VoxEU Column Monetary Policy

Influencing household inflation expectations

Although central banks have a natural desire to influence household inflation expectations, there is no consensus on how these expectations are formed or the best ways to influence them. This column presents evidence from a series of survey experiments conducted in a low-inflation context (the US) and a high-inflation context (Argentina). The authors find that dispersion in household expectations can be explained by the cost of acquiring and interpreting inflation statistics, and by the use of inaccurate memories about price changes of specific products. They also provide recommendations for central bank communication strategies. 

Expectations about macroeconomic variables play an important role in economic theory and policymaking. Household inflation expectations, in particular, are key to understand consumption and investment decisions, and ultimately, the impact of monetary policies. Although central banks have a natural desire to influence expectations, there is no consensus on how household expectations are formed or what the best way to affect them is (see Bernanke 2007, Bachmann et al. 2012, Coibion and Gorodnichenko 2013, and Armantier et al. 2014).

When measured using surveys, household inflation expectations tend to be much more heterogeneous than those of professional forecasters (Ranyard et al. 2008, Armantier et al. 2013). As an illustration, Figure 1 provides a comparison of households’ and experts’ expectations for the US and Argentina, the two countries in our study (see Nishiguchi et al. 2014 for similar evidence from Japan). Two main explanations have been given in the economics literature for this degree of dispersion in household expectations. Some authors attribute it to rational inattention, according to which individuals only partly incorporate information on topics such as inflation statistics, because acquiring that information is costly (Mankiw et al. 2003, Carroll 2003). This explanation is particularly convincing in contexts of low inflation like the US, where the potential financial cost of ignoring inflation is negligible for most households. Other authors argue that, in forming inflation expectations, individuals use information derived from their personal experience as consumers, which can be both diverse and inaccurate (Bruine de Bruin et al. 2011, Malmendier and Nagel 2013, Madeira and Zafar forthcoming). These explanations are hard to distinguish empirically because they are not mutually exclusive. Individuals may choose to be rationally inattentive and, at the same time, use their personal shopping experiences as a low-cost source of information about price changes.

Figure 1. Households’ and professional forecasters’ inflation expectations for 2013, US and Argentina

Note: Expected inflation for the period 1 January to 31 December 2013, reported in December 2012.
Sources: University of Michigan’s Survey of Consumers, December 2012 (household survey, US, N=502), Federal Reserve Bank of Philadelphia’s Survey of Professional Forecasters, fourth quarter of 2012 (professional forecasters, US, N=48), WP Public Opinion Survey (household survey, Argentina, N=777; see Cavallo et al. 2014 for details) and LatinFocus Consensus Forecast, January 2013 (professional forecasters, Argentina, N=16).

New evidence from survey experiments

In recent work (Cavallo et al. 2014), we present evidence from a series of survey experiments specifically designed to disentangle some of these effects. We randomly provided subjects with information related to past inflation, such as inflation statistics and the historical prices of specific supermarket products. On the basis of that experimental variation, we use a Bayesian learning model to infer how much weight subjects assign to a given piece of information relative to their prior beliefs about past inflation. Our methodology allows us to distinguish between spurious and genuine learning. In order to assess the role of the rational inattention model, we conducted similar experiments in a context of low inflation – the US, with an average annual inflation rate of 1.8% in the five years prior to our study – and in a context of high inflation – Argentina, where the average annual inflation rate over the same time period was around 22.5%.1

Households update their beliefs, especially in a low-inflation context

Our results indicate that information related to past inflation has a major impact on inflation expectations. We find that, when confronted with information about past inflation that is different from their priors, individuals will assign a weight of 50% to 80% to the new data to update their beliefs. This happens both when we provide information about aggregate inflation statistics and when we provide information about the historical prices of a few individual supermarket products. This evidence is consistent with the existence of largely inattentive consumers who learn from new information.

Furthermore, the results across countries suggest this inattention is rational. Relative to her prior belief, an individual in a low-inflation context assigns a weight of roughly 85% to the information on recent inflation statistics, whereas an individual in a high-inflation context assigns a weight of roughly 50%. The differences are similar when comparing the weights assigned to information about supermarket prices rather than inflation statistics. The fact that learning rates were 70% higher in the low-inflation context is consistent with the rational inattention model, which predicts that individuals in a context of higher inflation are more informed because the cost of misperceiving inflation is greater.

Inflation statistics vs. prices of familiar products

In one of our treatment arms, we provided individuals with information on inflation statistics and, simultaneously, with information on historical prices for a handful of supermarket products. Subjects still assigned significant weight to the prices of specific products – even a higher weight than that assigned to inflation statistics. In other words, subjects were more prone to incorporating information about the price changes of a few familiar products, such as bread and milk, than to statistics on the price changes of thousands of products. One possible interpretation, still consistent with rational inattention, is that it is less costly for individuals to incorporate information on individual prices because they are easier to understand.

