In his doctoral dissertation in Economics at the Jyväskylä University School of Business and Economics, Finland, Joni Heikkinen, who holds a Master of Science in Economics and Business Administration, examines household expectations and the role of communication in macroeconomics. The central result of the dissertation concerns the European Central Bank’s (ECB) monetary policy communication. The results show that under the previous monetary policy strategy (1999–2021), the ECB was more averse to inflation above the price stability target than to inflation below it. In addition, the dissertation shows that media tone is associated with inflation expectations across different population groups and that online search data provides useful additional information for forecasting short-term economic developments.

In his dissertation, Joni Heikkinen analyzes how central bank communication reflects monetary policy preferences and how households engage with economic information through the media and online searches.

Heikkinen’s research examines the ECB’s monetary policy communication and its tone as part of the conduct of monetary policy. The tone of communication was measured using text analysis methods, such as dictionary-based approaches and language models. The results indicate that the ECB displayed asymmetric preferences over inflation deviations during the period 1999–2021 under the previous monetary policy strategy. Heikkinen’s follow-up research extends the analysis to the period after the July 2021 strategy review, when the ECB refined its price stability objective to be symmetric.

“The analysis of ECB communication suggests that the central bank’s preferences over inflation overshoots and undershoots were asymmetric during 1999–2021, but following the 2021 strategy review, when the ECB revised its price stability objective, communication has become more symmetric according to preliminary results,” Heikkinen notes.

Media tone and inflation expectations

According to the dissertation, media tone appears to be associated with inflation expectations across demographic groups. Media tone was measured using text analysis methods, including approaches based on language models. The results show that traditional news media are particularly linked to the inflation expectations of older individuals, whereas social media align more closely with the expectations of younger and more highly educated individuals.

“In addition, the results suggest that the relationship between inflation expectations and specific news topics, such as inflation- and central bank-related news as well as cost-of-living topics, varies across demographic groups,” Heikkinen notes.

The related article has been published in the Journal of Forecasting.

Online search data and short-term forecasting

Heikkinen’s research also investigates the use of online search data to forecast short-term economic developments, using machine-learning methods. The results show that online search data contains useful additional information for forecasting economic indicators.

“We find that online search data provide valuable information for forecasting purchasing managers’ indices,” says Joni Heikkinen.

The article on forecasting purchasing managers’ indices has been published in the journal Applied Economics.