Inflation dynamics, as well as its interaction with unemployment, have been puzzling since the Global Financial Crisis (GFC). In this empirical paper, we use multivariate, possibly time-varying, time-series models and show that changes in shocks are a more salient feature of the data than changes in coefficients. Hence, the GFC did not break the Phillips curve. By estimating variations of a regime-switching model, we show that allowing for regime switching solely in coefficients of the policy rule would maximize the fit. Additionally, using a data-rich reduced-form model we compute conditional forecast scenarios. We show that financial and external variables have the highest forecasting power for inflation and unemployment, post-GFC.