Memories can be misleading

Given that inflation statistics are costly to interpret, individuals may substitute inflation statistics for their memories about price changes at retail stores. To better understand how past shopping experiences affect inflation expectations, we conducted a consumer intercept survey experiment at a supermarket chain in Argentina. We recorded consumers’ purchases by scanning the supermarket receipts of participants, which were linked to data on the actual historical prices of those same products at the same store. We also asked respondents to recall historical prices for a random set of items that they had just purchased, which allowed us to generate exogenous variations in the salience of their own price memories. We find evidence that individuals use their own memories of price changes for specific products when forming inflation expectations. However, their memories are orthogonal to the actual price changes experienced by the products recently bought by the subjects. Far from correcting a representativeness bias in aggregate inflation statistics, the use of price memories as inputs in the formation of inflation expectations seems to induce significant errors in inflation expectations.

Concluding remarks

Our findings are relevant for recent debates on central bank transparency and communication strategies. Central banks will often try to use information to affect household inflation expectations. At the time of writing, for example, Japan’s central bank is eagerly trying to increase household inflation expectations (see Baldwin and Gros 2013), while high-inflation countries such as Argentina and Venezuela are desperately trying to reduce them. Our evidence suggests that, in addition to the dissemination of aggregate statistics, central banks have an additional policy margin that consists of communicating how objective, precise, and representative their statistics are. For example, the ECB and the French statistical agency have made notable efforts to create online tools to convey this information and the way it is collected and processed in a user-friendly way. Moreover, in addition to publishing aggregate annual inflation rates, statistical agencies could provide tables with historical prices for individual products similar to the ones in our experiments, selecting sets of widely known products with average price changes that replicate the aggregate rate.


Armantier, O, S Nelson, G Topa, W van der Klaauw, and B Zafar (2014), “The Price Is Right: Updating of Inflation Expectations in a Randomized Price Information Experiment”, Review of Economics and Statistics, forthcoming.

Armantier, O, W Bruine de Bruin, G Potter, G Topa, W van der Klaauw, and B Zafar (2013), “Measuring Inflation Expectations”, Annual Review of Economics 5: 273–301.

Bachmann, R, T Berg, and E Sims (2012), “Inflation Expectations and Readiness to Spend: Cross-Sectional Evidence,” NBER Working Paper 17958.

Baldin, R and D Gros (2013), “Augmented inflation targeting: Le roi est mort, vive le roi”, VoxEU.org, 17 April. 

Bernanke, B (2007), “Inflation Expectations and Inflation Forecasting”, Speech at the Monetary Economics Workshop of the NBER Summer Institute, Cambridge, Massachusetts, 10 July. 

Bruine de Bruin, W, W van der Klaauw, and G Topa (2011), “Expectations of inflation: The biasing effect of thoughts about specific prices”, Journal of Economic Psychology 32(5).

Carroll, C (2003), “Macroeconomic Expectations of Households and Professional Forecasters”, Quarterly Journal of Economics 118(1).

Cavallo, A, G Cruces, and R Perez-Truglia (2014), “Inflation Expectations, Learning and Supermarket Prices: Evidence from Field Experiments”, NBER Working Paper 20576.

Coibion, O and Y Gorodnichenko (2013), “Is The Phillips Curve Alive and Well After All? Inflation Expectations and the Missing Disinflation”, NBER Working Paper 19598.

Madeira, C and B Zafar (2014), “Heterogeneous Inflation Expectations, Learning, and Mar-ket Outcomes”, Journal of Money, Credit, and Banking, forthcoming.

Malmendier, U and S Nagel (2013), “Learning from Inflation Experiences”, Working paper, Berkeley.

Mankiw, N G, R Reis, and J Wolfers (2003), “Disagreement About Inflation Expectations”, in M Gertler and K Rogoff (eds.), NBER Macroeconomics Annual 2003.

Nishiguchi, S, J Nakajima, and K Imakubo (2014), “Disagreement about inflation expectations: The case of Japanese households”, VoxEU.org, 2 May. 

Ranyard, R, F D Missier, N Bonini, D Duxbury, and B Summers (2008), “Perceptions and expectations of price changes and inflation: A review and conceptual framework”, Journal of Economic Psychology 29(4): 378–400.


1 We do not use official inflation statistics for Argentina, since these are widely discredited. We use instead an indicator compiled by the private sector, which is well known and trusted.

2,205 Reads