Does the Sharing Economy Increase Inequality Within the Eighty Percent?: Findings from a Qualitative Study of Platform Providers

Home , TaskRabbit, Turo (company)

Cambridge Journal of Regions, Economy and Society 2 0 1 7, 10, 263–279 doi:10.1093/cjres/rsw047 Advance Access publication 10 February, 2017

Does the sharing economy increase inequality within the eighty percent?: findings from a qualitative study of platform providers

Juliet B. Schor Department of Sociology, Boston College, 531 McGuinn Hall, 140 Commonwealth Avenue, Chestnut Hill, MA 02467, USA, [email protected]

Received on November 15, 2015; editorial decision on August 8, 2016; accepted on December 19, 2016

The sharing economy has generated controversy for its effects on labour conditions, wages and the distributions of income and wealth. In this article, we present evidence for a previously unrecognized effect: increased income inequality among the bottom 80% of the distribution. On the basis of interviews with US providers on three for-profit platforms (Airbnb, RelayRides and TaskRabbit), we find that providers are highly educated and many have well- paying full-time jobs. They use the platforms to augment their incomes. Furthermore, many are engaging in manual labour, including cleaning, moving and other tasks that are traditionally done by workers with low educational attainment, suggesting a crowding-out effect.

Keywords: sharing economy, income inequality, platform economy, Airbnb, TaskRabbit JEL Classification:J01, J46, Z13

Introduction costs for person-to-person exchange, in part The ‘sharing economy’ comprises a diverse set by crowdsourcing information from users but of platforms and organisations, including non- also via their sophisticated logistics software profits such as time banks, food swaps and mak- (Sundararajan, 2016). Some observers have erspaces, as well as for-profit platforms that offer gone so far to predict a “zero marginal cost income-earning opportunities, such as home society” (Rifkin, 2014), in which highly produc- and car rental and the sale of goods and labour tive technologies combine with users to remake (Schor and Fitzmaurice, 2015). The for-profits, economic relations. Investors are also optimis- which are often quite large, have attracted a tic about the sector, as recent valuations of for- great deal of popular attention, in part because profit companies in this sector have been high. they have the potential to yield economic The ride-sourcing platform Uber, which is the benefits by replacing conventional economic largest of all sharing economy companies, was activity with new technologies and innovative valued at US$50 billion in 2015, which at the business models. Proponents argue that tech- time made it more valuable than 80% of all nologically based disruptions will enhance eco- companies on the Standard and Poors index nomic efficiency, flexibility and autonomy for (Myers, 2015). Rapid growth in the two largest providers. Sharing platforms reduce transaction companies—Uber and Airbnb—also reveals

© The Author 2017. Published by Oxford University Press on behalf of the Cambridge Political Economy Society. All rights reserved. For permissions, please email: [email protected] Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

the attractions of these services to consumers, Despite its common-good rhetoric, the sec- who, we find, appreciate the low costs, conveni- tor is seen as part of an anti-labour offensive ence and branding of many of the platforms. from business which is expanding the precariat However, sharing economy platforms have and facilitating a larger risk shift onto workers also become objects of heated controversy (Hacker 2006; Standing, 2011). The for-profit around the world. To some extent this is because platforms are described as super-exploiters of many of them launched with a rhetoric of com- labour, as in Trebor Scholz’s evocatively titled mon-good claims (Schor, 2014). Most of the large book Uberworked and Underpaid (Scholz, companies in the sector have taken credit for 2016; Slee, 2015; Ravenelle, 2017). This is partly reducing ecological and carbon footprints, pro- due to their practice of classifying providers as viding opportunity for people who are struggling independent contractors rather than employ- economically, building social connections and, in ees, which absolves them of responsibility for the case of Airbnb, fostering cultural exchange. expenses, benefits and employment security. The Airbnb conducts local impact studies to show its sharing economy is seen as an ultra-free mar- positive effects on local communities, looking at ket which is resulting in a race to the bottom, or both economic and ecological outcomes (http:// what political economist Robert Reich termed a blog.airbnb.com/economic-impact-airbnb_ and “share the scraps” economy (Reich, 2015). http://blog.airbnb.com/environmental-impacts- While the sharing economy is frequently of-home-sharing/). Founders, consultants and conceptualised as sui generis, it is more use- many users argue that the sharing economy is a ful to consider it within its broader context force for social and ecological good, an alterna- and ask whether it strengthens or undermines tive to a dysfunctional and inefficient conven- larger economic trends. For example, in recent tional economy (Botsman and Rogers, 2010). As decades inequality has increased sharply in the companies grew, observers assessed these many countries. Most attention has been paid claims and found that many of the platforms to the concentration of income at the very top, were coming up short (Schor, 2014). including among the 1% (Piketty, 2014). Within Critics have assailed the sharing economy on the sharing sector, there has been attention a number of fronts. One issue is terminology and to the large fortunes being made by founders whether renting or providing labour services is and venture capitalists (Schneider, 2014; Schor, properly considered ‘sharing’. Anthony Kalamar 2014), which raises the question of whether (2013) has argued that these exchanges crowd the sharing economy is contributing to the out genuine sharing and that for-profit compa- increase in extreme inequality. Alternatively, nies are ‘sharewashing’, i.e., using the positive some argue that it is reducing inequality by associations of sharing to hide their self-inter- spreading opportunity and providing incomes ested activities. Ravenelle (2017) reports that to people at the bottom of the distribution the New York City providers she studied reject (Fraiberg and Sundarajan, 2015). However, the sharing designation, seeing themselves more these debates have had relatively little empiri- like workers. A related critique is that selling cal data to inform them. The platforms have slivers of one’s life (room, car, time, attention) is been largely unwilling to share their data, par- a commodification of daily life that will under- ticularly to independent researchers, and much mine genuine social connection and solidarity of this activity is not captured in government (Henwood, 2015; Morozov, 2013). surveys. As such, it is not possible to give defini- A second line of argument asserts that the tive answers to these questions. Furthermore, sharing economy is exacerbating neo-liberal the debate is not just about current practices, economic trends and policies which favour but also what the effects of the sector will be business and undermine the power of labour. as it grows.

264

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Does the sharing economy increase inequality within the eighty percent?

