Hypothetical—Moon Station Government
You are now in control of a habitat on the moon. It has no ties to any government; its creation was funded by a wealthy philanthropist who just wants people to emigrate from Earth. The cost of doing so is within the reach of a middle-class family if they sell their home; you can therefore expect a decent number of immigrants.
What sort of government do you establish? How do you go about ruling so that your new settlement on the moon will survive and thrive?
LessWrong… IN SPACE!
I’d try to join the US as a state.
Is the moon habitat self-sustaining, or will bad things happen if it annoys the Earth governments who export food and water and stuff?
Physics provides certain tactical advantages to moon colonists. (Citing fictional evidence I know but as far as I can see the advantages are likely to be real).
A ‘technocratic’ Bureacratic Despotism, with the Philanthropist technically in charge.
Before you say otherwise, realize that the OP has asked: “What sort of government do you establish?”, not “What sort of government WOULD you wish to establish?” Regardless of my personal views, the wealthy philanthropist has too much influence and control over the government (by being its sole source of funding) for it to be truly ‘democratic’. I also have to make sure that the system survive and thrive, so I cannot rely on testing out new and exciting ideas, lest the philanthropist sues me when they fail. The safest method for surviving and thriving then would be to rely on a bunch of technocratic experts to handle the government for the philanthropist, and if the government collaspes, it would be the technocrats who are blamed, and not me.
I’d establish a constitutional monarchy, where the monarch is the head of the judical branch and the executive branch would be a senate with elected senators. The legislative branch would be devided amongst judical and executive. There would be a constitution which can only be altered by direct democratic vote. The government would not have the ability to levy taxes or other fees. The position of monarch would be tied to land ownership of the majority of land. The senate would be unpaid. Police and military would be directly funded by the local community as they see fit. There are still a lot of bugs in this system because it’s not really fleshed out, just an initial thought.
Why are we assuming that we would want to leave the state of nature?
No one likes Futarchy?
What sort of tech level is available? Are we talking a very early colony with not much higher tech than we have now? (ie, need to be super strict because it would be very easy for a malicious or even clueless individual to seriously screw everyone) or way beyond that so that’s not so much a worry?
Hard science fiction, pretty much. Hydroponic farms, meat-vats, and algae-vats for food, more-or-less modern life support systems (possibly assisted by cracking lunar rocks for oxygen), et cetera. Good enough to be self-sustaining, but no magic nanobots, FAI, or anything like that.
Basically, my question was getting at “how easily could one person misbehaving epically screw the entire colony?” or, alternately “How many things to the colonists need to do ‘just so’ to keep it all viable?”
The more easy for an individual to screw things up for everyone, and the more ways the colonists need to all work together in very specific ways, the more “rigid” the system needs to be, but it should not be a shred more rigid than that.
That’s obviously not a complete answer to what sort of government, it’s more my initial thought about how one would start thinking things through.
I agree. The tech level is a pretty major consideration. Government is also essentially a set of rules. A set of rules that will depend on the technology level.
Giving the actual question some thought.
What are the colony’s foreseeable needs in the future? What are its threats? Any governance mechanism setup will have to be one which tries to rationally minimise dependencies and threats.
My personal preferences are for some kind of a mutual credit money system, so that people’s preferences dictate the flow of goods and services.
i’m also certain that David Brin’s transperent society idea is somewhere in the ballpark of right. Radical transperency will help.
Otherwise, not very sure of what structure might help. Some kind of debate graph kind of mapping for arguments with assumptions laid out and having a separation of values and methods to achieve them (like in futarchy) seems to be a good structure.
Currently? I’d start by winging it for a bit until we got a few people together then do something a long the lines of the following.
Startup condition: Get the members of the community to nominate someone else to have a unit of power* (UOP) in the new system.
General running: These people would bid UOPs for various posts in government (see this post ).
Some of these posts would be posts that got feedback from the populace. The populace would leave feedback based on how well the government was doing. Say a 0 if the system was doing poorly or 1 for it doing well. This number would be averaged and then given as UOPs to those people. They would then give units of power to the people that helped him run the moon base (and so would they, recursively), if they don’t they won’t get helped next time and the expected number of UOPs would reduce.
Then after a period of time the people would have to bid for the posts again. If people didn’t do very well running the country, they would lose UOPs and those that did well should be able to outbid them for control.
Or something like this that I can show would maximise the populaces utility function as discussed in linked post.
*A unit of power is just a currency specifically tied to the control of the system
It sounds like the sum of UOPs is constantly inflating, possibly by a large (but perhaps nearly constant?) amount. This might or might not be a problem, but it seems to me to at least be a barrier. I don’t know enough about economic modeling, especially for this particular case, to decide whether spending UOPs on bids negates this or not. It seems like it would, but I can’t justify that.
Also, selfish agents won’t willingly pay their underlings much relative to how much they earn. I imagine that the UOP distribution will be similar to the income distribution for a corporation in the absence of a minimum wage, although I expect actual income and UOP income to have an interplay that allows one to mitigate the other. Again, this might not be a problem, at least if you think that people at the top really matter all that much more.
If UOPs are a currency, how are exchanges regulated? Can one write them into a will to be inherited? Can one purchase them with regular money? Can one give them to another politician in order to negotiate a policy? Can one donate them to another politician in order to see him or her elected? What about demanding that donation be a loan, perhaps with interest?
If you regulate all these things away, I wonder if you can still call it a currency. Still, it could be effective. As long as voters are 1) well informed and 2) vote, I could see this being more effective than current democratic systems.
It depends how competitive the voting is. Lets say we have two people Jim and Bob. The average amounts of UOPs you can get from being in charge is 5000, let say they both have the same fixed costs for flunkies (2000), it makes sense to bid anything up to 2999. They will both still get a profit of 1 UOPS. If either of them think they can improve on the average they might bid more than 3000, and if they aren’t well calibrated they would lose UOPS.
The big worry is collusion to keep bidding low, so that they can both gain control.
I’m not sure. If the market is competitive I’m hoping the transfers won’t need to be too heavily regulated as no one should gain too much, so they won’t be able to give much away without sacrificing their ability to gain a post. The worse that might happen in willing things is someone incompetent might be given them, and then you have a time period of not very good governance, and that person loses most of their UOPs and can’t get control again. But that might happen with things like alzheimers or other brain injury.
These seem necessary for new politicians to be introduced into the system, so I wouldn’t regulate them too much.
Its design is loosely inspired by learning classifier systems where nepotism/collusion is not too much of a problem, so I don’t know how it will interact with live humans. It might be like communism, nice in theory but not practical.
I’m planning to try it out at some point, on the small scale.
I’m definitely interested in how you plan to try it out.
The biggest problem that I see is the potential for
Rich person buys UOPS.
Spends UOPS to get office (for him or a flunky).
Uses powers of the office to get richer (but lose potential UOPS).
Repeat as necessary.
The rich get richer, and the powerful gain more power; we require a lot of protections against this pattern in our own governments, which still work poorly. It seems like with a fungible currency, this pattern is that much more dangerous.
I hope it would be moderately self-regulating. Whether sufficiently self-regulating I’m not sure.
Lets say Fred sells his UOPs for money, and becomes known as such, other politicians might decide to not work with him, as he is hurting their ability to make UOPs by introducing bad elements into the market. Collusion is a problem in this situation again though.
I’ll briefly sketch out my latest slightly more complex thoughts. It was designed to avoid the free rider problem. If you had multiple politicians getting UOP directly one may not bother to allocate their UOPs well or at all (as they don’t give UOPs they have lots to bid, so are hard to dislodge).
Lets say you have 12 ministers each that get slightly different flavours of UOPs, each month you need a different flavour of UOP to bid on a political post, including ministerial seats.
