Saturday, January 02, 2010
Regular readers know that I object that the chattering classes -- including both politicians and the mainstream media -- do not hold government (or newspapers, for that matter) to the same standards of accountability that government imposes on businesses, sometimes at the risk of criminal liability. For example, how is it that climate models used by the government to shape policy are not constructed under the process controls that we require for software used in medical devices? Naturally, therefore, I was delighted to see the same question raised about the government's obvious mishandling of "material" information about the Christmas day panty bomber.
It occurs to me [CIA Director Michael] Hayden and his colleagues are getting a taste of what the private sector deals with under the securities laws.
Firms are supposed to disclose "material" information. The question of which of the zillions of bits of information they have will be deemed material is often assessed by a judge and jury in hindsight, after a given bit of information turned into a big problem.
CWCID: Tom Kirkendall's Facebook status.
"Holding government to the same standard as business"
But, they are the Government. There are no "same" standards.
The government routinely exempts itself from the laws that ordinary citizens/residents must follow, for if they didn't, they would be one of the largest segments of lawbreakers in the country.
"how is it that climate models used by the government to shape policy are not constructed under the process controls that we require for software used in medical devices?"
The climatologist don't have anywhere near the same budget for software.
Also, the software involved in control of equipment is completely different from that used for statistical analysis. In software for medical equipment, or cars or robots, a bug will have obvious and dangerous effects. It will be proprietary and closed-source.
The statistics software is likely to be standard stuff used all over the place and well tried and tested.
What you are complaining about is the interpretation of the results, not bugs in the software itself. If you want a medical example comparable to climate, you might look at research into the medical effects of cannabis. Hard to measure, noisy data, emotive topic.
"Hard to measure, noisy data, emotive topic"
That's the basic problem with climate "research". Generally, things that are hard to measure shouldn't be considered international emergencies...
unless there is a veiled purpose.