Decent work, capabilities and flourishing at work

Abstract

Purpose

This study aims to investigate the associations among decent work (DW), capabilities and the flourishing of employees in a South African context.

Design/methodology/approach

A cross-sectional survey was conducted with a convenience sample (N = 436) of early childhood development practitioners from two South African provinces. A demographic questionnaire, the Decent Work Scale, the Capability Set for Work Questionnaire and the Flourishing-at-Work Scale were administered.

Findings

Latent class analysis showed four capability sets: robust, relational, knowledge/skills and weak capability sets. Employees with a robust capability set were more inclined to report DW than those with knowledge/skills and weak capability sets. Employees with a weak capability set were significantly less inclined to report organisational values that complement family and social values than the other three capability sets. Employees with a robust capability set reported significantly higher emotional well-being (EWB), psychological well-being (PWB) and social well-being (SWB) levels than those with relational, knowledge/skills and weak capability sets. DW was significantly related to EWB, PWB and SWB.

Originality/value

This study contributes to the literature regarding DW, capabilities and flourishing of employees in a non-western, educated, industrialized, rich and democratic and non-POSH context. The study highlights the need for well-being policies that focus on DW and the capabilities of people in disadvantaged positions. These together would strengthen their agency for converting capabilities into well-being.

Citation

Ragadu, S.C. and Rothmann, S. (2023), "Decent work, capabilities and flourishing at work", Mental Health and Social Inclusion, Vol. 27 No. 4, pp. 317-339. https://doi.org/10.1108/MHSI-05-2023-0054

Introduction

Work matters to people and provides an important context for their well-being (Blustein, 2019; Clifton and Harter, 2021). However, the world of work is changing rapidly. Firstly, technologies are changing the nature of work and how people interact and do their jobs (Ghislieri et al., 2018). Secondly, due to global change processes, people and institutions face precarious conditions (Baart, 2021), which result in precarious jobs and lives (Blustein, 2019; Mumby, 2019). Thirdly, climate change, poverty, inequality and corruption severely threaten humanity (Goldin and Muggah, 2020). Fourthly, individuals are changing: they want to be involved, participate and feel empowered in work (van der Klink et al., 2016). Moreover, individuals’ identities and capabilities are becoming more important for individual and organisational performance and well-being (Grant, 2021; van der Klink et al., 2016).

Positive organisational studies, rooted in positive psychology, offer a lens for understanding the well-being and performance of people in work contexts (Gruman and Saks, 2019; Seligman, 2011). However, in their study of critiques and criticisms of positive psychology, Van Zyl et al. (2023) found that it neglects the role of the context in the flourishing of people. Wissing (2022) criticised the ignorance concerning contexts in positive psychological research by confirming the tendency to focus on research from western, educated, industrialized, rich and democratic (WEIRD) societies (Henrich et al., 2010). Furthermore, Gloss et al. (2017, p. 329) argued that research in industrial-organisational psychology has focused mainly on the POSH perspective (“Professionals who hold Official jobs in a formal economy and who enjoy Safety from discrimination while also living in High-income countries”), which reflects bias away from people living in multidimensional poverty. Gloss et al. recommended that the capability approach (CA; Sen, 1999, 2009) be used as a thrust for moral action in research and practice. In addition, van der Klink et al. (2016) recommended that the CA be used to understand what people value and what they are enabled and able to achieve, considering their social and working conditions.

With the rise of positive psychology, interest in the well-being and flourishing of people has grown (DeHaan et al., 2016). Individuals who flourish fully express their capacities and excellence (Ryan et al., 2013). According to Sen (2009), people’s well-being is best assessed through their ability to engage in what they value. More specifically, well-being requires individuals to have capabilities, i.e. the freedom to achieve valuable functionings. This study investigated the associations among decent work (DW), capabilities and the flourishing of employees in a South African context. Pouyaud (2016) suggested that the CA (Sen, 1985a, 1985b) could be used to understand DW effects. DW addresses the ideological conditions of work, and the CA focuses on the inherent conditions of a person’s capacity and freedom (Gloss et al., 2017).

Based on the CA, van der Klink et al. (2016) developed a sustainable employability model that is relevant to understanding the effects of DW. Sustainable employability involves employees’ capabilities to keep them employable in a specific job or field (van der Klink et al., 2016). Thus, the sustainable employability model is relevant for investigating individuals’ work, capabilities and well-being. Core elements of CA are based on a person’s freedom to choose what constitutes a flourishing life (Orton, 2011; van der Klink, 2019). The CA provides a social justice framework that includes values, enablement and achievement, which result in opportunities to be well (Sen, 1999).

The two approaches, DW and the CA, pave the way for a psychosocial perspective on the well-being of employees (Burchell et al., 2014). The CA provides a broad perspective of DW, capability sets and flourishing in the work environment (Molla and Nolan, 2019). Therefore, this study makes the following contributions to science: firstly, it adds to the literature on the operationalisation and measurement of work capabilities in a context characterised by poverty, unemployment and inequality (Gloss et al., 2017). Secondly, it provides scientific information about the association between capability sets and DW. Thirdly, it provides insights into the associations between DW, capabilities and well-being in a non-WEIRD and non-POSH work setting.

Decent work

The International Labour Organisation (ILO) (1999, p. 3) defined DW as “productive work for women and men in conditions of freedom, equity, security and human dignity.” Duffy et al. (2016) conceptualised DW as a central variable within the psychology of working theory (PWT). The PWT concerns the work experiences of individuals, focusing specifically on those close to or in poverty, those who face discrimination and marginalisation and those who face challenging work-based transitions for which contextual factors are often the primary drivers for the ability to secure DW (Duffy et al., 2016).

The concept of DW emerged as an institutional effort to combat the degradation of the labour market (Pereira et al., 2019). The freedom to choose to work is a core human activity essential to an individual’s being. For most people, their families and communities, decent and productive work is the main route out of poverty. It involves opportunities for a fair income, security in the workplace, social protection for families, better prospects for personal development and social integration, such as the freedom to express their concerns and to organise and participate in the decisions that affect their lives. Likewise, the United Nations (UN) has identified DW and economic growth as one of their envisioned 2030 sustainable development goals (UN, 2015).

According to Duffy et al. (2016), the PWT explicates the link between the ILO definition of DW and a psychological approach to working (Blustein, 2006, 2019). The PWT distinguishes between five characteristics of DW, namely, (Duffy et al., 2016, p. 130):

1. physical and interpersonally safe working conditions;

2. hours that allow for sufficient free time and adequate rest;

3. organisational values that complement social and family values;

4. adequate compensation; and

5. access to health care.

Social and contextual factors (e.g. economic constraints, marginalisation and a lack of social justice) limit access to DW. Furthermore, DW affects the need satisfaction of individuals, which impacts their well-being (Blustein, 2019; Duffy et al., 2016).

