Eric Lai

Fighting software piracy --- with shaky statistics

By Eric Lai
July 14, 2008 3:25 PM EDT
The IDC study released last week by Microsoft claiming partners will gain $5.50 for every $1 Microsoft earns from reducing software piracy isn't the first one that has attracted doubters.

The annual piracy study our sister firm conducts on behalf of the Business Software Alliance (BSA) has been accused by critics of shaky methodology and eye-popping numbers that defy common sense.

According to the IDC/BSA, the software industry lost $48 billion worldwide from piracy, $8 billion more than the prior year.

How did IDC arrive at that figure? First, it estimated through surveys and other research how much and what software should be installed on each PC owned and used in every country in the world.

IDC then calculated the amount of software sold legally in each country. But rather than basing this on actual, raw sales data from resellers, such as research firms like The NPD Group, IDC's analysts arrive at an estimate based on "interviews" with resellers and distributors in 70 countries that is then adjusted after consulting with sales figures provided by software vendors as well as their financial statements, if they are publicly-held.

IDC then subtracts its estimate of how many copies of software were sold in each country from its estimate of how much software should have been sold to arrive at the amount of pirated software.

That is then multiplied by the softwares' list price in each country, adjusted downward for software that comes to customers pre-installed on PCs, to derive a dollar figure for the retail value of pirated software.

Every step in IDC's methodology relies on assumptions -- most software would have been purchased at their list price, vendors' sales figures are not biased upward, etc. -- or estimates.

The Economist magazine, in a 2004 article headlined "Software Piracy: BSA or just BS?" was skeptical.

IDC/BSA's "figures rely on sample data that may not be representative, assumptions about the average amount of software on PCs and, for some countries, guesses rather than hard data," concluded the magazine.

Even shakier, according to the Economist, is that "the figures are presented in an exaggerated way" by the BSA and IDC. "They dubiously presume that each piece of software pirated equals a direct loss of revenue to software firms."

Ivan Png, a business professor at the National University of Singapore, agrees.

"Going from an estimate of piracy to an estimate of loss is a very big stretch," he said.

Anyone who's taken Econ 101 and remembers how supply and demand curves work can understand why.

For necessities such as food and gasoline, consumer demand remains basically the same even as price increases. In other words, demand is 'inelastic.'

But for most goods, demand is 'elastic' -- in other words, demand drops as prices rise. Cracking down on software piracy effectively raises the price of software in piracy-rife countries.

It seems dubious to argue that Mcrosoft Office or Adobe PhotoShop is a necessity like food or energy, especially in a third world country. Png, who published a journal article this spring questioning the BSA/IDC study's results, certainly doesn't buy that.

"Let's say that suddenly there was a death penalty for software piracy, so that the piracy rate dropped from 95% to 0% in some country such as Vietnam," said Png. "It's highly unlikely that the same amount of software that was once pirated will now be purchased at list price."

And the widespread availability of free alternatives such as open-source software such as OpenOffice.org and hosted alternatives such as Google Docs, also seem to make the argument absurd.

In a 2004 interview with the New York Times, IDC analyst John Gantz, co-author of the BSA study, seemed to backpedal from the stance.

Only about one in ten pirated software actually represents a lost sale, he told the Times, adding that he would've "preferred" to use the term "retail value of pirated software" instead of "losses."

But in an e-mail interview last week, Gantz said he continues to "stand by" IDC's claim that "every dollar of pirated software is a dollar of lost revenue to the software industry. The relationship is one to one."

-----------

What do you think of IDC/BSA's methodology and conclusions?