Is There a Right Way to Lay Off Employees in Times of Crisis?: The Role of Organizational Justice in the Case of Airbnb

Abstract

The COVID-19 pandemic has negatively affected the tourism and services sector. Using the example of Airbnb’s recent layoff of 25% of its workforce, we focused on the role of organizational justice in mitigating the negative psychological impacts of layoffs. Based on a unique survey of Airbnb employees who survived the layoffs, as well as those who left, we employed an ordinary least squares regression to show that employees’ perceptions of organizational justice were positively related to their job satisfaction and trust in management, while being negatively related to their emotional exhaustion and cynicism. We discovered the crucial importance of interactional justice (i.e., interpersonal and informational justice). The respect, dignity, and politeness shown by management (i.e., interpersonal justice), as well as truthful and adequate communication about the procedure (i.e., informational justice) were pivotal to successfully conducting layoffs, especially during unprecedented economic uncertainty.

1. Introduction

Tourism demand is vulnerable to external factors such as changes in climate, economic conditions, and political instability [1,2]. Many researchers have argued that the COVID-19 pandemic presents a transformative moment, threat, or opportunity that will change the world, especially the tourism sector, by impacting jobs [3,4]. Countries worldwide have enacted travel restrictions in response to COVID-19. These included stay-at-home orders for domestic travel, except for essential services, and closures of international borders, except for returning nationals and permanent residents [5].

Tourism and services employees were increasingly vulnerable to job loss during the COVID-19 pandemic, as it slowed down travel. The pandemic represented contemporary and unprecedented threats to job security in the industry. McKinsey [6] reported that, as a result of social distancing, the accommodation and food service sector was one of the most severely-affected sectors, accounting for 20% of all vulnerable positions. Reports on layoffs and bankruptcies in the tourism and services sector have followed since the breakout of COVID-19; Marriott International, the world’s largest hotel company laid off thousands of its employees in March, 2020 [7]. Furthermore, Banyan Tree laid off 10% of its 11,000 staff in June. Major airlines, including Air Canada, Air New Zealand, and Scandinavian Airlines, laid off thousands of workers [8]. Online travel booking agency Expedia also laid off 12% of its 24,000 full-time employees in February [9].

It is expected that many companies in the tourism and services sector will accelerate layoffs as COVID-19 disrupts the economy [9]. Companies lay off employees to survive, retaining productivity with a smaller workforce. Yet, there is growing evidence that layoffs create negative effects, such as lower performance levels, on the remaining employees, [10,11,12]. For those employees laid off, job loss can be both economically and psychologically devastating. For the remaining employees, the possibility of additional rounds of layoff and the shocks from the previous layoff may create anxiety and dissatisfaction towards their jobs. Therefore, understanding how to mitigate layoff shocks on employees can be critical to ensuring successful transitions and survival measures, especially during an unprecedented crisis such as the COVID-19 pandemic.

In this study, we focused on the role of organizational justice in mitigating layoff shocks on both the laid off and remaining employees. Specifically, we examined the impact of four components of organizational justice (i.e., distributive, procedural, informational, and interactional) on employees’ psychological states as measured by job satisfaction, trust, cynicism, and emotional exhaustion. Previous studies have reported positive effects of organizational justice on various outcome variables [13,14,15]. Organizational justice may be particularly important during the layoff process, as the organizations’ employees are likely to evaluate the fairness of the process and question the legitimacy of the layoff [16]. Employees’ perceptions of justice may also be important for protecting the organization’s image, reducing chances of employee lawsuits, and demonstrating corporate social responsibility.

The evidence for this study comes from Airbnb, one of the largest peer-to-peer tourism and services in the world. Specifically, we examined the relationship between employees’ perceptions of the layoff and their psychological states and attitudes. We analyzed these for both the so-called survivors and victims (i.e., those who stayed in the organization and those who left) to produce a compelling estimate. This short-term-rental startup’s market value significantly dropped in April 2020 after booking volume plummeted by 80%. In May 2020, the company’s CEO, Brian Chesky, announced its plan to lay off 25% of the 7500-strong workforce. No matter how much management attempts to make a layoff empathetic and transparent, employees may be still disappointed with how the company approaches the process [17]. Thus, this large downsizing provides a unique opportunity to observe how employees’ justice judgments influence their psychological states and attitudes, as measured by their perceived levels of job satisfaction, trust, cynicism, and emotional exhaustion.

The remainder of the article proceeds as follows. In the next section, we explain the context of this study and the layoffs in the tourism and services industry during the pandemic. Next, we review the previous relevant literature and present our hypotheses. In Section 4, we describe our methodology and explain how we constructed the dataset. In Section 5, we present the test of our proposed hypotheses. Based on the findings, we present our conclusions and discuss our findings and implications.

