For some time now, union membership has been steadily declining. In the United States, after reaching a peak in 1954 of 34.8 per cent for salaried workers (28.3 per cent for all employed workers), unionisation fell to 10.3 per cent for salaried workers in 2019 (and slightly less for all employed workers) – its lowest level in more than 75 years. In Australia, union membership has fallen from more than 50 per cent as recently as 1976 to just 14 per cent in 2016. And in the United Kingdom, union membership stood at more than 13 million workers at its peak in 1979; by 2014, it was down to 6.4 million, less than half the peak rate, and about where it was just before the Second World War. In many other liberal capitalist democracies, unionisation rates have been falling steadily as well.
In part, this decline comes from the success that unions have had in establishing better working conditions and wages for all. Joining a union simply seems less urgent now than it used to. In part, it’s due to changes in the relevant economies, where large numbers of unionised jobs in heavy manufacturing have been moved to less-unionised countries, and the jobs that remain are in industries that are more difficult to organise. Local outsourcing, where what used to be high-paying unionised jobs are moved to smaller, exploitative, non-unionised contractors, has also played a role. As have technological advances in automation, which have made many well-paying unionised jobs obsolete.
But this decline in union membership is also the result of decades of relentless attacks by the political Right against the very idea of unionisation – attacks that have been steadily increasing in both frequency and vigour, especially in the US. Union membership is now becoming sufficiently small that unions might soon lose their ability to adequately protect workers from economic and personal abuse, if they haven’t already. They might even lose their relevance as a political force.
Over the exact same period, we have also seen a dramatic rise in the share of income going to the top 1 per cent, and a lengthy stagnation in real wages for everybody else. Not surprisingly, studies show that the reduction in unionisation rates explains a good part of this phenomenon. These trends were then accentuated by the Great Recession of 2008. This caused a steep rise in unemployment, not only in the US but in other liberal capitalist democracies as well. And as unemployment came down over the following decade, employers replaced high-paying permanent unionised jobs providing good benefits with low-paying temporary non-unionised ones providing no benefits at all. Now that unemployment is soaring again because of the COVID-19 shutdown, the lack of unionisation also threatens public health. The overwhelming majority of essential workers are non-unionised, and they are, in many cases, being asked to work without adequate protective equipment, testing and physical distancing measures.
Joining a union has always been presented as part of the right of free association. That is, workers argue that they have a right to unionise; employers argue that they do not. But even when employers recognise the right of employees to unionise, they tirelessly place obstacles in front of those who would like to do so. Employers also try to starve unions of the funds that they need to operate.
Union supporters respond by emphasising that unionisation promotes the common good. That is, they argue that unions raise wages, improve working conditions, reduce income inequality, suppress invidious discrimination, and so on. Anti-union forces deny these claims, but these denials fall against the overwhelming weight of empirical evidence. The more successful anti-union argument is that compulsory unionisation is a violation or workers’ rights, just as a prohibition on voluntary unionisation would be. It is a violation, they claim, because the right to liberty gives workers the option to refuse to join a union if they don’t want to. Employers then lobby workers to vote against unionisation or to refuse to join even if unionisation passes.
But the claim that liberty protects workers against compulsory unionisation is perverse. Workers shouldn’t have to go through what amounts to a political campaign and vote to unionise before an employer is required to recognise the union and deal with it. Unions are a basic institution of a just society. And as a basic institution, their existence isn’t optional. They are subject to regulation, like any other basic institution, but every firm’s employees must have a union, like every community must have a government if it’s going to be in a position to be just. Except perhaps for very small businesses, firms must accordingly be unionised, no matter whether the firm’s employees have affirmatively voted to unionise or not.
This doesn’t mean that firms would all become what are called ‘closed shops’ – businesses where only existing members of the union can be hired. Instead, firms would be ‘union shops’. Hiring would proceed as it does now, and workers wouldn’t need to be members of the union to get a job. Once hired, however, the worker would automatically become a member of the relevant union.
This might sound like a radical position, but it’s not. Remember, capitalist societies reject the idea of central economic planning. Instead, they rely on the price mechanism of the free market to ensure that the economy allocates resources, including human resources, efficiently. But the free market doesn’t operate within the firm. In the firm, resources are deployed – like in a socialist economy as a whole – by centralised command. Every firm has a strict hierarchical structure. Notwithstanding the fact that shareholders might have some say in how the firm is managed (although only huge shareholders have any real influence; others simply have to choose whether to go along or sell their stock and move on), the firm functions like a mini socialist dictatorship. Except that in a socialist society, the guiding principle for how the firm should be run is the common good. In a capitalist society, in contrast, the guiding principle is the good of that individual firm, meaning, supposedly, the maximisation of its profit.
