Good morning, it's Paul here, with the placeholder for today's SCVR. Please add your comments in the usual way, and the main report should be finished by 1pm today, depending on how much news there is. Edit: predictably, I'm running late, so will finish at 3pm. Today's report is now finished.

In case you haven't seen it, Jack has published an interesting article here:

Multi-bag or bust? RTN, MAB, RBG, and OTB


OBR coronavirus reference scenario

This document published by the Office for Budget Responsibility has understandably caused a furore. It's a forecast of what impact coronavirus might have on the economy. Being sensible people, they are not calling it a forecast, but instead a "reference scenario".

Economic forecasting is widely discredited, because people repeatedly try to forecast, repeatedly get it (often wildly) wrong, but continue pretending that they actually can forecast the future with any accuracy, when they cannot. Former Bank of England Governor, Mervyn King, has written persuasively about this, urging economists and pundits to stop pretending that they can forecast the economy, because they can't.

There's a terrific book on this called "Super-forecasting: the art and science of prediction". This makes the same point. Dominic Cummings (Boris's brilliant adviser) spat out the name of this book to journalists doorstepping him, obliquely calling them "useless political pundits", and suggesting they read this book. It's a fascinating read, highly recommended if you haven't already seen it. The core theme in it, is that a long-running study showed that many experts & pundits are useless at forecasting, but nobody ever pulls them up on their serial failure.

The same useless forecasters repeatedly appear in the media, despite getting things mostly wrong. It also identified (through a scientific process) an unusual type of person, who presented with facts & figures and allowed to research economic or political issues independently, were proven to be amazingly good forecasters. Not experts, but smart laypeople. Yet these super-forecasters rarely had any influence or involvement in policy, or even media commentary.

The conclusion being that we need to measure the accuracy of pundits forecasts, and jettison the ones who are no good. Then recruit the unusually good super-forecasters to give us predictions instead. Very sensible, but it won't happen. Apparently, an engaging & confident manner is enough to convince media outlets that you talk sense.


Unlock the rest of this article with a 14 day trial

or Unlock with your email

Already have an account?
Login here