An anti-piracy lobby group has bowed to pressure and released a controversial report on the impact of internet piracy on the Australian economy, following howls of protest over its existence and veracity.
Earlier this year, the Australian Content Industry Group, an umbrella body for local entertainment companies, commissioned a report, The Impact of Internet Piracy on the Australian Economy, penned by a mysterious outfit called Sphere Analysis. It claimed the cost of illegal copying was $900 million last year, with a loss of 8000 jobs.
But until now, despite the best efforts of industry sites like TorrentFreak, the Sphere report and its methodology was nowhere to be found. Neither Sphere nor ACIG maintain a public website and their principals have been uncontactable.
This morning, an employee of Music Industry Piracy Investigations, a self-styled member of ACIG, confirmed to Crikey the report would be released on “either Friday or Saturday”, adding it would “send it through when it becomes publicly available”. (UPDATE: 3:15pm The study has been uploaded to the Business Software Analysis website here).
Ten days ago, Fairfax papers splashed hard with the Sphere report, reporting uncritically the $900 million “cost” of piracy in the music, film, television, software and video games sectors, which would apparently rise to $5.2 billion by 2016 and $18 billion between 2010 and 2016. The job loss figure of 8000 was also reported. But the freelance journalist who wrote the piece — and an accompanying feature — is understood to have been “told” of the raw numbers but was not furnished with the actual report or the associated methodology.
The stories followed a speech given last month by Attorney General Robert McClelland to the Blue Sky Conference on copyright law, in which he cited the ACIG study alongside a similar report by the Australian Federation Against Copyright Theft, which appears to contain wildly different assessments of the economic impact.
A domain name search under the inactive sphere.com.au domain name reveals the site is registered to a “Sphere Property Corporation” with a contact of John Harston, listed here as Sphere’s managing director. Another employee, Phil Nott, serves on the board of the not-for-profit Australian Communication Exchange and previously worked for the Queensland Industry Development Corporation (now Suncorp-Metway).
However, far from the world of software theft, Sphere Property’s main function appears to be conducting economic analyses of property developments. A NSW business names extract, published here on the TorrentFreak website, shows Sphere Analysis was only registered in late November and maintains offices in Bondi Junction. Crikey left a message on John Harston’s mobile phone but did not hear back before deadline.
Sphere’s numbers seem to be the subject of furious debate, even among the anti-piracy lobbyist raternity. As Bernard Keane wrote a week ago, in January AFACT, which represents the interests of Hollywood studios, released the Ipsos MediaCT report. That effort, which contains a helpful section called “how we arrived at these figures”, said movie piracy alone cost $190 million in forgone tax and 6000 jobs. Either AFACT or Sphere’s report would appear to be out by a large margin.
Crikey understands there is widespread discomfort within the industry, with one senior figure describing the release of the two contrasting reports so close to each other as “unfortunate”.
A comment on the TorrentFreak website from sometime Fairfax contributor Guy Cranswick said he had also failed to obtain the report and slammed The Age for publishing sight unseen:
“In my capacity as a researcher I requested the report from one of the organisations in the so-called copyright alliance but no reply has come which is very suspicious because normally these organisations are very happy to send their expensive studies to known researchers and other media contacts.
“As no one has see this report it goes to the core of journalistic credibility that a paper such as The Age could have published news about this uncorroborated survey.”
A spokesperson for the Attorney-General’s office told Crikey it had received the Sphere report, but that “we would have to speak to the authors” if we wanted to obtain it in full.
Sphere Analysis partner Emilio Ferrer contacted Crikey after publication to say that he completed the report a month ago but referred all queries about its initial non-release to his client, the Australian Content Industry Group.
It’s not “theft” (unless someone ran out of a store with a box under their arm), it’s copyright infringement.
Accurate terminology is important in any debate, especially one as important as this. Copyright infringement bears no resemblance to theft, in statute, action, or outcome.
In summary: they took the estimates of a similar consultancy which reported on the “retail loss to creative industries” in five EU countries and scaled the figures to match the Australian economy. Page 6 is where the rubber hits the road, and as usual with economics the most interesting action is in the footnotes.
Footnote 22 tries to carry forward the essence of the TERA report without carrying forward exchange rate fluctuations. Unfortunately, the methodology ends up relying upon the ratio of “creative” to “retail” jobs being constant. Something for which no evidence is offered, which would seem unlikely, and which is counterfactual in the one reported-upon country I checked (Spain) for a change in ratio between 2008 and 2010.
A classic error is made when the report averages the results of the five European countries. That is, the average implies that Australia is 20% like Germany, 20% like France, 20% like the UK, 20% like Italy and 20% like Spain. There is no justification at all for assuming that the weights of each country in the average are the same. Think about that — the report’s ratio requires that English speaking Australia is as much non-English-speaking Italy as it is like English-speaking UK. Despite the greater competition in English-speaking countries from English-speaking North American creative product.
“We have assumed that volume of internet piracy grows at the same rate as IP traffic” is argued in footnote 34. This footnote is key to the estimates in “Projections 2010-2016”. Firstly, that assumption ignores new non-infringing applications which the greater bandwidth of the NBN will enable. Videoconferencing is an obvious example where ADSL data rates currently constrain traffic. Secondly, the assumption requires consumers’ demand for unpaid cultural product to be unlimited, yet there is a strict upper bound to the number of hours in a day.
Obviously, any shortcoming in the TERA report examining the five EU countries also applies to the application of that report to Australia.
It worries me that a report where I can see major errors in methodology within minutes is being used to “inform” Australian parliamentarians.