Regulating the Fairness of Work Contracts in the Gig Economy

Abstract

The growth of the ‘gig economy’ has prompted debate about the regulation of arrangements to obtain work and income through digital labour platforms. For platform workers who are classified as freelancers or independent contractors, rather than as employees, one possibility is to invoke general laws on the fairness of contractual terms to challenge the inclusion of harsh or one-sided provisions in the contracts of adhesion typically drafted and imposed by digital platforms. To test the potential application of one such regime, in pt 2-3 of the Australian Consumer Law (‘ACL’), we systematically analyse the terms and conditions used by various platforms intermediating work performed in Australia, within and across different industry sectors. Our analysis uncovers many examples of terms that are designated in s 25 of the ACL as potentially unfair or that have been identified as potentially problematic by the Australian Competition and Consumer Commission (‘ACCC’). We also examine the practical difficulties confronting a worker seeking to challenge the fairness of their contract terms, against the background of recent reforms to enhance the efficacy of this regime.

I Introduction

The gig economy (or ‘sharing economy’ or ‘platform economy’, as it is variously termed) can encompass many different types of ‘peer-to-peer’ transactions, including the sale or lease of goods or land. ¹ But a good deal of the public and scholarly debate concerning the use of apps and websites to create digital marketplaces has revolved around arrangements for the supply of personal services. This is especially true when the services in question are sourced from workers who might otherwise have been regarded as employees of the ‘end-users’ who benefit from their labour, or (more likely) of the digital platforms who connect them. ² In two sectors in particular, personal transport and food delivery, there has been a great deal of both controversy and litigation concerning the potential application of minimum wage laws and other labour standards. ³ There are, however, many other types of labour market, including for the supply of personal care and support, that are potentially affected by the growing use of digital labour platforms. ⁴

In the absence of any significant legislative intervention, it seems likely that many of the ‘gig workers’ offering their services through digital platforms will, whether appropriately or not, be categorised as self-employed ‘freelancers’ or independent contractors. As such, whatever the application of broadly framed laws on work health and safety or discrimination, they may fall outside the protection of labour statutes such as the Fair Work Act 2009 (Cth). ⁵ The chances of platform workers in Australia successfully establishing that they are employees, in the face of contracts classifying them otherwise, have been significantly reduced by two recent decisions of the High Court. These insist that disputes over employment status are to be resolved by reference to the rights and obligations contractually agreed between the worker and the person or organisation engaging their services, not the reality of the relationship or what actually happens in practice. ⁶ In a recent case, a Full Bench of the Fair Work Commission made it clear that a food delivery rider would have been classified as an employee, had the tribunal been permitted to assess the substance of his relationship with the platform through which he was finding work. ⁷ This is not to say that digital labour platforms cannot choose to be employers. ⁸ As the law stands, however, most are free to avoid the costs and liabilities associated with having that status.

Against this background, the Albanese Government is planning to empower the Fair Work Commission to ‘make orders for minimum standards for new forms of work, such as gig work’, by extending its jurisdiction under the Fair Work Act 2009 to regulate ‘employee-like’ types of work. ⁹ But another and more immediate possibility, identified by a number of commentators, ¹⁰ lies in the use of general laws on the fairness of contractual terms. The potential unfairness of platform contracts has been examined from the perspective of end-users: that is, those who use platforms to obtain goods or services, whether as individual consumers, ¹¹ or businesses. ¹² It is also possible that workers who are not employees might be able to challenge the inclusion of harsh or one-sided provisions in the contracts of adhesion typically imposed as a condition of gaining access to digital labour platforms.

Our focus is on the unfair contract terms (‘UCT’) provisions in Part 2-3 of the Australian Consumer Law (‘ACL’), as set out in Schedule 2 of the Competition and Consumer Act 2010 (Cth). ¹³ To test the potential application of this regime, we analyse the terms and conditions used by various platforms intermediating work performed in Australia, within and across different industry sectors. Our analysis seeks to uncover examples of terms that are specifically designated in s 25 of the ACL as being potentially unfair, as well as terms that have been identified as potentially problematic by the Australian Competition and Consumer Commission (‘ACCC’). As the law currently stands (though this will soon change), there is no prohibition on the use of such terms. Nor do we suggest that the terms in question would necessarily be found by a court to be unfair if challenged. Nevertheless, the identification of such terms provides a useful way of assessing the possible value of the ACL as a means of protecting platform workers.

We also examine the practical difficulties confronting a party seeking to challenge the fairness of contract terms, against the background of recent changes to improve the efficacy of the UCT provisions. We highlight in particular the role that the ACCC could and should be playing in seeking to protect the interests of platform workers and the need for those workers to have access to cheap, quick and effective methods of dispute resolution.

II Contracts of Adhesion in the Gig Economy

In theory, the choice and flexibility offered by digital platforms should benefit both customers and providers, with intermediaries making peer-to-peer transactions safe and efficient through innovative mechanisms for enhancing trust. This in turn has led to calls for either self-regulation or light-touch regulation of platform operations. ¹⁴ In practice, platforms frequently use standardised contractual terms, as well as website architecture and algorithms, to leverage their power over users, and there is little sign of market-based mechanisms emerging to control such behaviour. ¹⁵

That view has been confirmed by a number of studies of online platform terms and conditions. ¹⁶ One detailed review of the terms used by 102 platforms for B2B (business to business) platform transactions, conducted for the European Commission, revealed that many were unclear and/or difficult to understand. They typically sought to disclaim platform responsibility, even for culpable behaviour. Imbalanced provisions were common, especially in giving platforms a unilateral right to change the conditions of use or to delist business offerings. ¹⁷

