iPhone Data Debunks Recording Industry's Report On How French Three Strikes Law Increased Sales
from the correlation-is-not-causation dept
The annual Digital Music Report (pdf) of the International Federation for the Phonographic Industry (IFPI) is a curiously conflicted production. On the one hand, it must celebrate "a healthy 8 per cent increase in our digital revenues in 2011 -- the first time the annual growth rate has risen since records began in 2004 "; on the other, it must continue to push the party line about how the industry is being destroyed by piracy.
The IFPI has a stab a reconciling that contradiction, writing: "The truth is that record companies are building a successful digital music business in spite of the environment in which they operate, not because of it." However, it desperately needs some proof of that statement, because otherwise the simplest explanation is that piracy is not a serious problem, and that the recording industry is thriving, just like the rest of the creative industries.
The IFPI probably thinks it has found some proof in the French HADOPI experience, which, according its report, demonstrates that introducing three-strikes measures against unauthorized sharing boosts digital sales.
A new academic study -- The Effect of Graduated Response Anti-Piracy Laws on Music Sales: Evidence from an Event Study in France, by Danaher et al -- has also found evidence that Hadopi has had a positive impact on iTunes sales in France. The authors studied sales of digital singles and album downloads on iTunes from July 2008, before the law was adopted, until six months after the start of notices. They developed an estimate of what French iTunes sales would have looked like in the absence of Hadopi by studying a control group of similar markets.
Taking a look at the study (pdf) provides some details of how the research was carried out:
The analysis found that French iTunes sales saw a significant uplift at exactly the period when awareness of Hadopi was at its highest, in
Spring 2009, when the law was being debated in the National Assembly. This effect was maintained throughout the period studied. French iTunes sales were 22.5 per cent higher for singles and 25 per cent higher for digital albums than they would have been, on average, in the absence of Hadopi.For this study, we obtained a panel of total weekly iTunes sales units for a number of European countries including France. Our data extend from July 2008 to May 2011, and we observe separately both track unit sales and album unit sales. The data were obtained directly from the four major music labels -- EMI, Sony, Universal, and Warner -- and aggregated to reflect total iTunes sales for the majors.
In an attempt to observe the effect of HADOPI, these sales were compared with a control group of five other European countries that didn't introduce similar legislation: the UK, Italy, Spain, Germany, and Belgium. By looking for differences between these two data sets, the researchers hoped to observe the effects of the three-strikes legislation on sales of digital music, using a Google Trends graph of searches for the word "HADOPI" as a proxy for awareness of that legislation, both before and after it was passed.
The graph of iTunes sales for France clearly diverges from that of the control group, lying consistently above it. The divergence begins around about the time that HADOPI was first presented to the French National Assembly, increases slightly, and then decreases a little after the first warning letters were sent out. From this, the researchers deduce that the discussion around HADOPI caused significantly increased sales of iTunes compared to the control countries:
these estimates indicate that French track sales units rose about 25.5% in the control group after March 1, 2009 but by 48% in France, indicating that French iTunes track sales were 22.5% higher on average than they would have been in the absence of HADOPI. Similarly, album sales units rose by 42% in the control group but 67% in France, indicating that HADOPI increased iTunes album sales an average 25% per week in France.
That's a plausible explanation if you believe that piracy is stopping people from buying digital music, but it's not the only one. The French newspaper Le Monde decided to use the same technique of comparing the rise in iTunes sales with Google Trends, but with a different search term. Since iTunes is intimately bound up with Apple's products, Le Monde thought to take a look at the trend for "iPhone" searches on Google.
What it found were five very pronounced peaks in the French searches that corresponded exactly with five (smaller) peaks in iTunes sales, and also to five well-defined external events: the launch of the Iphone 3GS and iPhone 4, and three Christmas seasons. The effect was so marked in France because it was starting from a lower base: according to the researchers, the average sales of iTunes in France were 450,000 per week, while in the UK they were 2,900,000 per week. So an alternative explanation for those impressive increases in sales is simply the uplift in iPhone ownership generated by new launches and the holidays in an immature market with plenty of room for growth.
The researchers do offer one other piece of evidence for the uplift in sales being due to the crackdown on piracy:
EMI surveys of French citizens show that that Rap and Hip Hop are the most heavily pirated genres, even relative to popularity in legal sales channels. While Rock and Pop experience average levels of piracy, the data also indicate that genres such as Classical, Christian, Folk, and Jazz experience significantly lower levels of piracy.
Therefore, they argued, if the increase in sales were due to reductions in piracy, they would expect "the increase in Rap sales to be larger than that for Rock and Pop and the increase for Classical, Christian, etc. to be quite low." And that is precisely what is observed. Conclusive proof? Maybe not.
As the Le Monde analysis points out, another explanation is that many recent iPhone purchasers are younger people, who are generally the most interested in acquiring the latest technology as soon as it comes out. And younger people, by and large, listen to more Rap than Classical or Christian music, which would explain the difference in the increase across genres.
Spending so much effort here on exploring one research report might seem excessive, but it matters. The IFPI is already branding this supposed increase in digital music sales -- quantified by the researchers at $18.6 million annually for France -- the "HADOPI Effect". In the months to come, you can bet that the recording industry's representatives and lobbyists will be visiting governments and showing them this "proof" that three-strikes really "works" -- and demanding they follow suit to "protect" the artists.
What's ironic is that the IFPI report spends many of its pages discussing a much more sensible way of reducing unauthorized sharing: offering high-quality music streaming services instead, as recent market research from Scandinavia indicates. Unfortunately, the recording industry is so obsessed with punishing pirates that it can't see that its future lies in promoting innovation, not legislation.
Follow me @glynmoody on Twitter or identi.ca, and on Google+
![Hide This Hide this](https://ii.techdirt.com/s/t/i/topclose.png)
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Reader Comments
Subscribe: RSS
View by: Time | Thread
Study vs Studies
[ link to this | view in chronology ]
Re: Study vs Studies
Not so fast. The 22% figure considers EIGHTEEEN (18) months. The result I got considers only 12 months. So you see, I can wait till hadopi is 2 years old and tell it made the sales increase , say, 35% and I wouldn't be really lying, it's just I won't say directly how I got to those results (yea, it is lying but you got the point). MAFIAA is pro at that and even though it's fairly easy to debunk the mainstream media will swallow it and vomit all over the ppl. And ppl don't just go digging and checking all data they receive. So it's pretty good Le Monde is raising awareness of the lies from the MAFIAA.
[ link to this | view in chronology ]
We'll see.....
[ link to this | view in chronology ]
Re: We'll see.....
The point this article was making is that there could be many reasons why sales increased. HADOPI may have had an effect, but so might normal retail buying trends like the introduction of new iPhone models and higher overall rates of spending normally associated with the holidays.
[ link to this | view in chronology ]
Re: We'll see.....
So your argument boils down to: "because there was a comet sighted, we must all believe that [insert inane statement]".
Your logical argument has me convinced. /sarc
[ link to this | view in chronology ]
Boom, end of profit for every piratebay wanna be.
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re: Re:
[ link to this | view in chronology ]
lame phone
[ link to this | view in chronology ]
Re: lame phone
This was a study done for the IFPI, not for Apple. The only thing that Apple have to do with this is that they provided digital sales data from iTunes. What was done with that data was up to the IFPI.
And as for the iPhone-bashing... I'm getting sick of it. I hate Apple as much as anyone, and think that too many people buy the iPhone because of the brand (and not on features or price), but I'm getting sick of the iPhone hatred all across the Internet. If Apple hadn't brought out the iPhone, the smartphone market would be years behind where it is now.
[ link to this | view in chronology ]
Re: lame phone
[ link to this | view in chronology ]
iphone peaks don't explain findings
I do see that each of 5 peaks in iTunes sales correspond to peaks in iphone searches (which may or may not correlate to actual iphone sales). This is really beside the point though, even though everyone seems to be focused on it. French iphone sales seemed to move the same as the authors' "control group" sales over time, right up until public awareness of Hadopi. Immediately following public awareness, itunes sales in France rise above the control group *and remain* elevated at that level. It's not just Christmas peaks in France. Or peaks resulting from iphone releases. It's a consistent elevation of French sales that begins exactly at the same time as Hadopi and is then maintained.
