Against inflation targeting
why monetary control via inflation targeting is past its sell-by date
Last week, the Brazilian president criticised both the high interest rates in Brazil (currently 13.75%) as well as the Brazilian central bank’s inflation target of 3.25% which, according to him, is too low for “Brazilian standards”. As the Washington Post explains,
Lula called current interest rates an “embarrassment” and said there’s “no reason” to hold them so high. He’s argued that the bank’s target of about 3% inflation isn’t appropriate for an emerging market such as Brazil and should instead be 4.5%. During Lula’s prior two terms in office, the target was closer to that. The extent to which central bankers were allowed to miss their target, known as the tolerance range, was also higher back then.
Rather predictably, investors around the world have reacted to this attack on the Brazilian central bank’s independence with concern. However, there are good reasons for Lula’s criticism of the current Brazilian monetary policy regime. Even the architect of Brazil’s inflation-targeting regime believes that the decision to progressively lower the inflation target by 0.25% every year since 2019 was a mistake.
However, the real problem with Brazilian monetary policy is not that the inflation target is too low. The problem is that inflation targeting itself is the wrong approach for the Brazilian economy as it is the wrong approach for most other economies in the world today. What Lula is really saying is that the 13.75% interest rate is too high relative to the inflation rate which is now less than 6% i.e. the real interest rate is too high. Although inflation may be temporarily lower thanks to Bolsonaro’s tax cuts last year, even an inflation rate of 10% would leave Brazil with a significant positive real interest rate.
As I explained in a now decade-old essay, central banks should target a ‘real interest rate’ range rather than an inflation target. Also, “excessive volatility in real rates must be avoided even if it is at the expense of a more volatile inflation rate”. What should this range be? In short, real rates should not be excessively positive or negative. Negative real rates of -3% or lower are damaging and risk bubbles in real assets. But positive real rates of +5% and above are also unnecessarily damaging to economic growth in a developing economy like Brazil.
This question is not just relevant in Brazil today. Given the strength of recent economic data across the United States and Europe, and especially the strength and wage growth in the services sector, it is entirely possible that inflation remains sticky at 3-4% even with rates at 5-6%. What should a central bank do in this situation? Should it increase rates to 7-8% to hit an inflation target of 2%? Intuition tells us that it should not and this intuition is correct for a simple reason - there is no justification for positive real rates of 3-4%, especially when economic actors do not suffer any loss of purchasing power and can invest their cash in deposits that pay more than inflation. Partly in anticipation of such a scenario, there is already a school of thought that argues that the Fed’s inflation target is too low and should be raised to 3%. But this is simply putting lipstick on a pig. The real problem is the obsolescence of the inflation-targeting strategy.
So am I arguing that inflation targeting was always the wrong tool? Not at all - All legible and codified control regimes in complex adaptive systems succeed at first, then the system and its agents adapt to the regime and circumstances change, and the regime fails. There is nothing unusual about this. In fact, this is the norm. Complex problems don’t have final solutions - all strategies work at first and eventually fail. Another way of looking at this is that Goodhart’s Law and system adaptation mean that all intermediate targets eventually become divorced and disconnected from the final aims of the system. And of course, all legible targets are intermediate targets.