Strengthening policy capability: New Zealand's Policy Project

Abstract Policy practitioners apply tools and frameworks to policy challenges in order to improve social, economic, and environmental outcomes. Could such tools and frameworks be applied to improve the quality of policy design itself? In 2014, prompted by evidence of widespread inconsistency in the quality of policy advice being produced across agencies, the New Zealand Government launched the Policy Project. It deployed policy analytic tools and frameworks to investigate current practice in policy design to improve the quality of policy advice across the whole of government. Through collaborative methods, the Policy Project identified and codified what quality policy advice looks like and the skills and processes needed to produce it. We review the context and creation of the Policy Project, its contributions, evidence of its impacts, and prospects for its replication across other public sectors.


Introduction
The rise of policy analysis as a profession has been facilitated by the on-going accretion of knowledge about effective approaches to policy development (Mintrom and Williams 2013;Radin 2000;Weiss 1980). Along the way, there have been efforts to outline appropriate content of university-based public policy programs (Wildavsky 1979) and various attempts to establish the set of technical skills required by policy analysts (Weimer and Vining 2017). Efforts have also been made to highlight the people skills and political nous required to support effective policy development and advising (Mintrom 2003;Rhodes 2016). Recently in New Zealand, a unique effort has been made to drive systemic improvement of policy capability across the whole of the government sector. This represents a novel contribution to the continuous effort to improve the quality of policy development and advising. The systemic ambition is especially noteworthy. The operating model and products of the Policy Project provide the focus of this paper. In terms of process, it offers lessons for public management reform that optimises engagement and buy-in. This contrasts with the command-and-control approaches typically employed by central agencies seeking to influence the behaviour and practice of line agencies.
To date, the Policy Project has produced three policy improvement frameworks, which were launched by the Prime Minister of New Zealand in 2016. 1 Here we describe the collaborative processes by which the Policy Project sought to drive systemic improvements in the quality of policy advising. We then review the three frameworks and tools within which the Policy Project has identified and codified what quality policy advice looks like and the skills and organizational capabilities needed to produce it. Following that, we consider emerging evidence of the Project's impact. Finally, we highlight lessons for other jurisdictions and consider the merits of other governments replicating the Policy Project.

The context of the Policy Project
The Policy Project was initiated in 2014 by the New Zealand Department of Prime Minister and Cabinet. Its purpose was to meet a key challenge: "lifting the policy game across the system". This was not the first time a central agency in New Zealand had attempted to improve the quality of policy advice across government. In the 1980s, The Treasury provided the intellectual core for the New Zealand Public Service. Through its influential post-election briefing papers, Economic Management (1984) and Government Management (1987), senior Treasury officials articulated coherent and consistent narratives on directions for policy change. This was highly influential in guiding policy directions (the "what" of policy) but did not promulgate an explicit strategy to raise the quality of policy advising across government (the "how" of policy). In the late 1990s, the State Services Commission (SSC) led a process to promote broad improvements in the quality of policy advice, which extended into people capability issues and case studies of policy units deemed to be high-performing. However, by 2010 lingering concerns about policy quality remained, including about the costs of policy advice, and resulted in the Review of Policy Expenditure and Advice, led by former Secretary of the Treasury, Graham Scott. In hindsight, neither the SSC nor the Scott Review, despite being analytically robust, were viewed as having delivered significant change, in part because their recommendations were never fully implemented. 2 New Zealand's highly devolved public management system gives Public Service Chief Executives and their departments considerable autonomy. This makes whole-ofgovernment reforms especially difficult; line agencies see reform as a challenge to their authority (IfG 2017). The context of New Zealand's Better Public Services reforms 3 , including a call for more joined-up government and the introduction of "functional leadership" 4 roles provided an opportunity to take a systemic approach to improving policy quality and capability. In 2014 there was a sense that: policy was of variable quality within agencies and across the system and that there was no shared view of good policy advice, good policy advisors or good policy organizations; a significant amount of policy advice was short on evidence, showed limited knowledge of user needs and was poorly informed on what had or had not worked in the past; skilled senior policy advisors were few and far between and agencies were competing to attract them; there was a focus on meeting ministers' immediate demands to the detriment of investment in policy capability for the future or "policy stewardship"; cross-government systems for collaboration, alignment and prioritization of both policy development and implementation were weak.
The Policy Project was sponsored by the Chief Executive of the Department of Prime Minister and Cabinet, who became New Zealand's first "Head of the Policy Profession". The role was mandated by the Head of State Services following a proposal to the then State Sector Reform Leadership Group (senior chief executives and functional leaders overseeing the Better Public Services reform agenda). While other functional leaders (in property, procurement and ICT) had received Cabinet mandates, it was felt that the chief executive of DPMC had enough influence already and that ministerial endorsement was both not required and could send the wrong message for a reform programme promoting collaborative change (not mandatory requirements).

A collaborative change process
The Policy Project has been deliberately managed to emphasize the importance of both good process and good results. The Project employed tools of policy analysis to diagnose the current state of policy quality and capability. It applied new methods such as design thinking to co-produce policy improvement frameworks and tools. In doing so, it served to demonstrate the value of those methods to the wider policy community.
Early in the life of the Project, a paper was produced entitled "Narrative and Direction of Travel". 5 This set out the original diagnosis of what needed to change in the policy system. The paper was the product of bringing together the best available evidence and data about policy practices and the workforce, as well as insights from across the New Zealand policy community. Policy Forums to diagnose "the policy problem" and to brainstorm potential "game changing ideas" were part of co-designing a roadmap and work programme for the Policy Project. These Forums involved respectively, policy leaders (deputy secretaries and their nominated thought leaders), 6 early career policy staff, 7 and policy managers. 8 The Forums were intended to tap into the diverse views of the community and to build momentum for action. They were also designed to demonstrate the value of a design-led collaborative process and techniques. Here, the key message was that policy should not be a solely desk-based activity. The views of senior ministers (the "customers" of policy advice) were also This collaborative approach to change management was influenced by the Stanford collective impact model (Kania and Kramer 2011). That model is usually associated with social innovation rather than applications within government. The Policy Project Programme Director felt that the model offered a set of well-defined steps and an infrastructure for bringing organizations and groups together in pursuit of a common goal. In this instance, the dedicated Policy Project team acted to catalyse, facilitate, and drive the co-production process, or act as the "backbone organization" according to the Stanford model. 11 In espousing the Stanford collective impact model, Kania and Kramer (2011) have argued that large-scale social change requires broad cross-sector coordination rather than the isolated interventions of individual organizations. The same can be said of large-scale public management change. The Policy Project's operating model 12of "doing things with not to people", of leveraging good practice and of collaborative leadershipwas influenced by the five conditions necessary for successful innovation enumerated by the Stanford model. These are: a common agenda, shared measurement, mutually reinforcing activities, continuous communications, and a backbone support organization (Kania and Kramer 2011). Next we discuss how each condition was designed into the Policy Project.

1.
A common agenda. The Policy Forums were used to create an agreed or "common" agenda. The overall conclusion from the Forums, reflecting the "collective wisdom", was that, while the policy system was not broken, co-ordinated action was required to meet future policy demands and address current deficiencies. From here, phases of the Project were established. Phase 1 of the Policy Project culminated in the policy improvement frameworks. These were designed to reflect a shared view of good policy advice, good policy advisors and good policy organizations. We review them below. Phase 2 focused on embedding those frameworks and areas for innovation (application of new methods and better use of evidence) and tackling people capability and workforce issues. Work on free and frank advice and "policy stewardship 13 " has been on-going, particularly though speeches made by the Head of the Policy Profession. 14 The early Policy Project narrative argued that while most attempts to improve policy quality and capability focused on the "supply side", the "demand side" was equally important, in particular ministers' willingness to expect and receive free and frank advice and to allow space in budgets and work programmes for capability building and policy stewardship. A parallel policy quality framework for ministers, to help them interrogate advice as "intelligent customers", has been drafted as the first foray into this space. 2. Shared measurement. This remains work in progress. The Head of the Policy Profession has articulated the "future state" as one where the business of providing advice in New Zealand is world-leading (see Box 1). It is a challenge to measure changes in policy quality and capability across the system. Measurable indicators of impact are in development (discussed below). They must be meaningful to the policy community so that they drive genuine improvements in policy practice and capability and do not result in compliance behaviour.

