Collective bargaining can help workers and companies adapt to the opportunities and challenges of a changing world of work. Even if to some people collective bargaining may sound like a relic of the past, it is an instrument to reach flexible and consensual solutions to regulate the use of new technologies and shape new rights. However, low levels of organization among workers, in particular non-standard workers, pose a serious challenge to collective bargaining. This partly reflects legal obstacles for workers classified as self-employed, for whom the right to bargain collectively may be seen as infringing competition law. To bypass these obstacles and extend bargaining rights to some non-standard workers, some OECD countries have either adapted the labor law or tailored competition regulation.

Andrea Bassanini, Stijn Broecke, Sandrine Cazes, Andrea Garnero, Chloe Touzet1


Workers for platforms such as ride-hailing or food delivery apps are increasingly demanding better working conditions and stronger regulation of the sector in both the United States and the European Union. Strikes, i.e. workers logging off from the app, have recently taken place in the United Kingdom, Belgium, the Netherlands, France, Germany, Australia, and Hong Kong. Despite remaining relatively limited in scope, platform work has become one of the most visible and controversial symbols of the “new world of work.”

While platforms are a recent development, triangular employment relationships are not new. For several decades, it has been commonplace for temporary agency workers to be hired by a staffing agency and assigned to work in another firm. However, a key difference between staffing agencies and platforms is that agency workers have an employment contract, while most platform workers are (rightly or wrongly) classified as self-employed. Having such a status means that those workers may not be covered by the standard rights and protections that employees enjoy. However, even if they are considered to be self-employed on paper, some of these platform workers, and several other workers in less visible but similar situations, have some of the characteristics of employees and, like them, might be in an unbalanced power relationship when employers have a higher degree of control over the employment relationship than they do.

Why are unbalanced employment relationships (which are not limited to platforms) a problem? Is it just a question of social justice? No, it is also an issue of economic efficiency. When workers have limited options (because they work in areas with few employers, or have limited possibilities to move, or their skills are not easily transferrable) and are unable to negotiate their working conditions, employers can inefficiently suppress wages and employment – a situation usually referred to as labor market monopsony.

Several public authorities, from national governments to city councils, are scratching their heads to find a balance between ensuring good working conditions for all workers and a level-playing field for all companies while not restraining new forms of business that offer job opportunities to workers that may have not found another one. Some are considering stricter regulation of working hours as well as minimum pay. Others are pushing platforms to self-regulate through codes of conduct or charters.

Tribunals are also struggling with the problem. Many court cases have weighed in on the debate in several countries, in particular to disentangle whether platform workers are really self-employed or just in a disguised form of standard, salaried employment. In several countries, platform workers have been re-classified as standard employees. In others, the self-employed status of platform workers has been validated. In Europe, the European Court of Justice (in Case C-434/15) even ruled on the nature of the business itself, holding that Uber acts as a transportation service provider rather than a mere technological intermediary and therefore has duties and responsibilities towards its workers that a simple intermediary would not have.


Beyond government interventions, social partners can also find solutions on their own via collective agreements. Even if for many people collective bargaining and organized labor sound like relics of the past, they provide a form of self-regulation which can prove more flexible and tailored than one-size-fits-all solutions through legislation. However, as shown by the figure below, non-standard workers are 50 percent less likely to be unionized, on average, than standard employees.

This lower level of unionization partly reflects the practical difficulties of organizing non-standard workers (who might be more fearful of retaliation when joining a union and/or be less attached to a particular workplace because they frequently change jobs), as well as the fact that collective bargaining historically developed around standard employees.

It also results from legal obstacles to collective bargaining for some non-standard workers such as the self-employed. If ILO Convention 98 on the right to organize and bargain collectively refers to “workers” in general, in practice, the right to bargain for non-salaried workers is subject to legal discussion as possibly infringing antitrust rules. As illustrated in the chart below, while salaried workers face only practical difficulties in exercising their collective rights, workers in the “grey zone” between the usual definitions of employee and self-employed, where genuine ambiguity exists about their employment status, as well as genuine self-employed workers, who might nonetheless be in unbalanced power relationships with their employer/client, may also be barred from bargaining collectively due to laws prohibiting cartels, which tend to consider them as “undertakings.” The argument that collective bargaining for self-employed workers is incompatible with competition law has also been used in the United States by Uber to challenge a 2017 ordinance by the City of Seattle that allowed drivers to unionize and bargain together. In the European Union, the European Court of Justice in 2014 ruled that only “false self-employed” workers are not to be considered undertakings for the purpose of competition rules while the “other self-employed” should continue to be seen as “enterprises” and therefore barred from bargaining collectively.

Access to collective bargaining for different forms of employment, current situation

Source: OECD Employment Outlook 2019, OECD Publishing, 2019.

The European Union Commissioner for competition, Margrethe Vestager, has recently called for “gig economy” workers to be allowed to collectively bargain for their rights. “We need to make sure that there is nothing in the competition rules to stop those platform workers from forming a union, to negotiate proper wages as you would do in any other business,” Ms. Vestager said.2 How can this be done in practice?

First and foremost, by effectively enforcing existing labor regulation. In some cases, some of these workers are simply “falsely self-employed,” i.e. they have working arrangements that are essentially the same as those of employees but contracts define workers as self-employed in order to avoid regulations, taxes, and unionization. In the United States, the Department of Labor has estimated that between 10-30 percent of workers are misclassified, and that this could have a significant impact on tax revenue.3 Enforcing the correct classification would also give access to collective bargaining to these workers. Fighting misclassification may require revising and/or harmonizing definitions of employee and self-employed statuses as well as strengthening regulatory enforcement by facilitating third-party actions and reducing the costs of filing classification complaints for workers.

However, no matter how effective enforcement is, there will always be workers who are genuinely difficult to classify and will belong to a grey zone between dependency and self-employment. Independent contractors and freelancers such as ride-hailing drivers and delivery workers often fall into this category. This occurs because they simultaneously share some of the characteristics of employees (e.g. they cannot set their own rates of pay, cannot be replaced in executing their tasks by someone else, are financially dependent on one main “customer”) and some others of the self-employed (e.g. they can choose when to work; they use their own equipment; they can work for competing platforms). As a consequence, the factors relevant to the tests used by courts to determine whether a worker is an employee or not point in different directions, which typically leads courts to conclude that the worker is not an employee, and therefore must be considered self-employed. Yet, these workers usually are in the same unbalanced power relationship with their “employer(s)/customer(s)” as employees, and evidence suggests that they are even more exposed to monopsony4 than standard employees. The consequences of power imbalances on pay and employment levels tend to be stronger when workers are unable to organise and bargain collectively. When workers negotiate individually (or do not negotiate at all), employers’ buyer power is usually not offset by sufficient bargaining power on the side of workers.


There is, therefore, a strong argument in favor of extending collective bargaining rights to workers in the grey zone. The main difficulty is to identify some criteria for providing access to collective bargaining to avoid giving unregulated freedom to own-account self-employed workers – that is, self-employed without employees – to form cartels (even small ones), as this could have clear negative consequences for consumer welfare. In particular, the challenge is to avoid that extending bargaining rights would threaten the effectiveness of the cartel prohibition in competition law – for example by risking giving de facto a blanket authorization to cartels of true business undertakings, which typically face multiple consumers (e.g. plumbers agreeing to divide a geographical market among themselves).5

There are two approaches to providing bargaining rights to workers who find themselves in an unbalanced power employment relationship. The first consists in tailoring labor regulations to define ex ante the categories of workers who have access to bargaining right; the second touches competition regulations.

Some OECD countries have followed the first approach by defining specific categories of workers in the “grey zone” with clear boundaries, such as the dependent self-employed (own-account self-employed who are financially dependent on one client). Since the mid-1960s in Canada, collective bargaining legislation at the federal level (and in several provinces) has explicitly defined “dependent contractors,” allowing for their inclusion in the same bargaining unit as permanent full-time employees and usually with the same collective agreements (while it is uncommon for dependent contractors to have a separate collective agreement from permanent employees, this is legally permissible). In a similar vein, dependent contractors in Korea, parasubordinati in Italy, Arbeitnehmerähnliche Personen in Germany, or TRADE in Spain are included in collective bargaining (or in the case of Spain they can sign specific “professional interest agreements,” acuerdos de interés profesional) even if not formally employees.

When the categories are too narrow, this approach risks leaving out most workers in unbalanced power relationships with their de facto employers. Other countries have thus developed categories with less defined boundaries, such as “person working for money” in Poland or “workers” in the United Kingdom, who are simply defined as individuals who work under a contract to provide a personal service. In these countries, these categories have a number of employment rights, including access to collective bargaining. While being more inclusive, this solution nonetheless leaves room for judicial uncertainty and may create opportunities for downgrading employees into such a third worker category.

