Wicked Problems and Institutional Courage - Testing Relational Justice in Veteran Entitlement Offsetting.
Proof of Concept Testing for a Veterans' pro bono Mediation Program
Acknowledgment
This reflection is offered with deep respect for the current and former members of the Australian Defence Force, and the communities and advocates who stand alongside them. Their experiences - shared through testimony, conversation, and enduring perseverance - have shaped this work. This model for a veterans’ mediation program was developed in response to their stories, and it remains grounded in the understanding that care, not theory, must be the proper centre of practice.
The Work in Progress
Over the past two months, I have been developing a dispute resolution model known as Accord Resolution - a values-based, post-law approach to resolving conflict through relational justice, care ethics, narrative and insight-led dialogue. Building from this foundation, I began designing a structured Veterans' ADR Program, to be offered pro bono to current and former members of the Australian Defence Force.
As work progressed, I documented design insights and reflections in a growing body of writing - Field Notes from Post-Law Peacemaking. These entries traced the evolution of our theory and practice of relational justice, with particular attention to the moral obligations that emerge at the point where legal remedies and administrative discretions intersect.
“Shadow Space”
One of our key areas of focus became the “shadow space” of:
the Act of Grace Scheme under the Public Governance, Performance and Accountability Act 2013 (Cth), guided by Resource Management Guide No. 401: Requests for discretionary financial assistance under the PGPA Act (RMG401); and
the Scheme for Compensation for Detriment caused by Defective Administration (CDDA Scheme), operating on the basis of authority conferred upon individual portfolio ministers under the executive power in sections 61 and 64 of the Constitution, and described in Resource Management Guide 409 (RMG 409).
This is post-law space where law no longer mandates outcomes, but moral discernment and discretionary repair remain possible. I began to explore how mediation, drawing on insight and narrative techniques, might open and hold relational space in which institutional courage could be exercised and moral repair might begin.
Concept Testing
Encouraged by the emergence of these post-law pathways, I moved to test the model through a proof-of-concept exercise, applying it to a challenging and contentious area of veteran entitlements: offsetting.
A Note on Offsetting
The Interim Report of the Royal Commission into Defence and Veteran Suicide (2022) noted at page 170:
6. …Entitlements are offset when a veteran has entitlements under more than one Act or receives a pension or lump sum under a superannuation scheme.
7. The complexity of the legislation leads to a complicated administrative system and has the potential to result in perceived or actual inconsistency. … this adds to the burden placed on veterans engaging with the system. The complexity also contributes to erroneous decisions, which contributes to more delay and more distress for some veterans. It is an urgent matter that needs to be addressed to reduce mental health impacts and suicide risk.
Approach
To conduct testing, I developed a hypothetical scenario grounded in:
findings from the Interim Report of the Royal Commission into Defence and Veteran Suicide;
testimonies from veterans and advocates; and
accumulated understanding of the structural, procedural, and emotional dimensions of entitlement reduction through offsetting.
The scenario centred on a veteran whose payments had been significantly reduced due to offsetting, after a troubled procedural history that included:
complex interactions between different legislative schemes
representation by multiple advocates;
conflicting and incomplete advice; and
a critical moment of incomplete departmental advice that led to a materially disadvantageous outcome.
The challenge presented was not theoretical. While the character was composite, the scenario reflected the lived experience of many veterans. It not only asked whether the system had functioned properly, but whether it had responded with care.
To test the model, I provided a generative AI tool with:
a fully developed hypothetical scenario;
curated extracts from the Royal Commission’s Interim Report on the topic of offsetting;
my complete anthology of Field Notes tracing the development of our relational justice framework; and
relevant policy materials, including section 65 of the PGPA Act and Resource Management Guides Nos. 401 and 409.
I then asked:
How might this model help a veteran in a matter now pending before the Veterans' Review Board?
Diagnostic
A quick diagnostic revealed several distinct features:
Administrative fragmentation: offsetting rules applied across three statutory schemes.
Procedural mishap: concurrent claim lodgement without informed consent.
Misattribution of impairment: psychological conditions treated as primary cause, despite clear physical incapacity.
Representation failures: no advocate had sustained a relational view of the whole matter.
Outcome: full offsetting and a veteran left materially and morally unacknowledged.
Confirmation and a Framing
The results affirmed something I hoped might prove true but needed to see demonstrated: the model held.
It did not deliver a solution. It held space. It allowed formulation of five foundational questions for dialogue:
What has the system failed to recognise?
What has the institution failed to do - not legally, but relationally?
What harm still lives in the story?
What could justice look like now - not in judgment, but in relational repair?
And what can our model offer that the law no longer can?
The Model Applied
From there, it was possible to sketch how the matter might be brought into a Veterans' Review Board Preliminary Conference without undermining its purpose:
by framing the engagement not as opposition, but as clarification with presence;
by using insight techniques to surface underlying cares and threats-to-cares;
by carefully deconstructing dominant narratives;
by proposing space for moral reflection, even if formal remedies remained unavailable; and
by signalling possibilities for institutional learning and discretionary review.
Perhaps most telling was the way the model engaged with departmental misadvice as a moment of relational disconnection. The question was no longer, who was to blame? The question became, what have we learned - and how will we carry that forward?
Discretion and the Limits of Procedure
In examining the ethical weight of harm in AB’s case, I also considered the frameworks established by the Act of Grace scheme (RMG 401) and the Compensation for Detriment caused by Defective Administration (CDDA) Scheme (RMG 409). Both confirm what veterans and advocates have long understood: not all harms suffered in contact with government are remediable through law alone.
