Previous Next

The Realities and Futures of Work


Flexibility, the ‘gig economy’ and the employment relationship

Take a moment to imagine the future world of work. It could be a world in which trucks and other vehicles drive themselves. A world in which all your shopping is without human interaction, either done online or by you walking into an unstaffed shop. A world in which queries you have about anything, from legal or financial to technical or customer service matters, are answered not by a person but by a bot that sounds, maybe even looks, perfectly human. A world in which energy is generated on the roofs of houses and shared in communities, ugly transmission towers having long ago been torn down. A world in which robots provide assistance to the aged, the disabled, the young. A world in which giant workplaces like mines (though not coal mines, there will not be many of them) or factories or construction sites are fully automated and remotely controlled from some air-conditioned office in Sydney, or Dallas, or Mumbai. A world in which we will all be freelancers, pursuing the next contract like the Uber driver pursues the next ride. Maybe.

If this libertarian vision of a world of freelancers frees us from the drudgery of employment, it is also a nightmare for those who worry about equity and security in the future. In another corner sit those (often also from conservative political camps) who reassure us that nothing really is changing, and it will all work out in the end. Who to believe?

In this chapter I look at the interaction of matters we discussed in the preceding two chapters; that is, I look at trends in work in light of changing management structures and strategies. I look in particular at management and risk, including how management has responded to risk, and other factors including cost pressures and the pressure for accountability. This chapter considers the forms of flexibility that are sought by employers, their origins and their effects on employees, for example through their impact on security and insecurity. It examines ‘precaritisation’ of work through the push for flexibility. A common feature of precaritisation is the growth of nonstandard employment, including through casualisation and independent contracting, in which workers become self-employed and take on the risks of the self-employed. Yet claims that casualisation and self-employment have not increased are often made to dispute widespread perceptions that insecurity at work is increasing.1 How can these things be reconciled? That is one of the matters this chapter will investigate. As will be seen, the best way to start to reconcile these facts is to understand two simple things about how large corporations work these days: (a) they want to minimise the costs and risks they face, and avoid accountability when things go wrong; and (b) they also need to maintain control over workers to ensure the end product has the features, including quality, that they want—and, to date, the best way to do that has been through the employment relationship.

In this chapter we also consider what aspects of digital change are affecting, and likely to affect, work and workplace relations. We investigate whether we are moving to a world of freelancers or the ‘gig economy’ (sometimes referred to as ‘Uberisation’). We test whether digital platforms are fundamentally changing or destroying the employment relationship. We do this through the lens of ‘not there’ employment—the attempts by core capital to centralise profits and minimise costs, risks and accountability (through such devices as contracting out, franchising and the use of digital platforms), and the surprising resilience of the employment relationship. So we consider the reasons, in the logic of capitalism, for employment’s resilience, and what this means for risk and insecurity.

Management, uncertainty and flexibility

Writing in Europe in 1987, Wolfgang Streeck captured the mood that prevailed at the time, and to varying degrees since, in his article ‘The Management of Uncertainty and the Uncertainties of Management’. Streeck emphasised how management faced the problem of finding ways of dealing with ‘an unprecedented degree of economic uncertainty deriving from a need for continuous rapid adjustment to a market environment that appears to have become permanently more turbulent than in the past’.2 He argued that the key concept for management had become ‘flexibility’, which he defined as ‘a general capacity of enterprises to reorganise in close response to fluctuations in their environment’.3 He identified both external (numerical) and internal (functional and temporal) flexibility (discussed below) and consequently two ways firms could seek to achieve organisational flexibility—through a ‘return to contract’ (what we call external flexibility) or ‘extension of status’ (internal flexibility). However, said Streeck, the ‘management of uncertainty’ remained incomplete due to the profound ‘uncertainties of management’ in this new regime as they faced other parties (especially trade unions) while they pushed for greater decentralisation in industrial relations. In doing this, Streeck explained the rising significance of flexibility for employers—it was a means by which management could minimise risk.

Management and risk

Risk is a consistent theme through management, and indeed the corporation can be seen as a device for limiting risk, in the interests of capital. In particular, the corporate form enables owners of capital to collectively undertake gambles (or risks) that have the potential to generate substantial returns while limiting the amount of liability that the owners of capital face if the activity goes pear-shaped. In the era since the end of the 1970s, several important changes in risk have taken place. As noted by Streeck, higher volatility in product markets has increased uncertainty for management.4 Financialisation, discussed in Chapter 2, has substantially shortened the period in which the ‘return’ from risks is meant to be retrieved (though there are some opposing forces as well, as we shall see in Chapter 9). Consequently, there have been efforts by capital to improve the rates of return associated with risk, by affecting both sides of the risk equation; that is, through transferring risk (uncertainty) from capital to labour, and increasing capital’s share of returns (altering the share of income growth going to labour and profits). The last has been achieved both through changes in employment practices of corporations and through lobbying of the state to introduce regulatory changes that either transfer risk or increase capital’s share of returns.

There are several ways in which risk-shifting has occurred. Researchers Mike Rafferty and Serena Yu identified changes in the world of work (the nature of work performed, the nature of demand for and supply of labour, the structure of jobs, the nature of bargaining, and pay and conditions); changes in working lives beyond the workplace (the way risk is experienced across a lifetime and how it connects with the world of work); and changes in the relationships between workers (the way workers experience and understand their lives, especially their shared experience of risk both at work and elsewhere).5 Here, I centre attention on that key element focused on by Streeck: the management of uncertainty and its link to flexibility, though I start with some other issues surrounding risk.

Risk, randomness and the illusion of understanding

Cognitive psychologist Daniel Kahneman, whose work influenced the development of behavioural economics, had a lot to say about the way we think, a little to say about management and not much to say directly about work. But what he says in several areas (summarised in his book, Thinking, Fast and Slow) has implications for work and the role of management within it.6 Of particular relevance here is the illusion of understanding.

Talking about risk or uncertainty is another way of saying that we cannot be certain that a particular event will happen, only that it has a (commonly imprecise) probability that can be attached to it. All sorts of random events influence this probability. But we love to think that we are in control. So we will often ascribe agency for something that actually happened due to random variation. We grab the illusion of understanding that event. Naturally, we will take the credit if things go right and blame others if things go wrong. So, often the fault is not simply in failing to give enough recognition to others when things go right, or failing to accept blame ourselves when things go wrong; it is in failing to recognise that many of the good or bad things that happen are simply a result of chance events—what some would call good or bad ‘luck’ (perhaps as if ‘luck’ was somehow a characteristic that ascribed to people). Hence managers will often be seen to be good (or bad) depending on whether random events led to outcomes that (perhaps temporarily) looked good or bad. It creates an illusion of control for the CEOs themselves, and certainly for those observing them, such that ‘when a particular executive fails, or is embroiled in a scandal, there is a quick defence that it is the individual who is at fault for not living up to the moral status of the position they hold’.7

It also means that you get to hear and see lots of heroic case studies, where management in seemingly difficult situations overcomes the odds to create a brilliant successful product through its own skill (or, alternatively, turns a good situation bad by making some unlucky errors). Kahneman gives the example of the story of Google, oft trumpeted as a major success of great owners and managers, when in fact a whole series of chance events happened to go their way.8 These heroic tales overflow the bookshelves in airports. Some even get into university courses. We love to read case studies because they tell a coherent, plausible story from which we too can learn how to overcome the odds and succeed.

The trouble is, we do not get to see the studies of how other managers faced similar choices, made similar decisions and yet ended face-down, or for whom little changed. ‘Stories of success and failure consistently exaggerate the impact of leadership style and management practices on firm outcomes.’9 So firms with revered CEOs, who are granted prestigious press awards, subsequently experience falls in share price and operating performance.10 This is because previous results (that led to their veneration) were simply unrepresentatively good and were followed by a ‘reversion to the mean’. Often ‘overcoming the odds’ is, as the term ‘odds’ implies, simply a matter of probabilities: someone had to get lucky, and it’s the ones who got lucky that we want to hear and learn from and idolise. If you read a case study of a successful firm or person, ask yourself: ‘How different might things have been if one or two random events had gone another way?’.

Sometimes it is just these random events that will determine which particular cases get highlighted in public discourse; but sometimes they are chosen because somebody wants to push a particular policy barrow. While it is often said that there are lies, damned lies and statistics,11 case studies are especially handy for those who wish to do this, as information can be selectively included or excluded when it is presented to an audience, and verifiability can be difficult. So, for example, an Australian employer lobby group in a widely publicised 1989 report made much of how the case studies it reported demonstrated both a 25 per cent productivity gap between Australia and the USA resulting from industrial relations regulations and the need for major industrial relations reform.12 Details of the cases were sparse13 and subsequent developments invalidated the claims. Two decades after most of the demanded reforms had been implemented, the 25 per cent ‘productivity gap’ remained.14 Yet the commissioning organisation experienced the ‘illusion of understanding’ to such an extent that they had come to believe their own rhetoric. The same organisation also commissioned independent academic researchers to undertake case studies of ‘excellence’ in 30 organisations, only to surprisingly find that their premise (that individual contracting was superior to collective bargaining) was wrong, as the academics found that ‘both union and non-union workplaces were excellent’.15

Lobbyists have learnt not to leave the commissioning of reports, on issues on which they want to push a policy proposal, to chance. They now typically commission consultancy firms (who will do exactly what is asked of them) rather than independent researchers for that purpose, or ‘think-tanks’ established to promote the view for which the group, from which the commissioning organisation comes, wants support.16

This is not to say that case studies have no value. They are used a lot in studies of work. But readers should always be wary of agendas, of overgeneralising, and of attributing agency in place of either luck or selection bias in the cases or the data.

Related to the illusion of understanding are several other aspects of human perception. One is the illusion of hindsight—we think we were better at predicting things than we actually were, plus we ‘believe that we understand the past, which implies that the future should be knowable, but in fact we understand the past less than we believe we do’.17 The other is a tendency to ‘overconfidence’ in our perceptual and decision-making abilities (hence the willingness of people, as mentioned, to believe their own rhetoric).

Put these things together and management can be overly optimistic about how it is perceived by employees—in some cases wildly so, especially where countervailing employee voice mechanisms are limited. An example of this occurred after the Employment Contracts Act 1991 (ECA) was introduced in New Zealand. A survey of employees showed that 12 per cent of employees reported an increase in their trust of management, and 30 per cent reported a decline, with no differences between employees in firms with and without new individual contracts. But a contemporaneous survey of employers found that, in firms without new contracts, 23 per cent thought employee trust of management had increased (only 4 per cent conceded a fall), while in firms with new (mostly individual) contracts 42 per cent claimed an increase in employee trust of management, and only 6 per cent a fall.18 Overconfidence, or management self-delusion, therefore was both common and strongest in firms where managers thought that individual contracts had brought employees closer to management (when the reality was often the reverse). Formal voice mechanisms were weakened (the ECA debilitated unions) and many managers believed, somewhat self-servingly, that individual contracts were a better form of voice.

Just as managers have an illusion of control, so they also seek to maintain or increase control. It does not feel good to be not in control, and it is not good to be seen as not in control. So it takes a lot of belief by managers and supervisors in the benefits of employee participation in decision-making, or in any social justice considerations, for them to follow a high-trust or a collectivist strategy. Alternatively, it takes strong employee organisation to force management to do those things. No doubt, in the future, there will be some employers following such paths, but if power shifts further from employees then managers will be less likely to pursue paths that employees seek.

Managers as employees

Finally, it is worth pausing to remember that managers are also employees, and sometimes they are subjected to the same contradictions and pressures as lower-level employees.19 A study of managers in call centres found their ability to manage was undermined by role conflicts, micromanagement and lack of development or strategic direction, and their coping strategies were a mixture of embracing and resisting.20 So they experience, in sometimes quite specific ways, a number of the issues facing employees that are discussed elsewhere in this book; for example, issues regarding culture, gender, emotional labour, work intensity and ethics. At times in later chapters we use examples from managers to highlight certain aspects of those issues. How much, and in what ways, they experience those issues is often a function of how close to, or distant from, the top of the organisation they are—and therefore how much power they possess. Management is stressful work, especially for those who, as individuals, are troubled by the conflict between financial considerations and human considerations.

Flexibility explored

One person’s flexibility is another person’s uncertainty—just as one person’s stability is another person’s rigidity. The underlying questions, according to Richard Hyman, that we ultimately need to consider with flexibility are: who gains and who loses; and whose interests are promoted and whose interests are neglected.21

For Hyman, flexibility is ‘more a rhetorical slogan than … an analytical instrument’.22 The language of flexibility is value laden: flexibility is invariably portrayed as a ‘good thing’. If someone is deemed not flexible, then they are portrayed as rigid—a ‘bad thing’.23 Economists and others frequently tell us that labour market flexibility is a good thing.

But we could, of course, give the term ‘flexibility’ the opposite connotation. For instance, we could say flexible = uncertain = a bad thing; and that rigid = stable = a good thing. In this interpretation, flexibility would be characterised as a social ‘bad’ as opposed to a social ‘good’. As you can see, we need to separate out the prescriptive (‘we should be flexible’) component of the term flexibility from the analytical component (e.g. numerical vs functional).

This rise of the idea that ‘flexibility’ by employees was a desirable thing coincided with the opening up of market forces and the movement of neoclassical economic ideas (‘market liberalism’) to ascendancy in policy debates and decisions. The ‘Great Recession’ (as Americans described it) or ‘global financial crisis’ (as Australians called it) showed that many of the key propositions behind the neo-liberal model can no longer be assumed to be true.24 For example, the highly ‘flexible’ US labour market generated worse increases in unemployment during the financial crisis than were generated by the less flexible European labour market, even though Europe suffered a larger fall in GDP.25 Since then, austerity programs implemented in Europe, but not the USA, have prevented Europe from making the recovery in unemployment that was subsequently seen in the USA. In economic terms, demand, not flexibility, has driven the course of unemployment.

First, though, let us differentiate between flexibility for workers and flexibility by workers. Flexibility for workers occurs when companies change work practices or working time to better suit worker needs. Allowing workers to take time off to attend school concerts, enabling job sharing, permanent part-time work, ‘48/52’ arrangements—these are all examples of employers being flexible for workers. Research shows these things are mostly effective in enabling better work–life balance, and are often aimed at increasing job satisfaction, attraction or retention of valuable employees.26

But when employers or politicians complain about a lack of flexibility, they usually mean flexibility by workers. Here researchers distinguish between two types of flexibility by workers. They are functional and numerical flexibility.

Forms of labour flexibility by employees

Functional flexibility is the employer’s ability to deploy employees between activities and tasks to match changing workloads, production methods or technology. It is often loosely associated with multiskilling and the expansion of worker autonomy. Numerical flexibility is the employer’s capacity to adjust labour inputs to fluctuations in output. Numerical flexibility in turn can be broken down into three forms. External numerical flexibility refers to nonstandard forms of employment—the ability to adjust the level of labour inputs, for example through use of casual labour, part-time workers, subcontracting, labour hire or fixed contract workers. Internal numerical flexibility (or temporal flexibility) refers to the ability to adjust the quantity and timing of work without engaging additional labour. ‘Financial’ or wage flexibility refers to the ability to adjust wages up or downwards in line with output or market circumstances, through mechanisms such as piece rates, gain-sharing, productivity bonuses, incentive pay schemes or pay cuts.27

Different types of flexibility are often associated with different management styles. For example, numerical flexibility tends to be associated with low-trust styles. It may be linked to individualism (in that contracting and casualisation make it harder for unions to organise, and sometimes this is a purpose of casualisation or contracting out). On the other hand, some aspects of numerical flexibility (in particular, profit-sharing or productivity-related pay) may go hand in hand with employee participation mechanisms that in turn require high trust (though it may not always be provided). Indeed, it appears that contingent pay schemes only really work effectively when they are associated with employee ability to influence work, which, when profit-sharing or productivity-related pay are involved, means employee participation in decision making.28 However, profit-pay and participation do not necessarily coexist. In short, it’s complicated.

By contrast, functional flexibility tends to be disproportionately linked to higher trust approaches (you want employees to stay in the firm as they are trained in multiple skills and tasks, and this requires the building of some trust). It may be linked to collectivism in that internal labour markets (ILMs), often tied to functional flexibility, are also more favourable to unionism, and may force management to recognise collective organisations. And it might be linked to structured management, in that rules and procedures also tend to be linked to ILMs and their greater use of training. But the case should not be overstated: smaller firms may seek functional flexibility because they do not have enough staff for detailed specialisation, but they may also be too small to offer ILMs with strong career prospects. So these can be seen as theoretical tendencies rather than corollaries.

Functional flexibility

Functional flexibility denotes the ability of employees to move flexibly between different job tasks and assignments. While it appears that many employees are demonstrating functional flexibility in other work, Karen Legge’s case-study evidence suggested that, in firms that sought to focus on functional flexibility, the degree of flexibility achieved was ‘more modest than celebratory accounts of multiskilled teamworking or craft-based flexible specialisation might imply’.29 Changes reportedly rarely involved radical skill enhancement. One study examined attempts at introducing full functional flexibility—basically everybody can do everything—at a Pirelli greenfield site. These attempts were eventually abandoned because of the importance of specialist knowledge that was lost with full functional flexibility; the importance of commitment that was associated with ‘ownership’ of a particular area; the cost and availability of training; and the problems of skill retention being weak if new skills were not constantly practised.30

A British study of collective agreements to achieve functional flexibility also detected problems. While most agreements contained promises of job security, they ‘almost always implied a reduction in the workforce, if not through redundancies then by so-called natural wastage’ resulting from anticipated rises in productivity.31 According to Legge, the objective was ‘increased managerial control over the deployment of labour and consequent labour intensification, with [quality of working life] outcomes secondary if not incidental’.32

Numerical flexibility

External numerical flexibility is concerned with the ease of increasing or reducing the number of employees. Broadly speaking, there are two ways in which this can be achieved: by making it easy to hire or fire ‘core’ or ‘permanent workers’; or by making it easy to have people working on ‘peripheral’ contracts—as casual employees, labour hire workers or dependent contractors.

One of the changes in employment relations from the mid-1980s was the restructuring of employment associated with the growth of nonstandard or atypical forms of employment, such as contract, part-time, casual, agency, homeworking, trainees and self-employment. Although it is common to include part-time in this list of nonstandard employment, this is not analytically very helpful, as part-time work can be permanent and is sometimes (not always) driven by the supply side (workers wanting part-time employment). The main exception—that is, the main instance where part-time noncasual work is precarious—is where underemployment is experienced.33 So we focus most on contract, casual, agency, homeworking, trainees and dependent self-employment (i.e. where the worker has the appearance of being self-employed but is actually dependent upon one organisation to provide him or her with work). Casual employment is a term used in Australia to describe employment contracts in which the contract lasts only for the period of the current shift, though in practice employees are usually expected to stay for weeks, months or sometimes even years. In most other countries, temporary employment is a term used to describe employees whose period of engagement is not ongoing. Time series data are not easy to obtain, and most Australian studies rely on data about people with no access to paid holiday or sick leave (as most casuals lack such access). Most Australian casuals have no access to protection against unfair dismissal, redundancy pay or long-service leave. The closest analogy in the UK is ‘zero hours contracts’ (you have a ‘job’ but no guarantee that you will work in any particular week) while the OECD usually uses ‘temporary employment’ as its indicator of precarity in employment and includes Australian casual employment in that measure. There has not been much change in the past decade in the overall proportion of Australian employees who have access to holiday and sick leave.

