That’s the title of a new working paper that you can find here. The topic is closely related to a previous post that I wrote here at the Sound Money Project. In that post, I questioned the validity of NGDP targeting. Surely, there is no such thing as the perfect nominal indicator, however, as long as different variables have different imperfections, it is wise to follow more than one rather than focus just on the NGDP. In this latest draft, I deal with a slightly different question. Looking at indicators like the Taylor Rule or productivity changes, there are signs that monetary policy from 2002-2007 was too loose for too long. Furthermore, there is a consensus that monetary policy was too expansive in the years before the subprime crisis (where did the housing bubble come from?). Some Market Monetarists and other defenders of an NGDP rule argue that the crisis was due to a fall in NGDP that the Federal Reserve failed to offset and not a rapid growth the years before the crisis, as the Taylor Rule suggests. Christensen (2011, p. 17, emphasis added) quotes the following passage by Wolsey (2011) to describe the position of Market Monetarism: “In my view, the reason for the Great Recession is simple – the Federal Reserve allowed an excess demand for money to develop, so GDP fell 14% below its growth path from the Great Moderation.” Can both positions be right? That monetary policy was too loose for too long before 2008 or that the problem was a simple fall that the Fed did not offset? If we look at the NGDP (and final sales to domestic purchases –FSDP), we see two series that grow at a fairly constant rate (approximately 5% per year). However, if we look at deviations from trend, deviations before and after the crisis can be observed. If deviation from trend matter, then ex ante both might be important to understand the crisis. Namely, there wasn’t a crisis because NGDP fell, but rather NGDP fell because there was a crisis. The Fed’s failure to offset the fall made things worse, but that might not have been the initial problem. Note that both deviations before and after 2008 have similar magnitudes. The difference is that the upward deviation before 2008 took place in around 3 years, while the downward deviation after 2008 is the sudden fall that took place in the 2008 crisis. This is also why in figure 1 only the second deviation is observable “just by looking.” If we were to observe the Gross Output (GO), which more closely tracks all of the transactions in the economy, then we would notice that the same pattern is observable but with trend deviations that are twice as big.