This research proposal is finalized to study the macroeconomic implications of monetary and fiscal policy interactions in an monetary model environment where private sector¿s beliefs are subject to bounded rationality schemes. The novelty of our theoretical approach emerges from the joint consideration of i) potentially time/state-varying individuals¿ learning rules and ii) fiscal/monetary policy interacting regimes that may switch according to Markov processes. We apply this model setting to test its consistency with two stylized facts currently debated in the macro literature: i) the one addressing regime-switching fiscal policy effectiveness and non-linear trend-cycle relations; ii) the one addressing asset price volatility and the emergence of financial bubbles. In the fist place, we estimate Markov-Switching Structural Vector Auto-Regressions (MS-SVARs) with novel Bayesian methods to produce fresh empirical evidence supporting the idea of a state-dependent effectiveness of both fiscal and monetary policy. Such a regime-switching effectiveness is evaluated conditional on set-identified sources of macroeconomic variability. In the second place, by taking appropriate extensions of the monetary model, we then calibrate its structure to replicate the non-linear dynamics emerging from the MS-SVARs. Calibrations are obtained by a modified impulse response matching strategy which is able to take into account regime changes. Technically, this requires using using a non linear numerical solution routine for rational expectations models jointly considering the calibration of both the parameters space and of the transition probability matrix for the Markov-switching model. The recent research showing substantial macroeconomic non-linearity and structural breaks in time series data provide an encouraging starting point for the results expected from this project