Pence and beer

A study called Violence-Related Injury and the Price of Beer in England and Wales offers support for the chancellor's decision to raise the tax on a pint of beer by 1p in the recent budget. The study's authors, based at Cardiff University, are scholars of pence and beer. Kent Matthews is the Sir Julian Hodge professor of banking and finance. Jonathan Shepherd, a professor at the dental school and director of that school's violence research group, has long campaigned for the mandatory use of non-glass "glasses" and bottles in drinking establishments. Shepherd's colleague Vaseekaran Sivarajasingham also took part.

The question "What causes violent behaviour?" is not simple. The trio relate some insights they gleaned from others' research. Apparently, people who are frequently drunk are less prone to commit violence than drinkers who are not used to being intoxicated. The study describes this in succinct technical terms: "[Those] with the lowest usual involvement with alcohol were subject to a higher elevation in their risk immediately after alcohol consumption compared to those who drank more heavily."

The authors caution that it can be difficult to establish the specific causes of violence. A 1994 study in the Annals of Emergency Medicine, which they do not mention, demonstrated this very point. The study, Impact of Yankee Stadium Bat Day on Blunt Trauma in Northern New York City, said: "The distribution of 25,000 wooden baseball bats to attendees at Yankee stadium did not increase the incidence of bat-related trauma in the Bronx and northern Manhattan. There was a positive correlation between daily temperature and the incidence of bat injury. The informal but common impressions of emergency clinicians about the cause-and-effect relationship between Bat Day and bat trauma were unfounded."

The new beer/violence analysis was fairly straightforward. Matthews, Shepherd and Sivarajasingham looked up sets of numbers for each region in England and Wales - assault rates reported in hospital emergency departments, the local price of beer, the local unemployment rate, and other likely suspects. Then they compared region against region.

Their report concludes that "the regional distribution of the incidence of violent injury is related to the regional distribution of the price of beer". It predicts, quite specifically, that a 1% increase in the cost of beer would result in 5,000 fewer cases of assault every year. With the UK-wide average price of a pint standing at about £2.25, the chancellor's penny-a-pint action might thus be expected to prevent about 2,200 assaults in the coming year.

But the study supplies the logic for an intriguing alternative method by which a chancellor of the exchequer might reduce the number of assaults: stop young people having jobs. As the researchers explain it, "there is a strong negative relationship between youth unemployment and violence-related injury. The higher is unemployment, the lower is youth disposable income, and the lower the consumption of alcohol and consequently the lower the incidence of violent injury."

· Marc Abrahams is editor of the bimonthly magazine Annals of Improbable Research (www.improbable.com) and organiser of the Ig Nobel Prize