Thursday, 14 January 2010

Costs of alcohol in New Zealand's antipodes

Chris Snowdon over at Velvet Glove points to a new Scottish study of the social costs of alcohol, reported on by Alex Massie of the Spectator.

A quick flip through the report suggests some pretty familiar problems: most prominently, counting as social costs a host of private costs. Where the Scottish piece suggests costs of about 3.5 billion pounds, it looks like an economic measure of external costs would be somewhere between a quarter and a third of that value. I'm not going to spend a month on it like I did with the BERL study (see here and here). But some cursory notes:
  • the costs of anti-alcohol initiatives are counted as a cost of alcohol: so if funding for alcohol treatment centres doubled, the cost of alcohol would increase
  • Costs of social care are based on the proportion of families where social workers think alcohol is a problem, not the cost reduction that could be achieved in the absence of alcohol (as alcohol is only one of many problems for many such families)
  • crime costs seem to be counted whenever survey suggests 'drunk as one of the reasons for crime', but we can never really tell what proportion of those would have occurred for the other reasons even if alcohol hadn't been there
  • it looks like they're counting all forgone wages as being social costs; they've made a passing note to that they're private costs if labour markets are "perfect", but then seem to go on to assume all costs can be counted as social because markets aren't "perfect" (rather than just counting a portion of them as social to reflect the degree to which markets aren't perfect; they also assume alcohol-affected workers are employed at average wage rates rather than being concentrated in the lower brackets.
Somebody who wanted to spend a month reverse-engineering their figures and applying sensible method could likely knock the costs back to somewhere south of a billion pounds as a rough estimate. We knocked the BERL study back by rather more than that, but the Scottish piece seems to have avoided some of BERL's sillier methods.

The bulk of the mortality costs would properly count as internal costs for drinkers (save for victims of crime or victims of drink drivers who weren't in the drink driver's vehicle), so that 1.46 billion would knock back to somewhere around 100m, most likely; productive capacity costs would likely knock back to about a tenth of the reported figure; crime costs for matters other than alcohol-specific offences should be knocked back by at least a third. A rough ballpark for an 'external cost" figure would be between a third and a quarter of what they're reporting. It could be lower than that, but I'm not going to spend a month on it.

I can't quickly find the Scottish excise tax take, but the take for the whole UK is 11.5 billion pounds as of 1999-2000 and Scotland is about 8% of the overall population, so ballpark excise tax revenues in excess of about a billion pounds (since they'll have risen in the last decade).(Updated, see below) So, again, external costs are roughly at the level of the excise tax take.

Of course, the Scottish government views the figures as plainly demonstrating the need for ever more stringent controls on alcohol.

Yet more cost reports whose function is agitprop.

Update: Chris (in comments) very usefully notes alcohol excise taxes of £768m (which I still think likely close to what a corrected figure on external costs would look like). I think that includes Scotland's proportionate share of excise duties collected on imported alcohol, but I'm not completely sure.


  1. Alcohol revenues in 2006/2007 for Scotland were £768m, 9.7% of the total UK alcohol revenues of £7.9bn. Scotland spends more per capita on alcohol and tobacco than the rest of the UK and accordingly has the highest levels of alcohol and smoking related deaths. I can't remember off the top of my head, but I think the figure for alcohol death rates (physical disease only) in Scotland is double that of England and Wales - 20 versus 10 per 100,000 persons, or something.

    That figure of £11bn in 1999 has come from the Brewers Society, who, you have to admit, have a vested interest in overstating their case. It's then been repeated by some halfwit in the Scottish Civil Service - there's plenty, take your pick of halfwits. According to this HMRC factsheet the total revenue for the UK was £6.4bn in 1999, and Scotland took around £600m.

    Because the drinking market has been deregulated in the UK - 24 hour licensing, no lower limits on the price of drinks, council rules on number of bars in an area relaxed or eliminated - cheaper alcohol has become more widely available. Alcohol prices have dropped while the CPI has risen and tax take hasn't risen enough to counter the problems. Interestingly, this is conflict between the Scottish Parliament, who want to control alcohol more effectively, and the New Labour government, who've set about deregulating the market without thought for the consequences. When 24 hour drinking licenses were introduced, there was a section that prohibited drinks promotions. That was removed, after industry lobbying.

    What the Scottish government's legislation is about is pricing cheap alcohol out of the equation. The minimum price legislation won't affect things like good whiskies or ales, but it will mean you won't be able to buy a bottle of vodka at £5 a go, or a four pack of lager for £1. It's not intended to stop drinking or reduce revenues - it's intended to stop the flow of cheap alcohol that's considered to be the big problem.

  2. Sorry, Eric - alcohol is also subject to VAT at 17.5% as it was at the time we're talking about, which is accounted separately on the balance sheet, as it's an end charge to the consumer and in additon to alcohol excise duties. The revenue stream would be a bit higher: if £1.5bn in annual sales is a conservative estimate, then gross that's an extra £260m or net £230m. That would bring it back up to £1bn or so.

  3. I'm disinclined to count GST or VAT as being a revenue benefit of alcohol as folks would be buying something else in the alternative and that something else would be taxed at the VAT rate. So I'm happier with the first number.