Redistributing power in the gig economy: The impact of cooperative ownership on gig platform transparency,working conditions and earnings

Abstract

Gig platforms have reshaped how work is organised, accessed and governed. Traditional, investor-owned platforms such as Uber and Deliveroo have been criticised for their lack of transparency, labour protections and reliance on algorithmic management, resulting in precarious working conditions and unstable income for platform workers. In response to these common criticisms, a countermovement has emerged: platform cooperativism. Because of their collective ownership structure, where platform workers are worker-members, cooperative platforms have the potential to rebalance platform power and improve working conditions, income stability and worker autonomy. While promising in theory, there is limited empirical evidence of their practical impact on platform work. This article addresses that gap with an exploratory empirical analysis, comparing cooperative and traditional, investor-owned gig platforms across three dimensions: transparency, working conditions and quality of earnings. Based on an empirical document analysis of publicly available platform materials, the study systematically evaluates differences in platform design and practice, and verifies the hypothesis that the cooperative platforms can effectively foster greater transparency, quality of earnings and working conditions than traditional platforms. Findings suggest that the cooperative platforms offer advantages in respect of working conditions, including reduced reliance on algorithmic management and a more worker-centric use of technology. Additionally, they provide higher quality of earnings, not necessarily substantively but primarily through more predictable compensation structures. Lastly, however, while the cooperatives use clearer and more readable language, they have no consistent advantage over the traditional platforms in terms of transparency.

1. Introduction

In the past decade, the labour market has seen a shift from standard employment to contingent work as more and more people are working through digital platforms; in 2025, the number of active platform workers in Europe was estimated to be 43 million. ¹ Digital labour platforms, such as Uber and Deliveroo, have reshaped how work is organised, accessed and governed. ² While these platforms have reduced barriers to entry and offered workers flexibility in choosing when and where to work, they have also introduced new forms of precarity, economic dependency and power imbalances. Traditional, investor-owned gig platforms have been widely criticised for limited transparency, ³ algorithmic control ⁴ and lack of labour protections. ⁵ By outsourcing employment risks to self-employed workers and using algorithms to manage the workforce and determine compensation, platform worker rights and social protection are often sacrificed on the altar of profit maximisation – a phenomenon referred to as ‘platform capitalism’. ⁶ Under the guise of flexibility and autonomy, the rise of platform capitalism within the gig economy results in weaker social protection and renders workers increasingly disposable. This precarity has been recognised by myriad legislators and judges, as there is an increasing amount of case law that requalifies self-employed platform workers as employees, subsequently granting these platform workers the related labour protections. ⁷ At the European level, the Platform Work Directive aims to address the frequent misclassification of platform workers and the related lack of social protections. ⁸ However, these judgments, including any future judgments handed down following the adoption of the Platform Work Directive, only apply inter partes, which means that the requalifications only affect the individual platform workers bringing the cases to court. Consequently, a more structural shift in protection has not (yet) been realised.

Hence, given the limitations of top-down legislative and judicial solutions, it is interesting to investigate how bottom-up initiatives such as cooperative platforms attempt to address these common issues in platform work. Platform cooperativism emerged as a countermovement in response to platforms’ concentration of power and platform capitalism. Platform cooperatives combine the digital infrastructure of platforms with the collective ownership and democratic governance of cooperatives, making them member-owned rather than investor-owned. While cooperatives as an alternative to capitalist enterprises have existed since the nineteenth century, their emergence in the platform economy is a more recent development. Platform cooperativism adapts long-standing cooperative principles of democracy, transparency and fairness to the digital labour market, where it has been promoted primarily as a worker-owned alternative to investor-driven gig platforms. ⁹ Because the platform workers are the owners of the platform, they can create their own terms and conditions including better wages and working conditions, as well as resolve any issues concerning their often contested self-employed status. ¹⁰ Platform cooperatives thereby seek to address the power imbalances and shareholder-centrism inherent in the gig economy. ¹¹

While the membership base can vary – from single stakeholder models, such as worker or consumer cooperatives, to multi-stakeholder cooperatives including multiple user groups – the focus in this article is on worker-owned platform cooperatives in Europe. Depending on the model, workers may operate as self-employed members in a producer cooperative or as employees in a worker cooperative. Even though they are often still for-profit enterprises, their democratic self-governance enables them to foster entrepreneurship without prioritising profit at the expense of decent wages and working conditions. In this article, the terms ‘cooperative platform,’ ‘platform cooperative,’ and ‘platform co-op’ are used interchangeably to refer to worker-owned platforms. Similarly, for the purposes of this article, ‘traditional platforms’ will refer to platforms that are investor-owned.

Hence, platform cooperatives have emerged globally as alternatives to traditional platforms and help tackle challenges such as power and information asymmetries, employment status and job quality. ¹² The emergence of these ‘democratic platforms’ is a promising development within the platform economy, as they redistribute the control to the users most impacted by it: the workers. ¹³ By putting workers in control of platform governance, these platform co-ops aim to counteract some of the structural imbalances of traditional gig platforms. While the theoretical benefits of this cooperative platform model are clear, there has been little evidence demonstrating their actual influence on platform work(ers) in practice. This article attempts to address this gap with an exploratory empirical study analysing a selection of both traditional and cooperative gig platforms in Europe across three dimensions to enable a comparison. The study focuses on the dimensions of platform transparency, working conditions and earnings, because they are widely recognised as central indicators of job quality and labour standards in the gig economy (infra, sections 2.1, 2.2 and 2.3). ¹⁴ Specifically, this article verifies the hypothesis that cooperative platforms can effectively foster greater transparency, quality of earnings and working conditions than traditional platforms. The methodology is discussed in greater detail in section 3. Before turning to the exploratory empirical study, this article first briefly outlines common issues associated with traditional gig platforms, such as (the lack of) transparency, precarious working conditions under the influence of inter alia algorithmic management, and insecure remuneration.