2. Airbnb and Tourism and Services Sector

This study focused on the layoff in the tourism and services sector which faced unprecedented challenges due to border closings, curtailment of travel, lockdown-measures, and closure of attractions and tourism facilities as a result of COVID-19 [18]. We selected Airbnb, one of the most successful peer-to-peer (P2P) platforms in the tourism and services sector, as a subject for investigation. Airbnb is headquartered in San Francisco, California, USA, has 27 offices across 19 countries, and has successfully become a popular alternative to traditional accommodation, despite regulatory challenges [19,20] This sharing economy business is expected to disrupt the hotel industry, as Airbnb has already become bigger than the world’s largest hotel chains, when measured by the number of rooms in the market [21,22].

Yet, Airbnb was not exempt from the devastating impact that the COVID-19 pandemic had on the travel industry. Airbnb’s CEO, Brian Chesky, acknowledged that Airbnb’s business had been hit hard, with 2020 revenue forecasted to be less than half of the revenue in 2019. In April 2020, Airbnb told its employees that the executives would take a 50% salary reduction, the employees were unlikely to get their 2020 bonus, and they would institute a hiring freeze and suspend marketing spending. The company raised USD 2 billion in new debt funding and cut costs across the business [23]. About a month later, in May, 2020, Airbnb CEO, Brian Chesky, announced, on Airbnb’s blog post, that Airbnb would lay off 25% of its workforce, amounting to 1900 of 7500 employees. At the time of the announcement, Airbnb informed its employees of the layoff plan with compassion and empathy and offered generous severance packages, generating positive media coverage [23].

We selected Airbnb as the subject of this study due to its massive layoff and its popularity, as well as its growth in business. Focusing on Airbnb for this study ensured that a diverse range of employees in different roles, functions, and teams, as well as seniority levels in multiple offices worldwide, would be considered. With its growing popularity and exponential business growth, Airbnb has been the focus of discussion among practitioners and scholars alike [24,25,26]. Belarmino and Koh [27] conducted a systematic review of previous studies that examined peer-to-peer accommodations and found that the human resources topic had not been explored in previous literature. This omission motivated this current research.

3. Literature Review and Hypotheses Development

3.1. Organizational Justice and Layoffs

Employee layoff is a corporate strategy to downsize the workforce, improve performance, and optimize capacity [10,28]. Many scholars have argued that layoffs can lead to increased shareholder value and net benefits to the firm [29,30], such as improvements in subsequent corporate performance [29,30,31,32]. However, other researchers have demonstrated that layoffs, if conducted ineffectively, can backfire by decreasing the productivity and morale of the remaining employees in a post-layoff environment [11,33,34,35,36,37]. This could lead to a decline in company earnings and employee productivity [38,39].

Previous studies have explored whether layoffs have psychological effects on the remaining employees [11,33,34,37]. For example, Brockner et al. [34] showed that post-layoff survivors become less committed. Others have shown that layoffs are associated with lower levels of well-being [11,37]. This result is not surprising, as layoffs can be perceived as a signal that the company no longer upholds its share in the psychological contract, which can be perceived as a breach of the psychological contract. Consequently, post-layoff survivors’ reactions are likely to be potentially harmful and negative, which can influence important work attitudes and psychological states, such as job commitment, satisfaction, and trust [10,11,40,41,42]. Furthermore, they may even respond to layoffs by engaging in dysfunctional behaviors, such as retaliation [43]; reactions such as anger, relief, guilt, and resentment are commonly experienced by post-layoff survivors [41,44].

Previous research has demonstrated the critical importance of organizational justice during the layoff process [16]. Organizational justice refers to employees’ perception of fairness in the workplace, as employees assess the fairness of their organizations’ actions during layoffs. Organizational justice theory suggests that how employees are treated and how decisions are made during layoffs may have significant impacts on their perceptions and behaviors [42,43,45,46,47,48]. For example, how a company conducts layoffs informs post-layoff survivors about how they can expect to be treated in the future [42,46,48]. Similarly, providing adequate information in advance and showing interpersonal sensitivity could reduce potential negative reactions [48,49], whereas impersonal and abrupt ways of executing the changes could cause resentment and retaliation [50].

In summary, past research has greatly enhanced our understanding of the impact of employee layoffs and the importance of organizational justice during the process. However, previous studies have largely focused on the impact of layoffs on post-layoff survivors, paying relatively little attention to how the laid off employees perceive the process (see [51,52]). This is a gap in the literature because the perception of laid-off individuals may also have important consequences for organizations. For instance, laid off employees who perceive the procedure as unfair are generally unwilling to recommend the organization to others [53]. Victims, employers, and the society thus have a shared interest in getting those affected by layoffs back to work. Further, we do not clearly know whether the critical importance of organizational justice reported in previous studies can apply to a layoff during an exceptionally historic moment, when downsizing is often regarded as inevitable.

To address this gap, this study explored the impact of organizational justice during a major layoff by taking advantage of the recent downsizing in the tourism and services sector. Specifically, we investigated the impacts of organizational justice on four psychological constructs: job satisfaction, trust, cynicism, and emotional exhaustion. These variables have been found to have important implications for organizational functioning [42,54,55,56,57,58]. We thus investigated whether and to what extent survivors’ and victims’ psychological states differed in terms of these measures, and then estimated the association between organizational justice and each of these measures.

