Biased Returns to Tenure: The Impact of Firm-Specific Shocks on Base and Non-base Earnings

This study examines returns to tenure using Mincer wage regressions and longitudinal employer-employee payroll data from Great Britain. We find a pervasive downward bias in estimates of returns to tenure that rely solely on match fixed effects to control for unobserved factors influencing wages and tenure. This bias stems from the co-movement of average wages and tenure within firms, as theorised and empirically shown by Snell et al. (2018). By addressing this bias with firm-year fixed effects, we find that tenure-wage profiles increase by up to 20% in Britain's largest private-sector employers. Further analysis reveals that the bias primarily originates from non-base earnings (e.g., overtime). These findings underscore the need for caution when interpreting tenure returns from wage regressions that omit firm-year fixed effects, particularly in samples where non-base earnings are present; even if base earnings are sticky, firms may adjust other earnings components in response to shocks that influence employment levels.