Content uploaded by Lindsey Cameron
Author content
All content in this area was uploaded by Lindsey Cameron on Dec 16, 2021
Content may be subject to copyright.
This article was downloaded by: [129.105.215.146] On: 15 December 2021, At: 06:49
Publisher: Institute for Operations Research and the Management Sciences (INFORMS)
INFORMS is located in Maryland, USA
Organization Science
Publication details, including instructions for authors and subscription information:
http://pubsonline.informs.org
“Making Out” While Driving: Relational and Efficiency
Games in the Gig Economy
Lindsey D. Cameron
To cite this article:
Lindsey D. Cameron (2021) “Making Out” While Driving: Relational and Efficiency Games in the Gig Economy. Organization
Science
Published online in Articles in Advance 14 Dec 2021
. https://doi.org/10.1287/orsc.2021.1547
Full terms and conditions of use: https://pubsonline.informs.org/Publications/Librarians-Portal/PubsOnLine-Terms-and-
Conditions
This article may be used only for the purposes of research, teaching, and/or private study. Commercial use
or systematic downloading (by robots or other automatic processes) is prohibited without explicit Publisher
approval, unless otherwise noted. For more information, contact permissions@informs.org.
The Publisher does not warrant or guarantee the article’s accuracy, completeness, merchantability, fitness
for a particular purpose, or non-infringement. Descriptions of, or references to, products or publications, or
inclusion of an advertisement in this article, neither constitutes nor implies a guarantee, endorsement, or
support of claims made of that product, publication, or service.
Copyright © 2021, INFORMS
Please scroll down for article—it is on subsequent pages
With 12,500 members from nearly 90 countries, INFORMS is the largest international association of operations research (O.R.)
and analytics professionals and students. INFORMS provides unique networking and learning opportunities for individual
professionals, and organizations of all types and sizes, to better understand and use O.R. and analytics tools and methods to
transform strategic visions and achieve better outcomes.
For more information on INFORMS, its publications, membership, or meetings visit http://www.informs.org
“Making Out”While Driving: Relational and Efficiency Games
in the Gig Economy
Lindsey D. Cameron
a
a
Wharton School, University of Pennsylvania, Philadelphia, Pennsylvania 19103
Contact: ldcamer@wharton.upenn.edu,https://orcid.org/0000-0002-5019-577X (LDC)
Received: January 15, 2020
Revised: December 30, 2020; June 20, 2021;
September 13, 2021
Accepted: October 2, 2021
Published Online in Articles in Advance:
https://doi.org/10.1287/orsc.2021.1547
Copyright: © 2021 INFORMS
Abstract. On-demand or “gig”workers show up to a workplace without walls, organiza-
tional routines, managers, or even coworkers. Without traditional organizational scaffolds,
how do individuals make meaning of their work in a way that fosters engagement? Prior
literature suggests that organizational practices, such as recruitment and socialization, fos-
ter group belonging and meaningfulness, which subsequently leads to engagement, and
that without these practices alienation and attrition ensue. My four-year qualitative study
of workers in the largest sector in the on-demand economy (ridehailing) suggests an alter-
native and more readily available mechanism of engagement—workplace games. Through
interactions with touchpoints—in this context, the customer and the app—individuals turn
their work into games they find meaningful, can control, and “win.”In the relational game,
workers craft positive customer service encounters, offering gifts and extra services, in the
pursuit of high customer ratings, which they track through the app’s rating system. In the
efficiency game, workers set boundaries with customers, minimizing any “extra”behavior,
in the pursuit of maximizing money per time spent driving and they create their own
tracking tools outside the app. Whereas each game resulted in engagement—as workers
were trying to “win”—games were associated with two divergent stances or relationships
toward the work, with contrasting implications for retention. My findings embed
meaning-making in what is fast-becoming the normal workplace, largely solitary and
structured by emerging technologies, and holds insights for explaining why people remain
engaged in a line of work typically deemed exploitative.
History: This paper has been accepted for the Special Issue on Emerging Technologies and Organizing.
Funding: The author’sworkwasfinancially supported by the Rackham School, University of Michigan.
Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2021.1547.
Keywords:workplace games •meaning/meaning-making •independent work •on-demand/gig economy •Uber •Lyft
TaskRabbit. DoorDash. Instacart. The on-demand or
“gig”economy is rapidly changing how work is orga-
nized, with digital platforms connecting workers and
customers for tasks that last only minutes or hours, and
algorithms performing functions previously carried out
by managers. Whereas emerging research explores how
algorithms and digital platforms restructure work (e.g.,
Lee et al. 2015,Curchodetal.2019, Kellogg et al. 2020,
Vallas and Schor 2020,Rahman2021), only a few stud-
ies have examined how such changes—notably, the
lack of traditional organizational scaffolds (e.g., sociali-
zation practices) and the replacement of human manag-
ers with automated algorithms—shape the relationship
between workers and their work. In a work setting de-
void of managers and coworkers, how do individuals
make meaning of their work, and how does that affect
the effort or engagement they devote to their work? Go-
ing beyond the dichotomy of “good job, bad job”
(Kalleberg 2011) that often characterizes independent
work, this paper explores how those working in the
socially isolated and algorithmically managed setting of
theon-demandworkplacemakemeaningofand,ulti-
mately, engage with their work.
Making meaning of one’ssurroundingsisafun-
damental human endeavor (Brief and Nord 1990,
Baumeister 1991,HallandMirvis2013), and from
the earliest days of capitalism, business leaders and
the popular press have encouraged workers to inter-
pret the significance of their work and its role in
their lives (Hurst et al. 2017). Traditionally, organi-
zations align workers’behaviors to organizational
objectives, scaffolding meaning-making by provid-
ing a shared mission, fostering group membership,
and infusing work with purpose. Organizations
with “strong cultures”(O’Reilly 1989) connect work-
erstosharedgoals,suchasputtingamanonthe
moon (Carton 2018), promoting a healthy lifestyle
(Besharov 2014), or achieving financial freedom
(Pratt 2000), and thus align workers’behaviors to
their objectives. Managers and coworkers offer cues
1
ORGANIZATION SCIENCE
Articles in Advance, pp. 1–22
ISSN 1047-7039 (print), ISSN 1526-5455 (online)
http://pubsonline.informs.org/journal/orsc
December 14, 2021
that help individuals interpret their surroundings
(Salancik and Pfeffer 1978, Wrzesniewski and Dut-
ton 2001)—for example, when managers normalize
the taint of dirty or illegal work by extolling its posi-
tive value (Anteby 2008,Ashforthetal.2017)or
when coworkers remind one another of shared val-
ues in difficult situations, such as tending to cranky
patients (Dutton et al. 2016).
Theon-demandworkplacelacksmanyofthese
scaffolds that traditionally support meaning-making,
relying instead on emerging technologies, particularly
algorithms, to facilitate work: a digital infrastructure
connects workers with customers for short-term assign-
ments in real time—hence the term, “on-demand.”Irre-
spective of “where”exactly the work is done (i.e.,
digitally by an MTurker or on the road by a Dasher),
the digital platform constitutes a defined work setting
by enabling worker–platform–customer interactions.
However, there is no “there”there. Individuals work
independently with little to no direct contact with man-
agers or coworkers, and algorithms take on managerial
functions such as hiring, directing, evaluating, and
disciplining workers (Lee et al. 2015, Cameron and
Rahman 2022, Rahman and Valentine 2021). Tasks are
microsized (Irani and Silberman 2013); algorithms set
pay rates (Rosenblat 2018); and automated bots trouble-
shoot workers’problems (Shapiro 2018). Moreover,
on-demand work has been construed as a “bad job”be-
cause of its precarious nature—the work is financially
uncertain, unpredictable in terms of scheduling, and of-
ten physically dangerous (Kalleberg 2011, Ravenelle
2019, Cameron et al. 2021b). This environment, which
offers limited opportunities to enrich group member-
ship or enhance work tasks, creates a context ripe for
alienation, in which workers become estranged from
the organization, their coworkers, and even themselves
(Braverman 1974, Pratt and Ashforth 2003). Indeed,
some accounts equate the resultant technically mediat-
ed workplace to an “assembly line in the head”(Bain
and Taylor 2000) where workers are subject to the
“tyranny of the algorithm”(Lehdonvirta 2018)andtoil
under “algorithmic despotism”(Griesbach et al. 2019).
1
In this paper, I ask, “How do on-demand workers make
meaning of their work in the face of these challenges,
and how does this affect their overall engagement with
the work itself?”
To answer these questions, I draw on a four-year
study of workers in the largest sector of the on-demand
economy, ridehailing. In this sector, which lacks tradi-
tional scaffolds, I find that two distinct touchpoints (in-
teractions or points of contact with the work), each in-
terpreted differently, undergird two workplace games.
In what I call the relational game,workerscreatemean-
ing by connecting with customers and crafting positive
customer service encounters, which they quantify and
track through the app’s algorithmically mediated rating
system. In the efficiency game,workersfind meaning by
completing work quickly at the highest pay rate and
managing customers by minimizing any extrarole be-
havior;but,unabletoaccuratelytracktheirefforts
through the app, drivers create their own tracking tools
and, at times, resort to manipulating the platform’s
algorithm. Whereas each game results in immediate
engagement or effort—as workers try to “win”—the
games are associated with two divergent stances or re-
lationships toward the work, either amicable or adver-
sarial, with contrasting implications for retention.
In contrast to when organizations and managers
scaffold meaning-making and bank on the common
perception of purpose to guide workers’behaviors,
the emerging technologies that scaffold on-demand
work cannot fully encompass the meaning-making
process, which thus becomes fragmentary. In other
words, as there is more room for workers to interpret
the touchpoints, there are more ways to incorporate
touchpoints into games, resulting in different kinds of
meanings. In contrast to prior research, which empha-
sizes how workers collectively define, refine, and rein-
force the rules of a single game (e.g., Burawoy 1976),
this study identifies two each with its own rendering
of the platform and mechanisms for fostering engage-
ment. Ultimately, I argue that the extent to which each
game’s meaning is supported by the digital platform
has divergent implications for workers’long-term
commitment to the platform.
Meaning-Making, Worker Engagement,
and On-Demand Work
Making meaning of work entails interpreting what
work signifies and the role it plays in one’s life (Brief
and Nord 1990,BaumeisterandVohs2002). According
to Pratt and Ashforth (2003), meaning is the output of
having made sense of something, such as an individual
coming to understand the place of work in the broader
context of their life. Meaning can be constructed indi-
vidually (from one's own values and perceptions), so-
cially (from norms or shared perceptions), or both (Pratt
and Ashforth 2003). Work that is deemed meaningful
(positive meaning) is associated with many benefits, in-
cluding performance (Wrzesniewski 2002), motivation
(Hackman and Oldham 1980), identification (Pratt et al.
2006), job satisfaction (Wrzesniewski et al. 1997), em-
powerment (Spreitzer and Quinn 1996), reduced stress
(Elangovan et al. 2010), and fulfillment (Kahn 2007)—in
short, elements that foster engagement.
Challenges in Socialization
Relationships with others in the workplace affect indi-
viduals’meaning-making and engagement, especially
when they perceive that these relationships provide a
sense of coherence between themselves and the work
Cameron: Making Out While Driving
2Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
(Dutton and Heaphy 2003, Kahn and Fellows 2013,Dut-
ton and Ragins 2017). Socialization practices often go
beyond a narrow focus on knowledge or skills to build
collective identities and foster belonging (Ashforth et al.
2000,Kunda2009). During NASA’slunarpursuit,man-
agement guided all workers, even secretaries, to con-
strue their jobs not as an assortment of short-term tasks,
but as actions supporting NASA’s long-term objective
of lunar landing (Carton 2018). Managers of individuals
who perform “dirty work”offer validation and cogni-
tive reframing, thereby neutralizing the job’sstigma
(e.g., without personal injury lawyers, manufacturers
would not be held accountable; Ashforth et al. 2017).
Many organizations with spatially distributed workers
employ similar socialization practices that emphasize
groupmembershipthroughregionalevents,routinized
scripts, and coordinating outfits (e.g., Hochschild 1983,
Leidner 1993). In the network marketing organization
Amway, distributors organize “dream-building sessions”
in which members of one’s“family”tree introduce new
distributors to the company’s culture and selling practices,
thereby fostering belonging and collective identification
(Pratt 2000). In building community among its members,
organizations foster a sense of group membership that en-
courages engagement in the work.
In recent years, one of the fastest growing areas of
employment has been in nonstandard work arrange-
ments, such as on-demand or “gig”work (Cappelli and
Keller 2013;KatzandKrueger2016,2019; Spreitzer et al.
2017), in which workers have peripheral and/or tempo-
rary group membership. Typically working indepen-
dently, on-demand workers are often physically and so-
cially isolated, having no regular in-person meetings
with managers or coworkers. Without true organiza-
tional membership, independent workers often struggle
to build a cohesive work identity or a sense of belong-
ing (Caza et al. 2018) and can expend significant effort
to forge their own practices to manage the existential
angst of being untethered to an organization (Petriglieri
et al. 2019).Andasthisworkisnotcloselytiedtoapro-
fession, individuals do not have access to professional
societies and associations, which can serve as avenues
of socialization (e.g., Barley and Kunda 2004,O’Mahony
and Bechky 2006). Even truck drivers and taxi drivers—
the closest analogue to ridehailing drivers—have
shared social spaces. Mandatory training for licensing
and vehicle pickups at corporate/leasing offices ensure
that drivers are aware of occupational norms and stand-
ards and that higher-ups are available if they have
problems (e.g., Luedke 2010, Viscelli 2016). On the road,
taxi drivers often run into one another at mechanic ga-
rages, road stops, and relief centers. One taxi driver-
cum-ethnographer describes the waiting lot at John F.
Kennedy airport as a “setting for drivers’sociability,”
with queues up to four hours long and 50 cars deep
(Occhiuto 2017, p. 279). While waiting, Occhiuto talked
shop,playedcards,ate,and,ofcourse,interviewed
other taxi drivers. Without a shared gathering space,
the potential of meaning-making through social me-
chanisms such as socialization and belonging for
on-demand workers is limited.
In addition, the nature of on-demand work itself
limits the ability to create rich interpersonal relation-
ships with customers. In contrast to relationship-based
service work in which parties have repeated interac-
tions that build trust and goodwill (e.g., hairdresser,
doctor; Rahman and Barley 2017), interactions with
customers in on-demand work are one-off, since the
matching algorithm makes it difficult to be matched
with the same customer repeatedly. These “pseudo-
relationships”are transactional and limited in trust,
mutual understanding, and goodwill (Gutek 1995).
Transactional service work highlights this instrumen-
tality, with one worker noting that “the ultimate con-
cern of sales is not the product or service—it’sabout
the prospect’smoney”(car sales; Oakes 1990,p.37).A
salesman that sold to those in his personal network
noted similar transactional relationships “All my ef-
forts were focused on getting to the next pin level. . . .
Friendship was only limited to what I could sell”(di-
rect sales; Butterfield 1985,p.66–8).Overall,without
organizational scaffolds to foster a sense of group
belonging, the burden lies on on-demand workers to
create their own systems to stave off alienation and
spur engagement in the work at hand.
Workplace Games as a Means of Fostering
Engagement
A long tradition of research has examined games in
the workplace. Early research describes how work-
ers would play poker or steal one another’s bananas
as a way to pass the time and relieve “the beast of
monotony”(Roy 1959,p.158;DeMan1928; Roethlis-
berger and Dickson 1939). Burawoy (1976)docu-
ments more sophisticated games that generate en-
gagement in the work itself. Defined as a set of rules,
a set of possible outcomes, and a set of outcome pref-
erences, games allow workers to make choices about
when and how much effort to exert in order to ob-
tain a desired outcome (Sallaz 2013).Inthegameof
“making out”in the manufacturing industry, for ex-
ample, machine operators synchronize their efforts
with production quotas and the piece-rate pay. In
the “tipping game”in the service industry, frontline
workers decide how much “emotional capital”to ex-
ert in completing tasks for customers in return for
tips (Sherman 2007, Sallaz 2009). More than just
helping workers pass the time or maximize income,
thesegamesproduceasenseofsocialandpsychological
achievement leading workers to enthusiastically engage
in the execution of their work. Games are widespread,
presenting across varied occupational groups, including
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 3
truckers (Ouellet 1994), lawyers (Pierce 1996), IT profes-
sionals (Barley and Kunda 2004), service workers in ca-
sinos (Sallaz 2009) and call centers (Sallaz 2019), “girls”
on the VIP party circuit (Mears 2015), and even the un-
employed (Sharone 2007).
A game is more than an assemblage of incentives; it
is a set of rules, strategies, and “wins”that take place
in a collective context. Unlike gamification, which re-
lies on rules designed by management to improve
workers’affective experience (Deterding et al. 2011,
Mollick and Rothbard 2014), games are a result of
organic interactions between coworkers. Indeed, be-
cause social interactions are considered a necessary el-
ement of games, some scholars have used work games
and social games interchangeably (e.g., Sherman 2007,
Sharone 2013). To be sure, organizations provide the
raw materials of the game; however, it is the workers
who determine how the game is played and who
is winning. Two aspects make games social. First,
workers typically agree on the objectives and rules,
with veterans instructing newcomers on how to play
(Sallaz 2009). In gratitude to an old-timer who taught
him how to improve his make-out rate, Burawoy gift-
ed him a ham (1976, p. 52). After a new concierge easi-
ly obtained a reservation at an exclusive restaurant, a
more experienced concierge advised her to “make the
guest think you worked hard”to secure larger tips
(Sherman 2007, p. 129). Second, games are social be-
cause workers receive feedback on their performance
from coworkers, often competing against one another
(Sallaz 2013). Among casino workers, breakroom con-
versations revolved around higher earners and their
strategies to increase tips (Sallaz 2002). And in the fac-
tory, Burawoy (1976, p. 63) remarked that even non-
work conversations referenced the game: When
“someone comes over to talk, his first question is, ‘Are
you making out?’followed by ‘What’s the rate?’”
Those who accumulate more wins garner status
among their coworkers, “strut[ting] around the floor”
(Burawoy 1976, p. 64) and getting praise for their
“hamminess”(Sherman 2007, p. 128), whereas those
who do not follow the rules are shunned and labeled
“rate busters”(Burawoy 1976, p. 145).
Thevariablenatureoftheoutcomeor“win”entices
workers to play games. In manufacturing work, unpre-
dictability is determined by the time-study man who
sets the piece rate for each machine. Some jobs are
“stinkers”(difficult), so workers are content with base-
line pay, whereas others are “gravy”(easy), so workers
take more frequent breaks or speed up the pace of
work, building up a kitty (Burawoy 1976). In service
work, customers’preferences and behaviors—from
what kind of service they prefer to their propensity to
tip well—generate uncertainty. Cab drivers used typol-
ogies such as “the sport,”“the blowhard,”and “the
lady shoppers”to classify customers’tipping predict-
ability (Davis 1959). Workers keep a close watch on
potential sources of unpredictability and adjust their
actions accordingly. Too much or too little variability
wouldleadthemtoquitplaying.Forexample,ifacus-
tomer was a never-tipper, such as a solo businessman,
workers exerted no extra effort, leaving him to carry
his own bags (Sherman 2007,p.130).Understanding
which elements of the game were unpredictable was
essential for workers to decide how to approach the
game or if they were even going to play at all.
Overall, the meaning-making literature has empha-
sized how organizations can encourage workers to
find meaning through their work by providing scaf-
folds that foster group belonging. To date, however,
the literature has not explored meaning-making in a
workplace with few scaffolds and minimal social in-
teractions, such as the on-demand workplace. The
lens of workplace games may provide an alternative
mechanism of meaning-making, with a focus on how
workers make meaning at work as opposed to through
the work. Whereas games are usually described as so-
cially immersive—as workers collectively strategize
how to control and “win”against the unpredictable
element of the game to garner status—I examine what
role games might play in understanding how individ-
uals construct meaning in a setting where work is soli-
tary and mediated by technology.
Research Setting, Data Collection, and
Analysis
Ridehailing Services
First launched in 2011, ridehailing services, such as
Uber, Lyft, and Juno, have disrupted the taxicab indus-
try. The core innovation that enables these services is
algorithms. Algorithms match drivers (independent
contractors working from their own cars) with custom-
ers within seconds, giving drivers block-by-block di-
rections. Behavioral suggestions or nudges direct
workers to take specific actions (Scheiber 2017,
Cameron 2021). Pop-up messages might encourage
drivers to continue driving for an extra ten minutes to
earn as much as yesterday or alert them to how much
more money they would make if they drove at differ-
ent times. Fares dynamically adjust based on demand,
and performance is evaluated by customer ratings and
driver acceptance and cancellation rates. Drivers have
little contact with company employees; even hiring
and firing, euphemistically called “activation or de-
activation,”is conducted virtually. Requirements may
vary, but most companies require clean driving re-
cords, no moving violations in the previous three
years, state vehicle inspections, and, increasingly,
despite industry protests in some cities, criminal
Cameron: Making Out While Driving
4Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
background checks. Once hired, which can take from
three days to three weeks, workers can go “online”
and begin.
The Workplace: RideHail—Technically Mediated
and Socially Isolating
In the ridehailing industry, drivers manage and
maintain their personal workplace, namely, the car
they are driving. Unlike taxi services, cars may be
personal vehicles, either purchased or leased (some
explicitly for work) or rented. In addition to being a
tool to earn income, cars often become a symbol of
personal expression, as workers choose the make
and model of the car and customize the interior with
trinkets, photos, or scents, for example. Two other
salient features of the workplace are the customers
and the app, both of which “manage”drivers. Cus-
tomer ratings are proxies for performance evalua-
tions, and drivers who continually receive poor rat-
ings or complaints are deactivated (Cameron and
Rahman 2021). Given the remote nature of the work,
interactions with customers are typically the only
human contact that drivers have while working.
2
Drivers begin a shift by choosing a location to open
their app and swiping right to go “online.”Once on-
line, the app coordinates the work cycle, including as-
signing rides, providing routing directions, setting
timers, and setting pay rates. Within the app, drivers
navigate between screens to get information specific
to their work—including maps of high-demand areas,
upcoming promotions, amount of time driving, rat-
ings, customer compliments/complaints, summary of
prior rides and fares, and account documentation
(e.g., ride history, car registration)—as well as to con-
tact driver support. Participants in my study were of-
ten active on several platforms, and switched between
them during a shift, having multiple apps open and
choosing the one that matched them with a ride first.
Because of the near-identical nature of the work, driv-
ers’windshields often displayed multiple company
decals (e.g., Lyft mustache and Uber decal), and,
when speaking, drivers often referenced their work
on the platforms interchangeably (e.g., using the Uber
term “surges”and the Lyft term “primetime”in the
same sentence to describe incentives). Further, core al-
gorithmic features across platforms were similar (e.g.,
rating systems). Thus, in this paper I use the terms
“ridehail(ing)”when discussing the act of driving and
“RideHail”when discussing the platform company,
and only name a specific company when it is neces-
sary to contextualize a comment.
Unlike taxi drivers, who attend training programs
and meet in leasing offices and garages, on-demand
drivers have no direct contact with managers or cow-
orkers, and most drivers do not know other drivers.
When I asked one informant if he ever saw other
drivers, he took a moment before replying thoughtful-
ly: “Yeah. When I check the passenger app and see
[icons of other] cars”(28). During my three years of
driving, I did not meet a single employee of RideHail,
and, apart from research purposes, I only met other
drivers in passing at gas stations and rest stops.
3
Even
drivers who worked in an area with a RideHail re-
source center tended to avoid them due to inefficient
service. Instead, the website and the app were the pri-
mary means of communication between drivers and
RideHail and outlined (un)acceptable behaviors. Vid-
eos and community guidelines, for example, encour-
aged drivers to keep their cars clean, treat customers
with respect, and follow safety laws. They also cau-
tioned that drivers could be deactivated for any
number of reasons, including failing to maintain a
minimum rating (no exact number was given), not fol-
lowing the app’s GPS directions (possible fraud), or
customer complaints. These materials were publicly
available, but drivers were not required to review
them before beginning to work, and, to the best of the
author’s knowledge, there was no way that RideHail
could confirm whether drivers had reviewed them.
Overall, drivers had few tangible scaffolds about the
organization’s culture, yet their actions while working
for RideHail were tightly scripted by the organiza-
tion’s algorithms.
Data Collection
Given the emerging nature of on-demand work and my
interest in theory development, I designed a multiple-
sourced qualitative study and spent four years collect-
ing data. I used three overlapping data sources, which
I triangulated to bolster validity (Eisenhardt 1989): par-
ticipant observation as a driver (n160 hours), partici-
pant observation as a rider (n112 rides), and longitu-
dinal semistructured interviews in 23 North American
cities (n136 interviews). Whereas these three data
sources form the basis of the analysis, I also collected
data from social and print media and company
websites.
Participant Observation. To better understand how
drivers constructed meaning at work, I participated in
the ridehailing industry as both a driver and a rider.
From 2016 to 2019, I worked as a driver in a major east-
coast U.S. city using both my personal car and a rental
car. I worked in several week-long bursts and varied
my driving times and routes to widen my range of ex-
periences: I drove the weekday morning commute
and the evening bar shift, timed my airport runs with
international flight arrivals, visited higher and lower in-
come neighborhoods, and worked major holidays, in-
cluding two New Year’sEves(thebusiestdayofthe
year). I also conducted mini experiments on myself:
Some days I tried to maximize my income by chasing
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 5
surges and bonuses, whereas other times I purposeful-
ly ignored surges and did not check my earnings until
the day’s end. Sometimes I manipulated the app, hop-
ing to confine my trips to a certain area, whereas other
times I let the app decide. To gain perspective on driv-
ers’experiences in different areas, I enlisted a research
assistant to drive in a Midwest U.S. city. Our ethno-
graphic notes included reflections on work perfor-
mance, busyness, ratings, surge pricing and bonuses,
pay, interactions with support services, breakdowns,
accidents, car care, and weather, traffic, and road con-
ditions. I also attended classes on defensive driving
and on my legal rights as a driver hosted by an orga-
nizing group. As a rider during the same time period,
I kept notes on nearly all rides taken (n112). Some of
these rides were personal, and some were specifically
for the sake of this research, such as an afternoon
taking rides around an unfamiliar city. Field notes in-
cluded how I hailed the ride, the car’scondition,app
malfunctions, and overall impressions of the ride in-
cludingmyrating.
Semistructured Interviews. To gain a broad under-
standing of ridehailing, in my first round of data col-
lection, I conducted 63 semistructured interviews
with drivers in 23 North American cities. Questions
focused on the individual level (e.g., why workers be-
gan driving and what a typical day looked like), the
situational level (e.g., interactions with customers and
the app), and the work system level (i.e., the social-
technical elements of RideHail’s business model and
the gig economy). In the second round (n44, 76% of
Round 1), which was conducted 18–24 months after
the initial interviews, I asked workers to describe any
changes since our last interview (e.g., schedule
changes or app updates) and their current financial
situation, and I followed up on themes that were not
addressed in the first interview. In the third round
(n36, 82% of Round 2), I focused on how drivers’
work lives were changing, with a focus on the
COVID-19 pandemic, and how drivers navigated and
solved problems (e.g., with customers, the app, the
platform company). The majority of interview data in
this paper is from the first and second rounds of data
collection. All interviews except one were conducted
in English, and all interviews except 10 were profes-
sionally transcribed.
4
In total, I conducted 136 inter-
views with 63 drivers, of which 19 (30%) were female.
Fifty (70%) reported driving as their primary source
of income, and all except one reported driving to
meet essential household expenses such as utilities.
Twenty-four drivers (38%) were active on at least two
apps, though not all participants lived in cities with
multiple ride-hailing companies. At the time of the
first interview, the amount of time that drivers had
been employed in the industry ranged from two
weeks (10 rides) to seven years (18,000 rides), with my
sample averaging drivers who had 14 months of driv-
ing and 1,800 trips completed, with a 4.87/5.0 rating.
Most drivers saw work as a means to earn a living,
and earning money was a common expressed motiva-
tion, with the exception of only one person.
5
(See the
online appendix for interviewee details.)
I used several sampling approaches to ensure maxi-
mum variation and participant anonymity. I met
roughly half my informants through hailing—either as
part of my everyday life (e.g., running errands) or
through expeditions to an unfamiliar area. To increase
participant anonymity, I often hailed rides from family
and friends’apps. The other half of my sample was re-
cruited from direct and online advertising (e.g., gas sta-
tions, forums), convenience sampling, and snowball
sampling. Snowball sampling was by far the least
effective method, as most drivers did not know other
drivers. Whenever possible, I tried to oversample on
drivers who were white or female, as these made up
the minority of drivers (Campbell 2018). Lastly, I col-
lected data across multiple cities because ridehailing it-
self as well as new features were introduced at varying
times. Shared rides, a feature that matches drivers with
riders traveling in the same direction, was introduced
in 2014 in San Francisco and, as of early 202, was only
available in larger cities. The sample included drivers
from cities where ridehailing was well established
(Boston, Philadelphia), nascent (Ann Arbor, Missoula),
eventually banned (Austin, Montreal), and facing pres-
sure from unions (New York City, Seattle). In sum, this
data represents a large swath of experiences.
Data Analysis
I analyzed data using a grounded theory approach
(Locke 2001, Charmaz 2006, Corbin and Strauss 2007),
with field observations, interviews, and artifacts col-
lected from drivers as my primary data sources.
Drawing from my ethnographic data, I identified key
features of the work environment (e.g., the app, cus-
tomers) that shaped drivers’experiences at work, and
I considered these the “touchpoints”for my induced
theory on workplace games. Drawing on marketing
literature (e.g., Zomerdijk and Voss 2010, Pantano and
Viassone 2015) that uses the term “touchpoint”as an
analytical tool to think about the customer’s experi-
ence with an organization, in this paper I define a
touchpoint as any “contact or interaction a worker has
with any part of the organization that shapes the
worker experience.”A soldier’s unit members, com-
mander, uniform, and the Soldier’s Creed are all ex-
amples of touchpoints in the military. Although any
workplace has an infinite number of touchpoints, not
all are theoretically salient; in the on-demand work-
place, where there are few traditional scaffolds, cer-
tain touchpoints (i.e., the customer and the app) are
Cameron: Making Out While Driving
6Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
particularly obvious—workers constantly interact with
them and they play an important role in facilitating
meaning-making. Below, I describe and summarize the
touchpoints and show their relationship to the induced
theory of workplace games.
Stage 1: Focused Coding. As already described, my
experiences driving sensitized me to the importance
of touchpoints. My first day driving—in the middle
of a heatwave in Washington, DC—was dramatic.
My overheated phone shut down twice, once mid-
ride, as I had not realized that I needed to buy a
holder and place it in front of my air-conditioning
vent. Befuddled by the app and its buzzing notifica-
tions, I missed my first ride because I dropped my
phone under the seat. Interactions with customers
were mixed. I had an energizing conversation with a
nurse whose daughter was at my alma mater, fol-
lowed by a ride in which I consoled a mother visit-
ing her son in chemotherapy. A few rides later, I was
trying to speak as little as possible, shrinking into
my seat as I drove three large men, one beside me,
down winding back roads against the setting sun.
Interviews confirmed my own experiences, and driv-
ers reported having both positive and negative expe-
riences with standard features in their environment.
Realizing that drivers’experiences centered around
these particular features—in part, because the work
was otherwise so socially vacuous—I created a list of
all the possible points of contacts, or touchpoints, in
the environment. Using touchpoints as my unit of
analysis, I catalogued every participants’comments
about every touchpoint over the three rounds
of interviews.
Stage 2: Axial Coding. In the next stage of analysis,
I began axial coding and iterating between the data
and existing theory to build “a dense texture of re-
lationships”around concepts (Charmaz 2006, p. 60).
Drawing on Corbin and Strauss’s(2007) suggestions
for early-stage coding schemas, I coded for each men-
tion of a touchpoint that corresponded with workers’
thoughts, feelings, or action. From this, I was able to
refine my analysis. With my touchpoint data laid out
neatly in front of me in a mass of index cards, I asked
myself: “What problems are touchpoints allowing
drivers to solve?”I focused on the customer touch-
point first, as the data were especially vivid and var-
ied, with one driver expressing delight at potentially
being forever remembered by a customer (31), in con-
trast to another driver who refused to even look di-
rectly at customers (35). I also noticed how touch-
points were related to one another. For example, a car
might be decorated in a way to spark conversations
with customers (e.g., covered in unicorn decals), and
then drivers would describe frequently checking their
customers ratings. This analysis led me to focus on the
touchpoints of the customer and the app and their re-
lationship to one another.
Stage 3: Theoretical Coding. In the final round of
analysis, theoretical coding, I developed relation-
ships between categories elicited in earlier stages to
“weave the fractured story back together”(Charmaz
2006,p.63).Startingwiththeconnectionbetween
the customer and the app, I went back to the litera-
ture for a theory to explain my observed phenome-
na. I found the literature on the sociology of work,
especially on workplace games, fruitful, as games
had been documented in many industries. After
identifying the building blocks of games (rules, feed-
back, variable outcomes), I reordered my touchpoint
data around these three aspects, paying particular
attention to when the existing theory did and did
not match my data. Moving between analyzing data,
drawing models, and writing memos, I abstracted
from these categories and relationships to identify
two games, ultimately explaining how they differ
from those previously examined and how these par-
ticular games affect work engagement.
Constructing Meaning in the On-Demand
Workplace
I present my findings on how workers make meaning
in the on-demand workplace and how this meaning
affects their engagement in two sections. In the first
section, I describe how interactions with the touch-
points are the building blocks of two distinct games:
the relational game and the efficiency game. The touch-
points of the customer and the app are central to both
games, but how workers interact with the touchpoints
and how they understand their relation to the game is
different in each game. In the second section, I discuss
the divergent implications of the two games and how
workers see themselves in either an amicable or ad-
versarial relationship toward the digital platform and
RideHail.
Engaging in the Work: Playing and
Winning Relational Games
The goal of the relational game is to connect with cus-
tomers and provide good customer service, and thus
drivers seek to curate positive service encounters. In
playing the relational game, drivers personalize their
interactions with customers by extending physical
and emotional support and using items in the car to
generate connection. Drivers then track the outcomes
of their efforts through the app’s rating system. In
part because workers are satisfied, in that they can
track their success via the ratings algorithm, they tend
to follow the algorithm’s other suggestions, seeing the
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 7
algorithmic management system as working on their
behalf. Below, I describe the two common ways that
drivers personalized service encounters (through be-
haviors and through the car itself), and then demon-
strate how drivers used the app for both immediate
and long-term feedback on their customer service.
Customers as an On-Demand Touchpoint:
Personalizing Service Encounters By
Extending Physical and Emotional Support. Drivers
personalized their service encounters with customers
by providing physical and emotional support creat-
ing, their own rules for how to connect with custom-
ers. Some drivers took note of customer details to
spark a conversation, such as a 76ers cap or Macy’s
shopping bags. Others kept small items such as mints
in the car for customers. One driver (31) kept three
chargers, water, an umbrella, two blankets, and gum
“just in case”(see Image 2 in the online appendix).
I was working football last Saturday, and two guys
got in the car late, and we were sitting in a traffic jam,
and the guy was like, “I’mdying,myGod,I’mso
thirsty.”Iwaslike,“There’s water in the back. Go
grab one,”and you would have thought they won a
pot of gold. They each had a water in their hand. And
they’re like, “God, I’mstarving,”and I’mlike,“Hey,
here’s some summer sausage, you guys should eat it,”
and they’re like, “Oh my gosh, you’re the best driver
ever!”[laughs]. You never know. It’s a dollar water
and a dollar sausage, and it made their day. They’re
going to go back and remember me and talk about
me for the rest of their lives. They won’tevenknow
who I am. (31)
By providing extra amenities to customers in
need, the driver seriously joked that he would be
forever immortalized by some customers, which was
also a source of professional pride. In addition to
providing physical comforts, drivers also offered
emotional support. As popularized on the HBO
show Taxicab Confessions,theclosequartersofthe
car and the transient relation with the driver make
the ride an easy setting for intimate sharing, with
drivers often stepping into the role of counselor and
confidant. Describing how she focused on uplifting
customers who were upset, a driver said, “You
know how Dr. Phil does 5-minute cures. I call mine
7-minute interventions. I let people vent. Your boy-
friend broke up with you? I tell you to find another
one! I make ’em laugh”(27). Emphasizing his empa-
thy and attuned listening another driver noted, “I’m
more of a listener, like a bartender. Sometimes
you’retheirpsychologistortheirsoundingboard.
I’m pretty good at being able to ask them questions
about what they do or what they’re interested in,
and then people will just go”(63). In more emotion-
ally charged situations, drivers may stop the ride to
console customers, such as when they share details
about experiencing addiction or domestic violence.
A driver explained helping a near-suicidal customer:
I look at them and try to talk with them. He was
about to cry and a little bit intoxicated. He told me he
doesn’t know—for this week, everything he touched
turned bad. He lost his job. He’s writing a book, and
he doesn’t know why. When he got home, he was
about to kill himself, just the end of his life. And I
talked to him, I told him a little bit about myself and
how in our lives sometimes things don’t go the way
you want—it’s just a chapter of a book, and you have
to go through anyway. He said thank you and that he
wished that he met me before. . . . Before he got out
of the car, he just shook my hand, said thank you to
me again. It was a good ride for me. (24)
In these situations, the rule is to prioritize the needs
of the customer, even if this is not to a driver’s imme-
diate financial or time benefit. Though these practices
are labor-intensive, requiring drivers to attune to a
rider’s emotions and respond appropriately, they sug-
gest that providing personalized service is a key com-
ponent of how some drivers see their role.
Tips were not mentioned as a motivating factor for
playing the relational game, and, indeed, those play-
ing the relational game did not report earning more in
tips than those not playing the game. Instead, beliefs
about customers—that they were “professional”(3)
and “high class people”(51)—predisposed drivers to
be friendly and open. Purposefully displaying a
friendly demeanor, a driver said, “Getting in a Ride-
Hail should always feel like you’re getting a ride by
your cousin’s friend to the airport. Why would you
try to do anything other than to make them feel like
they’re your buddy?”(18). In another illustration of
hospitality, a driver said, “I’m an ambassador of this
town . . . whether it’s the football team or the hockey
team or the student service center or this restaurant,
the people just love that stuff”(31). Once in the car,
drivers and riders swapped stories and discovered
mutual interests, sometimes even spending time to-
gether later at bars, casinos, or sporting events. A
driver described how a musical connection with a rid-
er blossomed into friendship: “I met this girl who was
also in a band, and she invited me to come to one of
her shows. I did, and their band is totally kick ass.
I’ve gone to fiveoftheirshows....IfeellikeImadea
buddy—I see her at [other] shows all the time, and we
always catch up”(50). Learning from one another and
finding shared hobbies strengthened the rapport be-
tween drivers and customers, making the work more
pleasurable and meaningful. In addition to develop-
ing personal relationships, several drivers reported
finding professional contacts (e.g., termite inspector,
babysitter, and book agent) from riders, and one driv-
er even found a customer an internship.
Cameron: Making Out While Driving
8Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
Using the Car to Vibe with Customers. Using the car
itself, or props within the car, was another way driv-
ers fostered connections with customers. In Ameri-
can culture, the car is an extension of the self and a
means of expressing personal and/or group identity
(Berger 2001). Popular media outlets, such as MTV’s
Pimp My Ride,Lowrider magazine, and Instagram
feeds dedicated to “VanLife,”showcase customized
cars that symbolize self-expression and individual-
ism. In ridehailing, drivers create their own space
with props to draw in customers. A car decorated
with unicorn decals and lights prompted customers
to squeal, “This looks just like my bedroom when I
was a kid!”(26;seeImage4intheonlineappendix).
On a Friday night, I got a ride with a so-called “party
car”:
The car was popping! Fairy lights on the floor. Tinsel
garlands on the backseat. A snow globe and glow
sticks on the dashboard. Top 40 music. I’d never seen
anything like it—“It’s a party car. I do it on Friday
and Saturday—it’s a hit on South campus [fraternity
row].”Seems the car has quite a reputation. On the
drive, he told me several stories about how excited
students were when they got inside and realized they
got the party car. (Field notes, September 2018; see
Image 5 in the online appendix)
In each of these examples, the decorations sparked
conversation, signaling that drivers were open to en-
gaging with customers—so much so that they spent
time and money to enhance customers’experiences.
Music was another way to foster connection with
customers. Music can be a powerful tool for tran-
scendence, making it possible for individuals to
communicate across gender, race, class, nationality,
and even language (Juslin and Sloboda 2011), and it
often spurred deeper, more meaningful exchanges
during a ride. I was surprised by a black driver’s
socially conscious music choices, including Tupac
Shakur’s“Brenda’sHavingaBaby,”asongabouta
12-year-old’s rape, and her pregnancy, homeless-
ness, and attempts at selling drugs and her body be-
fore being murdered. In a follow-on interview, the
driver provided more detail:
Driving started getting easy after I figured out what
music people liked, so we could all vibe together.
Honestly, I didn’t know what white people liked
[laughs]. I figured black people love Drake, so white
people might like Drake, so I played Drake’sViews al-
bum. The black community is starting to wake up
[become socially aware], so I’m going to play an artist
that’s a little woke. I played Chance the Rapper, and
a lot of people started vibing, and that’s when I really
started getting more comfortable. . . . [My favorite
thing is] teaching [white] people about black peo-
ple—you can’t do that at the workplace because peo-
ple get uncomfortable. (1)
SimilartoDriver1’ssocialawarenesscampaign
through music, other black drivers raised awareness
by sharing memorial cards featuring victims of po-
lice brutality (see Image 6 in the online appendix).
Drivers also used their props more generally to
spark conversations and create connections and had
items in the car such as pamphlets, flyers, products,
and, in one car, a dog that “hangs out and loves peo-
ple”(18). As a rider-cum-participant observer, I met
a gubernatorial candidate who handed and then
read me a copy of his stump speech, an author who
pitched his latest book, a fundraiser who solicited
donations for his after-school tutoring program, and
an energy healer who explained the metaphysical
purpose of the crystals and Tarot cards scattered
aroundthecar(seeImage7intheonlineappendix).
Music and these props turned the car from a space
that customers passed through to a place in which
rich social exchanges occurred. In summary, the
willingness of these drivers to offer emotional sup-
port and to have useful and/or creative items on
hand (charger, healing crystals) reflects how they
proactively sought to meet their goal of providing
good customer service.
The App as an On-Demand Touchpoint: Monitor-
ing and Following the Algorithm
Monitoring Customer Feedback. The app, the second
touchpoint, provided drivers with immediate feed-
back on how well they are doing at providing custom-
er service through the ratings algorithms. Scrolling
through customers’comments on the app, drivers can
see their overall rating and reminisce about specific
rides. Noting how the app reminded her of one such
positive encounter, a driver said, “One customer left a
nice comment because she was running for her train
and I recall making some moves to get her back in
time. She actually wrote and told me how much she
appreciated it. That’s always nice that people just real-
ly acknowledge and appreciate your customer serv-
ice”(28). During interviews, drivers often pulled out
their phones to show me their ratings or a recent cus-
tomer compliment. “Look—I have a ton of these. Peo-
ple love me,”a driver (32) exclaimed, looking at his
phone and scrolling through over 50 compliments, in-
cluding “Best driver ever! Driver is a real jokester and
made my ride to the airport fun”;“Cool car and good
music”;“Made my day—great conversation.”Others
kept memory books of past rides: “Oh, I remember
you!”one driver exclaimed, flipping through a spiral
notebook with more than 20,000 rides. “Look!”He
points to an entry. “You’re in here too—PhD student
at UM studying RideHail”(field notes, March 2017).
In another example, a driver described how she
checks the app repeatedly to remind herself of her
customer service skills:
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 9
I have all of the reviews to prove [I am a good driv-
er]. I can go on there and [see], “I just loved the con-
versation. Thank you for the ride. You put me in a
good mood.”“Your car is so clean. Your car smells
good; you’re a sweet person.”All of this stuff, it’s
wonderful. It makes you feel good and want to do
better. I always look at everything because I play
with my app a lot. I'll go in it and look at different
things or look at my ratings, see if it’s still a 4.86. I
just go in the app and touch all over it. It gives you
that motivation to continue. (33)
The app is a tangible reminder of a job well done,
and, as an app on one’s phone is always available, al-
lowing for an instant emotional boost. Many drivers
reported checking their apps constantly, even outside
of work shifts. As one driver reported, I “check it
more than [I] drive”(39).
Ratings were another indicator that drivers had suc-
cessfully managed interactions with customers. With-
out exception, the drivers interviewed rattled off their
rating to the hundredth-decimal point. Mine was 4.86.
Though only a rating in the mid 4’s was required by
RideHail, many drivers become obsessed about main-
taining near-perfect ratings, emphasizing their en-
gagement in the work. “It’s becoming addicting. . . . I
got an email from the head person in Detroit saying
I’ve made it in the top 10% of drivers in the past
week. My last week was 4.88, and any comments that
were made, [I can] tell you what those comments
were. Over the course of 13 months, I’ve only had one
negative feedback”(31). Returning after almost a
year-long break, the driver remained committed to
maintaining a near-perfect rating: “Ever since I’ve
been back on, I haven’t had anything less than a five-
star rating—it might even go up to 4.96 very shortly.”
The driver also reported his frustration that the app
didn’t provide more specific feedback to improve his
rating. Just for fun, another driver spent two months
trying to increase his rating from a 4.8 to a 4.9, saying
he “absolutely watched the app all the time”(3).
In addition to monitoring customer feedback, driv-
ers can also review their telemetric scores on accelera-
tion and deceleration to verify that customers had a
smooth riding experience. One driver noted, for ex-
ample, “It [the app] reports every day how smooth
you brake. The ratio is supposed to be—you can see
on the app—3:25, a smooth brake. I have a 3:23. . . . [It
signals that] the passenger can be comfortable with
you (26).”Overall, drivers saw high ratings and tele-
metrics as indicators of successfully managing the ser-
vice encounter, and constant feedback from the app
motivated them to continue exerting this effort.
Following the Benevolent Algorithm. In addition to
tracking their ratings on the app, when playing the
relational game, drivers usually followed the app’s
nudges on when and where to drive, trusting that
the nudges were aligned with their own interests—
even though they were unclear about how the algo-
rithms underlying the nudges worked. One driver
observed about the ranking system, “They [the app]
say I made the top 10 percent, but it doesn’ttellyou
in what. Maybe it’s the amount of hours you’re on
theroad.Maybeit’s the number of five-star com-
ments you get. I don’tknow.IhonesttoGoddon’t
know”(3). Unlike the Old Testament’sGodofwrath,
the algorithm was described more like the New Tes-
tament’s God of mercy in that drivers believed it al-
ways assigned a beneficial combination of rides: “I
really believe it’sGodandthealgorithm.Idon’t
know how it works, not at all . . . [but] at the end
of the day, it just works out. It’s really weird”(23).
Making money was easy because the algorithm was
“always pumping (42),”and even when drivers mis-
judged optimal driving times, they earned enough.
I call it lucky or blessed, but it seems when I start
late, I’ll get trips that are worth more money. I can’t
say that I can aim for that. . . . I get lucky all the time.
I catch a surge, a big surge halfway across town, and
then I catch another surge back across town, and then
I’ll be right back to the money where I would have
been working that whole morning. I say it’s a groove
because I keep getting that same luck [laughs]. I know
what it takes to get my 200 bucks a day. (19)
Calling himself blessed, the driver attributed his
success to an algorithm that helps him “get lucky”
with surges, later comparing the algorithm’s matching
skills to sex, noting that “it’s always good.”Another
driver (32) explained how the algorithm always as-
signed him interesting riders, including a local artist
and a popular athlete.
Drivers who used the app to confirm their good
customer service skills tended not to question the
app’s nudges around matching and incentives. They
trusted that the algorithmic management system was
aligned with their own interests. Describing the algo-
rithms as “fantastic”(50), “fair”(21), and “good”(63),
drivers had no reason not to follow the nudges around
matching and incentives. When logging on, drivers
would “check on the app first for any surging
demand”(13), because “RideHail has figured all that
stuff out”(16), and they would generally “accept
98%–99% of rides”(32). In describing a typical day, a
driver said “If I’m just starting out for the day, I pretty
much press yes to everything. . . . So right at the be-
ginning, I’ll say all right, let me just go”(13). Even
when drivers learned that the app presents varying
incentives to drivers, they seemed unconcerned about
possible inequity, emphasizing that they were earning
enough. “No. It doesn’t bother me. . . . I usually finish
[complete the incentive] . . . and then they give me
Cameron: Making Out While Driving
10 Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
[pay me] what they say”(6). In summary, when play-
ing the relational game, drivers were uncritical about
the matching and incentive algorithms, in part, be-
cause they believed that the app’s nudges were in
alignment with their own interests.
Summary: Rules, Feedback, and Wins in the Relational
Game. In the relational game, the primary goal of
drivers was to create connections with customers and
provide good customer service. To meet this goal,
drivers managed customers by curating positive ser-
vice encounters. What drivers classified as a “win”in
the relational game overlapped with RideHail’s goal
of customer satisfaction, such that the digital platform
supported the relational game. Ratings, compliments,
andbadgesprovidedworkerswithfeedback,confirming
that their efforts were valued and that they were, indeed,
“winning.”Drivers’actions constituted a game, as op-
posed to mere reaction to an organization’s attempts at
gamification, because workers went “above and be-
yond”the suggestions laid out in community guidelines.
There is no concrete benefit to drivers for increasing their
rating from a 4.8 to a 4.9; and decorating the car or pro-
viding water costs drivers’time, energy, and money. As
part of the relational game, once drivers confirmed that
they were winning, they paid less attention to other com-
ponents of the algorithmic management systems follow-
ing without hesitation the nudges from the matching
and incentive algorithms. As a result, drivers in the rela-
tional game viewed the digital platform as a facilitator,
and they described having an amicable relationship with
RideHail. Overall, drivers viewed themselves as repeat-
edly “winning”in a system that was working to
their advantage.
Engaging in the Work: Playing and
Winning Efficiency Games
The goal of the efficiency game is to minimize time
driving while maximizing income, and thus drivers
focus on completing rides quickly and efficiently. To
play the efficiency game, drivers depersonalize their
encounters with customers through emphasizing so-
cial and physical boundaries with customers, which
minimizes the length and depth of their interactions.
Unlike the relational game, in the efficiency game,
drivers are unable to track their wins on the app, and
they are constantly unsure whether they have been
assigned the most lucrative rides by the algorithm.
Because of this uncertainty about where they stand in
the game, drivers often take matters into their own
hands, keeping their own financial logs and at times
even countering the matching algorithm in order to
receive more lucrative rides. Below, I describe driv-
ers’strategies to create social and physical bound-
aries with customers and then consider how the app
keeps drivers from verifying their position in the effi-
ciency game.
The Customer as an On-Demand Touchpoint:
Depersonalizing Service Encounters
Enforcing Social Boundaries. In the efficiency game,
the goal is to get customers to their destination as
quickly as possible in order to maximize earnings,
and, therefore, drivers enforced social boundaries to
minimize the length and depth of their exchanges
with customers. Noting how her quest for efficiency
underscored her interactions with customers, a driver
said, “All I know is that I need to get you to your des-
tination ’cause that’s my mindset. Get you to your
destination in a fast, safe way so I can get my money
and you can get out of my car”(33). At best, custom-
ers are seen as faceless fares that need transport and,
at worst, as self-centered monsters that need to be de-
fended against. A driver described how he avoids en-
gaging with riders:
Twenty to thirty percent of people are nice, but I’m
not trying to establish a personal relationship. This
is a taxi. I get you where you need to go and go
about my business. I’ll talk, but I’m not trying to get
to know you. I don’tdomuchifI’m not getting paid
for it. Not going out of my way to help you. People
will take advantage of you. It’s the ones who do
three-minute rides and try to take two or three
treats. It’s one treat or get the hell out of here! People
get crazy, greedy, take more than they should.
You’re dealing with random people you don’teven
know, and you’re sharing your space with them
[laughs]....Afterawhile,mentally,youblackitout.
I’m not trying to develop a full-blown relationship,
hang out with them on the weekend, pow wow, and
all that stuff. It’sjustaride,andthat’sit[laughs]. Get
the fuck out. (35)
Believing that most customers are unpleasant, un-
appreciative, and sometimes downright greedy, driv-
ers merely tolerated their presence. By avoiding eye
contact, eschewing conversations, and refusing to of-
fer help, drivers created self-protective boundaries
that limited interactions, and even saw offering emo-
tional support as dangerous. A driver described his
defensive tactics, “I just be taking them where they
want to go. I had about five or six ladies in my car cry-
ing. You think you’re helping somebody and then you
get an email from RideHail saying you hurt them or
you harassed them”(44). Another strategy to avoid
talking to customers was to pretend they did not un-
derstand them when they reported problems. A driv-
er said, “If I see you’re about to say that you are not
happy, I just try to change the subject or talk about
something else. Or I’ll pretend I didn’t hear what you
said or that I just don’t understand and everything is
okay. When they get out, they’re like, ‘Bye, have a
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 11
nice day!’” (12). These drivers compartmentalized their
behavior, “keeping it business”(8) and “professional”
(3), making it clear that connecting with customers was
not their goal. Instead, their behaviors were focused on
minimizing social interactions so they could complete
their rides as quickly as possible.
Enforcing Physical Boundaries. Drivers enforced phys-
ical boundaries to protect themselves and their property,
further highlighting their desire to be efficient with their
time and energy. A common way to create this boundary,
and one that I frequently employed myself as a driver,
was to place a bag on the front seat in the hope of avoid-
ing the emotional labor of small talk and the physical
intrusion of having someone sitting so close (see Image 5
in the online appendix). More generally, drivers felt that
customers disrespected their property: “People forget it’s
my vehicle. They put feet on seats, fart in the car, do all
this weird stuff. They mess up your door handles. You
know they’ve got food on their hands. These people
don’tcare!”(35). To counter rider’s potential disrespect,
drivers prominently placed notices that customers were
being recorded by dashcams, and signs requesting riders
to wipe their feet and refrain from eating, drinking, and
slamming doors around their car to remind customers
to behave.
“What’s that?”I asked, pointing to a dry erase board
in the seat pocket. [The driver] quickly went into a
long rant, her voice getting louder and angrier. “I had
to make it ’cause people were misbehaving. They tri-
fling. This one woman spilled coffee all over my
backseat—and didn’t even say sorry!”[Swings arms
open. I get nervous as she’s driving.]“So, I made the
sign. But then people started touching the sign. And
messing it up. No.[Voice gets louder.] You don’t need
to touch the sign to read it, so I had to make a new
sign.”[The first line is “Don’t touch the sign.”]“Is it
okay if I touch the sign?”“Yeah, but don’t mess it
up.”(Field notes, September 2019; see Image 7 in the
online appendix)
Physical objects helped to enforce boundaries of ac-
ceptable and unacceptable behaviors in the car. In ad-
dition, drivers felt comfortable refusing customers’re-
quests that would decrease the profitability of a ride.
Taking customers by fast-food outlets was especially
problematic, as drivers not only lost time, but risked
incurring additional cleaning costs if customers
spilled anything. A driver described how he kept his
car clean and odor-free:
They want to go to McDonald’s and get something to
eat. If somebody gets in the car and says, “Hey, do you
mind going through McDonald’s?”Isay,“Absolutely
not. Why would I go through McDonald’s?”And they
say, “Well, I have other drivers that do it for me.”Isay,
“Well, if other drivers like to wait 15, 20 minutes sitting
in a drive-thru, that’sonthem.”Idon’twanttomake
17 cents a minute and drive you a mile down the road
and have my car smell like McDonald’s and have
you sitting back, eating fries, making my car smell like
fries. (61)
Similar to not comforting crying passengers for fear
that their help could be misinterpreted, drivers pro-
tected themselves from potential liabilities by not of-
fering more personalized services. A driver said:
I try not to overstep any boundaries. These people
are strangers. If it wasn’tfortheapp,yourproblem
wouldn’tbemine.Ihavepeoplecominguptothe
car, and they got real [a lot] luggage, and I say,
“I’m sorry, I’d really like to help you, but I can’t
touch your luggage.”You want my help now, but
the moment one of those straps breaks, you’re gon-
na be writing a complaint, and I’m gonna be respon-
sible. So I got to respect myself. I gotta respect my
boundaries. (44)
By not providing additional services, drivers pre-
served both their time and their energy, allowing
them to devote more energy toward driving. In the ef-
ficiency game, boundaries helped to limit interactions
that could be time-intensive, emotionally draining,
and financially unbeneficial. When drivers enforced
these boundaries, they asserted control over their
work and affirmed that their primary goal was simply
to get riders to their destination fast.
The App as an On-Demand Touchpoint: Monitor-
ing and Countering the Algorithm
Monitoring Financial Trackers. When playing the effi-
ciency game, drivers focused their efforts on making
the most amount of money in the shortest time possi-
ble; achieving this goal was complicated by the fact
that drivers did not trust the platform’s matching al-
gorithms and the incentive notices. Drivers described
being bombarded by alerts; receiving up to three per
day, one driver wondered, “Why is RideHail sending
me alerts for surge pricing? I never walked out of the
house or jumped out of the bed because of the surge”
(46). Similarly, one weekend I noted, “I really want
to rest . . . and RideHail has been sending me damn
[incentives] all week. Fine, I’ll drive”(field notes, Jan-
uary 2018). Drivers felt these messages were mislead-
ing, because often when they entered an area, surges
would disappear. Some drivers ultimately chose to ig-
nore incoming alerts—for example, “[I] don’t chase
the surge”(15); “I’m too smart to chase surges”(9)—
and some even drove in the opposite direction—for
example, “I never go where there is a surge”(29); “I
don’t go to them, I go away from them”(32). This
stance toward alerts sometimes led to a blanket disre-
gard of all messages as drivers chose to “ignore all the
texts and emails”(32). Instead, to get the best rides,
drivers positioned themselves in busy areas, scoured
Cameron: Making Out While Driving
12 Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
newspapers for large events, or drove around wealthy
neighborhoods in the early morning for the lucrative
airport pick-up. Yet even the savviest drivers who an-
alyzed local events and traffic patterns to find the
most efficient times and places to drive were frustrat-
ed, noting that driving for RideHail was “far more
complicated than being a taxi driver”(20).
Drivers also did not believe that the app accurately
accounted for their earnings and worried that it short-
changed them of their promised fares. To compensate,
drivers would compulsively check their app to verify
if they were paid correctly. A driver said, “I’m check-
ing my trips constantly. . . . Every time I complete a
trip, I look to make sure that I got paid. I see what my
money adds up to. [And] I call if there is a delay . . . in
the information getting to my phone”(19). Sending
“Correct my Fare”requests to RideHail was so rou-
tine that I simply logged it as “No boost [incentive]”
in my field notes. I wrote “I’ve gotten so many fares
incorrect. I normally just notice if I get [an incentive or
not], but sometimes it may not be calculated correct-
ly—especially when it’s a range [e.g., 1.6–1.9 surge].
How am I supposed to remember and keep track of
that?? So many ways that [we] could be cheating us!!
[emphasis in original]”(field notes, February 2017).
Not only did the app occasionally miscalculate fares,
it also did not reflect the true cost of driving, as it did
not account for gas, wait time, and car depreciation.
Drivers therefore tended to take matters into their
own hands, designing their own tools, such as Excel
spreadsheets and longhand ledgers, to accurately re-
cord their income, incentives, and expenses (see Im-
ages 8 and 9 in the online appendix). More than half
of my nearly six hours of interviews with one driver
focused on his extensive tracking system, as he de-
scribed his “really religious”accounting system. His
goal? To be a “savvy”driver with “no tax liability,”
because “RideHail doesn’t care about their drivers . . .
and the key to rideshare is to get what you can out of
them”(9).
Unlike in the relational game, in the efficiency game
drivers were not focused on customer ratings; instead,
drivers described ratings as another indicator of the
algorithmic management system’s crookedness. Frus-
trated about being blamed by customers for things
outside of their control, drivers said that the ratings
system was “unfair”(57), a “joke”(44), “Not right.
Not fair”(26), and, more generally, “sucked”(43).
Workers rejected the app’s attempt to gamify ratings,
such as the nonmonetary rewards (e.g., badges) for
high ratings. Emphasizing her alignment with the effi-
ciency game, and mocking Uber’s badges, one driver
said, “It’s like ‘Mommy! Mommy!! I got a badge. I got
a cool car badge, I’m cool.’It just bothers me. Give me
cash, I don’t want no stinking badges”(26). In sum-
mary, when playing the efficiency game, drivers
mistrusted the incentive information presented by the
app and instead devised their own methods to find
the most lucrative rides and track their earnings.
Countering the Malevolent Algorithm. Drivers were
regularly able to finesse their interactions with cus-
tomers to complete rides quickly but, as they were de-
pendent on the matching algorithm to assign and
price rides, they often resorted to manipulating the al-
gorithm to obtain desired rides. Enraged at not receiv-
ing enough rides during high-demand times, one
driver said, “[I] know exactly who is behind the algo-
rithm’s decisions, and it’s not God. The machine and
the software are set up by people, by humans. God??
No! Just humans made it”(40). Drivers frequently
claimed that the matching algorithm was not assign-
ing the best rides and that, indeed, the algorithm was
“out to get them.”
I swear there was a conspiracy because in the after-
noons—I logged in every single day at 4:00—I would
get a long ride that would take me out of the city in
the opposite direction towards the airport. And then,
right after I get out of the city, the city would light up
like a Christmas tree [on the heatmap display]! I
swore it was a conspiracy against me because I did
very little, if any, prime-time rides. Because I’m al-
ways sent in the opposite direction. (42)
Misleading projected wait times, not being assigned
another ride immediately after a high-paying one, and
long wait times between rides led drivers to believe
that they were being personally targeted. A driver de-
scribes how the algorithm “forced”him to work lon-
ger as he neared an incentive target:
I’ve had that weird thing happen outside a building
and didn’t get rides. I know what they’re doing. It’s
not rocket science. If a driver gets to 67 rides, and
they’ve got three to go [to reach an incentive], why
would you want them to get in the queue quicker
than everyone else? So, if I need three more rides, it’s
4:00 in the morning, I’ll get a ride in Bethesda, which
is 50 minutes [long], and then I’ll get a ride that’s an-
other hour. Well, all that time, I’m looking at the
clock going I almost don’t want to take this ride be-
cause I need to get quick [short] trips, so I can just get
the bonus. When you’re two rides away from the bo-
nus, you don’t care about [the fare]. You just want to
get the ride, so you can get that extra hundred bucks,
or 80 bucks, whatever it is. But I think [RideHail]
knows what they’re doing—they have it programmed
in to make it as hard as possible when a driver is
pushing the edge of the envelope to [get the incen-
tive]. (61)
The driver perceived the algorithm as penalizing
him, because it “knew”he was close to his target and
assigned longer rides to get more work out of him. In
contrast to viewing the algorithm as unknowable, as
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 13
in the relational game, here drivers saw the intention
of the algorithm as clear—to keep drivers on the road
longer at less pay, which was in direct conflict with
the goals of the efficiency game.
Because drivers believed the algorithm was work-
ing against them, they deployed tactics to counter
some of the algorithm’s decisions, thus reasserting
control in their work. Creating rules about which rides
were more profitable, drivers rejected rides they be-
lieved would pay less, such as those that suggested
that customers would not be traveling far (e.g., rides
starting at grocery stores). Many avoided unpaid trav-
el time: “If they want me to go somewhere that’stoo
far away from where I want . . . and it’s gonna take
me 30 minutes to go pick somebody up, or it’s rush
hour, I’ll usually decline—it’s not worth it”(53). An-
other driver declined anything “more than 10 mi-
nutes—it’s just too far”(46). If drivers declined too
many rides, they risked being temporarily suspended
or permanently deactivated.
A quest for efficiency also affected how workers
viewed the driver support hotline. The support phone
number was hard to find and frequently out of ser-
vice; conversations with support staff were frustrating
and often not helpful. Similar to creating expense
trackers, many drivers kept detailed logs of their
conversations with support to be sure issues were
handled. My interactions with support were equally
aggravating. After my first fare miscalculation, I spent
twenty minutes trying and failing to find their num-
ber. Taking matters into my own hands, I selected “I
have a safety issue”on the app to get someone to call
me back regardless of whether I had a safety issue. In
summary, in the efficiency game the algorithm was
perceived as not being aligned with drivers’inter-
ests—or even as conspiring against drivers—and thus
drivers deployed countering tactics to receive the
most lucrative rides.
Summary: Rules, Feedback, and Wins in the Efficiency
Game. In the efficiency game, the goal of drivers
was to complete their rides as quickly as possible,
ideally at the highest fare. To do so, drivers man-
aged customers by depersonalizing the service en-
counter, minimizing conversation, and not offering
additional services, such as carrying bags. Unlike in
the relational game, however, what drivers classified
as a “win”in the efficiency game did not overlap
with RideHail’s organizational goals; accordingly,
the digital platform was not designed in such a way
that it supported the efficiency game. Drivers be-
lieved that the app’strackersdidnotprovideaccu-
rate feedback, leading them to create their own tools
to track their efficiency. At times, drivers even ma-
nipulated the algorithm. As a result, drivers in the
efficiency game viewed the digital platform as an
opponent and described themselves in an adversari-
al relationship with RideHail. They viewed them-
selves as repeatedly trying to “win”in a system in
which they were disadvantaged.
Relating to One’s Work: Amicable and
Adversarial Stances
The two workplace games were associated with
different stances—that is, with how workers under-
stood their relationship to RideHail. The games dif-
fered in whether drivers knew for sure that they had
won, and the certainty or doubt of their position in
the game was in turn associated with their stance.
When playing the relational game, drivers were cer-
tain of their position in the game, as success could be
tracked via the app’s ratings system, which was as-
sociated with drivers seeing themselves as winning
and in a mutually beneficial relationship with Ride-
Hail. In contrast, when playing the efficacy game,
drivers were uncertain about their position in the
game, given that they could not track it via the app,
which was associated with drivers seeing them-
selves as losing and in an exploitative relationship
with RideHail. In the remainder of the findings, I de-
velop these stances and describe their relationship to
workplace games.
An Amicable Stance: Certainty and the Winning
Self
When playing the relational game, drivers could be
sure a win was a win—five-star ratings, compliments,
and badges confirmed that drivers were connecting
with customers and providing good customer service.
In accumulating “wins,”drivers saw themselves as
enjoying their work and as competent in it, and be-
lieved they were beneficiaries of the on-demand econ-
omy. Noting the enjoyment of driving, a driver said “I
like driving. This gets me out to see beautiful views of
the city, to hear cool things and cool stories, and meet
people from all around the world. This job is so much
freedom”(26). For many drivers who had previously
only worked lower-skilled and/or dangerous jobs,
these benefits were nontrivial. An immigrant from
Southeast Asia described himself as living the Ameri-
can Dream. RideHail was a step up, paying more than
his previous jobs as an overnight gas station attendant
and catering waiter, and helping him to support his
two children in college; it was “very comfortable,”he
said (4). RideHail helped a recent immigrant from
Central Asia to learn American culture and improve
his English. “At first, RideHail was hard [because of
language skills]. I was too tired to study at night [after
driving], but I learned [English] from RideHail—it’s
very good for me! In 2012, I was ESL Level 2. Now
[2016], I’ve jumped 4 or 5 levels”(14). From 2016 to
Cameron: Making Out While Driving
14 Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
2020, I could literally hear the driver’s English contin-
ue to improve; he eventually passed a test to take a
job as a municipal bus driver. Others emphasized
how RideHail “saved[them]when...inafinancial
bind”(17), such as during a layoff, allowing them to
make “good money”(60). A driver with multiple au-
toimmune diseases who had not worked a steady job
in more than 15 years credited RideHail with “building
[herself] back up to a person”(27). In her first year of
driving, she got a secured credit card, then her own car,
and then her own apartment. Others shared their suc-
cess stories with friends and family and recruited them
as drivers, sending texts to their new recruits “to keep
them encouraged”(19). Overall, in the relational game
drivers saw themselves as skillful in completing their
work, which in turn led them to see their relationship
with RideHail as mutually beneficial. Summing up his
relationship with RideHail, one driver said, “I think of
what a great idea, what a great company. It’s a plea-
sure to work with them. To work with [emphasis
added] them, because we’re partners, so I don’tsay
work for them. It’s a great partnership”(19). Optimis-
tic about the future, another cheerfully said, “[Ride-
Hail] is here to stay, and I’ll be doing it until I retire
(62).”He was in his mid-40s and had almost 20 years
until retirement.
An Adversarial Stance: Uncertainty and the
Losing Self
When playing the efficiency game, drivers could never
be sure that they had won—the app did not show if
drivers were being assigned the best rides and did not
accurately track their earnings. Indeed, every ride was
a crapshoot, as customers could be excessively de-
manding and the algorithm could assign a low-paying
ride, so drivers tried to manipulate the system to as-
sign them higher-paying rides. Amid this constant un-
certainty, drivers saw themselves as lacking compe-
tence and losing in the on-demand economy, forced to
do low-paid, low-status, physically punishing, and
sometimes dangerous work. Fluctuating rates were
“just a game and not a cool one either,”because each
time “they drop their price, a little bit of the joy of
driving for RideHail goes out the window”(32).
Trapped in a game they could not win, drivers felt so-
cially stigmatized for being part of the on-demand
economy. One driver said:
It’s not something I like to talk about a lot. . . . In
terms of what I do for preparation, what I do to ana-
lyze areas that are busy, the hours that I schedule it
around. It’s not something I’m proud of either. A
huge, huge part of the gig economy is they take ad-
vantage of people who don’t know any better. A lot
of people that drive aren’t that smart at all, and I
don’t want to be grouped in with the people that
don’t understand the difference. (20)
Many drivers spoke wistfully about when they
could quit and described their workdays as being
trapped on a hamster wheel of rude customers, weak
labor protections, and declining wages. Evoking
Adam Smith, a driver clearly saw himself as losing:
“With the gig economy, there is no future. There are
no illusions of grandeur. It is what it is. People using
the company for money. The company is using people
for money. It’s Adam Smith—the invisible hand. . . .
Our objectives are not the same”(20). And in perhaps
the most extreme metaphor of exploitation and degra-
dation, one driver compared his work to the world’s
oldest profession, in which RideHail “is the pimp, the
riders are the johns, and we just open our legs”(8).
For many who were accustomed to precarious
work, RideHail was simply the latest in a long string
of bad employers that had destroyed their bodies as
they eked out a living. Pushing themselves to exhaus-
tion, drivers complained of mental fatigue, “your
brain is always sharp”(51), dizziness, because the
“eyes [are] working all the time (51)”from “having to
be in a high state of attention”(58), and PTSD-like
symptoms (56). Company policies privileged custom-
ers, even if it meant exposing drivers to increased
physical risk. All drivers, even those with allergies,
were required to transport service animals—a policy
one driver called “horseshit”(31). And drivers, but
not riders, were required to provide evidence of
mask-wearing during the early days of the COVID-19
pandemic, another sign “that they [RideHail] just
don’t care”(41). Overall, when playing the efficiency
game, drivers were not able to see themselves as skill-
ful or successful in the work, as they were painfully
aware of the control wielded by RideHail and its algo-
rithms. This led drivers to describe their relationship
with RideHail as antagonistic and, at times, even de-
structive. Drivers saw themselves stuck in “a money
game, making money for them . . . a cog in a money-
making machine”(18). Deeply pessimistic, they envi-
sioned a bleak future. A driver said, “It’s hard to have
energy for the next month or next year—I am using
my body, my energy. Future is way black for me. I
can’t see anything good in the future in this type of
job”(14).
It is important to note that whereas these findings
show how workers invest in workplace games, this
does not necessarily translate into long-term commit-
ment, as turnover rates are high. Estimates using data
from RideHail and self-reporting from drivers place
annual turnover rates from 40% to 70%, with most
drivers lasting less than six months (Campbell 2018,
Hall and Krueger 2018) lower than this range. More-
over, whereas most drivers described being primarily
engaged with just one game, some drivers would stop
playing a game after an incident. One driver (19) who
was deeply invested in the relational game, calling
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 15
himself a “partner who worked with RideHail,”was
temporarily banned from the platform between our
first and second interview due to a dubious customer
complaint. Even though he was reinstated, he harbored
ill will toward the platform, remarking that he “had fi-
nally realized what type of situation he was in.”He
made no reference to playing either the relational or ef-
ficiency game in that interview, and, soon after being
reinstated, he left the platform entirely. These findings
suggest how important games are for workers to re-
main interested and engaged in their work.
Discussion
This research starts to unravel the puzzle of how
on-demand workers construct meaning in a setting
that scholars theorize is ripe for worker alienation.
Among ridehailing drivers, I found that instead of
experiencing alienation, workers construct meaning
through workplace games that revolve around the
touchpoints of the customer and the app. In the rela-
tional game, workers create meaning by connecting
with customers and crafting positive customer service
encounters, which they quantify and track through
the app’s rating system. In the efficiency game, work-
ers create meaning by completing work quickly at the
highest pay rate, optimizing the ratio of their time to
their pay; but, because drivers are unable to track their
“wins”on the app, they create their own tracking sys-
tems and, at times, manipulate the algorithm. Below, I
discuss how the identification and articulation of the
relational game and the efficiency game offer insights
for the literatures on meaning-making and workplace
games.
Understanding Workplace Games as a Form of
Meaning-Making
How workers make meaning or sense of their work en-
vironments has captured scholars’attention for more
than a century (e.g., Baumeister 1991). Much of the liter-
ature has focused on organizational practices, or scaf-
folds, that foster group belonging (e.g., Carton 2018)and
form an “encapsulated”(Pratt 2000) environment from
which meaning is created and sustained. Indeed, these
scaffolds are seen as so essential that the literature sug-
gests that in settings without them, such as on-demand
work, alienation ensues (e.g., Pratt and Ashforth 2003).
By contrast, I found in ridehailing that workers avoid
becoming alienated from their day-to-day work be-
cause they are able to create meaning through
workplace games. This is surprising, as the litera-
ture typically considers workplace games to be so-
cial games, generated through exchanges between
managers and coworkers (e.g., Sherman 2007). Rec-
ognizing workplace games as a crucial aspect of
how on-demand workers create meaning raises im-
portant insights about the nature of games in the
context of changing work structures and emerging
technologies—and ultimately about how the games
affect workers’engagement, commitment, and re-
tention in the on-demand workplace.
Emphasizing the reinforcing scaffolds that point
workers toward a particular game (“making out”in
manufacturingworkandthe“tipping game “in ser-
vice work; Burawoy 1976,Sherman2007), prior re-
search has consistently identified a single game in an
organizations. By contrast, in on-demand work, the
meaning-making process is fragmented, in part be-
cause emerging technologies do not scaffold the
work in the same way. The digital platform and its
underlying algorithms directs individual workers on
which rides to take and what route to follow but
does not provide much support on how to find pur-
pose in the work beyond the task of driving. Instead,
workers rely on interactions with touchpoints—in
this context, the customer and the app—to derive
their own meaning at work, either by building con-
nections with customers (the relational game) or by
earningasmuchaspossible(theefficiency game).
Two different games are thus played in the same or-
ganization. Whereas both games provide a source of
meaning for workers, how they construct meaning
differs depending on the game. This study high-
lights that in settings where meaning is only loosely
scaffolded by an organization and its technology,
the important questions become what touchpoints (fea-
tures) of the work are most central, how workers inter-
pret and interact with said touchpoints, and what might
indicate the quality of one’sinteractions.
Which game drivers play is significant for under-
standing the kinds of meaning they construct from
their work and how drivers view their relationship
with the platform company. Similar to the tipping
game (e.g., Sherman 2007), when playing the relation-
al game, workers focus on the social elements of the
work, creating meaning by building connections
with customers and providing good customer ser-
vice in order to secure high ratings. However, in the
tipping game, workers learn the rules from cow-
orkers, receive feedback directly from customers,
and jockey among each other for status, whereas, in
ridehailing, drivers learn the rules of the relational
game alone, given the solitary nature of the work.
They receive feedback not from each other, but from
the app's rating system, which allows them to track
“wins”and achieve rewards, such as priority dispatch
at airports. Experiencing the app as successful at quan-
tifying their efforts, drivers see it as a positive aspect
of their experience—that is, the app reflects aspects of
the work that they view as having meaning. The
Cameron: Making Out While Driving
16 Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
digital platform is thus rendered a facilitator in the re-
lational game.
When playing the efficiency game, workers aim to
complete each ride as quickly as possible. In contrast to
the tipping game and other service jobs in which peo-
ple seek to create personalized interactions (e.g.,
Wrzesniewski et al. 2003), in the efficiency game cus-
tomers were viewed as objects to be transported quick-
ly without the pretense of forming a connection. In the
making out game, machine Drivers focus on managing
the technology to optimize the ratio of effort to income,
similar to manufacturing workers operating the ma-
chines in the making-out game. Though piece rates
vary, machine operators know the rates and can adjust
their efforts for each task. In contrast, RideHail drivers
never know what ride the algorithm will assign or at
what fare. This opacity makes it impossible for them to
accurately track their “wins,”leading drivers to create
their own recording systems and, at times, manipulate
the algorithm. Experiencing the app as unsuccessful at
quantifying their efforts, drivers see it as a negative as-
pect of their experience—that is, the app does not sup-
port aspects of the work that these drivers view as hav-
ing meaning. The digital platform is thus rendered an
opponent in the efficiency game.
In a large range of today’s work settings, how work-
ers view the managing technology (as a facilitator or
as an opponent) could be a significant factor in work-
ers’meaning-making and engagement. Ranganathan
and Benson (2020) show in their study on seamstresses
that when electronic counters accurately captured pro-
duction, workers were more likely to embrace the
technology, which, in turn, motivated gamification
and productivity improvements. However, such im-
provements did not occur when work was more
complex and workers could not trust the algorithms’
quantification metrics. Derivative traders did not trust
the algorithms that replaced managers, making it com-
fortable for them to game the algorithm and take larg-
er risks, which contributed to the 2009 mortgage crisis
(Beunza 2019). As organizations are increasingly using
technology to track and rate workers, future research
should continue to explore how emerging technology
extends traditional conceptualizations of meaning-
making and workplace games.
Digital Platforms and the Implications for
Workplace Games
As I have noted, a critical aspect of these games in ri-
dehailing is how drivers view the role of the digital
platform. The notion of being in (or out) of alignment
with the platform is reminiscent of ongoing research
that examines how the configuration of actors in
the service triangle results in different structures that
mobilize shared interests (e.g., Biggart 1989,Macdon-
ald and Siranni 1996,Sallaz2013). Drivers that
described themselves as “winning”in part because
they were aligned with the “right”side of the digital
platform. This relationship echoes the “interest alli-
ance”between sales workers and their employer
when the former readily adopts the organization’s
scripts as a means to increase sales (Leidner 1993). Ex-
tending this analogy, this research suggests that
changes in the algorithmic management system can
support or undermine each game, with divergent impli-
cations for the platform company and its workers.
In the relational game, changes to RideHail’s re-
ward systems can strengthen a relationship that work-
ers already view as an alliance. In June 2017, Uber
launched a “180 Days of Change”campaign to im-
prove the driver experience. Changes included giving
drivers access to even finer-grained measurements of
their service encounters (i.e., badges for being a good
conversationalist). Such changes reinforced the rela-
tional game by amplifying the social aspects of the
work and making it more likely that drivers would
follow the algorithm and continue playing. By con-
trast, in the efficiency game, increasingly opaque algo-
rithms exacerbate a relationship that workers already
viewed as unsupportive. Over the years, changes to
RideHail’s algorithms have led many drivers to be-
lieve that the assignment of rides and offers of incen-
tives are never aligned with their own interests. Even
with financial incentives, drivers found their net pay
declining; they also found it increasingly difficult to
manipulate the algorithm in their favor, such as no
longer being able to match with customers who were
already in their car. Frustrating and unpredictable,
these technical changes undermine the meaning sys-
tems of workers playing the efficiency game and
make it less likely that they will continue to play the
game. The question arises, then, whether workers will
remain committed to an organization in which they
feel constantly disadvantaged. Future research should
continue to explore how consent and commitment are
manufactured with respect to these emerging technol-
ogies and independent contractor work arrangements.
Moreover, this study identifies how workers ac-
count for multiple sources of unpredictability in their
work environment. In studies within traditional or-
ganizations, scholars identified a single source of un-
predictability that workers had to take into account in
predicting their wins (i.e., piece rate, the customer); by
comparison, at RideHail drivers contend with two
sources of unpredictability: the customer and the
app’s algorithms. In both games, workers felt that
they had more control over humans than machines
(successfully managing the customer, but either fol-
lowing or unsuccessfully manipulating the algo-
rithm), pointing to a more inverse relationship be-
tween humans and machines than often considered
(cf. Suchman 1983,2007; Brynjolfsson and McAfee
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 17
2014). It is particularly important for scholars of work
to consider these interactive approaches for managing
unpredictability as work becomes more complex, in-
volving multiple actors and emerging technologies
(Kellogg et al. 2020).
Changes in the contemporary economy, such as the
rise of nonstandard work arrangements (Spreitzer
et al. 2017, Katz and Krueger 2019) and remote work
(Bloom 2020,Rhymer2020), suggest that fragmentary
meaning-making, divergent workplace games, and pos-
sibly higher attrition may become more prevalent.
More individuals than ever are working in contract, tem-
porary, or seasonal jobs, and, because of their limited ex-
posure to an organization’sscaffolds,theseworkersmay
create their own workplace games among other nonstan-
dard workers, viewing their managers and employee
counterparts as a touchpoint to be managed. These ques-
tions have become even more relevant and urgent in the
context of the COVID-19 pandemic, as unprecedented
numbers of workers have found themselves working re-
motely and relying on unfamiliar technologies to com-
plete their tasks (Cameron et al. 2021b). Being less embed-
ded in the social context of the organization may make it
harder for workers, especially new hires, to develop
shared meanings, which can result in workplace games
that may or may not align with an organization’sobjec-
tives. Overall, current changes to work arrangements sug-
gest that the contemporary workplace may become even
more fractured and “fissured”(Weil 2014).
Lower-Skilled Work and Workplace Games
Although scholars have recently been paying close at-
tention to the gig economy, there are still very few
management studies focused on gig workers, especial-
ly lower-paid workers. Prior research highlights how
those in “bad jobs”change their self-concept or identi-
ty in order to find (positive) meaning (e.g., Ashforth
et al. 2017, Cameron et al. 2021a; see also Rosso et al.
2010 and Brief and Nord 1990 for critiques on the liter-
ature’s preoccupation with meaningfulness). In con-
trast, this study finds that workers construct meaning
through the structural features of the work—that is,
finding meaning in how they work (i.e., enacting the
work), not why they are working (i.e., for money or a
mission). Although all drivers were, to some extent, fi-
nancially dependent on driving, worries about finan-
ces did not crowd out other sources of meaning and
were incorporated into larger schemas. Hitting earn-
ing targets was incorporated into the efficiency game,
and, in the relational game, the focus was on ratings
as opposed to tip amounts. These findings thus con-
tribute to discussions on how those in lower-paid
jobs, in which worries about finances are often promi-
nent, find additional ways of making meaning (e.g.,
Hodson 2001; Lamont 2002; Meuris and Leana 2018).
Limitations and Future Research
As in any research study, there are limitations to these
claims and opportunities for future research. First,
this study cannot answer why each worker played a
specific workplace game. A reasonable hypothesis
would be that workers pushed into driving by a shock
event (e.g., job loss) might be more likely to engage in
the efficiency game, whereas those pulled into and at-
tracted to driving (e.g., being lonely and wanting so-
cial interactions) may be more likely to play the rela-
tional game. My analysis comparing push and pull
factors did not support this claim. In part, this is be-
cause it was challenging to distinguish push and pull
factors, as workers’lives and expressed motivations
were fluid and complex. For example, a case that was
particularly hard to categorize involved a driver who
was forced to internally migrate due to a natural di-
saster, so he turned to driving, as opposed to other
work he was qualified for (as a pilot), because of the
high wages and the schedule flexibility that allowed
him to periodically return home to rebuild his house.
Over the four years of data collection, another driver
expressed multiple reasons for driving (e.g., to build
an emergency fund, to pay bills after losing a job, to
financially support her daughter’s dream of opening a
business). Further, I found no relationship between
work history or economic dependence and workplace
games. Other research designs and sampling strate-
gies (e.g., representative sample, longitudinal panels)
could provide additional insights on why and when a
worker may choose to play which game.
Second, as this is a paper for generating theory, it can
only propose a set of relationships as opposed to a caus-
al relationship—thus, this study does not suggest that
workplace games alone can generate consent or long-
term commitment. Economic dependence, lack of alter-
native employment options, and enjoyment of the work
are all reasons drivers may continue working. Although
my research finds that the app and the customer were
the most salient touchpoints in ridehailing, I acknowl-
edge there could have been others that did not come up
in the interviews. And in other contexts, certain touch-
points may be more salient: in relationship-based ser-
vice work (e.g., grocery delivery), customers may be the
more salient touchpoint, and in work conducted entire-
ly online (e.g., crowd work), it may be the app. And
whereasthiscontextfinds that the two touchpoints are
experienced in parallel (within the same ride), other
types of work may suggest a more sequential interac-
tion. Third, my recruitment of drivers resulted in a
sample that included drivers that were highly active
and who had longer than average industry tenure (Katz
and Krueger 2016). However, the sample also included
those with briefer work histories on the platform,
suggesting that even individuals with shorter tenure
Cameron: Making Out While Driving
18 Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
play games, although perhaps not to the same level of
involvement (e.g., only offering emotional support).
And finally, although diverse in age and prior work ex-
perience, drivers in this study were predominantly men,
all based in North America. Future research could explore
the generalizability of these findings to platforms that
have more female workers (e.g., grocery delivery on Shipt,
care work on Care.com), in which the relational aspects of
the work are even more heavily emphasized.
Moreover, this research seeks to understand these
games, not whether they are ultimately good or bad for
the workers or the company, which is actually quite a
complex question. Unequivocally, on-demand provides
income-earning opportunities to individuals who have
been shut out of the traditional workforce, often with
more schedule flexibility than in similarly skilled jobs
(Isaac 2019, Cameron and Rosenblat 2020). Whereas work-
ers playing the relational game described their
experiences as generative and enjoyable and some schol-
ars have pointed out the nonmonetary benefits of driving
(e.g., Raval and Dourish 2016; Kamaswaren et al. 2018),
critical theorists would categorize the relational game as a
form of control that “enchants”workers (Endrissat et al.
2015) by emphasizing the feel-good nature of pleasing
customers and reducing the inherent conflict between
workers and management. These same critical scholars
could describe the efficiency game—which is associated
with accounts of burnout, stress, and negative emotions—
as, ultimately, positive, because it may signal that dish-
eartened workers are beginning to question the terms of
the labor exchange ultimately undermining their long-
term commitment to a platform designed to take advan-
tage of workers. Indeed, the quest for efficiency has led
some workers to binge with life-threatening consequences
for themselves and their families (Smiley 2021). Future
research should continue to explore workers’relationships
to workplace games and their long-term commitment to
on-demand work.
Conclusion
This study has veered away from taking a normative
position on on-demand work and instead centers
workers’lived experiences. I find that workers take
matters into their own hands, creating two workplace
games—relational and efficiency—that create mean-
ing and generate engagement. But not all games can
be won, leading to divergent implications for workers
and organizations. Thus, the next time you’re in a car,
don’t ask your driver if they like the work or even
how much they are making; ask instead, “What game
are you playing, and are you winning?”
Acknowledgments
The author is extremely appreciative of the editors of this
special issue, especially Diane Bailey and Pam Hinds, and
the three anonymous reviewers for their constructive
feedback. The author thanks the workers at RideHail for
sharing their stories and Geras Artis, Sean Dew, Eli Gonza-
lez, Vidisha Hermani, Brandon Nguyen, Alexander Renaud,
Kalie Wertz, and Ruiling Wen for their research assistance.
The author also thanks Rose Ernst and Kristin McGuire for
their editorial suggestions. Previous versions of this article
benefited from comments from Michel Anteby, Beth Bechky,
Matthew Bidwell, Seth Carnahan, Drew Carton, Angelique
Davis, Jerry Davis, Julia DiBenigno, Jane Dutton, Tawanna
Dillahunt, Arvind Karunakaran, Katherine Klein, Kevin Lee,
Mike Maffie, Melissa Mazmanian, Nick Occhiuto, Ben Shes-
takofsky, and Julia Ticona. The author is also thankful for
the feedback and support from the participants of the May
Meaning Meeting, the NYU Qualitative Group, the NYU
Data Analysis Support Group, and the Remember to Exhale
Writing Group. The author is grateful for the feedback from
seminar participants at Carnegie Mellon University, Imperial
College, Stanford University, and the University of Texas, as
well as participants at conferences. Deep gratitude to the
workers at TeaHaus, the W. K. Kellogg Biological Center,
and the Labrys Resort for Women, where much of the first
draft of this manuscript was written.
Endnotes
1
While some workers have responded to this social isolation by join-
ing online forums, and scholars are enthusiastically looking at the po-
tential of these spaces to foster belonging (e.g., Rosenblat and Stark
2016, Wood et al. 2019), the extent to which these limited scaffolds
have affected the everyday meaning-making of workers—most of
whom have not joined these spaces—remains an open question.
2
Starting in 2015, Uber installed hubs in some cities (Isaac 2019).
Less than 15% of my informants had visited one and of those the
majority had only visited once. Starting in 2013, new Lyft drivers
were required to meet a mentor for 30 minutes; in 2017, this was
discontinued.
3
I did not meet a RideHail employee until the bulk of this research
was complete and then in an academic setting, during a talk.
4
One interview was conducted in French and transcribed by the au-
thor. One participant declined to be audio-recorded, and in the other
cases the audio files became corrupted. Participants’data were re-
corded and analyzed based on the contact summary sheet (Huberman
and Miles 1989) that was created immediately after each interview.
5
Driver 43, who worked full-time doing remote office work, drove
in the evenings after her job and during major events in order to get
out of the house and be social after being alone all day.
References
Anteby M (2008) Identity incentives as an engaging form of control:
Revisiting leniencies in an aeronautic plant. Organ. Sci. 19(2):
202–220.
Ashforth BE, Kreiner GE, Fugate M (2000) All in a day’swork:
Boundaries and micro role transitions. Acad. Management Rev.
25(3):472–491.
Ashforth BE, Kreiner GE, Clark MA, Fugate M (2017) Congruence
work in stigmatized occupations: A managerial lens on em-
ployee fit with dirty work. J. Organ. Behav. 38(8):1260–1279.
Bain P, Taylor P (2000) Entrapped by the ‘electronic panopticon’?
Worker resistance in the call centre. New Technol. Work Employ-
ment 15(1):2–18.
Barley SS, Kunda G (2004) Gurus, Warm Bodies, and Hired Guns: Itin-
erant Experts in the Knowledge Economy (Princeton Press, Prince-
ton, NJ).
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 19
Baumeister RF (1991) Meanings of Life (Guilford Press, New York).
BaumeisterRF,VohsKD(2002)Thepursuitofmeaningfulness
in life. Snyder CR, Lopez SJ, eds. Handbook of Positive Psy-
chology (Oxford University Press, New York), 608–618.
Berger ML (2001) The Automobile in American History and Culture
(Greenwood, Westport, CT).
Besharov ML (2014) The relational ecology of identification: How
organizational identification emerges when individuals hold
divergent values. Acad. Management J. 57(5):1485–1512.
Beunza D (2019) Taking the Floor: Models, Morals, and Management in
a Wall Street Trading Room (Princeton University Press, Prince-
ton, NJ).
Biggart NW (1989) Charismatic Capitalism: Direct Selling Organizations
in America (University of Chicago Press, Chicago).
Bloom N (2020) A snapshot of a new working-from-home economy.
Stanford News (June 29), https://news.stanford.edu/2020/06/
29/snapshot-new-working-home-economy/.
Braverman H (1974) Labor and Monopoly Capital (Monthly Review,
New York).
Brief AP, Nord WR, eds. (1990) Meanings of Occupational Work: A
Collection of Essays (Lexington Books, Lexington, MA).
Brynjolfsson E, McAfee A (2014) The Second Machine Age: Work, Pro-
gress, and Prosperity in a Time of Brilliant Technologies (W. W.
Norton, New York).
Burawoy M (1976) Manufacturing Consent: Changes in the Labor Pro-
cess Under Monopoly Capitalism (University of Chicago Press,
Chicago).
Butterfield S (1985) Amway: The Cult of Free Enterprise (South End
Press, Boston).
Cameron L (2021) (Relative) Freedom in algorithms: How digital
platforms repurpose workplace consent. Acad. Management
Proc. 2021(1):11061.
Cameron L, Rahman H (2022) Expanding the locus of resistance:
Understanding the co-constitution of control and resistance in
the gig economy. Organ. Sci. Forthcoming.
Cameron L, Rosenblat A (2020) Gig work used to be a recession-
proof safety net. Not anymore. Fast Company (August 10),
https://www.fastcompany.com/90537938/gig-work-used-to-
be-a-recession-proof-safety-net-not-anymore.
Cameron L, Chan C, Anteby M (2021a) Heroes from above and be-
low: Workers responses to the moralization of their work.
Working Paper.
Cameron L, Thomason B, Conzon V (2021b) Risky business: Gig
workers’and the navigation of ideal worker expectations dur-
ing the COVID-19 pandemic. J. Appl. Psych. Forthcoming.
Campbell H (2018) Rideshare Guide: Everything You Need to Know
About Driving for Uber, Lyft and Other Ridesharing Companies
(Skyhorse, New York).
Cappelli PH, Keller J (2013) A study of the extent and potential
causes of alternative employment arrangements. ILR Rev. 66(4):
874–901.
Carton AM (2018) “I’m not mopping the floors, I’m putting a man
on the moon”: How NASA leaders enhanced the meaningful-
ness of work by changing the meaning of work. Admin. Sci.
Quart. 63(2):323–369.
Caza BB, Moss S, Vough H (2018) From synchronizing to harmoniz-
ing: The process of authenticating multiple work identities. Ad-
min. Sci. Quart. 63(4):703–745.
Charmaz K (2006) Constructing Grounded Theory: A Practical Guide
Through Qualitative Analysis (Sage, Thousand Oaks, CA).
Corbin J, Strauss A (2007) Basics of Qualitative Research: Techniques
and Procedures for Developing Grounded Theory, 3rd ed. (Sage,
Thousand Oaks, CA).
Curchod C, Patriotta G, Cohen L, Neysen N (2019) Working for an
algorithm: Power asymmetries and agency in online work set-
tings. Admin. Sci. Quart. 65(3):644–676.
Davis F (1959) The cabdriver and his fare: Facets of a fleeting rela-
tionship. Amer. J. Sociol. 65(2):158–165.
De Man H (1928) The Psychology of Marxian Socialism (Transaction
Books, Piscataway, NJ).
Deterding S, Dixon D, Khaled R, Nacke LE (2011) Gamification:
Toward a definition. Proc. Conf. Human Factors Comput. Systems.
Dutton J, Heaphy E (2003) The power of high-quality connections.
Cameron K, Dutton JE, Quinn R, eds. Positive Organizational
Scholarship (Berrett-Koehler, Oakland, CA), 263–278.
Dutton J, Ragins B (2017) Moving forward: Positive relationships at
work as a research frontier: Building a theoretical and research
foundation. Dutton JE, Ragins BR, eds. Exploring Positive Rela-
tionships at Work (Lawrence Erlbaum Associates, Mahwah, NJ),
387–400.
Dutton JE, Debebe G, Wrzesniewski A (2016) Being valued and de-
valued at work: A social valuing perspective. Qualitative Orga-
nizational Research: Best Papers from the Davis Conference on
Qualitative Research (IAP Information Age Publishing, Char-
lotte, NC), 9–51.
Eisenhardt KM (1989) Building theories from case study research.
Acad. Management Rev. 14(4):532–550.
Elangovan AR, Pinder CC, McLean M (2010) Callings and organiza-
tional behavior. J. Vocational Behav. 76(3):428–440.
Endrissat N, Islam G, Noppeney C (2015) Enchanting work: New
spirits of service work in an organic supermarket. Organ. Stud.
36(11):1555–1576.
Griesbach K, Reich A, Elliott-Negri L, Milkman R (2019) Algorith-
mic control in platform food delivery work. Socius 5:23780231
19870041.
Gutek BA (1995) The Dynamics of Service: Reflections on the Changing
Nature of Customer/Provider Interactions (Jossey-Bass, San
Francisco).
Hackman JR, Oldham GR (1980) Work Redesign (Addison-Wesley,
Reading, MA).
Hall JV, Krueger AB (2018) An analysis of the labor market for
Uber’s driver-partners in the United States. ILR Rev. 71(3):
705–732.
Hall DT, Mirvis PH (2013) Redefining work, work identity, and ca-
reer success. Blustein DL, ed. The Oxford Handbook of the Psychol-
ogy of Working (Oxford University Press, New York), 203–217.
Hochschild AR (1983) The Managed Heart: Commercialization of Hu-
man Feeling (University of California Press, Berkeley, CA).
Hodson R (2001) Dignity at work. Working paper, Cambridge Uni-
versity Press.
Huberman AM, Miles MB (1989) Data management and analysis
methods. Denzin NK, Lincoln YS, eds. Handbook of Qualitative
Research (Sage Publications, Thousand Oaks, CA), 428–444.
Hurst CS, Baranik LE, Clark S (2017) Job content plateaus: Justice,
job satisfaction, and citizenship behavior. J. Career Development
44(4):283–296.
Irani LC, Silberman MS (2013) Turkopticon: Interrupting worker in-
visibility in Amazon Mechanical Turk. Proc. SIGCHI Conf. Hu-
man Factors Comput. Systems (ACM, New York), 611–620.
Isaac M (2019) Super Pumped: The Battle for Uber (W. W. Norton,
New York).
Juslin PN, Sloboda J (2011) Handbook of Music and Emotion: Theory,
Research, Applications (Oxford University Press, New York).
Kahn WA (2007) Meaningful connections: Positive relationships
and attachments at work. Dutton JE, Ragins BR, eds. Explor-
ing Positive Relationships at Work: Building a Theoretical and Re-
search Foundation (Lawrence Erlbaum Associates, Mahwah,
NJ), 189–207.
Kahn WA, Fellows S (2013) Employee engagement and meaningful
work. Dik BJ, Byrne ZS, Steger MF, eds. Purpose and Meaning in
the Workplace (American Psychological Association, Washing-
ton, DC), 105–126.
Cameron: Making Out While Driving
20 Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS
Kalleberg AL (2011) Good Jobs, Bad Jobs: The Rise of Polarized and
Precarious Employment Systems in the United States, 1970s–2000s
(Russell Sage Foundation, New York).
Kamaswaren V, Cameron L, Dillahunt T (2018) Support for social
and cultural capital development in real-time ridesharing serv-
ices. Proc. 2018 Conf. Human Factors Comput. Systems (ACM,
New York), 1–12.
Katz LF, Krueger AB (2016) The Rise and Nature of Alternative Work
Arrangements in the United States, 1995–2015 (National Bureau of
Economic Research, Cambridge, MA).
Katz LF, Krueger AB (2019) The rise and nature of alternative work
arrangements in the United States, 1995–2015. ILR Rev. 72(2):
382–416.
Kellogg KC, Valentine MA, Christin A (2020) Algorithms at work:
The new contested terrain of control. Acad. Management Ann.
14(1):366–410.
Kunda G (2009) Engineering Culture: Control and Commitment in a
High-Tech Corporation, revised edition (Temple University Press,
Philadelphia).
Lamont M (2002) The Dignity of Working Men: Morality and the
Boundaries of Race, Class, and Immigration (Harvard University
Press, Cambridge, MA).
Lee MK, Kusbit D, Metsky E, Dabbish L (2015) Working with ma-
chines: The impact of algorithmic and data-driven management
on human workers. Proc. 33rd Annual ACM Conf. Human Factors
Comput. Systems (ACM, New York), 1603–1612.
Lehdonvirta V (2018) Flexibility in the gig economy: managing time
on three online piecework platforms. New Tech. Work Employ-
ment 33(1):13–29.
Leidner R (1993) Fast Food, Fast Talk: Service Work and the Routiniza-
tion of Everyday Life (University of California Press, Berkeley,
CA).
Locke KD (2001) Grounded Theory in Management Research (Sage,
Thousand Oaks, CA).
Luedke T (2010) Driving lives: An ethnography of Chicago taxi
drivers. Working paper: 2010-03, Industry Studies Association,
Pittsburgh.
Macdonald CL, Sirianni C, eds. (1996) Working in the Service Society
(Temple University Press, Philadelphia).
Mears A (2015) Working for free in the VIP: Relational work and
the production of consent. Amer. Sociol. Rev. 80(6):1099–1122.
Meuris J, Leana CR (2018) The high cost of low wages: Economic
scarcity effects in organizations. Res. Organ. Behav. 35:143–158.
Mollick ER, Rothbard N (2014) Mandatory fun: Consent, gamifica-
tion and the impact of games at work. The Wharton School Re-
search Paper Series.
Oakes G (1990) The Soul of the Salesman: The Moral Ethos of Personal
Sales (Humanities Press International, Atlantic Highlands, NJ).
Occhiuto N (2017) Investing in independent contract work: The sig-
nificance of schedule control for taxi drivers. Work Occupations
44(3):268–295.
O’Mahony S, Bechky BA (2006) Stretchwork: Managing the career
progression paradox in external labor markets. Acad. Manage-
ment J. 49(5):918–941.
O’Reilly C (1989) Corporations, culture, and commitment: Motiva-
tion and social control in organizations. California Management
Rev. 31(4):9–25.
Ouellet L (1994) Pedal to the Metal: The Work Life of Truckers (Temple
University Press, Philadelphia).
Pantano E, Viassone M (2015) Engaging consumers on new integrat-
ed multichannel retail settings: Challenges for retailers. J. Retail-
ing Consumer Services 25:106–114.
Petriglieri G, Ashford SJ, Wrzesniewski A (2019) Agony and ec-
stasy in the gig economy: Cultivating holding environments
for precarious and personalized work identities. Admin. Sci.
Quart. 64(1):124–170.
Pierce JL (1996) Gender Trials: Emotional Lives in Contemporary Law
Firms (University of California Press, Berkeley, CA).
Pratt MG (2000) The good, the bad, and the ambivalent: Managing
identification among Amway distributors. Admin. Sci. Quart.
45(3):456–493.
Pratt M, Ashforth B (2003) Fostering meaningfulness in working
and at work. Cameron KS, Dutton JE, Quinn RE, eds. Positive
Organizational Scholarship: Foundations of a New Discipline (Ber-
rett-Koehler, Oakland, CA), 309–327.
Pratt MG, Rockmann KW, Kaufmann JB (2006) Constructing profes-
sional identity: The role of work and identity learning cycles in
the customization of identity among medical residents. Acad.
Management J. 49(2):235–262.
Rahman HA (2021) The invisible cage: Workers’reactivity to opa-
que algorithmic evaluations. Admin. Sci. Quart., ePub ahead of
print April 21, https://doi.org/10.1177/00018392211010118.
Rahman HA, Barley SR (2017) Situated redesign in creative occupa-
tions—an ethnography of architects. Acad. Management Discover-
ies 3(4):404–424.
Rahman HA, Valentine MA (2021) How managers maintain control
through collaborative repair: Evidence from platform-mediated
“gigs.”Organ. Sci., ePub ahead of print February 19, https://
doi.org/10.1287/orsc.2021.1428.
Ranganathan A, Benson A (2020) A numbers game: Quantification
of work, auto-gamification, and worker productivity. Amer. So-
ciol. Rev. 85(4):573–609.
Raval N, Dourish P (2016) Standing out from the crowd: Emo-
tional labor, body labor, and temporal labor in ridesharing.
Proc. 19th ACM Conf. Comput. Supported Cooperative Work Soc.
Comput. (ACM, New York), 97–107.
Ravenelle AJ (2019) Hustle and gig. Working paper, University of
California Press.
Rhymer J (2020) Location independent organizations: Designing
work across space and time. Unpublished doctoral dissertation,
University of Washington, Seattle.
Roethlisberger FJ, Dickson WJ (1939) Management and the Worker
(Harvard University Press, Cambridge, MA).
Rosenblat A (2018) Uberland: How Algorithms Are Rewriting the Rules
of Work (University of California Press, Berkeley, CA).
Rosenblat A, Stark L (2016) Algorithmic labor and information
asymmetries: A case study of Uber’sdrivers.Internat. J. Comm.
10:27.
Rosso BD, Dekas KH, Wrzesniewski A (2010) On the meaning of work:
A theoretical integration and review. Res. Organ. Behav. 30:91–127.
Roy D (1959) “Banana time”: Job satisfaction and informal interac-
tion. Human Organ. 18(4):158–168.
Salancik GR, Pfeffer J (1978) A social information processing ap-
proach to job attitudes and task design. Admin. Sci. Quart. 23(2):
224–253.
Sallaz J (2002) The house rules: Autonomy and interests among ser-
vice workers in the contemporary casino industry. Work Occu-
pations 29(4):394–427.
Sallaz J (2009) The Labor of Luck: Casino Capitalism in the United States
and South Africa (University of California Press, Berkeley, CA).
Sallaz JJ (2013) Labor, Economy, and Society (Polity Press, Malden, MA).
Sallaz JJ (2019) Lives on the Line: How the Philippines Became the
World’s Call Center Capital, 1st ed. (Oxford University Press,
New York).
Scheiber N (2017) How Uber uses psychological tricks to push its
drivers’buttons. The New York Times,2.
Shapiro A (2018) Between autonomy and control: Strategies of ar-
bitrage in the “on-demand”economy. New Media Soc. 20(8):
2954–2971.
Sharone O (2007) Constructing unemployed job seekers as profes-
sional workers: The depoliticizing work-game of job searching.
Qualitative Sociol. 30(4):403–416.
Cameron: Making Out While Driving
Organization Science, Articles in Advance, pp. 1–22, © 2021 INFORMS 21
Sharone O (2013) Why do unemployed Americans blame them-
selves while Israelis blame the system? Soc. Forces 91(4):
1429–1450.
Sherman R (2007) Class Acts: Service and Inequality in Luxury
Hotels (University of California Press, Berkeley, CA).
Smiley L (2021) He thought he could outfox the gig economy. He
was wrong. Wired (June 22), https://www.wired.com/story/
gig-economy-uber-lyft-doordash-jeffrey-fang/.
Spreitzer GM, Quinn RE (1996) Empowering middle managers to
be transformational leaders. J. Appl. Behav. Sci. 32(3):237–261.
Spreitzer GM, Cameron L, Garrett L (2017) Alternative work ar-
rangements: Two images of the new world of work. Annual
Rev. Organ. Psych. Organ. Behav. 4:473–499.
Suchman LA (1983) Office procedure as practical action: models
of work and system design. ACM Trans. Inform. Systems 1(4):
320–328.
Suchman L (2007) Human-Machine Reconfigurations: Plans and Situat-
ed Actions (Cambridge University Press, New York).
Vallas S, Schor JB (2020) What do platforms do? Understanding the
gig economy. Annual Rev. Sociol. 46(1):273–294.
Viscelli S (2016) The Big Rig: Trucking and the Decline of the American
Dream (University of California, Berkeley, CA).
Weil D (2014) The Fissured Workplace: Why Work Became So Bad for So
Many and What Can Be Done to Improve It (Harvard University
Press, Cambridge, MA).
Wood AJ, Graham M, Lehdonvirta V, Hjorth I (2019) Good gig, bad
gig: Autonomy and algorithmic control in the global gig econo-
my. Work Employment Soc. 33(1):56–75.
Wrzesniewski A (2002) “It’s not just a job”: Shifting meanings
ofworkinthewakeof9/11.J. Management Inquiry 11(3):
230–234.
Wrzesniewski A, Dutton J (2001) Crafting a job: Revisioning em-
ployees as active crafters of their work. Acad. Management Rev.
26(2):179–201.
Wrzesniewski A, Dutton JE, Debebe G (2003) Interpersonal sense-
making and the meaning of work. Res.Organ.Behav. 25:
93–135.
Wrzesniewski A, McCauley C, Rozin P, Schwartz B (1997) Jobs, ca-
reers, and callings: People’s relations to their work. J. Res. Per-
sonality 31(1):21–33.
Zomerdijk L, Voss C (2010) Service design for experience-centric
service