Why some say don’t regulate

The argument for not regulating is one favoured by supporters of free markets, and, of course, the gig firms themselves.

The first strand of this argument is that ‘gig work’, ‘independent contracting’ or ‘freelancing’ (Freelancers Union and Elance-oDesk 2013), is the way of the future – and nothing should be done to try to stem this inevitable tide. It represents a revolt by workers and firms against the restrictions imposed by the employment relationship and the regulations that go with it (Phillips Reference Phillips2008). Regulating it would be swimming against the tide of history and ignoring participants’ wishes to be free of the strictures of employment. One recruitment agency blogged, in 2016, that ‘by 2020, 40% of the US workforce is expected to be independent contractors’ (Trakstar Hire 2016).

A second line of argument is that regulating gig work would hamper innovation. Innovation is a driver of productivity, and so a third line of argument is that regulating gig work would hamper productivity, thus reducing the capacity for Australia to enjoy higher living standards.

It is also argued that gig work provides employment that would otherwise not exist, and so regulation of the pay and conditions in it will hamper employment. Another proposition is that regulating gig work would hamper the flexibility of labour. Closely related to this is the argument that gig workers themselves do not want regulation; in particular, gig workers do not want to be turned into employees, as they value the flexibility that their contractor status brings.

Why some say regulate as employees

Those who wish to see the gig work regulated focus on the vulnerability of gig workers, and the capacity of gig companies to use this vulnerability to undercut the pay and conditions of other workers, including by driving competitors out of business. Their vulnerability, and the threat to other workers, arise from the low pay afforded to gig workers and the difficult conditions under which they work, reflecting a power imbalance between them and the large firms.

As for how they should be protected, responses broadly fall into two categories. The first involves changing the employment status of gig workers from contractors to employees, as the control exercised by gig firms is similar in degree to the control exercised by employers over employees. The exclusion from employee status denies gig workers the protections and entitlements of employees, including annual and sick leave and protection from unfair dismissal. It is only through being classed as employees that they can receive adequate protection.

Why some say regulate as contractors

The alternative approach to regulation also takes, as a starting point, that vulnerability and low pay of gig workers, but questions whether this can only be done through the employment relationship. It argues that adequate regulation of gig workers can be achieved without their becoming employees. This is especially the case because of the attitudes of gig workers themselves, which are opposed to being turned into employees, and the greater political opposition that comes from doing so. Regulating gig workers as contractors, in this view, increases the opportunities for developing consensus between participants and hence for the sustainability of regulation. The experience of the regulation of road transport in New South Wales (via Chapter 6 of the Industrial Relations Act (NSW)) is cited as an example.

Way of the future

Is self-employment – the modus operandi of most gig work – inevitably increasing? Over the period from the 1980s to present, available OECD data for 36 countries (34 OECD plus Brazil and Russia) show that 15 countries had a fairly continuous decline in self-employment. Amongst these was the USA, in which self-employment averaged well below 10% in every decade. A further eight countries had general, but interrupted, declines in self-employment, and only four showed any sustained growth, while the remaining eleven had mixed outcomes or insufficient data. Within Australia, both Australian Bureau of Statistics (ABS) and Household, Income and Labour Dynamics of Australia (HILDA) survey data showed declines in self-employment both for those with employees and also (to a smaller extent) for those without employees. Amongst full-time workers, sole self-employment has declined. Yet amongst part-time workers, solo self-employment has grown, consistent with the gig economy’s need for part-time independent contractors (Peetz Reference Peetz2023).

The idea of self-employment as the way of the future is wrong because the employment relationship is still the most efficient way for capital to exert control over labour. Moreover, the gig economy is not as big as many people think: probably around 1%, maybe less, of the workforce are regularly engaged in it (Katz and Krueger Reference Katz and Krueger2016; Wilkins et al. Reference Wilkins, Vera-Toscano, Botha, Wooden and Trinh2022; Australian Bureau of Statistics 2003), though many others dip their toes into it and then take them out again (Chartered Institute of Personnel and Development 2017), the numbers are likely growing, but they are presently exceeded by the numbers of independent contractors and comparable workers who do not rely on apps for their work.

Innovation, competition, and productivity

Innovation is important in delivering improvements in living standards, either through higher productivity or new products that consumers prefer. Innovation to circumvent employment laws and reduce pay or conditions does not improve productivity (a reduction in labour costs will reduce, not increase, productivity over the medium term (Salter Reference Salter1969)), but it may produce a new product that consumers prefer (in which case the impact on living standards is ambiguous). Likewise, competition may increase productivity but alternatively it may reduce pay, depending upon whether, in the particular market, competition is on the basis of price (in which case, labour costs are driven down by competition) or product quality (in which case effort focuses on design or product features, and enhanced productivity). It is possible to innovate and promote competition in ways that improve consumer welfare without reducing pay and conditions. For example, an app that enables consumers to more quickly or efficiently access transport services or disability care, without reducing the incomes of those providing the services, would be an unambiguous improvement in living standards. If that cannot be achieved, it is not clear that an overall improvement in living standards has occurred. So, regulation should direct innovation to achieving social ends (higher consumer welfare without a reduction on worker pay and conditions).

Employment

It is unclear whether the availability of low-cost labour through an unregulated gig economy would lead to higher employment. The argument that it would do that is similar to the argument that removing minimum wage laws would have that effect. While this was a common view amongst economists in the 20^th century, it is seriously contested now. This rethinking commenced with Card and Krueger’s (Reference Card and Krueger1995) analysis of minimum wage changes at the state level in Myth and Measurement, and since then empirical evidence has not supported the ideas that reducing minimum wages and conditions would boost employment (Bishop Reference Bishop2018), that minimum wage increases in Britain have damaged employment there (Metcalf Reference Metcalf2008) or that minimum wage increases hamper economic performance (Winkler Reference Winkler2019).

Flexibility

If flexibility is a desirable outcome, we first need to ask at what level this is being sought. At the macro-level, ‘labour market flexibility’, may be a desirable objective to the extent that it leads to better economic outcomes (e.g. in unemployment or inflation). Indeed, the Organisation for Economic Cooperation and Development (1994) considered this to be a laudable aim of public policy up until the global financial crisis. After that, the OECD observed it found no evidence that structural reform policies to promote flexibility had made labour markets ‘less sensitive to severe economic downturns than was the case in the past’ (OECD 2009). This is not that surprising in light of the recent research on the absence of a convincing link between minimum wages and employment, as discussed above.

So perhaps the more pertinent level for examining flexibility is the micro level, that of the firm and the worker. Certainly, the gig economy provides flexibility for the firm, in the ability to vary the number of workers and their pay almost without limits and without conscious action. For the worker, it provides flexibility in starting and finishing times and hours worked that are beyond what most casual jobs would offer, as the direct consent of the employer is not required. This is portrayed as a major plus for gig workers, even by the president of the Ride Share Drivers Association of Australia (RSDAA), who was quoted as referring to the importance to gig workers of ‘the flexibility of working hours’ (Marin-Guzman Reference Marin-Guzman2018). Its significance should not be underestimated: in a UK survey of gig workers, half agreed that ‘People working in the gig economy make a decision to sacrifice job security and workers’ benefits for greater flexibility and independence’, and only 19% disagreed (Chartered Institute of Personnel and Development 2017, 32). We explore the implications of this for workers’ preferred status in the next part of this section.

What workers want

In 2014, the Resolution Foundation (D’Arcy and Gardiner Reference D’Arcy and Gardiner2014) published a survey of self-employed UK workers (of which gig workers are a subset). It asked them whether, taking everything into account, they would prefer to be self-employed or an employee. Excluding the ‘don’t knows’, 83% of self-employed said they would prefer to remain self-employed, and 17% said they would prefer to be employees. Amongst gig economy workers, it seems likely that interest in becoming employees is higher, but probably not constituting a majority. A US survey of Uber drivers by Berger et al. (Reference Berger, Frey, Levin and Danda2019) (two of the authors being Uber employees) claimed that 81% of drivers said they would rather be independent contractors than employees. Uber and Lyft have commissioned other opinion surveys, for example in the US and New Zealand, claiming similar majorities of its drivers prefer to be contractors (Benenson Strategy Group 2020; Moreno Reference Moreno2020; Tibshraeny Reference Tibshraeny2023), but there are questions about the objectivity of surveys in which the app firm itself is a participant. In the UK, an internal survey of minicab drivers using a ‘rideshare’ app, found that 31% of minicab drivers wanted ‘worker’ status, while 59% preferred to be independent contractors. Notwithstanding doubts about the validity of some existing surveys, and the general unevenness of data, I have been unable to identify credible studies in which a majority of gig workers, presently contractors, seek employee status. While the president of the RSDAA pointed out that their drivers would be better off as employees than contractors, she added that ‘the majority of drivers if asked would prefer to remain contractors due to the flexibility of working hours’ (Marin-Guzman Reference Marin-Guzman2018). This does not preclude the possibility that some workers decide against becoming gig workers because they cannot be employees.

