Do First Impressions Last? The Impact of Initial Assessments and Subsequent Performance on Promotion Decisions

Abstract

This paper examines whether and for how long managers’ initial assessments of employee ability influence promotion decisions. Using archival data from minor league professional baseball, we find that, controlling for performance, initial assessments are associated with promotion decisions for at least six years after the initial assessment was made. We also find that initial assessments are positively associated with future performance at the outset of a player’s career, but the association becomes insignificant after a player accumulates on-the-job experience. We show that the weight on initial assessments for promotion decisions declines as additional on-the-job performance signals are observed, reflecting the declining relative informativeness of initial assessments about future ability. We construct a proxy for relative informativeness based on coefficients from regressions of future performance on initial assessments and observed performance. When we compare the implied relative weight on initial assessments for promotion decisions to our proxy for relative informativeness, we find assessments receive greater relative weight than implied by informativeness overall and across experience and job-level partitions. Our results suggest managers update initial beliefs about worker ability slowly given available performance measures.

Keywords: Promotions, Performance Evaluation, Employer Learning, Dynamic, Ability

JEL Classification: G30, M40, M46, M51, M54