A recently-released report from the French government, Rapport sur les autorités publiques indépendantes (Report on the Independent Public Authorities), includes a section on HADOPI (Haute Autorité pour la diffusion des oeuvres et la protection des droits sur internet), the regulatory body set up to oversee France’s “graduated response” law for issuing warnings and potentially punishments to online copyright infringers.

The headline that most Anglophone writers took away from the 24 pages in this report that were devoted to HADOPI was “HADOPI’s budget to be cut by 23%.”  These writers took their cues from anti-HADOPI statements by various French politicians — including new French President Francois Hollande — and mischaracterized a statement about HADOPI by the French culture minister, Aurélie Filippetti.

Unfortunately, none of these people appear to have actually read the government report.  (Yes, it’s in French, but there is Google Translate.  I used it.)  HADOPI is not on the way out; not even close.

Let’s get the most obvious facts out of the way first.  Yes, HADOPI’s operating budget is being cut from €10.3 Million to €8 Million, but its headcount is being increased (from 56.2 to 65.2 FTE).  Apparently the budget cut reflects the fact that HADOPI’s ramp-up period is coming to an end in 2012, and the focus is being shifted to increasing operational efficiencies and cutting overhead.  Moreover, HADOPI’s purview is being expanded to include video games as well as music and video content.

Another bit of factual cherry-picking in the Anglophone press: HADOPI has merely sent out more than a million emails but only prosecuted 14 people and only fined one (less than €200), so therefore it must be a big waste of money.

On the contrary: all of the data in the report, as well as the conclusions it draws, point to an agency whose successes are outnumbering its failures and whose mission is quite properly being optimized.

As it turns out, HADOPI has several objectives, not just issuing warning notices to illegal downloaders.  Those other functions are where HADOPI does not look as successful as hoped.  One objective is to increase the number of legal content offerings in France.  To do this, it has put a labeling system into place, along with a website called PUR (Promotions des Usages Responsables, also an acronym for the French word for “pure”) that lists all of the labeled services.  Although the report cites a sharp increase in the number of such services in France over the past year, that increase is surely attributable to market forces and is no different from similar increases in other countries.

Another of HADOPI’s objectives is to regulate the use of DRM technology according to rules derived from the European Union Copyright Directive of 2001.  This means both ensuring that DRM systems don’t unduly restrict users’ rights to content and that DRM circumvention schemes (hacks) are prosecuted under the law.  So far, HADOPI has only been asked to intervene in two DRM disputes concerning users’ rights, and both reviews are ongoing.  This can’t be counted as a great success either.

Yet regarding HADOPI’s core “graduated response” function, the data in the report shows nothing but success so far.  Fining people (a maximum of €1500) and suspending their Internet access (up to one month) is not the objective; reducing copyright infringement is.  The number of people who have been fined or had their Internet access suspended is simply the wrong metric.

The good news is that HADOPI appears to be succeeding as an education program rather than as a punitive one.  In 2011, HADOPI reports that fully 96% of people who received a first warning message did not receive a second one; this number stayed about the same in 2012.  In addition, the percentage of people who received second notices but not third ones rose from 90% to 98% from 2011 to 2012.  (The legal steps that could lead to fines or suspensions begin after the third notice.)  To buttress this data, HADOPI has published results from four independent research reports that note significant decreases in illegal downloading in 2011.  No one has substantively debunked any of these findings.

Furthermore, HADOPI does not simply take complaints from copyright owners — which monitor the Internet and submit complaints to HADOPI — at face value.  For more than half of the users who received three warnings, HADOPI chose not to send the cases to French authorities for prosecution.

It is also interesting to note that the educational aspect of HADOPI appears to be succeeding despite the fact that it treats violations as misdemeanors, with small punishments, in contrast to the enormous criminal penalties associated with copyright infringement in France (as they are in the U.S.).  This points to the conclusion that online education is more effective than large statutory damages in curbing infringement.

Now let’s talk about the economics.  Ideally, this type of program would be funded by copyright holders — the ones with rights that they want protected.  France is funding HADOPI with taxpayers’ money, although copyright owners do pay for the monitoring services that detect allegedly illegal downloads and report them to HADOPI.

At the same time, €8 Million isn’t a bad deal.  France currently has about 50 million overall Internet users and about 25 million fixed broadband subscribers.  Let’s assume that the total number of French people who pay for Internet subscriptions is about 30 million.  In that case, HADOPI’s annual budget could be apportioned as a levy on Internet subscribers of about €0.27 (US $0.35).  This is two orders of magnitude smaller than the £20 (US $32) annual antipiracy levy on ISP subscribers that the Digital Britain Report proposed for the UK in 2009.

Furthermore, HADOPI measured the market impact of unauthorized downloading (not counting P2P) as €51 to 72.5 Million annually.  Although this figure can’t be taken as a magnitude of lost sales, the worst case break-even point for HADOPI’s cost-effectiveness would be that 16% of illegal downloads displace sales (one study attempted to measure promotional effects vs. sales displacement and suggested that about two-thirds of illegal downloads displace sales).

It’s still too early to proclaim HADOPI’s success or failure.  For example, the more determined infringers could move to ways of obtaining content that evade detection (e.g. HADOPI only deals with downloads and not streaming).  But the signs are encouraging enough that the French government has decided to keep the experiment going.

(By the way, if you would like to argue with me about this, I will be in Paris from November 7 through 11, speaking at the SNE conference “Les assizes du livre numeriques” on Thursday November 8.)