3.2. Hypotheses Development

In the early years of organizational justice literature, scholars focused solely on distributive justice, which mainly concerned itself with the fairness of decision outcomes (i.e., equitable allocation of resources among individuals). For instance, Adams [59] showed that perceived injustice in outcome allocations is associated with dissatisfaction, decreased efforts, and decreased turnover. Since then, scholars have expanded the concept of organizational justice to include the fairness of decision-making procedures through which outcomes are decided, termed procedural justice. For instance, Thibaut and Walker [60] argued that procedures in legal proceedings may be viewed as fair if disputants could voice their concerns to influence the verdict (i.e., outcome).

Distributive and procedural justice have constituted the primary components of justice for decades. However, later works have introduced another component, interactional justice, which focuses on the fairness of interpersonal interaction. Bies and Moag [61] argued that interactional justice is achieved if the management communicates procedural details in an adequate and respectful way and justifies the decision outcomes with honest and truthful information. The concept of interactional justice is further disaggregated into interpersonal justice, the quality of interpersonal treatment, and informational justice, the quality of information received [62]. Research has shown that the four components of organizational justice (i.e., distributive, procedural, interpersonal, and informational) may have independent effects on organizational outcomes, although they are closely related and interact with one another [14].

In this study, we examined the role of the four dimensions of organizational justice in the context of an employee layoff. To be clear, distributive justice refers to the appropriateness of layoff outcomes (i.e., the management’s decision regarding who stays or leaves the organization), and procedural justice refers to the appropriateness of the layoff decision process. On the other hand, interactional justice refers to the appropriateness of the treatment one receives from management during the layoff process. This justice concept has two aspects: interpersonal justice is about treating an employee with courtesy, dignity, and respect, and information justice is about sharing relevant information with employees during the layoff process.

We expect that these four dimensions of organizational justice will have significant impacts on employees’ psychological states and attitudes. We considered four outcome variables to measure employees’ reactions: two positive outcomes (job satisfaction and trust in management), and two negative ones (cynicism and emotional exhaustion). These four psychological constructs are chosen, as they are important indicators of employee attitudes and behaviors. We hypothesize that the four dimensions of justice will be positively related to job satisfaction and trust in management, whereas they will negatively affect cynicism and emotional exhaustion. We thus suggest:

H1-1: 

Four components of organizational justice—distributive, procedural, interpersonal, and informational justice—are positively associated with employees’ job satisfaction.

H1-2: 

Four components of organizational justice—distributive, procedural, interpersonal, and informational justice—are positively associated with employees’ trust in management.

H1-3: 

Four components of organizational justice—distributive, procedural, interpersonal, and informational justice—are negatively associated with employees’ cynicism.

H1-4: 

Four components of organizational justice—distributive, procedural, interpersonal, and informational justice—are negatively associated with employees’ emotional exhaustion.

Although we expect that the four dimensions of justice may all be positively related to these outcome variables, at least to some extent, their relative influence may vary significantly. Previous studies have shown that consideration of procedural and interactional justice is far more salient to employees experiencing the layoff than that of distributive justice [63]. That is, when employees perceive that their management has adhered to procedural, interpersonal, and informational justice, it can result in beneficial attitudes and behaviors [13,14,64]. As Natunann et al. [65] put it, “it is not the layoff per se, but the attention (or lack thereof) afforded to concern for the employees that determines individuals’ reactions.” Following this group of studies, we hypothesize that distributive justice will have the least significant effects on employees’ psychological states and attitudes during a layoff. Therefore, we suggest:

H2: Distributive justice is less strongly associated with job satisfaction, trust in management, cynicism, and emotional exhaustion than the rest of the three dimensions of organizational justice (i.e., procedural, interpersonal, and informational justice).

Although previous literature tends to agree that outcome fairness issues (i.e., distributive justice) are less important than the other three components during a layoff situation, the relative importance of procedural and interactional (i.e., interpersonal and informational) justice is ambiguous. Research focusing on survivors’ reactions has generally emphasized the importance of procedural justice by demonstrating that the negative impact of a layoff can be decreased by following a fair process [40,46,47,50]. From a practical perspective, procedural justice may be important to survivors because of its potential inference for future expected outcomes [46,60,66,67,68]. On the other hand, research on employees’ reactions to layoffs has paid greater attention to the role of informational and interpersonal justice [61].

Based on the literature, it is tempting to conclude that procedural justice matters to survivors, whereas interactional justice matters to victims. However, this conclusion ignores an important variable: the broader organizational and environmental context within which layoffs occur. In this study, we considered a layoff that occurred during an exceptionally historic moment, the COVID-19 pandemic. This context is special for the following reasons. Considering the catastrophic impacts of COVID-19 on the tourism and services sector, employees may expect that layoffs are inevitable. Further, even the layoff survivors may still be uncertain about their future, as there might be another round of downsizing, depending on the economic situation. For instance, an employee that we interviewed as a part of the survey explained,

“Even if almost 25% of employees were laid off, there is still uncertainty around whether we would be able to keep our jobs. I know that the CEO said that there won’t be a second wave of layoffs but when the company first laid off partners and contractors, the company told us that internal employees will not be affected. That is why I never thought that there would be even a first wave of layoffs and now I think we should definitely be worried about a second wave. So, I think that even if I work hard, I could eventually be laid off.”

Although survivors are not laid-off, they may also experience high stress levels similar to those who are laid off [69]. In such a highly uncertain context, survivors are also victims. In fact, previous research has suggested that interactional justice may matter to both victims and survivors after a downsizing operation [51]. Building on these findings, we hypothesize that interactional (i.e., informational and interpersonal) justice will have far more important impacts on the psychological states of both survivors and victims in such an exceptional context. The respect, dignity, and courtesy shown by the management (i.e., interpersonal justice), as well as truthful and adequate communication about the procedure (i.e., informational justice) are vital for successfully conducting layoffs in moments of uncertainty. Therefore, we hypothesize the following:

H3: Interactional justice (i.e., informational and interpersonal justice) is more strongly associated with job satisfaction, trust in management, cynicism, and emotional exhaustion than the other two dimensions of organizational justice (i.e., procedural and distributive), given the environmental context of the layoff.

4. Method

4.1. Empirical Method

To evaluate our hypotheses, we analyzed a recent layoff by Airbnb to examine the impacts of organizational justice as perceived by the company’s employees, along with their psychological states and attitudes. We surveyed both the survivors and victims to understand the role of justice perception in the two groups. To this end, we employed the following simple model, using STATA (V.13):

$$Outcome{s}_i=\alpha +\beta OrgJus{t}_i+\gamma X_i+{\nu }_i$$

(1)

Here, $Outcome{s}_i$ is the dependent variable, which is the psychological state and attitude of employee i. We examined four outcome variables: (1) job satisfaction, (2) trust, (3) cynicism, and (4) emotional exhaustion. These outcome variables are the summative scales generated from several survey items that are measured as a categorically ordered variable on a Likert scale from 1 (“To a very small extent”) to 5 (“To a very large extent”). We used the alpha command in STATA to produce the scale variables. These scales are continuous, unlike the categorical variables. Hence, we employed an ordinary least squares regression to test the direction and magnitudes of the hypothesized relationships among the four components of organizational justice and each of the outcome variables.

Here, $OrgJus{t}_i$ is the independent variable of our model, which is the level of organizational justice as perceived by employee i. The model includes all four components of organizational justice (i.e., distributive, procedural, interactional, and informational). We also include a set of control variables, $X_i$. Specifically, we controlled for the age, gender, educational level, marital status, years of experience in tourism and services, years at Airbnb, and an indicator of whether the survey respondent is a victim (i.e., whether they left the organization due to the layoff). We also estimated the model separately for survivors and victims to see whether the association between organizational justice and employees’ psychological states and attitudes are different between the two groups. The model includes the error term, ${\nu }_i$.

In this model, both the dependent and independent variables are measured from self-reported perceptual data, which may create the well-known common method bias [70,71]. Despite such concerns, we decided to use this data because access to such a unique set of employee perceptions is crucial for advancing our understanding of the role of organizational justice during a COVID-19-related employee layoff. These data are especially unique in that both the survivors and victims responded to the same survey, which makes it possible for us to compare the effects of organizational justice between the two groups. Moreover, previous studies have suggested that common source bias may not be as serious as is often claimed [72]. For example, Conway and Lance [73] report evidence that the coefficients estimated from self-reported data may not necessarily suffer upward bias. Consequently, self-reported perceptual data are commonly used in social science research, although one should be cautious in interpreting the results.

4.2. Data and Variables

We collected the survey data in June 2020, after the company laid off employees in most offices except for those in Europe where, legally, required redundancy consultation processes were taking place. To recruit a wide variety of functions and seniority levels across different offices worldwide and to ensure a representative sample, we relied on the convenience sampling and snowball sampling techniques. Because one of the authors had worked at Airbnb for four years until recently, we asked former Airbnb employees who were laid off and current employees to recruit potential research participants. Primarily, we used the Airbnb Talent Directory and LinkedIn connections to identify and recruit a representative sample of former and current employees. We explained the purpose of the research and asked them to fill out an online survey using Google forms.

To address any potential concern about the anonymity of participants’ responses, participants were made aware of the fact that the survey would be anonymous, and that all identifying information, such as job title and functions, as well as locations, would be removed from the dataset. A total of 117 employees participated in the study, yielding a 25.6% response rate; the sample consisted of 52 Airbnb employees who survived the layoff, as well as 65 former employees who were laid off during the process.

Once all data were collected, we examined whether our survey participants were somehow different from the population of potential respondents by comparing the demographic characteristics of our respondents with the characteristics of current Airbnb employees. Three current employees of Airbnb confirmed that our participants’ demographic profile mirrored the profile of the employees at the company. Here, it is important to acknowledge the uniqueness of this sample. The sample population was well-reflected in terms of position level, years within Airbnb, years in the hospitality industry, and educational background.

All measurement constructs were operationalized with multi-items on a 5-point Likert scale, ranging from 1–5. Each measurement construct was adopted from well-developed prior research, with some minor working changes to fit this study’s layoff context. For example, three items of organizational justice from Colquitt and Rodell [74], job satisfaction from Warr and Payne [75], emotional exhaustion from Maslach [76] trust from Gabarro and Athos [77] and cynicism from Dean et al. [78]. were used. For the emotional exhaustion construct, a license from Mind Garden, the copyright holder of the Maslach Burnout Inventory (MBI), was purchased via their website (MBI altered by the survey administrator. Copyright ©1996 Wilmar B. Schaufeli, Michael P. Leiter, Christina Maslach, and Susan E. Jackson. All rights reserved in all media. Published by Mind Garden, Inc., www.mindgarden.com accessed on 28 February 2022). The detailed measurement items are shown in

Table 1. Survey items and Cronbach’s Alpha.

Category/Question	Mean	SD	Cronbach’s Alpha
Job Satisfaction			0.95
I enjoy my job.	3.6	1.1	-
I am satisfied with my job.	3.4	1.1	-
I am happy with my job.	3.4	1.1	-
Trust			0.89
I believe my current employer has high integrity.	3.5	1.0	-
I can expect my current employer to treat me in a consistent and predictable fashion.	3.4	1.0	-
I believe my current employer is not always honest and truthful.	2.2	1.2	-
In general, I believe my current employer’s motives and intentions are good.	3.8	1.0	-
I do not think my current employer treats me fairly.	2.0	1.0	-
My current employer is open and up-front with me.	3.4	1.0	-
I am not sure I fully trust my current employer.	2.2	1.1	-
Cynicism			0.90
I believe that my company says one thing and does another.	2.4	1.1	-
My company’s policies, goals, and practices seem to have little in common.	2.1	1.0	-
When my company says it’s going to do something, I wonder if it will really happen.	2.6	1.2	-
My company expects one thing of its employees, but rewards another.	2.5	1.3	-
I see little similarity between what my company says it will do and what it actually does.	2.3	1.1	-
Emotional Exhaustion			0.92
I feel used up at the end of a workday.	3.1	1.1	-
I feel fatigued when I get up in the morning and have to face another day on the job.	2.7	1.2	-
Working with people all day is a real strain on me	2.1	1.1	-
I feel burned out from my work.	2.7	1.3	-
I feel frustrated by my job.	2.5	1.3	-
I feel I am working too hard on my job.	2.6	1.2	-
I feel like I am ‘at the end of my rope.’	2.2	1.3	-
Distributive Justice			0.95
Do those outcomes reflect the effort you have put into your work?	2.5	1.2	-
Are those outcomes appropriate for the work you have completed?	2.5	1.2	-
Do those outcomes reflect what you have contributed to your work?	2.5	1.3	-
Are those outcomes justified, given your performance?	2.4	1.3	-
Procedural Justice			0.82
I was able to express my views during the layoff procedures.	2.2	1.3	-
I was able to influence the decisions arrived at by the procedures.	1.4	0.8	-
The layoff procedures were applied consistently	2.7	1.3	-
The procedures were free of bias.	2.5	1.2	-
The procedures were based on accurate information.	2.6	1.2	-
I was able to appeal the decisions arrived at by the procedures.	1.5	1.0	-
The procedure upheld ethical and moral standards.	2.9	1.3	-
Interpersonal Justice			0.94
The company treated me in a polite manner.	3.9	1.1	-
The company treated me with dignity.	3.8	1.1	-
The company treated me with respect.	3.8	1.2	-
The company refrained from improper remarks or comments.	4.0	1.1	-
Informational Justice			0.89
My leader was candid when communicating with me.	3.6	1.1	-
My leader explained the decision-making procedures thoroughly.	3.0	1.4	-
My leader’s explanations regarding procedures were reasonable.	3.0	1.3	-
My leader communicated details in a timely manner.	3.4	1.2	-
My leader tailored communications to meet my individual needs.	3.1	1.4	-

. We also include a “sample profile” in

Table 2. Survey respondents: sample profile.

