Here's a good paragraph from Brad DeLong about why we're not getting more oomph out of monetary policy (emphasis added):
The Federal Reserve's announcement last December that it was switching from a time- to a state-based policy rule has not REPEAT NOT raised expectations of inflation over the next five, ten, or thirty years. Perhaps this is because Federal Reserve communicators have spent a lot of time telling people that the shift does not mean that the Federal Reserve will tolerate higher inflation. Perhaps it is for other reasons.
It is followed by a much less good paragraph about what this teaches us about the possible efficacy of monetary policy:
This is a powerful empirical piece of evidence that it is much harder to summon the Inflation Expectations Imp than economists like Greg Mankiw had thought. It is a point for the expansionary fiscalists in their debate with the expansionary monetarists.
We have empirical evidence of something here, but how can you read the italicized clause and conclude that it's evidence of the difficulty of increasing inflation expectations? The Federal Reserve is telling people, explicitly, that they should not expect a higher level of inflation over the five-, 10-, or 30-year time horizon. To know whether it's hard to raise inflation expectations, we would want to see what happens if the Fed told people they should expect a higher level of inflation over the five-, 10-, or 30-year time horizon.
Insofar as we have evidence of anything here, it seems to me that we have evidence that the Federal Reserve thinks it would be very easy to increase long-term inflation expectations, and that's why they really want to be clear that the long-term inflation target hasn't changed.
The fact is that we just don't have very good empirical evidence in this field. The theory behind monetary dominance seems strong to me; if you forced me to choose, that's the path I would choose. As a practical policymaker, I would try to hedge and do both. But as someone interested in finding empirical evidence, the first step would be for the Federal Reserve to start telling people they should expect higher inflation. As long as they're telling people to not expect higher inflation, we won't know very much beyond the fact that the Evans Rule seems to have been adequate to keep growth more or less on track despite substantial fiscal anti-stimulus.