The major problem of any large scale ID system is that it is inherently corruptible. This is not an issue of counterfeit cards but instead corruption of the sytem to produce false cards, that appear to be fully valid, by people on the inside.
As Canada and the US are finding out, passport levels of security work well only when few people have them. Universal IDs end up with DMV quality people and systems – that is exceptionally low. The higher the stakes behind an ID, the greater the incentive to corrupt the system.
Another system problem is that all current IDs are incredibly insecure. You can’t build a secure system off of insecure foundations, but you can’t fix our current problems, as any revalidation of the entire population is just as vulnerable to fraud. Truly secure IDs would require a 90 year program to create a solid foundation for all documents.
Another problem of “secure” IDs comes from the massive scale of the system. 300M people gives 3,000 errors for a 99.999% accurate system (5 nines, the goal of telecom). How many annual transactions would an individual have with this secure and trusted system? How high would the stakes be? 10 – 100 transactions a year seems pretty likely, if not a lowball estimate. So 300K people a year would have a high stakes, high value system fail if it was as reliable as the wireline telephone system is, which it won’t be.
With biometrics you run into problems that a substantial proportion (whole percents) of the population won’t have that body part or will have some problem with fitting into the parameters. What does glaucoma do to an iris scan? Lasik? Contacts? Fungal infection caused by contact lens solution? Biometric scans also fail at a fairly rate (for a large scale, high stakes system). 3-4% error per transaction wouldn’t be out of line, times the expected number of transactions…
Can you now see why people have reservations about these systems? Large scale, high stakes, high trust systems have inherent problems that simply can’t be resolved in any reasonable time frame for any reasonable budget. Make an ID cost $3k per person, and it’s achievable, but you’ll probably always have an annual maintenance cost of a few hundred dollars per person (identity systems have inverse economies of scale). Do you want to spend $1 Trillion to setup the system and $60 Billion annually to run it?
ID cards are just another static defense, easily surmountable by cheap, mobile forces. The money invested could likely be much better spent on offensive measures or high value security personnel. TSA screening costs similarly would create much more value by spending the money on air marshalls and special forces officers with global hunting licenses. Entrust not in walls but in legions!
