The Business Software Alliance is out with its latest guestimate of the loss that software vendors suffered in the past year because of piracy. The study methodology looks reasonable, but the message that the group read into the results and that most of the press accepts is basically flawed.The BSA is best known for its vigorous enforcement of software licenses by conducting audits of businesses. Based on their press releases, BSA seems to be quite successful in the audit business.In early July, the BSA released a "Global Software Piracy Study". The summary says "36% of the software installed on computers worldwide was pirated in 2003, representing a loss of nearly $29 billion." The first part of this claim I can believe, but I rather doubt the second part.The study was conducted by IDC (a sister organization to Network World). The report on the BSA Web page contains a description of the methodology IDC used to conduct the study. It is not as detailed a methodology description as it could have been but, based on what it does say, IDC approached the problem about as well as one can approach a problem where most of the actual data only can be estimated.But there is no way the data in the report can support some of the things the report says. In particular, there is no data that supports the assertion that the estimated amount of piracy cost software vendors "nearly $29 billion."There is no real problem with the BSA trying to figure out how much software is installed on computers around the world nor is there a problem with the BSA trying to figure out how much of that software was legitimately purchased. From these two facts the BSA can figure out what percent of the software has been pirated, and it does so. But the alliance cannot just multiply the estimate of the pirated software by the sales price of that software to come up with how much software vendors would have made if there had been no piracy. The BSA also needed to figure out how much of the software actually would have been purchased if the user had to pay the sales price. The BSA doesn't do this, or if it does, the group fails to mention doing so in its report. The $29 billion figure looks impressive, but I expect the actual number is a very small fraction of that.I can understand the BSA finessing the difference between actual losses and the value of pirated software but I don't understand the lethargy of the media that blindly repeats the bogus claims. The\u00a0initial report from Reuters\u00a0described the problem as "the global trade in pirated software" as if someone was making $29 billion from selling stolen software. I hope someday the press will stop taking industry pronouncements at face value and think before they print.Disclaimer: Even for Harvard, $29 billion would be real money, but I didn't ask its opinion on the reality of the money.