Box 1. A vision for quality policy advice.
Vision for the futureexcerpt from a speech by the Head of the Policy Profession 38 "Let me leave you with some thoughts on what success would look like in 5 years, my vision for the policy function: We provide free, frank, fearless and savvy adviceto support good government decision-making. And we are known for doing so.
Our advice is based on the best available evidence and user insightsan understanding of the real lives of real peoplewe know what works and we keep striving for new and better ways of doing things.
We can identify the 'big issues' (whether future, looming or cross-cutting) and can mobilize to collectively find solutions that improve the lives of New Zealanders.
We develop and can deploy a high-performing policy cadreso we have the right people in the right place at the right time.
We think system and not just agencywe are one policy community that supports the government of the day in the service of the people of New Zealand.
As Head of the Policy Profession I want to ensure that business of providing advicethe policy functionis the best it can be, world leading even." 3. Mutually reinforcing activities. The policy improvement frameworks reviewed below are mutually reinforcing. They contain common themes, such as better use of evidence, giving free and frank advice, and building diverse perspectives into policy. The notion of a "policy improvement infrastructure" highlights the systemic and interrelated nature of the Policy Project frameworks and tools. The Policy Project is a whole-of-government change programme where all parts of the policy system need to be aligned and engaged for success to occur. 4. Continuous communications. The Policy Project leadership routinely brings parts of the policy community together and subsequently communicates discussions from events and workshops. This is usually done though visual "conversation trackers" which are more accessible than a dry report. The goal has been to give the widest possible audience access to current issues, debates and methodological advances. The Project has also sought to cultivate "communities of practice" such as the cross-agency network of policy capability leads, 15 within which agencies are encouraged to share their good or new practices that others can "steal with pride". A dedicated website further supports building communication and conversation across the policy community. 5. A backbone support organization. Kania and Kramer (2011) argued that "The expectation that collaboration can occur without a supporting infrastructure is one of the most frequent reasons why it fails". They also noted that: "In the best of circumstances, … backbone organizations embody the principles of adaptive leadership: the ability to focus people's attention and create a sense of urgency, the skills to apply pressure to stakeholders without overwhelming them, the competence to frame issues in a way that presents opportunities as well as difficulties, and the strength to mediate conflict among stakeholders". The Policy Project team has acted as the catalyst, facilitator and driver of collaborative processes across the policy system. The original Programme Director insisted that the team is "owned" by the collective, not the agency that hosts it (the Department of the Prime Minister and Cabinet). This was done partly to emphasis the systemic focus ("all in this together"), and partly to mitigate any perception of a central agency taking a "command and control" approach to change. The Policy Project operating model reflects an agile and adaptive "create as we go" principle.

Policy project sponsorship and funding
Senior sponsorship from the chief executive of Department of Prime Minister and Cabinet and nominated Head of the Policy Profession, has been an important driver for the Policy Project as was having a programme lead dedicated to collaboration and co-production (Davison 2016). The original Programme Director led the Policy Project from inception in early 2014 until July 2017. In the first three years funding for the Project was based on voluntary piecemeal or yearly contributions from policy agencies. This created a degree of uncertainty and had implications for attracting and retaining project staff. Sustained funding (still voluntary) for a further three years has recently been secured. The search for the right type and level of governance, or "Goldilocks governance" (Washington 2016) has been a challenge, as it is for any cross-agency or whole-of-government initiative.

A policy improvement infrastructurethree Policy Project frameworks
The Policy Project has codified good practice for policy advice, the skills required of policy advisors, and the capabilities required of policy organizations. The three frameworks developed to date constitute a foundation "policy improvement infrastructure". Each of the frameworks and tools have been co-produced by sections of the policy community, and reviewed and agreed by senior policy leaders (deputy secretaries with policy responsibilities). They were subsequently endorsed by the Head of the Policy Profession and launched by the Prime Minister, who both communicated their expectation that agencies should use them. Figure 1 provides an overview of the three frameworks. 16 The frameworks are expected to change over time, and more could be added in future. The design principles underpinning current and future frameworks are that they should be: functional (scalable, adaptable, diagnostically sound); usable (easy to use, accessible, efficient and administratively painless); able to be maintained (iterated and improved, on the basis of use by early adopters and changes in the authorising environment or political context).
We here discuss the key features of each framework.

