Three types of "error"

It's often the case that sport can give us lessons for other walks of life. So it is with Mikel Arteta's recent reaction to Lee Mason's failure to draw the lines on the TV replay which would have shown that Brentford's equalizing goal was offside. He said:

That wasn’t a human error…it was a big, big, big, not conceiving or understanding your job. That’s not acceptable.

This distinction is a nice one. In environments where there is uncertainty, time pressure and incomplete information human errors are inevitable and thus forgivable (although the history of the airline industry shows they can be reduced). What's less forgivable, though, is just not understanding your job. We often see this in football; the player who mishits a backpass or misses an easy chance will usually be consoled by his team-mates whilst the one who fails to mark his opponent or track his runner will be criticized.

This distinction also applies in finance. The fund manager who picks a quality defensive stock might under-perform. But that's a human error. He at least understands his job - because we know that such shares in general on average do in fact beat the market. By contrast, the fund manager who buys a spivvy Aim stock and loses money is not even understanding his job, because Aim stocks on average under-perform badly, perhaps because investors systematically pay too much for lottery-type shares: the Aim index is lower now than it was in 1997 whilst the All-share index has almost doubled.

The failure of Neil Woodford's funds was, I think, a case of not understanding his job. Buying under-performing shares is mere human error. But Woodford's error was more culpable. He held illiquid stocks in an inappropriate vehicle - an open-ended fund - which meant that when investors wanted their money back he couldn't give it them.

This error was compounded by another failure of understanding the job - that the task of decision-makers must be to avoid egregious and well-known mistakes. One of these is that past success often makes people dizzy with success, with terrible effects. Woodford was following a pattern: Goodwin's success at taking over Nat West made him think he could do the same with ABN Amro, and Cameron's victories in the 2014 Scottish independence referendum and 2015 general election emoboldened him to think he could win a Brexit referendum. Both ended in disaster.

We should view policy mistakes through this same distinction. Some mistakes are inevitable and forgivable because of what Thomas Homer Dixon called an "ingenuity gap": there's a gap between the vast complexity and unknowability of the world and our puny knowledge.

I'd put some of the Covid reponses into this. Not locking down earlier when we first knew about the virus might well have been an error in retrospect. But it was a forgivable one: Covid was a new and unfamiliar threat.

Similarly, the Bank of England's failure to keep inflation down is also a partly forgivable error. It could not have foreseen the surge in oil and gas prices following Russia's invasion of Ukraine, and even if it had done so it could not have raised interest rates to the level needed to keep inflation at two per cent without inflicting massive damage on the economy. Yes, it might have raised rates in anticipation of mismatches between supply and demand driving up prices, but even had it done so inflation would now still be way above target.

What are not so forgivable, however, are fiscal policy errors. The purpose of fiscal policy is not to enact the voices in your head. It is to respond to economic facts on the ground. Failing to see this is to fail to understand the job of being chancellor. Which is what Kwarteng did in announcing a tax-cutting Budget at a time when markets were worrying about inflation.

And not just Kwarteng. Tight fiscal policy in the early 2010s when there were unemployed resources and interest rates were at their lower bound was also a misunderstanding of the job of chancellor - that looser policy is warranted in such circumstances.

Tory fiscal policy has been like Eric Morecombe's piano-playing; the right notes, but not necessarily in the right order*. Just as the one misunderstands the job of the pianist, so the other misunderstands the job of policy-making.

Useful as it is, Arteta's distinction is not a complete one. We might add a third type of error - that of professional deformation, that professionals are trained and selected to have a particular worldview. Politicians, lawyers and - yes - economists all have distorted views of the world by virtue of (and perhaps because of) understanding their job.

Journalism is an important example of this problem. A lot of what is wrong with it is due not to human error or to obeying the orders of billionaires. They are instead an inherent product of the profession of journalism itself: a bias towards individual agency and against emergence; towards human interest rather than statistics; towards individual events rather than slow-moving trends; and towards over-selling stories. Such biases exist even if the journalist is doing her job properly: journalism is not social science, and cannot give us an accurate picture of the world.

There's another problem here, inherent to our language. We apply words such as "error" and "mistake" to policy-making, forecasting or stock-picking as if we were teachers putting a red cross next to an exam answer. But this is sometimes wrong. In an uncertain world, nobody has the right answers except perhaps in hindsight. We need a different word, for a "mistake" which is inevitable and forgivable. Until we get one, we must ask other questions, the answers to which will vary from context to context. To what extent can human errors be reduced by better training and selection? Or might such training actually introduce other errors? Which errors would persist simply because the world is complex and our knowledge limited? And, if errors are inevitable, how can we ensure we are resilient to them, and which ones would we rather make - because in our complex world, this is sometimes all we can do?

* A pedant writes: technically, he wasn't, as he was playing in different keys.

Comments

[Rob]

You might add that Tony Blair though that he could sort out the Middle East with military intervention once NATO 'sorted out' the Balkans with a similar strategy.

[Graham]

Covid Responses:

No, not locking down earlier was an unforgivable error. I wrote to New Scientist on 8 February 2020 suggesting that the UK immediately close its borders to all incoming travellers. Sadly New Scientist chose not to publish my letter, but I'm sure their editor will confirm that I sent it.

Similarly, the BoE was powerless to control inflation caused by rising oil and gas prices since these rises were all caused by a reduction in supply on the international market. Boosting the value of the pound would have helped, if this were within the control of the BoE. Long term, the development of indigenous energy sources is the only protection against changes in external energy markets.

[Simon]

VAR is instructive because we can form a hierarchy of error...

Mason committed an error of negligence... so he didn't even get to make an error of measurement... let alone an error of judgement... so we didn't get to form an opinion on his taste. Mason thought his taste outweighed his duty so he shot from the hip and missed. He didn't even measure once before cutting.

Truss/Kwarteng are similar... negligence in failing to understand the problem compounded by negligence in failing to use the OBR meant they didn't even cut the right bit of wood.

[rsm]

《Boosting the value of the pound would have helped, if this were within the control of the BoE.》

What about boosting volume to meet money demand? If extraction is more expensive because capitalism has created more middlemen, why not simply print money to pay the middlemen?