In the United States, rising costs have made it hard for many people to find housing. This past summer, the median listed price for rent crossed $2,000 for the first time, according to the real-estate brokerage company Redfin. An affordability crisis has been building for years, and policy makers seeking to address it often focus on subsidies and lending programs targeting lower-income buyers.

But this tactic may neglect a crucial piece of the puzzle: the housing stock. Banks can be reticent to issue mortgages on properties that don’t fit neatly into appraisal models, according to University of Southern California’s Erica Xuewei Jiang and Chicago Booth’s Anthony Zhang. Thus, banks are less likely to issue mortgages for some of the very properties that attract lower-income households.

Jiang and Zhang constructed a statistical model to analyze as many as 62 million housing transactions from 2000 to 2020, taking into account sale values, mortgage amounts, dates, and locations, as well as property, mortgage, and borrower characteristics. They used data from real-estate information provider CoreLogic and the TransUnion Consumer Credit panel from Booth’s Kilts Center for Marketing.

To test price predictability, the researchers estimated the values of properties and compared their estimates with actual prices paid. They find that values were less predictable for houses that were older or differed substantially from the median size and distribution.

For banks, some houses are hard to value

It’s easy enough to put a value on a nice, 1,500 sq.-ft., two- or three-bedroom condo, Zhang says. It’s much harder to estimate the value of a really small house or a celebrity mansion. Quirky or historic properties and those with unusual features or dimensions may well find a buyer willing to pay top dollar, but that could take some time. A modern, standard three-bedroom condo is typically in more consistent demand. When a bank is weighing whether to provide financing, it considers what will happen if it needs to repossess and resell a property, making the issue of future value the central factor.

For lower-income families in the US who typically live in older or more run-down dwellings—precisely the kinds of nonstandard real estate that banks struggle to evaluate—securing a mortgage will be more difficult, the research suggests. Mortgage rejection rates were 10 percent higher in areas where housing values were hard to peg, the study finds.

Until now, policy makers mainly looked at affordability simply in terms of location: houses in San Francisco are more expensive than houses in Detroit, so the thinking goes. But the researchers point out that within a given area there can be significant differences in values assigned to collateral, depending on the quality and characteristics of the housing stock.

They find that in the neighborhoods they looked at with older, unusual buildings, the unpredictability of prices had a clear impact on homeownership. Fewer people bought in such areas, most likely because it was harder to get a mortgage, Jiang and Zhang suggest. Homeownership there was 1.4 percent lower than elsewhere, according to their study.

As price predictability makes it hard for private entities to lend to households with the greatest need for loans, policy makers should take into account the effects of housing policy on collateral values, the researchers argue.

“Maybe a better approach for local governments would be to rethink their housing policy in lower-income areas,” Zhang says. “If the goal is to boost affordability for struggling US families, one idea would be to incentivize developers to invest in modern, standardized, affordable housing schemes that retain a predictable value, such as you see in parts of Europe and Asia.”

More from Chicago Booth Review

More from Chicago Booth

Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.