University of Southern California Dissertations and Theses

Increasing capabilities or decreasing cost: Fairness perceptions of job displacement due to automation and outsourcing

(USC Thesis Other)

Increasing capabilities or decreasing cost: Fairness perceptions of job displacement due to automation and outsourcing

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Running head: FAIRNESS OF AUTOMATION AND OUTSOURCING
Increasing Capabilities or Decreasing Cost: Fairness Perceptions of Job Displacement Due to
Automation and Outsourcing
Jennifer Kim
Marshall School of Business, Management and Organization
Ph.D. in Business Administration
University of Southern California
FACULTY OF THE USC GRADUATE SCHOOL
Degree conferral date: December 1 2, 2018
ii

Abstract

Automation is a critical and timely source of job displacement, but little is known about how
observers react to automation-driven layoffs. In a series of six studies, I contrast reactions to
automation with reactions to outsourcing, a conceptually similar organizational decision. Studies
1a-1b demonstrate that people find layoffs caused by automation to be fairer than those caused
by outsourcing. Studies 2a-2b find a similar pattern when the issue of technology is moot:
respondents find replacement of a worker to be fairer when a more productive worker is hired as
a replacement (analogous to automation) than when a cheaper worker is hired (analogous to
outsourcing). Study 3 considers mediating mechanisms for these effects; potential for long-term
progress, worker devaluation and disenfranchisement, and perceived motivation emerged as
significant mediators. Study 4 examines how motivation framing and layoff cause interact to
shape fairness perceptions. I discuss the implications if these findings for theory and practice.

Keywords:

Automation, outsourcing, fairness, justice, layoffs, job displacement, productivity
iii
Table of Contents
TITLE PAGE………………………………………………………………………………… i
ABSTRACT………………………………………………………………………………….. ii
TABLE OF CONTENTS…………………………………………………………………….. iii
LIST OF TABLES…………………………………………………………………………… vi
LIST OF FIGURES…………………………………………………………………………... vii
INTRODUCTION……………………………………………………………………………. 1
Reactions to Layoffs by Various Stakeholders……………………………………………… 2
Automation Versus Outsourcing…………………………………………………………….. 7
Increasing Productivity and Decreasing Costs……………………………………………… 9
The appeal of productivity………………………………………………………………... 10
The unfairness of cutting costs…………………………………………………………… 14
Hypotheses …………………………………………………………………………………. 15
OVERVIEW OF STUDIES………………………………………………………………….. 16
STUDY 1A…………………………………………………………………………………… 17
Method……………………………………………………………………………………… 17
Results………………………………………………………………………………………. 18
STUDY 1B…………………………………………………………………………………… 18
Method……………………………………………………………………………………… 18
Results………………………………………………………………………………………. 19
STUDY 2A…………………………………………………………………………………… 20
Method……………………………………………………………………………………… 20
iv
Results………………………………………………………………………………………. 22
STUDY 2B…………………………………………………………………………………… 23
Method……………………………………………………………………………………… 23
Results………………………………………………………………………………………. 24
STUDY 3…………………………………………………………………………………….. 25
Proposed Mechanisms……………………………………………………………………… 25
Method……………………………………………………………………………………… 28
Results……………………………………………………………………………………… 31
STUDY 4……………………………………………………………………………………. 39
Method……………………………………………………………………………………… 39
Results………………………………………………………………………………………. 40
GENERAL DISCUSSION…………………………………………………………………… 44
Contributions………………………………………………………………………………… 45
Practical Implications……………………………………………………………………….. 49
Limitations and Future Directions………………………………………………………….. 51
Conclusion………………………………………………………………………………….. 54
REFERENCES………………………………………………………………………………. 55
APPENDICES………………………………………………………………………………… 68
A. Pilot study analyses……………………………………………………………………… 68
B. Study 1b additional measures and analyses………………………………………….…. 70
C. Study 2b results with outliers excluded…………………………………………………. 72
D. Study 3 results with outliers excluded………………………………………………....... 73
v
E. Study 4 scenarios presented to participants…………………………………………….. 76
F. Study 4 results with outliers excluded………………………………………………….. 77
G. Study 4 additional measures and analyses……………………………………………… 78
H. Additional studies similar to Study 4…………………………………………………… 81
I. Additional mediation study…………………………………………………………….. 83
J. Additional studies exploring the role of progress………………………………………. 86
K. Pre-registration links……………………………………………………………………. 88
vi

List of Tables
Table 1. Index of the Content of Appendices 17
Table 2. Descriptive and Difference Statistics for All Variables in Study 3, by 34
Sample
vii

List of Figures
Figure 1. Tables shown to participants in the productive replacement (a) and cheaper 22
replacement (b) conditions in Study 2a.
Figure 2. Tables shown to participants in the productive replacement (a) and cheaper 24
replacement (b) conditions in Study 2b.
Figure 3a. Multiple mediation model from Study 3 (Sample 1) whereby the effect of 36
layoff reason (0 = decreasing cost, 1 = increasing productivity) on Fairness
Perceptions is mediated by ratings of Devaluation of Worker,
Disenfranchisement of Worker, and Long-Term Progress.
Figure 3b. Multiple mediation model for Study 3 (Sample 2) whereby the effect of 38
layoff reason (0 = decreasing cost, 1 = increasing productivity) on Fairness
Perceptions is mediated by ratings of Devaluation of Worker,
Disenfranchisement of Worker, and Motivation to Increase Capabilities.
Figure 4. The means by condition with error bars representing 95% confidence 44
intervals for a) motivation to increase capabilities, b) fairness, and c)
willingness to invest from Study 4.
FAIRNESS OF AUTOMATION AND OUTSOURCING 1

Increasing Capabilities or Decreasing Cost: Fairness Perceptions of Job Displacement Due
to Automation and Outsourcing
Automation has been a major source of job loss in the past, and seems likely to emerge as
a disruptive force in the near future. While estimates vary, economists widely agree that we are
facing a major shift in how work is done. For example, Frey and Osborne (2017) estimate that up
to 47% of jobs in the U.S. are at risk of being automated. Estimates from the OECD and
McKinsey Global Institute are less dire, but still predict significant disruption, especially among
young people and low-skilled workers (Arntz, Gregory, & Zierahn, 2016; Manyika et al., 2017).
Observers argue that the advent of artificial intelligence will usher in a “second machine age,”
and that this coming change will have an even greater impact on our lives than the previous
industrial revolution (Brynjolfsson & McAfee, 2014). Such change is not limited to our direct
interaction with technology but has broad societal implications; for example, automation may
negatively impact social mobility, wiping out jobs that are often stepping-stones to better paying
careers, and leading to even greater disparity between the haves and have-nots (Boston
Consulting Group, 2017).
Polling data suggests that people are concerned about implications of technological
advances like artificial intelligence for job elimination, although consistent with self-serving
biases many people think their own jobs are less vulnerable to automation than jobs in general
(Gallup, 2018; Smith, 2016). However, less is known about fairness perceptions of job loss due
to automation. Judgments of fairness and justice are potent drivers of human behavior and at
times even compel individuals to act contrary to their self-interest in order to punish those
responsible for unfair behavior or to rectify an injustice (Camerer, 2011; Colquitt, Conlon,
Wesson, Porter, & Ng, 2001; Kahneman, Knetsch, & Thaler, 1986a; Rabin, 1993; Van Dijk,
FAIRNESS OF AUTOMATION AND OUTSOURCING 2

Sonnemans, & van Winden, 2002). In recent years, moral outrage among stakeholders and the
public have led to major upheavals within organizations, such as changes to leadership (e.g.,
Isaac, 2017a) and overhaul of data management practices (e.g., Bowles, 2018; Wakabayashi,
2018). Given that 1) corporate reputation, or “observers’ collective judgments of a corporation
based on assessments of the financial, social, and environmental impacts attributed to the
corporation over time (p. 34; Barnett, Jermier, & Lafferty, 2006),” is a valuable strategic
resource that can be leveraged to enhance organizational performance (Fombrun, 1996, 2001;
Gibson, Gonzales, & Castanon, 2006; Roberts & Dowling, 2002), and 2) we are likely to see a
proliferation of job displacement due to automation (Brynjolfsson & McAfee, 2014), I seek to
obtain answers to the following question: To what extent do people perceive automation-induced
layoffs as fair, and why?
The rest of this manuscript is organized as follows. I begin by briefly reviewing prior
literature on how various stakeholders respond to job displacement and outline my reasons for
focusing on impartial observers’ fairness perceptions of automation-driven layoffs. Next, I detail
how people’s reactions to outsourcing-driven layoffs can serve as a relevant comparison
condition in examinations of people’s reactions to layoffs that ensue from automation. Then, in
the following section I expand on a critical difference between automation and outsourcing,
namely the focus on productivity and cost, respectively, that may differentially affect fairness
perceptions of the layoffs that ensue from those decisions. I end this section with my hypotheses
and expected contributions before transitioning to an overview of the studies.
Reactions to Layoffs by Various Stakeholders

The employee downsizing literature suggests that layoffs are primarily implemented in
order to reduce costs and increase operating efficiencies (Budros, 1999; Cameron, Freeman, &
FAIRNESS OF AUTOMATION AND OUTSOURCING 3

Mishra, 1993; Cascio & Young, 2003), not just during times of declining demand but often
during times of robust demand and profitability (Cappelli, 2000; Cascio, 1993; S. J. Freeman &
Cameron, 1993). These layoffs have consequences for various groups of stakeholders, and
resultantly, there is a large body of work devoted to examining their reactions to layoffs (e.g.,
Brockner & Greenberg, 1990; Datta, Guthrie, Basuil, & Pandey, 2010; Robbins, 1999). In the
paragraphs that follow, I provide a brief review of the ways in which 3 stakeholder groups—
layoff victims, layoff survivors (i.e. those who witnessed layoffs but did not lose their jobs), and
investors—respond to layoffs, after which I present a case for examining impartial observers’
fairness perceptions of automation-driven job displacement.
The group of individuals who are most directly implicated during the layoff process are
layoff victims, who, predictably, do not respond positively to losing their jobs. However,
findings show that to the extent that they believe the procedure for implementing the decision
was fair (procedural justice; Leventhal, 1980; Thibaut & Walker, 1975) or they were given
adequate explanation for the layoffs (interactional justice; Bies, 1987; Tyler & Bies, 1990), they
perceive higher levels of overall fairness, report higher willingness to commit to a future
organization, and harbor less resentment toward, express less desire to sue, and exhibit higher
levels of endorsement for their previous employer (Brockner et al., 1994; Konovsky & Brockner,
1993; Konovsky & Folger, 1991; Wanberg, Gavin, & Bunce, 1999). Another stakeholder group
that is directly impacted by layoffs—at times more so than layoff victims (e.g., Burke, 2003;
Devine, Reay, Stainton, & Collins-Nakai, 2003)—are layoff survivors. These individuals, who
are left to process the aftermath of the layoffs, have been found to exhibit a host of negative
reactions, such as lower morale, depression, frustration, fear and anxiety, increased stress and
uncertainty, reduced motivation and commitment, and feelings of insecurity, betrayal, and
FAIRNESS OF AUTOMATION AND OUTSOURCING 4

distrust, which have been conceptualized as symptoms of “layoff-survivor sickness” (Noer,
2009; Robbins, 1999). As among layoff victims, however, researchers have also found that
negative reactions are shaped by perceptions of fairness: the degree to which layoff survivors felt
that they had received a clear rationale for the layoffs or that those who were laid off were
treated fairly was negatively associated with the extent to which they displayed reduced work
performance or effort, indicated turnover intentions, and/or reported decreased trust in, support
for, and commitment to the organization (Brockner, DeWitt, Grover, & Reed, 1990; Brockner &
Greenberg, 1990; Brockner, Grover, Reed, DeWitt, & O'Malley, 1987; Brockner, Wiesenfeld, &
Martin, 1995; Konovsky & Brockner, 1993; Mishra & Spreitzer, 1998). Broadly, extant evidence
suggests that justice perceptions, or the extent to which the procedures and outcomes were
perceived to be fair, figure prominently in layoff victims’ and survivors’ reactions to layoffs.
In contrast, investors (e.g., shareholders) are stakeholders who could potentially benefit
from layoffs that increase organizational efficiency and ultimately organizational performance.
The reactions of investors to layoffs have typically been measured as market reactions to layoff
announcements, and although the majority of findings indicate that layoffs have no significant or
negative effects on stock price (c.f. Chalos & Chen, 2002; Datta et al., 2010), several studies
reveal occasional positive effects (e.g., Brookman, Chang, & Rennie, 2007; Chalos & Chen,
2002; Nixon, Hitt, Lee, & Jeong, 2004; Palmon, Sun, & Tang, 1997). For example, Palmon et al.
(1997) found that when layoffs were implemented with the goal of enhancing efficiency (but not
when they were a response to declining demand), there were positive market returns. In addition,
Nixon et al. (2004) revealed a positive relationship between reallocation (i.e. “distribution of
employee reduction among the firm’s units”) and market valuation, suggesting that the more the
layoffs were selective and focused in certain units (as opposed to across-the-board), the more
FAIRNESS OF AUTOMATION AND OUTSOURCING 5

favorably investors reacted. These findings suggest that while investors generally react
negatively to announcements of layoffs, when the layoffs signal attempts that could potentially
improve organizational performance (e.g., through cuts that increase efficiency or focused and
deliberate reductions to less valuable units), investors react positively. While the above findings
speak to investors’ reactions to layoffs in terms of market reactions, less is known specifically
about how layoffs influence fairness perceptions of investors. However, given the wealth of
studies that document people’s willingness to undertake losses in order to punish and protest
injustices (e.g., Skarlicki & Folger, 1997), even when the injustice has little bearing on their own
self-interests (Porath, MacInnis, & Folkes, 2011), it seems unlikely that investors will react
favorably to a decision that they deem unfair. Put differently, I suggest that positive market
reactions to layoffs reflect optimistic valuations of a company that is believed to be operating
within the bounds of fairness and acceptability.
Considerations of various stakeholders’ reactions to job displacement make apparent the
possibility that what is judged to be fair depends on the stakes that each individual has in the
process and that what is fair to one individual may not be equally fair to another individual. For
example, an employee who is laid off with a small severance package may still believe that the
decision was highly unjust, even though investors, who could potentially gain from the decision,
construe the decision as legitimate and fair. Konow (2005) makes precisely this point, arguing
that evaluations of fairness are often conflated with self-interest and that people are prone to
making biased (i.e. ego-centric and self-serving) judgments of fairness. He suggests that “when
people are stakeholders, they tend to view their fair share of a reward as being greater than their
fair share in the view of others (p. 2),” a phenomenon that has been demonstrated in numerous
studies (e.g., Babcock, Loewenstein, & Issacharoff; Babcock, Wang, & Loewenstein, 1996;
FAIRNESS OF AUTOMATION AND OUTSOURCING 6

Konow, 2001; Robertson, Lamin, & Livanis, 2010; Uhlmann, Pizarro, Tannenbaum, & Ditto,
2009). Recognizing that the stakes and objectives of various parties can render conflicting
judgments about justice, Adam Smith (1759) proposed the concept of “impartial spectators,”
whose unbiased views should constitute the basis of legitimate justice claims. Because
organizational decisions about employee downsizing are a contentious arena in which different
stakeholders with competing objectives have much to gain and lose (e.g., employees who want to
keep their livelihood, managers who need to demonstrate their competence, shareholders who
benefit from efficiency and strong firm performance), an understanding of impartial observers’
fairness perceptions of such decisions can furnish a more objective reference point for informing
related discussions and decisions. It is important to note that impartial observers, or the public,
who have no current stake in an organization, may at one point become stakeholders who can
influence the outcomes of an organization. R. E. Freeman (2010) suggests that "A stakeholder in
an organization is (by definition) any group or individual who can affect or is affected by the
achievement of the organization's objectives (p. 46)" and because members of the public are
potential customers, employees, investors, or couriers of information and opinions (e.g.,
journalists) who could shape the ability of an organization to meet its objectives, they may be
classified as latent stakeholders (Mitchell, Agle, & Wood, 1997). Thus, I focus on the public’s
fairness reactions to layoffs due to their impartiality and their inherent power to compel change
as latent stakeholders.
In addition, I examine specifically the public’s fairness reactions to automation-induced
layoffs primarily because it is a timely and apposite issue of our time. From machines and
computer systems that can write news articles to those that can autonomously drive trucks across
the country, technological advances that may eventually eliminate the need for human labor have
FAIRNESS OF AUTOMATION AND OUTSOURCING 7

brought to the forefront concerns about mass job displacement in discussions about the future of
work (Arntz, Gregory, & Zierahn, 2016; Manyika et al., 2017). Estimates by McKinsey Global
Institute (Manyika, Lund, et al., 2017) suggest that approximately 50% of current work activities
are automatable and that 6 out of 10 occupations currently have more than 30% of automatable
activities. In addition, assuming that automation adoption occurs at the midpoint or fastest rate
estimated, 400 to 800 million individuals may be displaced around the world by 2030, including
75 to 375 million who may also need to change occupational categories; In the United States
alone, up to 53 million individuals, or up to 32% of the workforce may need to transition to a
different occupational category. There are variations in these estimates and it is unclear how
many jobs will be created to offset job loss (Acemoglu & Restrepo, 2017; Arntz, Gregory, &
Zierahn, 2016; Keynes, 1930; Manyika, Lund, et al., 2017; Nedelkoska & Quintini, 2018; Rifkin,
1995). However, one fact seems indisputable: automation will be a significant source of job
displacement. Of note, I limit my investigations to downsizing-related layoffs which do not
include terminations for cause, such as poor performance.
Finally, I center my investigations on fairness perceptions, because justice is a motivating
force that fashions much of human behavior not just within organizations (Cohen-Charash &
Spector, 2001; Colquitt, et al., 2001), but also more broadly within societies (Rawls, 1971).
When livelihoods and profit are at stake, arriving at a consensus about what is fair is likely to be
especially critical for preserving the social fabric.
Automation Versus Outsourcing

In examining people’s fairness perceptions of automation-induced layoffs, I compare
them in particular to people’s perceptions of outsourcing-driven layoffs. This is for two reasons.
First, automation is conceptually similar to outsourcing in that it entails the transfer of work to
FAIRNESS OF AUTOMATION AND OUTSOURCING 8

another entity—in essence, automation is a form of outsourcing, but to a machine instead of
another human worker. Both automation and outsourcing are actions that are taken in order to
increase efficiency. Second, outsourcing is another common cause of layoffs that has received
significant attention in public discourse. Defined as “the procurement of products or services
from an outside supplier” (Robertson, et al., 2010), outsourcing has been spotlighted alongside
automation as a potential source of job loss (Acemoglu, Autor, Dorn, Hanson, & Price, 2016;
Lee, 2016; Sen, 2018). For these reasons, I suggest that layoffs that ensue from the decision to
outsource constitute a relevant comparison condition for assessing people’s reactions to layoffs
that result from automation.
Considerations of how people may perceive the fairness of automation-induced layoffs
reveal two divergent possibilities. According to a Gallup (2017) poll, a majority (63%) of U.S.
adults expected artificial intelligence to widen the gap between the rich and poor, and economists
have made similar arguments regarding the role of technology in further driving inequalities
(Tyson & Spence, 2017). Given the potential for vast labor market disruption and increases in
inequality with automation, I might expect people to see job loss stemming from automation as
broadly unfair. This expectation is given further weight by the conceptual analogy of automation
to outsourcing, which is generally perceived by observers unfavorably (Lacity, &
Rudramuniyaiah, 2009; Global Sourcing Association, 2009). If automation is conceptually
equivalent to “outsourcing to a machine,” people may have similar negative reactions to
automation as they do toward outsourcing. Indeed, from an organization’s perspective both
decisions are similar in that they lower the cost per unit of production (i.e., increase efficiency).
On the other hand, prior research suggests that similar actions can be perceived
differently based on their framing and the intuitive reactions that this evokes within observers
FAIRNESS OF AUTOMATION AND OUTSOURCING 9

(e.g., Brockner et al., 1995; Kahneman, Knetsch, & Thaler, 1986b; Tversky & Kahneman, 1981;
Yoon, Oh, Song, Kim, & Kim, 2014). For example, while people tend to find decreases in wages
and salary in a situation with no inflation to be unfair, they find increases in salary that do not
keep up with inflation to be fair, even when the difference in “real wages” is equivalent in the
two cases (Kahneman, et al., 1986b). In a similar vein, while automation and outsourcing may
have a similar functionality from an organization’s perspective (i.e., in terms of expected cost per
unit), lay people may associate those decisions with different pieces of the efficiency calculus:
automation with increased production in a given amount of time (i.e. productivity), and
outsourcing with lowering of labor cost. These two objectives, in turn, may be seen as
differentially legitimate.
I argue for this second possibility, and propose that layoffs due to automation may be
construed as fairer than those caused by outsourcing because automation (compared to
outsourcing) is associated more with increasing productivity and outsourcing (compared to
automation) is associated more with decreasing cost. I outline these arguments in greater detail
below.
Increasing Productivity and Decreasing Costs

