Coulson: The art of painting pubs by numbers

This industry is quite fond of statistics and tables.

We have data coming out of our ears on beer consumption, catering, gaming machine expenditure, chief executives' salaries and the like.

We compile lists of the 100 top people, the 50 top pubs, and the 10 things we dislike most about the Government.

Whole TV channels are built on lists, usually historical or nostalgic, of favourite programmes, films or clips.

So to this must now be added, eventually, the published statistics on alcohol and entertainment licensing from the Department for Culture, Media & Sport (DCMS).

For commentators like me, they provide interesting reading, but by the side of the document is a large bowl of salt, from which I take a pinch from time to time. All is not as it seems.

The first interesting - and challenging - point about them is that they do not represent the whole picture. More than one in six of those licensing authorities contacted by the DCMS refused to give statistics or did not respond (17%).

Missing stats​

Among these was the London Borough of Camden, the capital's second-largest licensing division. Other large conurbations were also among the defaulters.

Why is this? You would have thought it in their interests to contribute to a complete analysis of licensing, especially as Camden has been among the most vociferous of authorities in moaning about the work it is doing on licensing without proper funding.

So all those pubs and clubs seething with people in north London do not figure in the statistics. Curious.

I take some of my approach to statistics from Professor Roy Light of the University of the West of England - something of a licensing expert himself. He suggests always reading out the other statistic to achieve the right perspective.

So when the Government says that 4% of young people commit offences while drunk after a night out, you should also focus on the idea that 96% of young people do not.

Extracting just one side of the story, as headline-writers are fond of doing - can fail to present the true picture.

Then there is what is called the "methodology". Remember that statistics are compiled by statisticians, not industry experts.

These mathematicians do not necessarily relate their questions to our overall picture of the trade.

A small example is the highlighting of the fact that a number of premises do not provide off-sales, when that may well have been an error during transition (I know of hundreds of clubs that did not realise they had to specifically request off-sale provision on the application form).

What surprises me is that the overall figures for licences are based on just a 70% return from authorities.

Under-estimate​

Apart from begging the question about what the missing 13% were putting on their returns, it does seem that the overall figures are a considerable under-estimate as a result.

They would suggest, for example, that there has been no increase at all in the number of licensed premises, even though we know that far more premises have required licences, and that the number of clubs has slumped in four years by 25% — from 20,000 in 2004 to 15,000 in March this year.

That is not credible, and it calls into question the basis of most of the figures in the document.

Much play has quite rightly been made about the comparative rarity of 24-hour licences and predominance of hotels in this list, to cater for residents.

From my own experience I know that 24-hour licences were granted mainly as stand-bys for other premises and have never been used for round-the-clock drinking.

It hardly demonstrates the rush to Bedlam that certain media predicted. So handle statistics with a great deal of care.