Tuesday, August 25, 2009

National ID

A big topic in small government circles, of late, is the current efforts in Washington DC to establish a national ID card. Such groups (such as Downsize DC) are generally opposed to this effort, for a variety of reasons. I, too, am opposed, though perhaps not for the same reasons.

Simply put, I fear large databases, particularly when participants have no choice but to make their lives dependent on these databases. I've done a lot of information processing over the years, and I've seen bad databases. I've seen databases filled with bad information. I've seen databases damaged by system transitions. I've seen databases with fields filled with information that had nothing to do with those fields because previous data-entry technicians didn't understand the purpose of those fields... or maybe they didn't care. These databases were so bad they were routinely ignored by the people who had to deal with them in some cases, and made people's work harder in other cases.

But at least these databases could be scrapped if necessary, only had to be dealt with internally, and, if sufficiently bad, could cease to exist when the company unwisely attempting to rely on them went out of business. The pressures of business competition ensured that databases improve or die.

Not so with a government program. Governments are no more infallible than private enterprises, and more susceptible to the long-term accumulation of non mission-critical (and sometimes outright malicious) side interests. (I specify "long term" since there are plenty of bad actors in private enterprise, but that the competitive environment limits the damage they can do and the length of time they can remain in a position of power if that power is abused.) Private actors can be replaced by their competitors. Government programs generally only improve if there is a sufficient amount of the right kind of political interest in them, and even then there are times when an institution becomes impervious to reform (generally the longer it remains in existence). Even the best intentioned, best implemented programs can become institutionalized in the long run, and replacing government programs is generally more problematic than replacing a private supplier.

To put it in more emotional terms, YOU could end up with bad personal information in the national database, resulting in a more difficult time seeking employment, applying for credit, or even simply verifying your own identity if biometrics become part of the process, but your information gets corrupted... with your only recourse being the sending of a form that ensures that in thirty days or more your entry MIGHT be fixed. On the flipside, YOU could end up having to rely on a bad database that is, thanks to taxpayer funding, the only game in town (or at least the only affordable one), for your own identity verification needs.

With the current system, there is, at least, some variety between individual states in how this problem is handled. Though they don't compete with each other directly, they do provide examples of good and bad ways to handle this issue, which can be imitated or avoided, as the case may be. If it gets handled truly horribly, at least the damage is contained.

Now, none of this is to say I am against the idea of a national level identification verification service; governments are hardly only group who have use for such a service. Pretty much any potential employer will ask for an ID card, and some even go so far as to issue their own card to their employees. Merchants properly ask for identification when accepting a check or a credit card. Banks want to make sure that the person attempting to withdraw money or apply for a loan is who he says he is. With all the business that is conducted across state lines, I could easily see a national level ID, valid no matter where you go, being quite useful.

But if there is to be a national identification system, there should be competition. No one institution should be given a state supported monopoly over this. It wouldn't even be necessary to do so; the majority of the value of an ID card would be network value, thus in any given market the tendency would be toward a single, main supplier of this particular service.

I say "any given market" because markets are not discrete units. The entire world is tied together in a "global market", the particular needs of a given nation result in many "national markets". Transportation costs mean that, for some goods, the "regional market" is more important than the national or global one, and that "regional market" probably isn't bounded by national borders. Then there are services, which typically, though not always, cater to a "local market." For minority populations united by particular interest but not necessarily by geography, there are "niche markets." This preponderance of possible markets ensures that no matter how strongly a particular industry tends toward a natural monopoly, that monopoly will never been complete, since nobody can ever please everybody all the time.

In identification, there is, obviously, a general need for an efficient way of establishing, to a reasonable degree of certainty, certainty about an individual's identity, with various information about their qualifications and abilities (driver licensing, security clearances, criminal record, etc.). Precisely what is meant by "reasonable degree of certainty", "efficient", and "necessary information" is up to interpretation, and it is along precisely those lines that different "markets" would be established.

The national market probably isn't huge at this time. Regional markets may be larger. The result of that could be at least two companies operating in a given area: a national company and a regional company. Some regional demand could transcend national boundaries; for example, an effective identification institution serving the American Southwest and northern Mexico could possibly prove useful. Then there's the fact that, for some applications, a greater degree of security or scrutiny (in a word, certainty) is needed than is provided by the larger companies; another company could arise to serve this niche market. A particular segment of the population may be, for reasons of tradition, religion, or whatever reason, leery about the practices of most companies; a special company that can establish identity to a reasonable degree of certainty without offending their sensibilities could serve such a niche. Whatever the need, somebody would be capable of meeting it, once it could be identified.

If any of these companies fail their market, whether through poor database administration, lax identity policies, insecure (easily forged or altered) documentation, or whatever reason, there is always the possibility that a group presently operating outside that market could expand into an ill served market. If a niche market arises that isn't yet accounted for, a multitude of companies could expand their operations, or a new company could arise. All this mutability comes from the optional nature of any form of ID verification: people would be free to refuse to utilize a particular company to verify either their own ID or those of others. Going completely without probably isn't an option for most people in this age, but even that option could be permissible, if not terribly viable for the one that chooses it.

Contrast that to a single, overarching monopoly institution, a central standard and a central implementation. Politicians establish it in exchange for votes and the money necessary to get votes. If the initial implementation is bad, participants may be able to amass sufficient capital to get it revised or revoked, but once it is running smoothly, it will continue running. It is not performance dependent, except to that degree necessary to remain politically viable (which is much less difficult than retaining customers). It is tax funded, so that even if alternatives are permitted, they are financially inviable. It can afford to ignore any niche that lacks a sufficiently powerful political voice, and it cannot be displaced if it fails its mission unless the failure is so spectacular as to attract political attention. By the time that happens, the damage has been done, and in the meantime there are no alternative institutions that can cover for them (by successfully taking their business) either until they get their act back together, or permanently.

Here, I find it necessary to address the reason the political establishment at large is seriously attempting to implement a change: Terrorism. While the proposed (and passed but not yet adopted) changes have been being proposed for many years, now, it is the new public awareness of the terrorist problem that has made these changes politically convenient. The issue, so far as I understand it, is the fear that the current ID regime is insufficient to keep potential terrorists out of sensitive areas, like airplanes, manufacturing facilities, nuclear plants, and so on.

So, okay, maybe airports and airlines need to reconsider their policies regarding who they allow onto their facilities. Perhaps an ID verification system, national in scope, is needed to verify the background of potential passengers. Such a system need only cover those who choose to fly, which is definitely not 100% of the population, and possibly not even 50%. This does not justify a single, central program that aims to cover 100% of the population, and a program designed to cover this specific need likely does not even require taxpayer funds to operate, and probably not even to establish.

No comments: