Thursday, November 07, 2013

When policies don’t work

I sometimes explain to people that I’m a computer scientist and so have – flying in the face of public opinion – very broadly transferrable skills.  The ability to make useful abstractions, systematic diagnoses and to understand complicated things as a whole, as parts and as the sum of parts are examples of basic training in CS. By-products of such training – if it’s done well – are critical thinking and sensitivity to when things don’t seem quite right.  Valuable. But there are side-effects too.We always have to worry about The Way Things Are Done.

Here’s a footling example that I enjoyed a little bit. It was our group meeting today and we were reminded  that the rules for expense claims now state that you can claim one alcoholic drink with a meal. You can’t claim more than one and you can’t claim even one unless there’s a meal.

The CS response to this is that it had better be a fucking big drink (I’m thinking those giant ornamental glasses they sometimes have behind bars to advertise some drink or other) and that if you want a drink on expenses but aren’t hungry, you should order a meal and not eat it, just so you can claim the drink on expenses. The taxpayer takes the hit for a meal and a drink instead of just the drink. Food is wasted. The world gets a bit poorer.

But there’s an even worse problem. I’ve been subject to it, although not recently. When you’re a young research fellow or PhD student at a conference and you go out for a meal with a bunch of important professors, your boss will invariably take the bill, ask you to pay it and claim it back.  But of course that only works if the profs are operating according to the same rules. Chances are they’re not and students certainly sometimes get saddled with bills they can’t claim back.

The lesson is this: if you’re going to write a policy, you need a computer scientist. Lawyers and administrators can’t seem to do it.

No comments:

Post a Comment