In this article, we explore a potential impact on interviews with 43 earners on three plat- of the sharing economy, which, to our knowl- forms (Airbnb, RelayRides and TaskRabbit). edge, has not yet been raised in the popular Interviews are semi-structured, range from press or the academic literature: how sharing 45 to 90 minutes and cover a range of topics, economy activity is affecting the distribution including participants’ life narratives, how they of income and opportunity within the bottom got involved in the platform, motives, attitudes 80% of the population. To answer this ques- toward risk and experiences. Interviews are con- tion, we use a sample of 43 providers on three centrated among people aged 18–34 because the platforms. To anticipate our results, we find innovators and early users of the sharing econ- that most providers are highly educated, with omy come from this age group. Almost all of the other sources of income. We also find that they interviews were conducted in 2013; however, in are engaging in activities that have tradition- 2015 we conducted follow-up interviews with 9 ally been blue and pink collar tasks. Our data TaskRabbit and Airbnb providers. is collected at the individual level, however, if Recruitment differed slightly by platform. our findings are generalisable, platform activ- In all three cases, we first eliminated users ity is likely exacerbating inequality within the who were obviously outside our age range. 80%, shifting more income and opportunity to Then we randomly sent communications via better-off households and providers. the platform. If we inadvertently contacted a We begin with a short description of our person outside the age range, we declined to methods and the platforms we are discussing in set up an interview. We also required that the this article. We then turn to definitions, because person had done at least five trades to be eli- there is considerable confusion about terminol- gible for the study. On TaskRabbit, we posted ogy and exactly what the ‘sharing economy’ is. the interview as a task, which readily yielded Next, we address the context of rising inequality informants. On Airbnb and RelayRides, we and the ongoing impacts of the 2008 financial queried providers via the platform, and once crash. We then discuss findings from interviews we made contact we let them know we were with providers on three for-profit platforms— interested in interviewing them. This method Airbnb, RelayRides (now renamed Turo) and yielded enough informants on RelayRides, TaskRabbit. We discuss the demographic char- but on Airbnb the platform repeatedly deac- acteristics of our sample of providers, their tivated our account when it realized that we earnings and the content of their work. Then were attempting to interview hosts. We then we consider how opportunities on these sites reverted to snowball sampling. We also faced vary between highly educated and low-income/ this problem at one point with TaskRabbit, low educational attainment providers. when they saw our posting and tried to stop us from interviewing. However, they did not de-activate us. To date, these platforms (and Methods others) have not made data available to The findings we report on in this article are part researchers, which has hampered our ability of a larger programme of research on the shar- to study them. (I twice had encouraging con- ing economy, which has been funded by the versations with Airbnb to gain access to their MacArthur Foundation (http://clrn.dmlhub.net/ data, but both were unsuccessful.)1 projects/connected-consumption). Since 2011, Overall, our sample consisted of 23 men and our research team has studied more than 10 shar- 20 women. Thirty-five (or 81%) classified them- ing economy initiatives, done approximately 275 selves as white, and eight were non-white. Of interviews and conducted hundreds of hours of those we had three Latino/as, four Asians and participant observation. In this article, we draw one Afro-Caribbean man.

265

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

Here, we offer a brief description of each product tester, or doing translation or online platform. Airbnb, largest of the three, is a hous- shopping. At the time of our first round of ing exchange that began in San Francisco in interviews, TaskRabbit used an auction model. 2008. Originally, hosts offered rented rooms in ‘Posters’ provided a description of the job they their own homes and apartments, but over time wanted done with a maximum price they were a much wider range of offerings became avail- willing to pay for the task. ‘Rabbits’ (hereafter able, including whole apartments and houses referred to as providers or workers) then bid that are unoccupied by owners. The site con- for the job and the poster opted for his or her sists of a set of listings, with photos, descriptions preferred provider. All providers are vetted by of the lodging, profiles of the hosts and other the company, and have profiles with pictures information. Prices are set by the host. Like and descriptions of themselves. In 2014, the almost all peer-to-peer sites, this one offers rat- platform undertook a radical redesign in which ings and comments about the hosts and their it eliminated auction pricing and set an hourly lodgings. At the time we conducted our inter- wage range for tasks. It also renamed the work- views, Airbnb was known for a rhetoric that ers ‘Taskers’, and shifted from an open format emphasized cultural exchange, meeting people in which clients could post any type of task, to and the homey-ness of its offerings. It calls itself one in which it offers tasks from a pre-set list. a community and has been at the forefront of All three platforms derive their revenue by the idea of “sharing” in this sector. taking a fraction of each completed transaction. RelayRides, a person-to-person car rental The percentage differs across the platforms, site, was founded in Boston in 2009. As with and varies with the nature of the exchange, Airbnb, owners list their cars on the site, with making it difficult to generalize. However, the pictures, descriptions of the car and profiles and range is large, from a current high of 35% for pictures of themselves. The site also calls itself a some transactions on TaskRabbit to 9–15% on community, although its rhetoric is more trans- Airbnb, combining both guest and host service actional and functional than Airbnb’s. Renters fees. At the time of our interviews, all three must have their identity verified, and the $1 platforms were growing. million insurance policy that RelayRides offers is prominently advertised on their site. Ratings of cars and drivers are also an important part of Defining the sharing economy: peer this site. RelayRides emphasizes convenience, to peer platforms value, selection and risk management. In 2015, What is popularly termed the ‘sharing econ- RelayRides rebranded itself as Turo as part of omy’ is a diverse sector. We have previously an attempt to orient its business toward out-of- (Schor and Fitzmaurice, 2015) identified five town renters. types of sharing economy sites and activi- TaskRabbit is a labour services site that spe- ties. The first, which is probably the category cializes in errands and relatively low skill tasks. most closely associated with the term sharing It was also founded in Boston in 2008, under economy, is sites that increase the utilisation of a different name (RunMyErrand), which was durable assets, via rental or free use. Examples changed to TaskRabbit in 2010. On this plat- include Airbnb and Couchsurfing. The second form, customers hire “Rabbits” to perform category is labour and service exchange sites, tasks such as house cleaning, delivery services, such as timebanks, TaskRabbit or Postmates. handyman work, computer tasks, pet sitting, The third is crowdfunding sites, such as moving and assembling furniture. Our inter- Kickstarter, Gofundme or Indigogo. The fourth viewees also reported engaging in non-man- is sites that facilitate the recirculation of goods, ual tasks such as being a virtual assistant or including the resale or gifting of used goods,

