Adopting a Universal Mandate on Platform Work: Balancing Contrasting Realities

1. Case Law

This research identified and reviewed 1,338 judicial and administrative decisions spanning 39 countries in the global North and South regions. Out of these 1338 cases, 892 were first-instance cases. The remaining cases correspond with further instances in the process.

The claims brought by various actors to judicial or administrative authorities exhibit substantial similarities and parallel characteristics across jurisdictions, indicating the relatively standardized employment practices of digital labour platforms worldwide. The literature has described such practices in depth as part of the business model of platform companies.²²

These claims were submitted before various bodies and procedures, including labour inspections, social security institutions, tax authorities, competition and data protection authorities, civil and commercial courts, specialized employment courts, state prosecutors, criminal courts, and constitutional courts.

The most common claim found in the mapping was the request for recognition of the employment relationship, followed by claims regarding social security entitlements (see Figure 1). In this sense, the majority of the cases are centered on misclassification, confirming the widespread use of regulatory arbitrage in the platform economy.²³

Figure 1

Some exceptions were identified, where platform workers were employed at a relatively early stage of their mobilization cycle or the use of regulatory arbitrage was much more nuanced, predominantly in the GN. Nevertheless, other conditions of precarity, such as a lack of work equipment or fragmented employment relationships, persisted, as illustrated by the examples below.

In Germany, for example, couriers filed two lawsuits against ‘Lieferando’ with the support of the ‘Gewerkschaft Nahrung-Genuss-Gaststätten’ – NGG. These couriers were engaged with the platform through an employment contract and demanded that the platform provide them with the necessary work equipment for the agreed-upon activity as a bicycle supplier. The ‘Federal Labour Court’ ruled in favour of the plaintiffs in both cases, stating that they were entitled to the defendant providing the necessary work equipment (i.e., a roadworthy bicycle and an internet-capable mobile phone). The contractual agreement, which stated that the plaintiff must use their own bicycle and mobile phone, was deemed ineffective by the judge.²⁴

Similarly, a lawsuit was lodged by the ‘Swedish Transport Workers’ Union’ against ‘Foodora’ for the dismissal conditions of a worker. In this case, ‘Foodora’ employed the bicycle courier on a fixed-term basis; however, when the worker started carrying out his deliveries on a motorbike, he was referred to sign a contract with an intermediary staffing company while continuing to work for ‘Foodora’ under substantially similar conditions. Thus, a dispute surfaced as to whether the worker should be considered to have remained employed by ‘Foodora’, whether the employment was for an indefinite duration, whether he had been dismissed from that employment, and whether there was a factual basis for dismissal. The Court decided in November 2022 that ‘Foodora’ had no employer liability, and that the intermediary should be considered a staffing company.²⁵

A wide range of secondary issues was often identified and addressed in the context of misclassification claims, as shown in Table 3. It is essential to note that the misclassification issue was not uniform or one-dimensional, as it involved various shades of disguised relationships, including self-employed workers, outsourcing chains, complex business structures, and temporary agency work arrangements. Hießl has thoroughly researched these multiparty relationships in platform work within a European context.²⁶

In turn, complaints related to social security contributions were the second most common claim globally. Although social security claims are closely connected to the issue of misclassification, it is essential to note that this category refers to lawsuits where the recognition of social security entitlements was the primary object of the claim, rather than procedures where social security contributions were recognized as a consequence of pursuing the recognition of employment status. These were almost entirely filed in the global North, as observed in Table 4. Nearly half the cases that pursued this claim (47.65%) were brought by state institutions and/or collective actors (see Table 5).

Platform companies themselves were also common actors in bringing this claim to court after administrative authorities—such as Labour Inspections, Social Security Institutions, and Tax Authorities—issued an administrative decision ordering the recognition and payment of tax or social security contributions to the platform company (Table 6).

This claim was not common in the GS. This may be influenced by welfare states that recognize social security entitlements to their citizens, which are predominant in the GN, together with the low involvement of state institutions/collective actors identified in the mapping and the limited coverage of social security systems in the GS,²⁷ leading to a low resort to litigation in this region.