Another option consists in inverting the logic of the multi-factor tests for specific protections. A worker is then considered an employee, with respect to those protections, except if it can be proved that she is a self-employed. This route is followed by a number of U.S. States as regards liability to unemployment benefit contributions and, less frequently, overtime pay and minimum wage. California, through the AB5 legislation,6 has recently enacted this principle for all protections defined in its labor code. Other countries are considering similar moves.

All these models for the extension of rights and protections through labour law draw a clear line between workers and business undertakings. Therefore, if used for extending access to collective bargaining, in particular at the company level, they have the advantage of not giving a blanket authorisation to cartels of undertakings facing multiple consumers.

A second, complementary, approach consists in tailoring competition regulations. Such objectives could be pursued either by adopting a pragmatic approach vis-à-vis groups of self-employed workers most exposed to unbalanced power relationships or by introducing explicit legal exemptions from the enforcement of the prohibition to bargain collectively.

In many cases, regulators and enforcement authorities have taken a case-by-case approach to avoid a strictly procedural analysis of cases involving those workers with little or no bargaining power and exit options. Moreover, in several countries (e.g. in France, Italy, and Spain), independent unions of platform workers de facto negotiate working conditions for their members even if they are classified as self-employed without any intervention from national antitrust authorities. The risk associated with this approach is that it potentially creates uncertainty since it could be reversed without any legislative reform.

Few OECD countries have introduced explicit exemptions to the cartel prohibition for certain forms of self-employed sectors or occupations. In 2017, the Irish Parliament amended the Competition Act to include voice-over actors, session musicians and freelance journalists among the occupational categories that have the right to negotiate. Furthermore, it also opened the possibility to access collective bargaining for “fully dependent self-employed” and not only “false self-employed” workers.7 Under Irish law, trade unions have to apply for the exemption, prove that the workers they want to represent fall in one of these two classes, and show that their request will have “no or minimal economic effect on the market in which the class of self-employed worker concerned operates,” and not “lead to or result in significant costs to the State.” The 2017 Irish amendment has attracted many criticisms and is currently being debated at the International Labour Organization (“ILO”). Irish employers as well as the International Organization of Employers, on the one hand, expressed their concern about the lack of clarity in the criteria used to identify “fully dependent” and “false” self-employed workers. They also contested the lack of employer consultation in determining the application of those criteria – currently the law states that the Government must make the decision in consultation with a trade union only. On the other hand, those in favor of extending bargaining rights to self-employed workers experiencing power imbalance find the dependency criteria too stringent (a platform worker can work for more than two platforms and still be economically dependent). The condition of “no or minimal economic effect on the market” may also be a potentially insurmountable practical limit for workers.

The Australian Competition and Consumer Act also allows businesses to collectively negotiate with suppliers or customers if the Australian Competition and Consumer Commission considers that collective bargaining would result in overall public benefits. The Australian Competition and Consumer Commission is currently undertaking a public consultation process regarding the creation of a class exemption for collective bargaining by small businesses (including independent contractors). A class exemption for collective bargaining would effectively provide a “safe harbor,” so businesses that met eligibility criteria could engage in collective bargaining without breaching competition law and without having to seek approval from the Australian Competition and Consumer Commission.

Legal exemptions for specific categories of self-employed workers also exist in other OECD countries. For example, in 1996, the U.S. Department of Justice and Federal Trade Commission jointly ruled that physician networks which “collectively agree on prices or price-related terms and jointly market their services” do not infringe anti-cartel regulation provided that “they constitute 20 percent or less of the physicians in each physician specialty in the relevant geographic market” – 30 percent if physicians are not prevented from belonging to multiple networks.

In practical terms, targeted exemptions by sector or occupation are not always easy to define and apply; the list may need frequent updating, and the potential reversal of exemptions is a source of legal uncertainty for workers and businesses alike. For instance, in 2010 in New Zealand, following an industrial dispute in the film industry, the government passed an amendment to the Employment Relations Act effectively preventing all workers in the film industry (considered independent contractors) to enter into collective bargaining. The current government has declared its intention to restore the right to engage in collective bargaining for film industry workers.

In addition, as outlined before, small cartels can induce suboptimal outcomes for consumers. For that reason, any exemption aimed at granting bargaining rights to self-employed workers in situations of power imbalance should be based on a comprehensive costs-benefits analysis. One way to focus on workers in real need of access to collective bargaining would be to prioritize exemptions to those groups of self-employed workers that are likely to have few outside options.

A more radical approach to ensure that all self-employed workers experiencing power imbalance have the right to negotiate their own terms of employment – with no precedent in OECD countries and in conflict with most existing regulations – is discussed in the academic literature8 and among trade unions.9 This consists in reversing the current presumption that self-employed workers do not only provide labor but also services by means of an independent business organization that they actually own and manage – which justifies their exclusion from collective bargaining. Under this approach, the burden of proof would be shifted onto those who propose the restriction, in particular regulation enforcement authorities. The main argument used in support of this approach is that “the right to bargain applies to all workers with the sole possible exception of those explicitly excluded by the text of ILO Convention No. 87 and No. 98” (notably, armed forces and the police) and “self-employed workers are not among those excluded and, therefore, the Conventions are deemed as fully applicable to them.”10 However, not all OECD countries have signed and ratified the ILO Conventions. Moreover, a reversion of the burden of proof would conflict with most existing antitrust regulations and it would likely increase the burden for antitrust authorities that would have to check ex post the validity of a large number of agreements. Moreover, while aimed at ensuring that all workers in unbalanced power relationship are covered, the inversion of the burden of the proof may be exploited more effectively by relatively stronger and more organized groups of workers.


In conclusion, the development of new forms of work offers the opportunity for legislators to reflect on possible adaptation to labor and competition legislation, which were designed in a world where the two opposed poles of salaried work and self-employment stood far apart. In large part, this issue is not new – but the platform economy has shone new light on it. Addressing the issue of worker classification to ensure that employment contracts match the real nature of the employment relationship is a first necessary step. However, regulators and enforcement authorities also need to reflect on how workers in the grey area between dependency and self-employment and those self-employed in situations of strong power imbalance vis-à-vis their client/employer can be empowered to negotiate and organize collectively.11

1Andrea Bassanini is a senior economist and editor of the OECD Employment Outlook at the Directorate for Employment, Labour and Social Affairs of the Organisation for
Economic Co-operation and Development (OECD) in Paris, France. Stijn Broecke is a senior economist at the OECD and leads the team on the future of work. Sandrine Cazes is a senior economist at the OECD and leads the team on collective bargaining. Andrea Garnero and Chloé Touzet are economists at the OECD in the team of collective bargaining.
This article is based on work done by the authors for two OECD publications: OECD Employment Outlook 2019 and Negotiating Our Way Up: The Role of Collective Bargaining and Workers’ Voice in a Changing World of Work. The views expressed here are those of the authors and cannot be attributed to the OECD or its member countries.

2 See

3 Brumm, F. (2016), Making Gigs Work The new economy in context, University of Illinois – Urbana Champaign.

4 M. Keith Chen, Judith A. Chevalier, Peter E. Rossi & Emily Oehlsen (2019), “The Value of Flexible Work: Evidence from Uber Drivers,” Journal of Political Economy, vol. 127, no. 6.

5 U.S. v. Joseph P. Cuddigan, et al.,

6 See

7 The Irish law defines precisely the two cases: A “false self-employed worker” is an individual who (a) performs for a person the same activity or service as an employee of the other person, (b) has a relationship of subordination, (c) is required to follow the instructions of the other person regarding the time, place and content of his or her work, (d) does not share in the other person’s commercial risk, (e) has no independence as regards the determination of the time schedule, place and manner of performing the tasks assigned, and (f) for the duration of the contractual relationship, forms an integral part of the other person’s undertaking. A “fully dependent self-employed worker” is an individual (a) who performs services for another person (whether or not the person for whom the service is being performed is also an employer of employees) under a contract (whether express or implied, and if express, whether orally or in writing), and (b) whose main income in respect of the performance of such services under contract is derived from not more than two persons (Competition (Amendment) Act 2017).

8 See, for instance, B. Creighton & S. McCrystal (2016), “Who is a ‘Worker’ in International Law?,” Comparative Labor Law and Policy Journal, Vol. 37/3, pp. 691-725 and V. De Stefano & A. Aloisi (2018), “Fundamental Labour Rights, Platform Work and Human Rights Protection of Non-Standard Workers,” in Bellace, J. & B. ter Haar (eds.), Labour, Business and Human Rights Law, Edward Elgar Publishing Ltd.,

9 Fulton, L. (2018), Trade Unions protecting self-employed workers, ETUC, Brussels.

10 Quotes from pages 14 and 15 in V. De Stefano & A. Aloisi (2018), “Fundamental Labour Rights, Platform Work and Human Rights Protection of Non-Standard Workers,” in Bellace, J. & B. ter Haar (eds.), Labour, Business and Human Rights Law, Edward Elgar Publishing Ltd.,

11 For further reading, see the OECD, “Employment Outlook, 2019,” available at, and OECD, “Negotiating Our Way Up: Collective Bargaining in a Changing World of Work,” available at