A Multivariate Analysis of Job Satisfaction of Ready-Made Garments (RMG) Workers in Bangladesh

There is a paramount importance of Job Satisfaction of the employees of an organization for the smooth functioning of its work. It has substantial impact on workers’ performance. RMG workers of Bangladesh are enormous in number and instrumental to the economic growth of Bangladesh. But the salary and other benefits of these workers are very low compared to that of other industrial sectors. Sometimes, the salary and benefits are not even paid on time to the workers. As a result, a serious concern has been raised by the national and international stakeholders about the job satisfaction of RMG workers. Therefore, this study aims at identifying the job satisfaction factors of the RMG workers and providing suggestions for the improvement of present situation. For analyzing data multivariate analysis techniques are used by using Smart PLS, and SPSS. Salary and benefits, supervisor’s behavior, work and family life, working condition and the working environment are identified as factors having influence on job satisfaction of RMG workers in Bangladesh. Results show that the salary and benefits, supervisor’s behavior, work and family life are significantly related to the job satisfaction of the workers. This study suggests that for ensuring job satisfaction of the RMG workers in Bangladesh the policy makers should focus more on salary and benefits, supervisors’ behavior, and work and family life of the workers.


Introduction
Ready-made Garment (RMG) industry is the biggest foreign exchange earner of Bangladesh with US$ 25.49 billion of exports (BGMEA, 2015). RMG started its journey in 1950 to export their products in the countries including western countries. In 1974, to control the imported RMG products from different developing and developed countries, an agreement called Multi Fiber Agreement (MFA) was made. As a result, there was an increase of 6% growth in every year from a developing country to a developed country. Bangladesh, as a developing country, started receiving investment in the early 1980s in this sector. Some of the Bangladeshis also got an opportunity to receive free training from different Korean Companies including Daewoo. At that period, there were only 50 RMG factories that employed only a few thousand workers in this sector. At present, there are about 4,500 manufacturing units operating in this sector. Currently, different countries comprising USA, UK, France, Germany and other European Union countries are importing RMG products from Bangladesh. About four million workers are working in this industry, of which 85% are rural illiterate women (Huffington Post UK, November 27, 2015). This sector accounts for more than 80% of the overall export earnings of the country and its contribution to the country's GDP is around 10% (BGMEA, 2015). As such, RMG industry is playing a very important role in creating employment opportunities of millions of rural people of Bangladesh. In spite of the epic growth of the RMG industry, and its bright prospects, this industry faces many challenges. Some of the challenges faced by this industry are to ensure workplace safety, working conditions and job satisfaction of the garment workers. Experts opined that the most important challenge in this sector is job satisfaction of the workers. Hence, this study aims at identifying the job satisfaction of RMG workers in Bangladesh.

