Comments

Joseph Gyourko: The first important point to make about Evenson and Wheaton's paper is that it deals with an extremely useful topic, because how we organize ourselves spatially has important socioeconomic implications for many aspects of our personal lives and for the communities in which we live. While the analysis of land use regulation in general and zoning in particular has a lengthy pedigree in economics, the fact is that theoretical models are far better developed than empirical analyses.⁴² Evenson and Wheaton set out toalleviate the primary reason for this—the lack of good data on land use controls.

Important new data based on a 1999 satellite-based survey of all open land in Massachusetts are matched with information on zoning ordinances in a large number of local towns and cities. A few other studies have amassed data on a specific type of land use control, but this is the first of which I am aware that allows one to observe multiple forms of land use controls simultaneously and do so over an area as extended as a state.

The goals of the paper are threefold: to document a new database, to report stylized facts about how towns actually do zone and regulate land use, and to produce results on certain key issues relating to how land use regulation varies with factors such as income. The paper is most successful in achieving the first two goals. Many of the empirical results are interesting in their own right, but more structure and even better data probably are needed to convince reasonable skeptics who propose alternative explanations for the findings. Even so, the work points the way to much future research for those interested in this issue and provides a valuable new data source with which researchers can try to satisfy their curiosity.

The paper includes a useful discussion of the existing studies on minimum lot size (and exclusionary zoning), on the provision of open space, on theeconomics of allowing commercial development, and on theimpacts of land use controls on housing prices. I would be interested in seeing the discussion and analysis extended to the issue of race. We know that there is substantial racial segregation in virtually all major metropolitan areas in the United States. The extent to which this might be correlated with how communities zone seems a potentially useful line of inquiry.

The data themselves are the best that urban scholars have ever had. As noted above, the key underlying source is a 1999 satellite survey of Massachusetts. The output for this survey is then merged into a variety of current land use files that allow the authors to know existing land use patterns as well as zoned land use patterns (that is, what the current law would allow for the future). The survey includes thirty-six detailed categories that range from residential lots of at least one-half acre to two types of bogs. These data are collapsed into four categories (residential, commercial, open space, and other) for the purposes of this paper.⁴³ I see no reason to require a more detailed breakdown for this initial foray with the data, but future work certainly should investigate the utility of using finer breakdowns. The zoning data themselves are current as of 1996 in virtually all cases (and are even more recent in most cases), so one can be reasonably confident that current land usage is matched with current zoning conditions.⁴⁴

A fairly complex set of calculations and mappings of one data set to another is employed to compute the amount of buildout permitted under existing land use regulations. I was convinced from the authors' presentation that the results were accurate because they could pin down the correct status of various detailed parcels (for example, Boston Common). However, this fact is not entirely clear from the text, and the paper would benefit from an expanded discussion, possibly with examples from a few places, so that readers are...