An Empirical Assessment of Hotel Departmental Managers Turnover Intentions: The Impact of Organizational Justice

This paper investigates the relationship between organizational justice (OJ) and turnover intentions (TI) among the lower and middle-level hotel departmental managers. Organizational justice, comprising three-dimensional measures of distributive justice (DJ), procedural justice (PJ), and interactional justice (IJ) was measured through inferential statistics. Distributive justice and procedural justice had a significant negative effect on managers’ turnover intentions while interactional justice did not support the proposed relationship. The perceptions of fairness of reward allocation and procedure received in organizations prompt the lower and middle hotel departmental managers to reciprocate their turnover intention behaviour. These research findings offer some insight for the hotel top management into how to prevent their valuable managers from leaving the organization.


Introduction
Most of the academic scholars share a common view on turnover intention (Hellman, 1997;Hom and Griffeth, 1995;Griffeth, Hom & Geatner, 2000;Price, 2001.It generally refers to an individual's perceived probability of leaving an employing organization, or the willingness of an individual to voluntarily permanently withdraw from the organization.Empirical studies have shown that turnover intentions are the best immediate predictor of actual turnover behaviour (Griffeth, Hom & Geatner, 2000;Price, 2001;Hemdi, 2006;Samad, 2006).There is, in fact a strong and significant positive relationship between turnover intentions and the actual turnover (Lambert, Hogan, & Barton, 2001;Mor Barak, Nissly, & Levin, 2001;Price;2001;Samad, 2006).This relationship supports the attitude-behavior theory, which holds that an individual's intention to perform a specific behaviour is the immediate determinant of the actual behaviour (Ajzen & Fishbein, 1980).Mor Barak et al. (2001) and Price, (2001) noted that many studies have used turnover intentions as a proxy in measuring actual turnover.
One of many reasons that cause an individual turnover intention is closely related to organizational justice which denotes the fairness and evaluation of treatment received by an individual in the organization (Cropanzano & Greenberg, 1997).It is, in other words, associated with the ways in which employees determine whether they have been treated fairly in their jobs or other work-related outcomes (Moorman, 1991).Considerable attention has been devoted to the dimensionality of organizational justice, leading to the conclusion that employees distinguish three forms of justice (Niehoff and Moorman, 1993) (1) distributive justice (fairness of the outcomes that the employee receives); (2) procedural justice (fairness of the procedures used to determine those outcome allocation decisions) and(3) interactional justice (fairness of communications or interpersonal treatment that accompanies an organization's formal procedures).
Distributive justice refers to the employees' perception of fairness in the workplace by comparing their inputs/outcomes received from the organization (Niehoof & Moorman, 1993).Greenberg (2001) said distributive justice highlights the employees' response to the fairness of the treatment or distribution of rewards such as pay, promotions, bonuses, job security and layoffs received from the organization.Distributive justice may also be important in predicting personal-level outcomes such as pay satisfaction (Linda & Ping, 1996;Sweeney & McFarlin, 1993), job satisfaction (Samad, 2006;Yusof & Shamsuri, 2006) and to have a significant negative influence on turnover intentions (Aryee & Chay, 2001;Dailey & Kirk, 1992;Lee, 