I have a new (fortnightly) column in the Guardian – the first instalment is here. I posted that link on Tumblr last night and The Lex left an interesting comment, which I’m going to excerpt:

“Illegal downloading is v much an elephant in the room here! Obviously there are no facts and figures but it’s so easy to find any given song for free – and almost all the time, way before it’s available to buy legally – that I really assume illegal downloads of singles vastly outstrip legal downloads. (And obviously there are plenty of demographic factors at play here too.) And that’s pretty much why I can’t take the charts seriously as a gauge of popularity.”

I know exactly what Lex means, but is he right to say “obviously there are no facts and figures”?

The question of how you measure the popularity of music is a knotty one. The charts have always been an imperfect solution – they don’t take how much buyers like a record into account, the separation of album and single makes it often hard to work out the popularity of a particular song, the sample of shops used to calculate them used to be grossly unrepresentative, and so on.

Now, as Lex says, with illegal downloading the charts simply can’t claim to be an accurate measurement of anything other than “tracks sold”. But measurements don’t need to be exact to be interesting – there’s no problem with an unrepresentative measurement if we can account for the ways it’s unrepresentative, and there’s no problem with an indicative measurement if we want to know the answer to a broad question. The question “Which is the most popular song in the country?” is fairly broad.

The problem of illegal downloading is that the taste profile of illegal downloaders and legal downloaders may be very different. Back in the days when only a certain number of sales outlets counted towards the charts, this was a real issue. The kinds of music being bought in, say, specialist dance music shops and mainstream shops were very different, and the charts only monitored the latter, on the assumption that they could generalise outward. A lot of the time, they could – the sales of most specialist records were miniscule. But sometimes sales weren’t miniscule, and it wasn’t uncommon for artists to report enormous hits through independent shops that were completely ignored in the actual charts.

Is something similar happening with downloading? Luckily, we have a couple of ways of checking.

We can, for example, look at what’s being downloaded on BitTorrent sites, which handily let you sort by number of seeders. Here the most popular acts are bands like Lady Gaga, Black Eyed Peas, Eminem – exactly the kind of global stars also doing well in the ‘real’ charts. Advance albums also do well here, of course – on the Pirate Bay Dr Dre is doing great “business” while Vampire Weekend’s new record isn’t getting a lot of traction.

BitTorrent stats suggest that at the top end the profile of illegal downloaders isn’t very different from the profile of legal downloaders. What about listening to songs, though? Here we can look at Last.FM for some data. Now, last.fm is problematic because it’s very clearly biased: its users are well known for deleting plays of particular tracks (mostly pop ones) from their libraries, and its safe to say that the service overweights rock and indie music. So for Last.FM what we’d need to do is look solely at a genre that is well-represented on the service, like rock, and see if its profiles differ much from how you see that genre behaving in sales charts. And here again we see acts like Muse, Coldplay, U2 and Depeche Mode – the big album and ticket sellers – coming out as most popular on the last.fm service.

So we can’t prove illegal downloaders have similar tastes to legal downloaders. But we can find information which points to that as a hypothesis at least. Big is big, whether paid for or not.

(Of course, our hypothesis only works at the top end, which can cope with wide margins of error. Further down the scale – asking, for instance, whether Joy Orbison’s “Hyph Mngo” is as ‘big’ as Little Boots’s “Remedy” – you do run into fog a bit.)