In this post (a response to one of my posts), M1EK writes:
Sprawl isn't the natural result of free-market processes; it's what the market gets forced into providing when regulations require fairly large minimum lot sizes and a ton of parking and subsidize single occupant automobile travel over other modes. Otherwise, we would have seen a lot more modern-style sprawl before the advent of zoning codes, parking minimums, lot size requirements, and government-subsidized freeways - all of which occurred long after most households had access to at least one automobile.
Although I rant about zoning as much as anyone, I actually disagree with this statement. I think the free market would still produce some sprawl even without zoning or subsidized roads. A city's land-use regulations cause sprawl only when the city has pricing power; i.e., the ability to use regulation to raise housing prices. That's true in San Francisco, Boston, to a lesser extent Austin. But it's not the case in a lot of places with sprawl.
Zoning and sprawl
By "zoning" I mean all restrictive land-use, development and building code regulations: lot size minimums, height limitations, FAR limits, minimum parking requirements, electrical codes, exaction fees, permitting requirements, etc.
Zoning looks like an obvious culprit for sprawl. Sprawl means large lots, lots of parking, and low-slung buildings, all the stuff that zoning mandates.
This correlation does not necessarily imply causation, however. Just because regulations are on the book does not mean they are "binding"; they might require people to do things they were already going to do. For example, federal regulations require that all cars sold in the United States be equipped with a seat belt for each seating position. Those regulations could be repealed tomorrow and nothing would change; automakers would continue to supply cars with seat belts.
Zoning regulations could, in some cases, merely dictate the same land use patterns we would see in a free market system. A good way to tell whether zoning regulations are binding is to see whether they're screwing with the market prices. (See the Glaeser & Gyorko papers in the sidebar.)
Let's suppose that in a given city, the free market would produce denser development if the city did not have zoning. This is just another way of saying that the market would supply more housing units than the city's zoning code permits. The zoning code thus creates a shortage of housing (relative to the market). Housing prices must rise to compensate. Binding zoning imposes a price premium.
But houses, in general, are pretty cheap in most places, even though virtually every city has strict zoning regulations on the books. In fact, Glaeser and Gyorko have shown that in most parts of the country, homes are sold at roughly the cost of construction plus a small fraction for the lot. The market cannot supply housing any more cheaply. Whatever zoning's effect in these jurisdictions, it is not to curtail the supply of housing. This means no effect on density.
I don't think it's hard to believe that the free market would still produce sprawl. All else being equal, on average people prefer larger lots to smaller lots. (Individual demand will vary, of course.) If a 20,000 square foot suburban lot costs the same as a 10,000 square foot city lot, then the average city dweller "pays" a 10,000 square-foot premium to live in the city. As average city lot sizes shrink (an inevitable consequence of densification), the "price" of city living goes up. More new buyers will choose the suburban lots.
All of this is just a complicated way of saying that some people prefer suburban living to city living.
When will zoning cause sprawl? When the suburbs are not a perfect substitute for the city. This will happen only when there is a particular reason to live in the city, such as when the city offers special amenities that can't be duplicated in the suburbs. Central Austin is like this. Zilker Park, Town Lake, downtown's bars and restaurants, and UT all make central Austin a unique draw. Austin's zoning regulations artificially restrict the supply of housing in central Austin (and add to sprawl), essentially transferring wealth to incumbent homeowners.
In most of middle America, there's no particular reason to live in the central city rather than the suburbs. Jackson, Mississippi has sprawl, and Jackson has typical restrictive zoning. Jackson wouldn't be denser if it repealed its zoning regulations, however. There is plenty of housing for people who want to live within the city limits. The 200,000 or so who live in Jackson's sprawling suburbs just don't want to live in Jackson.
Road subsidies and sprawl
I'm also skeptical of the road-subsidy argument. Sure, expanding highways to accommodate suburban traffic is expensive, as is maintaining the highways with the extra traffic. And these highways have historically been provided free to the suburbanites (well, not free, but suburbanites' taxes cover only a small fraction of the cost).
But it's a leap from "Suburbanites' highways are subsidized" to "Subsidies are a significant cause of sprawl."
I'm skeptical that the subsidy is worth that much. Suppose we convert a free highway from Suburb to City to a toll road and charge enough to cover the construction, operation and maintenance costs. Let's say a round trip toll is $5/day, $25 per week, $1,250 per year. Capitalized, that $1,250 per year is worth around $16,000, assuming a reasonable capitalization rate. When the toll is adopted, every home in Suburb will immediately drop a maximum of $16,000 (less, if the toll reduces congestion.)
If I'm a new home builder, I will still build new homes in Suburb as long as I can get cost plus $16,000. If City has no zoning, then people won't substitute to Suburb as quickly. I might have to wait a little longer to start building new homes. But $16,000 is a relatively small premium. City won't have to become much denser before people start substituting to Suburb because of its bigger lots. Suburb will begin to grow. We'll still have sprawl.
The only way to settle things is to look at the data (or have someone who knows what he's doing look at the data). How do home prices in suburbs linked by toll roads compare with home prices in suburbs linked by free roads? If free roads are a significant cause of sprawl, the "toll road" homes should not only be cheaper than the "free road" homes, they should sell for a large enough discount to discourage new home construction. I haven't seen any research one way or the other on this. I'd probably have to see something alone these lines to believe that road subsidies are a significant cause of sprawl.