2. Traditional gig platforms

2.1 Transparency

Transparency, or better, the lack thereof, remains a large concern in the governance and operation of traditional gig platforms. In the platform economy, platform workers are typically not employed under labour contracts but instead work as self-employed workers that merely need to agree to standardised contractual terms by accepting the platform's terms and conditions. As opposed to standard employment that brings with it social protections, sudden changes in the terms and conditions or seemingly arbitrary dismissals of platform workers are not uncommon. ¹⁵

Furthermore, there appears to be a ‘transparency gap’ between the information that platforms offer and the information platform workers believe they need in order to make informed decisions and maintain their well-being. ¹⁶ Essential information – such as on the calculation of fares, promotions, routes and task allocation – often remains undisclosed, increasing precarity and reducing workers’ capacity to make informed decisions. ¹⁷ Against this backdrop, the lack of transparency results in an information asymmetry between the platform and the worker that serves as a systemic barrier to worker rights and platform accountability.

Transparency and fairness have both already been on the European legislator's agenda in respect of platform contracts and relationships. For example, Regulation 2019/1150 on promoting fairness and transparency for business users of online intermediation services (the ‘Platform-to-Business Regulation’) explicitly seeks to address transparency concerns in platform-to-business user (i.e., self-employed platform worker) contracts. ¹⁸ It recognises that transparency is especially important considering the information asymmetry at both the pre-contractual stage as well as afterwards, and the fact that platform workers usually have little to no bargaining power. Platforms must therefore ensure that terms and conditions are easily accessible at all times, including before any contractual relationship begins, so that business users can understand the rules that would apply to them. ¹⁹ Furthermore, platforms need to set out in advance the grounds on which they might suspend or terminate a business user's account – an important safeguard given that sudden suspensions or removals, often without notice or explanation, can have severe economic consequences for platform workers. ²⁰

However, while these provisions establish baseline transparency obligations as a condition for platforms’ market access, their practical effectiveness remains limited. Low levels of compliance, exacerbated by a lack of (pro)active enforcement, undermines the Regulation's intended impact. ²¹ In practice, the provision of easily accessible and comprehensible terms and conditions for platform workers therefore remains contingent on the voluntary practices of individual platforms, rather than on proactive regulatory oversight.

2.2 Algorithmic management, geo-tracking and consumer rating systems

This lack of transparency becomes particularly consequential in the context of algorithmic management, which encompasses the digitalisation of traditional HR functions, ranging from the hiring of workers, coordinating almost all aspects of the labour process to terminating employment relationships. ²² Algorithmic management is widespread in the gig economy, as nearly all (traditional) platforms rely on algorithms for assigning tasks, deciding the remuneration and sanctioning unsatisfactory performance through performance/customer ratings. ²³ Building on the latter, data generated from customer ratings are often used as performance indicators to determine future work allocation and rewards – for example, drivers for transportation platforms may have their accounts deactivated when their average score falls below a certain threshold. Additionally, there is a mismatch between what consumers deem a good rating (such as 4 out of 5 stars), and what is considered a failing grade for most traditional platforms. ²⁴ Furthermore, if workers do not complete tasks as quickly as the platform desires or reject orders, they are warned and eventually sanctioned and kicked off the platform. ²⁵ As the European Commission has recognised, algorithmic management has a significant impact on working conditions in platform work, concealing the existence of subordination and control by the platform on the platform workers. ²⁶ Consequently, platform workers are fully dependent on the platform algorithm which exerts considerable control by not only dictating their workflow but also tracking and monitoring their performance, effectively placing the algorithm in the role of a de facto ‘boss’. ²⁷

Another concern around platform algorithms is so-called ‘algorithmic wage discrimination’, whereby platforms use data about the pay rates of all accepted and declined tasks and use it to predict and offer the lowest acceptable remuneration. ²⁸ Furthermore, as algorithms inherently lack human judgment or empathy, they are unable to consider certain drivers’ preferences or safety. As a result, female platform workers have raised safety concerns after being assigned tasks in isolated, suburban areas outside at night. ²⁹ Additionally, algorithms aggregate large amounts of data, where all platform worker actions are meticulously supervised, logged and compared. ³⁰ Couriers for Deliveroo, JustEat and UberEats have highlighted how the apps’ built-in GPS and geo-tracking features permit both the platforms and the consumers to track them in real-time, enabling them to know when they are at the restaurant, which route they take and when they are nearby. ³¹ This has increased pressure on workers, making them feel guilty about stopping to check directions, taking a quick break, or when there is heavy traffic.

This practice of algorithmic management eliminates any human oversight, reduces transparency and undermines workers’ autonomy and freedom. Through all these unilateral design choices, platform technologies predominantly tend to prioritise revenue maximisation over the well-being and autonomy of workers. In cooperative platform models, on the other hand, this platform technology can be reappropriated to facilitate user self-government rather than user domination (infra section 4.2.2). ³²

2.3 Remuneration

Since the majority of platform workers are classified as self-employed and consequently excluded from the scope of application of labour laws, they cannot fall back on social protection mechanisms providing a guaranteed minimum wage, paid sick leave, insurance, etc. The disparity between platform workers’ self-employed status and de facto subordination creates extreme wage vulnerabilities, resulting in many platform workers earning less than the local minimum wage or the lowest wage provided for under the relevant collective agreement in their respective jurisdictions. ³³

Despite having self-employed status, many platform workers lack the freedom and bargaining power typically associated with independents, such as the ability to negotiate their remuneration and working conditions. In platform relationships, the platforms unilaterally determine the rate of pay and task allocation procedures, while workers bear the costs of supplies, transport and unpaid waiting time between tasks. According to Human Rights Watch research, the cost of equipment for platform workers, such as vehicles, fuel and insurance, can add up to a significant amount– in some cases even amounting to 70% of their remuneration. ³⁴ Furthermore, due to task-based remuneration, fluctuating demand as well as the use of ‘dynamic pricing’ by platforms, platform workers’ income is often highly unpredictable. ³⁵ Both Deliveroo and Uber, for example, adjust their pay-per-delivery rate based on fluctuations in demand and supply, consequently leading to income volatility, with platform workers never knowing how much they will earn. ³⁶ It is therefore the exception rather than the rule that platforms pay their workers by the hour. This financial insecurity that platform workers often face stands in sharp contrast with the large amount of capital that gig platforms generate.

