本文探讨了工人合作社的运作模式及其相较于传统企业的优势。与传统企业老板说了算不同，工人合作社由工人共同拥有和民主管理，这被认为能提高工人积极性和生产效率。文章反驳了“搭便车”假说，指出重复互动和人性中的合作倾向能够解决激励问题。同时，合作社能更好地利用“地方知识”，克服传统企业信息不对称的局限。研究表明，工人合作社在实践中也确实表现出更高的生产力，并且其发展受限于信息传播、立法以及资本回报率等因素。尽管存在挑战，但合作社模式的长期潜力不容忽视，尤其是在结合更先进的治理机制后。

📈 **工人合作社的民主管理模式提升效率**：与传统企业不同，工人合作社由工人共同拥有和民主管理，这意味着工人拥有决策权，而非由单一所有者决定。这种模式被认为能激励工人更努力地工作，从而提高生产效率，因为他们能分享企业利润并对企业运营有发言权。

🤝 **克服“搭便车”假说与信息不对称**：文章指出，虽然理论上存在工人因利润分享而懈怠的“搭便车”问题，但现实中的重复互动和人类的合作本性可以有效解决。此外，合作社能够更好地利用“地方知识”，克服传统企业信息不对称的弊端，因为基层工人更了解具体情况，能为企业决策提供更准确的信息。

🚀 **合作社的实际效益与发展瓶颈**：研究表明，高程度的工人持股能够促使工人互相监督，从而提高生产力。合作社的工人可以通过投票来指导公司发展方向，确保组织目标与自身价值观一致。尽管如此，合作社的发展也面临信息传播缓慢、立法限制以及资本回报率问题，因为资本家可能更青睐回报率更高的传统企业，即使其生产效率较低。

💡 **合作社的未来潜力与改进空间**：文章最后提到，虽然合作社的发展速度不如传统企业，但其数量正在快速增长。作者认为，合作社可以通过采用先进的投票理论和治理机制来进一步提升内部协作和效率，尽管这需要时间来追赶传统企业在设计上的先发优势。

Published on August 9, 2025 2:47 PM GMT

Crossposted from my blog

 

Worker cooperatives are firms that, unlike traditional firms, are run democratically. This means that instead of the owner of the firm deciding who manages the workers, the workers become part owner and get a say in how the firm is run. This has some advantages, such as workers working harder and productivity appearing to increase.
 

Wait… how’s that even possible in theory?

You might be asking yourself: won't workers become lazy since the profit is shared with their colleagues, which means they only get a small proportion of the fruits of their individual labor? According to the free-rider hypothesis, rational and self-interested agents will always have an incentive to put in less effort and be a parasite to the efforts of others. That’s literally the first lesson of game-theory 101.

Yes, but this is solved by the second lesson of game-theory 101: repeated interactions. Once you can build up relationships with people, you get a chance to punish free-riders while rewarding the helpful. Not to mention, we’re not the self-interested rational agents of game-theory 101. We’re complex, we care about cooperation and kindness and all that other icky stuff.

It’s not just the behavior of humans that’s wrong in the game-theory 101 model. These simple economic models assume perfect information which, in reality-land, managers of traditional firms do not have access to. According to economist Friedrich Hayek, top-down organizers have difficulty harnessing and coordinating around local knowledge, and the policies they write that are the same across a wide range of circumstances don't account for the "particular circumstances of time and place".

One study, that looked at real firms with high levels of worker ownership, found that workers there start checking each other’s work more, which makes those workers more productive than those with low- or no worker ownership. This knowledge of what everyone is doing is also not going to waste, since the workers can steer the direction of the company bottom-up through their votes.

Those who make the top-down policies in a traditional company are different from those who have to follow them. In addition, those who manage the company are most often different from those who own the company. These groups have different incentives and accumulate different knowledge. This means that coops have two main advantages:

Workers can harness their collective knowledge to make running the firm more effective.Workers can use their voting power to ensure the organization is more aligned with their values.

   
…But also… these game-theoretic arguments strike me as putting the cart before the horse. We observe that people work harder in coops. I agree that the math of the economics-101 model is very pretty, but that shouldn’t be a reason to disregard our observations. If the model contradicts the data, we throw out the model.

Could this be selection bias?

Could it be that workers who are more motivated to work harder for a coop, are also more likely to start or join a coop? This would mean that we can’t use them as a measurement of how ‘normal people’ would perform in a coop.

If we only had observational studies this would indeed be a concern, but we’ve also tested this experimentally and found the same increase.

Okay, but what about firm productivity? You can’t easily test that in a lab, and it’s possible that while the workers are more productive, the firm itself isn’t.

That would be strange, but yes, it’s possible. Studies show that the cooperative firms themselves are also more productive, but these studies are indeed not done in a lab so could be subject to selection bias. Plus the effects are small and there are some studies that don’t show this increase. However, even if we revert to the null-hypothesis that they’re about equally as productive as traditional firms, the question would remain… where are they?

Why aren’t there more worker coops?

Here’s an old economics joke you might have heard before: Two economists are walking along the sidewalk. One sees a five-dollar bill and bows to pick it up. The other says, “Don’t bother, if that was a real five dollar bill someone else would’ve picked it up already”.

This joke is making fun of the assumption of perfect information. Unlike in the simplistic models, in the real world, market participants do not have all the necessary information. If you were to ask a random American what a coop was and how it worked, do you think they’d be able to answer that accurately? It takes time for innovations to spread. And we do observe that the number of coops are increasing rapidly, e.g., the number of worker coops in the US has tripled in the last decade.

So maybe a better question is: why did it take so long?

It could be that legislation constrained coops. Or maybe it’s because traditional firms have had more time to improve. There are, according to my research, approximately 4.7 bazillion books on how to run and manage your firm effectively while there’s less than 0.001 bazillion books on how to do the same for coops.^[1]
It could also simply be good old-fashioned propaganda that made people wary of collectivization.

I think all of these play a role, but I also think there’s another, more straightforward answer.

Return on investment

Capitalists tend to not like coops. You might object that capitalists only care about efficiency, so they would embrace coops if they were indeed more productive, but that’s not actually the case. Capitalists care about profits, which is not always aligned with productivity.

Say there’s a startup coop that can turn $10 of investment into $100, but will split that $50 with the workers and $50 with investors. Meanwhile there’s also a traditional capitalist startup that can turn $10 into $90 which goes purely to the investors. In this case the coop is more productive, but the capitalist firm gives the capitalists a higher return on investment, meaning the investors will invest in the capitalist startup even though it's less productive.
On top of that capitalists have less power over coops. Since workers have more protections/decision-power, capitalists aren’t micro-kings and have to actually work together with the workers, which is annoying.

The return on investment for society may be larger (especially if we take into account all the costs traditional firms externalize to society at large that cooperative firms internalize) but that doesn’t mean they have a higher return on investment for investors.

Studies show that capitalist firms do better with more capitalist firms around, while coops do better with more coops around. So we have a bit of a lock-in effect. It may be that to get to a place of greater long-term productivity we first need to give coop firms some time to find their footing.
 

1. ^{^}

   I know nothing about running a firm, but even I can see areas where coops could readily improve: e.g., coops are currently not using the state-of-the-art in voting theory. With better voting mechanisms the cooperation/coordination within the firm could drastically improve.
   It’ll probably take some time for coops to catch-up to this head-start in firm design that the capitalist firms have.

Discuss