. Third, the effect varies across city types and is more evident in downstream cities, medium-sized cities, ordinary prefecture-level cities, non-resource-based cities, and service-oriented cities. Fourth, the mechanism analysis shows that ecological compensation efficiency strengthens urban economic resilience by promoting green technological innovation and digital–real economy integration. Finally, ecological compensation efficiency generates significant positive spatial spillovers. Improvements in local ecological compensation efficiency not only enhance local economic resilience but also contribute to resilience improvement in neighboring cities. Taken together, these findings suggest that ecological compensation efficiency is not only an indicator of environmental governance performance, but also a policy-related governance capability through which cities transform ecological governance resources into adaptive and transformative economic capacity.

This study contributes to the literature by shifting the focus from the scale or intensity of ecological compensation to ecological compensation efficiency, linking ecological compensation with urban economic resilience, and identifying the mechanisms and spatial spillover effects through which ecological compensation efficiency affects resilience. These findings provide empirical evidence for understanding the relationship between ecological governance performance and resilient urban development at the watershed scale. Overall, ecological compensation should not be regarded only as an environmental protection instrument, but also as an important governance tool for enhancing economic resilience and promoting sustainable regional development. For policymakers, local governments, enterprises, and other stakeholders, improving ecological compensation efficiency, strengthening green innovation and digital transformation, and promoting cross-regional collaboration are essential for building more resilient and sustainable cities.

6.2. Implications

Theoretically, this study extends the literature on ecological compensation, payments for ecosystem services, and ecological fiscal transfers by shifting the focus from compensation scale and policy intensity to ecological compensation efficiency. It emphasizes whether ecological compensation inputs can be effectively transformed into ecological improvement, economic development, livelihood enhancement, and reductions in undesirable outputs. It also enriches the literature on urban economic resilience by identifying ecological compensation efficiency as a policy-related factor affecting the resistance, recovery, adjustment, and transformation capacity of urban economic systems. By incorporating green technological innovation, digital–real economy integration, and spatial spillover effects, this study further explains how efficiency-oriented ecological governance contributes to resilient urban development at the watershed scale.

Practically, ecological compensation should shift from input expansion to performance-oriented governance. Local governments should not only increase compensation inputs, but also strengthen the monitoring, evaluation, and performance management of compensation funds. More attention should be paid to whether these resources are effectively transformed into environmental improvement, livelihood enhancement, pollution reduction, and resilience building. Practices such as Gross Ecosystem Product accounting can provide useful references for linking ecosystem service valuation with performance-based ecological governance. Ecological compensation policies should also be tailored to local conditions. The heterogeneity results show that the effect of ecological compensation efficiency differs across regions, city sizes, resource endowments, and industrial structures. For cities with stronger absorptive capacity, ecological compensation can be combined with green innovation, environmental services, and industrial upgrading. For cities with weaker industrial foundations, ecological compensation should be coordinated with fiscal support, employment transition, and industrial transformation policies to reduce adjustment costs and enhance adaptability. Finally, cross-regional coordination and digital governance should be strengthened. The Xin’an River–Qiandao Lake Basin shows that horizontal ecological compensation can generate measurable ecological benefits when compensation standards are linked to environmental performance. Experiences from the Taihu Lake Basin and Costa Rica’s payment for environmental services program also suggest that ecological compensation mechanisms can be improved through monitoring systems, stable funding arrangements, and clear conservation targets. However, the international experience represented by Costa Rica’s payment for environmental services program should be adapted to the institutional and watershed governance context of the Yangtze River Economic Belt rather than being mechanically replicated. Therefore, ecological compensation should be more closely integrated with ecological value accounting, digital environmental monitoring, water quality early-warning platforms, and cross-regional information sharing, so as to reduce coordination frictions and transform watershed ecological governance into a source of regional economic resilience.

6.3. Limitations and Future Research Agenda

This study has several limitations. First, the analysis is constrained by data availability. Although the sample covers 108 prefecture-level cities in the Yangtze River Economic Belt from 2011 to 2023, Tongren and Bijie are excluded because of missing data, and interpolation may introduce measurement error. Second, the empirical framework mainly examines the average effect of ecological compensation efficiency on urban economic resilience, while nonlinear effects, threshold effects, and time varying spatial interactions remain insufficiently explored. Third, the findings should be generalized with caution, as the Yangtze River Economic Belt has distinctive watershed governance institutions, ecological compensation practices, industrial structures, and regional development conditions.

Future research can extend this study in several directions. First, more detailed and higher-frequency data can be used to improve the measurement of ecological compensation efficiency and urban economic resilience. Second, future studies may further examine nonlinear mechanisms, threshold effects, and dynamic spatial spillovers to better capture the complex relationship between ecological compensation efficiency and economic resilience. Third, comparative studies across different river basins, regions, or countries would help assess whether the conclusions of this study are applicable to broader institutional and geographical contexts.