Do Regional Integration Policies Promote Integrated Urban–Rural Development? Evidence from the Yangtze River Delta Region, China

Abstract

Regional integration policies play a crucial role in promoting coordinated regional development. However, it remains unclear whether the polices simultaneously take into account urban–rural integration to achieve a dynamic balance between efficiency and equity. Based on socioeconomic data from 250 cities in China between 2003 and 2019, we used a staggered difference-in-difference method to investigate the impact of the strategy for the integrated development of the Yangtze River Delta (YD integrated development) on integrated urban–rural development. Our results indicate that the YD integrated development effectively promotes integrated urban–rural development and this conclusion holds after conducting various robustness tests and heterogeneity analyses. Additionally, the YD integrated development can facilitate integrated urban–rural development through the following three main pathways: promoting economic growth, improving road transport links, and advancing technological progress. This paper offers new insights for advancing integrated urban–rural development. The next step could involve the further exploration of the connections between external regional integration policies and internal rural reforms, which will contribute to expediting the establishment of an integrated urban–rural pattern.

1. Introduction

In the development process of China for more than seven decades, the scientific management of the basic economic relationship between urban and rural areas has been a major strategic issue [1]. In seeking to address this, the direction of China’s urban–rural relations has also undergone a dynamic evolution, from “urban–rural duality” to “integrated urban–rural development”, to “urban–rural integrated development”, and to “urban–rural coordinated development” [2]. However, the approach of taking the urban area as the centre of attention and economic growth as the indicator of positive development has not fundamentally changed in practice, and the significant loss of production factors in rural areas has led to the gradual decline of rural production due to marginalisation. The imbalance between urban and rural development and the inadequate development of rural areas remain key structural incongruities in social governance, and the issue of fairness and justice between urban and rural areas has been continuously re-emerging through ongoing and increasing regional imbalance and spatial differentiation [3]. According to data from the National Bureau of Statistics of China, in 2023, the per capita consumption expenditure ratio between urban and rural residents was 1.47:1, the ratio of urban to rural employed persons was 1.58:1, and the per capita disposable income ratio was 2.39:1. Although the relative income gap between urban and rural areas in China has experienced a long downward trend (from 2009 to the present), the gap between urban and rural areas has not narrowed since the reform and opening up were initiated when measured in terms of the absolute difference in per capita income between urban and rural residents. Based on observational data from 2013 to 2023, this gap had increased from CNY 17,037 to CNY 30,130, with the absolute income disadvantage of rural residents relative to urban residents steadily increasing, which means that imbalanced and inadequate development between urban and rural areas in China remains a prominent feature. Therefore, in terms of the next stage of urban–rural integration and coordinated development, the question of how to address the current division between urban and rural areas to facilitate coherent integrated development involving a two-way flow of factor resources and rational optimal allocation remains challenging.

China’s highly coercive hukou (household registration) system has created an “invisible high wall” between urban and rural areas and is considered a key factor encouraging the current urban–rural separation structure [4,5,6]. However, at the same time, in the context of the slow progress of the reform of the household registration system, it needs to be noted that border barriers between regions and decentralisation have led local officials to pay more attention to the economic development and social well-being of their administrative regions. The “political tournament” competition between local officials for economic growth in their jurisdictions has exacerbated the phenomenon of segmentation between areas [7], which is not conducive to tapping the deep potential of a unified market. This phenomenon is widespread at all levels, that is, between countries, provinces, cities, urban and rural areas, and even between villages [8,9], and directly affects whether urban–rural integration and regional coordinated development can be achieved. In practice, to break down local protection and regional barriers, the integration and cooperation of different regions have become recognised as key to economic development and technological innovation.

Extensive theorising on regional integration [10,11], development and evolution [12,13], indicator measurement [14,15], and influencing factors [16,17] has been conducted, and in-depth mechanism analyses and empirical evaluations have been carried out on various socioeconomic effects derived from the integration process [18,19]. In practice, regional integration can be divided into international and regional integration within China. The former, such as that involving the European Union (EU), the Belt and Road (B&R), and the Asia Pacific Economic Cooperation Forum (APEC), works mainly through free trade areas, customs unions, common markets, economic alliances, and other measures that are intended to weaken or eliminate trade barriers between countries and promote improvements in production efficiency and capital accumulation in different countries or regions within their respective areas along with the free flow of production factors [20,21]. Some studies have found that, with the enhancement of exchanges and linkages between member states in the EU, existing market advantages of current member countries and factor cost advantages for new member countries are continuously improved with integration, such that there is a significant spatial pattern of “club convergence” within the EU [22,23]. Domestic integration in China mainly includes national-level regional development strategies such as the Beijing–Tianjin–Hebei coordinated development area, the Yangtze River Economic Belt, and the Guangdong–Hong Kong–Macao Greater Bay Area, which are profoundly affecting China’s economic development and reshaping the overall economic pattern of China and the world [24,25]. Integration policy is intended to promote the transformation of regional spatial patterns from a single-centre to a multi-centre focus, with the help of the joint effect of centripetal and centrifugal forces, and finally to achieve multi-directional selection and the optimal allocation of various production factors [26]. Specific economic and interest ties between different cities have also promoted the transformation of cities, in terms of a specialised division of labour, diversified development, and integrated collaboration, to achieve the intensive release of policy dividends, such as the continuous expansion of economic scale, the rationalisation of industrial structures, and the adaptation of multi-level employment needs [27].

It should be noted that various studies involving China and elsewhere have pointed out that the relationship between regional integration and socioeconomic effects is not static and that there may be differences or even negative effects on different stages, city types, and cooperation models [28,29,30]. In addition, various studies have examined the impact of urbanisation on the income disparity between urban and rural areas to understand the connection between regional development and urban–rural integration [31,32,33]. However, urbanisation does not necessarily mean regional integration. The urban–rural income gap is just one aspect of urban–rural integration. To fully understand this, it is important to analyse various factors.

Based on the foregoing, this study examines the case of the strategy for the integrated development of the Yangtze River Delta (YD integrated development) to explore both the role of regional integration in the integrated urban–rural development and the transmission channels. This study’s contributions are as follows. First, a combination of theoretical analysis and empirical testing is used to investigate whether China’s widely implemented regional integration policies can effectively promote fair and efficient development within cities while also fostering economic growth. This research extends beyond existing studies that have mainly focused on assessing the impact of integration policies at the overall or inter-city level. Second, in contrast to research that has primarily investigated the impact of rural land system reform on reducing the income disparity between urban and rural areas and achieving greater integration between them [34,35,36], this study begins with the external factor of regional integration. This approach expands the range of potential policies for promoting rural revitalisation and achieving common prosperity. Third, the channels through which YD integrated development affects integrated urban–rural development are explored to provide a more accurate and rigorous empirical basis for the continued promotion of integrated urban–rural development.

2. Theoretical Background and Hypotheses Development

The market-based allocation of factors of production such as capital, technology, and labour has long been hindered by the presence of trade barriers and the segmentation of regional markets, which have subsequently exacerbated the imbalance in regional development [37,38]. This is especially evident in the disparities between developed and underdeveloped regions, as well as between urban and rural areas. Two-way factor flows and equal exchanges have become a luxury because of various factors, including geography, economic fundamentals, the social environment, transport facilities, and institutional barriers. Over time, the gap in comparative advantages between regions has continued to widen, distorting the price relationships between factors and creating a vicious cycle of mismatched resources [39,40].

Regional integration policies (e.g., YD integrated development), which cover a wide range of areas, including industrial development, scientific and technological innovation, infrastructure, environmental protection, and public services, aim to remove local protection and border divisions through administrative intervention [41]. The objective is to open the channels for the free flow of factors at all levels of the regional scale through project cooperation, scientific and technological support, labour mobility, and funding [42]. Hence, a rational and effective regional integration policy should promote collaboration among cities and between urban and rural areas. It should effectively combine and coordinate the resources of different regions to maximise benefits and common interests. We can use borrowed size theory to explain the effects of regional integration policies from both performance and functional perspectives [43]. On the one hand, the high land rents, traffic congestion, and environmental pollution in large cities have led some enterprises and their staff to relocate to neighbouring small cities or towns (i.e., ‘borrowed function’). On the other hand, the positive externalities generated by this behaviour can provide more employment opportunities and increase labour productivity for the residents of small- and medium-sized cities and rural areas (i.e., ‘borrowed performance’). Therefore, we propose the following hypothesis:

Hypothesis 1:

The YD integrated development promotes the integrated urban–rural development.

As an administrative intervention oriented towards urban–rural relations, YD integrated development must be tested in practice to determine its development orientation and specific programs following its introduction. During the initial implementation, local government officials typically adopt a wait-and-see approach and maintain the original system, policy, and authority to prevent the broad spillover of local factor resources after the administrative boundary is removed [44]. Therefore, integrating urban and rural areas requires the establishment of supporting facilities, improved staffing, and the convergence of authority to approve various matters. These changes in the administrative system ultimately led to integrated urban–rural development at the spatial, economic, social, and environmental levels, but this process inevitably takes time. When the many constraints hindering urban development at the institutional level are removed, multiple explicit and implicit gains are revealed to show their more comprehensive strategic value and long-term practice [45]. Therefore, this study proposes the following:

Hypothesis 2:

The YD integrated development has a short-term lag and long-term positive impact on integrated urban–rural development.

