Ridesourcing, the sharing economy, and the future of cities

Cities xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Cities journal homepage: www.elsevier.com/locate/cities

Ridesourcing, the sharing economy, and the future of cities ⁎

Scarlett T. Jina, , Hui Konga, Rachel Wub, Daniel Z. Suia,c a b c

Department of Geography, The Ohio State University, 1036 Derby Hall, 154 North Oval Mall, Columbus, OH 43210-1361, USA Department of Economics, Yale University, New Haven, USA Division of Social and Economic Sciences, The U.S. National Science Foundation

A R T I C L E I N F O

A B S T R A C T

Keywords: Ridesourcing Shared mobility The sharing economy On-demand work Future cities

As an integral part of the emerging sharing economy, ridesourcing refers to transportation services that connect community drivers with passengers via mobile devices and applications. The spectacular growth of ridesourcing has sparked a burgeoning literature discussing how it affects the future of cities. This paper presents a systematic review of the existing literature concerning the impact of ridesourcing on the efficiency, equity, and sustainability of urban development. Ridesourcing has a positive impact on economic efficiency. It both complements and competes with public transit, but its influence on traffic congestions near city centers is still unclear. Regarding urban equity, ridesourcing further amplifies the issue of the digital divide and raises concerns over the issues of discrimination and data privacy and security. It is also hotly contested whether prosumers (producers/ consumers) are exploited by the sharing economy platforms, whether ridesourcing drivers are reasonably compensated, and how to better protect on-demand workers' rights. Even though ridesourcing has been promoting a green image, its true environmental impact has not been thoroughly investigated. According to the evidence reported in the literature so far, it is unlikely that ridesourcing will reduce private car ownership. Ridesourcing's impacts on energy consumption and greenhouse gas emissions are uncertain based on existing research. This paper outlines the danger of conceptual confusion and the methodological issues in the existing literature. Further research is sorely needed as the future of cities is indisputably tied to the sharing economy and its impacts on shared mobility.

1. Introduction The explosive growth of ridesourcing has sparked heated debate over its impacts on future urban development (Rayle, Dai, Chan, et al., 2016). Ridesourcing refers to transportation services that connect community drivers – people who drive private cars instead of commercial vehicles - with passengers via mobile devices and applications. There are many different terms used to describe this emerging transportation option. Academic transportation researchers use ‘ridesourcing’; practitioners describe themselves as ‘transportation network companies (TNCs) or mobility service providers (MSPs)’; the popular press calls it ‘ride-sharing’ or ‘ride-hailing’ (Shaheen & Chan, 2016). Ridesourcing services have been expanding rapidly across the world, with a number of successful TNCs, such as Uber and Lyft in the U.S., Didi Express in China, Ola in India, and even UberMOTO (for motorcycle rides) in Thailand. Ridesourcing is a particular form of shared mobility. Shaheen and Chan (2016) offer a classification of shared mobility based upon what is being shared (Fig. 1). Carsharing, motorcycle sharing, scooter sharing,

⁎

and bikesharing facilitate the sharing of a vehicle; whereas ridesharing, on-demand ride services, and microtransit enable the sharing of passenger rides. Ridesharing, including carpooling and vanpooling, allows drivers and passengers with similar origins and destinations to share a ride. Traditional ridesharing on a personal, organizational, or ad hoc basis has existed for decades. There are also mobile apps based ridesharing, such as BlaBlaCar in France and Didi Hitch in China. On-demand ride services involve the adoption of mobile technology to request and dispatch vehicles upon passengers' requests. There are three major forms of on-demand ride services. Ridesourcing is the largest segment in on-demand ride services. Ridesplitting is a variation of ridesourcing, which allows passengers with a similar route to share a ride and split the fare. Many TNCs operate both ridesourcing and ridesplitting. For instance, Uber has UberPOOL and Lyft has Lyft Line. E-hail is taxi service equipped with mobile apps. It is different from ridesourcing as the vehicles are taxicabs instead of private vehicles. Lastly, microtransit is a form of private transit enabled by mobile technology. It mainly provides commuting services that connect residential areas with urban and suburban working and commercial areas, such as

Corresponding author at: Department of Geography, The Ohio State University, 1036 Derby Hall, 154 North Oval Mall, Columbus, OH 43210, USA. E-mail address: [email protected] (S.T. Jin).

https://doi.org/10.1016/j.cities.2018.01.012 Received 19 October 2017; Received in revised form 9 January 2018; Accepted 10 January 2018 0264-2751/ © 2018 Elsevier Ltd. All rights reserved.

Please cite this article as: Jin, S.T., Cities (2018), https://doi.org/10.1016/j.cities.2018.01.012

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

paper contributes to a better and clear understanding of ridesourcing, as a first step to comprehend the convoluted relationship between the sharing economy and urban development. More specifically, this paper will conduct a systematic review of the existing literature concerning how ridesourcing affects the efficiency, equity, and sustainability of urban development – three key goals that urban scholars and policy makers aspire to achieve for our future cities. In addition to academic literature, this article also reviews a variety of materials such as news and magazine articles, blogs, websites, and reports produced by non-academic sources, considering that research on the sharing economy is in its infancy, and not many studies have been done specifically about ridesourcing. Given that Uber is the most successful TNC, it is used as an example by most literature analyzed in this paper. It is worth mentioning that, even though we didn't intend to limit our review to any specific country, the majority of the literature written in English focused exclusively on the U.S., which in turn constrained the scope of our review. That said, the goal of this paper is not to draw a broad conclusion on how ridesourcing may impact urban development, but rather to synthesize current understanding, identify problems, and point out directions for future research. The rest of this paper is organized as follows. Section 2 discusses how ridesourcing affects urban transportation efficiency and economic efficiency. Section 3 examines the impact of ridesourcing on urban equity. Section 4 analyzes how ridesourcing influences urban sustainability through altering private car ownership, energy consumption, and emissions. This is followed by a discussion section that outlines problems in existing research and prospects of future urban development influenced by ridesourcing and the sharing economy. The last section offers a summary and points out future research directions.

Carsharing

Motorcycle Sharing Sharing a Vehicle Scooter Sharing

Bikesharing Carpooling

Shared Mobility Ridesharing

Vanpooling

Ridesourcing

Sharing a Passenger Ride On-Demand Ride Services

Ridesplitting

Microtransit

E-Hail

Fig. 1. Categories of shared mobility. [Source: modified from Shaheen & Chan, 2016]

Chariot in the U.S. Shared mobility is a fast-growing sector of the emerging sharing economy. The sharing economy has a number of different names such as on-demand economy, gig economy, collaborative consumption, and collaborative economy. This paper uses “the sharing economy” to describe the economic activities and “on-demand work” to describe the type of jobs provided by the sharing economy. There have been an increasing number of studies discussing the social, economic, and environmental impact of the sharing economy (e.g., Cockayne, 2016; Frenken & Schor, 2017; Martin, 2016). Undoubtedly, the sharing economy is quickly becoming an important component of the economy, penetrating a growing number of economic activities (such as Airbnb, TaskRabbit etc.) in both developing and developed countries. The sharing economy is transforming business operations and models in many profound ways. With accelerated urbanization in the developing world, more and more people will be living in our crowded cities worldwide. How to make our cities more efficient, equitable, and sustainable has been the focus of interdisciplinary research during the past half century (e.g., Graham & Marvin, 2001; Vasconcellos, 2001). The rapidly growing sharing economy will inevitably exert its due impacts on the future of cities, and yet we have seen little work linking the sharing economy explicitly with urban studies, although most of the sharing economic activities are implicitly taking place in cities. More importantly, most studies examine the sharing economy in general, even though different types of sharing economy platforms or shared modes of transportation exhibit distinct characteristics and impose diverse impacts on cities. By focusing on one aspect of the sharing economy – ridesourcing – this

2. Urban efficiency To assess how ridesourcing changes urban transportation efficiency and economic efficiency compared to the status quo, it is essential to identify what mode of transportation ridesourcing is replacing – known as the modal shift (Light, 2017). Ridesourcing is most conspicuously competing with taxis. From March 2012 to July 2014, the number of taxi trips per month in San Francisco decreased by more than a half (San Francisco Municipal Transportation Agency, 2014). In Manhattan of New York City between April to June in 2015, taxi pickups dropped by 3.83 million compared to the same period in 2014 (Fischer-Baum & Bialik, 2015). Many scholars and the popular press made comparative studies on taxi and ridesourcing (Anderson, 2014; Bialik, Flowers, Fischer-Baum, et al., 2015; Glöss, McGregor, & Brown, 2016), all of which seemed to confirm this significant modal shift. Another major modal shift caused by the emergence of ridesourcing is from public transit (Rayle et al., 2016), even though there are also situations in which ridesourcing complements public transportation networks (American Public Transportation Association, 2016). In this section, we will examine these two modal shifts in detail and discuss how they

Table 1 Uber's impact on urban efficiency. [Source: Compiled by the authors] Transportation efficiency

Economic efficiency

Congestion

Accessibility

Safety

Ridesourcing vs. taxi

(?) Reduces or increases congestion in the city center?

(+) Reaches poor neighborhoods with insufficient taxi services

(+) Drivers and passengers feel safer than when in taxis (−) Insufficient driver training and insurance gap

Ridesourcing vs. public transit

(−) High-density area: competes with public transit

(+) Temporal: complements public transit at night and weekends (+) Spatial: serves as feeders for public transit

Note: “+” denotes positive impact on urban efficiency, “-” denotes negative impact, and “?” denotes uncertain impact.

2

(+) Better matching demand and supply (+) Reduces transaction cost

(+) Low-density area: cost efficient to substitute certain transit routes with ridesourcing

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

According to the interviews conducted by Glöss et al. (2016) with Uber drivers and passengers, both drivers and passengers felt safer because of Uber's rating and tracking system. Uber drivers, especially female drivers, perceived an increased sense of control and security, because customers were registered and rated, and the vehicle was tracked (Glöss et al., 2016). Uber passengers also gained an extra sense of safety and reliability knowing who the driver was and that the system was monitoring the real-time location of the vehicle (Glöss et al., 2016). Nevertheless, Edelman and Geradin (2016) contended that Uber drivers lacked sufficient training, even though there was no empirical evidence quantifying the correlation between lack of training with accident frequency. Edelman and Geradin (2016) also pointed out that Uber cars did not have mandatory regular inspections, and some of them were uninsured or underinsured.

affect urban efficiency. Table 1 provides a summary of the positive and negative impacts of ridesourcing on urban transportation and economic efficiency under these two primary forms of modal shift. 2.1. Ridesourcing vs. taxi Ridesourcing is believed to provide higher economic efficiency than taxis. Taking Uber as an example, its surge pricing (a dynamic pricing mechanism, in which fares increase when demand is high) efficiently meets the fluctuating transportation demand through the day (Edelman & Geradin, 2016). Uber also significantly reduces transaction costs through lowering the costs of searching for a ride, since passengers no longer need to call a dispatcher or stand on the street to hail a taxi, and instead, they can reserve a car from indoors (Rogers, 2015). Moreover, passengers can monitor the vehicles' real time GPS locations, which largely reduces the uncertainty and anxiety of waiting for a cab (Edelman & Geradin, 2016). Whether such economic efficiency can be translated into congestion reduction is another story. Anderson (2014) observed that ridesourcing brought more cars into the cities and has worsened congestion in the city center, as people living outside San Francisco drove to the city to work for ridesourcing, while taxi drivers could not do the same thing due to regulations. However, a study conducted by the popular media website FiveThirtyEight based on data of Uber trips in 2014 found that Uber was not substantially worsening the traffic congestion in New York during rush hours, because on average Uber created 682 additional trips between 4 p.m. to 5 p.m. per non-holiday weekday in Manhattan, which was a very small amount comparing to the total number of vehicles on the streets (Bialik, Fischer-Baum, & Mehta, 2015). Similarly, Nie (2017) estimated that ridesourcing imposed a relatively mild impact on the traffic congestion in Shenzhen, China, using taxi GPS trajectory data during 2013 to 2015. Given that ridesourcing is growing at an exponential speed, more research with updated data is needed to examine the impact of ridesourcing on congestions in central urban areas. It is also crucial to take into consideration the modal shift from public transit to ridesourcing, which is more likely to worsen congestions. In the popular media, ridesourcing has a positive image of improving transportation accessibility in low-income or low-density areas with insufficient taxi services (McArdle, 2015). This image is to a large extent fueled by a study commissioned by Uber. In this rider experiment study, riders were dispatched to low-income neighborhoods in Los Angeles and traveled along predetermined routes in pairs using UberX and taxi to compare costs and waiting time (Smart, Rowe, Hawken, et al., 2015). Their results indicated that UberX was cheaper and faster than taxis: on average, an UberX ride cost less than a half of a cab ride, and the waiting time of UberX was also less than half of taxis (Smart et al., 2015). FiveThirtyEight's study offered similar results: using data of nearly 93 million rides taken by Uber and taxi during April to September 2014, it found that 22% of the 4.4 million Uber trips started outside of Manhattan, in contrast to 14% of the 88.4 million taxi rides (Bialik, Flowers, et al., 2015). According to Pucher and Renne (2003) and Renne and Bennett (2014), low-income households used taxis more often than middle-income households in the U.S., probably due to a lower level of car ownership. If ridesourcing provides services similar to taxis but with lower costs and shorter waiting time in low-income neighborhoods, it could meaningfully improve the accessibility of lowincome individuals (National Academies of Sciences, Engineering, and Medicine, 2016). Nevertheless, it could also be argued that for low income neighborhoods, the availability of taxi services might be less important than the availability of public transit, considering the issue of affordability. Instead of promoting ridesourcing, it might be more worthwhile to improve public transit coverage and service frequency in low-income neighborhoods. The safety issue surrounding ridesourcing is still contested.

2.2. Ridesourcing vs. public transit Existing literature suggests that ridesourcing both complements and competes with public transit. On the one hand, ridesourcing fills in the temporal and spatial gaps in urban public transit networks and by doing so improves transportation accessibility. Temporally, ridesourcing is most commonly used during late night hours and weekends (American Public Transportation Association, 2016), when public transit services are less frequent or waiting for transit makes people feel unsafe. Spatially, ridesourcing might serve as feeders for public transits, alleviating the first-mile/last-mile problem. Notably, quantitative evidence for this claim mainly came from studies conducted by Uber's own policy research team, as they have the full access to Uber's rich trip data. For instance, one of their recent studies in Los Angeles found that Uber pickups surged when new train stations open, indicating that rail passengers used Uber to bridge between these new stations and their destinations (Williams, 2017). Their data showed that weekly Uber pickups within 100 m of the new stations increased significantly one month after the Expo Line extension opened on March 20, 2016 (Williams, 2017). However, without additional information on the actual trip origins, we cannot distinguish whether the trips were connecting to the new train stations or the new commercial or residential developments in the vicinity of the new stations. On the other hand, ridesourcing also competes with public transit, especially in high-density areas. One survey with ridesourcing users in San Francisco found that 33% of the ridesourcing users would have taken public transit if ridesourcing were not available, and that percentage (41%) was even higher for non-car owners (Rayle et al., 2016). Henao's (2017) study found that 22.2% of the passengers would travel with public transportation if Lyft/Uber were not an option. The American Public Transportation Association's (APTA, 2016) survey with shared mobility users in seven U.S. cities (including San Francisco) offered a lower percentage: if ridesourcing were not available, only 14% of respondents would use public transit. The difference amongst these three studies stems from their slightly different ways of asking the questions and their diverse sampling strategies. Clewlow and Mishra (2017), on the other hand, argued that the relationship between ridesourcing and public transit depended on the type of transit in question. They found that ridesourcing pulled people away from public bus and light rail, but complemented heavy rail (Clewlow & Mishra, 2017). In addition to ridesourcing, many TNCs have developed ridesplitting services to enlarge their market share in commuting trips. For instance, in 2016, Uber's ridesplitting service UberPool offered a summer deal for New York inhabitants that allowed riders to pay $49 for two weeks or $79 for four weeks of unlimited rides within Manhattan during commuting hours (Powers, 2016). UberPool's rate in this promotion campaign was lower than public transit because a 30-day unlimited MTA pass in New York City cost $116.50. Even though some cities and public agencies attempted to prohibit or regulate ridesourcing operations, many agencies started to cooperate with TNCs to supplement public transportation networks or even 3

