Riccardo was a kind and compassionate person, and also a kind and compassionate labour law scholar. However, this did not lead him to adopt a «one track mind» position in favour of always expanding workers’ rights. Given the inequality of bargaining power and subordination of employees to employers, some labour law scholars see the employment relationship solely as a power struggle, with the goals of labour law being to redistribute and give more power to employees. Riccardo certainly acknowledged these fundamental characteristics of employment and the redistributive goal of labour law, but complemented it with a rounded understanding of the complexity of employment relations, including the legitimate interests of the employer, the broader interests of society, and additional interests of employees.

If the traditional view conceives of the discipline «as a factor that curbed economic activity for the sake of social and more broadly humanistic values» (Del Punta 2015, 123), Riccardo has pointedly argued that «[i]t is pointless to conceive of labour law as opposed to economics, as if the former could still concern itself merely with the distribution of wealth and not also – especially at a time of low growth – with its production. Labour law should be law that fuses value claims and organisational and economic considerations» (Del Punta 2015, 135). This approach has led Riccardo to nuanced and pragmatic views that correspond to real-life complexity and rely on empirical economic studies alongside philosophical theories, for example in the context of dismissals or more generally «flexicurity» policies1. It can also explain Riccardo’s long-time interest in the capabilities approach and the potential it holds to reinvigorate labour law theory and policy debates. Instead of the traditional «protective paradigm», rooted in the «old ‘inequality of power’ narrative», Riccardo saw the capabilities approach as a useful answer to neoliberal policies that undermine labour law, while at the same time a new theoretical basis that can lead to much-needed internal rethinking about the goals of the field, and how it should respond to dramatic changes in the labour market and society more generally2.

The general idea of the capabilities approach – first developed by Amartya Sen – is that «capabilities» should be the main focus of our concern (rather than, for example, the allocation of resources) (Sen 1999; 2009). Sen’s theory of justice focuses on advancing capabilities as the best way to advance freedom as well as equality. People should be able to effectively pursue their plans of life, and in order to have this freedom to achieve what they want and value, they need capabilities. In Sen’s terms, what we need and want are «functionings». Rather than focus on providing people with those functionings directly, society should ensure that they have the capabilities needed to secure them. An important part of this idea is that people should «be left free to determine their own course after that» (Nussbaum 2000, 87). Alongside other leading labour law scholars – notably Simon Deakin (Deakin, Wilkinson 2005, 290 ff) and Brian Langille (2011, 101; 2019, 122) – Riccardo argued that these ideas can be used as a new theoretical basis for labour law. In line with his generally nuanced approach, he acknowledged that no one theory can provide justification for all labour laws; he considered it sufficient that the capability approach can be used to justify at least parts of labour law (Del Punta 2019, 87).

As Riccardo explained, some labour laws are designed to protect employees from abuse of power by the employer, and can be seen as advancing «freedom from» such abuse. He considered the capability approach to be compatible with such laws, although not adding much to other, traditional justifications focusing on inequality of power. However, he argued that other labour laws can be seen more directly as positively advancing capabilities, in the sense of promoting the «freedom to» do things that people value, whether within the work context (for example through workplace participation) or outside of it (by having the free time and basic resources needed for that) (Del Punta 2019, 94-96, 98-99; 2016). This is where he saw the capability approach as having a comparative advantage. In this regard, Riccardo saw promise in shifting some of the focus of labour law from «containing the stronger party» towards «an expansion of each individual worker’s capacities, skills, and opportunities, as well as contractual autonomy» (Del Punta 2019, 101).

The capabilities approach to labour law, if adopted as the main justification, can lead to less protections (Davidov 2007; 2014), but at the same time it is associated with a call for broadening the scope of the field to cover more people3. If our goal is to advance freedom (through capabilities), at work and from work, this appears to apply for self-employed people just like employees. Why not expand some of the same laws to cover all those who work for others? Indeed we can see this happening in some countries with regard to a few specific labour laws – mostly those dealing with non-discrimination and health and safety4. Supporters of the capability approach usually call for further expanding the scope, including by covering people whose work is not for any employer/client, such as the work of taking care of one’s own children (Fudge 2011, 120). Riccardo has also suggested that the capability approach can lead to broadening the scope of labour law, but at the same time he rightly acknowledged that at least some labour rights would still require the distinction between employees and independent contractors (and by implication, between people who perform work in general and those who work for others) (Del Punta 2016, 400-1).

