The authority overseeing France’s Creation and Internet law, Hadopi, recently released its third report on the implementation of the enforcement measures against copyright infringers contained in the law. Iptegrity’s Monica Horten, also an LSE visiting fellow, looks at the numbers in Hadopi’s report and questions whether such measures are really worth the cost.

Hadopi, the public authority charged with administering France’s 3-strikes anti-filesharing law, has just had its third birthday. To mark anniversary, it has released a report covering its activity to date. Interestingly, it reports 1 sole Internet disconnection in 3 years. It also outlines the underlying bureaucratic process, plus an issue with identifying subscribers. A close reading of the report raises questions about the scale and costs  of implementing 3-strikes measures to enforce copyright online. Can it really provide value for money?

Hadopi reports that a total of 1.912 million notifications were sent to French Internet subscribers in the three years of its operation – 2010-2013. That’s strike one of the process. It has sent out a 186,153 follow-up letters – that’s strike 2. There has been just one disconnection – that’s strike 3…

However, we can gain more insights by examining the supporting data. Hadopi actually got over 19 million referrals from the rights-holders who act as ‘agents’ and are tasked with informing Hadopi’s staff about alleged infringing subscribers.

The interesting question is how did 19 million referrals boil down to just one lonely disconnection?

Well, in order to be able to contact subscribers, the Hadopi has to go via the Internet Service Providers. Through that process, the 19+ million was reduced to 7.718 million. It seems that the reason was multiple allegations against the same subscriber. That is already a substantial reduction.

Hadopi then reports that 88% of the allegations were successfully matched against named subscribers. The French IT website Pc Inpact correctly asserts that this means 12% – 920,000 – could not be identified. Iptegrity has checked the maths and it stacks up.

Hadopi states that the difficulty in identifying individual subscribers is due to the practice by the ISPs of  using ‘natted’ IP addresses, where the same IP address is attributed to multiple subscribers.

That leaves 6.7 million identified addresses, out of which only 1.9 million were sent notifications.

Still, how do we get to 1 Internet disconnection?

Following the 186,453   letters sent for strike 2, there were 663 deliberations internally  by the Hadopi, as to whether or not to submit the cases to the court.  In 9 out of 10 cases, Hadopi says, it refrained from submission. Hence, there were only 51 submissions to the courts for penalties. Most of these appear to have incurred a fine of between $50-600. Just one got the disconnection penalty, for 15 days.

Within that process, there were 44 hearings by the Hadopi in 2012-13.

Then there’s all the supporting bureaucracy.  In 2012-13, there were 73 210 contacts by phone or email with Internet subscribers of which 81.73 per cent concerned with the content of the notification. According to another French technology website, Numerama, the reason is that Hadopi   notification does not state the name of the work that has allegedly been downloaded, so subscribers are contacting Hadopi to ask what they are being accused of downloading.

The French technology website Numerama suggests that Hadopi could include the name of the allegedly downloaded work in the notification as a cost-saving measure.

In addition, there’s the so-called ‘educational’ work that Hadopi is tasked with. This is a separate programme from the 3 strikes measures, and it entails taking the message into schools and educational establishments. Hadopi states that in the 12 months to June 2013, it met with 800 teachers, lecturers and librarians in 530 institutions.

Hdaopi’s budget for 2012 was €10.1 million, and a similar amount for 2011. The figure reflects cuts already made by the Hollande government:  Hadopi budget to be slashed as French review 3-strikes

The Hollande government wants to incorporate Hadopi within the Conseil supérieur de l’audiovisuel (CSA).  Effectively, it would no longer exist as an institution, but the measures it oversees could carry on. This could be in part a money-saving measure, getting rid of a layer of bureaucracy. Of course, it means that Hadopi loses face and influence, and Hadopi staff are fighting against it.

All of this will provide food for thought for the British government. Hadopi is a state-funded institution, with control over what it does. The consumer pays indirectly via taxation, of which it is a very small component.

When you split up those functions between different private actors, the administrative complexity increases and with that, the costs. I’ve previously argued that the costs in Britain could  be higher than originally predicted (see The 84 million-a-year bill for DE Act ). Who really wants pay for this? And how will it be passed on to the consumer?

This article originally appeared on 14 October 2013 on Iptegrity.com, and is reposted with permission and thanks. This article gives the views of the author, and does not represent the position of the LSE Media Policy Project blog, nor of the London School of Economics.