Monday, February 21, 2011

Ryan Avent on Density and Skyscrapers

He writes:

This comparison encourages a lot of people (not necessarily Richard Green) to decide that we don’t need skyscrapers. Defenders of the Washington height limit often fall into this category. But there are two points worth making in response to this. One is that it would be harder to build Paris in America than it would be to build Manhattan. Paris’ tiny streets are more hostile to the automobile than anything in the US, including Manhattan. And Paris has relatively tall buildings over a vast area; it’s easier for me to imagine Washingtonians tolerating 30-story buildings downtown than 10-story buildings in a central neighborhood like Brookland. If you need 10-story buildings in every little Brookland-like neighborhood to generate the same density you achieve with 30-story buildings in a central business district, then you can basically forget about generating high densities in American cities. The NIMBYs are just too strong.

The second point is that Glaeser isn’t directing people to go out and build skyscrapers. He’s not a planner. He’s merely saying that, yes, allowing developers to meet demand with supply will often yield tall buildings, and that’s a good thing. It will increase densities relative to the alternative, supply-limited case, and it will improve affordability relative to the alternative, supply-limited case. People who read Glaeser lauding density and who go on to tout the advantages of Paris get his argument precisely backwards. Because density is good, it’s costly — in terms of the metropolitan economy and affordability — to adopt Parisian limits on growth. Unless your city is one of the architectural jewels of the modern world, and if you live in America it isn’t, you should work very hard to avoid such constraints.

I think the comment about Brookland is very much on point.  At the same time, I can't help but wonder if Washington would be a more attractive and affordable city if it allowed a bunch of six floor buildings almost anywhere, instead of skyscrapers in certain districts.  BTW, I think Washington is close to being one of the architectural jewels of the modern world--its vistas are among its best features.

But now let us consider Los Angeles, another expensive city that has relatively few skyscrapers. Because of regulations arising from seismic concerns, it is very expensive to build steel-framed glass-curtainwall buildings here.   At the same time, while land is expensive in LA, it is nowhere near as expensive as Manhattan.  The combination of high construction costs and (lower) land costs leaves me skeptical about whether high-rise buildings could be economically feasible here.

On the other hand, we should use our land more densely and efficiently here.  Lots of four-to-six story buildings would help.  Maybe narrowing some minor street would help--particularly in the San Fernando Valley and Orange County.

As I said in my original post, high rises are almost certainly the right hammer for very large cities, such as Mumbai, Cairo, Mexico City and Sao [Paulo].  Singapore and Hong Kong have proved that cities that rely on the skyscraper can be beautiful and livable (although affordable is actually still an issue in those places).  But not every city is a nail.


ih said...

good post, but i believe you mean sao paulo, and not paolo.

Charlie said...

But Mexico City has very few skyscrapers at all, and only a handful of those are residential. In a city of 21 million they are a drop in the ocean, and are not an affordable option. Nonetheless, the city's real estate is quite cheap for a major metropolis.

Sao Paolo, on the other hand, which is filled with residential high rises, is one of the most expensive cities in Latin America, much pricier than Mexico City (despite Brazil being poorer, per capita, than Mexico.)

Hong Kong and Singapore meanwhile are both in the top 10 for most expensive real estate markets in the world.

There are very expensive low-rise cities too, but the relationship between skyscrapers and affordability does not seem to be that clear cut.