Research Proposal – Managing Psychosocial Hazards Under Australia’s WHS Duties

Introduction

Australian work health and safety (WHS) law has shifted decisively towards the prevention of psychological harm. Between 2022 and 2023, most Australian jurisdictions adopted amendments to the model WHS Regulations that place explicit, auditable duties on persons conducting a business or undertaking (PCBUs) to identify psychosocial hazards, eliminate or minimise the associated risks so far as is reasonably practicable, and review control measures (Safe Work Australia, 2022). The reform moved psychological safety from voluntary wellbeing initiatives into the same regulatory architecture that governs plant, chemicals and falls from height. What remains unresolved is whether this change in legal text has produced a change in organisational practice detectable in the conditions workers experience.

This proposal outlines a 12-month mixed methods study across six Australian organisations, examining how duty holders translate the amended psychosocial regulations into practice and whether the quality of that translation is associated with measurable differences in worker exposure and psychological outcomes, producing evidence for regulators, WHS practitioners and duty holders.

Background and Problem Statement

Regulatory reform and its expectations

The model WHS Regulations now define a psychosocial hazard as one arising from the design or management of work, or from workplace interactions and behaviours. The accompanying Model Code of Practice: Managing Psychosocial Hazards at Work sets out a four-step process of identification, assessment, control and review, and states that controls must address the source of the hazard rather than the resilience of the individual exposed to it (Safe Work Australia, 2022). Comcare (2023) has issued parallel guidance for Commonwealth employers, and ISO 45003 provides a compatible management-system framing adopted by many larger Australian employers for compliance (International Organization for Standardization, 2021). Related obligations sit outside the WHS statutes: the Fair Work Commission (2023) reports continued demand in its anti-bullying and sexual harassment jurisdictions, so the same conduct can generate simultaneous regulatory and industrial exposure.

The prevention gap

The regulatory case for reform rests on compensation data. Mental health conditions have become the most expensive category of accepted workers’ compensation claim in Australia, with median time lost and compensation several times the figures for all other injury types, and accepted claim volumes have risen steadily over the past decade (Safe Work Australia, 2023). Yet compensation data measure harm after it has materialised, and say nothing about whether hazards were identified in advance, controlled at the level of work design, or reviewed. The prevention gap is therefore both practical and evidentiary: regulators are enforcing a duty whose upstream implementation is largely unmeasured.

Problem statement

Two problems follow. First, organisational responses appear highly variable, from genuine redesign of workload and rostering to documenting existing employee assistance programs as if they were controls. Second, there is little Australian evidence linking the quality of documented risk management to the psychosocial conditions workers report. Without that link, regulators cannot distinguish substantive from paperwork compliance, and duty holders lack an empirical basis for prioritising investment. This study addresses both problems in one design.

Aim and Research Questions

The aim is to examine how Australian duty holders are implementing the amended psychosocial provisions of the model WHS Regulations, and to test whether variation in implementation quality and psychosocial safety climate is associated with worker-reported hazard exposure and psychological outcomes. Three research questions guide the study:

1. How do the six participating organisations identify, assess, control and review psychosocial hazards, and how closely does documented practice correspond to the four-step process in the Model Code of Practice?
2. To what extent do psychosocial safety climate and documented control quality predict worker-reported job demands, job resources, psychological distress and work engagement, after accounting for the clustering of workers within teams and organisations?
3. What organisational, regulatory and managerial factors do WHS and line managers identify as enabling or constraining substantive compliance with the psychosocial duties?

Literature Review

The job demands-resources model

The job demands-resources (JD-R) model provides the dominant account of how work characteristics produce psychological outcomes. Demands such as workload, emotional labour and role conflict initiate a health-impairment process culminating in strain, while resources such as job control, supervisor support and role clarity initiate a motivational process sustaining engagement and buffer the effect of demands on strain (Bakker & Demerouti, 2017). The model maps neatly onto the hazard categories in the Australian Code of Practice, which is one reason it suits regulatory research. Its relevance is sharpened by evidence that poor work design is self-perpetuating: organisations lacking the capability to design good work are least likely to detect the harm it produces (Parker et al., 2019).

Psychosocial safety climate

Psychosocial safety climate (PSC) describes shared perceptions of senior management commitment to psychological health, and the priority it receives relative to productivity. PSC is theorised as an antecedent of the JD-R pathways: high-PSC organisations set demands at sustainable levels and provide resources, so PSC predicts subsequent demands, resources and strain rather than merely correlating with them (Dollard & Bailey, 2021). It is measured with the validated 12-item PSC-12, for which Australian benchmarks distinguishing high-risk from low-risk organisations have been established (Bailey et al., 2015). As a climate construct, PSC is properly modelled at the group level.

Regulatory compliance behaviour

Regulatory scholarship warns that legal duties are mediated by organisational interpretation. Compliance is better understood as a continuum, from creative or symbolic compliance in which documentation is produced without substantive change, through to committed compliance driven by internalised norms and management systems capability (Gunningham & Sinclair, 2017). Process obligations like the psychosocial amendments are especially vulnerable to symbolic responses, because a risk register is easier to demonstrate and audit than the adequacy of the controls it records.

Synthesis and gap

The three literatures have developed largely in isolation: JD-R research rarely engages with regulatory obligation, PSC research seldom examines the documented artefacts through which commitment is enacted, and compliance research theorises organisational responses without measuring worker experience. No Australian study has yet combined document-based assessment of psychosocial risk management with validated measurement of climate, exposure and outcome in one multi-site sample, which is the contribution of this study.

Conceptual Framework

The framework integrates the regulatory risk management cycle with the PSC-extended JD-R pathway, as shown in Figure 1. Regulatory duty initiates the four-step cycle, and the quality of its execution, together with psychosocial safety climate, is theorised to shape the demands imposed on and resources available to workers, which in turn produce strain and engagement outcomes. The recursive arrow in Figure 1 shows monitoring data feeding the review step, so organisations with functioning review improve control quality over time.

Methodology

Design

A convergent mixed methods design will be used, in which quantitative and qualitative strands run in parallel and are integrated through a joint display (Creswell & Plano Clark, 2018). The survey measures what workers experience, the document review what the organisation has recorded, and the interviews explain the gap between the two.

Setting and sampling

Six organisations will be recruited through an industry reference group, purposively selected for variation in sector and regulatory exposure: two public health services, one local government authority, one construction contractor, one contact centre and one professional services firm, spanning both high-hazard operational settings and office environments. All workers in approximately 40 nominated teams will be invited to complete the survey, targeting about 400 responses at an anticipated response rate of 45 per cent, with a mean cluster size of about 10 workers per team. Eighteen semi-structured interviews, three per organisation, will be conducted with a WHS manager, a senior operational manager and a frontline supervisor.

Because workers are clustered within teams, the effective sample size is smaller than the raw count. Assuming an intraclass correlation of .08 for psychological distress, the design effect is calculated as:

Design effect = 1 + (m – 1) x ICC = 1 + (10 – 1) x 0.08 = 1.72

Effective sample size = 400 / 1.72 = 232.6

An effective sample near 233 gives roughly 80 per cent power to detect standardised coefficients of about .18, typical of PSC research.

Instruments and measures

Table 1 summarises the constructs, their level of measurement and the instruments selected, all with published Australian psychometric evidence supporting comparison against national benchmarks.

Table 1: Constructs, levels of measurement and instruments

Note. The control quality rubric scores each of the four steps in the Model Code of Practice against three criteria: evidence of worker consultation, specificity of the hazard or control, and an accountable owner with a review date.

Analysis

Quantitative data will be analysed using multilevel modelling to respect the nesting of workers within teams, with organisation entered as a fixed effect because six clusters are too few to estimate a third random level reliably (Hox et al., 2018). Null models will establish intraclass correlations; team-level PSC and organisation-level control quality will then be added as predictors of demands, resources, distress and engagement, with multilevel mediation testing the indirect pathways specified in Figure 1. Interviews will be transcribed and analysed using reflexive thematic analysis, coded initially around the four regulatory steps and then developed inductively (Braun & Clarke, 2022). Integration will occur through a joint display placing each organisation’s control quality score alongside its aggregated PSC score and dominant themes, identifying where documentation and experience diverge.

Ethical Considerations

Approval will be sought from the administering university’s Human Research Ethics Committee, and the study conducted in accordance with the National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council, 2023). Three issues require particular attention. First, worker confidentiality is at risk because employers control access; no individual responses will be shared with employers, and team-level results reported only where at least eight responses are received. Second, participation must be genuinely voluntary despite the employment relationship, so invitations will be distributed by the research team rather than managers, and completed in paid time without any requirement to report participation. Third, because questions about distress may cause discomfort, participants will receive contact details for their employee assistance provider and national support services. Site agreements will confirm in advance that researchers retain publication rights and that organisations will be de-identified, protecting against pressure to suppress unfavourable findings.

Project Timeline

The project will be completed over 12 months, as set out in Table 2. Parallel scheduling of the document review and survey fieldwork makes the convergent design feasible within a year.

Table 2: Twelve-month project timeline

Significance of the Study

For regulators, the study offers what the current evidence base lacks: an empirical test of whether documented compliance corresponds to improved conditions for workers. If control quality scores prove weakly related to worker-reported demands, the finding would support a shift in inspectorate practice from document audit towards verification of control effectiveness. For duty holders, the joint display will show which four-step elements carry the most explanatory weight, directing scarce WHS resources to the steps that matter. Theoretically, it extends PSC research by treating regulatory obligation as an explicit antecedent of climate, and compliance scholarship by attaching validated worker-level outcomes to the construct of symbolic compliance.

Limitations and Delimitations

Four constraints qualify the design. The cross-sectional survey cannot establish causal ordering between climate, exposure and distress, although the document review provides a partially independent measure that reduces common-method bias. Six organisations recruited through an industry reference group are likely more engaged with WHS obligations than Australian employers generally, so control quality estimates should be read as an upper bound, not a national average. Employer gatekeeping may also depress response rates in lower-climate teams, biasing results towards the null. Finally, the study is delimited to the WHS psychosocial provisions and does not examine parallel obligations under anti-discrimination or industrial law, despite their overlap in sexual harassment (Fair Work Commission, 2023). A longitudinal follow-up and a comparison with the Commonwealth scheme administered by Comcare (2023) are priorities for subsequent work.

Conclusion

The 2022 to 2023 psychosocial amendments have given Australian regulators a prevention duty of considerable reach, but the reach of a duty is not the same as its effect. This proposal sets out a feasible 12-month study that measures both sides of the compliance relationship at once: what organisations record in their risk registers, and what workers report about the demands and resources those registers are supposed to control. By combining the JD-R model, psychosocial safety climate theory and compliance scholarship, it aims to establish whether substantive and symbolic compliance can be distinguished empirically. That distinction is the practical precondition for enforcing a duty whose ultimate measure is not the quality of documentation but the psychological health of the Australian workforce.

References

Bailey, T. S., Dollard, M. F., & Richards, P. A. M. (2015). National standard for psychosocial safety climate: PSC-12. Journal of Occupational Health Psychology, 20(1), 15-26.

Bakker, A. B., & Demerouti, E. (2017). Job demands-resources theory: Taking stock and looking forward. Journal of Occupational Health Psychology, 22(3), 273-285.

Braun, V., & Clarke, V. (2022). Thematic analysis: A practical guide. Sage.

Comcare. (2023). Psychosocial hazards and factors: Guidance for Commonwealth employers. Comcare.

Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods research (3rd ed.). Sage.

Dollard, M. F., & Bailey, T. (2021). Building psychosocial safety climate in turbulent times: The case of COVID-19. Journal of Applied Psychology, 106(7), 951-964.

Fair Work Commission. (2023). Annual report 2022-23. Commonwealth of Australia.

Gunningham, N., & Sinclair, D. (2017). Smart regulation. In P. Drahos (Ed.), Regulatory theory: Foundations and applications (pp. 133-148). ANU Press.

Hox, J. J., Moerbeek, M., & van de Schoot, R. (2018). Multilevel analysis: Techniques and applications (3rd ed.). Routledge.

International Organization for Standardization. (2021). ISO 45003:2021 Occupational health and safety management: Psychological health and safety at work: Guidelines for managing psychosocial risks. ISO.

National Health and Medical Research Council. (2023). National statement on ethical conduct in human research. NHMRC.

Parker, S. K., Andrei, D. M., & Van den Broeck, A. (2019). Poor work design begets poor work design: Capacity and willingness antecedents of individual work design behaviour. Journal of Applied Psychology, 104(7), 907-928.

Safe Work Australia. (2022). Model code of practice: Managing psychosocial hazards at work. Safe Work Australia.

Safe Work Australia. (2023). Australian workers’ compensation statistics 2021-22. Safe Work Australia.