Similar observations have been made about the terms of service adopted by various ‘crowdwork’ platforms in two studies conducted by or on behalf of the International Labour Organization (‘ILO’). The first of these, published in 2018, noted that the ‘complexity, length, and sometimes undefined nature of these documents are compounded by the fact that most terms of service are presented to users at inopportune times for reading thoroughly, even if they had the time, desire or skill to comprehend them’. ¹⁸

A more recent study analysed the terms of service used by 31 major digital work platforms, 16 of which were online web-based platforms and the other 15 location-based platforms providing taxi or delivery services. The contracts in question, which tended to be lengthy, complex and legalistic, frequently restricted worker autonomy or flexibility. For example, this might happen through exclusivity or ‘non-circumvention’ clauses purporting to tie workers and their clients to the platform for up to two years, ¹⁹ strict stipulations as to methods of performance (such as routes to be used for transport or delivery) or the terms on which work may be accepted or rejected, and restrictions on worker/client communications. Platforms were also typically accorded a broad discretion to suspend or deactivate worker accounts. ²⁰

In Australia, a survey of over 14,000 internet users in 2019 revealed that 13.1% of participants had at some time undertaken digital platform work, with 7.1% doing so at the time of the survey or within the preceding 12 months. Of those who had performed such work, 38.7% had done so ‘in person at a specified location’, 28.2% had solely done ‘computer or internet-based only work’ and 33.1% had done both types of work. Consistent with findings from other parts of the world, only 2.7% of digital platform workers in Australia derived their total income from digital platform work, with many workers participating infrequently or to supplement an alternative income. ²¹

The survey also asked a series of questions about platform operations that provided interesting insights on platform workers’ knowledge of their obligations and the terms and conditions of using the platform to access work. Approximately one-third of platform workers (29.4%) were required to display the brand or logo of the platform when performing work, and 28.2% were required to ‘be available to work either at particular times, or for at least a minimum amount of time’. Over 40% indicated they were not able to contact other workers through the platform, sub-contract their work or do further work for a client they met on the platform without using that platform. Most workers (48.4%) had to pay a set amount or percentage of their earnings to the platform, 36.1% were charged a fee or subscription to simply access work, and 31.8% paid to receive priority work opportunities. If the standard of their work was unsatisfactory, 47.9% said access to work opportunities could be restricted by the platform. If they declined work, 22.8% said they were penalised by the platform, and 21.2% did not know if they would be penalised.

A substantial minority of respondents (between 14% and 33%) answered ‘I don’t know’ to statements about platform operations, suggesting that they were unaware of the conditions set by the platform. This may be because they had not read the terms and conditions, or because the terms and conditions were unclear, or because they had not yet experienced situations that would allow them to accrue the relevant knowledge. For example, 32% did not know if their platform had a dispute settlement procedure, and 17.7% did not know if they were required to supply an ABN (Australian Business Number). These are conditions commonly outlined in platform terms and condition documents.

Specifically regarding platform terms and conditions, almost one-third (29.8%) of workers said that the platform had ‘changed my contract or terms and conditions’. Of equal relevance, however, is that another 26.6% did not know whether their terms and conditions had changed, suggesting an awareness that contractual conditions might change without their knowledge. Platforms did, however, tend to provide information on regulations with which workers have to comply (such as tax) to 41.2% of workers.

Some of the problems identified with platform work contracts can also extend to the agreement (assuming there really is one) between the worker and the client or consumer for whom the services are performed. Once again, the terms of that contract are typically dictated by the platform. Prassl and Risak, for example, have urged legislators to prohibit ‘the notorious clauses that enable the [end-user] to refuse to accept a completed task without having to give a reason and refrain from paying the advertised remuneration or provisions that the result may be kept even in those cases’. ²² For present purposes, however, we focus on the question of possibly unfair terms in the contract between the digital platform and the worker.

III Challenging Unfairness in Platform Contracts

The European Commission’s study of B2B platform transactions mentioned above contains a useful review of existing legal frameworks for regulating the fairness of contractual terms across the EU, as well as in Australia, China, Mexico, Switzerland and the US. ²³ In some of these countries, there are detailed regulations for B2B contracts that parallel those applicable to consumer transactions. In many other countries, the only controls lie in general provisions or principles regulating unfair commercial practices, bad faith conduct or unconscionability.

A prominent example of that last doctrine being used to challenge the fairness of a platform worker’s contract occurred in Canada. In Uber Technologies Inc v Heller, ²⁴ an Uber Eats driver’s contract required him to pursue any dispute he might have with the platform through arbitration in the Netherlands. Just starting the arbitration process would cost the driver most of his yearly income—and he would likely need to travel to the arbitration forum as well, imposing further costs. Given the inequality of bargaining power that had caused him to agree to so improvident a bargain, a majority of the Canadian Supreme Court held that this provision was invalid on the grounds of unconscionability.