Other claims presented independently of the misclassification issue included the applicability of collective agreements, violations of data protection principles, and constitutional matters. Excluding the latter, these claims were preponderantly filed in the GN. (see Table 4)

Procedures requesting the applicability of a collective agreement tended to be Eurocentric. Given that social dialogue initiatives in the GS were much less common, this confirms that, as some literature suggests, collective agreements as an institutional tool are generally stronger in the GN (and more specifically in Europe).²⁸ This claim was most prominent in the Netherlands, Spain, and Italy. Often, it involved workers or collective actors—who were logically important actors in these procedures—requesting that a specific collective agreement be declared applicable to a platform company (for example, the collective bargaining agreement for the Taxi industry in the Dutch case). In most cases, the platform company was not applying any collective agreement. However, some Spanish examples were interesting because the actors requested the applicability of the collective agreement for logistics rather than for the courier sector, showcasing an interesting innovation: these workers were already covered by a collective agreement, which likely reflects the influence of the Riders’ Law in Spain.

Regarding violations of data protection principles, this claim mainly refers to the use of the existing data rights framework—primarily the GDPR—to address breaches of platforms’ obligations to disclose details about algorithmic management processes and automated decision-making. Therefore, it is not surprising that these cases are most common in Europe, as the GDPR has limited jurisdiction within the EU member states. In the absence of specific labor law provisions applicable to self-employed workers, the broad scope of data rights—which extend to all individuals regardless of employment status—has provided alternative legal channels for establishing protections against exploitative platform practices. At the time of writing, the number of cases involving this issue continues to grow within the European context, positioning itself as a promising tool to defend the working conditions of platform workers and the broader workforce. These decisions mark significant progress, reflecting a shift in focus from employment classification to data transparency and access to information as ways to better understand organizational models and limit the expansion of employers’ discretion.²⁹

In turn, concerning procedures that claimed constitutionality violations, these were most commonly presented against the enactment of a law specific to platform work (e.g., against Proposition 22 in the USA, the Loi relative à la relance économique et au renforcement de la cohésion sociale du 18 juillet 2018 in Belgium, or the law banning platform-based transportation in Brazil). Such procedures all requested that the law be declared unconstitutional/inapplicable. This claim was also commonly brought in response to the deactivation of workers’ account, mostly connected to the fundamental right to work.

A peculiar example is the Colombian jurisdiction, where, although only one case has reached a decision by the labour courts, a handful of decisions have been examined by the constitutional jurisdiction. This may be explained by the fact that the constitutional jurisdiction in Colombia is subject to an abbreviated procedure. Notably, in judgment T-109/2021,³⁰ the highest court in Colombia for constitutional matters declared the existence of an employment contract for an adult webcam content creator who provided her services through a digital labour platform. The court considered the adverse consequences imposed for not logging into the platform, the constant monitoring, the control over clients’ payment methods, and the power to review complaints about content uploaded as crucial elements in reaching this decision.³¹ Other decisions in the constitutional jurisdiction referred to the platform’s disregard for the right to due process when deactivating a worker’s account. (T-534/23³² and No. 11001418902120210087800³³). In addition to constitutional jurisdiction, lawsuits were also brought to criminal courts. Examples were identified in jurisdictions such as France³⁴ and Italy,³⁵ where criminal procedures were brought forward against platforms for the engagement of illegal migrants.

Lastly, procedural claims involve procedures where the main focus of the decision is of a procedural nature, such as jurisdictional conflicts, the enforceability of the arbitration clause, the certification of a class action, or other procedural issues. The debate over the enforceability of arbitration clauses was most prominent and supported by courts in Anglo-Saxon systems, such as the USA and Canada, often resulting in the arbitration clause being upheld. Jurisdictional disputes were also common, as platform companies frequently argued that the labour court lacked competence. This argument was often accepted in French and Brazilian courts, leading to many claims being dismissed. Overall, although some judges have moved beyond formalistic approaches and have argued against procedural tactics used by platforms, they still face significant obstacles, as seen with arbitration clauses in the USA and jurisdictional disputes in Brazil, which often prevent courts from addressing substantive issues.³⁶

Despite the clear prevalence of misclassification claims in general, when examining the cases brought to court by collective actors, as observed in Figure 2, over 53% of them involve other grievances. These include the applicability of the collective agreement, data protection, and claims related to algorithmic management, work and safety equipment, complaints against anti-union activity, the need to reactivate accounts during COVID-19 times to guarantee subsistence, or requests to guarantee their right to work when the platform was suspended from operations.

