Heterogeneity in Phillips Curve slopes among members of a monetary union can lead to biased to estimates of the union-wide ‘average’ slope in reduced form regressions. The intuition is that in a monetary union with heterogeneous regional Phillips curve slopes, the central bank, aiming at stabilizing demand shocks, will react stronger to shocks in regions with steep slopes compared to shocks in regions with flat slopes. Using a simple New-Keynesian model of a monetary union we show that when failing to account for this heterogeneity in the estimation, reduced form estimates of the union-wide ‘average’ slope suffer from a sizable bias. Empirically, we show that a similar bias exists in EMU data and slope estimates that adequately control for slope heterogeneity are steeper than those from reduced form OLS regressions.