The role of capitalists in capitalism

Many people do not understand what capitalists do within capitalism. Some suggest they ought to be removed, as they are just parasites who tax the productive elements of society (e.g. workers)


It is said, for example, that capital income (rents, interest, dividends, capital gains) does not come from producing anything, but from owning. It’s not from matching capital with labor and such, because it’s entrepreneurs who do that. To a point, it would be workers (CEOs, HRs, etc) too.

Elon Musk, or Amancio Ortega, or Jeff Bezos are both workers and capitalists. As workers, what they do is to manage the companies they also partially or totally own as capitalists. It is easy to see how an assembly line worker, working hard for hours, adds value to the product. It is somewhat less easy to see how an engineer, looking at a screen, scribbling unreadable things in a piece of paper, and producing models in 3D to be later used in manufacturing, adds value. The engineer is not sweating, after all. Yet the engineer earns more than the assembly line worker. The reason is that the engineer is more productive. Or technically, that the marginal productivity of engineers is higher than that of assembly line workers, in part due to there being more of the latter than of the former.

CEOs appear to have expensive lunches with other highly paid workers, sit around doing nothing, and ordering people around. They don’t even need to do the research to take decisions: they have other people to do that for them. A CEO can be sometimes seen as someone who pushes Yes/No buttons in response to corporate needs. The thing is, that seemingly easy task is incredibly complex. A wrong answer can cost millions, or even bankrupt the company. A mistake an assembly line worker makes can be, usually, easily fixed, and there controls in the factory to ensure quality. Quality is relatively easy to measure against well defined standards. But if a CEO approves a decision to build a new factory in India instead of in Iran, or if she decides to increase dividends rather than investment, there is no one to say she was wrong. Those mistakes can take years to make themselves evident.

It is sometimes said that there exists real uncertainty (e.g. Knightian) in markets. One key human skill that will take time to be replicated by machines is to aggregate uncertain information to craft productive plans that are able to succeed even when facing such an uncertain world. Accordingly, markets reward CEOs in proportion to their massive importance.

You could, in theory, run a company without workers, purely using machinery. But you can’t escape the fact that someone has to decide what to do with that equipment. What to produce? When? How? How to price it? Those decisions are orders of magnitude more important than soldering, riveting, programming, or assembling anything. People who think otherwise are reasoning from either wrong intuitive economics, or from wrong economic theories of value.

But CEOs are workers, and this post is about capitalists.


So what do capitalists do?º

According to some people, they do not produce anything. They just happen to own stuff, and they get money because that stuff is scarce. It is said that capital income is unearned plunder, just from owning things, not adding value to anything, just taxing the economy. If this is so, an economy without capitalists would, perhaps, be better than an economy with them.

Let’s then assume that capitalists are removed from the economy. Now, a corporation that was previously controlled by some thousand tiny capitalists that each owned a few shares and some few big ones is left uncontrolled by them. A couple of things can happen now:

  1. The CEO keeps acting as CEO, and becomes the new de-facto capitalist
  2. The company becomes a cooperative, and profits are distributed to workers by some mechanism. Now, workers become the new capitalists.
  3. Some society-wide entity receives the profits and does whatever some elected politician wants it do with it. This makes everyone a surrogate capitalist (a public bureau would be managing the money and profits). This entity would get a % of earnings from each corporation. The percentage would be set in such a way as to make sure whatever salaries are set by the company are adequate. In practise, the agency would have to calculate how much income generated by a corporation comes from labour, how much comes from entrepreneurial discovery, and how much comes from capital.

Option 3 is plagued with problems, and it supposes the end of capitalism, as there ceases to be a market for capital, and all the means of production become centrally owned. If you read economic history books about the Soviet Union, you know what to expect. We’ll explain the problems of this in more detail later on.

Option 1 is not what dislikers of capitalism have in mind, as it seemingly just keeps capitalists around, just reducing their number, and thus concentrating their power.

Option 2 is what they usually think of. But see that here you are not abolishing capitalism or capitalists. You are just forcing the capitalists of company X to be the workers of company X. You could still say that if, say, Apple is made into a cooperative, Apple’s workers are taking advantage of their position as owners of Apple, as they keep the profits to themselves rather than giving them away. They would be getting extra profits ‘just because’ they own Apple, in addition to the wages they make as workers.


Let’s look now at how worker-owned corporations work. Some examples are a few banks (like Crédit Agricole), Mondragon, John Lewis, The Cooperative Group. The largest right now is the relatively unknown Indian Farmers Fertiliser Cooperative Ltd.