While regional integration can facilitate integrated urban–rural development, inner-cities are complex systems comprising multiple dimensions, including economic, social, and environmental integration. Further exploration is thus needed to understand its internal mechanisms, including the following three channels of integration: basis, support, and dynamics (Figure 1).

(1) Integration basis: economic growth. Theoretically, stable high-quality economic growth is a prerequisite for promoting integrated urban–rural development, without which the pursuit of integrated development would be meaningless. Regional integration can enable the smooth flow and optimal allocation of production factors, forming integrated industrial, supply, and capital chains and other market resource allocation mechanisms. This can stimulate the development of strong social productive forces, reflecting the efficiency of the policy [46,47]. However, its effect does not simply mean that cities become ‘dominant’, but rather that the improvement in institutional mechanisms, infrastructure, and public services helps form a mutually open, interconnected, and interdependent organic whole involving both urban and rural areas. Based on the law of diminishing marginal returns, when the economy is growing stably, the disparity in per capita capital stock between urban and rural areas will lead to the continuous adjustment and optimisation of the industrial, supply and demand, and urban and rural structures. This will promote the transmission of social capital from developed to less developed regions, ultimately leading to cooperative sharing between urban and rural areas [48].

(2) Integration support: road transport links. In urban development, improving the transport infrastructure can overcome geographical constraints and spatial barriers, improve regional connectivity and accessibility, and facilitate the exchange and cooperation of regional economic activities. Denser transport networks have created new opportunities for developing previously marginal areas. This has prompted the greater circulation of the physical elements of the four major links of ‘social reproduction’—production, distribution, exchange, and consumption—between urban and rural areas. Consequently, domestic demand potential has been released and the income channels of urban and rural residents have broadened. The ‘siphoning’ or ‘trickle-down’ characteristic of regional economic activity lies behind the improved transport infrastructure [49]. Krugman’s core/periphery model notes that low trade costs can concentrate production factors from the periphery to the core in the early stages of the construction process, triggering the ‘siphoning effect’ of unbalanced regional development [50]. Since transport resources span the entire region and are shared by urban and rural areas, the gradual dissolution of geographical boundaries lowers the costs associated with the exchange of information, technology, capital, and other factors. Regional integration policies can accelerate this process. Simultaneously, the reciprocal exchange of urban and rural resources and the rational distribution of public resources have not only reduced traffic congestion, pollution, and resource shortages in urban areas, but have also met the educational, public service, and employment needs of rural residents. This has helped revitalise rural areas and achieve the ambitious goal of common prosperity through the trickle-down effect [51].

(3) Integration dynamics: technological progress. Technological progress and industrial transformation can drive integrated urban–rural development. Regional integration and development have weakened border barriers and reduced market segmentation. This has resulted in significant economies of scale and network effects, facilitating the diffusion and transfer of knowledge, skills, and innovation between cities and making technological progress an important driver of factor mobility and the synergy between urban and rural areas. Information technology, represented by big data and the digital economy, has overcome spatial constraints, providing convenient knowledge transfer and learning opportunities, and thereby promoting the cross-spatial flow and application of information technology such as e-commerce, live streaming, and online learning [52,53]. Second, scientific and technological progress can enhance the competitiveness and innovativeness of agriculture by pooling factors, diffusing skills and innovative mechanisms, and improving the balance, coordination, and ecologisation of the urban–rural industrial structure, thus achieving the integrated urban–rural development [54]. Based on this, the following hypothesis is proposed:

Hypothesis 3:

The YD integrated development promotes integrated urban–rural development by fostering economic growth, strengthening road transport links, and promoting technological progress.

3. Research Methods and Data

3.1. Models

3.1.1. Staggered Difference-in-Difference (DID) Model

To explore the impact of the YD integrated development on integrated urban–rural development, this study used the DID method to evaluate the policy effect from the two dimensions of time and region. The aim was to address any potential bias in the evaluation results by simply comparing the differences before and after a policy implementation while ignoring other relevant factors. Since the timing of the inclusion of cities in the YD integrated development has been inconsistent, the staggered DID method was used to test its effect on integrated urban–rural development. The baseline measurement model was set as follows:

$${URI}_{it}={\alpha }_0+{\alpha }_1{Policy}_{it}+{\alpha }_2{control}_{it}+{year}_t+{city}_i+{\epsilon }_{it}$$

(1)

where URI_it denotes the explained variable, namely, the level of the integrated urban–rural development; the subscript i is the city and the subscript t is the year; Policy is a dummy variable for joining the YD integrated development; control_it refers to the control variables; year and city represent the time- and region-fixed effects (controlling for the effects of changes over time and individual city-level differences, respectively); a refers to the parameters to be estimated; ε is the residual term. The coefficient a1, the parameter under focus in this study, reflects the difference in the level of the integrated urban–rural development between the experimental and control groups after the policy implementation.