It is important, in my view, to bear in mind the distinction between three different circles of responsibility: the state, and employer, and the individual. The fact that as a society we want to advance people’s capabilities does not mean that we should automatically do something about it through regulation or public funding. Rather, there is an additional decision to make about the allocation of various situations into the three above-mentioned circles. In one group of cases, we believe that individuals can take care of themselves – usually through the market – and leave the responsibility with the individual5. In another group of cases, we believe that there is a specific entity – an employer – that has to take responsibility. In such cases we impose liability on the employer through labour laws. Finally, in the third group of cases, society decides to take direct responsibility for people, by funding certain programs (such as public education or public healthcare) or by providing welfare payments.

This distinction – which can roughly be described as the distinction between private law, labour law, and welfare law/public services – remains necessary. Without a specific employer that can and should take responsibility for a specific employee, there is no justification for imposing labour law obligations. And given that public resources are always limited, at least in a capitalistic society the state cannot pay for every need of every individual, but rather has to focus on those who cannot secure their own needs through the market economy. Of course, the lines between the three systems are not fixed. In social-democratic welfare systems, the state takes more responsibility on itself compared to liberal systems who prefer to leave more to the market. And the definitions of «employee» and «employer» can be expanded to include more people within the scope of labour law, if society believes this to be justified. The lines can be moved, but they will still be needed. Which leads me to ask: can the capability approach help to determine how to set the lines?

In previous writings I have argued that the capability approach can provide justification for some labour laws (alongside other justifications). This is especially so for workplace equality laws, broadly conceived (including anti-discrimination, gender pay equity, sexual harassment laws, parental leave, and a right to request flexible work), as well as workers’ health and well-being laws (under this heading I include maximum hours, public holidays, vacation rights, health and safety, and limitations on child labour) (Davidov 2019, 42, 49). Somewhat less directly, the capability approach can also be used as a supporting justification for several other labour laws. At the same time, I expressed doubts in previous writings about the ability of justifications at this level of abstraction to assist in the specific task of determining the scope of labour regulations. This is true not only with regard to the capabilities approach, but also with regard to other justifications such as redistribution (or distributive justice), advancing democracy (or more specifically, voice), protecting dignity, and so on. These are clearly important goals (and justifications) of labour law, but they are not directly helpful for setting the boundaries of labour law and deciding who should be considered an «employee» or an «employer» (Davidov 2016, 118-19). For such purposes, I have argued before that the focus has to be on subordination (in the sense of democratic deficits) and dependency (in the sense of inability to spread risk). These vulnerabilities also provide justification for labour laws. They do not contradict the previous justifications, but rather explain the need for labour laws at a different level of abstraction, which seems to me more helpful for the task of delineating the scope (Davidov 2016, Ch 6). It is only when these characteristics exist that there is justification to impose liability on a specific employer, i.e. that the relevant circle is the one of labour law and not one of the other circles of responsibility.

Notwithstanding my previous sceptical position, I find myself drawn back to this question when thinking about the timely topic of on-demand platform workers, such as Uber drivers. This is because it is claimed that such workers have a relatively high degree of autonomy, in particular the freedom to choose how much to work and when to work, which is considered a key characteristic of independent contractors. Given the focus of the capabilities approach on substantive freedom, perhaps there is something to learn from it about the appropriate status of platform workers.

