The Impact of Rural Collective Property Rights System Reform on County-Level Urban–Rural Integration: Evidence from 1106 Counties in China

Abstract

The rural collective property rights system reform (RCPRSR) is a pivotal institutional innovation for revitalizing rural resources, optimizing factor allocation, and advancing urban–rural integration—a core goal of sustainable land use planning. This study evaluates the reform’s impact on county-level urban–rural integration using panel data from 1106 Chinese county-level administrative units during 2013–2020. Treating the staggered rollout of reform pilots as a quasi-natural experiment, we employ a multi-period difference-in-differences approach. The results show that the RCPRSR significantly promotes urban–rural integration, a finding robust to a series of sensitivity checks. The policy effects exhibit marked heterogeneity: the dividends of narrowing the urban–rural development gap are more pronounced in poverty-stricken counties and areas with lower baseline integration levels. Mechanism analysis reveals two pathways—population agglomeration and industrial structure optimization—through which the reform operates, specifically manifested as enhanced county population carrying capacity, accelerated tertiary industry development, and deepened secondary–tertiary industrial integration. These findings provide empirical evidence for optimizing rural property rights reform and advancing sustainable urban–rural development.

1. Introduction

Urban–rural integration (URI) concerns national long-term stability and universal common prosperity, bearing the historical mission of dismantling the urban–rural dual structure and promoting balanced and comprehensive urban–rural development. Its essence lies in the free flow of urban–rural factors with converging returns, functional complementarity, benefit sharing, and equalized development across economic, social, ecological, and other dimensions [1,2,3]. As the node connecting cities and serving rural areas, the county constitutes the core spatial carrier for the transition, connection, convergence, transformation, and integration of urban–rural flow spaces and serves as the pioneer and breakthrough point for achieving high-quality urban–rural integration development. However, at present, the persistent ills of urban–rural factor market barriers and low efficiency in resource allocation constitute the crux of the urban–rural development divide, leading to such challenges as large urban–rural economic and income gaps, the uneven distribution of public resources and services, and suboptimal urban–rural spatial patterns, posing severe challenges to the modernization process in vast rural areas [4,5,6].

In the broader international literature, rural–urban interactions, secondary towns, and rural transformation are likewise regarded as important channels through which spatial linkages shape poverty reduction, labor reallocation, and territorial development [7,8,9]. From this perspective, county-level urban–rural integration can be understood not only as a China-specific policy agenda but also as part of a wider debate on how institutions and rural–urban linkages jointly influence development outcomes.

Clear property rights have long been regarded as the prerequisite and foundation for promoting the efficient allocation and equal exchange of resource factors [10]. Beyond the Chinese context, the broad literature on land and property rights in developing countries has shown that better-defined and more secure rights can strengthen investment incentives, reduce transaction frictions, and reshape rural development outcomes, although the magnitude of such effects depends on the surrounding institutional setting [11,12,13,14]. For an extended period, one root cause of the constraints on urban–rural factor mobility has been the unclear ownership and incomplete rights over rural resource assets. Therefore, an explicit and sound property rights system is urgently needed to fundamentally dismantle urban–rural market barriers and open channels for bidirectional factor flows. Against this backdrop, the rural collective property rights system reform (RCPRSR), with “clear property rights” as its main thread and “improving farmers’ welfare” as its purpose, establishes a solid micro-institutional foundation for the equal exchange and efficient allocation of urban–rural capital, land, labor, and other factors through a series of institutional designs including asset verification, rights confirmation and empowerment, and shareholding cooperation, thereby manifesting the value of rural resource assets [15,16]. With the continuous development and strengthening of the rural collective economy, the policy effects of the property rights system reform have also attracted widespread scholarly attention. Most studies recognize the significant roles of the reform in increasing farmers’ income, narrowing income gaps within and between urban and rural residents, and boosting county-level economic and entrepreneurial activity [17,18,19,20,21]. Some studies have also empirically confirmed the reform’s promotion of farmers’ democratic participation and village governance effectiveness [22,23,24], as well as its role in improving the provision of village public goods [25]. Meanwhile, some scholars have found that confirming rights to individuals can optimize collective asset allocation and improve the operational efficiency of collective economic organizations [26,27].

Reviewing the literature reveals that research on reform effects has concentrated on three aspects: enriching farmers and increasing income, village governance, and strengthening collective economic development. Moreover, existing studies focus mainly on micro-level cases, qualitative discussions, and theoretical interpretations, while county-level large-sample evidence remains relatively limited. More importantly, the few county-level studies that are closely related to our topic mainly examine county economic development [20] or the county urban–rural income gap [21], whereas evidence on whether the reform promotes county-level urban–rural integration as a multidimensional development outcome remains scarce. In fact, a close connection exists between the RCPRSR and county-level urban–rural integration [28,29]. The clear definition of property rights structures and the standardization of institutional operations can significantly reduce information asymmetry in transactions [30,31]. When rural resource assets possess complete rights that are tradable and mortgageable, counties can more effectively conduct market-oriented operations of collective assets; activate dormant rural collective resources; attract capital, land, and talent; optimize urban–rural spatial layouts; and promote coordinated industrial development [32]. Thus, the rural property rights system reform has become a key institutional transformation that can reshape urban–rural relations and stimulate endogenous rural development. Meanwhile, because the reform is advanced at the county level, scientifically evaluating its actual effects on county-level urban–rural integration has important practical value for accurately grasping policy efficacy and distilling reform experience.

