Beer facts that aren’t clear facts
April 28, 2009
An article in the Liverpool Daily Post tells us that, “Beer sales fell more than 8% in the first three months of the year compared with 2008. The British Beer & Pub Association (BBPA) said 1.7m fewer pints were drunk daily from January to March than in the same period last year.”
So far, so good – that’s 1.7m fewer pints each day for 90 days (31+28+31), making a total of 153m fewer pints sold in the first quarter of 2009.
But then, a few paragraphs later, we learn this: “The BBPA’s UK Quarterly Beer Barometer showed a total of 68m fewer pints – more than 750,000 a day – were sold.”
What? This flatly contradicts the intro. So is the true figure 1.7m fewer pints daily or 750,000 pints? After all, the difference represents a hell of a lot of lager!
The answer is found on the BBPA’s own website, which reveals that the lower figure (68m pints) refers only to sales in pubs, bars and restaurants (the “on trade”, as opposed to the “off trade” of supermarkets and off-licences) – “68 million fewer pints were sold in the on-trade during January to March 2009 compared with the same period in 2008”.
But even here, all is not clear. The BBPA site confirms that total beer sales were down 8.2% compared with the same period in 2008, yet towards the end of this article, it says, “In total, beer sales for the year up to March 2009 fell 7.0 per cent on the year to March 2008. On-trade sales were down 8.9 per cent and off-trade sales fell by 4.5 per cent. ” (my emphasis). This set of figures is indeed confirmed by the supporting spreadsheet, which lists the total number of barrels of beer sold, broken down into on-sales and off-sales.
So the question is, if the BBPA spreadsheet shows total sales fell by 7.0% from Jan-March, why does the BBPA’s article start off with a higher figure of 8.2%?
The reason becomes clear when you look at the spreadsheet and see that it contains two separate sets of figures. One represents quarterly beer sales (the period Jan-March), while the second gives the moving annual beer sales (the 12-month period up to March 2009). The apparent anomalies have arisen because figures have been taken from the two different sheets.
This wasn’t at all obvious. The phrase “beer sales for the year up to March” was used when referring to the annual moving total. It would have been better to have written “beer sales for the 12 months up to March”, since the natural interpretation of the phrase “the year up to March” is that it means the first quarter of the year (ie, “the calendar year of 2009 up to March”).
So, armed with this detail, let’s revisit the Daily Post story.
That intro figure (“fell more than 8% in the first three months of the year compared with 2008”) is spot on. This is the quarterly beer sale figure, which was 6.298m in Q1 2008, down to 5.782m in Q1 2009. It’s a drop of 8.2%, to one decimal place.
Par 2 says “1.7m fewer pints were drunk daily from January to March than in the same period last year”. Now the BBPA spreadsheet measures beer sales in terms of barrels, so to convert to pints we need to know how many pints you get to the barrel. This isn’t in the Post story nor (as far as I could see) on the BBPA website, but a quick sum showed me the figure was around 290 pints (289.4, to one d.p.). The difference in barrels is 516,000 (or 5,733 per day) – which comes to 1,662,666.7 pints, reasonably enough rounded to 1.7m.
And now when we come to the sentence, the “Quarterly Beer Barometer showed a total of 68m fewer pints were sold”, we know this refers to the on-trade only (I make it 68,009,000 pints using the conversion figure of 289.4 pints per barrel).
Only one figure still puzzles me. On the BBPA website, it says that compared with the first quarter of 1999, 326m fewer pints were sold: “a decline of 3.6 million pints a day”. But there are 3652 days in a decade (including the two leap years, in 2000 and 2004), so that makes a decline of 89,266 pints per day.
Shurely shome mishtake? Hic!