In cities around the country, huge swaths of property in desirable locations house only empty warehouses, barely-used shipping facilities, and heavily subsidized industrial-age factories, often right across the street from high-end condos and office buildings. The reason is a widely-used, but poorly understood form of local industrial policy known as non-cumulative zoning. In contrast with traditional Euclidean zoning, in which manufacturing uses were prohibited in residential areas but not vice versa, areas that are zoned non-cumulatively allow only manufacturing uses and bar any residential (and sometimes even commercial uses) of property. The arguments for non-cumulative zoning are always the same: Cities seek to (a) reduce the degree to which urban manufacturers are held responsible for nuisance and (b) subsidize urban manufacturing by reducing the competition for land and hence reducing the price.
In this Article, we argue that non-cumulative zoning is an idea whose time has passed, if there ever was a convincing case for it at all. The two major justifications for non-cumulative zoning are flawed, and alternative means could achieve the same ends with fewer costs. The large number of nuisance claims engendered by urban manufacturing could be addressed by creating a “right to stink” in certain zones, allowing residential and commercial users to move into these zones but prohibiting them from suing manufacturers who are not violating regulatory laws. As for the second manufacturer-subsidizing justification, subsidies cannot be justified in terms of a subsidizing city’s own welfare unless the external “agglomeration” benefits of manufacturing exceed the cost of the subsidy to the city. Moreover, the broader social perspective also requires that some cities are better able to capture those agglomeration benefits than others, meaning that competition between jurisdictions could result in total increases in wealth. However, non-cumulative zoning is unlikely to achieve either local or broader social efficiency. Its scope is not closely tied to any theory of external benefit; it encourages the inefficient use of land and the substitution of land for other inputs; and it hides the true cost of urban manufacturing subsidies from the public. If urban manufacturing must be subsidized, a direct cash subsidy system would be preferable, particularly if it could be funded directly from taxes on the increased value of land caused by the removal of a non-cumulative zoning designation.