Global ripoff?

How bad is the international piracy problem? The software industry's two leading anti-piracy organizations, the Business Software Alliance and the Software Publishers Association, claim to know the answer almost to the last penny. In 1995, U.S. developers lost $13,191,630,000 to worldwide piracy, the two associations recently announced. And the two groups can even name the worst offenders: Vietnam, with a piracy rate of 99%; Indonesia, 98%; and Oman, China, and Slovenia, tied at 96%.

That's a loss equal to almost twice Microsoft's worldwide sales during 1995, so the global ripoff is clearly more than pocket change. Still, we couldn't help wondering how the BSA and SPA managed to uncover the exact amount of illicit copying that occurs in odd little countries like Paraguay (95%) and Mauritas (90%) that most Americans probably couldn't find on a map.

Well, it turns out the numbers come from a complicated, Rube Goldberg-like statistical model that strings together data about international PC shipments, U.S. software exports, inter-country transfer sales, penetration rates of PCs "compared to the number of white collar workers" in each country, home vs. business usage, and various "uplift factors" and adjustments. (One of the more curious of these adjustments was to ignore all consumer and educational software categories, presumably because the business-oriented BSA didn't want to waste money researching markets its members mostly ignore.)

Most of this underlying data is notoriously unreliable, so we asked the researcher who built the BSA/SPA model, David Fay of International Planning others give away software to generate demand for high-margin services and data. Overseas, there's even less of a bright line between legal and pirated software, especially in non-Western countries that often treat all intellectual property as a community asset. (Of course, deliberate counterfeiting is a separate problem, but that's not what the BSA and SPA are trying to measure.) Piracy simply isn't as easy to define as the lawyers and association executives would like it to be.

But the real problem to us isn't "ballpark" numbers that pretend to be impossibly accurate--it's the underlying notion that piracy itself is a quantifiable phenomenon. Certainly in the U.S., software marketers have been moving steadily away from the traditional "one license, one machine" standard. Major publishers now routinely saturate the market with trialware versions; others give away software to generate demand for high-margin services and data. Overseas, there's even less of a bright line between legal and pirated software, especially in non-Western countries that often treat all intellectual property as a community asset. (Of course, deliberate counterfeiting is a separate problem, but that's not what the BSA and SPA are trying to measure.) Piracy simply isn't as easy to define as the lawyers and association executives would like it to be.

In the end, the widespread disregard for shrinkwrap licenses in developing countries may be an important message by itself: When more than 90% of users choose to ignore U.S. licensing rules, it just might be time to invent new rules. That's a far harder job than cranking out semi-fictional statistics, but we suspect the real long-term solution to global piracy is to develop licenses that respect the way honest international customers want to buy and use software.

Business Software Alliance, 1150 18 St. N.W., Washington, D.C. 20036; 202/872-5500. David Fay, consultant, International Planning 206/868-8850.