It is far from clear that such reasoning could or would be applied in Australia. The courts here have generally taken the view that inequality of bargaining power alone is not sufficient to render a contract voidable under the equitable doctrine of unconscionable bargains. ²⁵ Nor could it be assumed that having terms imposed by a platform, more or less at will, would contravene the statutory prohibition on unconscionable conduct in s 21 of the ACL, given the fairly high bar for such findings set by the High Court. ²⁶

The more obvious option in Australia is to invoke a statutory jurisdiction to review the fairness of contract terms. In theory, there are a number of regimes that could potentially apply, but many of these have significant limitations. Some State industrial laws, for example, allow unfair terms in contracts relating to the performance of work (including by independent contractors) to be varied or set aside. ²⁷ But these laws cannot generally be invoked against incorporated businesses because of the pre-emptive provisions in s 26 of the Fair Work Act 2009 (Cth) and Part 2 of the Independent Contractors Act 2006 (Cth). There might be a similar problem with contractors trying to invoke the jurisdiction of the Victorian Civil and Administrative Tribunal (VCAT) to grant relief against unfair contracts in resolving ‘trader-trader disputes’ over sums of less than $10,000. ²⁸

Part 3 of the Independent Contractors Act 2006 superficially looks more promising. It allows a party to a contract for services that involves the performance of work by an independent contractor to complain to the Federal Court or the Federal Circuit and Family Court that the contract is ‘harsh’ or ‘unfair’. If the court agrees, it may vary or set aside all or part of the contract. The problem, however, is that to the extent that a platform worker is a contractor, the relevant contract for services may well be with the end-user, not the platform. ²⁹ It might still be argued that the agreement with the platform to facilitate that transaction is a ‘collateral arrangement that relates to a services contract’, which under s 5(4) of the 2006 Act may be taken to be part of that contract. But the scope of that concept has not been tested. ³⁰

The much more straightforward possibility, on which this article focuses, is that the fairness of a standard form contract between a worker and a platform may be reviewed under Part 2-3 of the ACL. Since being amended with effect from November 2016, ³¹ the UCT provisions have extended beyond consumer transactions to cover a ‘small business contract’. ³² At the time of writing, s 23(4) of the ACL defines this to include a contract for the supply of services to which one of the parties is a business employing fewer than 20 persons, provided the price payable for those services does not exceed $300,000 or for a contract of more than 12 months’ duration, $1,000,000. But a contract of employment is not reviewable, since the definition of the term ‘services’ in s 2(1) of the ACL specifically excludes ‘the performance of work under a contract of service’.

Given the unlikelihood of a platform worker employing 20 workers of their own, or earning more than $300,000 per contract, any work arrangement that does not involve employment should readily qualify as a small business contract. Importantly, this would be so whether the worker concerned was considered to be supplying services to the platform or acquiring services (in the form of technological support) from that business. Given too what has been said about the control typically exercised by platforms over their terms of use, it is also likely that the contract will satisfy the ‘standard form’ requirement in s 23(1)(b). ³³

A term in such a contract cannot be challenged, however, if it defines the main subject matter of the contract, sets the price payable for the services or is required or expressly permitted by a Commonwealth, State or Territory law (s 26(1)).

For a term to be treated as unfair, three requirements must be satisfied (s 24(1)). The term must (a) cause a ‘significant imbalance’ in the parties’ contractual rights and obligations, (b) not be reasonably necessary to protect the legitimate interests of the party advantaged by the term, and (c) have the potential to cause financial or other detriment if applied or relied upon. In applying those tests, a court must have regard to the extent to which the term is ‘transparent’ and to the contract as a whole (s 24(2)). A term is considered to be transparent if it is expressed in plain language, legible, presented clearly and readily available (s 24(3)). ³⁴ It is for the party asserting unfairness to establish the significant imbalance and detriment requirements, but there is a rebuttable presumption that a term is not necessary for the legitimate protection of the party seeking to rely on it (s 24(4)). ³⁵

Importantly for the design of our research, s 25 provides a non-exhaustive list of the kinds of terms that may be treated as unfair, though they are not automatically to be treated as having that character. ³⁶ In the case of a contract for services, these include terms:

(a) Permitting one party only to avoid or limit their performance;

(b) permitting one party only to terminate the contract;

(c) penalising one party only for breach or termination of the contract;

(d) permitting one party only to vary the terms;

(e) permitting one party only to renew or not renew the contract;

(f) permitting one party to vary the price payable for the services without the other being able to terminate;

(g) allowing one party to vary the characteristics of the services being supplied;

(h) permitting one party to determine the meaning of the contract or whether it has been breached;

(i) limiting a party’s vicarious liability for wrongdoing by their agents;

(j) allowing one party to assign the contract to the other party’s detriment, without their consent;

(k) limiting one party’s right to sue the other;

(l) limiting the evidence that can be adduced in proceedings relating to the contract; or

(m) imposing an evidential burden on one party in such proceedings.

In a 2016 review of standard form contracts used by large businesses (including Uber) in selected industries, the ACCC discussed independent contracting arrangements. In that context, it highlighted particular concerns with four types of term that it had encountered: ³⁷

• Unilateral powers to vary the terms of the contract;

• provisions unreasonably limiting one party’s liability, and/or requiring one party to indemnify the other in a wide range of situations, including for loss or damage caused or contributed to by that other party;

• clauses conferring ‘inappropriate’ powers of termination; and

• misleading statements about legal rights.

Pending the commencement of the amendments outlined in Part VI of this article, the remedies available in relation to terms identified under Part 2-3 of the ACL as being unfair are extremely limited. A party to the relevant contract, or the regulator (which for federal purposes means the ACCC), ³⁸ may seek a declaration from a court under s 250 that a particular term is unfair. The effect of such a declaration is to treat that term as void. The rest of the contract may remain in operation, provided it can operate without the unfair term (s 23(2)). Section 15 makes it clear, however, that merely including an unfair term in a contract is not taken to be a contravention of the ACL, so that the general remedies in Part 5-2 for contraventions of the ACL are not engaged. At most, an injunction or an order for compensation may be sought against a person that purports to apply or rely on a term after it has been declared unfair under s 250 (ss 232(3) and 237(1)(a)(ii)). Section 239 also permits the regulator to seek redress against a party advantaged by a term that has been declared unfair, on behalf of any class of persons not involved in the litigation who have been caused to suffer loss or damage by the term or who are likely to suffer such loss. A court may make any order it thinks appropriate to redress, prevent or reduce the loss or damage, though not an award of damages.

It is far from clear, however, how widely a declaration of unfairness can be considered to operate. In particular, it is uncertain whether a party could be penalised under the existing regime for using a slightly different version of the unfair term, or the same term in a slightly different contract. ³⁹

Despite these uncertainties, the ACCC has been fairly active in seeking to challenge unfair terms in at least some types of small business contract. ⁴⁰ It has sought a number of declarations of unfairness in the courts, ⁴¹ although not always with success. ⁴² It has also extracted enforceable undertakings from businesses to vary specified terms. ⁴³ Part 2-3 of the ACL has also been invoked in the course of private litigation involving B2B transactions, including at least one case involving a digital platform. ⁴⁴

IV Review of Platform Work Contracts: Methodology

In order to test out the potential for the current version of the UCT provisions in the ACL to be used to challenge the fairness of a contract between a digital labour platform and a worker using the platform to find work, we analysed the terms and conditions of 15 digital platforms operating in Australia. Each platform allows individuals to earn an income by working or offering services in person at a specified location or by doing work that is computer or internet based.

The terms and conditions of the 15 platforms were downloaded and saved offline in October 2021 and analysed for the inclusion of terms that fall within the categories designated in s 25 of the ACL as potentially unfair.

The 2019 national survey on digital platform work in Australia, to which reference was made earlier, identified over 100 different platforms through which workers in Australia were registered to undertake work, ⁴⁵ as set out in the Appendix to this article. The platforms most commonly used by platform workers to find work were Airtasker (34.8%), Uber (22.7%), Freelancer (11.8%), Uber Eats (10.8%), Deliveroo (8.2%), Ola Cabs (7.2%), Upwork (6.2%) and Fiverr (6.0%). ⁴⁶ These platforms were included in the sample for this study. Four are transport and food delivery services; however, the survey identified 11 other types of work being done via digital platforms in Australia. ⁴⁷ We thus extended our sample to select platforms that represented the variety of digital platform work and a mix off in-person and online-only services. We also chose to focus on businesses that were receiving significant media attention, were identified in the national survey or were large digital platforms recognised globally. Our first sample contained 16 platforms: Airtasker, Amazon Flex, Careseekers, Deliveroo, Fiverr, Freelancer, Guru, Helpling, hipages, Mable, Mad Paws, Ola, PeoplePerHour, Uber, Uber Eats and Upwork.

For each platform, we accessed the publicly available terms and conditions via the platform website. In each case the terms and conditions were downloaded, saved, read and checked for their applicability to workers. Hyperlinks to additional terms or policies were embedded in the terms and conditions of 8 platforms. These additional policies were subsequently downloaded and where relevant included in the analysis. Privacy policies, payment services terms, community or other ‘guidelines’ and intellectual property policies were excluded from the analysis.

During this process, we identified that the terms and conditions of Deliveroo related only to customers who were ordering services via the website/platform. Deliveroo, which has since ceased trading in Australia, ⁴⁸ was subsequently excluded from the sample because terms and conditions that applied to the worker were not publicly available. Additionally while the terms and conditions of Amazon Flex, Uber and Uber Eats applied to all users (customers and drivers), these terms and conditions also made reference to additional worker/driver agreements, which were not available without registering as a worker. The analysis of the terms and conditions of Amazon Flex, Uber and Uber Eats is therefore limited to terms that apply to all users. In the case of Uber and Uber Eats, the same set of terms and conditions applies to all users who ‘access any application, website, content, product or service by provided by Uber’, including UberX, Uber Eats and a range of other Uber services.

Table 1 summarises the final sample of 15 platforms against the types of work they offer.

Table 1.

Sample platforms by type of work.

Type of Work		Platforms
Transport and food delivery	For example, taxi services; food delivery; package or goods delivery	Amazon Flex, Ola and Uber/ Uber Eats
Professional services	For example, accounting; consulting; financial planning; legal services; human resources; project management	Guru, Freelancer, PeoplePerHour and Upwork
Odd jobs and maintenance work	For example, running errands; general maintenance duties; removalist	Airtasker and Helpling
Writing and translation	For example, academic writing; article writing; copywriting; creative writing; technical writing; translation	Airtasker, Guru, Freelancer, Fiverr, PeoplePerHour and Upwork
Clerical and data entry	For example, customer service; data entry; transcription; tech support; Web research; virtual assistant	Airtasker, Freelancer and Upwork
Creative and multimedia	For example, animation; architecture; audio; logo design; photography; presentations; voice overs video	Airtasker, Freelancer, Fiverr, PeoplePerHour and Upwork
Software development and technology	For example, data science; game development; app, software or web development; server maintenance; web scraping	Freelancer, Fiverr, PeoplePerHour and Upwork
Caring	For example, aged or disability care; pet services; babysitting; nanny services	Airtasker, Careseekers, Mable and Mad Paws
Skilled trades work	For example, carpentry; plumbing; electrical work	Airtasker and hipages
Sales and marketing support	For example, social media marketing; ad posting; lead generation; search engine optimisation; telemarketing	PeoplePerHour, Fiverr and Upwork

In October 2021, terms and conditions of each platform were reviewed and coded against the categories set out in s 25(a)–(m) of the ACL, aiming in each instance, to identify provisions that might potentially be regarded as unfair. Additional provisions that might arguably be seen as UCT, but which did not map onto any of those categories, were identified in a second stage of coding. In the final stage, the identified terms were further classified by reference to the four categories of concern identified by the ACCC in 2016: unilateral variation; limited liability and wide indemnities; termination powers; and misleading statements about rights at law. ⁴⁹ This iterative approach to coding produced some evidence of potential UCT in the terms and conditions of all digital platforms, including evidence of the four categories of terms identified by the ACCC as particularly concerning in relation to independent contracting arrangements.

