The role of decent work in the Zimbabwean retail sector: Testing a job engagement and turnover intention model

to the management of human capital in the retail sector.


Introduction
Zimbabwe has experienced economic depression, which has led big companies to either scale down or close operations (Manhiwa, Mapetere, & Mhonde, 2017). This has undoubtedly presented serious challenges and at the same time, limited opportunities for advancing human rights such as the right to decent work. Majoni, Matunhu and Chaderopa (2018) posit that the economy of Zimbabwe has now become deformalised so that four out of five jobs in Zimbabwe are informal, resulting in decent work deficits, namely poor quality, unproductive jobs, unsafe work and insecure income (UN, 2017). The term 'decent' denotes that work must be of acceptable quality in terms of income, working conditions, job security and rights (UN, 2017). Decent work is therefore a fundamental right or need that is common in all families and communities at all levels of development (Navajas-Romero et al., 2019).
The persistent need to curb these deficits and to contribute to sustainable development has, therefore, become an important goal under the decent work agenda (Di Fabio, 2018). It is worth mentioning that international organisations such as the United Nations (UN) have shown an interest in decent work and sustainability (UN, 2017). The UN put forth a framework with 17 goals, two of which directly tackled working conditions and commitment to decent work (Goal 8) as well as promotion of health and well-being (Goal 3) (UN, 2017). Looking at Africa, where economic growth has been buoyant, a realisation to also reap from the decent work agenda has not been ignored. Africa is characterised by a weakness both in the public and private sector and a huge size of the informal economy, making the decency of work questionable. No wonder, there is a greater need in Africa to ensure that workers' hard work is fully utilised and rewarded and that respect for fundamental rights and provision of social protection is ensured. Efforts to incorporate decent work tenets into national laws have been adopted by most African countries, including Zimbabwe (Ahmad, 2021).
Over the years, job engagement has gained attraction (Hoole & Bonnema, 2016). Scholars argue that job engagement is a product of workers' experiences of decent work. The idea is that when working conditions are proper and the work meets the survival needs, social connections and social determination, there is a higher chance of workers being fully engaged in their tasks and responsibilities (Duffy et al., 2017). The most common definition of job engagement comprises three elements: vigour, dedication and absorption. The interpretation here is that job engagement has to do with a worker's willingness to work, a reflection of high commitment and persistence, with a show of higher levels of energy and cognitive presence. In the current context, job engagement still matters; organisations are compelled to stay abreast of it to ensure that workers continue being a source of competitive edge. According to Glasson et al., (2013), organisations should ensure good working conditions, that is, safety, secure income, social protection and social dialogue to have an engaged staff. Thus, a link between decent work and job engagement cannot be ignored.
Academic research about decent work and job engagement has been more prevalent in the European context. The idea of decent work is a little different in developing countries as compared with developed countries (Ijalla, 2019). In the African context, research was carried out by Ijalla (2019), in Ghana, and the study also sought to investigate the relationship between decent work and job engagement. Despite the difference between the European and African contexts, the findings were consistent, as they also highlighted that a decent work environment is a major influencer of job engagement. A thorough look into turnover and turnover intention makes it possible for companies to cut costs. Turnover intention is defined as the likelihood that a worker will leave their job. Basically, it reflects the attitude that an employee has towards their job. Turnover intention is a key indicator of workers' perceptions and judgements about their work (Memon et al., 2014). It is important to observe that how workers' intentions to leave are characterised by working conditions. Turnover intentions come as a result of dissatisfaction with pay systems. This justifies the link between decent work and turnover intentions. Decent work deficits will obviously push a worker to scan for greener pastures.
Also, the link between job engagement and turnover intention is undeniable. According to Screw (2019), job engagement is said to predict turnover intention. The link between the two indicates that disengagement among workers results in a higher risk of considering other employment options (Belete, 2018). Moreover, a positive association is found between job engagement and intention to stay or leave. In addition, if the workers are more engaged in their work, they are less prone to actively seeking alternative employment opportunities.
This study seeks to assess the effects of decent work on worker engagement and turnover intention in the Zimbabwean retail industry (Namubiru, 2019). The existence of this gap in literature necessitated the need to carry out this study. The purpose of this study is to examine the concept of decent work and its relationships to job engagement and turnover intentions among employees in the Zimbabwean retail sector.

Context
Like most countries, Zimbabwe has not been spared from the environmental constraints, which have crippled most businesses and left the economy staggering (ZIMSTAT, 2017). According to Bakker (2017), Zimbabwe's business environment is a turbulent one and is highly unpredictable. There is no stability in terms of job security and income, as most organisations are surviving with little or no reserves at all (Mucheche, 2015). To make matters worse, the coronavirus disease 2019 (COVID-19) pandemic has further crippled the economy, leading to various organisations failing to cope with the new problem and eventually closing down. There is no doubt that all these challenges have presented a twist in the world of work. According to Booth (2019), work has taken a completely different form, whereby jobs are no longer secured and workers are not protected.
The retail sector, which is mostly dominated by supermarkets such as TM and OK Zimbabwe, is seriously constrained by several challenges (Bakker, 2017). As a way to curb these challenges, most employers in this sector across the globe are shunning permanent contracts and moving towards a gig economy, which is a labour market made up of short-term contracts and freelancers. A spokesperson for the then retail giant TM reported to The Daily News (2015) that although they were expanding, they were limiting the number of employees to the bare minimum. This undoubtedly makes workers vulnerable to exploitation: a common model of this sector. The retail sector is associated with abusive practices by the employer, long working hours, low pay, ill-treatment of subordinates, as well as inhumane conditions. Booth (2019) concurs, as he states that in some cases workers in this sector have reportedly been coerced to work overtime, with no extra pay. Considering the excessive supply of labour in the market, the bargaining power of employees in this sector has been reduced, placing them under the guillotine of the employer. The decency of work in this sector is, therefore, highly questionable. Workers have been reduced to an extent where they are forced to settle for anything, knowing that they can at least afford some of the basic physiological needs. Despite being a signatory to International Labour Organization (ILO) Decent Work Check, Zimbabwe's labour standards have deteriorated over the years. Perhaps this is because of resistance to being malleable in response to environmental forces or even bad corporate governance.

Literature review and hypothesis development Decent work
While it is not a new concept, decent work continues to take centre stage in the business world (Gensmer et al., 2020). Having gained international currency in recent years, decent work has been accepted as covering a wide range of issues, which are organised under the following four pillars: The first pillar is employment promotion. It encompasses policies that foster productivity and competitiveness within the world economy to promote employment growth targeted at the poor (Chyung et al., 2018;ILO, 2017). The second pillar is social protection. This relates to actions that help reduce people's vulnerability to chronic indigence through policies that promote fairness and social justice at work (Massoudi et al., 2018). Without these policies that foster social protection, neither work nor lives can be decent and all this undoubtedly contributes to economic prosperity (Pereira, 2019). The third pillar is social dialogue. There is nothing decent about work if it infringes the right to participate in decision-making or the freedom of association. This pillar is all about the extent to which employees are given the forum to express their workrelated issues and partake in making decisions about their well-being (Flores, Navarro, & Rasheed Ali, 2017). The overall objective is to promote the exchange of information, negotiation and consultation among workers, employers and the government. The fourth pillar is rights at work. This dimension relates to the promotion of human rights. The ILO (2018) defines these rights as the universal fundamental principles and rights at work and these must be followed by all states, whether they have ratified ILO conventions or not. These include rights such as elimination of child labour, abolition of forced labour and promotion of nondiscrimination in employment (Tayebwa, 2019).
Following the main tenets of decent work briefly discussed here, it is clear that decent work is a fundamental right. It determines a global evaluation of the quality of a given job to promote the physical and cognitive health of workers around the world, and it also addresses such needs as survival, self-determination and social connection. The business environment, which has globally continued to be unpleasant to organisations, has prompted many scholars to investigate the construct of decent work and its variables (Ferreira et al., 2019). Cruz et al., (2017) agrees that the concept has gained support and has been approved by the international community. Decent work has its origin in developed world formal working environments and most of the efforts to further understand it have been from the developed countries (Winchenbach, 2019). However, the ILO (2018) makes it clear that although a universal concept, decent work is not uniform, that is the policies of decent work should be based on national or sectorial contexts.