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

Hughes and MacKenzie's (2016) study tested whether people in areas with lower average income experienced longer waiting time for ridesourcing services than people in other areas, using the data of around 1 million estimated waiting times for UberX vehicles in Seattle in 2015. Their results indicated that waiting time for UberX was more dependent upon the population and employment density of the area rather than the income level, and on average, waiting time was actually shorter in areas with lower income. Nevertheless, most ridesourcing apps require users to have a credit card or at least a bank account to register and pay for the rides, which excludes the financially underserved population (National Academies of Sciences, Engineering, and Medicine, 2016). According to the World Bank Global Findex database (Demirgüç-Kunt, Klapper, Singer, et al., 2015), globally, 38% of adults did not have a bank account in 2014; in OECD countries, 6% of the adult population were unbanked; in developing countries, 45% of adults did not own a bank account. Credit card ownership is even lower: in 2014, 47% of adults in OECD countries reported not owning a credit card, a figure which rose to 90% in developing countries (Demirgüç-Kunt et al., 2015). Discrimination is another controversial issue surrounding the burgeoning TNCs. Ge, Knittel, MacKenzie, et al. (2016) carried out a study on racial and gender discrimination of Uber and Lyft through sending passengers to take nearly 1500 rides on controlled routes in Seattle and Boston in 2015. They identified discrimination against African American riders: in Seattle, African American passengers experienced longer waiting times, while in Boston passengers using African Americansounding names experienced higher cancelation rates (Ge et al., 2016). They also noticed that female passengers were given extended rides and hence excessive fares in Boston, confirming the anecdotes that female riders were exposed to profiteering and flirting (Ge et al., 2016). In addition to racial and gender discrimination, how ridesourcing serves people with disabilities also brings up concerns. There used to be lawsuits against TNCs for discriminating against disabled passengers in 2014 and 2015 (Rosenthal, 2014; Stempel, 2015; Wieczner, 2015). Nowadays Uber and Lyft offer special services for people with disabilities (Lyft, 2017; Uber, 2017). Nonetheless, discrimination against riders is not a practice invented by ridesourcing (Rogers, 2015). The taxi industry has long been criticized for racial or ethnic discrimination (National Academies of Sciences, Engineering, and Medicine, 2016). Still, the taxi industry is highly regulated and in many cities taxis need to follow non-discrimination requirements. It is argued that, since many TNCs receive monetary or non-monetary support from public agencies, the non-discrimination requirements should be extended to TNCs receiving such support (Shaheen, Cohen, & Zohdy, 2016). Yet if the nondiscrimination requirements are ineffective for taxis, extending them to TNCs does not make much sense. We should probably focus on making these requirements effective first. Lastly, data privacy and security have been key concerns of ridesourcing users, researchers and public agencies (Shaheen, Cohen, Zohdy, et al., 2016). TNCs maintain highly sensitive personal data of their users and drivers, such as names, phone numbers, email addresses, driver's licenses, credit card information, and trip information. Users and drivers worry about the possible monitoring, exploitation, and leakage of their personal data (Rogers, 2015; Shaheen, Cohen, & Zohdy, 2016). TNCs' data collection policies (such as location tracking) often receive criticism from users and privacy advocates, which constantly pressures TNCs to make adjustments on their privacy policies (Volz, 2017). Data leakage, on the other hand, draws more attention from public agencies. For instance, in response to Uber's data breach in 2016, which exposed personal information of around 57 million users around the world, the European Union privacy regulators recently announced plans to create a task-force to coordinate investigations (Fioretti, 2017).

substitute some bus routes with subsidized or free ridesourcing rides (Shaheen & Chan, 2016). For example, in August 2016, the Pinnellas Suncoast Transit Authority (PSTA) in Pinnellas County, Florida, began providing subsidies for Uber rides that end at nearly 20 designated transit stops in the county (Brustein, 2016). The PSTA also launched a separate program offering low-income residents free Uber rides after 9 p.m. when buses stop running (Brustein, 2016). In a similar example, in July 2016, Miami-Dade County in Florida applied for grants to subsidize Uber and Lyft rides connecting to two train stations (Brustein, 2016). Moreover, some transportation agencies saw the potential of cost saving in replacing bus routes with ridesourcing trips. Running bus routes in areas with low density and low ridership is particularly expensive, making ridesourcing an attractive and cost efficient alternative. For instance, Centennial, a suburb of Denver, Colorado, partnered with Lyft in August 2016 to replace a shuttle bus route connecting the Dry Creek light-rail station with free Lyft rides using public funds (Aguilar, 2016; Brustein, 2016). 3. Urban equity In this section, we examine the social impacts of ridesourcing from three distinct and interrelated perspectives. Firstly, from consumers' perspective, we analyze how ridesourcing serves different segments of the society and discuss discrimination and data privacy and security issues. Secondly, from the prosumers' (consumers/producers) perspective, we discuss how consumers are turned into co-producers and the meaning of such transformation. Finally, from producers' perspective, we further examine the nature of on-demand work, with a particular focus on ridesourcing drivers. 3.1. Transportation equity, discrimination, and data privacy and security It is acknowledged by many studies that younger, better-educated, and more affluent individuals are more likely to be ridesourcing users (Clewlow & Mishra, 2017; McGrath, 2015; Rayle et al., 2016; and A. Smith, 2016). This is not surprising as research found that young people were more interested in shared-based transportation services (Davis & Dutzik, 2012; National Academies of Sciences, Engineering, and Medicine, 2013). Such a user profile raises the question of transportation equity – who are enjoying this new transportation option and who are excluded from it. This question becomes more worrying when ridesourcing substitutes public transit, a public service that supposes to be available and accessible to all community members. The younger, better-educated, and more-affluent user characteristics remind us of the digital divide, which originally referred to the unequal access to information and communication technology (ICT) and the requisite skills to use it amongst community members (Selwyn, 2004) and nowadays expands to the unequal access to smartphone and mobile data access. According to a survey conducted by Pew Research Center in 14 advanced countries in 2016, nearly all people in these countries owned a mobile phone, but not everyone had a smartphone (Poushter, 2017). Amongst the countries surveyed, Sweden, the Netherlands, Spain and Australia had the highest smartphone ownership rates (79% to 80%). 77% of Americans owned a smartphone. The smartphone ownership rates in France, Japan, Poland, Hungary, and Greece were below 60% (46% to 58%). In addition, the survey identified sharp digital divide, as smartphone ownership rates were higher in people who were younger, with higher education, and with income higher than the country median (Poushter, 2017). Such digital divide also existed in developing countries, where smartphone ownership rate was much lower than the developed countries but was catching up swiftly (Poushter, 2016). There have been mixed perspectives regarding how ridesourcing serves low-income populations. As mentioned in the previous section, the study by Smart et al. (2015) showed that UberX offered cheaper and faster services than taxis in low-income neighborhoods in Los Angeles.

3.2. Prosumerism The term prosumer originates from Alvin Toffler. In his book The 4

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

peer-to-peer). An Uber-funded study observed that Uber drivers earned at least the same as and in many cases more than taxi drivers and chauffeurs through comparing their earnings per hour (Hall & Krueger, 2015). Nevertheless, Hall and Krueger's (2015) study didn't take into consideration the benefits (such as healthcare insurance and paid leave) and expense reimbursement (such as car insurance and gasoline) enjoyed by taxi drivers and chauffeurs but not by Uber drivers. Moreover, the fluctuation of wages is threatening the income stability of on-demand workers. For instance, when Uber launched price-cutting campaigns to enlarge market share and stimulate demand, drivers' earnings dropped significantly and even below minimum wages, which happened in multiple cities at multiple times (Kosoff, 2014; Lazzaro, 2016). The problem is Uber can easily make changes to its pricing mechanism without seeking opinions from the drivers, while taxi companies cannot easily change their drivers' wages. This problem also relates to our next question on flexibility and security. As independent contractors or micro-entrepreneurs, on-demand workers enjoy the flexible working schedule but sacrifice the benefits that come with employee status – minimum wage, health insurance, overtime payment, paid sick days, paid family leave, protection from discrimination, unemployment insurance, and workers' compensation and contribution to Social Security (R. Smith, 2016). Some agree that ridesourcing drivers do indeed enjoy the flexibility of arranging their own schedules (Glöss et al., 2016; Hall & Krueger, 2015), which enhances their adaptability to handle unpredictable shocks in everyday life, such as a sick child (Chen, Chevalier, Rossi, et al., 2017). Others question the magnitude of flexibility on-demand workers actually have. For instance, Chen and Sheldon (2015) noticed that Uber drivers tend to utilize the flexibility to increase their hourly earnings. Similarly, Cockayne (2016) found that drivers needed to work when the price is high, so they still worked on the schedule set by consumer demand and the algorithm, not themselves. Cockayne (2016) also emphasized that TNCs had strong control over drivers because they could regulate the behavior of drivers with the algorithm and the rating system and easily deactivate drivers that did not follow their instructions.

Third Wave, Toffler (1981) contended that during the pre-industrial era (the first wave) most people consumed what they produced by themselves, and thus might be called prosumers. It was followed by the second wave, in which the industrial revolution separated the functions of production and consumption and thereby gave birth to producers and consumers. The contemporary society is moving towards the third wave that brings back the prosumers. In the third wave, more and more companies are externalizing labor by asking customers to perform tasks that had been previously done by employees, and most importantly, customers do not get paid for doing them (Toffler, 2013). In ridesourcing, consumers are turned into prosumers by the rating system, which gathers ratings from consumers and tallies them up to monitor drivers' performance – consumers become TNCs' unpaid middle managers (Dzieza, 2015). There are mixed opinions upon whether prosumers are exploited: some view prosumers as liberating figures; whereas others consider prosumerism as a new form of exploitation. In management and marketing literature, prosumption is labeled as ‘value co-creation’ (Prahalad & Ramaswamy, 2004) or ‘service-dominant logic of marketing’ (Vargo & Lusch, 2004). Both labels center on the idea that consumers transform from passive recipients of information and commodities to active interpreters and co-producers of both, representing consumer empowerment and more equal relationships between producers and consumers. Another key argument contributing to the positive image of prosumerism is that prosumers engage in the co-producing activities voluntarily and they seem to enjoy the work (Ritzer & Jurgenson, 2010). In contrast, scholars on the opposite side offer a more critical assessment on the idea of putting consumers to work. Zwick, Bonsu, and Darmody (2008), drawing upon Foucauldian and neoMarxian theory, posit that the true meaning of co-creation is to exploit the creative and valuable consumer labor. Even though the idea of cocreation releases consumers' creativity and know-how, it also expropriates surplus value from unpaid consumer labor (Zwick et al., 2008). As consumer labor is recoded as an enjoyable consumption experience, the exploitative nature of prosumption is made less obvious (Zwick et al., 2008). Similarly, Pauwels (2015) asserts that prosumerism obscures power struggles and labor issues through replacing the old production model of coerced exploitative working conditions with a more insidious but just as exploitative model.

4. Urban sustainability Even though there is widespread belief that the sharing economy helps to improve sustainability, very little empirical evidence has been provided to support such belief (Schor, 2016). Ridesourcing imposes both direct and indirect impacts on urban sustainability. Indirectly, ridesourcing has a potential to alter consumer behavior towards possessing fewer private vehicles. Directly, ridesourcing might reduce the energy consumption and greenhouse gas emissions of urban transportation.

3.3. The nature of on-demand work There has been heated debate over the nature of on-demand work in both popular media and academia. Advocates view on-demand work as the rise of micro-entrepreneurs, a new generation of self-employed workers who enjoy flexible work schedules, ideal work-life balance, and family-friendly lifestyles; whereas critics perceive the on-demand economy as a disparaging race to the bottom, in which companies shift risks to individual micro-entrepreneurs, and workers work more hours for less money and receive minimal job security and benefits (Reich, 2015). There are two key issues embedded in the debate. The first one concerns compensation – whether ridesourcing and the sharing economy provide adequate compensation for workers compared to regular full-time jobs. The second issue is the tradeoffs between flexibility and employment status. Criticisms of the low compensation of on-demand work have been concentrated on one particular type of on-demand job – crowdsourcing. Crowdsourcing refers to the practice in which companies divide jobs into discrete micro-tasks and outsource these micro-tasks to workers via business-to-peer platforms, such as Amazon's Mechanical Turk. It is argued that the online marketplaces increase competition on a global scale and lower workers' wages, creating a new, digital generation of sweatshops that make the public ignore the often-desperate conditions of on-demand workers (Aytes, 2012; Ettlinger, 2016). Nonetheless, the compensation of on-demand work varies depending on the nature of work performed and on the nature of the platforms (business-to-peer or

4.1. Private car ownership The sharing economy is thought to be green or environmentally friendly because it can increase the use of pre-existing goods and reduce the demand for new products (Botsman & Rogers, 2010; Cervero, Golub, & Nee, 2007; Frenken & Schor, 2017). Following this line of thought, shared mobility is assumed to reduce private car ownership, as it encourages consumers to use shared mode of transportation rather than purchase private vehicles. This assumption has been confirmed for carsharing (Firnkorn & Müller, 2015; Martin & Shaheen, 2016; Martin, Shaheen, & Lidicker, 2010), but not for ridesourcing. Rayle et al. (2016) observed that the presence of ridesourcing might not affect car ownership behavior, based on the result that 90% of survey respondents did not change car ownership levels since they started to use ridesourcing, and those who did change ownership split into two even groups: one group owned more cars, the other owned fewer cars. Similarly, in Clewlow and Mishra's (2017) study, 91% of ridesourcing users did not make any changes to car ownership. Anderson (2014) raised that ridesourcing might even encourage private car ownership because some 5

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

ridesourcing: consumers might use the cost saving from transportation to purchase other high-impact goods or services (Schor, 2016).

of the drivers used the ridesourcing income to subsidize their use of private vehicles, or even to buy cars. 4.2. Energy consumption and greenhouse gas emissions

5. Discussions: problems and prospects Apart from private car ownership, researchers have paid attention to the comparison of energy efficiency and greenhouse gas emissions between ridesourcing and taxi. Using taxi and Uber trip data in five cities in the U.S., Cramer and Krueger (2016) found that Uber had a higher mileage-based capacity utilization rate (61%) than taxis (49.1%). The capacity utilization rate was calculated as the ratio between the miles drivers had fare-paying passengers in the vehicles and their total working miles. Notably, Cramer and Krueger's (2016) calculation did not account for the mileage traveled by drivers to and from the working area from home when the ridesourcing app was not turned on. In Anderson's (2014) study, 16 out of 20 drivers interviewed lived in the outer areas of San Francisco and drove deadheading long distances to and from the city to work as ridesourcing drivers. Under such circumstances, Cramer and Krueger (2016) might overestimate the capacity utilization rate of Uber. A later study conducted by Henao (2017) partially solved this problem by taking into consideration the commuting distance of drivers after logging off from the apps and achieved a slightly lower capacity utilization rate (59.2%). Yet Henao's (2017) calculation was based on his own rides as an Uber and Lyft driver and he had a careful research design that minimized the cruising time for ride requests, refused rides that required long pickup distance, and planed for conservative commuting distance at the end of shifts, making his driving experience less representative of the general driver population. More research is needed to study whether ridesourcing has higher energy efficiency. In addition, there is a lack of study measuring and comparing the exact level of emissions produced by ridesourcing and taxis. Instead, researchers express concerns over the lack of regulatory control on the emissions of ridesourcing vehicles comparing to taxicabs (Light, 2017). For instance, Anderson (2014) contended that in San Francisco, nearly 100% licensed taxis were “clean air” vehicles, owing to the local regulator's “Clean Taxi” program, in contrast to the estimation made by the San Francisco Police Department that only 17% of the ridesourcing cars reprimanded for dropping passengers at the airport were clean air vehicles. However, ridesourcing replaces not only taxicabs, but also other modes of transportation, such as public transit, walking, biking, driving, and ridesharing. It is essential to make distinctions between different forms of modal shift when evaluating how ridesourcing affects energy consumption and greenhouse gas emissions. Henao (2017) accounted for the different modal shifts through combining driving information with passenger surveys. Driving information keeps records of the total vehicle miles traveled (VMT), while a passenger survey indicates what mode of transportation ridesourcing is replacing. Through comparing the total VMT of ridesourcing and the modes of transportation being replaced, Henao (2017) found that ridesourcing lead to an 84.6% increase in VMT compared to the status quo. Existing literature also discusses the extent of rebound effects. In 1865, William Stanley Jevons raised the argument that improvement in energy efficiency will increase instead of reduce energy consumption (Jevons, 1906). This happens because there are various mechanisms – the so-called rebound effects – that offset the predicted energy savings (Sorrell, 2009). There are direct and indirect rebound effects. A direct rebound effect, in the case of transportation, indicates that improvement in energy efficiency reduces the marginal cost of travel, thus the quantity of travel may be expected to increase (Sorrell, 2009). This means ridesourcing might generate additional travel demands, rather than merely replacing existing transportation modes. Rayle et al.'s (2016) study found that 8% of the ridesourcing trips were induced travel, which is a small but not insignificant amount. Henao (2017) identified an even higher induced travel rate (12.2%). In terms of the indirect effects, the re-spending effect is most closely related to

As the business models of TNCs are becoming increasingly sophisticated, it is easy to confuse different forms of shared mobility, which might lead to misleading study results. Amongst various types of shared mobility, the most easily confused pair is ridesharing and ridesourcing. While ridesharing (both traditional and app-based) is proved to be reducing VMT and greenhouse gas emissions (BlaBlaCar, 2015; Jacobson & King, 2009; Minett & Pearce, 2011), the environmental impact of ridesourcing is uncertain. Ridesharing improves energy efficiency and reduces emissions because the drivers will travel to the destinations regardless of the presence of passengers, whereas ridesourcing has no such synergy effect, as drivers do not have destinations by themselves – they provide taxi-like transportation services to passengers. The popular media often use ridesharing to refer to the services provided by Uber and Lyft, which is harmless as long as the readers understand it. However, confusion of these two concepts in research can produce inaccurate results. For instance, mistakenly defining all trips of Didi Chuxing as ridesharing trips, Yu, Ma, Xue, et al.'s (2017) study assumed that Didi Chuxing trips cause no additional energy consumption and emissions because the drivers will drive to their destinations even with no request for ridesharing. Therefore, they calculated the environmental benefits of Didi Chuxing as the sum of energy consumption and emissions of the alternative travel modes that passengers would have chosen if Didi Chuxing were not available. However, as can be seen from Fig. 2, Didi Chuxing offers a great variety of shared mobility options, ranging from ridesharing, ridesourcing, ridesplitting, E-hail, to microtransit. Only trips made through Didi Hitch could be defined as ridesharing. Therefore, Yu et al.'s (2017) study largely overestimated the environmental benefits of Didi Chuxing in Beijing. Future studies should exercise special caution to distinguish between different types of shared mobility (especially those provided by a single TNC) when measuring their environmental impacts. In addition to concept confusion, methodological issues warrant caution over the conclusions made by existing literature. Studies on ridesourcing primarily utilize three types of data: big data, rider/driver surveys and interviews, and rider/driver experiment. Big data (trip related information) kept by TNCs are the most crucial and valuable data used by ridesourcing research, while data from government agencies (such as taxi and limousine commission) often serve as complementary information. However, TNCs are often reluctant to release their data for independent research due to concerns about privacy and trade secrets (Frenken, 2017). Many TNCs commissioned their own studies (e.g. BlaBlaCar, 2015; Hall & Krueger, 2015) or participated in the research team (e.g. Yu et al., 2017), but we need more independent studies, independent research questions, and objective results. Open access to high-quality big data is key to a better understanding of ridesourcing and the sharing economy. Nevertheless, the prevalence of big data raises concerns over the risk of a new positivist and quantitative turn that favors generality – universal laws independent of space Didi Chuxing

Didi Hitch Ridesharing

Didi Express Ridesourcing

Didi Chauffeur High-End Ridesourcing

Didi Taxi E-Hail

Didi Express Pool Ridesplitting

Fig. 2. Transportation services provided by Didi Chuxing. [Source: Compiled by the authors]

6

Didi Bus & Minibus Microtransit

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

and time – over particularity that varies across place and time (Schwanen, 2016). Location does matter in ridesourcing studies. For instance, FiveThirtyEight's study shows that Uber is not substantially worsening the traffic congestion in New York during rush hours (Bialik, Fischer-Baum, & Mehta, 2015), but such a conclusion cannot be applied to cities like Los Angeles, where the public transit network is far less sophisticated and car dependency is high. Rider/driver surveys and interviews could offer rich information for research on ridesourcing, but caution needs to be exercised on the data collection procedures and the implications made from the results. For example, Rayle et al. (2016) oversampled people who were likely to be present in the survey locations during night time and thus oversampled social and recreational trips. Lastly, rider/driver experiments face the problem of representativeness. For instance, Smart et al.'s (2015) rider experiment found that Uber provided faster and cheaper services than taxis in low-income neighborhoods in Los Angeles. However, in this study, riders were provided with mobile devices, trained to use Uber's app, and supplied with the ride fares to conduct the experiment, which made these riders different from residents in those low-income neighborhoods who might not have smartphones, know how to use the app, or have a credit card to use the ridesourcing app. Therefore, Uber might not serve the lowincome neighborhoods as well as the study depicts. Future studies should make careful design of the experiments so that the rider/driver's demographic profile is representative of the research population. Looking into the future of ridesourcing, two major trends might significantly change urban transportation configuration: the further integration of ridesourcing and public transit and the adoption of automated vehicles by TNCs. Regarding the first trend, a deepening public-private partnership in urban transportation networks or even a fully privatized public transit by TNCs would have a substantial impact on transit accessibility, equity, and reliability (Brustein, 2016). As for the second trend, autonomous driving technology is a key competition area for TNCs as robots can work tirelessly, do not demand a salary, and don't care for employment status or benefits (Dudash, 2017). Having fleets of driverless cars on the street will not only affect congestion and transportation accessibility and safety, but also have a strong impact on the livelihoods of ridesourcing drivers, and even drivers working for the traditional transportation and logistics industry. Considering the future of the sharing economy as a whole, Frenken (2017) articulates three possible scenarios depending on who will regulate the sharing economy and who will benefit from it. The first scenario is platform capitalism, a capitalist future in which monopolistic large-scale sharing platforms (such as Uber and Airbnb) integrate multiple businesses models to offer seamless services. The second one is platform redistribution, a state-led future which redistributes the gains of the sharing economy from capitalists to labor through taxation. The last one is platform cooperativism, a citizen-led future with cooperative ownership and democratic control over the sharing platforms. Each scenario has a distinct institutional logic, but they may coexist in different sectors or spatial contexts (Frenken, 2017).

increasing popularity of ridesplitting services that compete with public transit. Regarding the safety issue, ridesourcing increases trust between drivers and passengers but faces the problems of insufficient driver training and uninsured or underinsured vehicles. Even though ridesourcing improves the availability of transportation in poor and remote areas, it does not seem to provide equal access to all community members – people without a smartphone, mobile data access, a credit card, or at least a banking account are excluded from this transportation option. The problem of digital divide is not limited to ridesourcing and the sharing economy; it is closely related to the smart city mentality, as urban governance shifted its focus from social well-being towards economic competitiveness energized by technology advancement, which leads to social polarization and exclusion (Graham & Marvin, 2001; Hollands, 2008). Issues of discrimination against riders and data privacy and security have raised concerns amongst ridesourcing users, researchers, and policy makers. How to prevent discrimination and protect users' privacy is a continuing inquiry. The question of whether ridesourcing (and the whole sharing economy) empowers or exploits prosumers probably cannot be answered by empirical analysis, but it remains a critical question under the circumstances of a swiftly growing sharing economy. On the contrary, the question regarding the compensation level of on-demand workers can and should be answered empirically. In addition, the nature of the flexibility and the problems of insecurity in on-demand work should be scrutinized by not only researchers but also policymakers. As argued by R. Smith (2016), flexibility and employment status are not incompatible. Even though ridesourcing and the sharing economy have been promoting a green image, their environmental impacts have not been robustly quantified. According to the very limited evidence offered by literature, it is unlikely that ridesourcing can reduce private car ownership. Other forms of shared mobility such as carsharing or ridesharing have a higher potential to alter consumer behavior and discourage vehicle ownership. Ridesourcing's impact on energy consumption and greenhouse emissions is uncertain. The calculation of VMT and emissions should consider different modal shifts and the rebound effects. The availability of high-quality big data and rider surveys are key to the evaluating of ridesourcing's environmental impact. As can be seen from Table 2, many questions remain unanswered in existing literature, calling for further research into the current as well as the potential impacts of ridesourcing and the sharing economy on urban development. Empirically, future studies should investigate whether ridesourcing worsens congestion in central urban areas, quantify the relationship between ridesourcing and public transit, assess the compensation conditions of on-demand workers, and measure the energy consumption and emissions of ridesourcing. Theoretically, future research should critically analyze the issue of social exclusion in the sharing economy, the nature and future of prosumption, the precariousness of on-demand work, and the transforming nature of capitalism under the growing sharing economy. Lastly, as most existing literature on ridesourcing focused on what happened in the U.S., conclusions drawn by existing research might not be applicable to other contexts. Future research should pay attention to the development of ridesourcing in other countries to bring in international perspectives on the understanding of ridesourcing as well as the sharing economy. Perhaps even more importantly, we hope this paper will stimulate more discussions on our policy options in light of these empirical findings so that we can make our future cities towards achieving the goals of economic efficiency, social equity, and environmental sustainability with the growth of the sharing economy.

6. Conclusions In view of the substantial transformations that ridesourcing and the sharing economy bring to the everyday life of urban inhabitants and urban development, this paper systematically reviews the existing literature on ridesourcing's impacts on urban efficiency, equity, and sustainability. Table 2 provides a summary of the positive, negative, as well as uncertain impacts of ridesourcing on urban development. The impact of ridesourcing on urban efficiency is more straightforward than on the other two aspects. Ridesourcing improves economic efficiency. It reaches low-income neighborhoods with insufficient taxi services. Ridesourcing both complements and competes with public transit. It is uncertain whether ridesourcing worsens congestion in city centers, and if yes, to what extent, especially under the circumstances of the

Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. 7

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

Table 2 Impacts of ridesourcing on urban development. [Source: Compiled by the authors] Positive impacts Efficiency

Equity

poor neighborhoods with • Reaches insufficient taxi services public transit during late • Complements night hours and weekends economic efficiency • Improves trust between drivers and • Increases passengers

Negative impacts

Uncertain impacts (remaining questions)

with public transit, especially in high• Competes density areas concerns: insufficient driver training and • Safety uninsured/underinsured vehicle

ridesourcing worsen congestion in city • Does centers? what extent does ridesourcing complement or • Tocompete with public transit?

divide: some segments of society have • Digital difficulty accessing ridesourcing services

to prevent discrimination? • How to protect users' privacy? • How (the sharing economy) empowering • Isor ridesourcing exploiting prosumers? are on-demand workers paid? • How flexibility and security incompatible? • Are to protect on-demand workers' rights? • How ridesourcing reduce energy consumption • Can and emissions?

Sustainability

References

September 2017. Dzieza, J. (2015). The rating game: How Uber and its peers turned us into horrible bosses. In: The Verge. Retrieved from http://www.theverge.com/2015/10/28/9625968/ rating-system-on-demand-economy-Uber-olive-garden, Accessed date: 1 July 2017. Edelman, B. G., & Geradin, D. (2016). Efficiencies and regulatory shortcuts: How should we regulate companies like Airbnb and Uber? Stanford Technology Law Review, 19(2), 293–328. Ettlinger, N. (2016). The governance of crowdsourcing: Rationalities of the new exploitation. Environment and Planning A, 48(11), 2162–2180 (DOI: https://doi.org/ 10.1177%2F0308518X16656182). Fioretti, J. (2017). EU privacy regulators to discuss Uber hack next week. Reuters, 23 November, 2017. Retrieved from https://www.reuters.com/article/us-ubercyberattack-eu/eu-privacy-regulators-to-discuss-uber-hack-next-weekidUSKBN1DN1X4, Accessed date: 30 November 2017. Firnkorn, J., & Müller, M. (2015). Free-floating electric carsharing-fleets in smart cities: The dawning of a post-private car era in urban environments? Environmental Science and Policy, 45, 30–40. http://dx.doi.org/10.1016/j.envsci.2014.09.005. Fischer-Baum, R., & Bialik, C. (2015). Uber is taking millions of Manhattan rides away from taxis. In: FiveThirtyEight. Retrieved from https://fivethirtyeight.com/features/ Uber-is-taking-millions-of-manhattan-rides-away-from-taxis/, Accessed date: 1 July 2017. Frenken, K. (2017). Political economies and environmental futures for the sharing economy. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences. (Epub ahead of print 21 May 2017) https://doi.org/10.1098/ rsta.2016.0367. Frenken, K., & Schor, J. (2017). Putting the sharing economy into perspective. Environmental Innovation and Societal Transitions, 23, 3–10. http://dx.doi.org/10. 1016/j.eist.2017.01.003. Ge, Y., Knittel, C. R., MacKenzie, D., et al. (2016). Racial and gender discrimination in transportation network companies. Working paper no. 22776, October. Retrieved from http://www.nber.org/papers/w22776. Glöss, M., McGregor, M., & Brown, B. (2016). Designing for labour: Uber and the ondemand mobile workforce. 2016 CHI conference on human factors in computing systems, Santa Clara, California, USA, 7-12 May 2016 (pp. 1632–1643). . Graham, S., & Marvin, S. (2001). Splintering urbanism: Networked infrastructure, technological mobilities and the urban conditions. London: Routledge. Hall, J. V., & Krueger, A. B. (2015). An analysis of the labor market for Uber's driverpartners in the United States. The National Bureau of Economic Research. Working paper no. 22843, November. Retrieved from http://www.nber.org/papers/w22843, Accessed date: 5 July 2017. Henao, A. (2017). Impacts of ridesourcing – Lyft and Uber – on transportation including VMT, mode replacement, parking, and travel behavior (PhD Thesis)U.S: University of Colorado. Hollands, R. G. (2008). Will the real smart city please stand up? City, 12(3), 303–320. DOI https://doi.org/10.1080/13604810802479126. Hughes, R., & MacKenzie, D. (2016). Transportation network company wait times in Greater Seattle, and relationship to socioeconomic indicators. Journal of Transport Geography, 56, 36–44. http://dx.doi.org/10.1016/j.jtrangeo.2016.08.014. Jacobson, S. H., & King, D. M. (2009). Fuel saving and ridesharing in the US: Motivations, limitations, and opportunities. Transportation Research Part D: Transport and Environment, 14(1), 14–21. http://dx.doi.org/10.1016/j.trd.2008.10.001. Jevons, W. S. (1906). The coal question: An inquiry concerning the progress of the nation, and the probable exhaustion of our coal-mines. New York: The Macmillan Company. Kosoff, M. (2014). Uber drivers speak out: We're making a lot less money than Uber is telling people. In: Business Insider. Retrieved from http://www.businessinsider.com/ Uber-drivers-say-theyre-making-less-than-minimum-wage-2014-10, Accessed date: 1 July 2017. Lazzaro, S. (2016). Uber drivers plan boycott after fare cuts slash their earnings to below minimum wage. In: Observer. Retrieved from http://observer.com/2016/01/Uberdrivers-plan-boycott-after-fare-cuts-slash-their-earnings-to-below-minimum-wage/,

Aguilar, J. (2016). Centennial teams up with Lyft for free rides to light rail station. The Denver Post, 17 August. Retrieved from http://www.denverpost.com/2016/08/15/ lyft-centennial-team-up-for-free-rides-light-rail-station/, Accessed date: 1 July 2017. American Public Transportation Association (2016). Shared mobility and the transformation of public transit. Report, American Public Transportation Association. Retrieved from https://www.apta.com/resources/reportsandpublications/ Documents/APTA-Shared-Mobility.pdf, Accessed date: 1 July 2017. Anderson, D. N. (2014). “Not just a taxi”? For-profit ridesharing, driver strategies, and VMT. Transportation, 41(5), 1099–1117. http://dx.doi.org/10.1007/s11116-0149531-8. Aytes, A. (2012). Return of the crowds: Mechanical Turk and neoliberal states of exception. In T. Scholz (Ed.). Digital labor: The internet as playground and factory (pp. 79– 97). New York: Routledge. Bialik, C., Fischer-Baum, R., & Mehta, D. (2015). Is Uber making NYC rush-hour traffic worse? In: FiveThirtyEight. Retrieved from https://fivethirtyeight.com/features/isUber-making-nyc-rush-hour-traffic-worse/, Accessed date: 1 July 2017. Bialik, C., Flowers, A., Fischer-Baum, R., et al. (2015). Uber is serving New York's outer boroughs more than taxis are. In: FiveThirtyEight. Retrieved from https:// fivethirtyeight.com/features/Uber-is-serving-new-yorks-outer-boroughs-more-thantaxis-are/, Accessed date: 7 January 2017. BlaBlaCar (2015). BlaBlaCar contribution to climate change. Retrieved from https:// www.blablacar.in/blablalife/reinventing-travel/environment/blablacarcontribution-climate-change, Accessed date: 7 January 2017. Botsman, R., & Rogers, R. (2010). What's mine is yours: The rise of collaborative consumption. New York: Harper Business. Brustein, J. (2016). Uber and Lyft want to replace public buses. Bloomberg, August 15. Retrieved from https://www.bloomberg.com/news/articles/2016-08-15/Uber-andlyft-want-to-replace-public-buses, Accessed date: 1 July 2017. Cervero, R., Golub, A., & Nee, B. (2007). City CarShare: Longer-term travel demand and car ownership impacts. Transportation Research Record, 1992, 70–80. http://dx.doi. org/10.3141/1992-09. Chen, K. M., Chevalier, J. A., Rossi, P. E., et al. (2017). The value of flexible work: Evidence from Uber drivers. Working paper no. 23296, March. Retrieved from http:// www.nber.org/papers/w23296. Chen, K. M., & Sheldon, M. (2015). Dynamic pricing in a labor market: Surge pricing and flexible work on the Uber platform. Retrieved from http://www.anderson.ucla.edu/ faculty_pages/keith.chen/papers/SurgeAndFlexibleWork_WorkingPaper.pdf, Accessed date: 1 July 2017. Clewlow, R. R., & Mishra, G. S. (2017). Disruptive transportation: The adoption, utilization, and impacts of ride-hailing in the United States. Report. Davis: Institute of Transportation Studies, University of California (UCD-ITS-RR-17-07). Cockayne, D. G. (2016). Sharing and neoliberal discourse: The economic function of sharing in the digital on-demand economy. Geoforum, 77, 73–82. http://dx.doi.org/ 10.1016/j.geoforum.2016.10.005. Cramer, J., & Krueger, A. B. (2016). Disruptive change in the taxi business: The case of Uber. The American Economic Review, 106(5), 177–182. http://dx.doi.org/10.1257/ aer.p20161002. Davis, B., & Dutzik, T. (2012). Transportation and the new generation: Why young people are driving less and what it means for transportation policy. Retrieved from https:// uspirg.org/sites/pirg/files/reports/Transportation%20%26%20the%20New %20Generation%20vUS_0.pdf, Accessed date: 30 November 2017. Demirgüç-Kunt, A., Klapper, L. F., Singer, D., et al. (2015). The global Findex database 2014: Measuring financial inclusion around the world. World Bank Policy research working paper no. 7255, April. Retrieved from https://ssrn.com/abstract=2594973. Dudash, S. (2017). Why Uber can't lose the driverless car race. Forbes(10 July), Available at: https://www.forbes.com/sites/greatspeculations/2017/07/10/uber-has-poleposition-in-race-to-develop-driverless-cars/#6f00806c7ab0, Accessed date: 23

8

Cities xxx (xxxx) xxx–xxx

S.T. Jin et al.

news/houston-texas/houston/article/Texas-disability-advocates-sue-Uber-Lyft-305645520.php, Accessed date: 30 November 2017. San Francisco Municipal Transportation Agency (2014). Taxis and accessible services division: Status of taxi industry. Report, San Francisco Municipal Transportation Agency Board Meeting, 16 September. Retrieved from https://www.sfmta.com/sites/ default/files/agendaitems/9-16-14%20Item%2011%20Presentation%20%20Taxicab%20Industry.pdf, Accessed date: 1 July 2017. Schor, J. (2016). Debating the sharing economy. Journal of Self-Governance and Management Economics, 4(3), 7–22. Schwanen, T. (2016). Geographies of transport II: Reconciling the general and the particular. Progress in Human Geography, 41(3), 355–364 (DOI: https://doi.org/ 10.1177%2F0309132516628259). Selwyn, N. (2004). Reconsidering political and popular understandings of the digital divide. New Media & Society, 6(3), 341–362 (DOI: https://doi.org/ 10.1177%2F1461444804042519). Shaheen, S., & Chan, N. (2016). Mobility and the sharing economy: Potential to facilitate the first- and last-mile public transit connections. Built Environment, 42(4), 573–588. http://dx.doi.org/10.2148/benv.42.4.573. Shaheen, S., Cohen, A., & Zohdy, I. (2016). Shared mobility: Current practices and guiding principles. Report, U.S. Department of Transportation, Federal Highway Administration. Retrieved from https://ops.fhwa.dot.gov/publications/ fhwahop16022/fhwahop16022.pdf, Accessed date: 30 November 2017. Shaheen, S., Cohen, A., Zohdy, I., et al. (2016). Smartphone applications to influence travel choices: Practices and policies. Report, U.S. Department of Transportation, Federal Highway Administration. Retrieved from https://ops.fhwa.dot.gov/ publications/fhwahop16023/fhwahop16023.pdf, Accessed date: 8 January 2018. Smart, R., Rowe, B., Hawken, A., et al. (2015). Faster and cheaper: How ride-sourcing fills a gap in low-income Los Angeles neighborhoods. Report, BOTEC Analysis Corporation, July. Retrieved from http://botecanalysis.com/wp-content/uploads/ 2015/07/LATS-Final-Report.pdf, Accessed date: 1 July 2017. Smith, A. (2016). Shared, collaborative and on demand: The new digital economy. In: Pew Research Center. Retrieved from http://www.pewinternet.org/2016/05/19/thenew-digital-economy/, Accessed date: 1 July 2017. Smith, R. (2016). Flexibility in the on-demand economy. In: National Employment Law Project (NELP). Retrieved from http://www.nelp.org/publication/flexibility-in-theon-demand-economy/, Accessed date: 1 July 2017. Sorrell, S. (2009). Jevons' paradox revisited: The evidence for backfire from improved energy efficiency. Energy Policy, 37(4), 1456–1469. http://dx.doi.org/10.1016/j. enpol.2008.12.003. Stempel, J. (2015). Uber must face lawsuit claiming bias against blind riders. Reuters, 20 April. Retrieved from https://www.reuters.com/article/us-uber-blind-lawsuit/ubermust-face-lawsuit-claiming-bias-against-blind-riders-idUSKBN0NB1W020150420, Accessed date: 30 November 2017. Toffler, A. (1981). The third wave. New York: Bantam Books. Toffler, A. (2013). Revolutionary wealth. New Perspectives Quarterly, 30(4), 122–130. http://dx.doi.org/10.1111/npqu.11414. Uber (2017). Accessibility at Uber. Retrieved from https://accessibility.uber.com/, Accessed date: 30 November 2017. Vargo, S. L., & Lusch, R. F. (2004). Evolving to a new dominant logic for marketing. Journal of Marketing, 68(1), 1–17. http://dx.doi.org/10.1509/jmkg.68.1.1.24036. Vasconcellos, E. A. (2001). Urban transport environment and equity: The case for developing countries. New York: Routledge. Volz, D. (2017). Uber to end post-trip tracking of riders as part of privacy push. Reuters, 29 August. Retrieved from https://www.reuters.com/article/us-uber-privacy/uberto-end-post-trip-tracking-of-riders-as-part-of-privacy-push-idUSKCN1B90EN, Accessed date: 30 November 2017. Wieczner, J. (2015). Why the disabled are suing Uber and Lyft. Fortune, 22 May. Retrieved from http://fortune.com/2015/05/22/uber-lyft-disabled, Accessed date: 30 November 2017. Williams, R. (2017). How Uber works together with Southern California's growing rail network. In: Medium. Retrieved from https://medium.com/uber-under-the-hood/ how-uber-works-together-with-southern-californias-growing-rail-networkfa8b1aba03b4, Accessed date: 21 September 2017. Yu, B., Ma, Y., Xue, M., et al. (2017). Environmental benefits from ridesharing: A case of Beijing. Applied Energy, 191, 141–152. http://dx.doi.org/10.1016/j.apenergy.2017. 01.052. Zwick, D., Bonsu, S. K., & Darmody, A. (2008). Putting consumers to work. Journal of Consumer Culture, 8(2), 163–196. http://dx.doi.org/10.1177/1469540508090089.

Accessed date: 1 July 2017. Light, S. E. (2017). Precautionary federalism and the sharing economy. Emory Law Journal, 66(2), 333–394. Lyft (2017). Accessibility vehicle dispatch. Retrieved from https://help.lyft.com/hc/enus/articles/213584508-Accessible-vehicle-dispatch, Accessed date: 30 November 2017. Martin, C. J. (2016). The sharing economy: A pathway to sustainability or a nightmarish form of neoliberal capitalism? Ecological Economics, 121, 149–159. http://dx.doi.org/ 10.1016/j.ecolecon.2015.11.027. Martin, E., & Shaheen, S. (2016). The impacts of Car2go on vehicle ownership, modal shift, vehicle miles traveled, and greenhouse gas emissions: An analysis of five North American cities. Working paper, July. Retrieved from http://innovativemobility.org/ wp-content/uploads/2016/07/Impactsofcar2go_FiveCities_2016.pdf. Martin, E., Shaheen, S., & Lidicker, J. (2010). Impact of carsharing on household vehicle holdings. Transportation Research Record, 2143, 150–158. http://dx.doi.org/10.3141/ 2143-19. McArdle, M. (2015). Uber serves the poor by going where taxis don't. Bloomberg, 20 July. Retrieved from https://www.bloomberg.com/view/articles/2015-07-20/uberserves-the-poor-by-going-where-taxis-don-t, Accessed date: 7 August 2017. McGrath, F. (2015). The demographics of Uber's US users. In: GlobalWebIndex. Retrieved from https://www.globalwebindex.net/blog/the-demographics-of-Ubers-us-users, Accessed date: 1 July 2017. Minett, P., & Pearce, J. (2011). Estimating the energy consumption impact of casual carpooling. Energies, 4(1), 126–139. http://dx.doi.org/10.3390/en4010126. National Academies of Sciences, Engineering, and Medicine (2013). Millennials and mobility: Understanding the millennial mindset and new opportunities for transit providers. Washington, DC: The National Academies Presshttp://dx.doi.org/10.17226/22500. National Academies of Sciences, Engineering, and Medicine (2016). Between public and private mobility: Examining the rise of technology-enabled transportation services. Washington, DC: The National Academies Presshttp://dx.doi.org/10.17226/21875. Nie, Y. (2017). How can the taxi industry survive the tide of ridesourcing? Evidence from Shenzhen, China. Transportation Research Part C, 79, 242–256. http://dx.doi.org/10. 1016/j.trc.2017.03.017. Pauwels, M. (2015). Work and prosumerism: Collaborative consumption. In O. Frayssé, & M. O'Neil (Eds.). Digital labour and prosumer capitalism: The US matrix (pp. 66–84). Hampshire: Palgrave Macmillan. Poushter, J. (2016). Smartphone ownership and internet usage continues to climb in emerging economies. In: Pew Research Center. Retrieved from http://www. pewglobal.org/2016/02/22/smartphone-ownership-and-internet-usage-continuesto-climb-in-emerging-economies/, Accessed date: 29 November 2017. Poushter, J. (2017). Smartphones are common in advanced economies, but digital divides remain. In: Pew Research Center. Retrieved from http://www.pewresearch.org/facttank/2017/04/21/smartphones-are-common-in-advanced-economies-but-digitaldivides-remain/, Accessed date: 29 November 2017. Powers, M. (2016). Be jealous, New Yorkers are getting a month of unlimited Uber commutes for $79. The Washington Post, 12 July. Retrieved from https://www. washingtonpost.com/news/dr-gridlock/wp/2016/07/12/be-jealous-new-yorkersare-getting-a-month-of-unlimited-Uber-commutes-for-79/?utm_term=. 4f1d4b1aad3c, Accessed date: 1 July 2017. Prahalad, C. K., & Ramaswamy, V. (2004). Co-creation experiences: The next practice in value creation. Journal of Interactive Marketing, 18(3), 5–14. http://dx.doi.org/10. 1002/dir.20015. Pucher, J., & Renne, J. L. (2003). Socioeconomics of urban travel: Evidence from the 2001 NHTS. Transportation Quarterly, 57(3), 49–77. Rayle, L., Dai, D., Chan, N., et al. (2016). Just a better taxi? A survey-based comparison of taxis, transit, and ridesourcing services in San Francisco. Transport Policy, 45, 168–178. http://dx.doi.org/10.1016/j.tranpol.2015.10.004. Reich, R. B. (2015). The share-the-scraps economy. In: Robert Reich. Retrieved from http://robertreich.org/post/109894095095, Accessed date: 1 July 2017. Renne, J. L., & Bennett, P. (2014). Socioeconomics of urban travel: Evidence from the 2009 national household travel survey with implications for sustainability. World Transport Policy & Practice, 20(4), 7–27. Ritzer, G., & Jurgenson, N. (2010). Production, consumption, prosumption. Journal of Consumer Culture, 10(1), 13–36 (DOI: https://doi.org/ 10.1177%2F1469540509354673). Rogers, B. (2015). The social cost of Uber. Temple University Legal Studies Research paper no. 2015-28. Rosenthal, B. M. (2014). Texas disability advocates sue Uber, Lyft, 30 other entities. Houston Chronicle, 24 July. Retrieved from http://www.houstonchronicle.com/

9