3. Methodology

This article relies on an empirical document analysis to assess whether the selected cooperative platforms offer greater transparency, working conditions and earnings than the selected traditional investor-owned platforms. The empirical component of the study consists of a systematic analysis of publicly available platform materials, such as terms and conditions, privacy policies, frequently asked questions (FAQs) and all other publicly available information, enabling a comparative evaluation of each platform's transparency, working conditions, and remuneration structures. With this approach, the article aims to reach a more nuanced understanding of how platform models differ in practice.

A more detailed methodology for each platform dimension analysed is provided in the relevant subsections below.

3.1 Research questions and hypothesis

The research hypothesis is the following: cooperative gig platforms are able to provide platform workers with greater transparency, working conditions and quality of earnings than traditional gig platforms.

To test the veracity of this hypothesis, the research question consists of three sub-questions: to what extent do the selected cooperative gig platforms:

demonstrate greater platform worker transparency about vital information compared to traditional gig platforms;

provide better working conditions than traditional gig platforms and;

provide better quality of earnings than traditional gig platforms?

3.2 Selection of platforms

For this study, an exploratory selection of nine platforms was made, consisting of four traditional gig platforms and five cooperative gig platforms. To minimise the influence of varying legal requirements and regulatory environments, platforms that are active within Europe (incl. the UK) were selected.

Even though platform work can generally be both location-based and in the online sphere, this study focuses exclusively on location-based gig platforms operating in similar sectors (food delivery and transportation). Therefore, all nine platforms selected provide location-based services and have customers limited to certain local territories. This choice was made based on the lack of cooperative platforms for online gig work, due to the fact that legal, language and cultural differences across countries often make cooperative ownership and collective decision-making much more difficult compared to locally operating platforms. ³⁷

3.2.1 The selected cooperative gig platforms

The selection consists of three food delivery platforms, Bestellenbij, Kooglof and Wings; and two taxi platforms, Alpha Taxis and Taxiapp, as shown in Table 1 below. This selection represents a mix of both producer cooperatives (where the platform workers remain self-employed) and worker cooperatives (where the platform workers are employees). Given the limited number of cooperative gig platforms, the study includes cooperatives from various European countries, facilitating a broader understanding of cooperative models while maintaining the European focus.

Table 1.

Overview of selected cooperative gig platforms.

Name	Activity	Country	Type of co-op and platform worker status
Bestellenbij	Food delivery	Netherlands	Producer – self-employed platform workers
Kooglof	Food delivery	France	Worker – employed platform workers
Wings	Food delivery	UK	Worker – employed platform workers
Taxiapp	Transportation	UK	Producer – self-employed platform workers
Alpha Taxis	Transportation	France	Producer – self-employed platform workers

Bestellenbij, a Dutch food delivery co-op, allows its couriers to be self-employed entrepreneurs that manage and run the platform. ³⁸ Its model emphasises local cooperation and fair and sustainable business practices. Kooglof is a French food delivery co-op that aims to reduce job precarity by directly employing its couriers. ³⁹ Its focus on environmental sustainability is reflected in its use of (cargo) bikes for all deliveries. Wings, branded as ‘London's ethical food delivery service’, only delivers from independent local businesses and is committed to eco-friendly operations, using exclusively zero-emission vehicles for deliveries. ⁴⁰ Taxiapp is a UK-based cooperative and aims to ‘take back the trade and put every penny of every fare where it belongs – in your [the driver's] pockets’. ⁴¹ The fifth and final platform is Alpha Taxis, a French, Paris-based taxi platform. ⁴²

3.2.2 The selected traditional gig platforms

Four platforms with traditional platform models are also included to allow for a comparison: Uber and Bolt in the taxi sector, and Deliveroo and Takeaway in the food delivery sector, as shown in Table 2 below. To ensure a consistent framework for the analysis of traditional gig platforms that operate in multiple countries, the Belgian branches of these platforms are used as a focal point for the comparison.

Table 2.

Overview of selected traditional gig platforms.

Name	Activity	Country of origin	Platform worker status in BE
Uber	Transportation	U.S.A.	Self-employed
Bolt	Transportation	Estonia	Self-employed
Deliveroo	Food delivery	U.K.	Self-employed
Takeaway	Food delivery	Netherlands	Employee

Table 3.

Grading scale for transparency scores of platforms.

Grade	Label	Score range	Description
A	Excellent Transparency	85–100%	Information is very easy to find and easy to understand.
B	Good Transparency	70–84%	Information is generally accessible and understandable.
C	Moderate Transparency	50–69%	Basic transparency is present, but significant areas require attention to improve accessibility and clarity.
D	Low Transparency	30–49%	Transparency is poor, with key information either difficult to find or hard to understand.
E	Very Low Transparency	0–29%	Transparency is severely lacking; essential information is either missing or inaccessible.

Table 4.

Overview of traditional platforms’ transparency scores for platform workers.

	Accessibility	Readability	Overall transparency
Uber	E	D	E
Bolt	A	C	B
Deliveroo	E	C	D
Takeaway	C	C	C

Table 5.

Overview of cooperative platforms’ transparency scores for platform workers.

	Accessibility	Readability	Overall transparency
Bestellenbij	A	B	A
Kooglof	C	B	B
Wings	E	N/A	E
Taxiapp	A	A	A
Alpha Taxis	E	N/A	E

American platform Uber is one of the largest transportation platforms worldwide, demonstrated by its €9.9 billion revenue in 2025. 43 It uses self-employed couriers in Belgium, even though appeal judges very recently requalified Uber drivers as employees. ⁴⁴ Secondly, Bolt is an Estonian platform that uses self-employed drivers, active in 45 countries across Europe, Africa, Asia and Latin America. ⁴⁵

Deliveroo is a UK food delivery platform that – despite case law ⁴⁶ recognising the couriers’ de facto subordination – works with self-employed couriers, and is active in 10 countries worldwide. ⁴⁷ Lastly, Takeaway, also known as JustEat and Lieferando depending on the country, is an originally Dutch food delivery platform and is known as of the few platforms that employs most of its couriers in Europe, though is increasingly scaling back on its employment model.

4. Comparative review of the selected gig platforms

4.1 Transparency review

4.1.1 Methodology

By assessing the publicly available materials of both traditional and cooperative platforms, the study aims to determine whether cooperative platforms provide greater transparency than traditional platforms. This transparency is measured based on whether the information is publicly available and easy to find (accessibility), and whether the information is drafted in a plain and intelligible language (readability). By comparing these levels of transparency, the study can evaluate the extent to which platform co-ops can combat the information asymmetry that is often seen in traditional platforms. Furthermore, under the Platform-to-Business Regulation, all platforms with ‘business users’ (i.e., self-employed platform workers) are legally required to ensure that their users can easily find the terms and conditions (supra).

The study focuses on three key topics:

membership;

financial structures; and

order/service cancellation.

These topics were chosen because they reflect important governance mechanisms through which digital platforms structure labour relations. Membership rules determine who can access platforms and under which conditions, financial structures reveal how value is distributed between platforms and workers, and cancellation policies indicate how operational and economic risks are allocated. For membership, this study focuses on provisions that affect the formation, amendment, suspension and termination of the terms of service between users and platforms. Including the specific grounds for a suspension or termination in the terms and conditions is a requirement under the Platform-to-Business Regulation (supra). Regarding the financial structures, the study examines the transparency in respect of remuneration as well as commission fees, cancellation fees and any other fees that may arise. For cancellation policies, the study focuses on the extent to which information on the possibility to cancel services is made publicly available.

For each topic, the study examines two dimensions of transparency. Firstly, the study reviews the accessibility of information related to each topic, assessing whether the relevant information is made publicly available, regardless of whether it is included in the terms and conditions, platform policies, FAQs or other information found on the website. The study rates accessibility on a scale from 0 to 1: 0 represents not found, 0.5 means it is difficult to find, and 1 means it can be easily found. This rating is based on whether and how easily the information for each topic can be located, based on the amount of click(s) required to navigate the website and locate the information. ⁴⁸ Secondly, the study evaluates the readability of the information related to each topic on a scale from 0 to 1, with 0 being very hard to understand, 0.5 representing an average level of readability, and 1 demonstrating an active effort to increase readability. This rating reflects the extent to which the relevant information is drafted in plain and intelligible language by examining the length of sentences and words based on the commonly used Flesch-Kincaid readability test, ⁴⁹ the difficulty level of vocabulary used and the organisation or the display of content.

Finally, the study calculates the average transparency score of reviewed platforms on a scale from 0 to 100, considering all topics. The platforms were then classified into the grading categories shown in Table 1 below, ranging from A: Excellent Transparency to E: Very Low Transparency, as visualised below in Table 3.

4.1.2 Findings

4.1.2.1 For the traditional platforms

Table 4 provides an overview of all traditional platforms' transparency scores. Uber is the lowest-scoring traditional platform, receiving an overall transparency score of E. It performs poorly in both accessibility and readability, but loses the most points for accessibility, as there are no easily accessible terms and conditions for drivers on the website. Instead, a prospective Uber driver has to apply to become an Uber driver and complete information forms providing details, including full name, address, make of vehicle and licence plate, before finally gaining access to the relevant driver-specific information.

Bolt, on the other hand, is the highest-scoring traditional platform. Its website provides multiple and well-organised terms and conditions, including a separate set of terms and conditions for drivers, thereby making all information easily accessible within a single click. Strikingly, however, in the ‘membership’ information section, Bolt explicitly reserves the right to terminate the agreement at any time and for any reason, at its sole discretion. Not only is this a far-reaching clause that creates significant insecurity for platform workers, but it also violates the Platform-to-Business Regulation that requires platforms to set out in advance the specific grounds on which they might suspend or terminate a business user's account. ⁵⁰ This shows that transparency as a procedural value does not translate into substantive fairness in the platform relationship.

Deliveroo, similarly to Uber, receives the lowest score for accessibility, as it does not provide easily accessible terms and conditions for its couriers. Accessing the relevant information requires navigating through numerous clicks, scattered links and FAQs. In terms of readability, however, Deliveroo scores slightly better and provides relatively clear information, particularly regarding membership.

Takeaway earns a consistent C-score for both accessibility and readability. Interestingly, Takeaway provides separate terms and conditions for both consumers and restaurants, but not for its couriers. Across the website, however, there are several links to dispersed articles written by HR that contain relevant information for couriers. It is important to note that, unlike the other traditional platforms, Takeaway hires its couriers as employees rather than as independent (business) users. As such, there is no requirement under the Platform-to-Business Regulation to publish publicly available terms and conditions for these prospective employees – particularly as they (will) receive labour contracts that are more comprehensive than terms and conditions by default. Therefore, it can be reasonably assumed that the individual labour contracts contain all the relevant information, putting the Takeaway's courier transparency score into context.

4.1.2.2 For the cooperative platforms

Table 5 provides an overview of all cooperative platforms' transparency scores. Interestingly, only one out of five cooperative platforms investigated had publicly available terms and conditions for platform workers, i.e., Taxiapp. Therefore, despite the Platform-to-Business Regulation's requirement to provide business users with easy access to the terms that will apply to them in the pre-contractual phase, ⁵¹ it is the exception rather than the rule for cooperative platforms. Nevertheless, some co-ops do provide relevant information to business users in the form of FAQs sections. While it might be helpful, this approach should be supplemented with a centralised and comprehensive presentation of relevant terms. Including this information in easily accessible terms and conditions would reduce the need for frequently asked questions, since the answers could already easily be found.

Bestellenbij ranks as the second-best scoring cooperative platform, demonstrating active efforts to increase the readability of all available information. Although it does not provide formal terms and conditions for its couriers, the website has a separate tab for prospective couriers containing clearly explained FAQs, including information on remuneration, worker status, what it means to become part of the co-op, etc., hence its A-score for accessibility.

Surprisingly, both Kooglof and Wings do not provide any public terms and conditions for workers, or even for consumers. A possible explanation for this could be the fact that both Kooglof and Wings have made use of CoopCycle, ⁵² a federation of courier cooperatives that builds the software and application required for cooperative food delivery platforms at minimal cost. Hence, the only terms and conditions available for both Kooglof and Wings are the standard ones drafted by CoopCycle, which refer to the usage of the website and the application. The absence of worker-specific terms and conditions means that prospective couriers are not able to verify what rights (would) apply to them when delivering food for Kooglof and Wings. However, since these two platforms are both worker cooperatives, the fact that couriers are hired as employees – and thus protected under labour law – mitigates concerns about the low courier transparency, as with Takeaway.

