Behind the numbers

"Scotland is awash with alcohol," said Dr Brian Keighley this week. He is chairman of the British Medical Association north of the border, and is not alone in believing that Britain is drowning in booze.

Don Foster, the Liberal Democrat spokesman on licensing issues, blames the availability of cheap alcohol for encouraging "ever greater numbers of people to drink unsafe amounts". Andrew Lansley, the shadow Health Secretary, says Labour was irresponsible for rolling out 24-hour drinking.

Are they right? When everybody is in such easy agreement about an issue and a moral panic is under way, it's worth looking at what the statistics tell us. Data from the Office for National Statistics (ONS) shows that drinking levels have changed little since 1992, and have actually declined quite significantly in some groups in recent years. In 1992, 9 per cent of men aged 16-24 admitted drinking more than 50 units a week; in 2008, the figure was 7 per cent.

Binge-drinking levels also fail to show any clear trend. In 1998, 39 per cent of men said they had drunk more than eight units "on at least one day last week"; in 2008, it was 37 per cent. Among women, the trend is up, but the increase is entirely accounted for by a change in the way alcohol consumption is estimated.

This took place between 2005 and 2006 when the ONS changed the way it converted volumes of alcohol to units, assuming an average-sized glass of wine. The effect of the change was dramatic: in a single year, the proportion of women drinking more than 14 units a week soared from 12 per cent under the old measurement to 20 per cent using the new one. Among men, the proportion drinking more than 28 units rose from 23 per cent to 31 per cent.

Once this is allowed for, there is no evidence of a recent increase in drinking, and some persuasive evidence of a decline. The British may drink too much, but there's not a scrap of evidence this was caused by the change in licensing laws.

Of course, these figures come from surveys in which people are asked about how much they drink. There's a strong chance they tell lies. But unless the lies they tell have got bigger as the years pass, that wouldn't affect the trend.

There are alternative ways of measuring consumption, such as the volume of alcohol sold. That has increased since 2000, but not dramatically. In 2001-02 the volume of alcohol sold per adult was 11.09 litres; in 2007-08 it had increased to 11.53 litres. This is growth, but it's hardly an ocean of booze.

What about alcohol-related deaths? These have certainly increased, doubling between 1992 and 2008. But most of this increase happened in the 1990s, and the trend has now flattened off in women, less markedly so in men. Alcohol-related disease has risen fast, but from a low base, and little of it can be attributed to trends in drinking since 2000.

So why does everybody believe that drinking is growing at an explosive rate? Changed behaviour patterns, where some groups glory in getting drunk rather than being shamefaced about it, have made drunkenness more visible. There's not necessarily any more of it, but there's more of it on show.

Nigel Hawkes is the director of Straight Statistics

Comments