The platform economy
Consequences for labour, inequality and the environment
The current moment is a fateful one for the global community. We are witnessing a rise of authoritarian populism and the growth of anti-immigrant sentiment in many countries. Global temperatures are accelerating rapidly, bringing an intensification of dangerous impacts, such as melting permafrost, storms, wildfires and heat waves. Financialisation and the growth of extreme income and wealth concentration are destabilising societies that had relied on reasonable distributions of economic benefits. Democratic and other social institutions are under attack from these factors. Nuclear aggression is in the air.
In the midst of these perilous trends, studies of the ‘sharing economy’ may seem a bit beside the point. Companies such as Uber or Airbnb employ only very small fractions of the global workforce, and the sector as a whole only boasts a few real success stories. Yet the larger platform economy is but one part of another tidal wave that will be occurring around the world – rapid labour-displacing technological change. While the technological determinists who predict that artificial intelligence (AI) will wipe out massive numbers of the world’s jobs in a short time are surely overstating future impacts (see Arnold et al. this volume), it is undeniable that technological advances in digitisation and AI are proceeding quickly, and that they will have far-reaching consequences, even if we do not fully understand them yet. There will be big changes in labour markets, and it is likely that technological displacement will contribute to the growth of extreme inequality. Societies that shift to large scale use of AI are likely to experience significant social pressures and even dislocation. Perhaps most importantly, the historic safety valve for absorbing displaced labour – GDP growth – is much harder to achieve in wealthy societies now.
While economists do not fully understand why, low growth appears to be a new but persistent feature of the global north. Furthermore, even if rapid growth were magically to reappear, meeting the massive emissions reductions that are now necessary to address climate change will be very difficult, if not impossible, in that context. Some scientists have been forthright enough to argue that rich countries need 10% annual reductions in emissions, a number that is far beyond the range of current experience, and one that is nearly impossible to square with continued growth (Anderson 2012). Thus, the fourth industrial revolution is likely to collide with political instability, threats to democratic institutions, and climate change. The platform economy, which represents a large-scale reorganisation of many kinds of work, is at the cutting edge of one type of technological transformation.
What exactly is this sector? Broadly, it represents a set of platforms that use algorithms to match buyers and sellers in a range of goods and labour services. There are both consumer and business-oriented platforms. Amazon’s Mechanical Turk and Upwork are examples of the latter. This paper is concerned with the consumer-facing firms – in areas such as lodging, errands and tasks, durable goods rental and transportation.1 These are the key features of these platforms:
- They use sophisticated logistics software (or algorithms) for matching and payment.
- Providers on the platforms are independent contractors (rather than employees).
- There are very low barriers to entry for providers on most platforms.
- Trust is achieved via crowdsourcing of ratings and reputational data, typically on both sides of the market.
There’s a great deal of terminological dispute about the sector – with alternatives being collaborative consumption, sharing economy, on-demand economy and the gig economy. For reasons of space I will leave aside these disputes, which we have discussed elsewhere (Frenken and Schor 2017; Schor and Attwood-Charles 2017; Schor and Fitzmaurice 2015). There has also been a great deal of controversy about the sector, with supporters touting its efficiency and low cost and detractors claiming the platforms are exploiting workers, destabilising neighbourhoods, and acting illegally (Schor and Attwood-Charles 2017). While it is difficult to predict exactly how the sector will evolve, after conducting seven years of research (2011–2018), using a variety of methods, I am prepared to offer a number of findings, centring on three main issues.2
First, in contrast to frequent claims, platforms are less disruptors than reproducers of ongoing trends in labour markets. In particular, their impacts appear to be inequality-enhancing, rather than reducing, and racism appears to be endemic to their operation. Second, the growth of platforms is likely to be associated with rising carbon footprints, although this prediction awaits empirical testing. Finally, we ask whether the sector is sustainable, or whether its continued growth is contingent on a parasitic relationship to conventional unemployment.
Disruption or reproduction?
Platform companies and many observers claim that this form of economic organisation represents a disruptive and novel way to organise economic activity. Others emphasise continuities with ongoing trends. For example, the de-institutionalisation of labour markets is a decades-long trend across OECD countries. In the US the trend towards what has been called precarious, or non-standard, labour has been observed since the 1980s. As noted above, platform work is nearly always organised via independent contracting. Does the emergence of the platform economy signal ‘the end of employment’, as some would have it (Sundararajan 2016), or is it an unsustainable way to organise labour markets, as others suggest? Our research on the everyday functioning of platforms provides insight into this question, and supports the latter interpretation (Schor et al. 2017). A main finding is that the model of independent contracting is difficult to sustain on its own terms because to achieve a viable model of labour management, platforms are ‘free-riding’ on conventional employment. Indeed, they have yet to show that their model works as a stand-alone. While the platforms claim that their providers prefer the independent contractor model, with the flexibility and autonomy that it provides, we find that this is true only for those who use the platform for supplemental earnings, rather than to pay their basic expenses.
