But if nominal GDP is the magic bullet, then central banks all around the world must be doing it? Nope. As James Bullard, the Federal Reserve rate-setter said last year, to his knowledge not one central bank is using nominal GDP and in any case, “if you wanted to make the change during calm, successful times for the economy, that’s probably the time to do it”.
Why have central banks shunned it so far? Firstly, it is too difficult to measure. Inflation – except, that is, in a shutdown – is a simple case of deciding what is in the basket of goods and services, and then going out and collecting prices. Measuring nominal GDP on the other hand is a nightmare.
O’Neill’s confidence in high-frequency indicators and improved data analytics is such that he asserts that it “would not be too difficult to calculate GDP even on a weekly basis”. But I wouldn’t want to be the rate-setter steering the economy with our nominal GDP numbers.
To illustrate the point, look in the thickets of the Office for National Statistics website where none but the most committed eco-wonks dare venture. There you’ll find “revision triangles”, which show the difference between the number-crunchers’ first estimates and their latest ones.
In the first quarter of last year the ONS initially put quarterly nominal GDP growth at 1pc. Now, it’s 1.4pc. Or take the third quarter of 2019, where the ONS now thinks that it is twice as high as the 0.5pc initially estimated. Setting interest rates off these numbers massively increases the risks of over or under-steering, so we really do need a data revolution first.
Apart from the figures, there are other dangers. People broadly understand an inflation target. Would they get their heads around a nominal GDP target? As Mark Carney, the former Governor who flirted with the idea before being appointed, told MPs in 2013: “People must generally understand what the central bank is doing – an admittedly high bar.”
Another risk is around the setting of the target: pre-crash trend growth in nominal GDP was around 5pc a year, but that is no longer the case since 2008. Is it 3.5pc, or 4pc? The number settled on will have a significant bearing on the amount of inflation central bankers will be willing to tolerate.
On the line is the loss of central bank credibility and inflation expectations if the inflation target is no longer front and centre. Losing that anchor built up over decades in peoples’ minds could end up feeding into higher wage demands, if the amount of inflation to be tolerated varies constantly.