By focusing on county-level China, this study also speaks to international journal debates on rural–urban linkages and structural transformation, which emphasize the importance of intermediate territories and institutional conditions but seldom examine collective property rights reform as a driver of multidimensional urban–rural integration [7,8,9].

Accordingly, this study systematically examines the effect of the RCPRSR on county-level urban–rural integration. Utilizing a balanced panel of 1106 county-level administrative units in China from 2013 to 2020 and employing a multi-period difference-in-differences approach, we make three main contributions. First, unlike existing county-level studies that focus on county economic performance or the urban–rural income gap, we evaluate the reform from the perspective of multidimensional urban–rural integration, thereby extending the outcome dimension of reform assessment. Second, we anchor the empirical analysis at the county level as the key spatial unit linking cities and villages, which helps supplement the existing literature with systematic evidence from a large county-level sample. Third, we explore the policy transmission channels of the reform through population agglomeration and industrial structure optimization and further examine whether the effects vary with economic foundation and baseline integration level.

Based on the theoretical analysis below, we test three hypotheses: H1, the reform pilot significantly raises the level of county-level urban–rural integration; H2, the reform promotes county-level urban–rural integration partly by strengthening the concentration of population and economic activity within county jurisdictions; and H3, the reform promotes county-level urban–rural integration partly by optimizing the industrial structure, especially through relatively faster tertiary industry development and deeper secondary–tertiary integration.

The remainder of this paper is organized as follows. Section 2 develops the theoretical analysis and research hypotheses. Section 3 introduces the data, variables, and empirical strategy. Section 4 presents the baseline results and robustness checks. Section 5 and Section 6 report the heterogeneity and mechanism analyses, respectively. Section 7 concludes and discusses policy implications.

2. Theoretical Analysis and Research Hypotheses

Neo-institutional economics posits that in a market economy system, clearly defined property rights constitute the basic prerequisite for factor market transactions, while factor mobility serves as the core mechanism through which resources move toward more efficient uses [33,34]. This theoretical cornerstone lays the foundation for exploring how the RCPRSR promotes county-level urban–rural integration. In essence, the reform reshapes the incentive structure and transaction environment for factor allocation within county jurisdictions through reconstructed property rights arrangements and the market-oriented operation of collective resources. It breaks the original closure of collective assets, activates rural resources, and promotes the bidirectional circulation of urban–rural factors. Figure 1 summarizes the direct and indirect channels discussed below.

2.1. Direct Mechanisms

The policy implementation and efficacy release of the RCPRSR are highly dependent on the county level—a tier that possesses complete administrative structures and economic systems for both receiving urban radiation and driving rural development, serving as an ideal platform for integrating “reform dividends” with “integration needs.” Specifically, the reform empowers county-level urban–rural integration through three channels:

First, the allocative effect. A complete property rights system can effectively reduce uncertainty and transaction costs in asset allocation, and through full market transactions, it can release asset value, generating premiums and mutual wealth enhancement for both parties [35]. Prior to the reform, many resource and operating assets in rural collectives found it difficult to enter modern economic circulation because property rights were unclear and the entities responsible for ownership and asset management were not clearly defined [36,37]. Therefore, the reform of the rural collective property rights system became imperative. Through clarifying property rights structures and shareholding transformation, the reform stabilizes investor expectations, significantly attracting urban capital, technology, and other advanced factors to sink through the county platform; organically combining with rural land, ecology, labor, and other traditional factors; and giving birth to new industries and new business forms.

Second, the governance effect. The property rights system reform is not merely an economic reform; it also changes the internal governance structure of collective organizations. Once membership, shares, and decision rights are clarified, collective economic organizations can reduce opportunism and coordination costs, improve member participation, and support more stable collective action and resource management, a point consistent with the governance logic emphasized by Ostrom [38] and the incentive problems highlighted by Holmstrom [39]. In this sense, stronger collective governance may improve the capacity to provide member services and local public goods, partially compensating for fiscal constraints in the county-level equalization of basic public services. Meanwhile, the shareholding cooperative reform and the establishment of specialized collective asset management institutions can enhance farmers’ participation and sense of identity in collective economic development and social affairs management [15]. The optimization of public service levels and governance efficacy helps enhance county comprehensive carrying capacity and attractiveness, creating conditions for urban–rural social and spatial integration.

Third, the incentive effect. The reform effectively stimulates farmers’ enthusiasm for participating in market transactions by granting them clear and secured collective asset share rights. On the one hand, equity quantification enables farmers to directly share in collective asset appreciation returns, forming a dual incentive combining “more pay for more work” and “shared development,” which can enhance household organizational levels and improve the “dilemma of collective action.” On the other hand, clarified property rights stabilize farmers’ future income expectations, enhancing their internal motivation for human capital investment and employment entrepreneurship [40,41,42]. This positive incentive cycle—from clearer rights, to more secure returns, to stronger behavioral incentives—not only broadens farmers’ income channels and helps alleviate income inequality arising from differences in individual capabilities and resource endowments, but also stimulates the endogenous momentum for county-level urban–rural integration at the micro level. Accordingly, we propose Hypothesis 1 of this study:

Hypothesis 1 (H1).

After entering the reform pilot, county-level administrative units experience a significant improvement in county-level urban–rural integration relative to non-pilot counties and the pre-reform period.

2.2. Indirect Mechanisms

2.2.1. Population Agglomeration (PopAgg)

By reshaping farmers’ asset rights and identity, the RCPRSR can guide the reconfiguration of population and economic activity within county jurisdictions. Importantly, this mechanism does not imply a simple one-directional change in total county population. Some rural laborers may shift toward county towns or nearby cities while retaining collective rights, whereas return migration and local entrepreneurship may increase county attraction at the same time. The net effect may therefore appear less as a uniform population increase and more as a change in the spatial distribution and composition of people and activities within the county. In this study, such concentration is captured by nighttime lights, which proxy the agglomeration of population and economic activity at the county scale. Moreover, through equity quantification and income right guarantees, farmers are more willing to transfer employment to towns and seek employment opportunities while enjoying stable equity returns, forming a “leaving the land without losing rights” urbanization model [43]. Second, the reform promotes collective economic organizations to revitalize resources and develop characteristic industries, extend agricultural industry chains, and create more local non-agricultural employment positions [44]. Meanwhile, clarified property rights and optimized governance enhance external capital and talent confidence, attracting migrant workers to return and start businesses [45]. Third, the reform enhances public service provision capacity through collective returns, improving county-level education, healthcare, elderly care, and other basic conditions; enhancing attractiveness to rural populations; and promoting rational concentration within counties.

Hypothesis 2 (H2).

The RCPRSR promotes county-level urban–rural integration by strengthening the concentration of population and economic activity within county jurisdictions.

2.2.2. Industrial Structure Optimization (IndStr)

New Structural Economics theory points out that industrial structure upgrading should conform to factor endowment structures and comparative advantages [46,47]. The RCPRSR revitalizes dormant assets such as collective operational construction land, idle homesteads, and ecological cultural resources, enabling them to quickly and effectively connect with market demands and flow toward high-value-added industries after rights confirmation. Notably, at the county scale, the industrial structure evolution induced by the reform does not present simple linear replacement (Petty–Clark theorem) but rather exhibits integrated development characteristics of “tertiary industry leading, secondary industry optimizing, primary industry quality improving.” On the one hand, clarified property rights most directly activate tertiary industries sensitive to property rights such as rural tourism, homestay wellness, and e-commerce [48]. On the other hand, tertiary industry development can pull green food manufacturing, deep processing of agricultural products, and other rural industrial upgrades and in turn promote each county to cultivate advantageous agriculture and animal husbandry toward specialization and high efficiency based on local characteristic resources, enhancing value added and competitiveness [31,49], thereby forming a new county industrial system with deep integration and collaborative advancement of the three industries. This industrial structure optimization can not only improve total factor productivity but also create numerous local employment positions and economic growth points, thereby strengthening urban–rural economic and social integration. Accordingly, we propose Hypothesis 3:

Hypothesis 3 (H3).

The RCPRSR promotes county-level urban–rural integration by optimizing the industrial structure, especially through relatively faster tertiary industry development and a lower secondary-to-tertiary value-added ratio.

3. Materials and Methods

3.1. Data Sources

To ensure the robustness of our research conclusions, we strive to construct a county-level balanced panel dataset with the widest possible coverage in China. However, constrained by data availability at the county level and severe missingness in relevant indicators for many units before 2013 and after 2020, we determine the study period as 2013–2020. After excluding county-level administrative units with severe missingness in key indicators and linearly interpolating a small number of missing values, we obtain a balanced panel of 1106 county-level administrative units. The retained sample is distributed across 24 provincial-level regions—Beijing, Tianjin, Chongqing, Hebei, Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Jiangsu, Zhejiang, Fujian, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Hainan, Sichuan, Yunnan, Shaanxi, Gansu, and Ningxia—thus covering eastern, central, western, and northeastern China. Pilot counties implemented the RCPRSR in 2015, 2017, 2018, 2019, and 2020. A list of pilot counties and their batch-specific launch times are obtained from policy documents published on the official website of the Ministry of Agriculture and Rural Affairs. PM2.5 concentration data are sourced from the Atmospheric Composition Analysis Group at Dalhousie University. Other data are obtained from the China County Statistical Yearbook, EPS Database, and county-level statistical communiqués. All empirical analyses were conducted using Stata 17.0.

3.2. Variable Design

(1) Dependent Variable: The county-level urban–rural integration level serves as the dependent variable of this study. As scholars have not yet reached a consensus on its definition, we synthesize existing research [50]. Considering that urban–rural integration possesses the triple characteristics of process, state, and objective, we construct an indicator system from four dimensions: economic integration, social integration, spatial integration, and ecological integration. This system encompasses Momentum Indicators measuring the progress of urban–rural functional coordination, State Indicators revealing the overall level of regional urban–rural development, and Disparity Indicators reflecting specific urban–rural gaps, thereby comprehensively characterizing the true state of urban–rural integration. We retain grain output per capita in the economic dimension because grain production remains the most consistently reported and comparable indicator of basic agricultural output and food production capacity at the county level, whereas many other agricultural products show much stronger regional heterogeneity and less consistent county-level statistical calibers. The indicator system is reported in

Table 1. Indicator system for county-level urban–rural integrated development.

