The Fed, on top of promoting maximum employment, is in the business of setting expectations and anchoring rates. Setting interest rates expectations is a tricky endeavour but a very important one. Anchoring rates is necessary, especially in the presence of inflationary inertia. Japan and Argentina on either spectrum come to mind.
The prospect of high inflation in the US still strikes fear in the hearts of economists. In October 1979, the Fed under Paul Volcker faced inflation that was around 10%, with every indication of climbing higher. The Fed’s policy at that time failed to stimulate growth or made no improvements in employment. Doubt about the Bank’s policies was high and it took the Fed more than ten years to restore its credibility and bring inflation and inflation expectations down to the magic number, 2%.
However, this is not the story of our times.
We are at historically lowest rates, the inflation in the Fed’s preferred measure has averaged only 1.5% over the last decade. Alongside with high unemployment and poor growth, we can expect the Fed to commit to “low for longer” (LFL) policy rates, close to zero, and a large balance sheet for some time to come.
“In another moment Alice was through the glass… Then she began looking about, and noticed that… all the rest was as different as possible.”
– Through the Looking Glass, and What Alice Found There, by Lewis Carroll.
Having travelled to the other side of Alice’s looking glass, the Fed now has to be wary of deflationary pressures.
The dangers of deflation include:
- falling prices and wages,
- people will put off buying,
- negative rates policy is very hard to implement should it be necessary,
- it’s not good for debtors as even though prices and wages fall the value of your debt does not – just to list a few.
Therefore, it’s important that we have policies that help mitigate the adverse effects.
For the Fed, declining interest rates would mean less scope to cut interest rates to boost employment during downturns. In Chair Powell’s latest speech, he said, “We have seen this adverse dynamic play out in other major economies around the world and have learned that once it sets in, it can be very difficult to overcome. We want to do what we can to prevent such a dynamic from happening here.”
Inflation targeting policies work through the public’s inflation expectations. Two main measures of inflation expectations are the survey-based and market-derived (e.g. breakeven inflation expectation) measures. A recent Fed note by Anthony Diercks and Isfar Munir found that divergent signals can arise out of these two measures. So how would one interpret such divergent signals? They found that during the last 15 years, predictions based on market quotes tend to provide better forecasts of the Fed funds rate.
(Breakeven inflation is defined as the difference between interest rates on nominal Treasuries and TIPS, equal to headline CPI inflation expectations plus an inflation risk premium (which could in theory be positive or negative) minus the TIPS liquidity premium. Since the Fed’s 2% target is in PCE terms, a target-consistent level for breakevens would be 2% + CPI-PCE gap + inflation risk premium – TIPS liquidity premium.)
The idea of inflation targeting is very easy, it is to create a nominal anchor for monetary policy as evidence of the Fed’s commitment to price stability. New Zealanders were the first to adopt inflation targeting in 1990, followed by Canada in 1991, the UK in 1992 and Sweden and Finland in 1993. Inflation targeting was implemented after previous inflation controlling measures had failed (e.g. Canada and NZ) or after the removal of policies such as pegged exchange rates (e.g. Sweden, UK and Finland).
In “Twenty-five Years of Inflation Targeting in Australia: Are There Better Alternatives for the Next 25 Years?”, McKibbin and Panton write that inflation targeting strictly speaking involves achieving and maintaining low and stable inflation, with a base drift, without consideration for controlling deviations in the output level. Shocks that affect price stability, whether temporarily or permanently are accommodated by changes to the policy rates.
According to Frederick S. Mishkin, inflation targeting is a monetary policy strategy that encompasses five main elements:
1) the public announcement of medium-term numerical targets for inflation;
2) an institutional commitment to price stability as the primary goal of monetary policy, to which other goals are subordinated;
3) an information inclusive strategy in which many variables, and not just monetary aggregates or the exchange rate, are used for deciding the setting of policy instruments;
4) increased transparency of the monetary policy strategy through communication with the public and the markets about the plans, objectives, and decisions of the monetary authorities; and
5) increased accountability of the central bank for attaining its inflation objectives.
