Paul says

I re-estimated the same kind of rule (with equal coefficients on unemployment and core inflation) in 2010, and got somewhat different coefficients; here’s what the picture looks like with the updated numbers:

Not so close to exit from the liquidity trap, is it?

Some readers may already be asking whether it makes sense to say that the newer version of the rule is “right”; hold that thought for a minute.

Next, as Ryan Avent says, it’s far from clear that the numbers will actually move us closer to the crossover point in the near future. Core inflation, after rising for much of 2011, seems to be heading back down; unemployment may not fall much even if job creation is pretty good, because discouraged workers will come back into the work force.

But the main point is that using historical estimates of the Taylor rule is not a good way either to predict Fed policy or to recommend Fed policy at this point in our history.

Remember, a Taylor rule estimated on historical data is in effect an estimate of how the Fed thinks, not what it should be doing. So if your Taylor rule calls for a positive interest rate, it’s saying that this is what the Fed would do if it still thought the way it did on average over the period 1988-2008

I’m not sure that’s really correct.

The underlying presumption is that the Fed did a pretty good job at stabilizing the macro-economy in 1990s. This means whatever process they used for determining the Funds rate, whether it was smart analysis or throwing bones, stabilized interest rates in the vicinity of the natural rate.

If we then go back and say, well we think in fact the natural rate is a function of unemployment and inflation and then we say well what did they do relative to unemployment and inflation then we have a function that spits out an interest rate which stabilizes the economy around the natural rate.

Now, you might not want to do that. Indeed, I don’t think you want to do that. I think you need to run below the natural rate to catch-up lost growth. However, you could use this formula as a rough and ready guide to where the natural rate might be.

Thus suggesting that we the natural rate may be about to rise above zero – which is what Greg suggested in his original post.