In the past couple of weeks, I’ve written about proposed alternatives to current monetary policy: Nominal GDP Targeting and Monetary Aggregates Targeting. Knowing that these two are alternatives - what’s the status quo? What do central banks currently use to guide policy?
Serve the public trust, protect the innocent, uphold the law
In “The Role of Monetary Policy” (1968), Milton Friedman developed the general ideas at the core of the current thinking over the conduct of monetary policy. The Central Bank only has actual, direct control over nominal variables: nominal interest rates, the (nominal) supply of money, etc. This means that the Fed has control over how much money there is in the economy, but not over how much “stuff” or “jobs”. It doesn’t mean that the Fed has no control over real variables, just that, because they’re determined by supply and demand in a market driven by (real) factors, then it can’t fall under the nominal spell for too long.
This relates to the concept of monetary neutrality. Neutrality, roughly, means that “money is a veil” - changes in how much cash there’s going around doesn’t change how much stuff, or how many jobs, are going around, but do change things like prices. Is this true? The core facts of neutrality seem to be:
There is a strong correlation between money and inflation in the long term, but a weak one over the short term.
There is very little correlation between money and real output in the short term, but a weak one over the long term.
There is very little correlation between inflation and real output.
In the short term, the money-prices relationship is weaker and the money-output relationship is stronger, so money can affect output in the short term but cannot in the long term, and determines prices in the long term but not the short term. This means monetary policy doesn’t seem to be a good vehicle for say, promoting economic growth, but is a good tool to ameliorate recessions, or to control inflation.
Such broad empirical facts obviousy don’t tell you which policy tools are best, or how to use them. The very first idea about monetary policy is that it should be conducted according to rules. The reason why monetary policy rules are superior to discretion (i.e., doing whatever the Fed feels like) is, quite simply, that over time it makes policy less effective, which leads to more effort being required to achieve the same goals. If you promise 2% inflation and consistently allow a higher figure, people are going to stop trusting your promises, and you’ll need increasingly tighter policy to reach 2% the more untrustworthy you prove to be. A good rule has to be efficient (i.e. actually stabilize the economy), simple (being understandable and easy to communicate), precise (have clear targets and implementation), and have the central bank be accountable for its actions. So the question is, what rules should the Fed follow?
Most central banks have two mandates: price stability and full employment. This means, broadly speaking, to minimize inflation subject to not raising unemployment too much. Other banks also have other targets such as financial stability, and it’s becoming trendy to demand climate or inequality targets, but they’re not really the norm. But the target(s) you have don’t tell you what monetary rule to follow either, since it’s not really clear which mandate you have to prioritize or how to accomplish it. It’s necessary to add assumptions, or at least mechanisms, through which to determine how to best stabilize the economy.
I’d buy that for a dollar (+2%)
The way economic models normally account for monetary neutrality is through stickiness, also known as “nominal rigidity” or “nominal frictions”. Basically, for whatever reason, prices (or wages) don’t adjust perfectly and instantly, which means that changes in the supply of money affecting prices result in changes to output. For example, some businesses adjust prices early and make a profit, while others are saddled with higher costs and lose money. The same would happen to wages, where people’s purchasing power adjusts at different rates and therefore some people are worse off. Similarly, during recessions, companies either can’t cut prices or can’t cut wages, so they have to lower output and/or employment to stay in business. Monetary policy helps because it can change how much money is sloshing around the economy and prevent stickiness from messing up things.
If sticky prices are the main source of macroeconomic frictions, then that means that preventing changes in prices prevents major frictions. For instance, if inflation leads to relative price changes with real costs in output, then keeping inflation low and stable prevents price rigidity from having any real effects.
Thus, inflation targeting was born. Starting in 1980s New Zealand, the core of inflation targeting is to announce a given specific inflation rate that the government will keep at all times. Normally, inflation targets are announced as a range (for instance, 3% to 6% in the case of Uruguay), but the most common figure is 2%. The reason why is “some guy came up with it on live TV one time”, but also, while generally a 0% inflation target would be technically preferrable, there are large incentives to deviate from the target and create positive inflation. The rationale behind inflation being the primary focus of monetary policy makes sense beyond the short term, since money only really affects prices over the long run.
The exact number figure is not uncontroversial. A major criticism of the specific 2% inflation target, not inflation targeting as a whole, is that it leaves too little room to cut interest rates during recessions - if interest rates are low, or remain low, for a long time, then there’s not a lot to be done when rate cuts are needed unless more inflation is allowed. The biggest counter to this point (which was big in the second half of the 2010s) is that it came at a time when most major Central Banks weren’t even hitting their 2% targets: both the Fed and the ECB were often substantially below them, meaning it’s not even clear they’d actually get to 3% or 4% when they struggled to stay out of the 1% area. Plus, it’s not especially clear why anyone should expect the low rates episodes to be permanent - high inflation is currently a major concern, so nobody in their right mind would, retrospectively, think these points were very long lasting. Secondly, and in a related note, it’s not obvious that a higher inflation target would be able to stabilize expectations as well as a lower one - particularly because everyone might expect another upwards revision in the target later anyways. Lastly, central banks do actually have alternative tools to operate at this low interest rate area (known as the Zero Lower Bound) - which are less traditional but usually effective.
Stay out of trouble
What are the strenghts of inflation targeting? Its main benefit is that it is very good at keeping inflation stable if the Central Bank actually commits to it. This results in inflation being anchored over the long term, since the target is credible and strongly defended. Inflation targeting, especially the flexible variety, is much more able to adapt to circumstances than approaches such as the Taylor Rule too.
