While advancing urban digital transformation, smart city development exerts profound influences on the spatial configuration of cities. A nuanced understanding of its impacts on urban polycentric development is critical for formulating evidence-based urban spatial planning policies in the context of rapid digitalization. This study used a panel dataset of 269 prefecture-level cities in China covering the period 2003-2019 and employed multi-period spatial difference-in-differences (DID) models to systematically examine the direct effects and spillover impacts of national smart city pilot policies on urban polycentricity. The research further delved into the distance-decay thresholds of spillovers, city heterogeneity in policy responses, and the moderating roles of institutional factors in shaping these relationships. Key findings include: 1) Smart city construction had an inhibitory effect on the process of urban polycentric development. 2) Across three distinct spatial weight matrices—geographic proximity, economic similarity, and nested economic-geographic frameworks—smart city construction demonstrated negative spatial spillovers on neighboring cities' polycentric development. The geographic proximity matrix yields the most pronounced negative effects, indicating that geographic distance remains a key factor influencing resource flows between cities and the transmission of policy effects. 3) The spatial spillover effects of smart city construction exhibited characteristics of attenuation as geographic distance increased and spatial scale expanded, presenting a circular hierarchy of close-distance decay of pilot cities-provincial regions-urban agglomerations. The effects exhibited a trend of first strengthening and then weakening with increasing distance—indirect spillover effects were most pronounced in the 200-400 km distance range, reaching their peak at 400 km, and became insignificant beyond 600 km. This trend reflects the nonlinear transmission patterns of policy effects across different geographic distances and spatial scales. 4) Heterogeneity analysis revealed that, in terms of the division of economic zones, urban hierarchy, and dominant industry, the policy spillover effect was positively significant in cities in the western region, higher-level cities, and cities dominated by the tertiary industry. 5) Market integration levels and governmental governance efficiency played pivotal moderating roles. Lower market segmentation and higher governance capacity amplified the positive effects of smart city policies on polycentric development by facilitating resource flows and coordinated planning, while excessive administrative barriers weakened these impacts. The study underscored the need for spatially differentiated smart city development strategies that account for regional economic landscapes, urban hierarchies, and institutional contexts. Policymakers should adopt adaptive spatial planning approaches—such as promoting cross-city digital collaboration in urban agglomerations and enhancing governance coordination—to leverage smart city investments for optimizing urban internal structures and fostering balanced regional development in China's new-type urbanization process.

Since 2008, China's urban development model has undergone a financialized transformation, driven by a shift in the logic of land development from "commodification" to "assetization". Urban redevelopment processes epitomize such transformation. Existing research has primarily examined land financialization, focusing on land-collateralized borrowing, infrastructure construction, and the indebtedness of local governments, thereby elucidating the financing model of urban redevelopment. This study shifts the lens to the investment side, analyzing how mechanisms of commodification, assetization, and capitalization restructure property-led redevelopment and reconfigure the urban accumulation regime. Using Guangzhou City as a case study, the findings reveal that: 1) at the urban scale, redevelopment programs prioritize infrastructure investment, leveraging its capitalization effects to elevate the asset value of surrounding land and housing, thereby generating higher economic returns for local governments and developers; and 2) at the household scale, housing assetization reinforces middle-class families' property consciousness and investment motivations, with growing numbers of households engaging in property investment driven by optimism about urban development prospects. This asset-based urban accumulation regime, heavily reliant on real estate appreciation, fundamentally sustains the land-collateralized financing model and the land-driven fiscal regime. The article concludes by reflecting on the current housing market downturn and considers possible avenues for reframing the logic of assetization to guide new governance models of urban redevelopment.

In the context of urban renewal, it is of great value to understand the housing decision-making mechanism of residents in old urban communities to improve the suitability and effectiveness of old community regeneration. Based on the sampling data of Jiucheng District in the center of Chongqing Municipality in the 2023 Urban Physical Examination Social Survey of the Ministry of Housing and Urban-Rural Development, this study constructed a dual-model analysis framework of renovation intention and relocation intention, and used the random forest algorithm combined with explainable model to systematically explore the influencing factors and correlation effects of the two types of decision-making behaviors. The results indicate that residents in old urban communities showed a significant preference for in-situ improvement. Location is the primary factor that influenced housing decision making, renovation and relocation intentions formed completely different distribution patterns in urban space, and a key spatial threshold appeared at 10 km distance. In terms of housing characteristics, there was a positive correlation but with a decreasing margin between floors and relocation intentions, and a "U"-shaped relationship for relocation. The clarity of property rights strengthened the tendency to renovate and inhibited the relocation behavior; Housing quality was negatively correlated with both types of intentions. The length of residence had the opposite effect on the two. The social environment had a significant negative effect on the intention to renovate and the intention to relocate. In terms of individual characteristics, the middle-aged group rather than the elderly showed the strongest intention to renovate, and the elderly also had the most conservative tendency to relocate. Residents with higher income levels and education levels showed more positive attitudes in both renovation and relocation decisions, reflecting their decision-making flexibility. In addition, there was a significant substitution effect between renovation and relocation intentions. This study provides a theoretical support for targeted and differentiated urban renewal policies, and is helpful for realizing the urban renewal goal of "co-construction, co-governance, and sharing" of the people.