3.1.2. Parallel Trend Test

The DID approach necessitates the parallel trend test to ensure there is no substantial difference in the time trend between the experimental and control groups before the policy implementation. To confirm this, parallel trend and dynamic effect tests were conducted. The model was formulated as follows:

$${URI}_{it}=a_0+{\sum }_{j=-M}^N{\delta }_j{policy}_{i,t-j}+a_2{control}_{it}+{year}_t+{city}_i+{\epsilon }_{it}$$

(2)

where policy_i,t-j is a dummy variable such that, if city i implements a regional integration policy in period t-j, the variable value is 1 and 0 otherwise, while M and N represent the number of periods before and after the policy implementation, respectively. For analytical convenience, this study sets the number of periods before the policy implementation to 5 (M = 5) and the number of periods after the policy implementation to 5 (N = 5). For example, when j = 3, policy_i,t-j indicates that city i implemented a regional integration policy in period t-3, denoting the effect in the third year of the policy implementation. Therefore, δ_-5~δ_-1 indicates the policy effects before the policy implementation; if the regression coefficient of δ_-5~δ_-1 is not significant, there is no significant difference between the experimental and control groups before the policy implementation and the sample satisfies the parallel trend assumption. δ₁~δ₅ indicates the dynamic effect of the policy during and after its implementation. The remaining variables and parameters have the same meanings as those in Equation (1).

3.2. Variables

In Equation (1), the explained variable was the level of the integrated urban–rural development. Within the framework of the YD integrated development, the focus of integrated urban–rural development has transitioned to mutual integration and coordinated development, involving public services, factor allocation, industrial development, and environmental protection, rather than simply pursuing economic growth [55,56]. To measure the integrated urban–rural development, we considered indicators that would reflect both the differences between urban and rural areas and the results of integrated urban–rural development. Based on the principles of scientific evaluation, representativeness, accuracy, and availability, 16 relevant indicators were chosen from the four dimensions of spatial, social, economic, and environmental integration to evaluate the integrated urban–rural development. Furthermore, the entropy-weighted Technique for Order Performance by the Similarity to the Ideal Solution method was used to expand the measurement.

Table 1. Index system and meanings of the integrated urban–rural development.

Indicator Dimension	Indicator Name	Indicator Description	Indicator Attribute
Spatial integration	Population urbanisation level	Population in urban areas/resident population in the region	+
	Land urbanisation level	Built-up area/administrative district area	+
	Per capita car ownership in urban and rural areas	Private car ownership by urban and rural residents/total population	+
	Ratio of the efficiency of cultivated land/built-up land	Output of primary industry/cultivated land area/Output of secondary and tertiary industry/built-up area	+
Social integration	Proportion of students in higher education	Number of students in higher education/regional resident population	+
	Investment in education	Public education expenditure per capita	+
	Number of beds per 10,000 people	Number of beds in medical institutions/resident population	+
	The coverage rate of old-age insurance in urban and rural areas	Number of urban and rural residents enrolled in old-age insurance/number of permanent residents	+
	Coverage rate of unemployment insurance in urban and rural areas	Number of urban and rural residents enrolled in unemployment insurance/number of permanent residents	+
Economic integration	Binary comparison coefficient	Output of primary industry/number of employees in primary industry/Output of secondary and tertiary industries/number of employees in secondary and tertiary industries	+
	Efficiency of primary sector output	Output per person employed in the primary sector/GDP per capita	+
	Ratio of the per capita income of urban and rural residents	Per capita disposable income of urban permanent residents/per capita disposable income of rural permanent residents	-
	Ratio of the per capita consumption of urban and rural households	Per capita consumption of urban households/per capita consumption of rural households	-
	The ratio of Engel’s coefficient between urban and rural areas	Engel’s coefficient in urban areas/Engel’s coefficient in rural areas	+
Environmental integration	Level of urban and rural greening	Green coverage rate of built-up area	+
	Level of energy saving and pollution reduction	Total fertiliser consumption/primary industry output	-

explains these 16 indicators.

The explanatory variable was the YD integrated development policy. This study took the timing of the release of two documents, namely, the 2010 Regional Planning for the Yangtze River Delta Region and the 2016 Development Plan for the Yangtze River Delta Urban Agglomeration, as the corresponding time points to examine the policy effects. The variable was coded 0 for the time points before the publication of the policy and 1 for the time points in the year of the publication of the policy and thereafter. When selecting the samples for the experimental and control groups, we excluded the Beijing–Tianjin–Hebei region, the Pearl River Delta, and cities that underwent major administrative adjustments during the study period from the sample selection process following a previous study [57]. Finally, 26 core cities at and above the prefecture level as stipulated in the YD integrated development were selected as the experimental group, while 224 cities at and above the prefecture level were used as the control group (Figure 2).

In addition to the effects of the YD integrated development, the integrated urban–rural development is influenced by other socioeconomic factors. Therefore, the following control variables were selected for the analysis following the literature. Population density (LnPD) was measured as the logarithm of the resident population per unit area (individuals/km²); government fiscal size (Gov) was defined as the ratio of local government fiscal expenditure to GDP; the degree of openness (Open) was measured by the ratio of total foreign capital used to GDP, with the amount of foreign capital used in US dollars being converted into RMB according to the average exchange rate of the year; financial support (Finance) was defined as the ratio of financial institutions’ loan balances to GDP at the end of the year; the level of industrial structure (Indus) was calculated based on the percentage of value added from the secondary sector in GDP.

Table 2. Descriptive statistics of the variables.

Variable	Definition	Mean	Std. Dev.	Min	Max
URI	Level of the integrated urban–rural development	0.398	0.099	0.146	0.909
Policy	The strategy for integrated development of the Yangtze River Delta (YD integrated development) dummy variables	0.056	0.229	0.000	1.000
LnPD	The logarithm of the resident population per unit area (individuals/km2)	7.977	0.787	4.341	9.619
Gov	General budgets of local governments/GDP	0.184	0.110	0.031	1.485
Open	Total foreign capital used (in RMB)/GDP	0.013	0.018	0.000	0.376
Finance	Year-end balance of loans from financial institutions/GDP	0.873	0.551	0.075	9.622
Indus	Value added by the secondary industry/GDP	0.476	0.114	0.107	0.908

shows the descriptive analysis of each variable.

3.3. Data Processing and Sources

The socioeconomic data required for this study were primarily derived from the China Urban Statistical Yearbook, China Urban Construction Statistical Yearbook, China Statistical Yearbook for Regional Economy from 2003 to 2020, the statistical yearbooks of all provinces (municipalities directly under the Central Government), and the statistical yearbooks and statistical bulletins of prefecture-level cities. To eliminate the influence of price factors, this study refined the relevant socioeconomic data. The issue of missing data in relation to individual cities was addressed by interpolation using the mean value or linear trend method.

4. Empirical Analysis

4.1. Regression Results

To investigate the impact of the YD integrated development on integrated urban–rural development, a full sample estimation was first carried out using Equation (1).

Table 3. Regression results.

Variable	Model 1	Model 2	Model 3	Model 4
Policy	0.152 ***	0.117 ***	0.019 ***	0.014 ***
	(23.73)	(19.18)	(6.72)	(5.16)
Control	NO	YES	NO	YES
City	NO	NO	YES	YES
Year	NO	NO	YES	YES
_cons	0.390 ***	0.197 ***	0.307 ***	0.277 ***
	(296.11)	(11.71)	(83.35)	(24.61)
R2	0.127	0.381	0.951	0.954
N	5000	5000	5000	5000

Notes: Robust t-statistics are shown in parentheses; *** p < 0.01.

presents the regression results. According to the results obtained for Models 1–4, the estimation coefficients of the Policy variables were positive and significant at the 1% level; that is, the YD integrated development had a notable positive effect on the integrated urban–rural development. After incorporating the control variables as well as the time- and region-fixed effects, the estimated coefficient of the Policy variable was 0.014. This indicated that the YD integrated development played a significant role in promoting integrated urban–rural development, with a significant increase of 1.40% compared with control cities, supporting Hypothesis 1. Regional integration policy aims to break down the local protection and administrative barriers that exist between cities and achieve economic growth and the coordinated development of all cities through the free flow and optimal allocation of factors.

4.2. Robustness Testing

4.2.1. Parallel Trend and Dynamic Effect Test Results

The year before the policy implementation (N = −1) was chosen as the baseline period. Figure 3 shows the regression results obtained using Equation (2), and the estimation results of the key coefficient δ were used to test the parallel trend and dynamic effects of the policy. The main coefficients before the policy implementation were not statistically significant, suggesting no noticeable distinction between the experimental and control groups in the absence of the policy intervention, thus supporting the parallel trend assumption. Furthermore, the test results of the dynamic marginal effect indicated that the coefficient during the initial two years of policy implementation was low and not statistically significant, whereas the core coefficient in the third year after the policy implementation began to be significantly positive, indicating that the YD integrated development had begun to improve integrated urban–rural development, with a short-term lag and long-term impact. These empirical results confirmed Hypothesis 2 that the YD integrated development was oriented towards integrating cities into the region. Although the capital flow, industrial transfer, and economic ties in the Yangtze River Delta region have been enhanced to varying degrees by removing administrative barriers, these effects are not noticeable in the short term. However, over the longer term, objective and significant policy gains such as infrastructure construction, factor flows, and industrial output growth have begun to steadily promote integrated urban–rural development.