However, IT has several drawbacks. The first is that it might require the central bank from responding to a supply shock. This is a really big issue, because it means that the Central Bank might tighten policy in response to things it does not control - namely, the real supply and demand of goods and services. If, for any reason, there is simply less “stuff” in the economy, then it would both raise the prices of the stuff involved (but not of all goods and services - a big distinction), and show up as “inflation”
Secondly, it has an admittedly mixed record at actually reducing inflation - most countries that implemented it normally already had stabilized their inflation rates, especially in developing countries. This isn’t a major concern, but inflation targeting primarily keeps inflation low, but doesn’t actually lower it - at least not from levels significantly higher than the target.
Thirdly, it is that inflation targeting is not a rule and allows significant amounts of discretion. After downturns, it’s common for the inflation rate to accelerate, and under a strict definition of inflation targeting, the central bank would have to immediately raise rates and in all likelihood restart the recession, or at least cripple the recovery1. In practice, the central bank knows when to “take away the punchbowl” by looking at expectations and forecasts - once people think that inflation is going to go up soon, the bank knows it’s time to start tightening.
They’ll fix you, they fix everything
Regardless, the biggest issue regarding inflation targeting is, to quote DJ Khaled, suffering from success. Central banks with inflation targets normally garner credibility as inflation fighters, which means that whenever they have to stimulate the economy (and therefore raise inflation), people think they’ll stop it because inflation will go up soon, making ir harder (or even impossible) for them to actually do their job. This also harms flexible inflation targets, since nobody actually thinks the central bank will let inflation go over the target to make up from previous shortfalls. In this case, interest rates could be persistently low, policy could be persistently expansionary, and yet the economy could remain in a slump for a prolonged period of time - what’s known as a liquidity trap. While this issue is not unique to the IT regime, it’s most prevalent for it, since designing your whole monetary policy around inflation does lead to everyone expecting you to take away the punch bowl when the economy heats up. This last issue has been relevant in Japan since the 1990s, and became a major issue for the US and the EU in the 2010s. The issue here is that central banks can’t “commit to being irresponsible”, in Paul Krugman’s words.
An improvement on this last issue is what’s known as flexible inflation targeting, where the central bank commits to making up for excesses and/or shortcomings in their inflation target. The Federal Reserve committed to a flexible average inflation targeting regime known coloquially as FAIT in 2019, for instance. As an example, if one year has 1.9% inflation, the Fed would let inflation be 2.1% the next year, to smooth things over. The problem with FIT is that nobody really understands what it means, and that it’s not really clear over how long a period the Fed will make up for their inflation over or undershoot - especially important after inflation was four times the (unofficial) target of 2% last year.
There is one last alternative to inflation targeting that abandons the specifics but not the spirit: price level targeting. The main difference is that inflation is the (annual or monthly) change in the average price, while the price level simply is the average price. A major problem with IT is that you can’t make up from previous shortfalls in inflation, which consequently results on an inability to make up for previous periods of weak demand, and the long-term consequences thereof: lower investment, higher unemployment, less skilled workforce, etc. The difference with FIT and FAIT is that price level targeting actually has a clear numerical target attached - the central bank will allow inflation to run higher until prices make up the difference with the target. Of course, how to set the target is another question entirely; though in reality it’s just an inflation target with an upwards correction mechanism - if there is a shortfall in inflation, there is a very specific commitment mechanism to get it back up.
Conclusion
Is inflation targeting good or bad? Kinda. It’s main advantage is that it can keep inflation low and steady; however, the excessively detailed knowledge of the origin of price changes required for it to function, and the lack of self-correction mechanisms for excessive success at containing prices make it a suboptimal monetary policy agenda.
The first constraint is more important than ever: currently, there is a surge of inflation that is, to some extent or another, fueled by commodity prices and their relative scarcity. There is no reasonable way why fewer people having jobs will make more oil and wheat be obtained by rigs and farms - so the demand component of inflation is the only one monetary policy can affect. Nominal GDP Targeting is superior to inflation targeting in this regard because NGDP is not affected by supply shocks - both quantities and prices of either affected or non affected products adjust in the same direction, but total spending does not increase.
The second constraint was a major issue during the 2010s - monetary policy wasn’t sufficiently stimulative after the Great Recession, and a decade of sluggish growth followed, with pundits largely questioning if monetary policy could be effective at all. But the problem wasn’t that central banks were committing and couldn’t pull it off, it was that they weren’t - in the words of Paul Krugman, they couldn’t promise to be irresponsible and raise inflation, resulting in depressed expectations. Nominal GDP Targeting also proves superior on this issue, since it allows for credible irresponsibility, as the central bank would simply stimulate the economy until spending reaches the desired level.
Sources
Previous posts on monetary neutrality and Nominal GDP Targeting
Monetary policy 101
The Role of Monetary Policy
Goodfriend & King (1997), “The New Neoclassical Synthesis and the Role of Monetary Policy”
What is inflation targeting?
Bernanke & Mishkin (1997), “Inflation Targeting: A New Framework for Monetary Policy?”
Ben Bernanke (2017), “Monetary policy in a new era”
Drawbacks of IT
De Gregorio (2019), “Inflation Targets in Latin America”
Mark Carney, “A Monetary Policy Framework for All Seasons”, 2012 speech
Alternatives
Krugman (1998), “It’s Baaack: Japan’s Slump and the Return of the Liquidity Trap”
Richard Clarida, “The Federal Reserve's New Framework”, 2020 speech
This is, I must emphasize, not what is currently happening.