Labour hire (where the worker is not a direct employee of the company for whom s/he is working) is another form of casual employment, and the growth of labour hire is part of the reason for the relative growth of full-time casual employment.34 Employees paid by labour hire agencies are in uncertain ‘triangular’ relationships with the labour hire firm and the firm for whom they are doing work. Labour hire workers appear to have inferior occupational health and safety outcomes35 and to lack protection from unfair dismissal and redundancy.

In dependent contractor relationships, the worker is not even an employee, even though they are mostly only working for one organisation (that in other circumstances would be called an ‘employer’). The term ‘independent contractors’ is sometimes used in popular discourse to refer to both the above group and those who are genuinely independent and have many clients or customers.

Whereas external numerical flexibility allows the employer to change the number of workers they have at a point in time, internal numerical flexibility (or temporal flexibility) allows them to change when and for how many hours those employees work. Casual employment does both. Part-time employment contracts may (or may not) also allow the employer to vary hours per employee.

Another aspect of internal numerical flexibility is the capacity of employers to get more hours out of workers when needed, through paid or unpaid overtime. Over much of the last century, premiums or ‘penalty rates’ for working nights or weekends, and in particular premiums for overtime, reduced the incentive on employers to make existing employees work for extra hours (rather than take on additional workers). In the last quarter century, there has been a breakdown of the standard working-time model. A diminishing proportion of workers are employed for a standard 38 to 40 hours per week. A large proportion of the workforce are working either shorter hours (in part-time jobs) or longer hours.36

So, while part of internal flexibility concerns the number of hours people work, another part, one that employer organisations are putting more emphasis on, is about getting people to work when it best suits the employer—including on nights or weekends. Hence wage premiums for unsocial hours—that is, for working nights or weekends (referred to in Australia as ‘penalty rates’ and in New Zealand as ‘penal rates’)—are being challenged. These were established a long time ago, both as an effort to get people to work unsocial hours (i.e. to increase labour supply at those times) and to compensate people for the inconvenience of working those hours (part of what neoclassical economists would call a ‘compensating wage differential’). Sundays have (in recent centuries) been associated with religious observance in some countries, and so the decline of religion has been used by some employers to argue that Sunday wage premiums are outdated, though the concept of a rest day has a much stronger lineage.37

Although it is argued by some that the ‘24/7’ economy renders shift premiums obsolete, a counter is that it is still essential that people have time away from work on nights and weekends. In developed countries this is even more important now that two adults in a household are typically working, as otherwise the opportunities for a couple to spend time together would be very limited. Evidence indicates that people working at nights or weekends are more likely than others to be dissatisfied with the balance between their personal and family lives, and to report a deterioration in how well they get on with people at home. The partners of those working weekends are more likely than the partners of people who do not work weekends to want their spouses to get a different job and to complain about the way their social life is affected by their spouse’s work responsibilities.38

To the extent that flexibility is associated with uncertainty, it is numerical flexibility that is at issue. The growth of numerical flexibility since the mid-1980s has been seen as signifying the growth of ‘precaritisation’; that is, the creation of precarious jobs. People in casual jobs, including labour hire, fixed-term employees and dependent contractors, can be seen as in particularly precarious jobs. So too, however, may be a number of people in ‘permanent’ jobs, as they can still be made redundant or otherwise dismissed—especially in the context of the emergence of ‘not there’ employment, discussed below. Guy Standing coined a term, ‘the precariat’, to encompass what he considered to be a new and dangerous social class, evident globally, experiencing ongoing precarity in work and open to political extremism. It evokes, but does not mean the same as, Marx’s ‘lumpenproletariat’, described as a dispossessed group at the bottom of society, including ‘vagabonds … knife grinders, tinkers [and] beggars’ who lacked class consciousness.39 Precarity, however, affects more than just a particular class of people or the ‘low skilled’—the precarity of casual (‘sessional’ or ‘adjunct’) university teachers, often with PhDs, is illustrative of this.40

That growth in casualisation since the 1980s should not be seen as the emergence of a new phenomenon. It is better understood as the reemergence of an old phenomenon, reflecting a reversal of the gains workers had achieved in much of the preceding part of the twentieth century. In Australia, for example, casual employment was a dominant feature of working-class employment in the 1870s and 1880s. This reflected the high seasonality of rural and downstream employment, the unpredictability of demand in manufacturing, and employers’ ability to pass the costs of these things to employees (while maintaining an ILM for more trusted or important employees and relatives).41

Is risk about job insecurity?

We need to recognise that transferring risk need not be the same thing as making jobs insecure. Campbell and Burgess, in a significant review article, refer to precariousness as manifested not only in employment insecurity (including underemployment) but also in earnings insecurity and working-time insecurity (i.e. unpredictability of shifts).42 Risk can simply be about making the livelihoods of workers insecure. If, in relative terms, wages are lowered, and workers are juggling which bill they can afford to not pay in which week, then people’s livelihoods are precarious in a way that is not captured by data on the probability of unemployment or the rate of casualisation. Wages growth has slowed in many countries including Australia,43 and ‘in a majority of countries across the world wage growth in recent decades has lagged behind the growth of labour productivity’.44 Over the past three decades, median real wages in the USA have barely improved,45 and there an unexpected ill-health event can send even an insured family into penury.46

In economics, risk is typically correlated with returns (high risk implies high returns). In recent decades, large firms have shifted the risk component of that equation onto labour, but protected the returns for themselves (and so the labour share of national income has generally declined).47 So risk transference has been an asymmetrical process that, in order to maintain returns to capital, has imposed greater precarity of livelihoods onto parts of labour than would otherwise occur, at a time when average incomes in society have been increasing.

The flexible firm

As firm strategy changed again in the 1980s, one model that emerged to explain the new behaviours of corporations was the ‘flexible firm’ model. This model was principally developed by John Atkinson in the UK.48 In the flexible firm model, employers seek an optimal balance between functional and numerical flexibility through segmenting the labour force into two groups: the core and periphery groups. The core of full-time employees enjoys job security, career development, high wages and are multiskilled. They occupy the primary ILM within the firm. The core group is the principal source of functional flexibility. The periphery is the source of numerical flexibility. So, by contrast to core employees, the periphery has job insecurity, poor career options, lower wages and low multiskilling.

The occupation of the employees who might comprise these groups is to some extent influenced by the adoption and uses of technology. For example, blue collar workers might be part of the stable core if the tasks they do require the use of the employee’s discretion and judgment (e.g. in small batch production). From earlier, we could expect the core to be associated with high-trust management styles, perhaps structured management, perhaps collectivist approaches. By contrast the periphery would be associated with low-trust styles, more likely individualism, and possibly less structure. That in turn implies management may have different management styles for dealing with different parts of the workforce—something that, as we have seen, indeed happens.

There have been a number of criticisms of the flexible firm model.49 Some relate to the ways in which employees are categorised. Precarity is not restricted to the group Atkinson describes as the periphery. Legge claims that the flexible firm model suffers from confusion as to whether it is a prescriptive model—considering ‘what should be’—or a descriptive model, considering ‘what is’.50 At times, it appears to do both. Yet, if it is prescriptive it shows little evidence of being a conscious strategy. Company survey data in the UK suggested that moves to greater numerical flexibility were mainly piecemeal, opportunistic and ad hoc rather than a result of strategic intent.51 In Australia, casualisation appeared higher in workplaces with ‘unstructured’ rather than ‘structured’ management.52 Internationally, however, ‘one place where the core–periphery model undoubtedly has been strategically enacted is in the public services’.53 If, on the other hand, it is meant as a descriptive model, Legge asks how does it improve on preexisting models of labour market segmentation? In essence, segmentation theorists argue that ‘labour markets are characterised by noncompeting groups, and economic, social and political forces combine to differentiate the rewards of comparable groups of workers’.54

My concerns are several. The model treats part-time workers as part of the periphery, which is only true for some (mostly those underemployed or on temporary or casual contracts). It fails to explain why, in the face of all this flexibility, the employment relationship persists. It gives inadequate consideration to employer strategy towards employees and what determines that strategy—in particular, the importance for firms of avoiding accountability—and so has limited predictive value. In that sense, it privileges ‘flexibility’ above all other elements—hardly surprising as that is the focus of so much rhetoric from employers and policy-makers. But why doesn’t everyone just use independent contractors? To be fair, though, it is a theoretical model that emerged in the 1980s, so it is hardly surprising that modifications should be required.

‘Not there’ employment

Even if moves to the flexible firm model were initially piecemeal, ad hoc and opportunistic, its appearance in the literature signalled pressure for some fundamental changes to the way firms organise. More than just reduce risk and cost, however, firms have also sought to achieve ‘distancing’ through what can be called ‘not there’ employment—a method of corporate organisation whereby firms can avoid accountability for misbehaviour by affiliates or subsidiaries by denying responsibility: ‘We’re not there!’. The term ‘not there’ employment is deliberately ironic: of course workers are experiencing the capitalist employment relationship, so in that sense it is very much ‘there’; but it is also a device by which large corporations can say we are not there. So it is not necessarily employment that is ‘not there’, it is the lead corporation. At least, that is how it is meant to be in the eye of the intended beholder.

