Zoning by a Thousand Cuts: The Prevalence and Nature of Incremental Regulatory Constraints on Housing

Abstract

For a century, zoning — the local-government regulation of land use, structures, and lots through the assignment of lots to distinctly regulated districts — has dictated how and where we live. This groundbreaking project asks a simple question: how do zoning laws impact housing in Connecticut? To answer it, my team catalogued, coded, and analyzed housing-related zoning for all 2,622 zoning districts in Connecticut, where virtually all of the private land is subject to zoning. It is the first time the full variation of zoning districts in one state has been rendered at all, much less in such substantive detail and in the interactive graphic form of the Connecticut Zoning Atlas (the “Atlas”). The results are stunning. The data starkly confirms the expected, including the dominance of zoning for single-family housing (90.5% of zoned land) over three-or-more-family zoning (2.5% of the state), and the correlation of multi-family housing with the prevalence of people of color. It also confirms my hypothesis that zoning jurisdictions impose significant parking requirements on housing units, which has consequences for the cost of housing and the availability of land. But some results were surprising, including that two-family housing is allowed as-of-right in over a quarter of the land in the state, and that accessory dwelling units are allowed already in 92% of towns. In general, though, the data points to a somewhat nonsensical patchwork of housing-related constraints, across a variety of jurisdictions. This Article describes the basic legal mechanics of zoning and argues it is an important area of study. It reveals that despite the importance of zoning, very few studies have conducted rigorous multi-jurisdictional surveys of actual zoning codes. The Article describes the methodology for this research, including the universe of codes reviewed, the nature of the textual analysis, the process of collecting mapping information, and the manner in which the information was translated into an interactive public interface. And it concludes with key findings, both unsurprising and surprising. The Article concludes that these findings compel state action. This represents not the end of the questions that can be addressed through the data — but the beginning of what will be ever-deepening scholarly inquiries into a unique data set, one that has the potential to improve zoning decision-making in ways that improve our response to a variety of issues, from segregation to education, economic development to deindustrialization, and the environment to public health.