The Arrogance of Experts

Poor work gets echoed if it is sensational enough and suits ideological preconceptions..

In a December blog, David Farrar of Kiwiblog claimed that new incoming governments usually enjoy a huge honeymoon surge of post-election support. For instance he wrote

        ‘In 2008 National won 45% of the vote. The next poll had them at 56%. So an 11% bump from winning. So both recent changes of Government have seen an 11% bump for the winning party. This is because people like to give a new government a fair go.’

However, he says, the bounce failed to materialise for the 2017 Labour-led Government. Farrar goes on to suggest that in 2017, ‘there was no clear vote for change as happened in 1999 and 2008'. In arguing this he is consistent with National’s rhetoric that it was really a no-change election and they were robbed.

Farrar made even more extensive claims along the same lines. They were very effectively rejected by the blogger Swordfish. (He argues that the evidence even suggests the Labour-led Government did slightly better than was usual, although I am less sure because of the sampling errors that such surveys invariably suffer.) I’ll leave you to read the demolition job.

Swordfish attacks Farrar’s credibility, suggesting that he twists evidence to suit his overarching pro-National narrative. I shall not add to that story. (Perhaps Farrar made an unintentional error as a result of enthusiasm and ideological bias; as far as I know he had made no correction, explanation nor apology). This column is interested in the way the erroneous analysis was transmitted to the public.

Swordfish’s blog points to over a dozen public commentators, most of whom are very respected contributors to public discourse, quoting Farrar’s claim with hardly any doubts. In contrast only four were cautious: an anonymous writer for The Standard, Colin James and the RNZ pair Stephen Mills and Matthew Hooton. Thus the echo chamber of public opinion makers magnified the gross error into becoming the conventional wisdom (often with an ideological bias). As far as I know, not a single one has corrected their error or even commented on the Swordfish blog. Presumably they are relying on the mistake being forgotten, that a week is a long time in politics – they may well be right.

You can understand how it might happen. Writing a regular column or daily commentary often means grabbing some attractive bauble – perhaps a factoid – to hang one’s opinion on. The writers rarely have any time to do a thorough investigation or, if the truth be known, think. Instead, they depend on approved sources, deemed to be experts. Swordfish argues that this places the experts in a position of power, able to bend public discussion towards any bias the experts have.

Such behaviour is not confined to politics. How often do I see a headline which says ‘Economists say’ and my immediate reaction is ‘did I?’ Subs who put up the heads will defend the misrepresentation by saying that they did not have room to write ‘Some economists say ...’, let alone ‘A couple of economists we approached say ...’. The likelihood, though, is most readers thought it meant ‘All economists say ...’.

So most of our expert comment is filtered by journalists and public commentators who often go to as little trouble to evaluate the information they are transmitting as they did with the Farrar error.

Too often, too, the expert is as designated by the media transmitters. At least in economics, journalists often go to people I would rank low in professional expertise in the area (but they get high marks for presentability).

There is another source of bias captured in the Socratic dialogue, Apology. Socrates was told he was the wisest of men and, after thinking of all the possible ways this might be true, concludes that perhaps the claim arises because he knows how ignorant he is. Imagine the media going to Socrates and him saying that he cannot provide a definitive answer (although he might help you think about it). Down would go the phone and the journalist would ring someone else who, lacking the wisdom of knowing their ignorance, would give an answer that was short, simple, clear and wrong. It would almost certainly reflect an ideological (or employer) bias.

We saw this in the run-up to the British Brexit referendum. There were a number of forecasts by Remainer economists about the short-term impact if Britain decided to leave the EU. All the ones I recall grossly exaggerated the immediate downside. This does not invalidate the longer-term predictions that the British economy will suffer if it exits the EU but I do wish those forecasts had margins of errors. (The technical term for them is ‘confidence intervals’; in a way they are lack-of-confidence intervals.)

Given the earlier failure you can understand why many Leavers are sceptical of the experts’ long-term forecasts. To the best of my understanding their predictions are probably (roughly) right, although were I working on them I would put a lot more effort in the adjustment path to the long run (which is probably even more disastrous.)

In an  earlier column I commented that the Brexit referendum outcome seemed, in part, a rejection of experts, equating them with the hopes of those who have no expertise other than a belief in their rightness (and righteousness?).

As I have said, we need experts on tap, not on top. Keynes remarked that economists should be like dentists; they do a technical job – he may have had in mind it can be a painful one – but you make the go-head decision after, hopefully, you have had the technician explain the options and their possible consequences.

That did not occur in the Brexit referendum nor does it in many of our public debates. Instead the experts ignore the listeners’ concerns and with certainty – and arrogance – tell us what we should think (with an ideological bias often quite different from the listeners’).

It is compounded in New Zealand because we have so few competent experts that we often lack the robust, but cooperative, debates through which experts come to a consensus. Who cares? If you are a journalist you need a story, if a commentator you need an opinion.


PS. As I finished this, I came across an article which claimed that, in a ten-year period, consumption of sheepmeats in New Zealand fell 95 percent to about two chops per person per year. Is that credible? Even so, one talk-back host seized on the statistic: no doubt the session was lively but it was not informed.

It took me over two hours and approaching a real expert to understand why the figure is wrong; the figures failed to allow for inventory changes. (It is easier to explain why the original article made the mistake; it was grinding an anti-meat axe.) You can write an opinion piece in two hours. No wonder commentators prefer baubles.