Back to search

University of Glasgow

A system dynamics model of the labour market, gender norms, and mental health in Scotland

Abstract

dc:description.abstract

Background: The bidirectional relationship between the labour market and mental health is well established. Yet this evidence derives largely from individual risk factors studied in isolation; how these pathways aggregate into a population's prevalence of poor mental health over decades is far less understood. In Scotland, the prevalence of poor mental health among the working-age population has risen over recent decades, and women's prevalence has remained consistently higher than men's despite progress on gender equality indicators such as the employment rate and the gender pay gap. Understanding this trajectory, this thesis argues, requires focusing on the population-level interaction between these dynamic conditions: the gendered structure of work (the gender pay gap and unpaid work); the labour market (insecure work and unemployment); and population mental health. This raises three questions for Scotland: how reciprocal feedback between labour market conditions and population mental health has contributed to the rise between 1992 and 2019; why the gender gap has persisted; and which mechanisms dominate the system and can serve as leverage points for policy. Methods: The hypothesised causal structure, a system of interacting feedback loops, was assembled through an iterative literature review, in which identifying one causal relationship triggered a search for its reverse effect and any intermediate variables enabling the mapping of structure. The synthesis was expressed as a causal loop diagram, a system dynamics tool for capturing feedback mechanisms. It was refined through disconfirmatory interviews with ten interdisciplinary experts, designed to challenge the structure. Grounded in a complex realist philosophy that pursues explanatory verisimilitude rather than prediction, this conceptual model was then translated into a system dynamics simulation model to test whether the hypothesised structure could reproduce the trends observed between 1992–2019. Parameterisation prioritised published evidence and data, and the model was validated through structural and behavioural tests and sensitivity analyses. It was then analysed to identify which feedback loops drive behaviour over time, key contributors to the gender gap, and the policy combinations that most improve outcomes. Results: The model reproduced the observed trends endogenously, indicating that the structure can generate the observed trajectory. However, the magnitude of the rise in women’s prevalence post-2016 is lower than observed, pointing to forces beyond the modelled structure, notably the welfare system. The rise in poor mental health owed less to the demand shock of the 2008 recession than to the gradual rise of insecure work that preceded it. This rise quietly eroded the system's capacity to move people out of poor mental health. Consistent with this, counterfactual analysis found that holding insecure work at its 1992 level reduced the increase in poor mental health by more than removing the demand shock of the 2008 recession. The persistent gender gap is primarily driven by higher incidence among women, arising from their structural position, operationalised in the model through the gender pay gap and unpaid work. Structural decomposition of the gender gap attributed a smaller part to women's slower recovery. Levers focused on equality narrow the gender gap in prevalence, though childcare and norms levers do so partly by raising men's prevalence. This highlights a redistributive trap. Reducing insecure work lowers prevalence for both women and men but can widen the gap, because, given how the model specifies the effect of insecure work on recovery, men benefit more. The levers therefore act best in combination; expanding childcare alongside reducing insecure work reproduces almost the whole effect that deploying all four levers achieves. Discussion: The thesis makes four contributions. The first is a dynamic, population-level causal structure that combines health causation and health selection within one feedback structure and decouples incidence from recovery. The second is treating gender as a dynamic structural determinant rather than a static covariate. The third is the model as an explanatory thinking tool that highlights key structural stories such as the buffering role of labour-force exit at the population level. The fourth is methodological: the thesis demonstrates system dynamics applied to the upstream social determinants of mental health, where such models remain rare, and offers the feedback-driven review for future researchers to build on. For understanding population mental health, the modelling framework offers a structural explanation of how the gender gap can persist despite progress on equality indicators. For decision-making, it suggests asking whether a policy acts on incidence or recovery, and when, in what combination of levers, and in what macroeconomic state. Its policy relevance lies in framing poor mental health as a population-level system of interacting feedbacks: labour market, gender equality and childcare policy are core to preventing poor mental health, and they are better judged together.

Degree

thesis:*
Level dc:type.qualificationlevel
PhD
Grantor dc:publisher.institution
University of Glasgow
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Sonthalia, Shreya

Subjects

dc:subject × 2

Rights

Language dc:language
en

Chain of custody

source
Harvested from
University of Glasgow
Base URL
theses.gla.ac.uk/cgi/oai2
Last updated
2026-11-21
Source record
OAI-PMH GetRecord
citation

Sonthalia, Shreya. A system dynamics model of the labour market, gender norms, and mental health in Scotland. PhD thesis, University of Glasgow, 2026.