Being an employee is a serious threat to one’s liberty
So what? Well, the justification offered for allowing firms to form in a capitalist society, despite the internal central planning that this entails, is that doing so saves a lot of what economists call ‘transaction costs’. These are expenses that would otherwise be incurred while negotiating each individual business venture with all the relevant participants, thereby allocating resources internally as well as externally using the price mechanism. This would mean that no one could be called an employee – everyone would be an independent contractor, and could accept or refuse any assignment and haggle over the price every time something in a particular business venture needed to get done. This, in turn, would require countless time-consuming individual negotiations, giving the organiser of the business venture only limited ability to plan how a product would be designed, built, marketed, serviced and delivered over time. By adopting a strict hierarchical structure, we eliminate all the cost and delays of engaging in these repeated negotiations, and the uncertainty of not knowing what human resources are available in the medium to long term and at what cost. This leaves more resources to dedicate to actually producing things, making them cheaper and more accessible to larger numbers of people. It raises the general standard of living.
While some transaction costs are saved by recognising the firm as the basic business institution, doing so also incurs ‘social costs’. These are costs that arise from the production process but are typically more indirect, intangible and difficult to monetise. For example, few relationships are as fraught with opportunities for abuse, exploitation and mental, physical and economic domination as the employer-employee relationship. In other words, being an employee is a serious threat to one’s liberty. Just ask any employees what they fear most – it is the possibility that their manager will, for one entirely arbitrary reason or another, fire them, or do something to make their working life more physically or mentally onerous, less financially rewarding and, in any event, more psychologically distressing. So it’s ironic indeed that those on the political Right claim that compulsory (what I call universal) unionisation infringes on workers’ liberty. This is similar to those who are protesting lockdown and social distancing measures, effectively claiming that they have the liberty to infect themselves and others if they want to. But this is not the kind of liberty that a liberal capitalist democracy was created to protect.
Insisting that all employees join a union is an infringement of what philosophers call ‘negative liberty’ – that is, human interference with one’s ability to do something that one would otherwise have the capacity to do. But every liberal capitalist democracy interferes with everyone’s negative liberty in a massive number of ways every day. If I could beat you up, steal your stuff, defraud you, and enslave or even kill you if I wanted, I would have more negative liberty, and you would have more if you could do these things too. But this would make social life impossible. The motto is ‘Live free or die’, not ‘Live free and die or kill others.’
Everyone who is thinking clearly, then, realises that the mere interference with someone’s negative liberty doesn’t amount to a violation of rights, something you have an absolute power to prevent in a free society. Negative liberty is not a political theory – it’s not a theory about what ends we can pursue, and how to set priorities among competing ends. It’s an analytical theory – it tells us that interference with our capacity to act, by other human agents, is different from interference by things or animals or the laws of physics, and claims that, when there is such interference, this needs to be justified. It doesn’t tell us what counts as a justification or how strong that justification must be. To do that, we need a political theory – one that describes the ends we may pursue free from interference.
There are a number of theories that do this, but the one that’s most relevant here is called ‘republican liberty’. This theory is derived from how philosophers conceived of liberty back in the ancient republics of Greece and Rome. Their thinking was that someone isn’t free if they are subject to the arbitrary will of another. Being a slave is the most extreme example of a lack of republican liberty, given the absolute subjugation of the slave to the will of the master. But cases of absolute subjugation aren’t the only concern – even a single instance of subjugation to the arbitrary will of another is a violation of one’s republican liberty.
Given the degree to which workers lack autonomy and are at the mercy of arbitrary and capricious decisions by their employer, republican liberty is at risk when it comes to the employer-employee relationship. Without a union, employees are subject to all kinds of arbitrary treatment. With a union, employees have some protection against this. As long as the firm remains the basic institution of our economy – and I am not suggesting that this should change – then unionisation is a necessary countermeasure to the threat to republican liberty that the firm presents. Rather than being an infringement of the kind of liberty that we actually care about when we talk about freedom in a liberal capitalist democracy, compulsory unionisation is its protector.
Every society must have a basic structure. It must decide what kind of economic, legal, political and educational system it wants to employ. Once it makes those choices, it must then populate these systems with basic institutions. These are the institutions that operationalise the basic structure. They set the stage for ordinary social and political life to take place. They cannot guarantee justice – a great deal of post-institutional regulation is required to do that – but they can be designed so as to maximise the chances that justice might be achieved. That’s what guaranteeing ‘background justice’ requires.
Once we have embraced capitalism as our economic system, it’s clear that the primary form of business organisation must be the firm. It is, therefore, a basic institution. But because the firm also puts the republican liberty of workers at risk, we can’t rely on post-institutional regulation alone to address this risk. History shows us that, without unions, workers are too often subject to exploitation and abuse and arbitrary treatment even when there are laws prohibiting such misconduct. Given the expense, delay and risk of seeking recourse in the courts, the deterrent effect of these post-institutional legal prohibitions isn’t enough. A just society must do more to discourage the firm from using its power to threaten the freedom of its employees. This means that we need to recognise that unions are as much a basic institution as the firm itself, and make unionisation universal.
There’s a ceiling on what any rational union will ask for. There’s no floor beneath which employers won’t sink
There is already longstanding precedent for this in Germany and other northern European nations. There, as part of what these nations call ‘co-determination’, unionisation has been a prevalent fixture in the economy for almost the entire postwar period. And while the unionisation rate is dropping even there, it remains high compared with other liberal capitalist countries. In these countries, unions even get seats on many corporate boards. Given the quite impressive economic performance of these countries, the argument that increased unionisation in other liberal capitalist nations would retard their economic activity is not credible.
Nor does endorsing universal unionisation mean inviting widespread economic disruption. Workers wouldn’t constantly be on strike and firms wouldn’t be forced to pay salaries so high and provide benefits so great that businesses would quickly fail. Remember, unions have an incentive to be realistic in their demands. If their demands are excessive, the firm will go out of business and all the workers that the union represents will lose their jobs. There’s a natural ceiling on what any rational union will ask for. In contrast, there’s no floor beneath which employers won’t sink in their attempts to get more work for lower wages and fewer benefits. The more they can exploit their workers, the more profit they can make – at least, this is how many firms seem to act. In practice, employers pose a far greater threat to workers’ liberty than any created by compulsory unionisation.
Note, however, that when I speak of universal unionisation, I am speaking only of the core functions of unions. These are: the right of workers to collectively bargain with their employer; to monitor employer compliance with existing rules and regulations, and the terms of the collective bargaining agreement; to represent individual workers in disciplinary actions brought against them by their employer; and to lobby government for the enactment of laws, rules and regulations that promote their workers’ interests. These core functions don’t include the right to strike. Rather than being a core function, this is something that can be bargained for or relinquished in exchange for other benefits.
Note, also, that I have not addressed whether the same rules regarding universal unionisation in the private sector should apply to the public sector, especially to public employees who provide essential services. I happen to think the same rules should apply, but I don’t have room here to make an argument for this.
Finally, note that embracing universal unionisation wouldn’t resolve all the practical questions about how this would be implemented and what unions could and couldn’t do. All sorts of rules – about how unions could go about their business, how they would compete to represent the employees of particular firms, and so on – would be required. But again, these are post-institutional rules. All basic institutions are subject to such rules, and as long as these rules don’t undermine any of the unions’ core functions, these rules can be determined according to the normal political process, or by bargaining between the union and the employer. Taking unionisation away deprives workers of the very kind of liberty that liberal capitalist democracy was created to protect. By recognising unions as a basic institution, we would ensure that they couldn’t be taken away. Only then might workers reasonably consider themselves to be free.
But what about the rights of free speech and free association – wouldn’t universal unionisation violate these? The first thing to note here is that these are derivative rights – that is, they are derived from our general concept of liberty, and are not independent of it. They are merely applications of our general concept of liberty to specific kinds of activities. As such, they can’t be broader than the fundamental notion of liberty from which they’re derived. So, if our consideration of general liberty supports recognising universal unionisation, that conclusion can’t be reversed by considering the role of special liberties such as free speech and freedom of association.
Claims based on these particular notions, however, are popular, and their derivative nature hasn’t stopped anyone from raising them to successfully attack unionisation through the courts. So let’s consider them anyway. The free speech argument is an attack on the ability of unions to collect dues in order to finance their operations. Unions must represent all employees, even those who aren’t members, so workers who want to freeride have an incentive to refuse to join, as they get the same services whether or not they join. To prevent this, most states have rules that require non-members to pay ‘agency fees’ – basically the same dues as everyone else, less a small amount – to cover the cost of representing them. The free speech argument against this is that the involuntary payment of such a fee is a form of compelled speech, and therefore a violation of the freedom of expression.
We wouldn’t assume that taxpayers support everything a government does simply because they pay for it
But this is nonsense. Free speech might be infringed when we prevent someone from spending money to finance a particular point of view, as the US Supreme Court controversially held decades ago. But this does not imply, despite what the court has now also held, that the compelled payment of an agency fee is an infringement of free speech. In the former case, the money is being used to say something contributors want said, so precluding this is preventing them from publicising their views. But in the latter case, when someone is being compelled to pay a fee, what exactly is that person supposed to have been compelled into saying? The communicative significance of a compelled payment is ambiguous, at best. Indeed, the very fact that agency fees are known to be compelled means that their payment is unlikely to be interpreted as an expression of an individual employee’s beliefs. We wouldn’t assume that taxpayers support everything a government does simply because they pay for it. If an involuntary payment were compelled speech, then taxation and every fee the government imposes would be impermissible.
When it comes to freedom of association, the problem is this: freedom of association implies a liberty to associate with those you want to; it doesn’t imply a right not to associate with those you don’t. One cannot declare oneself not to be Australian or American or Texan or Californian if one otherwise meets the relevant criteria for being included in these groups. You can sever your ties if you want, and this might remove you from that grouping, but as long as the ties remain you remain a member. A person need not take a job if they don’t want one, but if they voluntarily become an employee, they automatically become a member of the union. No right to free association precludes that.