266

such as Yerdle and Freecycle. A final category and worker-provided tools and assets) have is a hybrid which combines both labour and been identified by the US Department of a tangible product, such as etsy, which offers Commerce, which has provided the first gov- handcrafts, and Feastly, a peer-to-peer site on ernmental definition of the sector, using the which aspiring chefs sell dinner spots at their terminology of ‘digital matching firms’ (Telles, homes or pop-up venues. 2016). However, this definition does not include Given this diversity of activities, it is not the many non-profit organisations that are typi- surprising that there has been a proliferation cally considered part of the sharing economy, of terms to describe the sector. These include such as food swaps and makerspaces, which collaborative consumption, on-demand labour, may not use matching software. They are part the gig economy, the peer economy, the access of the sharing economy but are not platforms. economy and the platform economy. (For one For the remainder of this article, we will use the discussion, see Botsman, 2015.) One termino- terms ‘platform economy’ and ‘sharing econ- logical issue is controversy about the appro- omy’ interchangeably, as we are only discussing priateness of the term ‘sharing’ (Schor and sites which do use platforms and apps. Attwood-Charles, 2017). Critics argue that We also reserve the terms ‘sharing economy’ monetized transactions on platforms such as and ‘platform economy’ for structures that are Airbnb or ride-sourcing apps like Uber are not organized via person to person, or Peer-to-Peer sharing (Kalamar, 2013; Slee, 2015). (P2P) exchange. The term P2P comes from the A second issue is that there is little analytic open source software movement, and refers to coherence or practical consistency to how open-access communities of collaborating indi- these terms are used (Schor, 2014). Amazon’s viduals (Benkler, 2006). In the sharing economy, Mechanical Turk, which is a digital labour many sites are organized as person to person, platform, is very similar to the errands site with the platform operating as a ‘middleman’ or TaskRabbit except that all work on Mechanical broker. This is the case for the three platforms Turk is digital and TaskRabbit includes both that we study in this article (Airbnb, RelayRides online and offline work. However, Mechanical and TaskRabbit). By this definition, Zipcar, Turk is almost never considered part of the which is considered by some to be the first sharing economy and TaskRabbit always is. car sharing company, is not part of the sharing Uber has never identified itself as belonging economy, because it owns the cars and is there- to the sharing economy, but Lyft, which pro- fore considered a Business-to-Peer (B2P) entity. vides a nearly identical service, always does. For this discussion, we exclude B2P companies Furthermore, the popular press nearly always because they are not sufficiently different from classifies Uber as a sharing economy company. conventional businesses. (Zipcar was originally However, there are some ways to differen- novel because it placed cars within neighbour- tiate among these labels. Collaborative con- hoods and rented them for shorter time periods sumption, the term used by Rachel Botsman than a day. Now it is less so.) Co-working spaces, (2010), mainly refers to sites that increase the another B2P model, are not functionally differ- utilisation of durable assets. On-demand and ent from conventional shared office space rental gig labour are used for labour services sites. models. We recognize that not all observers of Platform economy is broader, and we use it to the sharing economy agree with restricting the denote for-profit companies that use platforms definition to P2P models, however, we have fol- and apps, crowdsource ratings and reputational lowed this practice to highlight what is different data, and use digital technology to organize about this sector. exchanges. These characteristics, plus two per- It may be worth addressing the question of taining to labour conditions (flexible schedules whether Airbnb is a P2P or a B2P platform,

267

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

in view of claims that business entities offer- overstate quality (Overgoor et al., 2012). ing multiple properties are now prevalent on This is true of Airbnb, according to a recent the platform. Slee (2015) did a 2013 analysis study (Zervas et al., 2015). In general, we believe of listings in New York City, in which he found that users are likely overstating the accuracy of that while 87% of hosts have a single listing, the ratings and reputational data on these sites. the remaining 13% accounted for 40% of list- However, because there seem to be relatively ings. (This is a fluctuating number, because the few malfeasants on platforms at this time, that company periodically purges these high vol- overstatement may not be recognized. The data ume listers.) We have conducted a more recent may also be better at revealing certain kinds of study covering late 2015 and the first half of risk (e.g., poor quality) than others (e.g., safety 2016 in which we scraped listings from all US concerns). Metropolitan Statistical Areas with populations While there has been considerable debate of at least 500,000, a dataset which includes 319 about the sharing economy in the popular press, cities and more than 200,000 listings (Cansoy there are relatively few published academic and Schor, 2016). We find that approximately articles about this sector. As noted above, there 40% of all hosts have two or more listings and are a number of studies on the quality of rat- they comprise 37% of all entire home/apart- ings and reputational data. There are papers on ment listings. However, we find relatively few the motives and experiences of users, includ- large listers. In our data, 14% have five or more ing Airbnb users (Ikkala and Lampinen, 2015; listings and their properties comprise 16% of Lampinen and Cheshire, 2016; Lampinen et al., entire home/apartment lists. For 10+ properties, 2015; Möhlmann, 2015). There are a few unpub- the figures are 7.5% of listers and 10% of entire lished studies of racial discrimination on Airbnb home/apt properties. Thus, we conclude that (Cansoy and Schor, 2016; Edelman et al., 2016; while there is some movement toward turning Edelman and Luca, 2014). Cansoy and Schor Airbnb into a B2P platform, it is still largely a (2016) also look at educational attainment and P2P site. Among the providers we interviewed, find that Airbnb hosts are very highly educated. none offered more than one property for rent. There are also studies of other kinds of market A key difference between the P2P and B2P structures, such as B2P and non-profits which structures is that for the former, exchange consider a range of questions (Albinsson and occurs between unknown others, i.e., strangers. Yasanthi Perera, 2012; Bardhi and Eckhardt, Stranger exchange creates issues of incom- 2012; Bellotti et al., 2015; Dubois et al., 2014; plete information and, by extension, risk for Schor et al., 2016). A number of studies have would-be transactors. In conventional mar- involved platforms that do not use money, such ket transactions, brand reputation (in the B2P as Couchsurfing and HomeExchange (Forno context) or licensures (for professionals) are et al., 2013; Parigi et al., 2013; Parigi and State, used to reduce risk. The technological ana- 2014). Uber and Airbnb have funded their own logue in P2P economies is the crowdsourcing studies (Hall and Krueger, 2015) but they have of information from users in the form of ratings not made their data available to independent and reputational data. This data is believed to researchers. enhance the willingness of people to transact, by reducing the perceived risk of dealing with strangers. How much ratings and reputational Economic trends and the platforms data reduce true risk is as yet an unanswered A general question about the sector is whether question. There is a growing literature on the it is exacerbating or countering ongoing eco- quality of ratings and reputational data in nomic trends. In the US case, answering that online sites which suggests that current systems question requires attention to two factors:

268

the 2008 financial crisis and subsequent reces- such as Uber and TaskRabbit accounted for sion, and the trend toward extreme inequality. 0.5% of employment in 2015. A qualitative Airbnb and Uber were founded in 2008 and study of 1099 platform workers in New York 2009, respectively, and it is widely believed that City (Ravenelle, 2016) supports the idea that their success is due in part to the high unem- they labour under precarious conditions. ployment, indebtedness and difficult economic The second development is the growth situation that young people found themselves of extreme inequality of wealth and income in at that time. The 2009 recession was severe, (Piketty, 2014; Piketty and Saez, 2014; Saez and rivalled in the 20th century only by the 1930s Zucman, 2016). While social scientists have Depression. In the USA, GDP dropped more been writing about the growth of inequality for than 3%, measured unemployment rose to a decades, the Occupy Wall Street movement at high of 9.6%, and the employment-to-popula- the end of 2011 galvanized popular and politi- tion ratio fell to 54%, a drop from which it has cal attention to the growing mal-distribution largely not yet recovered (Council of Economic of income and wealth in the USA and else- Advisors, 2016). Youth, who have been the where. Occupy’s focus was on the concentra- innovators and first wave of users of these plat- tion of wealth at the very top. However, since forms, were especially hard hit, by a combina- the 1970s, the share of the top 20% has risen, tion of high unemployment and rising levels of at the expense of the bottom 80% (Mishel and education debt. Overall, youth unemployment Bivens, 2015). (defined as under age 25) after the crash rose to Our research suggests that the growth of 19.1%. High school graduates were the worst platforms since 2008 is contributing to an hit, but even among college graduates, the 9.9% intensification of the trend toward inequal- unemployment rate was far above previous ity, both as it relates to the 1–99% split and to experience (Davis et al., 2015). Elevated levels shifts within the broad middle class and work- of unemployment and underemployment have ing classes. The former effect is already widely persisted, even through the economic recov- recognized. Platform owners and their inves- ery. Among our respondents, we found quite a tors are appropriating large amounts of value few who were under- or unemployed. We also from users on both sides of the market. The co- found people were active on platforms in order founders of Airbnb became billionaires in 2015 to reduce their education-related debt. (Konrad and Mac, 2015) and Uber’s founder Recent studies support the idea of a precariat, is also likely to be in that exclusive group a term introduced originally by Guy Standing (Bertoni, 2014). Within the sharing community, (Standing, 2011). In the US labour market, par- the appropriation of wealth by founders and ticipants are increasingly likely to lack full-time venture capitalists has become a controversial employment and to be classified as independent issue (Schneider, 2014; Schor, 2014). contractors, or ‘1099 employees’ (a reference to The second effect, of increased inequal- the tax form that independent contractors are ity within the bottom 80%, has not yet been required to file). In the platform economy, most identified in the literature. On the basis of providers are classified as independent con- our research, we believe that platforms are tractors. A 2016 study of the rise of alternative increasing the incomes of the upper portion work arrangements (Katz and Krueger, 2016) of the bottom 80% of the income distribu- found that between 2005 and 2015 the fraction tion in two ways. The first is that well-off and of the labour force in non-standard work rose highly educated providers are using the plat- from 10.1 to 15.8%, and that non-standard work forms to increase their earnings. The second is accounted for the entire net gain in employ- that this group is doing work that is tradition- ment over this period. Online intermediaries ally done by people of low educational status.

269

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

White collar providers are engaging in blue and platforms to earn money. With a few excep- pink collar manual labour, in a ‘crowding-out’ tions, they all had at least college degrees, and effect. In 2014, the top income at the top of some had advanced degrees such as Masters or the fourth quintile (80%) was $112,262 (http:// law degrees. On TaskRabbit, we found highly www.taxpolicycenter.org/statistics/household- trained and accomplished people who had lost income-quintiles). We have only two provid- jobs in the technology sector, including one ers in our sample who earned more than that software engineer who had previously been in the year before we interviewed them. Only making $200,000 a year. Another TaskRabbit one averaged above that level in the previous provider was formerly in publishing. Airbnb and 5 years. Therefore, we believe that the fourth RelayRides had fewer unemployed providers; quintile is the group within which we are see- however, one unemployed former corporate ing this inequality-enhancing effect. We use the manager was managing a friend’s apartment as term ‘believe’ to describe our finding because an Airbnb rental and taking a cut of the earn- this is not a question that can be settled with ings, in addition to being active on other sites. qualitative data. Furthermore, it would be dif- One of the unemployed software engineers we ficult to show this effect with quantitative data interviewed explained: ‘…the economy’s just at the moment because the sector is so small. really tough right now…TaskRabbit adds a However, our findings point strongly in the little liquidity in an otherwise very thick situa- direction of an ‘inequality-enhancement’ effect. tion’. A number of the people who were active We turn now to discuss them. on RelayRides were also using the platform to pay for rent and basic needs. These included a recent college graduate who had not found Earning on the platforms a decent-paying job and an underemployed It’s, like, almost too good to be true (Shira, musician who typically had little work over Airbnb host). the summer. Although RelayRides was the least lucrative of the platforms and the earn- As noted above, the sites we are studying ings were low, these providers found the extra emerged after the 2008–2009 economic col- money to be essential. lapse, and they became a desirable option for For some of our respondents, student debt people who lost jobs or income in the crash, as was the spur to activity on the platforms, par- well as for recent college graduates who could ticularly Airbnb hosting. A number of the not break into the job market. Aiden was a younger hosts we interviewed used their plat- graduate with a 3.8 GPA who found himself form earnings to reduce debts. One couple, who unable to find a job after college. He turned to had earned $11,000 on Airbnb, used the money TaskRabbit hoping to earn some skills, make to pay off the husband’s college loans. Another, contacts and get a foot in the labour market. who also rented out a room in their apartment, Other TaskRabbit providers were also recent was using the money for the same purpose, pre- college graduates who were unable to find ferring Airbnb hosting to getting a permanent steady employment, and were piecing together roommate. different types of work. A number of these However, unemployment or precarity was graduates came from prestigious liberal arts not the motive for most of our sample. The colleges in New England, with experience majority of the providers we spoke with were and degrees that would have yielded full-time doing well economically, and for them, the employment in most years. We also interviewed appeal was to earn money to add to their full- a number of people who had lost jobs or hours time incomes. While we did have respondents of work during the crash and turned to the whose incomes left them barely able to meet

270

basic expenses such as rent and food, others using the platform to supplement their incomes. earned $100,000 a year. Among our sample About half were in lucrative professions (law- we have a lawyer, a political operative, man- yer, biotech scientist, accountant). Another five agement consultants, technology professionals, had part-time jobs and added TaskRabbit into medical researchers, teachers, an accountant, a the mix. Six reported no other type of employ- college teacher and a sales representative, and ment, although for most of them TaskRabbit other professionals. Many of the high earners seemed to be transitional—one person was were on the platforms because they saw a new between jobs, another had lost a job as a soft- way to earn money, although there was also a ware designer. Of the six who were working on group of high earners on Airbnb who reported the platform full-time, only one seemed to be doing it because they enjoyed the sociability. trying to build a career there. The woman dis- Respondents often described this economic cussed above, who wanted to be ‘productive’, opportunity as something novel, unlike other was an MIT graduate working in the life sci- activities they are involved in. This is one rea- ences who cleaned houses on the platform. son we think the platforms are resulting in Of the three platforms, Airbnb offers the increased inequality: they are adding to the highest earnings, by a wide margin. We asked incomes of high earners rather than just substi- providers to estimate their total earnings since tuting for prior kinds of off-platform earnings. they began activity on the platform. Median Participants also did not typically discuss other Airbnb earnings were $9,000, mean earnings ways they earn money outside of their jobs. The were $11,264. In our sample we have two indi- platforms, especially Airbnb, have emerged as viduals who had earned more than $30,000 on an easy new way to earn, using assets that peo- the platform—by renting out a single property. ple already possess. Shira, a single young woman whose family has In terms of working hours, we cannot say gotten into the Airbnb business, reported that with certainty how the platforms are changing she was expecting to earn $30,000 just in the the overall distribution. Because most platform year we interviewed her. She explained that providers already have full-time work, we think renting whole apartments on Airbnb yielded it most likely platform activity is intensifying between three and four times the income of a longstanding trend toward a more bi-modal ordinary renting. Indeed, the site had been so distribution of hours, in which a declining lucrative for her family that she was a bit sus- majority has rising hours and an growing picious: ‘Something’s going to happen, I know minority is underemployed (Schor, 1992). For that. Because it’s, like, almost too good to be example, one TaskRabbit provider explained true’. Other Airbnb hosts who vacated their that she liked the work because it gave her own homes to rent them were also able to earn something to do outside her regular full-time significant sums. One management consultant job, thereby allowing her to be ‘productive’ with reported charging about $350 a night for his her time. However, because the platforms also centrally located luxury apartment, and had offer opportunities to the unemployed and the already earned $34,000. under-employed, they also have an opposite, On RelayRides, earnings were much lower, equalizing effect on hours. with mean and median earnings at $600 and There is also the issue of the kind of work $643, respectively. Only two owners reported that is being done, which in the TaskRabbit case more than $1000 in total revenue. Economically, often involved high status professionals doing this group was probably the most diverse, as low status work. Six of the nineteen TaskRabbit it included some people with near-poverty providers we interviewed were people with incomes and others with $100,000+ a year full-time jobs or their own businesses who were salaries. Not surprisingly, their attitudes and

