Not all domestic surveillance and biometric technologies are created evil, says the Cato Institute's Wayne Crews, in an e-mail to privacy advocates.

Here's his framework for distinguishing between the mildly creepy surveillance efforts and the truly invasive:

1) BAD: Mandatory National ID cards encoded with biometric identifiers, or compulsory databases for data mining purposes.

2) NOT (NECESSARILY) BAD, but can be wholly abused and require extensive 4th amendment safeguards that do not yet exist: Gov't run face cameras (and related technologies like iris scanners) that ride on top of a database of criminals or wanted individuals. These should **not** collect data on individuals other than those already in the database (presumably there thru appropriate 4th amendment procedures). Incidental data collected on random individuals cannot be retained. Problem is the guarantee. This is where I think the real future fight lies, and the most risk for sensible evolution of these technologies.

3) GOOD: Countless private uses of biometrics that offer the opportunity for extraordinary security by preventing others from posing as us. This is where the market can shine. However, these must not be allowed access to data gleaned by gov't coercion, or they move into category 1 or 2 and give the entire industry (biometric or data mining) a black eye, and make it impossible to defend the industry from regulation. Let's keep it self-regulated.

Nutshell: (1) avoid mandatory databases (2) ensure 4th amendment protections even for public surveillance, and (3) avoid mixing public and private databases.

Agree? Disagree? Let's hear it.