Examining Determinants of Absenteeism among Blue Collar Employees in Petroleum Sector of Pakistan

The study focuses on the relationship between job satisfaction, job performance, and social / family compulsion and absenteeism in blue collar employees working in petroleum sector of Pakistan. 150 blue collar employees from refineries and oil marketing companies were included in the survey and responded the questionnaire, encompassing all variables. Results prove hypothesis 1 whereas, hypothesis 2 could not be proven. Results will help Pakistani organization to reduce absence rate in their blue collar employees by understanding the social and family compulsions.


Introduction
Absenteeism is a routine issue confronted by managers on daily basis. Causes of absenteeism are usually an area of interest to the supervisors and managers. There are many causes of absenteeism, each distinctive in nature and effect. Most of the research on absenteeism has been carried in the West and United States, effects envisaged by these researchers are to be reviewed in geographical, demographic and social context of Pakistan. An endeavor is made to explore reasons and causes of this counter-productive attitude in corporate sector in particular. HR scholars have focused the issue of absenteeism with varied perspectives keeping the management concerns in view, to suggest the workable solutions to improve productivity and cost cutting initiatives through controlling and managing this tendency to the level where its effects are negligible. There are different perspectives explored and researched by past and present scholars and practitioners on the core issue of absenteeism.
Frick & Malo (2008) found no effect of job protection on absence rate. Wolpin, Burke, Krausz & Freinbach (1979) defined that correlation of absenteeism and lateness prove that both may not be considered under same behaviors. Fins, Travaglione & Neill (2006) have explained absenteeism in rapid organizational changes and suggested that entitlement mentality employees resort to increase in certified leave just before their final departure whereas those who intend to leave an organization, had a high absenteeism record in the past. A significant negative relation between individual attitude towards job and intend of absence is explained by Remsey, Punnett & Greenidge (2008). Sanders & Nauta (2004) found negative relation in short term absenteeism and work ethics of employees and informal relationship with in a team. Sander & Hoekstra (1991) defined that closer the relation with in a group / department, the lesser the absence rate will be. Social variables are found to be strongly related to attitudinal measures and job performance by Steel, Shane & Kennedy (1990).
Absenteeism and punctuality are highly correlated with overall job performance (Roszkowski, Garrow, Spreat, Delavey & Davis, 2005). Engstrom & Holmlund (2007) analyzed the effects of higher sick pay on sick absence and found it to be positively related to each other. Ones, Viswesvaran & Schmid (2003) emphasized identifying absenteeism traits in a personality as part of selection strategy to cut down post-employment absenteeism. As relates to the self-reporting in absenteeism and lateness, Koslowsky & Berkovits (2001) suggested that absence is a relatively stable behavior whereas, lateness is a function of personal indicators. In this paper, relation of job performance, job satisfaction and family / social compulsions are studied as causes of absenteeism in blue collar employees in corporate sector of Pakistan with a focus on refining industry and oil marketing companies.

Absenteeism
Absenteeism is traced back into school days with future prediction on future employee (Lounsburg, Steel, Loveland & Gibson, 2004). Sickness is a cause of absence (Gafafer, 1948) and Pocock (1975) defines 'Blue Monday' where inception of absence period is observed higher on 'Monday'. Popp & Belohlow (1982) discussed that 'frequency of absence' can better be predicted than 'total days lost'. Lambert (2001) highlights absenteeism in correctional organizations which, he found, has more effect than commercial organizations whereas, Allen (1983) explained effects of absenteeism relatively small on both industrial production and employee wages.
Application of Monte Carlo analysis shows longer absence in large organizations (Winkelmann, 1996). Perceived unfairness has contribution to explain an employee's absence in future (Deboer, Ballker, Syrott & Schanfeli, 2002). Absenteeism, prediction of absence, long and short term absence speculation of temporal limits constraining employee absence prediction is somehow possible (Steel & Rentseh, 1995). As regard to long term absence rate, it is not found under the influence of informal network as the short term absence rate. Moreover, rate of mid-term and long term employee absence is considered to be less dependent on social environment of their respective department when analyzed with short term absence (Sanders & Hoekstra, 1998). Sick absence of day and shift employees of refinery differs considerably. Shift employees are recorded to have lower sickness rate than their counterparts in day (where annual average leaves 11 days for shift and 18 days for day worker). Three main groups are defined as frequency absence, long absence and no absence, major difference in groups is about attitude towards them and works (Taylor, 1968). Frequency of absence is more reliable than total day absence as measures absenteeism because past absenteeism helps predict future absenteeism of an employee (Ivaneevich, 1985). Brooke (1986) proposed causal model of absenteeism as modified form of Steers and Rhodes process model of employee attendance and found to have problem in operationalizing.
Absenteeism can be reduced if higher wage is offered. Work of Roza (1975) concluded that wage elasticity of demand is equal to or lesser than the wage elasticity of supply. A later endeavor found that efficiency wage is a useful, voluntary measure to control absence and check shirkers (employees who tend to absent by unacceptable sickness). Firms, operating assembly lines, have higher cost of absenteeism thus tend to pay their employees more than sick pay to arrest absence tendency (Bariuby, Sessions & Treble, 1994). Dougtas (1919) identified various causes of absenteeism and then recommended a few controlling measures like imposing fine for tardiness and not showing on work, may be equivalent to the time lost or even more than that to arrest this tendency. He advocates distraction between justifiable and unjustifiable lost. Gardiner (1992) in order to have control over abuse of sick leave, has suggested like a few initiatives are suggested. Inappropriate utilization balance sick leave by the employees can be identified by evaluation of the supervisors. Gauine & Sack (2001) proposed that feedback will not affect where absence is institutionalized as a remedial measure to manage physical and mental stress. Intervention by supervisions can help decrease tendency of absenteeism with above average and considered low at cost and the saving. A corporate initiative to control absenteeism suggest that profit sharing reduces absence by 7%, share ownership by 14% but profits and ownership sharing plan reduces it to 11% only. Further it testifies that negative relation exists between absences and wages (Brown, Frakhfakh & Sessious, 1999). Stowell (1931) suggest that medical examination before engagement and on re-joining after illness help identifying minor defects for which remedial measures can be suggests to avoid serious illness to employees and wasteful working hours to employer. A study of local government recorded loss of 8.7 days for each employee at an average, roots of which are found in managerial and organizational practices, hierarchy and autonomy (Bennett, 2002). Manager must be a role model for subordinate and co-workers, much can be done to reduce occurrence of employee deviant behavior (Everton, Jolton, & Mastranfolo, 2005). Another measure in absenteeism feedback interventions (AFI) which is found successful and useful in reducing both 'absent frequency' and 'total absence period" as high as above average but not in case where extreme absenteeism is found. AFI enhances the motivation to attend but fails to account for ability to attend the work.
It is suggested that in order to save, an organization has to spend for devising methods to control and check absenteeism just like any break down in plant equipment is corrected by paying a cost (Navarro & Bass, 2006). Poor organizational commitment and absenteeism have no significant relation (Gaurts, Schaufoli & Rutte, 1999). Higher the position in management, the lesser the rate of absenteeism (Rosenblatt & Shirom, 2006). Ehrenberg (1970) concluded that above zero absenteeism, stochastic nature of absenteeism is not always and every time found responsible for increasing the over time for each individual employee. Noisy work environment increases absenteeism and wages have no clear effect on absence showing more factors therein. An inverse relation exists between distance from work place and regularity of an employee to show on work (Knox, 1961).

2.2
Job satisfaction Goldberg and Waldman (2000) found no relation between job satisfaction and absenteeism whereas, Lau, Tang and Ho (2003) suggest that lesser the level of employee satisfaction (where his / her job), more the chances to get into counter-productive behavior (CPB) like absence, breakdown in work place and low productivity. When job boredom was analyzed by Vass, Vodanovich and Callender (2001), a significant relation was observed with job satisfaction, tenure and high absenteeism. Tharenon (1993) established a reciprocal relation in job satisfaction and absenteeism in line with Sagie (1998) who highlighted the possibility of more attendance by employees with high level of job satisfaction. Dagfinn (1962) extracted that high absence is expected as a result of low satisfaction with immediate working conditions even though highly satisfied with organization. Meta-analysis by Scott & Taylor (1985) bring out negative relationship in absenteeism (absence frequency) and job satisfaction. Watson.J, Driver & Watson.D (1985) raise questions on research methodology in statistical conclusion and validity H1: There is a negative relation between job satisfaction and absenteeism.

Job Performance
Job performance and attitude / behavior have been explored in multifarious dimensions. Sagie (1998) defines indirect indices of absences behavior (frequency and time lost) are weakly correlated to job attitude. Perceived job scope as significant relation to absenteeism (Mowday & Spencer, 1981) Melbin (1961) explained more the score of change (reassignment). Absence, more the chance of employee turnover is expected. Halpern, Shikiar, Rentz & Khan (2001) analyzed the positive effects on efficiency, quality and amount of work in case of ex-smokers.

H2:
There is a negative correlation between job performance and absenteeism.

Social and Family Compulsions
Interesting results are found in the study of absenteeism and lateness in the context of social and family compulsions and related factors. Jansen & Meintosh (2002) (2003) determines social influences as nonresponsible for notable differences in absence rate. Hendrix, Spencer & Gibsin (1994) stress upon the fact that absenteeism has to direct effect on job and life stress but an indirect effect exist, moreover employee believe that his / her absence / lateness to work place rates behaviors perceived strength of social expectations from them (Harrison & Price 2003).
Literature has elaborated in detail about group and absenteeism phenomenon in an organizational perspective and working conditions. There is no relation between group size and their absenteeism rate however, size and absenteeism can be related when within group perspective is applied (Mark ham & Alutto, 1982). Mason & Griffin (2003) elaborate that increased positive affective tone (more energetic and enthusiastic) shows a decrease in absence and a decrease in affective tone results into increase in absenteeism. An employee behavior in an organization has deep roots into his relations with referent peers. Thus forcing us to analyze social comparison beyond the formal organizational structure (Bamberfer & Bison, 2007). Blau (1995) emphasizes that there are three key referents (reference groups) affecting lateness: (1) the group, (2) the internal, and (3) the external (family and friends).

Theoretical Framework
Geographical relation of selected variables is explained in theoretical framework which is extracted from literature review. It is expression of relationship of selected variables. Absenteeism is considered to be a dependent variable whereas, three independent variables: (1) job satisfaction, (2) job performance, and (3) social and family compulsion. It will help us to understand the dynamics of these variables and testing of hypothesis to enhance and improve our understanding of the subject.

Research Methodology
Research objectives and research questions shape up the research methodology. Ary, Jacobs and Razavieh (2002), explained research approach into two broader categories: (1) the qualitative (to understand the social phenomenon from human perspective), and (2) the quantitative (makes use of objective measurements as well as statistical analysis of selected primary or secondary data) to understand a phenomenon. A mix of both the approaches is called hybrid or mixed approach. The research questions posed in this study are such that merit following quantitative methodology.

Instrument Used
The data was collected using 'questionnaire' that contains 36 items. Five point Likert scale is used with 1 representing 'strongly disagreed' and 5 representing 'strongly agreed'. Survey questionnaires are distributed to the respondents and explained with the purpose of study.

Sample
Population for this study includes blue collar employees (subordinates and supervisors) from petroleum industry of Pakistan including refineries and OMCs (oil marketing companies). The data was collected from 150 respondents for the study, by using random sampling. A total of 200 questionnaires were distributed and 150 were received back, making response rate 75%. The sample characteristic described in the table below.

Results
Results obtained from the survey are noted below in sub section 5.1 and 5.2 whereas, detailed discussion on the result in para 6 below.

Discussion
Above correlation and regression matrix indicates a negative relation between job satisfaction and absenteeism whereas, a strong negative relationship exists between job performance and absenteeism. Job satisfaction and job performance both are related to absenteeism however, job satisfaction is less related to absenteeism when compared with job performance.
Results conform the hypothesis 1 that negative relation exists between job satisfaction and absenteeism. Our results are in line with the research outcomes of Lau, et al., (2003) who suggest negative relation, Scott & Taylor (1985) who explained a negative relationship in absence frequency and job satisfaction. However, our result is not in accordance with the results of Kass, et al., (2001) who found a significant relationship in job satisfaction and absenteeism, Goldberg & Waldman (2000) who explained no relation in job satisfaction and absenteeism, Dagfinn (1962) who found high absence as a result of low satisfaction with immediate working conditions though highly satisfied with organization.
Our results indicate negative relationship in geographical context and existing social environment of Pakistan. A highly satisfied employee is expected to be very low in absence but social factors like strong family relationship not only to immediate relatives but also to the farther relatives may force him / her to absent. On the other hand such compulsion are both expected and accepted by the management in this geographical context being common to all. As regards to 'no relation' between job satisfaction and absenteeism, socio-economic factors are in place where switching jobs is not convenient due to lack of jobs and specially in sectors where specialization is more dominating, like in our case petroleum sector, that limits the options of employees thereby, a satisfied employee has no reason not to show up at work without sufficient cause being under heavy burden of not only immediate family i.e. wife and children or parents but also father family likes uncles, aunts and grandparents, etc. Though immediate working conditions are more related to absence than job satisfaction itself in other parts of the world but in underdeveloped world, seems to have more related to absence than immediate working conditions, may be due to economic conditions, socially related society and lack of health safety environment (HSE) practices.
Result proves strong negative correlation of job performance and absenteeism which replicate the results of Nowday & Spencer (1981) who conclude direct relation in job performance, perception and absenteeism, Keller (1984) who explained low performance and high absence as most appropriate predictors of turnover behavior. However, results are not as per the findings of other researchers like; Sagie (1998) who defines weak correlation in job performance and absence, Melbin (1961) who pronounced re-assignment as cause of absence being lesser in performance.
Probable causes of variation can be traced in geographical context as well as industrial base in Pakistan and specific skills to a certain industry. As regards to weak relation between job performance and absence, it can be explained that in our geographical context it is different from West and United States, may be due to lack of jobs in the job market an employee with good performance tends to show up more regularly at workplace to avoid being fired on more absence or absence due to insufficient reason. Moreover, those who get into a job and are performing well are more enthusiastic to be regular at work.
As regard to the relation of re-assignment and absence, it may not be the case here in this part of the world due more importance of being on job than being re-assigned for a number of reasons like: (1) economic situation, (2) social responsibilities, and (3) shrinking job market, however, if these reasons are not accounted, the result would have been same here in this part of the world.
Another element here is the size of industry and specialization. Mainly our study is related to petroleum sector (refineries and OMCs) that demands highly skilled and technical workforce. Highly skilled workers have limited choice due to two reasons: (1) high salary level, and (2) limited size of the industry.
Regression Matrix highlights the individual and combined impact of independent variables i.e. job satisfaction and job performance on dependent variable i.e. absenteeism. Job satisfaction has significant impact on absenteeism whereas, job performance also has significant impact on absenteeism but impact of job performance has more significance in our case as compare to job satisfaction. Value of R-square notifies the combine impact of independent variable which is job satisfaction and job performance, on the dependent variable that is absenteeism. Our regression analysis show the value of R-Square is 0.083 which is very low then the desired value (0.50). Probable cause of lesser value can be the number of variables affecting absenteeism are more whereas, variable considered in this study are very less. Other variables are work environment, attitude of management, relationship with supervisor and co-workers, pay and allowances, immediate job conditions, work complexity, distance form workplace, and social compulsion, etc.

Conclusion
Absenteeism and lateness are among the problems confronted by the managers and supervisors on day to day basis and need to be viewed in more details to help managerial issues. The study is conducted in Pakistani context that focuses on petroleum sector of Pakistan and highlights slight variations from the other studies. Thereby meaning that the studies carried out in the West and the United States can not be applicable in Pakistani context without testing their validation in the local context due to our culture, family system, well knitted society, and acceptance of certain deviations as relaxation. Future research may concentrate on the effects of different variables on absenteeism in geographical, demographical and social context of Pakistan to validate the applicability of the research.