Thursday, 14 August 2014

Liars, Damned Liars and Statisticians

Chris Huhne is a liar. It’s a matter of record, as it is for many an MP, but even a jail sentence hasn’t much hampered Mr Huhne’s ability to profit personally from his continuing political career. And if there was any such thing as religious justice (Hint: There isn’t) Tony Blair would eventually become (literally) a damned liar despite the riches he denies he has accumulated from spreading his tactical and strategic untruths, half-truths… and downright statistics. But all the lies of politicians through the ages are pale in comparison with those of their enablers, the statisticians.

In the news right now:  Employment is up but productivity is down, but more people are in work, yet pay hasn’t risen in line with inflation, but that’s okay because both unemployment and employment are at record levels in the right directions… but people are – I forget, is it worse off, or better off? And how would we know? The answer is, we don’t; we can’t. But that doesn’t stop the headline makers from stuffing column inches with numbers explained by incomprehensible vox-astutis soundbites.

Take me. Today I’m better off than I was last year, but in 2012 I was better off still, yet the ‘most-bestest-off’ I ever managed to be was in 1982. Then I joined the ranks of homeowners and profited… nowhere near as well as the media would have you believe I should have. Ten years later I was earning plenty more in real terms than when I was at my very-bestest-off, but my house had nowhere near kept up with the reported mahoosive rises in, it seemed, every other part of the country. So was I better off, or worse off? It’s hard to tell; tricky stuff this money.

One thing is for sure though, comparing your income with that of others is a poor indicator of, well, poverty. One man’s penury is another man’s self-reliance and a fortune for most of us might be squandered in a blur of immoderate hedonism by others. Nobody is ‘the norm’ so comparisons against that norm are pointless… and statisticians know this. It’s why they never get found out. A statistician can sell the same numbers to multiple buyers for fat fees and never be asked to justify those numbers because, knowing that the rest of us understand even less about mathematics than they do, the ‘facts’ that statistical surveys reveal can be bent to any shape you desire.

And it’s not just money. Thus a new car factory in Sunderland can spell variously: prosperity and much-needed jobs for the region, back-hand deals in smoky rooms, denial of employment in another European country (which these days is racism and therefore hateful and to be labelled fascism) more GDP, less GDP, a skills shortage, a skills surplus, an immigration problem, white flight, earthquakes, tornadoes, or god’s holy wrath, all depending on who is spinning the news and why.

The unofficial Law of Unintended Consequences and the application of Chaos Theory tells us that if a butterfly flaps its wings in the Stock Exchange, somebody wins the lottery in Panama and a Chinese fella gets laid off from a counterfeiting factory three miles outside Guangzhou. To a statistician this is solid gold – cut liberally with bullshit it can be sold for big bucks to dodgy dealers who will then further adulterate it and push it on street corners to the recreational stats-junky market with the cry: "Daily Mirror! Daily Mail!" They say that 20% of school-leavers in Britain today are functionally illiterate. They’re the lucky ones.


  1. To be fair to statisticians most of the problem is that statistics can be easily spun by those with an agenda to sell. Its a tricky mathematical discipline and easily misunderstood. I see this a lot these days with medical statistics where a headline figure "this procedure has a 1 in 1000 chance of killing you" may be statistically correct from the data but doesn't begin to reflect the actual risk

    1. But why would you want to be fair to statisticians? ;o)