piątek, 20 Październik 2017

(English) EC Diagnosed with © ‚Ostrich Syndrome’: Missing Study on Piracy

Przepraszamy, ten wpis jest dostępny tylko w języku Amerykański Angielski. For the sake of viewer convenience, the content is shown below in the alternative language. You may click the link to switch the active language.

German MEP Julia Reda (Greens/EFA) published a ‚new’ copyright study [PDF] from the European Commission (EC) titled „Estimating displacement rates of copyrighted content in the EU”.

Yes, you’re reading the 1st paragraph correctly, an MEP published a study from the EC. If this sounds weird to you, that’s about right, and we know the feeling. So, before we dive into sharing some of the study’s findings, let’s first give you some insights on why this study is not so ‚breaking’ as one would assume, and how and why MEP Reda needed to dig it from under the sand and publish it online herself.

The Ostrich Syndrome: © Evidence Buried in the EC’s Lost & Found Drawer

This story starts with the EC’s Directorate-General for the Internal Market (DG MARKT) issuing a tender mid-August 2013 for a study on ‚estimating displacement rates of copyrighted content in the EU’.

In January 2014, the contract award notice got published. DG MARKT attributed the study, for a value of €369 871, to Ecorys Netherlands. Back then, the EC’s copyright unit was still part of DG MARKT, and they were working hard on picturing the virtuous circle of the Internet ecosystem as a ‚tree’ in their early Impact Assessment drafts.

The time limit for completion of the contract was set to 8 months. Ecorys dutifully proceeded with the requested research, including conducting a survey in September and October 2014 of almost 30.000 citizens across 6 Member States,  and its final report was handed over to the EC in May 2015.

In the meantime, a new Commission, under President Jean-Claude Juncker, started out and some reshuffling happened, including the copyright unit moving to the Directorate General for Communications Networks, Content & Technology (DG CNECT), and DG MARKT being re-branded to DG GROW, the Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs.

Now, from here on reality starts to trump one’s imagination. At that time, the EC was still writing-up its Impact Assessment (IA) for its proposal on ‚copyright in the Digital Single Market’ [2016/0280(COD)] that was only published on 14 September 2016.

One would think that this expensive study would provide the EC with valuable evidence. This kind of evidence would indeed be a great input for evidence-based policy making, something which the EC claims it does, especially at a time when they were considering the awful censorship filter provision  (see our video) they ended-up proposing in Article 13 to solve the so-called ‚value gap’ (see our other video). However, when it comes to ‚evidence-based’, it seems the EC has a quite loose interpretation of the concept, and we have a slight hunch that they often inspire themselves from Dilbert’s philosophy to write-up their IAs.

Looking closely at this specific IA, no trace can be found of the Ecorys study. So, it appears Ecorys’ ‚output’ (you can’t really call it input when it gets buried away somewhere and hence doesn’t get ‚in’) got buried away rather quickly.

The extent to which digital consumption of pirated materials displaces legitimate purchases is of fundamental importance for EU copyright policy design. – Ecorys study p. 7

Could all of this have happened on purpose? One could think so, as the study starts by explaining that „the extent to which digital consumption of pirated materials displaces legitimate purchases is of fundamental importance for EU copyright policy design” (p. 7), and perhaps the EC preferred its own ‚design’ based on other starting points (such as protecting old business models).

On 22 September 2016 (less than 10 days after the publication of the EC’s legislative proposal), all of a sudden 2 civil servants from DG GROW’s Chief Economist team publish an academic paper on the SSRN network titled ‚Movie Piracy and Displaced Sales in Europe: Evidence from Six Countries’.

Guess which data is used for this paper: yes, the data collected by Ecorys, according to footnote 3. Data from a study that then still remains unpublished for the general public.

The authors explain that their paper „presents estimates of lost movie sales due to unpaid movie consumption”, and claim to be „the first to provide estimates that are recent, representative of the internet-using population, and cover multiple countries”. It’s not hard to be the first, when you make sure the others never get to the finish.

