No Rock And Roll Fun

WHELPING: It took them long enough, but the RIAA has finally got round to trying to rubbish the recent investigation which suggested that illegal downloading has no real effect on music sales. They wheeled out Amy Weiss, the "senior VP of communications" (quite a role in a little trade assocation, that) to blow the thing out the water. What have you got, Amy?

First, she dismisses it as "counter-intuitive" - in other words, it runs contrary to what the RIAA needs to believe. But a lot of people have been saying for a long time that downloading helps, not hinders, music sales, so it's not counterintuitive to what our gut instincts believe. Besides, even if it was - since when has a scientific study been without value because it doesn't prove an assumption? That's what science is for, isn't it?

Second, the report is "anomalous" because it doesn't agree with "five other studies of P2p activity." Now, we've spent some time poking about the RIAA website - it's actually online at the moment - and can't find any figures relating to this, so we're assuming that Weiss is talking about opinion poll type surveys, which aren't as rigorous as the sort of investigation undertaken by Oberholzer-Strumpf. (For example, the ludicrous BPI poll from a couple of weeks ago where a sliver-thin sample group, and a bunch of leading questions still produced "facts" that need to be spun more than buttermilk in a churn to try and help the BPi justify it's planned bullying of its customers.)

Then, there was this: "We look forward to what other academics will have to say about since it has not yet been peer-reviewed or published in a journal," Weiss writes. "We also look forward to understanding what the authors actually did in the study, since the text of the analysis is incomprehensible to the layman."

Um, Ms Weiss? It's a scientific article - you don't tend to write that in the sort of bright clunky language you get in, say, RIAA press releases. If you're having trouble understanding the words, why not ask the actual President of Communications to explain it to you? And if you really are too stupid to understand it, how can you claim its findings are counterintutive and anomalous? Either you're able to make out their case, or you're not. You can't say that something is rubbish if you don't even know what way to hold the pamphlet up.

But, seemingly, Weiss has managed to make out some of the shorter words:

The RIAA also says the study is skewed because the team used the fourth quarter of 2002 as a basis for its findings.
"It is not possible to examine record sales and downloading for 17 weeks and determine whether or not downloading has harmed sales over the last three years," Weiss writes.

Well, actually, you can - it's called sampling, and it's a fairly common process. The music industry uses it quite a lot when it's testing out its marketing campaigns, for example, or when it asks 1000 people about their habits with regard to downloading and then multiplies the results upwards to draw conclusions about the whole picture. I'm thinking, for example, of those surveys the RIAA released to "prove" that the first wave of lawsuits against customers had lead to a decrease in dowloading.

The study team says that a full explanation for the recent decline in record sales is "beyond the scope of this analysis," but then posits "several plausible" reasons, without providing statistical backup.
It suggests "poor macroeconomic conditions"; a reduced number of album releases; growing competition from other forms of entertainment; a reduction in music variety due to radio consolidation; the cost of independent promoter fees to gain airplay; "and possibly a consumer backlash against record industry tactics."

Not sure what this is meant to prove - they offer some possible explanations that would explain falling sales, but they're being rubbished because they don't "provide statistical backup"? But they're not setting out to explain the reasons for the fall in sales; they're explicitly investigating the effect of file sharing on the sale of music. To damn a study for not providing details about something outside its remit is at best stupid, and at worst very, very stupid.

Russ Crupnick, president of NPD Music, says several studies by his group, using multiple methodologies, have produced opposite conclusions.
"Everything we've looked at here sharply disagrees with the results" of the Oberholzer-Strumpf study, Crupnick says. A recent NPD study shows a 29% decline in units sold in 2003 due to P2P downloading.

NPD are basically a polling organisation. They might have used different "methodologies", but it mostly comes down to asking people questions. Comparing their methods with a proper scientific study is like putting a bloke holding his finger up in the wind with the Meterological Office. It's not to denigrate what NPD do, but they're an organisation designed to find information to sell to businesses. It's not surprising they find the sort of information that satisifies big businesses.

So, that's the best the RIAA has to offer - it's got big words in it, it doesn't agree with what we believe, it doesn't match what the pop science of market research says, it only covers 17 weeks. They don't, you'll notice, stand up and offer an equally rigorous set of data of their own. Because they're relying solely on their intuition. But by now, even Vice President of Communications must start to have their doubts that downloading is killing music.