A system leverage points approach to governance for sustainable development

Governments are inherently responsible for citizens' well-being. Given that achieving sustainable development ["Development that meets the needs of the present without compromising the ability of the future generations to meet their own needs"—(WCED in Our common future, Oxford University Press, New York, 1987)] is core to the attainment and maintenance of citizens' well-being, and increasingly understood to require major transformations in integrated social, technological and ecological systems (Sachs et al. in The decade of action for the sustainable development goals: sustainable development report 2021, Cambridge, 2021), it follows that governments have a significant role in shaping transformations. Muted progress on long-standing social, environmental, and economic challenges alongside spiralling public budgets and intergenerational debt suggests, however, that public governance systems are inadequate to facilitate the transformations urgently required. Conceptualising the practice of public decision-making as a complex system, this paper investigates whether known influences on public decision-makers can be linked to Meadows’ (Leverage points: places to intervene in a system, Sustainability Institute, North Charleston, 1999) leverage point framework. Finding meaningful connections, it further explores how the leverage point framework can be employed to engage decision-making influences as enablers of desirable public outcomes. It is contended that shifting decision-makers’ focus one step beyond currently prevalent leverage points will set in motion the transformations in governance required to facilitate sustainable development.


Introduction
The world faces an ever-evolving raft of complex, interconnected, enduring problems to address and consider. Planetary boundaries are being exceeded or increasingly tested (Rockström et al. 2009;Steffen et al. 2015), with direct and flow-on impacts between boundaries (Lade et al. 2020), including climate change (IPCC 2021) and biodiversity loss (FAO 2019;IPBES 2019) and their respective consequences for access to basic human needs, quality of life, and migration patterns (Kaczan and Orgill-Meyer 2020). While front of mind for many as we seek to manage and rebuild from the pandemic, the need to mitigate the risk of future zoonotic diseases (De Sadeleer and Godfroid 2020) is yet another long-standing significant challenge to face. Complex and wicked problems are also visible through rising inequalities across and within countries (Wilkinson and Pickett 2009;Stiglitz 2015;Balestra and Tonkin 2018;Alvaredo et al. 2018), which impact trust, shared visions of a desirable society (Bain et al. 2019), and the institutions of government responsible for delivering those visions (McGrath 2017).
The above concerns and many more are compounded by muted or insufficient progress on the international agreements intended to help address them, such as the United Nations Sustainable Development Goals (SDGs) (UNGA 2015). Similarly, Australia's Report on Government Services (RoGS) annually documents how slow and inadequate progress is in many areas of national and subnational social policy (such as education, justice, emergency management, health, community services like child protection, and housing and homelessness) (PC 2022). This is despite repeated attempts at reform. The SDGs and RoGS articulate and track progress on shared visions, but also demonstrate that merely coming up with plans and frameworks achieves little without their effective implementation. Thus, attention also needs to be given to what stops governments from delivering on shared visions, like the SDGs and RoGS, and steps taken to improve the effectiveness and efficiency of their efforts.
Every decision within government forms an aspect of public governance. Every actor within government, be they elected, members of the judiciary or public servants, forms part of that governance infrastructure. While elected officials may set the co-ordinates for where society is going through their promises or announcements, how the destination is reached is heavily influenced by public servants at all levels. Whether through influencing political decisions, or developing and implementing policy and legislation to provide for them, public servants are what power government action. That is, much of the governance of government action occurs through the public service-the worker bees of government. Hence, the capability and capacity of our public sectors heavily influence success in achieving the desired public policy outcomes. Correspondingly, in this paper, governance is specifically considered from the perspective of activities leading to and arising from public-sector decisions and, in particular, the role of public servants as public decision-makers.
Simply declaring that public servants need to do more or are somehow wilfully failing in their responsibilities is, however, unhelpful. Public servants are not working to actively impede enhancement or protection of our way of life; it is their way of life too. Indeed, many choose their careers to address complex problems more actively. They are, however, constrained in their attempts to do so by a multitude of influencing factors (Bolton 2020). Public decision-makers fail to consistently achieve stated objectives not because of corruption, laziness, or lack of will-though on occasion these will play a role-but rather the complexity of the operating environment they find themselves in.
Nevertheless, the repeated failure to deliver on community expectations across public policy domains is not something that can be brushed aside or accepted as 'the way things are', or 'the best we can do'. On the contrary, the pandemic has highlighted that we can and must do things differently to address shared problems quickly (WHO 2021; Apuzzo and Kirkpatrick 2020). If our way of life is to be enhanced or maintained within planetary boundaries, new governance approaches deliberately targeting weaknesses in current methods are needed.

