Psychological Capital , Emotional Labor and Counterproductive Work Behavior of Service Employees : The Moderating Role of Leaders ’ Emotional Intelligence

Based on the perspective of service employees psychological capital, this study adopts questionnaire survey, with every 218 employees of service industry of Jilin province as samples, to examine the Chinese services employees’ psychological capital influence on counterproductive work behavior, and explore the mediation role of emotional labor (surface and deep acting) and the moderating role of leaders’ emotional intelligence. The results show that the service sector employees’ psychological capital is significantly negative correlated with counterproductive work behavior, deep acting plays the mediation role between the two, and that leaders’ emotional intelligence plays the moderating role between deep acting and the counterproductive work behavior. This study provides valuable theoretical bases and feasible solutions for preventing and controlling staff’s counterproductive work behavior caused by psychological capital timely and effectively.


Introduction
With the depth of economic reform, China's economic structure is undergoing great changes and the services sector in the industrial structure is becoming more promoted.According to the National Bureau of Statistics data, the percent of the first half of 2014's GDP was 46.6%, which had already been more than the second industry.Seen as an emerging service industry, it is becoming a powerful force that enhances the country's economic strength.However, with the steady development of the service sector, the problems in the industry are also increasing.Thousands of teachers of Heilongjiang and Henan Province stroke in November and December of 2014, leading to a spate of city schools closed.Ku6.com had conflicts due to collective layoffs in 2011.A doctor of a Nanjing hospital was busy on his private affairs, so that a child who was over 5 months died in 2009.In 2008, China Eastern Airlines 18 flights returning make huge losses and so on [1].In addition to the large media coverage of these hot, employees play games, chat, lie down on the job, provoke the relationship among colleagues, and say bad things about leaders, and other small phenomena in work time are common.These destructive behaviors prevailing in the working environment can be called the "counterproductive work behavior (CWB)".
Counterproductive work behavior (CWB) is that employees violate the provisions of the organization, and threaten to organize and conduct the interests of members of other organizations intentionally and spontaneously, including theft, sabotage, miners, etc. [2].There will be a series of restricting business development issues that not only affect the degree of harmonious labor relations, but also will lead to lower production efficiency, corporate social image damage, etc. [2].For the service industry, its product is service, and its quality is even directly determined by the employee's behavior, so the behavior of employees is particularly important for service purposes.Compared with China, the foreign scholars studied CWB earlier and they had carried out extensive researches for businesses [3]- [5].Our research started late and most researches were for manufacturing enterprises [6] [7].Because CWB paid less attention to service workers, this paper studied CWB of service employees, analyzing influence factors of service employees' CWB by questionnaires and models.

Psychological Capital and Counterproductive Work Behavior
Psychological capital is positive mental state in personal growth and development to encourage individuals to have a positive organizational behavior.It contains four elements of self-efficacy, hope, optimism and resilience [8].The higher the level of active employees "psychological state (i.e., the higher the psychological capital), the stronger regulation of the internal working environment and interpersonal relationships, a positive impact on their own and others" work behavior generated greater; otherwise the impact of the small.For example, when faced with rude customers, employees who have high psychological capital will be more positive and optimistic to take the initiative to explore new initiatives and ways to change negative situations, rather than use CWB to deal with.In other words, employees who have high psychological capital is not easy to produce CWB [9].Despite the different subjects, James (2008,2010), Avey (2008), Steven (2010), Norman (2010), Bovine Kay Cloud (2012) and other scholars were also confirmed this result.Therefore, we propose the first hypothesis about CWB: H1: Psychological capital of service employees has a negative impact on counterproductive work behavior.

The Mediation Role of Emotional Labor
Service companies generally require employees to be enthusiastic and polite in the work process, but employees may be difficult to do this because of various pressures or personal feelings and other factors, these all will lead to emotional labor.Emotional labor refers to "the process in which employees try to adjust their emotional feelings and emotions expressed in order to achieve the stated objectives of the organization".In order to achieve the desired emotional expression by organization, part of the staff only change their outward expression of their emotions and difficult to change their inner true feelings immediately, and this is called surface acting; Another part of the staff try to adjust and change their inner feelings to make their own feelings and external performance consistent, that is deep acting [10].
Employees who have high psychological capital will use positive attitude to face difficulties, to adjust the mood actively, to explore the new plan and way to change the passive situation.On the other hand, for employees who have low psychological capital, it's difficult to adjust their emotions in a short time, so they often just false display of affection and muddle along.It can be inferred that in service industry, the level of psychological capital will influence emotional labor strategy.Employees who have high psychological capital are more likely to make deep acting, on the contrary, employees who have low psychological capital are more likely to make surface acting.At the same time, employees who make surface acting try to express positive emotions and suppress negative emotions.The inconsistency of their emotion and behavior will make the self-authenticity damaged, prone to mental fatigue, emotional exhaustion and other negative psychological.Thus it is easy to produce CWB [11].For employees who make deep acting, their feeling and emotional expression are consistent, so it will not promote CWB.Therefore, we put forward the second and third hypotheses: H2a: Surface acting of service employees plays the mediation role between the psychological capital and counterproductive behavior.
H2b: Deep acting of service employees plays the mediation role between the psychological capital and counterproductive behavior.