The Policy Quality Frameworka focus on policy advice
The Policy Quality Framework articulates the characteristics of good quality policy advice. It then sets forth what enables quality work ( Figure 2). The supports needed

Acid tests
Does this equip the decision maker to make fully informed decisions?
Is the real problem (or opportunity) exposed? How well do we enable and support the delivery of quality policy advice?
How well do we meet ons of ministers, customers, and other stakeholders?
How well do we ensure we have the right skills in the right place at me? Figure 1. A policy improvement infrastructure. Source: The Policy Project. https://www.dpmc.govt. nz/our-programmes/policy-project/policy-improvement-frameworks to produce quality work include consideration of multiple perspectives, good commissioning, quality assurance processes and work planning. A package of tools has been developed to assist practitioners at the outset of conducting policy work. These are captured in a "Start Right" package. 17 This set of ex-ante tools is complemented by a range of tools for ex-post evaluation of policy advice, including an assessment tool 18 to review and learn from critiques of policy papers, peer review templates 19 and guidance on internal quality panels, scoring and quality targets. 20 Together, these tools allow practitioners to gain a better understanding of what success looks like when it comes to developing and delivering policy advice. Bridging the ex-ante and ex-post tools, a Policy Methods Toolkit 21 is being developed. The intent is for the toolkit to provide an online searchable repository of information and guidance on analytical methods, including how and when to apply methods such as behavioural insights and design thinking, guidance on the use of data and analytics, and links to analytical frameworks such as Kapasa, 22 a framework for considering the impacts of policy on Pacific peoples (recently launched by the Ministry for Pacific Peoples). The intention is for the toolkit to be built for and by the policy community, whereby agencies will populate the toolkit in areas where they have particular expertise. For example, Statistics New Zealand will develop guidance on data and analytics for policy. Initial content for the Policy Toolkit has been drawn from Policy Project hosted workshops designed to expose the policy community to international experts and innovative methods and approaches to policy. These have focused on topics such as design thinking, 23 public value, 24 co-production, 25 and "people-centred policy". 26 While highlighting new methods, the Policy Project has tried to counter the fetishization of any one new method over another. The programme gives the message that skilled policy practitioners will increasingly need to know something about many methods and most importantly be able to tell what method to use, when, in what combinations, and with what level of capability and investment.
Many themes run across the policy quality tools, notably the need for advice to be "free and frank" and the need for the policy community to get better at collating and using evidence. Where evidence is not available they need to know how to generate it by tapping into user insightsfrom end users or those impacted by the policy, to those implementing policy, to up-stream decision-makers (senior leaders and ministers). While acknowledging the importance of data, the programme has cautioned against conflating correlation with causation, and rejected the suggestion that the sense-making part of the policy advisor's role can be reduced to an algorithm. This is in part to counter simplistic arguments that "the data will tell you everything", sometimes articulated by ministers energized by New Zealand's "investment approach" (Deloitte and NZIER 2016).
The Policy Project has also challenged the traditional stylised "policy cycle", insisting that the policy process depicted is more iterative than linear. As the process proceeds, those involved acquire growing levels of certainty, by regularly testing assumptions, having check-in points with decision makers, and making adjustments in content or direction in the light of evidence and knowledge. 27 This part of the Project remains in development. It requires more work to show policy design and delivery as an end-to-end process while acknowledging the messiness that characterises most policy processes in practice. 28

The Policy Skills Frameworka focus on policy advisors
The Policy Skills Framework describes the knowledge, applied skills, and behaviour expected of good policy advisors. Descriptions are broken down into levels from "developing" to "practicing" to "expert/leading", which loosely equate to analyst, senior analyst, principal analyst and above. Tools are available to enable individuals to articulate their policy skills profile 29 or credentials and for managers to map the skills make-up of their teams 30 to identify any gaps or overlaps. They can be used in recruitment, performance, and development processes or for organizational workforce planning. The components are underpinned by Public Service-wide foundation skills and knowledge related to things like an understanding of the machinery of government, key legislation, and the code of conduct (Figure 3).
The Policy Skills Framework was designed to address the lack of consistency in levels across the Public Service where the role (and salary) of a senior analyst or advisor in one department did not equate to those in another. (This was a symptom of New Zealand's highly devolved public management model and departmental autonomy). By spelling out the range of skills involved in policy work it provides the foundation for the development of future archetypes of policy practitioners (engagement specialists, deep data analysts, expert advisor/communicators) and future policy career paths.

