If you can follow the idea that price setting is a co-ordination game, you can see one of the reasons why monetary policy has real effects. When economists discuss “menu costs” as a major reason for prices not changing, they are primarily thinking of this strategic element (where there are multiple, pareto ranked, stable and state dependent eqm) – even though it isn’t really a traditional menu cost at all, and how it interacts with menu costs is of more interest.
Sure, when we think about the labour market we enjoy talking about the “relative price of labour” and using monetary policy to make “labour relatively cheaper” – in fact a co-ordination game argument can be used here. But the existence of a co-ordination game where the “relative price of goods” is held up too high is also important. As long as prices are “state dependent” monetary policy has traction (even if physically prices are completely flexible!).
I had never heard the daylight savings analogy before, but I like it.
Update: Economist’s View points to Romer’s graduate macro text where this is discussed. It is a useful rundown but:
If there were literally no cost to changing nominal schedules and communicating this information to others, daylight saving time would just cause everyone to do this and would have no effect on real schedules
This statement is true, but it is worth fleshing out. The “communicating” issue here is important methinks – as there are multiple Nash Equilibrium, and there may be uncertainty regarding the value other firms place on different ones. If firms could communicate, or had knowledge about each others payoffs, and if there was a single Pareto superior eqm – firms would turn around and set prices in this way.
There can be zero cost to changing prices – but the fact that we have a co-ordination game with uncertainty implies that we can slip into an inferior equilibrium. Furthermore, even with full information – if the equilibrium aren’t pareto ranked, but can be ranked on the basis of welfare, there can still be a justification for intervention.
I always felt that the fact that monetary policy has scope even in the face of fully flexible prices is something worth recognising.