Vedanta Dhamija, Ricardo Nunes and Roshni Tara
Inflation has been widely discussed in recent years, from supermarket aisles to newspapers. But what if what people think inflation is stems not only from grocery prices or energy bills, but from more? Our analysis in Dhamija et al (2026) shows house prices matter in this context, ie housing is salient. Using household surveys for the United States, we find that people tend to overweight their expectations about house prices when thinking about inflation with a coefficient of 25%–45%, significantly above the weight of house prices in the inflation index. Should central banks care about this? The short answer is yes.
Why expectations matter and why might house prices sneak into thinking about inflation?
Inflation expectations matter because they shape economic behaviour. When households expect prices to rise, they adjust their spending and saving decisions, as well as wage demands in ways that feed back into inflation itself. For this reason, central banks closely monitor measures of inflation expectations, and it has become increasingly important to understand how these are formed.
Several factors influence how households form their inflation expectations; this includes their prior beliefs, exposure to media, knowledge of monetary policy, cognitive abilities, and shopping experiences, among others (Coibion et al (2020)). However, it is not just frequently observed price changes, but also the less frequent, larger price changes that seem to matter. One such price is housing, irrespective of whether one is a homeowner or not.
House prices are widely reported, frequently discussed, and central to households’ financial well-being. Houses are typically the largest asset owned by a household and are associated with significant wealth and collateral effects. Housing is the largest expense for renters and homeowners alike. Changes in house prices are also highly salient as they often attract media attention and shape public debate about affordability and living standards. In the US, a large majority of the population are homeowners, and there is high geographic mobility, suggesting that house prices are closely monitored.
House prices are not directly included in headline inflation measures.
The consumer prices index (CPI) only reports the consumption part of housing services relevant to the cost of living index. In the US, housing services are captured through the CPI component Shelter, which accounts for approximately one-third of the index. The subcomponent of this attributed to homeowners is Owners’ Equivalent Rent (OER). To compute this, the Bureau of Labour Statistics surveys the rents in a region and weighs it by the proportion of homeowners. This is considered best practice and correctly reflects that the OER must represent the opportunity cost of rents at market value or the rent that homeowners implicitly pay to themselves to live in their home, not the asset-portfolio aspect of housing.
Chart 1: House price growth and CPI shelter inflation
Notes: This chart shows CPI shelter inflation and two sub-components: CPI-rent and CPI-OER from the Bureau of Labor Statistics. House price growth is the growth rate of the S&P/Case-Shiller US national home price index. The sample period runs from 1987 to 2022.
Since house prices are not directly part of the CPI, their influence is limited to indirect channels such as rents or OER. Chart 1 plots the S&P/Case-Shiller US National Home Price Index along with the relevant housing components of CPI from 1987–2022. Over this period, there have been some large swings in house prices, while the OER and other housing-related components of shelter are much more stable and have not kept up with the large house price swings. This shows that these indirect channels are likely to be small. As such, the impact of house price inflation on overall inflation is close to zero.
To capture this disconnect more precisely, we establish an ‘accounting benchmark’ to define how house price movements should, in theory, affect measured inflation. Using US data from 1987–2022, we regress actual house price growth on overall CPI inflation and its major components, including twelve leads and lags of house price growth. These coefficients are then weighed by their respective shares in the CPI. This gives the implied elasticity of overall inflation to house price inflation, and it ranges between 0.004 and 0.04 across different specifications, refer to Dhamija et al (2026) for details. That is, a one percentage point increase in house price inflation should raise CPI inflation by no more than 0.04 percentage points. Any estimated relationship substantially larger than that would imply overweighting by households. However, households as non-specialists may be unable to distinguish between the asset aspect of house prices and housing services. This could potentially lead to overweighting of house price expectations in overall inflation expectations.
But can households distinguish between houses as assets and housing services?
We use the Michigan Survey of Consumers (MSC) and the Federal Reserve Bank of New York’s Survey of Consumer Expectations (SCE) to examine household behaviour in the US. For each survey, we regress inflation expectations on house price expectations of households, controlling for individual demographics, region and time fixed effects, past house price growth, and rent expectations, among others. To further address potential endogeneity arising from common shocks and/or omitted variables, we instrument house price expectations with housing supply elasticity using the Wharton Land Use Regulatory Index and past expectations.
We find that a percentage point increase in households’ expected house price growth is associated with a 0.25 to 0.45 percentage point increase in their inflation expectations, holding all else equal. Relative to the benchmark, this indicates that households place disproportionate weight on house price expectations when forming expectations about inflation.
Our second identification strategy exploits variation in households across characteristics.
If households overweight house price inflation expectations, this bias should be less pronounced among individuals with stronger numeracy skills and those who are currently more attentive to housing market developments. We find that more educated households and those with higher numeracy skills place less weight on house price expectations when forming inflation beliefs. We also find that households that moved homes recently, and therefore potentially observed housing markets more prominently, overweight by more. Taken together, the results of both identification strategies provide strong evidence of individuals overweighting from house price expectations to their inflation expectations.
Does this household behaviour matter for monetary policy?
To address this question, in Dhamija et al (2026) we embed this household behaviour into a two-sector New Keynesian model where households assign disproportionate attention to inflation developments in one sector relative to its actual weight. The model provides a stylised framework representative of any two sectors such that it could be used more broadly to examine the monetary policy implications of overweighting any good. This also encompasses the results documented in prior literature, such as D’Acunto et al (2021) and Coibion and Gorodnichenko (2015) among others, related to groceries or fuel prices. We show that this overweighting behaviour distorts households’ intertemporal choices by creating a wedge between the actual and perceived expected inflation rate. This misperception carries through to consumption and saving decisions, generating a wedge between the true and perceived real interest rate, which can amplify or dampen the effects of monetary policy. This household behaviour, however, does not alter the firms’ price-setting. Deriving the welfare function or the central bank’s loss function shows that this overweighting does not introduce any new policy trade-offs for the central bank. This implies that it is sufficient for the central bank to set the nominal rate in line with the perceived expected inflation to stabilise any distortions from overweighting.
Chart 2: Optimal response to a markup shock in the overweighted sector in models with overweighting (black) and without overweighting (red dashed)
Notes: The chart shows how key variables change in response to a one percent increase in the markup in the overweighted sector. Values are shown as changes from normal levels (steady state). The interest rate is shown in percentage points. The solid black line is the version of the model which incorporates overweighting, and the red dashed line is the version without overweighting (the standard case).
To illustrate this, in Chart 2, we examine how a central bank responds when inflation increases due to a markup shock in the overweighted sector. A markup shock is an increase in firms’ profit margins that increases inflation and decreases output. Since people put extra weight on price changes in this sector, inflation expectations rise more than they would otherwise. To keep overall inflation on track, the central bank therefore needs to raise the policy rate by more. With an appropriately stronger response, the economy ends up on essentially the same path as it would if households did not place extra weight on that sector.
Conclusion and policy implications
Recent research on salience demonstrates that individuals disproportionately emphasise frequently observed prices and large price movements when forming inflation expectations, even when these items carry low weight in official inflation indices. In Dhamija et al (2026), we identify a novel channel through house price expectations. We further show that inflation shocks in any overweighted sector have outsized effects on expectations and macroeconomic outcomes.
The policy implications of our work are twofold. First, our results make a case for central banks to monitor the housing sector due to its salience; this is beyond the usual, very important, financial stability concerns. Second, the knowledge of this household behaviour is imperative for central banks as movements in expected inflation in overweighted sectors are disproportionately more important for monetary policy. When households overemphasise price movements in one (salient) sector, the perceived inflation rate deviates from actual inflation. This requires central banks to respond more strongly to such sectoral inflation shocks, ie set the nominal interest rate in line with the perceived inflation expectations to undo any distortions. Our results may also have implications for central bank communication, which could be explored in future research. Going forward, we plan to examine whether house price expectations receive disproportionate weight in the formation of inflation expectations in the UK and other countries.
Vedanta Dhamija works in the Bank’s Monetary Policy Strategy Division, Ricardo Nunes is a Professor of Economics at the University of Surrey and Roshni Tara works in the Bank’s Economic Outlook Division.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.