The cost of housing — of shelter — has been the problem child, one might say, of the Federal Reserve’s effort to tame inflation.

Housing inflation has been “sticky”— we’ve been reporting on it pretty much since 2021. It was up a seasonally adjusted 0.4% in May for the fourth month in a row. And because housing costs represent more than one-third of the consumer price index, they have been a major factor in keeping inflation above the Fed’s target of 2% annually.

Figuring out how much Americans spend on housing is actually a pretty complicated proposition. During his news conference this week, Fed Chair Jay Powell called it “one of the hardest things to think about” when you think about measuring inflation.

About a third of households rent, so it’s relatively easy to survey them to figure out what they’re paying. But most Americans own their homes. That comes with all kinds of costs, little and large.

So the Bureau of Labor Statistics — which calculates the CPI — uses a measure called owners’ equivalent rent, or OER. It accounts for around a quarter of the CPI. Let’s break down how it’s calculated, how it’s useful and how it’s maybe a little limited.

I’m one of those millennials who don’t own the roof over their head, so let’s use some guy named Bob as a hypothetical example. He owns a home, pays a mortgage, property taxes, insurance and upkeep. But the CPI doesn’t care how much that comes out to.

“What the economists of the government try to do is put everything on a level playing field by estimating what you would have to pay if you didn’t own your home and had to pay rent. Hence, owners’ equivalent rent,” said Connel Fullenkamp, an economics professor at Duke University.

In other words, what Bob would get if he rented out his home.

OER is such a big percentage of the CPI because the majority of people in this country own their homes. “It reflects the fact that, you know, housing is one of our main expenses in life,” Fullenkamp said.

But Bob is almost certainly paying less than the OER for his house every month. Because, for one thing, he may have locked in a low mortgage interest rate years ago.

So that OER doesn’t represent his actual cost of living. It’s inflated.

“If rent goes up and you say that his shelter costs are higher because of rent going up, I get why that sort of seems off,” said Steve Reed, an economist with the BLS.

But OER is not actually off, he said. That’s because it measures the market cost of living in the home.

“And so that opportunity cost of living in his house is going up, even if he’s not directly making payments for it,” Reed said.

There are some downsides to that. Bob might be living in a neighborhood that’s gentrified so much, he wouldn’t be able to afford it today.

That’s not the point, said Menzie Chinn, an economics professor at the University of Wisconsin, Madison.

“They’re really just price indices,” he said. “So you can say, well, what’s the price of a hamburger, irrespective of, you know, can you afford the price? Can you afford to buy that hamburger?”

Chinn said OER might be imperfect, but measuring homeowners’ housing costs any other way would put you off-track.

“People can debate about what’s a better approach, but you don’t want to just change your methodology in midstream because then your numbers won’t be comparable to the previous numbers,” he said.

So you won’t be able to measure change. Since that’s the biggest goal, OER is what we’ll keep using.