Wings does not even have a standalone website containing more information about the platform and/or its mission statement other than the food-ordering page, which explains its E-score for transparency. Unlike Wings, Kooglof has a separate website explaining its mission and values. While it also lacks formal terms and conditions, its use of accessible formats – such as bullet points and tables – earns it a C for accessibility and a B for readability.

Taxiapp is the highest-scoring cooperative platform, and the only one among the five cooperatives analysed that provides very accessible terms and conditions for their drivers. Not only does it provide all relevant information, but it also scores equally high for readability, showing active efforts to ensure everything is written in plain and intelligible language. This demonstrates a strong commitment to transparency for current and prospective platform workers.

Lastly, Alpha Taxis has a set of public terms and conditions, but limited solely to the use of the website and application and/or tailored specifically towards consumers, leading to the lowest E-grade for driver transparency. The website further lacks any specific information for (prospective) drivers, which means they are unable to find - let alone easily find - the terms and conditions that would apply to them in the pre-contractual phase, and therefore likely violates the Platform-to-Business Regulation.

4.1.3 Interim conclusion

The results of the first part of this empirical study suggest that cooperative gig platforms do not necessarily exhibit a higher degree of transparency compared to traditional platforms. While the initial hypothesis posited that cooperatives would provide better access to, and clearer communication of, platform information, the data reveals a more fragmented reality. Table 6 below ranks all investigated platforms according to their overall transparency scores, incorporating both accessibility and readability.

Table 6.

Overview of transparency scores of all gig platforms, both traditional and cooperative (from high to low).

Platform type	Platform	Accessibility	Readability	Overall transparency
Cooperative	Taxiapp	A	A	A
Cooperative	Bestellenbij	A	B	A
Traditional	Bolt	A	C	B
Cooperative	Kooglof	C	B	B
Traditional	Takeaway	C	C	C
Traditional	Deliveroo	E	C	D
Traditional	Uber	E	D	E
Cooperative	Wings	E	N/A	E
Cooperative	Alpha Taxis	E	N/A	E

Among all the platforms analysed, there is a wide range of transparency scores, ranging from as high as A to as low as E. Interestingly, the cooperative platforms make up both the best-scoring and the worst-scoring platforms, with the traditional platforms clustered in the middle. This distribution of cooperative platforms at both ends demonstrates that cooperative status alone does not guarantee a high degree of transparency.

A closer look reveals an interesting distinction between accessibility and readability. Traditional platforms tend to use more legalistic or vague language which leads to their overall lower readability scores. This creates barriers for workers and is especially problematic since workers must often accept contractual terms with minimal or no negotiation. Cooperative platforms, by contrast, when they do provide information, tend to offer clearer, more readable texts, consistently scoring in the A-B range for readability. Still, two of the cooperative platforms do not publish any relevant public information at all, undermining the assumption that the cooperative model inherently results in greater transparency.

Among the platforms that hire their couriers as employees (Takeaway, Kooglof and Wings), the lack of publicly available terms and conditions is partly mitigated as labour laws provide a comprehensive protective framework that combats the traditional information and power asymmetries. In such cases, the protective framework of the labour laws in the relevant jurisdictions partly substitutes for the role that extensive pre-contractual transparency would otherwise play.

As worker-governed entities, cooperative platforms may consciously decide not to make certain information publicly available, particularly if it is considered internal or strategically sensitive. Nevertheless, the Platform-to-Business Regulation introduces a legal requirement on all platforms with business users, regardless of size or scope, to ensure that they are able to find the terms and conditions easily, a requirement that many of the platforms analysed fail to meet. One possible explanation is the prevalence of multi-homing (i.e., users being able to simultaneously use and work for multiple platforms) in the gig economy, which may lead to co-ops consciously limiting access to certain information to avoid competitors accessing it. ⁵³ The opportunity – and often necessity – of multi-homing likely affects the extent of free information sharing and transparency within cooperative platforms.

Overall, these findings suggest that the approach of cooperative platforms to transparency as a value seems nuanced and context-related rather than structurally guaranteed. ⁵⁴ It is possible that cooperative platforms emphasise internal and operational transparency about, e.g., decision-making structures, the allocation of resources and revenue distribution, etc., rather than a pre-contractual and externally observable transparency. Hence, while the cooperative platforms analysed demonstrate a clear advantage in the readability of available information, their potential for increased overall transparency is not uniformly or evenly realised.

4.2 Working conditions

4.2.1 Methodology

Secondly, the quality of platform working conditions is examined by looking at the presence or absence of algorithmic management, geo-tracking and consumer rating systems. These indicators are selected as observable proxies for managerial control and are based on the OECD Job Quality Framework, which includes algorithmic management techniques such as progress tracking and performance assessment as indicators of the quality of the work environment. ⁵⁵ As these factors play a crucial role in shaping platform workers’ autonomy, job security and overall work experience, they serve as important parameters for evaluating the quality of working conditions across different platform models for this study. While consumer rating systems can be seen as a form of democratic feedback, this study focuses on their function within algorithmic management systems, as these ratings often have disciplinary consequences.

Information on the presence of these features was – to the extent available – drawn primarily from the platforms’ own materials, such as their terms and conditions, privacy policies or other relevant information made publicly available on their websites. Where such primary sources were unavailable or insufficient, secondary sources were consulted, including academic literature and empirical research specifically addressing the use of these elements in the examined platforms.

4.2.2 Findings

4.2.2.1 For the traditional platforms

Table 7 below provides an overview of the absence or presence of algorithmic management, geo-tracking and user rating systems in the traditional platforms investigated, followed by a more detailed description for each platform.

Table 7.

Overview of the absence or presence of algorithmic management, geo-tracking and user rating systems in the four traditional platforms investigated.