4.2.2. PSM-DID Test Results

Since the socioeconomic development of some cities in the Yangtze River Delta region is high nationally, we aimed to avoid control group selection bias affecting the robustness of the results. Hence, the PSM-DID method was used to further test the impact of the YD integrated development on the integrated urban–rural development. In addition, owing to the inconsistent timing of the policy interventions, year-by-year matching was chosen to screen the experimental sample more accurately.

Table 4. Results of correcting for selection bias.

Variable	Model 5	Model 6	Model 7	Model 8
	1:2 NNM	RM	KM	EAST
Policy	0.006 *	0.007 **	0.012 ***	0.005 *
	(1.70)	(2.18)	(4.44)	(1.73)
Control	YES	YES	YES	YES
City	YES	YES	YES	YES
Year	YES	YES	YES	YES
_cons	−0.267 ***	−0.097 ***	0.281 ***	0.348 ***
	(−3.49)	(−1.49)	(21.33)	(18.66)
R2	0.956	0.954	0.951	0.960
N	1188	1363	4421	1540

Notes: Robust t-statistics are shown in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

lists the 1:2 nearest neighbour matching (1:2 NNM), radius matching (RM), kernel matching (KM), and the manually screened ‘six provinces and one city’ in East China1. The estimated coefficients from various matching methods may vary slightly, but they all show significant positive results above the 10% level, indicating that the empirical results showing that the YD integrated development promoted integrated urban–rural development were relatively robust.

4.2.3. Placebo Effect Test

This section takes the approach of randomly generating an experimental group of cities to conduct placebo tests. This involved randomly selecting the same number of ‘pseudo-experimental groups’ as the original experimental group, assuming that these cities were those in which the YD integrated development was implemented, with the other cities serving as the control group. We repeated the sampling 500 and 1000 times, respectively. We then re-estimated the samples to obtain the experimental results for the ‘pseudo-policy dummy variables’ and subsequently compared the regression coefficients of the randomly selected samples with the baseline coefficients. If there was a significant difference between the two, we considered that the policy effects generated were due to the policy alone. Figure 4 reports the results of the placebo test. It can be seen that the regression coefficients were distributed centrally around 0, in stark contrast to the benchmark regression coefficient of 0.014, indicating no interference from other factors. Hence, the YD integrated development positively affected integrated urban–rural development.

4.2.4. Other Robustness Tests

In addition to the robustness tests mentioned earlier, we conducted three additional tests. These tests involved changing the explanatory variables, adjusting the sample period, and changing the timing of policy shocks. One of the tests involved changing the explanatory variable to the urban–rural income gap in the regression model 9. This was performed because economic factors are the most obvious and direct component of the integrated urban–rural development process. Another test involved setting the sample period to 2003–2019 in model 10. This decision was made to account for the potential interference of the COVID-19 factor on the socioeconomic data. The third test involved advancing the timing of policy shocks by 3–5 years in models 11–13. This was performed to test whether the policy impacts of YD integrated development could be attributed to other policies or stochastic factors. The results of the three robustness tests mentioned above are presented in

Table 5. Results of other robustness tests.

Variable	Model 9	Model 10	Model 11	Model 12	Model 13
Policy	−0.019 ***	0.015 ***	0.005	0.000	−0.009 **
	(−12.41)	(5.01)	(1.27)	(0.10)	(1.73)
Control	YES	YES	YES	YES	YES
City	YES	YES	YES	YES	YES
Year	YES	YES	YES	YES	YES
_cons	−0.102 ***	0.276 ***	0.277 ***	0.278 ***	0.280 ***
	(−9.20)	(25.15)	(24.61)	(24.45)	(24.37)
R2	0.883	0.951	0.953	0.954	0.954
N	5000	4250	5000	5000	5000

Notes: Robust t-statistics are shown in parentheses; *** p < 0.01, ** p < 0.05.

. They reveal that YD integrated development has a significant impact on reducing the urban–rural income gap. Furthermore, the coefficients show that adjusting the sample period to 2003–2019 yields significantly positive results, whereas advancing the timing of policy impacts by 3–5 years yields insignificant or negative results. This further validates the reliability of the benchmark regression results presented in this paper.

4.3. Mechanism Analysis

The above results show the robust relationship between the YD integrated development and integrated urban–rural development. With the analyses in the previous Section 2, we focused on the three aspects of economic growth, road transport links, and technological progress. First, in contrast to traditional socioeconomic indicators, night-light data are not influenced by price factors and were chosen as the proxy variable for economic growth. Second, a good transportation infrastructure can have a multiplier effect on resource reallocation and the density of transportation facilities in a city reflects its accessibility well. Therefore, we chose the ratio of highway mileage to the size of the administrative area to define the degree of road transport links. Finally, scientific and technological progress is a crucial driver of integrated urban–rural development, not only accelerating industrialisation and urbanisation, but also affecting the mechanisation and modernisation of agriculture significantly. Thus, we used the cities’ science and technology inputs and general budget expenditure to measure scientific and technological progress. The regression results for the three action channels indicate that the YD integrated development significantly fosters urban economic growth, road transport links, and technological progress (

Table 6. Mechanism analysis.