Digital platforms offer an extremely convenient service to consumers. One can easily open an app on the phone and book a ride, or a delivery, or some other service. The platform can hire employees to meet the demand and provide the service on its behalf, but many platforms (though not all) have chosen a different model: to rely on people considered by the platform to be independent contractors. This allows such a business to save the costs associated with labour laws, as well as to avoid other legal limitations on its managerial flexibility. Is this just another example of employee misclassification, a common and perpetual problem that labour law has to confront? Many courts around the world have concluded that this is indeed the case6. But they had to confront an obstacle on the way to this conclusion: platform workers are not committed to work on specific times or for any specified number of hours. They can log into the app and make themselves available to work whenever they want. This kind of autonomy is generally reserved to independent contractors, and seems (at least on its face) to contradict the characteristics of employees. Specifically, is there subordination (or «control») under these circumstances?

The answer depends on the facts of the specific platform, but at least in the case of Uber, in most countries courts have concluded that subordination exists, and the drivers are employees or at least part of an intermediate category subject to some labour rights7. One reason for this conclusion was the insistence on checking the level of autonomy as a matter of practice rather than theoretical possibility. In reality it turns out that drivers have quite limited autonomy regarding when to work and how much to work, if they want to secure reasonable compensation (which is not lower than the minimum wage). The rates determined by the platform, together with other incentives and disincentives, push people to work during the times they are needed rather than the times they prefer, and often to work more than they wanted. This is not to say that Uber drivers have no autonomy at all on these matters, but the level of autonomy is much lower than what might appear «on paper»8. A second reason that led to the conclusion of subordination in the relationship is the existence of other indicators pointing in that direction. A decision regarding status is based on a «multi-factor test»: an examination of multiple indicators, leading to a holistic assessment of all of them together. Some level of autonomy to decide when to work and how much to work is considered a sign pointing towards independent contractor status, but it has to be balanced against other indicators that might point in the opposite direction9.

It appears, then, that courts can find solutions to the new challenge of platform workers’ status within the confines of existing doctrines10. But the question remains, how can we tell to what extent autonomy is «real» and gauge the level of autonomy? And to what extent should autonomy be a factor determining employee status, at the normative level? Insights from the capability approach can be useful for both questions.

Focusing on the reality of the relationship is a crucial principle of dealing with misclassification claims. It is known in some countries as the «primacy of reality» or «primacy of fact»11. Given the inequality of power between the parties, it is quite easy for employers to unilaterally dictate the terms of the written contract. If we decide the status based on the written terms, that would give employers a simple and easy way to evade responsibility. Courts rightly stress that they decide the status based on the real characteristics of the relationship. But how far do they go to understand the reality?

The most minimal level is ignoring «sham» contractual provisions. For example, in the UK it is common for employers to insert a «substitution clause» which allows the worker to send others to do the work instead of him/her. This is intended to suggest that the relationship is not personal (a precondition for employment), but rather a contract to buy services, which the worker is free to provide either by themselves or through others. On paper, this appears to be a strong indication of independent contractor status; it paints the worker as having a small business with the ability to send their own employees to do the required work. But often it is easy to see that this written contractual provision has nothing to do with the actual agreement between the parties. Imagine, for example, that the worker is expected, through verbal statements (express or implied), to do all the work by themselves. In such cases, UK courts have recognized for some time that the written provision is a sham and should be ignored12.

What if a substitution clause is not a sham, in the strong sense of contradicting the true oral agreement between the parties, but still misrepresents the nature of the relationship? Imagine that the worker is formally allowed to send others to do the work, but both parties know that this is not going to happen, except maybe in very rare cases. The substitution clause was inserted only for the purpose of evading employer responsibilities. In the UK, the Supreme Court has been somewhat inconsistent about such situations. In the case of Autoclenz13, workers who provided car-cleaning services were ruled to be employees, despite a substitution clause and other provisions in their contract to the contrary. The Court made some ground-breaking statements about the need to understand what was the «true agreement between the parties»14, and in particular, acknowledged that «the relative bargaining power of the parties must be taken into account in deciding whether the terms of any written agreement in truth represent what was agreed and the true agreement will often have to be gleaned from all the circumstances of the case, of which the written agreement is only a part. This may be described as a purposive approach to the problem»15. With regard to the substitution clause, as well as another contractual provision allowing the workers to refuse work, it was concluded that they were «unrealistic possibilities that were not truly in the contemplation of the parties when they entered into their agreements»16. At the same time, the Court emphasised that the question was ultimately what was the «true agreement» and cautioned against focusing on the «true intentions» or «true expectations» of the parties, because of the risk that these might be only private intentions17.