A systems governance approach
One such approach could be to recognise public decision-making for what it is-a complex system-and bring systems thinking into the equation. We live in systems, we work in systems, and yet we try to solve problems within them by taking a siloed or linear approach. It is illogical to think this could work, and yet, we do it over and over again. Government especially is largely structured in siloes, from minister to street-level bureaucrat. Moreover, there can be a tendency to view public problems through the lens of what matters in the moment or a single policy domain, rather than to sit with and find ways to conceptualise the, at times, Escher-like whole. Recognising this, Meadows' work on leverage points (LPs) provides a useful framework for understanding where and why political and government decisions become 'stuck' (Meadows 1999(Meadows , 2008; see Box 1).
On balance, Meadows argued the levers can be ordered hierarchically: she considered shallower leverage points within the hierarchy (e.g. numerical parameters, the size and structure of buffers, and stocks and flows within the system) easier to change, but ultimately less impactful upon the overall functioning of the system-though also argued, this is where much of our public debate focuses; in the middle, she identified leverage points relating to the system's overall design and feedback mechanisms; finally, she argued deeper leverage points within the hierarchy (e.g. transcending paradigms, the mindset from which systems emerge, and the goals of the system) are harder to employ but more likely to lead to system transformation if successfully applied.
A significant caveat is that systems are complex and unpredictable. It is entirely possible that in some instances, the linearity implied by Meadows' framework and its application within this paper will not exist. Indeed, Meadows noted differing contexts may mean that positions within the framework shift on occasion (for example, delays may operate as deeper points of leverage where their length is able to be altered). Still, while recognising this and perhaps illuminating the difficulty of breaking from linear thinking, the leverage points are generally discussed in a hierarchal fashion by Meadows and much of the literature and that recognition and approach are also applied here.
An increasingly popular way to visualise the leverage points in recent years has been the introduction of the 'iceberg model' (Bosch and Smith 2007;Davelaar 2021). The iceberg represents lower-order leverage points as the visible ice (being shallower, more tangible, and easier to predict the behaviour of), while middle-to higher-order leverage points are represented by ice below the surface (being deeper, harder to conceptualise, and predict, but more impactful if engaged). Extending the iceberg metaphor to the public decision context, attention tends to be focussed on the overt and easily engaged decision or problem elements (the lower-order or shallow leverage points). However, as with icebergs, it is the more covert or less tangible elements (the higher-order or deeper leverage points) which hold the greatest potential for transformative impact.
For example, debates about the amount of social support or welfare provided to citizens are often topical and impassioned, but largely unresolved, as regardless of whether people are offered $x or $x ± y in welfare payments, the debate is focused on applying the lowest-order leverage points (numerical parameters) and the system will essentially continue to function as it has before. The value of 'y' will make a difference to some individuals, but the system itself will not change. A more significant change to the system would, for instance, be the introduction of a universal basic income which would more radically alter the structure and rules of the welfare system.
The value of systems thinking as a tool to enhance the achievement of sustainability has been drawing increased attention. Fischer and Riechers (2019) argue that Meadows' framework is an 'under-recognised' tool in the field of sustainability and propose that, 'conceptual, qualitative empirical or quantitative empirical work' drawing on the strengths of the framework may, 'yield both practical and theoretical advances'. Egerer et al. (2021) applied leverage points as a weighting system to understand and prioritise climate change adaptation measures within the Saxony agricultural sector. Further, a recent special issue on the topic identifies nine questions to help drive research and practice aimed at sustainability transformations through the application and consideration of leverage points (Leventon et al. 2021a). Relatedly, the Earth System Governance (2018) community 1 presented a research framework aimed at mobilising and coordinating research efforts reflecting the rapid evolution, emergence, and increase in complexity of the challenges humanity faces. This framework has four focal points: transformations, inequality, the Anthropocene, and diversity. Transformations are particularly relevant here, and the need for them is articulated by Burch et al. (2019) from three angles: (1) governance for transformations-decision-making that facilitates the conditions necessary for transformations to occur; (2) governance of transformations-decisionmaking that regulates or oversees transformations underway; and (3) transformations in governance-alterations to the how and what of the decision-making practice itself.
The complex problems flagged earlier suggest that successful governance for and of transformations has been lacking to date: The conditions for transformation have not been widely established, and where they have (e.g. in relation to altered food and energy systems to help address the climate crisis), the anticipated transformations have not been as swift or impactful as hoped. In recognition of this, perhaps our governance systems need to transform first so as to better position the lead out of more sustainable futures. Given the complex and contested governance operating space that exists, and the muted impact of the hundreds of billions spent to achieve desired public outcomes (PC 2021), it is hard to argue transformations in governance are not required.
The aforementioned multitude of influencing factors decision-makers must contend with are also complex, not only in number but also in function. Decision-making influences have both transformation-enabling and -inhibiting traits. Similarly, influences are both characteristics of the public decision-making system that leverage points can act upon and, pending the circumstances, themselves be expressions of leverage points. Further, influences can represent multiple levers, with the dominant lever expressed in any particular situation depending on the decision context and actors involved. For example: evidence 2 is acted on by and a product of the system, through which parameters are created and responded to; however, evidence may similarly act as negative (LP8) or reinforcing feedback loops (LP7), pushing the system in one direction or another.
The variable nature and role of decision-making influences, as structural or actor-based system elements, add additional complexity: some of the influences are relatively fixed in nature (e.g. the Institutions 3 within which decision-making occurs, Legislation 4 that imposes requirements upon decision-makers); some influences manifest in different ways pending the actions of individuals (e.g. the Personal characteristics of decision-makers, 5 the Engagement 6 approaches taken to communicate decisions, and indeed the Framing 7 of such engagement); and still others have the ability to reflect both structures and actors (e.g. Complexity 8 ). This latter variability in influence nature is perhaps what stymies some decision-makers from becoming policy entrepreneurs within their roles, as they do not realise the extent of their capacity as actors with the ability to influence system outcomes (Bolton 2020).
Considering the increased scholarly interest, the demonstrated need to accelerate sustainability and governance transformations, and the previously identified influences upon public decision-makers, this paper seeks to answer the following research question: can the influences on public decision-makers be linked to the leverage point framework? If so, what does that suggest about where efforts can be focused to drive transformations in governance for more sustainable outcomes?
Defining the system under consideration as the space and processes embodied by public decision-makers and the institutions within which they act to deliver optimal public outcomes, this paper commences by empirically exploring the relationships between decision-making influences and leverage points. With these in hand, consideration is given to how leverage points might be deliberately used to encourage enabling manifestations of the decision-making influences.
A key outcome of this consideration is the suggestion that decision-makers can and ought to apply tools that exercise leverage one point deeper in the leverage point hierarchy to drive system change. Finally, areas of caution are flagged, tempered by a recognition that the choices made through our governance systems today will heavily determine the nature of the Anthropocene 9 experienced by future generations.

