Benjamin Disraeli’s old adage about “lies, damned lies and statistics” is now spinning away from its original placement (he actually meant to say “expert witnesses” rather than pure statistics!) and is moving beyond oxymoron into cliche. But since the UK’s Publishers Association statistics book ( was published on 2 May, and sparked the usual irrelevant radio and newspaper commentary on its findings, I found myself pondering both our gross misuse of statistics in everyday life, and how increasingly the thirty or so trade bodies in the British Media, and their European confederations and US co-evals continuously mislead us by pretending to be a market measure when they are actually a symptom of change for producers, and an increasingly misleading one.

The statistic from the UK Publishers Association (PA) that got me itching in this rabid manner was this line: “Total physical and digital book sales have fallen from £3.5 billion in 2012 to £3.4 billion in 2013”. This got the chattering classes going at a furious rate. The Death of Reading? The Decline of Britain’s Place in World Culture? Collapse of Educational Standards in a once Great Nation? Well, I am all for firing the current Secretary of State for Education, re-opening the libraries and even, if it helps the struggling book trade, allowing prisoners in our teeming jails to receive books for reading purposes (currently forbidden in the UK), but, seriously, this is not down to the Government, or even the poor old publishers. It could be the start of a trend or a post-Fifty Shades statistical blip, or it could simply be a statistical error. It is not a “market” figure at all, but simply reflects the information collected by the Publishers Association from those publishers who happened in the years in question to be in membership of it. It indicates nothing and has no deeper significance until it becomes greater over time, so why do we fill so much airtime with premature discussion of things that have not yet happened?

The fact is, of course, that we build bricks from muddy statistics to support our own arguments. The wise men at the UK Office of Statistics reserve the right to change major announcements – GDP, cost of living, inflation – in the months following their monthly (absurd time interval) releases, as new and better evidence arrives. In the content, media and technology industries we should do the same, and rigidly differentiate between producer statistics, which reflect segments of sales data , and market statistics, which reflect consumption within a market. Thus the radio interviewer who I caught dilating on the PA figures was obviously unaware that the same report indicates that 43% of UK output from these publishers was exported, even further diminishing their usefulness as a guide to the literacy and reading habits of the Great British Public.

And then there is the Digital Thing. Here I must take issue with the Publishers Association CEO, while sympathizing with his need to sell as many of his reports as he can. The report, he says “shows revenues from books at an interesting equilibrium moment with the total growth up by the same figure as physical sales are declining, showing that digital books are fully pulling their weight in the market”. Apart from the thought that “equilibrium” is a strange word to apply in a market where digital is growing and print declining for the publishers indexed here, there is no acknowledgement of the Howey Thesis – that an examination of Amazon’s sales figures would quickly show that genre fiction sales of self-published works are far greater than publishers believe. It would reveal that, with these self-published elements added, between 70 and 90% of sales were digital. In the view of many, the fiction market has gone digital already, but we have no way of recognizing the change. The only short term hope for the major players is consolidation (Harper Collins and Harlequin this week). With self-publishing for initial publication becoming the order of the day publisher selection of the best for advanced marketing treatment creates a derivative business model. So what is this, Publishers Association, about “Publishers have a strong historical record in driving innovation, providing products and services appropriate to the digital age”. Well, the first part of the sentence would apply to Allen Lane, the last great innovator in British book publishing history, but the last part of the sentence strains credulity. True, independent players like Dorling Kindersley did wonderful things in multimedia in the 1990s, but that soon stopped when they were bought by a conglomerate, and replicating a print book in Epub3 is hardly a startling breakthrough, even for a publisher. But it does point to an issue that I have no statistical reason for asserting: if the UK consumer publishing industry does not quickly find a way of investing in and developing new forms and attributes appropriate to a networked age with mobile technology then it will, in the 2025 statistics, be shown to have expired – like the dodo, or the newspaper.

But, of course, we all love a figure. Buzzfeed is always full of nice stats for us to send to our friends and start an argument. But they are not entirely serious. In the same way I could point out, for example, that the UK publishers output in sales revenue terms have now climbed/fallen to 66% of Amazon’s book sales (believed to be $7.75 billion). And now that Amazon Publishing have sold over 1 million copies of each of two author’s works (Helen Bryan and Oliver Potzsch) surely it would make sense to recruit Amazon into the Publishers Association, and make the stats look vastly better in a single year (Headlines: “New Boost to literacy”, “Education Minister says strategies to get the nation reading have worked” etc).

And a final point. The UK Publishers Association launched, for the first time this year, their statistics on the UK academic journals market. I wonder why. Journals is essentially a global market. “Made in Britain” journals are great, but no greater than “Made in the Netherlands” or made anywhere else. The figure the PA comes up with for British journals revenue is £1.3 billion in sales, of which £850 million was digital. In other words, Reed Elsevier-owned Elsevier is not British, since its revenues are larger than this UK total. On the other hand Wiley does its journal publishing in the UK at Oxford and Chichester – but is presumably, because it is NYSE listed, a US company. In small, fully digital globalized markets this type of geographical output recording is next to useless. The PA would be well-advised to licence access to a good market research database and tell its members something much more valuable – whether the UK grew as a buyer of academic information last year.

Over 50 years ago I secured a vacation job at Gallup Poll to keep a penurious student going. After a frustrating first day I returned to the office and asked my controller whether anyone ever told the truth to pollsters, or had I encountered an unrepresentative sample of London liars. He thought, and replied “Well, the polls that derive from these interviews are surprisingly accurate – if you allow for a statistical error rate of 5% either side of the result” Quite.

« go back