Owners' Equivalent Rent in CPI

The Bureau of Labor Statistics doesn’t directly observe what homeowners “pay” for housing—so it estimates owners’ equivalent rent, the imputed rental value of owner-occupied homes. This single component often accounts for one-third of all CPI inflation, making it the most consequential number the BLS produces.

Why the BLS had to invent a number

The Consumer Price Index is supposed to measure the cost of living — what it costs to maintain the same standard of living month to month. For renters, this is simple: track the rent they actually pay. For the roughly 65% of Americans who own homes, it is not. Homeowners don’t pay rent.

The homeowner’s housing cost includes a mortgage payment, property taxes, insurance, utilities, and maintenance. But how much of the mortgage payment is “consumption” versus equity building? How do you compare a paid-off house to a mortgaged one? The BLS decided to sidestep these messy questions and instead estimate what homeowners would pay if they rented out their own homes — owners’ equivalent rent (OER).

This imputation is not a simplification; it is a massive bet on methodology. OER is so large and so politically sensitive that it has shaped Fed decisions, moved markets, and become the flashpoint in inflation debates.

How the BLS estimates OER

The Bureau of Labor Statistics uses a hedonic regression model. Here’s the rough process:

1. Survey renters. The BLS collects actual rent paid for a sample of rental units across the country each month.
2. Control for characteristics. It notes the size, age, location, amenities, and condition of each rental unit.
3. Estimate implicit rent for owners. For each owner-occupied home in the sample, the BLS finds comparable rental units and predicts what the home would rent for, given its characteristics.
4. Compare month to month. The predicted rent for an owner’s home in month one is compared to month two; the percentage change is the monthly OER inflation.

The model sounds rigorous in theory. In practice, it has blind spots. A homeowner in a gentrifying neighborhood sees their home’s market value soar 30%, but if comparable rental comps are slow to appear (because few owners in the neighborhood are renting), the model may show only 3% OER inflation. The lag can persist for years.

The shelter inflation boom: when OER lagged reality

The 2021–2023 inflation cycle exposed OER’s weakness. Home prices surged 20%+ in many markets, reflecting both pandemic demand and supply constraints. Rent also rose, but more slowly — in percentage terms, 5–8% annually while home prices were accelerating at 15–20%.

CPI shelter inflation, which is roughly 40% OER and 60% actual rents paid, showed that accumulated lagged rent growth. Owner-equivalent rent specifically was reported at 5–8% annually, trailing home price appreciation by miles. Yet a homeowner buying in that market faced a dramatic increase in effective housing costs (the price of the home, the size of the mortgage required).

This created a messaging problem. The BLS said housing inflation was 6%. Home buyers and renters said it was much higher. Both were right within their frame:

• The BLS measure captured the slow-moving hedonic adjustment of imputed rent.
• The market measure — actual prices paid — showed the true cost to enter the housing market today versus a year ago.

Why the lag matters for inflation policy

The Federal Reserve uses CPI to gauge inflation, and CPI is roughly 25% shelter via OER. When OER inflation is understated due to measurement lag, the entire inflation reading is understated. In 2022–2023, many observers argued that actual inflation in the “things people care about most” (housing, food, energy) was higher than the Fed’s policy rate. Part of that argument rested on the OER lag.

Conversely, OER eventually catches up. When home price growth slows or rents flatten, OER inflation decelerates sharply. In late 2023 and into 2024, OER inflation was the last shelter component to cool, but it did eventually reset. The lag is real, but it is typically temporary — measured in quarters to a year or two, not decades.

The debate: should OER even exist?

Some economists argue the BLS should abandon OER entirely and instead use actual home prices or mortgage payments. The logic:

• Home prices are more current. Market transactions reflect real cost immediately.
• Mortgage payments matter to households. What a homeowner actually pays in debt service is more relevant than an imputed rent.
• Hedonic models are fragile. They depend on comparable rental data that may not exist in owner-heavy neighborhoods.

Others defend OER:

• Mortgage payments conflate debt and consumption. A 30-year mortgage front-loads interest; the payment is not purely consumption cost.
• Home prices are asset values. They reflect expectations and liquidity, not the flow cost of housing services.
• OER captures real imputed cost. A homeowner does enjoy housing services; imputing a value for them is conceptually sound, even if the method is slow.

The Fed, in its official statements, has acknowledged OER’s measurement challenges while stopping short of recommending a rehaul. The index is used for policy because it’s what exists; the Fed also tracks other measures (like new home sales data and mortgage rates) to triangulate true shelter cost.

The forward-looking question: is OER broken?

OER inflation has historically moved in line with long-run rent inflation. But the pandemic-era divergence — home prices soaring while OER inched up — raised questions about whether the model was fit for purpose.

In response, the BLS has made micro-adjustments to OER methodology, attempting to weight owner-occupied comparables faster when new comps appear. But the fundamental constraint remains: OER is a regression model that lags because good rental comps can take time to materialize.

If the housing market stabilizes and rental growth settles, OER will eventually track more smoothly. If the next shock brings another asset price boom with thin rental data, the lag will recur.

Why this matters for savers and borrowers

OER’s motion through the CPI has real consequences. High shelter inflation (including OER) can keep headline inflation elevated even as goods prices cool. This delays the Fed’s rate cuts, which affects mortgage rates, bond yields, and savings deposit rates.

For a young renter watching interest rates stay high because of OER inflation they don’t directly experience, this feels unfair. For a homeowner whose housing cost surged 20% but OER only shows 5%, it feels like the BLS is lying.

Neither is quite right. OER is a legitimate imputation with a documented lag. The lag is frustrating, but it is not proof of intent. Understanding what OER is and how it works — rather than dismissing it as a statistical game — is the basis for informed debate about housing inflation measurement.

See also