Download data versus piracy claims: the figures don’t add up
Debunking the content industry's scare campaign
Comment First, a declaration of interest. Before I joined El Reg, I was working on an analyst project  (PDF/721 KB) with Sydney company Market Clarity led by long-time friend Shara Evans.
This project yielded a couple of data points that are relevant to claims about internet piracy in this country. The first is that while most broadband plans in Australia offer very high download allowances these days, household users still average only around 6 to 7 GB per month. The second, related to the first, is that we generally hang around the bargain bins when buying broadband: most of us buy on price rather than download allowances.
If you believe the claims made by the music industry – for example, in the now-notorious Sphere Analysis report  – then 4.7 million Australians are pirates and proud of it.
One reason I have trouble with this figure is that it either equals or slightly exceeds (depending on your data source) the number of household fixed broadband connections, which is roughly* 4.5 million (it's almost impossible to accept mobile broadband as a hotbed of piracy. Until recent months, mobile broadband has offered very skinny download allowances, it's relatively expensive, and performance is at best uncertain).
It makes the industry's claims sound like an old-style holy-roller, waving his arms over the crowd and yelling into the microphone "We are all sinners, every one!"
Any 1:1 relationship between broadband subscriptions and piracy suggests either that the estimate of pirates is too high or the definition is too broad.
Ahh, I hear you say, but there's usually more than one user per fixed broadband connection, isn't there?
Yes. Let's rework our figures a little, and allow a median of two individuals per household broadband connection. By that count, there are nine million users, more than half of which are pirates (by the Sphere Analysis definition). That's in line with the general industry claims, but there's still a problem.
As I said, Australian broadband users shop in the bargain bins, and we have frugal download habits. More than half of those 4.5 million subscribers consume less than the 6 GB average (the average is almost certainly dragged upwards by the outliers – the heaviest downloaders).
In other words, our 4.7 million pirates seems to encompass all the users on all the subscriptions that download more than 6 to 7 GB per month.
I'm back to a 1:1 relationship again, albeit of a slightly different kind. The content industry's claims suggest that all "above average" users are pirates. There is no heavy downloader whose usage is professional (running a home-based business, as I do; or regularly working from home and running a VPN to the office). There is no heavy downloader who is patronising legal content like the ABC's iView.
Heavy downloaders aren't using lots of video chat with friends, or playing lots of online games, or anything else that might legitimise their high consumption: all of them are pirates.
NBN scare campaign
I've spent a lot of words on the industry's assumptions about the number of pirates, because it is fundamental to the industry's scare campaign about the NBN.
The argument goes like this: if we wander around Australia connecting people to fibre willy-nilly, we'll turn them all into pirates.
People's behaviour tomorrow is almost certain to resemble their behaviour today: many users will still shop in the bargain bins, and they'll download less than their allowance.
In the case of the NBN, most customers aren't going to immediately leap on the fastest plan available. They'll start small and slow, with 12 Mbps or 25 Mbps download speeds and modest allowances.
If we want to make any valid extrapolation of today's behaviour into the NBN-connected world, we surely have to start with an accurate understanding of today's behaviour. It's easy – too easy – to scream "piracy", and much harder to make the numbers fit.
The mobile user
At the start, I ignored the mobile user. Here's why: according to the Australian Bureau of Statistics data , which covers the same time periods as the industry scare-stories, the average mobile user's download throughput was far less than the fixed user: around 2.5 GB per customer, per month.
Even if a million of those are substitutes – mobile customers with no fixed line – mobile broadband downloads are too small to sustain huge volumes of piracy in the customer base.
Call me a conspiracy theorist, but I don't believe in coincidence. It's no coincidence that the content industry is unveiling both a new lobby and a new set of numbers, at the same time as the American government is pushing for countries to adopt onerous IP protections under a new treaty.
It looks like cause and effect, to me: the Obama Administration's TPP proposals embody what the content industry wants, and here in Australia, the content industry is marshalling arguments that we need stronger laws – like those proposed under the TPP.
The tragedy in all of this is that Canberra has appeared so willing to embrace this kind of data. There are smart people on ministerial staff, believe it or not – I have met some of them – and they are quite capable of conducting the sort of reality-check analysis I've outlined above.
The other tragedy is the Internet industry's long history of being unable to open the right doors in Canberra. If they are to resist the content industry's push, Australia's ISPs need to become successful politicians – fast.
*For this article, my estimate of fixed broadband users was taken from the latest ABS data. I applied the business:household split to the mobile broadband user base, and subtracted this estimate, around 2.2 million non-business mobile users, to the total number of broadband customers (6.7 million).