Target Layer	Dimension Layer	Indicator Layer	Specific Indicator (Unit)	Type	Attribute
County-Level Urban–Rural Integrated Development Level	Economic Integration	Economic Development Level	S1 GDP per capita (CNY/person)	State	+
		Wealth Status	S3 Balance of urban–rural residents’ savings deposits (CNY)	State	+
		Grain Output Per Capita	S4 Total grain output/Regional population (tons/10,000 persons)	State	+
	Social Integration	Urban–Rural Income Gap	S5 Urban–rural income ratio (Urban per capita disposable income/Rural per capita disposable income) (%)	State	−
		Healthcare	X1 Hospital and health center beds per 10,000 persons (beds/10,000 persons)	Disparity	+
		Basic Education Security	X2 Primary and secondary school students/Regional population (%)	Momentum	+
	Spatial Integration	Science and Technology Innovation	X3 Patent grants per 10,000 persons (pieces/10,000 persons)	State	+
		Public Service Funding Support	S2 Per capita local fiscal budget revenue (CNY/person)	State	+
		Urban–Rural Market Circulation	Y1 Industrial enterprises above designated size per 10,000 persons (units/10,000 persons)	State	+
	Ecological Integration	Spatial Agglomeration Degree	Y2 Number of townships/County administrative area (units/square kilometer)	State	+
		Urban–Rural Air Quality	Z1 PM2.5 concentration (μg/m3)	Momentum	−
		Urban–Rural Vegetation Coverage	Z2 Mean NDVI of urban–rural areas	Momentum	+

.

(2) Explanatory Variable: The core explanatory variable of this study is the RCPRSR pilot. If a county was included in the reform pilot in a given year, it takes a value of 1; otherwise, it takes a value of 0.

(3) Control Variables: Referencing existing relevant research, we select the following variables to control for other potential factors affecting county-level urban–rural integration [50]: rural and urban residents’ income levels, population density, financial development level, residents’ consumption capacity, telecommunication infrastructure level, and government expenditure scale.

(4) Mediating Variables: Based on the theoretical analysis in Section 2 and relevant research, following [51], we employ DMSP/OLS nighttime light data—that is, regional average nighttime light intensity—to measure population mobility and agglomeration. Industrial structure optimization is measured using the ratio of the value added of the secondary industry to the value added of the tertiary industry [52].

The meanings and descriptive statistics of the main variables are shown in

Table 2. Definitions and descriptive statistics of main variables.

Variable Name	Variable Definition or Assignment	Observations	Mean	Std. Dev.
County-Level Urban–Rural Integration Level	Calculated by entropy method	8848	0.287	0.05
Rural Residents’ Income Level	Per capita disposable income of rural residents	8848	3392.433	2076.099
Urban Residents’ Income Level	Per capita disposable income of urban residents	8848	28,131.168	8238.932
Population Density	Total population/administrative land area	8848	0.037	0.032
Financial Development Level	Balance of deposits and loans at financial institutions/GDP	8848	0.713	0.423
Residents’ Consumption Capacity	Per capita retail sales of consumer goods	8848	14,769.102	10,434.461
Telecommunication Infrastructure Level	Fixed telephone subscribers/year-end total population	8848	981.285	776.774
Government Expenditure Scale	Fiscal expenditure/GDP	8848	0.256	0.200
Population Agglomeration	Regional average nighttime light intensity (DMSP/OLS)	8848	6.375	8.530
Industrial Structure Optimization	Ratio of secondary to tertiary industry value added	8848	1.239	0.845

Note: Population density, financial development level, residents’ consumption capacity, telecommunication infrastructure level, and government expenditure scale are reported as original values; logarithmic transformations are applied in subsequent regression analyses.

.

3.3. Model Construction

Given that the RCPRSR was rolled out in batches across regions, to examine the impact of the reform pilot on county-level urban–rural integration, this study treats it as a quasi-natural experiment, classifying pilot counties as the treatment group and remaining counties as the control group. We employ a difference-in-differences (DID) model for causal identification [53]. The specific model form is as follows:

$$Y_{it}=\alpha +\theta di{d}_{it}+{\beta }_{1}Control{s}_{it}+{\delta }_i+{\gamma }_t+{\epsilon }_{it}$$

(1)

In Equation (1), the dependent variable denotes the urban–rural integration level of county i in year t. The core explanatory variable indicates whether county i has implemented the reform in year t. The remaining terms denote the vector of control variables, county fixed effects, year fixed effects, and the error term.

To further identify the pathways through which the RCPRSR affects county-level urban–rural integration, we follow the classical procedure for mediation effects [54] and establish the following models for verification:

$$M_{it}={\rho }_0+{\rho }_{1}di{d}_{it}+{\beta }_{2}Control{s}_{it}+{\delta }_i+{\gamma }_t+{\epsilon }_{it}$$

(2)

$$Y_{it}={\sigma }_0+{\sigma }_{1}di{d}_{it}+{\sigma }_2{M}_{it}+{\beta }_{3}Control{s}_{it}+{\delta }_i+{\gamma }_t+{\epsilon }_{it}$$

(3)

In Equations (2) and (3), the mediating variable M_it encompasses population agglomeration and industrial structure, while other variable definitions remain consistent with Equation (1).

