My voice is my passport. Or is it my eyes?

There are several things I considered for today’s missive. For one thing, I thought about the fact that the headline “Housing Construction Hits 17 Year High” hides the fact that although housing starts were up 17.7% in the West and up 4.9% in the South, they dived 18% in the Northeast and 8% in the Midwest. Clearly, whether or not there is a lot of new housing going up near you depends on where you live. By extension, whether or not the economy is good enough that people are buying new houses depends on where you live.

Another thing I considered is that both Reuters and the Associated Press felt it important to note that unemployment among high tech types such as engineers and computer programmers is up. Over half a million such jobs vanished last year alone, with another quarter of a million expected this year. These were good paying, skilled, college-degree-required jobs.

Instead, I would like to alert you to a potential boondoggle. Biometrics, at its most simple, is identifying people based on their body: fingerprints; retina scans; voice matching. It has also been hyped as the up and coming thing in security, and is expected to have a $7 Billion market by 2007. US Homeland Security Secretary Tom Ridge is behind it, and is trying to get Europe to join him on the bandwagon. The Chinese think it’s a good idea too. Biometric data will be required in many passports by 2004.

Proponents say it is the best security device ever: no more forgotten passwords; no more forged identification; no more identity theft. But best of all, they say, biometric identification will allow frequent fliers to speed by that long security line at the airport.

That is just not going to happen.

First, experts admit that biometrics have “limited accuracy,” “would not stop people obtaining multiple personas,” and is “only part of the solution.” Indeed, a recent press release admits “Face recognition biometrics is striving to overcome performance inefficiencies and dispel skepticism regarding its ability to meet the demands posed by emerging applications.” In otherwords, they would like to make a product that works. USA Today quotes an expert as saying that “At best, fingerprint scanners are about 98% accurate.” That sounds pretty good until you realize that means it will fail 2% of the time. A minor annoyance if you have a hundred scans a day, a big mess if you have several thousand scans a day. For a tech-centric commentary on this article, click here.

Accuracy is further impaired by many things. Will a voice recognition system recongnize you if you have a cold? Can you use a thumb scanner if you have had an accident and injured your thumb? Will eyedrops effect your retinal scan? Do you dare risk a change in hair color or facial hair if your office uses face recognition? Will you still be recognised in a year, or if something about you changes, or if something around you is different?

As if the accuracy concerns were not enough to declare biometrics a not ready for prime time technology, there are privacy concerns. That same USA Today article says “Biometric systems store huge amounts of personal data. Some systems record the data on a computer inside the store. Others record it on computers at the biometric company’s headquarters. Critics say data will always be at risk.” What can you do if your fingerprint or retina scan is hacked? You can always get a new credit card number or a reissued drivers license number; but you can’t just replace your fingers and eyes.

Linked to but separate from privacy concerns are civil liberties concerns. That stored data in the previous paragraph may be shared with just about anybody, and at this moment there is no law to stop it. Maybe it doesn’t matter that you bought a bag of cat food, but is it really necessary to share that information? And share it with whom? Why? And what about that CIA project to identify people with biometric data at a distance? That isn’t from some bunch of tin-foil hat crackpots, that’s from the Australian Broadcasting Company. How much data about your travels and purchasing habits will end up stored with your biometric-enhanced passport?

And this last question brings us back to the airport security line. Americans want this to work, really. Americans are willing to pay a modest sum and have their privacy invaded for a bit of laminated plastic saying “NOT A TERRORIST” that will speed them past all the security lines. It will not happen. As long as people have luggage, it will need to be inspected. As long as people get on airplanes, they will need to go through metal detectors. As long as people have carry on bags, they will need to be x-rayed. Yesterday’s USA Today cites an expert who “predicts that within a decade or two, the technology will be sophisticated enough to monitor passengers at each step of the airport process — when they arrive, when they check their bags, when they pass through security and when they enter the plane. But he is less optimistic about the technology’s potential for slashing airport waiting times, saying the challenge for airports is to deploy the systems without increasing the wait for passengers.” That’s right. Experts think biometrics may well slow down the airport security lines.

Biometrics is not the end-all answer to identification and security. It is a whole new set of questions. Expensive questions.