It also articulates skills that appear commonly in job advertisements and role
Knowledge (what I know) Applied Skills ( Figure 3. The Policy Skills Framework. Source: The Policy Project. https://www.dpmc.govt.nz/ourprogrammes/policy-project/policy-improvement-frameworks/policy-skills descriptions but are rarely spelt out (such as "strategic thinking" or "political savvy"). The component skills were developed via a collaborative co-design process (see Box 2). This was to ensure that the list of skills reflected current practice requirements and signalled skills likely to be in growing demand in future (such as "engagement and collaboration"). The Policy Skills Framework provides a foundation for a future all-of-government policy workforce strategy.

Box 2. Policy Project practices
Co-designing the Policy Skills Framework 39 An initial co-design workshop brought together policy practitioners and managers with human resources and organizational development specialists to come up with a collective vision for what good policy people capability would look like. 40 Using design techniquespersonas, user-insights, building visual representational modelsthe group agreed that articulating the knowledge and skills required of policy practitioners would be the foundation piece of any policy people capability strategy. A number of subsequent workshops whittled down a long-list of skills (using techniques such as card-sorting) to arrive at the key components of the final Policy Skills Framework and their descriptions. 41

The Policy Capability Frameworka focus on policy organizations
The Policy Capability Framework is a diagnostic tool. It describes the key components of capability for a policy shop and offers "lines of inquiry" and potential indicators to review and build overall organizational capability to produce quality policy advice. The framework directs attention to organizational factors such as "stewardship" (investing in capability for the future), systems and processes for delivering quality advice, and being customer centric (with the customer defined as both citizens as users and ministers) (Figure 4).

POLICY QUALITY SYSTEMS
POLICY CAPABILITY ENGAGEMENT CUSTOMER-CENTRIC

PEOPLE CAPABILITY
How well do we focus on policy outcomes and build capability for the future?
How well do we enable and support the delivery of quality policy advice?
How well do we meet ons of ministers, customers, and other stakeholders?
How well do we ensure we have the right skills in the right place at me? Figure 4. The Policy Capability Framework. Source: The Policy Project. https://www.dpmc.govt.nz/ our-programmes/policy-project/policy-improvement-frameworks/policy-capability A range of tools support "light-touch" to "deep-dive" self-reviews of capability. Of these, the deepest-dive 31 offers "maturity ratings" that allow an organization or policy shop to assess the current state of performance, articulate desirable future performance, and then strategize how to get there. Guidance is provided on how to engage staff in the process, timing and communication, and options for including external peer reviewers. 32 Among other motivations, the Policy Capability Framework was devised in response to criticism that pre-existing periodic Performance Improvement Framework reviews of agencies were relatively light on policy performance and underpinning capability. 33

