Original URL: http://www.theregister.co.uk/2010/04/04/stats_policy/
Skewing statistics: Booze, money and sex
Policy-based evidence making at its finest
Comment Sorry to bang on about these things again but yes, they're lying to us. No, this isn't just the normal drink-induced paranoia common to such as I - we really are being lied to again and again by those who claim to be our lords and masters.
We've had Lord Adonis pompously announcing that he's minded to change the drink drive laws. Instead of the 80 mg limit we should move down to 50, possibly even 20 for learner drivers and HGV. We should also adopt the continental practice of having weaker punishments than the automatic 12 month ban for breaching that lower limit. He's even said he's waiting for a report so that he can consider the evidence. That evidence being that we've a lower accident rate than any EU country other than Malta, a place that hardly has any roads. What's really happening is that the EU has decided that drink drive limits must be harmonised, so harmonised they shall be.
No, this isn't now going to spin off into a rant about how appalling the EU is (despite the truth of such an assertion). It's going to carry on about how we're continually lied to by those who would rule us.
We've had innumerable reports over the years insisting that booze itself lays tremendous costs upon society. The latest ones claim that the costs are higher than any benefit that comes from either tax or booze consumption and therefore taxes must rise tremendously. Except - and the same is true of similar reports on the costs of baccy - they leave out the most obvious benefit: people enjoy smoking and boozing. That's why they do it. Quite how much that enjoyment is worth is difficult to calculate of course but we can put a lower limit on it quite easily.
No one forces the consumption of either tobacco or alcohol. The purchase of them is therefore a voluntary exchange and it's a very basic point about trade that voluntary exchanges only take place when both sides believe they are getting something of more value than what they are giving up to get it. I value my pint more than the £3 (£4 in London perhaps) than I'm giving up to get it. I value 20 tabs more than the £6.50 I have to earn to purchase them. The value in enjoyment of these things must, by definition, be more than people spend upon them. At which point all the calculations of their being more expensive than any benefits they bring fail.
To take just a few recent bits and pieces: it's now a criminal offence to pay someone who is “controlled for gain” for sex. The burden is put upon the punter to know that the dolly for hire is so controlled. Whether this is a good idea or not isn't my point - it's that the evidence presented to push for it was absurd.
We had the likes of the Poppy Project doing surveys of brothels in London. They asserted that the presence of lots of foreigners selling sex was evidence that lots of foreigners were held as sex slaves. Completely oblivious to the obvious point that - on the don't shit on your own doorstep principle - that most if not nearly all tarts do their tarting away from home. There is, after all, something of a social price to be paid for joining the oldest profession. With the rise of budget airlines doing it in another country is simply an extension of that movement away from one's immediate social circle before offering cock care for cash.
Well, if we thought that the likes of Julie Bindel (for yes, she was of course part of this) were bright enough to understand the point we would be accusing them of lying over this. But it was enough to get the law through Parliament and there we all are, stuck with it.
It isn't just booze and sex though: the Marmot Review on health inequality contained a similar howler. Their brief was to look at the confluence of income inequality and health inequality in the UK. Now you and I, bright people and fair-minded as we are, would start from the assumption that poverty contributes to health inequality. We would also note that ill health itself will contribute to poverty. Being struck down with a chronic illness during your supposedly working years is going to reduce your earnings - this seems pretty obvious. So one of the things that we would be trying to do would be to tease out the effects of each cause. To what extent is health inequality caused by income inequality and to what extent does health inequality contribute to income inequality?
Marmot and mates simply looked at health inequality and assumed all was caused by income inequality. This could be, if we were willing to make such heinous accusations, described as lying.
Or how about the National Equality Panel? It report led to those headlines about how the wealth gap was 100:1. The 9th decile of households have assets of £880,000 while the 1st decile have only £8,000, leading to shrieks of "Lor', lummee and good grief!"
Except, of course, the statement is false. When we look at income inequality we quite rightly do not look at market incomes alone. We look at incomes after the influence of the tax and benefit system. This is, outside the US at least, how all income inequality statistics are presented. For of course we don't want to run around looking at how unequal things are before we try to do anything about it. What we actually want to do is look at it after whatever it is that we do to reduce it, to see how much more we might desire or need to do.
We should of course be doing the same thing with wealth inequality. And the report most certainly did not do that. For example, private pensions were included in the calculation of assets. State pensions were not. No reason was give but the inclusion of the State pension changes those numbers quite considerably. The State pension is essentially an index linked annuity - a stream of payments until death - and can thus be valued as one. At age 65 for a man this would give a capital value of perhaps £75,000. Add that to both sets of assets and we have £955,000 odd to £83,000 or so. Sure, this is still a large gap but it's nowhere near 100:1 any more. It's more like a little over 10:1.
Owner-occupied houses were included as wealth (as they obviously are) but not the subsidy received by those living in social housing. One recent report said that the subsidy between housing association and market rents in Westminster, to take an obviously extreme example, was for a three-bed house over £20,000 a year. Such tenancies last for life, and indeed in some cases they are inheritable. While they can't be sold (at least not legally), the rights to occupy such housing is clearly and obviously part of wealth. Indeed, a very substantial part. Valuing it again as an annuity (for the obvious reason that it's a stream of income lasting until death) would give a capital value of what, £400,000 or so? It depends upon how many years and what the interest rate used is of course, but £20,000 a year for 30 or 40 years is worth that sort of amount.
So the wealth gap, in this already admittedly extreme example, comes down to something like 2:1, which is a very long way indeed from our original claim of 100:1.
Now yes, I've been extreme here, but the basic principle still stands. There's no one at all in this country that has “wealth” of £8,000 or less. We've all got vastly more than that, for we've all got a pot of wealth which will provide us with an income stream. It's called the welfare state.
This is, sadly, how we're governed now. The way used to be simply to lie directly - vote for me and I'll make the country a better place. Now we have the indirect lie - look, here's a report from the experts saying I should do what I already want to do anyway. What the report actually says is decided before it is written by skewing the things that are considered. Foreigners selling sex is evidence of sexual slavery, we'll ignore that people enjoy booze 'n' baccy, people getting ill doesn't reduce their income and we'll entirely ignore what we already do to reduce inequality when measuring inequality.
In short, we'll lie to get our way. Or as it is more politely put these days, we now have policy-based evidence making.
It's something of a pity that the upcoming ballot papers will not have a “hang them all” box to tick. ®