Unlike Randal O'Toole, I was delighted by Tyler Cowen's New York Times article on the high cost of free parking. And indeed, if I'm reading O'Toole's post right, it sounds like Cowen and O'Toole don't actually disagree on the policy issue: both agree that business owners should be free to decide how much parking to supply.

The debate so far has focused on whether parking mandates push the price of parking below the market rate. But I think the more important effect is on the geography of cities. Parking mandates (and other regulations) preclude developers from catering to people who want to live in pedestrian-friendly neighborhoods.

Parking mandates necessarily mean that every large building is surrounded by a large parking lot. And for someone who doesn't own a car, a parking lot is just a nuisance: a big, empty space he must walk across to get anywhere. Regulations that effectively require a parking lot around every store and restaurant almost guarantees that walking to them won't be practical.

As Jane Jacobs pointed out, pedestrian traffic is highly sensitive to density. Even a modest reduction in the density of a neighborhood can have a big effect on pedestrian traffic. And as the number of pedestrians falls, so too will the number of businesses that cater to pedestrians. So it's probably true, as O'Toole says, that charging for parking spaces wouldn't dramatically reduce the amount of driving people do. But this is partly because the proliferation of parking lots has made walking impractical. Fewer free parking spaces wouldn't just raise the price of car ownership; it would make car non-ownership more pleasant and convenient.

Government regulations often have subtle unintended consequences. Parking regulations have been on the books so long that the results have come to seem perfectly natural to us. But free markets are unpredictable. If developers had the freedom to decide how much parking to supply, the results might surprise us.