Modernising risk equalisation: A low-risk, high-impact pathway for Australia’s PHI system

Australia’s Risk Equalisation (RE) system plays a critical role in sustaining community rating, ensuring that insurers with older or higher-risk members are not disadvantaged. While this article focuses on RE, it is important to recognise that RE is only one part of the private health insurance (PHI) system. Any reforms to RE must align with broader PHI objectives.

Modernising the system should be seen as evolution rather than disruption. The idea of a pathway reflects the need to avoid rigidity and instead support learning and structured iteration.

The gateway: Problem or opportunity?

Transformation in systems begins not with solutions but with a shared problem worth solving (Trist 1983 ¹ ). This might be a “wicked mess” or a compelling opportunity that goes beyond the scope of any single organisation.

There are reviews and current debates about Gold products ² , mental health, and obstetrics. Recognising this gateway, whether as a shared challenge or an emerging opportunity, is essential. Without it, even the best-designed reforms risk becoming disconnected from the realities they are intended to address.

We ought also to also acknowledge that there can be losers as a result. However, naming the problem and framing the opportunity is not a side task. It is the starting point for purposeful, low-risk reform.

Design elements for system reform

Research on multi-stakeholder systems highlights a set of foundational elements that enable coordinated and adaptive change (Ainsworth 2016 ³ ). These elements can serve as design scaffolding for RE reform:

1. Shared goals: A broadly agreed and shared purpose is essential. Without goal clarity, stakeholders can drift in different directions, unsure what success looks like or what actions to take.
2. Measurement: Measurement enables reflection, adjustment, and accountability. It is also more effective when linked to defined goals.
3. Governance: Coordinated action requires deliberate processes. These include shared decision-making structures, clear roles, adequate resources, and mechanisms for collective leadership. Governance here is less about regulation and more about enabling function.
4. Legislation: Legislation provides the authorising environment for making deliberate and consistent system decisions over time
5. Information and cash flow: Beyond measurement, systems require the ability to share timely information and manage real-time transactions, especially where financial transfers are involved.
6. Trust: Trust is both a precondition and an outcome. It grows when stakeholders take small risks and receive consistent, supportive responses. Without trust in the reform process, change will be difficult to implement.

These design elements provide an important lens to evaluate current RE structures and to inform reform efforts in a multi-stakeholder system.

A phased pathway: Three practical phases

Rather than attempting a single major overhaul, a practical and lower-risk pathway is to modernise RE in phases. This three-phase transition delivers early improvements, builds capability for continuous refinement, and positions the sector for long-term adoption of a modern hybrid RE model.

Note on sequencing: These phases are presented in an order, but they do not need to be strictly sequential. Establishing a platform for improvement (Phase 2) could occur before or in parallel with initial coverage changes (Phase 1). What matters is that each phase reinforces the others and contributes to a cohesive, evolving system.

Phase 1: Align with Current PHI Priorities

Stakeholders often seek reforms that produce short-term, visible improvements. One option is to begin by aligning RE with issues already receiving attention in the broader PHI system. Topics like Gold policies, mental health, and obstetrics are not problems created by RE. They are current focal points where RE could make a constructive contribution, not work in isolation.

A practical first move could be to include Mental Health and Obstetrics in the RE pool while keeping the total shared amount constant. This could be achieved by adjusting the attachment point, the sharing proportion, or reallocating within the existing pool. In effect, RE coverage would expand, but the size of the pool would not.

This involves trade-offs. Expanding coverage for some areas means reducing support for others. Still, it sends an early and credible signal that RE reform is both feasible and manageable.

Phase 2: Establish a “Platform for Improvement”

Ideally, this could even be the first phase.  If the sector proceeds with mental health and obstetrics as an early move, it is essential that these are implemented in a way that also supports longer-term system readiness. The order is less important than ensuring each change contributes to a coordinated, adaptive platform for future evolution.

A key limitation of the current RE system is the absence of a mechanism for ongoing, evidence-based refinement. Historically, changes have been infrequent. What is needed is a platform that supports regular review, shared understanding, and coordinated development.

This platform can be built on three core components:

a) Clear system goals: Publishing explicit goals provides a reference point for future adjustments. A possible core goal could be: 

This goal aligns with community rating. Under community rating, insurers face predictable gains on young and healthy members and predictable losses on older or chronically ill ones. RE exists to address this imbalance.

Further discussion of goal setting can be found here:

• Purposeful Risk Equalisation in Private Health Insurance
• Measuring Success in PHI Risk Equalisation

b) A coordination mechanism: A designated group should be responsible for:

• circulating research and discussion papers
• convening technical working groups
• managing modelling and simulation
• coordinating stakeholder engagement
• supporting improvements in data quality

This coordination mechanism acts as the system’s learning engine. It ensures that RE evolves in a continuous and responsive way, rather than through sporadic and reactive changes. The goal is not to add a new regulator but to ensure design elements are aligned and adaptive in a multi-stakeholder environment.

c) Governance that enables regular adjustments: Governance may include adapting legislation to establish the overall RE framework while empowering updates to:

• adjust attachment points and sharing proportions
• update risk adjuster values
• conduct regular reviews and incorporate new data sources

This flexibility ensures the system can adapt to emerging evidence and sector capability. Potentially, with the other design elements, such as agreed goals, the elements of the system could be periodically reviewed and adjusted by an appropriate mix of system stewards, not requiring full government involvement.

Phase 3: Introduce risk adjusters and prospective elements through test-and-learn

The long-term vision for RE in Australia is a hybrid model that combines retrospective high-cost claims sharing with prospective risk adjustment. This combination allows for enhanced alignment of incentives while protecting against outlier risks.

A structured test-and-learn approach supports this shift.

a) Offline simulation: Before implementation, a “shadow” version of the hybrid model should be run using real claims data. This would allow stakeholders to assess:

• expected transfer impacts
• distributional effects across insurers
• incentive changes
• potential premium implications
• data quality issues

Publishing quarterly shadow results would build transparency and confidence.

b) Phased introduction of risk adjusters: Initial adjusters should be based on data that is already collected and reliable, such as:

• Age (5-year bands)
• Sex
• State

As data quality improves, health-based adjusters (such as hospitalisations, Diagnosis-Related Groups and clinical indicators) can be added

c) Gradual rollout of prospective components: If simulations show that the hybrid model produces benefits, the prospective elements can be introduced gradually. Even with long-term advantages, some stakeholders will experience short-term losses. Certain insurers may see reduced profits or even losses. Some consumers may face higher premiums for the same coverage. A phased rollout allows time for the benefits to emerge and for trust in the model to grow.

This approach gives the sector the ability to move toward a more enhanced RE model while managing implementation risk.

Conclusion

Australia’s Risk Equalisation system is not broken. That alone, however, is not a reason to stand still. Rather than waiting for failure, we can ask a better question: if the system can better meet its goals, why wait? Improving fairness, supporting community rating, and aligning incentives are all possible. That is, increasing effectiveness while minimising disruption.

This is a shift from maintaining what works to evolving what could work better. By using existing momentum, building system capability, and testing changes safely, we can move toward a more adaptive and future-ready RE model.

References

^[1] Trist, E. (1983). Referent Organizations and the Development of Inter-Organizational Domains. Human Relations, 36(3), 269-284. https://doi.org/10.1177/001872678303600304 (Original work published 1983)

^[2] Actuaries Institute. (n.d.). Time to rethink risk equalisation to save Gold hospital. https://www.actuaries.asn.au/research-analysis/time-to-rethink-risk-equalisation-to-save-gold-hospital  

^[3] Ainsworth, Dale & Feyerherm, Ann. (2016). Higher order change: A trans organizational system diagnostic model. Journal of Organizational Change Management. 29. 769-781. https://doi.org/10.1108/JOCM-11-2015-0209