It is important to remember that platforms have very low barriers to entry and attract a wide array of earners. While the companies do not release much data on their workforces, national surveys in the US, for example, find that less than 30% rely on their platform earnings as their means of subsistence. The vast majority have other sources of income. While this varies by platform, it remains a general rule almost everywhere (see Huws et al. this volume). Many in the supplemental earner category already have full-time jobs. Those in the partial-dependence category (some reliance on platforms to pay basic expenses) have significant alternative income. What we have termed ‘dependent’ workers rely solely on the platform. However, few in this group earn more than poverty wages. They experience extreme precarity, have less job satisfaction and autonomy, and are unlikely to persist if viable alternatives appear for them. If AI reduces conventional employment, we cannot expect the platform model to be a successful alternative without major changes to its methods of managing labour (see Petropoulos this volume).
Another main theme in the disruption–reproduction debate concerns inequality. Are platforms reducing wage inequality by providing new opportunities, particularly for less advantaged workers, as some claim (Fraiberger and Sundararajan 2015)? Or are they exacerbating existing patterns of privilege? While there is little question that platforms are offering new opportunities to the middle class, our research on the US suggests that these are mainly being taken advantage of by more privileged members (Schor 2017). We have argued that they are fostering an upward redistribution of opportunity and income within the top 80%; platform workers are disproportionately well educated, with majorities of college-educated providers on most platforms (Schor et al. 2017). In our qualitative work, we find that manual, often ‘dirty’ work, like housecleaning and driving, is being done by college-educated providers, who are displacing less educated workers. For example, TaskRabbit, Airbnb and Uber substitute for traditional housecleaning, hotel chambermaids and taxi drivers: as demand for the former expands, this has a negative impact on the latter. Chambermaids are unlikely to have apartments to rent on Airbnb. Although some taxi drivers have switched to ride-hailing apps, anecdotal evidence suggests their incomes plummet with the switch. Furthermore, many taxi drivers lack access to the latest vehicle models necessary for driving on platforms. Overall, we suspect that the additional income earned by high educated platform providers worsens the distribution between them and lower educated persons at the bottom of the income scale.
A second dimension of inequality is around race. In the US there is growing evidence that the platform economy is fostering racial discrimination, via the peer-to-peer structure of the exchanges. Every study we have seen confirms the existence of racially based discrimination. It is taking place on both sides of these platforms – customers are discriminating against providers and providers are discriminating against customers. In our research we find that in areas with high proportions of non-white residents, prices are lower, revenue is less and ratings are reduced (Schor et al. 2017)In an audit study, Harvard researchers found that Airbnb hosts were 16% more likely to refuse to rent to guests with African-American sounding names (Edelman, Luca and Svirsky 2017). This research received wide press coverage, a series of responses by the company, and attempts to create a new, ‘noir’ Airbnb for people of colour. Similarly, a field experiment of Uber and Lyft found that drivers cancelled on riders with African-American sounding names twice as often as riders with white sounding names and that African-American named customers had to wait longer. This study also found women were cheated more on these apps (Ge et al. 2016). A study of TaskRabbit found that Taskers were unwilling to provide services in areas with heavy concentrations of non-white residents (Thebault-Spieker, Terveen and Hecht 2015). Thus, on balance, it appears that rather than eliminating, or ‘disrupting’, racial inequalities, they are being transported into the ‘sharing economy’.
Changing consumer patterns
While there has been considerable criticism of platforms’ labour practices, and of the impacts of Airbnb on the availability of housing and neighbourhood quality, there has been less debate about how these companies are affecting consumers. But here too there are worrisome impacts, especially over the longer run.
A key part of the appeal of the platforms is new services and low prices.3 The benefits vary by platform: on Airbnb, much lower prices, local neighbourhoods and personalisation are key to consumer satisfaction. On Uber, factors include low prices, convenient payment, availability and ease of use. On TaskRabbit, middle-class consumers get middle-class providers, in contrast to the informal ‘errands and tasks’ market. For other services, such as the delivery of food or consumer goods, the platforms make ‘servant labour’ available at relatively low cost. These platforms allow middle-class and upper-middle-class consumers to access services previously reserved for the wealthy.
What of the worrisome trends? The most serious is environmental impacts. A major theme in the early years of the sharing economy was that these new services were more environmentally beneficial than existing businesses, in part because they were using ‘idle resources’; Airbnb claimed it would reduce new hotel construction. Ride-sharing apps like Uber and Lyft were expected by many to reduce car ownership, increase the number of passengers per ride, and reduce carbon emissions. However, it has been difficult to assess these claims because the companies will not provide their data to independent researchers. But there are strong reasons to believe that platforms are increasing, rather than reducing environmental impacts, and especially climate emissions.
