Peter Hoskin: We don’t need model politicians; we need anti-model politicians

It’s 6th November and everyone is talking about America. So what else to do but wish “Happy Birthday” to an American man who was born on this day in 1851? That man, who died 110 years ago, is Charles Dow, the co-founder of the Wall Street Journal and the brain behind the Dow Jones Industrial Average, a stock market index that is still among the most significant in the world. In both enterprises, he was guided by a belief that “pride of opinion has been responsible for the downfall of more men on Wall Street than any other factor.” Facts, honesty and objectivity were his lodestars.

These were important sentiments then, and they remain important sentiments now. It’s not without reason that we implore our politicians to look beyond the preoccupations of Westminster’s chattering class (which, as an opinion writer for hire, I guess I’m a member of) and to the facts that actually matter. Forget what Andrew Mitchell did or didn’t say to a policeman, and look at the growth figures. Forget Ed Miliband’s scattergun use of the phrase “one nation”, and look at how he polls with the public. Forget Twitter’s reaction to the latest session of PMQs, and look to 2015. Look, look, look.

Except, as admirable as this sentiment is, there’s often a problem with it in practice: the higher “facts” that we appeal to can be inadequate in themselves. Dow himself was alive to this problem. He intended his Industrial Average — originally, an average taken from the closing stock prices of twelve companies — to be nothing more than an indicator of market conditions. As with all financial models, it was just a simplification that the wise investor would use in conjunction with numerous other indicators. And, even then, the investor couldn’t be sure whether he’d got it right.                                                

You’d think that this was pretty basic stuff. But, in the century since Dow’s death, there have been countless instances of overconfidence in financial models. From Gordon Brown’s “no more boom and bust,” to the risk equations that underpinned the financial crisis, plenty of people suggested that they had it all covered when they didn’t. The whole hubristic mess was summed up by Alan Greenspan in 2008: “A Nobel Prize was awarded for the discovery of the pricing model that underpins much of the advance in derivatives markets … The whole intellectual edifice, however, collapsed in the summer of last year.”

Some edifices have remained upstanding, however. Despite the popularity of such we-don’t-know-what’s-around-the-corner theses as Nassim Nicholas Taleb’s Black Swan, policymakers have largely continued as before. Take George Osborne’s Budgets, which I’m broadly in favour of. Each of these contains forecasts for the public finances based on economic forecasts that are themselves based on sprawling models for productivity, policy effects, human behaviour and so on. It’s models upon models upon models. And yet Mr Osborne had enough confidence in this pyramid of models to found five-year fiscal rules upon it. In doing so, he made himself a hostage to the harsh truths those models don’t account for.

Okay, I know: the Chancellor’s fiscal rules were and are important insofar as they signal the Coalition’s deficit-reducing intent, and keep the credit-rating agencies from doing their worst. But this is another area where the models have proved inadequate. These were, after all, the same agencies that gave a collective nod to all the crazy debt packages — the CDOs and MBSs — that lay behind the subprime collapse. And now, as they downgrade country after country, it turns out that investors don’t much care. In many cases, ratings downgrades haven’t brought about dangerous increases in interest rates. The facts on the ground have marched away from expectations.

But models don’t even need to fail to distort our politics. Some models become so sacrosanct that — even if they are right and worthwhile in themselves — they can have a detrimental effect on public debate. This is something that I’ve mentioned before, in the case of Britain’s quarterly GDP figures. Westminster gets very excited about the fluctuations that emerge from the Office for National Statistics’ great model of our economy. Is it growth? Is it shrinkage? Call Osborne! Call Balls! Yet, all the while, we neglect the fact that some parts of the country have effectively been in recession for decades. 

And we have seen something similar in America over the past week. The quality of Nate Silver’s model for aggregating Barack Obama’s and Mitt Romney’s opinion poll scores has become one of the fiercest talking points ahead of today’s vote. Myself, I find Mr Silver’s work useful, and suspect — with what limited knowledge I have of these things — that it’s accurate. But still, when the polling model becomes the story, it feels like the election campaign is consuming itself.    

There is some hope left, however, thanks to one or two bursts of anti-model thinking. All of the polls suggest that Pennsylvania is a no-hope state for Mitt Romney — and yet that’s where he went campaigning at the weekend, whether out of desperation, or mindful of what the public might actually think of politicians who concentrate exclusively on swing areas. And then there’s the forthcoming Autumn Statement over here, during which Mr Osborne may have to jettison at least one of his fiscal rules. This will no doubt be sold as major embarrassment for the Chancellor, but if it represents a more careful attitude towards rules and forecasts and models, then it could be one of the best things this Coalition does.

And what will we be left with then? Hopefully, a generation of politicians that is more aware of its own fallibility. Perhaps they can then work towards a better understanding of how the country operates, but I don’t know. I’m reluctant to suggest a model solution, not least because I’ve got nothing to be sanctimonious about on this score — so let’s just go with healthy scepticism for now. Happy Birthday, Mr Dow.