Why AFACT’s piracy statistics are junk

Yesterday, the Australian Federation Against Copyright Theft (let’s call them AFACT or perhaps ‘Big Content’ for short) lost their appeal in the long-running and important copyright infringement suit against Australian ISP iiNet. As usual, some of the best commentary can be found by Stilgherrian (who really does need a second name, don’t you think?):

If you came in after intermission, you’ll pick up the plot quick enough. AFACT said iiNet’s customers were illegally copying movies, which they were, but iiNet hadn’t acted on AFACT’s infringement notices to stop them. AFACT reckoned that made iiNet guilty of “authorising” the copyright infringement, as the legal jargon goes. iiNet disagreed, refusing to act on what they saw as mere allegations. AFACT sued.

In the Federal Court a year ago, Justice Dennis Cowdroy found comprehensively in favour of iiNet. It was a slapdown for AFACT. AFACT appealed, and yesterday lost. Headlines with inevitable sporting metaphors described it as  two-nil win for iiNet.

But read the full decision and things aren’t so clear-cut.

One of the three appeals judges was in favour of AFACT’s appeal being dismissed. Another was also in favour of dismissal, but reasoned things differently from Justice Cowdroy’s original ruling. But the third judge, Justice Jayne Jagot, supported the appeal, disagreeing with Justice Cowdroy’s reasoning on the two core elements — whether iiNet authorised the infringements and whether, even if they had so authorised them, they were then protected by the safe harbour provisions of the Copyright Act.

There’s plenty of meat for an appeal to the High Court, and that’s exactly where this will end up going. Wake me when we get there.

As I argued today, also in Crikey, it’s ironic that Big Content seems to be about the only business lobby group in the country arguing for more regulation and red tape.

But the copyright case also comes in the wake of an interesting little micro-controversy about piracy statistics, released by AFACT late last week. Aided by an economics consultancy and a market research firm, AFACT released an impressive-seeming report that claimed that movie piracy was costing Australia $1.4 billion and 6,100 jobs a year.

Electronic Frontiers Australia made some pretty valid criticisms of the research, including the following:

1. The assumption that 45% of downloads equal lost sales is unproven and insufficient evidence is provided to support it. The survey method cited is better than assuming 100% of downloads are lost sales, but there is better analysis in other studies – for example this piece by Lawrence Lessig. If the study was correct, sales of DVDs and attendance at cinemas would be much more reduced than the reported industry figures. In fact, the movie industry is making record profits.

2. It can’t be ignored that downloads have an advertising effect both on the product downloaded and future releases. To the extent sales may be lost, these must be offset against other gains from advertising.

3. Gross revenue is not the relevant metric, due to variables such as investment in capital, distribution and costs of sales. Many of the movies downloaded may not have been available to view or buy in Australia. Profit is the metric of importance, but this is never studied.

4. Flow-on effects to other industries are wholly speculative, and lost tax on profits assumes the entities pay Australian company tax on sales pro-rata to revenue, which is not intuitive or evidenced. It also assumes that money not spent on movies is lost to the economy, instead of helping to create jobs in other sectors.

5. Peer to peer file sharing is merely the latest in a sequence of technologies since the 19th century which have been claimed to be the ruin of the creative arts. See chapter 15 “Piracy” by Adrian Johns (University of Chicago Press 2009) – the copyright owners said the same thing about copies of sheet music, tape recorders, every iteration of personal recording system and indeed public radio. However, “home piracy” acts not only as a loss to industry but also as a boon to distribution, bypassing censorship and limitations on sales by official outlets.

6. The report suffers, as have other industry-funded studies, from “GIGO”. With an assumption that “downloads = losses” unproven, all conclusions estimating the size of the loss are equally unproven. What if a vibrant sharing culture increases total sales for media respected as quality by consumers, but reduces sales of hyped media? (Research has shown that the biggest downloaders in fact spend more on entertainment than non-downloaders.)

7. The call-to-action of this report is obviously to “crack down on piracy”, shifting the cost of file-sharing from the industry to the taxpayer via increased law-enforcement. No industry, let alone the foreign-dominated entertainment industry, deserves a free ride for its business model. If instead, the industry noted that the report says 55% of downloads created a market for sales, much of which is unsatisfied due to current restrictive trade practices, then its future profitability would be in its own hands.

8. Repeated studies have demonstrated that the entertainment industry vies for money and commitment of time with all other forms of entertainment. The Internet, computer games and mobile telecommunication applications take “eyeballs and dollars” away from DVD and CD sales, but also sports arenas, sales of board games and printed works. Magazines are also suffering from a reduced value proposition with the Internet, and some forms of entertainment and some businesses in the industry will no doubt find it difficult to remain vibrant. Change is consumer-driven, and it’s futile for the industry to try to hold fast to a business model and methods of content distribution which are dying with or without fierce law enforcement of copyrights.