Several studies have shown that DW depends on privilege and access to resources that promote career freedom (Duffy et al., 2020). Studies related to work and health issues have been performed and have focused chiefly on job stress, job control, the (im)balance between job demands and rewards (e.g. compensation), job security and occupational hazards (Landsbergis et al., 2014). Duffy et al. (2019) explored how key features of DW can predict health outcomes. Their study showed relationships between DW and mental and physical health.

The capability approach

Initially conceptualised by Sen (1999, 2009) as an ethical framework, the CA covers a broad spectrum of well-being and individual performance qualities. The approach comprises capabilities, functioning and agency (freedom). Firstly, capabilities reflect the different combinations of functionings individuals can achieve, depending on their particular circumstances and the various combinations of what they can do or be. Secondly, functionings represent the circumstances and actions that signify beings and doings people have reason to value (Sen, 1999). Thirdly, agency refers to the freedom to do and be. By integrating a contextual view of agency, the CA observes that people must feel secure and free to initiate and sustain effective action (Blustein, 2019). Freedom is valuable because it gives people opportunities to pursue their valued objectives and protects the process of choice itself (Sen, 2009).

In the CA, the focus shifts from means (resources and goods that people can access) to ends (what they can do and be with such resources and goods). As a result, resources and goods alone cannot guarantee that people can convert them into actual beings and doings (Robeyns, 2017; Sen, 2009). Therefore, equity of capabilities is more important than equal means or resources (van der Klink, 2019). Significant barriers exist for many people in internalising work as an affirming aspect of their lives – they often lack the freedom to choose the work they want (Sen, 1992). The CA moves away from income-led evaluation methods to focus on people’s ability to achieve the things that they value (Frediani, 2010) while at the same time establishing common normative principles, such as inclusiveness, fairness, equality and justice (Ibrahim and Tiwari, 2014).

The CA contends that in a just society, the focus should be on developing and supporting the capabilities of each member of society (Sen, 2009). It should provide freedom and genuine opportunities for individuals to choose what they want to do and how they want to do it and alter their life plans accordingly (Sen, 1999). Understanding the nature and sources of capability deprivation and inequity is central to removing injustices (Sen, 2009). Concerning work contexts, Gloss et al. (2017) argued that it is essential to study injustice and intervene to promote not only procedural justice but also distributive justice (Lefkowitz, 2009). The focus on social justice aligns with Prilleltensky and Prilleltensky’s (2021) argument that without fairness, people will not experience mattering to its fullest extent, resulting in poor well-being.

Capabilities

Capabilities within the realm of the CA refer to an individual’s freedom to practice what he/she regards as meaningful. According to Robeyns (2017), capabilities are the actions and beings people can achieve when they want to. It is also referred to as real or substantial freedom. Therefore, capabilities refer to the idea that our activities are doings, and the type of person one can be is the being (derived from human beings). According to Sen (1999), capabilities are the real freedoms individuals have to achieve their potential beings and doings. Freedom does not imply conventional freedom to do or be something but enough opportunity to achieve it. Thus, if a person has the opportunity, it is only a capability if the opportunity is used.

Sen (1999) does not present a list of capabilities that comprise an individual’s capability set (Claassen, 2018). Instead, Sen (1999) has suggested five types of instrumental freedoms that are necessary for all capabilities: political freedom, economic resources, social opportunities, transparency guarantees and security of protection; these are the factors that are most likely to facilitate individuals in achieving their objectives (Blustein, 2019). In contrast, Nussbaum (2011) has proposed a set of central human capabilities that can be used as a conceptual bridge between the CA and the world of work, including working as a human being, forming meaningful relationships with other employees and exercising practical reason. Abma et al. (2016) identified seven work capabilities, i.e. valued aspects of work that are enabled and can be achieved:

1. using knowledge and skills;

2. developing knowledge and skills;

3. involvement in important decisions;

4. building and maintaining meaningful contacts at work;

5. setting own goals;

6. earning a good income; and

7. contributing to something valuable.

A single capability is essential, but the real value lies within a set of capabilities (De Wet and Rothmann, 2022; van der Klink, 2019).

Based on CA literature, Gloss et al. (2017) identified six characteristics of capabilities:

1. counterfactual – individuals’ actual choices cannot provide a complete understanding of their opportunities and choices;

2. multi-dimensional – there are multiple important capabilities, but one capability is not convertible to another;

3. differentially prioritised – people will differentially order capabilities in terms of their relative importance;

4. multivocal – individuals may attach different meanings to the same capability;

5. element of objectivity – individuals’ subjective self-report of capabilities is likely to be influenced by their relative position in society; and

6. individualistic – they need to be anchored to an individual to guard against the nullification of individuals’ fundamental rights by groups.

The CA examines what is valuable for and valued by people and how these values can be achieved (van der Klink et al., 2016). Conceptualised by Sen (1985a, 1985b), the CA provides a basis from which various factors concerning the performance and well-being of persons can be assessed. The CA argues that more emphasis should be placed on advancing the capabilities of individuals in society, as these capabilities are used to create, follow and adapt their livelihood. The approach pertinently places a dual responsibility on society and the individual to expand and accelerate their capability set to enable and achieve value. Van der Klink et al. (2016) describes a value as part of the capability set if a worker identifies the work value as important, is enabled to do it, and is able to achieve it.

Functionings

Functioning is defined as a person’s actual being and what the person is successfully doing or achieving (Robeyns, 2017). According to Robeyns, functionings can be building blocks of individuals’ well-being (e.g. being physically fit and experiencing good social relationships). A functioning differs from a capability: the opportunity to be well is a capability, whereas being well is a functioning.

Fabricio et al. (2022) summarised the reasons for studying optimal well-being. Firstly, the existence of good and the absence of psychopathology are both markers of well-being. Secondly, well-being provides individual and collective benefits (e.g. reduced absences from work). Thirdly, studies demonstrate that the prevalence of optimal mental health is relatively low. Individuals’ subjective well-being, or the extent to which their lives are fulfilling (Diener et al., 2015), is similar to the concept of capabilities, focusing on core aspects of their welfare (Gloss et al., 2017). As with capabilities, subjective well-being can also be measured individually. However, Gloss et al. (2017, p. 352) assert that “capabilities appear to be more counterfactual and objective than subjective well-being, and capabilities tend to focus on particularistic evaluations as opposed to the more holistic evaluations of subjective well-being.” It is also possible to have a wide array of capabilities but to have a fundamental lack of subjective well-being or to lack capabilities but have high subjective well-being. In this regard, Sen (2009) argued that assessing individuals’ capabilities is not necessarily a good guide to their well-being. The reason is that agency and well-being, as well as freedom and achievement, are distinctive concepts that could be combined to yield four advantages: well-being achievement, agency achievement, well-being freedom and agency freedom.