V Examples of Potentially Unfair Terms

The analysis uncovered examples of terms designated in s 25 of the ACL as potentially unfair. The findings from the coding and analysis of the terms in all 15 platforms are presented below.

It is also notable that for all food and transport platforms in our sample, and for the platform Helpling, the terms and conditions specific to workers were not immediately or easily accessible. Additional interrogation of the platform website was required to locate these terms, and in some cases, the terms could not be located at all without commencing the process to register as a worker. This suggests that individuals must first provide the platform with personal information and a tacit commitment to commence work as a driver or rider before they are able to receive and fully review the terms and conditions which will govern their work arrangements. ⁵⁰

A Section 25 Examples Found

As Table 2 shows, the terms and conditions of all 15 platforms contained provisions that would likely be regarded as falling within the categories designated in s 25 of the ACL as potentially (though not necessarily) unfair.

Table 2.

Number of platforms with likely examples of s 25 terms.

	Likely	Unlikely
s 25(a)—permits only one party to avoid or limit performance	15	0
s 25(b)—permits only one party to terminate	3	12
s 25(c)—penalises one party for breach or termination	15	0
s 25(d)—permits one party to vary terms	14	1
s 25(e)—permits only one party to renew or not renew	0	15
s 25(f)—permits only one party to vary upfront price	12	3
s 25(g)—permits only one party to vary characteristics of services supplied	11	4
s 25(h)—permits only one party to determine breach or interpret meaning	13	2
s 25(i)—limits one party’s vicarious liability for wrongdoing by agents	13	2
s 25(j)—permits one party to assign contract without consent	9	6
s 25(k)—limits one party’s right to sue	14	1
s 25(l)—limits evidence one party can adduce	0	15
s 25(m)—imposes evidential burden on one party	0	15

The most common terms to appear were clauses that permit only one party to avoid or limit performance (s 25(a)) and clauses which penalise one party for breach or termination (s 25(c)). The terms and conditions of all 15 platforms contained such provisions. A range of terms and conditions were detected that had the effect of limiting performance in either the provision of services advertised as offered by the platform (such as worker verification) or limiting the availability of the site or features provided by the platform.

In all cases, platform terms and conditions provided the platform with the power to terminate (or suspend) a worker’s use of the platform and to determine the material circumstances in which this occurred. Termination by the platform was usually accompanied by an additional penalty, especially if termination was the result of a breach of the terms and conditions. For example, Freelancer’s terms permitted a fine of up to US$3,000 to be imposed for each breach, as a ‘reasonable pre-estimate or minimum estimate’ of the loss likely to be suffered by the platform. It was also stipulated that the user could bring ‘no claim whatsoever’ in relation to any suspension or termination of their account.

Appearing less frequently were examples of s 25(b). Only three platforms contained clauses that permitted only one party to terminate (that party being the platform). These clauses and those relating to unilateral variation are discussed in more detail below.

The only categories not to appear in any of the platform terms and conditions analysed were permits only one party to renew or not renew (s 25(e)), limits on adducing evidence in proceedings concerning the contract (s 25(l)) and the imposition of evidential burdens on one party (s 25(m)).

Table 3 summarises our findings on s 25 by platform. The terms and conditions of Mable, Guru, Freelancer, Ola, PeoplePerHour, Uber and Uber Eats contained examples of nine or more s 25 categories. No fewer than six s 25 categories were found in any of the platforms reviewed.

Table 3.

Likely s 25 examples by platform.

In the sections that follow we discuss some of the examples we found, by reference to the ACCC’s four categories of concern.

B Unilateral Variation

The ACCC noted that ‘[t]he most common clauses of concern in independent subcontracting agreements were clauses that allowed the large business to unilaterally vary the agreement at any time’. ⁵¹ The concern is not limited to contracts of this type. In many of the cases to date in which contract terms have been found to be unfair under Part 2-3 of the ACL, such provisions have been among those successfully challenged. ⁵²

All but one of the digital labour platform contracts in our sample (Helpling being the exception) contained clauses that enabled the platform to unilaterally vary the terms and conditions of use. The terms used by some platforms, however, did stipulate they would provide notice of any change.

C Limitation of Liability and Wide Indemnities

All the platform contracts we reviewed contained broad exclusions or limitations on liability. Seven platforms had provisions that capped their liability in various ways. For example, Careseekers imposed a limit of ‘the lesser of the total amount paid by you (if any) for any chargeable service or feature on the Website or $500’. Airtasker's liability was likewise limited to the total paid by the user during the 12 months prior to the incident causing liability, or AUD$50.00, whichever was greater. The remaining platforms did not limit their liability to a set amount but sought to exclude it completely, at least in certain respects. Some of these exclusions were exceptionally broad, as in the case of Fiverr:

In no event will Fiverr, its affiliates or their licensors, service providers, employees, agents, officers or directors be liable for damages of any kind, under any legal theory, arising out of or in connection with your use, or inability to use, the website, any websites linked to it, any content on the website or such other websites or any services or items obtained through the website or such other websites, including any direct, indirect, special, incidental, consequential or punitive damages, including but not limited to, personal injury, pain and suffering, emotional distress, loss of revenue, loss of profits, loss of business or anticipated savings, loss of use, loss of goodwill, loss of data and whether caused by tort (including negligence), breach of contract or otherwise, even if foreseeable.