Figure 2

This suggests that collective actors are resorting to litigation for issues beyond employment status, possibly using it strategically to enforce smaller claims. As argued by Gaudio, collective litigation is advantageous over individual one as it enables the enforcement of small claims where many workers have been affected homogenously by the same or analogous decision-making processes, as often happens with algorithmic management devices. In these situations, damages may be small for each worker, and bringing an individual claim would be meaningless for them. However, since the sum of individual damages may be significant if the rights of many workers are infringed, this potential gap may be filled through collective litigation.³⁷

In sum, misclassification was the main issue leading to litigation in both the GN and the GS. Another common claim in both regions involved platform workers requesting the reactivation of their blocked accounts. Social security issues were much more prevalent in the GN and were mostly brought forward by the state and collective actors. This claim was most prominent in welfare states such as Switzerland, Spain, France, and South Korea, where social security institutions likely had to bear some costs because these workers were not covered by insurance. Conversely, it was rare in the GS. This aligns with the fact that, although social security systems in the Global South are expanding, coverage remains limited (confined to small sections of the population that are in formal employment), leaving billions without protection.³⁸ Regardless, landmark cases such as the lawsuit filed by the ‘Sindicato dos Trabalhadores com Aplicativos de Transporte Terrestre Intermunicipal do Estado de Sao Paulo,’ which requested safety equipment for workers during the COVID pandemic, were identified.³⁹

Other claims, such as GDPR, the applicability of collective agreements, and the exercise of information and consultation rights, have mostly been the subject of litigation in the GN. This may be attributed to the fact GN jurisdictions have data protection laws, like the EU’s GDPR. These issues were not frequently brought to court in the GS, likely because several—though not all—GS countries lack suitable laws for the protection of personal data. In this region, litigation primarily involved the reclassification of employment contracts or constitutional claims, seeking the immediate protection or restoration of fundamental rights. Examples include reactivating workers’ blocked accounts to guarantee the right to work, protecting the fundamental right to freedom of association, accessing medical care due to an industrial accident, or challenging the validity of laws that specifically regulate or ban platform work.

Regarding the outcomes of litigation, when assessing the direction of the decisions analysed per jurisdiction, the results are scattered, as observed in Figures 3 and 4. Specifically, regarding the issue of employment status (see Figure 3), a group of countries exhibits a tendency towards recognizing some labour rights for platform workers. This is achieved either by acknowledging them as employees or by categorizing them as a third category, as specified in national law – e.g., “workers” in the UK, or “lavoro etero-organizzato” in Italy-. A second group of jurisdictions displays a certain inclination towards the status of independent contractors. Finally, there are jurisdictions where the decisions are so dispersed, and the ratio is so close that it is not possible to establish a tendency. (A more comprehensive table is available as an addendum to this paper.)

Figure 3

Figure 4

Workers were more successful in bringing their claims in countries like Spain, Italy, and the Netherlands. In turn, judicialization of the conflict resulted in less favourable outcomes in Brazil and Australia. No clear trend can be identified in countries such as Austria, New Zealand, Nigeria, and South Korea.

Examining the courts’ positions exclusively regarding the classification of platform workers reveals a marked tendency towards employment status in the Netherlands, France, Spain, and Argentina. It should be noted that, at the time of writing, some decisions issued by the Dutch courts following the last update of the mapping have been less favourable to the Dutch trade union (e.g., the Uber and Temper cases), marking a step back from the precedent that recognized platform workers as employees.

In contrast, a tendency in favour of independent contractor status is seen in countries such as Australia and Brazil. However, the trend appears to be shifting in Brazil –and is presumably going to continue to do so with decisions like the one from the Supreme Court favouring the employment status of Uber drivers in 2022–.⁴⁰ Finally, courts have favoured an intermediate category in jurisdictions such as Italy and the UK.

These ongoing developments show that this topic remains a moving target, and even in jurisdictions where the position once appeared settled (such as the Netherlands or Brazil), the judicial debate over the misclassification of platform workers is still unfolding.

There are also jurisdictions where a clear trend could not be identified due to a scarcity of decisions combined with a lack of uniformity (e.g., Belgium or Chile). The absence of cases may hinder a shift in the judges’ position. Examples in jurisdictions with a higher number of cases (NL, ES, IT, BR) show that, although initially judges favoured the classification of platform workers as independent contractors, the evolution led to a shift in the judiciary’s opinion.