Flourishing at work can be described as a pattern of positive emotions, behaviours and attitudes (Keyes and Annas, 2009; Noble and McGrath, 2018; Rothmann, 2013). According to Rothmann et al. (2019), flourishing means high levels of well-being. Schulte and Vainio (2010) define well-being as a summative concept that embodies the quality of working lives. It is a critical individual, organisational and societal productivity factor. Rothmann et al. (2019) conceptualised flourishing as a subjective state of emotional well-being (EWB), psychological well-being (PWB) and social well-being (SWB). EWB includes a cognitive component (the extent to which employees judge their jobs as favourable and satisfying) and an affective component (positive affect, which reflects the extent to which individuals’ needs are satisfied). PWB in work contexts includes autonomy, competence, relatedness, meaningful and purposeful work, engagement and learning (Rothmann, 2013). SWB is focused on social tasks that individuals encounter in work contexts. SWB includes five dimensions: social integration, social actualisation, social acceptance, social contribution and social coherence. Redelinghuys et al. (2019) found that teachers flourish when embedded within meaningful environments. De Wet and Rothmann (2022) found that specific work capabilities and the capability set of secondary school teachers in South Africa significantly predicted their EWB, PWB and SWB. The effect sizes of specific work capabilities were medium, whereas the effect sizes of the capability set were large. Notably, having a good income did not significantly predict the well-being dimensions.

Current study

Carr et al. (2016) and Sehnbruch et al. (2015) indicated that DW relates to capabilities and functionings as conceptualised in the CA. The concept of DW encompasses the availability of job opportunities and considers the aspects that people value and the conversion factors that affect their capabilities. The CA furthers an understanding of DW because it considers individuals’ work values, representing a subjective work experience and contextuality (i.e. individuals’ experiences of their work situation). CA is essential in research on DW because it includes personal and social factors that affect individuals’ capabilities to convert work characteristics into functioning. Additionally, the CA transforms the DW concept from a means-based to an ends-based concept, where individuals, rather than their income or work, become the ends (Sehnbruch et al., 2015). This study addresses the following question:

Q1.

What are the associations between DW, capabilities and well-being?

The following hypotheses are proposed for this study:

H1.

DW is positively associated with work capabilities.

H2.

DW is positively associated with employees’ EWB (a), PWB (b) and SWB (c).

H3.

Work capabilities is positively associated with employees’ EWB (a), PWB (b) and SWB (c).

Method

Setting and participants

Quality early childhood development (ECD) is essential to ensure the sustainable development of countries (Cumming and Wong, 2019). As a result of poverty, inequality, inadequate resources and inaccessibility, ECD programs remain inaccessible in South Africa (Visser et al., 2021). Total expenditure on ECD in 2020/2021 was equivalent to 0.5% of total public expenditure and 0.15% of Gross Domestic Product (Dulvy et al., 2022). The average annual salary of ECDPs is estimated at R31 000, which is five times less than that of Grade R educators (ECDPs in a preschool year). Seventy-eight per cent of ECDPs lack the appropriate training to teach at the preschool level. Job demands are high because of a high children-to-teacher ratio (1:31). Teachers must constantly deal with a lack of resources, poor infrastructure at centres and children who grew up in deprived circumstances (Baloyi and Makhubela, 2018). Additionally, poor management of government services exacerbates the problems related to early childhood education (Smit et al., 2021; Visser et al., 2021). The factors mentioned above necessitate research regarding the well-being of early childhood teachers (Cumming, 2017; Hall-Kenyon et al., 2014).

A relatively equal sample of participants was sourced from two provinces in South Africa (N = 436), namely, the North West (48.50%) and Gauteng provinces (51.10%), with 0.7% of the participants not indicating their province. Table 1 describes the participants of the study.

Table 1 shows that most participants were females (96.1%). The age of the participants ranged between 17 and 73 years. Most participants regarded themselves as full-time employed (83.95%); 6.19% were part-time employed; 5.50% regarded themselves as casual workers, whereas 3.21% were self-employed, and the missing values were approximately 1.15%. The tenure in their current position revealed that most participants (61.7%) worked at their current institution for less than five years, whereas approximately 39.7% indicated that they were employed at their current school for less than two years. Approximately 17.70% of the participants indicated that they had worked at their current workplace between 6 and 10 years, and 18.60% indicated that they had worked at their current place of work for more than 10 years. Participants with Grade 12 as their highest qualification were the largest group and made up 25.50% of the sample.

Measuring instruments

The following measuring instruments were used in this study:

The Decent Work Scale (DWS; Duffy et al., 2017) was used to measure DW. The DWS consists of 15 items and measures five dimensions: safe working conditions (e.g. “At work, I feel safe from emotional or verbal abuse of any kind”); access to health care (e.g. “My employer provides acceptable options for healthcare”); adequate compensation (“I am rewarded adequately for my work”); hours that allow for free time and rest (“I have no time to rest during the work week”); and organisational values that complement family and social values (“The values of my organisation match the values within my community”). The items are rated on a Likert-type scale varying from 1 (strongly disagree) to 7 (strongly agree). Duffy et al. (2017) established that the total score of the DWS correlated with work-related well-being measures. Buyukgoze-Kavas and Autin’s (2019) study supports the measurement invariance of the DWS for gender, income and social class groups. Masdonati et al. (2019) confirmed the validity and measurement invariance of the DWS for different language groups in Switzerland. The internal consistency for the total score was 0.90, whereas the internal consistency of the subscales ranged between 0.88 (safe conditions) and 0.97 (access to healthcare) (Duffy et al., 2017).

The Capability Set for Work Questionnaire (CSWQ; Abma et al., 2016) was used to assess the work values of participants, their enablement and their achievement concerning the following dimensions: use of knowledge and skills; development of knowledge and skills; involvement in important decisions; building and maintaining meaningful contacts at work; setting own goals; having a good income; and contributing to something valuable. For each aspect, it was determined whether or not it is part of a capability set, that is when the aspect is considered valuable, is enabled for work and is realised (for example, “It is important for me to be able to use my knowledge and skills in my work”; “I have enough opportunity in my work to use my knowledge and skills”; and “I regularly succeed in using my knowledge and skills in my work”). Response options range from 1 (not at all) to 5 (very much so). Studies have supported the reliability (α = 0.77; Murangi et al., 2022a) and the convergent, predictive and incremental validity of the CSWQ (Gürbüz et al., 2022).

The Flourishing at Work Scale – Short Form (FAWS-SF; Rautenbach and Rothmann, 2017) measured flourishing at work. The FAWS-SF consists of 19 items, rated on a scale varying from 1 (never) to 6 (every day). The participants must indicate how frequently they have experienced specific elements of flourishing in the last month. The scale contains three dimensions: EWB, PWB and SWB. EWB includes positive affect (three items) and job satisfaction (one item, e.g. “During the past month at work, how often did you experience real enjoyment in your work?”). PWB includes ten questions concerning six dimensions: autonomy, competence and relatedness satisfaction, learning, meaningful work and engagement (e.g. “During the past month at work, how often did you become enthusiastic about your job?”). SWB consists of five dimensions: social acceptance, actualisation, coherence, contribution and integration, measured by five items (e.g. “During the past month at work, how often did you feel you really belonged at this school?”). Acceptable reliabilities ranging from 0.92 for the full scale and 0.81–0.86 for the subscales of the FAWS-SF were found in a study of teachers in South Africa (Redelinghuys et al., 2019). Rautenbach and Rothmann (2017) found alpha coefficients for the FAWS-SF varying from 0.82 to 0.90.

Research procedure

The Economic and Management Sciences’ Ethics Committee at the North-West University provided ethics clearance for the study (Ethics Number: NWU-00918-21-A4). Permission for the study was granted by the Department of Basic Education in Gauteng and North West provinces. The questionnaire was finalised, and an expert meeting was convened. The researcher waited for permission from the gatekeepers before commencing the data collection phase. The researcher adhered to the provisions and requirements of the gatekeepers. The researcher contracted research assistants for data collection to overcome possible language barriers. The research assistants were trained and collected data with the researcher under her supervision. Data collection was conducted by distributing hard copies electronically where possible. Participation required the participants to give informed consent, and participants were given assurance that the data would be kept confidential and anonymous.

Statistical analyses

Data were analysed using Mplus 8.8 (Muthén and Muthén,1998-2017) and SPSS 28 (IBM Corp., 2021). Confirmatory factor analysis (CFA) was used to confirm the factor structures of the variables in this study. Maximum likelihood estimation with robust standard errors was used to conduct CFA. The following fit indices were used for assessing model fit: the chi-square statistic (a measure of absolute model fit), the standardised root mean residual (SRMR), the root mean square error of approximation (RMSEA), the Tucker-Lewis index (TLI) and the comparative fit index (CFI: West et al., 2023). The TLI and CFI values must be greater than 0.90; scores greater than 0.95 indicate excellent fit. Moreover, RMSEA and SRMR values less than 0.08 indicate a good fit between the model and the data (West et al., 2023). The Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to compare alternative measurement models. The AIC and BIC, with the lowest values, indicate the best-fitting model.

Latent class analysis (LCA) was used in Mplus 8.8 (Muthén and Muthén,1998/2017). Several models with increasing latent classes were tested. A model was retained when it significantly improved over the reference model. Models were compared using the AIC, BIC and adjusted Bayesian information criterion (ABIC) values, Lo-Mendell-Rubin (LMR LR) test (Lo et al., 2001), the adjusted LMR LR test (ALMR) and the bootstrapped likelihood ratio test (BLRT: Wang and Wang, 2020). Latent class analyses use entropy to assess classification quality; a value of 1 indicates a good classification – a value lower than 0.60 is unacceptable (Geiser, 2013).

Descriptive statistics and Pearson correlation coefficients were computed using SPSS 28 (IBM Corp., 2021). Scale reliability estimates were computed, and a cut-off value of 0.70 was used for scale reliability (Nunnally and Bernstein, 1994). Using the automatic Bolck, Croon and Hagenaars (BCH) method, the mean of distal continuous outcome was calculated across latent classes (Asparouhov and Muthén, 2014; Bakk and Vermunt, 2016).

Results

Capabilities of participants

The frequencies and percentages of capabilities of participants are reported in Table 2. Table 2 indicates that most participants regarded the values as important, with the results ranging between 67.7% (involvement in important decisions) and 89.1% (use of knowledge and skills). The results for enabling values ranged between 51.6% (involvement in important decisions) and 72.5% (building and maintaining meaningful relationships). The results regarding the level of achievement ranged between 40.1% (earning a good income) and 71.1% (building and maintaining meaningful relationships). The highest capabilities of ECDPs were: building and maintaining meaningful relationships (62.8%), setting own goals (61%) and use of knowledge and skills (56%). Table 2 shows that the lowest capabilities of ECDPs were: earning a good income (26.8%), involvement in important decisions (37.8%), contributing to the creation of something valuable (49.3%) and developing new knowledge and skills (52.3%).

Confirmatory factor analysis of the Decent Work Scale and Flourishing at Work Scale

CFA was used to test the measurement model of the DWS and the FAWS. The models were specified as follows: Model 1 consisted of one second-order factor (i.e. DW) and one first-order factor (i.e. flourishing). DW consisted of five (first-order) latent variables (each with three observed variables):

1. safe working conditions;

2. access to health care;

3. adequate compensation;

4. free time and rest; and

5. complementary values.

Flourishing at work (as a latent variable) consisted of 17 observed EWB, PWB and SWB variables. Model 2 consisted of one first-order factor (i.e. DW) and one second-order factor (i.e. flourishing at work). DW consisted of 15 observed variables concerning safe working conditions, access to healthcare, adequate compensation, free time and rest and complementary values. Flourishing at work (as a latent variable) consisted of three latent variables: EWB (four observed variables), PWB (nine observed variables) and SWB (five observed variables). Model 3 consisted of one latent variable indicated by 32 observed variables (i.e. all DW and flourishing at work items). Model 4 consisted of two higher-order factors: DW and flourishing at work. DW consisted of five latent variables (each with three observed variables):

1. safe working conditions;

2. access to health care;

3. adequate compensation;

4. free time and rest; and

5. complementary values.

Flourishing at work consisted of three latent variables: EWB (four observed variables), PWB (nine observed variables) and SWB (five observed variables).

Comparing the AIC and BIC values shows that Model 4a fits the data significantly better than Model 1 (ΔAIC = –480.04, ΔBIC = –427.11), Model 2 (ΔAIC = –1,862.16, ΔBIC = –1,772.16) and Model 3 (ΔAIC = –3,238.57, ΔBIC = –3,124.59). The fit statistics of Model 4a are as follows: χ² = 1,065.80, df = 467, p > 0.01; CFI = 0.89; TLI = 0.87; RMSEA = 0.05, p < 0.01 [0.05, 0.06]; and SRMR = 0.08. However, some of the fit statistics are lower than the recommended values (CFI < 0.90, TLI < 0.90) (Table 3).

Model development

Given the unacceptable fit of Model 4a, it was necessary to modify the model. Inspection of the factor loadings revealed that two items of the DWS, namely, DW9 (“I am rewarded adequately for my work”) and DW12 (“I have free time during the work week”), had low standardised regression coefficients (β = –0.15, p = 0.030 and β = 0.15, p = 0.019, respectively). Therefore, it was decided to respecify the model without items DW9 and DW12 (Model 4b). Comparing the AIC and BIC values shows that Model 4b fits the data significantly better than Model 4a (ΔAIC = –3,167.14, ΔBIC = –3,191.57). The fit statistics of Model 4b are as follows: χ² = 937.14, df = 406, p > 0.01; CFI = 0.90; TLI = 0.88; RMSEA = 0.05 [0.05, 0.06]; and SRMR = 0.08. However, the TLI of Model 4b is lower than the recommended value (<0.90). To identify other areas of model misfit, modification indices (MIs) were inspected. An SWB item, FAWS17 (“During the past month at work, how often did you feel you had something important to contribute to this organisation?”), also loaded on EWB (MI = 60.61) and PWB (MI = 91.73). It was decided to further modify the model by removing FAWS17 (Model 4c). Comparing the AIC and BIC values shows that Model 4c fits the data significantly better than Model 4b (ΔAIC = –1,396.38, ΔBIC = –1,408.60). The fit statistics of Model 4c are as follows: χ² = 761.78, df = 377, p < 0.001; AIC = 40,319.06; BIC = 40,799.4; CFI = 0.92; TLI = 0.91; RMSEA = 0.05 [0.04, 0.05]; and SRMR = 0.06.

The factor loadings of the items were as follows: DWS – safe working conditions (3 items): λ = 0.65–0.79, mean = 0.73; DWS – access to health care (3 items): λ = 0.88–0.94, mean = 0.91; DWS – adequate compensation (2 items): λ = 0.84–0.90, mean = 0.87; DWS – free time and rest (2 items): λ = 0.79–0.84, mean = 0.82; DWS – complementary values (3 items): λ = 0.86–0.96, mean = 0.91; FAWS – EWB (4 items): λ = 0.57–0.76, mean = 0.69; FAWS – PWB (9 items): λ = 0.53–0.78, mean = 0.67; FAWS – SWB (4 items): 0.56–0.89, mean = 0.78.

Descriptive statistics, reliabilities and correlations of the scales

Table 4 shows the descriptive statistics, reliabilities and Pearson’s correlations of the variables. Table 4 indicates that the reliability coefficients for the three scales used in this study are acceptable at ω > 0.70 (Nunnally and Bernstein, 1994).

Concerning the distribution of DW scores, Table 4 shows that the mean score for safe work conditions is the highest, whereas mean scores for access to health care, free time and rest are the lowest. Concerning flourishing at work, the mean score for PWB is the highest, whereas the mean scores for EWB and SWB are lower.

Two DW dimensions: safe work conditions and complementary values (both medium effects), are statistically significantly related to the capability set. The capability set is significantly related to EWB, PWB and SWB (all medium effects). Regarding DW and well-being (flourishing at work), all the DW factors, except adequate compensation, correlate positively and significantly with EWB, PWB and SWB. The adequate compensation variable is positively and significantly related only to SWB (small effect). Therefore, H2a, H2b and H2c are partially accepted.

Latent class analysis

A LCA with Mplus 8.8 was applied to the seven capabilities. Different models with an increasing number of latent classes were tested. The results of the different models are reported in Table 4.

The results in Table 5 show that the two-class solution has a better fit than the one-class solution (ΔAIC = –623.38, ΔBIC = –590.86 and ΔABIC = –616.24). The three-class solution is significantly better than the two-class solution (ΔAIC = –45.92; ΔBIC = –13.39; ΔABIC = –38.77). Furthermore, the four-class solution shows a better fit than the three-class solution (ΔAIC = –35.24; ΔBIC = –2.71; ΔABIC = –28.11). However, the five-class solution was not better than the four-class solution (ΔAIC = 3.26; ΔBIC = 35.79; ΔABIC = 10.40). The LMR LR (p < 0.01), ALMR LR (p < 0.01) and BLRT (p < 0.01) of Class 2 were statistically significant. The comparison of Classes 1 and 2 shows that the Class 2 model fits the data significantly better. Although the LMR LR and ALMR LR values for Class 3 are not statistically significant, the BLRT is statistically significant (p < 0.01). At the same time, all these indexes are statistically significant (p < 0.01) for the four-class solution.

The class proportions were acceptable for Classes 1, 2, 3 and 4. The entropy value of 0.76 was acceptable. The average latent class probabilities were also acceptable: Class 1 = 0.85, Class 2 = 0.73, Class 3 = 0.91 and Class 4 = 0.92. The entropy value was 0.76, indicating an acceptable classification (Wang and Wang, 2020). The bivariate standardised residuals were not statistically significant, indicating that the assumption of local independence within classes was met (Sterba, 2013). The four classes are illustrated in Figure 1 .

The four classes in Figure 1 can be described as follows:

1. Latent Class 1 – robust capability set (33.18%). Participants in this class show high probabilities of endorsement of capabilities, except for earning a good income.

2. Latent Class 2 – knowledge/skills capability set (17.17%). Participants in this class show a high probability of endorsing two capabilities: using knowledge and skills and developing new knowledge and skills. Participants in this class show a moderate probability of endorsing building and maintaining meaningful work relations but a low probability of endorsing the other four capabilities.

3. Latent Class 3 – weak capability set (30.39%). This class shows a low probability of endorsement of the seven capabilities.

4. Latent Class 4 – relational capability set (19.26%). This class is highly likely to endorse setting goals, building and maintaining meaningful relationships at work and contributing to something valuable. The class shows low probabilities of endorsing the other four capabilities.

Latent classes and distal outcomes

This study calculated the mean of a continuous distal outcome across multiple latent classes by applying the automatic BCH approach (Asparouhov and Muthén, 2014; Bakk and Vermunt, 2016). The BCH method was used to analyse the auxiliary variables to estimate their means across the latent classes. The differences between the distal variables of the four different capability classes are displayed in Figure 1.

The results in Table 6 show that statistically significant differences existed between the dimensions of the different capability classes: DWS – safe working conditions (χ² = 72.85, p < 0.001); DWS – sufficient to access health care (χ² = 22.57, p < 0.001); DWS – sufficient free time (χ² = 10.62, p = 0.014); DWS – low complementary values (χ² = 72.86, p < 0.001); FAWS – EWB (χ² = 87.94, p < 0.001); FAWS – PWB (χ² = 56.44, p < 0.001) and FAWS – SWB (χ² = 63.56, p < 0.001).

Table 6 shows that regarding safe working conditions, individuals with a robust capability set scored significantly higher on safe working conditions than those with relational, knowledge/skills and weak capability sets. Consequently, individuals with a relational capability set scored significantly higher on safe working conditions than those with a weak capability set. Concerning sufficient access to health care, practitioners with a robust capability set scored significantly higher than those with knowledge/skills and weak capability sets, whereas individuals with a relational capability set scored significantly higher on access to healthcare than those with a weak capability set. Regarding sufficient free time, practitioners with a robust capability set scored statistically significantly higher than those with knowledge/skills and weak sets. Concerning complementary values, individuals with a robust capability set scored significantly higher on complementary values than those with relational knowledge/skills and weak capability sets. Practitioners with relational and knowledge/skills capability sets scored significantly higher on complementary values than those with a weak capability set. Therefore, H1 is accepted.

