4 Comments

Jim - some immediate concerns/questions …

Presumably agents are individuals (not pieces of software), in which case how many of them do you envisage - one for every 10/100/1,000/10,000/more people? How do they keep track of the people they are representing? And how do they represent diametrically opposing views and keep the confidence of their ‘constituents’? (Oh, and how are they funded?).

Your proposal appears to be IT heavy when not everyone has computer/internet access, or the inclination to get it. Are you not in danger of disenfranchising a significant part of society? You need to cater for Brenda from Bristol (https://www.theguardian.com/politics/video/2017/apr/18/does-brenda-from-bristol-have-the-best-reaction-to-the-election-news-video).

Obviously I’d like to see more democratic representation but am struggling to see how a panocracy could work in practice.

Cliff (being very mundane)

Ps maybe one day software will be the answer - when we can trust it!!!

Expand full comment

Hi Cliff and thanks for the comment. I'll try to address your questions in the same order as you raised them.

I see each agent as an organisation - some big, some small. Some may handle 1000 clients, some 100000. Their size will be set by the usual factors: cost, quality of service, reputation, ... There may be agents with a very small number of wealthy clients and there may be bargain basement agents with lots. Like any other organisation, they'll use software systems to keep track of their clients. Their clients may well use software - I imagine most will - but some may not wish to.

Agents don't represent views in the way that I think you mean. They merely present the positions of their clients to each vote. It's conceivable that 50% of the votes by a particular agent's clients are in favour of some issue and 50% against. Diametrically opposing views will effectively cancel each other out as they do now.

How they're funded is a political decision and therefore rests with the panocracy itself. It could decide that they're all private and paid for directly by their clients or it could decide that they're paid out of a general levy, or indeed some mixture of the two.

Yes, it is IT heavy because operating a panocracy is a big and very, very boring task and we have developed computers to do precisely that. We didn't have IT up until recently and that's one reason why panocracy is now a viable proposition.

People who don't have access to IT will still have access to an agent. They would have to go in person or talk on the phone or correspond by mail (if there still is such a thing). Agency costs might have to make allowance for a small percentage of their clients doing things this more expensive way but not having a laptop or a mobile phone or whatever will not bar access to the system to anyone.

Ha! I feel Brenda-from-Bristol's pain, too. That's why the panocracy is set up the way it is. Most people will be able to ignore the day to day votes secure in the knowledge that their views are *explicitly* being taken into account (and they can check). If they're interested then they can take a much closer look but I would expect most people to trust the system to just get on with it as they do now. The difference would be that their goodwill and trust couldn't be usurped by the ideologue, the incompetent or the vested interest as it has been.

I hope I've clarified some of my proposals. Please keep the comments coming!

Expand full comment

I feel there needs to be an audit function built in. I may be confident my agent is casting my vote as I would cast it myself, but how can I be sure a) my agent is being equally scrupulous with all their other voters b) all agents are doing so? The audit function should be independent of all agents, and should be able for example to sample 5% of all votes cast for / against a particular measure / policy / law, and then confirm with a random selection of that 5% directly with the voters, that they are happy with how their vote was cast by their agent. If the number of dissatisfied voters exceeds a certain threshold then the vote is declared unsafe. The threshold would vary according to the margin by which the vote was won / lost. So a very close result would require a relatively low threshold of mis-cast votes to be declared unsafe. Where a result is much more clear cut a correspondingly higher threshold of errors would be needed.

Expand full comment

Thank you for your suggestion - which I support. An audit function will be necessary across the administration. It must be independent and I believe that means it must be open and strictly procedural to avoid pressure by vested interests. This may mean it can be automated.

Also I expect that interested parties in the electorate would be quick to flag problems. We've seen lots of people from many disciplines point out the problems with the Covid response - despite widespread censorship. Their early views have been mostly vindicated as evidence accumulates.

I very much like the idea that the thresholds should be dependent on the closeness of the vote. I think it will be possible to use probability theory to set the thresholds (i.e. percentage of voters sampled versus closeness of the vote) to achieve a specified degree of statistical confidence but the details need to be worked out.

It would be easy to 'hold' votes considered 'unsafe' until the audit had been completed. I guess they would be re-run.

This does beg the question of what would happen if an agent was found to be incompetent or misrepresenting their clients. The Law and its enforcement is a big issue and I've shied away from it so far! I (or someone else) will have to bite the bullet at some point!

Expand full comment