As a model of firm behaviour, it has some similarities to the ‘flexible firm’ model, but focuses less on forms of ‘flexibility’ and more on capital, accountability and transaction costs. ‘Not there’ employment facilitates a higher rate of noncompliance with labour laws. It is thus more akin to the process of ‘fissuring’ that has often been used to describe modern workplaces.55 The terms ‘fractured’ and ‘fragmented’ have also been used.56 And yet, the employment relationship persists. Many corporations are seeking to make greater use of ‘flexible’ labour, and engage in ‘not there’ employment to minimise costs and risk and to avoid responsibility for some of the labour costs they would otherwise incur and possible breaches of the law. They typically contract to others who in turn hire employees (or who subcontract to others, who in turn hire employees). So part of their workforce is ‘flexibly’ deployed. But they also retain an ILM for their ‘core’ employees. This bit, then, is what Atkinson referred to as the ‘flexible firm’ model in the 1980s. The details have changed a bit since then, but the central idea—many firms simultaneously operating an ILM-based core and a periphery participating in an external labour market—remains true.

So ‘not there’ employment is the process by which centres of capital (we might call these ‘lead firms’ or ‘core capital’) fragment what would otherwise be corporate structures in ways that maintain high control, minimise labour costs and risk, maximise centralised profits and minimise accountability for externalities. Thus, ‘not there’ employment is the key feature of an emerging ‘not there’ capitalism, in which the concentration of capital within product markets grows.57

More precisely, the key methods of ‘not there’ employment are: control is retained by a central entity (e.g. the ‘lead firms’ in supply chains); production is undertaken within smaller entities (the ‘dependent firms’—we might call them ‘peripheral capital’), which are formally separated from the lead firms; core and dependent firms are linked by contract; and labour is ostensibly and directly controlled by the peripheral capital in dependent firms. That labour may be classed as ‘employees’ or as ‘contractors’, according to the context. Fundamentally, though, the employment relationship—hence ‘not there’ employment—persists, despite all the fissuring pressures.

The manifestations of this phenomenon vary between sectors and industries. In the public sector, the urge to minimise accountability is manifested as privatisation, or contractual relationships (such as ‘public–private partnerships’) made opaque through contracts that are ‘commercial in confidence’. In the private sector, variations between industries may occur according to the nature of the product in that industry, the degree of competition, the share of costs represented by labour costs, the geographic scale of production, the nature of labour and labour organisation including exposure to secondary labour markets, occupational norms and the ideology and strategy of employers, among other factors. If there is one major difference with the casualised employment of the late nineteenth century that we saw, it is this: whereas at that time much peripheral labour was directly hired by the responsible employer,58 under ‘not there’ employment a growing part of the responsibility for hiring peripheral labour rests with dependent firms or peripheral capital.

What determines the activities that are done ‘in-house’ and those that are ‘not there’—undertaken in franchises, contractor firms or some other form of peripheral capital? Transaction cost economics, a form of institutional economics associated with writers such as Coase and Williamson, suggests that the transaction costs (including supervision costs) of having work done determine whether it is done in-house or by ‘the market’.59 This, correctly, puts the emphasis on control. As Gramsci said, ‘discipline and a good state of production is only possible if there exists at least a minimum of constitutionality, a minimum of consent on the part of the workforce’.60 An aspect of this is being able to deliver, cheaply, products or services that might otherwise be considered too expensive: getting a product to a customer in time, or having a service available at an abnormal time.61 In addition, the desire to minimise risk and to avoid accountability (to minimise the appearance of control) are also factors in deciding whether to internalise or externalise particular aspects of production. In effect, each corporation makes an assessment of whether the combination of these factors—cost, control, risk and accountability—makes internal or external production more attractive. Some of these things lend themselves to calculation more than others. Cost is the simplest thing to express as dollars while, at the other extreme, the benefits or disadvantages of minimising accountability may be very difficult to express as dollars, and so judgements about that (expressed as ‘sticking to our core business’) may be taken at a higher level in the organisation.

The employment relationship and the future

The rise of the ‘gig economy’, discussed in more detail below, has raised a debate about whether the employment relationship has a future at all. Yet despite all the predictions of the ‘death of employment’ (including a 2008 book by that title),62 it has not happened and most likely will not happen.

The meme that the employment relationship is dying may be seen in a few ideas, not just the emergence of ‘not there’ employment. The first is the idea that, instead of the employment relationship, we are witnessing the rise of ‘freelancing’ or the ‘gig economy’ as workers become self-employed and contractors instead of employees. Second is the idea that workers will not want to stay with organisations any more—they want not one career, but lots of careers, even ‘boundary-less careers’ or ‘portfolio careers’. The third one is the idea that firm size is decreasing, and large firms are no longer as important as they used to be.

These ideas are clearly interrelated. They are also mostly either exaggerated or plain wrong. I address them in reverse order.

First, evidence shows that firm size is not decreasing.63 Nor is workplace size declining structurally. Such measures vary with the state of the business cycle (a more prosperous economy means bigger businesses, on average), but there is not a clear long-term trend.

Second, while the employer demand for increased ‘flexibility’ is genuinely ‘a thing’, self-employment, by contrast, is not growing. Table 6.1, using OECD data, shows self-employment declining between 2000 and 2014 in 26 countries for which data were available, and increasing in only 11. Across those 37 countries, the unweighted average rate of self-employment fell by 2.2 percentage points over the period. Across the OECD (a smaller grouping), the weighted average rate of self-employment fell by 1.9 percentage points between 2000 and 2011. The death of employment is an international myth without international substance.

The mythology of the ‘portfolio’ career, as if workers like to be shunted from industry to industry over their lifetime, hides the likelihood that workers are more insecure than before. Some, of course, may prefer changes of career, but for others career changes are a euphemism that is forced upon them by the impermanency of work.

Table 6.1: Percentage change in self-employment, 37 countries, 2000–14

Self-employment share



Change, 2000–2014



34 .0






















New Zealand












































































































Great Britain




Czech Republic












Unweighted average (all)




OECD average




Notes: * 2013 ** 2011 # 2001

Source: OECD database accessed 23 August 2016. The 3-digit country codes can be found at

Globally, only about half of all employed people are employees. That’s mostly because, in developing countries, a lot of work is in the ‘informal’ sector outside the scope of employment regulation. In sub-Saharan Africa and South Asia, only about one in five employed people are employees earning a wage or salary. This is nothing new. As countries become more industrialised, the informal sector declines and the use of the employment model grows, as capitalism organises production through employees and the need for control of workers grows.

So there is another paradox: the rise of new models like Uber, with its use of contract labour rather than employees, exists side by side with the continuing importance of employment. That paradox is explained by two things: the resistance of employees to greater insecurity; and the ongoing efficiency of employment as a means to control worker behaviour, alongside the ongoing urge of capital to cut costs.

The first of these has already been discussed. The second warrants elaboration. The employment contract is open-ended, and it is impossible to put into it every aspect of what an employee must do. That problem is multiplied many times over when you move from an employment contract to a contract for service.

Contracting is a way of reducing costs and risk, increasing profits and avoiding accountability, but it is not effective for maintaining control, so firms may use a combination that involves contracting out to others (I call them ‘mid contractors’) that in turn often (not always) hire employees.

For most owners of capital, and for most CEOs, the corporation is the most efficient form of economic organisation and it most efficiently operates through employing people. This is fundamentally because the transaction costs in controlling workers through an employment relationship are lower than the transaction costs of alternatives, such as drawing up and implementing contracts with them as independent contractors. So, for example, a firm can more closely supervise employees than contractors, it can manage uncertainty better, it can minimise the risk of opportunistic behaviour, and it need not waste time and money renegotiating contracts with them when they turn up for work each day. In The Circle, Mae is an employee, even though her job (the equivalent of an Instagram ‘influencer’ of the 2010s) is in effect held by the self-employed at present; she is an employee of The Circle because that is how best the company can exercise control over her.

Even when firms contract out part of their operations, the people who end up doing the work are often still employees, because that is the most efficient way for a firm to run its business. For example, when a mining company gets what are known in the industry as ‘contractors’ in to work in a mine, they are usually employees of a contracting firm, that supplies just the workers (a ‘labour hire’ firm) or both workers and equipment. They work on the mine site, but are employees—hired on a casual basis and at much lower cost to the mines.

Through global value chains, an organisation can control a very large number of workers, while having very few as employees within its own organisations. The rest may be employees in contracting forms. In 2013, over 700,000 people were part of Apple’s global value chain, but only 80,000 of them were employed directly by Apple itself.64

Another, growing, example is franchises: a large organisation like 7-Eleven or McDonald’s may contract out business to many franchisees, but they in turn hire employees to work in the stores. Franchising is a way by which the owners of capital (at the top or peak of the hierarchy) contract out to ‘franchisees’ who run the outlets (fast food, retail, etc.) rather than the large firms running the outlets themselves. But most of the people who work in the outlets are employees—of the franchisees, rather than the capital at the top of the hierarchy.

The franchisors still control the product, but franchisees have responsibility for employment. The franchisees are also small businesses, and so less likely to comply with labour regulations than large firms.65 Hence the franchisors gain the financial benefit from franchisees’ low-cost way of operating. This model cuts costs and transfers risk down the chain—which means jobs, or at least personal finances, are more insecure, and the opportunities for underpayment of workers (discussed in Chapter 10) are increased. Franchising has been growing in importance, and employment in franchisees as a share of total employment has likewise grown.66

Many organisations have sought to adopt the Uber platform model. I discuss that shortly. Others firms have tried to turn employees into faux ‘independent contractors’ (such as at Kemalex, in Australia, a few years ago) to avoid paying leave, superannuation and workers’ compensation. Workers and unions have resisted this ‘sham contracting’.67 Most employees do not actually want to be contractors. Like employers, they prefer the stability of the employment relationship.

Meanwhile, a lot of the people who end up as the ‘mid contractors’ are probably people who would otherwise have been small-business owners or self-employed anyway. Circumstances differ between industries. For example, people who love trucks might buy one. They will then become owner-drivers—and sit at the bottom of the road transport industry supply chain. In the textiles, clothing and footwear industries, workers were forced to be contractors (‘homeworkers’) and as recently as a decade ago worked in Australia for pay averaging as little as $6 per hour with some receiving half that or less.68

That has been the pattern in those industries for a long time. Most new manifestations of ‘not there’ employment take one of two forms. In most industries new forms of ‘not there’ employment continue to ultimately rely on employees. In some, though—mostly associated with the ‘platform economy’—core capital contracts to people, many of whom would have been self-employed anyway or, like many Uber drivers, have a ‘main’ job as an employee.69 The sources of worker insecurity vary between industries, and a lot of variation between industries and periods can be hidden by national aggregates. Franchising has grown in retailing; labour hire in mining; outsourcing in the public sector; second jobs in manufacturing; spinoffs in communications; casualisation in education and training; and global supply chains send jobs internationally to low-paid, often dangerous workplaces in a number of industries.

Even ‘casual’ or ‘permanent’ employment status (as proxied by access to paid leave) can be an inadequate indicator of worker insecurity. Employees at the bottom of the chain in many of these industries might have access to annual leave and sick leave. Offering leave is better for attracting labour and might be cheaper than paying the casual loading. There is also no need to hire someone on a casual contract if you can make them redundant when the work dries up—if, for example, you lose your contract with the main firm. If your firm can go bankrupt, then you often will not even have to pay redundancy benefits.

Other sources of worker insecurity might arise from weaknesses in the operation of labour-market institutions. In Australia, for example, firms have discovered they can terminate collective agreements, or get a handful of workers to sign new agreements that then apply to everyone, as a means of cutting pay and conditions.70 Some firms put employees onto contrived arrangements that make them out to be contractors, often illegally.71 The growing insecurity of workers, including through underemployment, and the concentration of economic power help explain why wage growth has been stagnating. Indeed, with wage growth low across most developed countries, that successful ‘war on wages’ may be the biggest single sign of worker insecurity.72

The ‘platform’ or ‘gig’ economy

Technology is often seen as fundamentally changing workplace relations through the ‘platform economy’, the ‘gig economy’ or ‘freelancing’. These terms are often used interchangeably, as if to indicate that the growth of the gig economy is synonymous with ‘freelancing’. The term ‘freelancing’ is used very loosely and often in multiple ways, and so some large estimates of the freelancing workforce are made. For example, the Freelancers Union, a lobby group (not a trade union) in whose interest it is to maximise the number of ‘freelancers’, defined them as including employees who at any time in the previous year had held a fixed-term job or a second job as a contractor, and so it claimed that freelancers made up 34 per cent of the US workforce.73 Using a different definition, Kitching estimated freelancers at just 6 per cent of the UK workforce.74 ‘Freelancer’ is such a poorly defined concept that we do not use it in the remainder of this chapter.

To get a clearer understanding of the platform economy, also known as the gig economy, we turn to Valerio De Stefano:

The gig-economy is usually understood to include chiefly two forms of work: ‘crowdwork’ and ‘work on-demand via apps’ … The first term is usually referred to [sic] working activities that imply completing a series of tasks through online platforms … ‘Work on-demand via apps’, instead, is a form of work in which the execution of traditional working activities such as transport, cleaning and running errands, but also forms of clerical work, is channelled through apps managed by firms that also intervene in setting minimum quality standards of service and in the selection and management of the workforce.75

Platforms involve many workers. Uber, for example, has a global workforce of around 160,000; its US competitor Lyft has 50,000. Amazon Mechanical Turk (AMT) mediates crowdwork for up to 500,000 people internationally and Crowdflower 10 times that number.76

Still, this is only a small portion of the overall workforce. In the UK, the estimates of platform economy work are fairly consistent, probably reflecting similarities in data methods. The UK Chartered Institute of Personnel and Development (CIPD), drawing on its own online survey of 5,000 adults in December 2016, estimated 4 per cent of adults had been ‘gig workers’ at some time in the previous 12 months.77 An online survey for the UK Department for Business, Energy & Industrial Strategy (BEIS) estimated that 4.4 per cent of the population in Great Britain had worked in the gig economy in the preceding 12 months.78 Of those, just slightly under one-third provided services this way at least once a week. That is, roughly 1 per cent of British workers regularly participated in the platform economy.

American estimates are more variable. An August 2016 online survey by the Pew Research Center of 4,500 adults in the US found only 8 per cent of adults had used digital work or task platforms in the previous year.79 It appeared to pick up especially more crowdworkers than the UK surveys. A more indirect estimate, by J.P. Morgan based on money transfers, put the number much lower: a cumulative 0.9 per cent of US adults had ever participated in the gig economy according to it.80 US economists Lawrence Katz and Alan Krueger, based on 2015 data from an official labour force survey, estimated only 0.5 per cent of US workers had identified customers through an online intermediary, and noted this was close to two other, independently derived estimates.81

Another study (not aimed at estimating its penetration) was undertaken by the ILO in 2016 and involved surveying 1,800 workers in Crowdflower and AMT.

It is likely that online surveys, by their nature, will overestimate the incidence of ‘platform economy’ workers relying on online apps, potentially by quite a bit, but that alone does not account for the discrepancies between estimates. Question wording also matters. Another important factor in interpreting these estimates is time-related: at any single point in time, only a small number of people will be employed in the gig economy, but a much larger number of people will have worked in it at some time over the previous year to supplement otherwise low incomes, sometimes only once and with minimal effect. There may be international differences, but it seems likely that 1 per cent or less of the workforce is regularly part of the platform economy in developed countries, many others dip their toes into it and then take them out again, the numbers are likely growing, but they are presently exceeded by the numbers of contractors and comparable workers who do not rely on apps for their work.

Despite its limited size, the ‘platform’ economy is a major area where digital technology has allowed a new form of work organisation to develop. There are many stories, ranging from exciting to horrifying, about work in the gig economy.82 Some of the key points about the ‘gig’ economy from the above research and another ILO study83 are as follows.

The use of digital work or task platforms is more common among younger workers. In the Pew survey, 17 per cent of 18–29 year olds, compared to 10 per cent of 30–49 year olds and just 4 per cent of 50–64 year olds, had earned money through digital work or task platforms.84

The use of digital work or task platforms is more common among lower income earners. In the Pew survey, the proportions of adults who had earned money through digital work or task platforms were 10 per cent in households with annual income less than $30,000, compared to 8 per cent with income of $30,000–$75,000 and 4 per cent in households with incomes above $75,000.85 The incidence was also higher among blacks (14 per cent) and Latinos (11 per cent) than whites (5 per cent), and lower among those who had college (university) qualifications (6 per cent) than those with lower qualifications (9 per cent).

‘Gig economy’ work provides essential income for a minority of participants. In the CIPD survey, only 25 per cent of ‘gig’ workers said it was their main job.86 In the Pew survey, 29 per cent of platform workers said that the income they earned from it was essential for meeting basic household needs, while 27 per cent said it was ‘important’ and 42 per cent ‘nice to have, but not essential’.87 In the BEIS survey, only 14 per cent of gig workers earned half or more of their total income from gig work.88

Many gig-economy workers have other jobs. In the CIPD survey, 58 per cent of gig-economy workers with other jobs had permanent jobs (compared to 78 per cent in the population at large). Only a minority (20 per cent) were self-employed.89 Another survey by the ILO confirmed that the otherwise self-employed were only a minority of ‘gig’ workers, and only a minority had been running a business before taking up ‘gig’ economy work.90

A majority of ‘gig’ workers are chronically underemployed. In the CIPD survey, only 26 per cent said that, overall, they got enough work on a regular basis in the gig economy.91 In the ILO survey, the most common reason given, when respondents were asked why they were not doing more crowdwork (or non-crowdwork), was that there isn’t enough work.92

Expressed as an hourly rate, pay in the gig economy is low and often below minimum wages. In the CIPD survey, median earnings for transport or delivery was £6 per hour, and for short-term jobs was £7 per hour, below the then living wage of £7.20 per hour and the minimum wage for 21–24 year olds of £6.95 per hour.93 In the BIES survey, only 36 per cent of those who disclosed earnings received less than £7.50 per hour but 85 per cent of those who were involved in gig work from one to three times a week in the gig economy had earned less than £5,000 from their work in the preceding 12 months.94 In the ILO survey, median pay for AMT workers was US$4.65 per hour, well below the minimum wage.95 A separate analysis of 3.8 million tasks on AMT found that the median pay was closer to US$2 per hour, much less than requestors paid, due in part to time between tasks (e.g. time spent searching for tasks) and tasks ‘returned’ (not completed, perhaps because the requestor’s description was inaccurate).96 In Australia, Unions NSW calculated that rates received via Airtasker for data entry, cleaning and sales were between AU$3 and AU$9 below the relevant minimum award rates.97 Hence the Twitter user named ‘Fair Gig for all’ tweeted on a regular basis about vacancies on Airtasker for jobs paying below the minimum wage.98

Although platform economy workers are doing it tough, they also seem surprisingly optimistic, perhaps reflecting an optimism bias identified by Kahneman,99 related to ‘adaptive preferences’.100 In the CIPD survey, only 49 per cent of ‘gig workers’ (compared to 56 per cent of other workers) said they were ‘living comfortably’ or ‘doing alright’. However, 46 per cent of platform workers (compared to 26 per cent of other workers) expected their economic situation to improve over the coming year. And while only 57 per cent of platform workers (compared to 67 per cent of other workers) were saving for retirement through a pension plan, fully 33 per cent of platform workers (compared to 21 per cent of other workers) were confident they could live comfortably when they stopped working.101

Platform economy workers did not think highly of their training opportunities. Nor did their thoughts readily turn to trade unions when asked to whom they would take complaints.102 For some ‘platform economy’ workers, like Uber drivers or their equivalent in other industries, who have a main job elsewhere, their attachment to work may be low and their willingness to organise resistance to their conditions of employment may be low. But for those who are reliant on gig-economy earnings as their sole source of income, their responses to attempts to organise to improve conditions of employment may be quite different. There are many media reports of such workers taking collective action, sometimes strike action, in support of improvements in pay or conditions, and some instances where they have won treatment equivalent to that of employees.103 The whole idea of them being independent contractors, not employees, has been challenged in several jurisdictions, sometimes successfully,104 and sometimes not (e.g. in Australia).105 Uber says it is not a transport company, it is a technology company, acting as a client to drivers, but it is clearly in competition with transport companies and celebrates providing that competition.106

There have been reports of a new labour-hire app called ‘Squaddle’, described as the ‘Uber of hospitality.’ This app allows ‘contractors’ to bid for work at restaurants and bars to fill last-minute staffing gaps.107 Many firms are trying this approach, and app developers see it as a way of making lots of money. Platform firms both supply and discipline low-paid blue-collar contractors, for use by warehouses or other corporations, much like the day labour of the past.108 They may even pretend to represent the interests of those contractors by compulsorily enrolling them as members in associations in which they have no voting rights.109 It is a model that has already spread to many industries and occupations and that will likely find its way into almost any area of work that lends itself to casual employment or contracting, including some where those modes of employment have not been used before—because it provides for a new mechanism of control that partially takes the place of the employment relationship in exercising control, especially through the ‘rating’ functionality. That said, many US firms following the Uber model—like Cherry (car washes), Prim (laundry), SnapGoods (gear rental), Rewinery (wine) and HomeJoy (home cleaning)—have failed. One author argues, ‘they provide crummy jobs that most people only want to do as a very last resort. These platforms show their workforce no allegiance or loyalty, and they engender none in return’.110

In effect, there is a major labour supply problem that many of these firms have been unable to reconcile. Uber itself has never made a profit. Despite the low wages, and frequent law-breaking,111 its prices are too low, and its financiers are waiting for a return on their investment. It might appear that it aims to drive competitors (taxis) out of business and grow its market share and prices over the long run. Yet, if it does that, other platform-based ride services (such as Lyft) would step in and out-compete it. Perhaps it is hoping to hold on until driverless cars are available, in which case it would no longer have to worry about payments to drivers, but unless it hires those cars from owners it would face huge capital costs that it has so far avoided. Its hopes likely extend beyond driverless cars.

The biggest challenge to the employment relationship is in the ‘platform economy’. This is because virtual platforms provide a new, cheap way of control that may replace the need for the employment relationship. It is not only the availability of electronic control and surveillance that makes this possible. It is also that the rating systems in many platforms minimise the costs of monitoring quality. Not only does the platform reduce the need for an employment relationship, it also reduces the need for intermediate or peripheral capital to organise labour on behalf of core capital.

However, there are still limits to the use of cost cutting and of platform control. The gig economy will grow, but it will not overtake the employment relationship.


Flexibility is not a concept restricted to the gig economy and has been important in work for over three decades. The language of flexibility, we have seen, is value laden. The platform (or ‘gig’) economy is a major area where digital technology has allowed a new form of work organisation to develop. It is an area of heightened flexibility by labour, income insecurity, minimal employee control, and barriers to labour organisation.

The emergence of a new phase of capitalism, featuring ‘not there’ employment, signals some major challenges to labour and to regulators. But it does not signal the demise of the employment relationship and its replacement by a horde of freelancers. This is because cores of capital want to minimise costs and accountability, but they do not wish to minimise control. Their share of product markets increases. The paradox of the rise of new ‘gig economy’ models like Uber, and the continuing importance of employment, is explained by the ongoing efficiency of employment as a means to control worker behaviour, alongside the ongoing urge of capital to cut costs. The employment contract is open-ended, and it is impossible to put into it every aspect of what an employee must do. Contracting is a way of reducing costs and risk, increasing profits and avoiding accountability, but it is not effective for maintaining control. The contradiction discussed here is at the heart of ‘not there’ employment and its resolution plays out differently in different industries, but in the end the employment relationship provides a form of control that no other contractual relationship can achieve. Even when critical activities are outsourced to peripheral capital, the labour used by peripheral capital is often (not always) kept under control through the employment relationship, because this enables peripheral capital to control labour (and to minimise labour turnover costs) in the way it needs, to meet its obligations to core capital. The platform economy reduces the need for peripheral capital as an intermediary between core capital and labour.

1 Mark Wooden, ‘Factcheck: Has the Level of Casual Employment in Australia Stayed Steady for the Past 18 Years?’. The Conversation, 23 March 2016,; Robert Sobyra, ‘Australian Jobs Aren’t Becoming Less Secure’. The Conversation, 17 July 2018,

2 Wolfgang Streeck, ‘The Uncertainties of Management and the Management of Uncertainty’. Work, Employment and Society 1, no. 2 (1987): 281–308.

3 Ibid., 290.

4 Ibid.

5 Mike Rafferty and Serena Yu, Shifting Risk: Work and Working Life in Australia. A report for the Australian Council of Trade Unions (Sydney: Workplace Research Centre, University of Sydney, September 2010).

6 Daniel Kahneman, Thinking, Fast and Slow (New York: Farrar, Straus and Giroux, 2011).

7 Peter Bloom and Carl Rhodes, CEO Society: The Corporate Takeover of Everyday Life (London: Zed, 2018), 49.

8 Kahneman, Thinking, Fast and Slow, Chapter 9.

9 Ibid., Chapter 19, citing Philip Rosenzweig, ‘The Halo Effect and the Challenge of Management Inquiry: A Dialog between Phil Rosenzweig and Paul Olk’. Journal of Management Inquiry 19 (2010): 48–54.

10 Kahneman, Thinking, Fast and Slow, Chapter 24, citing Ulrike Malmendier and Geoffrey Tate, ‘Superstar CEOs’. Quarterly Journal of Economics 24 (2009): 1593–638.

11 This phrase was attributed, apparently incorrectly, to Disraeli by Mark Twain. For more information on the problem referred to, however, see Darrell Huff, How to Lie with Statistics (New York: W.W. Norton & Company, 1954).

12 Business Council of Australia, Enterprise-Based Bargaining Units: A Better Way of Working. Part 1 (Melbourne: BCA, 1989); Business Council of Australia, ‘The Report That Will Change the Way We Work’. Business Council Bulletin 59 (1989): 38–39.

13 Stephen J. Frenkel and David Peetz, ‘Enterprise Bargaining: The BCA’s Report on Industrial Relations Reform’. Journal of Industrial Relations 32, no. 1 (1990): 69–99.

14 David Peetz, ‘Does Industrial Relations Policy Affect Productivity?’. Australian Bulletin of Labour 38, no. 4 (2012): 268–92.

15 Daryl Hull and Vivienne Read, Simply the Best: Workplaces in Australia. Working Paper 88 (Sydney: ACIRRT, University of Sydney, December 2003), 8. See also David Peetz, ‘Hollow Shells: The Alleged Link between Individual Contracting and Productivity’. Journal of Australian Political Economy 56 (2005): 32–55.

16 On consultants, see Ross Gittins, ‘Dark Art of Econometrics’. Sydney Morning Herald, 27 August 2007; Peter Martin, ‘Is Tony Abbott’s “Cop on the Beat” Worth $6 Billion?’. Sydney Morning Herald, 28 August 2013; Michael West, ‘Ignore Page Count, Just Note Number on the Front’. Sydney Morning Herald, 8 March 2014. One organisation concerned later gained notoriety for suppressing results of a survey it had undertaken of its members that undermined its own position on reducing company tax: Laura Tingle and Phillip Coorey, ‘Secret BCA Survey Does Not Back Tax Cuts Going to Jobs, Wage Rises’. Australian Financial Review, 26 March 2018, On think-tanks, see A. Salas-Porras and G. Murray, eds, Think Tanks and Global Politics: Key Spaces in the Structure of Power (Singapore: Palgrave Macmillan, 2017).

17 Kahneman, Thinking, Fast and Slow, 201–2.

18 David Peetz et al., ‘Workplace Bargaining in New Zealand: Radical Change at Work’. In Workplace Bargaining in the International Context, ed. David Peetz, Alison Preston, and Jim Docherty. Workplace Bargaining Research Project (Canberra: Department of Industrial Relations and Australian Government Publishing Service, 1993), 290–93.