A few years later, in the case of Uber, the UK Supreme Court further developed its purposive approach, acknowledging that the ultimate question when deciding if one is an employee was whether the purpose of labour laws was to cover this type of worker. Uber drivers have the vulnerabilities of subordination and dependency and are therefore employees. To reach this conclusion, the Court noted that focusing on the written contract as a starting point when asking if someone is an employee (or «worker» under UK law) would be inconsistent with the purposive approach, given that «an employer is often in a position to dictate such contract terms»18. The Court acknowledged that the drivers «had in some respects a substantial measure of autonomy and independence», because «they were free to choose when, how much and where» to work19, but then went on to detail several other facts showing that there was ultimately a high degree of control over the drivers20.

The Uber judgment took a step further in moving beyond the contract and looking at the «true nature of the relationship»21. This means not only setting aside «sham» contractual provisions, or giving preference to verbal/implied agreements when the written contract is inconsistent with them. Rather, the Uber judgment suggests that courts should give little weight (if at all) to the written contract; instead they should focus on the socio-economic characteristics of the relationships that put the worker in need of protection. Unfortunately, in a recent judgment dealing with Deliveroo riders22, the UK Supreme Court took a step backwards, giving significant force to a substitution clause inserted into work contracts. Although the Court reiterated the «need to focus on the reality of the situation»23, it inexplicably ignored the fact that, as a matter of practice, substitution was very rare24, and concluded that Deliveroo riders where not entitled to labour rights merely because of a substitution clause25. While the Deliveroo judgment purported to follow the same principles adopted in the Uber case, it appears to be entirely inconsistent with it, leaving the legal situation in the UK unclear.

It seems to me that a focus on the purpose of labour laws when interpreting the term employee is key. In Deliveroo this was not the Court’s main focus. But there was also a more specific problem: the Court gave significant weight to the perceived autonomy of the workers to send someone else to do the work. Once the Court concluded that this was not a sham – that riders were really allowed to send someone else if they wanted to – it did not bother to examine whether it made any sense for them do to so, and whether they really had characteristics of an independent business, even though they almost never used the substitution right. A focus on substantive freedom, which is central to the capability approach, could have been helpful here. The judgment in Deliveroo is a good example of giving too much weight to formal freedom. Giving Deliveroo riders a right to send others to do the work is similar to giving a homeless person a right to free speech. It is rather meaningless if one does not have the capabilities necessary to make use of the right. For the homeless person, a shelter, food, health services and otherwise basic economic security are preconditions to be able to have actual freedom of speech – the capacity to be able to think about issues other than bare survival, to formulate opinions and voice them in a manner that can be heard by others. For a Deliveroo rider, some characteristics of an independent business are necessary as precondition for being able to make use of the formal autonomy to send others to work for you. No such indicators where present in the case.

The suggestion proposed here to use the capability approach is not at the level of a grand theory justifying labour law or showing the way for its future development. It is a much more specific use, but one that can prove very helpful. It complements other ideas – notably the purposive approach and the «primacy of reality» – and provides guidance on how to implement existing tests. To be seen as an independent contractor, the autonomy of the worker has to be real. The freedom should be substantive rather than formal. And when looking at some contractual right (such as the right to send others to do the work), it helps to ask whether it actually creates capabilities which enhance freedom26. The overall circumstances and socio-economic situation can assist in answering this question.

The previous section focused on whether a freedom that a worker appears to have is «real». Another question raised in the context of on-demand platform workers is what types of autonomy are important for the decision of whether one is an employee or not. A worker can be free to choose when to work. Or free to choose how much to work. Or free to decide to send others to do the work. Or free to set the price of the service (vis-à-vis the customers). Or free to refuse specific work. Or free to connect directly with customers and build a client base. Or free to work for several platforms concurrently. Or free to choose what route to take when doing deliveries. Or free to decide what to wear during work. And so on. These are all indicators considered as part of the multi-factor test. But in the context of platform work, special attention is given to the freedom to choose when and how much to work, because this is the main unique characteristic of on-demand platform workers that separates them from «regular» employees.