The Moderating Role of Leaders' Emotional Intelligence
Emotional intelligence is a kind of ability which people use to effectively manage their emotions and feelings to guide their thoughts and actions [12].Leaders' emotional intelligence in the organization, namely refers to the ability which leaders performance out of the understanding, control their own, evaluation and management of others' (especially subordinate) emotion and psychology and behavior with emotion related in the process of work and social interaction [13].
Specifically, for leadership of the high emotional intelligence, it's easier to understand and control of their own and employees' emotions and behaviors, and be able to communicate and help employees to solve the problem in the process of work.When employees choose surface acting in the working process, degree of their psychological exhaustion will be high.The leadership of the high emotional intelligence will perceive their exhaustion, adjust their mood and relieve their mental fatigue to reduce their CWB.When employees at work choose deep acting, high emotional intelligence leadership will perceive employees' sincerity and dedication, they will praise and care about them timely.Praise and understanding from leadership will also inhibit employees' CWB.Conversely, leaders who have low emotional intelligence cannot feel employees' emotions timely and accurately, so the mediation will be weakened and the influence will be small.Because in service industry, goods and services are the same, the service quality of employees is crucial.High emotional intelligence of leadership is precisely the powerful security services of employees.Therefore, we add leaders' emotional intelligence as the moderator into the model to explore the interactive effects of emotional labor and leaders' emotional intelligence on counterproductive work behavior in service industry and then we propose the forth and fifth hypothesis: H3a: In the high emotional intelligence organization, positive effect on the CWB from surface acting is weak; in the low emotional intelligence organization, positive effect on the CWB from surface acting is strong.
H3b: In the high emotional intelligence organization, negative effect on the CWB from deep acting is strong; in the low emotional intelligence organization, negative effect on the CWB from deep acting is weak.
Above all, we construct a theoretical model of our study, as shown in Figure 1.

Sample
Samples selected for this study were derived from service industry of Jilin Province, including financial, food, communication, basic education and other industries, mainly distributed in Changchun, Baicheng, Siping, Yanbian Korean Autonomous Prefecture.The electronic questionnaire issued by email, Internet chat tools, chat client mobile phone and so on.Participants were asked to complete questionnaires anonymously.We collected information for three times, issued a total of 300 copies of questionnaires, recovered 227 copies of questionnaires.The response rate is 75.67%.Then we removed some invalid questionnaires and recovered 218 copies of valid questionnaires finally.
In our survey, the proportion of men and women generally flat: the proportion of men accounted for 46.8% and women was 53.2%.Age is mainly distributed in the 21 -30 years old (37.2%) and the second distribution in 31 -40 (28.4%) and 41 -50 years old (22.5%).Seen in this sample there are much younger workers.Sample average age of employees work mainly in five years (63.3%).Employees and managers (including grassroots managers, middle managers and senior managers) accounted for 50.4% and 49.6%.In addition, for industry where samples belong to, basic services (including communication and information services) account for 15.1%, production and market services (including finance, intermediary and information, e-commerce and other professional services) account for 28.4%, consumer services (including accommodation, catering, entertainment music, tourism, real estate) account for 31.7%, and the public service industry (including the government's public management, medical, basic education, public information services, etc.) account for 24.8%.

Scale
In order to ensure the reliability and validity of the scale, we choose scales which are mature and referenced repeatedly by scholars.According to the content and purpose of this study, we refer to the expert' opinion to delet and modify some specific questions and then we test the reliability, validity of the revised items.1) Psychological capital scale Psychological capital questionnaire (PCQ-24) we used is written by Luthans (2007), and translated by Li Chaoping (2008).This scale is the widely used in the study of the psychological capital.It has 24 items, using 5 point Likert scoring (1-strongly disagree, 5-strongly agree) [14].Getting higher scores shows that the participant has higher psychological capital.Reliability coefficient (Cronbach's alpha) of the scale is 0.910 (>0.7).Except for 19th item, factor loading value of each item is greater than 0.5.Therefore, after removing the first 19th question, the average variance extracted (AVE) value is 0.542 (>0.5) and composite reliability (CR) value is 0.947 (>0.7).All these indicate that psychological capital scale has good reliability and validity.
2) Emotional labor scale Emotional labor questionnaire we used is written by Grandy (2003), and translated by Lang Peijun (2003).The scale measures two dimensions: surface acting and deep acting.There are a total of 11 questions items [15].On the basis of our research, we revised the scale and got a total of 6 items of the scale, of which 2 items to measure the surface acting, 4 items to measure deep acting.It is 5 point Likert scoring (1-strongly disagree, 5-strongly agree).Getting higher scores shows that the participant is more possibility to do surface acting or deep acting.Reliability coefficients (Cronbach's alpha) of the surface acting and deep acting scale are 0.712 and 0.860 (>0.7) and factor loading value of each item is greater than 0.5.The average variance extracted (AVE) values are 0.604 and 0.510 (>0.5) and composite reliability (CR) values are 0.753 and 0.806 (>0.7).All these indicate that these two scales both have good reliability and validity.
3) CWB scale CWB scale we used is developed by Rees had based on the scale of Robinson and Bennett (2000), which has 12 items [16].According to our situation and on the basis of this study, we deleted 4 items.Using Likert scale 5 point scoring method to measure employees CWB frequency (1-never, 5-always).Higher scores indicate that participants have counterproductive behavior more frequently.Reliability coefficient (Cronbach's alpha) is 0.863 (>0.7) and factor loading value of each item is greater than 0.5.The average variance extracted (AVE) value is 0.553 (>0.5) and composite reliability (CR) value is 0.867 (>0.7).All these indicate that the CWB scale has good reliability and validity.
4) Leaders' emotional intelligence scale Leaders' emotional intelligence scale (WLEIS) we used is developed by Wong and Law(2002), which has 16 items [17].According to our situation and on the basis of this study, we put the "I" to "my leadership", and removed some redundant items.Finally, we get a revised version of the scale which has 14 items.It is 5 point Likert scoring (1-strongly disagree, 5-strongly agree).Getting higher scores shows that the leaders have higher emotional intelligence.Reliability coefficient (Cronbach's alpha) is 0.877 (>0.7) and factor loading value of each item is greater than 0.5.The average variance extracted (AVE) value is 0.6461 (>0.5) and composite reliability (CR) value is 0.9156 (>0.7).All these indicate that the leaders' emotional intelligence scale has good reliability and validity.

Data Analysis and Results
Refer to the method to test the "moderated mediator" model by Wen and Zhang [18], we use SPSS statistics software to analyze data to explore the mediation role of emotional labor (surface and deep acting) and the moderating role of leaders' emotional intelligence.The test results are shown in Table 1 and Table 2.
Table 1 shows: In model 1, we set CWB as dependent variables, control variables (gender, age, etc.), independent variables (psychological capital) and moderator (leaders' emotional intelligence) as the independent variable, and make regression analysis.The coefficient of independent variables (psychological capital) is significant (β = −0.348,p < 0.001), showing that the service employees' psychological capital has a significant negative impact on CWB and it supports the H1.In model 2, we set surface acting as dependent variables, control variables (gender, age etc.), independent variables (psychological capital) and moderator (leaders' emotional intelligence) as the independent variable, and make regression analysis.The coefficient of independent variables (psychological capital) is significant (β = 0.567, p < 0.001), showing that the service employees' psychological capital has a significant positive impact on surface acting.In model 3, we set CWB as dependent variables, control variables (gender, age, etc.), independent variables (psychological capital), moderator (leaders' emotional  intelligence) and mediator(surface acting) as the independent variable, and make regression analysis.The coefficient of mediator (surface acting) is not significant (β = −0.030,p > 0.05), showing that the service employees' psychological capital doesn't have a significant negative impact on surface acting, suggesting that for service workers, effect of surface acting on CWB is not significant.So the mediation role of surface acting between psychological capital and CWB is not significant, against H2a.In model 4, we set CWB as dependent variables, control variables (gender, age etc.), independent variables (psychological capital), moderator (leaders' emotional intelligence) and interaction item (surface acting * leaders' emotional intelligence) as the independent variable, and make regression analysis.The coefficient of interaction item(surface acting*leaders' emotional intelligence) is not significant (β = 0.261, p > 0.05), showing that interaction item cannot have a significant impact on CWB, suggesting that for service workers, effect of the moderating role of leaders' emotional intelligence between surface acting and CWB is not significant which is against H3a.We believe that the reason why the above assumption doesn't hold in this study is that participants selected are mostly young workers who are in the early career.They work more enthusiastic, and are eager to get someone else's appreciation, which will minimize their negative behavior in the work process.Thus, despite doing surface acting, they will not produce more CWB.This is the reason why H2a doesn't establish.For the same reason, younger workers are more attention to their own future career development.Their behavior is more influenced by themselves, so the moderating role of leaders' emotional intelligence is not significant, which is why H3a not established.In addition, we selected samples are distributed in Jilin Province, taking into account that there are different culture and values in the different regions [19], so for the other provinces and cities if the study conclusions are equally applicable should be further verified.
Table 2 shows: In model 5, we set deep acting as dependent variables, control variables (gender, age, etc.), independent variables (psychological capital) and moderator (leaders' emotional intelligence) as the independent variable, and make regression analysis.The coefficient of independent variables (psychological capital) is significant (β = 0.376, p < 0.001), showing that the service employees' psychological capital plays a significant positive impact on deep acting.In model 6, we set CWB as dependent variables, control variables (gender, age.etc.), independent variables (psychological capital), moderator (leaders' emotional intelligence) and mediator (deep acting) as the independent variable, and make regression analysis.The coefficient of mediator(deep acting) is significant (β = −1.038,p < 0.001), showing that the service employees' deep acting has a significant negative impact on CWB, suggesting that for service workers, effect of deep acting on CWB is significant.So the mediation role of deep acting between psychological capital and CWB is significant, supporting H2b.In model 7, we set CWB as dependent variables, control variables (gender, age etc.), independent variables (psychological capital), moderator (leaders' emotional intelligence), mediator (deep acting) and interaction item (deep acting * leaders' emotional intelligence) as the independent variable, and make regression analysis.The coefficient of interaction item (deep acting * leaders' emotional intelligence) is significant (β = 11.842,p < 0.001), showing that interaction item have a significant positive impact on CWB, suggesting that for service workers, effect of the moderating role of leaders' emotional intelligence between deep acting and CWB is significant, supporting H3b.

Conclusions
The purpose of this study is to explore the antecedent variables of counterproductive work behavior in service industry.In particular, we delved into the mediation role of emotional labor (surface and deep acting) between psychological capital and CWB, and the moderating role of leaders' emotional intelligence.From this study, we find: 1) The service employees' psychological capital has a significant negative impact on CWB.Employees with high psychological capital are more difficult to produce CWB; on the contrary, staff who with low psychological capital will be prone to produce CWB.
2) The mediation role of surface acting between psychological capital and CWB is not significant.
3) The mediation role of deep acting between psychological capital and CWB is significant.Employees whose psychological capital is high are more likely to be deep acting and more difficult to produce CWB.Conversely, low psychological capital employees are unlikely to choose deep acting and still be prone to produce CWB.
4) The moderating role of leaders' emotional intelligence between surface acting and CWB is not significant.
5) The moderating role of leaders' emotional intelligence between deep acting and CWB is significant.For the services employees, deep acting produces a stronger negative effect on the CWB in high emotional intelligence leadership organization.Conversely, the effect will be weaker.That is, the higher the leaders' emotional intelligence is, the higher the tendency to do deep acting and suppress CWB in services industry will be.

Management Suggestions
1) For service companies, during the staff recruitment, candidates' psychological level of capital should be focused on and tested.Companies should try selecting candidates who have high degree of psychological capital [20].The level of employees' psychological capital will impact them work behavior.Thus, on the basis of a high degree of personal, job and organization matching, companies should choose one who is more healthy, positive and doesn't afraid of setbacks, on the purpose of reduce the possibility to produce CWB and to disorganize the organization in the future.
2) For service companies, in training process, they can develop additional staff trainings about psychological capital [21].Since the employees' psychological capital not only affect the generation of CWB directly, but also affect CWB by way of deep acting indirectly.Therefore, for employees, the development of psychological capital is essential.Companies could add additional training to develop employees' psychological capital, such as sharing successful experiences, quality development, the establishment of a learning organization, optimistic attribution training, role models and other methods to make them more positive attitude.Then they will work more positively and reduce the possibility to produce CWB.At the same time, they will enhance the quality and productivity of work.For enterprises, it will create more substantial in the future.
3) For service companies, leaders should pay more attention to human care in the work process [22].They should improve the ability to manage employees' emotional, psychologic and behavior.In the high emotional intelligence organization, negative effect on the CWB from deep acting is strong, the higher the leaders' emotional intelligence, the higher the tendency to do deep acting and suppress CWB.Therefore, the leadership in the work process should always understand and care about staff to make them feel the concern from company.More attention and understanding will help them to make their business objectives as their responsibility, to do their own feelings and external performance consistency, thereby to reduce the probability of employees' CWB and to promote the company healthy, effective and sustainable development.

Table 1 .
The mediation role of surface acting and the moderating role of leaders' emotional intelligence.

Table 2 .
The mediation role of deep acting and the moderating role of leaders' emotional intelligence.