(…) most damages of movie piracy could therefore potentially be prevented with well targeted policies. – EC Civil Servant’s Academic Paper

They also claim that their „findings have important implications for copyright policy”, but guess what: they come out just after the publication of the EC’s legislative proposal. One of the conclusions they drew is that „lost sales due to movie piracy are substantial” (we’ll get back on that). However, they point out that „virtually all the lost sales of movies are due to a very small group of individuals, most damages of movie piracy could therefore potentially be prevented with well targeted policies„.

This ‚well targeted policies’ point is quite interesting, but somehow the suggestion tdoes not seem to have reached their colleagues from the copyright unit, as they clearly left aside any attempt at a ‚targeted’ effort and went full speed for the general monitoring obligations put forward in Article 13.

Then all goes quiet again about this study.

Until the summer of 2017, when MEP Julia Reda (Greens/EFA, Germany) submits a freedom of information (FoI) request in an attempt to find out more about the fate of this study (and of the substantial amount of money allocated to it). As with almost all FoI requests, especially those related to copyright, the EC plays hard to get, by pushing back with all kinds of evasive administrative manoeuvres they have at hand (‚it’s hard too find’, ‚the scope of the request is too broad’, ‚there’s a signature missing from the hierarchy’, etc.).

Now fast-forward to 20 September 2017, when MEP Reda ends-up obtaining a copy of the study, and  did what was long overdue from the EC’s side, namely putting it online. We still have to wait for the EC’s Publications Office to officially give ‚birth’ to it, which could now perhaps happen sooner than later thanks to the pressure MEP Reda put on the EC [update: the EC suddenly published the study on 22 September 2017].

‚Forgetting’  about (or hiding) a study of almost €400.000 in a drawer is quite baffling! The doctor would call this a typical case of the ‚ ostrich syndrome’. In behavioural finance, this is described as „the attempt made by investors to avoid negative financial information”. Like the proverbial ostrich with his head in the sand, the EC duck away when the bad news hit them in the face that this kind of money actually did not deliver the results they had hoped for to support their policy decisions. So, the Ecorys study ended-up getting buried deep down in the EC’s ‚lost and found’ drawer. Until, MEP Reda decided to pull the EC’s head back out of the sand.

Funnily enough, when asked by POLITICO about the whole matter today, an EC spokesperson told Morning Tech that “this academic study provided important — even if statistically limited — estimates on the displacement rate of copyright infringement … The study concluded that, taking a sample of top 100 blockbuster movies in 6 countries, the displacement rate of illegally downloaded content that would be replaced by legal downloads had illegal content not been illegal is estimated at 40 percent.”… So after cherry-picking the results a first time, the EC did it again?

Now let’s move on to what the study actually has to tell, and what the EC doesn’t want us to find out.

Bad News for Some, Good News for Others: Online Piracy ≠ Lost Sales


The study analyses the relationship between online piracy and sales of copyright protected content. More concretely, the research tried to find out what people would do if online piracy would no longer be possible. If someone would have accessed the content through other ‚legal’ means, this would equate to a lost, or displaced, sale. If they wouldn’t have bothered about the content at all, then actually there would be no displaced sale.  To avoid ‚socially desirable’ responses, the researchers did not put that question forward as bluntly as that, but tried to estimate the displacement rate based on the number of legal and illegal transactions that online users engaged in (p. 31).

(…) the results do not show robust statistical evidence of displacement of sales by online copyright infringements. – Ecorys study p. 4

Spoiler alert: the research found that „the results do not show robust statistical evidence of displacement of sales by online copyright infringements„, so piracy does not always equate to lost sales (p. 4). The EC must have had an ‚OMG’ moment when reading that conclusion, as it was probably not what they had hoped for nor what they are constantly being told (you know that ‚Piracy = Theft’ myth, that’s linked to subliminal messaging, such as, ‚You wouldn’t steal a car’).

