The Data You Actually Have

The Number You Decided On Was Wrong

Not falsified, not fabricated — revised, years later, by enough to change what you should have done.

12 min140 XP available4 challenges

Run the model

Set policy twice on the same economy.

The same rule from "Setting the Rate", run on two different estimates of the same output gap: the one available at the meeting, and the one the statisticians settle on years later. Inflation is not revised much. The gap is revised enormously — it is not measured but inferred, from a potential output nobody observes. Move the two apart and read the bottom number as what it is: the policy error you would have made, at the time, following the rule correctly.

Policy error at the time

-2%

negative means policy was too loose

Rate the rule gave you then: 7%
Rate it would give with hindsight: 9%

Inflation at the time6%

Output gap as estimated then-4%

In the 1970s this was believed to be deeply negative

Output gap as later revised0%

Revisions of several points are ordinary

Reproduce the 1970s: a gap believed deeply negative that was later revised away, leaving policy at least 1.5 points too loose

Try -8% — not yet
Reproduce the 1970s: a gap believed deeply negative that was later revised away, leaving policy at least 1.5 points too loose — done

Answer to continue.