Benevolent government data mining, or Big Brother?

The New York Times recently ran a fascinating story called, "U.S. mines personal health data to find the vulnerable in emergencies." It raises provocative issues that will become more and more relevant over time, as technology allows for ever-improving data mining and data analytics.

Here are the facts of the story:

There are a growing number of local programs that use government data sources, such as Medicare insurance claims, to locate people who are especially vulnerable to problems during emergencies such as hurricanes, and then to contact them to make them aware of problems they could face. Examples include people on kidney dialysis who might be advised to seek early treatment because a natural disaster would close dialysis centers, or people using electronic breathing devices who lack battery backups and could run into a huge problem if power goes out.

Good idea or not? As government becomes more sophisticated about how to mine and analyze large databases, this issue will start coming up more and more, limited only by the ingenuity and creativity of dedicated civil servants. (Or contractors trying to sell an app!)

"There are a lot of sensitivities involved here," the article quotes Kristen Finne of the Health and Human Services Department as stating. "When we started this idea, there was a lot of 'are you crazy'?"

The sensitivities are obvious in a privacy-obsessed U.S. society -- particularly in the wake of the revelations regarding the National Security Agency's capture of communications long assumed to be private. So let me give my opinion, but then ask blog readers to weigh in on this.

I would rank myself as below average in sensitivity to privacy. (My wife notes that, to her shock, the bathroom doors in my parents' home were typically left open, so this relative indifference may run in my family.) I would also definitely rank myself as way below average in paranoia about government abuses, at least in the United States and other robust democracies. While of course we must always remain vigilant about abuses, I am about as far from the psychology of the black-helicopter conspiracy crowd about government as you can get.

So my intuitive reaction to these examples is basically, "What's the problem?" Intelligent people should be able to make a distinction between activities safely on this side of abuse -- such as these disaster assistance examples – and illegitimate government snooping on individuals. (It should be noted, of course, that people are free to say "no thank you" to information or heads-ups they may get.)

That said, let me cite two other examples from the article that are at least somewhat more problematic. One -- the less-troublesome of the two -- involves sending people whose kids have not been vaccinated an email message pointing this out and urging them to do so. The second, and more worrisome, involves informing insurance carriers of patients who make unusually large use of ambulance services paid for by Medicare.

The first example has more of a ring of paternalism than the problems with dialysis and batteries, which people are unlikely to have thought about. Vaccinations are something most parents know about, or should be expected to, and some deliberately choose to opt out. In the other example, the use of the data mining is not to help the individual but rather as a control function.

Having said this, however, I guess I would argue that such interventions are justified -- even the second one, where the government has a legitimate interest in reducing unnecessary taxpayer funded spending.

But I would love to hear how readers react to these interventions. We're going to be seeing more and more of these proposed as time goes by.

Let me also take my hat off to the civil servants at HHS who thought somewhat out of the box on this, looking for ways better to serve customers and taxpayers. I hope, whether or not we like these ideas, we can cherish the interest in innovation they represent.