In the run-up to the UK General Election, numbers will be knocked around harder than the balls in a Djokovic-Wawrinka tennis match.
Each side will choose the stats – and the definitions and the contexts and the baselines – to support their argument.
A Guardian editorial bemoans the state of number-based debate.
The Royal Statistical Society (RSS) has responded. They reassure us that the numbers can – usually – be trusted. They’re right. The UK has an official statistical system which is among the world’s best. The statistics it produces are guided by a code of practice, and tested against it. Those found good enough are awarded National Statistics status – a kind of statistical kitemark which we call can trust.
The RSS also makes important recommendations that could improve political debate using numbers. Not least is the need for politicians and civil servants to have a good level of statistical literacy.
Some – many? – will feel faint at the thought of calculating a mere percentage let alone carrying out a t-test.
But statistical literacy isn’t just about the calculations. It’s as much, if not more, about knowing what the numbers – counted or calculated – can and cannot tell you.
I feel, though, that we shouldn’t just be asking politicians to do statistics better.
The Science Media Centre has a philosophy: “The media will DO science better when scientists DO the media better.”
So I’d suggest that politicians will DO statistics better when statisticians DO politics better.
That’s not about fiddling figures and drawing dodgy bar-charts – it’s about ‘political literacy’.
Why is the politician using the numbers inappropriately or unfairly? What is their motivation? What change are they trying to secure? What point do they want to make?
Do they even know that their use of the statistics in a particular way is incorrect?
Week in, week out, MPs meet people in troubling, sometimes desperate, circumstances. In the chamber of the House of Commons, they will want to make their points powerfully. The statistics of the many are employed to get change to solve the problems of the individual.
Might that misused poverty statistic not be political point-scoring but a proxy? Might it come from wanting to help real parents who are struggling to keep their children fed on handouts from a foodbank?
So I hope that any programme to educate the politicians statistically will be one of dialogue and mutual learning. Wouldn’t it be great to see government statisticians seconded as a matter of course to individual MPs’ offices – not just in Westminster but in their constituencies, too?
Who knows, some statisticians may find they want a career change … Now, there’s a thought – a House of Commons with a critical mass of MPs from a professional statistical background!