Although organizations both automate and outsource in order to increase efficiency, the
features that people associate with each differ markedly. In the current study I use the term
“efficiency” to signify multifactor productivity, or the ratio of the output to the combined inputs
(e.g., labor, capital, energy, materials) used to produce the output (“Multifactor Productivity,”
n.d.), and the term “productivity” to represent labor productivity, or the output produced per hour
FAIRNESS OF AUTOMATION AND OUTSOURCING 10

of work (“Labor Productivity,” n.d.).
1
Automation involves the use of machines, which are
generally more reliable, consistent, and predictable, and less prone to errors or emotional whims
than human workers (Kaplan, 2015). This allows organizations to produce more goods in a given
amount of time. Consequently, the most salient aspect of automation is likely to be the
accompanying increase in productivity. Outsourcing, however, is most strongly associated with
cutting costs (Global Sourcing Association, 2015).
To provide further support for the idea that people tend to associate automation with
increasing productivity and outsourcing with decreasing costs, I directly measured these
associations in a sample of 92 Mturk workers. Participants were asked to rate the extent to which
they believed automation and outsourcing (order counterbalanced) were associated with
increasing productivity (“increasing productivity” and “producing more goods in a given amount
of time”) and decreasing costs (“the company saving money” and “paying workers less for the
same amount of goods produced”). As expected, people associated automation more with
increasing productivity than outsourcing and outsourcing more with lowering organizational
costs than automation (see the Appendix for details). Despite automation and outsourcing being
two strategies for enhancing efficiency, due to the different features that are prominent in each—
increased productivity for automation and decreased labor costs for outsourcing—I predict that
they will have divergent effects of perceptions of fairness. Below, I explain the theoretical basis
for this rationale in more detail.
The appeal of productivity. Prior research in the fields of time consumption, well-being,
and values underscore the primacy of productivity in guiding human experience. In an emerging

1
The BLS only calculates labor productivity for human workers. For the purposes of comparing automation and
outsourcing, in this paper I apply the concept of labor productivity to machines, and conceptualize productivity as
the amount of output produced for each hour a machine is worked.
FAIRNESS OF AUTOMATION AND OUTSOURCING 11

body of work on time consumption, researchers have found that people value staying busy and
productive (e.g., Hsee, Yang, & Wang, 2010; Hsee, Zhang, Cai, & Zhang, 2013; Wilson et al.,
2014), and Yang & Hsee (2018) have even proposed that people pursue goals in order to stay
productive (instead of stay productive in order to pursue goals). Put differently, goals may not
motivate behavior as previously thought (Austin & Vancouver, 1996; Emmons, 1996), but
rather, people may be innately geared toward staying occupied in a meaningful way. For
example, in a clever demonstration of people’s need for justifiable busyness, Hsee, Yang, and
Wang (2010) gave participants a choice between submitting a completed survey and receiving a
piece of chocolate at a nearby location, which would result in 15 minutes of idle waiting before
the next task, or a location further away that would take 12-15 minutes roundtrip. Whereas most
participants (68%) who were told that the same chocolate would be offered at both locations
chose the closer location, most participants (59%) who were told that the two locations would
each offer a different type of chocolate went to the further location. Furthermore, those who
visited the further location, regardless of whether or not it was by choice, were happier than their
counterparts who visited the closer location. In another study, T. D. Wilson et al. (2014) found
that many participants (67% of men, 24% of women) who had indicated that they would pay
money to avoid aversive electric shocks chose to self-administer these shocks to avoid being
alone with their thoughts for a 15-minute “thinking period.” These studies converge on the
notion that people are averse to idleness and desire to be productive.
Having observed an intriguing preoccupation with maximizing productivity, as part of an
unrelated project, I asked participants to note one way in which they try to be efficient and the
reasons why they do so (Kim & Wakslak, 2017). Efficiency in this study was defined as aiming
to be as productive as possible, with minimum wasted time or effort. Analyses of the responses
FAIRNESS OF AUTOMATION AND OUTSOURCING 12

yielded a substantial redundancy in people’s answers: 58% of the reasons given cited elements of
being productive (e.g., “save energy,” “save time,” “gain time,” “do more”). With saving and
earning money also classified as elements of productivity, the proportion of responses in this
category rose to 74%. The most common reason for trying to be productive, was essentially, to
be productive. Thus, productivity per se may be a powerful motivator of human behavior.
The literature on well-being also provides support for the basic human need to stay
engaged and productive. Eudaimonic conceptions of happiness suggest that the good life, or true
human flourishing, materializes when striving to realize one’s potential and living with purpose
(Ryan & Deci, 2001; Ryff, 1995; Waterman, 1993), and studies have shown that the pursuit of
personally meaningful goals lead to well-being (Brunstein, 1993; Emmons, 1996; Ryan & Deci,
2000; Sheldon & Elliot, 1999; Sheldon & Kasser, 2001). Not surprisingly, involuntary
unemployment, which reduces opportunities for self-actualization strivings, predicts a variety of
negative effects on psychological and physical health, such as higher rates of depression,
mortality, and cardiovascular disease (Jin, Shah, & Svoboda, 1995; Mathers & Schofield, 1998;
Roelfs, Shor, Davidson, & Schwartz, 2011; Strully, 2009; S. H. Wilson & Walker, 1993). This
relationship between productive engagement and well-being has also been documented at the
cognitive level: Killingsworth and Gilbert (2010) found that people were happiest when their
minds were not wandering (i.e. they were engaged in the current activity) than when their minds
were wandering (even to pleasant thoughts).
The literature and findings described thus far depict making good use of one’s time as a
fundamental human value that motivates behavior and confers well-being. Indeed, productivity-
related terms appear on various lists of foundational human values. For instance, Schwartz
(2012)’s list of basic values, which are theorized to be universal, include “achievement,” and
FAIRNESS OF AUTOMATION AND OUTSOURCING 13

Rokeach (1973)’s list of terminal values include “a sense of accomplishment.” The values in
these lists represent desirable end-goals, the ordering of which may differ across individuals
(Schwartz & Bilsky, 1987). However, in light of Yang and Hsee (2018)’s suggestion that people
dislike idleness and pursue goals in order to stay busy, I raise the possibility that these
foundational human values may in fact constitute various means for staying productive. For
example, power and hedonism are also basic human values posited by Schwartz (2012). It may
very well be that seeking power affords those who value power a highly attractive and
meaningful option for staying productive, and for those who value hedonism, the pursuit of
pleasure may be a fulfilling way to keep occupied.
I propose that for many, staying busy and productive, whether by pursuing power,
hedonism, achievement, or any other value, may be the ultimate goal. Productivity may be the
standard by which we implicitly measure the worth of our own lives as well as others’ lives.
Corroborating this idea, a neuroimaging study revealed that pictures of the homeless and addicts,
who are among the least productive from society’s vantage point, did not activate regions of the
brain that are associated with social cognition and person perception (Harris & Fiske, 2006). The
productivity-as-worth notion is also reflected in the Protestant Work Ethic (Furnham, 1984;
Weber, 1958), which is a lay theory that pervades much of the United States (and other Western
nations) and suggests that hard work, the productive use of time, and frugality will lead to
economic prosperity. It seems clear that productivity is a value reified by individuals and
embraced by industrialized Western society. Given this, I predict that individuals will bestow
more charitable judgments on any action taken by an organization for the sake of increasing
productivity. More specifically, I predict that people will perceive automation-driven layoffs,
which signal increasing productivity, to be more legitimate than outsourcing-driven layoffs,
FAIRNESS OF AUTOMATION AND OUTSOURCING 14

which generally do not convey attempts to increase productivity (Global Sourcing Association,
2015).
The unfairness of cutting costs. In contrast, laying off workers in order to cut costs is
likely to be viewed as highly unfair. This argument is based on several findings. First, in an
examination of fairness perceptions of various profit seeking actions by businesses, Kahneman et
al. (1986b) demonstrated that a majority of people (75-77%) viewed attempts by a profitable
organization to increase their profits by lowering worker wages as unfair. This suggests that,
although organizations are in the business of making money, it is not seen as legitimate for them
to achieve this aim by lowering costs in a manner that comes at the expense of their workers.
This point was echoed by Van Buren (2000), who proposed that organizations that downsize
“merely to increase an already-adequate rate of profit…are likely to be held more culpable for
such actions (p. 214)” than organizations that do so due to declining returns, because downsizing
when there is no threat to the viability of the organization violates psychological and social
contracts. Similarly, Schulz and Johann (2018) found that organizations that gave efficiency-
related, or more specifically cost-cutting reasons for downsizing (as opposed to external reasons
about declining demand), were not viewed favorably and suffered decreases in firm reputation;
here, firm reputation was a composite of ratings by senior executives, outside directors, and
financial analysts in the respective industry on eight criteria, such as quality of management,
wise use of corporate assets, and responsibility to the community and the environment.
In sum, cutting labor costs increases efficiency but it does so at the expense of
employees, which may generate a sense of injustice and wrongful harm (e.g., Kennedy, 2008).
For this reason, I predict that people will perceive outsourcing-driven layoffs, which
communicate a primary concern with reducing costs, as less acceptable than automation-driven
FAIRNESS OF AUTOMATION AND OUTSOURCING 15

layoffs. To be clear, from a business owner’s perspective, lowering labor costs is a valid method
to increase productivity: I can hire two workers if I pay them each half as much. My argument,
however, is that people do not spontaneously make these calculations, which would highlight the
moral equivalence of the respective decisions (cf., Kahneman et al., 1986b); rather, I suspect that
different strategies call to mind different objectives that are seen as differentially legitimate.
Hypotheses
Having established that there is a basic difference in how people think about automation
and outsourcing, I hypothesize that people will perceive automation-induced layoffs to be fairer
than outsourcing-induced layoffs. Moreover, I hypothesize that direct manipulations of focus on
increasing productivity and saving cost will produce similar effects; specifically I predict that
people will judge layoff decisions that appear to increase productivity (analogous to automation)
as fairer than layoff decisions that appear to center on cutting costs (analogous to outsourcing). I
expect the findings of the present investigation to contribute to the literature on layoffs and
justice in two ways. First, although prior research and theory suggest that layoffs implemented
by an organization that is not facing threats to survival is likely to be seen as unfair (Bies, 1987;
Kahneman, et al., 1986; Shore & Tetrick, 1994; Van Buren, 2000), I reveal when and how such
layoffs can be deemed fair: when the layoffs are motivated by a desire to automate, or more
broadly a desire to increase productivity and capabilities. In doing so, I raise the possibility that
the pursuit of productivity can comprise an additional basis for justice perceptions. Extending
previous research indicating that productivity is a prized goal among individuals, I demonstrate
that individuals apply this deep-seated partiality for productivity to judgments of the fairness of
organizational actions, rendering more charitable judgments for organizations that seek to
increase productivity.

FAIRNESS OF AUTOMATION AND OUTSOURCING 16

Overview of Studies

I test my ideas in six studies that explore public attitudes toward automation versus
outsourcing, and, more generally, job displacement that enhances productivity versus saves
money. As previously done by other researchers examining impartial third-party observers’
perceptions of fairness (Kahneman, et al., 1986b; Konow, 2003, 2005; Pfeifer, 2007), I use
hypothetical scenarios. Given the arguments presented above, I predict that although automation
and outsourcing both entail the transfer of work to an external entity, people will perceive layoffs
ensuing from decisions to automate to be fairer than those following from decisions to outsource.
I establish that this is the case in Studies 1a and 1b. Studies 2a and 2b then consider whether a
productivity-enhancing versus cost-saving divergence could drive this difference by keeping
constant the human-replacement element, but varying whether this human replacement is more
productive (analogous to automation) or less expensive (analogous to outsourcing). In Study 3,
to better understand what specifically about productivity increases and cost-saving impacts
fairness perceptions, I consider a number of mediating mechanisms that may drive this
relationship. Finally, Study 4 considers the effect of motivation framing, exploring whether
describing the decision as being motivated by enhancing capabilities or decreasing costs can
impact perceptions of automation and outsourcing. All studies were preregistered on
Aspredicted.org; the Appendices include materials, alternative analyses, and pre-registration
links, as well as descriptions of several additional studies that are not included in the text (see
Table 1 for a summary of the Appendices content). Data and syntax files are available on OSF
(http://osf.io/rjqfm).
Table 1.
Index of Appendices
Section Contents
Appendix A. Pilot study analyses
Appendix B. Study 1b: additional measures and analyses
Appendix C. Study 2b: results with outliers excluded
FAIRNESS OF AUTOMATION AND OUTSOURCING 17

Appendix D. Study 3: results with outliers excluded
Appendix E. Study 4: scenarios presented to P's
Appendix F. Study 4: results with outliers excluded
Appendix G. Study 4: additional measures and analyses
Appendix H. Additional studies: variations of Study 4
Appendix I. Additional study: mediation
Appendix J. Additional studies: the role of progress
Appendix K. Pre-registration links

Study 1A

In Study 1a I tested whether individuals would view layoffs caused by automation as
more legitimate than layoffs caused by outsourcing.
Method

Participants. One hundred ninety-nine participants were recruited through Amazon’s
Mechanical Turk (Mturk). In this and all further studies I set target sample sizes at or around 100
per cell, as noted in the pre-registration. I used this sample size because it provides 80% power to
detect effects of d = .4, the typical effect size in social/personality psychology (Richard, Bond &
Stokes-Zoota, 2003). Any deviation from the pre-registered number was due to Mturk over- or
under-filling slots. Four observations with duplicate IP addresses were excluded due to suspicion
that those individuals had participated more than once.
2
The final sample consisted of 195
participants (39% female, 61% male; mean age = 34.82, SD = 10.72). Sixty-six point two percent
of respondents reported being employed full-time, 11.8% employed part-time, 13.8% self-
employed or an Mturker, 3.6% full-time homemaker, and 4.6% retired or other.
Procedure. Participants were randomly assigned to read a simple hypothetical scenario
either about job displacement caused by automation or job displacement caused by outsourcing.

2
In this and Study 2a I accidentally failed to include this exclusion criteria in the pre-registration. However, reported
results do not appreciably change when analyses are run without excluding these respondents.
FAIRNESS OF AUTOMATION AND OUTSOURCING 18

In the automation condition, participants read the following text: “A manufacturing company
decides to lay off 100 employees because it will be automating some of its operations.” In the
outsourcing condition, participants read the following text: “A manufacturing company decides
to lay off 100 employees because it will be outsourcing some of its operations.”
Dependent variable. A large justice literature has identified many different specific
components of fairness (for a review see Colquitt, et al., 2001). My interest was not in
differentiating between these different judgments, but rather measuring broad perceptions of
fairness and legitimacy. I measured such perceptions with four items: “I find this decision
acceptable,” “It is fair for the company to do this,” “The company has the right to do this,” and
“It is ethical for the company to do this” (α = .86). Items were measured on a 7-point scale going
from “strongly agree” to “strongly disagree” and were recoded before analysis so that higher
scores signified greater fairness perceptions.
Results

Consistent with My hypothesis, respondents perceived layoffs caused by automation (M

= 4.55, SD = 1.48) to be more fair than layoffs caused by outsourcing (M = 3.60, SD = 1.34),

t(193) = 4.68, p < .001, mean difference = .95, 95%CI = [.55, 1.34], d = .67.

Study 1B

Study 1b sought to replicate the results of Study 1a, and further compare automation-
related judgments to a third condition: layoffs ensuing from a company merger. In this new
condition the layoffs are a byproduct of an organizational decision that is not directly about
altering employment arrangements; I selected this as an alternative comparison because it seems
on the surface a highly legitimate reason for layoffs.
Method

Participants. Three-hundred four participants were recruited through Amazon’s
FAIRNESS OF AUTOMATION AND OUTSOURCING 19

Mechanical Turk. I excluded 9 observations with duplicate IP addresses due to suspicion
that those individuals had participated more than once. The final sample consisted of 295
participants (42% female, 58% male; mean age = 34.72, SD = 10.40). Sixty-two point four
percent reported being employed full-time, 9.5% employed part-time, 21.1% self-employed or
an Mturker, 4.1% full-time homemaker, and 3.1% retired or other.
Procedure. Participants were randomly assigned to read a simple hypothetical scenario
about job displacement caused by: 1) automation, 2) outsourcing, or 3) a company merger. I used
the same scenarios from Study 1a for the automation and outsourcing condition, and in the
merger condition, participants were presented with the following text: “A manufacturing
company decides to lay off 100 employees because it recently merged with another company and
there are departments with overlapping functions.”
Dependent variable. I measured respondents’ fairness perceptions with three of the four
items used in Study 1a that yielded the highest reliability: “I find this decision acceptable,” “It is
fair for the company to do this,” and “It is ethical for the company to do this” (α = .92). Items
were measured on a 7-point scale going from “strongly agree” to “strongly disagree” and were
recoded before analysis so that higher scores signified greater fairness perceptions.
Results

An ANOVA on fairness perceptions across the three conditions yielded marginally
significant differences across conditions, F(2, 292) = 2.75, p = .066, η2 = .02. An LSD post hoc
test revealed that, consistent with my prediction and finding in Study 1a, fairness perceptions in
the automation condition (M = 4.05, SD = 1.42) were higher than in the outsourcing condition (M
= 3.63, SD = 1.69) at a marginally significant level, p = .069, mean difference = -.42, 95%CI =
FAIRNESS OF AUTOMATION AND OUTSOURCING 20

[-.87, .03]. Fairness perceptions in the merger condition (M = 4.12, SD = 1.69) were also
significantly higher than in the outsourcing condition (M = 3.63, SD = 1.69), p = .030, mean
difference = -.50, 95%CI = [-.94, -.05]. There was no statistically significant difference in
fairness perceptions between the automation and merger conditions, p = .738, mean difference =
-.08, 95%CI = [-.53, .38], suggesting that the decision to lay off workers in order to automate is
perceived to be just as acceptable as laying off workers as a consequence of redundancies from a
merger. The lack of difference points to a fascinating perceived legitimacy of job loss due to
automation. I attempted to unpack this phenomenon further in subsequent studies.
Study 2A

In Studies 1a and 1b I established that people see job displacement due to automation as
more acceptable than job displacement due to outsourcing. Studies 2a and 2b explore one
potential reason for this divergence. Although automation and outsourcing both can increase an
organization’s efficiency and decrease the cost of production, automation is more generally
associated with increased productivity (organizations turn to automation to be able to produce
more goods in a given amount of time) and outsourcing with cost saving (organizations turn to
outsourcing to be able to pay workers less for the same amount of goods produced). In Studies 2a
and 2b I sought to see if this difference on its own would lead to differences in fairness
perceptions. To this end, I removed any mention of automation or outsourcing per se (labels that
are associated with a variety of differences and connotations), and instead described a situation
where a worker is replaced with a more productive worker (analogous to automation) or a
worker is replaced with a cheaper worker (analogous to outsourcing).
Method

Participants. One hundred ninety-seven participants were recruited through Amazon’s
FAIRNESS OF AUTOMATION AND OUTSOURCING 21

Mechanical Turk. Six observations with duplicate IP addresses were excluded due to
suspicion that those individuals had participated more than once. The final sample consisted of
191 participants (42.9% female, 57.1% male; mean age = 33.87, SD = 9.62). Fifty-seven point
six percent reported being employed full-time, 8.4% employed part-time, 22% self-employed or
Mturker, 6.8% full-time homemaker, and 5.2% retired or other.
Procedure. Participants were randomly assigned to read a scenario in which an
organization decided to lay off a worker to hire a more productive worker (analogous to
automation) or to hire a cheaper worker (analogous to outsourcing). Note that I did not include
any mention of automation or outsourcing per se. In the productive replacement condition,
participants read the following information: “A manufacturing company decides to lay off a
worker who was making 10 widgets/hour and getting paid $20/hour, in order to hire a worker at
$20/hour who is able to make 20 widgets/hour. In other words, the company is firing a worker in
order to hire a replacement who works at a faster rate than that worker. In this way, the company
will be able to produce more for the same cost.” In the cheaper replacement condition,
participants read the following scenario: “A manufacturing company decides to lay off a worker
who was making 10 widgets/hour and getting paid $20/hour in order to hire a replacement at
$10/hour who will make the same 10 widgets/hour. In other words, the company is firing a
worker in order to hire a replacement who works at a cheaper rate than that worker. In this way,
the company will be able to produce more for the same cost.” The information for the relevant
condition was also presented to participants in a table (see Figure 1). Of note, across scenarios
the replacement worker’s pay-per-unit rate was identical ($1 per widget); this was done to
explore whether people might see these decision as differentially fair, even though the actual
FAIRNESS OF AUTOMATION AND OUTSOURCING 22

efficiency of the decision was similar from the company’s perspective (as also reflected in the
final statement of both scenarios that the company would be able to produce more for the same
cost).
a.
Worker Output Capability Pay
Old worker 10 widgets/hour $20/hour
New worker 20 widgets/hour $20/hour

b.
Worker Output Capability Pay
Old worker 10 widgets/hour $20/hour
New worker 10 widgets/hour $10/hour
Figure 1.
Tables shown to participants in the productive replacement (a) and cheaper replacement (b)
conditions in Study 2a.

Dependent variable. I measured respondents’ fairness perceptions with the same three
items used in Study 1b (α = .94). Items were administered on a 7-point scale going from
“strongly disagree” to “strongly agree.”
Results

As predicted, respondents perceived layoffs that were executed in order to hire a more
productive worker to be more fair (M = 4.80, SD = 1.66) than layoffs that were executed in order
to hire a less expensive worker (M = 3.42, SD = 1.70), t(189) = 5.67, p < .001, mean difference =
-1.38, 95%CI = [-1.86, -.90], d = .82. These results parallel the pattern that I found in Studies 1a
and 1b. Notably, I obtained these results with two scenarios that were equivalent in terms of the
efficiency level of the new workers, with 1 widget being produced for every dollar spent.
However, the two scenarios were dissimilar in terms of the level of output produced by the new
worker (cheaper replacement condition: 10 widgets/hour, productive replacement condition: 20
widgets/hour). Arguably, an organization could use the cost-savings from hiring a cheaper
employee to hire an additional employee and keep production constant. I therefore ran Study 2b
FAIRNESS OF AUTOMATION AND OUTSOURCING 23

in order to see if the effect would still be obtained if total output levels remained constant across
condition.
Study 2B

Study 2a revealed that layoffs deriving from the decision to hire more productive workers
(analogous to automation) were seen to be fairer than layoffs caused by the decision to hire
cheaper workers (analogous to outsourcing), even though the company’s cost-per-unit was
equivalent across condition (1 widget/$1). Study 2b sought to replicate this finding with the
following modification: I specified in the cost-saving condition that two workers would replace
the old worker, explicitly keeping total output the same across condition. I also measured
willingness to invest in the company as another dependent variable.
Method

Participants. Two hundred participants were recruited through Amazon’s Mechanical
Turk. One observation with a duplicate IP address was excluded due to suspicion that this
individual had participated more than once. The final sample consisted of 199 participants
(41.7% female, 58.3% male; mean age = 34.01, SD = 10.66). Fifty-five point three percent
reported being employed full-time, 11.6% employed part-time, 22.6% self-employed or Mturker,
5.0% full-time homemaker, and 5.5% retired or other.
Procedure. As in Study 2a, participants were randomly assigned to read a scenario in
which an organization decided to lay off a worker in order to hire a more productive worker
(analogous to automation) or to hire a cheaper worker (analogous to outsourcing). In the
productive replacement condition, participants read the same scenario that was used in Study 2a.
In the cheaper replacement condition, participants read the following scenario, which noted the
hiring of two workers in place of the old worker: “A manufacturing company decides to lay off a
FAIRNESS OF AUTOMATION AND OUTSOURCING 24

worker who was making 10 widgets/hour and getting paid $20/hour in order to hire 2
replacements at $10/hour who will each make the same 10 widgets/hour. In other words, the
company is firing a worker in order to hire replacements who work at a cheaper rate than that
worker. In this way, the company will be able to produce more for the same cost.” The
information for the relevant condition was again presented to participants in a table (see Figure
2). In these scenarios, the unit per pay rate was identical for each worker and the output produced
by the replacement workers was also equivalent across conditions.
a.
Worker Output Capability Pay
Old worker 10 widgets/hour $20/hour
New worker 20 widgets/hour $20/hour

b.
Worker Output Capability Pay
Old worker 10 widgets/hour $20/hour
2 new workers 10 widgets/hour each $10/hour each

Figure 2.
Tables shown to participants in the productive replacement (a) and cheaper replacement (b)
conditions in Study 2b.

Dependent variables.

Fairness. I used the same 3-item scale used in Study 1b and Study 2a to measure fairness
perceptions (α = .94).
Willingness to invest. I also included a one-item measure of willingness to invest in the
company: “To what extent would you be willing to invest in this company?” This item was
administered on a 7-point scale going from “Not at all likely” to “Very likely.”
Results

Fairness. As predicted, respondents perceived layoffs that were followed by the hiring of
more productive workers to be fairer (M = 5.08, SD = 1.62) than layoffs that were accompanied
FAIRNESS OF AUTOMATION AND OUTSOURCING 25

by the hiring of cheaper workers (M = 3.90, SD = 1.65), t(197) = 5.10, p < .001, mean difference

= -1.18, 95%CI = [-1.64, -.73], d = .72.

Willingness to invest. Similar to fairness perceptions, respondents were more willing to
invest in the company if workers were fired in order to hire more productive workers (M = 4.77,
SD = 1.62) than when the same action was taken to hire cheaper workers (M = 3.63, SD = 1.95),
t(194) = 4.50, p < .001, mean difference = -1.14, 95%CI = [-1.64, -.64], d = .64. Levene’s test
indicated unequal variances (F = 12.21, p = .001), so degrees of freedom were adjusted from 197
to 194.
Study 3

In Studies 1a and 1b people evaluated automation-induced layoffs to be more acceptable
than outsourcing-driven layoffs and Studies 2a and 2b demonstrated a similar effect with
scenarios in which workers were laid off and more productive workers were hired (typifies
automation) or cheaper workers were hired (typifies outsourcing). Together these findings
suggest that people see actions that increase productivity as fairer for a company to pursue than
actions taken in order to save costs. What is still unclear, however, is why this is the case; what
exactly about firing workers to hire more productive workers versus cheaper workers makes it
seem more legitimate? I suspected that this effect would be multiply determined, and in Study 3
considered several potential mediating mechanisms.
Proposed Mechanisms

Devaluation and disenfranchisement of workers. Studies on psychological contracts,
or the individual beliefs about reciprocal obligations between employees and employers,
suggests that employees expect job security and advancement opportunities in exchange for
loyalty and hard work (e.g., Rousseau, 1990). Although all layoffs that are not related to
FAIRNESS OF AUTOMATION AND OUTSOURCING 26

discharge for cause (e.g., poor performance) implicate violations of psychological contracts, the
salience of the cost-saving dimensions is likely to highlight the expendable nature of workers in
the face of profit. Because human life is generally considered sacred (e.g., Zelizer, 1978), and
people are averse to trading off sacred values for secular ones, such as money (Tetlock, 2003), a
cost-saving (compared to a productivity-increasing) focus should enhance perceptions of the
commodification (i.e. devaluation) of workers. In addition, the cost-saving focus, by virtue of
spotlighting the loss of employees’ livelihoods in exchange for mere profit by the firm (versus
the lofty pursuit of productivity) should heighten the perception that the employees have lost
something rightfully belongs to them (i.e. disenfranchisement). I predicted that these ensuing
perceptions of devaluation and disenfranchisement of the worker would in turn decrease
perceptions of fairness.
Progress and quality. In addition to concerns about the workers’ treatment and rights, I
predicted that beliefs about the consequences of productivity would also mediate the relationship
between layoff condition and fairness perceptions. It is possible that hiring a more productive
worker, rather than a less expensive worker, would signal better potential performance outcomes
(i.e. progress) for the company, both in the short- and long-term. Also, given the common
tendency to make inferences about quality from price (Rao & Monroe, 1989; Shugan, 1984;
Zeithaml, 1988), replacement with a more productive worker as opposed to a cheaper worker
should lead to higher output-quality assumptions. I anticipated that these judgments of progress
and quality would in turn influence fairness perceptions.
Benefit to company over other stakeholders. Given that the participants were impartial
observers who did not have any stake in the described (hypothetical) company, I anticipated that
they would be sensitive to the relative distribution of outcomes to those initiating the layoffs (i.e.
FAIRNESS OF AUTOMATION AND OUTSOURCING 27

the company) and other stakeholders, such as customers and society. In addition, because layoffs
are a downsizing strategy that is employed in order to cut costs and increase efficiency (Cameron
et al., 1993; S. J. Freeman & Cameron, 1993), they should be construed as benefiting the
company more than other stakeholders. However, I predicted that layoffs with a focus on
increasing productivity (versus decreasing costs)—the fruits of which could be translated into
more products for customers and ultimately society—would lead to decreased evaluations of
preferential benefit to company over customers and society, which would ultimately results in
higher fairness evaluations.
Motivation and inevitability. Finally, I included measures of beliefs about the nature of
the decision as potential mediators. There have been some studies suggesting that the conveyed
motive behind layoff decisions differentially affects outcomes for the firm. For example, the
effect of layoffs on the firm’s stock performance was more negative when the downsizing was
attributed to declining external demand versus a desire to increase efficiency (Palmon et al.,
1997) and the effect of layoffs on corporate reputation was more negative when the layoffs were
attributed to concerns about enhancing efficiency versus external decline in demand (Schulz &
Johann, 2018). In these studies, however, “efficiency” was operationalized as cost-cutting, which
is only 1 dimension of the efficiency calculus. Increasing productivity is the other dimension,
and layoffs that highlight this aspect as opposed to the cost-cutting dimension are likely to be
construed as having been more motivated by the desire to increase capabilities. Increasing a
firm’s capabilities entails improving its effectiveness and ability to meet its goals and given
people’s preoccupation with productivity and making good use of their time, to the extent that a
firm seems motivated by upgrading their capabilities, the layoffs should seem more fair.
Conversely, layoffs with cheaper (versus more productive) replacements should be construed as
FAIRNESS OF AUTOMATION AND OUTSOURCING 28

having been more motivated by cost-reduction, which in turn should decrease perceptions of
fairness. I note that perceptions of the organization’s motivation to increase their capabilities or
decrease cost is closest to the manipulation itself, but distinct in its measurement not of what the
organization did on a behavioral level, but the perceived underlying motivation for this behavior.
Lastly, due to the ubiquity of productivity pursuits, productivity-enhancing efforts may
seem more inevitable than cost-cutting efforts. Given research showing that people justify and
endorse to a greater degree systems that seem more inescapable (Kay & Friesen, 2011; Laurin,
Shepherd, & Kay, 2010), I predicted that layoffs signaling increases in productivity (versus
reductions in cost) would increase judgments of inevitability and ultimately fairness perceptions.
Method
We ran this study twice. The first iteration (Sample 1) included measures of all of the
proposed mediators listed above, with the exception of perceived inevitability and the underlying
motivation measures. In the second iteration (Sample 2), which took place 9 days later, I
included these three additional potential mediators.
Participants. Participants were recruited from Amazon’s Mechanical Turk(NSample1 =
202, NSample2 = 201, NCombined = 403). Several observations with duplicate IP addresses were
excluded due to suspicion that the individuals had participated more than once (Sample1: 8,
Sample2: 1). The final samples consisted of 194 participants for Sample 1, 200 participants for
Sample 2, and a total of 394 participants across the two samples (36.5% female, 62.7% male,
and .8% other; mean age = 34.89, SD = 10.74). Sixty-three point two percent of respondents
reported being employed full-time, 8.6% employed part-time, 18.8% self-employed or an
Mturker, 3.6% full-time homemaker, and 5.8% retired or other.
FAIRNESS OF AUTOMATION AND OUTSOURCING 29

Procedure. Participants saw scenarios and tables similar to those from Studies 2a-2b,
describing a hypothetical manufacturing company that either decided to lay off a number of
workers in order to hire more productive workers (analogous to automation) or to hire cheaper
workers (analogous to outsourcing). Three small change from Studies 2a & 2b were made:
Layoffs were described as impacting a number of workers rather than focusing on the decision to
lay off a single worker, the company was called “ALY Industries,” and the payment and output
levels were adjusted such that the old workers made 20 widgets/hour for $30/hour and the new
workers made 30 widgets/hour for $30/hour (more productive worker condition) or 20
widgets/hour for $20/hour (cheaper worker condition). These changes were made to ensure that
the effects are robust across such small surface-level changes. The payment/output levels were
specifically adjusted to make sure that the prior effects were not driven by the cheaper workers
getting paid an amount that could be perceived as unfairly low ($10 an hour).
Dependent variables. I measured fairness perceptions using the same scale from Studies
2a & 2b (α Sample1 = .94, α Sample2 = .95, α Combined = .95).
Proposed mediators. All mediator items were measured on 7-point scales going from
“strongly disagree” to “strongly agree.”
Devaluation of workers. Three items measured perceived devaluation of workers: “This
decision devalues workers,” “This decision treats people like a commodity,” and “This decision
objectifies workers” (α
Sample1
= .89, α
Sample2
= .93, α
Combined
= .91).

Disenfranchisement of workers. Three items measured disenfranchisement of workers:
“This decision violates an implicit working agreement between the original workers and ALY,”
“This decision disenfranchises workers from jobs that rightfully belong to them,” and “This
FAIRNESS OF AUTOMATION AND OUTSOURCING 30

decision deprives the original workers of something to which they are entitled” (α Sample1 = .93,

α Sample2 = .91, α Combined = .92).

Short-term Progress. Four items assessed the degree to which respondents believed the
decision was conducive to organizational progress in the near future: “ALY is likely to grow in
the upcoming year,” “ALY is likely to become a more successful company in the upcoming
year,” “ALY is likely to become more profitable in the upcoming year,” and “ALY is likely to
make substantial progress toward its goals in the upcoming year” (α Sample1 = .91, α Sample2 = .92,
α
Combined
= .91).

Long-term Progress. Four items assessed perceptions of long-term progress; these were
the same items used to measure short-term progress but with “in the upcoming year” replaced
with “in the next 10 years” (α
Sample1
= .95, α
Sample2
= .96, α
Combined
= .96).

Quality of output. Three items measured projected quality of the output: This decision...
“...will cause the quality of widgets to improve,” “...may result in diminished quality of widgets
(r),” and “...may jeopardize the quality of current widgets (r)” (α Sample1 = .73, α Sample2 = .64,
α Combined = .69).
Benefit to company over society. Three items measured perceptions of benefit to the
company over society: This decision ultimately... “...provides more benefits to ALY than
society,” “...is more in the interest of ALY than society at large,” and “...reflects greater concern
for itself (ALY) than society” (α Sample1 = .93, α Sample2 = .93, α Combined = .93).
Benefit to company over consumer. Three items measured perceptions of benefit to the
company over the consumer: This decision ultimately... “...provides more benefits to ALY than
its customers,” “...is more in the interest of ALY than its customers,” “...reflects greater concern
for itself (ALY) than its customers” (αSample1 = .89, αSample2 = .87, αCombined = .88).
FAIRNESS OF AUTOMATION AND OUTSOURCING 31

Inevitability. In addition to the mediators listed above, participants in Sample 2 evaluated
the seeming inevitability of each decision on three items: “ALY’s decision was inevitable (it was
bound to happen),” “The world is moving inevitably in the direction of ALY’s choice,” and
“ALY’s decision reflects a larger trend that is unstoppable” (α Sample2 = .89).
Motivation. Sample 2 participants also completed two scales assessing the motivation
behind ALY’s decision: increasing capabilities and decreasing costs. Three items assessed the
extent to which respondents viewed ALY as having been motivated by the desire to enhance
their capabilities: ALY’s decision was motivated by... “...a desire to achieve excellence,” “...a
desire to improve their capabilities,” and “...a desire to continuously improve” (α Sample2 = .91)
and three items assessed the degree to which respondents believed ALY had been motivated by
cutting costs: ALY’s decision was motivated by… “...a desire to increase their bottom line,” “...a
desire to be as cost-efficient as possible,” and “...a desire to save money” (α Sample2 = .71).
Results

In my analyses of main effects of layoff reason on fairness perceptions and the mediating
variables that were administered in both samples, I report the results from the combined sample
(collapsed across Sample 1 and Sample 2). Table 2 displays the analyses conducted within each
of the individual samples, including the means and standard deviations by condition and the
difference statistics (t-statistic, p-value, mean difference, 95% confidence intervals, and effect
size) for all of the measures that were collected in the respective samples. Mediation analyses are
reported separately for each sample, given that not all measured mediators were present in both
samples.
The Effect of Layoff Reason on Fairness. Consistent with Studies 2a and 2b, fairness
perceptions were higher when workers were laid off and more productive workers were hired (M
FAIRNESS OF AUTOMATION AND OUTSOURCING 32

= 4.41, SD = 1.73) than when workers were laid off and less expensive workers were hired (M =
3.27, SD = 1.86), t(391) = 6.29, p < .001, mean difference = -1.14, 95%CI[-1.49, -.78], d = .64.
Levene’s test indicated unequal variances (F = 4.28, p = .039), so degrees of freedom were
adjusted from 392 to 391. This difference in fairness perceptions was reliable across the two
samples (see Table 2).
The Effects of Layoff Reason on the Mediators. Collapsing across the two samples,
devaluation of workers and disenfranchisement of workers showed significant differences across
condition. Devaluation of workers was higher in the cheaper replacement condition (M = 5.73,
SD = 1.36) than the productive replacement condition (M = 5.18, SD = 1.57), t(392) = 3.77, p
< .001, mean difference = .55, 95%CI = [.26, .84], d = .38. Disenfranchisement was similarly
higher in the cheaper replacement condition (M = 4.88, SD = 1.62) than the productive
replacement condition (M = 4.10, SD = 1.70), t(392) = 4.66, p < .001, mean difference = .78,
95%CI = [.45, 1.11], d = .47. These differences were also significant within each of the
independent samples (see Table 2).
Other mediators measured in both samples were significantly different across condition
in the combined sample, although did not reach significance within both independent samples
(see Table 2). Respondents expected the quality of goods to be higher in the productive
replacement condition (M = 3.35, SD = 1.22) than the cheaper replacement condition (M = 3.08,
SD = 1.11), t(392) = 2.33, p = .021, mean difference = -.27, 95%CI = [-.50, -.04], d = .23.
Conversely, they perceived greater benefit to the company over society in the cheaper
replacement condition (M = 6.00, SD = 1.13) than the productive replacement condition (M =
5.61, SD = 1.39), t(374) = 3.03, p = .003, mean difference = .39, 95%CI = [.14, .64], d = .31.
Levene’s test indicated unequal variances (F = 7.91, p = .005), so degrees of freedom were
FAIRNESS OF AUTOMATION AND OUTSOURCING 33

adjusted from 392 to 372. There were marginal differences in ratings of short-term progress, with
ratings higher in the productive replacement condition (M = 5.23, SD = 1.11) than the cheaper
replacement condition (M = 5.01, SD = 1.21), t(392) = 1.91, p = .056, mean difference = -.22,
95%CI = [-.45, .01], d = .19. Finally, there were no differences by condition in the combined
sample for the long-term progress or benefit to company over customer measures (although the
effect for long term progress was significant in Sample 1; see Table 2).
Three mediators were collected only in the second sample: Perceptions of the
organization’s motivation to increase their capabilities and their motivation to decrease their
costs, and decision inevitability. The organization’s motivations were perceived differently
across conditions. Perceived motivation to increase capabilities was higher in the productive
replacement condition (M = 4.82, SD = 1.63) than the cheaper replacement condition (M =
3.33, SD = 1.71), t(198) = 6.32, p < .001, mean difference = -1.49, 95%CI = [-1.96, -1.03], d
= .89, and perceived cost-saving motivation was lower in the productive replacement condition
(M = 5.91, SD = 1.12) than the cheaper replacement condition (M = 6.21, SD = .95), t(198) =
2.06, p = .040, mean difference = .30, 95%CI = [.01, .59], d = .29. There was no difference in
perceptions of decision inevitability (productive replacement: M = 4.80, SD = 1.48; cheaper
replacement: M = 4.54, SD = 1.59), t(198) = 1.23, p = .221, mean difference = -.27, 95%CI =
[-.70, .16], d = .17.

Table 2.
Descriptive and Difference Statistics for All Variables in Study 3, by
Sample
Sample 1
(N=194)
Sample 2
(N=200)

t p-value Mean Difference 95% CI of the Difference Cohen's d

Replacement Replacement Replacement Replacement Sample 1 Sample 2

over customer

over society

FAIRNESS OF AUTOMATION AND OUTSOURCING 34
Cheaper Productive Cheaper Productive
Sample

Sample

Sample

Sample

Sample

Sample Sample Sample
M (SD) M (SD) M (SD) M (SD)
1 2 1 2 1 2

1 2

Fairness 3.58 (1.84)

4.48 (1.65)

2.97 (1.84)

4.34 (1.81)

-3.59

5.32

<.001

<.001

-0.90

-1.37

[-1.40, -0.41]

[-1.88, -0.86]

0.51

0.75

Devaluation 5.55 (1.33)

5.07 (1.51)

5.90 (1.38)

5.27 (1.64)

2.33

2.96

0.021

0.003

0.48

0.63

[0.07, 0.88]

[0.21, 1.05]

0.34

0.42

Disenfranchisement 4.82 (1.59)

4.12 (1.63)

4.93 (1.64)

4.08 (1.78)

3.03

3.52

0.003

0.001

0.70

0.85

[0.25, 1.16]

[0.38, 1.33]

0.43

0.50

Short-term progress 5.08 (1.17)

5.24 (1.09)

4.93 (1.25)

5.22 (1.14)

-0.96

1.72

0.337

0.087

-0.16

-0.29

[-0.48, 0.16]

[-0.63, 0.04]

0.14

0.24

Long-term progress 4.80 (1.30)

5.18 (1.27)

4.91 (1.40)

4.96 (1.37)

-2.05

0.27

0.042

0.784

-0.38

-0.05

[-0.74, -0.01]

[-0.44, 0.33]

0.30

0.04

Quality 3.23 (1.10)

3.40 (1.24)

2.93 (1.10)

3.30 (1.20)

-1.02

2.29

0.310

0.023

-0.17

-0.37

[-0.50, 0.16]

[-0.69, -0.05]

0.15

0.32
Benefit to company
5.56 (1.28)

5.35 (1.50)

5.69 (1.36)

5.60 (1.31)

1.04

0.51

0.301

0.608

0.21

0.10

[-0.19, 0.60]

[-0.28, 0.47]

0.15

0.07
Benefit to company
5.85 (1.20)

5.55 (1.42)

6.15 (1.04)

5.66 (1.37)

1.54

2.81

0.125

0.005

0.29

0.48

[-0.08, 0.66]

[0.14, 0.82]

0.23

0.40
Motivation to
increase capabilities

3.33 (1.71)

4.82 (1.63)

6.32

<.001

-1.49

[-1.96, -1.03]

0.89
Motivation to
decrease costs

6.21 (.95)

5.91 (1.12)

2.06

0.040

0.30

[0.01, 0.59]

0.29

Inevitability

4.54 (1.59)

4.80 (1.48)

1.23

0.221

-0.27

[-0.70, 0.16]

0.17

FAIRNESS OF AUTOMATION AND OUTSOURCING 35

Mediation Analysis. I report below the mediation analysis separately for Sample 1 and
Sample 2. In each analysis, I include only the proposed mediators that were found to be
significantly different across condition in that respective sample (see Table 2). Because the
independent variable (layoff reason) is a dichotomous indicator variable, I report the
unstandardized coefficients.
Sample 1. For Sample 1, I tested whether the relationship between layoff reason and
fairness ratings was mediated by perceptions of 1) long-term progress, 2) devaluation of workers,
and 3) disenfranchisement of workers, using bootstrapping procedures for estimating direct and
indirect effects with multiple mediators (PROCESS macro, Model 4; Hayes, 2012; Preacher &
Hayes, 2008).
Consistent with results reported in Table 1, layoff reason was a significant predictor of
ratings of long-term progress, B = .38, p = .042, 95%CI = [.01, .74], devaluation, B = -.48, p
= .021, 95%CI = [-.88, -.07], and disenfranchisement, B = -.70, p = .003, 95%CI = [-1.16, -.25].

Regression of fairness on the proposed mediators and layoff reason indicated a significant effect
of devaluation, B = -.24, p = .005, 95%CI = [-.41, -.08], disenfranchisement, B = -.51, p < .001,
95%CI = [-.66, -.36], and long-term progress, B = .28, p < .001, 95%CI = [.14, .43]. The results
indicated that the total effect of layoff reason on fairness (B = .90, p < .001, 95%CI = [.41,
1.40]) became nonsignificant when the mediators were included in the model (direct effect of
layoff reason on fairness = .32, p = .080, 95%CI = [-.04, .69]). Tests of mediation with 5,000
bootstrapped samples revealed that the total indirect effect of layoff reason on fairness through
the 3 mediators was significant, with a point estimate of .58 and 95% bootstrap confidence
interval of [.25, .94]. In addition, there were significant indirect effects through each of the
proposed mediators: devaluation (point estimate = .12, 95%CI = [.01, .27]), disenfranchisement
FAIRNESS OF AUTOMATION AND OUTSOURCING 36

(point estimate = .36, 95%CI = [.12, .65]), and long-term progress (point estimate = .11, 95%CI

= [.00, .25]), suggesting mediation by each of these variables. The mediation model is shown in
Figure 3a.

Figure 3a.
Multiple mediation model from Study 3 (Sample 1) whereby the effect of layoff reason (0 =
decreasing cost, 1 = increasing productivity) on Fairness Perceptions is mediated by ratings of
Devaluation of Worker, Disenfranchisement of Worker, and Long-Term Progress. Values
represent unstandardized regression coefficients and the values in parentheses represent the
standard errors. The values before the slash show the total effect and the values after the slash
show the direct effect. * p < .05. **p < .01. ***p < .001.

Sample 2. For Sample 2, I tested whether the relationship between layoff reason and
fairness ratings was mediated by perceptions of 1) devaluation of workers, 2) disenfranchisement
of workers, 3) the quality of widgets, 4) benefit to company over society, 5) motivation to
increase capabilities, and 6) motivation to decrease costs, again using bootstrapping procedures
(PROCESS macro, Model 4; Hayes, 2012; Preacher & Hayes, 2008).
As reflected in Table 1, layoff reason was a significant predictor of devaluation, B = -.63,

p = .004, 95%CI = [-1.05, -.21], disenfranchisement, B = -.85, p < .001, 95%CI = [-1.33, -.38],
FAIRNESS OF AUTOMATION AND OUTSOURCING 37

quality, B = .37, p = .023, 95%CI = [.05, .69], benefit to company over society, B = -.48, p

= .006, 95%CI = [-.82, -.14], motivation to increase capabilities, B = 1.49, p < .001, 95%CI =
[1.03, 1.96], and motivation to decrease costs, B = -.30, p = .040, 95%CI = [-.59, -.01]. In
addition, when I regressed fairness on these 6 variables and the independent variable, I found a
statistically significant effect of devaluation, B = -.25, p = .004, 95%CI = [-.43, -.08],
disenfranchisement, B = -.39, p < .001, 95%CI = [-.53, -.26], and motivation to increase
capabilities, B = .36, p < .001, 95%CI = [.23, .48]. Quality, benefit to company over society, and
the motivation to decrease costs did not exert a statistically significant effect on fairness, B =
-.01, p = .931, 95%CI = [-.20, .18]; B = -.06, p = .520, 95%CI = [-.23, .12]; and B = .07, p = .509,

95%CI = [-.13, .26]; respectively.

The results indicated that the total effect of layoff reason on fairness (B = 1.37, p < .001,
95%CI = [.86, 1.88]) became marginally significant when the mediators were included in the
model (direct effect = .33, p = .088, 95%CI = [-.05, .72]). Tests of mediation with 5,000
bootstrapped samples revealed that the total indirect effect of layoff reason on fairness
perceptions through the six mediators was significant, with a point estimate of 1.04 and a 95%
bootstrap confidence interval of [.64, 1.45]. The analysis also yielded significant specific indirect
effects through devaluation (point estimate = .16, 95%CI = [.03, .35]), disenfranchisement (point
estimate = .34, 95%CI = [.12, .59]), and motivation to increase capabilities (point estimate
= .53, 95%CI = [.26, .87]), but not through quality (point estimate = -.00, 95%CI = [-.09, .08]),
benefit to company over society (point estimate = .03, 95%CI = [-.05, .11]), or motivation to
decrease costs (point estimate = -.02, 95%CI = [-.08, .03]). The mediation model is displayed in
Figure 3b.
FAIRNESS OF AUTOMATION AND OUTSOURCING 38

Figure 3b.
Multiple mediation model for Study 3 (Sample 2) whereby the effect of layoff reason (0 =
decreasing cost, 1 = increasing productivity) on Fairness Perceptions is mediated by ratings of
Devaluation of Worker, Disenfranchisement of Worker, and Motivation to Increase Capabilities.
Values represent unstandardized regression coefficients and the values in parentheses represent
the standard errors. The values before the slash show the total effect and the values after the slash
show the direct effect. * p < .05. ** p < .01. *** p < .001.

Across the two samples, devaluation and disenfranchisement of workers were consistent
mediators of the relationship between layoff reason and fairness judgments. In addition, long-
term progress was a significant mediator in Sample 1 (but not Sample 2) and the motivation to
increase capabilities was a significant mediator in Sample 2 (having not been measured in
Sample 1). Although the progress measures were created to tap the likelihood that the
organization would improve from the status quo (the focus is on outcomes) and the increasing
capabilities measure was intended to assess the extent to which the organization seemed
motivated by a desire to enhance their capabilities (the focus is on motivation), they both pertain
to growth, which suggests that organizational growth and improvement may be an important
factor that affects judgments of fairness in the context of layoffs. In the final study, I built on
these findings by specifically considering whether manipulations of the motivation underlying a
FAIRNESS OF AUTOMATION AND OUTSOURCING 39

layoff decision—either to increase capabilities or to decrease costs—could shape fairness
perceptions.
Study 4

Study 3 suggested multiple mediating paths supporting the relationship between layoff
reason and fairness perception. Study 4 focused more closely on the motivation behind the layoff
decision, returning to the automation/outsourcing distinction and exploring whether framing
automation/outsourcing decisions as having been motivated by increasing capabilities versus
decreasing costs would impact perceptions of the fairness of these decisions.
Method

Participants. Four-hundred twelve participants were recruited through Amazon’s
Mechanical Turk. Four observations with duplicate IP addresses were excluded due to suspicion
that the individuals had participated more than once. The final sample consisted of 408
participants (mean age = 36.05, SD = 11.83; 40.4% female, 59.6% male). Fifty-nine point six
percent of respondents reported being employed full-time, 10.5% employed part-time, 21% self-
employed or Mturker, 3.9% full-time homemaker, and 4.9% retired or other.
Procedure. I used a 2 (layoff reason: automation or outsourcing) X 2 (motivation
framing: increasing capabilities or decreasing costs) between-subjects design. Participants read
about job displacement due to automation or outsourcing. In the increasing capabilities
condition, the decision (to either automate or outsource) was described as having been motivated
by a desire to increase capabilities; in the decreasing cost condition it was described as motivated
by a desire to lower cost. The four scenarios are provided in full in the Supplemental Material
(see Table 1). Of note, in the outsourcing condition the work was specifically described as
FAIRNESS OF AUTOMATION AND OUTSOURCING 40

outsourced to a different company in the same state; this was done to make sure any effects
would not be due to negative assessments of globalization.
Dependent variable. I used the same 3-item measures of fairness perceptions that I used
in Studies 2-3 (α = .91). In addition, I used the same 1-item measure of willingness to invest in
the company that I used in Study 2b.
Manipulation check. I measured the extent to which respondents perceived the layoff
decision as having been motivated by the desire to increase capabilities with the same scale used
in Study 3, but with the first item changed to “ALY was motivated by a desire to be good at what
they do” (α = .96). I assessed the degree to which respondents perceived the layoff decision as
having been motivated by the desire to decrease costs with the same scale used in Study 3 (α
= .90).

Other variables. I included the same scales that I used in Study 3 to measure benefit to
company over customer (α = .92), benefit to company over society (α = .92), devaluation of
worker (α = .91), disenfranchisement of worker (α = .92), inevitability of the decision (α
= .89), and the quality of the goods (α = .81), in order to explore whether the effect of layoff
condition on these variables found in Study 3 would replicate. Because these variables are not
the focus of this study, I only present the means and standard deviations by condition (and
associated statistics) in the Supplemental Material (see Table 1) and do not elaborate on these
results further.

Results
Manipulation check. I conducted a 2 (layoff reason: automation or outsourcing) X 2
(motivation: increasing capabilities or decreasing costs) ANOVA on each of the 2 manipulation
check measures.
FAIRNESS OF AUTOMATION AND OUTSOURCING 41

Increasing capabilities. Results of the 2 x 2 ANOVA on increasing capabilities indicated
that there was a main effect of motivation: the degree to which the decision seemed motivated by
the desire to increase capabilities was higher in the increasing capabilities condition (M = 5.09,
SD = 1.55) than the decreasing cost condition (M = 3.69, SD = 1.67), F(1, 404) = 74.31, p < .001,
η2 = .16, in line with the intended manipulation. In addition, there was a significant main effect
of layoff reason such that the extent to which the decision seemed motivated by the desire to
increase capabilities was higher in the automation condition (M = 4.73, SD = 1.61) than the
outsourcing condition (M = 4.04, SD = 1.83), F(1, 404) = 13.74, p < .001, η2 = .03. The
interaction between motivation and layoff reason was marginally significant, F(1, 404) = 3.42, p
= .065, η2 = .01. This pattern of results suggests that, while I successfully manipulated
perceptions of the organization’s motivation to increase their capabilities, I did not do so in a
way that fully equalized automation and outsourcing on this measure: although the increasing
capabilities (versus decreasing costs) framing produced higher perceptions of the motivation to
increase capabilities both for automation (increasing capabilities: M = 5.22, 95%CI = [4.92,
5.52]; decreasing costs: M = 4.16, 95%CI = [3.84, 4.48]), mean difference = 1.07, p < .001,
95%CI = [.63, 1.50], and outsourcing (increasing capabilities: M = 4.93, 95%CI = [4.61, 5.25];
decreasing costs: M = 3.28, 95%CI = [2.99, 3.58]), mean difference = 1.65, p < .001, 95%CI =
[1.21, 2.09], the difference was larger for the automation condition. This finding raises the
possibility that there may be benefits of automation (e.g., consistency, reliability, predictability)
that may seem inherently more conducive to increasing capabilities than outsourcing. The means
by condition are represented in graphical form in Figure 4.
Decreasing costs. The 2 x 2 ANOVA on decreasing costs also showed a main effect of
motivation: in line with the intended manipulation, the degree to which the decision seemed
FAIRNESS OF AUTOMATION AND OUTSOURCING 42

motivated by the desire to decrease costs was higher when the decision was framed as having
been motivated by decreasing costs (M = 6.20, SD = 1.01) than when it had been described as
having been motivated by increasing capabilities (M = 5.88, SD = 1.25), F(1, 404) = 8.00, p
= .005, η2 = .02. Here I did not find a main effect of layoff reason (automation: M = 6.03, SD =
1.09; outsourcing: M = 6.04, SD = 1.20), F(1, 404) = .03, p = .858, η2 < .001, suggesting that
respondents did not perceive the organization to have been differentially motivated by decreasing
costs across layoff reason conditions. In addition, the interaction between motivation and layoff
reason was not statistically significant, F(1, 404) = .24, p = .627, η2 = .001.
Fairness Perceptions. I next conducted a similar 2 X 2 ANOVA on fairness perceptions,
the dependent variable. Consistent with my expectations, I found a main effect of motivation:
respondents judged the decision to be fairer when it was framed as having been motivated by
increasing capabilities (M = 4.20, SD = 1.65) then when it was described as having been
motivated by decreasing costs (M = 3.77, SD = 1.65), F(1, 404) = 6.29, p = .013, η2 = .02. There
was no main effect of layoff reason on fairness (automation: M = 4.09, SD = 1.65; outsourcing:
M = 3.88, SD = 1.67), F(1, 404) = 1.25, p = .265, η2 = .003. The interaction effect between
motivation and layoff reason was marginally significant, F(1, 404) = 3.74, p = .054, η2 = .01.
The means by condition are represented in graphical form in Figure 4.

We followed up this marginal interaction effect with simple effects analyses to better
understand the data patterns. Within the outsourcing condition, respondents rated the decision as
more fair when it was framed as being about increasing capabilities (M = 4.27, 95%CI = [3.94,
4.60]) than when it was framed as being about decreasing costs (M = 3.55, 95%CI = [3.24,
3.85]), mean difference = .72, p = .002, 95%CI = [.27, 1.18]. There was no such difference for
the automation condition (increasing capabilities: M = 4.14, 95%CI = [3.83, 4.44]; decreasing
FAIRNESS OF AUTOMATION AND OUTSOURCING 43

costs: M = 4.04, 95%CI = [3.71, 4.38]), mean difference = .094, p = .684, 95%CI = [-.36, .55].

At a broad level, this pattern is consistent with the effects of the manipulation on perceptions of
motivation to increase capabilities: the manipulation was more effective for the outsourcing
condition than the automation condition, and the effects on fairness perception mirror this
pattern.
Willingness to invest. I also conducted a 2 x 2 ANOVA on willingness to invest. As with
fairness, there was a significant effect of motivation on willingness to invest, with ratings higher
in the increasing capabilities condition (M = 4.18, SD = 1.96) than in the decreasing costs
condition (M = 3.44, SD = 1.80), F(1, 404) = 14.61, p < .001, η2 = .04. There was also a main
effect of layoff reason, which indicated that respondents were more willing to invest in the
company if the layoffs were due to automation (M = 4.06, SD = 1.94) than if they were a
consequence of outsourcing (M = 3.56, SD = 1.88), F(1, 404) = 5.72, p = .017, η2 = .01. The
interaction effect between motivation and layoff reason was not statistically significant, F(1,
404) = 2.36, p = .125, η2 = .01, although the basic pattern is consistent with that for fairness
judgments (and the increasing capabilities motivation measure). See Figure 4 for the graph of the
means by condition.

a)
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1
Decrease Costs
Increase Capabilities
Outsourcing Automation
Motivation to Increase
Capabiltiies
FAIRNESS OF AUTOMATION AND OUTSOURCING 44

b)

c)
Figure 4.
The means by condition with error bars representing 95% confidence intervals for a) motivation
to increase capabilities, b) fairness, and c) willingness to invest from Study 4.

General Discussion

Six studies suggest that automation- and outsourcing-driven layoffs, by virtue of the
different dimensions of efficiency they highlight (i.e. productivity and costs, respectively), exert
disparate effects on fairness perceptions. In Studies 1a and 1b, respondents perceived layoffs
stemming from the decision to automate to be fairer than those arising from the decision to
outsource. Additionally, in Study 1b, fairness perceptions of automation-induced layoffs were
not significantly different from fairness perceptions of merger-induced layoffs, in which the
company did not intentionally set out to alter employment arrangements but displaced workers
Willingness to Invest Fairness
5
4.5

4

3.5

3
Decrease Costs
2.5
Increase Capabilities
2

1.5

1

Outsourcing Automation
5

4.5

4

3.5

Decrease Costs
3

2.5
2
Increase
Capabilities
1.5

1

Outsourcing Automation

FAIRNESS OF AUTOMATION AND OUTSOURCING 45

due to the circumstances. In Study 2a and 2b, I removed any mention of automation and
outsourcing per se and instead presented scenarios in which workers were laid off and more
productive workers were hired (analogous to automation) or cheaper workers were hired
(analogous to outsourcing). Findings indicated that layoffs deriving from the decision to hire
more productive workers were deemed more legitimate than layoffs caused by the decision to
hire cheaper workers, and this was true despite the efficiency level (cost/unit) being the same
across condition. Study 3 supported the role of several factors, such as progress, devaluation and
disenfranchisement of workers, and motivation to increase capabilities, in driving this effect. In
Study 4, I focused more closely on one of these factors—the perceived motivation behind the
layoffs—and found that when the layoffs were caused by the decision to outsource, framing the
decision as having been motivated by increasing capabilities provided a buffer and prevented a
negative effect on judgments of fairness. I was less successful at changing the way automation
was perceived than changing the way outsourcing was perceived, most likely because people
already perceive significant potential for improving capabilities with automation. Taken together,
the results illustrate a decided magnanimity toward an organization that decides to lay off
workers in order to improve, grow, and pursue excellence (increase its capabilities).
Contributions

These findings contribute to the literature on layoffs and fairness in two ways. First,
expanding on prior studies that have suggested that the degree to which individuals are given
clear and adequate reasons for organizational decisions shape perceptions of fairness (Bies,
1987; Bies & Moag, 1986; Bies & Shapiro, 1987), my findings shed light on the specific content
of accounts for employee downsizing decisions that individuals deem adequate and fair. Prior
research broadly distinguishes between layoffs that are implemented as a competitive strategy in
FAIRNESS OF AUTOMATION AND OUTSOURCING 46

the absence of threats to organizational viability (i.e. proactive downsizing; Freeman &
Cameron, 1993), and those that are executed in order to circumvent organizational decline (i.e.
reactive downsizing; Freeman & Cameron, 1993); as a whole, this body of work suggests that
proactive downsizing decisions are likely to be perceived as less fair than reactive downsizing
decisions because organizations are held less accountable for circumstances that are beyond their
control (Bies, 1987; Bies & Moag; 1986; Brockner, 1992; Charness & Levine, 2000; Kahneman,
et al., 1986b; Shore & Tetrick, 1994). For example, Kahneman et al. (1986b) found that whereas
77% of respondents believed that employee wage cuts were unfair when the company was
making money, 68% of respondents believed that the same wage cuts were fair when the
company was losing money. In a similar vein, Shore and Tetrick (1994) describe how threats to
organizational survival can mitigate the effect of psychological contract breaches (e.g., layoffs)
on perceptions of injustice:
People make attributions of responsibility when they make judgments of fairness. If an
organization appears to break the psychological contract voluntarily, judgments of injustice may
be greater than when the organization is not held fully responsible. For example, a psychological
contract representing organizational obligations of job security in exchange for employee
obligations to be loyal, which is broken (e.g. when an employee is fired or part of a layoff) may
be viewed as only a partially broken contract if an economic downturn caused the organization to
be unable to fulfill the obligation (p. 104).