Evidence of impact on policy practice and quality
Theoretically, various approaches could be used to assess the Policy Project's impacts on policy practice and quality. A comprehensive evaluation approach would involve objectively measuring and comparing the quality of policy advice in agencies before and after uptake of frameworks promulgated by the Policy Project. This could be augmented with assessment of feedback from Ministers and other key stakeholders regarding their perceptions of the quality of policy advice and adjustments through time. Policy agencies are required to report policy quality scores in their annual reports, but until now there has been no consistent method of analysis. Some agencies had their own assessment regimes while others contracted their assessments to the New Zealand Institute of Economic Research. Some individual agencies have traced quality scores over time (Barker 2016;Tyson 2014). With a common measurement tool in the PQF, more systemic evaluation and provision of formative and summative feedback is possible in future.
Has the Policy Project made a difference? It is a challenge to measure changes in policy quality and capability across a whole system. It is even harder to measure and then plausibly attribute any change to the Policy Project. Qualitative feedback from chief executives and their deputies is being used to monitor "customer satisfaction" related to both the Policy Project's products and its collaborative operating model. 34 One chief executive has observed, "The enhancements we are making to the oversight of our advice are based on the continuous quality improvement ideas of the Policy Project". Some interim indicators of impact are based on the assumption that use and uptake of the frameworks will lead to improvements in quality and organizational capability.
At the time of this writing, the three Policy Project frameworks have been operational for less than a year. They will take some time to embed and show results. Following a strategy of making them "easy to use" and "hard to avoid", new tools and guidance have been developed in response to demand and to support agencies to "pick up and use" the tools in their day-to-day work. Further, the frameworks are now referenced in central agency accountability processes and guidance (e.g. for fouryear plans, annual reports, Performance Improvement Framework departmental reviews, Cabinet Office guidance for cabinet papers, the revised Cabinet Manual).
So far, around a third of policy agencies are using the Policy Quality Framework for assessing the quality of their policy advice including the three central agencies (the Department of Prime Minister and Cabinet, the State Services Commission, and the Treasury). Around a third of agencies are also using the Policy Skills Framework for various purposes from recruitment, to performance measurement processes, to assessing teams' skills gaps. Steady and growing website traffic accessing the frameworks and tools is another indicator of awareness and interest, with a monthly average of around 3500 hits (the policy workforce is 2500 people).
At least three agencies have already used the Policy Capability Framework for a comprehensive self-assessment of their policy capability and a restructure of their policy function. Another, a more operational department with a small policy function, has used the framework to guide a discussion with the chief executive and leadership team about the role of the policy function within that organization. In addition, the eight largest ministries have pledged voluntary funding for the programme for the next three years. This suggests senior leaders have confidence in the Policy Project.
Looking to the future, a set of measurable indicators of impact will need to be developed. Improvement in the overall quality of policy advice can be measured in changes to agencies' policy advice quality scores, which they are required to report in annual reports against appropriations. DPMC's Strategic Intentions (2017-2021) document 35 lists how it will measure progress, including improvements over time in: Agencies policy advice quality ratings Policy Project engagements with agencies to deploy the frameworks Reported use rates of the Policy Project frameworks and tools The ex-post assessment tool from the Policy Quality Framework, and expectations set by the Head of the Policy Profession related to scoring and quality targets 36 provide, for the first time, a common measurement methodology and suggested improvement trajectory. Recent analysis of data on the policy workforce, 37 current agency human resources practices and "pain points" could provide benchmarks for some related indicators such as a reduction in "unplanned staff turnover", managing the pipeline of new analysts, or number of collective workforce training initiatives potentially as part of a whole of government policy workforce strategy. Finding indicators that are measurable, meaningful, and easily utilized is a challenge that will need to be met.

