Skip navigation.

Interview: Power in the Firm

Dan Chandler, Elizabeth Anderson

In her 2017 book, Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It), Elizabeth Anderson argues that workplaces give employers authoritarian powers over workers. She shows how free market discourse has blinded us to the authoritarian nature of the workplace, and explores ways to democratise work.

Daniel Chandler (DC): There is a lot of interest in the UK regarding questions about power and control in the workplace. You are working with a new think tank, Common Wealth, on these issues. There are also policy ideas coming out of the Labour Party in particular about different models of ownership and workplace democracy. In your book you draw a parallel between the power that the state exercises over citizens and the power that employers exercise over their workers. Could you say a bit more about what you mean by ‘private government’?

Elizabeth Anderson (EA): I am trying to get at the idea that the workplace is a little government with authority relations – that bosses have authority over workers. Once we can see the workplace as a form of government, we can ask what the constitution of that government is. In American workplaces, and I think to a substantial degree in workplaces in the UK, the constitution of that government is a dictatorship. It is fundamental to it that workers don’t have a constitutive voice in how things are run. They are subject to the authority of their workplace government without having a voice within it. That’s what I call private government. Private government is government that is not accountable to the governed, because the governed don’t have authority within that form of government.

DC: There is a debate within economics about the nature of power within the firm. One objection to the parallel which you draw is that workers can leave firms more easily than citizens can exit their countries. What is your response to that?

EA: I think this is only a matter of degree. Within the EU it’s actually quite easy to exit your country! You have guaranteed rights to enter other countries in the EU.

The power to leave an employer, meanwhile, declines over time. Suppose one’s occupation is such that one is powerless and vulnerable to abuse at work. So, for instance, the overwhelming majority of workers in the United States who work in restaurants suffer from sexual harassment by bosses, by customers and by co-workers. Well, if they want to stay in that industry where are they going to exit to in order to avoid this? You don’t know that you are going to be subject to sexual harassment until you enter that job and experience it. It’s hard to tell from the outside. But someone who accumulates a record of quits is going to look less and less employable. Under American law people who quit also can’t collect unemployment insurance. So, there are huge costs to exit.

DC: There hasn’t been much discussion of power relations between bosses and workers in recent decades, even on the left. Why do you think this is?

EA: I think the dominance of free market discourse plays a role. When we conceive of the employment relation as a matter of freedom of contract it presents a superficial appearance of ‘no power’, because the two parties enter into a contract as independent parties. There is supposedly a meeting of minds, and they work out the details in mutually acceptable terms. But that is illusory.

The employment contract is not something that’s been independently devised and negotiated between the parties. In fact, the default employment contract is much like the marriage contract. Its default terms have been set by the state. In the nineteenth century, the marriage contract cancelled out a woman’s independent legal existence. She was rendered by the terms of the contract entirely subordinate to her husband. Today the state has altered the marriage contract to make things more equal. But it hasn’t altered the employment contract to equalise the power of workers and employers.

The employment contract, by default, hands nearly all power to the employer, reserving almost none for employees. Yes, you can negotiate from that baseline, just as under the old law of coverture wives could negotiate from a powerless baseline for a pre-nuptial agreement to retain some rights, say to keep their own property. But in reality, when you have dealt all the authority cards to one side, what incentive does that side have to deal any of those cards back to the other side? Very little.

Ordinary workers really only have the power to get authority cards back if they are organised as a collective in a union. Then they have some power, but as individuals generally not.

DC: Some economists argue that it doesn’t really matter what is written in the contract – that the eventual relationship will just reflect the relative bargaining power of bosses and workers. What do you make of this?

EA: To a certain extent this is true. The market will deliver better conditions for people with more rare skills. But we also see enormous market pressures to commodify labour – that is, to define tasks in such a way that workers are highly replaceable. Then the employer doesn’t need to care about massively high turnover due to abusive working conditions.

That’s what we see among the masses of lower-paid work, and with the rise of the precariat I think this is getting more and more extreme. There has been a decline in the relative number of permanent jobs, particularly for young workers. More and more people are doing gig work, temporary work, zero hours contracts, all kinds of unstable employment arrangements – or they are considered (often falsely) as independent contractors.1 There are more and more devices being deployed to deny workers any kind of bargaining power or ability to avoid abusive treatment.

DC: I want to step back for a moment. In your book you talk about how in the eighteenth and early nineteenth centuries, especially in the United States, markets were associated with more egalitarian social relations. The market was seen as a sphere where citizens could interact as equals. Support for free markets was part of a wider movement to challenge existing social hierarchies – monarchy, church, patriarchy, guilds and so on. I think this idea might come as a surprise to a lot of people on the left. Can you say a bit more about this?

EA: The original vision of free market theorists was that eventually everyone would be self-employed. They might have a temporary few years of employment under a boss, but then they would be able to save enough money to go into business for themselves.

This was a very American idea. Due to the ethnic cleansing of Native Americans, there were mountains of free capital available for distribution. So it wasn’t a crazy idea to think that free markets, especially in land, would lead to a condition of universal self-employment, at least after the abolition of slavery.

This was Abraham Lincoln’s explicit vision – free labour meant self-employed labour. Lincoln was anti-slavery. He also enacted the Homestead Acts, which literally gave away free farms out in the Great Plains and further West to anyone 

willing to farm them. He really did promise that everyone could be self-employed. If everyone is self-employed, then everyone can only profitably own as much capital as they are able to work. That was a condition of claiming a farm under the homestead act: you couldn’t just hold the land, you had to farm it. So every family would be roughly equal in their holdings, because one person can’t plough that much more than another. That is a plausible model of how you could have a free society of equals under a market society.

The advantage was that, with a free market in land, everybody could get their piece. This was in contrast with England, where so much land was tied up by the monarchy, aristocracy and corporate institutions. I just learned on a punting expedition that Trinity College alone owns enough land so that you can walk from there to London without ever leaving Trinity College property! That was one of the big complaints that Adam Smith had about land holding in England: it’s monopolised. He thought a free market in land would force the lazy landlords to break up their estates and sell their land off in little pieces, providing opportunities for all kinds of enterprising yeoman farmers to take it over and make something of it.

DC: What changed? Why didn’t this come to pass?

EA: The industrial revolution proved that economies of scale are real. Once you get large-scale industrial organisations – corporations that have a lot of employees – then hierarchy is pretty much inevitable. Within an organisation, just to make it governable, you are going to get a hierarchy of offices. There is no other plausible way to run a large-scale organisation. Participatory democracy, once you scale up, becomes unworkable, and so that’s why you get hierarchy. The hierarchies required by economies of scale do not entail, however, that occupants of higher offices within the firm must have dictatorial or unaccountable power over those beneath. Nothing stops us from devising a constitution in which the lower-level workers have a real voice and can hold their bosses to some account.

DC: Is there any way in which the recent rise of self-employment, gig work and so on could be seen as moving the labour market back in the direction envisaged by Lincoln, Smith and others?

EA: While gig work suits some people, especially those who desire flexible working time, for a lot of others it just puts them into a situation of precarity where they can barely pay the bills. They have no stability of income, and hence can’t plan ahead for big projects like buying a house or raising children. That’s terrifying.

The capital of firms in our high-tech era has more and more shifted to intellectual property. Apple, for instance, has few direct employees, and these are very highly skilled. They are tech workers, software engineers and designers. Everything else has been outsourced. The people who make the hardware are in China labouring under sweatshop conditions. But they are not Apple employees. Immense riches are available to a tiny number of people, while the vast majority of work is outsourced to people who are paid very little because they are just commodified workers who are infinitely replaceable.

These outsourced workers are powerless. Gig workers like Uber drivers are probably employees under American law. Yet Uber classifies them as independent contractors, even though it regulates minutely the way these workers conduct their business. That is what makes them employees and not true independent contractors. But Uber persuaded the drivers not to press that case, so that they are not subject to all of the obligations of the employment relation.

DC: I want to come to what we could do to democratise the firm. Are there principles that you would use to think about how to go about democratising the firm, or what an ideal firm would look like?

EA: I’m not committed to any particular constitution of workplace government. I think this is an area ripe for experimentation. Ideal theorists on the left often think that the only acceptable model of the firm, or the absolute ideal model, would be a pure workplace democracy where workers both own and democratically manage the firm. Well, fine, go ahead – you can try that, it’s a great idea. But the idea that no alternative to that is acceptable is problematic. We need to experiment. There are some risks to worker-owned firms because if you have to own the firm, that’s a huge investment of capital. Not everyone wants to concentrate their savings in one enterprise, which they also rely on for their salary. There are good reasons to diversify investments, and there are also some challenges that arise with workplace democracy.

We can see this in the famous Mondragon cooperative. This is probably the leading example of a successful workplace democracy, of a worker-managed and worker-owned set of firms. Yet, when they expanded production abroad, they didn’t extend democracy to overseas workers. We see time and time again that there are internal incentives to these organisations to create insiders and outsiders. The outsiders don’t have the same voice as the insiders. So these models, while they are wonderful and deliver a lot of benefits, aren’t necessarily as stable as one might think in terms of their principles, and the extension of their principles across everybody who is involved. That gives us reasons to explore additional models of offering voice to workers in the firm.

One idea that I find attractive, particularly since it could be implemented rapidly and without massive alterations in ownership, is the model of co-determination pioneered in Germany. Under co-determination there are several levels of governance in the firm. Workers have seats on the board of directors. What might be even more important is works councils, which are on the shop floor and directly regulate the conditions of work.

I think in many ways it’s a more important form of worker voice than having seats on the board of directors, because most of the time the board of directors is looking at overall business strategy. It is thinking about approving plans for new products, marketing strategies and things like that. Not all workers have direct stakes in those kinds of decisions; although of course they do have stakes in questions like are you going to shut down a plant in Germany so that you can set up one in Vietnam?

On the shop floor there are lots of questions about, say, the safety of equipment, the pace of work, flexible working time – the kinds of things that directly affect the day to day experience of work, and workers’ ability to have families and to have a life outside of work. Those kinds of decisions are usually handled at the works council level.

DC: Is it a question of removing barriers to different models of worker participation and ownership, or should the state actively support or require different models?

EA: Because the state already sets the default terms of the employment contract, unless they make the default contract one in which workers have some kind of guaranteed representation, it will be very difficult to get there. One way to get there would be to strengthen the labour union model. In the United States the model of unionisation is extremely fragmented. It’s almost impossible to organise a labour union anymore. That’s been true for decades, because you have to organise shop floor by shop floor and have a protracted election. Even after a union wins an election, the employer still has many powers to resist good faith-bargaining. Employers exploit their powers to the hilt. The US model hasn’t been very effective for organising workers for decades.

There are other ways to organise unions. In many European countries – for example France – the percentage of unionised workers is not that different from the United States, but unions have the power to negotiate on an industry-wide basis, so people still have enormous coverage and protection from unions. But there are difficulties of representation there too, because if you are not a member you can’t vote for these people, and the rules that are put in place by the unions are almost inevitably more attentive to the members than to people who aren’t members. 

So I’m not going to say this is a perfect model, although it delivers greater protections for workers who actually have jobs. A bigger problem is how do people even get into that system? In France it’s really hard for many people to get a permanent job because then it’s almost impossible to fire you. Many employers don’t want that and so they just don’t hire people on permanent contracts, and that’s creating a precariat.2

DC: So it’s not obvious what the right arrangement of rules and policies is. The problem is we haven’t really been grappling with these questions, and we need to start trying out different models.

EA: Yes. Suppose we made the default something that automatically incorporates a worker voice. The state could just say this is the way things are going to be. It could even offer a variety of models and then the corporation could select a model. You could start off with one model, but then permit deviations produced by negotiation between parties. These would now be more equal in their power, and better able to encourage experimentation.

DC: Do you think there is a trade-off between firm efficiency and worker voice and participation? And if so, how do we deal with this?

EA: If there are trade-offs, they are modest. 3 We should accept them in the name of protecting workers’ dignity and interests. Note, too, that sometimes empowering workers improves productivity. Sometimes worker co-determination empowers workers to speak up about dysfunctional work processes and make them more efficient. In other cases, workers want to slow down an insane pace. Maybe profits will go down, but only against a baseline that disempowers workers.

DC: In the book and in this discussion, we have mostly been talking about spreading power within the firm – shifting the firm from a kind of private government to a more public type of government, where workers have a say over decisions which affect them. It seems to me that there is another broad strategic alternative, which is to reduce the power of bosses over workers by making it easier to exit the firm. I’m thinking of policies like a Universal Basic Income (UBI) or even Universal Basic Services – policies which are also sometimes described as ways of transforming the bargaining power of workers with respect to employers. Do you see a role for these policies too?

EA: I think they could play a role, but I wouldn’t be too optimistic that they would have very powerful effects. The UBI will never be so generous as to make quitting costless. Suppose the UBI provides a bare subsistence, so that you could actually live on a UBI without working. That’s probably the most generous conceivable UBI. But it would still be bare subsistence, so if you have a mortgage you still can’t quit with impunity or have other plans that require more than subsistence. There would still be a serious cost of exit. At the same time, high quit rates, which might be encouraged by a UBI, are consistent with the further commodification of labour into infinitely replaceable units. So yes, it could help the precariat. It would provide an important safety net for people who already belong to the precariat, and that’s important. But will it actually discourage the expansion of precarious employment? If anything, it might even encourage it. Why even hire people for jobs? We can just hire them task by task. Everybody is now like a ‘Task Rabbit’, and then you’re scrounging around to get a living beyond subsistence.

DC: Turning to politics, is there interest in the US in these sorts of questions?

EA: In the United States, the decline of the traditional model of union organising has led to a lot of creative thinking and action on the part of workers concerning ways to get their voices heard outside of the traditional union structure. So we see alternative models of campaigning – for example, the ‘fight for fifteen’ campaign to raise the minimum wage to $15 per hour. Workers are expressing their grievances through social media, and in some cases this has inspired consumer complaints. Consumers don’t want to buy from firms who treat workers so horribly.

I was heartened by the fact that Senator Elizabeth Warren, now running in the Democratic Primary for the Presidential nomination, has put co-determination on the legislative agenda. While there’s no chance it will be passed until Democrats sweep a national election, it’s a sign of the seriousness with which new ideas about empowering workers are being taken in the United States. Polling on co-determination in the US shows that it is surprisingly popular, even across party lines. It hasn’t been identified as one of those liberal wonky ideas that Republicans have to hate. If you poll workers and don’t tell them that this is Democratic Senator Elizabeth Warren’s idea, they think ‘Hmm, yeah that sounds really good!’. So I think the moment is ripe, because workers feel disempowered, and they haven’t been getting a fair share of gains of productivity.

Ultra-elite workers have been getting virtually all the gains for the past several decades. People are eager to hear alternatives. There is a lot of interesting practical experimentation and legislative proposals that could be brought about. Of course, there are very high barriers in the United States because of our constitution. You would need Democrats in control of both houses of Congress and the Presidency to get anything passed. But it could happen!

DC: If you had to pick two policies that would make the biggest difference to this agenda what would they be?

EA: One of them would be co-determination. Warren wants to require it for larger corporations. Even in Germany you have to reach a certain size before these requirements kick in. Another policy might be giving unions powers to operate their own forms of insurance, especially unemployment insurance. Unions could also run workers’ compensation schemes, which have been severely degraded in the United States, because employers are responsible, and they lobby states to minimise the requirements. An alternative funding method would be to run them through labour unions. That’s the Danish model. This would be an enormous spur to labour union membership and power.

Elizabeth Anderson is John Dewey Distinguished University Professor of Philosophy and Women’s Studies at the University of Michigan, Ann Arbor. She has written widely about issues relating to democratic theory, equality in political philosophy, racial integration and the ethical limits of markets.

Daniel Chandler is a doctoral researcher in economics at the LSE and a commissioning editor for Renewal.

Further reading

Elizabeth Anderson, Private Government: How Employers Rule our Lives (and Why We Don’t Talk about It), Princeton University Press, Princeton, NJ, 2017 Elizabeth Anderson, ‘The great reversal’, IPPR Progressive Review, Vol 25 Issue 2, 2018.


1. Kathleen Thelen, ‘The American precariat: US capitalism in comparative perspective’, Perspectives on Politics, Vol 17 Issue 1, 2019.

2. Clémence Berson, ‘Fixed-term contracts and labor market duality in France’, De Economist, Vol 166 Issue 4, 2018.

3. Steffen Müller and Jens Stegmaier, ‘The dynamic effects of Works Councils on labour productivity: first evidence from panel data’, British Journal of Industrial Relations, Vol 55 Issue 2, 2017.