Initially, judges had a superficial understanding of the platform business model and how algorithms work;⁴¹ however, they later gained knowledge and improved their understanding, as shown in the reasoning behind their decisions.⁴² Examples of this shift can be seen in jurisdictions worldwide, both in the GN and GS.⁴³ Naturally, this change is not exclusively shaped by the judiciary and is also influenced by other sources of power, such as legislative reforms that will be discussed below (institutional power) and societal pressure to improve the working conditions of this workforce (ideational power).

However, this shift requires cases to reach the courts, which, in several jurisdictions, platforms have hindered by employing a practice that could be termed “strategic conciliation” or “manipulative conciliation.” ⁴⁴ This practice was identified by national experts such as Palomo in Chile⁴⁵, de Sena et al in Brazil⁴⁶ and Cherry in the USA.⁴⁷ Although several lawsuits were lodged, they never reached a judgment, as digital labour platforms are alleged to have resorted to settlements on repeated occasions to avoid adverse legal precedents from arising. This practice seemed pivotal in avoiding the consolidation of an adverse judicial precedent in some jurisdictions, such as Brazil, Australia, Chile, and the USA, where the general direction of the jurisprudence remains unfavourable or unclear at best.

2. Social Dialogue

The mapping conducted identified 95 social dialogue initiatives in the platform economy. This is significant considering the widespread use of regulatory arbitrage in the platform economy,⁴⁸ among other obstacles faced by platform workers that foster a barren environment for collective representation and social dialogue.⁴⁹

As observed in Table 1 (supra), the higher presence of institutional responses in the GN is most noticeable in social dialogue, where over 88% of the initiatives identified are located in the GN, primarily in Europe. In the GS, although scarce, most initiatives were identified in Latin America.

Regarding the material scope of the agreements, Figure 5 shows the topics addressed and key provisions included in the different social dialogue initiatives.

Figure 5

Remuneration was the most frequently regulated issue in social dialogue. It has been a highly contested issue in some agreements, for instance, in the French and Italian systems, as it has raised questions about whether it has been disadvantageous for the workers.

The French example refers to an income agreement reached on April 20, 2023.⁵⁰ This agreement was finalized between the ‘Association des Plateformes d’Indépendants’ (API), which represents platform companies, and several new workers’ organizations.⁵¹ This income agreement has faced heavy criticism from traditional unions such as ‘CGT,’ which argues it sets the delivery drivers’ hourly pay at 11.75 euros per hour of actual work. Representatives of ‘CGT’ claim this rate applies only during the ordering process. However, waiting times are not considered part of working hours. A representative from ‘Uber Eats’ clarified that waiting at the restaurant and the travel time to the restaurant are included.⁵² Nonetheless, this does not account for all the waiting time workers experience, as over half (52%) of delivery workdays involve waiting because the platform does not assign enough orders.⁵³ Therefore, delivery workers often wait long periods for orders, and this waiting time is not covered by the agreement. In Colombia, a very similar pay scheme, included in an agreement between Rappi and Unidapp, stipulates a minimum pay of 3,050 Colombian pesos per delivered order.

In turn, in the Italian case, the ‘Assodelivery’ and ‘UGL’ collective agreement recognized in Article 7 that riders were classified as self-employed. Moreover, Article 10 stated that riders will be compensated based on the deliveries completed. This was agreed upon despite platform workers being covered by employment protections, including a statutory minimum hourly wage (Law No. 128/2019, known as the Riders’ Decree). This agreement was considered unlawful by the Court of Florence, ruling No. 781/2021, as the court found ‘UGL’ acted as a “yellow union”.

These collective agreements have raised questions as they were deemed detrimental to the workers. The underlying issue in these agreements is that of unpaid labour performed by platform workers, as the agreements establish a pay-per-delivery instead of a minimum hourly wage that recognizes the entire time it takes the worker to perform their job.

Following remuneration, although not related to a specific issue, agreements aimed at regulating the entire sector were the second most common. Examples include the ‘Logistics National Collective Bargaining Agreement’ that applies to ‘Just Eat’s’ workers in Italy, the LO and Foodora sectoral collective agreement in Sweden, the agreement reached by the ‘Sindicato de Comercio’ with ‘Pedidos Ya’ for drivers to be covered under the collective agreement for catering services in Argentina, the sector-based Memorandum of Understanding for ride-hailing companies in Kenya, and the sectoral agreement for delivery drivers adhered to by JustEat in Denmark.

