If a random firm were to increase its wages, would that decrease the firm’s churning (“excessive” worker reallocation)? Although the trade-off between wage and churning costs has received attention in both the labour and HRM literatures, there seems to be no evidence about the causal impact of wages upon churning. This paper seeks to fill that gap by considering detailed Portuguese matched employer-employee panel data and different identification methods. After presenting comprehensive evidence about job and worker flows and churning, we find that even models based on within-firm time differences do still generate the negative association between wages and turnover found in most research. However, that result no longer holds when we consider instrumental variables based on minimum wages determined by collective bargaining arrangements. One possible interpretation of our finding is that workers’ effort may not be sufficiently sensitive to wages: employers may replace workers priced out of the labour market with more skilled individuals, so that churning does not fall.