Balancing spatial logic and governance objectives for street vending: An optimizing model for dynamic formalization zones

Abstract

Street vending, a key component of the informal economy, presents significant challenges for urban governance due to its informal nature, mobility, and conflicts with formal urban activities. While formalization policies aim to regulate street vending by relocating vendors to designated zones, conventional approaches often fail to balance the interests of vendors, residents, and urban authorities, leading to reduced vendor profitability, resistance, and social conflicts. This study addresses this critical gap by proposing a spatial optimization model for dynamic formalization zones, leveraging deep learning and ant colony optimization (ACO) to reconcile spatial governance objectives with vendor preferences. Using Shenzhen as a case study, the research employs street view images and social sensing data to analyze vendor distribution patterns and stakeholder preferences, delineating dynamic formalization zones with permanent “permitted areas” (270,760 m²) and time-specific “restricted areas” (497,610 m²) after 500 iterations. The results demonstrate the model's ability to balance vendor location preferences, resident eviction preferences, and management costs, offering a sustainable solution that preserves vendor mobility and flexibility. This research contributes by introducing a novel method for balancing stakeholder interests, proposing a dynamic governance approach that adapts to the inherent mobility of street vending, and providing actionable policy recommendations for urban managers, emphasizing inclusivity, efficiency, and adaptability in street vending management. By addressing the limitations of conventional formalization strategies, this study offers a practical and scalable solution for promoting inclusive and efficient urban governance.