The Taylor rule is widely seen as a good summary of what the Federal Reserve does. Though the rule cannot easily be fitted to actual data as subsequently revised, at least for a full postwar sample, it can be fitted to real-time data (i.e., data as seen at the time), as shown by earlier work by Orphanides (2003). But in practice the Fed’s Federal Open Market Committee (FOMC), if it is using a Taylor rule, will look at its own forecasts or projections. Orphanides and Wieland (2008) examine whether a Taylor rule can be fitted to the FOMC’s own projections since 1988. They find that it can with appropriate parameters that satisfy the Taylor principle - that is, that give a unique stable solution under rational expectations. Furthermore, they find that the rule works better with these projections and resolves various puzzles regarding the data on outcomes.