ANZSOG Research discovers ways to improve Public Sector Integrity

New ANZSOG research looks at how government agencies can build a pro-integrity culture, using four examples from the Victorian public sector to outline practical changes that public sector agencies and leaders can make. 

The report finds that an effective integrity culture cannot be realised through top-down directives or training alone, but is a collaborative, ongoing and relational practice, However, there are general principles around successful pro-integrity projects that agencies can tailor to their own circumstances. 

The project was a partnership between ANZSOG, the Victorian Public Sector Commission and the University of Sydney, The final report  – Improving Public Sector Integrity through Action Research – by Assel Mussagulova, Susan Goodwin and Alicia Wang Sin, documents promising examples of integrity-as-practice and pro-integrity interventions developed across four Victorian government agencies. 

The project used ‘action research’ methodology - a participatory, iterative approach to inquiry that positions employees as co-investigators rather than subjects of study.  

Who was involved 

Across four agencies, practitioners worked in Communities of Practice to identify their own integrity challenges, co-design interventions, and test them in real organisational settings, generating knowledge that was both grounded in local context and immediately actionable. 

The four action research projects were developed from the ground up from within four Victorian departments. Each project was distinctive in scope and focus: 

 • In order to understand ‘speak up’ culture across a large and dispersed school workforce, the action research project in the Department of Education elicited school leaders and school staff perspectives to generate insights to inform future resources and policy. 

 • With the goal of promoting everyday integrity practices, the project run in the Department of Health sought to embed integrity discussions into existing meeting structures across two work areas over a trial period, finding that normalising integrity as an everyday topic shifted how staff understood and engaged with it. 

 • The action research undertaken by the Department of Justice and Community Safety trialled an integrity risk self-assessment tool with the Sheriff's Office of Victoria, working through integrity risks together in cross-level groups to integrate knowledge and frank reflections. 

 • In the Department of Transport and Planning, a network of voluntary integrity champions distributed across the organisation was established, with the research finding that peer-based approaches extended the reach of the central integrity team and began to shift integrity from a centralised function to a shared organisational responsibility. 

Central finding: integrity is relational and requires shared responsibility 

The research’s central finding is that integrity is best understood as a relational practice, that involves all employees working together. Where structured spaces for dialogue were created, participants moved integrity from an abstract compliance requirement to a lived, shared responsibility which was built through everyday interactions, mutual accountability, and collective sense-making. 

Integrity emerges from the ways people communicate, make sense of challenges, and support one another to act ethically, not through individual knowledge or compliance alone. Even using the word ‘integrity’ in meetings was enough to open up substantive conversations and helped participants discuss issues often left unspoken. 

Across the four projects, several other overarching lessons emerged about what is required to creating the conditions form which a pro-integrity culture, unique to each individual agency, can emerge. 

Six key findings 

Systems and culture are sequential, not competing. Attention to systems and processes is not an alternative to culture change, it is its foundation. The evidence from the CoPs suggests that structured engagement with systems and processes builds the trust and shared understanding from which more challenging conversations about behaviour and culture can follow.  

Individual knowledge is necessary not sufficient. Awareness-raising and training are important starting points but they are not end points. The action research consistently showed a gap between individual knowledge acquisition and the capacity to apply that knowledge in ways that produce collective culture change.  

Leadership is enabling, not merely endorsing. Executive commitment provides mandate and signals organisational seriousness, but momentum requires mid-level and peer leadership to translate that mandate into everyday practice. The evidence from the CoPs points to integrity leadership as a distributed, relational responsibility rather than a top-down function. Executives set the tone; managers make it real; and local champions and peers sustain it through ongoing conversation and modelling. . Where any of these levels were absent or disengaged, the impact of the intervention was diminished. 

Context determines what works. Locally adapted approaches consistently outperformed generic frameworks. Effective integrity work requires investment in understanding local integrity risk profiles, vocabulary, and cultural dynamics before designing interventions. What resonates in one business area or organisational setting may fall flat in another.  

Sustainability must be designed in. Several projects demonstrated that integrity conversations and practices, once initiated, risk falling off the agenda without deliberate structural support. Initiatives that rely solely on voluntary goodwill or the momentum of an initial trial are unlikely to endure. Building integrity into existing structures - as standing agenda items, formal network roles, or performance and development goals - is more likely to produce lasting culture change than time-limited projects operating alongside business as usual. 

Reflection is itself integrity practice. The willingness to acknowledge what did not work, learn from it, and share that learning openly is both a method and a manifestation of pro-integrity culture. The CoPs that reflected most honestly on the limits of their interventions produced the most useful insights for future practice. 

The report, plus an earlier report from the research project can be found on the ANZSOG website