Ilsøe and Söderqvist analysed two of these sectoral agreements, namely JustEat in Denmark and Foodora in Sweden, when examining the Nordic model in the platform economy. They assessed that there is still a long way to go as most platforms do not adhere to the agreements, and choose to remain outsiders to the system. The scholars acknowledge that both food delivery platform companies hired platform workers as employees, engaged in collective bargaining, and voluntarily shared rulemaking capacity with organized labour, ultimately becoming first movers and shaping future sectoral norms, thereby exercising significant rulemaking capacity. However, they also noted that a sectoral agreement with a single signatory company does not meet the requirements of full Nordic IR-system integration. Consequently, a pertinent challenge remains integrating competitors such as Wolt, who may insist on continued IR-system evasion due to the existing collective agreement being a poor fit for their business model.⁵⁴ This particular analysis can be applied (with saved proportions) to other sectoral agreements found in the platform economy, as most of them share the condition of being concluded with a single signatory company, with multiple competitors choosing to remain outside the system. Therefore, these agreements, in general, although exhibiting some progress in the working conditions of workers on specific platforms, remain with many challenges ahead for the sector.

The different social dialogue initiatives also frequently included provisions on occupational health and safety, automated decision-making, social dialogue, social security, and working conditions.

Geographically speaking, most agreements identified in the GS were in Latin America, mainly in Chile, Colombia, and Argentina. The main issues addressed in said agreements were remuneration, social security, deactivation procedures, representation, and OSH.

Traditional and grassroots unions were important actors in negotiating collective agreements in the GS, the latter having a role just as significant as the former. The state also had a prevalent role in this regard in the GS vis-à-vis the GN. Although the state participated and promoted social dialogue in the platform economy in the GN as well -e.g. the Charter of Fundamental Rights of Digital Labour in the Urban Context with the intervention of the municipality of Bologna in Italy and the Voluntary Charter that included ten road safety principles that aim to keep motorcycle couriers and other road users safe in the UK, with the participation of Transport for London- in relative numbers, it displayed a minor role compared to the GS. In the GS, the state’s participation was decisive in the agreements signed in Kenya, Colombia, and Argentina. This exhibits that collective actors and industrial relations systems in the GS are more reliant on state-given institutional power, whereas in the GN, social partners and IR systems retain a higher degree of autonomy.

The (even greater) lack of power and autonomy of platform workers and their representatives within social dialogue processes in the GS is reflected both in the voluntary nature and content of the scarce initiatives concluded. For instance, the memorandum of understanding between riders and several ride-hailing companies in Kenya was a non-binding instrument that did not result from a collective bargaining process.⁵⁵ In turn, in Colombia, the platform workers’ trade union ‘Unidapp’ was founded in 2020 with the support of the traditional organization ‘Central Unitaria de Trabajadores’. The delivery platform ‘Rappi’ had been refusing to recognize or negotiate with ‘Unidapp’, leading to ‘Unidapp’ filing a complaint before the Ministry of Labour.⁵⁶ The intervention of the Ministry of Labour, which facilitated the parties’ bargaining, ultimately led to the conclusion of a first agreement in February 2024 addressing issues of minimum remuneration, account deactivation, and union recognition, despite the workers still being regarded as independent contractors. In the country’s first-ever agreement in the platform economy, ‘Rappi’ recognized ‘Unidapp’ as a trade union. The involvement of the Ministry of Labour was a crucial factor in reaching this agreement.⁵⁷ Both the Kenyan and the Colombian agreements classify the workers as self-employed. Similarly, the conditions under which a company-level collective agreement was reached in Chile suppose a rare case, unlikely to be replicated, as it was concluded under the wings of a domestic start-up (’Cornershop‘) that employed its workers and fostered collective dialogue before ‘Uber’ -which holds an opposite position- acquired the company.⁵⁸

Another notable development in the social dialogue scene is the role of global unions, such as the International Transport Workers’ Federation (ITWF), which has signed two framework agreements with Uber: the Memorandum of Understanding and the Global Safety Charter for Couriers. Said agreements paved the way and subsequently led to national affiliates of the ITWF signing national agreements at the domestic level, namely in Belgium, the UK, and Spain.