Unsurprisingly, AFACT  have responded, attacking EFA’s arguments.

Notably, AFACT replies that:

“The study does not assume that ‘downloads = losses’. As stated above, some 32 per cent of respondents said that they viewed an authorised version of a movie after watching the pirated version. As a result, 32 per cent of ‘all pirate views’ were removed from the ‘lost revenue’ calculations and were treated as ‘sampling’.”

This is a valid argument. AFACT has indeed removed these later viewings from their lost revenue calculations. But, as I’ll explore below, this doesn’t mean that AFACT’s methodology is sound.

AFACT’s other replies are far less persuasive. Take this line:

“It should be clearly noted that in almost all of these cases government or technology provided a barrier to prevent continued rampant infringement. In the case of public radio, legislation provided statutory copyright royalties. VHS and cassette tape may have been efficient technologies for recording, but in terms of cost and quality (analog degrades with time) they proved not to be efficient for distribution at that time. Laws were also designed to prevent mass distribution of pirated VHS tapes. Solutions, whether legislative, technological or otherwise are currently required to prevent or deter the unfettered digital distribution of pirated versions of copyrighted content.”

Not to put too fine a point on it, this is a rubbish argument. Statutory copyright royalties for broadcasters were not barriers to listeners – they were income streams to publishers. And, in fact, as EFA point out, radio proved to be such a powerful marketing tool for music labels that record companies regularly resorted to payola and other measures to get their songs on high-rating radio stations. This argument is a classic tautology: because AFACT believe that regulatory barriers are necessary to prevent infringement, they argue that the reason previous technologies didn’t lead to “rampant infrignement” was because they were strictly regulated. You don’t need a degree in logic to spot the flaw in this argument.

So who’s right?

On the whole, EFA has the better of the exchange. Indeed, there are plenty more holes you can pick in AFACT’s methodology if you wish. To start with, let’s examine their laughable “Annex 1” in the full report. This purports to explain how ABS input-output tables are used to generate a final figure for total piracy impact in terms of lost sales and job losses.

I’d like to say I carefully checked their methodology for its econometric accuracy. Unfortunately, I can’t – because the authors at Oxford Economics and Ipsos don’t publish their equations; nor do they publish their raw data.

Just as an exercise, I downloaded the ABS input-output tables and attempted to match the ABS data to the AFACT report. It’s impossible. The data tables in the AFACT report which might allow that kind of scrutiny are missing.

What Annex 1 does tell us is that Oxford Economics and Ipsos have made all sorts of behind-the-scenes calculations to do with the exact value of the multipliers they use and the precise allocation of various ABS industry data to various categories of their assumptions. But they don’t tell us how these figures were arrived at. To get a flavour of the opacity of the modelling, here’s their full explanation of two of the the multipliers they use:

Type II multipliers of 2.5 (Gross Output) and 1.1 (GDP) were estimated. This covers activity in the Australian motion picture exhibition, production and distribution industries as well as TV VOD, internet VOD, downloads of motion pictures and the retailing of these motion pictures

There is no further explanation of how the numbers of 2.5 and 1.1 were “estimated” and no equation which shows us what they multiply. Hence, it is literally impossible to verify, cross-check or otherwise scrutinise these figures. Indeed, the full report contains no true methods section. In other words, the academic credibility of these figures should be zero.

This rubbish is just another example of how lobby groups use consultants-for-hire to create vocal scare campaigns based on fictitious figures. It’s junk modelling, ordered up for the express purpose of industry rent-seeking.

Crikey’s Bernard Keane explained it helpfully for us in relation to climate lobbying in 2010:

This what you do:

  1. Commission a report from one of the many of economics consultancies that have broken out like a plague of boils in the past decade.  This should feature modelling demonstrating the near-apocalyptic consequences of even minor reform.  Even if your industry is growing strongly, you should refer to any lower rates of future growth as costing X thousands of jobs, without letting on that those jobs don’t actually exist yet, and might never exist due to a variety of other factors.
  2. Dress up the report as “independent”, slap a media-friendly press release on the top and circulate it to journalists before release, with the offer of an interview of the relevant industry or company head.
  3. Hire a well-connected lobbyist to press your case in Canberra.  When the stakes are high, commission some polling to demonstrate that a crucial number of voters in crucial marginal seats are ready to change their vote on this very issue.