Reader Joe Wilson points me to Witold Rybczynski's review of a book on sprawl, wherein Witold offers:
What this iconoclastic little book demonstrates is that sprawl is not the anomalous result of American zoning laws, or mortgage interest tax deduction, or cheap gas, or subsidized highway construction, or cultural antipathy toward cities. Nor is it an aberration. Bruegmann shows that asking whether sprawl is "good" or "bad" is the wrong question.
Witold hints that there is a right question. Yet he neglects to suggest — and I know that he knows — its precise formulation. Of course readers of this blog know the question. And even a bit of the answer. So no point repeating it...
UPDATE: Benjamin Hemric asks and answers nicely, in comments. Here's my own test:
Suburban expansion is ok so long as what you build out there at the urban fringe is something you can walk to and in.
UPDATE 2: The more I consider the test, above, the better I like it.