	Algorithmic management	Geo-tracking	User rating system
Uber	✓	✓	✓
Bolt	✓	✓	✓
Deliveroo	✓	✓	✘
Takeaway	✓	✓	✘

Uber employs numerous forms of algorithmic management, a factor that has recently been explicitly used in the reasoning of the Brussels Court of Appeal in a judgment that requalified Uber's drivers as employees. ⁵⁶ The judges recognised that the way Uber uses algorithms for assigning rides, evaluating performances and determine fares, amounts to a modern form of employer authority. When it comes to performance evaluation, Uber drivers stated in an empirical study that passenger ratings constitute the most significant performance metric, despite having limited control over how passengers interact with the rating system or how Uber assesses it. ⁵⁷ According to Uber's own terms and conditions, drivers who fall below the minimum average rating for their city (around 4.6 out of 5 stars) ⁵⁸ risk losing access to the platform. Lastly, through its application, Uber has access to drivers’ GPS and gyrometer and acceleration sensors in each driver's phone to detect speeding or sharp braking. This form of control through geo-tracking is often justified in terms of consumer protection rather than as a key aspect of employer control. ⁵⁹

Bolt does not explicitly mention the use of algorithms, but in a data science interview, it reveals that it uses algorithms for dynamic pricing and allocating tasks in order to optimise rider and driver matches. ⁶⁰ In terms of algorithmically tracking workers’ productivity, the most explicit example comes from Bolt, which uses a metric called Driver Score, based on their activity in accepting, declining, not responding to and completing transportation service requests. ⁶¹ Furthermore, Bolt's terms and conditions are transparent about the fact that Bolt uses a passenger rating system in order to guarantee a high-quality service and provide additional reassurance for passengers. ⁶² The terms and conditions further state that drivers must establish and maintain a minimum average rating, and if drivers do not increase this rating or activity score within a prescribed time period, their accounts will be automatically suspended. ⁶³ If drivers consequently fail to meet this minimal rating or activity score, Bolt is entitled to immediately terminate agreements without any advance notice. ⁶⁴ Lastly, the privacy policy explicitly states that Bolt uses geolocalisation when Drivers are using the Bolt app, in order to analyse the geographical area. ⁶⁵

Deliveroo states, in its Rider Privacy Policy, that it uses automated systems to allocate orders, and states this allocation happens solely based on riders’ location and vehicle type to ensure efficiency. ⁶⁶ Furthermore, it uses algorithms to measure and supervise the work, and has real-time GPS and geo-tracking information on each worker. ⁶⁷ Unlike Uber and Bolt, there is no consumer rating system in place at Deliveroo, as the consumer only has the option to add a tip after the delivery. However, Deliveroo does use internal rating systems to measure worker performance through its algorithmic management and extensive data extraction. For example, a Deliveroo courier reported that after he was no longer physically able to accept long-distance deliveries and had to reject certain deliveries, Deliveroo's system rated him as one of the worst-performing riders in the city, and consequently terminated his contract. ⁶⁸

Takeaway's approach to algorithms varies depending on jurisdiction and model. In Belgium, where the couriers are employees, their income does not depend on the number of deliveries they are allocated. Nonetheless, algorithmic performance management remains embedded in the system: courier metrics – such as punctuality and delivery completion – are tracked internally and influence shift allocation and bonus eligibility (such as incremental pay increases after consistent performance over several months). ⁶⁹ While these performance evaluations are embedded within an employment framework (and not visible or rating-based as with Uber), the presence of algorithmically driven control mechanisms adds a layer of internal surveillance and conditionalism. This dual structure of employee status, coupled with algorithmic oversight, limits the applicability of the Platform-to-Business Regulation while still emphasising the necessity for transparency in internal performance criteria. Lastly, Takeaway uses geo-tracking for its couriers, but only for delivery logistics, not for monitoring performance.

4.2.2.2 For the cooperative platforms

Table 8 below provides an overview of the absence or presence of algorithmic management, geo-tracking and user rating systems in the cooperative platforms investigated, followed by a more detailed description for each co-op. A question mark indicates that the public resources available did not provide sufficient information to verify the presence or absence of the feature.

Table 8.

Overview of the absence or presence of algorithmic management, geo-tracking and user rating systems in the five cooperative gig platforms investigated.

	Algorithmic management	Geo-tracking	User rating system
Bestellenbij	✓	✓	✘
Kooglof	✘	✓	✘
Wings	✘	✓	✘
Alpha Taxis	?	?	?
Taxiapp	?	✓	✓

Bestellenbij serves as an interesting example of how platform technology can be reappropriated to serve workers rather than control them. At present, the cooperative is small enough for couriers to allocate incoming orders themselves; orders are visible to every courier via a shared app, and allocation is arranged through informal consultation – often via WhatsApp – based on location, availability, and mutual agreements. This system relies on trust, peer coordination, and a strong sense of collective responsibility. Looking ahead, Bestellenbij does plan to introduce algorithmic tools to support workforce management, but will explicitly design these as user-driven algorithms that propose options based on member input and predefined human criteria, rather than issuing commands. The goal is to preserve worker autonomy by ensuring that final decisions rest with those performing the work, thereby reversing the dynamic of traditional platforms where opaque algorithms dictate task allocation, deadlines and remuneration.

Kooglof and Wings are both cooperatives part of the CoopCycle federation (supra). As a result, both Kooglof and Wings use no algorithmic management to coordinate their platform workers since the CoopCycle software does not incorporate such technology. Instead, the software allows for the management to be performed by a human co-worker who manually assigns customer requests to available riders. ⁷⁰ In order to organise this, the dispatchers are in contact with the active riders through common messenger apps, outside of the CoopCycle platform system. Additionally, the absence of algorithmic management results in the freedom for both couriers and restaurants to decline certain tasks without implications for their ranking, status, income, etc. ⁷¹ All the couriers under the CoopCycle wing have made the active design choice to have manual task assignment rather than task assignment algorithms, since humans can make better context-related decisions. ⁷² Initially, CoopCycle did not include geo-tracking in its software and application, yet the couriers themselves approved the presence of geo-tracking to support the dispatch process by a majority vote. ⁷³ However, details of the couriers’ locations are only available to the dispatcher when they are active in the dispatcher role and are not visible to anyone else. The location data is not collected, stored, or otherwise analysed. In conclusion, the geo-tracking function is limited to its declared purpose – i.e., support of the dispatch process – and not (ab)used for worker surveillance or algorithmic management. Lastly, through the CoopCycle software, neither Kooglof nor Wings have a rating system.

