Back-of-the-envelope low-quality musings: #takes

Longform stuff I've put time and thought into is at #effortposts.

To hear from academics and policymakers, #interviews.

Compiled scribbles on what I read can be found in #notes.

Topic-related tags include #economics, #coronavirus and #miscellaneous.

# The Fickle Finger of FAIT

12 Jun 2021

Yeah, I don’t know what to say
Since a twist of FAIT when it all broke down
And the story of U.S. looks a lot like a tragedy now
- Taylor Swift in The Story of Us (maybe)

On Thursday, the Bureau of Labour Statistics released its monthly inflation data. Prices as measured by the Consumer Price Index jumped by 5% in May compared to the same time last year. Some of this can be accounted for by a base effect, with low prices last May making this increase look larger. But even stripping away base effects and focusing on core CPI, which excludes variable food and energy prices, saw a 2.6% inflation rate that landed above the Federal Reserve’s 2% target.

The important question is whether this represents a transitory phenomenon as the economy comes to life and supply constraints bottleneck the economy, or whether it is a symptom of a more persistent trend. The latter would be concerning, with Larry Summers arguing that “overheating is now the largest risk in the near term US macroeconomic outlook”. One possible reason for this danger is the Fed’s new monetary policy framework.

Previously, it had a Flexible Inflation Targeting regime, aiming for 2% inflation every year while giving itself room to be flexible. The new approach is called Flexible Average Inflation Targeting. The main difference is that it looks to average an annual inflation rate of 2% across a few years. So if it misses the 2% in any year, it has to makeup for that instead of simply letting bygones be bygones as it used to. Some economists view this change as being dangerous, and there are three main lines of criticism levelled against FAIT.

1. The secular decline in $r^\ast$ which made FAIT attractive is less important now due to expansionary fiscal policy.
2. The alternative of a higher inflation target dominates FAIT as a solution.
3. FAIT is difficult to implement in practice and may unanchor inflation expectations.

Let’s address each in turn: whether the problem exists, whether there are better solutions and whether the consequences of FAIT are positive. The reason FAIT was adopted is the fall of the natural rate of interest $r^\ast$ from 3% at the turn of the millenium to under 1% now. This is the interest rate expected to prevail when the economy is at full output and inflation is stable.

The Fed tries to set the interest rate above $r^\ast$ if it wants to cool the economy and below $r^\ast$ if it wants the economy to expand. The way it does this is by setting a nominal interest rate $i$, which equals the real interest rate $r$ plus inflation $\pi$. Since it is difficult for the Fed to set $i<0$ without banks just taking out their money, they face a zero lower bound on the nominal interest rate, which means the real interest rate is at minimum $-\pi$. In that case, a lower $r^\ast$ means they have less room between $r^\ast$ and $-\pi$ to be expansionary in a recession.

One view is that the Biden administration’s fiscal policy have driven up $r^\ast$ by making loanable funds scarcer. I think this is not a good reason to assume that the $r^\ast$ problem has been taken care of, given many other causes of a low $r^\ast$ remain. Slow population growth and the low capital-intensity of our modern economy keep the demand for investment weak, while aging populations and increased income inequality continue to provide a large supply of savings. A one-off infrastructure plan will not resolve these issues, so it is likely this constraint remains. Even if it doesn’t, the design of a monetary regime should be aimed at adaptability if $r^\ast$ changes in the future, rather than just being for the here and now.

If $r^\ast$ is expected to stay low, the response of monetary policy to demand-side shocks is almost always involve raising inflation expectations one way or another, since that lets real interest rates fall further. Under FIT, the Fed raised inflation expectations by forward guidance, promising to keep interest rates lower-for-longer and to accept inflation above its 2% target for a while afterwards. In theory, this should incentivise people to spend now. With FAIT, the exact same effect occurs, except that the promise of future inflation above 2% is not a promise to deviate from its rules but to follow those rules. Output and inflation fall together in demand-side recessions, and so inflation must be made up for in the future, ensuring these higher inflation expectations are a FAIT accompli.

A proposed alternative is raising the inflation target. By the same logic as before, this should give the Fed more room to maneuver. Unfortunately, this forces the economy to incur the costs of higher inflation at all points in time, in return for a gain during a limited post-recession period. By contrast, FAIT implies those costs are only imposed temporarily, with average inflation staying at 2%. This change would also mark a significant shift for the Fed, which has expended a huge amount of effort since the Volcker disinflation to anchor the 2% figure into people’s minds. FAIT does allow occasional deviations from 2%, just as FIT does - but the overarching number to keep in mind is still 2%. By contrast, if people thought the Fed were willing to change its target willy nilly at the first signs of monetary weakness and raise the average rate of inflation, inflation expectations would be much less anchored. And a look at the various market, forecast and survey measures of inflation since FAIT show that expectations remain anchored at 2%.

Finally, two difficulties of controlling FAIT in practice have been noted. The first is that the window within which inflation must average 2% is vague, making it less clear when the Fed will eventually clamp down on inflation. The second is the idea of time inconsistency. FAIT relies on overshooting inflation in the future. Thus the Fed would have an incentive to promise higher inflation during the recession, but not allow inflation to rise once the recession were over and it had already accrued the benefits of higher expected inflation. If consumers and firms thought this would happen, they would not believe the Fed’s promise of higher inflation to be credible and so the policy would fail to begin with.

These are valid concerns. However, any sort of flexible monetary policy which deploys forward guidance faces both of these issues. Under FIT, the Fed looked through and ignored certain inflationary periods, allowing inflation to rise above 2% if it thought inflation came from a transitory shock where cooling the economy would cost more in output than gained in lower inflation. The way it managed this was by a clear communications strategy, and this can be deployed to ensure people understand FAIT. As for time inconsistency, the incentive to keep its promises is actually larger with FAIT, since not doing so would contradict its own monetary framework and incur a cost to its reputation. This makes it more believable than with FIT, where the monetary framework would imply controlling inflation afterwards. Because credibility is currency that central banks trade on, consumers and firms are more likely to believe the Fed will let inflation go over 2% when they know it has a reputation to protect.

This is not to mention the other advantages available from FAIT, compared to both FIT and FIT with a higher inflation target. For example, FIT suffers from the base drift phenomenon. If the Fed undershoots its 2% target for a number of years, the path of the price level would be lower than if it had met the annual targets. This is because FIT requires the Fed not to makeup for past errors. By contrast, FAIT would not face this issue, and so people planning for the future would have a more reliable sense of the price level across a long time span.

FAIT would also let monetary policymakers fix mistakes. The reality is that the abstract values of $r^\ast$ and “full output” are impossible to perceive for certain. Where FIT bakes in past mistakes, FAIT compels policymakers to make up for them. So if they underestimate the level of full output and keep the economy at some degree of slack, they have to correct for the associated failure to bring inflation to 2%, rather than simply being able to sidestep those errors.

The construction of monetary policy needs to be robust to a variety of circumstances while still being easy to communicate. Although there are undoubtedly teething problems in any regime change, the new Fed framework is one that will keep long-run inflation expectations well-anchored while giving it more punching power in recessions. As such, we should feel comfortable leaving our fate in FAIT.