Abstract
Recent theoretical advances have called the invariance of sales shares with respect to productivity differences into question, arguing that sizable higher-order macroeconomic effects may emerge when sales shares react to productivity shocks. Here we propose a parsimonious model for the adjustment of sales shares in response to productivity shocks, and operationalize it to readily quantify the impact of idiosyncratic shocks on aggregate fluctuations. Using input-output data for the European (EU28) economy, we find that our model significantly outperforms recently suggested specifications that rely either on the granularity of sales shares or the heavily skewed distribution of connections in production networks. While our results confirm earlier findings in the sense that microeconomic shocks are an important driver of macroeconomic fluctuations, we demonstrate that previous approaches substantially underestimate their relative impact because they fail to account for the interaction of granularity and network effects. Our main empirical finding is that idiosyncratic shocks to about three percent of industries already explain more than two thirds of the business cycle once we account for this interaction. ∗We thank Simone Alfarano, Herbert Dawid, Sebastian Gechert, Alexander Hempfing, Tim Hagenhoff, Tomasz Makarewicz, Philipp Mundt, Christian Proa˜no and Frank Westerhoff, as well as participants at the Workshop on Heterogeneous Interacting Agents in London in 2019 and Ancona in 2026, the First Behavioral Macroeconomics Workshop in Bamberg in 2019, the conference of the Forum for Macroeconomics and Macroeconomic Policies in Berlin in 2019, and the International Workshop on Economic Complexity and Macroeconomic Dynamics in Valencia in 2025, for helpful comments and discussions. We gratefully acknowledge financial support from the Hans-B¨ockler-Foundation through grant PK045. 1