Each scheme offers a discretionary pathway when loss, confusion, or disadvantage arise outside strict legal entitlement. Yet each, in its own way, tends to frame harm through the logic of administrative fault. This can constrict the very space in which relational harm and moral responsibility most need to be addressed.
The CDDA scheme, in particular, casts the matter in the language of defect, detriment, and causal proof. That framing can make it harder for institutions to step into ethical reflection. When the first question is, “Did we err?”, the likely response is procedural silence or defensive rebuttal. The deeper questions - “What was experienced? What mattered? What remains to be acknowledged?” - may never be reached.
This is where Accord Resolution opens complementary space. It does not displace or duplicate the CDDA process. Rather, it holds a different frame:
From defensive posture to ethical presence;
From proof of harm to acknowledgment of care at stake;
From eligibility determination to shared learning and discretionary courage.
When the CDDA framework is implicitly present - even without a formal claim - it can shape the conversation in ways that narrow it. Words like “error,” “defect,” and “entitlement” begin to crowd out terms like “recognition,” “repair,” and “trust.” In these moments, the mediator’s role includes gently externalising the procedural discourse, offering a reframing such as:
“It sounds like we’re being drawn into a question of fault. That might be important later. But for now, can we stay with what was experienced? What mattered? What might still be possible?”
This shift is not cosmetic. It is foundational. It reorients the conversation from risk management to relational attentiveness. It creates a setting where institutions are not asked to concede liability, but to act with presence, care, and discretionary integrity.
Accord Resolution, in this sense, does not remove the need for compensation. But it widens the aperture through which remedy can be understood. It recognises that in complex veteran matters, harm is rarely just financial. It is also relational. And it is there - in that space of meaning and moral repair - that the mediation begins.
In the scenario, offsetting had produced a quantifiable loss. But it also had ethical weight. The model helped surface that weight with care rather than with accusation.
Discovery Set 2: Reflective Practice
Encouraged by this, I posed a second set of questions.
What does the model do well?
What might it still need to do differently?
This was a necessary phase in the work. A moment to assess whether the structure was merely functional, or whether it was also faithful to the principle of relational justice which had been woven into its design.
What the Model Does Well
It holds relational and ethical space where law might not.
It anticipates trauma and honours insight.
It respects formal process while walking beside it.
It invites institutions not into blame, but into integrity.
What the Model Might Do Differently
While the model has shown its strengths, there are areas where its structure could be improved.
It may need a clearer way to respond when veterans have received and act on advice that turns out to be wrong or incomplete. One approach could be to offer a structured process, which we are calling a Without Prejudice Harm Review - to help all parties reflect on what happened and decide what can be done now, without becoming defensive.
It may need to help participants understand what happens after a mediation, especially when no immediate decision or remedy follows. A short guide - perhaps titled If You Still Need to Be Heard - could help people know what their options are, and what steps they can still take.
It may need to recognise that people don’t just step into and out of a conversation about harm. Emotions build and shift across time. A quiet follow-up - a call, a note, a check-in - can help people feel that what was said still matters. We’re calling this a TI Check-In (trauma-informed check-in). It is a light-touch follow-up offered a week or so after mediation or facilitated dialogue. Its purpose is not to revisit the matter, but to affirm that what was shared still matters - and to support emotional closure where needed.
Finally, it may need a way to notice patterns. When the same issues come up across different cases, they shouldn’t be lost. A simple system - what we’re calling a Systemic Insight Register - could record these patterns (with consent), so that learning isn’t just personal, but institutional too.
A Deeper Realisation
Beyond providing a framework for mediation, the model offers a structure for trauma-informed presence.
Where other models seek closure, it seeks continuation - because care is rarely finished when the conversation ends.
Where other models settle, it sustains.
Where other models end, it listens for what still needs to be said.
Mediation is more than an event
The process demonstrated itself as more than a single event:
Intake became orientation.
Pre-mediation became narrative anchoring.
Dialogue became midpoint.
Outcomes became waypoints.
Follow-up became continued care.
Learning not Blaming
The model demonstrated its own ethic - and in doing so, it reminded us:
We do not blame.
We learn.
We respond with care before correction - and never with coercion.
Those precepts were shaped by principles of significant learning that had been designed into the foundations of the model in alignment with L. Dee Fink’s call to attend to the human dimension, to caring, and to learning how to learn.
Lessons from Root Cause Analysis
It resonated, too, with the logic of root cause analysis in medicine - where the goal is not to assign fault, but to understand how systems can do better. That opened the way to consider incorporating a Relational Learning Loop, and with it, a new type of mediation outcome: the Learning Statement - a co-authored reflection that documents what happened, what mattered, and what will now guide ethical response.
Repair May Precede Remedy
In many systems, the sequence is clear: first prove entitlement, then receive remedy.
But in this work, the sequence is often reversed:
First, the harm must be acknowledged.
Then, the relationship must be repaired.
Only then, can any remedy be meaningfully received.
This insight does not displace legal order - it extends the frame of what responsibility requires.
A Final Reflection
What began as a proof of concept became something more: a demonstration of a practice, finding its voice in the midst of complexity and pain.
The model did not seek to settle or compromise the veteran’s case – or that of the Department. It did something else.
It held space.
It surfaced meaning.
It honoured care.
And it engaged with the story—without rushing it to an end.
And it engaged with the story—without rushing it to an end.
There is no fanfare in that. I hope there is integrity. And in this kind of work, perhaps that is the most we can ask for - and the most we can offer.