In a socialist society, should workers have democratic control over their own workplaces? For many socialists, the answer is obvious: of course, almost by definition. But among economists who have developed proposals for how a socialist system might work, the answer is much more controversial. Some of the most well-known models do not involve democratic workplaces.

Workplace democracy is not a feature of one major vision published in Catalyst, John Roemer’s “sharing economy,” though workers share in firm profits.¹ In his seminal earlier proposal, A Future for Socialism, Roemer explicitly rejected labor management, at least as part of socialism’s “first step,” and possibly even as part of an ideal vision.² This is in line with the famous neoclassical socialist model of Oskar Lange, which also features labor markets and conventional employment relations.³

Workplace democracy is not a module that can be plugged into a socialist model without affecting the rest of it. It is tangled up with questions of finance and labor mobility. It brings challenges around risk, income distribution, and efficiency. Activity at any workplace must be coordinated within the broader division of labor across the whole economy. Roemer’s agnostic position is understandable:

My preference for the managerial proposals is based on conservatism, namely, that it is best to change features one at a time, if possible. The biological metaphor is apt: an organism with one mutation is more likely to survive than one in which two mutations occur simultaneously. I think it is more important to change the private nature of the financing of firms than the management structure as the first step.⁴

For many of us, however, workplace democracy is not to be given up lightly. It is central to the other proposal for “Socialism for Realists” published in Catalyst, by Sam Gindin.⁵ Gindin presents an appealing model, and the present essay is broadly in line with his suggestions. But he does not respond there to the problems identified by skeptics of workplace democracy — sympathetic or otherwise.

Those problems are real. Their root is the fact that modern standards of living depend on a vast and complex division of labor, in which the individual workplace is a small node. These nodes must be coordinated to meet social needs efficiently, but if each workplace must play a part, what is the scope for local democratic decision-making? Any system must constrain the choices made within the workplace, and structure those choices with incentives to bring local interests in line with broader social needs. That is easier said than done. The critics have rightly pointed out serious flaws in some of the standard models, and this has convinced many that worker management is simply unviable.⁶

The aim of this essay is to face the problems squarely and suggest a set of institutions that can plausibly cope with them — a model that is hopefully appealing, meeting socialist aspirations for equality, democracy, sustainability, care, and personal freedom. The problems cannot be resolved simply at the level of the workplace — they are systemic problems and so require the whole system to be built around them. In particular, workplace democracy has fundamental implications for finance, the labor market, and the relationship between the public sector and the sector of commodity-producing, democratic firms. This essay therefore gives a lot of room to the broader systemic architecture before focusing on the democratic firm itself.

The proposed system includes a large market sector of commodity-producing firms, as in both Gindin’s and Roemer’s models, as well as other prominent socialist proposals of recent decades.⁷ As Gindin argued, planning and markets are not opposites: the grid of prices and incentives emerging from and organizing market activity is a vital source of information and tools for planners. But having accepted a substantial role for markets, socialists must resolve their tensions with egalitarianism and democracy. Roemer’s proposals concentrate on the egalitarian side and reserve democracy to the state, and so they have a more limited set of tensions to resolve.

The next section discusses the entangled relationship between the commodity-producing private sector and the noncommodified public sector. It also explains the ways in which a democratic government can engage in effective planning and policy in and through the market sector. In addition to the public agencies and policy tools familiar from capitalism, the state will have two powerful sites for intervention: a public banking system, and a labor board setting benchmark wages and conditions.

There are many possibilities to expand public agency in new ways, as well as facilitating new kinds of community provisioning. But my attention in the rest of the essay is on the special problems of reconciling efficiency and democracy in socialist commodity production. Efficiency is sometimes seen as a conservative goal, but it simply means avoiding the waste of resources, including people’s time and effort. It need not imply the pursuit of maximum output; efficiency is just as important in maximizing leisure and minimizing environmental impacts for a given output. Waste and misallocation played a big part in the problems of state socialism in the twentieth century, and the social sustainability of any future socialism depends on a reasonable degree of efficiency.

I explain the core problems raised by economists about collective firms: the so-called horizon and common property problems, and related problems of risk management, that can keep democratic firms from investing and employing members in a way that efficiently meets social needs. In response, I propose treating investment as a partnership between public banks and democratic firms, allowing for systemically rational investment decisions and an appropriate spreading of risks and returns.

There is something to be said for Roemer’s caution in presenting a minimum viable socialism. But there is also something to be said for proposals that push further down the road to utopia, while trying not to actually be utopian by ignoring predictable problems. By wrestling with an expanded set of problems, they build the plausibility of an attractive and viable socialist alternative. We can never, of course, know if we have correctly anticipated all the problems a real program would run into — there will always be “unknown unknowns.” But by giving at least possible answers to the “known unknowns,” we make an appealing socialist alternative seem more conceivable. The point of sketching models, as Gindin stresses, is not to write recipes for the cookshops of the future — nothing would force them to use our recipes if they no longer suit the tastes or solve the problems of the time. It is to convince people in the present that there are workable recipes that make socialism worth pursuing.⁸