Study does not show that disconnection threats terrorized France into using iTunes

IFPI, the international trade group for the record industry, has trumpeted a study that allegedly shows that France saw a surge in iTunes sales following the institution of a mass-scale regime of "disconnection warnings" -- threats to remove you and your family from the Internet if you don't stop downloading. These warnings are the first step of the controversial HADOPI system, which is the first of a series of global "three strikes" laws pushed for by IFPI.

TorrentFreak had a look at the study, which was written by researches at Wellesley College and Carnegie Mellon, and they found that none of the benefits claimed by the record industry were in its conclusions: "What the researchers found is that in France, compared to five other European countries, more music was sold through iTunes. Looking at the graph below (from the report), it’s clear that the “uplift” in France before Hadopi was introduced (March 2009) is actually much sharper than the two years after."

“We also estimated the model for the 6 months before and after September 2010, as this was the first month that HADOPI began sending out first notices. In this case, the resulting coefficient was close to zero and statistically insignificant.”

Indeed, when the three-strikes warnings were actually sent out, there was no effect on iTunes sales compared to the control countries. This is unusual, because you would expect that the hundreds of thousands of warnings that went out would have had more of an impact than the ‘news’ that this could happen in the future.

In addition, if we look at the search trends for Hadopi and The Pirate Bay we don’t see a drop in interest for the latter, suggesting that the interest for pirated goods remained stable.

Anti-Piracy Warnings Have No Effect on iTunes Sales

• [pKp]

  HADOPI is a joke. I know no one here that takes it seriously, and among the “pirates” I know, not one of them has diminished its illegal downloading. Most of the “piracy” these days is streaming anyway (although now that they’ve closed Megavideo, people might return to good ol’ torrents).

• [Toxa]

  A bit of wishful thinking this time Cory. March 2009 is exactly when France started to distance itself from the control group…

  • [Mordicai]

    Which is what the article said…comparing that to HADOPI’s “start date” of September 2010.

    • [penguinchris]

      Right, but its pending existence was known well before September 2010. A logical guess might be that it started getting a lot of press around March 2009.

      Well, that’s not even a guess because that’s what the Google Trends line on the graph shows.

      I won’t claim there are no problems with any of this, but based on the provided data you can’t say that the introduction of HADOPI to the public’s consciousness is not correlated directly with an otherwise anomalous increase in iTunes sales in France.

      This is the internet, and we all know that correlation does not necessarily imply causation. It’s far from an invalid interpretation in this case, however.

      Note that I don’t defend HADOPI and such laws. I’m on TorrentFreak’s side here. But, I don’t necessarily agree with their interpretation of the data and the assumptions they’ve made to arrive at their conclusion.

    • [Toxa]

      I stand corrected, thanks.

      However, there is a clear correlation between HADOPI and iTunes’ growth that cannot be dismissed just because nothing happened in the exact moment the system was implemented – anticipation plays a major role in human behavior ( not exactly related, but check http://boingboing.net/2011/10/10/what-reward-does-your-brain-actually-seek.html ).

      • [cheevey]

        Did you go to the link to the full article and read the correlation to Spotify ?

        • [Alan Twelve]

          Yeah, anyone who wants to argue the effectiveness of HADOPI using that graph is going to have a job explaining away the Spotify correlation.

          Spotify was made available for free and without invitation in the UK and Spain in Feb 2009 – just before the gap with France appears.  The same service wasn’t available in France till Feb 2010, since when, the gap appears to have been narrowing, despite the first HADOPI warnings being sent out in Sept 2010.

          The fact that the HADOPI warnings appear to have had no effect is pretty damning too.

• [Fnordius]

  I think the biggest proof that copyright cartels like IFPI don’t understand the internet is how they think they can still make stuff up as if no one could call them on their bullshit.

• [Steve]

  I had a look at the original article, and found an interesting fact. They make some guesstimate of the total amount of money the HADOPI effect provided to iTunes. Let’s read the sentence in the original paper (p. 21):

   ”we believe that the €13.8 ($18.6) million per year increase in French iTunes revenues suggested in our results should be taken as a lower bound on the total effect that HADOPI has  had on music industry revenues” 

  What was HADOPI budget for 2011? €13,79 milion.

  PS: The paper: http://electronlibre.info/IMG/pdf/HADOPI-IFPI-FINAL.pdf and PC Inpact article about HADOPI budget: http://www.pcinpact.com/news/61778-hadopi-budget-2011-offre-legale.htm

  • [Paul Renault]

    Thanks, Steve.  I like to keep track of  instances of government spending that is, essentially, a means to transfer taxes to ‘them that has’.

• [Daniel]

  I use iTunes, I also get music by “other” means – often because iTunes refuses to sell to me.

  I discovered a Swedish singer (Ulrik Munther) on Youtube – his videos are filmed in Los Angeles. He has an album out, all songs are in English. I went to his website and clicked on the link to buy the album through iTunes. It tells me I have to connect to the iTunes store for Sweden. Okay, no big deal, charge my account in Kronors and I’ll still get my credit card charged. No deal, iTunes will only let me buy things from the American iTunes store. So I downloaded the album by other means and the artist and the label are out the money that I was willing to pay.

  Same issue with Ronan Parke from last summer’s Britain’s Got Talent – again discovered via Youtube. Simon Cowell released his album but is only selling it to people in Britain. iTunes has it, but won’t sell it to American accounts. Again the artist and the label are out the money.

  It seems to me that iTunes needs to start thinking globally and realize that national boundaries don’t mean much to the Youtube generation. Start selling globally and stop driving customers such as myself to the pirates.

  • [Jer_00]

    I’m not one to defend Apple in general, but I doubt that they are the ones making this decision.  The labels have always enjoyed the ability to have different distribution contracts for different countries and to squeeze every single penny they possibly can out of those contracts.  They still haven’t gotten used to the idea that someone in Sweden can just buy something offered for sale in Brazil without the need to go through a local Swedish reseller.  And they refuse to adapt because that would deprive them of a revenue stream that they’ve gotten used to tapping.

    I imagine that if it were up to Apple they’d offer everything in their catalog to everyone in the world (where allowed by local law, of course).  Apple is a company that more than anything else likes to make lots and lots of money (as do all for-profit companies), and unless someone can point out a way that they might even fantasize about making money by not allowing sales on a country-by-country basis I just don’t see them being behind this particular brain-dead decision.

• [Simon Papon-Jousset]

  Le Monde has better infographics, showing that songs sold on itunes correlates to the numbers of iphone sold in France but not to the Hadopi law:

  http://www.lemonde.fr/technologies/article/2012/01/24/hadopi-source-de-la-croissance-d-itunes_1633919_651865.html

  • [Alan Twelve]

    I was just about to post asking if anyone knew if the report compared iDevice ownership in France and in the control group.

• [RuthlessRuben]

  As my old boss used to say: “Never trust any statistics unless you’ve falsified them yourself.”

  It’s really pretty amazing how, in the academic environment, integrity commissions x-ray the most minute detail of any study for incongruities, when in the “real world” that everybody rants about so much, statistics and studies are manipulated so blatantly that even a marmoset with congnitive disorder could see through.

  Problematically, according to the average-crowd-IQ rule of thumb, the average IQ of the market as a whole might actually be below that of a marmoset that stopped breathing a month ago.