A Stitch in Haste

Wired has a a good op-ed piece by a private data-miner on how massive government data-mining of the sort advocated by Richard Posner simply cannot work as a matter of technological and statistical fact.

Much of the number-crunching goes over my head (if I want to hear about "trillions," then I'll stick to the federal budget or the national debt). But this part deserves emphasis:

Data mining works best when you're searching for a well-defined profile, a reasonable number of attacks per year and a low cost of false alarms. Credit-card fraud is one of data mining's success stories: all credit-card companies mine their transaction databases for data for spending patterns that indicate a stolen card.

Many credit-card thieves share a pattern ... and data mining systems can minimize the losses in many cases by shutting down the card. In addition, the cost of false alarms is only a phone call to the cardholder asking him to verify a couple of purchases. The cardholders don't even resent these phone calls -- as long as they're infrequent -- so the cost is just a few minutes of operator time.

True, but there's also another point underlying this observation: having a credit card is a voluntary act. If I don't want my purchases or other personal information to be data mined, then I can simply pay by cash whenever possible. Government data mining is by definition involuntary; there of course cannot be an "opt-out" provision.

So the questions of effectiveness, efficiency and efficacy are all well and good, especially for those of us opposed to government data-mining, but let's not lose sight of the far more fundamental reason to oppose such programs, whether they work or not: because they are an unacceptably intrusive infringement of our privacy.

It's quite simple really:

"costs too high" + "benefits too low" = "not a good idea"

Jack Balkin links, not unreasonably, the Spitzer scandal to the War on Terror:

These events offer a window into a much larger phenomenon, the National Surveillance State, in which the state increasingly identifies and solves problems of governance through the collection, collation and analysis of information. Governments have always used information, but today's techniques are made more powerful and more prevalent by lower costs of computing and data storage.
...
If computing power increases enough, there is no reason why governments might not lower the threshold for reporting of suspicious transactions, or, indeed, require that every transaction over 100 dollars be reported. All this information could later be sifted through by data mining programs, in order to spot patterns of suspicious activity. The only limit is the technology and the manpower that law enforcement is willing to devote to analysis of financial transactions.

Of course, there ought to be another "limit," namely a rational basis for collecting such data.

It's one thing to suggest that there is a reasonable inference that a transaction involving over $10,000 in currency is likely, perhaps almost certain, to involve illegal activity, and that therefore the government has a rational basis to demand disclosure of all such transactions. That's not the libertarian conclusion, but it's not a facially absurd premise.

Demanding disclosure of all currency transactions involving over $100, however, is facially absurd. Demanding disclosure of all transactions of any kind over $100 would leapfrog over "facially absurd" and straight into Orwell. A terrorist can't bring down a skyscraper with one $100 bill or with one prepaid $100 phone card. The fact that the government can now technologically and economically perform such micro-surveillance still does not mean that it can constitutionally perform it. "You don't want another 9/11, do you?" is simply not enough to justify turning over every ATM, credit card, debit card and EFT transaction, no matter how small, to the government.

Even conceding (which libertarians ought never do anyway) that some (indeed most) laws are subject to mere rational basis review does not mean that a law that fails rational basis review (i.e., an irrational law) should not be decried — and judicially invalidated — as such.