The Hadopi - the French public authority in charge of the 3-strikes measures - released statistics last week, claiming a massive success. Indeed, the figures for the apparent reduction in unauthorised copying have been praised by President Sarkozy himself. But analysis by the French newspaper Le Figaro, and the technology website Numerama, suggests that the numbers have been massaged to support the President's political position prior to the elections.

In essence, whilst Hadopi claims that unauthorised downloads have significantly reduced, the evidence reveals that users have simply switched.

The analysis by Le Figaro shows that there is an inverse relationship between peer-to-peer traffic and streaming traffic. As peer-to-peer declines, so streaming grows.

Le Figaro takes one set of figures from Mediametrie. According to these figures, there were 4.5 million file-sharers in France in December 2010, which reduced to 3 million in December 2011. It then takes an aggregate of visitor numbers to streaming sites to show an increase from 6.5 million to 8.3 million visitors over the same period.

When the numbers are placed on a graph, the intersection if the two traffic streams occurs in October 2010, just after the first hadopi warning notices were sent out.

The Figaro analysis is supported anecdotally by information gleaned last week by iptegrity from a representative of France Telecom.

The French telecoms operator is seeing a relative decline in peer-to-peer traffic, but not an absolute decline. That is, as the overall traffic grows, peer-to-peer represents a smaller proportion of the total, but it does not show a decline in respect of the previous traffic levels. At the same time, France Telecom has seen a dramatic increase in streaming traffic.

France Telecom has also noticed a marked increase in levels of encrypted traffic since the Hadopi notice-sending began. This could be an indication that users are choosing to use encryption to hide their unauthorised downloads.

This certainly suggests food for further analysis and leads to a serious questioning of the Hadopi report.

Numerama correctly pointed out that the Hadopi data is funded by rights-holder organisations such as the IFPI and the ALPA ( a French anti-piracy organisation which includes the US film studios and the recorded music companies). These organisations have a direct interest in the "success" of the Hadopi measures, and therefore they have an interest in showing that 3-strikes has led to a reduction in peer-to-peer traffic.

But these rights-holder lobbying groups have absolutely no interest in showing any other change. They want to use these figures to convince policy-makers in other countries to adopt 3-strikes measures.

The political nature of the Hadopi "success" is reflected in a Presidential press release from Sarkozy himself, which claimed that the Hadopi results were 'indiscutable' which translates as 'indisputable'.

I would add a couple of further observations of my own. Hadopi claims that 95% of the subscribers who received a first notice (strike one), did not get a second notice. However, this is not the same as saying that 95% stopped downloading unauthorised material. As the traffic patterns show, it may just mean that they didn't do it using peer-to-peer systems which they know are being monitored.

Hadopi cites four different figures for the decline in peer-to-peer traffic. These figures range from a 17% decline in an IFPI- funded study, to 66% decline in an ALPA-funded study. The huge disparity merits further examination.

Contrary to President Sarkozy's PR, the Hadopi's claims for "success" are indeed very much disputable.

You may re-publish my article under a Creative Commons licence, but you should cite my name and provide a link back to iptegrity.com. Media and Academics - please cite as Monica Horten, Hadopi - has it massaged the numbers? www.iptegrity.com 30 March 2012 . Commercial users - please contact me.