Abstract:
There is nothing misleading in the fact that different filtering techniques lead to different facts about macroeconomic time series. The fact that economists use a large number of filters to extract the 'cyclical' and 'trend' components of time series simply means that these concepts do not have unique meaning among them. Alternative filters provide different windows through which economists can examine their models and data. It is an open question as to whether some of these windows are more or less interesting to look through. The fact that some economists restrict themselves to a small set of filters is an issue to the extent that they thereby induce a lack of power. Here, I argue that a commonly used method of testing business cycle models induces no such lack of power.
