Preview

PDF, English Download (8MB) | Terms of use

Abstract

This dissertation contributes to the empirical macroeconomics of inflation and expectations, with a particular focus on how information and inflation regimes shape price dynamics. Across three chapters, it investigates state dependence in the transmission of shocks and examines how media content influences households’ beliefs. Two of the empirical analyses draw on a large corpus of German newspaper articles spanning 2006-2025, classified with large-language-model methods, and on household inflation expectations from the European Commission and European Central Bank consumer surveys. This rich combination of text and survey data enables a granular view of belief formation across outlets and readerships. The first chapter documents that the pass-through of supply shocks to consumer prices is regime-dependent: using a Markov-switching model and instrumental-variables local projections, shocks have larger and more persistent effects in high-volatility inflation states. A simple model with endogenous price flexibility rationalizes these findings. The second chapter constructs outlet-specific news indices that summarize the inflation outlook conveyed by newspapers and links them to household expectations. Weighting by readership composition shows that newspapers best predict the expectations of their own audiences. The third chapter measures fiscal narratives, such as whether expansion today implies future consolidation, and studies how these narratives shape expectations and interact with government spending shocks. Readership-weighted fiscal narrative indices raise inflation expectations when expansion is portrayed as inflationary and amplify the effects of fiscal shocks in local projections. Together, the chapters show that both the inflation regime and the content of public information are central to understanding price dynamics. The results highlight the value of integrating text-as-data with macroeconometric identification and point to policy implications for communication and forecasting.