Comparing Inflation and price-level targeting
This chart from VoxEU, shows how expectations adjust according to the regime in place. If the Fed commits to a 2% inflation target and inflation unexpectedly rises to 3% in period 3, the public will expect future inflation to be 2% at periods 4 and 5. However, if the regime is a price-level target, the public will expect future inflation to be at 1% in period 5. In the face of a shock, the price level jumps in the inflation targeting regime. A slight difference, but good to note at this point is that price-level targeting may give different outcomes compared to inflation targeting.
The effective lower bound (ELB) is the rate below which it becomes profitable for financial institutions to exchange central bank reserves for cash. Practically, the lower bound for nominal interest rates is not zero but negative due to cash storage costs. Hence, it is an “effective” lower bound rather than “zero” lower bound.
In a world without ELB, a successful discretionary policy should counteract the effects of a demand and/or supply shocks as well as anchoring inflation expectations at the desired target. However, at the level where the ELB is binding, interest rates move will not be as responsive to shocks, leading to lower inflation and a larger output gap.
The bad news is that economists believe that the economy will frequently hit this lower bound for the foreseeable future. The Fed’s new monetary policy framework can be affected by the frequency, depth and duration of this bounce off the ELB.
“Opportunistic reflation” is a good phrase to describe Fed’s recent aspiration. In Governor Brainard’s latest speech she said, “By committing to seek inflation that averages 2 percent over time, flexible average inflation targeting (FAIT) means that appropriate monetary policy would likely aim to achieve inflation moderately above 2 percent for a time to compensate for a period, such as the present, when it has been persistently below 2 percent. Consistent with this, I would expect the Committee to accommodate rather than offset inflationary pressures moderately above 2 percent, in a process of opportunistic reflation.”
Mertens and Williams, in “Monetary Policy Frameworks and the Effective Lower Bound on Interest Rates”, reinforce this idea of mitigating downward bias in inflation by following an AIT framework by aiming for above-target inflation when not at the ELB. Raising inflation expectations when inflation is low can both anchor expectations at the target level and further reduce effects of the ELB on the economy.
Unlike the conventional inflation targeting method where expectations can end up being anchored at a level below the inflation target, which in turn exacerbates the negative effects of the ELB, AIT hopefully can correct for the downward bias in inflation expectations.
Historical dependence is a very important component of calculating AIT. What do we mean by historical dependence? It is how far back we look in time to calculate the target. The lag in the AIT equation or the size of the time window we use to calculate the average is not disclosed by the Fed.
In fact, Chair Powell has specifically said that the Fed will not commit to any AIT formulas, “In seeking to achieve inflation that averages 2 percent over time, we are not tying ourselves to a particular mathematical formula that defines the average. Thus, our approach could be viewed as a flexible form of average inflation targeting.”
This sentiment is reiterated by the Vice Chair Richard Clarida, “nor is it a commitment to conduct monetary policy tethered to any particular formula or rule…at the risk of repeating myself, let me restate it verbatim: “… following periods when inflation has been running persistently below 2 percent, appropriate monetary policy will likely aim to achieve inflation moderately above 2 percent for some time.” FULL STOP.” (Bold and capital letters inserted by me).
But just for the sake of discussion, what would the AIT formula even look like?
It could look something like this:
(φ is weight on the targets)
For good measure, included are inflation targeting rule and price-level targeting rule (PLT) from “Average Is Good Enough: Average-Inflation Targeting and the ELB” by Amano, Gnocchi, Leduc and Wagner for us to notice and compare the slight differences.
All three work through raising the inflation rate above the target rate when policy is not constrained. All three also depend on influencing private sector expectations so policy credibility and public understanding are important in ensuring each policy’s success.
Alternatively, the formula could be such as the one from Reifschneider and Wilcox’s paper (two former top Fed staff economists), “Average Inflation Targeting Would Be a Weak Tool for the Fed to Deal with Recession and Chronic Low Inflation”, where they assume the Fed would simply modify a “balanced approach” Taylor rule to put some weight w on the trailing inflation gap:
Policy interest rate = r* + current inflation + 0.5 * (current inflation – 2%) – 2 * (unemployment rate – structural unemployment rate) + w * (trailing inflation gap)
The lookback window for the trailing gap varies for different policies. Traditionally, the lookback is over last year’s inflation. For price-level targeting, it may be the full inflation gap since the start of the policy. For AIT however, it could be the whole gap since the start of ELB binding or, the lookback could be an intermediate window, e.g. 5 years. The weight w could be the same as the year-on-year gap weight that conventional policy uses, which is 0.5.
