[jerf]

I see I may have skipped showing my work on too many steps here for my point to have made sense initially. Sorry, my fault (no sarcasm). So, when you try to concretely manifest the sortition idea, you find yourself stepping in an awful lot of the same problems you get with juries. With a twist, but very similar. I'd intend to have advocates on hand for the various positions, and if you think the bureaucracy won't be stepping in here with procedures and rules, probably enforced by a process moderator that looks an awful lot like a "judge", along with formalized rules for the other parts of the process, you don't know bureaucracies. Surely we don't expect the participants to all be auto-didacts? And yet, we can't just have one person teaching the relevant issues, or that one person has too much of the power.

And I don't say this often, but the bureaucrats would be right here. They would be necessary.

There's no way sortition is going to be "randomly choose 100 people, stick them in a room with nothing but a problem statement, accept their decision". It shouldn't be, either; it needs more structure than that. Right down to things like shielding the chosen participants from undue influences, and a lot of really interesting questions around transparency... again, quite like a jury.

[dragonwriter]

> I'd intend to have advocates on hand for the various positions, and if you think the bureaucracy won't be stepping in here with procedures and rules, probably enforced by a process moderator that looks an awful lot like a "judge", along with formalized rules for the other parts of the process, you don't know bureaucracies.

In a US-like system, if you replace election of leaders with sortition, there is no superior bureaucracy that can impose a judge-like supervisor on the sortition-based bodies.

And if you have an bureaucracy that can impose supervision on the elected bodies you are considering replacing, then you don't have a democracy to replace with sortition to start with, you have bureaucratic aristocracy, and talking about problems of elections and democracy is kind of irrelevant to dealing with your system of government.

> There's no way sortition is going to be "randomly choose 100 people, stick them in a room with nothing but a problem statement, accept their decision".

There's basically a couple of possible models of sortition; there's a legislature-like model (where sortition is how you choose legislators for a term, and then they are replaced with a new body chosen by the same method.)

And there's a case-by-case model where there is some method of proposing an issue, and then there is a "jury" chosen for an adversarial "trial" over the issue. (One way you could do this would be to replace one house of a bicameral legislature with case-by-case sortation, so that the remaining house would propose legislation and provide advocates for [and against, usually] that legislation, who would argue in front of a sortition-chosen panel.)

In the latter, case-by-case, approach, there would probably be some bureaucracy supervising the process, and the sortition-based body would be very much like a trial jury in relation to that bureaucracy. But then you aren't replacing the elected body with sortition, you are replacing it with a bureaucracy with sortition serving some function within that bureaucracy.)

In the legislature-like approach, though, a supervising bureaucracy is fundamentally inconsistent with the concept.

> Surely we don't expect the participants to all be auto-didacts?

I would expect members of a legislature chosen by sortition to educate themselves about issues the same way as any other legislature. This probably involves a subordinate bureaucratic infrastructure (e.g., staff employed by each member, and bodies answering to the body as a whole, like the GAO), but not a supervising infrastructure.