A controversial report purporting to show the cost of music, movie and software downloading to the Australian economy has been revealed as the application of a discredited European report on piracy to Australian data.
Copyright industry lobby group the Australian Content Industry Group commissioned the report and provided selected highlights from it to Fairfax, which featured the report, complete with a colour piece on the individual costs of piracy, in the Sunday papers last week. The online news reports were savaged by angry readers until Fairfax quickly closed comments.
I provided the history of copyright industry’s efforts to conjure up “independent modelling” showing the impact of piracy last week. Today, Crikey’s Andrew Crook delved into the mystery around the ACIG report, which remained unreleased ten days on from the Sunday paper splash and which allegedly revealed a $900m a year cost from piracy.
But problematically for ACIG, another copyright industry group, AFACT, had released its own report on movie downloading in February, and its figures were entirely inconsistent with those reported of the ACIG report. The AFACT report, by IpsosMediaCT/Oxford, has been criticised by Electronic Frontiers Australia, but laid out its methodology and the assumptions underpinning its modelling for scrutiny.
Today Andrew Crook managed a feat that so far eluded the rest of us, convincing ACIG’s representatives to provide a copy of the report. The ACIG report, by Sphere Analysis’s Emilio Ferrer, an economist and former Labor staffer, doesn’t attempt to model downloading levels in Australia and their impact, or try to establish, as AFACT’s report did, what proportion of downloads equal lost sales, but applies the results of a similar European study, by TERA Consultants, to Australia and recalculates the European costs of downloading according to local numbers, in effect replicating the European experience – or TERA’s version of that experience – using local data.
But the 2010 TERA report, which looked at a range of European countries, has been savaged by critics. Torrentfreakdissected it, revealing a number of flaws, including simplistic correlations about linear connections between growth rates of internet usage and piracy, assumptions about the negative economic impacts of piracy, ignoring redirection of money “lost” from piracy to other areas of the economy, and projected trends that would eventually mean the loss from piracy was greater than the entire output of the copyright industry.
And blogger Tom Koltai at the Koltais Whatif blog has devoted posts running in to double figures tearing the report apart from an statistical perspective, identifying such problems as how, using the industry’s own figures, a small drop in the amount of time consumers spent listening to music had apparently cost it $23 billion; how the report inflated the size of the industry (something Australian reports all do as well) and how its purported job losses in Europe didn’t match any actual reported losses.
Even ignoring the flaws in the TERA report, its application to Australia appears to have a number of problems. The Sphere Analysis suggests the TERA report, which concentrates on European countries, is more useful than US reports for Australian comparative purposes, but overlooks that most of the countries in the TERA report are not Anglophone countries. According to the Institute of Criminology, Australia actually has a lower incidence of illegal downloading than other countries except in one area where we outperform: downloading foreign television shows from the US and the UK. This completely skews how the value “lost” to piracy can be calculated given many downloaded TV shows are not able to be watched legally here and have to be priced differently to movies and music.
Nonetheless, the report takes the TERA numbers, calculates a number for retail loss as a proportion of content industries’ GDP share in Europe, then simply applies it to Australia to calculate cost, job losses and tax losses.
Having calculated piracy cost figures with virtually no reference to what is happening in Australia, the report then uses projected growth rates to expand the cost figures massively. It uses assumptions such as that internet penetration will reach 90%, because the NBN is intended to provide high-speed broadband to over 90% of the population, and that piracy will grow exactly proportional to internet traffic (rather than for example, involve downloading of bigger files like Blu-ray versions of movies, rather than more movies) to calculate that the costs of piracy will grow to an extraordinary $5b a year.
In short, it’s consistent with the usual quality of internet piracy reports.
Update: Glyn Moody (@glynmoody) tells me via Twitter that the TERA report methodology has also been debunked at SSRC.