So the different ministers would have to trade UOPs between them, if they wanted to have long term control. If one minister was perceived not to be distributing them well (selling them for example) by the others, they might not be allowed to trade or get bad trade terms.
This is supposed to be a way of having the populace be able to give different feedback signals to different politicians without them having to figure out who did what at a specific time. Not perfect, I really want a theoretical framework to allow me to test out these ideas.
Numbers/times scales are for ease of thought only.
I’m going for an non-profit company devoted to trying to spread the word about this style of system, run by those same ideas. So definitely eating its own dog food.
So the organisation would try and get donors and fund research into the maths behind what would be stable, run experiments when people are encouraged to try to subvert the system (harmlessly) and build web applications/phone/facebook apps to manage the currencies. Hopefully some people will find it a novel and fun enough idea to want to take part/donate some and gain some control *.
If nothing else I’m willing to provide some enough money for web hosting or something and we can fight for control over that.
I’ve got a course to finish off first though.
This clause triggers my “too complex for human agents” switch.
I’m approaching this problem from the perspective of finding flaws exploitable by selfish agents, rather than reviewing it as a better method of organization between agents that desire to cooperate but have imperfect information; perhaps that’s not the point, and you’re not trying to make a system that is robust against corruption?
Fungible commodities are money, and I’m not sure whether I believe a government can work where politicians are paid based on their approval rating, and elected based on their worth times their desire to govern. Dividing up the UOPs into several buckets in order to force increased trade only increases barriers to trade and thus increases the need for a common currency, and the likelihood that all politicians will start exchanging UOPs for dollars (because they need to exchange dollars for different kinds of UOPs).
In our world, not liking someone is usually a small factor in negotiating trade terms, and I don’t see why moon-world is different in this regard.
I actually expect this to work much better for non-profits, because they are small and full of people who want to be there and support a common goal; I see an internal currency exchange as a signaling mechanism that might help group coordination. I think it’s worth pursuing, but I don’t feel like it’s analogous to running a government.
I was thinking about why I didn’t have an explicit problem with selling UOPs for money.
I see giving someone UOPs as equivalent to delegating some political authority. I’ve read a bit in a computer security field that has had long running arguments with the main stream security field about delegation.
They tend to argue that if bad things happen with explicit delegation, they can easily happen with implicit delegation as well. So your example of buying UOPs for money and then using the UOPs to get political power to enact something, why not just buy off the politician directly to enact something? I do propose that the transactions of UOPs be transparent so that people can see who gave UOPs foolishly/unwisely as well. So the delegation would be known.
You made me wonder if the dislike of explicit delegation is a bias. And if so why do we have it?
However the system doesn’t have any checks and balances for implicit bias. Somehow we have to align the politicians incentives with the long term good of the organisation. Huge potential payoffs X years after they leave office under some conditions? I’ll have to think about this some more.
You might be right.
That is the point, but I would also like to root out laziness. I.e. people doing nothing but managing to get UOPs somehow.
The assumptions it was originally designed with might not be compatible with politics, I’m unsure. If nothing else it might inspire other people to think of other systems that are compatible with politics.
In other words, a karma system?
Not quite. Systems differ on how the value signal is created and transferred.
When I upvote you on lesswrong I create a new bit of karma and give that to you. I could sneakily go through all someone’s old posts and upvote them all, in order to artificially inflate their karma (if it was important people would do this, they would also increase their post count subtly so that they could get more different things to upvote).
In the system I am describing there would be limited numbers of posts (as in positions in government, rather than posts on a message board) that you can get UoPs for serving on.
There is also not much you can spend karma on, so there is not much incentive to game the system. In a system where you need UoPs to get a post, there would be a lot of incentive, so the karma system would need to be robust to cheating.
Also I can’t give you some of my karma. It does not form a currency, which would be needed for governance with it.