In this paper we use a time varying parameter vector autoregression (TVP-VAR) to assess the importance of the learning component in the US postwar economy. The random coefficients are assumed to follow a mean reverting process around an unconditional mean that can be interpreted as the estimates of the coefficients from the reduced form of a rational expectation equilibrium model. The deviations from the unconditional mean are attributed to learning of the agents about the value of the coefficients which regulate the economy. We estimate a monetary model for the post WWII U.S. economy including inflation, output growth and the federal funds rate. We document the presence of learning dynamics and find that the importance of the learning mechanism is somewhat limited for real activity but it is substantial in explaining the dynamics of inflation and interest rate.