The effects of 2 years of Hadopi "3 strikes and you're out" in France.
Reminder: Hadopi is a dissuasive automatic punishment system that sniffs the bit-torrent network (and other peer-to-peer) for IP addresses transmitting certain contents such as copyrighted songs or movies. It then sends a request to the ISP to give the name corresponding to said IP and finally sends a warning to the person who rents the connection. 3 warnings and you're banned from internet. No need for proof of correspondence between the IP and the person. Note: A late amendment requires the 3rd strike to be reviewed by a legal authority, virtually rendering the systematic nature of the system useless, but it was passed anyway.
It should be clear, from that description, that anyone reaching the banning point is either a tech illiterate, either doesn't check his snailmail.
After a 2 years of operation, it is time to look at the numbers.
Indeed, noboby, nobody reached the GTFO point. I suppose this is due to the extreme slowness and yet unestimated but probably significant cost of enforcing the 3rd strike through a legal procedure. However, in Sept 2011, 500,000 1st strike had been blown, as well as 20,000 2nd warnings. There was, at this date 60 cases being processed for 3rd strike [source (French)].
This all comes at a cost. The automated nature of the process permits significant savings (in particular when there is no legal process), and Hadopi has only cost 23 M€ to the French citizens in its 2 years of existence. On their side each ISP has had to adapt its infrastructure, to a cost of roughly 80 M€, cost that either eats their margins or is forwarded to the customer. which is the same thing from the French economy's perspective. The cost of legally processing the 3rd strike candidates is yet unestimated, however it is likely to be high since it must go through court. We'll round it up to 100 M€, discarding the legal costs for now. For the sake of perspective, this is equivalent to ten million music albums, each costing 10€. [source (French)]
The authorized music sales market in France amounts to 600 M€, CDs, paid downloads and authorized streaming lumped together. Of which, 420 M€ for the sales of CDs [source (French)]. The authorized video market (DVD + Bluray) reaches 1,500 M€ [source (French)]. The total is then 2,100 M€. So, in order for Hadopi to generate profit, it would need to boost the sales by at least 5% to generate more money than it costs the public. However, let us note that half (roughly) of the money channeled through music albums and movies gets out of the French economy, mostly in the direction of the USA economy. So, to start benefiting the French economy, it actually needs to boost sales by 10% yearly. The USA majors, on their side, start raking in at the first percent, since they only had to pay a few 1000s in lobbying to help it through.
note: by "boost the sales", I mean "compared to its previous performance". So it means that if the market is losing 10% yearly, a year without loss is considered a boost of 10%
This is taking the point of view of the French economy only but note that, from the point of view of the citizen, it's a lose-lose scenario, where he pays for the stick that gets him to pay for his content. From the label's point of view, any increase is a definite win, of which a portion is expected to be forwarded to the artist it represents.
However, this point is moot. The sales of plastic disks keep decreasing at the same steady pace of 3% to 5% a year, as it already was before Hadopi, the market for authorized download keeps rising as well, but at the same steady pace as it used to before Hadopi. The video sales market has been fairly stable in the past years [source (French)] and has taken an actual loss of 3% in 2011 [source (French)].
The system has had no effect on the profit charts, and its cost would have been put to better use in pretty much anything else.
But it was not without side effects. As the representative of the system were proudly bragging, the 10 most downloaded movies have lost 60% of their sharers in the course of 2010. Which they hoped would be misinterpreted into a proof of success.
The real consequences of Hadopi
It seems that the people making torrent trackers of unauthorized torrents have greatly overestimated the impact of the system on the people's behaviour, and have stopped serving French-market-related torrents. Instead, a greatly diversified offer of unauthorized direct download has sprung up to supplement it. This is bad news for the pipes of the internet, but good news to trigger-happy enforcers, since it is possible to take down central servers, difficult as it may be. This affects little non-French-centered cultural items torrent availability such as non-French music albums, that are easily found and downloaded fast. Those amount to more than half the demand in music in France (French mostly listen to Anglo-Saxon pop). The same goes for non-French movies, but only for the few French people that speak languages or the ones hardy enough to spend 5 minutes looking for subtitles.