Decent work versus job engagement
Evidence shows that a working environment with favourable conditions is positively associated with work engagement.
Another study by Navajas-Romero et al. (2019) analysed the state of working conditions, if decent or not, as well as their effects on work engagement and the results broadly confirmed decent work as a strong determinant of job engagement. The same study mentioned job security as a configured dimension of decent work that directly contributes to making workers elicit higher levels of vigour, dedication and absorption. Tayebwa (2019) examined the influence of decent work on job engagement among traditional employees at Kiruhura district local government in Uganda. This study revealed a positive association between the two. According to Glasson (2013), organisations should ensure the decency of work to cultivate engagement among workers. A cross-sectional study by Namubiru (2019) carried out at Kampala Capital City Authority also affirmed that employees must work in a decent work environment to remain engaged. However, the results indicated a weak positive relationship as it was found that decent work does not necessarily result in increased job engagement because engagement is more of an intrinsic factor and cannot be affected by the physical environment. Decent work could be an added advantage, but engagement is innate. Based on the given literature, it can be hypothesised that: H 1: Decent work is positively associated with job engagement.

Decent work versus turnover intention
People work to meet certain needs and if those needs are not met, they are likely to search for other job opportunities, especially when their work negatively impacts their psychological well-being, leading to turnover intentions (Blustein et al., 2019;Taskiran, 2017). Most organisations, especially in Zimbabwe, have acute decent work deficits but they might not be experiencing actual turnover because of the high rate of unemployment. It does not necessarily mean that these workers do not harbour any intentions to leave.
Research carried out in Turkey established that all elements of decent work predict turnover intention. The findings reflected that those workers are more likely to experience turnover intentions in an unsafe workplace, offering insufficient compensation and not offering career growth opportunities. A related study by Griffeth, Hom and Gaertner (2010) examined the relationship between pay, a decent work indicator and turnover intentions and concluded that jobs that provide adequate income were more likely to retain workers that exhibit positive behaviours. The other conclusion from this study was that insecure pay had a negative correlation with turnover intentions. Hence, the link between decent work and turnover intentions cannot be denied. Based on the given literature, it can, therefore, be hypothesised that: H 2: Decent work is negatively associated with turnover intentions.

Job engagement versus turnover intention
A relationship between job engagement and turnover intention is a significant one that all organisations should investigate. Job engagement is associated with positive perceptions towards a job, optimism and confidence, and all these are crucial in predicting workers' willingness to stay or leave. An organisation's capacity to raise the engagement levels of its employees would motivate employees to stay, thereby curbing turnover intention. Saks and Rotman (2006) found job engagement to be a strong predictor of turnover intentions. This link indicated that more disengagement among workers resulted in a higher risk of considering other employment options. Santoni and Harahap (2018) observe that when employees are disengaged, they make decisions to leave. A related study by Agoi (2015) in the public sector in Kenya indicated that job engagement has surfaced in the world of work as a construct that has a significant influence on employee turnover intentions. It is, however, important to note that there is a serious gap in literature on studies relating to the correlation between job engagement and turnover intention in Africa and specifically in Zimbabwe. However, based on the aforementioned literature, it can be hypothesised that: H 3 : Job engagement is negatively associated with turnover intentions.

The psychology of work theory
Work plays a pivotal role in people's lives (Aburumman, 2020). Following an attempt by traditional career theorists to expound on the role of work and how individuals make career choices, Blustein et al. (2006) realised how these theorists did not capture the concept of work fully. Instead, they excluded marginalised groups. The career development and work experiences of individuals with insufficient financial access and social capital were not captured in these theories (Duffy et al., 2017). Also, important to note is how these theories failed to appreciate the role played by contextual factors in shaping the work lives of people amid a dynamic labour market affected by globalisation. Considering all these weaknesses of the traditional career theories, the psychology of work theory (PWT) is developed: a fully realised conceptual structure widening the scope of traditional career choice and development theories. Efforts are still underway to generate interrelated and applicable assumptions to this theory.
The PWT argues that the importance of work in meeting people's needs and well-being captures work as a crucial aspect of life and an indispensable element of mental health (ILO, 2018). A key aspect of this theory is that everyone involved in work, regardless of sex, race or gender, should be afforded social justice by securing decent work, which makes a significant component of well-being and access to opportunity (Duffy et al., 2017;Isik, Kozan, & Isik, 2018;Thompson, 2017). Another eminent tenet of this theory is that sociocultural factors should be treated as primary in understanding career decision work experiences of all people. Di Fabio (2018) builds on this view, stating that access to work is constrained by powerful social, economic and political forces. Generally, all these are captured in what the PWT expresses as predictors and outcomes of securing decent work (Duffy et al., 2019).

Outcomes of decent work
Psychology of work theory argues that when people obtain decent work, their fulfilment and well-being are enhanced through the satisfaction of the following three basic needs: 1. Survival and power: People require basic needs to survive and PWT proposes that performing decent work will help individuals to meet these survival needs (Kline, 2016). 2. Social connection: Humans have an inherent need for attachment and belonging. Meaningful work will help individuals to meet their need to link with others as it provides opportunities to build those connections (Ribeiro, 2017;Rodriguez, Reise, & Haviland, 2017). This, undoubtedly, gives workers the idea that they are contributing to the larger context. 3. Self-determination needs: These provide meaning, a sense of identity and accomplishment (Gansler et al., 2020). As a result, it works again fulfils the need to partake in meaningful and value-congruent activities through experiencing autonomy (Pouyaud, 2018).

Research design
This study adopted a descriptive survey design (Saunders et al., 2009). The descriptive survey design is a quantitative research technique in which a researcher uses some sort of survey or questionnaire on a sample to describe and interpret attitudes, experiences, current status of individuals and events. This design seeks to establish the possibility of a causal relationship among the variables in question. In calculating the proposed model, a multiple regression with partial least squares (PLS) was utilised. This was conducted using a structural equation model (SEM) approach. Such an approach allows for the building of the research models by establishing latent variables. Latent variables as those that are not observed directly but inferred from other observed variables. Such an approach and technique is useful when working with small samples, estimating complex models and making predictions and explanations (Hair et al., 2019). There is an observation of the suitability of such approaches and techniques when seeking to ascertain success factor research or when exploring the source of a competitive advantage.

Research participants
A total of 103 participants from the major retailers in Bulawayo Urban were conveniently sampled. The participants comprising both till operators, supervisory and managerial staff from three mainstream retailers took part in the study. A combination of these participants provided for data triangulation. The researchers employed the convenience sampling approach.
http://www.sajhrm.co.za Open Access Convenience sampling is a non-probability sampling method in which participants are chosen because of their convenience, which could be proximity or accessibility to the researcher (Yin, 2018). The researchers chose this sampling approach because according to Saunders et al. (2009), this approach is simple, efficient and makes it easy to collect data in a short space of time from a large sample. Furthermore, the lack of employee database because of the reluctance by most HR managers to avail them to the researchers, made convenience sampling even more appropriate.

Research instruments
The researchers used a structured questionnaire. A questionnaire allows the collection of large amounts of data from a huge number of participants in a short period. The structured questionnaire required the respondents to select answers from a set of choices offered explicitly by the researchers and it simplifies data analysis. As a result of the COVID-19 pandemic, the researchers distributed the questionnaires online through e-mails and social media platforms, to reduce contact with respondents. The questionnaire had four sections, namely Sections A, B, C and D.
Part A measured decent work and consisted of bio-data involving five items, namely age, sex, marital status, level of education and duration of employment. Part B comprised 21 items that measured decent work. The items were adapted from the 31-item Decent Work Questionnaire, with a reliability of 0.70. A 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) was used.
Part C measured job engagement. A 19-item Utrecht Work Engagement Scale, with a reliability level of 0.70 was adapted. The study, however, adapted 14 items from this scale. The scale is a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). A sample item is as follows: 'When I get up in the morning, I feel like going to work'.
Part D measured turnover intention using the 14-item Turnover Intention Scale. From this scale the study only adapted 11 question items. The instrument has a high level of reliability of 0.80. The scale is a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Statistical analysis
The quantitative data were inputted into the IBM Statistical Packages for Social Sciences (SPSS). Structural equation modelling techniques were used for data analysis (Hair et al., 2019). The SEM data analysis involved dealing with missing data using the multiple imputation method. Item analysis was used to remove poor items in the different scales used for data collection. Integrated analyses checked were item-total correlations, the squared multiple correlations, checking the subscale or scale reliability and variance when an item or items is deleted, inter-item correlation and the item means and standard deviation. To assess the unidimensionality of scales and sub-scales of the different instruments used in the study, dimensionality analysis was also performed. Exploratory factor analysis and the principal component analyses (PCA) were used. The PCA analyses all variances in the correlation matrix using a cut-off value of 60% as recommended by Hair (2014).

