We now have scientific justification for McMansion regulations: Infill McMansions: Style and the psychophysics of size, by Jack Nasar, a professor of planning at Ohio State, and Arthur Stamps, a something at the Institute of Environmental Quality in San Francisco. The abstract (I couldn't find an ungated version):
The number and size of infill oversized houses, or McMansions, has increased in the U.S. and the world. To maintain desirable neighborhood appearances, communities should know what constitutes “too big.” This paper reports six studies that used color simulations of blocks of houses to examine compatibility and visual appeal in relation to attributes of the infill house and its context. Following psychophysical findings, the relative size of the infill house should have larger effects on response than its actual size. The studies confirmed that the infill ratio had more weight than size for perceived compatibility. Compatibility did not always translate into visual appeal. For visual appeal, infill style had larger effects than did size or size ratio, height had larger effects than width, and, for larger sizes and ratios, infill ratio had larger effects than actual size. To maintain visual quality in relation to infill houses, communities should first seek an architecturally compatible style, and then try to control its relative size (the infill façade area). Tests of responses to real infill houses can show how well these findings generalize.
Translation: Relative size matters more than absolute size; architectual style matters more than either relative size or absolute size; and height matters more than width. (Based on these "findings," Austin's McMansion ordinance regulates the wrong things.)
According to this account of their research, the scientists' methodology consisted of showing computer-generated simulations to college students and asking them which they liked and which they disliked. (See the above example.)
I question whether we should base our land-use regulations on a bunch of college students' gut reactions to a photo line-up. But what really bothered me was Nasar and Stamps' casual assumption that there are no trade-offs to regulating architectural design and relative size.
Groups tend to be deeply conservative — or reactionary — simply because they are groups. They tend strongly to favor current fashions in architecture and planning; it is the only way they can reach consensus, which, by their nature, they must. In the process, they prohibit the revolting and the revolutionary (it’s often hard to know which is which without the benefit of a time machine). And so, in my neighborhood (recently covered by a historic preservation ordinance), they routinely approve lavish “Tuscan” villas and faux Tudor mansions, while turning their noses up at “modern” things — some of which might be awful and some quite extraordinary; time will tell.
By now, you get the point, or at least the question: How much discretion do we really want to vest in groups, which are likely to prohibit both the sublime and the ridiculous? In Los Angeles, I think the answer ought to be: “Just a little.” There are good, quasi-objective economic reasons to prohibit a landowner in a residential neighborhood from building a ten-story office building. Requiring people to trim their trees, and to build homes that will withstand earthquakes, makes sense. But in matters aesthetic . . . well, I’ve chosen to live in what’s left of the Wild West, as opposed to Haussmann’s Paris or Celebration, Florida. I hate strip shopping centers and their signage as much as the next guy. But if they are the price I must pay for the next Hollywood sign, or for the bizarrely beautiful little addition I built onto my house (which my neighbors undoubtedly would have forbidden if they had had the power to do so) — I can live with them. Criticize. Debate. Shame people, where appropriate. But let’s pause before continuing down the road of making design decisions by committee (even by a well-meaning and educated committee). Because we all know how well that works.