Another variant is put forward by Goldman Sachs, which suggest two possible alternatives of AIT formulas where for the first one, the policy rule raises the inflation target temporarily when inflation has persistently run below 2%. The second formula adds instead the cumulative inflation shortfall, perhaps since the start of the recession, as an additional term to the policy rule.
Looking at all these formulas it’s clear that no matter which form it takes, the lookback is a very important aspect of AIT.
Questions about the speed of adjustment are difficult to answer, mainly because of the flexible element of this monetary framework. Having said that, the seminal work of Brainard cautions us against too quick of an adjustment, considering the uncertainty in economic data and the reaction of participants to this unconventional policy.
The gradualism approach tells us that policy calibration is not a one-off process but rather an iterative one, with what should be plenty of feedbacks. The Fed needs to listen to the market and the market needs to be educated and informed at every step of the way. This feedback loop will inform the efficacy of the policy and its transmission mechanism. Prudence should be the basis of every Fed’s decision or as Yellen used to say, “data dependent”.
The exception to this would be when the degree of inflation persistence is likely to be high and risks de-anchoring inflation expectations. At that point, policy anti-attenuation should be the rule of the day, where the Fed responds aggressively to inflation (or deflation).
Flexible in the “flexible AIT” means that AIT is applied in conjunction with considerations of unemployment, output and general domestic and global conditions. Chair Powell in his statement on the 27th said, “In seeking to achieve inflation that averages 2 percent over time, we are not tying ourselves to a particular mathematical formula that defines the average. Thus, our approach could be viewed as a flexible form of average inflation targeting. Our decisions about appropriate monetary policy will continue to reflect a broad array of considerations and will not be dictated by any formula. Of course, if excessive inflationary pressures were to build or inflation expectations were to ratchet above levels consistent with our goal, we would not hesitate to act.”
In other words, the ultimate decision will be more subjective than it is objective.
Perhaps the connection between unemployment and inflation today can be summed up by this paragraph from Joe Weisenthal’s newsletter: “While many people say that the problem for the Fed is that inflation has consistently come in beneath its 2% target, that’s not really it. The real issue is that it has consistently underestimated how strong the labour market could get, without triggering inflation. In the wake of the Great Financial Crisis, the Fed’s first hike came at the end of 2015 when the unemployment rate was still at 5%. Over the next few years, the unemployment rate just kept falling without triggering sustained inflation, indicating that there was far more “slack” left in the labour market than they had appreciated back in 2015 when they first raised rates.”
Thus, there seems to be plenty of slack left in the economy, but what happens when the slack starts to decline? Declining slack significantly boosts wage growth. According to Goldman Sachs, the rule of thumb is that a 1 pp fall in the unemployment rate gap raises wage growth by 0.3-0.4 pp. The unemployment rate has statistically highly significant and very robust effect on wage growth in the modern era of low and stable inflation expectations.
The effect of the unemployment rate on wage growth is also statistically significant when looking only at data from the current expansion. Most datasets suggest that the Phillips curve is non-linear, strengthening at low unemployment rates. All this underscores that slack is a key driver of wage growth.
Domestic cost inflation is one of the main drivers of medium-term inflation pressure with wage inflation as the largest component of domestic cost inflation. When other temporary factors such as import prices affect inflation, wage inflation is a very useful indication of the underlying inflation pressures once the temporary factor goes away.
From the recent Fed’s press release, “On maximum employment, the FOMC emphasised that maximum employment is a broad-based and inclusive goal and reports that its policy decision will be informed by its “assessments of the shortfalls of employment from its maximum level.” The original document referred to “deviations from its maximum level.””
Another way to say this is that it will be a high bar for the FOMC to tighten just because the unemployment rate goes below the estimated natural rate of unemployment.
The “broad-based and inclusive goal” in the press release is an effort to avoid opportunity cost of reducing accommodation in terms of the employment of minorities. In her latest speech, Governor Brainard said, “The decision to allow the labour market to continue healing after the unemployment rate effectively reached the 5 percent median Summary of Economic Projections (SEP) estimate of the normal unemployment rate in the fourth quarter of 2015 supported a further decrease of 3-1/2 percentage points in the Black unemployment rate and of 2-1/4 percentage points in the Hispanic unemployment rate, as well as an increase of nearly 3 percentage points in the labour force participation rate of prime-age women. It also created conditions for the entry of a further 3-1/2 million prime-age Americans into the labour force, a movement of nearly 1 million people out of long-term unemployment, and opportunities for 2 million involuntary part-time workers to secure full-time jobs.” In this there are lessons for the carrying out of future Fed decisions to be more inclusive.
The AIT policy framework is not without its disadvantages. The presence of lags between policy to the inflation outcome as well as the current parameter fuzziness means that there will be weak accountability on the Fed’s side. If applied more rigidly AIT also brings about the possibility of passing through big shocks to the economy in cases where maintaining the average would run counter to the immediate needs of the economy, as well as the question of whether AIT will affect the exchange rates.
Referring once again to Reifschneider and Wilcox’s paper mentioned before, the criticism they have for AIT is that the policy might not realistically be able to deliver the large inflation overshoot needed to reach the average 2% target. At the other extreme where the ELB is binding, AIT will not be able to influence expectations very much and therefore, is not an effective tool in fighting recession or boosting the economy.
In their own words, “The combination of the ELB constraint and the flatness of the Phillips curve prevents monetary policy from raising inflation back to 2 percent quickly; with the return instead taking many years, the average inflation gap is thus modest by the time actual inflation is close to 2 percent again. As a result, the rules do not call for a materially easier stance of policy at that point compared with the balanced approach rule and so do not support the tight labour market conditions needed to push inflation materially above target.”
To reiterate, the case today is such that a moderate application of the AIT will mean a very long time for a return to the 2% average, especially since the Phillips curve is flat. (Let’s not get into the discussion of whether the curve is dead or alive).
In addition, AIT cannot reduce the unemployment gap in a very bad recession. Reifschneider and Wilcox find that none of the AIT rules markedly improve simulated labour market conditions or check disinflationary pressures during the first five years after the onset of a recession. There are other concerns outlined in the paper that may be of interest to some.
According to “From Taylor’s Rule to Bernanke ‘s Temporary Price Level Targeting (TPLT)” by Hebden and Lopez-Salido, strict adherence to Taylor-type rules would usually lead to early departures from ELB episodes, but with a high likelihood of a relatively prompt return to the bound.
In the paper, Hebden and Lopez-Salido compute the efficient policy frontiers that trace out the best minimum obtainable combinations of output and inflation volatility given the ELB constraint on the policy rate and the assumed structure of the model economy. Each point on a frontier corresponds to an optimal trade-off for a central bank with a particular preference for minimising inflation volatility relative to unemployment volatility.
They found that an inflation-averse policymaker will have to tolerate greater and greater increases in unemployment volatility to achieve negligible improvements in stabilising inflation. The frontier also indicates that an unemployment-averse policymaker will face a steep trade-off in accepting higher inflation volatility if exclusively concerned with minimising unemployment volatility. In a nutshell, policies would fare better when aversions to inflation and unemployment volatility are relatively balanced.
AIT is neutral in terms of fiscal policy. Whereas ten years ago, the concern with inflation targeting was that it cannot prevent fiscal dominance, today’s problem is that Congress may not be doing enough in giving substantial fiscal policy support for the economy.
In a recent NPR interview Chair Powell said, “So Congress did respond very strongly. I sense that there’s – you can see that there is pretty widespread agreement on both sides of the aisle that something needs to be done. I guess there are differences, it appears, as to what needs to be done and how big it needs to be…But those sorts of things are for Congress to decide. But I do think there’s agreement that something needs to get done. And my guess is that in time, more will be done. And certainly, I think more will be needed.”
The credibility of this policy rests on the efficacy of the AIT model itself, and how the Fed’s policy framework is explained and understood by the public. For now, the public still has a murky understanding of it – not only because there are some aspects of the policy that is kept hidden from the public such as the actual not-a-model used and its parameters, including the list of additional criteria that define its “flexibility” – but also with the way the Fed is explaining the policy to the public.
Perhaps the Fed should keep in mind that the public can deal with any uncertainty thrown its way, as long as they know what they are supposed to be uncertain about.
EightAteEight 