Wendell Cox's Voodoo Economics
So, yet again Wendell Cox – a leader of the anti-anti-sprawl crowd -- has trotted out an impressive-looking quantitative report that purports to prove that certain metropolitan regions have high home prices because of "more restrictive land use regulation". In his New Geography piece last week, which linked to a report on his web site, Cox seemed to attribute virtually all the variation in home price around the country to land use regulations – just as he has done in the past. But – as usual – Cox’s analysis is based on the assumption that sprawl is the natural state of affairs and any deviation from sprawl must therefore be caused by regulation.
He does extensive quantitative analysis to prove that all difference in home price is due to regulation. But it’s not too surprising that he reaches that conclusion, considering that his analysis assumes that any difference must be due to regulation.
And even Cox himself apparently recognizes that he can’t quite make an airtight connection. As he said in the New Geography piece: "Nearly all of this difference is in costs other than site preparation and construction, which indicates rising land and regulation costs."
Note the language: Indicates, not proves. And that’s not the only defect in his methodology. There are lots more assumptions piled on top of assumptions – some contradictory -- that make the numbers come out the way he wants them to.
To summarize Cox’s latest analysis, he compared home prices in 11 metropolitan areas. Home prices in six metropolitan areas that he categorizes as having "less restrictive land use regulation" (Atlanta, Dallas, Houston, Indianapolis, Raleigh-Durham, and St. Louis) are lower than home prices in five metros he characterizes as having "more restrictive land use regulation" (Minneapolis-St. Paul, Portland, San Diego, Seattle, and Washington, D.C.-Baltimore).
Not too surprising on the face of it, but let’s unpack Cox’s methodology, most of which is contained in a separate document.
First, he asserts that more restrictive regulation raises home prices and disrupts normal market functioning in a variety of ways. There’s obviously a lot of truth in that.
Second, he identifies six specific regulatory approaches that, he claims, are characteristic of "more restrictive land use regulation policies with potential to increase land costs and house prices". These are:
1. Urban containment (urban growth boundaries, urban service districts, restrictions on physically developable land, infill quotas.)
2. Large-lot zoning in urban fringe & rural areas.
3. Geographical growth steering
4. Housing building moratoria or limits
5. High development fees & exactions
6. Mandatory regional or county planning.
Where did he get this list? Well, in some cases he got them from the 2000 HUD report The Cost of Sprawl. In other cases he references himself, a la Mike Davis. But he doesn’t differentiate among these policies; he simply asserts that they all fall into the category of restrictive regulation. Never mind, for example, that he lumps together UGBs (which encourage higher density) and large-lot zoning (which encourage lower density). He also asserts that no matter what the policies are called – for example, smart growth or growth management -- they are basically all the same.
Next he categorizes the 11 metros based on the presence or absence of these six criteria. In the case of the five "more restrictive" metros, he finds the presence between two and four of the criteria. He’s on target in many cases – clearly, Portland has a UGB. But he gives them equal weight, even though his discussion of the six admits that not all of them have the same effect.
And, amusingly, he finds that none of these criteria are present in any of the six "less restrictive" metros. Apparently there’s large-lot zoning in Minneapolis but not in Atlanta or Raleigh or St. Louis? He gives no indication as to how he decided this.
Once he comes up with these categories, he then goes through an exhaustive – and, frankly, mostly well-executed – quantitative analysis about home construction cost, accounting for variation in construction costs in each metropolitan area.
But then he assumes that construction cost is typically 80% of the advertised home price, meaning 20% is attributable to land cost and regulation. He says he has data to prove this but doesn’t provide references; he simply assumes that 20% is what land and regulation cost should be in a less restrictively regulated market. And then he simply assumes that if the difference between home price and construction cost is more than 20%, then all the difference must be due to regulation.
For example, if construction cost is $80,000, the sales price of the house should be $100,000. Under Cox’s methodology, if the cost of the house is more than $100,000, anything over that price is either due to excess regulation or due to high land cost that is caused by excess regulation.
In his 11 metros, he compares the "expected finished land and regulation cost" with the actual difference between construction cost and home price. This is, miraculously, zero in his six "less restrictive" metros, and it’s a lot more in the other metros – ranging from $28,000 per unit in Minneapolis to $221,000 in San Diego.
To give a more detailed comparison, Cox calculates that construction cost of the average home in Atlanta and Washington-Baltimore is about the same -- $128,000. Based on his 80/20 rule, this means the average home price in each metro should be about $160,000 (with $32,000 for land and regulatory cost). The average home price in metro Atlanta is $161,000 – right on target. But the average home price in metro Washington/Baltimore is $235,000. That’s $75,000 more than he thinks it should be – so obviously all of the increase MUST be due to regulation!
The list of other things that could account for this difference is long indeed, but Cox doesn’t even give lip service to any of them. To begin with, there’s demand – and, in particular, the psychology of any given real estate market. Real estate booms and busts are common, as we saw have learned once again here in California in the last few years. Then there’s the income of the people buying the houses; that’s a factor because, in practical terms, home prices depend not only on how much houses cost to build but also on how much you can afford to pay. The more you can afford to pay, the higher the prices will be. (Median incomes in Washington-Baltimore are about 10% higher than median incomes in Atlanta
And where’s developer profit, which will rise in a boom market and drop in a bust market? Cox’s formula doesn’t seem to acknowledge profit at all. Poor guys.
I could go on, but you get the idea.
I respect the anti-anti-sprawl researchers – Cox, Randall O’Toole, Sam Staley – and I try to stay on good terms with them. But what drives me crazy about this stuff is that the self-fulfilling assumptions undermine the valid points and make it difficult to reach consensus about what’s going on.
There’s no question, for example, that UGBs do increase home prices at least a little and also create a "bounce" effect, as my colleague Rolf Pendall and I have acknowledged in a piece of research Cox cites as part of his source material. But UGBs do not, by any means, account for all home price variation.
Instead of acknowledging the complexity of the situation, however, Cox lumps all public policy into the same category of intrusive regulation and then ascribes all variation in home price to that regulation. (O’Toole does the same thing all the time.) I’m no academic snob, but I can’t believe this would pass any serious peer-review muster.
Unfortunately, Cox’s stuff is the Fox News – or MSNBC, if you prefer – of land use research. One point of view always wins out. The possibility that another point of view may have merit is simply never entertained. Instead of moving toward greater understanding about land use policy and how to use it, we are pushed deeper into our own separate world views – and further away from a useful policy debate.
-- Bill Fulton
Bill Fulton's new book on economic development, Romancing The Smokestack, is available here.