4. Empirical Results and Analysis

4.1. Baseline Regression Results

Table 3. Baseline Regression Results.

Variable	County-Level Urban–Rural Integrated Development Level
	(1)	(2)
RCPRSR Pilot	0.002 ***	0.002 ***
	(0.001)	(0.001)
Rural Residents’ Income Level		0.000 ***
		(0.000)
Urban Residents’ Income Level		0.000 ***
		(0.000)
Population Density		0.007
		(0.006)
Financial Development Level		0.000
		(0.000)
Residents’ Consumption Capacity		0.000
		(0.000)
Telecommunication Infrastructure Level		−0.000
		(0.000)
Government Expenditure Scale		−0.000
		(0.000)
County Fixed Effects	Yes	Yes
Year Fixed Effects	Yes	Yes
Constant	0.286 ***	0.251 ***
	(0.000)	(0.005)
Observations	8848	8848
R2	0.936	0.939

Notes: (1) *** denotes statistical significance at the 1% level; (2) standard errors clustered at the county level are reported in parentheses.

reports the baseline regression estimation results of the impact of the RCPRSR on county-level urban–rural integrated development. Column (1) presents the results controlling only for county and year fixed effects without additional control variables, while Column (2) incorporates the full set of control variables. The comparison reveals that regardless of whether control variables are included, the reform pilot variable is significantly positive at the 1% level, thereby validating Hypothesis H1—that is, the reform exerts a positive impact on county-level urban–rural integrated development.

Regarding the control variables, both rural and urban residents’ per capita income levels are significantly positive at the 1% level, indicating that higher income levels among rural and urban residents promote county-level urban–rural integrated development.

In economic terms, the coefficient in Column (2) implies an increase of about 0.002 points in the county-level urban–rural integration index. Relative to the sample mean of 0.287, this is about 0.70% of the average level. Although the annual marginal effect is modest, it is meaningful for an institutional reform whose impacts accumulate over a multi-year pilot cycle and are expected to materialize gradually through factor reallocation, population concentration, and industrial restructuring.

4.2. Precondition Tests

The validity of our DID identification strategy relies on two critical assumptions: the randomness of policy assignment and the parallel trends condition.

4.2.1. Policy Randomness Test

In investigating the net effect of the reform policy, ensuring the randomness between the treatment and control groups is crucial, as this can effectively exclude the interference of non-policy factors—that is, whether the RCPRSR selects counties for policy implementation in a random manner. From a policy perspective, the reform implementation is the result of collaborative promotion by the central and local governments, and it can be regarded as an external shock variable with exogenous characteristics. To statistically verify the randomness of pilot county selection from an econometric perspective, we construct a binary Logit model for testing [55]: using the sample before reform implementation, with the county-level urban–rural integration level as the core explanatory variable and “whether it is a pilot county” as the dependent variable to examine whether the county-level urban–rural integration level affects the selection of policy reform pilot counties. The results are reported in

Table 4. Logit Results for Pilot Selection Randomness.

Variable	Pilot County Selection (Logit)
Baseline Urban–Rural Integration Index (2013)	0.570
	(3.220)
Control Variables	yes
County Fixed Effects	yes
Year Fixed Effects	yes
Constant	7.480 ***
	(1.796)
Observations	2212
Sample Period	2013–2014

Notes: (1) *** denotes statistical significance at the 1% level; (2) standard errors clustered at the county level are reported in parentheses.

.

The results show that the county-level urban–rural integration level is uncorrelated with the selection of reform pilot counties, further demonstrating that the selection of reform counties basically satisfies the random assignment assumption.

4.2.2. Parallel Trends and Dynamic Effects Test

The identical trends in the outcome variable between the treatment and control groups prior to policy implementation constitute the prerequisite for unbiased DID estimation results. To verify the parallel trends assumption and identify the dynamic effects of the RCPRSR on urban–rural integration levels, we employ an event study approach. Specifically, given that the majority of pilot batches were concentrated after 2017, with limited sample size beyond three years post-implementation—which may lead to insufficient estimation degrees of freedom after including control variables—we follow [53,56] and appropriately aggregate the periods before and after the concentrated pilots: data three years after pilot initiation are collapsed into period +3, and data four years before initiation are collapsed into period −4. We construct the following model (4) for testing:

$$Y_{it}={\omega }_0+{\displaystyle \sum _{s=-4}^3{\omega }_s}di{d}_s+{\beta }_{4}Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(4)

In Equation (4), the estimated coefficients capture the impact of the reform pilot on urban–rural integration. The event time indicator denotes periods relative to the pilot implementation year, taking values in {−4, −3, −2, 0, +1, +2, +3}; the period immediately before implementation is omitted as the reference group. The remaining terms are defined as in Equation (1).

