Abstract
Purpose - The purpose of this paper is to estimate the strength of the relationship between job performance and intentions to quit (ITQ), identify moderators to this relationship, and calculate the direct and indirect effects that job performance has on ITQ and turnover. Design/methodology/approach - Data from 65 studies (n=17,918) were meta-analyzed to estimate the performance-ITQ relationship. This overall sample was separated into subgroups for moderator analyses. Meta-analytic path analysis was used to test the hypothesized model of turnover. Findings - Supervisor ratings of performance had the strongest relationship with ITQ (ρ=-0.16), followed by self-ratings (ρ=-0.14), and objective measures (ρ=-0.02). Employee nationality and job type also acted as moderators. Poor performers are more likely to quit even after controlling for job satisfaction and turnover intentions, indicating that they are more apt to engage in unplanned quitting. Good performers were slightly more likely to intend to quit after controlling for job satisfaction. Research limitations/implications - Limitations on the number and type of studies available prevented a test of how performance level acts as a moderator to the job performance-turnover relationship and may cause some of the moderator analyses to be unstable. Practical implications - The findings provide for a better understanding of how employees' job performance affects their turnover decisions and how organizations can control turnover. Originality/value - This is the first meta-analysis to estimate the relationship between performance and ITQ and to test a meta-analytic path model of the job performance-job satisfaction-ITQ-turnover relationships.