Archival or perceived measures of environmental uncertainty? Conceptualization and new empirical evidence

•We compare archival (AEU) and perceptual (PEU) data on Environmental Uncertainty.•Result: longitudinal AEU and PEU correlate just moderately but significantly.•Causes of differences: decision unit, predictability, use of leading indicators.•Conclusion: AEU and PEU are proxies for each other but not perfect substitutes.•Data: archival and survey from 60 large, listed German companies/top executives.

Environmental uncertainty is a major determinant in many managerial decisions and evaluations. Yet, there is no consensus what constitutes a valid measure of uncertainty to which executives can respond. This paper explicates conceptual and methodological differences between Archival Environmental Uncertainty (AEU) and Perceived Environmental Uncertainty (PEU).Conceptually, we discuss the controversial development of these measures in the literature. We propose a framework showing that AEU and PEU differ due to the specificity of the decision unit, the predictability of volatility, and the use of leading indicators.Empirically, we are the first to investigate the statistical association between prevailing measures of AEU and PEU. Our analysis combines archival data on AEU (annual reports) with survey data on PEU from top executives of the 110 largest listed German companies (55% response). As predicted, AEU and PEU correlate moderately on a significant level (r = 0.257 to 0.374). Also, AEU indicators (mainly volatility in sales and earnings) explain over 26% of the PEU measure. Yet, adjustment of the AEU measure for predictable volatility does not improve its relationship to PEU. Overall, our findings imply that AEU and PEU are not perfect substitutes, but valid proxies for each other if the relevant limitations are considered.