Figure 2 presents the parallel trends test results for county-level urban–rural integration based on 95% confidence intervals, with the period immediately before reform implementation (s = −1) set as the base period. The results show that all time windows prior to reform implementation exhibit no significant impact on county-level urban–rural integration—that is, no systematic significant difference exists in urban–rural integration levels between the experimental and control groups prior to reform, indicating that the parallel trends assumption is satisfied. After reform implementation, the coefficients in each time window gradually rise and become significantly positive, peaking at period +3, indicating that the RCPRSR significantly enhances county-level urban–rural integration levels and that policy effects exhibit significant time lags. This pattern is highly consistent with the 2-to-3-year reform cycle stipulated by the pilot policy.

On the one hand, during the initial reform period, the transmission from central policy dissemination to local implementation scheme formulation, and then to comprehensive grassroots rollout, inherently requires a process; moreover, foundational tasks such as asset verification, membership identification, and equity quantification cannot be accomplished overnight, and the operational mechanisms of collective economic organizations require gradual improvement. On the other hand, farmers need time to comprehend and adapt to the new policy, with limited participation enthusiasm in the short term, compounded by mutual adaptation processes for business models and collaborative mechanisms under new organizational forms. When phased reform objectives are completed, the deeper effects of removing barriers to urban–rural factor flows and enhancing resource allocation efficiency begin to manifest, forming “cumulative circular” development momentum through the sustained accumulation of positive factors, steadily promoting county-level urban–rural integration.

4.3. Robustness Checks

To ensure that our baseline estimates capture the causal effect of the RCPRSR rather than spurious correlations or model dependence, we conduct a battery of robustness tests.

4.3.1. Placebo Test

To exclude the influence of potential random interference and omitted variables on the core estimation results, we employ a placebo test to enhance the reliability of our research conclusions. Specifically, we randomly select “pseudo-treatment groups” and virtual policy implementation time points from all sample counties, then re-estimate the DID model to obtain spurious estimated coefficients of the core explanatory variable. Repeating this process 500 times, we visualize the estimated coefficients. As shown in Figure 3, the regression coefficients from random assignment follow a unimodal distribution centered at zero, with the vast majority of p-values exceeding 0.1, significantly distinct from the baseline regression estimate. This finding indicates that the RCPRSR indeed promotes the convergence of urban–rural integration gaps, verifying the robustness of our research conclusions.

4.3.2. Alternative Timing Specification

We advance the reform implementation time by one year, assuming that the treatment counties implemented the reform in 2014 instead of the actual year. Based on this, we reconstruct the interaction term as the core explanatory variable and employ a counterfactual test to corroborate the policy effectiveness of the RCPRSR. The validity of the multi-period DID method relies on the absence of systematic and significant differences in county-level urban–rural integration levels between the treatment and control groups prior to reform implementation. Accordingly, if the interaction term with the advanced reform timing exhibits a significantly positive coefficient, this would suggest that the baseline results may be contaminated by potential confounding factors or random interference. Conversely, if the reform variable is insignificant, this indicates that the improvement in county-level urban–rural integration levels stems from the practice of the property rights system reform. Column (1) of

Table 5. Additional robustness tests.

Variable	(1)	(2)	(3)
	Alternative Timing	Lagged Dependent Variable	Exclusion of Leading Regions
Lagged URI		0.565 *** (0.021)
RCPRSR Pilot	0.001 (0.001)	0.001 *** (0.001)	0.002 *** (0.001)
Control Variables	Yes	Yes	Yes
County Fixed Effects	Yes	Yes	Yes
Year Fixed Effects	Yes	Yes	Yes
Constant	0.233 ***	0.124 ***	0.251 ***
	(0.004)	(0.006)	(0.005)
Observations	8488	7742	8104
R2	0.937	0.961	0.925

Notes: (1) *** denotes statistical significance at the 1% level; (2) standard errors clustered at the county level are reported in parentheses.

shows that the regression coefficient with the advanced reform timing fails to pass the significance test, consistent with the expectations of the counterfactual test, confirming that the core conclusion regarding the promotion of county-level urban–rural integration by the property rights system reform possesses robustness.

4.3.3. Lagged Dependent Variable

To alleviate potential endogeneity concerns and account for the dynamic continuity of the socioeconomic environment, we introduce a one-period lag of the dependent variable into the baseline regression model to capture the influence of unobservable time-varying factors on the baseline estimation results. Column (2) of Table 5 shows that the pilot interaction term remains significantly positive at the 5% level even after including the lagged dependent variable, verifying the robustness of our baseline regression results.

4.3.4. Exclusion of Leading Reform Regions

Considering the pilot nature of the reform process in certain regions, we conduct a special sample exclusion test. Although the reform pilot was officially launched in 2015, Beijing and Zhejiang had essentially completed village-level reform tasks prior to this (coverage rates: exceeding 95% in Beijing and reaching 75% in Zhejiang by the end of 2014), far surpassing the less than 10% level in other provinces. That is, the actual reform progress in Beijing and Zhejiang differed significantly, which may interfere with the estimation results. Additionally, municipalities rely on administrative and economic resources (such as Beijing’s collective construction land marketization pilot and Chongqing’s land ticket system) to form unconventional reform paths, whose experiences may not be generalizable to ordinary provinces. To exclude these potential estimation biases, we remove samples from Zhejiang and the municipalities (Beijing, Tianjin, and Chongqing) and re-conduct regression analysis. Column (3) of Table 5 shows that the RCPRSR remains significantly positive at the 1% level, indicating that our research conclusions are not disturbed by the advanced reforms in Beijing and Zhejiang or the particularities of municipalities, and they thus possess universal robustness.

5. Heterogeneity Analysis

5.1. Economic Foundation

Focusing on whether the impact of the RCPRSR on county-level urban–rural integration exhibits heterogeneity across economic foundations, we divide national counties (cities, districts) into poverty-stricken and non-poverty-stricken groups based on the list of national poverty counties published by the National Rural Revitalization Administration in 2014 and conduct grouped regressions.

Column (1) of

Table 6. Heterogeneity analysis results.

Variable	(1)	(2)	(3)	(4)	(5)
	Poverty-Stricken Counties	Non-Poverty-Stricken Counties	Low Integration	Medium Integration	High Integration
RCPRSR Pilot	0.003 **	0.001	0.005 ***	0.00165	0.000126
	(0.001)	(0.001)	(0.001)	(0.001)	(0.002)
Control Variables	Yes	Yes	Yes	Yes	Yes
County Fixed Effects	Yes	Yes	Yes	Yes	Yes
Year Fixed Effects	Yes	Yes	Yes	Yes	Yes
Constant	0.250 ***	0.258 ***	0.223 ***	0.251 ***	0.270 ***
	(0.001)	(0.007)	(0.008)	(0.010)	(0.012)
Observations	2768	6080	2952	2952	2944
R2	0.856	0.941	0.854	0.793	0.923

Notes: (1) *** and ** denote statistical significance at the 1% and 5% levels, respectively; (2) standard errors clustered at the county level are reported in parentheses.

shows that the regression coefficient of the RCPRSR pilot on urban–rural integration is significantly positive at the 5% level, indicating that the reform significantly accelerates the urban–rural integration process in poverty-stricken counties. This may be because poverty-stricken counties generally face constraints such as poor factor mobility, unclear property rights definition, and weak collective economies, while the RCPRSR effectively activates idle resources and collective asset income distribution mechanisms through measures such as rights confirmation and certification, shareholding cooperation, and trading platform construction, thereby significantly improving urban–rural factor allocation efficiency. In contrast, non-poverty-stricken counties have either completed preliminary property rights reforms earlier or possess more complete market systems, leading to diminishing marginal returns from the reform [57], resulting in an insignificant coefficient.

5.2. Initial Urban–Rural Integration Levels

To examine whether reform effectiveness varies with initial conditions, we take the year immediately prior to the pilot rollout of the RCPRSR (2014) as the base period and classify counties into three categories—high, medium, and low—according to their initial urban–rural integration levels. Columns (3) to (5) of Table 6 show that the effect of the reform in promoting the convergence of county-level urban–rural integration gaps is significant in counties with low urban–rural integration levels, while it is insignificant in counties with medium and high levels. That is, policy dividends are concentrated in counties with low initial urban–rural integration levels.

The underlying mechanism likely stems from differences in regional development foundations: in low-integration areas where the urban–rural dual structure is pronounced, problems such as dormant collective assets, blocked factor flows, and large urban–rural income gaps are particularly severe. The property rights system reform generates institutional breaking effects through the “three transformations” mechanism (resource-to-asset, asset-to-equity, and equity-to-income), dismantling the urban–rural dual system. In contrast, high-integration counties already possess high development capacity and urban–rural integration levels, with relatively mitigated inherent urban–rural dual structure contradictions and more mature urban–rural industrial linkage pathways and bidirectional population flow mechanisms. Naturally, the marginal returns and “institutional breaking” effects that the reform can generate are limited.

6. Mechanism Analysis

Based on the theoretical discussion in Section 2.2, we select population agglomeration and industrial structure adjustment as mediating variables and construct a mediation effects model to progressively estimate Equations (2) and (3), thereby exploring the mechanisms through which the RCPRSR promotes county-level urban–rural integration.

6.1. Population Agglomeration Mechanism

Column (2) of

Table 7. Mediation Results for Population Agglomeration and Industrial Structure Optimization.

Variable	(1)	(2)	(3)	(4)	(5)
	Baseline Regression	Population Agglomeration	URI with PopAgg	Industrial Structure Optimization	URI with IndStr
RCPRSR Pilot	0.002 *** (0.001)	0.786 *** (0.111)	0.001 (0.001)	−0.045 ** (0.019)	0.002 *** (0.001)
PopAgg (Night Light)			0.002 ***
			(0.000)
IndStr (Second/Tertiary)					0.006 *** (0.001)
Constant	0.251 ***	−0.123	0.251 ***	0.835 ***	0.244 ***
	(0.005)	(0.878)	(0.005)	(0.154)	(0.005)
Observations	8848	8848	8848	8848	8848

Notes: (1) *** and ** denote statistical significance at the 1% and 5% levels, respectively; (2) standard errors clustered at the county level are reported in parentheses.

shows that the RCPRSR pilot has a significant positive effect on population agglomeration. However, when population agglomeration is added on the basis of Equation (2), the effect of the reform on county-level urban–rural integration becomes insignificant, while population agglomeration remains significantly positive (see Column (3)). This indicates that, in the pathway through which the reform affects county-level urban–rural integration, population agglomeration possesses full mediating effects—the reform significantly promotes county-level urban–rural integration through population agglomeration within counties.

