Inflation expectations: How they are formed, why they matter, and how we measure them
Ask a central banker what keeps them up at night, and inflation expectations will rank near the top. The reason is that expectations of households and firms can influence the very inflation they anticipate. If workers expect prices to rise, they ask for higher wages; higher wages raise firms' costs, and firms that expect their costs to keep rising may pass them on in higher prices. Through these channels, expectations can make an inflation shock more persistent.¹ Their influence also extends beyond prices. What households and firms expect about the future already shapes what they decide today: how much to spend and save, whether to buy a car now or later, and whether to hire, borrow, or invest.²
The importance of forward-looking behavior was highlighted again by the 2026 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, awarded to Robert E. Hall and Michael Woodford “for microeconomic foundations of macroeconomic stabilisation.” Their work helped establish how expectations about the future influence households’ consumption decisions, firms’ pricing behavior, and the transmission of economic policy. Yet, while expectations had long played a central role in economic theory, economists had relatively limited direct evidence on what households actually expected. For much of the twentieth century, expectations were primarily inferred from observed behavior or represented through theoretical assumptions rather than measured directly.
This post sets out why expectations matter, how their measurement has evolved, and two strands of survey work that underpin the research agenda of the European Expectations Center (EEC) at ESMT: survey experiments on the NielsenIQ consumer panel in the United States and the Global Subjective Expectations Survey (GSES).
Why expectations matter
The modern case for taking expectations seriously goes back to Edmund Phelps and Milton Friedman, later also awarded the Nobel Prize in economics in 2006 and 1976, respectively.³ Their point was simple: a government cannot keep unemployment low for good just by letting prices rise faster. Once people come to expect higher inflation, workers ask for higher wages and firms raise their prices, so the extra jobs disappear and only the higher inflation remains. The rational expectations revolution, for which Robert Lucas received the Nobel Prize in 1995, then helped make expectations a central element of modern macroeconomic models.⁴
In standard models, expectations also matter because they enter decisions taken today. A household that expects prices to rise faces a lower expected real interest rate if nominal rates stay unchanged, which makes saving less attractive and buying now rather than later more appealing. The same logic underpins a range of policies: if a central bank or government can raise expected inflation when interest rates are stuck at zero, it can in principle stimulate spending today by lowering perceived real interest rates.⁵ Firms face similar trade-offs when they set prices, hire, and invest. But whether households and firms actually respond as these models predict is ultimately an empirical question.
Recent advances in survey research have made it possible to test these predictions directly, linking individuals' reported expectations to their economic decisions. A few examples, each of which will be the subject of a forthcoming post on this blog:
-
Consumption. Germany announced in late 2005 that VAT would rise by three percentage points in January 2007. Compared with observationally similar households in other countries, German households raised their inflation expectations and became more willing to buy durable goods before the tax took effect.⁵ Germany’s temporary VAT cut in the second half of 2020 provides complementary evidence: households that perceived a greater reduction in consumer prices increased their spending on durable goods more than those that perceived a smaller reduction.⁶
-
Who responds. The response is far from uniform. In Finland, where cognitive ability is measured in military entry tests taken by men, only men with high cognitive ability raised their propensity to consume when they expected higher inflation, as standard models predict.⁷
-
Uncertainty. Not only the level of expectations matters, but also how uncertain people are. In a randomized experiment in the euro area, households whose uncertainty about future growth rose after receiving information treatments cut their spending in the following months, including on large durables.⁸
-
Learning over time. Randomized experiments across countries show that belief formation changes with the inflation environment: as inflation rose, households and firms paid more attention to inflation news and became better informed, so additional information about inflation and monetary policy produced smaller revisions in their expectations.⁹
The experience of 2021 to 2023 made these questions pressing. As inflation reached levels not seen in four decades in many advanced economies, policymakers watched survey measures of household expectations closely for signs that a temporary price shock was turning into a persistent one.¹⁰
From experts to households
But whose expectations should policymakers monitor? For decades, the measures most prominent in macroeconomic analysis and policy discussions came from professional forecasters and financial markets.¹¹ Household surveys existed for even longer, led by the University of Michigan's Surveys of Consumers founded in 1946, but received far less attention.¹² Yet neither forecasters nor markets tell us what the households and firms whose spending, work, pricing, and investment decisions drive the economy actually believe. Over the last fifteen years, central banks have built a new generation of household and firm surveys that follow the same people over time and ask far more detailed questions.¹³
Households form inflation expectations partly from everyday shopping: they focus on price changes of frequently purchased goods and give more weight to price increases than to equally large decreases,¹⁴ while often recalling past prices as lower than those they actually paid.¹⁵ The answers look very different from professional forecasts. Household inflation expectations are typically higher and much more dispersed than professional forecasts, often exceed realized inflation, and vary systematically with gender, education, and cognitive ability, among other factors.¹⁶ These patterns initially raised concerns that household survey responses contained mostly measurement error. The evidence now points the other way: however imprecise as predictors of future inflation, subjective beliefs help explain differences in consumption, saving, and borrowing, and why similar households respond differently to the same policy.²
Two strands of survey work behind the EEC's agenda
Studying these relationships requires more than asking households to forecast inflation. Researchers also need to observe how beliefs change, how they relate to actual behavior, and whether the same patterns hold across different economic environments. Two survey programs are central to this research agenda: the NielsenIQ survey experiments in the United States and the Global Subjective Expectations Survey (GSES). The table below summarizes their main features, with the following sections providing more detail on their design, coverage, and use in research.
|
NielsenIQ survey experiments |
Global Subjective Expectations Survey (GSES) |
|
| Coverage | United States | Advanced and emerging economies |
| Population | Households in the NielsenIQ consumer panel | Adults aged 18–70 in participating countries |
| Period | Quarterly; surveys fielded since 2015 | Annual waves since 2025 |
| Sample size | Varies by wave, typically about 10,000–40,000 households | About 1,000–2,000 respondents per country, depending on wave |
| What is measured | Expectations, economic attitudes, financial behavior | Expectations, perceptions, central-bank knowledge, information sources |
| What makes it distinctive | Survey responses can be linked to detailed household purchases and prices; some waves include randomized information treatments | Harmonized questionnaire across countries, allowing internationally comparable measures of subjective expectations |
Further information on the survey data, questionnaires, and data access is available here.
Survey experiments on the NielsenIQ consumer panel
U.S. household surveys linking subjective expectations to detailed consumer purchase data and, in selected waves, randomized information treatments.
The NielsenIQ Homescan panel follows tens of thousands of U.S. households that record detailed purchases of consumer goods, mostly groceries and other everyday items, by scanning them at home. Surveys fielded to this panel combine three elements that are rarely available together: households' stated expectations, their actual purchases and the prices they paid, and randomized information treatments. This makes it possible to move from correlation to causation. A few important findings using this survey:
-
Where beliefs come from. Households put heavy weight on price changes of goods they buy often, react more to increases than to decreases, and are unaffected by prices of goods on the same shelves that they do not buy.¹⁴ This helps explain why household perceptions and expectations can differ substantially from official inflation statistics.
-
How communication works. In 2018, nearly 20,000 participants received different pieces of information about inflation and monetary policy. Reading the actual FOMC statement had about the same average effect on expectations as simply being told the Federal Reserve's 2% inflation target; reading news coverage of the meeting moved expectations by about half as much. The effects weakened substantially over the following months.¹⁷
-
From beliefs to behavior. The changes in expectations induced by the experiment affected subsequent household spending, as recorded in both scanner and survey data.¹⁷ In related experiments, information about near-term interest rates moved expectations, while guidance about the more distant future added little.¹⁸
The Global Subjective Expectations Survey
An annual cross-country survey measuring household expectations, perceptions, information sources, and knowledge of economic institutions using a harmonized questionnaire.
The NielsenIQ experiments provide detailed evidence on how households form expectations and respond to information, but this evidence comes from the United States. Whether similar patterns hold across countries is an important question. Most high-quality survey evidence still comes from a handful of advanced economies, yet inflation histories, central bank independence, and media environments differ widely across the world. The environment also shapes how people learn: the same information moves expectations much more when inflation is low and attracts little attention than when it is high and already in the news.⁹
The GSES builds on an earlier global survey led by Francesco D'Acunto and EEC Director Michael Weber, which reached 46,285 consumers in 47 countries, covering about 90% of world GDP, in April and May 2023.¹⁹ It showed that many consumers rely heavily on local sources of information, such as their own shopping, utility bills, and conversations with people they know, which are not representative of the economy as a whole. Reliance on these sources goes along with more biased macroeconomic expectations, and distrust of governments and central banks helps explain why people turn to them.
Building on this experience, the Global Subjective Expectations Survey (GSES) was launched in 2025. The first wave was fielded in March and April 2025 in 32 advanced and emerging economies, with about 1,000 adults per country.²⁰ The second wave, fielded in April and May 2026, covers 32 countries and more than 64,000 adults, representing around 65% of the world's population and more than 85% of global GDP. Each round combines a common core of questions on inflation expectations, perceptions of prices and wages, knowledge of central banks, and information sources with a special module that changes from round to round. The 2026 module focuses on the economic effects of artificial intelligence.
Country samples are drawn from online survey panels and are designed to be broadly representative of each country's adults aged 18 to 70 by age and gender, following a harmonized sampling approach across waves. In 2026 the survey was fielded by Cint, respondents were screened for age and attention, and the questionnaire was translated into each country's official languages using a common protocol. The samples are representative by age and gender, although they are not probability samples of the full population.
The first wave illustrates why these differences in perceptions and knowledge matter. Household expectations are roughly double professional forecasts in advanced economies, and the gap is wider in emerging markets. Households believe they lost about 6% in real wages after the pandemic, although official data suggest wages broadly kept pace with prices. And while most people recognize their central bank's name, only about half know that price stability is among its goals; those who do report markedly lower inflation expectations.
Looking ahead
Seventy years ago, one university survey and a journalist's poll of economists were nearly all we had. Today it is possible to follow the same households over time, link their beliefs to what they buy, test which messages change their minds, and compare all of this across countries. The picture that emerges is not one of fully informed, identical agents, but of households whose beliefs reflect the information, prices, and experiences they encounter in everyday life. For central banks the lesson is practical: speaking clearly to markets is not enough when the public, whose spending and wage demands matter most, may not be listening.
This is why tracking household expectations closely, and across countries, remains essential. The EEC will continue to field its surveys and report new findings on this blog.
Data and supporting materials for the surveys discussed in this post are available here.
Notes
-
Weber, M., D'Acunto, F., Gorodnichenko, Y. and Coibion, O. (2022), "The subjective inflation expectations of households and firms: Measurement, determinants, and implications", Journal of Economic Perspectives, Vol. 36, No 3, pp. 157–184.
-
D'Acunto, F. and Weber, M. (2024), "Why survey-based subjective expectations are meaningful and important", Annual Review of Economics, Vol. 16, pp. 329–357.
-
Phelps, E. S. (1967), "Phillips curves, expectations of inflation and optimal unemployment over time", Economica, Vol. 34, No 135, pp. 254–281; Friedman, M. (1968), "The role of monetary policy", American Economic Review, Vol. 58, No 1, pp. 1–17.
-
Muth, J. F. (1961), "Rational expectations and the theory of price movements", Econometrica, Vol. 29, No 3, pp. 315–335; Lucas, R. E. (1972), "Expectations and the neutrality of money", Journal of Economic Theory, Vol. 4, No 2, pp. 103–124.
-
D'Acunto, F., Hoang, D. and Weber, M. (2022), "Managing households' expectations with unconventional policies", Review of Financial Studies, Vol. 35, No 4, pp. 1597–1642.
-
Bachmann, R., Born, B., Goldfayn-Frank, O., Kocharkov, G., Luetticke, R. and Weber, M. (forthcoming), "A temporary VAT cut as unconventional fiscal policy", Review of Economic Studies.
-
D'Acunto, F., Hoang, D., Paloviita, M. and Weber, M. (2023), "IQ, expectations, and choice", Review of Economic Studies, Vol. 90, No 5, pp. 2292–2325.
-
Coibion, O., Georgarakos, D., Gorodnichenko, Y., Kenny, G. and Weber, M. (2024), "The effect of macroeconomic uncertainty on household spending", American Economic Review, Vol. 114, No 3, pp. 645–677.
-
Weber, M., Candia, B., Afrouzi, H., Ropele, T., Lluberas, R., Frache, S., Meyer, B. H., Kumar, S., Gorodnichenko, Y., Georgarakos, D., Coibion, O., Kenny, G. and Ponce, J. (2025), "Tell me something I don't already know: Learning in low- and high-inflation settings", Econometrica, Vol. 93, No 1, pp. 229–264.
-
D'Acunto, F., Charalambakis, E., Georgarakos, D., Kenny, G., Meyer, J. and Weber, M. (2024), "Household inflation expectations: An overview of recent insights for monetary policy", NBER Working Paper No 32488.
-
Among the longest-running are the Livingston Survey of economists (since 1946) and the Survey of Professional Forecasters (since 1968), both now conducted by the Federal Reserve Bank of Philadelphia. Market-based measures are derived from inflation-indexed government bonds.
-
University of Michigan, Surveys of Consumers.
-
Examples include the Bank of England Inflation Attitudes Survey (1999), the Federal Reserve Bank of Atlanta Business Inflation Expectations Survey (2011), the Federal Reserve Bank of New York Survey of Consumer Expectations (2013), the Deutsche Bundesbank Online Panel – Households (2019) and the ECB Consumer Expectations Survey (2020). See Armantier, O., Topa, G., van der Klaauw, W. and Zafar, B. (2017), "An overview of the Survey of Consumer Expectations", Federal Reserve Bank of New York Economic Policy Review, Vol. 23, No 2, pp. 51–72; Georgarakos, D. and Kenny, G. (2022), "Household spending and fiscal support during the COVID-19 pandemic: Insights from a new consumer survey", Journal of Monetary Economics, Vol. 129, pp. S1–S14.
-
D'Acunto, F., Malmendier, U., Ospina, J. and Weber, M. (2021), "Exposure to grocery prices and inflation expectations", Journal of Political Economy, Vol. 129, No 5, pp. 1615–1639.
-
D'Acunto, F. and Weber, M. (2023), "Memory and beliefs: Evidence from the field", Working paper.
-
D'Acunto, F., Malmendier, U. and Weber, M. (2023), "What do the data tell us about inflation expectations?", in Handbook of Economic Expectations, Academic Press, pp. 133–161; Malmendier, U. and Nagel, S. (2016), "Learning from inflation experiences", Quarterly Journal of Economics, Vol. 131, No 1, pp. 53–87.
-
Coibion, O., Gorodnichenko, Y. and Weber, M. (2022), "Monetary policy communications and their effects on household inflation expectations", Journal of Political Economy, Vol. 130, No 6, pp. 1537–1584.
-
Coibion, O., Georgarakos, D., Gorodnichenko, Y. and Weber, M. (2023), "Forward guidance and household expectations", Journal of the European Economic Association, Vol. 21, No 5, pp. 2131–2171.
-
D'Acunto, F. and Weber, M. (2026), "Information and macroeconomic expectations: Global evidence", NBER Working Paper No 35511.
-
D'Acunto, F., De Fiore, F., Sandri, D. and Weber, M. (2025), "A global survey of household perceptions and expectations", BIS Quarterly Review, September.
-
D'Acunto, F., De Fiore, F., Sandri, D. and Weber, M. (forthcoming), "Inflation, Households, and Central Banks Around the World", Journal of Monetary Economics.
-
Bruschi, C., D'Acunto, F., De Fiore, F., Sandri, D. and Weber, M. (2026), "Will AI Help or Hurt Work and the Economy? What People Around the World Expect", Working paper.
