File sharing not killing music after all, it turns out

It's not Girls Aloud coming to a graceful end; that other music industry trouper, the claim that file sharing hurts music sales, has pretty much been laid to rest as well: Extensive research from the European Union confirms what a lot of us have thought all along: unlicensed listening doesn't really take money from anywhere else:

The goal of this paper is to analyze the behavior of digital music consumers on the Internet.Using clickstream data on a panel of more than 16,000 European consumers, we estimate theeffects of illegal downloading and legal streaming on the legal purchases of digital music. Ourresults suggest that Internet users do not view illegal downloading as a substitute to legal dig-ital music. Although positive and significant, our estimated elasticities are essentially zero: a10% increase in clicks on illegal downloading websites leads to a 0.2% increase in clicks on legalpurchases websites. Online music streaming services are found to have a somewhat larger (butstill small) effect on the purchases of digital sound recordings, suggesting complementaritiesbetween these two modes of music consumption. According to our results, a 10% increase in clicks on legal streaming websites lead to up to a 0.7% increase in clicks on legal digital purchases websites. We find important cross country difference in these effects.

So - it differs a bit from country to country, but broadly speaking, there's a tiny positive effect, and - oh joy - even legal streaming leads to a small rise in actual sales.

Even if you accept that the upswings are as small as to be zero, that's still no negative effect.

You don't, of course, take down one of the RIAA-IFPI shibboleths with data and expect the music industry to admit they've been wrong. So they try a response to tear the research to shreds. It doesn't quite work:

IFPI believes the JRC study is flawed and misleading. The findings seem disconnected from commercial reality, are based on a limited view of the market and are contradicted by a large
volume of alternative third party research that confirms the negative impact of piracy on the
legitimate music business

Let's just look at that a little closer, shall we?

"Flawed and misleading" is quite a big claim - one suggesting incompetence and one duplicity. You better have something strong and convincing to follow that up.

"The findings seem disconnected from commercial reality". And, really, they don't. "This survey doesn't confirm what I believe, so it must be wrong" is just piffle.

"are based on a limited view of the market" - this seems to be the objection that the survey takes no account of legal streaming. Except it does. Look, that bit we quoted up there makes it clear that the impact of legal streaming on the download market is part of the thing they were investigating. Bit misleading to complain that research hasn't investigated the effect of legal streaming on legal streaming, isn't it?

"and are contradicted by a large volume of alternative third party research that confirms the negative impact of piracy on the legitimate music business" - nice to see the music industry calling itself a legitimate business. There's a pleasing Corleone ring to that.

I think this existence of contradictory data is upon which the IFPI are building the claim of the survey being misleading.

Trouble is, two of the three surveys are referenced by the EU survey - their findings detailed, and proper references given; the intention being that people read them alongside the EU findings.

To not have included the research which contradicts their findings would have been misleading. You know, like if the music industry issued a press release which only cherry picked research which agreed with them and ignored the research which took an opposing position. Like this one.

So what of the flaws? The IFPI details a "key example" - which apparently is so key they need no further examples:

A key example of this problem is the treatment of iTunes ‘clicks’ by Nielsen. iTunes is a major legal music service and an essential data point in establishing legal music consumption. Nielsen measures use of the iTunes application, which involves any activity around iTunes - such as a user simply plugging their iPhone into the PC (which launches the iTunes application), a user listening to music via iTunes, a user synchronising their Apple device with their PC, a user renting a film on iTunes, or downloading a game app. Each one of these instances are counted as an iTunes ‘click’ and considered as legal music behaviour by the JRC. This severely impacts the results and is not a good proxy for legitimate music consumption

Does Nielsen really count someone launching a desktop application as a click on a website? If that's true, then the methodology flaw is Nielsen's. (I've dropped an email to the IFPI to check where they're getting this definition of a click from.)

Even if we take this at its word - although people who still sync iPhones with PCs are hardly going to have much of a grasp on modern technology - there's no indication of how "severely" the impact of this is. (Really? Every time someone starts and stops a song and starts another one of their desktop iTunes player Nielsen adds a click on iTunes? That's extraordinary if true. Is it?) The IFPI might be right, this could undermine the findings. But rather than give us any data which could be set against the research, the IFPI just shrug and say they reckon it's just enough to know it could be flawed.

If they were interested in the truth, the IFPI would be pointing to a better source of the music interaction data. That they don't suggests their takedown is a panicky attempt to get everyone to look the other way.

[UPDATE: Took a while, but finally got a confirmation from Nielsen that their methodology can, indeed, include people recharging their iPhones as a visit to iTunes. I don't think that fatally undermines the EU report, but it does weaken it to the point where it's worth approaching the findings with a bit more caution. I've written more about this here.]