Taxiapp, according to its terms and conditions, operates solely as a booking and communication intermediary between licenced black cab drivers and passengers. ⁷⁴ There is no mention of algorithmic dispatch or performance-based task assignment, so it is unclear whether a form of automated system is used to match clients with available drivers, or whether this happens manually. Furthermore, the platform does not employ a customer rating system for drivers, thereby avoiding the common feedback loops that can impact workers’ access to future tasks on other platforms. Taxiapp does collect and process precise location data, as confirmed in its privacy policy and mobile app metadata, in order to enable live map functionality and efficiently connect nearby taxis with passengers. ⁷⁵ This geo-tracking is explicitly limited to real-time logistical purpose and not tied to performance evaluation. Overall, Taxiapp's approach reflects a conscious decision to preserve driver autonomy by limiting geo-tracking to its intended purpose and avoiding ratings, aligning with its cooperative ethos and commitment to driver control.

Alpha Taxis does not explicitly disclose the use of algorithms, geo-tracking or performance ratings in its terms and conditions or anywhere on the website, which is unsurprising given its E-score for transparency. Similar to Taxiapp, it is unclear whether a form of automated system is used to match clients with available drivers or a manual allocation. The website does state that taxi fares are calculated based on regulated taxi meters, indicating there is no dynamic pricing such as, e.g., with Uber and Bolt. ⁷⁶ Geo-tracking is implicitly necessary for the service to function, particularly for real-time pick-up coordination, but it is not transparently addressed in the platform's legal documents by lack of a Privacy Policy. Moreover, it is unclear whether Alpha Taxis implements a consumer rating or feedback mechanism for drivers.

4.2.3 Interim conclusion

Related to and following cooperative platforms’ lower transparency scores, there is less data to judge the concrete working conditions of their platform workers. Yet, across the platforms analysed, there are clear structural differences between traditional and cooperative models in the use of algorithmic management, geo-tracking and consumer ratings.

All four traditional platforms integrate both algorithmic management and geo-tracking as core operational tools, with consumer ratings commonly used among taxi services (Uber, Bolt) but absent in food delivery (Deliveroo, Takeaway). The use of geo-tracking is universal, as it is essential for location-based platform services, but extends beyond logistics in many traditional platforms, doubling as a mechanism for surveillance and performance control. Furthermore, although the use of consumer rating systems can serve as a valuable, bottom-up tool for service quality control, many platforms penalise workers for receiving ratings that may appear acceptable to consumers (e.g., 4 stars), without clearly communicating the thresholds used. Increasing consumer awareness of these rating consequences could mitigate unfair outcomes and foster a more constructive feedback culture. In sum, in the traditional platforms, algorithms allocate tasks, monitor performance, and may influence workers’ future access to work, often through opaque systems that leave workers in a passive rather than participatory position.

By contrast, the cooperative platforms demonstrate greater variation and, in some cases, deliberately choose to get rid of these control mechanisms. Two co-ops (Kooglof and Wings) have opted to forego algorithmic management entirely, relying instead on human dispatchers and peer coordination. Even when geo-tracking is used, as with Kooglof, Wings and Taxiapp, its scope is consciously limited to real-time logistical support, with safeguards to prevent data storage or repurposing for surveillance. In other cases, such as Bestellenbij, algorithmic tools are being developed but framed as ‘user-driven,’ designed to support rather than dictate worker decision-making. This contrasts with the traditional platforms and underscores the broader dynamic that platform governance choices can shape the quality of working conditions and worker agency. In this sense, overall, the examined cooperative platforms seem to provide better working conditions than the traditional platforms.

4.3 Quality of earnings

4.3.1 Methodology

Lastly, the quality of earnings of platform workers on cooperative and traditional platforms is compared. This aspect encompasses both the wage (either calculated per hour or per task, depending on the platform) as well as any platform commission that might be charged to the workers. Due to the limited availability of specific data on platform remuneration, this section contains a more general rather than substantive assessment of compensation structures among the traditional and cooperative platforms investigated.

4.3.2 Compensation structures across gig platforms

As Figure 1 below illustrates, a wide variety of compensation structures can be seen across all gig platforms. Almost all cooperative platforms investigated assert on their websites that they care for and guarantee ‘fair remuneration’ for their platform workers. ⁷⁷ Whether the platform workers are paid by the hour of by the number of completed tasks, varies based on the legal status of the platform workers.

Figure 1.

Overview of various worker statuses and compensation structures among the investigated platforms.

Cooperative platforms Kooglof and Wings are both worker cooperatives that employ their couriers, and consequently pay them by the hour. Kooglof couriers receive the ‘salaire minimum interprofessionnel de croissance’ (SMIC), which is the legal minimum hourly wage in France. Wings goes beyond the minimum wage and has undertaken to pay every rider the London Living Wage, ⁷⁸ a voluntary initiative that takes into account the actual costs of living in a city like London. Takeaway is the only traditional platform investigated that employs its couriers and consequently pays them the minimum wage per hour, including for the usually unpaid waiting time.

All other cooperative platforms investigated (Bestellenbij, Alpha Taxis and Taxiapp), as well as traditional ones (Uber, Bolt and Deliveroo), require the platform workers to be self-employed, which involves payment per completed task instead of per hour. As previously mentioned, Uber and Bolt use algorithms for dynamic pricing and calculating the fares, consequently impacting drivers’ income. Deliveroo, too, uses a variable fee-per-order system where algorithms compute the most efficient fee for each delivery, based primarily on distance and traffic. This means that platform workers for Uber, Bolt and Deliveroo receive a different fee for each task they complete, leading to significant income uncertainty. Bestellenbij couriers, on the other hand, are paid according to a fixed formula consisting of a set amount per order, supplemented by an additional payment per kilometre travelled, which means the payment per completed delivery varies only depending on the distance of each delivery. Alpha Taxis pays their drivers per ride according to the regulated Paris taxi meter, rather than through platform-specific or dynamic pricing systems. Fares are calculated based on official prefectural tariffs, determined by time and distance. Thus, despite still being subject to customer demand and waiting times, the fixed compensation formula of Bestellenbij and Alpha Taxis offers the workers a certain amount of predictability.

