NEGATIVE — CONSUMER/INTERNET — SIGNIFICANCE� 395

NO SIGNIFICANCE TO ERRORS IN DATABASES

THERE ARE VERY, VERY FEW ERRORS IN PRIVATE DATABASES ABOUT CITIZENS

Solveig Singleton, director of information studies at the Cato Institute, January 22, 1998 Cato Policy Analysis No. 295 PRIVACY AS CENSORSHIP: A Skeptical View of Proposals to Regulate Privacy in the Private Sector http://www.cato.org/pubs/pas/pa-295.html // acs-EE2001

Many complaints about private databases surface when people find errors in their credit reports. But the evidence suggests that, on the whole, rates of error in credit reports are low. Two highly publicized but biased studies misleadingly report high rates of error in credit reporting (from 30 to 50 percent).(47) A 1991 study by Consumers Union relied on its own employees and their acquaintances to review their own credit reports and report "inaccuracies." Consumers Union did not check whether those claims of inaccuracy were true or false, however, or try to identify the source of the errors.(48) Ralph Nader's Public Interest Research Group also failed to select a random sample, instead estimating an error rate from a sample of consumers who had paid to review their credit reports--people who probably had reason to suspect they would find errors.(49) A more rigorous study of 15,703 consumers, conducted by Arthur Anderson & Co., showed that the true error rate is probably as low as 1 percent.(50)

ERRONEOUS DATABASES ARE NOT A VIOLATION OF INDIVIDUAL RIGHTS

Solveig Singleton, director of information studies at the Cato Institute, October 18, 1999 http://www.cdt.org/privacy/FTC/profiling/singleton.htm// acs-EE2001

Outside credit reporting, however, erroneous data in databases rarely will amount to a violation of someone's rights. I don't care if I'm accidentally listed somewhere as a hockey fan. The vast majority of transactional data that is collected is trivial and harmless; access mechanisms would simply be an absurdity that would prevent new information services from coming into existence.

COMPANIES HAVE A STRONG INCENTIVE TO MAKE THERE DATABASES ACCURATE, AND THE ERROR RATE IS VERY LOW

Solveig Singleton, director of information studies at the Cato Institute, October 18, 1999 http://www.cdt.org/privacy/FTC/profiling/singleton.htm// acs-EE2001

And, finally, no one wants their database to be inaccurate. So companies in the database business have a strong incentive to get it right. (the two methodologically flawed studies showed rates of error in credit reports as high as 30 percent notwithstanding--Arthur Andersen's properly conducted survey of a truly random sample of consumers showed rates of serious error are probably around 3 percent).