From the start of 2021 to the end of 2023, owner-occupied house prices grew 17 percent more than rents and interest rates on new 30-year fixed-rate mortgages rose from 2.7 percent to 6.6 percent.

This surge in both house prices and borrowing costs has drawn attention from policymakers concerned about housing affordability, and defied analyst expectations – for example, a 2022 Moody’s Analytics forecast predicted that higher interest rates would reduce national house prices by 10 percent due to lower homebuyer demand.

Why did a house price boom coincide with rapid financial tightening? In a recent working paper, Robert Minton and I study a possible contributing factor: rising rates reduce the supply of existing homes due to mortgage “rate lock,” where owners with low-interest, fixed-rate mortgages have a financial incentive to keep their homes and their valuable low rates, rather than selling.

The result, Federal Reserve Chair Jerome Powell argued in 2023, is the “supply of existing homes is really tight [and] keeping prices up.”

Not an Obvious Connection

We start by observing that rate lock does not obviously impact house prices.

By reducing the number of home sellers, it also reduces the number of buyers because there are fewer current owners shopping for new homes. If potential sellers would have bought another home in the same housing market, then supply and demand fall equally and prices remain unchanged.

However, if sellers would have bought in a different market, rate lock affects relative demand and hence relative price.

Aggregating across all homeowners, if rate lock reduces moves by owners who would have otherwise exited to the rental market, the overall supply of owner-occupied homes declines, increasing house prices relative to rents.

Quantifying Impact of Rate Lock-in

Using household-level data from the Federal Reserve system, we quantify the impact of rate lock on house prices, home sales and residential investment.

We first show that the rate lock incentive increases local price growth by reducing exit from the local housing market. This analysis faces a challenge: The rate lock incentive is high in markets where more mortgage originations occur in low-rate periods, but such places have specific characteristics that likely relate to price growth.

We overcome this challenge using a series of natural experiments. Our main approach uses unexpected moves due to homeowners having children that induce moves at times when mortgage rates happen to be high or low. An especially clean approach uses year-to-year changes in the number of local twin births, which we estimate creates changes in the number of local moves and generates quasi-random differences in timing of moves across markets.

Our analyses imply that rate lock significantly increases prices: a 1 percentage-point decrease in the average outstanding mortgage rate in 2021 increased nominal house price growth by 8 percentage points between 2021 and 2023. Moreover, rate lock affects prices much more in markets where the supply of new housing is constrained.

Rate Lock Explains 40 Percent of Price Jumps

We next quantify how much rate lock offsets the negative price effects of reduced housing demand when interest rates rise.

We show that rate lock increases aggregate house prices, relative to rents, by reducing moves from owning to renting. Rate lock explains 40 percent of the gap between the predicted decrease in prices and the observed price growth between 2021 and 2023.

Finally, we analyze the price impact of rising interest rates under two recent mortgage contract proposals: portable mortgages that borrowers can keep when they move, and mortgages with a buyback option, where borrowers can prepay remaining mortgage payments at the present-value cost given current market rates.

These reduce frictions in moving between owner-occupied housing units, but still incentivize owning, rather than renting, when rates rise. Tightening thus has similar aggregate price effects as with existing fixed-rate mortgages.

What Policymakers Need to Know

Our research has two policy implications.

First, for central bankers and regulators monitoring house prices for signs of credit market exuberance, our findings show that rate lock can help explain elevated house prices relative to traditional indicators like rents and interest rates.

The price impacts of rate lock also imply that interest rates have a path-dependent impact on house prices: After a long period of low rates, most outstanding mortgages have low rates, and prices fall by less when rates subsequently rise.

Finally, we show that the home sales decisions of existing owners only impact house prices in places where supply is constrained. This emphasizes that policies to improve housing affordability should focus on expanding housing supply.

By contrast, policy interventions that affect home sales choices by changing mortgage structure, such as portability, would have little impact on aggregate house prices.

Justin Katz is a PhD student in business economics at Harvard, and will join the Boston College Carroll School of Management as assistant professor of finance in July 2026. He was a 2024 Meyer fellow at the Harvard University Joint Center for Housing Studies. His column is being reprinted with permission from the center’s Housing Perspectives blog.