Rethinking regions and companies

[orthecreedence]

A region is a collection of shared assets in a geographic location. However, after some discussion this might morph a bit.

It's interesting because this concept of shared assets makes sense in some places and not in others, and a discussion on reddit (first comment below (ah, hell, I'll just link it again here)) really changed some of my viewpoints.

There are two things that immediately come to mind when thinking about shared assets:

• Things specific to some producers
  • Large equipment (construction, farming, etc)
  • Various "real" property (farmland, forests)
• Things general to everybody (mostly)
  • Housing
  • Commercial space
  • Transportation

If I live in regional housing and use a regional office space, does that give me the ability to vote on how many tractors the region owns? The answer, really, comes down to whether the capital pool is shared between me and the farmers in the region. If yes, then I should obviously get a say. If not, and this is where it gets interesting, then no. And we're starting to get into networks-within-networks territory here. What if the farmers in my geographical area had their own network and capital pool and it managed their farmland and various machinery, and they solely managed it themselves? This would be closer to the picture of workers managing the MoP based on use.

Then thinking about things like housing, maybe that would be its own local network that everyone who wants socialized housing is a part of, and they all make decisions along those lines.

This breaks the geographical regional model and turns it more into an emergent network, and these networks are free to form, grow, shrink, and dissolve as they see fit. Each one would have its own shared set of assets and capital pool. If I'm a member of a company that is participating in the system, I can request to join any number of regions in the system, and hether or not I am accepted would be dependent on criteria (what companies am I a member of? where do I live? etc). In other words, the idea of one member, one region is more and more seeming brittle and straitjacketed.

This brings up a lot of questions, some of which were previously unsolved to some extent, but some of which are new and complicated:

• Geographic grouping makes sense in the case of pooling money together to buy things like housing for everyone. If regions are just sub-networks, the farmer "region" and construction "region" all will have siloed capital pools. How do these capital pools help other regions, like the local housing region?
• If regions are no longer geographical, what stops us from having a region-per-company (effectively capitalism)?
  • Should there be defined limits in the network? Members of a region only get at-cost pricing if there are >3 companies in that region?
  • Note that this was somewhat unsolved already in the current notion of regions.
• Having some grouping of people by area does make sense in the context of housing, commercial property, transportation, etc. Would this form via emergent behavior, or would it need some kind of systemic guideline?
• What does membership in various regions look like? Obviously joining requirements are similar (must be a member of a member company), but if regions are less rigid then would they have other requirements, like "must be a member of a farming company" or "must be a member of a healthcare company"? Seems each region would have its own membership requirements.
• How does UBI function (UBI #79) if regions are more nebulous groupings/networks?
  • Leans toward it being systemic (global), but maybe there's some strategy for pooling resources among regions here.
• Naming for regions? Region implies area/geography. "Group" is too general. Nothing immediate comes to mind.
• How does this affect things like banking (Banking updates #73)? If I cash out my credits, who pays me? Maybe the region my company is a member of? In which case, credits need to track where they came from.

[orthecreedence]

See https://www.reddit.com/r/DebateAnarchism/comments/hbhxnk/the_governance_challenge_of_blockchain_ecosystems/fvdc51c/

A very interesting high-level conversation about regions and their scope.

[orthecreedence]

removed the relation with #71

[orthecreedence]

Thinking this through more and stewing on it for quite a few days, I'm becoming more and more convinced this is the way to move forward.

If we generalize the "company" object such that a company is really just an agent (VF already stipulates this) then it becomes clear that a company can really just be a grouping of agents and assets. This means a region could really just be a company of members and companies. How companies use assets and the permissions they exercise on them is extremely important and so far the goal is to have these permissions determined democratically. But if you let companies be members of other companies, you can effectively define a region in the exact same way you would define a company.

[orthecreedence]

More to the above points:

• A company is a collection of agents. An agent is a person or a company.
• A company can own assets, and members of the company can have access to those assets based on permissions that are decided democratically.
• When making decisions for a company, the atom members (ie, users) are the ones who vote. Companies themselves (as members of other companies) have no agency: only their constituent users do.
  • The goal of this is to avoid creating bureaucratic hierarchies but instead allow all user-members to exercise their individual power independently of the various companies they are members of.

For instance, if User A and B work at Company X, User C, D, and E work at Company Y, and Company X and Company Y are members of Company Z, then when deciding issues for company Z, A, B, C, D, and E all have decision making power, even though their membership to Z is assigned through companies X and Y. In other words, X and Y are conduits for decision making power, not aggregators or representatives. If A, B, C, D or E wish to delegate their voting power to various members of companies X or Y via liquid democracy, they are free to do so, but X and Y exist only as a membership link and not as power structures.

It's important to note that User F might be a direct member of Company Z, and as such would be allowed voting just like A B C D and E on issues of Company Z, but F would have no voting rights to Company X or Y unless F joined their ranks as a member.

[orthecreedence]

There are some other considerations here that are worth discussing.

With regions, the region is effectively a cell in a larger body. The cell is autonomous: it has a bank, the ability to conduct business with the outside world, federate with the other cells (regions), and acts as a geographical container for companies and members.

Breaking this into more general pieces breaks some of this. Individual companies would get access to their own capital pools (originally reserved just for regions). I view this as bad, because now any company can produce individually for profit. There might need to be some rules about capital pools only being allowed for companies that contain other companies.