That said, while many gig workers do not want to become employees, it is very likely that this attitude varies substantially between sectors. For example, while Uber has made use of opinion surveys to assert that the majority of ‘rideshare’ drivers do not want to become employees, this may not be the case in the care sector. A majority (over two-thirds) of National Disability Insurance Scheme (NDIS) workers surveyed by the NDIS Quality and Safeguards Commission considered that they were employees, not contractors (NDIS Commission 2023, 8, plus Data Supplement – Appendix B – Survey Analysis, spreadsheet ‘Roles’ tab.). There is no evidence that a majority of gig workers in disability care would prefer to be contractors than employees.

There is a recent history of some groups of self-employed workers, especially those in the gig economy, undertaking collective action to protect their incomes or conditions and advance their interests (Burns Reference Burns2014; Griswold Reference Griswold2014; Kilhoffer et al. Reference Kilhoffer, Lenaerts and Beblavý2017; Minter Reference Minter2017; O’Donovan Reference O’Donovan2017; Sansom Reference Sansom2017). Globally, across four platform sectors – ride-hailing, food delivery, courier services and grocery – at least 1,271 worker protests were publicised between January 2017 and July 2020 (Ioulia Bessa et al. Reference Ioulia Bessa, Neumann, Stuart, Trappmann and Umney2022). Widespread collective action strongly suggests these workers seek protection.

Survey evidence confirms this interpretation. The British survey that showed many gig workers wanted flexibility also found that 63% agreed that ‘The Government should regulate the gig economy so that all working in it are entitled to receive a basic level of rights and benefits (for example Living Wage/ holiday pay)’, and only 11% disagreed (Chartered Institute of Personnel and Development 2017, 32). Gig workers’ likely desire to remain contractors does not stop a majority of gig workers from wanting regulatory intervention for protection. While they might not see regulation as leading to employee status, they may see it as leading to their protection from capricious behaviour by large corporations and establishing minimum standards to protect their pay and conditions, discussed next.

The views of the affected workers are never determinative of whether regulation is appropriate: we do not eschew minimum wage laws merely because some workers are willing to work for sub-minimum wages. Such standards protect not only the workers directly involved but also minimise a ‘race to the bottom’ that could adversely affect the conditions for numerous other workers. However, the attitudes of those involved can legitimately shape the form that regulation takes, as they can influence the implementation and politics of regulation.

Vulnerability

As contractors, gig workers lack the protection provided through labour law (and through the internal rules typically created within firms regarding employees). They can be terminated with little or no notice, and without recourse to unfair dismissal laws if they are dismissed harshly, unjustly or unreasonably. They receive no compensation if dismissed.

They have very low power compared to the typically large gig firms that hire them. They therefore have little say over the income they receive or the conditions under which they work. Most are underemployed, that is, they are after more hours of work than they are offered by the gig firms (Berg Reference Berg2016; Chartered Institute of Personnel and Development 2017).

Many gig workers receive incomes that, after expenses are taken into account, are below the relevant award wage they would receive if they were employees (Campbell Reference Campbell2019). Outside Australia, many receive incomes equivalent to amounts below the minimum wage in the relevant country (CIPD 2017). They receive fewer training opportunities than other workers. There are also adverse health and safety implications for many gig workers.

As platform technology has developed, gig workers’ vulnerability has increased, the nexus between fares and driver payments has been broken, and in the US Uber’s average ‘take rate’ (the proportion of the fare that it extracts) has risen from 21% over 2015–2021 to 28% in 2022 (Smith and Horan Reference Smith and Horan2023). As the opportunities for large digital corporations to extract surplus from workers increase, the vulnerability of the latter intensifies and the argument for regulation strengthens.

Control

A key matter in determining whether a person is an employee or contractor has been the form of control exercised by the firm over the worker. Prior to recent High Court of Australia cases, criteria (or ‘indicia’) had been developed by tribunals, the courts and even the Australian Taxation Office (ATO) to determine whether persons were employees or independent contractors. The High Court did not overturn the indicia approach but said it must be applied to the written contract, not the contract as it is performed.

Control featured heavily in those indicia, and control over working hours was one of the most important. Amongst mainstream employers, organisational control of employees’ working time has become less important over recent decades than organisational control of the product employees generate for the employer (Fear Reference Fear2011; Kenyon Reference Kenyon2016), yet control of working time remained one of the indicia used to determine whether someone is an employee or a contractor. A Philadelphia US District Court decision acknowledged that Uber could: terminate a driver’s access to the Uber App; deactivate a driver for cancelling trips, failing its background check policy, falling short of the required 4.7-star driver rating, or soliciting payments outside of the Uber App; make deductions against a driver’s earnings; and limit the number of consecutive hours that a driver may work. Yet against the more traditional indicia, Uber was not an employer and its drivers were independent contractors.Footnote ² After finding that the relevant Uber driver was a contractor, the Court commented that Uber and Lyft ‘present a novel form of business that did not exist at all ten years ago’ and added ‘With time, these businesses may give rise to new conceptions of employment status’.Footnote ³ Virtual platforms provide a new, cheap way of control that may replace the need for the employment relationship. Electronic systems enable control and surveillance, rating systems in many platforms minimise the costs of monitoring quality.

The law in Australia has provided no expanded conceptions of employment status. In Australia, recent High Court decisionsFootnote ⁴ have made clear that virtually all gig workers would now be treated as contractors, provided the contracts drafted by the gig firm indicated (when the indicia approach was applied to them) that the workers were contractors. While reducing uncertainty (of which there was much), this further restricted the potential avenues for redress for vulnerable gig workers.

Internationally, the outcome of trying to define gig workers as employees has been mixed, even in the UK (e.g. Gall Reference Gall2016; Grierson and Davies Reference Grierson and Davies2017; Moylan Reference Moylan2017). This is partly because of different interpretations by courts, tribunals and other bodies, but also because of the strong political resistance from platform firms to attempts to define their workers as employees, leveraging the fact that quite a number of gig workers like to conceive of themselves as independent, self-employed people, as well as customers’ preference for cheap services. Even when a rule is devised to interpret the contracts that gig workers sign as employment contracts, gig firms could (and do) amend their contracts to get around that (Bell Reference Bell2021). In the end, though, platform firms can afford regulation (Fang Reference Fang2022) and they end up adhering, grudgingly, to most standards that are imposed on them – other than defining their workers as employees. Thus Uber, for example, has accepted training requirements in Quebec (despite first threatening to quit the Canadian province) (Hughes Reference Hughes2017), fare regulation in Massachusetts (Schoenberg Reference Schoenberg2016) and driver accreditation requirements in several jurisdictions (intellinews 2018).

An ambitious approach was taken in California under Bill AB5. That law was enacted in 2019 after the California Supreme Court had adopted an ‘ABC’ test, in which a worker was assumed to be an employee unless the firm proved all of three criteria: the worker is free from direction and control; the work is performed away from employer premises; and the worker is customarily engaged in a similar but independently established occupation or business.Footnote ⁵ It was the subject of an expensive and successful oppositional campaign by Uber and other rideshare firms that led to AB5’s defeat through ‘Proposition 22’ (Conger Reference Conger2020; Murphy Reference Murphy2020). The AB5 approach would capture many gig workers (otherwise gig firms would not have campaigned against it), but not those where all three criteria could be successfully argued – for example, most Airtasker workers and some food delivery and rideshare drivers, especially those who multi-app. Gig firms would arrange their contracts to maximise the number of uncovered workers, which would plausibly be quite successful in the above areas and in fields such as aged and disability care.

The potential for protection as contractors

The main alternative approach is not to redefine self-employed workers as employees (regulated through the award system), but instead to regulate payments and selected other aspects of work outside the employment relationship. Any regulation needs to be tailored to the circumstances of the industry where it occurs. In some cases, regulation may be expressed in a minimum hourly rate but in others as a piece rate (as per New York taxi and ‘rideshare’ drivers (International Transport Forum 2019)). Each industry is different, and it is difficult, probably impossible, to express minimum standards for gig workers or other relevant self-employed people, in national, across-the-board terms. But with a national minimum wage (and, in Australia, minimum award wages), there are already useful benchmarks from which specific industry standards could be drawn.

While regulation of contractors may seem difficult, there are important precedents. In 1979, New South Wales legislated to allow the New South Wales Industrial Relations Commission (NSW IRC) to regulate minimum terms of contracts for owner-drivers of trucks and other ‘contract carriers’. Now known as ‘Chapter 6’ of the New South Wales Industrial Relations Act,Footnote ⁶ this statute enables the NSW IRC to issue ‘contract determinations’ that specify minimum standards for the drivers concerned. That state’s Industrial Relations Commission can regulate pay rates, union recognition and dispute settlement, for owner-drivers. These standards are analogous to ‘awards’ in NSW labour regulation, namely minimum terms and conditions set by industry. For example, the ‘General Carriers Contract Determination 2017’ establishes, for various types of owner-drivers of trucks, minimum rates of remuneration, comprising a per-kilometre rate, an hourly rate (both varying by truck size and type), allowances and a minimum earnings guarantee, and formulae for adjustment of these, plus entitlements to annual leave, rest breaks and various minimum standards of work and obligations. That Determination also establishes union representation rights, where sought by workers. The NSW IRC can also approve ‘contract agreements’, analogous to ‘collective agreements’, between owner-drivers and firms.

Regulation has avoided treating the owner-drivers as employees. This was important for obtaining support from the owner-drivers, many of whom identify as entrepreneurs rather than employees. Classifying owner-drivers as employees would likely be resisted by both the owner-drivers and the corporations.

The Chapter 6 legislation has survived changes of government and sustained periods of conservative rule, including from 1988 to 1995 and, again, from 2011 to 2013. It avoids transport operators being forced to contract out work to compete against unsustainable contracting practices and low rates. Key companies support the system if they are able to obtain certainty through minimum standards and minimisation of ‘unfair’ competition.

The Chapter 6 experiment has led to a demonstrable improvement in occupational safety for road transport drivers (and other road users) in NSW (Peetz Reference Peetz2022), along with improvements in pay (through the raising of the floor). Drivers have greater control over working time because of less pressure to deliver goods within a defined period, and in that sense have more autonomy. The maintenance of contractor status has also avoided the losses of autonomy that owner-drivers feared (and about which corporations fear-mongered) could accompany employee status.

There are limits, however. The Chapter 6 framework provides little supply chain accountability, which would otherwise have shaped corporate decisions at the initial contracting stage. There are areas that could be regulated if the workers were classed as employees, but which are not regulated through the Chapter 6 approach, such as training.

The Chapter 6 framework very clearly formed the model for the drafting of legislation in Queensland to provide protection for independent courier drivers – in the form of Chapter 10A of the Queensland Industrial Relations Act. At the time of writing, that Chapter had not been proclaimed to take effect as it would be largely made redundant by the Loopholes Bill.

Political sustainability

More than anything else, employment status is the issue that motivates expensive oppositional campaigns by the affected firms. Those campaigns frequently co-opt the workers themselves, many of whom do not want to be employees. The ability of industry to capture regulators would potentially prevent any policy action from taking place if there was not a legislative obligation for action. Rideshare firms have shown they have the skills for policy capture (Borkholder et al. Reference Borkholder, Montgomery, Chen and Smith2018).

That said, not all attempts at regulating gig workers or the self-employed are doomed to encounter determined political opposition. In New South Wales, Chapter 6 regulation of independent contractors involved in road freight transport (owner-drivers) has been successfully achieved for over four decades, since the late 1970s. As mentioned, successive conservative Coalition governments in NSW have traditionally supported Chapter 6. Chapter 6 has also survived a conservative federal government’s 2006 takeover of industrial relations which would have effectively put an end to the system, had it rescinded the state governments’ jurisdictional power to regulate standards for independent contractors. This framework survived because it was able to appeal to the ‘small business’ ethos of some politicians, as ‘small business’ people are seen as a natural constituency of conservative parties (Knott Reference Knott2022). It also mobilised owner-drivers along with employee transport workers to protect the legislation. Indeed, in 1994, the scope of Chapter 6 was extended to encompass owner-drivers using other vehicles (including bicycles), and goodwill provisions were inserted. A Coalition government amended the NSW legislation, establishing a contract of carriage tribunal to deal with disputes over goodwill. Even in 2016 (when the federal Road Safety Remuneration Tribunal (RSRT) was centre stage) many firms sought to retain the NSW system to avoid low-cost operators taking advantage of an unregulated system, favouring the greater certainty of the Chapter 6 system. The conservative NSW Government chose not to undermine the system due to its significance for small business owner-drivers.

The sustainability of Chapter 6 was a stark contrast to the fate of the RSRT, which was established as a separate tribunal (not part of the Fair Work Commission (FWC)) to perform somewhat similar transport regulation functions at the federal level. The RSRT only operated for too short a time for any proper evaluation of its impact to be made. It was abolished by a conservative government three years after a Labor government had established it. This followed a political campaign by major national trucking and logistics organisations (more hard-line than NSW equivalents) and after the federal Liberal Party opposed it. A narrative of ‘freedom’ was popular amongst owner-drivers and used to great effect by the large corporations to garner support from owner-drivers, especially as there were technical weaknesses in the RSRT’s key determination concerning the treatment of ‘backloads’. While the different institutional structures may have mattered (i.e. establishing a separate tribunal vs operating within an existing industrial tribunal) the different politics surrounding the RSRT and Chapter 6 played a major role in explaining the different longevities of the two systems (Peetz Reference Peetz2022).