Variables	Model 14	Model 15	Model 16
	Lnlight	Road	Technology
Policy	0.183 ***	0.126 ***	0.022 ***
	(14.42)	(7.34)	(16.20)
Control	YES	YES	YES
City FE	YES	YES	YES
Year FE	YES	YES	YES
_cons	0.864 ***	−0.691 ***	0.006 *
	(11.77)	(−6.50)	(1.65)
R2	0.977	0.902	0.732
N	5000	5000	5000

Notes: Robust t-statistics are shown in parentheses; *** p < 0.01, * p < 0.1.

). Hypothesis 3 is tested.

4.4. Heterogeneity Analysis

Based on the regression results from the full sample, the YD integrated development increased the integrated urban–rural development markedly. However, the policy effects are likely to differ across types of cities owing to their different economic fundamentals, technological levels, development orientations, and geographical locations. Therefore, we compared the differences in the policy effects from the perspectives of population size and city classification in a subsample regression.

First, for population size, we subdivided large cities (i.e., those with more than one million people) and small- and medium-sized cities (i.e., those with fewer than one million people), according to the Notice of the State Council on Adjusting the Classification Standards of Urban Size (Guo Fa [2014] No. 51). The results for models 17 and 18 in

Table 7. Heterogeneity analysis results.

Variable	Population Size		City Classification
	Model 17	Model 18	Model 19	Model 20
	Large Cities	Small- and Medium-Sized Cities	High-Level Cities	General-Level Cities
Policy	0.014 ***	0.007 *	0.015 **	0.010 ***
	(3.78)	(2.13)	(2.43)	(4.35)
Control	YES	YES	YES	YES
City	YES	YES	YES	YES
Year	YES	YES	YES	YES
_cons	0.589 ***	0.266 ***	0.592 ***	0.269 ***
	(28.30)	(23.35)	(16.68)	(25.49)
R2	0.952	0.947	0.935	0.949
N	1520	3480	860	4140

Notes: Robust t-statistics are shown in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

show the effects of the YD integrated development on the integrated urban–rural development for large and small- and medium-sized cities, respectively. The Policy coefficients of both types were statistically significant at the 10% level and above, with the coefficients and significance levels for large cities significantly higher than those for small- and medium-sized cities, perhaps because population size is directly related to resource endowment, transport links, and economic growth at the city level. Assuming that a city is geographically remote and materially under-resourced, its economic development is likely to be relatively backward, with population migration then inevitable. Therefore, for large cities, the weakening of administrative barriers and local protection brought about by regional integration policies could augment the ‘agglomeration effect’. This would lead to a better flow and optimal allocation of factors between cities and urban and rural areas, thereby promoting the integrated development of both regions.

Second, for city classification, we defined provincial capitals, sub-provincial cities, and cities with legislative power as high-level cities and the others as general-level cities, as there are large differences in policy support across the cities in China, which determine their future development direction. The results for models 19 and 20 show that the effects of the YD integrated development were significantly positive for both city classifications, but that the effect for high-level cities was higher than that for general-level cities. This may be because high-level cities benefit from easier resource allocation, resulting in more government subsidies, financing options, skilled labour, and lower local tax burdens [58]. By contrast, development started later in general-level cities and the average development level is relatively low, and there is still a long way to go for integrated urban–rural development.

5. Discussion

The promotion of high-quality sustainable socioeconomic development is a focus of attention in countries worldwide. In the context of this goal, governments at all levels have increasingly become committed to deepening regional cooperation to foster new drivers of development, initiating regional integration development strategies at different levels, including the EU, APEC, B&R, and the YD integrated development. While studies have already explained the extent to which regional integration policies affect economic growth, effective integration must address not only underdevelopment but also unbalanced development to foster common prosperity. Unlike countries with mature market economies, transition economies such as China’s economy have unique institutional advantages in that they can promote a better combination of a well-functioning government and an efficient market through strong administrative intervention. We find that regional integration policy not only directly affects the whole region and its cities, but also indirectly affects the integrated urban–rural development. Also, the findings of this study can be corroborated with the conclusions of previous studies. For example, from the perspective of urban agglomeration, scholars have confirmed that regional integration policies affect cities’ economic resilience significantly and evaluated the effects of mechanisms such as urban size, industrial structure, and innovation capacity [18,59]. One study found that urbanisation in China has enhanced urban–rural relations and reduced the gaps between urban and rural areas [60]. Furthermore, scholars have employed a singular indicator, specifically the disposable income ratio, to quantify development disparities and found that regional integration reduces the urban–rural income gap through the mechanisms of transaction costs, knowledge diffusion, and functional division [61]. These theoretical and practical discussions have provided clarity regarding the direction for our next step in promoting integrated urban–rural development.