The evidence is hiding in plain sight: lower prices lead to more demand. In the lodging sector, cheap accommodation increases miles travelled and trips taken. Furthermore, Airbnb enables hosts to rent out their homes when they travel, so that lodging is essentially free. (We also find some hosts travel specifically to rent, to take advantage of price arbitrage – they can rent out their homes at a higher rate than the places they stay at.) Similarly, in the US ride-hailing apps appear to be taking people away from lower-carbon modes of transport. A recent study based on survey data finds that had there been no transportation app, 49–61% of ride-hailing trips would have either not been made at all, or been taken via walking, biking or transit (Clewlow and Mishra 2017). Furthermore, this study finds that there is no reduction in car ownership as a result of ride-hailing. The authors conclude that these services are likely to increase rather than reduce vehicle miles travelled. In the US, at least, if the transportation apps continue to grow it seems likely that they will further strain public transportation budgets by reducing ridership and weakening public support. This would have disastrous carbon consequences, as transportation is already the largest source of greenhouse gas emissions in the US and is a major contributor in many countries.
Has the sharing economy peaked?
In 2017, a research report suggested that growth in the US sharing sector had peaked (Farrell and Greig 2017). The torrid expansion of the previous few years appeared to be over. Platform incomes were even falling in some cities, labour turnover was extremely high (with more than half of all participants dropping out after a year), and the strengthening of the conventional labour market was reducing the pool of interested workers. While our research has found that employed workers are more likely to be satisfied on the platforms, this data also shows that they are less likely to stay. These findings suggest an obvious point that much of the discourse has failed to recognise: the platform economy remains tethered to the conventional labour market. The precarious model of independent contracting without benefits, protections or guaranteed income is unlikely to be preferred, except for those who can command superior market positions.
A number of other developments also suggest that the ‘end of employment’ future may be little more than fantasy. They concern the only two very large platforms in this sector: Airbnb and Uber. San Francisco recently enacted strong regulations to curtail Airbnb hosting, which has dramatically reduced the number of hosts who are eligible and registered. One report found that only 15% of hosts have registered (Brinklow 2017). Stricter regulations are being debated and enacted in cities around the world, and they will predictably reduce the growth of lodging platforms, particularly Airbnb. In the case of Uber, whose valuation recently dropped by one-third, the lack of a viable business model may be a more serious constraint than regulation. Independent analysis suggests that Uber will have lost $5 billion in 2017 and that passengers are paying only 41% of the cost of their rides, with Uber’s investors subsidising the remainder in the hopes of achieving market domination (Smith 2016). However, competition in this market may be increasing. Furthermore, even if Uber were to best its competition, there is reason to believe consumers will balk at a more than doubling of fares, particularly since so many trips would not otherwise be taken.
Policies for the future
Our findings, at least to date, suggest that letting algorithms drive economic activity will further privilege the privileged. Furthermore, the platform sector fails to provide adequately for those without secure alternative sources of income, and exacerbates existing forms of social inequality. If we want to reap the benefits of platforms (and there are many) without this dark side, higher levels of regulation and new patterns of governance will certainly be necessary. Our research finds that platforms work best when workers have alternative means of support, and participate freely and without compulsion. One option is to reduce workers’ dependency through broad social measures: stronger welfare support for the unemployed, a basic income, or more collective provisioning of basic needs could reduce desperation for platform providers with no other sources of income (see Palier this volume). Alternatively, platforms could be required to provide regular benefits and protections for workers who are essentially full-timers, who work over a certain number of hours a week. In that scenario, the independent contractor status would be reserved only for those who work below that threshold.
A second issue is that platforms must begin to monitor and take responsibility for their carbon impacts. The world cannot afford a dynamic new sector with a high carbon footprint. Commitments to data transparency are essential, in order to craft environmentally positive policies. Possibilities include carbon taxes on lodging (eg Airbnb) stays and ride-hailing services (eg Uber, Lyft and others). In many localities, platform companies are getting preferential regulatory treatment. In return they should commit to a strong sustainability agenda to control and reduce their environmental impacts.
Finally, platforms must tackle discrimination head-on. Racial discrimination in public accommodation has been outlawed in many societies. Consumers have legal rights to access to lodging, transportation and labour services. While the rise of a person-to-person alternative means of exchange has many beneficial aspects, it must not be allowed to re-inscribe racial and other forms of discrimination. Here what is required is a combination of legislation outlawing discrimination and new policies by the companies. Solutions include eliminating or de-emphasising pictures that show skin colour and company monitoring and punishment of discriminatory behaviour.