Seemingly, then, there is nothing stopping worker-owned companies from spreading.

A brief literature review (Pencavel, 2012, Pérotin 2016,) highlights some problems for cooperatives, among which there is the ‘degeneration’ argument: coops will become capitalist enterprises. The original workers will just hire non-owner workers and in the long run, they will be able to stop working, living from profits alone. In Mondragon, for example, 10% of workers were non-owners in 1990. In 2007, 70% are in that situation. It is rare for a co-op to employ only and exclusively owner-workers.

Other argument is that equal division of revenues may provide weak work incentives. Managers can only stay as managers as long as workers approve of them, which may generate the short-term problems present in modern-day politics, and the company may increase its aggregate time preference (i.e. increase wages at the expense of investment). It is also said that co-ops will be slower to adopt innovations that may save labour, in an effort to keep every member employed.

Another argument is that if the firm is managed by a direct-democratic process, it may generate perverse incentives, similar to the ones described above. Whether this will be superior to a regular CEO-managed and shareholder-owned corporation will depend on the relative talent of CEOs and workers in managing companies. The democratic process may also introduce conflicts between heterogeneous preferences that are less salient in regular companies: the CEO has an overarching goal of maximising profits for shareholders, and profits are easy to measure. This goal is compatible with others (e.g. Corporate Social Responsability), nonetheless.

There are also arguments for worker ownership: If workers own the machines and equipment they use, they may care better for it, and perhaps there will be less need for supervision relative to a capitalist company.

Another argument is that profits will accrue to workers, so their effort becomes higher compensation in a more direct way, which is always a good incentive.

Some facts from the empirical literature:

  • Cooperatives tend to be larger than regular companies (greater median and average number of employees). This is partly because cooperatives are born larger than other companies (many companies start with just 1 or 2 employees, while cooperatives begin with 4 on average)
  • There are no obvious sectors where cooperatives are better than regular companies, or viceversa.
  • Cooperatives survive as long or longer than other firms, controlling for industry, starting wage, size, and year of creation.
  • Capital intensity (fixed assets per employee) is lower than in conventional firms, but there are some studies that show that this is not always the case.
  • Worker cooperatives are more productive, on average
  • Cooperatives tend to adjust pay, rather than employment, in response to demand shocks. That is, they hire and lay off less.

Some suggest (Schwartz, 2011)  that the reason why cooperatives are not the predominant form of firm in the market is not because they are worse, but because their formation suffers from a collective action problem: individual workers do not know about worker cooperatives (they are rare) and how to set them up, and do not have the resources to start them on their own. Capitalists could be seen as market entrepreneurs who profit from solving the problem of bringing together resources.

However, the fact that, historically and for very long periods of time capitalist ownership has been the norm, suggest that it does have some advantages. If worker ownership is better overall, and given that technology tends to reduce transaction costs, and thus make collective action problems easier to solve, then we would expect worker cooperatives to become more prevalent in the future.

Finally, it can also be the case that there were cooperatives are more efficient, they appear, so that the social optimum of cooperatives is well below 100% of total firms. [1]


The surprising effectiveness of worker cooperatives is not problematic for capitalism: instead of having shareholders, you have workers acting as capitalists. You still have capital markets, and varied ownership of capital goods. It does show that in some circumstances, expropriating shareholders and giving the ownership to workers may be a non-negative-sum game, at least from a static point of view, if those workers do a better job at controlling the firm than shareholders. But from a dynamic perspective, and economy where expropriation and redistribution towards workers is expected will work differently than one in which it doesn’t happen, so even if an individual case may seem economically efficient to expropriate, it may not be so in the long term. This discussion glosses over the ethical case for respecting private property, independent of efficiency, which I consider important.

Worker ownership would not, by itself, dispel the critique that there are people earning income without doing anything to earn it, ‘just from owning’. Workers in a hypothetical Apple co-op would get more profits than the equally hard working owners of Mondragon. People might think it is different because they would see that only people (workers) who do something visible are acquiring income, but it’s just that the roles of capitalist and worker are overlapping.

If worker ownership is not enough to end capitalism, one would have to turn to Option 3. Here you have the problem of there being just one ultimate power that decides over every capital allocation in the market. The Public Board of Investment would receive petitions from workers to give them resources to start new cooperatives, and it would decide who to give control power over capital to. There would be no stock market, but companies could post their value as their discounted expected future revenues. This would miss input from anyone who might want to change that figure. Apple might say that it’s worth a certain amount, but others can disagree in the stock market and short it. In the end, those with better prediction skills are rewarded in the stock market, which tends to ensure that information gets aggregated relatively efficiently.

You can imagine this as a world in which there is only equity-based venture capital (where the venture capitalists takes 100% equity), and the only venture capitalist is a Public Board.

Why would this underperform capitalism?


Basically, because the Board would not be challengeable. If one has an idea that has been rejected by the board, you can’t implement it, period. There is no other agent in the entire economy who can finance you.

So maybe can you fix this by having thousands of analysts, and each one of them can decide? If one rejects your proposal for funding, you can take it to another analyst of the Public Board and hope they accept it.

This takes the problem one step further. The analysts will have a budget allocation from the Public Board, which will be decided for them by a manager.

Okay, so let’s have lots of managers to distribute money!

But someone would have to assign money to the managers mow. Say, you have 1000 analysts, and 100 managers, so each manager gets 1/100 of the total money earned by the Public Board and uses it to hire/fire and reward its 10 analysts. But why 1/100 per manager? Why not give more to the more successful managers.

The system is then improved with an algorithm that rewards managers according to their success. But which criteria to use for the algorithm? Whoever designs it is again in the position of being the sole decisor for the whole economy.

The problem is not solvable the way I’m trying to solve it here. Yo really need separate capital ownership so that every piece of capital equipment can be potentially owned by everyone.This can be approximated adding epicycles to the basic market socialist model to make it more and more like capitalism. This discussion ignores incentive problems, and assumes everyone will be well behaved, but those problems would be there.


Capitalists do three useful things: to refrain from consumption, and invest instead, to allocate capital, and to bear risk. Even in my market socialist example this still happens, only that the whole society is made a collective ‘capitalist’.

That is: First, you decide to set apart a fraction of your resources for future use. Then, you decide where to invest those resources in. Then, you benefit from the upside, or suffer from the downside of that decision.

Even something as simple as owning a pension or investment fund is a valuable activity. First, you are not consuming the money, so you are not demanding resources that others can use. Then, you are investing the money in a particular fund, so you are favouring those productive plans vs others in an economy. The money you invest goes to buy shares of a set of companies. So first, selling shareholders are rewarded for their past performance. Then, the performance of the value of those newly acquired shares will determine your future benefits and losses.

And also, the value of the shares themselves will serve for others to measure the performance of the company, which is useful when new shares are issued.

In capitalism, no one is rewarded ‘just for’ owning. I own a few books, and I’m not earning anything for them. Not even ownership of the means of production automatically generates rewards. You can buy an Airbus A380, park it in a field, and wait as much as you want, and you are not getting income from there anytime soon. The same goes for a landlord: a crumbling house in the middle of nowhere is almost worthless. A well furnished, nicely looking flat in the middle of London can cost a fortune.

Capitalists are rewarded for performing useful functions for the rest of the economy. And even marxists end up admitting this, as pointed out at the end of this talk. Key paragraphs:

Let us take, first of all, the words “proceeds of labor” in the sense of the product of labor; then the co-operative proceeds of labor are the total social product. From this must now be deducted: First, cover for replacement of the means of production used up. Second, additional portion for expansion of production. Third, reserve or insurance funds to provide against accidents, dislocations caused by natural calamities, etc. […]

Before this is divided among the individuals, there has to be deducted again, from it: First, the general costs of administration not belonging to production…(Marx, 1875)

It has often been contended that in the future society everyone will have the right to the full product of his labour. ‘What you have made by your labour, that you will receive.’ This is false. It would never be possible to realize it fully. Why not? For this reason, that if everyone were to receive the full product of his labour,there would never be any possibility of developing, expanding, and improving production. Part of the work done must always be devoted to the development and improvement of production. If we had to consume and to use up everything we have produced, then we could never produce machines, for these cannot be eaten or worn. But it is obvious that the bettering of life will go hand in hand with the extension and improvement of machinery. It is plain that more and more machines must continually be produced. Now this implies that part of the labour which has been incorporated in the machines will not be returned to the person who has done the work. It implies that no one can ever receive the full product of his labour. (Bukharin, 1920)

Someone will be dealing with the consumption/investment tradeoff, and someone will be choosing where to invest, and someone has to bear risk for those decisions. In communism, the entire society, as one entity, bears risk, and very indirectly, decides how much and where to invest. Aggregation of decisions happens via voting, or comittees. In capitalism, every person decides all of that for themselves, with the possibility of letting others take more fine grained decisions (e.g. invest on a pension fund, and then the fund picks companies). Aggregation of decisions work via prices, which, as we know from economics, convey signals far better.

So can you have an economy without capital and capitalists? Yes, you can, but worker-owned firms do not suffice. You have to socialize the entire capital stock. But the functions that capitalists perform will still be performed by someone, so those you cannot get rid of without consequences, for those functions are economically useful and necessary.

Further questions

  1. What about the so called ‘sovereign wealth fund socialism’.
    1. That will only work within a market. Full market socialism will end like regular central planning socialism.
    2. How large can a SWF that invests on its own economy be?
    3. What do the empirical literature say about this?

EDIT: Meta-analysis that shows that worker-owned companies perform slightly better than non-worker-owned companies.

[1] I’m not very familiar with the literature on worker cooperatives, so my confidence in section III is not very high. I remember there were theoretical arguments why they are not prevalent, but the History of Economics is the story of how theoretical arguments fail and institutional innovations succeed. apparently, the key book on the matter is Dow’s Governing the Firm: Worker’s control in theory and practice., which I haven’t read.

This entry was posted in Blog. Bookmark the permalink.

6 Responses to The role of capitalists in capitalism

  1. Arjun says:

    Excellent write-up–and I say that as a Marxist/anti-capitalist, too! Too many of my comrades seem to have a very superficial understanding of the actual functions of capitalists/managers/investors, and the requirement for some kind of surplus value generation for society to develop and progress.

    I think the biggest point of contention that I would have is the lack of an “Option 4” for socialism, that combines Option 2 and 3, in a way that sees ownership of enterprises and means of production and whatnot socialized in a decentralized fashion, that puts both workers and the larger community in charge. So, not just one public entity (a la USSR), but along a more confederated and municipal structure. This model appears to be getting implemented by certain contemporary radical movements, particularly Zapatiasta-controlled territories of Chiapas, Mexico, and the northern Syrian region of Rojava, and on a smaller-scale in some local areas in Colombia, Venezuela, and Bolivia.

    • Artir says:

      Interesting. Do you have some sources on that? I wasn’t aware of it!

      So if I understand you well, the system would have separate capital ownership, but it would be tied to different communities rather than particular people. That would ensure that a) decisionmakers would retain some degree of knowledge over the investments, if those are made within communities b) Investments would be made according to the needs of that particular community.

      Or even, within communities, subcontrol given to coops over certain domains, and be subject to audits from the community, or something like that.

      Rather than central planning, this would be decentralised planning under diverse control (like in capitalism), but by communities and workers (unlike in capitalism). In principle, people who wanted to be capitalists in this system would have problems trying to be so, as perhaps the nicer communities would follow the described system. They would have to go to some island and set up their capitalism there, or something. Like what happens now with taxes: Perhaps one dislikes taxes, but prefers living in Sweden than in Nigeria. Or maybe one dislikes direct democracy, but prefers living in Switzerland, because it works well.

      • Arjun says:

        I can’t think of any specific articles on the empirical examples off the top of my head, but I’d recommend looking at articles on Rojava in ROAR Magazine, an autonomist/libertarian Marxist magazine which has developed close links with local activists and theorists. Also, I’d recommend the essay “Chasing Utopia” by Sam Gindin, published last month in Jacobin Magazine, to get a lengthy Marxist socialist critique of cooperatives which echoes many of the points you made in this article (although I’m not quite sure he supports the “decentralization” perspective).

  2. Pingback: Saturday Interesting Links | Jnana Yuddha

  3. Ethon says:

    Is (V) really that problematic? Wouldn’t a few iterations make things work good enough? It seems to me that the main problem is one of information, the analysis of such an vast range of possible investment opportunities and their relations can’t be scrutinized simply by some analysts (even assuming all of the information is available… and the tools for data processing and analysis of the analysts, that maybe would be also used by others, would also be an investment choice!).

    Since prices stem from exchange, in a monetary economy with a well developed capital market, prices signal profitable investment opportunities and they incorporate almost all if not all public information available. Not surprisingly, a very important result of market design is that the price system, under competitive assumptions (but experimental results show that way more lax conditions deliver competitive results), is the most efficient mechanism that economizes on information:
    In the price system, only prices must be known to reach en efficient allocation of resources, nothing more.

    Also, some other problems for a coop world, due to only workers of an enterprise being allowed to own ‘shares’ of that enterprise, would be: inefficient risk sharing (tons of idiosyncratic risk in the agent’s portfolios and risk bearing would not be necessarily associated with risk preferences) or maximization of average benefit per worker instead of total benefit,

  4. ohwilleke says:

    A few observations:

    * One of the main historical functions of cooperatives (aka “mutual companies”) was to have financial institutions (banks and insurance companies mostly) owned by consumers rather than shareholders. Prior to FDIC imposed reserve requirement and tough state insurance regulation of reserves, banks and insurance companies routinely went bankrupt during recessions (i.e. like 50% of all banks every recession). Mutual companies, in contrast, almost never went under. The difference was a simple matter of incentives. Shareholders multiplied their gain with leverage, but didn’t bear the loss in an undercapitalized company because of limited liability. Mutual companies, in contrast, cared about downside losses and well as upside gains and chose to be less leveraged as a result. Mutual companies remain common in the insurance industry and in professional liability insurance companies (where accurately estimating claims histories is important and customer-owners are more likely to cooperate in taking steps to reduce premiums by reducing risk).

    * The defining feature of almost all successful worker owned ventures is that each worker’s contribution is either the same, or is comparable by some simply metric (e.g. hours worked or units produced). Worker owned ventures tend to thrive under these conditions (e.g. Pacific NW plywood plants, taxi companies, law firms, accounting firms), even when capital investment requirements are substantial and worker income is modest. In contrast, when worker owners are heterogeneous in the kind of value that they add to the business, worker ownership fails unless it is restricted to a subset of owners are are homogeneous. For example, a worker owned taxi company can work, but only if the only owners are taxi drivers and other kinds of employees (IT managers, mechanics, etc.) are non-owner employees of the firm.

    * Rural America is full of cooperatives, usually of people who buy from, or sell to, some homogeneous good or service to a firm. Rural cooperatives tend to work because farms that join them are relatively homogeneous (producing or consuming identical fungible goods, albeit in different proportions), because of the social capital that arises from a very stable population of owners, because those owners often have experience in government and non-profit as well as farm governance (because rural areas need the same number of elected officials per jurisdiction despite having much smaller populations), and because economies of scale juxtaposed against a limited demand means that the community can often only afford to support one supplier. Here, cooperatives provide the benefits of economies of scale while mitigating the abuses associated with a monopoly in the absence of effective public monopoly regulations via either anti-trust laws or regulated utilities.

    * State ownership of capital generally involves public ownership of natural resources such as mineral rights, water rights, rights to generate hydropower, vacant land that is difficult to farm or develop due to terrain (mountains and swamps) or climate (deserts or near deserts or flood plains) Allocating ownership of pre-exploration mineral rights to surface property owners is nothing more than a raffle with tickets coming free with the purchase of surface rights in land, and the need for mineral right lessors to correctly determine the owners of mineral rights (which tend to be split up over multiple generations of inheritance rather than sold) and compensate them accurately according to their interests imposes profound dead weight loss transaction costs on mineral exploration and development that are eliminated with public ownership of mineral rights.

    Private ownership of water rights is also a quite expensive to administer property rights system and it turns out in practice to lead to thin markets because water rights are complicated and of limited value to most users at the point where they must be utilized. The high transaction costs of transferring water rights, in practice, tends to make the effective cost of water far too cheap for the predominantly agricultural users that usually hold the most senior water rights in the Western U.S., and far too expensive for other water users who put water to much more economically productive uses. If water rights were instead auctioned on a regular basis by a public owner of the water, this problem could be eliminated and water would be used much more efficiently even without formal government regulation during droughts. For example, if farmers paid a market price for their water, they would find it worth their while to use more efficient drip irrigation systems which cost more but use far less water per agricultural product produced.

    FCC auctions of the airwaves in the first instance (rather than using some equivalent to homesteading) has proved quite successful. Auctioned mineral rights and grazing rights in the West also work quite smoothly and in the case of grazing rights facilitate the sharing of a scarce grazing resource by ranchers. Public ownership of wildlife through hunting and fishing licenses has worked much better than private ownership at managing wild animal populations to prevent them from going extinct.

    * Econometrically, the biggest benefit of capitalism (defined as private ownership of the means of production) has far more to do with the fact that capitalism forces failing enterprises to discontinue operations or go bankrupt and reorganize (something that often does not happen for decades in money losing state owned enterprises), than it does with having a better system for deciding who should receive the proceeds from the use of scarce capital assets.

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s