Households often report inflation perceptions and expectations that differ markedly from official measures. This column proposes a new measure of perceived inflation, which uses disaggregated inflation data with indicators for household attention to capture the prices that people actually notice. The difference between perceived and official inflation varies over time and becomes particularly pronounced during periods of economic stress. Heightened attention to inflation is persistent and can remain elevated even after official inflation measures improve. As central banks recognise the importance of household beliefs for macroeconomic outcomes, it becomes important to also understand the gap between measured and perceived inflation.
Central banks around the world have spent the last several years battling the aftermath of the post-pandemic inflation surge. While headline inflation has fallen substantially in many countries, policymakers continue to face a puzzle: households often report inflation perceptions and expectations that differ markedly from official measures. This divergence matters because consumers make spending, saving, and borrowing decisions based on the inflation they believe they face, not necessarily the inflation measured by statistical agencies.
Recent research has highlighted the importance of household inflation beliefs for macroeconomic outcomes. Surveys consistently show that households’ inflation expectations differ from those of professional forecasters and central banks (Coibion et al. 2022, Weber et al. 2022). A growing literature has also documented that consumers place disproportionate weight on salient prices such as groceries, gasoline, and other frequently purchased goods when forming beliefs about inflation (D’Acunto et al. 2021). Recent VoxEU columns have surveyed the implications of household inflation expectations for monetary policy (D’Acunto et al. 2024), discussed how consumers learn about inflation (Malmendier et al. 2022), and examined how exposure to frequent price changes shapes inflation beliefs (Malmendier et al. 2019).
These findings raise an important observation: if households do not process all prices equally, we should expect them to perceive inflation differently from the Consumer Price Index (CPI).
In recent research (Francis 2026), I propose a new measure, Perceived Inflation (PI), to quantify this difference. The central idea is simple: CPI measures changes in the cost of living, while perceived inflation measures changes in the prices that people actually notice.
Official inflation statistics are designed to track the cost of a representative consumption basket. The CPI assigns weights to categories according to their expenditure shares and aggregates price changes across thousands of goods and services.
Households, however, do not experience inflation in this way.
Consumers pay more attention to some prices than others. A sudden increase in gasoline prices may attract considerable attention because it is visible and frequently encountered. Grocery prices are similarly salient because households observe them regularly. In contrast, many prices receive little attention, even when they represent a significant share of total expenditure.
Behavioural economics provides several explanations for this pattern. Rational inattention models suggest that individuals face cognitive limits and cannot monitor all available information. Salience models emphasise that unusually large price movements attract disproportionate attention. Other frameworks highlight that consumers selectively focus on information that appears most relevant for their decisions.
The implication is that inflation perceptions are shaped not only by actual price changes but also by how attention is allocated across goods.
To measure perceived inflation, I combine disaggregated CPI data with several indicators that proxy for household attention.
The first is price salience, which captures how large or unusual recent price changes are. The second is price volatility, reflecting the tendency of consumers to pay more attention to categories whose prices fluctuate frequently. The third is media coverage, measured using counts of news references to specific consumption categories. The fourth is online search activity, which captures active information seeking by households.
Together, these indicators are used to estimate a latent attention index for each consumption category. Categories that attract greater attention receive larger weights in the perceived inflation measure than they receive in the official CPI.
In effect, the approach allows inflation weights to vary with attention rather than remaining fixed at expenditure shares.
The resulting measure preserves the information contained in official price statistics while incorporating evidence on how consumers process those prices.
The difference between perceived inflation and CPI inflation can be summarised by what I call the perceived inflation wedge:
Perceived inflation wedge = Perceived inflation − CPI inflation.
The wedge is not constant over time because perceptions frequently change.
Using monthly US data, I show that perceived inflation closely tracks CPI inflation over the long run, but it exhibits substantially greater short-run volatility. The divergence becomes particularly pronounced during periods of economic stress, including the Global Crisis 2008-09, the COVID-19 recession, and the inflation surge of 2021–22 (see Figures 1 and 2).
Figure 1 Perceived inflation and CPI inflation
Figure 2 The perceived inflation wedge
These episodes share a common feature: heightened attention to prices.
When inflation becomes a prominent topic in news coverage and daily conversations, households devote more attention to price information. As attention shifts, some categories receive greater weight in inflation perceptions than they do in official statistics. The result is a time-varying wedge between measured and perceived inflation.
The evidence also suggests that attention itself is persistent. Once consumers begin paying close attention to inflation, that heightened awareness tends to remain elevated for some time, even after inflation starts to decline. This may help explain why inflation concerns often linger after official inflation measures have begun to improve.
At first glance, differences between perceived and measured inflation may appear to be merely a psychological curiosity. But they can have important economic consequences.
Economic decisions depend on perceived real returns. When households decide whether to save, consume, or borrow, they compare nominal returns with their beliefs about inflation. If perceived inflation differs from actual inflation, households may misjudge the real value of financial decisions.
To illustrate this mechanism, I embed the estimated perceived inflation wedge into a standard heterogeneous-household model. In this framework, households make decisions using perceived inflation rather than official inflation.
The results suggest that attention-driven inflation perceptions can meaningfully influence household behaviour. Changes in the perceived inflation wedge affect saving decisions, consumption choices, wealth accumulation, and the prevalence of liquidity-constrained households. The effects are particularly pronounced among lower-wealth households, whose financial decisions are more sensitive to perceived changes in purchasing power.
These findings are consistent with a broader literature showing that households act on their inflation beliefs even when those beliefs differ from official statistics.
The perceived inflation wedge also carries implications for monetary policy communication.
Central banks typically communicate using aggregate inflation measures and forecasts. Yet households often form beliefs from a very different information set. They encounter prices at grocery stores, gas stations, and online marketplaces. They are influenced by news coverage, social media, and conversations with friends and family.
As a result, changes in official inflation may not translate immediately into changes in household perceptions.
This communication challenge has become increasingly important in the wake of the recent inflation episode. Evidence suggests that households pay more attention to inflation when inflation is high and become less attentive when inflation is low (Pfauti 2026). This asymmetry implies that perceptions may react strongly during inflation surges but adjust only gradually during disinflation.
Understanding the perceived inflation wedge can therefore help policymakers interpret survey measures of inflation expectations and design more effective communication strategies. If households respond to salient prices rather than aggregate statistics, successful communication may require engaging with the sources of attention that shape inflation perceptions.
The CPI remains the appropriate benchmark for measuring changes in the cost of living. The goal of perceived inflation is not to replace official statistics. Rather, it is to complement them by providing information about how households experience and interpret price changes.
The distinction matters because economic decisions are driven not only by objective conditions but also by perceptions.
Inflation measurement tells us how prices change. Perceived inflation helps explain how those changes are noticed.
As central banks increasingly recognise the importance of household beliefs for macroeconomic outcomes, understanding the gap between measured and perceived inflation may become as important as measuring inflation itself.
Source : VOXeu
Walk into a hospital in Nigeria, where one doctor serves every 9,000 patients--fifteen times the…
The vision of the World Bank is to end extreme poverty and boost shared prosperity on a livable planet. The…
In many developing countries, two people can earn the same income and pay very different…
Every month, millions of migrant workers send a portion of their earnings home to support…
Defence policy is usually discussed as a national security issue, but its economic effects are…
The use of sanctions has grown steadily over the past seven decades, with well-documented effects…