Although prior discussions have compared across the broad categories of reactive and
proactive downsizing decisions, there are very few studies that have investigated the types of
accounts given for proactive downsizing decisions and the ramifications for justice perceptions.
One such study by Charness and Levine (2000) examined the fairness perceptions of layoffs that
were manipulated on a variety of dimensions including cause (e.g., declining demand, increased
productivity due to new technology, increased productivity due to employees’ suggestions, end
of project) and found that in a sample of Canadian respondents, fairness perceptions were among
FAIRNESS OF AUTOMATION AND OUTSOURCING 47

the highest when the layoffs were attributed to declining demand. Contrary to their expectations,
they also found that in a sample of Silicon Valley residents, the fairness perceptions were highest
when the layoffs were ascribed to higher productivity arising from new technology, suggesting
that proactive downsizing decisions may not always seem unfair. I sought to systematically
assess the impact of two common causes of proactive downsizing decisions—automation and
outsourcing—and their impact on justice perceptions. My findings revealed that voluntary layoff
decisions by profitable organizations may still be construed as fair if the decision seems to have
been motivated by automation or a desire to increase productivity and capabilities. The mean
fairness ratings in the automation or productive replacement condition were above the midpoint
of the scale (4) for all studies and significantly so for Studies 1a-3, suggesting that proactive
layoffs motivated by enhancing productivity were seen to be legitimate and just. However,
because I did not explicitly state in my scenarios that the organization was profitable (although
there was no apparent indication of threat to organizational viability), I recognize that with the
addition of that detail the means on average could have been lower than the midpoint. At
minimum, my findings show that a productivity-related account of layoffs can attenuate the
negative effects of proactive layoffs on fairness judgments. Given the increasing use of proactive
downsizing strategies among organizations (Cameron, Freeman, & Mishra, 1993; Cascio, 1993;
Freeman & Cameron, 1993), my studies provide important insight into the types of downsizing
accounts that can increase fairness perceptions of this strategy.
Secondly, I extend the literature on justice perceptions by showing that productivity
pursuits can be a basis of fairness judgments. Konow (2001) suggests that there are broadly 3
principles of justice on which distributive justice evaluations are based: the Accountability
Principle, Needs Principle, and Efficiency Principle. The Accountability Principle “calls for
FAIRNESS OF AUTOMATION AND OUTSOURCING 48

allocations to be in proportion to volitional contributions, e.g. a worker who is twice as
productive as another should be paid twice as much if the higher productivity is due to greater
work effort (p. 138),” and “deals with the relative size of allocations across individuals.” The
Efficiency Principle, in contrast, “captures a concern about the total size of the allocable variable
(p. 145), or maximizing the total allocation to all parties involved. Finally, the Needs Principle
dictates that “a just allocation is simply one that is sufficient to meet each individual’s basic
requirements for life (p. 139).” The scenarios used in the present study described workers who
were laid off by an organization who had decided to automate/outsource operations or hire more
productive/cheaper workers. Assuming that the good to be allocated is the revenue generated by
the organization (from which workers get paid), concerns about the livelihood of workers who
were displaced is likely to have called into operation the Needs Principle. In addition, the desire
to see pay that is commensurate with performance is likely to have elicited the Accountability
Principle, and the issue of maximizing total revenue, which is determined by the number of
widgets produced, should have activated the Efficiency Principle.
However, I found differences in fairness ratings across condition despite having held
constant the information relevant to these three principles: the number of workers laid off in
Studies 2a-3 (relevant to the Needs Principle), the compensation and efficiency level of the new
workers reflected in the pay/hour and units/hour in Studies 2a-3 (relevant to the Accountability
Principle), and the total units produced per hour by the 2 new workers in Study 2b (relevant to
the Efficiency Principle). The only difference was the productivity level of each replacement
worker, as measured by the number of widgets produced per hour. In line with Konow (2001)’s
argument that fairness judgments are contingent on the context and how relevant information is
weighted and construed, I propose that the decision to lay off workers and hire more productive
FAIRNESS OF AUTOMATION AND OUTSOURCING 49

(versus cheaper) workers were interpreted as reflecting a desire to pursue productivity and
increase capabilities, which in turn influenced fairness perceptions. This would suggest that
people not only evaluate fairness based on the actual outcomes (e.g., loss of livelihoods for
workers and revenue level for the organization) but also on the manifest intentions of various
actors, including the organization. This notion is consistent with agent-centered deontological
views of morality, which locate “the moral quality of acts in the principles or maxims on which
the agent acts and not primarily in those acts' effects on others” (Alexander & Moore, 2007). In
the present study, decisions to lay off workers that seemed to be initiated by a desire to improve
and grow from the status quo were deemed fairer than those that were not. Thus, the findings of
this study intimate an additional basis of fairness judgments: the motivations and actions taken
by an actor to further productivity. Building on prior studies that chronicle the natural pull of
productivity for individuals, I raise the possibility that individuals’ attraction to productivity and
betterment are so deeply ingrained that they grant others who display such pursuits, including
faceless entities such as organizations, much clemency in the domain of justice. Essentially, the
pursuit of productivity and betterment may be a moral mandate, not only for individuals but also
for organizations.
Practical Implications

There are a number of practical implications that stem from the findings of the present
investigation, which demonstrate more lenient fairness judgments for an organization that is
perceived to be motivated by increasing productivity (versus reducing costs). One is that people
could be galvanized for or against decisions that harm current workers, depending on the
motivation that is emphasized. For example, reframing decisions around capability-enhancement
and improvement might allow organizations to better prevent negative fallout when their
FAIRNESS OF AUTOMATION AND OUTSOURCING 50

decisions impact current workers. On the opposing side, employee advocates may be able to
reframe automation as primarily a cost-saving technique that has limited effects on increasing
productivity (i.e. total value of output produced / number of hours worked), as suggested by
numbers on declining productivity across nations (e.g., Manyika, Remes, Mischke, & Krishnan,
2017), in order to decrease the legitimacy of automation-caused job displacement. Those with
such intentions may also benefit from highlighting examples in which over-reliance on
automation and renunciation of human labor curbed productivity and the ability of an
organization to meet its goals, as was the case with the electric car manufacturer Tesla (e.g.,
Hull, 2018). The cost-saving focus alone in Study 4 did not create movement on perceptions of
fairness for the automation-induced layoffs, but it seems plausible that limited-productivity
framings may be effective. Previous research has distinguished between the differential effects of
attributing layoffs to declining demand versus the desire to increase efficiency, operationalized
as reducing costs (e.g., Palmon, Sun, & Tang, 1997; Schulz & Johann, 2018), but none have
examined attributions to improving the positive element of efficiency (i.e. productivity). The
results of the studies suggest that the two can have drastically different effects on judgments of
fairness.
Conversely, individuals may benefit from taking note of the leniency that they may afford
organizations seeking to increase productivity and increase capabilities. Numerous start-up
companies (e.g., Uber, Theranos) received vast amounts of funding from investors with promises
of disrupting industries with technologies that would enhance both organizational and individual
capabilities. Arguably, the potential of the technologies to allow individuals to do more and do
so in a shorter period of time contributed to the favorable treatment of these companies and
permitted unethical behaviors to go unnoticed or even disregarded. For instance, Uber, a ride-
FAIRNESS OF AUTOMATION AND OUTSOURCING 51

hailing services company, evaded regulatory authorities across the globe using fake versions of
their mobile application, promoted an unrestrained workplace culture that condoned harassment
and discrimination, and exhibited a host of other questionable behaviors, in the name of making
quicker progress toward their goals and disrupting the existing ride transportation industry
(Isaac, 2017b; 2017c). Their ethical infractions, however, did not come to light and persisted for
several years before the founder and C.E.O. was ousted in 2017 (Isaac, 2017a). Similarly, the
previous C.E.O. of Theranos, a consumer healthcare technology company, claimed to have
created a simple blood test that would revolutionize healthcare and succeeded in deceiving and
defrauding countless hopeful investors, patients, and doctors (Abelson, 2018). It is evident that
the lure of productivity and progress is real, often causing people to overlook unethical shortcuts
or make optimistic bets on organization that aim to contribute to this lofty goal; simply being
aware of this proclivity to prioritize productivity and progress may temper its adverse effects.
Limitations and Future Directions

In this section I address possible limitations of the present investigation. First, my
findings were based on a survey of Mturk workers, who I believed to be a reasonable sample
given my goal of understanding the public’s fairness perceptions of automation-induced layoffs.
Although Mturk samples cannot be taken as being representative of the general population in the
United States, studies have shown that Mturk samples are at least as, if not more, diverse and
reliable as undergraduate subject pools and other internet samples (Buhrmester, Kwang, &
Gosling, 2011; Paolacci & Chandler, 2014; Paolacci, Chandler, & Ipeirotis, 2010). This prompts
the question of whether I would have found the same results in different samples. For example, if
the results obtained were a product of participants’ difficulty in making the (equivalent)
efficiency calculations across conditions (e.g., Crump, McDonnell, & Gureckis, 2013), the
FAIRNESS OF AUTOMATION AND OUTSOURCING 52

effects would be mitigated in a sample of participants who have higher levels of educational
attainment or advanced degrees.
Also, given that perceptions are shaped by cultural context and precedent (Konow, 2001;
Young, 1995) and people make self-serving judgments of fairness (Konow, 2005), it is possible
that individuals who have something to gain from outsourcing (e.g., those who reside in
countries that are often blamed for taking American jobs) would have deemed outsourcing-
driven layoffs to be fairer than those in the current sample. Along similar lines, cultural values
and precedents regarding the employer-employee relationship that are different from those in the
United States may produce different results. For example, Nègre, Verdier, Cho, and Patten
(2017) distinguished between code-law countries (e.g., the United States) that are characterized
by a stakeholder governance model that allows for greater flexibility in the hiring and firing of
employees, and common-law countries (e.g., France) that are characterized by powerful labor
unions which make firing employees much more difficult. They found that the typically positive
market reactions to downsizing implemented proactively (i.e. when the firm is profitable) as
opposed to reactively (i.e. when profitability is under threat) in the United States was reversed
for France, suggesting that individuals residing in such countries may react more negatively to
all proactive layoffs, regardless of whether they were caused by automation or outsourcing.
Another relevant example is the study by Charness and Levine (2000) which found that
Canadian respondents rated layoffs resulting from new technology lower on fairness than those
caused by declining product demand, whereas the opposite was true for respondents in Silicon
Valley, a region known for its pursuit of innovation and productivity. Further research on
different samples is necessary to determine the boundary and moderating conditions of the found
effects.
FAIRNESS OF AUTOMATION AND OUTSOURCING 53

Moreover, it is possible that the scenarios in the two conditions for each study were
interpreted differently. Because I did not specify that the company was altering employment
arrangements despite being profitable, participants may have interpreted outsourcing-induced
layoffs or the hiring of cheaper workers as having been motivated by a concern with remaining
profitable. However, if this were the case, the fairness perceptions would have been higher in
those conditions, which would have mitigated the effects. I suggest that my study designs
allowed for a conservative test of my hypotheses. In addition, existing research has demonstrated
that outsourcing, which does not necessarily involve the relocation of work overseas, is often
confused with offshore outsourcing, or “the procurement of products or services from an outside
supplier located in a foreign country (p. 169; Robertson, et al., 2010)” (Global Sourcing
Association, 2015). Although the term “outsourcing” may have evoked negative reactions due to
a dislike of foreign “others” (e.g., Mansfield & Mutz, 2013), the findings of studies 2a-3, in
which I had removed any mention of the word “outsourcing” itself, provided stronger support for
the effect being due to the motivations attributed to the organization. Further lending credence to
this point, Study 4, which specified that the outsourcing was to another company in the same
state, also revealed higher fairness ratings in the automation than outsourcing condition
(although the effect did not reach significance) and a significant main effect of motivation
framing with higher fairness ratings in the increasing capabilities than decreasing cost condition.
Finally, the type and characteristics of automation technology involved may also
moderate the effects found in the present study. I focused specifically on automation in the
context of manufacturing, but with the development of artificial intelligence, widespread
automation of services may also become a reality in the near future (e.g., Davis, 2016; Knight,
2017; Mathiassen, Fjellin, Glette, Hol, & Elle, 2016). For services and work that involve human
FAIRNESS OF AUTOMATION AND OUTSOURCING 54

interaction (e.g., caretakers), which people may believe to be the sovereign dominion of human
beings, automation and the resulting job displacement may elicit more negative reactions. This
may be potentially be even more so for machines that resemble humans, which have been found
to create discomfort and a sense of eeriness (MacDorman, 2006; Mori, 1970; Mori, MacDorman,
& Kageki, 2012). Thus, a closer examination of the features and types of machines that replace
human workers and the resulting fairness perceptions may provide an interesting direction for
future research.
Conclusion

My findings speak directly to a timely and important issue: understanding people’s
responses to automation. While automation is evoking concern and worry, its association with
increasing capabilities appears to underlie its perceived legitimacy. This may be why people’s
growing concern is not triggering related outrage, despite automation’s potential to dramatically
impact society in a way that may harm many people. Developing a nuanced understanding of
perceptions of automation is critical for those hoping to shape public consciousness around this
issue. It may be plausible, for example, to directly question the legitimacy of productivity
pursuits and call attention to a potential “dark side” of capability-enhancing efforts.
Alternatively, it may be more effective to acknowledge that accelerating productivity and
progress are prized goals when engaging in any debate around automation, and make salient the
role this has in making technological advances feel legitimate. Given the major potential
automation has to alter organizational and societal landscapes (Brynjolfsson & McAfee, 2014), I
hope the studies reported here lead to further research on the way automation is understood and
perceived.
FAIRNESS OF AUTOMATION AND OUTSOURCING 55

References

Abelson, R. (2018, June 15). Theranos Founder Elizabeth Holmes Indicted on Fraud Charges.
The New York Times. Retrieved from
https://www.nytimes.com/2018/06/15/health/theranos-elizabeth-holmes-fraud.html

Acemoglu, D., Autor, D., Dorn, D., Hanson, G. H., & Price, B. (2016). Import competition and
the great US employment sag of the 2000s. Journal of Labor Economics, 34(S1),
S141-S198. https://doi.org/10.1086/682384

Acemoglu, D., & Restrepo, P. (2017). Robots and jobs: Evidence from US labor markets.
NBER Working Paper, No. w23285. Retrieved from
https://ssrn.com/abstract=2941263

Alexander, L., & Moore, M. (2007). Deontological ethics. In Stanford Encyclopedia of
Philosophy. Retrieved from https://plato.stanford.edu/entries/ethics-deontological/

Arntz, M., Gregory, T., & Zierahn, U. (2016). The risk of automation for jobs in OECD
countries: A comparative analysis. OECD Social, Employment, and Migration
Working Papers, 189, 1-35. https://doi.org/10.1787/5jlz9h56dvq7-en

Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure, process,
and content. Psychological Bulletin, 120(3), 338-375. doi:10.1037/0033-
2909.120.3.338

Babcock, L., Loewenstein, G., & Issacharoff, S., Camerer, C. (1995). Biased Judgments of
Fairness in Bargaining. American Economic Review, 85(5), 1337-1343. Retrieved
from http://www.jstor.org/stable/2950993

Babcock, L., Wang, X., & Loewenstein, G. (1996). Choosing the wrong pond: Social
comparisons in negotiations that reflect a self-serving bias. The Quarterly Journal of
Economics, 111(1), 1-19. https://doi.org/10.2307/2946655

Barnett, M. L., Jermier, J. M., & Lafferty, B. A. (2006). Corporate reputation: The definitional
landscape. Corporate Reputation Review, 9(1), 26-38.
https://doi.org/10.1057/palgrave.crr.1550012

Bies, R. J. (1987). The predicament of injustice: The management of moral outrage. In L. L.
Cummings & B. M. Staw (Eds.), Research in Organizational Behavior (Vol. 9, pp.
289-319). Greenwich, CT: JAI Press.

Bies, R. J. and Moag, J. F. (1986) Interactional Justice: Communication Criteria of Fairness.
In: Lewicki, R. J., Sheppard, B. H. and Bazerman, M. H., Eds., Research on
Negotiations in Organizations (Vol. 1, pp. 43-55). Greenwich, CT: JAI Press.
FAIRNESS OF AUTOMATION AND OUTSOURCING 56

Bies, R. J. & Shapiro, D. L. (1987). Interactional fairness judgments: The influence of causal
accounts. Social Justice Research 1(2): 199-218. https://doi.org/10.1007/BF01048016

Boston Consulting Group (2017). The state of social mobility in the UK. Retrieved from
https://www.suttontrust.com/wp-content/uploads/2017/07/BCGSocial-Mobility-
report-full-version_WEB_FINAL.pdf

Bowles, N. (2018, April 12). After Cambridge Analytica, Privacy Experts Get to Say ‘I Told
You So’. The New York Times. Retrieved from
https://www.nytimes.com/2018/04/12/technology/privacy-researchers-facebook.html

Brockner, J. (1992). Managing the effects of layoffs on survivors. California Management
Review, 34(2), 9-28. https://doi.org/10.2307/41166691

Brockner, J. (1994). Perceived fairness and survivors' reactions to layoffs, or how downsizing
organizations can do well by doing good. Social Justice Research, 7(4), 345-363.
https://doi.org/10.1007/BF02334861

Brockner, J., & Greenberg, J. (1990). The impact of layoffs on survivors: An organizational
justice perspective. In J. S. Carroll (Ed.), Applied Social Psychology and
Organizational Settings (pp. 45-75). Hillsdale, N J: Erlbaum.

Brockner, J., DeWitt, R. L., Grover, S., & Reed, T. (1990). When it is especially important to
explain why: Factors affecting the relationship between managers' explanations of a
layoff and survivors' reactions to the layoff. Journal of Experimental Social
Psychology, 26(5), 389-407. https://doi.org/10.1016/0022-1031(90)90065-T

Brockner, J., Grover, S., Reed, T., DeWitt, R., & O'Malley, M. (1987). Survivors' reactions to
layoffs: We get by with a little help for our friends. Administrative Science Quarterly,
32(4), 526-541. doi: 10.2307/2392882

Brockner, J., Konovsky, M., Cooper-Schneider, R., Folger, R., Martin, C., & Bies, R. J.
(1994). Interactive effects of procedural justice and outcome negativity on victims and
survivors of job loss. Academy of Management Journal, 37(2), 397-409.
https://doi.org/10.5465/256835

Brockner, J., Wiesenfeld, B. M., & Martin, C. L. (1995). Decision frame, procedural justice,
and survivors′ reactions to job layoffs. Organizational Behavior and Human Decision
Processes, 63(1), 59-68. https://doi.org/10.1006/obhd.1995.1061

Brookman, J. T., Chang, S., & Rennie, C. G. (2007). CEO Cash and Stock-Based
Compensation Changes, Layoff Decisions, and Shareholder Value. Financial Review,
42(1), 99-119. https://doi.org/10.1111/j.1540-6288.2007.00163.x
FAIRNESS OF AUTOMATION AND OUTSOURCING 57

Brunstein, J. C. (1993). Personal goals and subjective well-being: A longitudinal study.
Journal of Personality and Social Psychology, 65(5), 1061-1070.
http://dx.doi.org/10.1037/0022-3514.65.5.1061

Brynjolfsson, E., & McAfee, A. (2014). The second machine age: Work, progress, and
prosperity in a time of brilliant technologies. WW Norton & Company.

Budros, A. (1999). A conceptual framework for analyzing why organizations downsize.
Organization Science, 10(1), 69-82. https://doi.org/10.1287/orsc.10.1.69

Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk: A new
source of inexpensive, yet high-quality, data? Perspectives on Psychological Science,
6(1), 3-5. https://doi.org/10.1177/1745691610393980

Burke, R. J. (2003). Survivors and victims of hospital restructuring and downsizing: who are
the real victims? International Journal of Nursing Studies, 40(8), 903-909.
https://doi.org/10.1016/S0020-7489(03)00039-7

Camerer, C. F. (2003). Behavioral game theory: Experiments in strategic interaction. New
York, NY, US: Russell Sage Foundation.

Cameron, K.S., Freeman, S.J., and Mishra, A.K. (1993). Downsizing and redesigning
organizations. In Huber, George P. and Glick, William H. (Eds.), Organizational
Change and Redesign. New York: Oxford University Press.