19 Keith Townsend and Sue Hutchinson, ‘Line Managers in Industrial Relations: Where Are We Now and Where to Next?’. Journal of Industrial Relations 59, no. 2 (2017 ): 139–52.

20 Maeve Houlihan, ‘Managing to Manage? Stories from the Call Centre Floor’. Journal of European Industrial Training 25, no. 2 (2001): 208–20.

21 Michele Salvati, ‘A Long Cycle in Industrial Relations’. Labour 3, no. 1 (1989): 42–72.

22 Richard Hyman, The Political Economy of Industrial Relations (London: Macmillan, 1989).

23 Salvati, ‘Long Cycle’.

24 John Quiggin, Zombie Economics (New Jersey: Princeton University Press, 2010).

25 David Peetz, Stephane Le Queux, and Ann Frost, ‘The Global Financial Crisis and Employment Relations’. In The Future of Employment Relations: New Paradigms, New Approaches, ed. Adrian Wilkinson and Keith Townsend (Basingstoke: Palgrave Macmillan, 2011), 193–214.

26 Natalie Skinner and Janine Chapman, ‘Work–Life Balance and Family Friendly Policies’. Evidence Base 4 (2013); Jason Hess, ‘Ignore Flexible Working at Your Peril’. Real Business 5 February 2014,

27 Karen Legge, Human Resource Management: Rhetorics and Realities. Management, Work & Organisations Series (Basingstoke, England: Macmillan Business, 1995).

28 John R. Cable and Felix R. Fitzroy, ‘Co-operation and Productivity: Some Evidence from West German Experience’. Journal of Economic Analysis and Workers Management 14, no. 2 (1980): 163–80; Douglas Kruse, Employee Ownership and Employee Attitudes: Two Case Studies (Norwood Editions: Pennsylvania, 1984); Bjørne Grimsrud and Torunn Kvinge, ‘Productivity Puzzles—Should Employee Participation Be an Issue?’. Nordic Journal of Political Economy 32, no. 2 (2006): 139–67; General Accounting Office, Employee Stock Ownership Plans: Little Evidence of Effects on Corporate Performance (Washington DC: GAO, March 1987).

29 Legge, Human Resource Management, 162; Tony Elger, ‘Task Flexibility and Intensification of Labour in UK Manufacturing in the 1980s’. In Farewell to Flexibility, ed. Anna Pollert (Oxford: Basil Blackwell, 1991).

30 J. Clark, ‘Full Flexibility and Self-Supervision in an Automated Factory’. In Human Resource Management and Technical Change, ed. J. Clark (London: Sage, 1993).

31 Income Data Services, Flexible Working. Study 407 (London: IDS, April 1988); Anna Pollert, ‘The Flexible Firm: Fixation or Fact?’. Work, Employment and Society 2, no. 3 (1988): 281–316.

32 Legge, Human Resource Management, 169.

33 Between 2010–11 and 2016–17, the number of hours sought, but not worked, by underemployed people in Australia grew by 31 per cent, five times the total growth in hours worked. See Australian Bureau of Statistics, ‘Labour Account Australia, Quarterly Experimental Estimates’, September 2017. Canberra 6150.0.55.003.

34 If we take the liberty of labelling people without leave as ‘casuals’, then the number of ‘casual’ full-timers grew in Australia by 38 per cent between 2009 and 2017. So-called ‘permanent’ full-timers (that is, those with annual leave) grew by just 10 per cent. On the other hand, some organisations have found their reliance on part-time casuals has been counterproductive, as it meant workers had no commitment and were unreliable. So some large retailers now use ‘permanent’ part-timers rather than casuals. So-called ‘casual’ part-timers grew by just 13 per cent between 2009 and 2016. ‘Permanent’ part-timers grew by 36 per cent. See Australian Bureau of Statistics, ‘Characteristics of Employment’.Canberra 6333.0.

35 E. Underhill, ‘Changing Work and OHS: The Challenge of Labour Hire Employment’. Paper presented at the New Economies: New Industrial Relations conference, Noosa, February 2004, 544–51.

36 For example, Australian Bureau of Statistics, ‘Working Time Arrangements, Australia’, November 2012. Canberra 6342.0.

37 David Peetz, Scott Bruynius, and Georgina Murray, ‘Choice and the Impact of Changes to Sunday Premiums in the Australian Retail and Hospitality Industries’. Journal of Industrial Relations (2019),

38 David Peetz et al., ‘Quality and Quantity in Work–Home Conflict: Nature and Direction of Effects of Work on Employees’ Personal Relationships and Partners’. Australian Bulletin of Labour 37, no. 2 (2011): 138–63.

39 Karl Marx, The Eighteenth Brumaire of Louis Bonaparte (London: The Electric Book Company, 1852; repr., 1998), 78. See also Robert L. Bussard, ‘The “Dangerous Class” of Marx and Engels: The Rise of the Idea of the Lumpenproletariat’. History of European Ideas 8, no. 6 (2012): 675–92.

40 For example, Robyn May, David Peetz, and Glenda Strachan, ‘The Casual Academic Workforce and Labour Market Segmentation in Australia’. Labour and Industry 23, no. 3 (2013): 258–75.

41 Jenny Lee and Charles Fahey, ‘A Boom for Whom? Some Developments in the Australian Labour Market, 1870–1891’. Labour History—A Journal of Labour and Social History 50 (1986): 1–27.

42 Iain Campbell and John Burgess, ‘Patchy Progress? Two Decades of Research on Precariousness and Precarious Work in Australia’. Labour & Industry 28, no. 1 (2018): 48–67.

43 Andrew Stewart, Jim Stanford, and Tess Hardy, eds, The Wages Crisis in Australia: What It Is and What to Do About It (Adelaide: University of Adelaide Press, 2018); Joe Isaac, ‘Why Are Australian Wages Lagging and What Can Be Done About It?’. Australian Economic Review 51, no. 2 (2018): 175–90.

44 International Labour Office, Global Wage Report 2016/17 (Geneva: ILO, 2016), xvi.

45 Drew Desilver, ‘For Most U.S. Workers, Real Wages Have Barely Budged in Decades’. Fact Tank, Pew Research Center, 7 August 2018,

46 Kimberley Amadeo, ‘Medical Bankruptcy and the Economy’. The Balance (financial advisers), 3 February 2019,

47 Jim Stanford, ‘The Declining Labour Share in Australia: Definition, Measurement, and International Comparisons’. Journal of Australian Political Economy 81 (2018): 11–32; Loukas Karabarbounis and Brent Neiman, ‘The Global Decline of the Labor Share’. Quarterly Journal of Economics 129, no. 1 (2013): 61–103; David Autor et al., ‘Concentrating on the Fall of the Labor Share’. American Economic Review 107, no. 5 (2017): 180–85.

48 John Atkinson, ‘Manpower Strategies for Flexible Organisations’. Personnel Management (1984): 28–31; John Atkinson, Flexibility, Uncertainty and Manpower Management. IMS Report No. 89 (Brighton, UK: Institute for Manpower Studies, 1985). You may find out more about this model by reading Pollert, ‘The Flexible Firm: Fixation or Fact?’. A graphical representation of the model is there at p. 284.

49 See the studies cited in Legge, Human Resource Management.

50 Ibid., 154.

51 Ibid., 160.

52 Ron Callus et al., Industrial Relations at Work: The Australian Workplace Industrial Relations Survey (Canberra: AGPS, 1991).

53 Legge, Human Resource Management, 161.

54 Peter Brosnan, David Rea, and Moira Wilson, ‘Labour Market Segmentation and the State: The New Zealand Experience’. Cambridge Journal of Economics 19, no. 5 (1995): 667–96.

55 David Weil, The Fissured Workplace: Why Work Became So Bad for So Many and What Can Be Done to Improve It (Cambridge: Harvard University Press, 2014).

56 Tony Dundon, ‘The Fracturing of Work and Employment Relations’. Labour and Industry 29, no. 1 (2019): 6–18.

57 Andrew Leigh and Adam Triggs, ‘A Few Big Firms’. The Monthly, 17 May 2017; John Van Reenen, ‘Research: The Rise of Superstar Firms Has Been Better for Investors Than for Employees’. Harvard Business Review, 11 May 2017; Data Team, ‘Corporate Concentration’. Economist, 24 May 2016; Autor et al., ‘Concentrating on the Fall of the Labor Share’.

58 Lee and Fahey, ‘Boom for Whom?’.

59 Ronald Coase, ‘The Nature of the Firm’. Economica 4 (1937): 386–405; Ronald Coase, ‘New Institutional Economics’. American Economic Review 88, no. 2 (1998): 72–74; Oliver E. Williamson, Markets and Hierarchies: Analysis and Antitrust Implications (New York: Free Press, 1975); Oliver E. Williamson, Institutions of Capitalism (New York: Free Press, 1985).

60 Antonio Gramsci, La Construzione Del Partito Comunista 19231926 (Turin: Einaudi, 1971), 6. Translated and cited in Anne Showstack Sassoon, ed., Approaches to Gramsci (London: Writers and Readers Publishing Cooperative Society, 1982), 103.

61 The extent to which consumer demands for timely delivery influence outsourcing in Australian road transport is presently being investigated by Kaylee Boccalatte.

62 Ken Phillips, Independence and the Death of Employment (Ballan, Vic.: Connor Court Publishing, 2008).

63 See, for example, Stefan Hajkowicz et al., Tomorrow’s Digitally Enabled Workforce: Megatrends and Scenarios for Jobs and Employment in Australia over the Coming Twenty Years (Brisbane: CSIRO, 2016), 41, Figures 14 & 15. The figures show declining shares of small business, even though the accompanying text suggests the opposite.

