Monday 19 December 2016

Are we drinking 42% more than we did in 1980?

Public Health England had an awkward moment earlier this month when they press released their report on alcohol with the claim that people in Britain are drinking twice as much as they did in the early 1980s. When asked to give their source for this assertion, they removed it from their website and replaced it with this:

Between 1980 and 2008, there was a 42% increase in the sale of alcohol.

This statistic is also highly dubious. The only question is whether it is a flat-out lie or is merely grossly misleading. It appears in the report like this:

According to Her Majesty’s Revenue and Customs (HMRC) duty and tax receipts, alcohol sales in England and Wales have increased by around 42% from roughly 400 million litres in the early 1980s, peaking at 567 million litres in 2008 and have since declined (Figure 6) (45). 

You might expect reference 45 to be a link to the HMRC data. It is not. Instead, it is a reference to a BMJ article by two minimum pricing campaigners, Nick Sheron and Ian Gilmore, both of whom were co-authors of the PHE report.

(45) Sheron N, Gilmore I. Effect of policy, economics, and the changing alcohol marketplace on alcohol related deaths in England and Wales. BMJ. 2016;(1860):353. 

The article does indeed make the following claim...

'UK alcohol sales increased from around 400 million litres in the early 1980s, peaking at 567 million litres in 2008, an increase of around 42%...'

It cites three sources for this: the British Beer and Pub Association's Statistical Handbooks of 2009 and 2013, and this HMRC webpage.

The most recent HMRC dataset does not go back further than 1999 but there is an archive which contains some figures going back to 1980. In Table 6 of the 2013 edition, it can be seen that 567 million litres of pure alcohol were sold in 2007/08, just eclipsing the previous high of 566 million litres recorded in 2004/05.

All good so far. However, these particular figures only go as far back as 1986/87 when 431 million litres were sold. There are no figures for 1980 that I can see. It is quite possible that only 400 million litres were sold in one of the recession years of the early 1980s, but I can see no evidence of it in the HMRC source provided.

But on the very same page, HMRC shows the amount of alcohol bought per adult. This is the relevant figure since the population has grown substantially since the 1980s. The figures are as follows:

1986/87: 9.53 litres
1990/91: 10.01 litres
2000/01: 10.50 litres
2004/05: 11.73 litres
2007/08: 11.47 litres
2012/13: 9.65 litres

Put in their proper context, it is clear that the recent peak in alcohol consumption arrived in 2004/05, not 2007/08. This is significant because it is central to Sheron and Gilmore's argument that consumption was rising until the alcohol duty escalator was introduced in March 2008, and Public Health England use exactly the same argument to call for higher alcohol taxes. It wasn't rising.

It also shows that current rates of alcohol consumption are very similar to those of thirty years ago. This is in direct contradiction to the headlines that were inspired by Public Health England's report, such as 'Britons are drinking 42% more than they were 30 years ago' (Daily Mail).

Other than propaganda, there is no justification for picking 2008, rather than the most recent year, as the benchmark date, nor is there any justification for looking at total alcohol sales rather than per capita sales. It is true that a record 537 million litres of alcohol were sold in 2007/08. It may be true that only 400 million litres were sold in 1980. But neither of these statistics mean anything unless they are put in the context of the fastest population growth in recent times.

To repeat: the per capita figures are on exactly the same page of the HMRC spreadsheet as the total sales. It is inconceivable that Sheron and Gilmore did not see them.

As for the question of how much we are drinking compared to 1980, we must turn to the BBPA statistical handbooks cited in Sheron and Gilmore's BMJ article. The most recent edition I have to hand (2014) gives the following figures for alcohol consumption per adult:

1980: 9.4 litres
1990: 9.8 litres
2000: 10.4 litres
2010: 10.1 litres
2013: 9.4 litres

We are, therefore, drinking exactly the same as we did in 1980. In fact, the figure dipped again in 2014 so we are drinking less. (Note that the BBPA define an adult as anyone 15 or over whereas HMRC define it as anyone 16 or over, hence the slightly different figures. HMRC also used financial years rather than calendar years.)

There is no particular reason to compare 1980's figure to 2008 but, for the record, the BBPA recorded 10.8 litres per adult in 2008, a rise of 15 per cent on the 1980 total, not 42 per cent.

At best, Public Health England have used an inappropriate measure of alcohol consumption to mask the fact that alcohol consumption today is lower than it was in 1980.

No comments: