The outbreak of the financial crisis of 2007 has generated a lively debate on the real or alleged faults of the Federal Reserve (Fed). Some economists argue that in the period 2002-2005 the U.S. central bank has taken its target interest rate below the level implied by monetary principles that had been followed for the previous 20 years. One can characterize this decision as a deviation from a policy rule such as a Taylor rule. This behavior determined the end of the Great Moderation and gave birth to the Great Recession. In this paper I challenge this view. I show how the deviations from the Taylor-rule's hypothetical interest rate can be explained by the ambiguity on inflation indicators to use. I also explain how the Great Deviation was instead caused by an error in the estimate of one of the fundamental components of the Taylor rule, i.e. the natural rate of interest. Too expansionary monetary policy of the Fed was therefore not due to discretionary choices, but to a structural problem of the Taylor rule. Finally, I show how an adaptive rule based only on observable variables would have avoided the huge gap between short-term rates and natural rates