Partitioned Sortition: Ensuring Proportional Representation despite Random Selection

Posted on November 19, 2020 by Yoram Gat

This post was withdrawn at the request of the author. You can still read the post on the author’s blog.

Filed under: Elections, Sortition, Theory |

« »

2 Responses

  1. Conall Boyle, on November 20, 2020 at 4:00 pm said:

    It is great to see a comment on the mechanics of random selection (and a very nice web-site to go with it!)

    I think the #1 principle ought to be Keep it Simple. The ‘punters’ being selected randomly are having to cope with a strange mechanism as it is.

    Your suggested solution is similar to the Stratified Sampling Technique used by opinion pollsters. We should heed their experience, they wouldn’t do it it if it didn’t work for them. Polling (randomly) on subgroups seems to ensure they seldom get it badly wrong, but at the expense of a degree of accuracy.

    But dividing the population based on M/F, Black/White, sexual orientation is a hazardous endevour in these days of ‘wokeness’! And it’s not really necessary. Sure, we can envisage rare pathological outcomes in the event of a one-off selection for a Citizens-Jury.

    But in the Sortitionist democracy we anticipate there will be repeated selections for juries. The Law of Large Numbers tells us that eventually all categories will be fully represented in the right proportion.

    More importantly for the acceptability of lottery selection is transparency. So many lotteries — the US Green Card springs to mind — conduct the draw in secret, ‘using a computer’ as if that should re-assure people of its fairness. If we want citizens to accept the results of lottery selection it has to be done in public. A mechanical system with balls is much more re-assuring than any computer program!

    Liked by 1 person

  2. keithsutherland, on November 20, 2020 at 6:53 pm said:

    Why was this post withdrawn?

    Like

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.