Modern central banking rests on a claim that is difficult to prove and expensive to abandon.
The claim is that inflation depends heavily on what people expect inflation to be.
If firms expect prices to rise, they raise prices; if workers expect it, they ask for higher wages.
The expectation is therefore partly , and it becomes a policy objective in itself.
Independence, inflation targets and published forecasts are all instruments for managing belief rather than money.
The framework performed well during two decades in which shocks were mostly on the demand side.
A poses a harder problem, because raising rates cannot produce energy or semiconductors.
The bank can only demand until it matches the reduced supply, which means creating unemployment deliberately.
Whether to do so depends on whether the shock passes through into , which cannot be observed directly.
Surveys of are noisy, and market measures expectation with a risk premium.
Central banks therefore act on an estimate of an unobservable variable, using an instrument that works with a lag of over a year.
Under those conditions, is not a luxury but the entire mechanism.
A bank that is believed can tighten less; a bank that is doubted must tighten more to achieve the same result.
Credibility is accumulated slowly and is destroyed by a single episode in which the target is quietly abandoned.
The consequences receive less attention than they deserve.
The unemployment created by tightening does not fall on the households whose spending caused the pressure.
It falls on recent entrants to the labour market, which is a transfer nobody voted for.
Acknowledging that openly would not change the decision, and it would change how the decision is discussed.