2000;Loi, Ngo, & Foley, 2006).Procedural justice on the other hand reflects the extent to which employees perceive that the outcome allocation decisions were fairly made in relation to the organization's formal procedures and treatment given in enacting those procedures (Moorman, 1991).These procedures not only offer employees control over the outcomes they received but ensure those outcomes are fair.Roch and Shanock (2006) argued that procedural justice consisted of procedures associated with determining promotion, terminations, performance ratings, bonuses or any valued items that organizations make available.In turn, fair procedures can bring benefits to organizations in the form of effective job behaviours and positive work attitudes (Cohen-Charash and Spector, 2001).Past research evidence has revealed that procedural justice may have strong effects on employees' attitudes (Khatri, et al., 2001;Samad, 2006;Tremblay, Sire, & Balkin, 2000;Yusof & Shamsuri, 2006).Tremlay et al (2000) noted that procedural justice would result in stronger attachment to the organization, particularly for those who experience being respected by the organization.Loi et al (2006) proved that there is a strong positive relationship between procedural justice and effective commitment and this is significantly related to turnover intentions (Khatri et al., 2001;Dailey & Kirk, 1992).Interactional justice, on the other hand, refers to the fairness the employees' perceive when they have been communicated with in a sensitive and respectful manner and are treated with courtesy and dignity in their organizational practices (Cohen-Charash & Spector, 2001).Interactional justice has been commonly associated with personal-level outcomes such as supervisory relationships (Cropanzano, Prehar, & Chen, 2002), organizational citizenship behaviour (Niehoof & Moorman, 1993), job satisfaction (Masterson, Lewis, Goldman, & Taylor, 2000), and job performance (Cropanzano et al., 2002).The above explanation clearly indicated that the three dimensions of organizational justice have an influence on employees' work attitudes and behaviours (Colquitt, Conlon, Wesson, Porter, & Ng, 2001;Kernan & Hanges, 2002;Shaw, Wild, & Colquitt, 2003).They act as a predictor of turnover intentions and have received significant attention among the researchers in various organizations including hotels (Aryee & Chay, 2001;Khatri et al., 2001;Loi et al., 2006Carbery et al., 2003;Hinkin & Tracy, 2000;Lam et al., 2001;Pizam & Thornburg, 2000).Despite this, studies of the predictors of turnover intentions within the hotel organization have mainly been conducted in the western world (Carbery et al., 2003;Hinkin & Tracy, 2000;Lam et al., 2001;Pizam & Thornburg, 2000) with very limited analyses carried out in Malaysian hotels.Previous studies merely focused on the operational hotel employees (Hemdi, 2006;Khalid, 2006) (Hemdi, 2006;Hemdi & Nasurdin, 2003) with no attempt made to study operational departmental managers' behavioural intentions.Because of that gap, this study investigates the relationship between organizational justice and turnover intentions among the lower and middle-level hotel managers in the Malaysian hotel setting.In other words, the extent to which the operational departmental hotel managers' perceptions of organizational justice influence their behavioral intentions to leave the organization is investigated.The criterion variable of interest is "turnover intentions" and the predictor variable is the "organizational justice dimensions" (Distributive Justice (DJ), Procedural Justice (PJ) and Interactional Justice (IJ).The theoretical framework is illustrated in Figure 1and this is followed by the main and sub-hypotheses of the study.