Adendum: the little availability of French-centered torrent might also be explained by the low level of tech knowledge in France.
So, the nation-wide effect of Hadopi, besides scaring a few hundred thousand clueless citizens, is that French users either move to direct download (bad for internet), which is less practical than peer-to-peer but still more practical than getting your ass to the CD retailer, either move to iTunes or Amazon Music Store, which is neither good nor bad for French interests, but great for a few massive foreign corporations. They also might be phasing out French cultural items because their foreign counterpart are more accessible (bad for French artists).
The suggestion that people used to get content online would revert to plastic disks is just preposterous, and the statistics given earlier seem to agree. Many do not own plastic-disk-reading devices anyway.
What the world has learned
As predicted during its inception by anyone with some understanding of netizens behavior, Hadopi failed its intention. It wasted significant amounts and earned a bad name for France abroad in the process. However it's been an interesting experiment with some surprising consequences that I'll try to highlight.
It was successful in curbing peer-to-peer infringement.
It must be reminded that infringement was simply transfered to another medium (direct download and unauthorized streaming mostly) AND that curbing doesn't mean killing. But it is still quite an achievement. Many critics of Hadopi during its inception were of the opinion that it would not dent peer-to-peer usage, and that people would just educate themselves to secure and anonymous ways to keep sharing such as VPN or would devise new anonymous networks altogether. This did not happen. People instead moved their lazy asses to a non-anonymous method that simply sidesteped the repression.
One way law enforcement could react to that is by coordinatingly taking down streaming and direct download servers, which are much more manageable targets than peer-to-peer networks, as shown recently with the takedown of Megaupload. My prediction is that this would work the same way as the successful takedown of Napster or Kazaa back int the day. Alternatives would just appear faster than they can be dealt with. And those alternatives will be harder to uproot.
Another way would be to outlaw direct download and streaming along with peer-to-peer in the same fashion. However it must be noted that it is a much trickier legal feat and reminded that Hadopi already challenges ethics pretty badly. Still, I believe it is more feasible than the first option.
In any case: It is possible to police the internet given sufficient technical and legal tools. China stands as a tremendous success in that respect.
You don't actually need to use the stick
Remember? Nobody was disconnected. The amendment that prevented the original prospect of 1,000 disconnections a day [source (French)] was rushed at the last minute in order for the system not to be unconstitutional, as it had been deemed so by a commission of experts. It rendered it virtually toothless, but it still worked. Because only 60 people out of 500,000 don't panic and pull the plug on the download program. I have good hopes that they will never be convicted. All you need to do is scare them with the stick and they'll all submit. All you need to do is suggest that there might be a stick. As, really, there is none.
As far as I am concerned, this is the most important lesson Hadopi taught us. It's rather scary and does show the people for the flock of sheep they are, and I'm a very sad about it, but it's there and denying it would be stupid.
It was not successful in affecting sales.
Which was the whole point of it. You can curb peer-to-peer all you want, if it doesn't translate into more monies, the content industry just doesn't care. What does generate sweet $$ is selling the public stuff it wants to buy. You might think it goes without saying, but I don't see much effort on the side of labels in that direction. iTunes is not an invention of labels but of the tech industry. The very industry the content industry is so keen on despising. And digital sales created a new market, instead of stealing market from plastic discs merchants. It is just not the same type of people that buy plastic disks and that buy digital. The kind of people that buy digital are actually the ones that seed torrents like crazy. Just provide them with a service they can use, and they'll throw bucketloads of cash at you. Of course, once their content-budget will be exhausted, they'll keep downloading like maniacs, but that doesn't cost anything to the content industry, does it? The content industry can't claim more than what the people are willing to spend, can they? And they are willing to pay. All it takes is giving it to them in a shape they want. I repeat : many of them don't own a plastic-disk reader.
NB: The elements to write this article were obtained through a couple of search engine query, no law was broken in the writing of this article, no animal was harmed either.
Enough already with blog articles?