Taxiapp provides no concrete information on how driver pay is calculated or the exact amounts earned per job, but it does state it uses a surge-free pricing model, where passengers pay the Transport for London-regulated metered fare plus a flat booking fee of £2. It used to be one of the only platforms that did not charge its drivers any commission at all, but recently changed this policy to charge 5% commission, which is – according to Taxiapp – ‘the lowest commission on the road’. ⁷⁹ This change was introduced when Taxiapp formed a partnership to attract investment to improve the app, while simultaneously keeping the commission low. On its website, it assures that Taxiapp is still worker-owned and that the commission cannot be increased without a unanimous vote of the drivers. ⁸⁰ Hence, drivers’ pay will be equal to the exact fare on the taximeter, minus 5% commission. Lastly, Taxiapp also advertises a Drivers Reward Scheme, under which all drivers share in profits based on the number of jobs completed.

4.3.3 Interim conclusion

Among platforms that do not pay their workers by the hour but instead use a (variable) compensation formula, it is not clear whether these platform workers receive a living wage equal or higher to the legal minimum wage in their respective countries, particularly given the lack of transparency and publicly available information on remuneration. Consequently, it is important to note that there is a lack of insight on actual take-home pay across the platforms investigated.

For employed platform workers, the minimum wage is a legal safeguard guaranteeing them a predictable basis income for the hours they have worked. Some cooperative platforms, such as Wings, show the potential to go beyond minimum legal requirements by instead paying the workers a living wage that takes into account the actual cost of living in a capital city like London.

Among the self-employed platform models, a clear divide emerges between the traditional and cooperative platforms. Within all traditional platforms analysed, platform worker pay fluctuates due to variable, algorithm-controlled rates, making earnings unpredictable. By contrast, the cooperative platforms using the self-employed model-based remuneration with fixed parameters – such as a set fee per order plus a per-kilometre supplement or official taxi meter tariffs – provide for reduced earnings volatility and offer a higher degree of predictability. Nonetheless, wage predictability alone does not guarantee wage adequacy: predictable earnings can still lead to (predictably) low wages, if not aligned with an actual liveable income.

To ensure more predictable wages for platform workers, even within a self-employed, pay-per-task model, Uber's model in the UK could serve as inspiration. Since 2021, Uber drivers are no longer qualified as self-employed but rather as ‘workers’, a status unique to the UK, entitling them to a wage floor, holiday pay and pension contributions while remaining self-employed for tax purposes. Even though it might be unfeasible to introduce a similar third worker status across the EU, the underlying principles behind this remuneration system could be applied more broadly. As Uber explains on its website, drivers get paid for their ‘engaged time’, i.e., the time between accepting the trip and completing the trip, and if a driver's weekly earnings fall below the UK National Living Wage based on the engaged hours, Uber will top up the driver's earnings to meet that minimum. ⁸¹ This model, while maintaining drivers’ autonomy and flexibility to choose when and where to work, offers a basic level of financial security that traditional self-employment models often lack.

Taken together, these findings suggest that the cooperative platforms that employ their workers can match the traditional platforms in providing stable, predictable earnings – and, in some cases, exceed them through living wage commitments – while those operating with self-employed workers are able to improve the quality of earnings, though solely by reducing the uncertainty inherent in algorithm-driven pay systems. However, the extent of this improvement is constrained by the absence of guaranteed minimums.

5. Conclusion

This article set out to test the hypothesis that cooperative gig platforms provide platform workers with greater transparency, working conditions and quality of earnings than traditional gig platforms. Drawing on a comparative document analysis, the findings confirm this hypothesis in some respects, though suggest a mixed picture across the three dimensions when assessing the performance of cooperative gig platforms against their traditional counterparts. Given the exploratory nature of the study and the limited number of platforms analysed, these findings should be interpreted as illustrative rather than representative of the platform economy as a whole.

On transparency, the cooperative platforms do not consistently perform better than the traditional platforms. While they score higher on readability, using clearer and more comprehensible language, this is offset by significant gaps in the availability of information. Two cooperative platforms provide no publicly accessible terms and conditions or worker-specific information at all, challenging the assumption that cooperative platforms inherently guarantee higher transparency. Therefore, transparency does not appear to be fully integrated as a core value of cooperative platforms in terms of their outward-facing practices. This may be because cooperative platforms place greater emphasis on internal transparency – ensuring openness and accountability within the platform and among its worker-members – rather than on external transparency, which is the focus of this study through a review of publicly available materials.

In terms of working conditions, the study suggests that the cooperative platforms offer advantages compared to the traditional platforms. Where the traditional platforms rely on algorithms and geo-tracking for allocating tasks and monitoring worker performance, the cooperative platforms appear to approach the use of technology more consciously, either by reframing technology as a supportive tool rather than an instrument of control, or abolishing it altogether. The absence of algorithmic management or the exclusive use of user-driven algorithms designed to support rather than control workers, combined with geo-tracking features limited to their operational purpose rather than used for surveillance, fosters a less intrusive and more worker-centric environment.

Not enough specific information was publicly available on the quality of earnings dimension to substantively compare the remuneration levels, yet a more procedural assessment of compensation structures suggests that cooperative platforms provide a better quality of earnings, in the sense of predictability. Where the traditional platforms use dynamic and algorithmic pricing, cooperative platforms rely on fixed compensation formulas or regulated fares, thereby ensuring more predictable earnings for self-employed workers. Where workers are employees, the statutory minimum wage applies equally to both cooperative and traditional platforms. However, the example of Wings illustrates how a cooperative may go beyond the minimum wage, committing to pay the London Living Wage. This combination of predictability and the potential to exceed minimum legal standards illustrates the capacity of cooperative models to deliver more equitable remuneration

In sum, the findings of this exploratory comparison suggest that cooperative platform models may result in better outcomes for platform workers. While traditional platforms often face criticism for their lack of transparency, precarious working conditions, unilateral decision-making and insecure income, the cooperatives examined demonstrate that another way is possible, with the strongest advantage in working conditions, a notable but narrower advantage in earnings quality, and an inconsistent advantage in transparency. This illustrates how alternative governance structures may shape platform practices in different ways and shows it is possible to organise digital labour in a way that prioritises workers’ preferences and income stability over profit maximisation. By placing workers at the centre of governance and decision-making, cooperative platforms demonstrate that digital labour can be organised in more democratic and worker-oriented ways. While further research with larger samples and additional data sources would be required to assess the broader generalisability of these findings, the platforms examined in this study suggest that cooperative platform models may represent a promising avenue for rethinking the governance of platform work.