Tuesday, 10 April 2012

Upwardly Mobile: Are you living in the wrong city?

I'm not a big fan of city-rankings. Part of the problem is the the most simple rankings usually only consider one or two aspects of cities, which don't capture the complex trade-off faced by households and firms. In contrast, once you consider a lot of factors, you need to somehow collapse these down to get one index. Then the problem is what weight should you give to each of the factors?

A piece in Atlantic City reports on a new app that would appear to get round these problems: "Upwardly Mobile, a new tool from Sunlight Foundation, can take your career information and your spending priorities and figure out where it makes the most sense to be a library archivist with children in daycare and cars to gas up."

In other words, you provide your personal weightings and the app then 'tells' you were to live. While that sounds great, there's still one major problem - in practice, would people have any idea what weight they should impose on the different components? Urban economists would tend to figure out the weights that people apply in practice by looking at the decisions they make about where to live. That is, they would tend to focus on 'revealed' preferences (revealed because economists are backing them out from actual decisions) rather than the stated preference (what people say they pay attention to). The economist approach may break down, however, if people don't have good information about other places on which to base their decisions.

That suggests that the raw information underlying this new app might be far more useful than the computer generated ranking. Of course, that's a lot of information to digest. This is why most people would use broad brush information to narrow down the possibilities and then look in detail at the various trade-offs they would face. Economists assume that this process - using price signals, information on amenities and personal intuition - is likely to get the trade-off roughly right. People often argue that economists assume to much but, at least when it comes to where to live, I'd still trust my personal decision making 'algorithm' over any computer.

No comments: