Crystal Ball or Rear-View Mirror? Danish Consumer Expectations

In macroeconomic theory from the 1960s and 1970s, inflation expectations play a central role (Friedman 1968). In adaptive expectations models, agents form expectations about future inflation partly from past inflation. This matters because expectations can influence wage-setting, price-setting, consumption decisions, and therefore the inflation process itself. Central banks also monitor inflation expectations because well-anchored expectations can make inflation easier to stabilise. For further reading about how real interest rates are affected by inflation expectations, see the Danish National Banks overview (in Danish)

In this analysis, I use Statistics Denmark’s consumer survey on perceived and expected price developments to study two questions. First, do Danish consumers form expectations adaptively, by linking expected future price developments to recently perceived price increases? Second, are Danish consumers future-predicting prophets and wizards, or do their expectations have limited forecasting power?

The results suggest a clear “rear-view mirror” pattern. Actual year-over-year CPI inflation is strongly associated with consumers’ perceived price increases, meaning that consumers do register periods of high inflation. Perceived price increases are also positively related to expected future price developments, suggesting some degree of adaptive expectation formation. However, these expectations are a weak and unstable predictor of future inflation. Across several forward-looking horizons, the correlation between expected price developments and later CPI inflation is small, and the relationship remains unclear even when major shock periods such as the financial crisis and the 2021–2023 inflation surge are excluded. Overall, Danish consumers appear better at recognising inflation after it has happened than predicting where inflation is going.

This analysis combines Danish consumer survey data with official inflation data from Statistics Denmark between 2001 and 2026. Consumer perceptions and expectations are taken from the FORV1 survey, which asks households how they perceive recent price developments and how they expect prices to develop over the next 12 months. Actual inflation is measured using CPI data from Statistics Denmark and converted into year-over-year inflation rates. The dataset therefore allows a comparison between actual inflation, perceived past price developments, and expected future price developments.

The consumer survey variables are balance indicators, not actual percentage inflation rates. The survey variables are measured on a scale from minus 100 to 100. For perceived past prices, a value of 50 means that consumers on balance report prices as somewhat higher than one year earlier, while −50 means that prices are perceived as roughly unchanged. For expected prices, 50 means that consumers expect prices to rise at about the same pace as currently, while −50 means that they expect prices to remain unchanged. Higher values therefore indicate stronger perceived or expected price pressure, but they should not be interpreted as exact inflation forecasts. Further information can be found here.

The strategy is descriptive as I compare actual inflation with perceived inflation to test whether consumers recognise inflation after it occurs. I then compare perceived price developments with expected price developments to assess whether expectations appear adaptive. Finally, I compare expected price developments with future inflation at different horizons to evaluate whether consumer expectations contain predictive information.

Figure 1 shows that Danish inflation is usually quite stable, but with clear extreme periods around the financial crisis and the Corona/inflation shock. Figure 2 then shows that consumers do realise when inflation has risen: actual year-to-year inflation is clearly connected with perceived price development. Figure 3 shows a positive relationship between perceived recent price development and expected future price development. Consumers who report stronger recent price increases also tend to expect stronger future price pressure. However, this should be interpreted cautiously because the two survey scales are not directly identical: a value of 0 in perceived inflation means that prices are “a little higher”, while a value of 0 in expected inflation means that prices are expected to “increase more slowly than now.” The figure is therefore best understood as evidence of a positive correlation, rather than a precise measure of adaptive expectations.

However, Figure 4 shows the main result: consumers are not future-predicting prophets and wizards. Expected price development has no clear or stable relationship with actual future inflation across the different lags. Even when the extreme shock periods are removed, the relationship is still weak. Overall, Danish consumers seem much better at noticing inflation after it has happened than predicting where inflation is going.