This month we have to choose our medical insurance provider for the next year. For non-American readers, there are three basic alternatives: An HMO (Health Maintenance Organization) which is a company that contracts to insure employees of other companies, and in turn hires contracting companies that are small collections of doctors to do the work; A self-contained variation of the HMO, where doctors and whole hospitals are run by the company retained by your employer to insure you, thus having one fewer layers of contractors but a more restrictive choice of practitioners; or finally a company that has very little control over your choice of doctor and is more like a simple insurance company. This last gives the widest choice (virtually all doctors who charge the standard fee and agree to communicate on billing matters seem to be available) but is the most expensive.

We at Agilent in California have about four choices of HMO. One of these, the one we are currently using, appears this year to be cheaper by a factor of two or three than the other alternatives. (They were about equal a year ago.) On paper, these HMOs are almost indistinguishable except for the "customer satisfaction rating" in which our HMO has seen a marked fall from 90%+ to 60%+, the others seeing no significant fall.

What goes on? The only scenario I can see goes like this: The employer subsidises the employee payments, so that one typically sees only a few percent of the premium deducted from your (before-tax) income. Call that s% for example. If this HMO of ours has managed to tighten its belt by only a few percent, x% say, (with consequent incidents of dissatisfaction), it has the chance to say to Agilent "look, if your employees use us, you will save x% of big sum... so you can afford to bribe them to do so by charging only 0.5*s% and you will be ahead". A little algebra shows that if you want to offer a factor of two in premiums to the employee, and the employee normally pays 10% of the real cost, the HMO needs only to cut costs by only just over 5% to make this work. I like to think that Bill and Dave would not have approved, but I bet Agilent thinks it is great.

What can an HMO do to cut costs? HPR seems to overload doctors, to refuse specialist visits where the need is not unarguable, and give staff no jurist diction to use common sense.

Mid way through composing this reflection (overdue through the not-unexpected arrival of the vice-widget) I was assailed by emails bringing the news that Radiata has been bought for about US$280,000,000.00 by Cisco. This carefully-plotted scenario ("go straight to millionaire's row, do not ever again drink wine costing less than $200 a bottle") is to the Californian engineer the Nirvana-dream... not unlike working for Hewlett-Packard was to an undergraduate in an unaware, would-be third-world place like Sydney in the 1970s. Still, it could not happen to a nicer guy than Dave, except possibly me of course.

We had an "intruder" in the garden Sunday night last. We heard a clatter, and vaguely saw a large shape blunder its way through the branches of the tree that I had that day pruned but not yet chopped up. The clatter proved to be cans scattering, the cans from our recycling crate escaping the bag in which I had carried them out. The shape must have been human, and it must have very rapidly leaped the fence on the far side of the garden to effect an exit before I switched on the garden flood lights.

We know that the official recycling truck is usually preceded on a Friday morning by someone filtering the crates in the street for aluminium cans. However, you know that margins are thin when they have to climb fences and do runners with your cans mid-week, or bribe you to join a medical insurance program that has slightly less generous policies.