Reliability analysis
The composite reliability (CR), Cronbach's alpha test (CA) that test internal consistency and average extracted variance (AVE) for the variables decent work (DW), job engagement (JE) and turnover intention (TI) are presented in Table 1.
It can be observed from Table 1 that all the CR values and the CA values are above the recommended threshold value of 0.7. Furthermore, the AVE values are all above 0.5. According to Hair (2014), these results imply the existence of internal consistency in the variables. Cross-loadings and factor loadings of the variables that are used to test convergent validity are presented in Table 2.
The results presented in Table 2 show that all the crossloadings and factor loadings of the variables are above 0.6. This means that convergent validity exists. The factor loadings that were below 0.6 were eliminated from the model, hence convergent validity was met. The interconstructs correlations used to test discriminate validity are shown in Table 3. Table 3 supports the existence of discriminant validity because all the square root of the AVE values, which are the diagonal elements for the factors are all above the corresponding correlation coefficient values of the factors. A proposed SEM is presented in Figure 1.  The following hypotheses were derived from the SEM presented in Figure 1: H 1 : Decent work is positively associated with job engagement H 2 : Decent work is negatively associated with turnover intentions H 3 : Job engagement is negatively associated with turnover intentions To statistically examine the stated hypotheses, the partial least squares (PLS) technique was used and SmartPLS software was used to calculate the path coefficients. The path coefficients are shown in Table 4.
The first hypothesis sought to examine the relationship between DW and JE. According to the results displayed in Table 4, DW had a positive effect on JE because of the coefficient value of 0.294. The relationship between DW and JE is statistically significant at a 1% level of significance because of the probability value, which is less than 0.001. These results suggest the acceptance of the H 1 which says that decent work is positively associated with job engagement.
The second hypothesis sought to examine the relationship between DW and TI. It can be observed from Table 4 that DW had a positive effect on TI because of the coefficient value of 0.026. However, the relationship is not statistically significant because the probability value of 0.286 is greater than 0.05. The results suggest the rejection of H 2 , which says DW is negatively associated with TI. Regarding the third hypothesis, which sought to examine the relationship between JE and TI, it was found that JE had a positive effect on TI as indicated by a positive coefficient value of 0.961. The relationship between JE and TI is statistically significant at a 1% level of significance because of a probability value less than 0.001. Despite the negative effect of JE on TI, hypothesis H 3 that say job engagement is negatively associated with turnover intentions is accepted. This means that improvements in JE will be reflected by a decrease in TI.

Goodness of fit model
The results in Table 4 suggest that H 1 and H 3 are supported while H 2 is not supported. Figure 2 provides a summary of the fitted model containing coefficients and factor loadings.
R² is useful in determining the goodness of fit of a model and indicates that the value must be above zero. The R 2 (R 2 = 0.939) exceeds the threshold value of zero. According to Figure 2, 93.9% (R 2 = 0.939) total variability in TI was being explained in the model by the independent variables (DW and JE). The R² for the fitted model suggests good levels of predictive accuracy of the model. Table 5 shows some of the goodness of fit results.
According to Table 5 results, the Standardized Root Mean Square Residual (SRMR) value for the fitted model is 0.073, which is less than the threshold value of 0.08, suggesting that the model can be accepted. Furthermore, the NFI value for the model is 0.915 that is slightly above the recommended threshold value of 0.9. These results suggest that the fitted model is a good model.

Discussion of findings Decent work versus job engagement
The study established a statistically significant relationship between decent work and job engagement because of a probability value less than 0.001. This led to the acceptance of the alternate hypothesis H 1 , which says decent work is positively associated with job engagement. The implication here is that improvements in decent work, through an organisation's ability to provide social protection, income security, safety and health as well as enhancing social dialogue, prompt an increase in the level of job engagement. As is the norm, people work to earn a living and to provide for those who depend on them. If they feel these needs are well-catered for, they are likely to be highly engaged in their various organisations (Namubiru, 2019). This helps to confirm that decent work has a positive relationship with employees' levels of job engagement.
Furthermore, the findings of this study are in tandem with the results of a study carried out by Kittler et al. (2020). The study found that working environment with favourable conditions is positively associated with high level of work engagement, especially when the dimension of decent work is present. In a country like Zimbabwe where the business environment is turbulent, job engagement is a critical ingredient for the survival of firms. To attain job engagement, organisations should be able to provide work that offers safety and health, sufficient compensation, time and relief from work (Navajas-Romero et al., 2019).  However, the findings of this study were incongruent with those established by Namubiru (2019). The study found a weak positive relationship between decent work and job engagement. The indication was that the provision of a decent work environment does not necessarily mean that workers will be engaged. The explanation was that work engagement is an innate state and therefore it is less likely to be influenced by the external factors in the workplace. Nonetheless, it should be observed that the relationship between decent work and job engagement is contextual and different findings will emanate from different organisations, nationals or even continents.

Decent work versus turnover intentions
This study established an insignificant relationship between decent work and employee engagement because a probability of 0.286 is greater than 0.05. The results suggest the rejection of H 2 , which says DW is negatively associated with TI. Strange as it seems, the findings reflect those employees who considered their job to be decent were actively searching for other jobs. A decent work environment does not necessarily curb turnover intentions, nor does it stop employees from harbouring turnover intentions (Namubiru, 2019;Taskiran, 2017). The work might be decent but that cannot tie employees to it.
The weak relationship between decent work and turnover intentions (shown by a coefficient of 0.026) can be explained by age demographics. Very few respondents (5.3%) were aged 50 years and above. The age demographics revealed that the retail industry is mainly dominated by younger workers probably because of their higher levels of labour intensity (Taskiran, 2017). As a result of the long hours and unfavourable conditions (Griffeth et al., 2010), young workers can better bear such conditions when compared with their older counterparts. This explains why the retail sector should come up with measures that are meant to include older workers (Spencer et al., 2021).

Job engagement versus turnover intention
The third objective of the study was to examine the relationship between job engagement and turnover intention. The proposed hypothesis was H 3 : Job engagement is negatively associated with turnover intentions. A positive coefficient value of 0.961 revealed a strong association between job engagement and turnover intention. The probability value less than 0.001 suggests the acceptance of the alternative hypothesis. The implication is that job engagement has a greater bearing on turnover intentions. Engaged workers are more attached to their job and have a lower chance of leaving. Workers who answered to being deeply immersed and dedicated to their work did not voice turnover intentions.
The results of the study also indicated that the psychological attachment that workers had towards their jobs discouraged  any development of turnover intentions. Workers with a positive mindset are not likely to withdraw from their work. Hence, it is of great importance to improve engagement among employees because it reduces turnover intentions. Workers who found their work meaningful and full of purpose -the most influential factor of job engagementexhibited no signs of seeking alternative employment. When workers are disengaged, they make decisions to leave (Santoni & Harahap, 2018).

Limitations and direction for future studies
The study was cross-sectional and conducted among the retail sector in Bulawayo province. More robust studies including national samples could give more generalisable findings. Research on decent work in Africa and Zimbabwe in particular, is still in its infancy, hence the need to explore more on this subject (Perreira et al., 2019).

Conclusion and recommendations
The study established that decent work significantly affects job engagement. An environment characterised by secure pay, social protection and social dialogue goes a long way in creating an engaged workforce. It was found that a decent work environment does not necessarily curb turnover intentions. An organisation might be offering the best as far as decent work is concerned, but that will not stop employees from harbouring turnover intentions. The study also established a strong association between job engagement and turnover intention. Job engagement has a greater bearing on turnover intentions. Engaged workers are more attached to their job and have a lower chance of leaving.
Considering the conclusions, it is recommended that given the importance of decent work in the retail sector, employers should strive to prioritise all dimensions of decent work. Decent work should therefore be seen as a fundamental right for all employees. Managers should craft policies that create a favourable environment for job engagement to curb employees' intentions to quit. These policies could address issues such as social protection, social dialogue and income security, among others.