So, once again, the Business Software Alliance (BSA) has come out with its annual report on software piracy around the world, entitled "Shadow Market" [.pdf]. And, once again, it makes all the same methodological mistakes – it's almost as if the BSA hasn't been reading my critiques of last year and the year before....
The fundamental error made in this latest survey is the usual one:
The commercial value of this shadow market of pirated software climbed from $58.8 billion in 2010 to $63.4 billion in 2011, a new record, propelled by PC shipments to emerging economies where piracy rates are highest.
Notice the tell-tale words right at the start of that: "commercial value". A later section explains this choice as follows:
The commercial value of pirated software is the value of unlicensed software installed in a given year, as if it had been sold in the market. It provides another measure of the scale of software piracy and allows for important year-over-year comparisons of changes in the software piracy landscape.
Well, that's true – it is one measure of how the claimed level of software piracy is changing. But as the people who put this report together surely know, it is not a true measure of the losses caused to software companies by piracy, for the simple reason that not every pirated copy corresponds to a lost sale. Often, people use pirated software because they cannot afford the retail price of authorised copies – there is no way that they could be converted to paying customers. So a proportion of pirated copies are not lost sales at all, simply a reflection of pricing problems.
As it happens, that issue is even more central for this year's report, for the following reason:
Emerging economies, which in recent years have been the driving force behind PC software piracy, are now decisively outpacing mature markets in their rate of growth. They took in 56 percent of the world's new PC shipments in 2011, and they now account for more than half of all PCs in use.
Frequent pirates in emerging economies install nearly four times as many programs of all sorts per new PC as do frequent pirates in mature markets. Among infrequent pirates – those who say they rarely acquire unlicensed software – there is a greater than two-to-one gap in the total number of programs they install.
What that means is that the piracy in emerging countries is becoming an even greater proportion of the overall software piracy:
Software piracy rates in emerging markets meanwhile towered over those in mature markets: 68 percent, on average, compared to 24 percent. Emerging economies thus continue to account for an overwhelming majority of the global increase in the commercial value of pirated software.
But it is precisely in those countries that piracy is driven by pricing issues, as was revealed in the seminal report "Media Piracy in Emerging Economies":
Based on three years of work by some thirty-five researchers, Media Piracy in Emerging Economies tells two overarching stories: one tracing the explosive growth of piracy as digital technologies became cheap and ubiquitous around the world, and another following the growth of industry lobbies that have reshaped laws and law enforcement around copyright protection. The report argues that these efforts have largely failed, and that the problem of piracy is better conceived as a failure of affordable access to media in legal markets.
This means that the BSA's conclusions about the real impact of software piracy are even less valid than usual: even fewer of those unauthorised copies represent lost sales, at least at the market prices currently being charged. In fact, the latest BSA figures simply confirm the continuing failure of software companies – mostly Western software companies – to price their goods appropriately for emerging markets.
BSA's own figures suggest that software piracy is already relatively unimportant in "mature" economies – undermining calls for harsher copyright enforcement measures in things like ACTA or IPRED. As for emerging economies, it wouldn't be such an issue there if more realistic pricing were adopted. The real problem is Western software houses' attempt to charge Western prices that in real terms represent vast swathes of individuals' earnings, or of companies' IT budgets.
It's particularly unreasonable since the actual marginal cost of software is close to zero, unlike hardware, say – so there are no good economic reasons why differential pricing couldn't be adopted. Apparently, Western software houses would rather sell a few copies at high prices than many copies at much lower prices more appropriate for emerging economies.
I said that the report was the same old nonsense as previous years, but in fact there is one novelty: we get to find out the actual question posed to those taking part in the survey:
"How often do you acquire pirated software or software that is not fully licensed?"
Now, I fully accept that people are likely to under-report how often they acquire pirated software, which will tend to produce an under-estimate of the true level. But it's that second phrase that intrigues me: what exactly does "not fully licensed" convey to you?
For example, is free software "fully licensed"? Although I know that it is licensed to you provided you comply with its terms, I wonder how many people using Firefox or LibreOffice really get that. Might it not be the case that they think they are allowed to use free software without a licence? It would be a natural interpretation of what open source and free software meant – the idea that you can just use it, and don't have to worry about licences.
So that leads me to wonder how many survey respondents in both mature and emerging economies "admitted" that they acquired "not fully licensed" software when what they meant was non-proprietary stuff like free software? That's important, because it would mean that the figures claimed by the BSA study would be over-estimates of piracy. Moreover, as open source becomes more widely used, it might be that this will become a more significant distorting factor.
Finally, it's worth noting that the BSA is up to its old tricks when it comes to trying to "prove" that people just love the chains of intellectual monopolies. This year it has some new phrasing, but the same misleading dichotomy:
By a wide 71-percent to 29-percent margin, respondents aligned themselves with the
idea that "it is important for people who create new products or technologies to be paid for them, because it provides an incentive to produce more innovations. That is good for society because it drives technological progress and economic growth."
Computer users around the world rejected the alternative proposition: "No company or individual should be allowed to control a product or technology that could benefit the rest of society. Laws like that limit the free flow of ideas, stifle innovation, and give too much power to too few people."
These are completely separate questions: one is whether people should be rewarded for creating new software, and the other is about control. I doubt whether anyone would be against rewarding coders for the work they do, although that's not necessarily a question of money. Some programmers want to get paid for their programming, and they often join companies to do that. Others might be horrified at the idea that people would try to pay for what is gift - but wouldn't mind having some recognition instead .
Similarly, I doubt whether anyone in the free software world thinks that nobody should be allowed to control a technology – that's certainly not true for projects like Linux. There are all kinds of benefits in having people who lead projects and control the direction of development. But the key question isn't about control, it's about ownership: the problem is not when people direct software projects, it's when they claim ownership over ideas – software patents, for example.
So, once again, the BSA survey brings very little real value to the debate about unauthorised copies of software. Instead, it's clearly designed to produce headlines about software piracy "costing" $63.4 billion in 2011 – even though the report is careful never to say that. However, the 2011 survey's main Web page does claim "Shadow Market of Pirated Software Grows to $63 Billion", which is highly misleading, since by the report's own admission it is simply one measure that has grown to $63 billion, not any meaningful figure.
As I've indicated, the reality is that unauthorised copies are not only "costing" Western software companies far less than that, but that this loss is almost entirely self-inflicted because of inappropriate pricing strategies in the emerging economies. BSA members could turn a problem into an opportunity overnight if they so wished, but they choose not to – that's the real story, not the big, scary numbers that are churned out each year but that actually mean nothing.