The within-year walkdown of analysts' earnings forecasts has largely been attributed to analysts' incentives to curry favor with managers. We appeal to cognitive psychology literature on motivated reasoning and propose that forecasting difficulty interacts with such incentives to yield the observed walkdown. Higher forecasting difficulty generates a wider range of outcomes from which analysts can justify optimistically biased forecasts. In regression analyses, we find that the interaction between analysts' incentives for optimism and difficulty exhibits the strongest effect on earnings walkdowns. We also examine revenue forecasts as a benchmark of lower forecasting difficulty and find that revenue walkdowns are relatively diminutive. However, when analysts forecast losses, revenue forecasts are more critical and exhibit markedly steeper walkdowns. Our results suggest that analyst forecast walkdowns are better characterized by an interactive effect between analysts' strategic incentives for optimism and forecasting difficulty. JEL Classifications: G17; M41. Data Availability: Data are available from public sources identified in the text.

This study examines the effects that two balanced scorecard framework (BSF) elements, causal linkages between strategic objectives (“causal linkages”) and time delay information (“delays”) in a strategy map, have on long-term profit performance in a dynamic decision-making environment. Using a computer-based simulation task, we conduct a 3 × (4) experiment (control group; causal linkages without delays; causal linkages with delays; four simulation rounds) and find that managers presented with causal linkages without delays generate greater long-term profit compared to a control group. For managers presented with causal linkages with delays, long-term profit generation is higher than the control group, but is not significantly different from the causal linkages without delays treatment. Those managers presented with causal linkages with delays, however, demonstrate learning across the four simulation rounds. In contrast, learning is found to plateau for the causal linkages without delays treatment and is not present for the control group. We also examine the cognitive mechanism through which these two BSF elements impact performance, by measuring the accuracy of two components of managers' mental models. Data Availability: Experimental materials are available upon request from the authors.