But it also breaks the capital pool model in other ways: with the region, all capital goes to the region, ie the closest capital pool. With the general model, if Company A is member of B is member of C and A sells a widget to walmart at a profit of $100, where does the $100 go? B's capital pool or C's? Is it a function of the assets A uses from both B and C, ie if A uses a tractor from B but farmland from C, do we compare the value of these to determine the distribution of the profit? How is this value determined without markets?

One idea that has been swimming in my head is the idea of a per-member (user member, not company-member, ugh need better terminology) distribution setting. Basically, for each company you're a member of, you decide how your share of what would be profits are distributed.

For instance, if you're a member of the Super Tech company and the San Francisco Housing company, you might have 60% of profits generated go to Super Tech and 40% to housing. Now, if Super TEch has 10 members, you control 1/10 of the profit, and if SF housing has 1000 members, you control 1/1000th. So if Super Tech makes $8000 in one month, you control $800 of that, and per your distribution Super Tech keeps 60% ($480) and the rest, $320, goes to SF housing. Now if SF housing makes $10000 in profit in a month, same thing: you control $10, $6 goes to Super TEch, $4 goes back to SF housing.

So each user-member sets distribution of what would be individual profits to be invested back inot the companies they are members of.

Concerns with this model:

• Complicates membership
• Companies are encouraged to compete with each other over distribution
• Distribution settings would necessarily need to be private, lest a member be outed for their decisions.
• Effectively acts as a tax, although the capital would be off-limits to members anyway.

Fixes for above concerns:

• Set all distribution points to 1 (all companies get an equal distribution). Don't allow members to change distribution values yet (or ever? or set minimums/maximums?)
  • This still doesn't fix the problem of a company wanting members to NOT be members of other companies, but maybe it would be fine because they would get distributions from the other companies as well. In other words, everyone shares with everyone, as opposed to a taxation system where some central authority collects and redistributes.
  • This, I must reiterate, only matters when the network is not yet critical mass.

[orthecreedence]

Next concern: companies and capital pools. There are two ideas here:

Don't let producing entities control capital pools

The idea here is that companies shouldn't be able to make a profit, and that profits they do make should go to what used to be the region (one level up) so it could be used for other things (housing, MoP). The goal is to eliminate the idea of producing for profit.

Pros:

• Enforces profitless production

Cons:

• A larger leap from our current system (harder to onboard companies)
• Complicates the system (ie some companies can have capital pools but not others and needing to draw the dividing lines between them)

Let producing entities control capital pools

The idea here is that companies should be able to make individual profit, but would not be able to distribute it (ie, must spend it on reinvestment). This is very close to the original idea of Basis (ie "regional socialism"). That said, the profit balancing scheme outlined above would still apply, so profit realized by a company would be spread across the members' companies (we know the profit because we track costs so carefully).

Pros:

• Simpler model: all companies operate the same way.
• Closer to current system and easier to transition existing co-ops
• Logically consistent: if a company can own assets, then it follows capital would just be another asset a company can have command over.
• Allows companies "capital autonomy" from the previous concept of a region

Cons:

• Muddies the "profitless" water somewhat. That said, if a "region" can make profit and enforce the inability to distribute that profit, it follows that a company should be able to do that as well.

[orthecreedence]

I think the final piece to this then is banking. Its inner workings were tenuous at best with the regional model, but now this throws things into disarray again. Or maybe some simplistic solution will come from this new reorganization.

The main questions:

1. Who pays who (capital) when resources with a currency cost flow across company boundaries?
2. Who pays for conversion of credits into currency?

Who pays who (capital) when resources with a currency cost flow across company boundaries?

In the old model, a widget produced in Region A with a currency cost of $10 that was ordered by a company in Region B meant that the bank in Region B would send $10 to Region A. Pretty simple.

With the company model, one thought that comes to mind is the idea of "nearest bank" or "nearest capital pool." The idea here is that when Company B orders the widget, the bank/capital pool closest to B would pay into the pool closest to A. So if B has a capital pool of its own, that would be used. If B has no pool but is in a larger company (C) that does have a capital pool, C's pool would pay the nearest capital pool to A.

Problems:

• A company can be a member of multiple higher companies, each with its own capital pool. What is the "nearest pool" in this case?

Who pays for conversion of credits into currency?

This one is a doozy. It was difficult with the regional model specifically because of the federation (ie, how do banks all fix the credit value to a peg?) but with the company model it's even more convoluted. The capital has to come from somewhere. A few thoughts (spitballing):

• Set up a global company every member is a member of. This directs profits to the global level, and the global level would pay out the conversion.
  • Centralization (brittle/attackable)
• When credits are converted to currency, spread the withdrawal across all capital pools in the system.
  • Equal by pool? Pool A and B have $1000 and $10000, a withdraw of $100 leaves Pool A at $950 and B at $9950.
  • Percentage based on pool value? Pool A and B have $1000 and $10000, so a withdrawal of $100 would leave pool A at $990 and pool B at $9910
• Pull from pools user is a member of?
  • Would take money from housing/MoP
• Pull from nearest pool?
  • A member can have multiple "nearest pools" just like a company
• Don't set up a global pool, but rather a separate one specific to withdrawals that companies contribute to voluntarily to maintain the peg.
  • Good luck with this one! (like I said, spitballing)

Not seeing any options I like very much...

[orthecreedence]

mentioned in commit core@46dc04d2bcfee25eced4f8e21fd9b3d8af047888

[orthecreedence]

#88 is a better discussion of the dollar peg.

[orthecreedence]

Ok, I think the ideas here have been solidified enough and pulled into their respective issues (and, now, written into the paper) that this particular issue can be closed.