The Data-Driven Home Search
Using Data to Find a New York Suburb That Fits
By LISA PREVOST JULY 18, 2014
… Other real estate websites are supplying home buyers with loads of hyper-specific community data, including racial makeup, percent of married households and education level. Because these sites, if not actually brokerages, are linked to home sales, they have attracted the attention of the Department of Housing and Urban Development, which is charged with enforcing fair housing laws.
Ms. Bernstein said her service abides by the same ethical guidelines as any real estate agency would, but her staff can be more objective. They aren’t necessarily licensed agents because, she said, “having a license really conflicts with what we’re trying to do.” …
Real estate websites — the largest being Zillow and Trulia — also are elevating the importance of “location, location, location” to new heights by offering data that allows home buyers to scrutinize communities or neighborhoods in any way they like.
Want to find a “family-friendly” community within 20 miles of Boston with a high Asian population, a low poverty rate and a median home value of $400,000? On NeighborhoodScout.com, you can plug in these preferences (and many more) on the subscription-only “Advanced Search” page and get a ranked list of options. …
This trend raises some thorny questions. The growing accessibility of highly detailed demographic data plays into the natural tendency of home buyers to look for “people like us,” which is as old as the subdivided hills. Indeed, some suburban communities were created in the late 19th and early 20th centuries specifically with this in mind, some with discriminatory policies written into leases and deeds.
But Bill Bishop, a Texas journalist and the author of “The Big Sort: Why the Clustering of Like-Minded America Is Tearing Us Apart” (Houghton Mifflin, 2008), argues that this tribalism is a major driver of the country’s deepening political polarization. Over the last 30 years, he says, greater mobility, laws enforcing racial equity and prosperity have given Americans even more choice about where to live. Will Internet-enhanced abilities to scout out communities intensify that sorting effect?
And what about the impact on segregation? The National Association of Realtors’ code of ethics prohibits Realtors and associates involved in a sale from volunteering information regarding the racial, religious or ethnic composition of any neighborhood, lest they run afoul of the Fair Housing Act, which prohibits the steering of clients to or away from neighborhoods out of bias. But many nonbrokerage real estate websites that act as referral generators for agents readily offer such information.
“I’ve heard some concerns from Realtors — ‘is that a violation of the Fair Housing Act?’ ” said Fred Underwood, the N.A.R.’s director of diversity and community outreach programs. He said he did not feel qualified to say one way or the other, but given that the housing act was passed in response to racial segregation and discrimination, Mr. Underwood said it was at least worth questioning the purpose of providing such information.
The sites are simply making it easier for home buyers to access data that is already available, much of it through the Census Bureau, says Peter Goldey, the chief information officer and chief knowledge officer for Onboard Informatics, a provider of local content data and lifestyle search products for real estate sites. “We think the consumer has a right to information as long as the information is factual,” he said.
HUD may yet weigh in on the question. The department declined to make someone available for an interview, but in a prepared statement, Gustavo F. Velasquez, the assistant secretary for fair housing and equal opportunity, said: “We are aware of the issue and are reviewing it. It would be premature for us to comment while the review is underway.”
“The problem comes when real estate professionals have product to move in a location that has issues or is not attractive to people because the schools aren’t good, the crime rate is high, or the appreciation rates aren’t moving in the right direction,” Mr. Schiller said. “That’s one of the potential conflicts with the real estate industry.”
He compares his service to a sort of Consumer Reports for communities. (The site doesn’t sell houses, but refers home buyers to agents, who then pay NeighborhoodScout.com a fee if they sell a home.) Subscribers may choose from more than 360 search criteria, using as many or as few as they like. The most detailed category is demographics, which offers choices of population size, family type, education levels and income, and also asks for preferences as to the level of diversity, the percentage of foreign-born residents, languages spoken and specific ethnicities.
If some home buyers use that information to eliminate communities with certain racial or ethnic types — a use Mr. Schiller said he would find “despicable” — so, too, might they use it to find places they hadn’t considered.
