The Impact-Oriented Collaborative Governance Model:A Systems Approach to Evaluating Cross-sector Collaboration

Background There is evidence to suggest a collaborative approach to governing complex programs and projects can deliver better outcomes for beneficiaries. However, governments, the community sector and industry continue to find it difficult to establish a structure and culture that fosters collaboration, as well as find alignment in incentives and goals—the result being inconsistent outcomes of collaborative governance. Aims This research evaluated collaborative governance implementation to generate insights for practitioner guidance by answering: Under what conditions (or combination of conditions) does collaborative governance lead to successful outcomes, and which conditions are most impactful? Successful outcomes, or performance, was measured in terms of effectiveness, legitimacy and adaptability of the collaboration. Methods In line with the pragmatist approach of critical systems thinking, a multi-perspective multi-methodological design was used to tackle the complexity of collaborative governance. This included Bayesian networks and qualitative comparative analysis (QCA) from the mechanistic systemic perspective, and Critical Systems Heuristics from the societal/environmental systemic perspective. These were applied in a global medium-N study of 60 cases from the Collaborative Governance Case Database and a small-N study of two Australian cross-sector collaborations of national significance. Findings Through Bayesian network modelling, the six variables of collaborative governance with the highest impact on performance were determined to be: (1) interdependence, (2) convening relevant actors, (3) incentives to collaborate, (4) conflict resolution, (5) operationalising joint goals and (6) inclusivity. The Bayesian networks analysis also revealed the relative importance of accountability, which was not emphasised in the leading Collaborative Governance Model. The Bayesian networks analysis informed a proposed Impact-Oriented Collaborative Governance Model. The QCA research validated the Impact-Oriented Collaborative Governance Model based on cases from the Collaborative Governance Case Database and provided other valuable insights. Performance could be achieved via four pathways: (1) the combination of leadership and starting conditions, (2) leadership and process and design, (3) accountability and design and (4) accountability and process. However, the frequency and consistency of achieving performance was increased when more conditions of collaborative governance were present, so practitioners should seek to implement as many conditions as possible to increase the chances of success. The QCA findings also emphasise the importance of accountability. Two original empirical studies of nationally significant collaborative governance cases in Australia were studied, and served as out-of-sample tests supporting the validity of the Bayesian networks and QCA models. First, the Emergency Relief National Coordination Group served as an exemplar for implementing collaborative governance with high fidelity leading to successful outcomes. Second, the Consistent Gambling Messaging Project under the National Consumer Protection Framework for Online Wagering was an example of a successful intervention based on collaborative governance, implemented in a short timeframe to move forward a collaborative project that was experiencing difficulty. Ethical considerations of collaborative governance were explored by applying Critical Systems Heuristics. This informed issues such as the potential for certain stakeholders to be excluded from decision-making, power imbalances inherent in cross-sector collaboration and self-interested stakeholders such as gambling companies. Furthermore, it was determined that collaborative governance was not immediately appropriate for use in First Nations contexts without first being decolonised, and that alternate culturally appropriate collaboration frameworks already existed. Significance The research makes a theoretical contribution by incrementally refining existing leading frameworks to develop and test an Impact-Oriented Collaborative Governance Model. The model and associated guidance also constitutes a practical contribution by equipping scholars and practitioners with tools to rethink collaboration in systemic, ethically grounded and outcome-focused ways. The research demonstrates methodological innovation in the application of Bayesian networks, QCA, and Critical Systems Heuristics in creative combination, which was effective. Finally, the research makes a contribution to the limited Australian empirical collaborative governance literature through the rich description of two original cases of nationally significant cross-sector collaboration: (1) the Emergency Relief National Coordination Group and (2) the Consistent Gambling Messaging Project. Further research opportunities arising from the thesis include: (1) case studies and comparisons to more comprehensively test the utility of the Impact-Oriented Collaborative Governance Evaluation Tool; (2) studies grounded in Critical Systems Heuristics to inform the Australian Government’s Regulation Impact Analysis process regarding the treatment of input from self-interested stakeholders, as well as how stakeholders more generally might be limited by the framing of issues and policy options presented in Regulation Impact Statements; and (3) studies focusing more on the external impacts to cross-sector collaboration such as political influence, coercion and inclusivity.