Solves many, but not all problems

Solves many, but not all problems

My Land use casebook, like most of them, mentions New Urbanist zoning and planning techniques, but does not dwell on them.  In order to teach New Urbanist concepts such as Form-Based Codes, SmartCode, and the Transect, I had to develop my own materials, as well as shamelessly stealing a couple of Powerpoint presentations from a friend who works at Smart Growth America.

What's the cause of this gap?  Is it because land use professors have a thing about Euclidean zoning?

I doubt it.  A quick check in the Westlaw "ALLCASES" database yields only one result for the phrase "Form-Based Code" and none of the results for "transect" has anything to do with the New Urbanist land use concept.  That means that it is very difficult actually to find cases that reflect aspects of New Urbanism.

One can understand that in several ways, I suppose.  You could infer that New Urbanism just leaves less room for legal disputes than traditional Euclidean zoning.  For example, there is no need to worry about non-conforming uses, use variances, or conditional use permits with Form-Based Codes because those codes do not regulate uses to begin with.  Certainly many advocates of New Urbanism might make this argument; they would argue that New Urbanist codes, based upon building form and the transect, are more certain than traditional Euclidean zoning and also more protective of private property precisely because they leave more discretion in the market.  They are right about the general point, but it is hard to argue that any legal framework simply eliminates legal disputes -- if it did, it would either be the first such system to do so or rely upon a sort of coercion wholly at odds with New Urbanism.  (Not too many land use disputes under Stalinist land use, but that surely did not reflect an advantage of the system).

Alternatively, you could argue, as many critics of New Urbanism do, that its land use philosophy is essentially a boutique product, suitable for Berkeley, Boulder, or Austin, but not for "real" American places.  Thus, it is not prevalent enough to generate cases.  This argument runs aground on facts.  New Urbanism does not work everywhere, but it is hard not be impressed by the wide variety of American communities that are adopting it.

Instead, it seems to me that New Urbanism squeezes a lot out of the uncertainty and over-regulation out of the Euclidean system, but cannot get rid of it.  Euclidean zoning might be necessary to handle these uncertainities.

Consider an example from the casebook.  Where does one put a private school?  A specific "private school zone" would suffer from a lack of demand -- maybe the private school doesn't want to be there.  It doesn't quite belong in your standard commercial zone, which might pose traffic safety problems for children.  Making it a permitted use in a residential zone would generate a homeowner revolt.  But ignoring uses does not solve the problem -- it just shoves it under the rug.  Put it in areas with a lot of commercial uses will lead to the same traffic safety problems, and putting it in an area with predominantly residential uses will still cause a homeowner revolt, putting it in mixed use areas might lead to both problems, and it doesn't really matter if local officials say, "it's not our concern."  It is their concern, and if they say that to their constituents, they will soon find themselves out of a job.

In other words, while New Urbanism coding can serve as a replacement for a lot of Euclideanism, it cannot eliminate it entirely -- not because we are addicted to Euclidean forms, and not because we are dumb, but because lots of the world is uncertain, and cities will have to grapple with that.  Conditional use permits and variances aren't about the flawed ideology of early-20th century planners, but about the world's inevitable uncertainty.  If this is right, then land use casebooks will still emphasize Euclidean zoning, because that's where the disputes are and necessarily will be.

A problem set with form-based codes would be nice, though.  Just sayin'.