Twelve of the platforms (Amazon Flex, PeoplePerHour and Fiverr being the exceptions) imposed requirements on workers to indemnify the platform against certain types of loss or damage. Some of these were simple party–party indemnities, covering loss caused by any breach by the worker of their obligations. ⁵³ But many others were expressed to operate in a much broader range of situations, which might not necessarily involve any wrongdoing on the part of the worker. For instance, workers are required to hold Mable and its affiliates harmless from any claims, liabilities, losses or expenses arising from, among other things, use of the platform (including any information posted or transmitted), or the provision of care services.

By contrast and unlike the majority of platforms, the indemnity required by hipages was expressed not to apply to the extent that any loss, claim or liability arose out of gross negligence or wilful default on the part of the platform or its representatives.

D Termination Powers

All platforms reviewed had broad powers to terminate contracts by terminating access to the platform’s website and/or app. Three platforms did not provide the user (whether a worker or client) a reciprocal right to terminate. These were Amazon Flex, Fiverr and PeoplePerHour.

Even where a reciprocal right to terminate was provided, contracts varied between permitting termination for any reason whatsoever and detailing a range of circumstances that might warrant termination. Common reasons included violation of the terms and conditions, violation of laws, cancellations of accepted tasks/bookings and ‘problematic’ ratings and reviews. In these cases, the contracts provided the platform with absolute discretion to determine whether the situation had occurred and to suspend or terminate the contract.

Some platforms provided detailed lists of potential violations. For example, Fiverr listed 11 potential violations that could result in termination. Ola’s terms and conditions for drivers provided two pages of specified service levels, which included a ‘Zero Tolerance Policy’. Seven of the fifteen platforms provided, in a separate document, an additional ‘Code of Conduct’ or ‘Community Guidelines’ for all users of the platform. Contracts could be terminated for breach of these community standards.

E Misleading Statements About Legal Rights

The presence of misleading statements about legal rights is necessarily harder to detect, given that the characterisation of a provision as having that effect is likely to involve contestable assumptions. One seemingly obvious candidate for examination is a term denying that a worker is an employee of the platform through which they are finding work. Eleven platforms had such terms. However, by definition such a provision cannot be an unfair term for ACL purposes, at least to the extent that it seeks to characterise the platform–worker relationship. If the term were found to be misleading, because the worker was in fact employed by the platform, then as explained earlier the contract would not be one for ‘services’ and hence would not be covered by the ACL! ⁵⁴

A more likely candidate for challenge on this ground might be terms seeking to characterise the relations between the parties involved in ways that do not necessarily accord with legal or practical reality. For example, Ola’s terms insisted that it was ‘merely an intermediary providing online marketplace services’ and that its portal was ‘only a platform where the Driver may offer the Transport Services to Customers’. Likewise, Mad Paws insisted not merely that it was not a party to any agreement between the sitters and owners using its platform, but that it ‘has no control over, and is not responsible for’ the actions of those persons (or indeed their pets). Based on previous court rulings as to the nature of a ridesharing platform’s business, ⁵⁵ the Ola terms in particular might be susceptible to legal challenge. However, we did not consider for the purposes of our review that it was possible to form any clear opinion on the matter.

A further type of provision identified as potentially falling within this category by the ACCC in its 2016 review was an entire agreement clause. If a set of written terms were presented as exhaustively setting out the parties’ rights and obligations, this might ‘mislead a subcontractor in instances where they may have additional rights at law (for instance, based on pre-contractual representations made by the larger business)’. ⁵⁶ Our review uncovered examples of such provisions in 10 of the 15 contracts. For instance, Mable’s terms were stated to supersede ‘all prior oral or written understandings, communications or agreements’.

Terms that restrict the jurisdiction of legal proceedings or require legal proceedings in a foreign court might also constitute a misleading statement about legal rights. Examples of such clauses were identified in the conditions imposed by eight of the platforms under review. As a general proposition, there is nothing improper about the parties to a contract agreeing that their dealings are to be regulated by the laws and/or courts of a country in which only one (or even neither) of the parties resides. ⁵⁷ It is, however, a different matter with ‘mandatory’ statutes—such as the ACL itself—which are either expressed or can be construed to have effect on parties or transactions within a particular jurisdiction, regardless of any choice of law to the contrary. ⁵⁸ Hence it is possible that a platform contract might be misleading in suggesting the non-application of local laws of that sort, or in seeking to deny the jurisdiction of a court or tribunal in the same jurisdiction to resolve disputes over their effect.

VI Practical Difficulties in Challenging Unfair Terms

In November 2020, Commonwealth, State and Territory consumer affairs ministers agreed to improve the remedies available in respect of UCT. ⁵⁹ The necessary amendments to the ACL have now been passed, as part of Schedule 2 to the Treasury Laws Amendment (More Competition, Better Prices) Act 2022 (Cth). ⁶⁰ As from 9 November 2023, these will:

• expand the definition of a ‘small business contract’ to cover a contract made by a business with fewer than 100 employees or an annual turnover of less than $10 million, and with no cap set by reference to contract price;

• make it clear that repeat usage of a contract must be taken into account in determining whether it is in a standard form;

• clarify that a contract may still qualify as in ‘standard form’, even if the other party is allowed to negotiate minor or insubstantial changes, or can select a term from a range of pre-determined options;

• prohibit the inclusion or use of a term subsequently found to be unfair, with contraveners exposed to penalties of up to $10 million (or in certain cases potentially more) per breach; and

• broaden the power of the courts to grant redress for loss suffered or likely to be suffered through the use of unfair terms, including by prohibiting the use of similar terms in future contracts. ⁶¹

These reforms will not just enhance the scope and effectiveness of the relief available to those who challenge the fairness of terms in consumer and small business contracts. In theory, they should increase the incentives to take action against UCT. However, what the announced changes do not address are the potential barriers to accessing any new and enhanced remedial regime.