Prospects for replicating the policy project
Many commentators around the globe agree that policy advising needs updating and improvement. Today's policy problems and political systems are complex. Policy challenges increasingly cut across traditional administrative and portfolio domains and new policy partners (including citizens) demand to be more involved in the decisions that affect them. Improvements in the quality of policy advising could be raised by interventions in the professional development of those in advisory roles, and in the infrastructure that supports them.
Effort anywhere to raise overall quality in a system requires ambition and determination. Writing about organizational change, Robert E. Quinn (1996) observed: "The land of excellence is safely guarded from unworthy intruders. At the gates stand two fearsome sentriesrisk and learning" (p.165). The pursuit of excellence requires divergence from the status quo. This carries risk and is often resisted. The pursuit of excellence also requires learning, including unpacking what actions worked and what did not. Emerging evidence suggests that the Policy Project has challenged members of the policy advising community in New Zealand to be more reflective about their practices and has supported them to drive up the quality of their work. From the point of view of pursuing excellence, the Policy Project is making a significant collective contribution, in particular by facilitating common views about what high quality policy advice looks like and how it can be developed. By doing so, the Policy Project has provided the tools and signposted the journey towards excellence in terms of policy skills, policy organizations and policy advice itself.
But will it stick? In 2014 the UK Institute for Government (IfG) produced an assessment of successful civil service reforms in the United Kingdom (Panchamia and Thomas 2014) and the ten factors that support reforms at different stages in their development and implementation. The factors are: 1. Clarity around the reform idea and purpose 2. Personalized leadership 3. The right degree of political support 4. Ambitious purpose connecting to departmental priorities 5. The support, or at least the permission, of the Treasury 6. A dedicated and diverse team to drive the vision and the model 7. Balance of compulsion with collaborative values 8. The right use of accountability and governance 9. Effective management of critical transitions 10. Creation of a lasting coalition of leaders around reform A recent IfG report on Professionalising Whitehall (McCrae and Gold 2017), added factors that can act as a brake on reforms, including leadership turnover, constraints on leadership, insufficient resources, and lack of stable funding.
Assessed against these two sets of factors, the Policy Project exhibits considerable potential to have a sustained impact on policy capacity, design, and delivery in New Zealand. The Project has a clear purpose. That purpose is consistent with broadly shared support for improving the quality of the public service. Policy Project leadership has been visible and collaborative, and considerable efforts have been made to establish a lasting coalition of supporters. This operating model will need to be maintained as the Project team and leaders change, as it has done recently. The three policy improvement frameworks provide a foundation infrastructure for improved advice and capability, which have the blessing of the three central agencies including the Treasury. The Project also complements and leverages current systems of accountability, connecting it to departmental capability priorities and likely to result in increasing levels of uptake of the Policy Project frameworks (despite the absence of formal compulsion to do so).
Future impact will depend on the Policy Project being a constant source of fresh material on policy methods and their application, on it actively supporting departments in their policy improvement journeys, and on it continuing to nurture and facilitate networks across the policy community. Senior support and active involvement in cross-agency governance, currently embodied in the Head of Policy Profession Board, will need to be maintained, including to give weight to the capability side of policy advice. Senior policy advisors tend to be more interested in substantive policy issues, sometimes paying less attention to their stewardship responsibilities and investment in the capability (people and organizational infrastructure) required to produce good advice.
The IfG report (2017) suggested that Heads of the Policy Profession should be fulltime. Both the UK and the New Zealand Heads of the Policy Profession have the role on top of permanent secretary and chief executive roles respectively. The New Zealand HoPP has scant time dedicated to that role. Leaning in to the "demand side", to involve ministers in the quest for improved policy, to nudge them to request and expect free, frank and evidence-informed advice, is another challenge.
To date, no other jurisdiction has tried such a systemic intervention as New Zealand's Policy Project. The United Kingdom has an active Head of the Policy Profession and has also pursued various efforts to raise the quality of policy advice. However, it has focused on the specialism itself (policy skills and professional standards) like other UK functional leaders (McCrae and Gold 2017). The two jursidictions have regular communications to learn from each other's approach.
The Policy Project offers a sound model of systemic intervention that policy leaders in other jurisdictions could readily emulate to achieve better results for their governments and their citizens. The policy improvement frameworks reviewed here are highly transferable and could be adopted and adapted in other contexts (local government, NGOs, the private sector, or anywhere else where advice is provided to decision-makers).
An international network to share strategies, frameworks and tools, as well as lessons about what works in terms of modernising policy advice, the policy profession and organizational policy capability would be highly desirable. Various mutually supportive initiatives could contribute to that goal. We mention just three. First, governments could do more to encourage cross-jurisdictional interactions and collaborations among policy leaders and teams; an international network of Heads of the Policy Profession could be developed. Second, schools of government could do more to forge international interactions and promote greater scholarly and practitioner dialogue around ways to improve policy advising and policy capability. Finally, international organizations could also contribute. Potentially useful approaches include creating regular forums for sharing innovative policy practices, and using various means to communicate lessons across jurisdictions.

Conclusions
Governments play a powerful role in shaping the wellbeing of their citizens through the vehicle of policy. Given this, efforts to raise the quality of the advice tendered to decision-makers can have significant multiplier effects. New Zealand's Policy Project has presented a collective, catalysing approach to doing exactly that. Evidence suggests the Policy Project has the potential to deliver longer-term benefits to New Zealand. Further, the Policy Project offers a sound collaborative and agile operating model for systemic intervention that policy leaders in other jurisdictions could emulate to achieve better results for their governments and their citizens.