Self Management and Parecon?
If not authoritarian relations, why not concensus? Isn’t that the opposite, and best, for a participatory approach?
Participatory economics does not institutionally prejudge what procedures should be used for decisions made in workplace or consumer councils. It does not say you have to use majority rule or consensus or any other particular procedure. It could be that in a real parecon, workers and consumers opt for consensus decision-making all the time, much of the time, or rarely. That is a choice for them. What parecon prescribes is that people should ensure, as best they are able without investing excessive time and energy, that each actor has an impact on outcomes in proportion to how much he or she is affected.
As potential participants in a participatory economy, however, do we ourselves think it would make sense for workers and consumers to conduct all their decisions via consensus? No. We think consensus makes very good sense for some decisions, but not for others. There are two key but quite different aspects to consensus decision- making that bear on this perspective. One is about process. The other is about formal power.
The process of consensus decision-making emphasizes respect for all parties and the use of diverse methods of information preparation and dissemination and subsequent discussion and exchange to ensure that each person’s input is appropriately accessed and addressed. It is important to realize, however, that techniques for how information is gathered and addressed are one thing, and for how power is allotted are another. That is, the same methods of being sure that information gets out, preferences are expressed, issues are addressed, etc., as are used in contemporary consensus decision-making can be utilized when decisions are being made by one-person-one-vote majority rule, or by one-person-one-vote two-thirds needed for a positive outcome, or by other norms. Indeed, it would probably simplify debate about these matters if we had two concepts or names: one for the method of mutual discussion and information exchange, we could call this participatory preparation, and one for requiring unanimous consent, which we could call consensus.
At any rate, the second component of contemporary consensus decision-making is that for a decision to be settled, all must agree with it or at least refrain from blocking it. Each actor has a veto they can employ. The theory is that people (whether individually or in groups) will not veto options unless the impact of the choice on them is so great that they ought to have the right to block it. In other words, the implicit and sometimes explicit logic of consensus decision-making is that it permits each person to determine, relative to the others, the degree to which they are affected, and to then submit or withhold their expressions of opposition in accord with their best estimate of their own situation relative to the reported preferences and situation of others. If one actor or a group together among the people making a decision is sufficiently affected that they believe their rejection of the decision should dominate the outcome, then he, she, or they will oppose or block it. If they do not like it, but they do not think they should dominate the choice, then they will abstain or otherwise avoid blocking it. In this sense, when used as intended by actors who are attuned and respectful of one another, consensus decision-making works perfectly. Only individuals or subgroups that dislike an outcome and would be in sum sufficiently affected by it to warrant dominating the outcome, will opt to impede decisions. Working thusly, when consensus decision-making fails, imperfections derive not from having established an inflexible and inappropriate procedure for making decisions, but due to mis-estimates of each other’s feelings or the impact felt, or to abuses of the unfolding process by individuals in the group. So the question becomes, how likely are we to have good interaction and outcomes rather than problematic ones, and are the prospects for the latter low enough, in all contexts, to warrant using consensus all the time? Or do the prospects differ for different situations and decisions, so that in some cases using other approaches will be more likely to yield the best results with the least hassle?
Consider hiring a new worker for a small workplace, or adding one to a small work team. Suppose we collectively assess this type of recurring decision in our workplace and decide that in light of who we are, the time we have for this type of decision, our general situations relative to decisions of this type, etc., this is a situation where the impact on each person of a choice to hire someone that they don’t like is huge, whereas the impact of hiring someone they do like on any actor is much less. Everyone has to work in close proximity with a new person day in and day out, and if anyone really doesn’t like him or her, that will potentially be a far more serious problem for that person than it is a plus that everyone else favors the hire.
So in our workers’ councils, we decide that for each new hire to our small workplace where everyone works in close proximity and knows one another well, everyone involved is entitled to a veto. The voting guideline might be that you need three quarters to approve someone for that person to be hired, but that anyone who is strongly enough opposed can block any proposed hire no matter how many others favor it. The voting rules aren’t reworked for every new hiring situation, but nor do they imply a universal rule that applies to all other types of decision. Instead, this is a pre-agreed rule specifically about hiring decisions.
And note, it is chosen because it makes life easier, not harder, in that it approximates most closely what we generally think will be appropriate input for each person involved and thereby reduces the complexity of arriving at the desired result once we begin our deliberations. The person who is highly upset over a new hire doesn’t have to convince everyone of the validity of her concern and get them to vote her way as well. She is concerned, period. She doesn’t have to explain why. She gets a veto because being strongly opposed to hiring trumps favoring hiring. There is no need for everyone to engage in fancy mutual calculations to decide if they have the right to trump, though of course, as with any procedure, we can include diverse methods for communicating feelings, etc.
But suppose we had instead adopted a one-person one-vote majority rule approach to hiring decisions. Now the person who feels her life would be made miserable by the new person’s entry must convince a majority of others to respect her strong feelings and vote her way. If she fails, her strong feelings will not have their appropriate impact on the final decision.
Something interesting characterizes the above comparison. In this particular type of decision, it turns out that the consensus approach (not the communications methods but the voting system itself) can yield proper results even with less mutual empathy and less communication of preferences and compromise than simple majority rule voting would entail. In this case it is the one-person one-vote approach that would fail to yield the appropriate influence for each actor, unless, due to an extensive process of discussion, the actors mediated very constructively on behalf of one another.
What is the lesson of all this for decision making?
The lesson is clear,isn’t it? Good process is always good to have, of course, though one can spend more time on communication and mutual exploration than warranted by a decision’s importance. But different decision procedures will put more or less weight on having a perfect process and will arrive at better or worse representations of the proportionate will of the actors involved more or less quickly and more or less easily. Some might achieve proportionate say almost automatically as compared to others achieving it only with great difficulty and due to very precise jockeying by each actor in light of knowledge of the others’ views and willingness to bend toward their stronger preferences. The irony is, if consensus advocates want to say that consensus is good because it forces actors to mediate their choices in accord with their mutual assessments of one another rather than merely consulting their own preferences, then they should in fact opt for one-person-one-vote majority rule, not consensus, for a decision like hiring. The second irony is, this would precisely reverse the type of logic that we think a council should employ in choosing decision procedures.
In our view the upshot is that the processes we settle on to prepare for, debate, and finally make decisions should be chosen to maximize an appropriate level of give and take, exploration, and mutual understanding, as well as appropriate influence for the importance of the decision and the time available. Communication should not be coerced by choosing a procedure that will fail miserably if communication falls short of optimal forcing people to spend more time deliberating than another procedure would require. In other words, the voting procedure used in decisions should approximate as closely as we can arrange to directly facilitating proportionate say, so that if the supporting process doesn’t work perfectly the procedure is least distorted by the communicative inadequacies.
Those who favor using consensus all the time presumably feel, instead, that we should opt for the approach that so demands good process that we must expend great effort in having good process all the time, or we will get horrible results. For that matter, the folks who advocate ubiquitous use of one-person-one-vote majority rule are presumably saying something like, let’s have a middle of the road orientation. But why should we have any single orientation at all? Sometimes one procedure is better, other times a different one is better. Why prejudge the choice universally, as compared to settling it differently, if appropriate, for each different venue?
The differences between always favoring consensus or favoring one-person-one-vote majority rule or some other option, or favoring different procedures for different situations, are not simple to see, experience shows. So let’s consider a different kind of decision, to clarify a bit more.
Let’s say we have to make choices about investment options in a workplace. We might imagine workers in a workers’ council considering a consensus approach for this type of decision but opting against it, because in application it would be cumbersome and any errors could easily lead to harmful outcomes. For investment, non consensus procedures would be easier to enact and less likely to diverge from optimal choices due to errors or bad faith by anyone involved.
For example, suppose there is a proposal to put in a new heating system. After discussion there needs to be a decision. With a consensus approach anyone can block a choice for any reason, but if you are considering doing so, how do you know whether you have the moral right, given the scale of the decision’s relative impact on you, to block it or not? In the context of the debate you have to decide yourself if it is warranted for you to veto a choice given the intensity of your feelings and those of others. With a relatively few trusting people and enough time, and with thorough information flow, consensus may be optimal, but without these features working nearly perfectly, using consensus for this type of decision is asking for trouble.
With that in mind, workers might decide it is better to prejudge that in cases of investment choices they should opt for the abstract approach that each worker gets a vote and majority rules, but also allow any strongly dissenting minority to put off a decision for further discussion, at least twice. The point is, the workers might decide that something other than consensus (which allows for individual veto) comes closer to correct apportionment of influence and for that reason leaves the actors less difficulty in choosing to moderate or to strongly express their preferences to attain proper proportionate input for all.
Now, nothing is perfect. So (to make the point graphically), suppose there is a worker who will die if the temperature goes down to 68 degrees but is fine at 70 and above. Obviously, with consensus he will have no problem manifesting his intense preference even if the mutual exchange of information is faulty. In the one-person- one-vote majority rule approach, for the decision to come out properly the debate (or perhaps overarching rules about disabilities) needs to to give that person his extra due. But the view of a group opting for majority rule for investment decisions is that the degree of sensitivity required for the chosen approach when deciding investments and the harm that errors due to poor process will most often be less than the degree of sensitivity required and the harm that would arise from errors were the algorithm for investment decisions consensus.
The point of all this is to see that decision-making procedures and communication methods are flexible and not goals in and of themselves. They are a means to the desired end of proportionate, informed, participatory, and efficient influence. It follows that we should be principled about the goals, but not about the means.
Something that emerges from this is that in all modes of decision-making, if everyone operates ideally after a full exchange of relevant information and feelings, they will reach ideal decisions. Perfect process plus perfect people plus any decision-making system at all yields perfect decisions.
Consider the case of decision-making by a single leader. The leader hears everybody, calculates all impacts and preferences perfectly, and decides the perfect outcome, incorporating into her choice each actor’s will in proportion to how they will be affected by the outcome. In a one-person-one-vote majority rule framework everyone has access to the same information and is able to freely express themselves, then modulate their vote so that the sum of all yeas and nays is appropriate. Or, of course, this same thing occurs in a consensus framework, with each person coordinating their choice to advocate or to block an outcome in light of impact on self and on others.
In other words, in any setup, if all the actors are able as a result of a free exchange of information and feelings to determine perfectly accurately their own relevant input and that of all other actors, and then in hearing the preferences of the others, if each actor decides accurately and justly whether those in the overall yea camp should carry the day and if yes maintains their yeas and if not rescinds them, all choices will come out ideally and unanimously, regardless of the voting procedure used.
In this sense, assuming our norm of self-management, in any system the abstract situation is identical. That is, those involved have to assess feelings, preferences, and information, and then decide what to do to collectively reflect every actor’s cumulative will in accord with the norm that decision-making input should be in proportion as one is impacted. In all cases, with perfect process and choice, final dissent or assent is not solely a singular decision based on one’s own feelings but depends on whether those assenting or dissenting see their joint appropriate influence level as warranting their choice. If so, they persist in it. If not they retire from it.
So is it just convention that determines which system we use for settling outcomes, the only important consideration being the process of exchange of information, feelings, and preferences, and the willingness of actors to support and respect one another’s depth of feeling and opinion in pursuit of proper proportionate influence for each?
No. Instead, in the real world it makes sense to prejudge certain types of decisions and decide that they would best be handled with certain decision-making processes, and not to rely on continually reassessing each, or, even worse, on using some fixed approach for everything. Why?
The primary reasons for preferring a flexible approach are:
1 It is desirable to come as close as one can to determining in advance how best to give each person involved in a decision appropriate impact on it, so that the need for each actor to bend their expressed vote in light not only of their own preferences but the preferences others have is minimized and the entire process is simplified. And it is also a truism that no one can know my interests as well as I do—unless I’m a child or deranged.
2 It is desirable to minimize the extent to which any actor can inappropriately distort decisions from ideal proportionate say whether whether this is due to honest mistakes, preset biases, or even dishonest manipulations.
We do not always opt for having a perfect communication process plus the smartest and most perceptive person present making the final decision unilaterally, or for a randomly chosen person doing so—and surely no consensus advocate would favor this. But why not? It involves as good a pre-vote process as we can muster. And if we say that by such processes everyone always arrives at perfect estimates of their own and all other people’s proper input, then everyone is in position to make the right decision. So why not let anyone do it? Well, we do not do that for four very good reasons.
1 It is not true that everyone is always going to accurately know everyone else’s situation perfectly, nor that they could, and obstacles can be a matter of benign lack of under- standing or less benign self-interest and bias.
2 Even if people did know everyone else’s desires and the relative impact of all options under consideration, it is not true that everyone will always behave honorably.
3 By having the ultimate decision made by one person, whether he or she is randomly chosen or otherwise, there is no record of dissent from the decision. We just have the ultimate yea or nay. We have no lasting feeling for or permanent record that we can consult of the existence of a minority and its views, and there is no tendency to empower the minority to try other alternatives or even to remember that the minority exists, should difficulties with the decision emerge down the road.
4 In practice, we know unilateral decision-making would devolve into steadily reduced participation and a divergence from real self-management.
But this rejection of one person making the final decision by fiat tells us that different approaches have different merits for different situations, which is why parecon does not prejudge how decisions should be made, but only the broad norm or goal regarding self-managing input and participation.
We like to think advocates of consensus favor it precisely because if there has to be only one method elevated above all others they are seeking the method that will at least in modest-sized groups most promote participation and permit the emergence of appropriate influence. Our response to this is that there doesn’t have to be only one approach, and there shouldn’t be.
So the bottom line, however complex the diverse cases and their specific logic can turn out to be, is simple. To facilitate and organize worker and consumer decision-making in keeping with the goal of self-management, parecon incorporates councils at diverse levels, from the smallest work team or family to the largest industry or state, and beyond. The actors involved need appropriate information and need to be properly confident, empowered, and skilled. They should utilize decision-making procedures and communication methods in their councils as they see fit, adapting these as best they can to the time and hassle involved and to the possibilities for error and abuse, and seeking to attain appropriately informed decision-making influence in pro- portion to the degree each person is affected by decision-making outcomes.
Surely it is impossible to perfectly apportion influence so there is self management as you define it – does that upset your vision?
Not at all. Self management is a social concept, arrived at by a social process. It isn’t some kind of mathematical variable that you either get to the fifth decimal place or your bridge falls or plane crashes. What is necessary in seeking to apportion influence to actors in proportion as outcomes will affect them is really just two things. 1) That all actors agree that the apportionment has been fair and sensible, reflecting available knowledge (which of course could be wrong) and taking into account the cost in time or hassle of trying to achieve too great precision. And 2) that whatever deviations due occur whether by error or due to settling rather than expending excess energy for minimal gains, are random rather than one-sided. They should not accumulate, in other words, but rather average out over time.
Once one understands that what self management means is correct apportionment in the eyes of those involved and deviations that don’t snowball into lasting injustices, the mystery and complexity dissipates quickly.
In most workplaces it will mean, for example that there are different categories of decisions – all known and categorized in advance by the council, perhaps revisited once a year. For some there will be majority rules, for others consensus, and still others might have a different voting pattern. Some will involve all workers, some will involve subsets of the whole though often operating within broad policy guidelines.
It is true that if a group, whether in a workplace or a community, were to fetishize precision and argue over every tiny detail, it would waste inordinate time and likely achieve little greater accuracy, as well. But why would anyone bother Much easier to agree on broad norms regarding information transfer, transparency, deliberation, and tallying of prefernces for different kinds of decisions, then acting in accord with each as necessary.