H1: There is a negative relationship between perceptions of organizational justice and turnover intentions.
H1a: There is a negative relationship between perceptions of distributive justice and turnover intentions.

H1b: There is a negative relationship between perceptions of procedural justice and turnover intentions.
H1c: There is a negative relationship between perceptions of interactional justice and turnover intentions.

Instrument
For the instrument, the predictor variable of organizational justice (OJ) was measured using a 25 items scale developed by Niehoof and Moorman (1993).The justice scale consists of three dimensions measuring perceptions of distributive justice (DJ) with 5 items, procedural justice (PJ) 6 items and interactional justice (IJ) with 9 items respectively.The criterion variable, turnover intentions (TI) was measured via a 5-item scale adapted from Hom and Griffeth (1995).Responses to all items were made on a 7-point Likert scale ranging from (1) "Strongly Disagree" to (7) "Strongly Agree".In addition, respondents' demographic profiles such as age, gender, marital status, ethnicity, educational level, job position, tenure and turnover experience using nominal scales were also asked.

Samples and Data Collection Process
Participants for this study consisted of lower (supervisor) and middle-level departmental hotel managers (line/outlet managers).Thirty medium-sized hotels (3 &4-stars) located in the Klang Valley, Kuala Lumpur were identified from the Malaysian Accommodation Directory published by the Malaysian Tourism Promotional Board.The Klang Valley of Kuala Lumpur area was chosen due to its having the highest distribution of medium-sized hotels in the country.As the survey was not under the researcher's supervision, the cooperation of the human resource manager from each participating hotel was sought for distribution of the instrument.The human resource managers were briefed to randomly distribute the questionnaires to the managers in two major operating departments; the room department (front office and housekeeping) and the food and beverage department (production and service).A total of 600 questionnaires (in mini booklet form) were sent to the respective hotel human resource managers.Respondents were given two weeks to complete the questionnaire and return it to their human resource department.The completed questionnaires were returned directly to the researcher by the respective human resource managers via mail.In total, 281 (47%) questionnaires were received.After reviewing the collected questionnaires, 27 were found unusable, 14 were answered by accounting, human resources, sales and marketing staff, while 13 came from operational employees.The 254 (42%) usable questionnaires were then coded for analyses.

Respondents' Profiles
The sample profiles have been analyzed and results are tabulated in Table 1.As can be seen from the table, the sample comprised 68.9% (n=175) of male managers compared to 31.1 % (n=79) of female managers.The age range was between 20 to 51 years with a mean age of 32.3 years.52.0% (n=132) were married compared to 42.5% (n=108 who were single, divorced and 0.4% (n=1) widowed.It was found that 50.8% (n = 129) of the managers had obtained a diploma qualification, 21.3% (n =54) had the Malaysian Higher Certificate of Education or the Malaysian Certificate of Education, 16.5% (n= 42) had a Bachelor's Degree and 11.4% (n=11) possessed their certificate qualification.As for organizational tenure, 1.6% (n=4) of managers had less than 6 months working experience in their current hotel, 16.7% (n=17) had worked between 6 months and 1 year, 33.1% (n=84) had worked between 2 and 3 years, while 23.2% (n=59) had worked between 4 and 5 years, and the remaining 35.4% (n=90) had worked more than 5 years in their current hotel.

Confirmatory Factor Analysis of the Study Variables
Prior to hypotheses testing, the assessment of confirmatory factor analysis (CFA), using a covariance matrix (LISREL 8.80) was undertaken.This test is to confirm a pre-specified relationship and to evaluate the distinctiveness of the measures used.As previously mentioned, 25 items were used to tap the respondents' feelings.Five items reflected organizational justice, six items measured procedural justice, nine items related to interactional justice and five items measured turnover intentions.
Table 2 depicts the iteration process that took place in achieving the perfect-fit model of the organizational justice and turnover intentions construct.As can be seen, the initial model of distributive justice had a perfect fit and no additional modification was required (GFI = 1.00,RMSEA = 0.000) while the initial model of procedural justice indicated an acceptable GFI.Nevertheless, the RMSEA was high indicating a possibility of improvement to the measurement model (GFI = .86,RMSEA = 0.225).Jöreskog and Sörbom (2006) suggested that only one item can be altered at a time to avoid excessive model modification and this iterative process continued until the model parameters and key fit indices met the recommended criteria.Iteration 1 was then carried out by dropping item PJ1 since the R 2 value was the lowest (R 2 = 0.19), and the result showed that the RMSEA value did not improve.Item PJ2 was subsequently dropped in iteration 2 owing to the low R 2 value (R 2 = 0.38), and the model showed a significant improvement with a perfect model fit (GFI = 1.00,RMSEA = 0.000).
Similarly, the initial model of interactional justice had the acceptable GFI of 0.74, but a relatively high RMSEA of 0.235.The model however presented a perfect fit (GFI = 0.98, RMSEA = 0.010) after iteration when item IJ4 was dropped owing to the lowest R 2 value (R 2 = 0.30).On the turnover intentions construct, the initial model had a poor fit with an acceptable GFI of 0.94 and a considerably high RMSEA of 0.159.Item TI1 was then dropped in iteration 1 owing to lowest R 2 value (R 2 = 0.33) and the result finally showed a perfect model fit (GFI = 1.00,RMSEA = 0.000).
Consequently, 3 items (PJ1, PJ2, and IJ4) were eliminated from the organizational justice constructs and 1 item (TI1) from the turnover intentions construct, and the final constructs comprised 21 items, 17 items measuring the respondents' perceptions of organizational justice and 4 items measuring turnover intentions.This can clearly be seen in Table 3.

Reliability and Validity Analysis
After the evaluation of the model, the measurements of each construct were assessed for their reliability and validity.The reliability coefficient for each measure was examined by computing the Cronbach's alpha values.The items included in the test are based on the results of confirmatory factor analysis.As shown in Table 4, the reliability of the organizational justice dimensions of distributive justice, procedural justice and interactional justice were .89,.83,and .93respectively.The dependent variable of turnover intentions had an excellent reliability coefficient of .91.With that, the internal consistency of all study variables was considered high and acceptable.
The validity of the instruments, particularly construct validity was assessed.Unidimensionality analysis was carried out by identifying the measurement model for each construct.A comparative fit index (CFI) value for the study variables exceeded .90 for the model which indicated strong evidence for unidimensionality.In addition, convergent validity using normed fit index (NFI) was also assessed.All scales had an NFI value ranging from .90 to 1.00 which implies a reasonably strong indication of convergent validity.The CFI and NFI values for all scales are summarized in Table 4.

Descriptive Statistics
The descriptive statistics and the correlation among the study variables (predictor and criterion variable) are shown in Table 5.
From the results, it can be said that no serious multicollinearity exists among the study variables as the strength of the correlations was all below .90.

Hypothesis Testing
A two-step hierarchical regression was conducted to assess the relationship between three dimensional measures of organizational justice and turnover intentions.The five demographic variables (age, gender, marital status, education level and organizational tenure) were statistically controlled and entered into the first step of the regression equation.In the second step, all three model variables of organizational justice were entered.Table 6 summarized the results.
As can be observed from Table 7, the control variables managed to explain 7.0 % of the variance in turnover intentions (R 2 = .07,F-Change = 3.49, p < 0.01).Of the five control variables, only respondents' gender ( = .19,p < 0.01) and organizational tenure ( = .21,p < 0.01) were significantly related to turnover intentions.Adding the three model variables of organizational justice, the R 2 increased to .45.This indicated that organizational justice was able to explain an additional 38% (R 2 Change = .38,p < 0.01) of the observed variations in turnover intentions and the effects of demographic variables.Of the three organizational justice dimensions, only two dimensions namely distributive justice and procedural justice significantly and negatively predicted turnover intentions.Indeed, distributive justice had the most impact on the prediction of turnover intentions ( =-.56, p < 0.01), followed by procedural justice ( =-.26, p < 0.01).In other words, distributive justice was found to be more influential in explaining the variation in turnover intentions.Conversely, the interactional justice dimension (H1c) had no significant relationship with turnover intentions.In sum, these findings only support H1a and H1b.Thus, the hypothesis is partially supported.