Identification of system variables
Public decision-making system variables or influences were identified previously through inductive thematic analysis of interviews conducted with 35 current or former public servants associated with the Victorian Public Sector (VPS). The VPS is the subnational civil service responsible for supporting the State of Victoria, one of Australia's six federated states. Participants ranged in seniority from frontline or street-level bureaucrats to organisational leaders, and collectively represented all 2017-18 Victorian government departments. Interview topics included decision-making approaches and considerations, definitions of evidence and sustainable development, awareness of the SDGs, and participant's suggestions of changes needed to enhance public decision-making. The interviews did not explicitly reference system leverage points or thinking, nor were participants explicitly asked to identify factors influencing their decisions which, as stated, were subsequently identified through inductive thematic analysis. The full list of decision-making influences and their definitions is provided in Appendix A.

Analysis of influence-leverage point relationships
The 40 decision-making influences identified through the prior thematic analysis were each considered in terms of their potential to act as each of Meadows' 12 system leverage points. This involved: (1) repeatedly reading the descriptions Meadows (1999Meadows ( , 2008 provides, in conjunction with the author-developed definitions for each decision-making influence and the interview text coded to those influences; (2) making annotations as to why a potential match was considered to exist or not; and (3) repeating the latter steps 4-7 days later to confirm the decision, until no further changes were recorded (this occurred after a fifth review). The final rationale for the matches made is included in Appendix B, and further debate on the matches identified through this process is welcomed.
The results of this latter process enabled simple calculation of the number of intersecting decision-making 9 Crutzen (2002) introduced the idea of the Anthropocene as a new geological age, arguing mankind's impact on the planet has become a 'significant geological force' which has shifted planetary functioning from the relatively stable functioning of the Holocene to unknown territory.
Footnote 3 (continued) functioning. Further, how these impact who has authority to consider and make a public decision and also government inertia (designed and unintentional) in responding to perceived needs for public decisions. 4 Legislation-the legal requirements acting as opportunities and barriers, such as inbuilt policy resilience. 5 Personal characteristics of public decision-makers-the skills, experience, attributes, and personal capabilities of public decisionmakers. This includes their values and motivation, willingness or perceived ability to be frank and fearless, and self-perceived ability to influence public decisions. 6 Engagement-how (and if) communication with stakeholders occurs and the framing of that messaging. 7 Framing-discussed in the sense raised by Lakoff (2014), and Tversky and Kahneman (1981), and how it is used to present ideas more or less favourably. 8 Complexity-the array of considerations within and of decisions, and how this leads to increased uncertainty and public decision-makers feeling overwhelmed. influences per leverage point (LP). Considering the number of matches as a marker of opportunities for the leverage points to be applied, the latter process similarly enabled consideration of which leverage points have the most potential to be active within the public decision-making system in Victoria.
Recognising the frailty of using a total numbers approach only, the literature was searched for other leverage point ranking methods and, finding nothing of relevance at the time, novel alternate ranking approaches were developed and tested. The most meaningful of these, a reverse linear weighting, applied Meadow's heuristic of a hierarchy to provide a comparative value to each leverage point. Leverage point 1, the power to transcend paradigms, having the most power to alter a system was given twelve points. LP2, the mindset out of which systems arise, being the second most powerful was given eleven points, and so on, down to LP12, constants, parameters, numbers, which, having the least power, was given one point. The number of intersecting decision-making influences per LP was then multiplied by these corresponding weights to better reflect Meadows' hierarchy within the ranking of leverage point prevalence in the VPS (i.e. leverage point 1: 12 weighted points × 7 influences intersected = 84).
While still a simple measure and, as aforenoted, in some contexts the linearity it implies may not be reflective of systems functioning, for the purposes of a general comparative approach to the total number of matches, this weighted approach was found to be valuable. Assigning values to leverage points to enable ranking is also an approach taken by Egerer et al. (2021). Table 1 provides an overview of the influence-leverage point relationships identified. It illustrates that all decision-making influences have the potential to operate as multiple leverage points and vice versa. It further shows a universal relationship between influences and reinforcing feedback loops (LP7), and a near universal relationship between decisionmaking influences and the power to alter system structures (LP4).

Influences-leverage points relationships
An annotated rationale for each of the 220 identified relationships is included in Appendix B. The results of this analysis are likely to have applicability to other jurisdictions for two reasons: (1) participants spanned the gamut of roles, responsibilities, and policy areas, ranging in seniority from frontline or street-level bureaucrats to organisational leaders across the sector; (2) the Victorian Public Sector operates within a Westminster system of government and serves a population of approximately 6.7 million people (ABS 2020), attributes which are likely to be reflected elsewhere.

Priority leverage points in the Victorian Public Sector
Ranking leverage points by the total number of related decision-making influences (Table 2, column 2, 'influences intersected') further demonstrates that LP7, reinforcing feedback loops (40/40 matches), and LP4, the ability to evolve or change the system (39/40 matches), are the most accessible leverage points within the public decision-making system in Victoria. 'Accessibility' is considered from the perspective of the number of opportunities to effect change on the system, as it could be argued that an increased number of opportunities to intervene makes a decision-making influence or leverage point more likely to be used and applied within governance processes, and, therefore, more practically valuable to decision-makers. The latter leverage points, LP4 and LP7, each have almost double the potential number of influence-leverage point interactions as the next, LP11, the size of buffers (22/40).
When the reverse linear weighting is applied, the power to alter system structures (LP4) and reinforcing feedback loops (LP7) remain the most dominant leverage points, followed by LP3, the origins of paradigms (see Table 2, column 4 'weighted ranking'). That is, when considered through the lens of the 40 influences decision-makers must contend with, the dominant leverage points under both frequency and weighted analysis are the ability to evolve or change the system (LP4) and reinforcing feedback loops (LP7).

Leverage points as catalysts for enabling influences
This analysis found that the decision-making influences upon public decision-makers can clearly be linked to system leverage points. Where they are, one of the most striking things is the universal relationship between decision-making influences and reinforcing feedback loops (LP7), and the almost universal relationship between influences and the leverage point of self-organisation or system evolution (LP4). This is not to suggest that every influence reinforces the status quo or alters the system structure in the same way-a reinforcing feedback loop for Ministers 10 would be different  (6) 6. Information flows (7) 5. Rules of the system (8) 4. Ability to evolve or change the system (9) 3. System goals (10) 2. Origins of paradigms (11) 1. Ability to transcend paradigms ( to a feedback loop for Evidence 11 or Risk. 12 Nevertheless, reinforcing feedback loops exist for all of the influences. These commonalities shift consideration beyond which influences are best placed to stimulate system change, to which leverage points have the most potential to do so. That is, knowing the decision-making influences within this decision-making system, and that they can have both positive and detrimental impacts, we can turn our minds to which leverage points can be applied to encourage positive influence expression. The benefit of this is that, rather than focusing on determining which decision-making influences are most impactful and determining how to individually master all forty of them to activate transformations, efforts can instead be applied to particular leverage points to simultaneously drive change across multiple influences and throughout the decision-making system. For example, if a concerted effort were made to identify, confirm and, where necessary, alter reinforcing feedback loops within the public decisionmaking system, the behaviour and outcomes of multiple if not all influences within that system would be altered. In a way, this approach is trying to achieve the same outcome as Abson et al. (2017). In exploring the potential to group leverage points based on shared characteristics (intent, design, feedbacks, and parameters), they mused that further research is needed to determine if there is a differentiated effect between a single or combined leverage point focus. The suggestion here is that a single lever focus reflecting an aggregation of influences upon public decision-making may be more transparent and impactful. That is, a single lever approach may make it easier to identify existing path dependencies and the likely flow-on impacts of deliberate system change. A single leverage point focus may also see attention on the whole system rather than subcomponents within it (Kim 1999), giving rise to greater appreciation of the overall context and synergies. Similarly, a deliberate choice to apply a single leverage point across all or many decision-making influences may streamline the focus of system reformers enhancing efficiency through reduced need to identify and corral the 'energy for change' recognised as necessary by Birney (2021).
But which leverage point ought to receive this attention? Applying a purist approach, one would adopt the leverage point hypothesised to be most impactful, transcendence of paradigms (LP1). However, transcending paradigms within public decisions is arguably out of reach for many public servants and thankfully so, as some might question the legitimacy of non-elected officials seeking to drive transcendence of paradigms within public decisions (Leventon et al. 2021a). Returning to the findings here, one could apply the leverage points that all or most of the decision-making influences are interacting with, the power to alter system structures (LP4) and reinforcing feedback loops (LP7). However, given that any change made within these leverage points would still be operating at the level of the existing dominant Based on the total number of decision-making influences intersected; *weighting calculated by multiplying the attributed LP weight (i.e. assigning a score of 12 to LP1, a score of 11 to LP2, etc.) by the number of decision-making influences intersected system dynamics, it is arguable that the system will respond by seeking to restore its current equilibrium.
Perhaps instead, efforts could be focussed on the leverage points one step deeper than each of those considered to have the potential to be universally active within the current decision-making system (i.e. LP4 -> LP3; LP7 -> LP6) (see Fig. 1). As is discussed in the following sections, applying a deeper leverage point in this way could drive system change by effectively disturbing the status quo just enough to override it. Hence, a one-deeper approach may balance the practical constraints and considerations of decision-making within public institutions with the need for transformation in governance, in a democratically sound way.

