OK, this might be the longest one – but it gets much snappier after this, I promise.
(Caveat: like the previous post, in this one I can’t offer a ‘true’ figure. The intention is to raise reasonable doubt about the figures that have been quoted, and how they were arrived at.)
The BMA and the Health Select Committee cite not one, but two figures for the cost of alcohol to society. The first is £20 billion. The second is £55.1 billion.
£20bn or £55.1bn? So, er, which is it then?
The huge discrepancy doesn’t just mean one figure is more likely than the other – you can’t just say “let’s call it £20bn to be conservative then”. It means both figures are suspect.
Incredibly, the HSC Report seems happy to quite both calculations, simply stating in the Executive Summary, “In 2003 the PM’s Strategy Unit estimated the total cost of alcohol to society to be £20bn; another study in 2007 put the figure at 55bn,” as if the huge discrepancy between them is simply not an issue.
In any business I’ve worked in, if someone was asked to find the cost of something and came back with two different costs, one almost three times the other, with no recommendation as to which was right, questions would be asked about their ability to do their job.
But of course, by simply putting both figures out there with no comment whatsoever as to which might be the closest to the truth, the HSC allows the anti-alcohol lobby to simply quote the higher figure.
Think that’s paranoid?
In its coverage of the report on the day of its release, the BBC website simply says, “It is estimated alcohol abuse in England and Wales kills 40,000 people and costs the economy £55bn every year.”
It doesn’t dispute the figure. It doesn’t even mention the lower figure. And so the higher figure simply becomes fact.
(Rereading that, linking back to my previous post, see also how “alcohol consumption is probably a significant factor in 30 to 40,000 deaths per year” has magically become alcohol “kills 40,000 people” – a colossal misrepresentation of what the HSC Report actually says, even if we were to accept the report’s figures as accurate – which they aren’t.)
Let’s get out calculators out. Oh hang on, let’s just guess instead
The £20bn figure was calculated in 2004 by the Prime Minister’s Strategy Unit, and includes:
- Healthcare ‘Up to’ £1.7bn
- Alcohol-related crime and public disorder ‘Up to’ £7.3bn
- Loss of productivity/profitability in workplace ‘Up to’ £6.4bn
The £55.1bn figures was calculated in 2007 by the National Social Marketing Centre (A Department of Health sponsored research company) and includes:
- Cost to individuals and families/households (eg loss of income, informal care costs) £21bn
- Public health services/care services £2.8bn
- Cost to other public services (eg criminal justice system costs, education and social services costs) £2.1bn
- Cost to employers (eg absenteeism) £7.3bn
- Human costs (DALYs)[1] £21.9bn
The health care cost is at least taken from the studies quoted in the previous post, and needs no further discussion here.
Loss of productivity in the workplace is also roughly comparable and consistent.
But look at the estimated cost of crime. In the lower figure, ‘crime and disorder’ costs come in at £7.3 billion. In the higher figure, ‘costs to other public services’ – which includes social services as well as costs to the criminal justice system – comes in at only a third of that figure, at £2.1bn. I’m sorry, but given that these figures are available to civil servants, such a huge discrepancy between the two casts doubt on the credibility of both reports.
If we look closer at the higher figure, £42.9 bn – 78% of the total – is made up of intangibles – cost to individuals and families, cost in terms of loss of life or healthy years – that would seem extremely difficult to put an accurate figure on. And that’s before we start to dispute the correlation/causation argument when alcohol is calculated as a factor in hospital admission, crime etc.
Alcohol and crime
I haven’t been able to look in detail at every element of these costs, but let’s take alcohol-related crime as one example.
The Reports claims that “Alcohol-related crime and anti-social behaviour have increased over the last 20 years.”
However, Home Office data shows that there has been a 41% fall in alcohol related violent incidents since 1995. This data may not be perfect – there’s a huge difference between crime that occurs, crime that’s reported, and crimes that are prosecuted. Critics argue – perhaps rightly – that the police are now deciding not to prosecute alcohol related crimes, but issuing cautions instead. However, home office data shows consistent decline in crimes reported, number of arrests/prosecutions, and number of cautions. The data has its issues. But if you’re going to use the data when it shows an increase, you also have to acknowledge when the same data shows a decrease. The HSC does not.
And then there’s that old correlation/causation argument again. The HSC says:
“Examples of offences which are often committed by people under the influence include physical and sexual assault, breach of the peace, criminal damage and other public order offences.”
Yes, those offences are also examples of offences committed by people not under the influence of alcohol. Your point is?
Hang on, there’s more:
“According to the British Crime Survey (BCS) 2005-06, 44 per cent of violent offenders in England and Wales were perceived by their victims to be under the influence of alcohol. This corresponds to a decrease in the number of violent incidents where the victim believed the offender or offenders to be under the influence of alcohol from 1,659,000 in 1995 to 1,029,000 in 2005/06. The offender was judged to be under the influence of alcohol in 54 per cent of incidents of stranger violence, 44 per cent of incidents of acquaintance violence and 21 per cent of incidents of mugging. Eighteen per cent of violent offenders between the age of 10 and 25 reported being under the influence of alcohol only, and three per cent under the influence of drugs and alcohol, at the time of the offence. Thirty-two per cent of young people surveyed reported being under the influence of alcohol when committing criminal damage offences and 27 per cent were under the influence of drugs and alcohol while being involved in vehicle related thefts.”
There’s so much to attack here:
‘perceived by their victims’ to be under the influence of alcohol? How can you tell? Can you be sure when you’re being attacked?
And you’re admitting that the number of such incidents fell by a third between 1995 and 2006, but don’t deem this worthy of comment?
18% of 10-25 year old violent offenders admitted being under the influence of alcohol – that seems pretty low, especially given the temptation to use intoxication as an excuse for your actions.
Overall, there are two levels of reasonable doubt here: firstly, the figures consist of people who seem to be drunk – at a time when it is hard to make such cognitive judgments – or people who claim they were drunk when committing a crime – which they may be likely to do in search of mitigation.
Secondly, there is no distinction between correlation and causation. Did this person commit this crime because they were drunk, as is being implied here? If that were true, then the more I drink the more violent and/or criminally minded I will become. This simply does not occur among the majority of us who are not criminally minded to begin with. And global studies show no direct correlation between alcohol consumption and crime.
What is the net cost?
If we want to produce an accurate figure for the cost of alcohol to society, we must surely look at the net cost. There is not a single accountant on the planet who would argue that looking at expense without looking at income provides an accurate financial picture.