How does the public-private cooperation network affect property value? Analyzing Manhattan's business improvement districts

Business improvement districts (BIDs) have been adopted since the 1970s as innovative mechanisms for urban governance. However, limited research has quantitatively examined the complex cooperation dynamics within BIDs. To address this gap, this study examines 26 BIDs in Manhattan, New York City, leveraging 2763 news and promotion texts. Drawing on urban regime theory, it integrates natural language processing (NLP) and social network analysis (SNA) to construct a public-private cooperation network spanning from 2003 to 2023. Employing hedonic price and two-way fixed effect models, the impact mechanisms of the cooperation network on property values are analyzed across three dimensions - cooperation stakeholders, modes, and projects - and at two scales: overall and individual BIDs. The study discovered the evolution stages and typical types of cooperative networks and summarized four urban regime types. The results also show that stakeholder participation, public and private, demonstrates both linear and nonlinear, as well as positive and negative effects on property values. Furthermore, a balance in public-private participation may maximize overall property values. Specifically, opposition modes within the cooperation network significantly enhance property values, while excessive stakeholder participation in projects may have adverse effects. The findings of this study further support the formulation of BID governance strategies. Moreover, the identified power-law distribution underscores the effectiveness of integrating NLP and SNA in analyzing spatial cooperation networks.