Here’s an interesting paper by researchers who modelled the behaviour of a two-party or coalition parliament and showed that such a parliament is most effective (in terms of number of laws passed and social welfare achieved) if a certain proportion of the legislators are randomly selected. The proportion of randomly selected legislators needed depends on the size of the majority of the ruling party or coalition. (The results are interesting, but you don’t actually want to read the paper unless you like lots of numbers in your reading.)
Randomly selected politicians would, of course, have another benefit: fair representation. The United States and Canada, for example, are ostensibly more or less democratically governed, but, of course, we all know that numerous classes of people need not bother running for office. It’s true that they can vote, but they can’t vote for anyone who is like them. Select politicians randomly from the population at large, on the other hand, and then everyone stands an equal chance. This is precisely why that ancient democracy, Athens, insisted on random selection.
The traditional objection to this is that random selection is wholly unbiased and we in fact want some bias. More precisely, we want some bias in favour of the smart and skilled. But … have you seen the recent GOP debates?
Maybe it’s time to reintroduce the drawing of lots.