How much weight should be given to this autonomy? I have already argued in the previous section that we should ask whether the freedom is real, and how significant it is in practice, in light of the overall circumstances. At the same time, we can ask more generally about each type of autonomy/freedom to what extent it can enhance capabilities. Imagine that most of the indicators suggest that a worker is an employee; should we change this conclusion if they are free to choose when and how much to work? It could be helpful to ask, what kind of capabilities are enhanced by this particular freedom?

The ability to decide how much and when to work gives workers the possibility to do other things with their time. When pay is very low, as in the case of Uber drivers for example, this freedom is not real but illusionary. In order to make a living, even at a basic level, you have to work more hours and you have to choose the times that pay more. But let us assume another hypothetical platform, where the wages are higher and the flexibility regarding when and how much to work is real. This kind of arrangement supports the capability to do other things with your life. This seems to suggest that there is no need to apply working time laws, or a right to annual leave. The worker can arrange those for themselves. But for other labour laws, the freedom to choose how much and when to work does not seem very relevant. For example, workers who have this freedom can still suffer from discrimination, which diminishes capabilities – thus raising the need to apply anti-discrimination laws. They still suffer from unequal bargaining power, raising the need to allow them to bargain collectively. And so on.

These are just preliminary thoughts. More research is needed to decide which specific labour law protections should apply when workers have real freedom to choose how much and when to work. And the same is true with regard to other freedoms (for example, if the worker is given some discretion to decide on the price of the service). The point I wish to make is merely that it is helpful to ask to what extent a specific freedom enhances capabilities that are relevant for specific protections. A separate question is whether we want to create a group of workers that is subject only to parts of labour law; such an intermediate group has some advantages but also disadvantages-1. But even for those who prefer a single group of employees that enjoy the same group of rights, an understanding of which rights are relevant for each type of work would help to decide who should be included in this single group.

Among his many academic contributions, Riccardo Del Punta was part of a group of world-leading labour law scholars who advanced the capability approach to labour law. His writings on this topic are sophisticated, convincing and balanced. My own approach is somewhat more sceptical when it comes to adopting the capability approach as a justification for labour law (although I do agree that it provides an additional justification for some specific labour laws). However, in this short paper I took inspiration from Riccardo’s work and argued that the capability approach can be useful for the timely and important task of deciding the status of platform workers. A focus on substantive freedom and on capabilities can help to understand whether the autonomy given to platform workers is «real», and whether it really makes a difference in terms of their vulnerabilities. This does not come instead of the need to apply a purposive approach and look for the vulnerabilities of subordination and dependency, but rather as a supplementary tool to assist in this inquiry.

Adams, Z., et al. 2021. Deakin and Morris’ Labour Law. Oxford-Portland: Hart Publishing.

Aloisi, A., Cherry, M. 2024, “Tertium non Datur: Intermediate Employment Categories.” In The Oxford Handbook of the Law of Work, a cura di G. Davidov, B. Langille e G. Lester. Oxford: Oxford University Press

Bogg, A. 2012. “Sham Self-Employment in the Supreme Court.” Industrial L.J. 41: 328 sgg.

Caruso, B., Del Punta, R., Treu, T. 2022. Manifesto for a Sustainable Labour Law.

Davidov, G. 2007. “The (Changing?) Idea of Labour Law.” Int. Lab. Rev. 146: 311 sgg.

Davidov, G. 2014. “The Goals of Regulating Work: Between Universalism and Selectivity.” U. Toronto L.J. 64: 1 sgg.

Davidov, G. 2016. A Purposive Approach to Labour Law. Oxford: Oxford University Press.

Davidov, G. 2017. “The Status of Uber Drivers: A Purposive Approach.” Spanish Lab. L. & Emp. Rel. J. 6, 1-2: 6 sgg.