No research is perfect. This study has some limitations. First, it used documents issued at the national level as time points for identifying the policy implementation, while the experimental group was limited to the YRD region, which may reduce the generalizability of the study’s findings to other regions. Second, while this study sought to comprehensively measure the integrated urban–rural development level from different perspectives, considering all the indicators used in previous studies was impossible owing to issues with the integrity of the data and multicollinearity between the variables. In future research, more variables should be used to draw urban–rural comparisons and the structure of the various indicators should be changed to enhance the reliability of the research findings. Finally, integrated urban–rural development involves both exogenous and endogenous dynamics. This study only examined the exogenous impact of the YD integrated development on integrated urban–rural development. Future studies should explore more extensive endogenous dynamics and exogenous impacts under this new stage of integrated urban–rural development.

6. Conclusions and Implications

This paper constructed a comprehensive evaluation system for integrated urban–rural development from the following four dimensions: spatial, social, economic, and environmental, and explored the impact effect of the YD integrated development on integrated urban–rural development with the panel data of 250 cities from 2003 to 2022, and drew the following conclusions: First, the YD integrated development significantly promotes integrated urban–rural development, which is a valid conclusion after parallel trend tests and a series of robustness checks. Second, the policy effects of YD integrated development vary across different groups. The policy impacts of large and high-level cities are significantly higher than those of small- and medium-sized and general-level cities. Third, YD integrated development promotes integrated urban–rural development through the three paths of urban economic growth, enhanced road transport links, and improved technology. Our conclusions can provide a reference for the next steps in promoting integrated urban–rural development and achieving common prosperity.

China remains on a distinctive development pathway, with unbalanced and inadequate socioeconomic development in most regions. The integrated urban–rural development emphasised in this study does not erase the functional distinctions between urban and rural areas or imply the need for uniform egalitarianism. Instead, it focuses on enhancing the complementary roles of urban and rural regions, facilitating efficient economic flows, and attaining high-quality integrated development by establishing a new form of urban–rural relationship and spatial organisation. In the future, policymakers must clearly understand the main lines of integrated urban–rural development, strengthen overall planning and top-level design, overcome the institutional drawbacks of urban–rural duality, and harness the synergy of external driving forces and endogenous capacity. Specific policy implications are as follows:

First, this study confirms the positive effects of YD integrated development on integrated urban–rural development. In the future, the experience and practices of the YD integrated development should be further summarised and extended to other city clusters and metropolitan areas in China. It strengthens connections and cooperation between provinces, cities, and urban and rural areas through regional integration policies. This includes eliminating local protection and administrative fragmentation and promoting high-quality development through a unified, open, competitive, and orderly market economy.

Second, differences in the level and stage of development have resulted in policies being more effective in large cities (high-level) than in small- and medium-sized cities (general-level). In the future, it is necessary to take full account of natural and socioeconomic locations and implement precise and differentiated policy supplements for different cities. On the one hand, further encouragement should be given to enhance the comprehensive carrying capacity of regional centres such as Shanghai, Nanjing, Suzhou, Wuxi, Hangzhou, and Ningbo to effectively bring into play the effectiveness of borrowing scale and drive the overall development of the regions. On the other hand, we must be highly alert to the phenomenon of agglomeration shadows, such as environmental pollution, population loss, and efficiency loss, faced by regions under integrated development, and clearly define the functional positioning and layout of the industrial division of labour through the formulation of tilted and oriented policies so to achieve overall integrated planning and dislocated development between centres and peripheral cities, and between cities and villages.

Third, improvements in the overall economic efficiency of cities, the enhancement of road transport links, and technological progress can all facilitate the effect of YD integrated development on integrated urban–rural development. To improve the overall economic efficiency of cities, the rural property rights system must be reformed to improve the circulation and transfer mechanism of rural land, revitalise rural land capital, enhance the vitality of rural development, and improve the efficiency and level of factor allocation in both urban and rural areas. Concerning road transport links, different types of transportation should first be identified and developed accordingly. For road construction, promoting the balanced development of urban and rural roads (i.e., both quantity and quality) is crucial, as is considering the integration of the environment and landscape. Concerning railroad construction, the coverage of inter-city and express railroads must increase, and an efficient and convenient urban and rural passenger and freight transportation network must be established together with highways. In terms of technological progress, the use of new-generation technologies such as artificial intelligence, big data, and 5G must be increased, while policy support and guidance in areas such as e-commerce, livestreaming, and digital finance must be provided to low-income rural communities.