Theoretical Framework
Job Satisfaction of workers helps to determine the sense of achievement and success of workers and it is directly related to their productivity and personal well-being (Miller & Ross, 2002). Job Satisfaction also shows the worker's interest and liking towards their work that applies to RMG sector of a Bangladesh (Spector, 1997).
It can also be defined as the workers overall effectiveness of mind attaining from the consent of all areas of their work (Hossan, et. al. 2012). It may be the evaluation between the workers' wants and their real gain. Contentment of the employees can be defensible as worker's all assessment of work as favorable or unfavorable (Locke, 1976). Employees' satisfaction is an expression which is sometimes used to explain the attitude of a worker towards the job and roles and responsibilities. Satisfied worker shows a positive attitude towards the work, on the other hand dissatisfied worker does not (Pushpakumari 2008). It is often related with employees' needs and worker's conditions too. Numerous classes of needs are concerned to the job satisfaction of the workers including physical, social and egoistic needs (Dubrin 2007). Satisfaction phase of workers toward their work differs with specific dimension of the work. It sometimes relies mostly on payment of their job (Shabnam and Sarker, 2012). Satisfaction towards their job can also be taken care of through the maternity benefit with full payment and healthcare facilities of the workers.
Bangladeshi people have a very low per capita income since it is a least developed country. Hence, salary can work as an important factor to improve the workers satisfaction and their productivity. But it is seen that when the workers are very dissatisfied with their salary, they are not satisfied with other benefit packages they have had (Zohir, 2007). Satisfaction towards job also relies on sufficient provident fund and encouraging financial benefits (Taylor, 2008). Garments Workers of garment sector in Bangladesh are satisfied with their overtime wages and other privileges but not with regular wages and promotion policy (Sadrul H., et. al. (2011). So, low compensation, irregularities in payment and low wage are the factors of the dissatisfaction of RMG workers in Bangladesh (Islam, M. S., Faruk, M. O., Khatun, R., and Esfaqur, M., 2014). Without their regular and full payment of salary or timely payment, workers' often become worried and get anxious regarding their future that ultimately results in low effectiveness and high dissatisfaction (ILO, 2005;Morshed, 2007).
It is found that financial factors like salary, bonus, incentives' and other financial benefits, like heath care, sick leaves, etc. can improve the workers satisfaction in RMG sector but other nonfinancial factors like work description, work orientation, unbiased evaluation of work performance, training, job security, efficient and supportive line manager, good working environment etc. can facilitate it further (Sarker, A. R., & Afroze, R., 2014). To balance work and family life, employers should concentrate on friends and family related benefits like annual picnic, transportation facilities for workers, daycare facilities, subsidized meal, and so on.
The male and female workers of garment industry in Bangladesh work about 12 hours a day and there is absence of leave facilities in Bangladesh. Although, the workers were paid for overtime on weekly holiday, no alternative holidays were given to the workers (Paul-Majumder and Begum 2000). Sometimes, they are found to work whole month at a stretch neglecting a single day off, in contravention of Factory Act 1965 which is applicable in Bangladesh, mentioning that no employee should work more than ten days consecutively without a break (Jamaly and Wickramanayake 1996;Ahmad 1996;and Hossain, et al. 1990) Working condition has impact on job satisfaction of the RMG workers. It includes the space provided for workers, medical facilities, lighting, toilet facilities, and safety arrangements, such as, first aid boxes, etc. The underperformed working conditions have been aggravated by the building collapse like Rana Plaza accidents and outbreaks of fire at the Tazreen Garment in December 2012 (Miller 2012). Working conditions can influence job satisfaction since workers consider the comfortable physical working environment which results in more positive stage of job satisfaction of the workers (Robbins 2001).
Working hour is another factor concerned with job satisfaction of the workers of RMG sector in Bangladesh. Kumar (2006) focused on the facts that workers of garments sector are not satisfied with lengthy working hours, frequent consecutive shifting duty, individual risky working environment, very poor working conditions, low wage and discrimination at workplace. Nancy and Katherine (2002) identified that healthcare and disease, education, attachment with labor unions are important for the workers, Islam and Zahid (2012) found that the social status, overtime, savings, working hours, working rights and fringe benefits are the influencing factors for the job satisfaction of female garment workers in Bangladesh. Tiotangco and Nunag (2012) identified that food, water and sanitation, healthcare and disease, transportation, housing and accommodation and hygiene are the important factors for the workers for their socio-economic lives that are connected with the job satisfactions of the workers.
The quality of supervision has impact on the satisfaction of job of the workers in garment industry of Bangladesh. Supervision is another factor having a correlation between job and the satisfaction gained from it (Wright, et.al. 2003) There are many factors which have influence on the satisfaction of the workers among them five factors comprising supervision, relationship with coworkers, present pay, nature of work and opportunity for promotion that are directly related with job satisfaction (Smith, Kendall and Hulin, 1969). Sometimes workers level of satisfaction might be enhanced by ensuring the participations of workers in the decision making process (Gürbüz 2009). Some negative factors like fear, discouragement, dissatisfaction, resentment and turnover among employees create negative feelings among the workers (Elangovan and Xie, 2000). When the supervisor guides the workers in a decent way, they are pleased, on the other hand, inadequate supervision may lead to the dissatisfaction of the workers, resulting high turnover (Marchington and Wilkinson, 2008).
Work-family balance is very important where the workers are equally engaged in and satisfied with his or her role regarding family and work (Greenhaus and colleagues 2003). Workers quality of work-life is dependent on different bonus packages including festival bonus, attendance bonus and wage increment of a worker (Zohir, 2007). Conflict between work and family act as a negative factor in case of satisfaction (Kossek and Ozeki 1998) Expression of rights is another factor of job satisfaction of Bangladeshi RMG workers. Labor rights are usually overlooked in RMG industry of Bangladesh. Common problems and inadequate practices in this sector include the absence of trade unions, excessive work, informal recruitment, irregular payment, sudden termination, wage discrimination, and abusing child labor (Ahamed F., 2012).
Nature of work and overall comfort is very important for job satisfaction of RMG workers. If workers' find the work interesting, promotion process is smooth, his or her supervisor is cooperative, and co-workers are helpful, then a situational approach leads to predict workers satisfaction at work (Taylor, 2008). Improvement in RMG sector is not a single man task. Therefore, all the stakeholders including government, international community should work together for improving working conditions in the RMG sector in Bangladesh by establishing a HRM unit or Personnel Management unit in each RMG company (Ahamed, F., 2014). The factors related to job satisfaction of the ready-made garment workers identified by the scholars and researchers at different times are summed up in Figure 1.

Objectives of the Study
The main objective of this study is to identify the factors of job satisfaction of RMG workers in Bangladesh. The specific objectives are: (i) to identify the significant factors concerned with job satisfaction of RMG workers of Bangladesh (ii) to analyze the factors that affect satisfaction of garments workers (iii) to identify the relationships of the factors with the overall satisfaction of the RMG workers, and (iv) to give some suggestions for the improvement of the satisfaction level of the workers in Bangladesh.

Hypothesis of the Study
To fulfill the objectives of this study, the following hypotheses have been constructed comprising the five important factors of job satisfaction of RMG workers in Bangladesh including salary and benefits, supervisor's behavior, work and family life, working condition and working environment. H 05: Working Environment has no significant impact on the overall Job Satisfaction of RMG workers.

Methodology
This is an empirical study for the identification of the job satisfaction factors of RMG workers of Bangladesh. For data collection structured questionnaire with 5-point scale was used. Convenience sampling method was used for the selection. After data collection, unusual data like incomplete and biased or abnormally answered data were avoided thorough scrutinizing process. By using SPSS software, the reliability of 37 items used in the questionnaire has been tested and the Alpha Coefficient was identified as 0.746 which is at the acceptable limit as per Nunnally (1967 and1970). To analyze data both descriptive and inferential statistics were used. A multivariate analysis technique like Partial Least Square (PLS) was used to identify the significant job satisfaction factors from the factors identified through factor analysis. The theoretical framework of job satisfaction of RMG workers of Bangladesh is shown in Figure 2.

Data
Relevant data for this study were collected from both primary and secondary sources. Primary data were collected from the interview of 455 workers from 43 leading garment companies located in Dhaka, Savar and Gazipur districts of Bangladesh. The secondary data were collected from the journals, periodicals, annual reports of BGMEA, research monographs, BGMEA website, etc. Both qualitative and quantitative data were gathered for the research.

Population
In this research study, there were an estimate of 4.2 million garment workers working in 4,500 factories (Islam & Chowdhury, 2014) and thus it is taken as population.

Sampling Technique
This study used simple random sampling method for collecting sample respondents. Slovin's formula was used to justify the number of sample size.

Sample
The sample of the study consisted of workers from 43 selected garment factories situated in and around Dhaka city, Savar and Gazipur district. 455 workers were surveyed of which 51.2% were female and 48.8 % were male. Workers of different age groups were chosen. The minimum age of the respondents is 15 and maximum is 55. Sample distribution of the study has been shown in Table 1.

Sample Selection
Data were collected from a range of 8-15 workers from each RMG factory. A close-ended structured questionnaire with 37 items in the questionnaire has been used for data collection. Workers from four departments: cutting, sewing, finishing and packaging were chosen for data collection. The following is a table showing the number of sampled worker taken from different departments of the companies under study (Table 2).

Statistical Techniques Used
Both descriptive and inferential statistics were used to analyze the data. Inferential statistics like Factor Analysis (FA) was used to separate the factors related to job satisfaction of the ready-made garment workers of Bangladesh. Partial Least Square (PLS) method was used to identify the significant factors from the factors identified through factor analysis.

Convergent Validity
Whenever many items are utilized to measure single construct, the item (indicator) convergent validity should be one of the main concerns to the researcher. In this article, the model was tested for convergent validity to measure the extent to which different items are in agreement (Kinnon, 2008). When all the factor loadings for the items used in the same construct are statistically significant, convergent validity is tested (Anderson and Gerbing 1988). In addition, it could also be accessed through factor loadings, composite reliability and the average variance extracted (Hair et al. 1998). The findings of the model used in this study (Table 3) show that the factor loadings for all items exceeded the recommended value of 0.50 (Hair et al. 1998). Composite reliability (CR) values in this study are ranged from 0.720 to 0.833 which exceeded the acceptable value of 0.70 (Hair et al. 1998).

Discriminant Validity
The discriminant validity indicates the degree to which variables of a given model vary from variables of other variables in the same model (MacKinnon, 2008). To conduct PLS analysis, the important thing for discriminant validity is its construct share more difference with its variables than that of other constructs in a specific model (Hulland, 1999). In this study the discriminant validity of the instrument has been tested. It has been evaluated by examining the correlations between the measures of potentially overlapping constructs. Factors loads are stronger on their own constructs in the model and the square root of the average variance extracted for each construct is greater than the levels of correlations involving the construct (Fornell and Larcker, 1981). The square root of the average variance extracted for each construct is greater than the items on off-diagonal in their corresponding row and column, thus, indicating the adequate discriminant validity (Table 3). The inter-construct correlations demonstrate that every construct shares greater variance values with its own measures than with other measures. In fine, the model confirmed adequate convergent validity and discriminant validity.

Average Variance Extracted
All values of the Average Variance Extracted (AVE) that deals the variance captured by the indicators relative to measurement error were greater than 0.50 which indicate acceptability of the constructs (Fornell & Larcker, 1981;Henseler, Ringle, & Sinkovics, 2009). Table 3 shows that these indicators satisfied the convergent validity of the constructs.

Test of Reliability
To analyze the reliability (internal consistency) of the variables, this study used the Cronbach's alpha coefficient and composite reliability (CR) value. Table 3 shows all Cronbach's alpha values are above 0.60 cut-off values as suggested by Nunnally and Berstein (1978). Standardized Cronbach's alpha formula is given below.

= . ̅ ̅ + ( − 1). ̅
Here, N is equal to the number of items, C-bar is the average inter-item covariance among the items, and V-bar equals the average variance.

Coefficient of Determination
The reliability also finds that the coefficient of determination (R square) is 0.366 for the dependent variable i.e., workers' job satisfaction (Table 3). This means that the five independent variables are Salary and Benefits, Supervisor's Behavior, Work and Family Life, Working Condition and Working Environment moderately explain 36.6% of the variance in workers' job satisfaction.  (Fornell & Larcker, 1981;Henseler, Ringle, & Sinkovics, 2009), Composite Reliability>0.70 (Hair et al. 1998), Cronbach's alpha>0.60 (Nunnally and Berstein , 1978) Generally, a global fit measure (GOF) was conducted for path modeling. It is defined as the geometric mean of average communality and average 2 (especially endogenous variables) (Chin, 2010) (see the formula below). In this study, GOF value was 0.46 ( 2 = 0.366, average AVE = 0.5674 for overall workers' job satisfaction). So, the value of GOF exceeded the largest cut-off value (0.46) and it was indicated that the proposed model of this study had better explaining power than that based on the recommended value of GOF small= 0.1, GOF medium= 0.25, and GOF large= 0.36 (Akter et al., 2011).

Results of Factor Analysis
A survey on a total of 455 respondents was conducted. Exploratory Factor Analysis was used in analyzing the data which is a widely utilized and broadly applied statistical technique in social science. The factor analysis technique has been applied to identify the factors that affect the job satisfaction of readymade garment workers in Bangladesh. A total of 37 variables were identified for job satisfaction of workers through literature review. The variables were categorized into five factors which have found from rotated factor matrix (Table 4).The factors are named as (i) Salary and Other Benefits, (ii) Supervisor's Behavior, (iii) Work and Family Life, (iv) Working Condition, and (v) Working Environment.

Salary and Benefits
This factor includes variables like amount of salary, medical facilities, first aid facilities, paid vacation time/sick leave, maternity leave with benefits which are the major components of salary and benefits. It is the most important factor concerned with job satisfaction of RMG workers of Bangladesh as it contains highest eigenvalue.

Supervisor's Behavior
The second important factor of job satisfaction is supervisor's behavior that included variables like freedom to express my opinion, behavior of supervisor, good working relationship with supervisor are found to be the major components of Supervisor's Behavior.

Work and Family Life
The third important job satisfaction factor is work and family balance that included variables like good working relationship with colleagues, happiness with job, spouse employment option are the major components of work and family life.

Working Condition
This factor includes variables like bonus to the workers, flexibility in work scheduling which are the major components of Working Condition.

Working Environment
This factor includes variables like fire protection facilities, adequate spaces at workplace, chaos and confusion of company as major components of working environment.

Results of Multivariate Analysis -Pertial Least Square (PLS)
A multivariate analysis technique, by using 'SmartPLS', was used to identify the significant job satisfaction factors from the factors identified through factor analysis. Path Diagram of Job satisfaction of RMG workers of Bangladesh suggested that the salary and benefits (0.368) has the strongest effect on worker job satisfaction. The hypothesized path relationship between salary and benefits and workers' job satisfaction is highly significant at 1% level of significance. The hypothesized path relationship between Supervisor's Behavior and workers' job satisfaction is statistically significant at 5% level of significance. The hypothesized path relationship between Work and Family Life and workers' job satisfaction is highly significant at 1% level of significance. On the other hand, the Working Condition and Working Environment are not significantly related to the workers' job satisfaction (Figure 2).

Figure 2. Relative Importance of the Job Satisfaction Factors and their Relationship with Overall Job Satisfaction of RMG Workers
The analysis also shows relationships of the variables constituted the job satisfaction factors and their relative importance and relationships with the factors and the overall job satisfaction of the RMG workers of Bangladesh ( Figure 2).
The path coefficients of the factors concerned with job satisfaction of RMG workers show that salary and benefits is the most important factor of job satisfaction (0.368) followed by work and family life (0.186), supervisor's behavior (0.094), working environment (0.069) and working condition (0.053) ( Table 5).   Table-6 shows that all of the T-Statistic are larger than 1.96 at 5% level of significance, that means the outer model loadings are highly significant. So, our SEM model is accepted in this study. Table 6 shows that T-values of outer loading are significant at 5% level of significance. The hypothesis testing was carried out by examining the path coefficients (beta) between latent constructs and their significance. To test the significance of the path coefficients the bootstrapping technique was utilized with a re-sampling of 500 (e.g., Bradley et al., 2012). The R 2 value of endogenous latent construct illustrates the predictive relevance of the model. Table 7 presents the results of hypothesis testing. The findings show that the hypotheses H 1 , H 2, H 3, were rejected on the basis t-values which is higher than 3.3 at the 0.1% level of significance. But H 4 , and H 5 were not rejected the null hypothesis on the basis of t-values which is not more than 1.96 at the 5% level of significance. Table 7 reflects the determination of the respective hypotheses. This is visually depicted in the conceptual model in Figure 2. The discussion, below, considers the outcome of each hypothesis in turn. Working Environment has no significant impact on overall Job Satisfaction of RMG workers. Accepted

Discussions
In this study, five important factors have been identified through factor analysis for job satisfaction of ready-made garment workers of Bangladesh. But all the factors are not equally important for job satisfaction. Sarkeret. et.al., (2014) mentioned that workers satisfaction in RMG sector is not only dependent on financial factors like salary, bonus, incentives healthcare, sick leaves, etc., but also dependent on some nonfinancial factors like work description, work orientation, unbiased evaluation of work performance, training, job security, efficient and supportive line manager, good working environment etc. Another author Smith, et. al., (1969) stated that there are five factors that have direct impact on satisfaction of workers i.e. supervision, relationship with coworkers, present pay, nature of work and opportunity for promotion. Compared to previous literature, the study has concluded that some of the factors like salary and benefits, supervisors' behavior, work and family life are the most influential factors for employees' job satisfaction while working environment and working condition are found to be less important factors for job satisfaction of RMG workers in Bangladesh.

Conclusions and Recommendations
In a developing country like Bangladesh, where most of the garment workers are illiterate and are very poor in comparison to workers of other developed countries, salaries and benefits are prime concern for them. The salary and benefits are very poor of these countries. The reasons are also concerned with the on time payment of salary and benefits to the workers. These workers are migrated from the rural areas to urban areas for job and the poor salary does not guarantee them to survive them well. As a result, salary and benefits is the most important factor of the workers in this sector especially in Bangladesh. As the workers work under the supervision of their supervisors, their satisfaction directly depends on the supervisors' behavior. If they behaved well by their supervisors, they feel comfortable with their jobs and duties and thus become satisfied at work. Family and personal life balance also plays a very important role in job satisfaction of the RMG workers of Bangladesh. This study finds that if the workers are happy in their families and personal lives, they feel better and thus become satisfied with their jobs. On the other hand, RMG workers hardly care about working condition or working environments which were not found significant in this study.
This study suggests that the garment owners are concerned about their earnings that can be achieved by making the workers satisfied at their jobs. This study also suggests that if workers are paid better salaries and provided them with adequate benefits like medical facilities, first aid facilities, paid vacation time/sick leave, maternity leave they will be more satisfied with their jobs. Supervisors' behavior and relationship with the colleagues should also be improved for enhancing the satisfaction level of the RMG workers in Bangladesh.

Scope for Further Research
This study only concentrates on job satisfaction of ready-made garment workers of Bangladesh. So, the factors identified are based on the lower level workers' job satisfaction, not the job security and or external environmental factors concerned with job satisfaction. Therefore, there is an ample scope to conduct further study by taking diversified variables into account for more accurate results of job satisfaction, job security, and external environment of RMG workers in Bangladesh.