How to understand shelter inflation in CPI
Shelter is the largest single component of US CPI at ~36%, yet it is measured using a survey methodology that lags actual market rents by 12–18 months. Understanding this lag is essential for forecasting CPI — and for understanding why headline inflation stayed so high in 2022–24 even after market rents had already peaked.
Shelter costs — primarily rent and owners' equivalent rent (OER) — represent approximately 36% of the US CPI-U basket, making shelter the single most important component for understanding US inflation. In the UK, housing costs (including private rents) make up around 17% of CPIH. In the euro area, HICP excludes owner-occupied housing entirely, making cross-country comparison complex.
Shelter inflation has a unique property that distinguishes it from almost every other CPI component: it is measured as an average of all current rents — both long-term leases and new leases — rather than just new-market transactions. This creates a structural lag between what new tenants are paying and what the CPI reports. Understanding this lag is the key analytical challenge in shelter inflation.
Understand OER: the dominant US shelter component
Owners' equivalent rent (OER) is the hypothetical rent homeowners would pay to rent their own home. It is a survey-based estimate, not a market transaction. OER makes up approximately 26% of CPI-U — a quarter of the entire index. Why? Because US CPI is designed to measure the cost of living for all urban consumers, including the 65% who own their homes. Using house prices directly would make CPI much more volatile and would reflect asset price changes rather than consumption costs. The BLS methodology paper explains the survey design in detail.
OER is measured by asking homeowners what they think their home would rent for. The answer is based on their perception of the rental market — which in turn is anchored by the average of the stock of existing leases, not just new market rents. This is why OER lags market rents.
Learn the "flow vs stock" distinction
The critical concept is the difference between new-lease rents ("flow") and the average rent paid by all tenants ("stock"). Market rent indices — Zillow ZORI, Apartment List, ApartmentData.com — measure flow: the rent on leases being signed today. CPI shelter measures stock: the average rent paid by all renters, including those on leases signed 1–3 years ago.
When market rents surge (as in 2021–22), flow rents rise immediately, but stock rents catch up slowly as older leases renew. This is why CPI shelter was still rising at 8% in early 2023, even though Zillow ZORI had already peaked and was falling. Conversely, when new-lease rents fall (as in 2023–24), CPI shelter continues to rise while the old-lease "stock" catches up to the new lower level. The lag is mechanically 12–18 months on average, but varies by how many leases are on 12-month vs 24-month terms.
Use market rent indices as leading indicators
Three market rent indices are freely available and update monthly: Zillow Observed Rent Index (ZORI), Apartment List National Rent Report, and ApartmentData.com. These indices lead CPI shelter by approximately 12 months, meaning today's ZORI reading predicts where CPI shelter will be in about a year. As of October 2026, all three show new-lease rent growth at approximately 2–2.5%, down sharply from their 2021–22 peak of 15–16%. If they remain in this range, CPI shelter should fall to ~2.5–3% by mid-2027.
For a practical forecast: if ZORI is at 2.1% today and CPI shelter is at 4.8%, the path of disinflation is clear — not the direction, but the timing. The BLS has published research confirming this leading-indicator relationship, including a widely read paper by economists Bolhuis, Cramer, and Summers on the shelter lag. See our US shelter inflation analysis for the current data.
Compare UK and euro area treatment of housing costs
The UK's main inflation target measure (CPI) treats housing differently from the US. UK CPI includes actual rents paid by tenants (in its "private rental" component) but excludes owner-occupied housing. CPIH adds back OOH costs using a rental equivalence approach — similar conceptually to OER, but less survey-intensive. Private rents in UK CPI make up approximately 8% of the basket; OOH in CPIH adds roughly 17%. This means UK CPIH is more complete than CPI for measuring housing affordability, but the BoE still formally targets CPI.
Euro area HICP excludes owner-occupied housing entirely — a notable methodology gap compared to the US and UK. The ECB has proposed adding OOH costs to HICP (as "HICP+OOH") but this is not yet in the target measure. The practical implication: euro area HICP tends to run lower than US CPI in periods of rising house prices, partly because the housing cost channel is not fully captured. The HICP glossary entry covers this in detail.
Watch rental vacancy rates as the cyclical indicator
Rental vacancy rates are the best indicator of near-term rental inflation pressure. The US Census Bureau publishes quarterly rental vacancy rates. A rate above 7% is associated with weak rent growth (landlords compete for tenants). A rate below 5% is associated with rapid rent growth (tenants compete for apartments). In October 2026, the US vacancy rate stands at approximately 6.2% — healthy, and consistent with cooling market rents. A sustained fall toward 5% would be a warning signal that shelter disinflation may stall.
Supply is the other side of the vacancy equation. New housing completions feed directly into rental supply — a surge in apartment completions (as occurred in 2024–25 in many US Sun Belt cities) can drive vacancy rates up and rents down even with strong demand. Watch the Census Bureau's New Residential Construction data monthly for the supply pipeline. Combine with vacancy rates to get the full housing market picture.