The Fed changes its inflation policy

Earlier this year, the US central bank, the Federal Reserve or Fed, changed its monetary policy framework. Previously, the Fed has had an inflation target of 2%. It also targeted maximum employment consistent with the price target – the so-called ‘dual mandate’. The new policy is one of average inflation targeting (AIT) and deviations from maximum employment. This article will look purely at the inflation side of the policy[1].

‘Pure’ or orthodox inflation targeting, as practiced by most central banks including the Bank of England, sets an annual target for inflation. Whether that target is achieved or not is only relevant for that specific year, and in effect each year is treated as a separate target period; what happened in the past is water under the bridge. By contrast, AIT looks at the target from a longer perspective. The goal is to reach a certain average over a period of time, be it the business cycle or some other measure. Under AIT, the Fed would seek to compensate for an over- or undershoot in one year by adjusting its policy for the rest of the target period.

In theory, this makes a lot of sense. Much of economic policy plays out over longer time periods than a single year, and some central banks (eg the Reserve Bank of Australia) already practice AIT. It certainly increases the central bank’s policy flexibility.

But the policy also has some weaknesses. The first, and perhaps most immediate, problem is deciding what the target period is. It is essential for this to be known, and known in advance, or else inflation and so the success of the central bank’s policy cannot be measured against it. For instance, if the central bank aims to reach its inflation target over a period, when does it start the period and when does it stop it? How long is the relevant period and is it susceptible to manipulation – ie, can the bank extend the period to give itself more time to attain its target? Would this not be akin to someone marking their own examination papers?

As many governments are finding in the very different challenge of combatting the pandemic, clarity of policy is important for establishing public confidence in the authorities and their actions.  The Fed may yet come to regret ‘muddying the waters’ in this way.

The second issue is one of credibility. The credibility of the Fed in its targeting of inflation is anyway not high at the moment:  if we define ‘meeting your target’ as a period of twelve consecutive months during which inflation is at or near the target, then the Fed has undershot its target every year since early 2012.  It is far from alone in this, and indeed far from the most unsuccessful central bank, as the Bank of Japan has not attained its target more than once or twice in the last 30 years! But it does not bode well for its new policy of AIT, the main effect of which will be to raise the immediate target above 2% to compensate for the years of undershooting.

What makes Fed policymakers so sure that they can reach a new, higher target going forward when for the last 8 years they have failed even to achieve 2% inflation? And if they cannot, is not the main effect of the change to AIT to increase the degree by which they fail?

Next, AIT implies a commitment to compensate for over- or undershooting the target over the extent of the target period. However, to be credible, this assumes that one central bank leadership can commit its successors to a certain policy (Fed chairmen serve a four year term, though it is renewable). History shows that this is very unlikely to work for longer than a fairly short period.

In practice, what is more likely to happen is that deviations from the target will be accepted and ignored, provided they are in the desired direction. Which this direction is, will vary according to circumstances at the time. Currently, inflation almost everywhere is undershooting central banks’ targets. Hence, deviations above target will be accepted, and, it is safe to say, not be compensated for. Potentially, they can be presented as compensation for current and previous undershooting. By contrast, continued undershooting will be seen as an incentive to increase the efforts to reach the target.

The upshot of this is that – always assuming that the Fed can reach its new target – inflation going forward will be higher than we have seen for the past 10-15 years. But how much higher? Markets generally seem to feel that the Fed’s new target equals 2.5%, in other words, somewhat higher than the previous 2% target but not unduly so. This could be dangerously complacent.

Consider: As noted above, the Fed has actually on the whole undershot its inflation target since April 2012. If we compare the actual level of the PCE index with the one that would have been reached if inflation since then had constantly been on target, the cumulative shortfall is 5.5%. (In other words, the actual PCE index was 111.1 in July; the ‘ideal’  one would have been 117.4). So there is about 6% of inflation foregone to compensate for.

How long would this take? That is of course anyone’s guess. (For that matter, so too is the 2012 starting point). But the compensation period cannot be too long. First, because if it is, there can be other events that throw it off its tracks. Second because it involves one Federal Open Market Committee attempting to bind its successors. Third, because the membership of the FOMC and, crucially, its priorities may well change, in fact some of it must change over the period. And fourth, because the longer inflation remains at any one level, the more that level becomes entrenched as the new normal.

All this means that any compensatory overshoot of inflation would have to take place over a period probably not exceeding four or at most five years. And therefore, with a cumulative shortfall of 6% to make up for the Fed would have to aim at an annual average over this period of more like 3.5%!  And this starts to become non-negligible (ie, noticeable by the man or woman in the street) – with the distinct added risk that once one reaches 3.5% inflation, it is not stable and can quite rapidly become 5%, 8% and higher as employees seek higher wages in compensation.

Nor will the fact that economies are likely to have considerable spare capacity and high unemployment necessarily restrain this:  the 1970s showed that it is quite possible to have high inflation even if the economy is not working at full capacity.

With all these potential downsides, we have to ask, why has the Fed made this move?  Well, apart from the fact that it gives them more flexibility (though at the price of less clarity of policy), we think there is another rather different reason.  And that is that it is actually quite likely that inflation will start to increase over the next year or two, whatever the Fed does or says.

As Milton Friedman famously said, “Inflation, over any sustained period of time. is always and everywhere a monetary phenomenon”. The huge monetary stimulus of the last six months has pushed US money growth to its highest peace-time rate ever[2], and the renewed flare-up of the pandemic implies that this rate will not slow anytime soon. Based on past experience, the rapid monetary growth should be enough – always assuming even a modest economic recovery as well – to produce inflation that reaches and indeed overshoots the Fed’s inflation target.

So this is we think what is behind the Fed’s new policy.  They can see inflation coming, whatever they do, and would like us to believe that it is under their control and all part of their master plan.  Nor is this exclusively an American issue: at a conference recently the Governor of the Banque de France, M. Villeroy de Galhau, whose voice is influential at the European Central Bank, discussed the ECB’s ongoing strategic review along very similar lines, including the comment that “We might be ready to accept inflation higher than 2% for some time”.

Central bankers are warning us that inflation is coming. They would like us to believe that this is by design, that they have not only permitted it but engineered it.  But they are likely to be as unable to influence, let alone control the higher inflation of 2021-23 as they were the abnormally low inflation of the last decade.


This is an abridged and edited version of two articles that appeared first on the author’s website:


[1]              This is not to say that the employment target is flawless. For instance, ‘maximum (or full) employment’ is not a fixed number but varies substantially over time, while employment is affected by a number of factors that are not always possible to influence by monetary policy. All this adds extra moving parts to what the Fed is trying to do.

[2]              “Money” here refers to broad money; specifically, the equivalent of the (discontinued but still relevant) M3 measure.