The only way that iphone sales could explain this is if the French started buying way more iphones right around March 2009 and these iphones drove itunes sales sales (and iphones increase sales of rap more than other genres).
To see the point, look at December 2008. That's Christmas, but it was before Hadopi. The French peak is no higher than the control peak. The French Christmas peaks are only higher than the control after Hadopi. So this Christmas argument seems to miss the point.
I don't think the study is absolute proof that Hadopi increased sales, but I don't think the Le Monde criticism is a valid one and I think the study is more powerful than the "copyleft" is giving it credit for.
Thoughts? Am I missing something?
[ link to this | view in chronology ]
Re: iphone peaks don't explain findings
Check the "Le Monde" article and you will see that indeed the French became more interested in iPhones than the rest of the control group. And they became more interested about at the same time (but still slightly before = consistent) as the iTunes sales got higher in France than in the control group.
And when you buy the media to buy on iTunes, then you keep it and you keep on buying on iTunes (or at least, you are more likely to).
These two studies prove only one thing: When the market is indeed changing and a new media or technology or trends makes its apparition, anything else that is studied during that unstable period will produce non-reliable and fake results.
Unfortunately for them, it is undeniable that the iPhone (and then other smartphones) changed many things (micro softwares, games...) about the existing economy and markets.
[ link to this | view in chronology ]
Re: Re: iphone peaks don't explain findings
[ link to this | view in chronology ]
Surely people can see countries that have zero protections in a matter of fact growing in spite of granted monopolies that obstruct growth.
[ link to this | view in chronology ]
Thanks, France.
I guess you really enjoy living dangerously...
[ link to this | view in chronology ]
[ link to this | view in chronology ]
[ link to this | view in chronology ]
[ link to this | view in chronology ]
There's a hundred factors that could account for the differences in sales that have nothing to do with HADOPI. In addition to those mentioned above, were there major hits by French artists during that period? A few major albums in the French language that wouldn't appeal to a UK or German audience would skew the figures (that is, most French music). What about retailers? Major supermarkets in the UK quite often have offers where they sell, say, two �10 iTunes gift cards for �15. A few of those offers in French stores could account for a sudden rise in sales.
Overall, it's the same story told over and over again here. There's a correlation that's mild at best, and all the pro-corporate drones rush to it as though it's hard proof that we need to destroy our rights rather than have corporations change to fit the modern marketplace. The correlation falls apart the moment you apply logical reasoning to it, of course.
[ link to this | view in chronology ]
[ link to this | view in chronology ]
Ahhh...
Really does pop up often, doesn't it?
[ link to this | view in chronology ]
The data while not conclusive is very persuading. The Le Monde arguments are weak, inconclusive and may infact only demonstrate a conincidental relationship to music sales. From my unbiased perspective it seems that awareness of the three strikes law might actually have reduced piracy, if for no other reason than it made people aware that there were punishments for pirating music.
[ link to this | view in chronology ]
Re:
Lol
To paraphrase:
"I don't like the way that the conclusion goes, so I'll ignore it and call it a weak correlation. I'll go off and accept my preferred narrative without question, even though it's equally as flawed and inconclusive"
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re: Re:
[ link to this | view in chronology ]
For example, people say that a few French hits could have caused this. If you read the study, the authors say that this same effect/pattern is observed for each of the 4 music labels individually. Clearly the originator of that critique never even read the study, and just wants to debunk it.
The Christmas argument doesn't seem to hold water since the 2008 Christmas has no French peak (over the control group) but the 2009 and 2010 do.
Everyone keeps saying "correlation is not causation" - that is exactly why the authors use a control group and then also bring in their genre analysis.
Is the proof conclusive that it is causal? No. Is the proof much better than the theological copyleft concludes? Absolutely - they just don't seem to understand this paper. Did most of the people critiquing the paper read it? It doesn't look like it.
I think it's very reasonable to say that this study is not conclusive proof. I think it's also reasonable to say that this study is not pro-Hadopi and simply measures the Hadopi effect (which may or may not justify legislation). But I think pointing out Christmas peaks or making statements about French artists having a few hits completely misses the point of what the authors were doing.
[ link to this | view in chronology ]
iPhone 5s/5c recovery software
[ link to this | view in chronology ]