Cappelli, P. (2000). Examining the Incidence of Downsizing and Its Effect on Establishment
Performance. In Neumark, D. (ed.), On the Job: Is Long-term Employment A Thing of
the Past? New York, Russell Sage. doi: 10.3386/w7742

Cascio, W. F. (1993). Downsizing: What do we know? What have we learned? The Academy
of Management Executive, 7(1), 95-104.
https://doi.org/10.5465/ame.1993.9409142062

Cascio, W. F., & Young, C. E. (2003). Financial consequences of employment-change
decisions in major U. S. corporations, 1982–2000. In K. P. De Meuse & M. L. Marks
(Eds.), Resizing the organization: Managing layoffs, divestitures, and closings (pp.
131–156). San Francisco, CA: Jossey-Bass.

Chalos, P., & Chen, C. J. (2002). Employee downsizing strategies: Market reaction and post
announcement financial performance. Journal of Business Finance & Accounting,
29(5-6), 847-870. https://doi.org/10.1111/1468-5957.00453

Charness, G., & Levine, D. I. (2000). When are layoffs acceptable? Evidence from a quasi-
experiment. ILR Review, 53(3), 381-400.
https://doi.org/10.1177/001979390005300302
FAIRNESS OF AUTOMATION AND OUTSOURCING 58

Cohen-Charash, Y., & Spector, P. E. (2001). The role of justice in organizations: A meta-
analysis. Organizational Behavior and Human Decision Processes, 86(2), 278-321.
https://doi.org/10.1006/obhd.2001.2958

Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O., & Ng, K. Y. (2001). Justice at the
millennium: a meta-analytic review of 25 years of organizational justice research.
Journal of Applied Psychology, 86(3), 425-445. http://dx.doi.org/10.1037/0021-
9010.86.3.425

Crump, M. J., McDonnell, J. V., & Gureckis, T. M. (2013). Evaluating Amazon's Mechanical
Turk as a tool for experimental behavioral research. PloS one, 8(3), e57410.
https://doi.org/10.1371/journal.pone.0057410

Datta, D. K., Guthrie, J. P., Basuil, D., & Pandey, A. (2010). Causes and effects of employee
downsizing: A review and synthesis. Journal of Management, 36(1), 281-348.
https://doi.org/10.1177/0149206309346735

Davis, L. K. (2016, March 9). Hilton and IBM pilot “Connie,” The world’s first Watson-
enabled hotel concierge robot. Retrieved from
https://www.ibm.com/blogs/watson/2016/03/watson-connie/

Deci, E. L., & Ryan, R. M. (2014). The importance of universal psychological needs for
understanding motivation in the workplace. In M. Gagne, The Oxford handbook of
work engagement, motivation, and self-determination theory (pp. 13-32). New York,
NY: Oxford University Press.

Devine, K., Reay, T., Stainton, L., & Collins-Nakai, R. (2003). Downsizing outcomes: Better a
victim than a survivor? Human Resource Management, 42(2), 109-124.
https://doi.org/10.1002/hrm.10071

Emmons, R. A. (1996). Striving and feeling: Personal goals and subjective well-being. In P.
M. Gollwitzer & J. A. Bargh (Eds.), The psychology of action: Linking cognition and
motivation to behavior (pp. 313-337). New York, NY, US: Guilford Press.

Fombrun, C. J. (1996). Reputation: Realizing Value from the Corporate Image. Boston:
Harvard Business School Press.

Fombrun, C. J. (2001). Corporate Reputation as Economic Assets. In M. A. Hitt, R.E.
Freeman, & J. S. Harrison (Eds.), The Handbook of Strategic Management (pp. 289-
312). Oxford, UK: Blackwell Publishers.
https://doi.org/10.1111/b.9780631218616.2006.00011.x

Fombrun, C., & Shanley, M. (1990). What's in a name? Reputation building and corporate
strategy. Academy of Management Journal, 33(2), 233-258.
https://doi.org/10.5465/256324
FAIRNESS OF AUTOMATION AND OUTSOURCING 59

Freeman, R. E. (2010). Strategic management: A stakeholder approach. Boston: Pitman.

Freeman, S. J., & Cameron, K. S. (1993). Organizational Downsizing: A Convergence and
Reorientation Framework. Organization Science, 4(1), 10-29. doi:10.1287/orsc.4.1.10

Frey, C. B., & Osborne, M. A. (2017). The future of employment: how susceptible are jobs to
computerisation? Technological Forecasting and Social Change, 114, 254-280.
https://doi.org/10.1016/j.techfore.2016.08.019

Furnham, A. (1984). The Protestant work ethic: A review of the psychological literature.
European Journal of Social Psychology, 14(1), 87-104.
https://doi.org/10.1002/ejsp.2420140108

Gallup (2018). Optimism and Anxiety: Views on the impact of artificial intelligence and higher
education’s response. Retrieved from
https://www.northeastern.edu/gallup/pdf/OptimismAnxietyNortheasternGallup.pdf.

Gibson, D., Gonzales, J. L., & Castanon, J. (2006). The importance of reputation and the role
of public relations. Public Relations Quarterly, 51(3), 15-18.

Global Sourcing Association (2015). The public perception of outsourcing. Retrieved from
http://www.gsa-uk.com/files/129.pdf

Hackman, J. R., & Oldham, G. R. (1980). Work redesign (Vol. 72). Reading, MA: Addison-
Wesley.

Harris, L. T., & Fiske, S. T. (2006). Dehumanizing the Lowest of the Low: Neuroimaging
Responses to Extreme Out-Groups. Psychological Science, 17(10), 847-853.
https://doi.org/10.1111/j.1467-9280.2006.01793.x

Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable
mediation, moderation, and conditional process modeling [White paper]. Retrieved
from
https://pdfs.semanticscholar.org/e9bb/7b23993113a73ee1ff6cde5ff9a4164f946e.pdf

Hsee, C. K., Yang, A. X., & Wang, L. (2010). Idleness aversion and the need for justifiable
busyness. Psychological science, 21(7), 926-930.
https://doi.org/10.1177/0956797610374738

Hsee, C. K., Zhang, J., Cai, C. F., & Zhang, S. (2013). Overearning. Psychological science,
24(6), 852-859. https://doi.org/10.1177/0956797612464785

Hull, D. (2018, April 13). Musk Says Excessive Automation Was ‘My Mistake’. Retrieved from
https://www.bloomberg.com/news/articles/2018-04-13/musk-tips-his-tesla-cap-to-
humans-after-robots-undercut-model-3.
FAIRNESS OF AUTOMATION AND OUTSOURCING 60

Isaac, M. (2017a, June 21). Uber Founder Travis Kalanick Resigns as C.E.O. The New York
Times. Retrieved from https://www.nytimes.com/2017/06/21/technology/uber-ceo-
travis-kalanick.html

Isaac, M. (2017b, February 22). Inside Uber’s Aggressive, Unrestrained Workplace Culture.
The New York Times. Retrieved from
https://www.nytimes.com/2017/02/22/technology/uber-workplace-culture.html

Isaac, M. (2017c, March 3). How Uber Deceives the Authorities Worldwide. The New York
Times. Retrieved from https://www.nytimes.com/2017/03/03/technology/uber-
greyball-program-evade-authorities.html

Jin, R. L., Shah, C. P., & Svoboda, T. J. (1995). The impact of unemployment on health: a
review of the evidence. CMAJ: Canadian Medical Association Journal, 153(5), 529.

Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1986a). Fairness and the assumptions of
economics. Journal of Business, 59(4), S285-S300. Retrieved from
https://www.jstor.org/stable/2352761

Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1986b). Fairness as a constraint on profit
seeking: Entitlements in the market. The American Economic Review, 76(4), 728-741.
Retrieved from https://www.jstor.org/stable/1806070

Kaplan, J. (2015). Humans need not apply: A guide to wealth and work in the age of artificial
intelligence: Yale University Press.

Kay, A. C., & Friesen, J. (2011). On social stability and social change: Understanding when
system justification does and does not occur. Current Directions in Psychological
Science, 20(6), 360-364. https://doi.org/10.1177/0963721411422059

Kennedy, A. A. (2008). The end of shareholder value: Corporations at the crossroads. New
York: Perseus Publishing.

Keynes, J. M. (1930). Economic possibilities for our grandchildren. In J.M. Keynes, (Ed.),
Essays in persuasion (pp. 358-373.). New York: W.W.Norton & Co.

Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind.
Science, 330(6006), 932. doi: 10.1126/science.1192439

Kim, J., & Wakslak, C. J. (2017). Study examining why people pursue efficiency.
Unpublished raw data.

Knight, W. (2017, November 6). Is Technology About to Decimate White-Collar Work? MIT
Technology Review. Retrieved from https://www.technologyreview.com/s/609337/is-
technology-about-to-decimate-white-collar-work/
FAIRNESS OF AUTOMATION AND OUTSOURCING 61

Konovsky, M. A., & Brockner, J. (1993). Managing victim and survivor layoff reactions: A
procedural justice perspective. In R. Cropanzano (Ed.), Series in applied psychology.
Justice in the workplace: Approaching fairness in human resource management (pp.
133-153). Hillsdale, NJ, US: Lawrence Erlbaum Associates, Inc.

Konovsky, M. A., & Folger, R. (1991). The effects of procedures, social accounts, and benefits
level on victims' layoff reactions. Journal of Applied Social Psychology, 21(8), 630-
650. https://doi.org/10.1111/j.1559-1816.1991.tb00540.x

Konow, J. (2001). Fair and square: the four sides of distributive justice. Journal of Economic
Behavior & Organization, 46(2), 137-164. https://doi.org/10.1016/S0167-
2681(01)00194-9

Konow, J. (2003). Which is the fairest one of all? A positive analysis of justice theories.
Journal of Economic Literature, 41(4), 1188-1239. doi:
10.1257/002205103771800013

Konow, J. (2005). Blind spots: The effects of information and stakes on fairness bias and
dispersion. Social Justice Research, 18(4), 349-390. https://doi.org/10.1007/s11211-
005-8566-6

Labor Productivity and Costs (n.d.). Retrieved from https://www.bls.gov/lpc/

Laurin, K., Shepherd, S., & Kay, A. C. (2010). Restricted emigration, system inescapability,
and defense of the status quo: System-justifying consequences of restricted exit
opportunities. Psychological Science, 21(8), 1075-1082.
https://doi.org/10.1177/0956797610375448

Lee, D. (2016, June 30) Behind shrinking middle-class jobs: A surge in outsourcing. The LA
Times. Retrieved from http://www.latimes.com/business/la-fi-outsourcing-middle-
class-jobs-20160630-snap-story.html

Leventhal, G. S. (1980). What Should Be Done with Equity Theory? In K. J. Gergen, M. S.
Greenberg, & R. H. Willis (Eds.), Social Exchange: Advances in Theory and Research
(pp. 27-55). Boston, MA: Springer US.

MacDorman, K. F. (2006, July). Subjective ratings of robot video clips for human likeness,
familiarity, and eeriness: An exploration of the uncanny valley. In Toward social
mechanisms of android science. Symposium conducted at the meeting of International
Conference of the Cognitive Science Society, Vancouver, Canada.

Mansfield, E. D., & Mutz, D. C. (2013). US versus them: Mass attitudes toward offshore
outsourcing. World Politics, 65(4), 571-608.
https://doi.org/10.1017/S0043887113000191
FAIRNESS OF AUTOMATION AND OUTSOURCING 62

Manyika, J., Lund, S., Chui, M., Bughin, J., Woetzel, J., Batra, P., . . . Sanghvi, S. (2017). Jobs
Lost, Jobs Gained: Workforce transitions in a time of automation. McKinsey Global
Institute. Retrieved from https://www.mckinsey.com/mgi/overview/2017-in-
review/automation-and-the-future-of-work/jobs-lost-jobs-gained-workforce-
transitions-in-a-time-of-automation.

Manyika, J., Remes, J., Mischke, J., & Krishnan, M. (2017). The Productivity Puzzle.
McKinsey Global Institute. Retrieved from
https://www.mckinsey.com/~/media/mckinsey/featured%20insights/employment%20a
nd%20growth/new%20insights%20into%20the%20slowdown%20in%20us%20produ
ctivity%20growth/mgi-the-productivity-puzzle-discussion-paper.ashx

Mathers, C. D., & Schofield, D. J. (1998). The health consequences of unemployment: the
evidence. The Medical Journal of Australia, 168(4), 178-182.

Mathiassen, K., Fjellin, J. E., Glette, K., Hol, P. K., & Elle, O. J. (2016). An Ultrasound
Robotic System Using the Commercial Robot UR5. Frontiers in Robotics and AI,
3(1). doi:10.3389/frobt.2016.00001

Mishra, A. K., & Spreitzer, G. M. (1998). Explaining how survivors respond to downsizing:
The roles of trust, empowerment, justice, and work redesign. Academy of
Management Review, 23(3), 567-588. https://doi.org/10.5465/amr.1998.926627

Mitchell, R. K., Agle, B. R., & Wood, D. J. (1997). Toward a theory of stakeholder
identification and salience: Defining the principle of who and what really counts.
Academy of Management Review, 22(4), 853-886.
https://doi.org/10.5465/amr.1997.9711022105

Mori, M. (1970). The uncanny valley. Energy, 7(4), 33-35 (in Japanese).

Mori, M., MacDorman, K. F., & Kageki, N. (2012). The uncanny valley [from the field]. IEEE
Robotics & Automation Magazine, 19(2), 98-100.

Multifactor Productivity (n.d.). Retrieved from https://www.bls.gov/mfp/home.htm

Nedelkoska, L., & Quintini, G. (2018). Automation, skills use and training.

Nedelkoska, L., & Quintini, G. (2018). Automation, skills use and training. OECD Social,
Employment and Migration Working Papers, No. 202.
https://doi.org/10.1787/2e2f4eea-en

Nègre, E., Verdier, M.-A., Cho, C. H., & Patten, D. M. (2017). Disclosure strategies and
investor reactions to downsizing announcements: A legitimacy perspective. Journal of
Accounting and Public Policy, 36(3), 239-257.
https://doi.org/10.1016/j.jaccpubpol.2017.03.003
FAIRNESS OF AUTOMATION AND OUTSOURCING 63

Nixon, R. D., Hitt, M. A., Lee, H. U., & Jeong, E. (2004). Market reactions to announcements
of corporate downsizing actions and implementation strategies. Strategic Management
Journal, 25(11), 1121-1129. https://doi.org/10.1002/smj.423

Noer, D. M. (2009). Healing the wounds: Overcoming the trauma of layoffs and revitalizing
downsized organizations. San Francisco: Jossey-Bass.

Palmon, O., Sun, H.-L., & Tang, A. P. (1997). Layoff announcements: Stock market impact
and financial performance. Financial Management, 54-68. doi: 10.2307/3666213

Paolacci, G., & Chandler, J. (2014). Inside the Turk: Understanding Mechanical Turk as a
participant pool. Current Directions in Psychological Science, 23(3), 184-188.
https://doi.org/10.1177/0963721414531598

Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on amazon
mechanical turk. Judgment and Decision Making, 5(5), 411–419.

Pfeifer, C. (2007). The perceived fairness of layoffs in Germany: Participation, compensation,
or avoidance? Journal of Business Ethics, 74(1), 25-36. doi:10.1007/s10551-006-
9217-z

Porath, C., MacInnis, D., & Folkes, V. S. (2011). It’s unfair: Why customers who merely
observe an uncivil employee abandon the company. Journal of Service Research,
14(3), 302-317. https://doi.org/10.1177/1094670511404393

Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and
comparing indirect effects in multiple mediator models. Behavior Research Methods,
40(3), 879-891. doi: 10.3758/BRM.40.3.879

Rabin, M. (1993). Incorporating fairness into game theory and economics. The American
Economic Review, 83(5), 1281-1302. Retrieved from
https://www.jstor.org/stable/2117561

Rao, A. R., & Monroe, K. B. (1989). The effect of price, brand name, and store name on
buyers' perceptions of product quality: An integrative review. Journal of Marketing
Research, 26(3), 351-357. doi: 10.2307/3172907

Rawls, J. (1971). A theory of justice. Cambridge, MA: Harvard University Press.

Richard, F. D., Bond Jr, C. F., & Stokes-Zoota, J. J. (2003). One Hundred Years of Social
Psychology Quantitatively Described. Review of General Psychology, 7(4), 331.

Richard, F. D., Bond, C. F., Jr., & Stokes-Zoota, J. J. (2003). One Hundred Years of Social
Psychology Quantitatively Described. Review of General Psychology, 7(4), 331-363.
http://dx.doi.org/10.1037/1089-2680.7.4.331
FAIRNESS OF AUTOMATION AND OUTSOURCING 64

Rifkin, J. (1995). The end of work: The decline of the global labor force and the dawn of the
post-market era. New York: Putnam.

Robbins, S. P. (1999). Layoff-survivor sickness: A missing topic in organizational behavior.
Journal of Management Education, 23(1), 31-43.
https://doi.org/10.1177/105256299902300104

Roberts, P. W., & Dowling, G. R. (2002). Corporate reputation and sustained superior
financial performance. Strategic Management Journal, 23(12), 1077-1093.
https://doi.org/10.1002/smj.274

Robertson, C. J., Lamin, A., & Livanis, G. (2010). Stakeholder perceptions of offshoring and
outsourcing: The role of embedded issues. Journal of Business Ethics, 95(2), 167-189.
https://doi.org/10.1007/s10551-009-0353-0

Roelfs, D. J., Shor, E., Davidson, K. W., & Schwartz, J. E. (2011). Losing life and livelihood:
a systematic review and meta-analysis of unemployment and all-cause mortality.
Social Science & Medicine, 72(6), 840-854.
https://doi.org/10.1016/j.socscimed.2011.01.005

Rokeach, M. (1973). The nature of human values. New York: Free press.

Rousseau, D. M. (1990). New Hire Perceptions of Their Own and Their Employer's
Obligations: A Study of Psychological Contracts. Journal of Organizational Behavior,
11(5), 389-400. https://doi.org/10.1002/job.4030110506

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic
motivation, social development, and well-being. American Psychologist, 55(1), 68-78.

Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of research
on hedonic and eudaimonic well-being. Annual Review of Psychology, 52(1), 141-166.
https://doi.org/10.1146/annurev.psych.52.1.141

Ryff, C. D. (1995). Psychological well-being in adult life. Current Directions in Psychological
Science, 4(4), 99-104. https://doi.org/10.1111/1467-8721.ep10772395

Schulz, A. C., & Johann, S. (2018). Downsizing and the fragility of corporate reputation: An
analysis of the impact of contextual factors. Scandinavian Journal of Management,
34(1), 40-50. https://doi.org/10.1016/j.scaman.2017.11.004

Schwartz, S. H. (2012). An overview of the Schwartz theory of basic values. Online readings
in Psychology and Culture, 2(1), 11. https://doi.org/10.9707/2307-0919.1116

Schwartz, S. H., & Bilsky, W. (1987). Toward a universal psychological structure of human
values. Journal of Personality and Social Psychology, 53(3), 550-562.
FAIRNESS OF AUTOMATION AND OUTSOURCING 65

Sen, C. (2018, January 24). Remember Outsourcing? Tech Companies Do. Bloomberg.
Retrieved from https://www.bloomberg.com/view/articles/2018-01-24/remember-
outsourcing-tech-companies-do

Sheldon, K. M., & Elliot, A. J. (1999). Goal striving, need satisfaction, and longitudinal well-
being: the self-concordance model. Journal of Personality and Social Psychology,
76(3), 482-497. http://dx.doi.org/10.1037/0022-3514.76.3.482

Sheldon, K. M., & Kasser, T. (2001). Goals, congruence, and positive well-being: New
empirical support for humanistic theories. Journal of Humanistic Psychology, 41(1),
30-50. https://doi.org/10.1177/0022167801411004

Shore, L. M., & Tetrick, L. E. (1994). The Psychological Contract as an Explanatory
Framework in the Employment Relationship. In C. L. Cooper & D. M. Rousseau
(Eds.), Trends in Organizational Behavior (pp. 91-109). New York: Wiley.

Shugan, S. M. (1984). Price-quality relationships. In T. Kinnear (ed.). Advances in Consumer
Research, 11, 627-32.

Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The roles of distributive,
procedural, and interactional justice. Journal of Applied Psychology, 82(3), 434-443.

Smith, A. (2016). Public predictions for the future of workforce automation. Pew Research
Center. Retrieved from http://pewresearch.org/wp-
content/uploads/sites/14/2016/03/PI_2016.03.10_Workforce-Automation_FINAL.pdf.

Strully, K. W. (2009). Job loss and health in the US labor market. Demography, 46(2), 221-
246. https://doi.org/10.1353/dem.0.0050

Tetlock, P. E. (2003). Thinking the unthinkable: Sacred values and taboo cognitions. Trends in
Cognitive Sciences, 7(7), 320-324. https://doi.org/10.1016/S1364-6613(03)00135-9

Thibaut, J. W., & Walker, L. (1975). Procedural justice: A psychological analysis. Hillsdale,
NJ: Erlbaum.

Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice.
Science, 211(4481), 453-458. doi: 10.1126/science.7455683

Tyler, T. R., & Bies, R. J. (1990). Beyond formal procedures: The interpersonal context of
procedural justice. In J. S. Carroll (Ed.), Applied Social Psychology and
Organizational Settings (pp. 77-98). Hillsdale, NJ: Erlbaum.

Tyson, L., & Spence, M. (2017). Exploring the effects of technology on income and wealth
inequality. In H. Boushey, J. B. DeLong, & M. Steinbaum (Eds.), After Piketty: The
agenda for economics and inequality. Harvard University Press.
FAIRNESS OF AUTOMATION AND OUTSOURCING 66

Uhlmann, E. L., Pizarro, D. A., Tannenbaum, D., & Ditto, P. H. (2009). The motivated use of
moral principles. Judgment and Decision making, 4(6), 476-491. Retrieved from
https://escholarship.org/uc/item/9g60v0vq

Van Buren, H. J. (2000). The bindingness of social and psychological contracts: Toward a
theory of social responsibility in downsizing. Journal of Business Ethics, 25(3), 205-
219. https://doi.org/10.1023/A:1006158022338

Van Dijk, F., Sonnemans, J., & van Winden, F. (2002). Social ties in a public good
experiment. Journal of Public Economics, 85(2), 275-299.
https://doi.org/10.1016/S0047-2727(01)00090-1

Wakabayashi, D. (2018, June 28). California Passes Sweeping Law to Protect Online Privacy.
The New York Times. Retrieved from
https://www.nytimes.com/2018/06/28/technology/california-online-privacy-law.html

Wanberg, C. R., Gavin, M. B., & Bunce, L. W. (1999). Perceived fairness of layoffs among
individuals who have been laid off: A longitudinal study. Personnel Psychology,
52(1), 59-84. https://doi.org/10.1111/j.1744-6570.1999.tb01813.x

Waterman, A. S. (1993). Two conceptions of happiness: Contrasts of personal expressiveness
(eudaimonia) and hedonic enjoyment. Journal of Personality and Social Psychology,
64(4), 678-691. http://dx.doi.org/10.1037/0022-3514.64.4.678

Weber, M. (1958). The Protestant ethic and the spirit of capitalism. New York: Scribner.

Wilson, S. H., & Walker, G. M. (1993). Unemployment and health: a review. Public health,
107(3), 153-162. doi: 10.1016/S0033-3506(05)80436-6

Wilson, T. D., Reinhard, D. A., Westgate, E. C., Gilbert, D. T., Ellerbeck, N., Hahn, C., . . .
Shaked, A. (2014). Just think: The challenges of the disengaged mind. Science,
345(6192), 75-77. doi: 10.1126/science.1250830

Winnick, E. (2017) Lawyer-Bots Are Shaking Up Jobs. Retrieved from
https://www.technologyreview.com/s/609556/lawyer-bots-are-shaking-up-jobs/

Yang, A. X., & Hsee, C. K. (2018). Idleness versus Busyness. Current Opinion in Psychology,
26, 15-18. https://doi.org/10.1016/j.copsyc.2018.04.015

Yoon, S., Oh, S., Song, S., Kim, K. K., & Kim, Y. (2014). Higher quality or lower price? How
value-increasing promotions affect retailer reputation via perceived value. Journal of
Business Research, 67(10), 2088-2096. https://doi.org/10.1016/j.jbusres.2014.04.017

Young, H. P. (1995). Equity: in theory and practice: Princeton, New Jersey: Princeton
University Press.
FAIRNESS OF AUTOMATION AND OUTSOURCING 67

Zeithaml, V. A. (1988). Consumer perceptions of price, quality, and value: a means-end model
and synthesis of evidence. The Journal of Marketing, 52(3), 2-22.
doi:10.2307/1251446

Zelizer, V. A. (1978). Human values and the market: The case of life insurance and death in
19th-century America. American Journal of Sociology, 84(3), 591-610. Retrieved
from https://www.jstor.org/stable/2778256
FAIRNESS OF AUTOMATION AND OUTSOURCING 68

APPENDICES

Appendix A
Pilot study analyses

Ninety-seven Mturkers responded to a survey about automation and outsourcing. Five
observations were removed due to suspicion that the individuals had participated more than once
(indicated by duplicate IP addresses), resulting in a final sample of 92 respondents (mean age =
33.20, SD = 9.65; 32.6% female, 67.5% male). Participants were asked to rate the extent to
which they believed the following items were relevant to automation and outsourcing:
“Increasing productivity,” “Producing more goods in a given amount of time,” “The company
saving money,” and “Paying workers less for the same amount of goods produced.” The first two
items (r automation = .67, p < .001; r outsourcing = .67, p < .001) were averaged to compute the increasing
productivity composite and the last two items (r automation = .36, p = .001; r outsourcing = .49, p < .001)
were averaged to compute the decreasing cost composite. Analyses of both the composites and
individual items revealed that respondents believed automation was associated with increasing
productivity to a greater extent than outsourcing, and outsourcing was associated with decreasing
costs to a greater degree than automation. The t-tests of the composites and individual items are
reported in Table A1 below.
FAIRNESS OF AUTOMATION AND OUTSOURCING 69

Table A1.
Descriptive and Difference Statistics for Measures in the Pilot Study

Measures
Outsourcing
M (SD)
Automation
t df p

M (SD)
Mean
Differenc

95%CI
Cohen'
s d
e
Increasing
productivity
(composite)
Decreasing
costs
(composite)
4.40 (1.79) 5.90 (1.23) 7.00 93 <.001 -1.49 [-1.92, -1.07] 0.73

5.84 (1.31) 5.41 (1.39) 2.90 93 .005 0.42 [0.13, 0.71] 0.31

Increasing
4.41 (1.96) 5.85 (1.40) 6.09 91 <.001 -1.435 [-1.90, -0.97] 0.64
productivity
Producing
more goods in
a given
amount of
time
The company
4.39 (1.97) 5.95 (1.30) 6.76 91 <.001 -1.554 [-2.01, -1.10] 0.71
saving money
6.1 (1.23) 5.77 (1.40) 2.33 91 .022 0.326 [0.05, 0.60] 0.25
Paying
workers less
for the same
amount of
goods
5.58 (1.79) 5.05 (1.95) 2.20 91 .030 0.522 [0.05, 0.99] 0.23
produced
As in all of our studies, we screened for outliers (data points more than 1.5 interquartile ranges
below the first quartile or above the third quartile). Here, two observations qualified as outliers
on the composite measures. When these were excluded from the respective analyses the results
were similar to those reported above.
FAIRNESS OF AUTOMATION AND OUTSOURCING 70

Appendix B
Study 1b: Additional measures and analyses

We also included measures of long-term and short-term progress judgments (outlined in greater
detail in Study 3) in an exploratory fashion because we had planned to examine these variables
as potential mediators in subsequent studies. ANOVA’s on the short-term and long-term
progress measures revealed no significant differences across conditions, F(2, 292) = 1.97, p
= .141, η2 = .01; F(2, 292) = 2.12, p = .122, η2 = .01, respectively. However, to better
understand the data patterns, we conducted LSD post-hoc tests to examine pairwise comparisons
of progress judgments. Analyses revealed that participants perceived greater potential for both
short-term and long-term progress from layoffs due to automation (ST: M = 5.11, SD = 1.23; LT:
M = 5.26, SD = 1.24) than from layoffs caused by outsourcing (ST: M = 4.81, SD = 1.26; LT: M
= 4.91, SD = 1.25) at a marginally significant and significant level, p = .080, mean difference =

-.30, 95%CI = [-.64, .04]; and p = .041, mean difference = -.35, 95%CI = [-.68, -.02],

respectively. Respondents also perceived greater potential for short-term progress with
automation-induced layoffs (M = 5.11, SD = 1.23) than merger-induced layoffs (M = 4.82, SD =
1.10) at a marginally significant level, p = .091, mean difference = .29, 95%CI = [-.05, .63]. No
such difference was found for long-term progress (Automation: M = 5.26, SD = 1.24; Merger: M
= 5.07, SD = 1.04), p = .257, mean difference = .19, 95%CI = [-.14, .53]. There were no
statistically significant differences in short-term and long-term progress judgments across the
outsourcing and merger conditions, ps > .340.

We detected 5 outliers for short-term progress and 2 outliers for long-term progress. After
excluding outliers, the overall effect of condition on short-term progress remained similar, p
= .104, but the difference between the automation and outsourcing condition became non-
FAIRNESS OF AUTOMATION AND OUTSOURCING 71

significant, p = .240, mean difference = -.19, 95%CI = [ = -.50, .12], and the difference between
the automation and merger condition became statistically significant, p = .034, mean difference
= .33, 95%CI = [.03, .64]. The overall effect of condition on long-term progress and the
difference across the automation and outsourcing conditions remained significant, p = .030 and p
= .009, respectively. However, the difference across the automation and merger condition moved
from non-significance to marginal significance, p = .094, mean difference = .28, 95%CI =
[-.05, .60].
FAIRNESS OF AUTOMATION AND OUTSOURCING 72

Appendix C
Study 2b: Results with outliers excluded

In Study 2b we detected six outliers for fairness perceptions. When these were excluded from the
analyses the results were similar to the reported results, with fairness perceptions higher in the
more productive replacement condition (M = 5.33, SD = 1.32) than the cheaper replacement
condition (M = 3.90, SD = 1.65), t(190) = 6.71, p < .001, mean difference = -1.44, 95%CI = [-
1.86, -1.01], d = .96. Levene’s test indicated unequal variances (F = .70, p = .006), so degrees of
freedom were adjusted from 191 to 190.
FAIRNESS OF AUTOMATION AND OUTSOURCING 73

Appendix D
Study 3: Results with outliers excluded

T-tests:

In Study 3, Sample 1, we detected 3 outliers on short-term progress, 4 outliers on long-term
progress, 1 outlier on benefit to company over customer, 2 outliers on benefit to company over
society, 1 outlier on quality, and 2 outliers on devaluation of workers. After excluding the
respective outlier(s) for each measure, the significance level of the differences across condition
were similar to those reported in the main manuscript (Table 2), with slightly higher p-values.
Condition differences were now only marginally significant for two of the mediators that had
shown significant differences in the full sample: long-term progress (productive replacement: M
= 5.23, SD = 1.20; cheaper replacement: M = 4.92, SD = 1.13), t(188) = 1.80, p = .073, mean
difference = -.30, 95%CI = [ -.64, .03], d = .27, and devaluation of worker (productive
replacement: M = 5.16, SD = 1.40; cheaper replacement: M = 5.55, SD = 1.33), t(190) = 1.97, p
= .050, mean difference = .39, 95%CI = [ -.00, .78], d = .29.

In Study 3, Sample 2, we detected 2 outliers on short-term progress, 3 outliers on long-term
progress, 1 outlier on benefit to company over customer, 3 outliers on benefit to company over
society, 2 outliers on quality, 9 outliers on devaluation of workers, and 7 outliers on motivation
to decrease costs. With the exclusion of the respective outlier(s) for each measure, t-tests
revealed results generally similar to those reported in the main manuscript (Table 2). The two
exceptions are as follows: quality of goods was perceived to be higher in the productive
replacement condition (M = 3.22, SD = 1.09) than the cheaper replacement condition (M = 2.93,
SD = 1.10), but instead of this difference being significant as in the full sample it was now only
marginally significant, t(196) = 1.91, p = .058, mean difference = -.30, 95%CI = [-.60, .01], d
FAIRNESS OF AUTOMATION AND OUTSOURCING 74

= .26, and the difference between condition in perceptions of short-term progress (which had
been marginally significant in the full sample) became statistically significant (productive
replacement: M = 5.30, SD = 1.03; cheaper replacement: M = 4.93, SD = 1.25), t(196) = 2.24, p
= .026, mean difference = -.37, 95%CI = [-.69, -.04], d = .32.

In the combined sample, we found 7 outliers for short-term progress, 8 outliers for long-term
progress, 2 outliers for benefit to company over customer, 5 outliers on benefit to company over
society, 2 outliers on quality, and 6 outliers on devaluation of workers. After excluding the
respective outlier(s) for each measure, the significance level of the differences across condition
were similar to those reported, with the following exception: the cross-condition difference of the
long-term progress ratings (productive replacement: M = 5.15, SD = 1.20; cheaper replacement:
M = 4.93, SD = 1.24), which had not been statistically significant before, became marginally
significant, t(384) = 1.75, p = .081, mean difference = -.22, 95%CI = [ -.46, .03], d = .18.

Mediation:

For each sample we ran a similar bootstrapping mediation analysis as described in the main
manuscript, considering multiple mediation via the mediators that showed significant condition
differences in the respective samples when outliers were excluded (see above). In Sample 1, after
excluding outliers, we had found that the difference across condition was nearly significant for
devaluation (p = .050). Thus, for the mediation analysis, we entered devaluation and
disenfranchisement (for which we had not detected any outliers) as proposed mediators, after
excluding the two data points that were outliers on devaluation. The analysis revealed a
significant total indirect effect through the two measures (b = .46, 95%CI = [.12, .82]) and a
significant specific indirect effect through disenfranchisement of workers (b = .36, 95%CI =
FAIRNESS OF AUTOMATION AND OUTSOURCING 75

[.10, .66]). We did not find a significant specific indirect effect through devaluation of workers
(b = .10, 95%CI = [-.002, .26], in contrast to the result with the full sample.

In Sample 2, a total of 18 data points qualified as an outlier on at least one of the proposed
mediators that exhibited significant differences across condition in the t-tests reported above:
short-term progress, devaluation, disenfranchisement, benefit to company over society,
motivation to increase capabilities, and motivation to decrease costs. Mediation analysis after
excluding the 18 data points revealed results that were consistent with the results reported in the
main manuscript. The total indirect effect through the 6 measures was statistically significant (b
= 1.03, 95%CI = [.63, 1.44]), as well as specific indirect effects through devaluation (b = .21,
95%CI = [.05, .45]), disenfranchisement (b = .28, 95%CI = [.09, .51]), and motivation to
increase capabilities (b = .43, 95%CI = [.16, .73]). We did not find significant specific indirect
effects through short-term progress (b = .05, 95%CI = [-.03, .15]), benefit to company over
society (b = .05, 95%CI = [-.06, .17]), or motivation to decrease cost (b = .01, 95%CI =
[-.11, .13]).
FAIRNESS OF AUTOMATION AND OUTSOURCING 76

Appendix E
Study 4: Scenarios presented to participants

Table E1.
Full Scenarios Used in Study 4
Layoff
reason
Motivation: Decreasing Costs Motivation: Increasing Capabilities

ALY Industries is in an increasingly
competitive landscape, and decides to
outsource some of its operations to a
different company in the same state.

As a result of this decision to outsource,
ALY fires 100 employees.

ALY Industries is in an increasingly competitive
landscape, and decides to outsource some of its
operations to a different company in the same
state.

As a result of this decision to outsource, ALY fires
100 employees.

Outsourcing
The reason for this decision is simple: they
want to decrease their costs as much as
possible.

With outsourcing, the work will be performed
by workers who are a less expensive option.
The workers will be more cost-efficient than
the current workforce, substantially reducing
the cost of operations and making ALY a
more profitable company as a result.

Ultimately, outsourcing would substantially
reduce the cost of ALY’s operations.
The reason for this decision is simple: the
organization wants the best production process
possible.

With outsourcing, the work will be performed by
workers who are a more effective option. The new
workers will be more productive, more reliable,
and less prone to mistakes than the current
workforce, making ALY a stronger company as a
result.

Ultimately, outsourcing would greatly enhance
ALY’s capabilities.

ALY Industries is in an increasingly
competitive landscape, and decides to
automate some of its operations.

As a result of this decision to automate, ALY
fires 100 employees.
ALY Industries is in an increasingly competitive
landscape, and decides to automate some of its
operations.

As a result of this decision to automate, ALY fires
100 employees.

Automation
The reason for this decision is simple: the
organization wants to decrease their costs
as much as possible.

With automation, the work will be performed
by machines that that are a less expensive
option. The machines will be more cost-
efficient than the current workforce,
substantially reducing the cost of operations
and making ALY a more profitable company
as a result.

Ultimately, automation would substantially
reduce the cost of ALY’s operations.
The reason for this decision is simple: the
organization wants the best production process
possible.

With automation, the work will be performed by
machines that are a more effective option. The
machines will be more productive, more reliable,
and less prone to mistakes than the current
workforce, making ALY a stronger company as a
result.

Ultimately, automation would greatly enhance
ALY’s capabilities.

FAIRNESS OF AUTOMATION AND OUTSOURCING 77

Appendix F
Study 4: Results with outliers excluded

In Study 4, two data points were outliers on the increasing capabilities manipulation check
measure and 21 data points were outliers on the decreasing costs manipulation check measure.
We ran the same analyses as described in the main manuscript with these outliers excluded from
the respective analysis. Results were similar to those reported in the main manuscript, except for
the interaction effect between layoff reason and motivation framing on increasing capabilities,
which moved further away from statistical significance, F(1, 402) = 2.64, p = .105, η2 = .01.
FAIRNESS OF AUTOMATION AND OUTSOURCING 78

Appendix G
Study 4: Additional measures and analyses

We collected ratings of the additional mediators that we had considered in Study 3. We intended
to explore ratings of these measures across the automation and outsourcing conditions (see t-tests
in Table G1 below). We did not find significant differences in the benefit to company over
society, devaluation, or disenfranchisement measures, as we had found in previous studies,
although there were significant differences in perceived quality and inevitability, and marginal
difference in benefit to the company over the customer.
Table G1.
Descriptive and Difference Statistics for Additional Variables Included in Study 4

over customer
over society

Note.
a
Levene's test indicated that the variance was significantly different across condition, F =
5.21, p = .023 so the degrees of freedom were adjusted from 406 to 394.
To further understand effects on these measures, we also considered whether our framing
manipulation may have influenced these ratings. Indeed, 2 x 2 ANOVAs on the measures
revealed significant main effects of motivation framing and interaction effects between layoff
reason and motivation (in addition to main effects consistent with the t-tests described above in
Table G1). More specifically, ratings of benefit to company over customer, benefit to company
over society, and devaluation were higher in the decreasing cost condition than the increasing
Measures
Automation Outsourcing
t df p
Mean
95% CI
Cohen's
(M, SD) (M, SD)

Difference

d
Benefit to company
5.02(1.43)

5.27(1.49)

1.73

406

.085

-0.25

[-0.54 , 0.03]

0.17
Benefit to company
5.63(1.26)

5.70(1.35)

0.56

406

.579

-0.07

[-0.33 , 0.18]

0.05

Quality 4.25(1.24)

3.66(1.47)

4.42

406

<.001

0.59

[0.33 , 0.86]

0.43

Inevitability 5.33(1.18)

4.59(1.49)

5.49 394
a

<.001

0.73

[0.47 , 0.99]

0.55

Devaluation 5.41(1.42)

5.53(1.45)

0.83

406

.408

-0.12

[-0.40 , 0.16]

0.08

Disenfranchisement 4.19(1.64)

4.38(1.63)

1.20

406

.230

-0.19

[-0.51 , 0.12]

0.12

FAIRNESS OF AUTOMATION AND OUTSOURCING 79

productivity condition. Conversely, ratings of quality were higher in the increasing productivity
condition than the decreasing cost condition. There were also marginally significant or
significant interaction effects between layoff reason and motivation framing for benefit to
company over customer, quality, devaluation, and disenfranchisement (see Table G2 for means
by condition and relevant statistics). The general pattern of results was as follows: when the
layoffs were framed as having been motivated by decreasing costs, the difference across the
layoff cause conditions resembled the main effect of layoff cause, but when the layoffs were
framed as having been driven by increasing capabilities, the difference was mitigated or even
reversed. The interaction effect was not significant for benefit to company over society (p
= .173), although the pattern was similar, or inevitability (p = .100), for which the pattern was
reversed: difference across layoff condition was greater in the increasing capabilities condition
than the decreasing cost condition. This analysis was highly exploratory, but suggests that future
work might consider the interplay between the different potential mediators we have identified
and the possibility of serial mediation (e.g., layoff cause influences perceptions of company
motivation, which affects other proposed mediators, which in turn shapes fairness perceptions).