Davidov, G. 2019. “The Capability Approach and Labour Law: Identifying the Areas of Fit.” In The Capability Approach to Labour Law, a cura di B. Langille. Oxford: Oxford University Press.

Davies, A. C. L. 2009. “Sensible Thinking About Sham Transactions.” Industrial L.J. 38: 318 sgg.

De Stefano, V., et al. 2021. “Platform Work and the Employment Relationship.” ILO Working Paper No. 27, 31–3.

Deakin, S., Wilkinson, F. 2005. The Law of the Labour Market: Industrialization, Employment and Legal Evolution. Oxford: Oxford University Press.

Del Punta, R. 2015. “Labour Law on the Edge: Neoliberal Decline of Regeneration?” In Labour Law and Social Rights: An Evolving Scenario, 121 sgg.

Del Punta, R. 2016. “Labour Law and the Capability Approach.” Int. J. of Comp. Lab. L. & Ind. Rel. 32, 383: 403 sgg.

Del Punta, R. 2019. “Is the Capability Theory an Adequate Normative Theory for Labour Law?” In The Capability Approach to Labour Law, a cura di B. Langille. Oxford: Oxford University Press.

Del Punta, R. 2022. “The Regulation of Dismissal in Italy.” King’s L.J. 33: 188 sgg.

Freedland, M., Kountouris, N. 2011. The Legal Construction of Personal Work Relations. Oxford: Oxford University Press.

Fudge, J. 2011. “Labour as a ‘Fictive Commodity’: Radically Reconceptualizing Labour Law.” In The Idea of Labour Law, a cura di G. Davidov e B. Langille. Oxford: Oxford University Press.

Gamonal, S., Rosado Marzán, C. 2019. Principled Labor Law: U.S. Labor Law Through a Latin American Method. Oxford: Oxford University Press.

Hiessl, C. 2021. “The Classification of Platform Workers in Case Law: A Cross-European Comparative Analysis.” Comp. Lab. L. & Pol’y J. 42: 465 sgg.

Katsabian, T., Davidov, G. 2023. “Flexibility, Choice, and Labour Law: The Challenge of On-Demand Platforms.” U. Toronto. L.J. 73, 3: 348 sgg.

Langille, B. 2011. “Labour Law’s Theory of Justice.” In The Idea of Labour Law, a cura di G. Davidov e B. Langille. Oxford: Oxford University Press.

Langille, B. 2019. “What is Labour Law? Implications of the Capability Approach.” In The Capability Approach to Labour Law, a cura di B. Langille. Oxford: Oxford University Press.

Nussbaum, M. 2000. Women and Human Development: The Capabilities Approach. Cambridge: Cambridge University Press.

Routh, S. 2014. Enhancing Capabilities through Labour Law: Informal Workers in India. London: Routledge.

Sen, A. 1999. Development as Freedom. Oxford: Oxford University Press.

Sen, A. 2009. The Idea of Justice. Harvard: Harvard University Press.

Waas, B. 2017. “Comparative Overview.” In Restatement of Labour Law in Europe–Vol I: The Concept of Employee, a cura di B. Waas e G. Heerma van Voss, xxvii sgg. Oxford-Portland: Hart Publishing.

Guy Davidov, The Hebrew University of Jerusalem, Israel, guy.davidov@huji.ac.il, 0000-0003-1716-3271

Referee List (DOI 10.36253/fup_referee_list)

FUP Best Practice in Scholarly Publishing (DOI 10.36253/fup_best_practice)

Guy Davidov, Platform Workers, Autonomy, and the Capability Approach, © Author(s), CC BY 4.0, DOI 10.36253/979-12-215-0507-8.20, in William Chiaromonte, Maria Luisa Vallauri (edited by), Trasformazioni, valori e regole del lavoro. Scritti per Riccardo Del Punta, pp. -13, 2024, published by Firenze University Press, ISBN 979-12-215-0507-8, DOI 10.36253/979-12-215-0507-8