Altering system structures with altered system goals
As noted, almost all influences have the potential to alter system structures (LP4), thus LP4 is a lever with the potential to be highly impactful within current governance arrangements. Focussing on the next leverage point deeper in the hierarchy to change or clearly restate system goals (LP3) will provide opportunity to deliberately construct a holistic narrative for action around which the system will respond and likely shift. If this approach were desired, then a focus on the decision-making influences identified as having the potential to alter system goals (i.e. Economics, 13 Role of Government, 14 and Commitment to Concepts 15 ) could help.
For example, while our policy-making commitment to the concept of sustainable development is questionable at times (Bolton 2021;Sachs et al. 2021), advocation and affirmation of it continues. Furthermore, agitation for an altered focus on economics and the role of government are increasingly visible in both academic and government circles. looking to change economic paradigms, and reaffirm the role of government to enhance efforts toward a just and prosperous life for all. Similarly, regulatory practice has been seen to evolve from prescription to co-and self-regulatory regimes (Sparrow 2020) and even general duties (Edwards et al. 2020). Hence, it seems there is some recognition of the systemic impact altered expression of the decisionmaking influences of Economics, the Role of Government, and Commitment to Concepts has upon system goals, and a preliminary willingness to deviate from current paths. Revisiting system goals may also have additional benefits, such as stimulating a virtuous cycle to alter or at least revisit the acceptability of the mindsets which led to our current system. However, while noble and potentially quite effective, as demonstrated by the incorporation of such considerations in  14 Role of government-conceptions of the purpose of public decision-makers and government as a whole, and the impact this has on licenses to act and individual decisions to influence (or not) particular outcomes. 15 Commitment to concepts-acceptance and application of sustainable development or the SDGs, reflections on the application and impact of these concepts/tools on public decision-making.
formal public decisions (New Zealand Government 2019), few public servants will feel comfortable driving changes in system goals without already having some indicative authority to do so from elected or senior officials. That is, even if it is theoretically possible for public servants to alter system goals, individuals' core beliefs or wider societal norms on the role of the public sector may prevent them from doing so (Sabatier 1987), suggesting this leverage point is also inaccessible to many public decision-makers.

Addressing reinforcing feedback loops through altered information flows
The association of recurring reinforcing feedback loops with each decision-making influence in this system explains why switched-on, well-meaning, self-efficacious, public decisionmakers may give up on achieving stated objectives. Having repeatedly hit up against these loops they reason there is nothing more they can do. To use the iceberg analogy, reinforcing feedback loops sits just below the water line-If we look, we can see them but, we do have to look, and, when we do, we may struggle to see beneath them with our existing tools. Encouragingly, considering the relationships between influences and leverage points (  20 and Framing, 21 have the potential to alter information flows. Similarly, a focus on altered information flows is also an approach that aligns with popular and ever-increasing calls for evidence-based or informed decision-making to be an integral part of public decision-making processes (OECD 2020;Head 2008). Moreover, practical examples of an altered information flow approach already exist. For example, cross-government networks that facilitate information exchange and learning (both across and within jurisdictional boundaries), staff-led initiatives to focus attention on desired objectives (Bryant and Thomson 2021), and investment in improved data management systems (EPA 2013, 39;VAGO 2013, 19-20). There is also increasing interest and appetite for the use of technology as potential decisionmaking aids (PC 2020).
As fields which excel in collecting and presenting information in novel ways, artificial intelligence (AI) and data science provide an array of example tools which could sit within an information flow altering toolbox. By providing new pathways for existing or previously uncollected data to reach decision-makers in novel and status quo-disrupting formats, these technologies can shift the governance focus above reinforcing feedback loops (Miller 2020).
Brenner (2012) notes a risk of 'drowning in a sea of data and thirsting for some theoretical framework with which to understand it'. AI and advanced data science tools can help avoid this trap. These tools excel at creating novel information flows to connect information (new or pre-existing) and decision-makers in new ways. These flows draw attention to hereto unrecognised knowledge and enable robust predictions and assessment of possible futures. In doing so, they better position decision-makers to make informed, holistic policy improvements. For example, Bayesian networks, a form of causal probabilistic modelling, can cut through intractable data-collection loops to identify and rationalise priority interventions. This enables decision-makers to shift their focus from data collection and analysis to the business cases for one to two calculated front runners. Further, applying approaches in a modelled context first, per the Bayesian Network example, may provide additional comfort to decision-makers looking to employ LP-informed approaches. Such modelling provides timely, but safe opportunities to test and explore solutions before implementing them at scale or directly within the community. At a grander scale, the UK Ministry of Justice has commenced linking tens of millions of data records to better understand interaction patterns within their social and criminal justice systems to enable a more holistic approach to identifying 'what works' (ADR UK 2021; Office for National Statistics 2021). If successful, this approach could mark a shift towards more fully understanding the impact public institutions and point in system decisions have on individual's life trajectories, and where changes in well-meaning but ultimately poor decisions and processes are required.
Examples such as these ought to provide confidence to public decision-makers that enhancing information flows (LP6) is an accessible, yet systematically deep, leverage point which can be adopted or at least piloted more widely. However, while AI and advanced data science hold much promise, the latter example applications are far from the norm, and ultimately a suite of tools that support enhanced and novel 16 Cognitive biases defined here as, "Heuristics, personal anecdotes and pain points that influence decisions, and their associated unintended consequences." 17 Collaboration defined here as, "How actors work together, particular public decision-makers and parliamentarians." 18 Engagement defined here as, "How (and if) communication with stakeholders occurs and the framing of that messaging." 19 Evaluation defined here as, "Both consideration or focus on outcomes and evaluation of public decisions". 20 Evidence defined here as, "Consideration of evidence or information in decision-making, what 'counts' as evidence (e.g. qualitative and quantitative), and availability of data." 21 Framing defined here in the sense raised by Lakoff (2014), and Tversky and Kahneman (1981), and how language is used to present ideas more/less favourably.

3
information flows is needed to cater to the varied needs and appetites of differing decision-makers in differing contexts.

The need for caution
There is, of course, a need for caution when intervening in systems and to be particularly cognizant of proverbial butterfly wings in generating change elsewhere. Any plans to intervene may benefit from a third-party reference group or steering committee, providing advice but not approvals, to ensure those leading systems work do not lose sight of critical connections and consequences.
Still, systems are constantly incrementally changing, whether by design or in response to shifts within the other systems they are a part of. Further, part of our governance considerations when choosing whether or not to 'dance' with system change must be whether existing systems are fit for purpose and operating as desired. As articulated earlier, this is arguably not the case in relation to governance for and of sustainability transformations.
Further, consideration of what our institutions need to be and do to facilitate a good Anthropocene must focus not just on the how and the what of our institutions, but also the when. The sixth IPCC Report (2021) notes both, "climate change is already affecting every inhabited region across the globe with human influence contributing to many observed changes in weather and climate extremes", and that, "global surface temperature will continue to increase until at least the midcentury under all emissions scenarios considered", because, "there are already substantial committed changes associated with past greenhouse gas emissions". More succinctly, there is little time to act. In the context of the Anthropocene and the planet's threatened tipping points, there is a need for wellconsidered and efficient catalysts of change now.
Given the need for caution in stimulating transformative system change, limited time in which to undertake those transformations before further adverse consequences are 'locked in', existing decision-making influences inhibiting those necessary transformations, and, public servants' (dis) comfort with their role in such transformations, it would seem prudent to focus on holistic, efficient, transparent, and accessible mechanisms for change to our governance systems. Altering information flows can meet these criteria.

Making space for public decision-maker agency
Hypothetically, enhancing information flows (LP6) overcomes a critical driver of disappointing public decisions, reinforcing feedback loops (LP7). Enhancing information flows also appears to be a more accessible lever to everyday bureaucrats and therefore more likely to be mainstreamed within business-as-usual decision-making than changes to system goals (LP3): it is one thing to challenge reinforcing feedback loops within your decision-making sphere, but quite another to question dominant societal paradigms.
Starting with more accessible levers may act as a gateway to alter public decision-maker mindsets and empower them to see their influence and role-modelling potential within the system (Nielsen et al. 2021). This may also enable greater consideration of the how and why of decision-maker, community, and organisational values (Horcea-Milcu et al. 2019), allowing the 'inner dimensions' described by Woiwode et al. (2021), as well as more traditionally focussed upon techno-scientific advances, to begin to influence sustainability transformations.
In time, by demonstrating the value and pathway to employing deeper leverage points, introducing novel information flows (LP6) may engender greater comfort and authorisation for public decision-makers to embrace deeper leverage points. It may also see greater application of what Newell et al. (2021) refer to as 'spiral' scaling of transformation, where the pathway forward involves dynamic shifts between the use of shallow and deep scale interventions, pending the outcomes of public decision outcomes along the way-a leap-consolidate-repeat approach to transformation. That is, applying a one deeper approach now may help stretch us beyond the use of systems thinking as a framework for navigating current paradigms toward a broader realisation of the nested nature of the systems within which we live and work (Leventon 2021a, b).
In the meantime, creating a toolbox of approaches that can be employed across a range of public decision-making settings may empower public decision-makers to exercise their individual agency to apply tools one step deeper. Doing so may disrupt the reinforcing feedback loops that have prevented the achievement of collective aspirations and set transformations in governance in motion.

Conclusions
Humanity is at an inflection point: business as usual is not possible if the well-being of current and future generations is to be secured. Continuing to careen down a road of unsolved complex issues risks a significant question for democracy and public institutions: if governments and societies do not endeavour to improve imperfect things for the betterment of all, what is their role?
This paper sought to explore how transformations in governance might be enabled. The suggestions here are not posited to usurp or upend existing democratic processes, quite the opposite. The core objective of this research is to enable democratically expressed aspirations to be realised within the realities of our governance and planetary systems. It couples: prior research identifying barriers to public decision-making for and of the sustainability transformations global, national and subnational governments have repeatedly reaffirmed; calls to increase empirical research on applications of Donella Meadows' leverage point framework within sustainability science; and the recent Earth System Governance transformation agenda.
Influences upon public decision-makers were repeatedly linked to Meadows' framework, enabling a better understanding of which leverage points are currently dominant within the public decision-making system of Victoria. It is clear in this assessment that two moderately to highly powerful leverage points are routinely present within Victoria's governance system: reinforcing feedback loops and the ability to alter system structures.
With this knowledge in hand and reflecting on Meadows' core leverage point argument (that higher-order influences are harder to apply but also more impactful), this paper contends that employing tools reflecting leverage points one step deeper in the hierarchy may enable decision-makers to disrupt current system machinations.
It would be remiss not to acknowledge that there is danger and unpredictability in system change, certainly there is a danger that the leverage point framework cannot be applied with the linearity implied. However, it is also imperative to consider the very real and well-recognised risks of not altering the unsustainable trajectory current governance systems enable. While deploying deeper leverage points would require careful planning and orchestration to anticipate and mitigate perverse outcomes, doing so would likely prove transformational and deliver efficient, transparent, change in a time frame better reflecting the urgency of the complex challenges facing decision-makers today. Moreover, if little else, enhancing information flows ought to help drive improvements in the evidence-base of our public decisions.
Employing an approach which enhances information flows within public decision-making processes through the increased use of existing and novel approaches appears to be a relatively accessible and benign way to achieve stated public objectives. A future research agenda could look to confirm or dispel the one step deeper model through the identification and piloting of a suite of information flow enhancing tools. Having a suite would enable the uptake of deeper leverage points in a way that best fits the decision-making context and personal agency of the decision-makers involved. Enabling public decision-makers to counteract dominant feedback loops in democratically sound ways may just be the lever needed to master the influences currently holding them, and us, back from achieving our goals for sustainable development and a good life for all.

Appendix A: Influences and their definitions
Definitions of the 40 influences (identified through inductive thematic analysis of 35 public servant interviews) are as follows:

