Why Kydland, Prescott and Sims got the Nobel and why it isn't weird

This is really just a comment on Noah Smith's blog post about the oddity of the Nobel Committee awarding the Economics prize to Kydland and Prescott for finding that the business cycle was caused by technology shocks, and then later to Sims for finding that it was caused by monetary policy shocks.  I tried to post the comment a couple of times, but was obviously fended off by some clever algorithm that filters microfounded macro fans, but poses as a device to stop spamming. [:)]

One reply to this is just to note that the KP prize came first and the Sims prize came second.  By the time Sims' prize came along, the Committee had seen the light, and, had they known what they knew later, they would not have given the prize to KP.  However, this isn't the case.  We knew most of what Sims had told us that got him the prize by about 1995, by which time he'd explained how he would identify monetary policy [and some other] shocks.  Moreover, we knew that the KP claim was literally spurious long before that.  For example, a nice paper by Ingram, Kocherlakota [recent freshwater-basher] and Savin made the point that an economy that produces multiple time series with independent random components cannot be explained by one shock, technology or otherwise.

In his post Noah highlights the KP citation as focusing on 'the driving forces of business cycles' to suggest that the prize wasn't awarded just for methodology, which would rescue the Nobel committee from incoherence.  Reading back those lines, I think that the language leaves enough room for us to translate along the following lines 'KP showed us how to articulate a claim about the dominant causes of business cycles conditional on an entire laboratory economy, something that had not been done before and itself is deserving of the prize;  the substantive part of their claim, that the dominant driver of those cycles was technology shocks, has since been refined, overturned, restored and overturned many times, but it nevertheless now has to be seen as kicking off an ever sharpening conversation about those causes, where echoes of the original claim still bounce around even the very latest papers.  [If you will excuse the outrageous presumption of mentioning my own work in this conversation, see the exchange of papers between Christiano and coauthors and myself and coauthors on 'risk shocks'].  This itself is also deserving of recognition in the Nobel.'  I conjecture that 90 per cent of those in the RBC/DSGE microfounded macro literature would agree with this view of their contribution.  Since the Nobel Committee take extensive soundings - for example they even ask around the central banks who should get it and why - I am sure that they would have heard this view made many times over.

Final point.  Part of Sims' Nobel was no doubt the contribution towards identifying monetary policy shocks, and measuring their impact on real variables, and, therefore, their potential role as drivers of the business cycle.  But although his own substantive work measured significant effects of monetary policy on real variables, this did not mean on its own that they were a significant driver of business cycles.  In fact it's common to find that the contribution of monetary policy shocks themselves is quite small.  One of the bugbears of central bank policymakers [and I remember that Charlie Bean used to make this point whenever such VARs came up in conversation] is that they don't inject monetary policy shocks.  Why should they?  They should be doing good systematic policy.  If they were, Sims and his acolytes would be searching for something that wasn't there.  There is a perfectly reasonable defence of this search:  monetary policymakers get imprecise information;  personnel at the top change for reasons unconnected with the business cycle...  But nevertheless the point remains that although we might measure that monetary policy is not neutral on the real economy, nevertheless it may still not be even a major cause of business cycle movements.  So in principle, awarding Sims the prize in part for measuring those non-neutralities in a rigorous way was not in contradiction with KP's original claim.

A weightier issue to raise might be the fact that the econometric toolkit Sims devised gave rise to ways to agnostically identify technology shocks in VARs.  First came Blanchard and Quah (1989), and then Gali (1999) with 'long run restrictions' [ways of identifying technology shocks that rely on asserting things like 'in the long run, monetary factors should not affect real factors'], and later Faust (1998), Uhlig (1998), Canova and de Nicolo (2002) and Uhlig (2005) with 'sign restrictions' [that rely on asserting things like 'technology shocks that increase output should reduce prices at the same time'].   Depending on your perspective, this collective body of work [all springing from Sims' own] must be read as dramatically narrowing the apparent contribution of technology shocks relative to the original KP claim.  However, I rest on my earlier point, that this conversation would not be going on if it weren't for KP.  Indeed, most of the restrictions in these VAR models appeal to properties of the original KP model or its direct descendants.

There are plenty of reasons to think that Nobel prizes are not going to be perfect.  But I think that giving it to both KP and Sims isn't one of them.