64 Direct employment estimate from Bloomberg, quoted in Ricarda McFalls, The Impact of Procurement Practices in the Electronics Sector on Labour Rights and Temporary and Other Forms of Employment, (Geneva: International Labour Office, Working Paper WP313, 2016). Global value chain data implied from a statement by Tim Cook, Apple CEO, on a 2014 BBC Panorama program quoted in Thomas Clarke and Martijn Boersma, ‘The Governance of Global Value Chains: Unresolved Human Rights, Environmental and Ethical Dilemmas in the Apple Supply Chain’. Journal of Business Ethics 143 (2017): 111–31.

65 Weil, The Fissured Workplace. Haroon Bhorat, Ravi Kanbur and Natasha Mayet, ‘Estimating the Causal Effect of Enforcement on Minimum Wage Compliance: The Case of South Africa’. Review of Development Economics 16, no. 4 (2012): 608–23; Tim H. Gindling and Katherine Terrell, ‘Minimum Wages, Globalization, and Poverty in Honduras’. World Development 38, no. 6 (2010): 908–18; Eric Strobl and Frank Walsh, ‘Minimum Wages and Compliance: The Case of Trinidad and Tobago’. Economic Development and Cultural Change 51, no. 2 (2003): 427–50; Jeremy Buultjens, ‘Labour Market Deregulation: Does Small Business Care?’. International Journal of Employment Studies 2, no. 1 (1994): 132–57.

66 Lorelle Frazer, Scott Weaven, and Anthony Grace, Franchising Australia 2014 (Brisbane: Griffith University, 2014).

67 Anthony O’Donnell, ‘Reality Bites for Bosses of Nothing’. Age, 29 September 2005,

68 Mindy Thorpe, ‘Outworkers’. In Pay Equity in Queensland, ed. David Peetz and Rosemary Hunter (Brisbane: Centre for Research on Employment and Work and Socio Legal Research Centre, Report to Department of Employment, Training and Industrial Relations, Queensland Government Submission, Pay Equity Inquiry, Queensland Industrial Relations Commission, No. B1568 of 2000), 99–114; Emer Diviney and Serena Lillywhite, Ethical Threads: Corporate Social Responsibility in the Australian Garment Industry (Melbourne: Brotherhood of St Laurence, 2007), 4.

69 Justin Fox, ‘Your Uber Driver Probably Has Another Job’. Bloomberg, 20 February 2016,

70 Josh Butler, ‘Streets Look to Cut Workers Pay and Ice-Cream Fans Are Furious’. Huffington Post, 29 August 2017,; Emma Field, ‘Hundreds Rally in Gippsland at Longest Industrial Dispute in 40 Years’. ABC News Online, Australian Broadcasting Corporation, 28 June 2018,

71 ‘Factory Workers Strike over Contract Plan’. The Age, 3 May 2005,; Australian Government, ‘Unfair Contracts and Sham Contracts’ (Department of Innovation, Industry and Science, 2018),

72 Mike Seccombe, ‘The Truth About Wage Stagnation’. Saturday Paper, 12 May 2018,; Anna Patty, ‘War on Wages: Australians Are Working Harder and Going Backwards’. Sydney Morning Herald, 5 August 2018,

73 Freelancers Union and Elance-oDesk, Freelancing in America: A National Survey of the New Workforce (2013).

74 John Kitching, Exploring the UK Freelance Workforce in 2015 (London: Association of Independent Professionals and the Self-Employed, 2016); Kitching’s data indicate that the estimate would more than double if a broader occupational definition of freelancers were used.

75 Valerio De Stefano, The Rise of the ‘Just-in-Time Workforce’: On-Demand Work, Crowdwork and Labour Protection in the ‘Gig-Economy’. Conditions of Work and Employment Series No. 71 (Geneva: International Labour Office, 2016), 1.

76 Ibid.

77 Chartered Institute of Personnel and Development, To Gig or Not to Gig? Stories from the Modern Economy (London: CIPD, March 2017).

78 Katriina Lepanjuuri, Robert Wishart, and Peter Cornick, The Characteristics of Those in the Gig Economy (London: Department for Business, Energy & Industrial Strategy, 2018).

79 Aaron Smith, Gig Work, Online Selling and Home Sharing (Pew Research Center, 17 November 2016).

80 Diana Farrell and Fiona Greig, The Online Platform Economy: Has Growth Peaked? (New York: JPMorgan Chase Institute November 2016).

81 Lawrence F. Katz and Alan B. Krueger, The Rise and Nature of Alternative Work Arrangements in the United States, 19952015. NBER Working Paper No. 22667 (Washington DC: National Bureau of Economic Research, September 2016),

82 For example, Chartered Institute of Personnel and Development, To Gig; Steven Hill, ‘Good Riddance, Gig Economy: Uber, Ayn Rand and the Awesome Collapse of Silicon Valley’s Dream of Destroying Your Job’. Salon, 28 March 2016,; A.J. Wood et al., ‘Good Gig, Bad Gig: Autonomy and Algorithmic Control in the Global Gig Economy’. Work, Employment and Society 33, no. 1 (2018): 56–75.

83 This was undertaken by the ILO in 2016 and involved surveying 1,800 workers in Crowdflower and AMT. Janine Berg, Income Security in the on-Demand Economy: Findings and Policy Lessons from a Survey of Crowdworkers (Geneva: Inclusive Labour Markets, Labour Relations and Working Conditions Branch, International Labour Office, 2016).

84 Smith, Gig Work.

85 Ibid.

86 Chartered Institute of Personnel and Development, To Gig.

87 Smith, Gig Work.

88 Lepanjuuri, Wishart, and Cornick, Characteristics.

89 Chartered Institute of Personnel and Development, To Gig.

90 Berg, Income Security.

91 Chartered Institute of Personnel and Development, To Gig.

92 Berg, Income Security.

93 Chartered Institute of Personnel and Development, To Gig.

94 Lepanjuuri, Wishart, and Cornick, Characteristics.

95 Berg, Income Security.

96 Kotaro Hara, Abi Adams, Kristy Milland, Saiph Savage, Chris Callison-Burch and Jeffrey P. Bigham, ‘A Data-Driven Analysis of Workers’ Earnings on Amazon Mechanical Turk’, paper presented to CHI 2018 conference, Montreal, 21–26 April 2018,

97 Unions NSW, Innovation or Exploitation: Busting the Airtasker Myth (Sydney: Unions NSW, 2016).

99 Kahneman, Thinking, Fast and Slow.

100 Steven Lukes, Power: A Radical View, 2nd ed. (London: Palgrave, 2005), 124.

101 Chartered Institute of Personnel and Development, To Gig.

102 Ibid.

103 Caroline O’Donovan, ‘Uber Rallies Drivers against Teamster Unionization Efforts with Podcasts and Pizza Parties’. Buzzfeed News, 10 March 2017; Kate Minter, ‘Negotiating Labour Standards in the Gig Economy: Airtasker and Unions New South Wales’. Economic and Labour Relations Review 28, no. 3 (2017): 438–54; Anna Sansom, ‘Bicycle Couriers Protest against Takeaway Food Service Deliveroo’. France24, 12 August 2017,; Zachary Kilhoffer, Karolien Lenaerts, and Miroslav Beblavý, The Platform Economy and Industrial Relations: Applying the Old Framework to the New Reality. CEPS Report No. 2017/12 (Brussels: Centre for European Policy Studies, 2017); Rebecca Burns, ‘The Sharing Economy’s “First Strike”: Uber Drivers Turn Off the App’. In These Times, 22 October 2014,; Alison Griswold, ‘Uber Just Caved on a Big Policy Change after Its Drivers Threatened to Strike’. Slate, 12 September 2014,; Adam Carey, ‘Uber Strike: Group Calls for Drivers to “Log Off” in One-Day Protest over Pay and Conditions’. The Age, 12 March 2017.

104 BBC, ‘Bike Courier Wins “Gig” Economy Employment Rights Case’. BBC News, 7 January 2017; Jamie Grierson and Rob Davies, ‘Pimlico Plumbers Loses Appeal against Self-employed Status’. Guardian, 11 February 2017; Gregor Gall, ‘Is Uber Ruling the Beginning of the End for Bogus Self-Employment?’. The Conversation, 5 November 2016.

105 ‘Uber Employee Question Not Settled: Expert’, Workplace Express, 9 February 2018, (subscription required), referring to [2017] FWC 6610 (21 December 2017).

106 Jenna Castle, ‘UberX vs Taxi—Which Is Best?’. Choice, Australian Consumers Association, 14 August 2017,; Stephen Corby, ‘Uber vs Taxi’. CarsGuide, 31 October 2017,; Caspar, ‘Sydneysiders Have Spoken—and They Choose Ridesharing!’. Uber Newsroom, Uber, 10 October 2015,

107 Charlie Lewis, ‘What Does the Gig Economy Do to Workers’ Rights?’. Crikey, 13 April 2017; Laura Gartry, ‘“Uber for Hospitality” App Hopes to Shake up on-Demand Job Market’. ABC News, Australian Broadcasting Corporation, 6 April 2017,

108 Venee Dubal, ‘The Digitalisation of Day Labor as Gig Work’, OnLabor, 7 May 2019,

109 Ibid.

110 Hill, ‘Good Riddance’.

111 Benjamin Edelman. ‘Uber Can’t Be Fixed — It’s Time for Regulators to Shut It Down’. Harvard Business Review, 21 June 2017,

Previous Next