As the European Commission has noted in the context of regulating digital platform contracts, even large businesses that are heavy users of platform services will rarely commence litigation to resolve disputes about the fairness of their terms of use, given the cost and uncertainties involved. ⁶² Those problems are likely to be magnified in the case of self-employed platform workers. This is especially true given that gig workers are disproportionately likely to be young and/or migrants working on temporary visas, two groups who tend to have limited knowledge about their rights, and how to defend them. ⁶³

In a number of countries, proceedings relating to UCT can be initiated by professional associations or government bodies. ⁶⁴ But that does not guarantee that action will be taken. The ACCC, for example, has been willing to investigate the fairness of terms imposed by digital platforms on some users, such as restaurants producing food to be delivered to customers. ⁶⁵ In 2019, its concern about the imposition of unfair terms by digital platforms in relation to both consumers and business users led it to recommend (or, more accurately, reiterate its previous support for) the introduction of a prohibition against the use of UCT. ⁶⁶ In the course of its earlier review of UCT in small business contracts, it had also engaged with Uber over its standard Driver Agreement, persuading the platform to limit the circumstances in which it could terminate the contract. ⁶⁷ That seems to have been a rare example of the ACCC concerning itself with platform work contracts. The agency has not otherwise demonstrated any interest in seeking to protect self-employed platform workers, even though, as explained above, they clearly seem to fall within the unfair term regulations in the ACL. It has reported receiving very few inquiries from platform workers, suggesting that they are more likely to channel any concerns to bodies such as the Fair Work Commission. ⁶⁸

In 2020, the Victorian Government published a report of an inquiry into on-demand work conducted by the former Fair Work Ombudsman, Natalie James. Responding to preliminary findings from our research, as well as other evidence as to contractual arrangements for such work, the report addressed the adequacy of existing regimes for remedying unfair terms. In relation to the ACL, it observed that while the ACCC might be ‘highly visible in supporting small businesses … platform workers do not see this agency as a natural “go to” in seeking help’. ⁶⁹ Given the potential confusion about the interaction of different UCT regimes, including the one administered by VCAT, ⁷⁰ the report concluded that the ‘pathways’ in question were ‘not suitable to ensure fairness or transparency for platform workers’. ⁷¹

The report recommended, among other things, that governments ‘clarify, enhance and streamline’ the relevant laws so that they are accessible to ‘low-leveraged workers’ and enable ‘system-wide scrutiny’ of platform work contracts. ⁷² It also proposed that an appropriate agency be ‘sufficiently resourced to provide effective support to self-employed platform workers and to prioritise actions against systemic deployment of unfair contracts involving these workers’. ⁷³ This might be an existing agency (such as the ACCC or the Australian Small Business and Family Enterprise Ombudsman, or an equivalent body at State level) or a new ‘Streamlined Support Agency’. ⁷⁴

In response to this recommendation, the Victorian Government has made it clear that since ‘[t]he Commonwealth controls many of the levers to change laws that are relevant to on-demand work’, its priority will be to advocate for national reform on this and other issues. ⁷⁵ Nevertheless, it has established a Gig Worker Support Service to allow Victorian gig workers to ‘obtain information about conditions they should expect to be offered by platform companies’. ⁷⁶

Another step the Victorian Government has taken is to promulgate a set of Fair Conduct and Accountability Standards for digital platforms engaging ‘non-employee on-demand workers’. ⁷⁷ These will initially be voluntary, but the government is ‘planning to introduce laws in 2023 to ensure compliance’. ⁷⁸ Among other things, a standard on ‘bargaining power’ encourages platforms to ensure that the terms of any applicable contract are ‘clear and able to be understood’ and ‘consistent with the nature of the actual engagement’. ⁷⁹ Platforms should consult with workers over the fairness of any proposed terms, for example, as to liability for damage arising from the performance of platform work, and be willing to amend the contract or consider other ways of fairly distributing that or other risks. There should also be a process for workers to challenge platform decisions that affect their terms and conditions, or that find them to be in breach of their obligations. ⁸⁰ The new Support Service will be tasked with maintaining a register of platforms that sign up to the Standards. ⁸¹

It remains to be seen how and to what extent both the Victorian Government’s plans to regulate in this area and the role of the ACL’s UCT provisions are affected by the election of a federal Labor government that is far more interested in protecting gig workers than its conservative predecessor. As previously mentioned, ⁸² the Albanese Government has indicated that it will seek to empower the Fair Work Commission to regulate wages and working conditions for at least some types of platform worker. That change, or more especially what the Commission does with its augmented powers, might well ‘crowd out’ any new form of State regulation. It might also, conceivably, reduce the scope and utility of any litigation against UCT. But it seems more likely that there will still be opportunities under that regime to address issues with contractual provisions on matters that are not covered by any new Commission standards.

