Shaping the Coast with Permits: Making the State Regulatory Permitting Process Transparent with Text Mining

This paper examines the California Coastal Commission's permitting process. Using several text mining techniques, including web scraping, information extraction, and supervised classification, I demonstrate how to retrieve empirical data from unstructured texts, namely public meeting agendas and staff reports. Contrary to the concern that the Commission routinely delays or rejects permitting requests, the data reveal that outright rejection of permit applications is rare. On average, eight of ten applications were approved. Single-family homes and commercial development projects were approved about 80% of the time; the rates were about 70% for seawalls and retaining walls, and 60% for land-use changes. Most applications were processed swiftly, with a median application length of 3 months. The agency's influence comes primarily from negotiating each application. Qualitative study of 50 cases pertaining to single-family home construction reveals that the agency adopts a “managed development” approach, that is, allowing development but scrupulously managing various aspects of development. These case studies illustrate how the agency interprets the broad, abstract state laws and translates the mandates into enforceable actions as permitting conditions. In areas where the state mandates conflict, particularly over development in receding shorelines, the agency has the largest leverage in creating and implementing its preferred policies. The text mining techniques demonstrated in this paper can be applied to study any governmental agency. These techniques help to extract information from a massive volume of papers and organize them into a database for analyses. The empirical data extracted from texts can significantly increase bureaucratic transparency.