In recent years, this model (depending on which inflation measure is used) has signaled the need for a negative federal funds rate. At the worst of the crisis the Taylor Rule said rates should have been negative 5% or 6%. And some versions of the model signal the need for negative rates right now. This has been the impetus behind Quantitative Easing (QE).  If interest rates should be negative, but clearly can’t be so, then some other form of easing is necessary, right?

The problem with this is that the Taylor Rule can be, and in our opinion has been, wrong.

In 1993, at virtually the same time John Taylor was building his rule, Brian Wesbury discovered that using a simple two-year annualized rate of growth of nominal GDP provided a robust target for the federal funds rate. This model explained the inflation of the 1970s and the disinflation of the 1980s and 1990s. If the Fed holds rates below nominal GDP it is too loose. If it holds rates above nominal GDP it is too tight.

Back in 2004, this nominal GDP model showed that Alan Greenspan’s policy of 1% interest rates was inappropriate. The model said interest rates should have never fallen below 4% or so. No wonder there was overinvestment in housing. Interest rates were misleadingly low – they fooled people into thinking credit was cheaper than it really was.

There is also a good article in today's Wall Street Journal (Fed's Low Interest Rates Crack Retirees' Nest Eggs) about how low interest rates are hammering retirees whose incomes have dropped dramatically on their savings.  If Wesbury and Stein are right, these people will soon be hit with a second whammy, higher inflation, which will make their savings worth less.