BeyondDC provides his thoughts on my thought experiment. His response is given, as others were, as though this were a policy proposal rather than a thought experiment, but he nonetheless gives me the opportunity to make a few useful comments. The first thing I’d like to address is this:

Other buildings matter. Cities without height limits (or with limited ones) often see a handful of very tall buildings surrounded by empty or dramatically underused lots. Any model to allow taller buildings in Washington should be structured in a manner that reduces the likelihood of this happening. This would be possible (but cumbersome) to do under Ryan’s proposed model by adjusting the extra floor tax rate based on the height of buildings on all properties with a certain radius, say 3 blocks. For example, the tax rate to build extra floors would be much higher if there’s a parking lot across the street than if there’s a bunch of 12 story buildings, because we really don’t want that empty lot to stay empty.

This is an intuitive way of thinking about the issue, but it’s also wrong. Economists have a name for a similar error of logic when it’s made in the context of labor market discussions — the lump of labor fallacy. It’s very tempting to think that at any given time an economy has a set number of jobs that need filling, such that a good way to increase employment within an underemployed group is to reduce employment among a fully employed group. So you might try to boost youth employment, for example, by encouraging early retirement or cutting the length of the workweek. But this tends not to work. Why? Well, those older workers were earning an income in return for their productive labor, and then using that income to buy other goods and services, in the process creating additional employment opportunities. If you reduce labor supply, you also end up reducing labor demand.

In the passage above, Dan is doing something similar; we could call it the lump of buildings fallacy. He assumes that there’s only so much economic activity to be housed at any given time, and so more occupied space on one lot necessitates less on another — potentially reducing the value of nearby lots so much that surface parking is the most attractive use for them.

But you can see the problem. If you limit building supply, then you limit the number of firms that can locate in an area. But firms don’t work in isolation. They create demand for other businesses (accountants, consultants, lawyers, ad firms, or design firms, for instance). Additional supply, by making it affordable for more businesses to operate in an area, attracts additional ancillary uses, which will occupy still more space.

Now, if there is falling demand for an area in the first place — if some structural trend, like suburbanization associated with rapid increases in car ownership, or like the collapse of Midwestern industry, is undermining the very purpose behind an economic node — then adding a lot of new supply might accelerate decline. When rents are falling and vacancy rates are rising, local leaders often think that a bold new project may turn an area around. But this is mistaken; it will add to supply and accelerate price declines, potentially pushing rents below building maintenance costs and generating pressure to tear down existing structures.

But this is precisely the opposite of what we observe in the economically dynamic, skill-driven cities in which supply limits are so enormously costly. And in these cities, it is quite rare to see buildings next to surface parking lots (and getting rarer). If anything, construction of a few new buildings in places with lots of surface parking accelerates the pace at which the rest of the land is developed. Economic activity attracts economic activity, and limits to economic activity drive economic activity away.

A second point:

Land use matters. Ryan doesn’t care if new construction is residential or commercial, but I do. I want to incentivize different land uses in different areas, so I want a model that takes land use into account. There should be options to apply a different tax rate to different land uses.

I explicitly left land use out of the initial experiment for a couple of reasons. One is that the experiment relates to new construction. You can say that a new building is this or that, but if the rest of the built space in the area is somewhat fungible, and it is, then it ends up not mattering very much; other uses may shift. You can say, “I want x new, tall building to be residential.” Well, ok. Unless you also assume all sorts of things about the rest of the space in the neighborhood, you can’t say anything about whether that changes the total share of residential in the area.

But the second and more important reason was that the experiment was an attempt to gauge the value that folks like Dan place on the height limit. It didn’t occur to me that Washingtonians might object to their views being blocked by some things but not by others. But apparently this is the case.

What Dan is telling us in his post is that he values the height limit a great deal to the extent that it encourages a diversity of uses within a neighborhood. Can you see the trouble with this? That’s right — the height limit is a powerful force for the reduction of diversity of use within a neighborhood. Given strict limits on supply, existing space will tend to be cater to the high-value use. In Washington, that’s high-end office space, and height limits in Washington act to crowd out lower margin commercial uses (which include a lot of the independent and start-up businesses that both invigorate an economy and please urbanists), as well as residential uses in places where housing isn’t explicitly zoned in. Obviously, the city government could increase diversity by changing zoning from high-end commercial to residential, but that’s generally an unattractive proposition, given the importance of the commercial cluster to the city’s economy and tax base.

In other words, if Dan wants a diverse District, he’s favoring exactly the wrong proposal. The best way to accommodate lots of stuff in an area is to let folks build. Then, they can satisfy demand for high-end commercial in tall buildings until the margin on that kind of space has been squeezed and there’s plenty of profit opportunity in and demand for buildings that cater to other uses (which may then rise on nearby land, ensuring that none lies fallow, as argued earlier).

Now, accommodating all this demand will mean taller buildings. But Dan has already demonstrated that he doesn’t mind height if it helps achieve his goal of an active, diverse neighborhood. It seems, then, that we’re finally in agreement on the need for an end to this costly policy.