Discussion
As hypothesized, distributive justice has a significant negative influence on turnover intentions.This result is consistent with the findings of Aryee & Chay (2001) and Lee (2000).What could be said from this result is that if the managers feel contented with what they receive compared to those of a referent other, they are more likely to be productive and remain with the organization.In fact, managers who perceived fair prospects of different work outcomes, particularly relating to pay satisfaction, work schedule, workload, rewards, and job responsibilities may view their current organization favourably and might not see outside alternatives as attractive.Therefore they are less likely to have the intention to leave their current organization.Similarly, as hypothesized procedural justice appeared to significantly and negatively predict turnover intentions.This finding is congruent with that of Khatri et al. (2001).This indicates that high perceptions of fairness of the formal decision-making procedures would influence the managers to stay in the organization.In other words, managers are less likely to experience turnover intention if they received, or were provided with fair and consistent organization policies and practices related to procedure, such as determining promotion, termination, performance ratings, bonuses or any values the organization makes available.In fact, many commentators argued that fair procedures can bring benefits to the organization in the form of effective job behaviours and positive work attitudes.
Besides the above findings, interactional justice was found to have an insignificant relationship with turnover intentions.This was contrary to the recent work of Hubbell and Chory-Assad (2005), who found that if managers' perceived fairness and quality of communication as being received in their organization, they were less likely to leave the organization.One possible explanation for this finding is that managers in this study sample may have perceived the interpersonal treatment that they received to be unfair (e.g. they were treated with dishonesty and disrespect) and, in turn, they progressively seek more options of better job positions through the external labour market.In other words, as the external labour market becomes more attractive, managers tend to engage more in job-search behaviours, thus increasing intentions to leave the organization.In other words, there may be a lack of interpersonal relationships between the lower and middle managers and the top management.Departmental hotel managers in this study may have less interaction with the top management, since according to their job descriptions they are performing more operational tasks rather than management decisions (Dittmer, 2002).In addition, the "culture" or the "nature" of human beings to continually seeking job promotion and job security could be one of the reasons to leave the organization as reported by Iverson and Deery (1997).Despite this, the overall findings of this study support most of the previous similar studies, although undertaken in different settings or industries.

Conclusion
Based on the aforementioned discussion, it clearly proven that the perception of fairness of reward allocations and procedures received in the organization prompts the lower and middle departmental hotel managers to reciprocate with their behavioural intentions.Thus, it is logical to predict that when managers perceive that they are treated fairly in terms of outcome and procedures; they tend to stay in their current organization and do not have the intention to seek a fairer alternative.Conversely, if they perceived the interpersonal treatment received to be unfair the intention of leaving or seeking an alternative is higher.The relevant hotel authorities concerned with reducing turnover intentions among their lower and middle departmental managers therefore should realize and be consistent, and focus their attention on providing to at least a decent amount of fairness in terms of reward allocation, formal procedures, and interpersonal treatment.By considering these, the likelihood of quitting or leaving among the lower and middle managers can slowly be reduced.Some recommendations for future research could be made from this study.First, the use of a larger sample within the same industry, or from other industries would improve the generalizability of the findings.Second, given that there may be other organizational factors (e.g.psychological contract violation, organizational politics etc.) that influence managers' turnover intentions, these factors therefore could also be used to predict the hotel managers' turnover intentions.Finally, given the better ability to prove a cause-effect relationship between the predictor variables and the criterion variables, the experimental or longitudinal approach could be used instead of the cross sectional approach.Sweeney, P. D., & McFarlin, D. B. (1993).Workers evaluations of the 'ends' and the 'means': an examination of four models of distributive and procedural justice.Organizational Behavior and Human Decision Processes, 55 (3), pp. 23-40. Tremblay, M., Sire, B., & Balkin, D. B. (2000).The role of organizational justice in pay and employee benefit satisfaction and its effect on work attitudes.Group and Organization Management, 25 (3), pp. 269-290. Yusoff, A. A., & Shamsuri, N. A. (2006).Organizational justice as a determinant of job satisfaction and organizational commitment.Malaysian Management Review, 41 (1), pp.12-21 Zikmund, W. G. (2003).Business Research Methods (7 th Ed.).Ohio: South-West as soon as possible Actively look for a new job next year Often think of quitting current job Leave as soon as can find a better job Figure 1.The Theoretical Framework of Hotel Managers' Turnover Intentions

Table 2 .
Description of the Modification Process of Unidimensionality and Convergent Validity of the Organizational Justice and Turnover Intentions Factors

Table 6 .
Results of Hierarchical Regression of Perceptions of Organizational Justice on Turnover Intentions