271

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

specific economic motives varied. For a number We asked providers how their earnings on of them, having their cars sit unused is an irri- the platforms compared to their full-time jobs tation, because they have monthly bills associ- (or if they did not have a full-time job, at other ated with the vehicles. For others, the site made relevant paid employment). On Airbnb, nearly it financially viable to buy the vehicle. Another 60% reported earning more on the platform, group was just pleased to be able to pay off about a third earned less, and fewer than 10% their car loans or expenses with this incremen- reported earning the same. On RelayRides, only tal income stream. As noted above, we also had 30% earned more and 60% earned less. On a few who were just scraping by financially. TaskRabbit, only 20% earned more on the plat- Among the TaskRabbits, median earnings form, nearly half earned less, and a third earned were $2500, and mean earnings were $6819. the same. This was also a diverse group in terms of their To summarize, we find that most providers situations and how they used the platform. In on the platform are highly educated. A large 2013, the company estimated that 10% of pro- majority are supplementing their incomes with viders were using the platform for full-time platform activity, thereby boosting their incomes work (Newton, 2013). As noted, our rate was relative to non-participants. The platforms considerably higher than that. For some, the appear to be a new source of income that are not flexibility provided by the platform was the replacing prior supplemental earnings. Among biggest draw, as they were either starting busi- the TaskRabbit providers, some are highly edu- nesses or had family obligations. Some were cated unemployed or under-employed who highly enterprising types, who preferred not to would probably have been earning less if the spend their free time unproductively. Overall, platforms had not been available. the hourly wages on this platform compared favourably to other market opportunities. We found that even people who only had com- Blue and pink collar work for white monly available skills (such as cleaning, driving, collar providers putting together Ikea furniture or doing prod- It’s manual labour in person (Jed, TaskRabbit). uct testing) were able to make at least twice the minimum wage, and many were able to get The second way in which platform labour is a wage of $20–25 per hour or more. However, inequality-enhancing is that highly educated, few had worked enough hours to make signifi- white collar providers are doing manual work cant sums—only two reported total earnings of that has traditionally been done by people $10,000 or more, and most had earned less than without college degrees. We also find that $5,000. We also found that a number of provid- most of these providers are racially ‘white’ and ers were using the platform entrepreneurially. native-born, in contrast to the people of colour One man began getting transcription jobs and and immigrants who disproportionately do this outsourced them at lower wages to people off manual work in the conventional economy. We the platform. (However, in a follow-up inter- begin with the educational credentials of our view two years later he reported that he had sample and then move on to discuss the kinds gotten into trouble by taking on more jobs than of work they are doing on the platforms. he could handle and was no longer active on Among our 43 providers the lowest educa- the platform.) A few had started online busi- tion level is ‘some college’. Only four are in this nesses, as personal assistants and digital work- category, and they are all on TaskRabbit. For ers. However, as we note in the next section, three of the four this reflects not a final educa- much of the work on TaskRabbit was manual tional level, but the fact that they are currently labour. either in college or doing college courses to

272

complete their educations. The fourth person is ours is also likely relevant, a topic we return a software developer who did not complete col- to below. lege but founded a company and has significant Turning to the kinds of work providers are technical training. Twenty-two of our inform- doing, we find that a good deal of it is low- ants have a college degree, sixteen have an MA skilled and blue and pink collar. On Airbnb, degree and one has a Juris Doctor (JD), a law there is a mix of tasks involved in hosting. There degree. Overall, this is a highly educated group. is the labour of making the initial arrange- They are also from highly educated families. ments and afterwards the generally more time- Only three come from homes where neither consuming task of cleaning the rooms and parent has a college degree. Sixteen have at least apartments and making them ready for the one parent with a degree beyond the BA/BS, next guests. In hotels and motels that work is that is, either a Masters or an MD/JD/PhD. Five done by desk clerks and chambermaids. Few of have two parents with Bachelors degrees and our Airbnb providers mentioned using clean- seven come from parentage with two advanced ing services or domestic labourers to do room degrees. In terms of the social class they identify preparation work. Quite a few discussed doing with, six called themselves upper middle class, it themselves, explaining how they handle the seventeen said they were middle class and ten cleaning. This was even the case for some of the said lower middle class. Nearly all the provid- hosts with degrees from Ivy League or prestig- ers who reported being lower middle class were ious schools, who had high paying professional TaskRabbit providers. The high levels of edu- jobs. On RelayRides, the work involved is mini- cation we found in our TaskRabbit sample are mal, mainly the handoff of the keys (when that typical of that platform nationally. In 2013, the is done in person), keeping the car clean and company reported that 70% of their workforce parking and servicing the car. held a Bachelor’s degree, 20% had a Master’s On TaskRabbit we see a fuller range of types Degree and 5% had a PhD (Newton, 2013). of labour. While some providers were engaged Among the 19 TaskRabbit providers that we in white collar or online labour, much of it was interviewed, seven had completed Bachelor’s low skill. One person discussed a task where degrees and five had graduate degrees. Four she was asked to find a type of sunglasses for had ‘some college’, as noted above. a certain price. App testing is a frequent task. It is worth noting that our sample and some Beth, who has an MA, said that often her tasks of our findings differ from those of Ravenelle were ‘mindless work’. Some providers were (2017), who, in 2015, interviewed 87 provid- hired as staff at ‘events’, working registration ers on four platforms (Airbnb, TaskRabbit, desks or dressing up in costume. Kitchensurfing and Uber). Ravenelle’s sam- However, the more commonly discussed ple is considerably less educated (42% college tasks were what Jed described as ‘manual graduates) and less white (58.5%) than ours. As labour in person’. Common examples include we discuss below, she has some divergent find- house-cleaning, driving, moving, putting ings, for example on the extent to which provid- together Ikea furniture and office organizing. ers feel that the work is stigmatizing. We believe Valeria, an immigrant and a student who does a there are three reasons for the differences. One lot of cleaning on the platform, explained that is the difference in platforms: drivers and cooks doing this kind of work has been a challenge: are less likely to be college educated than providers on the first two platforms. The second is Task Rabbit has also been a journey to learn that Boston has a more favourable labour mar- new skills, to develop new things that were ket than New York City. And finally, the fact not there before I started…In the beginning that Ravenelle did her research two years after I sucked at cleaning. I sucked. People were

273

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

leaving bad reviews, like, ‘Oh, she’s okay. doing cleaning. Handyman work is male. On She’s not awesome’. Because back at home RelayRides, an automotive site, respondents are I didn’t even make my bed, you know? There three-quarters male. For other, less gendered was a cleaning person in my home. types of work, such as lodging services, we find a gender mix. While we do not include these Drake, a former software engineer who lost his interviews in this article, our sample of Uber job, found himself doing a variety of manual and Lyft drivers is also largely male. However, jobs. He did handyman work, snow shovelling we do also see some erosion of gender segrega- and food delivery. He even discussed scrubbing tion. We had reports of women on TaskRabbit toilets. He described the work as ‘backbreaking’. who were putting together Ikea furniture. And The range of tasks he has been involved in also on the ridesourcing apps such as Uber and Lyft, suggest a ‘servant economy’, in which highly priv- while the labour force is predominantly male, ileged people use platforms to save themselves it seems to be less so than conventional taxi the trouble of doing simple things, like picking drivers. up food or drinks. Drake discussed tasks where How does the movement of white collar he would pick up supplies for students’ parties. workers into manual labour affect the dis- Others talked about being asked to buy groceries tribution of income? The first effect we have or make other kinds of deliveries, sit for pets, put already discussed: by disproportionately pro- cheap furniture together, act as personal assis- viding earning opportunities for people who tants and help people with parties. Moving is a are already well-educated and relatively well- popular activity that some of the men do a lot of. off, platforms increase inequality. The second Cleaning is the modal activity for our pathway is via reduced demand for the services TaskRabbit providers, especially for the women. of non-platform businesses, a crowding out or Some of the most professionally successful substitution effect. If Airbnb reduces demand women in our sample do a great deal of clean- for hotels that employ low-wage workers as ing. For Kate, who had a stable job as an admin- maids, food service and other manual workers, istrator at a prestigious local university, one task income will shift from those workers toward the turned into a long term cleaning arrangement. higher income platform providers. The same is Rachel, who was on the platform full-time, did a true of car rental companies who employ clerks lot of cleaning, including some residences which and cleaners. If people turn to TaskRabbit to were very filthy. The lawyer in our sample mostly have their homes cleaned it may reduce the does cleaning. Overall, we find that blue and pink demand for immigrant cleaners and others who collar labour, similar to what domestics and serv- have been in this market. One issue is whether ants do for wealthy patrons, comprises a large consumers prefer to contract with platforms portion of activities on these platforms. That this that do background checks and provide highly work is being done by highly educated profes- educated service workers more like them- sionals represents a departure from the past. selves. Of course, it is possible the platforms are Although our sample is too small to make increasing demand overall, which would miti- claims about subgroups, it may be worth raising gate the size of this labour substitution effect. the question of gender. While one might have One factor which will affect the extent to which expected that the emergence of a new institu- the platforms increase, rather than substitute tional setting attracting highly educated young for demand, is their relative prices. Airbnb has people would yield a less gendered distribution made travel less expensive and is likely to be of tasks, we find that the platform economy is increasing demand overall. TaskRabbit’s 2014 not radically different from the conventional price increase suggests it is less likely to expand labour market. Women are more likely to be total demand significantly.

274

De-stigmatizing, but up to a point who worked on TaskRabbit after graduating from law school explained: I don’t feel like I’m demeaning myself…It’s fine (Veronica, TaskRabbit). That was very, very humbling. That was actu- ally the one thing that would bother me The movement of high status people into low sometimes doing TaskRabbit. So I put in my status work begs an explanation. Why are our profile that I went to law school and every- providers willing to do tasks that would tradi- thing, because, like, you know, I wanted to tionally be considered demeaning or degrading look more credible. But, you know, people for people with their levels of education and sometimes that would hire me to come over accomplishments? A key part of the answer and clean, would almost make comments is that the platforms have been able to de- almost pitying me for having to clean their stigmatize the types of tasks and work they apartment, having gone to law school, and organize people to do. What we have found is I hated that....they would be, like, “Oh, it that technological novelty, the branding of the sucks you have to do this.” Like, “Yes, I know platforms and the demographic composition of it sucks. You don’t have to remind me.” early adopters are important parts of how that de-stigmatisation has occurred. For Veronica, who had an MA in a science From the beginning, the platforms presented field, as well as a full-time job, the work was themselves as technologically advanced, a new, mostly okay. ‘It doesn’t make me feed bad…I cool thing. That made them feel upscale rather don’t feel like I’m demeaning myself…It’s than downscale. Furthermore, the platforms fine. I try to pick stuff that’s like normal to do’. prominently espoused a common good rheto- However, she notes that she draws the line at ric that emphasized doing something beneficial some tasks: ‘I saw one that was “get me a latte for society—sharing—rather than just making from Starbucks and I’ll pay you $8’”…Like no, money (Schor, 2014). Quite a few of our par- get off your butt and get it yourself. Because ticipants explained their motives in these com- that’s lazy... I don’t want to be, like a servant’. mon good terms, especially on Airbnb, but also The sentiments of these providers are more in on the other two platforms. They were doing line with what Ravenelle (2017) reports, as she something green, building social connection, argues that some of her informants are embar- helping others or fostering cultural interchange. rassed about the work they are doing and hide Even our most money-oriented providers usu- their participation. None of our respondents ally appreciated some common good aspect of expressed that concern. the platform, and quite a few disavowed their interest in making money. (Most were credible in that disavowal, a few not.) This discourse has Can low-income providers prosper played an important role in de-stigmatisation, on the platforms? perhaps because people are willing to do a It takes money to make money (Kiran, wider variety of work in the service of an ideal Airbnb host). than they are just for money. Finally, the demo- graphics of users contributed to de-stigmatisa- While the platforms themselves argue that tion. Early users were white, young and highly they have operated as a cushion in bad eco- educated, on both sides of the market. That said, nomic times and are helping to spread wealth, there were moments in our interviews when the the story is more complicated. One issue is low-status nature of the labour or the inequal- whether low-income, less-educated people will ity of relations with the customers arose. Katy, be able to prosper on these platforms. To date,

275

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

there is little research to address this question. with highly valued skills, such as the person However, one working paper by economists who started the translation business. Fraiberger and Sundarajan (2015) is worth dis- The experiences of our unemployed provid- cussing, especially as Sundarajan has been a ers also makes us sceptical of the claim that low- prominent voice in the debate about the shar- income people will benefit disproportionately. ing economy. The paper argues that platforms For them, especially the few who are not recent will help the poor more than other groups. The college graduates, attempting to make a living on authors use a simulation technique calibrated the platform is very difficult. As one TaskRabbit, with data from Getaround (a RelayRides-type an older man who lost his job explained: platform) and conventional economic assumptions to predict that below-median income I mean like there are many times that you do users will benefit disproportionately, via both this and you think, I’d be way better off work- the opportunity to rent out vehicles and lower ing at McDonalds because I’d make the same cost rides. Their method does not study actual amount of money and I’d have free fries… outcomes, but simulations. Our qualitative anal- Working for TaskRabbit is just a fantastic ysis leads us to be sceptical of their assumptions way to always stay at the poverty level, right? and conclusions, as we believe they have missed But at least you can pay your phone bill and important aspects of this market, namely the you can buy some food and the landlord isn’t ways in which it is difficult for low-income peo- upset with you. ple to benefit from these platforms. On RelayRides, we find that relatively few In our 2015 interviews with TaskRabbit provid- low-income car owners are participating and ers they report more frustration with the plat- the few who are have high educational attain- form, and feel the company is more concerned ment. Their cars are old, as is common among with customer than worker satisfaction. Overall, low-income car owners. This means their daily we are sceptical of the idea that the sharing rental rates are low, so they do not earn much. In economy will disproportionately aid economi- addition, unlike many low-income people, our cally and educationally disadvantaged providers. providers live in middle class neighbourhoods. Location is important in this market—cars which are sited in low income or poor neigh- Conclusion bourhoods are likely to receive fewer requests While the sharing economy has raised many because consumers are more wary of those questions, this article highlights one effect that locations. Unlike with ridesourcing, where driv- has not yet been identified: how participation ers are not confined to the areas around their in for-profit sharing platforms may be influenc- own residences, with peer-to-peer car rentals, ing the distribution of opportunity and income the cars are parked near the owners’ homes. within the bottom 80% of the population. We Similarly, with Airbnb, earning requires com- find that sharing economy participants are ing to the market with valuable assets. As noted highly educated, often professionals, and that above, some hosts are earning $20,000–35,000 a they are using the platforms to increase their year from a single property. But achieving that earnings. We believe that their activity is crowd- level of revenue requires either owning a nice ing out, at least to some extent, less advantaged, home or apartment or having enough earn- lower educational attainment workers who ing power to obtain expensive leases. It also have traditionally done much of manual work requires access to alternative living quarters that more privileged sharing providers are now while their places are rented. On TaskRabbit doing. In one sense this is not surprising. At some of the most successful earners were those times when employment and income are scarce,

276

standard economic reasoning expects a cascade platforms will be and in which sectors. Many effect in labour markets, as more educated peo- aspects of their technology can be (and are ple take jobs and opportunities that they would being) adopted by legacy businesses, includ- not accept in better times. We believe the coming the convenient payment systems. However, mon-good discourse of the sharing economy as a counterpoint to observers who think this reduced cultural barriers that might impede model will eliminate conventional employ- this downward slide, and lengthened the sta- ment (Sundarajan, 2016), it is worth remember- tus distance that middle class whites have been ing that the most successful platform (Uber) willing to travel for opportunity. Indeed, as we entered a highly regulated, dysfunctional indus- argue, platform providers are now doing some try with huge economic rents. Labour platforms of the least desirable urban work—cleaning have been much less successful, with many and moving. going defunct or (as with TaskRabbit) repeat- Overall, the providers we interviewed edly changing their business model. expressed strong feelings of satisfaction. In the controversies about the sharing econ- However, whether this attitude will endure is omy, most discussions of inequality, power and an important question. It is possible that con- adverse outcomes have focused on the creation ditions are changing as the platforms expand of sharing economy billionaires, the exploitation and attract a less educated and more exploit- of labour, regulation and taxation, and ecologi- able group of providers. Ravenelle (2017), who cal and social impacts. Our findings suggest that did her research two years after ours, paints a there is another important issue to study: how far more pessimistic picture. How much this is a relatively more privileged middle class has attributable to the aforementioned growth, the used this technological innovation to expand harsher economic environment of New York opportunities for itself. Occasionally, this issue City, the difference in platforms she studied, or has been raised under the guise of ‘access’, with the demographic differences between her sam- concern about the relative whiteness and afflu- ple and ours, we cannot presently determine. ence of the user base (Schor, 2014). However, However, a comparison of her results with the ways in which ‘accessing’ the platforms is ours also suggests an important insight which affecting larger trends in income distribution, has not been sufficiently recognized in the lit- employment and work has not yet been rec- erature: the sharing economy cannot be sepa- ognized. We hope this article will begin that rated from the labour market context in which conversation. it operates. While most discussion of the sector has considered it in isolation, platforms’ ability Endnote to attract providers will depend significantly on alternative labour market opportunities. 1 Airbnb’s aggressiveness in trying to stop research- Many platforms launched during the period ers from finding informants on the platform is par- when financial collapse and recession domi- ticularly frustrating, given its size and importance. nated local labour markets, which undoubtedly While there are a few researchers who have been increased their available labour pool. Should granted access to their data, it is our understanding that they are required to sign agreements that give labour markets continue to tighten, they may Airbnb the right to prevent publication of results. have to improve earnings and terms of con- tracts to assure a robust provider base. Of course, if platforms significantly displace Acknowledgements legacy businesses, they will have more influence This article was produced as part of the Connected over the labour markets in which they oper- Consumption and Connected Economy project of ate. It is impossible to predict how successful the Connected Learning Research Network of the

277

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018 Schor

MacArthur Foundation, and has been generously sup- Dubois, E., Schor, J., and Carfagna, L. (2014) New ported by the foundation. I would like to thank the PhD cultures of connection in a Boston time bank. students who worked on this research: Luka Carfagna, In J. Schor and C. Thompson (eds.) Sustainable William Attwood-Charles, Robert Wengronowitz, Lifestyles and The Quest for Plentitude: Case Connor Fitzmaurice, Mehmet Cansoy, Isak Ladegaard, Studies of the New Economy, pp. 95–124. New Haven, CT: Yale University Press. Emilie Dubois and Taylor Cain, and undergradu- Edelman, B. G., Luca, M., and Svirsky, D. (2016) ates Alison Wawrzynek, Nathan Schwan, Leá Oriol, Racial Discrimination in the Sharing Economy: Elizabeth Jiang, Hahnsol Kang and Carolyn Ruh. Evidence from a Field Experiment (Research Paper Series). Harvard Business School Negotiation, References Organizations and Markets Unit; Social Science Research Network. Albinsson, P. A. and Yasanthi Perera, B. (2012) Edelman, B. and Luca, M. (2014) Digital discrimina- Alternative marketplaces in the 21st century: tion: the case of Airbnb.com. Harvard Business building community through sharing events: alter- School Working Paper, (14–054). Available online native marketplaces in the 21st century, Journal of at: http://www.west-info.eu/files/airbnb_research. Consumer Behaviour, 11: 303–315. pdf Bardhi, F. and Eckhardt, G. M. (2012) Access-based Forno, F., Garibaldi, R., Scandella, E., and consumption: the case of car sharing, Journal of Polini, I. (2013) My house is yours: a world- Consumer Research, 39: 881–898. wide study on home exchangers’ profiles and Bellotti, V., Ambard, A., Turner, D., Gossmann, C., motivations. University of Bergamo. Available Demkova, K., and Carroll, J. M. (2015) A Muddle online at: https://www.academia.edu/4020708/ of Models of Motivation for Using Peer-to-Peer My_House_is_Yours._A_Worldwide_Study_ Economy Systems, pp. 1085–1094. New York: on_Home_Exchangers_profile_and_motiva- ACM Press. tions_with_Roberta_Garibaldi_ [Accessed 29 Benkler, Y. (2006) The Wealth of Networks: How December 2016]. Social Production Transforms Markets and Fraiberger, S. and Sundararajan, A. (2015) Peer-to- Freedom. New Haven, CT: Yale University Press. peer rental markets in the sharing economy. New Bertoni, S. (2014) Uber CEO Kalanick Likely a bil- York University. Unpublished paper. lionaire after $18.2 billion valuation. Forbes, 6 June. Hacker, J. S. (2006) The Great Risk Shift. New York: Botsman, R. and Rogers, R. (2010) What’s Mine is Oxford. Yours: the Rise of Collaborative Consumption. Hall, J. V. and Krueger, A. B. (2015) An Analysis of New York: Harper Business. the Labour Market for Uber’s Driver-Partners Botsman, R. (2015) Defining the Sharing Economy: in the United States. Available online at: https:// What is Collaborative Consumption—And s3.amazonaws.com/uber-static/comms/PDF/ What Isn’t? Available online at: https://www.aca- Uber_Driver-Partners_Hall_Kreuger_2015.pdf demia.edu/4020708/My_House_is_Yours._A_ Henwood, D. (2015) What the sharing economy Worldwide_Study_on_Home_Exchangers_pro- takes. The Nation. 27 January. file_and_motivations_with_Roberta_Garibaldi_ Ikkala, T. and Lampinen, A. (2015) Monetizing net- [Accessed 29 December 2016]. work hospitality: hospitality and sociability in the Cansoy, M. and Schor, J. B. (2016) Who Gets to Share context of Airbnb (pp. 1033–1044). Presented in the “Sharing Economy”: Understanding the at the CSCW’15 Proceedings of the ACM 2015 Patterns of Participation and Exchange in Airbnb. conference on Computer supported cooperative Boston College, Unpublished Paper. work. New York: ACM. Council of Economic Advisors. (2016) Economic Kalamar, A. (2013) Sharewashing is the new Report of the President. Available online at: greenwashing. Available online at: http://www. https://www.whitehouse.gov/sites/default/files/ opednews.com/articles/Sharewashing-is-the- docs/ERP_2016_Book_Complete%20JA.pdf New-Gr-by-Anthony-Kalamar-130513-834.html Davis, A., Kimball, W., and Gould, E. (2015) The class [Accessed 29 December 2016]. of 2015: despite an improving economy, young Katz, L. F. and Krueger, A. B. (2016) The Rise and grads still face an uphill climb. Economic Policy Nature of Alternative Work Arrangements in the Institute Briefing Paper #401. Available online at: United States, 1995–2015. Available online at: http:// http://www.epi.org/files/2015/the-class-of-2015-re- krueger.princeton.edu/sites/default/files/akrueger/ vised.pdf files/katz_krueger_cws_-_march_29_20165.pdf

