When at the zero bound for interest rates, the central bank can lower the real interest rate by raising the inflation target. Some have argued that this is politically unacceptable, and hence central banks may get stuck in what Paul Krugman called an “expectations trap”, an inability to convince markets that it intended to be “irresponsible”, i.e. unwilling to allow higher inflation.

Some advocates of this view advocate fiscal stimulus as an alternative method of boosting aggregate demand. But why would a central bank that is unwilling to raise its inflation target allow fiscal stimulus to boost the inflation rate? One answer is that perhaps they are willing to tolerate higher inflation, but not willing to announce a policy of higher inflation. In that case, fiscal stimulus might be a way of boosting aggregate demand.

But fiscal stimulus is not the only way of raising inflation, even at the zero bound. To see why, let’s go back to 2009, when inflation was roughly zero, i.e. about 2% below the Fed’s implicit target. Also assume that in order for the economy to recover quickly from the recession it was necessary to make up for the 2% inflation undershoot with an equal overshoot of the inflation target, perhaps spread over 5 years. (In fact, inflation continued to mostly undershoot 2% throughout the 2010s.)

How can the Fed achieve an overshoot if it’s politically infeasible to raise the inflation target from 2%? The answer is simple, stop targeting inflation and switch to a price level target. Promise that the average inflation rate during 2008-13 or 2008-18 will be 2%. Because inflation was zero in 2009, that means above 2% inflation in the post-2009 years. That policy would also lead to substantially lower real interest rates, compared to the policy the Fed actually implemented. There is no zero lower bound for real interest rates.

One argument against my proposal is that implementing the expansionary policy would require tools that are politically controversial.  Hence the Fed could not achieve its average inflation target over five or ten years, even if it wished to.  But that argument is wrong, as it confuses cause and effect. The worry about policy effectiveness is based on the mistaken assumption that the greater the QE the more expansionary the monetary policy.  Thus the worry that an effective QE policy might be too large, too controversial.  But this view is wrong for reasons analogous to assuming that low interest rates are easy money.  It’s reasoning from a quantity change.

In fact, QE programs are mostly endogenous.  The amount of QE done by central banks around the world is negatively correlated with inflation.  Apart from cases of hyperinflation, the biggest QE programs typically occur in countries with the lowest inflation rates, such as Japan and Switzerland.  Thus central banks should never refrain from a policy of boosting inflation because they fear it would lead to a larger balance sheet, which is politically controversial.  Just the opposite is true.  It is low inflation policies that lead to bigger central bank balance sheets as a share of GDP.

Although it may sound counterintuitive, a central bank that says “we will do as much QE as necessary to raise inflation to target over 5 years” will generally end up doing far less QE than a central bank that says “we will not do as much QE as necessary to raise inflation up to our target.”  Without an expansionary monetary policy, inflation and interest rates will fall to ultra-low levels, forcing central banks to do large amounts of QE.  Compare the Eurozone and the US post-2013.

If we want to achieve a robust recovery from a demand side recession, we must allow inflation to overshoot the target, in order to make up for below target inflation during the recession.  That means a successful policy requires something like average inflation targeting or level targeting.  But even those policies are not enough if the central bank is unwilling to do whatever it takes to achieve the target.

One reason I prefer price level targeting to average inflation targeting is that a price level target commitment is more precise.  Because it is more precise, there is a greater degree of central bank embarrassment when they fall short of the target.  Average inflation targeting is just vague enough to allow plausible excuses if they fall short.  In order to have an effective monetary policy, it is essential that central banks become embarrassed when they fail to hit their target.  Fear of embarrassment spurs them to do whatever it takes to hit the target.  Ironically, if markets believe the central bank is sincerely willing to do whatever it takes, then it can get by doing much less than otherwise.  Thus Australia avoided recession in 2009, despite not instituting a QE program and not cutting interest rates to zero.

Elsewhere I’ve argued that the public (and especially pundits) can help to make Fed policy more effective if they treat the Fed’s vague average inflation target as a firm commitment to push the PCE price level to 135 in January 2030.  We can all do our part to make monetary policy more effective.

Although it might seem that we are being cruel by threatening Fed officials with embarrassment if they fail to achieve a 135 price level in 2030, we are actually doing them a favor.  Pressure from society will make their promises more credible, and (counterintuitively) they can then hit those same targets with less effort, less QE.