If you've looked for an apartment in San Francisco or Boston recently, the lack of affordable places to rent is obvious. But when the government tries to track trends in rental prices, the data it uses tends to be both dated and inaccurate. A new study demonstrates how planners can use a better source—Craigslist.
The study pulled data from 11 million Craigslist ads across the U.S., cleaned up multiple listings and obvious spam, and analyzed the results. Though the study looked at ads from May and June 2014, the same process could be used to look at trends in real time.
Rents are higher in some cities now, but even then, typical tenants ended up spending more than 30% of their household income in New York City, San Francisco, Los Angeles, Miami, Boston, and San Diego. The map shows the highest rents per square foot clustered on the coasts, but also in places like North Dakota, where the oil boom drove up prices.
In some cities, only a tiny percentage of listings fell under what the U.S. Department of Housing and Urban Development considers "fair market rents"—the value tied to many housing voucher programs. In theory, the fair market rent should be the 40th percentile of rent for typical apartments in an area, meaning 40% of apartments should be cheaper. In New York and Boston, listings cheaper than the fair market rent were in the single-digit percentages.
Right now, when city planners or housing advocates study rent prices, they often look at census data, which can give snapshots of rents across a metropolitan area once a year. For specific census tracts, rents from the last five years are rolled together, so it's hard to see current trends. In the East Bay Area, for example, where rents are surging upward, HUD recently lowered the fair market rent—because it was looking at a five-year average, not the most current data.
Planners also often look at commercial databases, which don't include smaller landlords. "They'll leave out a lot of the informal market—so somebody who owns a four-unit building in downtown Berkeley, or somebody who has a little cottage behind their house that they're renting out in their backyard," says Geoff Boeing, one of the co-authors of the study and a PhD student at the University of California-Berkeley. "Those really aren't tracked very well right now."
Craigslist, on the other hand, can give cities a better picture of reality. "It's very fine-grained data, at a very fine-grained spatial scale and time scale," he says. "The data's really fresh."
Planners could easily start using archived Craigslist posts, using sites like Internet Archive, to get a much more accurate sense of trends in any city at any time. It's the type of tool that Boeing hopes cities start to use more often.
"I'm very much all about bringing technology into city planning," he says, though he points out that it has to be used knowledgeably. "It becomes very easy to throw around terms like 'big data' as a buzzword, or just act like you can throw data at any problem and find an answer to it, and that's false—it takes a lot of nuance and a lot of complexity."
Have something to say about this article? You can email us and let us know. If it's interesting and thoughtful, we may publish your response.