278

Konrad, A. and Mac, R. (2014) Airbnb cofounders Reich, R. B. (2015) The Share-the-Scraps Economy. to become first sharing economy billionaires as Available online at: http://robertreich.org/post/ company nears $10 billion valuation. Forbes, 20 109894095095 March. Rifkin, J. (2014) The Zero Marginal Cost Society: the Lampinen, A. and Cheshire, C. (2016) Hosting via Internet of Things, the Collaborative Commons and Airbnb: motivations and financial assurances in mon- the Eclipse of Capitalism. New York: St. Martin’s etized network hospitality. Presented at the CHI’16 Press. Proceedings of the annual conference on Human Saez, E. and Zucman, G. (2016) Wealth inequality factors in computing systems. New York: ACM. in the United States since 1913: evidence from Lampinen, A., Huotari, K., and Cheshire, C. (2015) capitalized income tax data, Quarterly Journal of Challenges to participation in the sharing econ- Economics, 131: 519–578. omy: the case of local online peer-to-peer exchange Schneider, N. (2014, December 21) Owning is the in a single parents’ network. Interaction Design New Sharing. Available online at: http://www. and Architecture(s), Special Issue on Peer-to-Peer shareable.net/blog/owning-is-the-new-sharing Exchange and the Sharing Economy, 16–32. Scholz, T. (2016) Uberworked and Underpaid: How Mishel, L. and Bivens, J. (2015) The State of Working Workers are Disrupting the Digital Economy. America. Cornell, NY: ILR Press. Cambridge: Polity Press. Möhlmann, M. (2015) Collaborative consumption: Schor, J. (2014) Debating the sharing economy. Great determinants of satisfaction and the likelihood of Transition Initiative. Available online at: greattransi- using a sharing economy option again, Journal of tion.org/publication/debating-the-sharing-economy Consumer Behaviour, 14: 193–207. Schor, J. B. (1992) The Overworked American: the Morozov, E. (2013) The “sharing economy” under- Unexpected Decline of Leisure. New York: Basic mines workers’ rights. Financial Times, 14 October. Books. Myers, C. (2015) Decoding Uber’s proposed $50B Schor, J. B. and Attwood-Charles, W. (2017) The valuation (And What It Means For You). Forbes, Sharing Economy: labour, inequality and socia- 13 May. bility on for-profit platforms, Sociology Compass. Newton, C. (2013) Temping fate: can TaskRabbit go forthcoming. from side gigs to real jobs? The Verge, 23 May. Schor, J. B. and Fitzmaurice, C. J. (2015) Collaborating Overgoor, J., Wulczyn, E., and Potts, C. (2012) Trust and connecting: the emergence of the sharing econ- Propagation with Mixed-Effects Models. Stanford omy. In L. Reisch and J. Thogersen (eds.) Handbook University, Unpublished paper. Available online of Research on Sustainable Consumption, pp. 410– at: http://web.stanford.edu/~cgpotts/papers/ 425. Cheltenham: Edward Elgar. OvergoorWulczynPotts.pdf Slee, T. (2015) What’s Yours is Mine: Against the Parigi, P., State, B., Dakhlallah, D., Corten, R., Sharing Economy. New York: OR Books. and Cook, K. (2013) A community of strangers: Standing, G. (2011) The Precariat: the New Dangerous the dis-embedding of social ties, PLoS One, 8: Class. New York: Bloomsbury. e67388. Sundararajan, A. (2016) The Sharing Economy: the Parigi, P., and State, B. (2014) Disenchanting the End of Employment and the Rise of Crowd-Based world: the impact of technology on relation- Capitalism. Cambridge, MA: MIT Press. ships. In L.M. Aiello and D. McFarland (Eds.) Telles, R. Jr. (2016) Digital Matching Firms: A New Social Informatics, LNCS 8851, pp. 166–182, 2014. Definition in the “Sharing Economy” Space, ESA Available online at: http://link.springer.com/chap- Issue Brief #01-16, U.S. Department of Commerce, ter/10.1007/978-3-319-13734-6_12 [Accessed on 29 Economics and Statistics Administration, Office of December 2016]. the Chief Economist, Washington, D.C. Available Piketty, T. (2014) Capital in the 21st Century. online at: http://www.esa.gov/sites/default/files/dig- Cambridge, MA: Belknap Press. ital-matching-firms-new-definition-sharing-econ- Piketty, T. and Saez, E. (2014) Inequality in the long omy-space.pdf, [Accessed on 29 December 2016]. run, Science, 344: 838–843. Zervas, G., Proserpio, D., and Byers, J. (2015) A first Ravenelle, A. (2017) Sharing economy workers: sell- look at online reputation on Airbnb, where every ing, not sharing, Cambridge Journal of Regions, stay is above average. Social Science Research Economy and Society, 10: 279–293. Network. January 28.

279

Downloaded from https://academic.oup.com/cjres/article-abstract/10/2/263/2982086 by Curtin University Library user on 06 March 2018