The ‘sharing’ or consumer-facing portion of the platform economy has proven to be an attractive option for consumers and many providers. To date, it has proven to be neither the earth-shattering innovation its proponents claim nor the absolute dystopia its detractors have asserted. However, negative impacts are already significant, and if it continues to grow, these harms will as well. To preserve the potential benefits, it will behove government at all levels to craft legislation and regulation that controls the impacts it is already having on labour, climate and public goods.
1 This is the portion of the platform sector typically referred to as the ‘sharing economy’.
2 For more detail on our project, and copies of our papers, see: https://www.bc.edu/bc-web/schools/mcas/departments/sociology/connected.html.
3 There are also ideological, or ‘moral’ appeals that matter to users, as we detail in Domesticating the Market: Moral Exchange and the Sharing Economy (Fitzmaurice et al. 2018).
Anderson, K. (2012), ‘Climate Change Going Beyond Dangerous: Brutal Numbers and Tenuous Hope’, Development Dialogue: What Next, 3: 16–40, http://www.whatnext.org/resources/Publications/Volume-III/Single-articles/wnv3_andersson_144.pdf.
Brinklow, A. (2017), ‘Airbnb Says Online Registration Will Legalize all SF hosts’, Curbed San Francisco, 6 September, https://sf.curbed.com/2017/9/6/16263986/airbnb-online-registration-san-francisco.
Clewlow, R. R. and Mishra, G. S. (2017), Disruptive Transportation: the Adoption, Utilization, and Impacts of Ride-Hailing in the United States, Davis, CA: Institute of Transportation Studies, University of California.
Edelman, B., M. Luca and D. Svirsky (2017), ‘Racial Discrimination in the Sharing Economy: Evidence from a Field Experiment’, American Economic Journal: Applied Economics, 9(2):1–22.
Farrell, D. and F. Grieg (2017), The Online Platform Economy: Has Growth Peaked?, JP Morgan Chase Institute, New York City and Washington, DC: JP Morgan Chase Institute.
Fitzmaurice, C., I. Ladegaard, W. Attwood-Charles, L. B. Carfagna, M. Cansoy, J. Schor and R. Wengronowitz (2018), ‘Domesticating the Market: Moral Exchange and the Sharing Economy’, Socio-Economic Review, forthcoming.
Fraiberger, S. and A. Sundararajan (2015), Peer-to-Peer Rental Markets in the Sharing Economy, NYU Stern School of Business Research Paper.
Frenken, K. and J. Schor (2017), ‘Putting the Sharing Economy into Perspective’, Environmental Innovation and Societal Transitions, 23: 3–10.
Ge, Y., C. R. Knittel, D. MacKenzie and S. Zoepf (2016), Racial and Gender Discrimination in Transportation Network Companies, NBER Working Paper 22776.
Schor, J. and C. Fitzmaurice (2015), ‘Collaborating and Connecting: the Emergence of the Sharing economy’, in L. A. Reisch and J. Thørgensen (eds), Handbook of Research on Sustainable Consumption, 410–25. Northampton, MA: Edward Elgar.
Schor, J. (2017), ‘Does the Sharing Economy Increase Inequality within the Eighty Percent? Findings from a Qualitative Study of Platform Providers’, Cambridge Journal of Regions, Economy and Society, 10: 263–279.
Schor, J. and W. Attwood-Charles (2017), ‘The Sharing Economy: Labor, Inequality and Sociability on for-Profit Platforms’, Sociology Compass, Under Review.
Schor, J., W. Attwood-Charles, M. Cansoy, I. Ladegaard and R. Wengronowitz (2017), ‘Dependence and Precarity in the Platform Economy’, unpublished paper, Boston College.
Smith, Y. (2016), ‘Can Uber Ever Deliver? Part One: Understanding Uber’s Bleak Operating Economics’, Naked Capitalism, 30 November, https://www.nakedcapitalism.com/2016/11/can-uber-ever-deliver-part-one-understanding-ubers-bleak-operating-economics.html.
Sundararajan, A. (2016), The Sharing Economy: the End of Employment and the Rise of Crowd-Based Capitalism, Cambridge, MA: MIT Press.
Thebault-Spieker, J., L. G. Terveen and B. Hecht (2015), ‘Avoiding the South Side and the Suburbs: the Geography of Mobile Crowdsourcing Markets’, in Proceedings of the 18th ACM Conference on Computer Supported Cooperative Work & Social Computing, https://dl.acm.org/citation.cfm?id=2675278.
Warning: count(): Parameter must be an array or an object that implements Countable in /home/policyne/public_html/wp-content/themes/Policy Network/inc/article/more.php on line 4