Evidently, clarified property rights not only facilitate the free transfer of labor toward high-productivity sectors within counties but also create substantial local non-agricultural employment generated by the development of collective economies and new business forms, thereby enhancing county population carrying capacity and economic density [40]. The interaction of this “siphon” and “diffusion” effect ultimately manifests as intensified nighttime lights, reflecting the tendency toward concentration, networking, and scaling of population and economic activity within county jurisdictions. Hypothesis H2 is thus verified.

6.2. Industrial Structure Optimization Mechanism

Column (4) of Table 7 shows that the coefficient of the reform on industrial structure optimization is significantly negative at the 5% level. Since this variable is measured as the ratio of value added in the secondary industry to value added in the tertiary industry, a decrease in the ratio indicates a relative increase in tertiary-industry value added compared with the secondary industry. In other words, the reform is associated with relatively faster tertiary-industry development, which is consistent with industrial structure optimization. Column (5) shows that after introducing industrial structure adjustment, the coefficient of the reform on county-level urban–rural integration remains significantly positive at the 1% level, indicating that the reform does not suppress the secondary industry but rather more vigorously stimulates the potential of the tertiary industry with “light assets and fast returns,” whose growth rate often surpasses the traditional secondary industry requiring substantial fixed asset investment in the short term. Through demand pull and service empowerment, this in turn promotes quality upgrading and efficiency improvement in the secondary industry, thereby effectively driving industrial structure adjustment and optimization and ultimately serving as an important channel for promoting county-level urban–rural integration [58,59]. This occurs because the tertiary industry responds more rapidly and sensitively to property rights reform, causing county industries to exhibit not “leapfrog” upgrading but rather a pattern of “secondary industry quality improvement and efficiency enhancement, tertiary industry accelerated rise.” In summary, Hypothesis H3 is verified.

7. Conclusions and Policy Implications

Based on panel data from 1106 county-level administrative units in China from 2013 to 2020, this study treats the RCPRSR as a quasi-natural experiment and systematically examines how the reform affects county-level urban–rural integration. The findings show that the RCPRSR significantly promotes county-level urban–rural integration, and this conclusion remains robust after a series of stringent tests. The effect is more pronounced in poverty-stricken counties and in counties with lower baseline levels of urban–rural integration. Mechanism analysis further shows that population agglomeration and industrial structure optimization are important transmission channels through which the reform improves county-level urban–rural integration.

These results are consistent with the logic of new institutional economics: clearer and more enforceable property rights arrangements can reduce transaction costs, improve collective governance, and reallocate factors toward more integrated county development. At the same time, the economic magnitude of the baseline estimate should be understood as cumulative rather than instantaneous, because institutional reforms usually operate through gradual adjustments in resource mobilization, spatial concentration, and industrial upgrading. The Chinese setting also has distinctive institutional features—especially collective ownership, county-based governance, and the policy framework of rural revitalization—so the broader applicability of the results should be interpreted with appropriate caution. In addition, because the county-level balanced panel relies on the continuous availability of multiple indicators, the evidence reported here is necessarily limited to 2013–2020.

This study also adds to international discussions on rural–urban linkages and land governance by showing that institutional reform can affect territorial integration through local agglomeration and industrial restructuring, rather than only through income or productivity effects [7,8,9].

(1)

Consolidating Reform Achievements, Deepening Rights Empowerment, and Facilitating Urban–Rural Factor Circulation.

Within the current policy framework of rural revitalization and county-based urban–rural integration, reform should continue from the stage of clarifying rights to the stage of activating rights. On the basis of completing asset verification and shareholding reform, policy should focus on improving rural property rights trading markets and prudently exploring more forms of rights realization, such as compensated exit, mortgage, guarantee, and inheritance of collective asset shares. Improving valuation, mortgage, and transaction mechanisms can help reduce institutional costs in urban–rural factor transactions and promote the more efficient matching of rural resources with urban capital, technology, and talent.

(2)

Adhering to Local Conditions, Implementing Differentiated Regional Policies, and Precisely Addressing Integration Challenges.

Policy design should fully consider regional heterogeneity and adopt differentiated guidance. For poverty-stricken counties, counties with weak urban–rural integration foundations, and many counties in central and western China, policy priority should be placed on strengthening reform implementation, improving infrastructure and public services, and reinforcing complementary fiscal support so that the reform’s institutional effects can be more fully released. For counties with stronger economic foundations and higher integration levels, the policy focus can shift toward deeper reforms, such as diversified market entry for collective construction land, rural inclusive finance, and digital governance of collective assets.

(3)

Strengthening the Coordination of Industrial Quality Improvement and Population Agglomeration to Enhance County Development Vitality.

County development policy should strengthen the coordination between industrial upgrading and population agglomeration. On the one hand, local governments should support tertiary sectors that can rapidly create employment on the basis of local resource endowments while also encouraging their linkage with processing, manufacturing, and other secondary industry activities. On the other hand, they should continue to improve education, healthcare, and elderly care services in county seats and key towns so that more rural residents can achieve non-agricultural employment and stable settlement within county jurisdictions. In this way, reform can better translate institutional change into sustained county-level urban–rural integration.