Alignment of sustainable development and public decisions-reflects discussion on assumptions about alignment between sustainable development (SD) or the sustainable development goals (SDGs) and public decisions (PDs), as well as considerations and perspectives on the principles of SD.
Appetite for change-talks to expectations, willingness or calls for change, i.e. alignment with social values, how compatible SD is with other values, and the recognition of need for the SDGs within Australia.
Businesses/non-government actors-applications of ESD by actors with influence outside of government and lobbying of government by those actors.
Central and review agencies-the role of central and and pain points that influence decisions, and their associated unintended consequences.
Collaboration-how actors work together, particularly public decision-makers and parliamentarians.
Commitment to concepts-acceptance and application of sustainable development or the SDGs, reflections on the application, and impact of these concepts/tools on public decision-making.
Complexity-within and of decisions, how this leads to increased uncertainty and public decision-makers feeling overwhelmed.
Culture-considerations of culture, people or frameworks, and their impact on public decision-making practices.
Economics-understanding and application of different schools of thought, growth as a goal, externalities, monetary/ financial costs, and Maslow's hierarchy of needs.
Election cycles-the impact of elections on decisionmaking and actor's behaviours.
Engagement-how (and if) communication with stakeholders occurs and the framing of that messaging.
Evaluation-both consideration and focus on outcomes and evaluation of public decisions.
Evidence-consideration of evidence or information in decision-making, what 'counts' as evidence (e.g. qualitative and quantitative), and availability of data.
Framing-covers framing in the sense raised by Lakoff (2014), andTversky andKahneman (1981), and how it is used to present ideas more/less favourably. 22 Funding-availability of economic resources to facilitate public decisions to be made and implemented.
Governance-arrangements within and across public decisions to ensure their rigour, such as accountability, KPIs and transparency. 22 Lakoff, George. 2014 Implementation-putting decisions into practice, including whether stakeholders have the capabilities necessary and expected of them to achieve the desired outcomes.
Institutions-characteristics of the machinery of government, such as the Victorian Public Sector and departments and agencies within it as individual and combined institutions, as well as the structure of those institutions and administrative tools to support their functioning, plus how these impact who has authority to consider and make a public decision and also government inertia (designed and unintentional) in responding to perceived needs for public decisions.
Jurisdiction-consideration of where the Head of Power for a decision sits across jurisdictions, as well as how that impacts willingness to act.
Leadership-covers the concept of leadership as well as the need or demonstration of leadership by individuals, senior officials, ministers and/or organisations.

Legislation-legal requirements acting as opportunities and barriers, such as inbuilt policy resilience (inertia).
Mandate-the provision or lack of authority (e.g. crises, external scrutiny, political/party driven, expectations and responsibilities conferred on organisations) to make a decision in a particular area.
Media-presence, use and impact of the media in shaping public decisions.
Ministers-the position, interests, incentives, and capabilities of ministers (grouped, as ministers are not the primary focus of this research).
Paradigms-the impact of established 'rules' within a society that govern the way it thinks and acts in determining what is possible within public decision-making, including the presence of luck or serendipity, and focus on the shortterm (n.b. growth is covered separately under economics).
Public decision-makers' understanding-public decision-makers' understandings (including definitions) and awareness of sustainable development and the SDGs, including confusion with one another and the Millennium Development Goals (MDGs).
Public decision-making considerations-success of prior or parallel projects, availability and reliance on technology, and government priorities (including overall public decisionmaking goals, competing priorities across portfolios, the comparative priority of ESD and SDG impacting their status as goals within decisions). Also includes less commonly mentioned factors considered in making public decisions.
Public decision-making processes-the stages and act of making public decisions, from understanding and follow through of the whole public decision-making process to individual components such as problem definition, options identification and analyses, and recognition of the importance of separating stages to retain objectivity. Also includes the transparency of the process and how it is communicated within briefing notes.

Personal characteristics of public decision-makers'-
the skills, experience, attributes, and personal capabilities of public decision-makers. This includes their values and motivation, willingness or perceived ability to be frank and fearless, and self-perceived ability to influence public decisions. For example, dichotomies were present between those felt they could have meaningful impact, compared to those who felt their impact was tightly constrained and/or that it was not their role to try to influence outcomes.
Politics-the impact of political beliefs, gameplay, party dynamics, power struggles and allegiances.
Public awareness-public understanding, awareness of and support for SD or the SDGs.

Relationship between bureaucracy and ministers-how public decision-makers and their ministers interact and view each other.
Relationship between Public decision-makers and community-how public decision-makers and the community interact and view each other, the trust between them and the impact of this on connections between them.
Resources-capability/capacity-the amount of full time equivalent (FTE) staff available and the relevance of their skills and experience to the task at hand, as well as the use of consultants and citizen scientists to undertake work considered to be within the remit of public decision-makers. Also, a general catch all for where 'resources' are mentioned outside of the context covered in other resource-relevant influences (e.g. around data, funding, time, institutions).
Risk-appetites for taking decisions outside of tried-andtrue approaches, and behaviours driven by an avoidance of criticism.
Role of public decision-makers/government-conceptions of the purpose of public decision-makers and government as a whole, and the impact this has on licenses to act and individual decisions to influence (or not) particular outcomes.
Scale-the size of problems requiring public decisions. Strategic planning-proactively utilising processes to anticipate future policy needs and drive consideration of a more strategic than reactive view, culminating in reports that set agenda and measurable goals.
Time-both as a capacity-limiting resource and as a consideration within public decisions.

Appendix B: Influences and leverage points
Detailed annotations as to why influences were deemed to act or have the potential to act as leverage points. Connections are based on the question, Does the influence act as (or have potential to act as) as a Leverage Point in public decision-making for sustainable development?