Variable	Category	Survivors	Victims	Total
Gender	Female	29	35	64
	Male	23	30	53
Marital Status	Unmarried	32	44	76
	Married	20	21	41
Age	20–24		2	2
	25–29	20	17	37
	30–34	15	29	44
	35–39	7	6	13
	40–45	10	7	17
	45 and above		4	4
Job Functions	Communications	5	1	6
	Customer Support	16	17	33
	Engineering	1	8	9
	Human Resources	1	5	6
	Marketing	1	6	7
	Operations	5	4	9
	Public Policy	15	7	22
	Sales	4	15	19
	Trust and Safety	4	2	6
Level of Education	Juris Doctor		1	1
	Ph.D.		1	1
	Master’s Degree	12	17	29
	Bachelor’s Degree	34	43	77
	Associate Degree		1	1
	Vocational-Hotel Management	1		1
	High School Diploma	5	2	7
Years of service at Airbnb	Less than 1	1	16	17
	1–2	7	17	24
	2–4	20	13	33
	4–6	14	15	29
	More than 6	10	4	14
Years of service in the industry	Less than 1		12	12
	1–2	6	12	18
	2–4	17	13	30
	4–6	12	10	22
	6–8	12	10	22
	More than 8 years	5	8	13
Annual Income	USD 100,000–150,000	5	9	14
	USD 150,000–200,000	5	5	10
	USD 200,000–250,000	3	1	4
	USD 30,000–50,000	5	11	16
	USD 50,000–100,000	17	17	34
	<USD 30,000		1	1
	(blank)	17	21	38

, in which we report survey respondents’ gender, age, marital status, job functions, level of education, years of service at Airbnb, years of service in the industry, and annual income. Lastly, we present the summary statistics for all variables included in the analysis in

Table 3. Descriptive statistics.

Variables		Mean	S.D.	1.	2.	3.	4.	5.	6.	7.	8.
1.	Job satisfaction	3.46	1.07	1.00
2.	Trust	1.10	0.82	0.45	1.00
3.	Cynicism	2.39	0.97	−0.39	−0.71	1.00
4.	Emotional exhaustion	2.55	0.99	−0.57	−0.55	0.57	1.00
5.	Distributive justice	2.44	1.18	0.06	0.13	−0.01	−0.10	1.00
6.	Procedural justice	2.25	0.81	0.21	0.38	−0.38	−0.26	0.38	1.00
7.	Interactional justice	3.86	1.03	0.22	0.55	−0.45	−0.32	0.08	0.49	1.00
8.	Informational justice	3.22	1.06	0.37	0.53	−0.42	−0.31	0.25	0.54	0.44	1.00

.

5. Results

5.1. Mean Difference between Survivors and Victims

Before we examine the impacts of organizational justice on the dependent variables, we first demonstrate whether the key variables differ between survivors and victims. In

Table 4. Mean differences between survivors and victims.

Variables	Survivors (N = 52)		Victims (N = 65)		Difference
	Mean	S.D.	Mean	S.D.
Dependent variables
Job satisfaction	3.571	1.030	3.369	1.093	0.201
Trust	1.198	0.668	1.018	0.915	0.180
Cynicism	2.496	0.792	2.308	1.083	0.188
Emotional exhaustion	2.434	0.916	2.637	1.037	−0.203
Organizational Justice
Distributive justice	2.817	0.958	2.135	1.252	0.683 **
Procedural justice	2.286	0.905	2.220	0.732	0.066
Interactional justice	3.750	0.988	3.950	1.063	−0.200
Informational justice	3.419	1.011	3.074	1.089	0.345 *
Control variables
Age	3.154	1.195	3.123	1.193	0.031
Male	0.462	0.503	0.462	0.502	0.000
Educational level	1.288	0.498	1.262	0.477	0.027
Marital status	0.615	0.491	0.692	0.465	−0.077
Years in hospitality	3.827	1.184	3.308	1.676	0.519 *
Years at Airbnb	3.462	0.999	2.584	1.249	0.877 **

* p < 0.10, ** p < 0.05.

, we report the mean differences between the two groups for all the variables included in the model. Three items are immediately apparent from Table 4. First, the employees’ psychological states and attitudes (i.e., the dependent variables) are not significantly different between survivors and victims. On average, while survivors expressed a higher job satisfaction, higher trust in the company, and lower emotional exhaustion, they also expressed a higher level of cynicism regarding the employer. However, none of these differences are statistically meaningful.

Second, survivors and victims expressed different perceptions about the level of organizational justice. Survivors reported significantly higher levels of distributive and informational justice than victims. The difference in the levels of procedural and information justice is not significant. Third, employees reported higher levels of interactional and informational justice than levels of distributive and procedural justice. Overall, employees seem to perceive that the employer provided adequate explanation and information for the layoff and demonstrated concern for the victims.

Lastly, we tested the difference in the covariates between survivors and victims. The two groups were not noticeably different in terms of age, gender, marital status, and educational level. However, we found that the survivors had worked longer at Airbnb than had the victims. The former group had more extensive professional experience in the tourism and services sector than the latter. These differences in work experiences were statistically significant. Although we cannot guarantee that our sample is truly representative, we have a reasonable level of confidence that we would find similar results if we surveyed all the employees who remained at the organization and all those who left.

5.2. The Impacts of Organizational Justice on Employees’ Psychological States and Attitudes

We now present the regression results of Equation (1). In

Table 5. The impacts of organizational justice on job satisfaction and trust.

	Dependent Variable: Job Satisfaction			Dependent Variable: Trust
	(1)	(2)	(3)	(4)	(5)	(6)
	Total	Survivors	Victims	Total	Survivors	Victims
Distributive justice	−0.006	−0.152	0.070	0.018	0.063	0.022
	(0.099)	(0.128)	(0.122)	(0.047)	(0.102)	(0.057)
Procedural justice	−0.004	0.029	−0.272	0.026	0.036	−0.048
	(0.160)	(0.110)	(0.335)	(0.106)	(0.081)	(0.252)
Interactional justice	0.127	0.081	0.344 **	0.370 **	0.371 **	0.377 **
	(0.105)	(0.170)	(0.119)	(0.076)	(0.083)	(0.126)
Informational justice	0.343 **	0.586 **	0.092	0.222 **	0.172	0.241 **
	(0.087)	(0.183)	(0.136)	(0.064)	(0.115)	(0.090)
N	117	52	65	117	52	65
R2	0.208	0.393	0.204	0.456	0.443	0.487

Note: Standard errors, clustered by office, in parentheses, ** p < 0.05; all models include the employees’ age, gender, educational level, marital status, years of experience in hospitality and tourism, and years at Airbnb.

, we report the associations between organizational justice and the two positive outcomes (i.e., job satisfaction and trust), whereas in

Table 6. The impacts of organizational justice on cynicism and emotional exhaustion.

	Dependent Variable: Cynicism			Dependent Variable: Emotional Exhaustion
	(1)	(2)	(3)	(4)	(5)	(6)
	Total	Survivors	Victims	Total	Survivors	Victims
Distributive justice	0.067	0.192	−0.024	−0.017	0.162	−0.075
	(0.101)	(0.137)	(0.100)	(0.122)	(0.146)	(0.165)
Procedural justice	−0.202	−0.311 **	0.056	−0.129	−0.086	0.120
	(0.176)	(0.100)	(0.309)	(0.174)	(0.140)	(0.347)
Interactional justice	−0.263 *	−0.187	−0.291	−0.267 **	−0.346 **	−0.261
	(0.138)	(0.172)	(0.186)	(0.094)	(0.116)	(0.176)
Informational justice	−0.210 **	−0.070	−0.284 *	−0.129	−0.236	−0.118
	(0.092)	(0.166)	(0.135)	(0.088)	(0.145)	(0.164)
N	117	52	65	117	52	65
R2	0.311	0.355	0.392	0.177	0.404	0.157

Note: Standard errors, clustered by office, in parentheses, * p < 0.10, ** p < 0.05; all models include the employees’ age, gender, educational level, marital status, years of experience in hospitality and tourism, and years at Airbnb.

, we show the association between organizational justice and the two negative outcomes (i.e., cynicism and emotional exhaustion). Here, a higher score for either job satisfaction or trust is a positive sign for the organization, whereas a higher score for either cynicism or emotional exhaustion is a negative sign. For each dependent variable, we estimated three models: (1) the full sample model, in which we include both survivors and victims, (2) the survivors model, in which we analyzed only the survivors, and (3) the victims model, in which we analyzed only the victims. In each model, we controlled for the covariates: the employees’ age, gender, educational level, marital status, years of experience in hospitality and tourism, and years at Airbnb. The standard errors were clustered at the office level because employees in different offices may hold different perceptions about the studied measures.

In the first column in Table 5, we present the association between the four components of organizational justice and job satisfaction. As can be seen, of the four components, informational justice is significantly associated with job satisfaction. The second and third columns in Table 5 separately examine the association, for survivors and victims. The results, with survivors in column 2, confirm the importance of informational justice in improving job satisfaction among employees after the layoff. On the other hand, the results, with victims in column 3, showed a different story; among the employees who were laid off during the process, we find a significant association between the perceived level of interpersonal justice and their job satisfaction. In the rest of the three columns in Table 4, we report how employees’ justice judgments are associated with their trust in management. In column 4, we find that the employees’ perceived levels of both informational and interpersonal justice have a meaningful association with their trust. As can be seen in column 5, interpersonal justice matters more to survivors. On the other hand, to the victims, both informational and interpersonal justice were positively associated with trust level. Taken altogether, the results showed the importance of informational and interpersonal justice during layoffs; we find that employees’ perceived levels of informational and interpersonal justice are positively associated with job satisfaction and trust in management.

In Table 6, we report the association between the four components of organizational justice and the negative psychological outcomes (i.e., cynicism and emotional exhaustion). In column 1, we find that both informational and interpersonal justice have meaningful negative associations with cynicism. In column 2, we show that procedural justice matters to the survivors. On the other hand, as can be seen in column 3, informational justice is more important to the victims. In columns 4, 5, and 6, we show whether and to what extent employees’ emotional exhaustion is associated with their justice perceptions. This time, only interpersonal justice show significant associations with the outcome; the survivors’ perceptions about interpersonal justice are negatively associated with their emotional exhaustion. Thus, we find once again that informational and interpersonal justice during the layoffs have critical effects on the psychological states and attitudes of the employees (i.e., both survivors and victims). We found that distributive justice is not associated with any of the outcome variables, and procedural justice has a meaningful association for only one specification (in column 2 of Table 6).

6. Discussions

In this study, we explored the role of organizational justice in mitigating layoff shocks on both the laid off and remaining employees. For this, we conducted a unique survey of Airbnb employees who survived the layoffs, as well as those who did not.

We summarize our findings as follows. First, the results of our analyses suggest that employees’ perceptions of organizational justice are positively related to their job satisfaction and trust in management and negatively related to their emotional exhaustion and cynicism. These associations between employees’ justice perceptions and their psychological states were observed among both the survivors and victims. Consistent with past research, this result demonstrated the importance of organizational justice in mitigating the negative psychological shocks of layoffs [51,52,53,79,80,81].

Second, we found that, for both the survivors and victims, interactional justice (i.e., informational and interpersonal justice) is most strongly associated with employees’ psychological states and attitudes. Procedural justice was associated only with the survivors’ cynicism. We found little evidence that distributive justice is associated with any of the psychological constructs we tested in this study, i.e., job satisfaction, trust in management, emotional exhaustion, and cynicism. This result is consistent with previous findings; consideration of procedural and interactional justice is far more salient to employees who experience the layoff than that of distributive justice [52,63,82].

This finding, that interactional and procedural justice matters more than distributive justice, merits further explanation. We found negligible impacts of distributive justice, which is somewhat inconsistent with previous research [51,69,83].

We argue that this result may be driven by the unique circumstance of the layoff. Specifically, most interviewees acknowledged that layoffs were inevitable given the company’s plummeting revenues during the pandemic. In such exceptional circumstances, when a large layoff is triggered by a factor beyond their control, we argue that distributive justice may carry less weight. That is, distributive justice matters less because employees all recognize that the layoff is caused by an event beyond their control (i.e., COVD-19) rather than their poor performance outcomes. In such circumstances, employees care less about distributive justice, which is concerned with how rewards and costs are distributed across group members, because they perceive the layoff as inevitable, regardless of their contributions to the organization.

Further, our finding that interactional justice is the most important for both survivors and victims also merits attention. Previous research has generally reported that different types of justice matter to survivors and victims. Puzzlingly, however, we found little difference between survivors and victims, regarding the impact of justice. We argue that this apparent puzzle is also caused by the unique circumstance of COVID-19. Specifically, survivors may have been almost equally threatened by the layoff decisions due to the extremely uncertain future situation. Thus, we note that readers should be cautious in generalizing our findings to normal times; our findings may be specific to a global macro crisis.

7. Conclusions and Implications

This study extended several strands of previous scholarly works. First, our research is especially meaningful, as we offered a comprehensive testing of the impact of organizational justice during a major COVID-19 layoff. Second, this study reported timely evidence about the impacts of the pandemic on the tourism and services sector. This is important because professionals in the tourism and services sector face greater threats to their job security than those in any other industry in the wake of COVID-19 [84]. Third, the results of this study extended the tourism and services literature in the service sector by presenting evidence of how various components of organizational justice affect employees’ perceptions in distinct ways.

Before we conclude, we should acknowledge that our research has limitations. First, we relied on cross-sectional self-reported survey data to examine our research query. As a result, we acknowledge that it is difficult to claim causality from our analyses. However, our dataset in unique in that it allows us to test the impact of organizational justice during a major COVID-19 layoff. Our findings also provide a point of comparison for future work aimed at increasing the understanding of the impact of organizational justice during large downsizing events. Second, our sample size is relatively small, with 117 respondents—52 Airbnb employees who survived the layoff, as well as 65 former employees who were laid off during the process. This sample size is relatively small for drawing a decisive conclusion. Our small sample may also limit, to some extents, our power to obtain statistically significant results. Despite the small sample size, however, our data allows us to report timely evidence about the impacts of the pandemic on the tourism and services sector.

To conclude, our findings support the hypothesis that, in times of great uncertainty, a layoff process that embodies interactional justice (i.e., informational and interpersonal justice) becomes crucial to mitigating the negative psychological consequences of the event. The central premise of this argument is that, in times of great uncertainty, survivors are also victims. This is because, in such a context, the management’s promise that there will be no additional layoffs may be perceived as “cheap talk;” the survivors may also perceive that their future is still extremely uncertain and suffer psychological anxiety.

Our findings also exhibit important practical implications for human resource management. Organizational downsizing, redundancies, and layoffs are frequent occurrences in today’s society, and communicating the organizational change is, therefore, an important regular responsibility for leaders and managers. We conclude that the respect, dignity, and politeness shown by the management, as well as truthful and adequate communication about the procedure, has become ever more important during the COVID-19 recession.