Concerning well-being, Table 6 shows that individuals with a robust capability set scored statistically significantly higher on EWB, PWB and SWB than those with relational, knowledge/skills and weak capability sets. Practitioners with relational and knowledge/skills sets scored significantly higher on EWB than the weak capability set. Moreover, individuals with a relational capability set scored significantly higher on PWB than those with a weak capability set. Finally, practitioners with relational and knowledge/skills sets scored significantly higher on SWB than those with a weak capability set. Therefore, H3a, H3b and H3c are partially accepted.

Discussion

This study investigated the association between DW, capabilities and the flourishing of employees in a South African context. ECD practitioners were the focus of this study because they face stressful work contexts affected by poverty and inequality. While most participating ECDPs perceived all seven work values as important, they reported a lack of enablement of various work values, notably earning a good income and involvement in important decisions. A similar pattern was observed regarding the achievement of work values.

The results show that those participants who felt they were enabled to achieve values ranged between 34.9% (earning a good income) and 72.5% (building and maintaining meaningful relationships). A total of 84.9% of the participants valued earning a good income, but only 34.9% felt that they were enabled to earn a good income. Earnings are not the only determinant of well-being; however, it influences employee dissatisfaction as income impacts their livelihood and opportunities, i.e. realised capabilities (Frediani, 2010). Regarding the level of achievement, the percentage of individuals that felt they achieved capabilities ranged between 40.1% (earning a good income) and 71.1% (building and maintaining meaningful relationships).

The strongest capabilities were found to be building and maintaining meaningful relationships (62.8%), setting own goals (61%), using knowledge and skills (56%) and developing new knowledge and skills (52.3%). The weakest capabilities were earning a good income (26.8%), involvement in important decisions (37.8%) and contributing to the creation of something valuable (49.3%). The importance of capabilities vastly exceeded the opportunity and achievement of capabilities. These findings align with previous studies’ results (Abma et al., 2016; De Wet and Rothmann, 2022).

Table 7 shows the capabilities of ECDPs compared to employees in three other jobs in South Africa and Namibia (Barnard et al., 2023; De Wet and Rothmann, 2022; Murangi et al., 2022a).

Table 7 shows that ECDPs and emergency nurses showed two low capabilities: earning a good income and being involved in important decisions. Also, ECDPs had lower capabilities regarding using knowledge and skills and developing new knowledge and skills. People must be above a threshold on all capabilities to optimise their potential (Nussbaum, 2011; Prilleltensky and Prilleltensky, 2021). When individuals are below a threshold on a particular capability, they might find it difficult to achieve other capabilities (Mullainathan and Shafir, 2014). For example, a poor capability to earn a good income may affect the capability to develop new knowledge and skills. Likewise, a low capability to be involved in important decisions may affect capabilities such as developing new knowledge and skills and contributing to something valuable, resulting in poor well-being. However, lacking or abundant capabilities might disrupt or compensate for other capabilities (Prilleltensky and Prilleltensky, 2021).

This study established four latent classes for capability: robust, knowledge/skills, weak and relational. Firstly, individuals with a robust capability set (33.18%) showed a high likelihood of endorsement for six of the seven capabilities; only one capability, earning a good income, had a lower probability of endorsement. Earning a good income is a global problem in the ECD environment (Baloyi and Makhubela, 2018). Secondly, the knowledge/skills capability set (17.17%) was likely to endorse using knowledge and skills and the development of knowledge and skills, with a moderate likelihood of building meaningful relationships at work and low capabilities to contribute something valuable, set own goals, be involved in important decisions and earn a good income. Thirdly, the relational capability set showed a high likelihood of endorsement for setting goals, building meaningful relations and creating something valuable, along with a low chance of endorsement for using knowledge and skills, developing knowledge and skills, involvement in important decisions and earning a good income. Finally, employees with a weak capability set showed a low probability of endorsing all seven capabilities. A concern is that about 30.39% of ECDPs measured low on all seven capabilities, negatively affecting their sustainable employability (van der Klink et al., 2016).

Regarding DW and work capabilities, our results showed that employees with a robust capability set (compared to the other three capability sets) were significantly more inclined to report safe working conditions and organisational values that complement family and social values. These employees were also significantly more inclined to report sufficient access to health care and sufficient free time than those with knowledge/skills and weak capability sets.

Employees with a relational capability set (compared to those with a weak capability set) were significantly more inclined to report safe working conditions, access to health care and organisational values that complement family and social values. Employees with a knowledge/skills capability set were significantly more inclined to report that the organisational values complement their family and social values. In line with PWT (Blustein, 2019; Duffy, 2020), DW is associated with work capabilities of individuals.

Concerning well-being, DW (except for adequate compensation) was positively related to EWB, PWB and SWB (Blustein, 2019; Duffy et al., 2016). Considering the strength of relationships, safe work conditions and organisational values that complement family and social values of employees has the strongest effects. Interestingly, these two dimensions of DW also had the strongest effects of the capability set. Therefore, the results support the relationships between DW (especially safe work conditions and organisational values that complement family and social values of employees), the capability set and well-being.

Employees with a robust capability set showed significantly higher EWB, PWB and SWB levels than those with relational, knowledge/skills and weak capability sets. Employees with a relational capability set scored significantly higher on EWB, PWB and SWB than those with a weak capability set. Employees with a knowledge/skills capability set scored significantly higher on EWB and SWB than those with a weak capability set. These findings align with findings that employees flourish when engaging in a meaningful environment (Redelinghuys et al., 2019) and findings that work capabilities predict well-being at work (Barnard et al., 2023; De Wet and Rothmann, 2022; Murangi et al., 2022a).

Research by Murangi et al. (2022b) showed that the capability set of employees is related to meaningful work (which is also a dimension of flourishing work). Sayer (2009) argued that inequalities in the availability of meaningful work limit individuals’ opportunities, such as the extent to which they can develop their knowledge and skills and find fulfilment, respect and self-esteem and, therefore, mattering (Prilleltensky and Prilleltensky, 2021). In this regard, Sayer (2009) argued that thinking about justice is overwhelmingly distributive (i.e. concerned with what people get in terms of resources and opportunities) and that considering contributive justice (i.e. what people are expected and able to contribute) is essential.

Employees with a weak capability set are vulnerable, and their preference may highlight the dire need to increase investment, expand support and create DW conditions. This outcome indicates a need to scrutinise and evaluate this vulnerable group’s DW conditions and well-being. A better understanding of the DW conditions will enhance the enablement and achievement of employees and develop sustainable employability and flourishing at work (Orton, 2011; Pouyaud, 2016; van der Klink et al., 2016; van der Klink, 2019). On the one hand, the results showed that DW is associated with attaining capabilities for employees and enabling their flourishing in the workplace. On the other hand, the weak capability set shows how the lack of DW conditions can affect capability and result in employees not functioning optimally in their workplace.

Our results suggest that to reduce inequalities in DW and capabilities and to promote well-being, policies and interventions should focus on DW and the capabilities of people in disadvantaged positions. These together would strengthen their agency for converting capabilities into well-being (Mäki-Opas et al., 2022). The need for general support of disadvantaged people should also be balanced with the need to respond to their diverse capabilities.

Limitations and recommendations for future research

This study had various limitations. Firstly, a convenience sample was used, and the data were only collected in two South African provinces. However, the findings confirm findings in other WEIRD and POSH contexts (Barnard et al., 2023; De Wet and Rothmann, 2022; Murangi et al., 2022a). Secondly, a cross-sectional survey was used, and a longitudinal study is recommended for future studies as it may provide more information on employees’ capabilities and DW conditions. Thirdly, a quantitative approach was used in this study. Future studies could benefit from qualitative and mixed-methods approaches. Future studies should consider differentiating the capabilities of employees based on demographic variables. Future studies that use a quantitative research design and embark on LCA should follow up with interviews to verify the description and details of each capability set.

Conclusions

This study investigated DW and capability sets’ influence on workplace flourishing. Four distinctive capability classes were derived from employees who participated in the study. These classes provide greater insight into the capabilities of employees in challenging contexts. The results showed that the capability sets better indicate the functioning of employees than individual capabilities. It also revealed that DW conditions impact the capability sets and effectively influence the employees’ ability to flourish in the workplace. The quality of teaching and sustainable employability of employees in South Africa may be compromised due to the absence, in most cases, of DW conditions, and in turn, this may cause a lack of work capabilities, such as earning a good income, participating in important decisions, contributing to something meaningful and developing new skills.

Figures

Figure 1

Latent profiles of capability sets of ECDPs

Table 1

Characteristics of the participants (N = 436)

Variable	Grouping	n	%
Gender	Female	419	96.10
	Male	4	0.90
	Missing values	13	3.00
Province	North West	210	48.20
	Gauteng	223	51.10
	Missing values	3	0.70
Race	Black	285	65.40
	White	103	23.60
	Coloured	39	8.90
	Indian	2	0.50
	Other	4	0.90
	Missing values	3	0.70
Tenure at current school	Less than 6 months	62	14.20
	6 months to 2 years	111	25.50
	3–5 years	96	22.00
	6–10 years	77	17.70
	More than 10 years	81	18.60
	Missing values	9	2.10
Highest qualification	Lower than Grade 12	65	14.90
	Grade 12	111	25.50
	1-year vocation cert.	44	10.10
	Diploma/degree (or higher)	153	35.10
	Other	56	12.80
	Missing values	7	1.60
Union	Yes	98	22.5
	No	329	75.5
	Missing values	9	2.1
Age	17–19	6	1.38
	20–29	83	19.04
	30–39	96	22.02
	40–49	83	19.04
	50–59	63	14.45
	60–65	10	2.29
	Older than 65	2	0.46
	Missing value	93	21.33
Employment status	Casual work	24	5.50
	Full-time	366	83.95
	Part-time	27	6.19
	Self-employed	14	3.21
	Missing value	5	1.15
Job level	Top management	23	5.30
	Supervisor	14	3.20
	ECD practitioner	249	57.10
	Assistant ECD practitioner	78	17.90
	ECD Admin	17	3.90
	Other	50	11.50
Institution type	Public School	160	36.70
	Private School	164	37.60
	Non-government organisation	69	15.80
	Other	35	8.00
	Missing values	8	1.80

Source: Authors’ own work

Table 2

Capabilities of participants

Capability	Code	Importance	Opportunity	Achievement	Capability	Combined
Use knowledge and skills	0	8.9	30.0	29.1	42.6	Not capable
	1	89.1	68.8	68.8	56.0	Capable
	MV	1.4	1.1	2.1	1.4
Develop new knowledge and skills	0	13.3	32.1	31.9	45.9	Not capable
	1	85.1	66.3	67.0	52.5	Capable
	MV	1.6	1.6	1.1	1.6
Involvement in important decisions	0	30.5	46.6	49.5	60.4	Not capable
	1	67.7	51.6	48.6	37.8	Capable
	MV	1.8	1.6	1.8	1.8
Build and maintain meaningful relationships	0	11.5	26.1	26.1	35.6	Not capable
	1	86.9	72.5	71.1	62.8	Capable
	MV	1.6	1.4	2.8	1.6
Set own goals	0	10.6	26.8	27.1	37.4	Not capable
	1	87.8	71.6	70.6	61.0	Capable
	MV	1.6	1.6	2.3	1.6
Earn a good income	0	13.5	63.5	57.8	71.6	Not capable
	1	84.9	34.9	40.1	26.8	Capable
	MV	1.6	1.6	2.1	1.6
Contribute to something valuable	0	17.7	36.9	37.4	49.1	Not capable
	1	80.7	61.2	60.8	49.3	Capable
	MV	1.6	1.8	1.8	1.6

Notes:

Code = 0 if the participant’s score was ≤ 3; Code = 1 if the participant’s score ≥ 4; MV = missing value

Source: Authors’ own work

Table 3

CFA of the DWS and the FAWS

Model	χ2	df	AIC	BIC	CFI	TLI	RMSEA	SRMR
1	1,464.45*	480	45,362.62	45,826.68	0.81	0.80	0.07* [0.06, 0.07]	0.08
2	2,520.47*	489	46,744.74	47,172.17	0.62	0.59	0.09* [0.09, 0.10]	0.12
3	3,534.31*	495	48,121.15	48,524.16	0.43	0.39	0.12* [0.12, 0.12]	0.11
4a	1,065.80*	467	44,882.58	45,399.57	0.89	0.87	0.05* [0.05, 0.06]	0.08
4b	937.14*	406	41,715.44	42,208.01	0.90	0.88	0.05* [0.05, 0.06]	0.08
4c	761.78*	377	40,319.06	40,799.41	0.92	0.91	0.05 [0.04, 0.05]	0.06

Notes:

*p < 0.01; χ²: chi-square statistic; df: degrees of freedom; AIC: Akaike information criterion; BIC: Bayesian information criterion; CFI: comparative fit index; TLI: Tucker–Lewis index; RMSEA: root mean square error of approximation; SRMR: standardised root mean square residual

Source: Authors’ own work

Table 4

Descriptive statistics, reliabilities and correlations of the scales

Variable	ω	Mean	SD	1	2	3	4	5	6	7	8
1. DW: SW	0.77	5.48	1.28	–	–	–	–	–	–	–	–
2. DW: AH	0.93	3.67	2.11	0.33**	–	–	–	–	–	–	–
3. DW: AC	0.86	4.85	1.60	–0.07	0.26**	–	–	–	–	–	–
4. DW: FT	0.80	3.89	1.48	0.20**	0.04	0.35**	–	–	–	–	–
5. DW: CV	0.94	4.93	1.76	0.51**	0.30**	0.05	0.02	–	–	–	–
6. Capset	0.81	0.50	0.33	0.38**	0.22**	0.09	0.16**	0.40**		–	–
7. EWB	0.77	4.75	1.05	0.57**	0.24**	– 0.07	0.17**	0.47**	0.45**	–	–
8. PWB	0.87	5.20	0.84	0.46**	0.21**	−0.02	0.17**	0.31**	0.36**	0.74**	–
9. SWB	0.86	4.76	1.23	0.5**	0.39**	0.13**	0.15**	0.47**	0.40**	0.63**	0.57**

Notes:

DW = decent work; SW = safe work conditions; AH = access to health care; AC = adequate compensation; FT = free time and rest; CV = complementary values; EWB = emotional well-being; PWB = psychological well-being; SWB = social well-being, **p < 0.01; *p < 0.05; r < 0.30 = small effect; 0.30 < r < 0.50 = medium effect; r > 0.50 = large effect

Source: Authors’ own work

Table 5

Comparison of capability classes

Class	AIC	BIC	ABIC	LMR LR test p-value	ALMR LR test p-value	BLRT p-value	Entropy	Smallest class membership
1-Class	3,984.05	4,012.52	3,990.30	n/a	n/a	n/a	–	–
2-Class	3,360.67	3,421.66	3,374.06	0.001**	0.001**	0.001**	0.82	197
3-Class	3,314.75	3,408.27	3,335.29	0.233	0.238	0.001**	0.72	134
4-Class	3,279.51	3,405.56	3,307.18	0.002**	0.002**	0.001**	0.76	74
5-Class	3,282.77	3,441.35	3,317.58	0.290	0.298	0.429	0.78	52

Notes: AIC = Akaike information criterion; BIC = Bayesian information criterion; ABIC = adjusted Bayesian information criterion; LMR LR = Lo-Mendell-Rubin test; ALMR LR = adjusted Lo-Mendell-Rubin test; BLRT = bootstrapped likelihood ratio test; **p < 0.01; *p < 0.05

Source: Authors’ own work

Table 6

Equality tests of means across latent profiles of ECDPs

Variable	Mean	SE
DW safe working conditions
Robust set	0.56	0.08
Knowledge/skills set	–0.27	0.15
Weak set	–0.47	0.10
Relational set	0.13	0.13
Chi-square tests	χ2	p
Overall test	72.85	0.001**
Robust vs weak	66.80	0.001**
Knowledge/skills vs weak	1.17	0.279
Weak vs relational	11.39	0.001**
Robust vs knowledge/skills	21.60	0.001**
Robust vs relational	7.413	0.001**
Knowledge/skills vs relational	3.407	0.065
DW adequate compensation
Robust set	0.12	0.13
Knowledge/skills set	–0.09	0.19
Weak set	–0.12	0.12
Relational set	0.07	0.21
Chi-square tests	χ2	p
Overall test	1.95	0.582
Robust vs weak	1.74	0.187
Knowledge/skills vs weak	0.01	0.923
Weak vs relational	0.51	0.477
Robust vs knowledge/skills	0.74	0.389
Robust vs Relational	0.05	0.825
Knowledge/skills vs relational	0.27	0.602
DW complementary values
Robust set	0.93	0.13
Knowledge/skills set	–0.19	0.25
Weak set	–0.82	0.17
Relational set	–0.04	0.23
Chi-square tests	χ2	p
Overall test	72.86	0.001**
Robust vs weak	68.77	0.001**
Knowledge/skills vs weak	4.10	0.043*
Weak vs relational	6.59	0.010**
Robust vs knowledge/skills	13.90	0.001**
Robust vs relational	12.38	0.001**
Knowledge/skills vs relational	0.16	0.688
Psychological well-being
Robust set	0.35	0.05
Knowledge/skills set	–0.12	0.12
Weak set	–0.37	0.09
Relational set	0.14	0.09
Chi-square tests	χ2	p
Overall test	56.44	0.001**
Robust vs weak	51.98	0.001**
Knowledge/skills vs weak	2.73	0.098
Weak vs relational	15.14	0.001**
Robust vs knowledge/skills	12.02	0.001**
Robust vs relational	4.36	0.037*
Knowledge/skills vs relational	2.75	0.097
DW sufficient access to health care
Robust set	0.60	0.19
Knowledge/skills set	–0.30	0.28
Weak set	–0.56	0.17
Relational set	0.24	0.29
	χ2	p
Overall test	22.57	0.000**
Robust vs weak	21.52	0.000**
Knowledge/Skills vs weak	0.57	0.452
Weak vs relational	5.06	0.024*
Robust vs knowledge/skills	6.23	0.013**
Robust vs relational	1.01	0.314
Knowledge/skills vs relational	1.54	0.214
DW sufficient free time
Robust set	0.34	0.15
Knowledge/skills set	–0.35	0.22
Weak set	–0.24	0.13
Relational set	0.15	0.21
	χ2	p
Overall test	10.62	0.014**
Robust vs weak	8.47	0.004**
Knowledge/skills vs weak	0.17	0.678
Weak vs relational	2.21	0.137
Robust vs knowledge/skills	5.91	0.005**
Robust vs relational	0.49	0.483
Knowledge/skills vs relational	2.36	0.124
Emotional well-being
Robust set	0.51	0.06
Knowledge/skills set	–0.05	0.13
Weak capability set	–0.52	0.10
Relational capability set	0.05	0.11
	χ2	p
Overall test	87.94	0.001**
Robust vs weak	82.96	0.001**
Knowledge/skills vs weak	8.15	0.004**
Weak vs relational	13.44	0.001**
Robust vs knowledge/skills	14.04	0.001**
Robust vs relational	12.24	0.001**
Knowledge/skills vs relational	0.32	0.570
Social well-being
Robust set	0.37	0.06
Knowledge/skills set	–0.07	0.11
Weak set	–0.38	0.08
Relational set	0.05	0.11
	χ2	p
Overall test	63.56	0.001**
Robust vs weak	60.99	0.001**
Knowledge/skills vs weak	4.96	0.026*
Weak vs relational	9.45	0.002**
Robust vs knowledge/skills	11.25	0.001**
Robust vs relational	6.63	0.010**
Knowledge/skills vs relational	0.53	0.466

Notes:

*p < 0.05; **p < 0.01; DW = decent work

Source: Authors’ own work

Table 7

Percentages of capabilities in different occupations

Cap	Emergency nurses (N = 204)	Special education teachers (N = 200)	Secondary school educators (N = 144)	ECDPs (N = 436)
UKS	67.7	66.0	78.7	56.0
DKS	62.8	63.0	69.5	52.5
IID	36.8	52.0	54.3	37.8
MRW	60.3	64.5	75.0	62.8
SOG	53.9	67.0	73.0	71.6
EGI	23.0	45.0	50.4	26.8
CSV	45.1	62.0	73.0	49.3

Notes:

UKS = using knowledge and skills; DKS = developing knowledge and skills; IID = involvement in important decisions; MRW = meaningful relationships at work; SOG = setting own goals; EGI = earning a good income; CSV = contributing to something valuable

Source: Authors’ own work