1. Introduction

In the workplace, it is common for employees to be at a disadvantage in terms of salary and promotion compared with their colleagues. Management psychologists and organizational behavior scholars use the concept of workplace envy to reflect employees’ negative emotions in face of upward social comparison (Foster, 1972) . Envy is an unpleasant emotion that individuals experience when others have what they desire but lack (such as excellent qualities, achievements or possessions, etc.), which includes inferiority, desire, resentment and hostility towards the envied (Parrott & Smith, 1993: p. 906) .

A large number of studies have discussed its harmful consequences to organizations. For example, some studies have found that individuals who experience envy tend to be hostile and resentful, thus reducing their organizational citizenship behaviors (Kim et al., 2010) . Envy can affect group satisfaction and lead to absenteeism, social undermining and counterproductive work behavior (Duffy, Scott, Shaw, Tepper, & Aquino, 2012; Khan, Quratulain, & Bell, 2014) ; envy can also indirectly reduce group performance by increasing social undermining and reducing group effectiveness and cohesion. In addition, high performers are more likely to be targets of group members’ envy and victims (Kim & Glomb, 2014) ; then high performers may reduce their efforts when they realize they are being envied. Envy may lead to abusive behaviors of leaders towards high-performing subordinates (Yu, Duffy, & Tepper, 2018) . Some scholars use malicious envy to describe such hostile envy (van de Ven, 2016; van de Ven et al., 2009) .

Because of its harm to the organization, scholars began to explore the mechanism and influencing factors of envy, trying to reduce workplace envy. Unfortunately, little relatively research has been done. These studies focused on the influence of factors such as the personal characteristics of the jealous person and the object of envy, and the personality traits of the jealous person such as dispositional envy and attachment style (Lange, Crusius, & Hagemeyer, 2016; Lange & Crusius, 2015; Smallets, Streamer, Kondrak, & Seery, 2016) ; the characteristics of the object of envy, such as similarity, warmth and so on (Schaubroeck, & Lam, 2004; Tai, Narayanan, & McAllister, 2012) . There are differences about the effect of self-esteem on workplace envy: one theory said that people with higher self-esteem are more likely to be jealous; another theory said that individuals with higher self-esteem are less likely to be envy. To enrich and extend envy study, this study based on social comparison theory and cognitive conflict theory, explores the common influence of gender identity and self-esteem on workplace envy in upward social comparison.

2. Research Hypotheses

To be specific, this paper first assumes that employees will experience envy in the context of upward social comparison. The relative status produced by social comparison makes individuals to evaluate themselves, low relative status means low self-evaluation (Festinger, 1954) ; one may feel jealous when confronted with low self-worth information (Silver & Sabini, 1978) . Therefore, noticing another person’s comparative advantage—making upward social comparisons with others—can lead to envy (Smith, 2000) .

This paper then assumes that employee gender identity will enhance the effect of upward social comparison on envy. Specifically, compared with employees with lower gender identity, employees with higher gender identity have more workplace envy in the face of upward social comparison with colleagues. Identity within the group increases self-worth, individuals with high gender identity have higher self-evaluation. When faced with situations that threaten self-evaluation, individuals with high sense of identity will have strong cognitive conflicts. The upward social comparison scenario is one that threatens self-worth, therefore, this paper believes that compared with employees with low gender identity, employees with high gender identity are more jealous when making unfavorable social comparison.

Finally, this paper assumes that employees’ self-esteem increases workplace envy in the face of social comparison. Employees have a high self-evaluation of themselves and think that they are excellent individuals in the organization, when employees have a high explicit self-esteem. However, when an employee with a high self-evaluation encounters upward social comparison, the employee finds that other colleagues are better than him; then he feels a low sense of self-worth. At this time, the high self-worth brought by the high explicit self-esteem and the low self-worth brought by the unfavorable social comparison formed the cognitive conflict, which further deepened the workplace envy. The research model of this paper is shown as Figure 1.

2.1. Upward Social Comparison and Workplace Envy

Upward social comparisons can lead to envy experienced by employees. Because there will be strong negative emotions such as inferiority and envy when individuals have negative self-evaluation (Festinger, 1954) . Social comparison often has a powerful effect on self-evaluation, because people usually evaluate themselves by making social comparisons with others. According to the social comparison principle, employees use other people as reference points to evaluate their own performance (Festinger, 1954) . In the organization, employees keep track of the performance development of themselves and their colleagues through regular performance reviews and publicly visible rewards in the organization (such as employee of the month sales tournament, etc.). Employees actively compare these attributes to their peers in an upward social way (Brown, Ferris, Heller, & Keeping, 2007) . When the excellent performance of others threatens the self-evaluation of individuals, individuals will be jealous (Silver & Sabini, 1978) . This kind of social comparison that has a negative effect on the self is called upward social comparison. Individuals will make upward social comparison, especially when they pay attention to the comparative advantages of other colleagues. Such upward social comparison will have a negative impact on self-evaluation, which will lead to negative emotions such as envy (Smith & Kim, 2007) .

Therefore, we propose hypothesis 1: upward social comparison induces workplace envy.

2.2. The Moderating Effect of Gender Identity

We further infer that the effect of upward social comparison on envy is mediated by employee gender identity. First, individuals pay attention to those with more similar attributes, such as gender, age and social class, and make social comparisons with them (Gastorf & Suls, 1978; Goethals & Darley, 1977) . Gender is a social structure that may serve as an important basis (other examples may be nationality, religion, social status, family member, etc.), so gender factors can influence people’s behavioral responses to social comparisons. As an important inherent attribute of individuals, gender is an important part of individuals’ self cognition. The identity of individuals to their own gender will affect themselves’ value evaluation. However, workplace envy is caused by upward social comparison that affects individuals’ self-evaluation. Therefore, gender identity can interactively affect individuals’ self-assessment and thus affect the generation of workplace envy.

Second, identity within the group increases self-worth, individuals with high gender identity have higher self-evaluation. When faced with situations that threaten self-evaluation, individuals with high sense of identity will have strong cognitive conflicts. The upward social comparison scenario is one that threatens self-worth, therefore, this paper believes that compared with employees with low gender identity, employees with high gender identity are more jealous when making unfavorable social comparison.

Therefore, we propose hypothesis 2: gender identity regulates the relationship between social comparison and workplace envy.

2.3. The Moderating Effect of Self-Esteem

Finally, we predict that the higher the level of employees’ self-esteem, the more workplace envy will be generated in the face of upward social comparison. The self-esteem of this paper adopts the definition of explicit self-esteem.

Employees have a high self-evaluation of themselves and think that they are excellent individuals in the organization, when employees have a high explicit self-esteem. However, when an employee with a high self-evaluation encounters upward social comparison, the employee finds that other colleagues are better than him; then he feels a low sense of self-worth. At this time, the high self-worth brought by the high explicit self-esteem and the low self-worth brought by the unfavorable social comparison formed the cognitive conflict, which further deepened the workplace envy. However, employees with low explicit self-esteem originally had a low self-worth evaluation. The low self-worth generated in the face of adverse social comparison was consistent with the low self-worth generated by low self-esteem, which did not form cognitive conflict and would not deepen workplace envy. Compared with low self-esteem, employees with high self-esteem will form strong cognitive conflicts and generate higher workplace envy when they face the unfavorable social comparison with their colleagues.

Second, individuals have a motivation to maintain a positive self-evaluation (Pyszczynski, Greenberg, Solomon, Arndt & Schmiel, 2004; Tesser, 2000) , as a result, most people place their self-worth more in areas where they think they’re doing well. When there is a social comparison that is not conducive to positive self-evaluation in the field of self-confidence, the individual’s self-worth will be more threatened, thus generating more envy.

Therefore, we propose hypothesis 3: self-esteem moderates the relationship between social comparison and workplace envy. Compared with low self-esteem individuals, high self-esteem individuals produce more workplace envy in the face of adverse social comparison.

3. Method

3.1. Sample

We used snowball method to recruited 543 employees from 27 provinces in China as participants online. The experimental process is voluntary and participants can stop or quit at any time (data is treated as invalid).According to the answers to polygraph questions, 14 invalid answers and 529 valid data were screened, and the effective rate of the questionnaire was 97.42%. Participants were aged 19 - 60 years, with an average age of 31.39 years (SD = 9.81), among which 270 were males (accounting for 51.00%) and 259 were females (accounting for 49.00%). Participants would be paid 9 RMB after completing the experiment.

3.2. Experimental Design and Procedures

We used a univariate intergroup design in which we manipulated the independent variables by priming upward social comparisons of the subjects. Gender identity and self-esteem were measured as situational envy of the target co-worker.

After signing the informed consent form, the subjects filled in their gender and trait envy scale (control variable). In order to control the influence of gender, we will conduct a balanced treatment on the gender of objects of envy in the study materials. The subjects were then randomly presented with one of four scenarios.

Upward social comparison. We manipulated upward social comparisons by presenting participants with descriptive situational priming material. In the upwardly social comparison group priming scenario, a colleague would perform well (appreciated by the leader, perform well, and receive a high performance bonus), while his own performance was average, thus manipulating the participants’ upwardly social comparison; In the control group, participants were presented with no advantage in the text context, and both colleagues and themselves were average, with the same information as in the upward social comparison group.

3.3. Measures

Envy (situational) was measured by Cohen-charash (2009) and Schaubroeck and Lam (2004) . For example, I have a grudge (resentment, bitterness) against X. The scale was scored at 6 points, with “1” being “very inconsistent” and “6” being “very consistent”. In this study, the α coefficient of the scale was 0.78.

Unfavourable social comparisons were measured by Cohen-Charash (2009) and Schaubroeck and Lam (2004) 4-item social comparison scale in situational envy. The sample items are as follows, “X has things going better for him/her than I do”. The scale was scored on a 6-point scale, with 1 representing “very inconsistent” and 6 representing “very consistent”. In this study, the scale’s α score coefficient was 0.79.

The situational envy scale and the upward social comparison scale were based on exploratory factor analysis results and subject content decomposition of Cohen-charash (2009) and Schaubroeck and Lam (2004) scales. The load of four items of situational envy scale is 0.83, 0.80, 0.75, 0.69, respectively. The load of the four items in the adverse social comparison scale was 0.84, 0.82, 0.77, 0.64.

Gender identity is a revised version of Schmader’s (2002) gender identity scale. There are seven sample items, for example, “Overall, being a woman is an important part of my self-image”. The scale was scored on a 6-point scale, with 1 representing “very inconsistent” and 6 representing “very consistent”. In this study, the scale’s α score coefficient was 0.76.

The explicit self-esteem scale of Rosenberg (1964) was adopted for self-esteem, and there were 10 items in total.

Control variables the study controlled for the effects of trait envy, which was measured on Smith’s (1999) 8-item scale.

4. Data Analysis

4. 1. Confirmatory Factor Analysis

As shown in Table 1, the results of Confirmatory Factor Analysis (CFA) showed that the fitting effect of the four-factor model was significantly better than that of the other three competing models, trait envy, situational envy (emotional dimension), gender identification and self-esteem have good discriminative validity (χ² = 1318.66; χ²/df = 3.55; RMSEA = 0.07; NFI = 0.87; IFI = 0.90; CFI =0.90). The fitting index of the three-factor model (trait envy + situational envy, gender identification, self-esteem) of the competition model is as follow: χ² = 1694.11;

χ²/df = 4.53, RMSEA = 0.08, NFI = 0.83, CFI = 0.86; The fitting index of the two-factor model (trait envy + situation envy + self-esteem gender identity) of the competition model isas follow: χ² = 2546.45; χ²/df = 6.77, RMSEA = 0.10, NFI = 0.74, CFI = 0.77; The fitting index of the single-factor model is as follow: χ² = 3590.50; χ²/df = 9.52; RMSEA = 0.13; NFI = 0.64; IFI = 0.66; CFI = 0.66. The above data show that the common method deviation is not serious and the discriminant validity is good.

4.2. Descriptive Statistics and Correlation Analysis

The mean value (M), standard deviation (SD) and correlation coefficient (r) between variables are shown in Table 2. Social comparison was positively correlated with workplace envy (r = 0.50, p < 0.01), dispositional envy is negatively correlated with self-esteem(r = −0.50, p < 0.01), gender identity (r < −0.18, p < 0.01), dispositional envy is positively correlated with situational workplace envy (r < 0.24, p < 0.01), there was a significant positive correlation between self-esteem and gender identity (r < 0.27, p < 0.01).

4.3. Manipulation Effects Checking

To test the manipulation effect of upward social comparison situational materials, we used Cohen-Charash (2009) and Schaubroeck and Lam (2004) scales to conduct one-way anova on the dimensions of upward social comparison. The results showed that participants in the upward social comparison situational group scored significantly higher on the upward social comparison scale than participants in the non-upward social comparison group, (F_{(1, 527)} = 22.53, p < 0.001; M_{upward social comparison group} = 2.44, SD_{upward social comparison group} = 0.87, N = 265; M_{non-upward social comparison group} = 2.10, SD_{non-upward social comparison group} = 0.81, N = 264), the results show that the manipulation of the adverse social comparison scenario material is effective. (Notes：N = 529, *p < 0.05, **p < 0.01, ***p < 0.001.)

4.4. Hypothesis Testing

Hypothesis 1 predicts that upward social comparison will lead to workplace

Note: *means p < 0.05; **means p < 0.01.

envy. Compared with the group without upward social comparison, workplace envy generated by upward social comparison group is higher. Since workplace envy is affected by dispositional envy, we took trait envy as a control variable and analyzed the main effect by one-way anova. The results showed that the main effect of social comparison on workplace envy was significant (F_(1,527) = 173.71, p < 0.001), the envy level of the upward social comparison group was higher than that of the non-upwardly social comparison group (M_{upward social comparison group} = 4.31, SD_{upward social comparison group} = 0.74; M_{non-upward social comparison group} = 3.27, SD_{non-upward social comparison group} = 1.05). Therefore, hypothesis 1 is supported.

Hypothesis 2 predicted that gender identity moderates the influence of upward social comparison on workplace envy. When employees’ gender identity was higher, upward social comparison leads more workplace envy. First, we put age, gender and dispositional envy into the equation as control variables, and then centralize the independent variables (upward social comparison) and moderating variables (gender identity) into the regression equation, as shown in Table 3 and Figure 2, the interaction between social comparison and gender identity significantly affected workplace envy (β = 0.08, p < 0.05), the stronger the individual’s gender identification, the more workplace envy will be generated.

The simple slope analysis is shown in Figure 2: when employees’ gender identity is high, social comparison will bring more workplace envy. Hypothesis 2 has been supported.

We further tested hypothesis 3, which inferred that employee self-esteem enhanced the effect of social comparison on workplace envy, and put the independent variable upward social comparison and the moderating variable self-esteem centralization into the regression equation, as shown in Table 4 and Figure 3,

the interaction between upward social comparison and self-esteem significantly affected workplace envy (β = 0.07, p < 0.05) the higher the level of individual self-esteem, the more workplace envy will be generated by upward social comparison.

The simple slope analysis shows that social comparisons lead to more workplace envy when employees have higher levels of self-esteem. Hypothesis 3 has been supported.

5. Discussion

This study explored the influencing factors of employees’ workplace envy, discussed the influence of upward social comparison, gender identity and self-esteem on workplace envy, and conducted an empirical study on 529 employees. The results show that: 1) when employees make upward social comparison in their work situations, they will experience workplace envy; 2) when employees have a higher gender identification, they will generate more workplace envy in the face of upward social comparison, which does not show gender difference; 3) Employees with higher levels of self-esteem (verse low self-esteem employees) are more likely to experience workplace envy in the face of upward social comparisons.

5.1. Theoretical Contributions

First, this study enriched relevant studies on influencing factors of social comparison. Existing studies found that upward social comparison would cause employees’ workplace envy (Festinger, 1954; Smith & Kim, 2007) , but we further explored under what circumstances the effect would be stronger. Employees with different gender identification and self-esteem have different workplace jealousies in the face of upward social comparison. Specifically, employees with higher gender identities or self-esteem have more workplace jealousies.

Secondly, the study enriched the antecedent variables of workplace envy. Previous studies on workplace envy have focused on the consequences of workplace envy, such as the fact that workplace envy can lead to social undermining and leaders’ abusive management behaviors (Duffy, Scott, Shaw, Tepper, & Aquino, 2012; Reh, Tröster, & van Quaquebeke, 2018; Yu, Duffy, & Tepper, 2018) . However, there are few empirical studies on the influence mechanism and boundary conditions of the antecedent variables of workplace envy.

Thirdly, this study expands the research on workplace envy in gender identity. This paper extended the study of gender identity, the study of gender identity mainly concentrated on the gender stereotype and social role, etc. in the past, it was once believed that men value work more than women. But the research of this paper found that both men and women, as long as the employees with higher gender identity will produce higher feelings of workplace envy when facing upward social comparison.

5.2. Management Implications

In the process of enterprise management, in order to motivate employees, managers will reward or promote excellent employees to achieve the goal of motivation. However, when rewarding excellent employees, those who are not rewarded will be jealous in the workplace, which will have a negative impact on the organization. This paper finds out the factors that influence workplace envy, and puts forward the following management enlightenment for alleviating the generation of workplace envy:

Firstly, in the process of talent recruitment and employee training, enterprises should focus on the gender identity of employees, mature gender identity scales can be used in recruitment to screen employees who do not have a higher gender identity; try to avoid too many employees with high gender identity when building a team; for on-the-job employees with high gender identity, the enterprise should strengthen the performance-oriented atmosphere through the traction of the corresponding internal management mechanism, and stimulate the positive self-improvement behavior of employees, for example, employees with higher gender identity should be guided to reduce their malicious envy behavior tendency and increase their self-improvement behavior tendency.

Secondly, the organization needs to consider the ways to motivate employees and conduct diversified incentives, not limited to a single assessment standard. While motivating employees, the organization should carefully consider whether it will lead to the malicious envy of other employees. In terms of award design, the organization can adopt more detailed award categories, so that employees with high self-esteem can always find their own strengths in self-related (work) to achieve psychological balance, thus weakening the negative impact of workplace envy.

6. Conclusions and Limitations and Prospects

Firstly, this paper discussed the antecedent variables and regulating mechanism of workplace envy, but lacked the discussion on the outcome variables and mediating variables of workplace envy. What factors will affect employees’ subsequent behavioral tendency after workplace envy occurs? Will there be more covert or explicit aggression between different genders? Whether employees with a high degree of work center will have self-promotion motivation can be considered in the future to carry out research on workplace envy.

Secondly, this paper only adopted the method of situational experiment. Although the independent variables were well manipulated and the operational test results are significant, the ecological validity of the research results may be deficient. Future research can use questionnaires to survey the real work experience of employees to increase the ecological validity of the research and make it more consistent with the management environment.