The impact of the COVID-19 shock on response distributions in business tendency surveys: An application of the Kullback–Leibler divergence
The impact of the COVID-19 shock on response distributions in business tendency surveys: An application of the Kullback–Leibler divergence

Summary/Abstract: This paper applies the Kullback-Leibler divergence (KLD) to quantify discrepancies between expectations and subsequent assessments in business tendency surveys for Polish manufacturing firms. Using monthly survey data from May 2004 to January 2026, it examines how closely firms’ expectations align with subsequent assessments across key economic variables and how this relationship evolves during the COVID-19 pandemic. The results indicate that, under normal conditions, expectations and assessments remain closely aligned, suggesting a high degree of consistency. However, during the pandemic, the KLD increased across all variables, with statistically significant rises observed for some of them, indicating a widening gap between expectations and assessments. The strongest effects are found for aggregate economic indicators, particularly the general economic situation and, to a lesser extent, firms’ financial standing. On the other hand, employment and inventories exhibit relatively greater stability. Overall, the results demonstrate that the KLD provides a useful distribution-based measure for identifying regime shifts in expectation formation and for assessing the information content of survey data under changing economic conditions.

• Details
• Contents