Sortition for worker-owned firms: a conversation with Simon Pek

The video above is a recording of a conversation I recently had with Simon Pek.

Simon is a professor at the University of Victoria, Canada. He has recently published a paper called “Drawing out democracy: The role of sortition in preventing and overcoming organizational degeneration in worker-owned firms” in the Journal of Management Inquiry. In the conversation we discuss this article, the notion of “organizational degeneration” which is central to the paper, as well as how and why Simon became interested in sortition, and related matters.

15 Responses

  1. I think its not a good idea to choose executives bij lot. We beter can choose them in an assesment procedure, because then we can judge them on specific qualifications, like coordinative skills.

    Like

  2. Sortition can play a vital role in maintaining democracy in co-ops and large worker-owned enterprises (for example, imagine if Uber were owned by its drivers or users, how could they function democratically?).

    However randomly selecting a tiny body like a board of directors or worse, a chief executive, is nonsense. However, a much larger, short duration random member assembly could recruit, hire and fire a chief executive, or nominate a slate for a board of directors. Allowing a current board to appoint a nominating committee for future board members is a recipe for corruption or capture.

    Here is a brief article I wrote on an appropriate use of sortition in co-ops.
    https://blog.p2pfoundation.net/a-better-co-op-democracy-without-elections/2017/04/19

    Liked by 1 person

  3. Terry,

    > However randomly selecting a tiny body like a board of directors or worse, a chief executive, is nonsense

    IIUC, this is based on the idea that such a body – a single decision maker – is “obviously” too small to be representative. But why? What makes a body “too small”? And if such a body is too small, why would a professional body or decision maker be any better?

    In general, the discussions of body size tend to be unreasoned – leading to the common use of “1,000” a some sort of big-enough magical number (or as you put things here “much larger”). What makes a body too small? What makes a body “big enough”? (Indeed, it should be clear that there should also be a notion of “too big” – otherwise, no sampling, and no representation would be needed at all.)

    > However, a much larger, short duration random member assembly could recruit, hire and fire a chief executive, or nominate a slate for a board of directors.

    Such an arrangement is obviously anti-democratic. At best, the short term body, just like voters, would be able to make crude “overall” estimates of the quality of governance and hire and fire accordingly. This would in no way translate into a democracy, where all the details of policy are supposed to serve the interests and values of the members. In a democratic system, any professional body must be monitored closely on an ongoing basis by an allotted body.

    Like

  4. Depending on the tasks at hand, and what traits might be beneficial, numbers can matter a whole lot. For some things (such as adopting laws) representativeness matters a lot. But for some tasks it is the impartiality (the “blind break”) of sortition that is more crucial, and small groups might be fine. An assertion of democracy is that civic judgment is broadly distributed through the population and the principle of political equality means that a lottery is a good procedure. However when representativeness matters (the body is making the sort of decision the population as a whole would make if they had the same information and time), selecting too small a body increases the risk that by random chance a very UNrepresentative sample will be drawn. This may be okay if the decision is fixable (reversible) by a subsequent jury due to the regression to the mean, but for other decisions the body needs to get it right the first time (so a larger sample is needed).

    A single person or small body that has some executive functions, needs an additional set of traits. Just as the Classical Athenians did not select their generals by lot (and Aristotle noted that the pilot of a ship should not be selected by lot). The passengers or a random sample of passengers should decide WHERE they want to sail to, and could select a pilot… but selecting the pilot by lot is not a requirement of democracy.

    Like

  5. Every citizen has the quality of political judgement and must have the right to legislate. But executive tasks must be done by special experts. The classical Athenians were indeed already aware of that fact and choosed their officials. I think that we, in our modern society, must make a first selection of candidates by a special board and secondly an election between them by an allotted group.

    Like

  6. Election shares out a minuscule amount of power to everyone, but limits the opportunity for substantive power to a small political class and excludes everyone else. Sortition don’t offer any power at all to the majority, but gives everyone an equal chance at wielding substantive power.

    It seems to me that using a randomly selected body to choose executives is the worst of both worlds. It creates a leadership class from which most people of excluded, but it also denies the majority any say at all in the matter, not even a vote.

    Now, all this is slightly orthogonal to Pek’s paper. He doesn’t mention executives at all. He refers to worker representatives in “bodies such as boards of directors and councils”. These bodies can offer some statistical representation, even if not tremendously accurate.

    But on the matter of executives, there’s an important point here: A technique for preventing organisational degeneration that Pek doesn’t explicitly mention is decentralisation: Devolve decision making to smaller functional units which look after their own affairs wherever possible and co-ordinate with others wherever necessary. This lets you scale up an organisation without meetings taking over everything. It’s also democratic in that the workers retain a lot of power over their workplace and their work.

    You can contrast decentralisation with bureaucracy. In bureaucracy, power emanates from a single chief executive and flows down through a pyramid of managers, being elaborated at each step until it is applied to workers who are little more than mindless instruments. Here’s the idea that, as Ronald says, “executive tasks must be done by special experts”.

    I suspect that simply appending any democratic representatives, whether elected or randomly chosen, to the top of a bureaucracy, is going to lead to organisational degeneration.

    On the other hand, coordination becomes an issue as you try and scale up decentralised organisations. So decentralisation and sortition seem complementary: A randomly selected council can get a coherent view of the whole organisation and help co-ordinate things, and more autonomous elements balance out any statistical inaccuracies you might get from having a small council.

    Like

  7. Liam,

    > Sortition don’t offer any power at all to the majority, but gives everyone an equal chance at wielding substantive power.

    This way of conceiving things focuses on formal decision making power and rather than on the substance of power-wielding and is therefore not useful, in my opinion. The reason sortition is a good democratic tool is not because it “gives everyone an equal chance at wielding substantive power” – that is at best merely a device. The reason sortition is a good democratic tool is because it produces (or can be expected to produce) outcomes that serve the public interest (i.e., that would meet with widespread approval).

    Liked by 1 person

  8. Liam,

    > decentralisation: Devolve decision making to smaller functional units which look after their own affairs wherever possible and co-ordinate with others wherever necessary

    It seems to me that decentralization is one of those standard tools in the Anarchist toolset that have a superficial intuitive appeal (“I don’t want those politicians in Washington to be making decisions about my community”), but do not stand scrutiny. In fact, this idea is quite similar to the classic Liberal idea of individualism, which is associated with opposite side of the Left-Right political spectrum. According to the individualist idea, society is made of free-acting individuals and government’s role is “merely” to regulate interactions between those individuals. In the localist ideology, society is made of free-acting small scale units, and government merely regulates the interactions between those units. Neither of those ideas are in fact democratic.

    Like

  9. Yoram,

    >”This way of conceiving things focuses on formal decision making power and rather than on the substance of power-wielding”

    With sortition, the majority of people have neither substantive nor formal decision making power. Things they approve of may well happen, but that’s because people similar to them are making decisions, not because they themselves are doing so. Whether or not that’s a problem is a complex and important issue, but not relevant to the point at hand.

    Obviously sortition does many things beyond those I mentioned. Obviously it can’t be fully described in a single sentence. But I didn’t need to describe it fully. I only needed those particular features to criticise the idea of using a randomly selected body to elect executives.

    >”decentralization is one of those standard tools in the Anarchist toolset”

    I’m not sure I understand the point of this comment. Is it meant to be a criticism? Because though it contains several accusations, I don’t see any real arguments there.

    You say that decentralisation does not stand scrutiny, but fail to scrutinise it.

    You say that is has a superficial appeal, but your accusation consists of nothing but two superficial summaries of different ideologies, each of which — only superficially — resemble each other, and a superficial mention of the left-right spectrum.

    You say decentralisation is an anarchist idea (since we’re on a roll here, a superficial observation), but then describe it as a government regulating interactions between units.

    You talk about governments regulating interactions on society, even though the entire context here isn’t about government and society, but the management of organisations in general and worker owned firms in particular.

    Like

  10. Liam,

    (I hope I am wrong, but if I am reading your response correctly, it seems that the discussion has taken an acrimonious turn. That was certainly not my intention – I was, and still am, aiming to make substantive points.)

    > With sortition […] Things [the majority of people] approve of may well happen

    That is the substance of having power. Voting and allotment mechanisms are, at best, tools for attaining such substantive power.

    >> decentralization

    > I don’t see any real arguments there.

    My argument was that the arguments that are presented for decentralization are the same ones that are presented for Liberal individualism (“the night watchman state”). Yet the two ideologies are contradictory. Thus, the arguments presented are obviously lacking a way to tell which of the two (if any) is valid.

    Like

  11. Yoram,

    Thank you that aside. It’s very magnanimous of you.

    I think we may share the same trait of, when in a debate, bypassing the diplomacy and going straight for the logic and arguments. That approach has some benefits, but it can lead to misunderstanding. My response was a tiny bit grumpy. So thank you for clearing that up. Be assured that I have no ill will towards you.

    > “That is the substance of having power.”
    I can’t agree with that. It seems at odds with our intuitive understanding of individual agency.

    If something happens and I like it, that isn’t necessarily agency. If I want something and it happens, that isn’t necessarily agency either. Only if I want something, take actions to make it happen, and it happens, do I have agency.

    Is that a point against the democratic potential of sortition? I don’t think so. (Though this is a complex and subtle area and I’m still working things out. I’ll have to get back to you on this matter when I’ve thought a bit more.) Democracy in general is insensitive to individual agency. It only reads collective preferences. Sortition can do that very well.

    > “Yet the two ideologies are contradictory. Thus, the arguments presented are obviously lacking a way to tell which of the two (if any) is valid.”

    If an argument implied two strictly contradictory conclusions, then it would indeed be a bad argument. But that’s not the case here. The issue is rather more complex because, (1) there are many different arguments for decentralisation, (2) there are many ways of implementing decentralisation, (3) there are different contexts in which one could implement decentralisation, (4) two ideologies being on different ends of the political spectrum is not strictly a logical contradiction, because the political spectrum communicates very little detail. (All of this, incidentally, applies to sortition too.)

    In addition, though advocates of liberal individualism sometimes say it leads to more decentralisation, that doesn’t happen in practice. The most recent turn towards it, neoliberalism, has led to much greater centralisation of power in a few giant corporations. Decentralised power structures are nowhere built into liberal individualism.

    Like

  12. Liam,

    > I think we may share the same trait of, when in a debate, bypassing the diplomacy and going straight for the logic and arguments.

    Yes – thanks. That sums things up nicely. That seems like the proper practice, as long as we maintain an a-priori assumption of good faith, and make sure we our discourse justifies this assumption.

    >> “That is the substance of having power.”
    > I can’t agree with that. It seems at odds with our intuitive understanding of individual agency.

    We can go into this in more detail if you want – I find this topic interesting. However, I am not sure we need to because, whether we refer to this as “power” or not, it seems to me that getting the outcomes you want is really what matters in government. So if we have a system that consistently provides good outcomes, I would call that good government and aim for that.

    Regarding decentralization:

    Very abstractly, a democratic society must reflect in its basic governance structure (as in all other aspects) the values and interests of its members, weighted equally. I don’t see any a-priori reason to believe that decentralization (which, as you write, as a rather vague concept) is “a good thing” in a sense that it would tend to promote this reflection of popular values and interests. It seems like a constraint on what systems we consider, and this constraint may turn out to rule out good arrangements. If someone tells me, for example, that instituting nation-wide pollution rules is undesirable because it is a form of centralization, would that be true or false? Why does it matter? If the rules serve the public as the public sees things, then it is a good thing – decentralization just does not seem like a useful generic objective.

    In practice, don’t think there is good reason to think our society is over-centralized, or that centralization is a major source of our system’s oligarchical outcomes.

    (Maybe this would be useful: I had a similar discussion back in 2008.)

    Like

  13. Discussions in comment sections tend to jump around a bit much… but returning to worker-owned enterprises or co-ops, and my suggestion for using a large jury to select a board of directors or a chief executive, Liam wrote
    >”It seems to me that using a randomly selected body to choose executives is the worst of both worlds. It creates a leadership class from which most people of excluded, but it also denies the majority any say at all in the matter, not even a vote.”

    Yet, I persist in thinking this is the best and most democratic way to govern a large organization (and many organizations and members would suffer if forced to decentralize). This is the only method I can imagine that can avoid Michels’s “iron law of oligarchy,” and expressly PREVENTS the development of a “leadership class.”

    Take my Uber example (if owned by its drivers). How does one conduct an informed election among countless part-time workers who don’t know each other? The oligarchic nature of governmental elections and rational ignorance would only be magnified when members have even less investment in the outcome than in government elections. The typical co-op or large worker-owned enterprise pr4ocedure is to let the existing board constitute a nominating committee to select their own replacements (or renew their own terms), with the “election” being a mere formality. This same elite body picks an executive. Instead, a representative sample of members given the time, resources and incentive to deeply explore options and recruit an optimal mix of new board members, or an ideal chief executive, avoids oligarchic capture, and protects democratic equality of members.

    Like

  14. Apologies for taking so long to get back to this.

    Yoram,

    On the matter of agency and power, in principle we don’t really need to get into it at all, if the goal is argue for sortition. We’re all in favour of that. But to the degree that we want to explain and justify sortition as a sort of democracy — either for our own philosophical peace of mind or to convince the undecided — your approach has little in in favour. It requires we (a) redefine crucial words like agency and power in strange ways, (b) largely ignore objective properties of sortition and the causal processes involved even when they look egalitarian, (c) focus solely on outcomes rather than processes, even though the outcomes haven’t been empirically verified, “good outcomes” is ill-defined, and outcome-based arguments could easily be (and often are) used to justify attacks on democratic principles.

    I can understand why you’ve taken this approach. Much of sortition theory isn’t very good. There are a lot of muddled analogies between statistical representation and electoral representation, confusions about accountability, and the like. But we can do better, I think, without having to focus solely on ill-defined and unverified “good outcomes” and without having to redefine useful notions of agency and power. How, exactly? I’m still working on that. When I’ve got something clearer, I’d be happy to share it with you.

    Regarding decentralisation:

    Interestingly, when you write “a democratic society must reflect in its basic governance structure (as in all other aspects) the values and interests of its members, weighted equally”, you’re contradicting your earlier point about focusing solely on outcomes. That’s a point of principle about the structure of a society.

    That aside, as a thought experiment, consider your argument, but substituting “sortition” for “decentralisation”. There too, there is (if misinterpreted) a constraint on the systems we can consider which might rule out other arrangements. Any idea, if considered rigidly and abstracted from the real world, leads to absurd examples.

    On the other hand, if we treat the matter sensibly, which is to say we think of these ideas as options and possibilities to be evaluated in given situations rather than arbitrary demands, the worry about constraints never arises.

    Given all that, you’ve failed to offer any criticisms of my use of decentralisation. You’ve made up an imaginary interlocutor who holds silly ideas about pollution, but that person is not me.

    That’s on your failure to offer any points against decentralisation. What about my responsibility to offer points in favour? You repeat yourself three times in your reply that you don’t see any points in favour. There are many good arguments in favour of decentralisation, both conceptual and empirical, which I haven’t mentioned. That’s because I never intended to mount a full defence of decentralisation in general – and because you have so far entirely ignored the one argument I did use. So I will repeat the point:

    A decentralised organisation has less need for hierarchical executives. Hierarchical executives, being a minority who are able to order about the rank-and-file, are a source of authority; and being individuals with specific roles, aren’t suitable for selection by lot and remain a source of authority even if answering to a randomly selected council. Therefore, decentralisation *in this context* helps remove an authoritarian structure that sortition can’t.

    Terry,

    I think you’ve got your finger on the scales with your choice examples. Uber is very much an outlier in terms of corporate structure. Indeed, it was engineered ot be so – by making its employees look like the self-employed, it could get away with exploitive labour practices. Things that work for the majority of worker owned businsses won’t necessarily work for Uber and vice versa.

    The typical co-op structure you give may well be exactly that. But if the existing solution is a poor one in many ways, it’s a low barrier for your proposal to clear. If you want to show the benefits compared to a general class of solutions (like elections), you should compare your proposal to the best possible example of that class you can come up with. There’s some scope for disagreement about what the best example might be, of course, but it certainly doesn’t include letting the current board have power to select their own replacements.

    As for your proposal, you say it will PREVENT – in caps no less – the development of a leadership class. But I can’t see your reasoning. I can see how a randomly selected board would do that, because anyone could end up on it. But your proposal is, literally, an election. The rank-and-file workers choose, or elect, their leaders. The only difference is that the electorate is a statistical sample.

    In that case, how will you prevent the same problems that afflict every system of election? The fact that some people will have skills that will make the more suitable in the eyes of the electorate? That the structure of a competition favours those who have the resources and confidence to compete? That irrelevant traits like networking, charisma, popularity, beauty, egoism, and position in the social hierarchy will influence the mini-electorate? That a competitive frame will encourage more aggressive behaviour in the candidates? That those who win the election, having power, could use that power to favour people like themselves in the next election?

    The closest thing I can see as a defence that your election system won’t generate an elite is an invocation of rational ignorance. I have some doubts about the power of this principle, given that it’s based on an empirically discredited model of human psychology. But I don’t need to make a stand there, because we know that hiring committees regularly make biased decisions, despite being much smaller (I assume) than your random electorate. And that’s not all, because even if we were to remove all possibility of bias from the electorate, that would do nothing to fix a selection effect on the applicants.

    So it’s not at all clear how this form of election would overcome the problems with all others forms of election.

    Like

  15. Liam,

    I won’t take a deep dive into the downsides of decentralization (I assume you concede there are both advantages AND disadvantages). I hope you agree that SOME important functions are better done by larger organizations (for economies of scale, societal efficiency and coordination reasons, etc.).

    But turning to my point about using a randomly selected recruiting body… you stated that “The only difference is that the electorate is a statistical sample.” No. It is totally different. Even if you reject the economics cost/benefit basis for “rational ignorance,” it is an indisputable fact that most people put less time and effort into researching things when there “vote” will be swamped by a vast sea of other votes. However, a representative sample of the membership can be paid to spend the time and effort to dig deep, and become well-informed. I don’t propose they simply vote for candidates, but rather recruit members to serve on a board or to serve as executive director. Indeed, I agree with Douglas Adams that wanting an office where one gets to have authority over others should be a disqualification from getting that office. So, no candidates and no competitive campaigns. The organization does not need to be vast to benefit from a jury recruitment process. If I am a member of a credit union (and don’t know hardly any other members) a large food co-op, a large worker-owned business, etc., I am simply not going to know much at all (nor spend the time to learn) about any candidates running in any election for any of these entities. If I was selected in a lottery and given time, resources and paid to take responsibility on behalf of all members, I would spend the time, and so would most people. Even the BEST elections of larger member-owned entities are either token (an elite nominating committee selected by the current board makes the real decision), or bitter and divisive that tear the organization apart.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: