Scaling Loomio, Functional Decision size?

@alanna done :)

http://blog.loomio.org/2015/03/09/some-thoughts-about-large-scale-decision-making/

Hello. Apologies in advance for wall of text. I’ve been thinking about these issues for a while.
It seems to me that there are two problems of scaling. If anything the less serious one is how to make a decision involving huge numbers of people. The more serious one to me is, over millions of people there are too many decisions. Not everyone can be involved in all of them. How do we scale the use of Loomio-like direct democracy mechanisms to allow all the decisions to get made? How can we make all the decision-making direct and accountable without everyone spending 57 hours a day making them?
A few thoughts about the first problem. First, it seems to me that while on a smaller scale there are good reasons to favour consensus approaches to decision-making, when you get to a decision with huge numbers of people participating, while it would be good to retain some of the sensibilities of consensus approaches I don’t feel that actual consensus is a feasible way to do things. If you have twenty people you can work things out so you get a result everyone can live with, or even work things out so you have an everyone who will be able to work consensually together. If you have two hundred thousand, I suspect there will always be some who disagree.

Second, taking that as a background assumption, I also think that with that many people it becomes a bad idea to structure things so that voting is on one proposal. There will be many ideas, some will not be able to be assimilated into a single proposal or persuaded to back off. A mechanism for decision-making on a large scale would ideally accommodate voting among multiple options suggested by discussion participants. Otherwise the first mover would control the debate, and the structure would be very vulnerable to the whole “Something must be done. This is something. Therefore, this must be done!” fallacy. I tend to favour a ranked ballot type setup for deciding among the multiple options, but there are lots of possibilities.
For really large decisions, this would also likely become unwieldy. With enough participants, you could end up with hundreds of proposals to choose among on a single decision topic. Many of them would be largely redundant. There would have to be some way of winnowing them down before the real decision was made. I envision a couple of things, first having some moderators who would look proposals over, try to figure which ones were functionally about the same and get the proposers together to draft a common unified version. And some kind of pre-decision dealie perhaps, where only the (some number of) proposals with the most “likes” or whatever actually got decided among in the end.
For me this problem, while difficult, has been less of an Achilles’ heel for egalitarians than the second problem, of decision overload.
The current approach divides decision-making hierarchically and among things like ministries, departments and so forth, such that top-level political representatives make very broad decisions and set the tone for those further down, who make decisions on less broad issues within their particular silos, and set the tone for those still further down in further-subdivided silos and so on. Eventually you get down to the majority of people who make very trivial decisions in very narrow contexts, or effectively no decisions at all. Those higher up can generally countermand the decisions of those below. Overall, an awful lot of people are involved in making decisions of some sort, and on average few are involved in each such decision. All of this is a result of a lot of practical compromises with the core organizational goal. In hierarchical decision-making, the ideal for those at the top would be for them to be in complete control, make all the decisions, and for everyone else to make none. This is physically impossible, but they do try, as terms like “Taylorism” and “micromanagement” attest.
One key objection to serious popular involvement generally in decision making is that if you radically expand the number of people involved in each decision on average, the total amount of time and effort devoted to the making of decisions greatly increases. And indeed, increases even more than the number of people, because of the very problems we’re discussing here; the more people who might have something to say, the more there is for each of the people involved to assimilate and potentially respond to, and so on. It’s a process of communication, and more people means more happens; without care it can be exponential.
But I think a lot of the problems with the ways decisions are currently made are more about the hierarchy and the silos than they are about the number of people actually making the decision. Hierarchy means those deciding are different from either those the decision impacts or, much of the time, those with knowledge of the issue, both of whom are generally lower on the totem pole than those deciding. Thus hierarchy also means those affected and those with knowledge cannot change the decisions. And of course people higher up, deciding what will happen to others, typically have different interests from those of the people they’re deciding for. Silos, the idea that a particular group owns a kind of decision and is allotted sole power over it but has no involvement in or power over other kinds which may be related, make all this worse—they further limit information and accountability, and widen divergences of interest.

None of these problems have much to do with the sheer number of people involved in the decision. It’s just that we think of it in a binary opposition: The opposite of elite/autocratic decision-making is decision-making by “the people”, and so we think in terms of everyone in general deciding. That would be the ultimate. I think for an egalitarian approach to decision-making, the practical compromises should not and need not involve a little limited re-introduction of hierarchy. Rather, we still need to split up decision-making but the split can be done horizontally. What’s needed is distributed decision-making, with groups making decisions about particular kinds of things, but flat groups open to entry on a voluntary basis by anyone who cares to, perhaps also deliberately inviting people with relevant expertise. Further, groups could nest, be broader or narrower. There will be a need for some decisions to see wider participation. Rather than someone higher in a hierarchy being able to countermand decisions from lower down, a broader group should be able to countermand the decisions of narrower groups that are part of it. And, a narrow group contemplating a decision should have a way for people in it to decide it’s too big and needs to be kicked out to the wider group.
So in this sense, the model I’m proposing remains the opposite of the hierarchy model in a sense. Rather than the big, wide decisions being made by the few at the top with decisions going narrower as they go down to the masses, the few big wide decisions would be made by pretty much everyone, the public at large, with narrower decisions made by smaller groups according to their interests (in both senses of that word).
We see something like distributed decision-making today in Free/Open Source software production; people decide what projects they will involve themselves in, if any, and a huge ecosystem of software has grown up this way. It has its shortcomings, though; many important but unnoticed niches end up with crucial work undone. In an actual full-blown society trying to make decisions this way, I might expect compensation through something analogous to jury duty—everyone might have to be involved in some minimum number of decision-making groups, and in addition to the ones you choose, there might be some sort of random assignment so that necessary but unsexy things got some attention.
(On a side note—Richard Bartlett mentioned Parpolity. Back when I first heard of it, I was expecting great things from Parpolity because I’m quite a fan of Parecon. Then I read through the basic proposals, and it strikes me as just a delegate model pretending not to be one. Worse, it’s multi-layer delegation, making those in the “top” layers even more distant than they are now. It seems much like the way the more unwieldy, unresponsive large union federations operate today. Not an improvement, not immune to capture by powerful interests, not the direction I’d like to see at all. This sort of committee-of-committees-of-committees model has been proposed many times, I think because starting from a single group that seems like the obvious way to generalize, and perhaps without technology it would be all that could be done. But I don’t think it stands up to serious thought and I am certain that we can do much better nowadays.)
Well. That’s my, rather more than two bits especially at today’s prices. I hope it’s in some way useful or thought-provoking to somebody.

A great lot of thoughts @purplelibraryguy. I hope you don't mind that I reposted it. The group is open if you wish to join and I am happy to remove it if you don't approve.

https://www.loomio.org/d/LKOhVmqr/large-scale-decision-making

Don't mind at all, to the contrary flattered.