Decentralization: Disservice to the Poor or More Equitable Urban Arrangment?

Stephen Smith recently analyzed how off-center employment clusters end up being hurtful towards the poor.  It serves as a continuation on an earlier post of his, which explains how European cities have relegated skyscrapers to the periphery in order to preserve historic city cores.  He makes the argument that these off center skyscraper districts are the result of restrictions on central city growth (in Europe's case this is for historical preservation), and follows up by saying that these off center employment districts are not desirable.

This analysis ends up holding true with a few assumptions:  there is an existing monocentric city with a defined downtown, transportation is arranged to get people from the periphery to the center, and there is a relatively small number of off-center employment clusters.  Under these assumptions, it's absolutely correct that off-center employment clusters cause a disservice to the poor.  It will take longer to go from one side of the periphery to another, and the formation of off-center clusters will produce least favored quarters on the opposite side of the city.  Stephen frames this as a problem that is caused by restraints on development in the center of the city, and concludes that if only these constraints were lifted a monocentric city would emerge, and that city would be more beneficial to the poor and therefore more desireable.

I agree that restraining growth in the central city leads to inefficient land use, and that most cases you will look at can be captured in the above example.  However, I believe his conclusion that monocentric cities are desirable overall and the main government problem overall is restriction of central city development is a wrong one.

Stephen touches on the fact that most existing transit systems are built to get users from the periphery to the center, and that it's an interesting thought experiment to consider hypothetical decentralized transit system. I would actually argue this sort of transit arrangement is a major failing of the monocentric city.  I previously wrote about differing urban forms and how equitable they are, and came to the exact opposite conclusion as Stephen: decentralized cities are more equitable.  My assumption is independent of the skeleton you start out from: if you were looking at lots of different theoretical cities evolving over hundreds of years, those that evolved more decentralized would be more geographically equitable.

There is a similar issue when looking at the nature of these employment clusters.  I'd prefer to think of this in more continuous terms, and finding out where on the scale between centralized and completely evenly distributed employment a city is.  Aside from existing conditions, there's no reason why one sector would be preferred over another.  Given the right conditions, instead of having quadrants of the city with different fortunes, there will be no anchor for land prices to get bid up around, and an even distribution of rich and poor neighborhoods.

If you look at the article that spurred my analysis for example, you can see how it might be dangerous to follow Stephen's line of reasoning.  It may seem like a good thing to build up Downtown LA towards Manhattan style densities, but is this where we should be encouraging development?  LA developed mostly with the car in mind, and became really decentralized as a result.  While downtown LA suffered the same fate as the downtowns of older monocentric cities, it might not have been as grave a misinvestment as in other cities.  Because it's a newer city, LA developed it's decentralized "skeleton" that older cities with already established forms did not develop.  I'd argue that trying to centralize LA would lead to problems just as decentralizing older cities would.

Of course, my argument isn't simply that framing matters.  I believe that there if you were ever given the choice between decentralization and centralization, decentralization is more equitable.  While you run into problems trying to force multiple centers on an older centralized city, I think we should move in that direction whenever given the opportunity.  Using New York as an example, we should be building the Triboro RX instead of the Second Avenue Subway.

In my opinion it comes down to a timescale issue.  In the short term, do what's best given the existing urban form.  In the long term, where decisions can change the urban form, try to slowly but surely decentralize that form.  While restriction of development is definitely a problem, the more pressing concern is how hard it is to invest towards these long term goals.