The research did find that things were a bit different for movies, but we’ll dive in to that in a second, first lets look at the research questions and approach.


Getting the Findings: Research Questions & Approach


The study’s starting-point were these 2 main research questions:

  1. How do online copyright infringements affect sales of copyrighted content?; and,
  2. How much are online copyright infringers willing to pay for copyrighted content?
See p. 27

To answer these questions, the researchers focused on four types of creative content: (1) music, (2) audio-visual material, (3) books and (4) games. Their research approach relied on a review of the literature, interviews with national authorities and content providers, as well as a survey (p. 35).


The survey collected input from almost 30.000 people (5.000 per country) in September-October 2014 who answered roughly 40 questions. The survey group also included minors in the 14 to 17 year age range, who got a tweaked version of the questionnaire. The researchers point out that this group was actually over-represented to ensure their replies would be representative (p. 12). The survey gathered data from 6 Member States (see p. 7):

  1. Germany;
  2. The United Kingdom;
  3. Spain;
  4. France;
  5. Poland; and,
  6. Sweden.

To ensure that the survey was as concrete as possible the questionnaire even referred to explicit examples of file sharing and hosting services, depending on the most popular illegal sites in the respondent’s country (p. 31).


It’s so 2014: Putting the Findings in Context


We mentioned above that 2 civil servants from DG GROW’s Chief Economist team published an academic paper on the SSRN network on ‚Movie Piracy and Displaced Sales in Europe’, based on the data Ecorys collected in its survey for this report. One of the conclusion they drew is that „lost sales due to movie piracy are substantial”, and we promised that we would get back on that.

Well indeed, the Ecorys survey results found a 40% ‚displacement’ rate, as „for every ten recent top films watched illegally, four fewer films are consumed legally”. However, we thinks this finding should be put in a more actual context based on today’s market reality:

  1. The survey data is so 2014: It’s almost 2018, and thanks to the Internet, content is becoming more and more easily accessible from legal sources. However, in 2014, we still had a long way to go. Netflix only started its European expansion in early 2012, making its arrival in most parts of the EU in September 2014, and it actually only became available in Southern Europe in October 2015. Amazon’s Prime video service, as another example, only rolled out to more than 200 countries less than one year ago, in December 2016. So, the impact of the developing online offers could not be properly reflected in findings dating from 2014, as they were almost non-existing in most EU Member States.
  2. Clinging on to old business models: The research does not discuss the fact that the movie industry is desperately clinging on to old business models, with an inflexible system of media chronology, or so-called ‚release windows’. A subscription video-on-demand (SVoD) provider (i.e. Netflix) in France has to wait 36 months (3 years) after the release in cinemas to include a movie in its catalogue. With the arrival of Netflix in France in 2014 the Government proposed to ‚shorten‚ this period to 24 month, yes that is still 2 years, but due to fierce resistance from the cinema industry no agreement has been reached yet. Compare this to the 12 week period in the US, or 6 months in Germany (check out this interesting write-up on the different periods). So, if people across Europe and the globe can communicate with each other with the snap of a finger, but are hindered to access the content they want to watch due to absurd ‚release windows’, whilst other can already do so, then it shouldn’t come as surprise that people look for other means to access content if there is no legal offer available.
  3. Willingness to pay: The researchers assessed people’s ‚willingness to pay’ and found that, back in 2014, „for films and TV-series current prices are higher than 80 per cent of the illegal downloaders and streamers are willing to pay” (p. 8). More concretly, the average ‚willingness to pay’ in 2014 was €6.90, which was then „slightly below the average market price” (p. 170).  The researchers note that „the low willingness to pay any price suggests that if the film or TV-series were no longer illegally available, they would not have downloaded the film or TV-series from a pay site”, in other words, they wouldn’t have cared about the content anyway and therefore there wouldn’t be a lost, or displaced, sale (p. 170). Again this finding should be put in context, as the researchers only assessed the willingness to pay for the last piece of content illegally accessed online, so they explain that „this is not representative for the average willingness to pay” (p. 32). They also note that differences in purchasing power or in unemployment rates across countries play a role, as they found that „piracy is negatively correlated with employment status” (p. 114). This last factor, can also play a role, seeing that a lot of minors were survey, and their purchasing power is of course not similar to that of adults. Next to this, it should be pointed out that prices to access movies have dropped in the meantime below the  average ‚willingness to pay’ back in 2014, seeing that you can now rent a movie on Google Play at €4.99 in HD quality.

