City Research Online

Abstract

Rationale and Aims: Standard analyses of the labour market tend to focus more on job quantity than work and employment (QWE), however, there is increasing interest at national and international levels to highlight QWE as a salient social and labour market policy issue. This, though, has been hindered by a lack of consensus on a definition, and inherent challenges in the conceptualisation and operationalisation of QWE, evidenced by the substantial number of measurement instruments in the literature. Although there is consensus about the concept being multidimensional, it cannot be directly measured, and there is no consensus on what attributes to include in the measure, how these should be aggregated and weighted, or whether to report overall QWE and/or different dimensions. Furthermore, the measurement of QWE is limited by the lack of a single source of data capturing all the relevant attributes. At the same time, there is also a lack of evaluation of the measurement equivalence of the instruments, which is a prerequisite for between-group comparisons. This study aimed to make theoretical contributions to the conceptualisation of QWE, as well as methodological contributions in the measurement of QWE. The study also sought to make substantive contributions by investigating how overall and/or dimensions of QWE varied by demographic, socio-demographic, and socio-economic characteristics in the UK employee population.

Methods: The study used data from the UK Household Longitudinal Study (UKHLS) and applied item response theory (IRT) modelling to develop a measure of QWE. Competing measurement models (unidimensional, correlated-factors, second-order factor, and bifactor models) were estimated and compared, and psychometric properties of the measurement instruments were examined. Differential item functioning (DIF) was used to evaluate the measurement equivalence of the instruments between different groups. Multiple group analysis and multiple indicators multiple causes (MIMIC) models were used to investigate the effects of the demographic, socio-demographic, and socio-economic characteristics on overall QWE and other dimensions of QWE.

Results: The theoretical framework of QWE consisted of six dimensions: economic compensation, training and progression, employment security, working conditions, work-life balance (or work-time scheduling), and social dialogue. The social dialogue dimension was measured by a single item and in the subsequent IRT modelling, responses to this item were not sufficiently explained by the model resulting in the exclusion of the item. Results suggested that the measurement of QWE was better represented by a bifactor model. An evaluation of the psychometric properties of the measure of QWE suggested that training and progression and employment security were not a good representation of these latent traits and were excluded in subsequent analysis. DIF analyses indicated that while some items measuring overall and other dimensions of QWE exhibited differential performance between some groups, the magnitude of DIF was negligible and between-group comparison was feasible.

Substantively, results from the study suggested that demographic or socio-demographic characteristics did not explain much of the variation in overall or other dimensions of QWE, with the effect sizes either small or negligible. In contrast, socio-economic characteristics explained more of the variation in the latent traits. The bifactor model also provided a more nuanced understanding of differences in QWE between some groups that would otherwise not be feasible with other methods. The study found that females had better working conditions than males, while younger employees were more aware of and had better access to other forms of work-time scheduling than older employees. Results suggested that employees with longstanding illnesses or disability had poorer economic compensation and working conditions but were more aware of and had better access to forms of work-time scheduling than those without a longstanding illness or disability, while there were no differences in overall QWE. The study highlighted longstanding regional disparities in the labour market with advantages for those in the London and Southern England regions. However, it also highlighted better working conditions and more awareness of and better access to other forms of work-time scheduling, along with comparable economic compensation for employees in Scotland relative to those in London, but employees in London had better overall QWE than those in Scotland. Results supported evidence of better outcomes for employees in public sector organisations than those in private sector organisations. Employees in public sector organisations had better economic compensation, working conditions, and more awareness of and better access to forms of work-time scheduling but poorer overall QWE than those in private sector organisations.

Conclusions: The study contributed to the conceptualisation of QWE by developing a theoretical framework for measuring QWE and made methodological contributions by applying IRT modelling to address some of the shortcomings of existing measures. These shortcomings related to the aggregation and weighting of items measuring QWE, including whether to report overall and/or other dimensions of QWE, as well as evaluating measurement equivalence of the instrument. The study presented new knowledge that suggested the measurement of QWE was better represented by a bifactor model. The study also made some substantive contributions, which suggested that socio-economic characteristics explained more of the variation in overall or other dimensions of QWE than demographic or socio-demographic characteristics. While results from IRT modelling largely replicated those of other methods, there were some discrepancies, and the bifactor model provided a more nuanced understanding of differences in QWE between some groups.

Publication Type:	Thesis (Doctoral)
Subjects:	H Social Sciences > HD Industries. Land use. Labor
Departments:	School of Policy & Global Affairs > School of Policy & Global Affairs Doctoral Theses School of Policy & Global Affairs > Sociology & Criminology Doctoral Theses

Export

Downloads