VII Conclusion

The concept of fair contracts for platform workers is one of the five ‘pillars’ of the Fairwork Principles adopted by the Fairwork Foundation, established by the Oxford Internet Institute in partnership with a large number of organisations around the world. ⁸³ Besides calling for terms to be ‘transparent, concise, and provided to workers in an accessible form’, the Foundation has identified the need for contracts to clearly identify the party contracting with the worker and to be free of ‘clauses which unreasonably exclude liability on the part of the platform’. ⁸⁴

In Europe, a major initiative from the European Union has sought to respond to this challenge. The Platform to Business (‘P2B’) Regulation ⁸⁵ imposes detailed obligations in relation to the terms and conditions governing the relationship between online intermediation services and ‘business users’, a term which is broad enough to include self-employed workers. ⁸⁶ Among other things, terms must be ‘drafted in plain and intelligible language’, be easily available to users (including before any contract is formed), and set out the grounds for terminating, suspending or restricting use of the intermediary’s services (Art 3(1)(a)–(c)). Changes must be clearly notified, with users given at least 15 days to decide whether to terminate the contract (Art 3(2)). Decisions to restrict, suspend or terminate use of the platform must be communicated in writing and users given an opportunity to respond (Art 4). The criteria for ‘ranking’ goods or services (i.e. the relative prominence given to one user’s offerings over those of others) must be clearly outlined (Art. 5). So too must any differentiation between users (Art 7), any restrictions on the ability of users to offer the same goods and services to customers by different means (Art 10), and the right of users to terminate the contract (Art 8(b)). Online intermediation services are also generally prohibited from making retroactive changes to terms and conditions (Art 8(a)).

At present, Australia has no equivalent to those provisions. In the absence of such explicit controls, or of effective forms of collective bargaining over the terms on which platform workers are engaged, ⁸⁷ the ACL’s UCT regime may have an important part to play in controlling the use of superior bargaining power by digital labour platforms to impose one-sided and unfair conditions.

As our analysis has shown, each of the 15 platform work contracts we reviewed contained multiple instances of provisions likely to fall within the categories of potentially unfair terms set out in s 25 of the ACL. Many of those contracts also exemplified the concerns expressed by the ACCC in its 2016 review, concerning powers of unilateral variation, broad indemnities or limitations on liability, inappropriate powers of termination, and (though less clearly) misleading statements about legal rights. We stress that our analysis does not and cannot suggest that the terms we have identified would necessarily be regarded by a court as unfair, once full account is taken of the context and of the need to satisfy the criteria set out in s 24 of the ACL. What is potentially unfair may not actually be unfair. That said, it is hard to imagine that, at the very least, the unilateral powers of variation found in all but one of the contracts we examined would not be struck down if challenged.

The changes that have been introduced to the UCT regime in the ACL, not least the provision for the imposition of penalties for merely proposing unfair terms, are likely to create greater incentives for legal challenges to the fairness of platform work contracts. But what seems to us to be of equal importance is pursuing three other reforms.

One is to impose a positive obligation on platforms to ensure that workers can access and understand the contractual terms that will govern their use of the platform and the performance of their work. Our research revealed that many platforms spread their conditions across multiple documents, some of which were not easily accessible by those still considering whether to sign up for work. In our view, transparency is not just a matter to be considered in determining the fairness of contractual terms, as it presently is under s 24(2)(a) of the ACL. Making all terms transparent, easily accessible and understandable for workers should be a regulatory objective in its own right.

A second goal should be to ensure that an appropriate regulator is empowered and resourced to bring the same scrutiny to bear on platform work contracts that the ACCC has done for many other types of small business contract over the past six years. If the ACCC is not willing or able to perform that role, it should be given to another agency—possibly the Fair Work Ombudsman.

A third reform should be to provide platform workers with access to cheap, quick and effective dispute resolution over issues concerning the fairness of their terms. The Fair Work Commission is now empowered to resolve disputes over bullying and sexual harassment for self-employed workers as well as employees. ⁸⁸ There is no reason why the Commission could not be made an option for the resolution of unfair contract claims as well. ⁸⁹ It would also be appropriate to allow trade unions or others to commence class actions on behalf of workers, as can now, for instance, be done under enhanced provisions for representative applications in relation to the breach of federal anti-discrimination laws. ⁹⁰ The disparities of power and resources that create unfairness in contract terms in the first place must surely also be considered when designing appropriate mechanisms for relief.

VIII Postscript

Since this article was written, the Albanese Government has introduced the legislation foreshadowed in Part I. The Fair Work Legislation Amendment (Closing Loopholes) Bill 2023 (Cth) (‘Closing Loopholes Bill’) would empower the Fair Work Commission to set minimum standards for ‘employee-like’ workers who are engaged through digital labour platforms, as well as to hear complaints from such workers (if earning less than a prescribed income threshold) about ‘unfair deactivation’ from a platform. It is also proposed to add a new Part 3A-5 to the Fair Work Act 2009 (Cth). This would allow independent contractors earning below the threshold to seek relief from the Commission against any unfair contractual term dealing with what, in an employment context, would be a ‘workplace relations matter’. However, the contractor would need to be working under a ‘services contract’, and the term(s) being challenged would need to be either in that contract or a collateral arrangement. For platform workers, that might raise the same issues we outlined concerning the operation of Part 3 of the Independent Contractors Act 2006 (Cth), on which the proposed new provisions are closely modelled. It is also unclear whether some of the potentially unfair terms discussed in this article, such as overly broad indemnities or powers of unilateral variation, would fall within the concept of a workplace relations matter. To the extent then that platform workers might still need to rely on the ACL to challenge the fairness of the terms on which they are permitted to seek work, the Closing Loopholes Bill does not do enough to address the concerns we have outlined in Part VII of the article.