The paper from the DG GROW civil servants did acknowledge that „virtually all the lost sales of movies are due to a very small group of individuals, most damages of movie piracy could therefore potentially be prevented with well targeted policies„. If we put the research findings into today’s context, the situation has dramatically evolved in a positive direction, as legal offers develop, perhaps not as fast as they actually could due to the movie industry sticking to old ‚windowing’ concepts, and prices to access the content drop.

Therefore, well targeted policies, such as harmonising ‘notice-and-take down’ procedures (suggested by both the German Government and the Max Planck Institute), would have been more than sufficient to tackle any concerns of rightholders.

Instead, the EC opted for the most destructive policy possible, as Article 13 mandates that online platforms put in place an automated censorship filter (see our video) to block potential copyright-infringing material under the pretence of solving the ‘value gap’ (see our video on that myth). This proposal unscrupulously brushes aside the e-Commerce Directive and the EU Charter of Fundamental Rights, and freely reinterprets case law from the Court of Justice of the European Union (CJEU).


But but but … What about Music?!


Actually, the movie industry has been very quiet when it comes to Article 13, compared to the music industry, which is the most vocal cheerleader of the EC’s effort. Rightholders complain about the so-called ‚value gap‚, and see this upload filter as the ‚magic wand’ in their get-rich-quick plan. A plan which they sold to the EC as if it would ensure creators getting fair remuneration.

However, it’s actually all about the rich getting richer. This is evidenced by an open letter issued on 6 September by the ‘Fair Internet Coalition’, a group representing European performers, and which clearly states that “[Article 13] will mostly benefit the corporate ‘creative’ business who sign up performers and have the upper hand with respect to how they should be paid in return (…) Transparency measures may only be of use for a small minority of performers with enough leverage to negotiate ongoing payments. However, this is not the case for the overwhelming majority of performers who simply sign buy-out contracts”. The whole discussion seems to be about consolidating a few big ones, to the detriment of all others, including creators.

The researchers actually found that (p. 149):

For music, the overall displacement rate estimated in this study is zero. In particular, the displacement  of physical sales (though with a large error margin) is compensated by a significant positive effect of illegal streams on live concerts.”

Again, a fact that the EC tried to hide, rather than acknowledging it in its policy decisions.


More Good News: The Upside of Piracy


Now to end on a positive note, the research actually found that for games the estimated effect of illegal online transactions on sales is positive – implying that illegal consumption leads to increased legal consumption„. They explain that this could be attributed to the fact that this industry has made actual efforts in converting ‚pirates’ into paying customers by, for example, offering extra perks or levels for those who pay (pp. 14-15).

Also for books, the study concludes that „the estimated displacement rate of illegal book downloads is insignificant„. The researchers also remark that „people seem willing to download or stream books legally but do so illegally because the book is not available online on legal sites” (p. 170), again here the lack of a legal offer and access to content plays a big role.

Finally, the authors also point out that previous studies have shown that there can be a so-called ‚sampling effect’, whereby „people explore creative content via illegal sites before deciding whether to purchase the content legally”, so „this exploration makes people more aware of creative content and illegal online transactions therefore increase sales” (p. 41).

What others are saying


Herman Rucic is Senior Policy Manager in the secretariat of the Copyright 4 Creativity (C4C) coalition. He is Senior Policy Manager at N-square Consulting since September 2010.