It is worth noting, however, that these agreements are not always concluded through the traditional collective bargaining procedure, and furthermore, the issue of employment status is not typically discussed. Both parties to the agreement acknowledge they do not see eye to eye on the issue of the classification of the platform workers but express their commitment to work for the improvement of other aspects, such as the right to freedom of association and bargaining, flexibility, and open access to the uber application, occupational health and safety, working conditions, social protection and dispute mechanisms which include deactivation and data rights.

Despite the progress made by this worker’s organization in reaching a macro-level agreement, these efforts have so far only produced effects in the GN. Affiliate organizations to the ‘ITWF’ in Spain,⁵⁹ Australia,⁶⁰ the UK,⁶¹ and Belgium⁶² have reached agreements at the domestic level. This has been much harder to accomplish in some jurisdictions in the GS, as exemplified by an African case. Notably, in Kenya in 2018, members of the ‘ITWF’ attempted to unionize Uber drivers together with the Digital Taxi Forum. However, they were not successful, as drivers preferred the grassroots organizations that had spontaneously formed in the community.⁶³ Literature on platform workers’ organization in the GS shows that the Kenyan example is part of a larger trend of informal collective action, grassroots organizing, and even the creation of alternative platforms in this region.⁶⁴ This informal collective action partially explains the lesser degree of collective bargaining, as research suggests that there is little grassroots action in the platform economy, resulting in practical, long-term improvements in workers’ working conditions.⁶⁵

The minor presence of social dialogue outside of Europe, which was most pronounced in the GS, together with the struggles identified when structures attempt to foster it, such as in the Kenyan case, contributes to reinforcing arguments elucidated by Nowak that stress the Eurocentric focus of traditional trade unions and collective negotiations, which do not have an equally prevalent role outside of Europe.⁶⁶ This also results from the existing legal barriers (observed in both GN and GS regions) that restrict platform workers and the broader self-employed population from accessing structures of worker representation, such as trade unions. (See e.g., APP trade union in Argentina or The App-Based Drivers Union of Bangladesh).⁶⁷

This strategy by global unions such as the ‘ITWF’ managed to get Uber to the negotiation table—a task that itself is an achievement—and resulted in agreements with them, so far within different jurisdictions of the GN. The agreements addressed issues related to working conditions but did not include the issue of misclassification. These agreements can be classified as “staircase agreements.” This concept, initially developed by Ilsøe, Jesnes, and Munkholm in the Nordic context and later recaptured by Lammanis on a broader European scale, refers to agreements designed as a first step, leaving open the possibility for gradual improvements in protections through subsequent renegotiation.⁶⁸

However, not all affiliates and organizations agree with this strategy. Some organizations argue that platform workers are wrongly classified and, therefore, should be granted full employment status with all the consequent rights that come with it. Thus, they consider negotiating below this threshold as watering down the rights of platform workers. A notable example is the FNV in the Netherlands. The biggest trade union in the Netherlands has assumed a position, which became apparent in the aftermath of the collaboration agreement signed between the hospitality division of the ‘FNV’ –‘FNV Horeca’- and ‘Temper’ in 2018.⁶⁹ Subsequently, reports of tensions emerged between ‘FNV Horeca’, ‘FNV ABU’, and ‘FNV Flex’. The latter two departments of the ‘FNV’ perceived the collaboration agreement as undermining the union’s efforts to fight bogus self-employment, as they claim ‘Temper’ operated as a temporary employment agency instead of as a freelance platform, underpaying workers who are falsely defined as freelancers.⁷⁰ It should also be noted that the Dutch industrial relations system is known for its widespread use of the extension mechanism in collective agreements. This instrument provides a solid basis for trade unions to refuse to bargain for lower standards, especially within a sector where a generally applicable standard exists. Arguably, these factors led to a clear focus on the FNV’s use of strategic litigation to combat misclassification, which has so far been successful in its results, as seen in Section 2.1. Nonetheless, it should be noted that at the time of writing, some decisions issued in the Dutch case after the mapping was last updated have been less favourable to the Dutch trade union (e.g., Uber and Temper cases).

Other actors who have engaged with these so-called staircase agreements, only in rare circumstances, can be seen in the Nordic systems. This was the case of the ‘Hilfr’ collective agreement, which was conceived as a first step, leaving the door open to a gradual improvement in protections through subsequent renegotiation. However, as explained by Ilsøe and Söderqvist the exceptional background to the agreement created the premises for this model, such as the fact that the cleaning market in private households was unregulated and affected by a large amount of undeclared work.⁷¹ This is believed to have made it easier for the union to conclude an agreement with less protection than usual (and make it acceptable), ‘a “lesser” agreement being preferable over no agreement’.⁷² However, generally speaking, this constitutes the exception rather than the rule in the Nordic model.

Nevertheless, it must be noted that systems such as those in the Nordic and Dutch regions are known for their strong presence of social dialogue and social consensus.⁷³ Therefore, the higher degree of power of these collective actors and the surrounding systems allows social partners to reject gradual routes, compared to systems assembled differently, predominantly but not exclusively in the GS, where this “staircase approach” could be an interesting alternative.

Overall, reflecting on the approaches observed and the content of the agreements reviewed, it can be assessed that a large number of agreements concluded within the platform economy are negotiated outside traditional collective bargaining systems and procedures. As observed, at times, this led to tensions emerging on whether these gradual approaches are better than nothing at all or, on the contrary, contribute to watering down existing employment protections. It is valid to question whether these agreements are materially improving the working conditions of platform workers in a significant way, or if they are merely temporary fixes that contribute to these companies remaining in the “exceptionalism” on which their business model has been based so far. Although this analysis does not provide a definitive answer to this question, it does propose that the industrial relations system and broader context in which these agreements are taking place are relevant factors for setting a strategy forward.

3. Legislative Reform

As observed in Table 1, legislative reform initiatives were significantly less common in the GS. For instance, Africa has close to no regulation on this matter. Most African nations have not enacted specific laws pertaining to the gig economy, even though there are an increasing number of allegations of abuse linked to the sector.⁷⁴

This contrasted with the global North, where not only more initiatives were identified, but some of them had a wider reach. A quintessential example of this can be observed in Europe, where the PWD has been approved and will be implemented across the 27 EU member states.

Regarding the material scope, among the diverse legislative initiatives identified, the employment status of platform workers was the most frequently addressed subject, followed by initiatives related to social dialogue and workers’ representation. Working conditions, occupational health and safety, and social security were other frequently identified matters. (See Figure 6)

Figure 6

Regarding the status of platform workers and how it has been regulated, the trend is far from uniform, and different stances have been adopted by the countries regulating it.

A block of jurisdictions have introduced presumptions of an employment relationship. These are Belgium,⁷⁵ Spain,⁷⁶ Portugal,⁷⁷ Croatia,⁷⁸ Malta,⁷⁹ and most recently, the European Union.⁸⁰ Although later rendered inoperable with Proposition 22, the state of California introduced the AB5.⁸¹ Argentina has approached this stance; nonetheless, the law has not yet been effectively implemented, and several legal actions have been taken against it.⁸² In turn, Finland’s approach introduced some criteria to distinguish self-employment from dependent employment more clearly.⁸³

In contrast, in other regulatory frameworks, platform workers are classified as independent contractors. This includes several states in the USA, including Utah,⁸⁴ Iowa⁸⁵ and California with Proposition 22,⁸⁶ and the province of Mendoza in Argentina.⁸⁷ Countries such as Greece⁸⁸ and Chile⁸⁹ have introduced two categories of platform workers: those who perform work under an employment contract and those who perform it as independent contractors. Thus, they hold a similar stance and are open to the classification as self-employed.

Finally, there is a group that creates an intermediate category for platform workers, extending them certain rights. France,⁹⁰ Italy⁹¹ have adopted legislation in this sense. However, it remains to be seen how some EU member states, such as Greece and France, will adapt their current regulations in response to the implementation of the new EU directive on platform work.

Broadly speaking, there is a greater emphasis and will to regulate employment status in the GN, especially when contrasted with the GS. On the other hand, the issues regulated in the GS so far pertain more frequently to other matters not directly related to employment status, such as social security (India,⁹² Uruguay⁹³ and Colombia⁹⁴), occupational health and safety -especially during Covid- (Brazil⁹⁵, Peru⁹⁶), working conditions, AI/AM and data protection (Brazil,⁹⁷ Colombia⁹⁸, Nigeria⁹⁹), anti-harassment and discrimination, and even an ordinance that incentivizes platform work (Pernambuco, Brazil¹⁰⁰). Although some legislative projects have been identified in the Latin American region, aiming to regulate employment status, such as in Mexico,¹⁰¹ Colombia,¹⁰² and Uruguay,¹⁰³ the necessary political will has not yet been sufficient for their enactment. The larger presence of informality and unemployment rates in the GS accounts for this regulatory trend to some extent, as regulations seem to adopt a more neoliberal approach. At times, not only by allowing it in the terms proposed by platform companies but by incentivizing it, as seen with Pernambuco.

Another trend identified from a regulatory perspective is the intention to advance algorithmic management, AI, and the rights of platform workers to access their personal data. The EU is leading this matter, with the PWD, which includes an entire chapter establishing new algorithmic management and data rights (most of which also cover self-employed workers)¹⁰⁴ and the AI Act. It could be argued that the former places platform workers in a privileged position, even in comparison to traditional workers, concerning algorithmic management rights. Although research widely exhibits platform workers face considerably more precarious conditions than traditional workers in a large set of labour rights, with regards to algorithmic management, the opposite is the case within the EU, as this Directive provides specific rights such as to receive adequate information about the algorithms used to hire, monitor and discipline them, a right to request human review, inter alia, applicable specifically to persons performing platform work. There is also a growing interest in this issue, as identified in the GS, with examples in Brazil,¹⁰⁵ Nigeria¹⁰⁶ and Colombia.¹⁰⁷ The ILO has included this as a key issue in the Draft Convention (Sections X, XI, XII) and Recommendation (Sections VII and VIII), acknowledging the significant implications of the use of automated systems based on algorithms or similar methods for the working conditions of digital platform workers. This is certainly one of the most promising aspects that the international labour standard could bring forward in terms of new protections for this workforce, especially in the GS, where data protection regimes are still in their infancy.

Finally, a common pattern observed in the mapping of both the GN and GS was that platform work was more prone to regulation in specific types and sectors, predominantly in on-location platform work. Close to half of the regulatory initiatives identified aim to address the conditions of platform workers performing services in specific sectors, such as ride-hailing, delivery, and, to a lesser extent, care, cleaning, and household services. All these sectors are included within the realm of “location-based” platform work.¹⁰⁸ (see Figure 7)

Figure 7

This regulatory focus on location-based platform work was even more noticeable in the GS (see Figures 8 and 9). Thus, online platform workers are not covered by a large percentage of the legislative initiatives adopted.

Figure 8

Figure 9

In general, this suggests a tendency for legislative reform to focus on specific companies and sectors in location-based platform work, particularly those with prominent visibility. Several authors have previously highlighted this trend in different national systems.¹⁰⁹ Meanwhile, the exclusive focus on riders has inadvertently overshadowed other forms of platform work, which at times represent the majority of individuals working via digital platforms – e.g., crowd workers.¹¹⁰

Thus, social policy and legislative reform so far remain narrow in scope, targeting specific sectors in which platform work is most ‘visible’, such as ride-hailing and delivery services. This exhibits tunnel vision on how challenges are tackled.

This more prominent focus on location-based platform work in the GS is of concern, considering there are major players in the platform economy—especially in terms of labor recruitment in crowdwork and click work—located in the GS, such as India and the Philippines.¹¹¹ Therefore, to address the precarious conditions in online platform work, it is necessary that the scope of regulation be broad enough to consider all forms of platform work and that these regulations are not geographically limited to the GS.

Unfortunately, regulations with limited geographic scope within countries in the GN cannot improve working conditions for those working offshore, which is a significant portion, and this trend is likely to continue due to the phenomenon of labour deterritorialization. Moreover, a great portion of the regulations enacted in the GS leave online platform workers outside the scope of protection.

The ILO’s commitment to adopt an international labour standard on platform work could be a timely instrument to shed light on this issue, broadening the scope of protection to a wider range of platform workers. In fact, the ILO appears to seize the opportunity provided in the Draft Convention Article 1(a)(ii), which states that it applies to digital labour platforms “regardless of whether that work is performed online or in a specific geographic location.”

Finally, while this paper advocates a tailored approach that accounts for regional contexts, this should not be confused with excessive flexibility or vagueness that could compromise the instrument’s effectiveness. Provisions such as Article 10 of the Draft Convention—which states, ‘Each Member shall take measures to ensure the correct classification of digital platform workers in respect of the existence of an employment relationship’— risk being too broad, potentially hindering enforcement. To ensure meaningful progress, more concrete measures are needed (see, for example, the concluding remark on page 32.)