, which is owned by RealtyTrac, can type in a specific address, and pull up not just the property details, but also a wealth of community data. In addition to pinpointing nearby sex offenders (whose photos are provided) or indicating areas at high risk for tornadoes and earthquakes, the site provides demographic facts such as dominant religions, political affiliations and races.
Property listings are advertised elsewhere on the site. But Jamie Moyle, the president and chief executive of RealtyTrac, said Homefacts.com is less about listings than about telling every property’s “story.” And when it comes to data, “we put up anything that we think consumers are going to be interested in,” Mr. Moyle said. “If it’s part of the home-buying process, than it’s part of home values. We don’t have an agenda with the data. The data is our agenda.”
In a report last year for brokerages called “Why online consumers love Zillow and Trulia more than you,” Todd Carpenter, then an industry technology consultant and now N.A.R.’s managing director of data analytics, noted that nonbrokerage portals are well aware of the freedoms they enjoy by not being real estate agents. And one of those is “being free of fair housing rules,” he wrote. Trulia’s crime heat maps, for example, might be a problem for a broker navigating fair housing laws, but as third parties, “Zillow and Trulia have more freedom to answer these questions.”
Some full-fledged brokerage sites, like Movoto.com, are pushing the online envelope by putting ZIP code data on race, languages and foreign-born households directly on their sites. But most brokerage sites provide only links to outside sources of information (if at all) for areas that might fall under fair housing rules. This is an acceptable practice under the N.A.R.’s code of ethics, Mr. Underwood said.
Redfin, for example, links to a third-party provider, Onboard Informatics, for demographics, although the Redfin link does not include information about race and religion. Bridget Frey, the company’s vice president of Seattle engineering, said Redfin is more focused on developing proprietary features that give buyers an edge, such as the “Hot Homes” alert of listings most likely to quickly go under contract.
Mr. Goldey of Onboard maintains that real estate websites should be free to post all kinds of demographic data, regardless of whether they are selling directly to buyers, as long as they stick to the facts. The Fair Housing Act, he says, “is only supposed to keep what are potentially nonfactual observations or perspectives by a single person like a real estate agent from influencing a person like a home buyer.”
Legal questions aside, the growing accessibility of so much demographic data has the potential to fuel the segregation that is already increasing along a number of lines — economically, racially, ideologically. Mr. Bishop, the author of “The Big Sort,” argues that as other forms of community have gone away or weakened, Americans are increasingly reordering themselves around shared values and areas of interest. “Given a choice,” he said, “people choose to segregate themselves into these places where they can surround themselves with people like themselves.”
This self-segregating boosts people’s sense of well-being by satisfying the need to belong, says Mr. Motyl, who studies ideological migration. But the resulting decrease in contact with anyone who thinks differently serves to heighten partisanship. “It allows us to become more extreme in our own ideas,” Mr. Motyl said, “and is one explanation for why our system has become so gridlocked.”
This seems like a business — using data to predict gentrification — that has been underinvested in. My wife and I went with a Graham-Buffett “Value Investing” approach to forecasting gentrification in Chicago in 1988, buying a well-made 1923 condo near the beaches, on the grounds that the main very long run feature distinguishing one neighborhood from another in Chicago is proximity to Lake Michigan (and all the amenities that come with it).
But, that wasn’t a fashionable view in Chicago up through 1997, so if we had wanted to sell during our first eight years, we would have barely gotten what we paid for the place. The people who did better in 1988-1997 were the Momentum Investors, the proto-hipsters who flocked together to the the dumpy inland neighborhood of Bucktown (Liz Phair’s “Guyville”). Because my wife and I are pretty contrarian by personality, we missed out on that. As Keynes remarked, markets can sometimes stay irrational longer than you can stay solvent.
Anyway, it seems like some sabermetrician-types should stop playing around with baseball statistics and start churning massive amounts of data to figure out underlying patterns, if any, in gentrification.