Research on the identification and evaluation of inefficient land use is the premise of revitalizing and reusing the stock land. Existing studies focused on the comprehensive evaluation of inefficient land use, but lack interactive analysis of a variety of inefficiency factors, and ignore the excavation of inefficient high potential parcels. In addition, existing inefficient land use identification models usually use linear methods such as entropy weight method, which do not consider the nonlinear relationships between the attribute characteristics and the degree of land use. Therefore, this study took land parcels as the spatial unit of research and proposed a dual dimensional interactive framework for identifying and evaluating inefficient land use. Specifically, BP neural network was used to identify the internal development intensity of land parcels, entropy weight method was used to evaluate the external locational conditions of land parcels, and the interaction between internal development intensity and external locational conditions was analyzed to identify low development intensity-favorable locational conditions land parcels with synergistic effects. An empirical research was conducted using Beijing as an example. The results show that the BP neural network model can better take into account the nonlinear relationships between the internal attribute characteristics and development intensity. Through the coupling analysis between inefficient land use and urban regeneration districts, we found that most relatively severe/severe low efficiency-superior/favorable locational conditions parcels are distributed in key urban renewal blocks, with a high degree of coincidence. In addition, there are also a few relatively severe/severe low inefficiency-superior/favorable locational conditions parcels outside the key blocks, mainly distributed in Yangfangdian and Qinghe Communities of Haidian District, Lugouqiao Community and Nanyuan Community of Fengtai District, and Jinsong Community and Shibalidian area of Chaoyang District. The research results can provide a reference for the fine-scale governance and dynamic improvement of urban renewal blocks.

In the context of urbanization transition from growth to stock retention, optimizing and enhancing land use functions to revitalize inefficient urban stock space is of significant theoretical and practical importance for urban ecological protection and livability improvement, high-quality sustainable development, and improving governance levels. However, current research on inefficient urban stock space mainly focuses on social, economic, and built environment dimensions, with few scholars addressing the issue of inefficient urban stock space from a functional perspective using detailed and quantitative research methods. Therefore, this study adopted a multifunctional land use perspective, integrating multi-source data such as statistics, surveys, remote sensing, and point-of-interest (POI), along with theories and methods from the systems theory, spatial governance theory, and GIS spatial analysis, to explain the formation mechanism of inefficient urban stock space. It also developed a precise identification method for urban inefficient stock space and governance optimization path, using Xi'an City as a case study for empirical research. The study found that: 1) The spatial efficiency of urban stock space of Xi'an showed a "high in the center-low at the periphery" distribution, gradually decreasing from the city center to the outer areas. 2) The production-living deficient function was the main obstructive function for one-third of the area in Xi'an City. 3) The dual-function improvement zone contained the largest number of towns, followed by the triple-function improvement zone, and the single-function improvement zone had the smallest number of towns. Enhancing and optimizing the structural functions at the overall urban level is the primary task to comprehensively improve the efficiency of inefficient urban stock space. Moving forward, Xi'an should adhere to the principles of planning guidance and regional optimization, focus on improving deficient functions, and optimize the entire inefficient urban space to promote the healthy development and protection of territorial space, thereby achieving the goal of high-quality development. This study provides theoretical, methodological, and empirical references for optimizing inefficient urban stock space in Xi'an and other megacities.

As an important instrument of high-density urban area micro-renewal, pocket parks significantly enhance green space accessibility, advance public welfare, and elevate residential environmental standards by repurposing underused urban spaces. During the strategic scaling phase of pocket parks, elucidating their spatial configuration mechanisms is essential for optimizing their site selection and spatial distribution. This study leveraged geospatial data from pocket parks in Shenyang City and integrated multi-source geographic datasets, employing a repeated random sampling-verification framework combined with interpretable machine learning algorithms to systematically decode the complex drivers of pocket park distribution through factor prioritization, nonlinear relationship analysis, and interaction effect quantification. The optimal algorithm modeling revealed that: 1) The repeated random sampling-verification model substantially mitigates subjectivity and stochastic bias in negative sample selection, thereby improving the reliability of machine learning outputs. 2) Compared to socioeconomic indicators such as GDP, spatial determinants including green space proximity and public transit accessibility exerted disproportionately stronger influences on distribution patterns, with housing costs demonstrating a slight negative marginal effect, land use diversity showing statistically negligible effects, and notable synergistic interactions emerging in park service overlap zones. 3) Threshold-dependent response dynamics (positive/negative/nonlinear) among key drivers establish quantitative guidelines for precision-oriented placement strategies. This research advances the theoretical foundations of pocket park planning while offering novel conceptual frameworks and practical implications for deciphering the spatial interplay between high-density urban configurations and micro-green infrastructure.