Table G2.
The Effects of Motivation on the Additional Variables Included in Study 4

Measures Decreasing Cost Increasing Productivity Main Effect of Motivation Interaction Effect

over customer
over society
FAIRNESS OF AUTOMATION AND OUTSOURCING 80
Automation Outsourcing Automation
Outsourcing
F (1, 404) p η2 F (1, 404) p η2
M (SD) M (SD) M (SD) M (SD)

Benefit to company
5.26 (1.26)

5.72 (1.22) 5.51 (1.26) 4.82 (1.54) 25.07 <.001 0.06 3.53 .061 0.01
Benefit to company
5.77 (1.17)

5.98 (1.09) 5.88 (1.13) 5.51 (1.33) 11.37 .001 0.03 1.86 .173 0.01
Quality 3.94 (1.13) 2.98 (1.11) 3.43 (1.21) 4.52 (1.27) 68.00 <.001 0.14 12.90 <.001 0.03
Inevitability 5.13 (1.24) 4.63 (1.4) 4.86 (1.35) 5.49 (1.11) 1.20 .274 0.00 2.72 .100 0.01
Devaluation 5.5 (1.39) 5.87 (1.16) 5.7 (1.28) 5.33 (1.44) 10.63 .001 0.03 4.02 .046 0.01
Disenfranchisement 4.09 (1.64) 4.69 (1.52) 4.41 (1.6) 4.27 (1.64) 2.28 .132 0.01 7.18 .008 0.02

FAIRNESS OF AUTOMATION AND OUTSOURCING 81

Appendix H
Additional studies similar to Study 4: 2 (layoff cause: automation/outsourcing) x 2 (motivation:
increasing productivity/decreasing costs) design

Prior to running Study 4, we ran several similar studies in which we sought to manipulate the cause
of layoffs (decision to automate or outsource) and motivation driving the decision in order to
examine the effects of motivation framing on perceptions of fairness. In these original
implementations, we focused the motivation manipulation quite narrowly on increasing productivity
(versus decreasing cost), operationalizing this as the motivation to increase literal production (i.e.
output levels). We conjectured that this discrepancy was the reason we did not find a main effect of
motivation framing on fairness perceptions, despite having successfully manipulated motivation to
increase output and productivity (indicated by manipulation check measures). In reflecting on the
broader significance of increasing productivity and the finding that replacing workers with more
productive workers was deemed fairer than replacing workers with cheaper workers (Study 2a), we
realized that increasing productivity reflected more broadly an increase in capabilities. This led us
to collect the second sample reported in Study 3 (so that we could include motivation to increase
capabilities as a mediator), and, in Study 4, to focus on framing layoff decisions as having been
motivated by increasing capabilities or decreasing costs. Descriptions of each of these additional
earlier studies, along with the findings, are summarized in Table H1 below.
FAIRNESS OF AUTOMATION AND OUTSOURCING 82

Table H1.
The Results of the Studies Similar to Study 4
Study
Difference from
other studies

Fairness:
Main Results Other Measures and Results
Willingness to invest in the company:

a
Mturk
(N =
397)

Instead of using
the term
"outsource," we
described a
decision to
"transfer
operations to
another part of
the U.S. where
labor costs are
less expensive"
Main effect of layoff type was
marginally significant,
Automation: M = 4.37, SD = 1.43;
Outsourcing: M = 4.06, SD = 1.69;
F(1, 393) = 3.66, p = .056, η2
= .01.

Main effect of motivation was n.s.,
F(1, 393)<.001, p = .984, η2<.001.

Interaction between layoff type and
motivation was n.s., F(1, 393)
= .04, p = .849, η2<.001.
Main effect of layoff type was significant,
Automation: M = 4.13, SD = 1.72; Outsourcing: M
= 3.66, SD = 1.73; F(1, 393) = 7.22, p = .008, η2
= .02.

Main effect of motivation was n.s., F(1, 393)<.001,
p = .996, η2 = .00.

Interaction between layoff type and motivation was
n.s., F(1, 393) = .38, p = .539, η2 = .00.

No manipulation check included.

b
Mturk
(N =
397)

c
Mturk
(N =
381)

Attempted to
strengthen
manipulation of
motivation by
introducing
normative
aspect ("Most
companies
automate/outso
urce in order
to…")

Attempted to
strengthen
manipulation of
motivation
without the
normative
aspect
Fairness:

Main effect of layoff type was
significant, Automation: M = 4.22,
SD = 1.47; Outsourcing: M = 3.47,
SD = 1.59; F(1, 393) = 23.27, p
< .001, η2 = .06.

Main effect of motivation was n.s.,
F(1, 393) = .09, p = .768, η2<.001.

Interaction between layoff type and
motivation was n.s., F(1, 393) =
1.45, p = .229, η2 = .004.

Fairness:

Main effect of layoff type was
significant, Automation: M = 4.07,
SD = 1.56; Outsourcing: M = 3.37,
SD = 1.56; F(1, 377) = 19.53, p
< .001, η2 = .05.

Main effect of motivation was n.s.,
F(1, 377) = 1.21, p = .273, η2
= .003.

Interaction between layoff type and
motivation was marginal, F(1, 377)
= 2.80, p = .095, η2 = .01.
Willingness to invest in the company:

Main effect of layoff type was significant,
Automation: M = 4.21, SD = 1.63; Outsourcing: M
= 3.35, SD = 1.75; F(1, 393) = 26.15, p < .001, η2
= .06.

Main effect of motivation was n.s., F(1, 393) =
1.87, p = .172, η2 = .01.

Interaction between layoff type and motivation was
n.s., F(1, 393) = 1.76, p = .186, η2 = .004.

Manipulation check:
Analyses of the manipulation check items indicated
that we had successfully manipulated motivation.
Willingness to invest in the company:

Main effect of layoff type was significant,
Automation: M = 4.07, SD = 1.56; Outsourcing: M
= 3.37, SD = 1.56; F(1, 377) = 19.31, p < .001, η2
= .05.

Main effect of motivation was n.s., F(1, 377) = .43,
p = .512, η2 = .001.

Interaction between layoff type and motivation was
n.s., F(1, 377) = 1.78, p = .183, η2 = .01.

Manipulation check:
Analyses of the manipulation check items indicated
that we had successfully manipulated motivation.
FAIRNESS OF AUTOMATION AND OUTSOURCING 83

Appendix I
Additional mediation study

Prior to running Study 3, we ran a similar study in which we examined whether there was an
indirect effect of layoff cause (automation or outsourcing) on fairness perceptions through short-
term progress, benefit to company over society, benefit to company over customer, devaluation of
worker, disenfranchisement of worker, and quality of widgets produced. We replicated the main
effect of layoff cause on fairness perceptions found in our other studies. Short-term progress,
benefit to company over customer, devaluation of worker, and disenfranchisement of worker
exhibited significant differences across layoff condition (see Table I1 below for means and
difference statistics) and were tested as mediators using bootstrapping. There was a significant total
indirect effect of condition on fairness perceptions through the proposed mediators (b = .81, 95%CI
= [.48, 1.15]), as well as specific indirect effects through short-term progress (b = .31, 95%CI =
[.16, .49]), benefit to company over customer (b = .15, 95%CI = [.03, .32]), devaluation of worker
(b = .18, 95%CI = [.02, .36]), and disenfranchisement of worker (b = .17, 95%CI = [.05, .37]). The
mediation model is also presented below in Figure I1.

In this study we measured benefit to company over other stakeholders (customers and society) by
measuring the perceived benefit to each entity separately on similar 3 item scales and subtracting
the ratings of benefit to customer and society from the ratings of benefit to company (i.e., benefit to
company over customer was defined as the difference score between responses to the benefit to
company scale and benefit to customer scale; benefit to company over society was likewise the
difference between those respective measures). In addition, we realized in retrospect that several
items that were included to measure benefit to the company conceptually overlapped with items
intended to measure company progress. Due to concerns about the use of difference scores and the
FAIRNESS OF AUTOMATION AND OUTSOURCING 84

redundancy of some of our measures, we ran Study 3 with revised and improved measures and
report the improved study in the main manuscript.
Table I1.
Descriptive and Difference Statistics of all Variables Included in the Study

p
Diff.

Difference d

Benefit to Society
Benefit to Consumer
Worker
of Worker

Measure
Cheaper
Replacement
Productive
Replacement

t (195)
Mean 95% CI of the Cohen's
M (SD) M (SD)
Fairness 2.83 (1.71) 3.76 (1.54) 4.00 <.001 -0.93 [-1.39, -.47] 0.57
Short-term Progress 4.57 (1.48) 5.64 (0.85) 6.18 <.001 -1.07 [-1.41, -.73] 0.88
Benefit to Company 5.44 (1.22) 5.74 (0.99) 1.86 .064 -0.30 [-.61, .02] 0.27
Benefit to Society 3.00 (1.57) 3.65 (1.48) 2.99 .003 -0.65 [-1.08, -.22] 0.43
Benefit to Consumer 3.86 (1.54) 4.61 (1.29) 3.72 .001 -0.75 [-1.15, -.35] 0.53
Benefit to Company -
2.44 (1.92)

2.09 (1.68) 1.38 .170 0.35 [-.15, .86] 0.19
Benefit to Company -
1.59 (1.60)

1.13 (1.24) 2.23 .027 0.46 [.05, .86] 0.32
Devaluation of
5.78 (1.13)

5.40 (1.29) 2.23 .027 0.38 [.04, .72] 0.32
Disenfranchisement
4.96 (1.39)

4.35 (1.45) 3.01 .003 0.61 [.21, 1.01] 0.43
Improve Quality 2.97 (1.07) 3.10 (0.99) 0.86 .391 -0.13 [-.42, .16] 0.13

FAIRNESS OF AUTOMATION AND OUTSOURCING 85

Figure I1.
Multiple mediation model whereby the effect of layoff reason (0 = cheaper replacement, 1 =
productive replacement) on Fairness Perceptions is mediated by ratings of Short-term Progress,
Benefit to Company over Consumer, Devaluation of Worker, and Disenfranchisement of Worker.
Values represent unstandardized regression coefficients and the values in parentheses represent the
standard errors. The values after the slash show the direct effect after controlling for the indirect
effect. * p < .05. **p < .01. ***p < .001.
FAIRNESS OF AUTOMATION AND OUTSOURCING 86

Appendix J
Additional studies exploring the role of progress: 2 (layoff cause: automation/outsourcing) X 2
(progress specified/unspecified) design

In the additional mediation study described in Supplement 9, progress (short-term) emerged as a
significant mediator of the relationship between layoff cause and fairness perceptions. As a result,
we ran a series of 3 studies investigating whether holding constant the level of progress following
the layoff decisions would mitigate differential evaluations of automation and outsourcing. We
anticipated an interaction between layoff cause and a progress specified/unspecified condition, such
that we would replicate the layoff cause effect when progress was not specified (similar to our
earlier findings), but find no difference when progress was held constant across conditions. We also
included measures of long-term progress as an exploratory measure in these studies because we
were concerned that manipulations attempting to equalize perceptions of short term progress across
conditions might not extend to perceptions of long-term progress (and that automation might
therefore still be associated with greater progress, despite explicitly providing equivalent short-term
progress information in both layoff type conditions). Across the 3 studies, contrary to what we had
expected, we found that there was no interaction effect between layoff and progress condition. We
did, however, consistently replicate the main effect of layoff cause, consistent with our other
studies. Analysis of the manipulation check and long-term progress measures revealed a main effect
of layoff condition in which short-term and long-term progress judgments were higher with
automation than with outsourcing. This suggested that despite our description of progress as equal
across the automation and outsourcing conditions, people persisted in the general belief that
automation is conducive to greater progress. Descriptions of each study, along with the findings, are
summarized in Table J1 below.
FAIRNESS OF AUTOMATION AND OUTSOURCING 87

Table J1.
Results of the Studies on the Role of Progress

Study
Difference from
other studies
Main Results Other Measures and Results

a
Mturk
(N =
383)

b
Mturk
(N =
386)

c
Mturk
(N =
388)
In the progress
unspecified
condition, no
additional
information was
given regarding
company progress,
similar to the
studies reported in
the main
manuscript; this
feature was
consistent across all
three studies. In the
progress specified
condition, P's were
told: "QZT will still
be producing the
same amount of
goods as before."

To make the level
of progress more
salient, in the
progress specified
condition, P's were
given concrete
numerical estimates
of growth/progress:
"According to an
estimate produced
by analysts at the
company, this
decision should
increase production
by 15% in the next
5 years and 20% in
the following 10
years."

To further heighten
the salience of
progress, the
decision was
presented as having
been made 5 years
prior and in the
progress specified
condition, P's were
given concrete
numerical estimates
of growth/progress
in the past tense:
"According to an
outside estimate,
this decision
increased
production by 15%
in the past 5 years."
Fairness:

-Main effect of layoff type
was significant, Automation:
M = 3.96, SD = 1.63;
Outsourcing: M = 3.14, SD =
1.55; F(1, 379) = 25.09, p
< .001, η2 = .06.

-Main effect of progress
condition was significant,
Specified: M = 3.73, SD =
1.66; Unspecified: M = 3.38,
SD = 1.62; F(1, 379) = 4.21,
p = .041, η2 = .01.

-Interaction between layoff
type and motivation was n.s.,
F(1, 379) = .47, p = .496, η2
= .001.
Fairness:

-Main effect of layoff type
was significant, Automation:
M = 4.21, SD = 1.59;
Outsourcing: M = 3.44, SD =
1.59; F(1, 382) = 16.52, p
< .001, η2 = .04.

-Main effect of progress
condition was significant,
Specified: M = 4.14, SD =
1.63; Unspecified: M = 3.50,
SD = 1.58; F(1, 382) = 8.68,
p = .003, η2 = .02.

-Interaction between layoff
type and motivation was n.s.,
F(1, 382) = 1.52, p = .219,
η2 = .004.

Fairness:

-Main effect of layoff type
was significant, Automation:
M = 3.98, SD = 1.57;
Outsourcing: M = 3.38, SD =
1.52; F(1, 384) = 14.36, p
< .001, η2 = .04.

-Main effect of progress
condition was significant,
Specified: M = 3.89, SD =
1.57; Unspecified: M = 3.47,
SD = 1.55; F(1, 384) = 6.29,
p = .013, η2 = .02.

-Interaction between layoff
type and motivation was n.s.,
F(1, 384) = .01, p = .934,
η2<.001.
Manipulation check (short-term progress):

-Main effect of layoff type was marginally significant, Automation: M
= 4.73, SD = 1.19; Outsourcing: M = 4.52, SD = 1.31; F(1, 379) =
2.93, p = .088, η2 = .01.
-Main effect of progress condition was significant, Unspecified: M =
4,77, SD = 1.19; Specified: M = 4.48, SD = 1.29; F(1, 379) = 5.22, p
= .023, η2 = .01.
-Interaction between layoff type and motivation was n.s., F(1, 379)
= .04, p = .846, η2<.001.

Manipulation check (long-term progress):

-Main effect of layoff type was significant, Automation: M = 5.05, SD
= 1.22; Outsourcing: M = 4.64, SD = 1.39; F(1, 379) = 9.23, p = .003,
η2 = .02.
-Main effect of progress condition was n.s., F(1, 379) = .65, p = .422,
η2 = .002.
-Interaction between layoff type and motivation was n.s., F(1, 379)
= .13, p = .718, η2<.001.
Manipulation check (short-term progress):

-Main effect of layoff type was significant, Automation: M = 5.41, SD
= 1.07; Outsourcing: M = 4.90, SD = 1.13; F(1, 382) = 11.64, p
= .001, η2 = .04.
-Main effect of progress condition was significant, Specified: M =
5.49, SD = 1.03; Unspecified: M = 4.81, SD = 1.12; F(1, 382) = 29.53,
p < .001, η2 = .07.
-Interaction between layoff type and motivation was n.s., F(1, 382)
= .06, p = .815, η2 = .00.

Manipulation check (long-term progress):

-Main effect of layoff type was significant, Automation: M = 5.53, SD
= 1.15; Outsourcing: M = 4.87, SD = 1.20; F(1, 382) = 17.70, p
< .001, η2 = .04.
-Main effect of progress condition was significant, Specified: M =
5.62, SD = 1.05; Unspecified: M = 4.77, SD = 1.23; F(1, 382) = 41.21,
p < .001, η2 = .10.
-Interaction between layoff type and motivation was n.s., F(1, 382)
= .11, p = .738, η2<.001.
Manipulation check (short-term progress):

-Main effect of layoff type was significant, Automation: M = 5.19, SD
= 1.06; Outsourcing: M = 4.77, SD = 1.34; F(1, 384) = 11.53, p
= .001, η2 = .03.
-Main effect of progress condition was significant, Specified: M =
5.47, SD = .94; Unspecified: M = 4.50, SD = 1.29; F(1, 384) = 71.31,
p < .001, η2 = .16.
-Interaction between layoff type and motivation was marginally
significant, F(1, 384) = 2.99, p = .084, η2 = .01.

Manipulation check (long-term progress):

-Main effect of layoff type was significant, Automation: M = 5.20, SD
= 1.09; Outsourcing: M = 4.85, SD = 1.13; F(1, 384) = 9.07, p = .003,
η2 = .04.
-Main effect of progress condition was significant, Specified: M =
5.23, SD = 1.10; Unspecified: M = 4.81, SD = 1.12; F(1, 384) = 13.54,
p < .001, η2 = .03.
-Interaction between layoff type and motivation was n.s., F(1, 384)
= .00, p = .960, η2<.001.
FAIRNESS OF AUTOMATION AND OUTSOURCING 88

Appendix K
Pre-registration links

Studies in the main manuscript:

Pilot: http://aspredicted.org/blind.php?x = ri28kd
Study 1a: http://aspredicted.org/blind.php?x = rt23gu
Study 1b: http://aspredicted.org/blind.php?x = 76de8k
Study 2a: http://aspredicted.org/blind.php?x = yy675z
Study 2b: http://aspredicted.org/blind.php?x = u8dp6u
Study 3 (sample 1): http://aspredicted.org/blind.php?x = s3jw99
Study 3 (sample 2): http://aspredicted.org/blind.php?x = pj3em9
Study 4: http://aspredicted.org/blind.php?x = 9s3w8u

Additional studies described in Appendix H:

Study a: http://aspredicted.org/blind.php?x = 5ni7gs
Study b: http://aspredicted.org/blind.php?x = 6cf2fw
Study c: http://aspredicted.org/blind.php?x = eu6q48

Additional study described in Appendix I:
http://aspredicted.org/blind.php?x = d59q28

Additional studies described in Appendix J:

Study a: http://aspredicted.org/blind.php?x = cq6fs9
Study b: http://aspredicted.org/blind.php?x = 6v6md8
Study c: http://aspredicted.org/blind.php?x = 6tq6eg

Abstract (if available)

Abstract Automation is a critical and timely source of job displacement, but little is known about how observers react to automation-driven layoffs. In a series of six studies, I contrast reactions to automation with reactions to outsourcing, a conceptually similar organizational decision. Studies 1a-1b demonstrate that people find layoffs caused by automation to be fairer than those caused by outsourcing. Studies 2a-2b find a similar pattern when the issue of technology is moot: respondents find replacement of a worker to be fairer when a more productive worker is hired as a replacement (analogous to automation) than when a cheaper worker is hired (analogous to outsourcing). Study 3 considers mediating mechanisms for these effects

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The curse of loyalty: interdependent self construal and support for corruption

PDF

Technology, behavior tracking, and the future of work

PDF

An examination of outsourcing student services for online graduate students

PDF

The morality of technology

PDF

Better now than later: the cost of victims' delayed accusations

PDF

Novelty versus familiarity: divergent effects of low predictability and low personal influence

PDF

The effects of accounting performance and professional relationships on promotion, dismissal, and transfer decisions in a conglomerate

PDF

Executive compensation: the trend toward one size fits all

PDF

Workplace organization and asset pricing

PDF

Mutual fund screening versus weighting

PDF

Voluntary disclosure responses to mandated disclosure: evidence from Australian corporate tax transparency

PDF

A framework for research in human-agent negotiation

PDF

Expectation dynamics and stock returns

PDF

The importance of rigor and engagement on student achievement and mastery: an evaluation study

PDF

Public opinion and international affairs: a multi-method approach to foreign policy attitudes

PDF

Essays on econometrics

PDF

Effectiveness of the SEC’s comment letters in initial public offerings

PDF

Dividend policy, earnings announcements, and analysts' earnings forecasts

PDF

Scale construction and effects of brand authenticity

PDF

Social determinants of physiological health and mortality in China

Asset Metadata

Creator Kim, Jennifer (author)

Core Title Increasing capabilities or decreasing cost: Fairness perceptions of job displacement due to automation and outsourcing

School Marshall School of Business

Degree Doctor of Philosophy

Degree Program Business Administration

Publication Date 12/04/2018

Defense Date 07/09/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag automation,Fairness,job displacement,Justice,layoffs,OAI-PMH Harvest,outsourcing,productivity

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wakslak, Cheryl (committee chair), Appel, Gil (committee member), Fast, Nathanael (committee member)

Creator Email jennk677@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-110353

Unique identifier UC11675785

Identifier etd-KimJennife-6995.pdf (filename),usctheses-c89-110353 (legacy record id)

Legacy Identifier etd-KimJennife-6995.pdf

Dmrecord 110353

Document Type Dissertation

Format application/pdf (imt)

Rights Kim, Jennifer

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA