An infrastructure for building policy capability – lessons from practice

Abstract The Covid-19 pandemic highlighted the importance of good systems for policy and decision-making. An effective policy system depends on robust policy capability. This article articulates key dimensions of policy capability based on the practical experience of policy practitioners from a range of jurisdictions. It briefly draws on the literature on policy making and organizational capability before situating the key components of policy capability as mutually reinforcing parts of a policy capability infrastructure. These include “supply side” components of leadership, policy quality systems, people capability, and effective internal and external engagement, as well as the “demand side” component of the political administrative interface that shapes and is shaped by policy capability in the public service. This framing of policy capability as an infrastructure broadens the definition of policy capability from a narrow focus on people and skills to a systemic approach that includes the range of systems and processes that enable and support good government decision-making. The article argues that the policy capability infrastructure could serve as a useful and generic analytical framework for describing, assessing, and improving policy capability in teams, organizations, or across an entire public service. Policy leaders are invited to test the framework and share their insights and results, including with colleagues in academia. If it works in practice, it might also work in theory.


Introduction
Governments around the world recognize that their policy making systems need a step change. 1 The imperative is well rehearsed and the challenges eerily similar. They include concerns about poor quality policy advice, or proposals not backed up by evidence; shortages of well-trained policy advisors; weak systems for prioritization, collaboration and alignment between organizations and portfolios (the silo problem); and a focus on short-term demands to the detriment of longer-term policy needs (like climate change, inter-generational poverty, aging populations). Despite common capability challenges, organizations and jurisdictions have taken different approaches to addressing them.
The UK has a long-standing Policy Profession Unit with recently refreshed standards for policy professionals and a policy curriculum (Policy Profession 2021), while Australia has a "Delivering Great Policy" programme in the Department of the Prime Minister and Cabinet with a policy hub (Australian Government 2022) and policy training (now falling under the auspices of the Australian Public Service Academy). Canada has a policy community of practice supported by a Policy Community Partnership Office. The Irish government is currently working with the OECD on a project to enhance its policy making system. At a whole-of-government level Aotearoa New Zealand has taken arguably the most comprehensive approach in the Policy Project in the Department of the Prime Minister and Cabinet (Washington and Mintrom 2018). Within jurisdictions multiple organizations are also working to improve their policy capability and advice to decision makers. The Australia New Zealand School of Government (ANZSOG) is working with a range of departments across its national and state jurisdictions on their various quests to improve their policy capability. There is demonstrable action and evidence from across Western democracies that policy capability is a concern and requires attention. The Covid-19 pandemic put systemic weaknesses into stark relief.
Analysis of action on the policy capability front suggest that most organizations and jurisdictions adopt a piecemeal approach, focusing on a piece of the policy problem rather than taking a system view of how the bits of the policy puzzle fit together. Most often the focus is on people capability reflecting an assumption that better trained policy staff will lead to better policy advice and outcomes. Even the initial diagnosis in Aotearoa New Zealand in 2014 revealed the main identified "pain point" for senior leaders was not being able to find skilled policy staff. But the problem is much more complex than having smart well-trained people in policy roles. It requires a system shift and action on multiple fronts.
So how should public organizations and jurisdictions go about improving the quality of their advice to decision makers and the capability that underpins that work? What can we learn from the literature on policy making and on broader organizational capability? And what can we learn from practitioner experience? This article briefly scans the literature on both policy and organizational capability to examine important pieces of the puzzle. It then draws on lessons from across organizations and jurisdictions to highlight the focus areas considered by practitioners to be most important for building policy capability. Following a more detailed outline of those key components the article situates them as parts of a mutually reinforcing Policy Capability Infrastructure. It argues that the concept of a policy capability infrastructure adds to the literature on both policy development and the capability of public sector organizations and systems. As such, it could serve as a useful and generic analytical framework for describing, assessing, and improving policy capability.
For this article, policy capability is used in its broadest sense; including capacity issues (resources and bandwidth to get the job done) but focusing on the skills, tools, processes and institutional enablers to support the design and delivery of policy advice.
1.1. What lessons for policy capability from the literature on policy and on organizational capability?
The academic literature offers little in the way of guidance for improving an overall policy system. There is a body of literature focused on policy analysis and methods (see Mintrom 2012, the policy process (Althaus, Bridgman, and Davis 1998;Colebatch 2009;Bardach 2012), and on policy design (Howlett 2014;Howlett et al. 2015) as well as on the critical success factors for good policy or "policy successes" (Luetjens, Mintrom, and 't Hart 2019). Other literature unpicks the range of skills required by policy analysts (Weimer and Vining 2017) with some digging deep into certain aspects of the policy skills repertoire, such as the inter-personal skills and political nous required to give effective policy advice (Rhodes 2016;Hartley et al. 2019). Recently scholars have taken to describing new types of policy practitioners such as the "policy entrepreneur" (Cairney 2018; or analyzed new policy methods (Blomkamp 2022) and the institutions set up to apply them, like policy labs (McGann, Wells, and Blomkamp 2021;Wellstead, Gofen, and Carter 2021). However, most of this literature focuses on how to do policy rather than how to build the capability to produce it.
There have been a few attempts to take a more systemic perspective on policy capability. Alford and O'Flynn (2021;O'Flynn 2019) conceptualized capability as incorporating three different levels: competencies of individual public servants; structures and processes in public sector organizations, and the wider enabling public sector environment (such as budgets, rules and risk appetites). Wu, Ramesh, and Howlett defined a conceptual framework for policy capability that combines multiple elements, including at the organizational and system level. Building on the policy and public management literature, including Mark Moore's strategic triangle (Moore 1995) the authors identify three sets of skills and competencies (analytical, operational and political) and three levels of resources or capabilities at three different levels-individual, organizational and systemic to come up with a "nested model of policy capacity" which they argue provide "policy-makers and commentators with a better idea of how capacities can be built" (Wu, Ramesh, and Howlett 2015). These are important steps toward thinking of policy capability as a system made up of multiple components.
The literature on overall organizational capability adds some practical pointers for organizations and governments wishing to improve their policy making systems. Thomas (2020) proposes a simple and insightful definition of organizational capability as "a product (or not) of the interaction between the skills, experience and methods of an individual-and the culture, structures, processes of the organisation they work in." ANZSOG (Speagle et al. 2021) has recently reviewed various models for reviewing capability in government organizations, drawing on experience from New Zealand, the United Kingdom and Canada to come up with a generic agency capability review framework. However, while this model includes policy relevant items and lines of inquiry such as "clarity of purpose vision and strategy," "engagement with ministers," "citizen focus," and "collaboration with other agencies," they appear under different rubrics and may not go deep enough or be joined-up enough for organizations whose core business is policy. Indeed, analysis of the NZ Performance Improvement Framework (the capability review regime that formed part of the ANZSOG research) concluded that it was light on policy. 2 Public sector leaders and practitioners on the front line of the policy capability challenge are seeking more detailed and practical roadmaps for change. What lessons can be drawn from their efforts?

What lessons for policy capability from practice?
The following sections draw on recent work by the author across multiple jurisdictions to help organizations and the wider public service build their policy quality and capability. It also draws on the NZ Policy Capability Framework (PCF) of which the author was the chief architect (see box). Taking an inductive approach, the key components of capability have been derived from practitioner experience, structured conversations, and critical reflection on how jurisdictions have approached the policy capability challenge. It reflects what practitioners consider to be the key aspects of policy capability. Those components are articulated below and positioned as mutually reinforcing parts of a "policy infrastructure." On the supply side (organizational capability) it includes four mutually supporting infrastructure components-leadership, policy quality systems, people capability and engagement-which are then broken down into related sub-components. A crucial element is added. While most efforts to improve capability focus on the "supply side" of the policy equation, the policy infrastructure adds the "demand side" into the mix-the vital political administrative interface which shapes and is shaped by policy capability in the public service.

The New Zealand Policy Capability Framework (PCF)
The PCF was co-designed with government policy leaders and other organizational and policy experts. It responded to policy leaders' demand for a framework that could be used for a "deep-dive review of policy capability" and for policy leaders to "facilitate constructive and courageous conversations with staff on the current state of the policy unit's capability, strengths and weaknesses, improvement strategies, priorities, and sequencing … to support an aspirational improvement process" (Policy Project 2016).
The co-design process, including several forums involving multiple agencies, asked participants to articulate what components of policy capability they considered most important. The process also drew on earlier research on what were deemed "high-performing policy units" to uncover critical success factors or what made them high-performing (State Services Commission 1999). After prototyping and testing, the resulting PCF was endorsed by the Tier 2 Policy Leaders group (Deputy Secretaries with policy responsibilities) and launched by the NZ Prime Minister in 2016 (New Zealand Government 2016).
The PCF has different versions, one with optional maturity ratings. Related user guidance includes processes for self-review, critical friend peer-review or external review, and approaches to rating performance and reporting results. The PCF has been used by departments in Aotearoa in a range of different ways. For example, one department uses it for regular 6-monthly health checks of its policy capability (Policy Project 2020).
The concept of a policy infrastructure has been tested with a range of senior government officials and practitioners (including those attending various ANZSOG forums) and found to be a practical and useful tool to guide assessments of, and efforts to improve, policy quality and capability.

The policy capability infrastructure
On the supply side (Figure 1), the key components of policy capability include: Leadership-to ensure the organization (or team or whole-of-government system) is focused on policy priorities while at the same time investing in the ongoing health of the policy function and the capacity to provide future advice. Policy quality systems-the frameworks, tools, evidence base, analytical methods, and processes to ensure the provision of quality advice to decision makers. People capability-making sure policy professionals have the right skills and are supported in their work and development: right people, right place, right time. Engagement-ensuring that a diverse range of people and perspectives are involved in the development of policy advice, that ministers are well-served as the "customers of policy advice" and that decisions meet the needs of the people it was designed for.
The demand side component is about supporting ministers to be "intelligent customers" of advice and ensuring the political administrative relationship is functioning well. This speaks to the authorizing environment and what Moore's strategic triangle refers to as "legitimacy and support" (Moore 1995). 3 Targeted "lines of inquiry," indicators, and maturity ratings can be attached to each of these components and sub-components to diagnose capability and subsequently design actions toward improvement. Some key insights and focus areas are described below.

Leadership to focus on outcomes
Leaders of any government organization set the scene and create the environment for good policy ( Figure 2). Policy leaders need to be able to articulate a clear vision and direction of travel with a focus on outcomes, including where the organization can contribute to meeting overall government goals (including cross-cutting goals). They need to be able to translate this into strategy and priorities and communicate those down the organization so that all staff know what they are there to achieve. They need to build an enabling innovation culture, by curating openness to new ideas, encouraging constructive challenge and debate and ensuring that policy is informed by insights on the latest research, evidence, and trends. Leaders also need to mitigate the tendency toward short-termism by encouraging a focus on medium to long-term policy challenges that will bite in future if not dealt with today. That means negotiating space in budgets and work programmes to focus on "policy stewardship." In Aotearoa New Zealand, the Public Service Act 2020 requires chief executives to produce regular longterm insights briefings (Speagle et al. 2021). This is partly an attempt to encourage capability in foresight. A lack of foresight capability has been identified as a shortcoming common to many jurisdictions to the detriment of their overall policy performance (OECD 2018). Countries like Singapore with a mature and dedicated foresight capability are the exception to that rule; Singapore's response to the covid pandemic and insights on how it would shape future transformation drew on this superior capability (Centre for Strategic Futures 2020).

2.2.
Policy quality systems-systems, tools, and processes to ensure the delivery of quality policy advice Underpinning good policy design and good policy advice are frameworks, tools, and processes to guide policy practitioners in their day-to-day work (Figure 3). Practitioners tend to develop their own bespoke methods or ways of working over their careers, which is why policy is often described as an "art" or "craft." But increasingly organizations and jurisdictions are calling for scalable and repeatable models for how to do policy, to ensure consistency of approach (without stifling innovation) and greater consistency in the quality of policy outputs. This means designing generic processes and practical policy tools-including for quality assurance-to support policy practitioners and to avoid continually reinventing the wheel.
Poor commissioning is a common theme across jurisdictions, as is the lack of good planning and project management. Project management methodologies developed for IT projects are not well suited to iterative policy projects. The New Zealand Policy Project Start Right tool (Policy Project 2017), which has been adapted by other jurisdictions (Get-to-Go in the South Australian Department for Education Strategic policy model) was designed as a guide to set a policy project up for success. It is a policy initiation tool that includes "commissioning conversations" and a set of initial explorations to assess the dimensions of a challenge-including who needs to be involved, what analytical lenses to apply, the likely governance and management required-to understand the scale and scope of the change required. It stresses that all policy is not created equally. Large-scale, complex, transformational change requires more wrap around effort and resources than small transactional changes to the status quo (Figure 4).
Start Right forms part of a wider policy methods toolbox which includes guidance on other methods like design-thinking, behavioral insights, and foresight methods. Policy toolkits also feature in the Australia (Policy Hub) and the UK (Open Policy making toolkit) (UK Government 2022) as a system-wide support for policy professionals, with a notable level of cross-referencing between them. Guidance on Cabinet requirements (Cabinet manuals) and more traditional analytical tools like regulatory impact analysis, and cost-benefit analysis are also included. Having the tools and guidance in one place (within an organization or a jurisdiction) makes them visible and accessible to policy practitioners, more likely to be used, and more likely to drive consistent performance.
Setting out what good performance looks like is also important. New Zealand has articulated detailed quality standards for policy advice in the Policy Quality Framework. The PQF is supported by quality assurance tools like peer review templates and a quality scoring regime. Government departments with policy advice appropriations must now report annually on their policy advice quality scores. Other, albeit less comprehensive, attempts to set out what good policy advice should encompass include Australia's Delivering Great Policy model and Policy Tests originating in the UK Department for Education (UK Government 2013). Quality standards set a benchmark for what policy practitioners should be aiming for in their advice.
Quality policy advice also depends on quality evidence, whether from the literature, international experience, results of evaluation of past efforts, administrative and other data (integrated data infrastructures and data sharing regimes support this) or from insights from frontline staff and user-experience. Knowledge management to ensure evidence is curated and accessible including for future users (as institutional knowledge) is important. For jurisdictions with indigenous populations the notion of evidence is being reframed to also include indigenous knowledge and lived experience.
But gathering and generating data and information only gets you halfway. Policy practitioners need to be able to make sense of evidence from the variety of sources to inform their advice and need to be able to apply different analytical lenses to evidence to uncover the potentially differential impacts of policy options on different population groups (women, ethnic groups, people with disabilities, LGBTQI, different age cohorts etc.). Policy needs to be both inclusive and targeted. Providing collateral (frameworks, tools, guidance) to support policy staff in their work is vital.
And what of those staff and their skills and competencies? Again, it is important that these are well articulated so that policy practitioners know what is expected of them and how they can develop their policy muscles.

People capability-right people, right skills, right place
How do policy leaders ensure they have the right people with the right skills in the right place at the right time? As policy design gets more complex, policy practitioners require proficiency with a range of newer methods (user-centred design, behavioral insights, data analytics) as well as being skilled in more traditional analysis (cost-benefit, regulatory impact analysis) and advisory tasks (briefing notes, Cabinet submissions). This goes beyond generic public service or departmental competency frameworks to see policy as a profession with a defined skill set ( Figure 5).
Both Aotearoa New Zealand and the UK Policy Profession have articulated the skills required of policy practitioners. The UK has recently upgraded its Policy Standards which are supported by a curriculum and comprehensive certified training options (UK Government 2021). The OECD has also attempted to define policy skills, albeit as part of a wider package of skills for a high performing civil service. It highlighted NZ Policy Skills Framework (PSF) in its work (OECD 2017). What is unique about the NZ framework is that it was designed for and by the NZ policy community and includes knowledge ("what I need to know"), applied skills ("What I need to be able to do") and behaviors ("how I need to act") and provides detailed descriptions about what those skills mean at different levels of expertise. While skills like "strategic thinking" and "political savvy" have always been seen as critical for policy analysis and advising they are rarely described in a practical "what does this mean for me" way.
The New Zealand PSF has practical application through supporting tools for individuals to map their skills and for managers to map the skills makeup of their teams. A development pathways tool offers advice on how individuals can build their policy muscle in each skill area. Setting out required skills offers a foundation for performance and development conversations, for training ("where do we need to upskill individuals or the team?"), recruitment to specification (where a manager can identify and fill gaps in the team or organization), the future articulation of policy archetypes (deep data expert, ministerial advisor, engagement specialist etc) and for workforce strategies ("what skills are we going to need in future and how do we develop or recruit them?"). Being deliberate about the balance of buy versus grow in terms of building people capability is crucial, although leaders in many jurisdictions are being encouraged to think of people resources as system, not single agency assets. Jurisdictions like Singapore have built people capability based on a "one united public service" platform (Prime Minister's Office, Singapore 2022). Others like Aotearoa have a more siloed approach to people development, albeit changing under the "unified public service" mantra of the Public Service Act, 2020 (Public Service Commission 2020). A "one system" mindset can help mitigate concerns about staff churn or fear of losing staff to other departments. Well trained staff moving around the system could catalyze knowledge sharing and overall skills development. It also enables deployment of capability to challenges that are cross-cutting. New Zealand's response to covid was underpinned by an "all-ofgovernment" team, seconded from across the public service (Cameron 2020).
In managing people capability policy leaders and managers need to think about how to enable staff. This means clarity about where decision rights sit, avoiding overly hierarchical authorization regimes (a handbrake on efficiency and a staff demotivator) and offering staff opportunities to stretch and grow their skills and experience. This links to notions of work allocation and maintaining balance across a team or organization, ensuring the policy "stars" avoid burnout and less experienced staff get a crack at working on challenging, rewarding and skillsenhancing policy projects. It also means more senior staff taking responsibility for guiding, coaching, and mentoring less experienced team members. Staff should also be encouraged to network with colleagues in other departments, especially those with a substantive policy or sectoral interface.

Engagement-investing in relationships and joining up the policy ecosystem
Policy is no longer a desk-based activity. Designing and delivering good policy depends on relationships and being engaged with the wider policy ecosystem. The lack of community engagement or understanding the needs of citizens and the end-users of public policies and programmes was recently identified by the former UK Head of the Policy Profession as the key shortfall in the UK policy-making system (Slater 2022). Engagement with end-users and stakeholders ranges from generating insights about and "walking in the shoes" of those who will be affected by policy (empathy tools), to actual co-production of policies and programmes with those groups. Many policy organizations and jurisdictions have attempted engagement guidelines, often drawing on the International Association for Public Participation's spectrum of public participation (IAP2 2018). The key thing is that policy practitioners know when to engage, who with, for what purpose, and how. Indigenous populations are a specific focus for engagement in Australia, and especially in Aotearoa New Zealand where Te Arawhiti, the Ministry for Crown M aori Relations has issued a framework for engagement with M aori as part of new public service responsibilities (enshrined in the Public Service Act 2020) to honor the Treaty of Waitangi (Te Arawhiti 2022). New Zealand's overall effective response to the covid pandemic was initially jeopardized by a failure to understand, engage, and work with indigenous M aori as well as Pacifica communities and health providers (Megget 2022).
Policy practitioners need to build relationships with other parts of the ecosystem including the academic and research community, relevant sector groups, and think tanks, all of whom are sources of insights, evidence, and feedback, and can serve as critical friends. Links to others in the policy value chain, including in other government agencies as mentioned above, and other parts of the same department (legal, finance, communications, operations) are also vital. Bridging the policy-operations or delivery divide is the holy grail of policy yet policy teams often still perceive their work to be done once a decision is taken and the responsibility is thrown over the fence to operational colleagues to implement. Policy staff need to involve delivery colleagues up front in the policy process and keep involved once the baton is handed over for implementation-like a relay, they should remain keenly interested in the final result.
Cultivating relationships is not a set and forget task or only activated when some specific information or support is required. It requires a shift from transactional to relational engagement with internal and external stakeholders. Investing in relationship capital means building an asset that can be drawn on when needed. Policy teams need to be deliberate about building ongoing relationships across the policy ecosystem.

The forgotten "demand side"-the role of ministers in policy advice and capability
With the concentration on the supply side of policy performance, the role of ministers-or the "demand side"-in the good policy equation sometimes gets forgotten. Too often policy practitioners see being responsive to ministers as simply implementing ministers pre-baked solutions. That is not surprising when many politicians treat their officials as their delivery wing, not a source of evidence-informed advice on how to solve policy problems or to identify policy opportunities. Engaging with decision makers requires policy practitioners to be politically astute and to develop good working relationships across the political administrative divide. While this is often seen as something organic or dependent on alchemy between politicians and officials, the dimensions of these relationships and how to curate them can also be codified (Washington 2021a(Washington , 2021b. The foundations of those relationship include ( Figure 6): Figure 6 The Demand Side -the role of ministers in policy advice and capability.
Agreeing a policy programme-reflecting government priorities and departments' capabilities to deliver them, including leaving space in budgets for longer-term policy stewardship. Setting ground rules for commissioning advice and for adjusting the policy programme while ensuring policy intent is understood and not lost in communication down the chain to people doing the policy analysis and design. Determining an operating model for engaging with policy advisors-mutual understanding about who is in the room for discussions, encouraging officials to be courageous and innovative, and free and frank in their advice ("tell me what I need to hear, not what you think I want to hear"). Developing processes to ensure advice tendered is of high quality-supporting ministers to be intelligent customers of advice including by agreeing policy advice quality standards and building feedback mechanisms to drive continuous improvement.
While each of the components of policy capability, on the supply and demand side, are important, it is how they fit together that counts. They need to be seen as a mutually reinforcing infrastructure, with common themes referenced across the components, like a commitment to policy advice that is evidence informed, free and frank, user-centric, and impactful. But where to start? How can policy leaders and managers assess their team or organizational capability and where should they start on the improvement journey? (Figure 7).

2.
6. An improvement trajectory-determine where you are at and where you are going Participants in a recent ANZSOG forum (ANZSOG 2022) shared their insights about how to catalyze and maintain an improvement programme. There was general agreement that connecting-up the pieces of the policy improvement puzzle is important for sustainable reform. Improvement initiatives that are mutually reinforcing-"all the elements aligned and moving in the same direction"-add up to more significant and longer-lasting change. They also agreed that "top-down" approaches are less successful than collaborative change processes which catalyze "coalitions of the willing." But endorsement and support from leaders is vital. Leaders "needs to be present and connected at all levels" to sponsor, champion, and enable purposeful change, by giving staff a clear sense of purpose and direction (ANZSOG 2022).
The first step toward an improvement trajectory-in policy or any other public administration domain-is knowing where you currently stand. Participants at the above discussion agreed that being clear about the current state is an important starting point in the improvement journey, noting that it is helpful to "hold a mirror up to ourselves" to understand the challenge, and to catalyze and guide action toward improvement (ANZSOG 2022).
The policy infrastructure offers a framework for assessing policy capability when supported by related indicators, maturity ratings and improvement targets. Like the NZ PCF it could be applied or easily adapted to other jurisdictional contexts. It can be applied to various units of analysis-team, organization, or wider whole-of-government system. Rather than contradicting the literature on policy making, it provides an organizing framework within which the themes in the literature-policy methods, skills for policy, policy design, analytical frameworks-can sit. It adds to the literature on organizational capability in the public sector by providing a deep-dive into key aspects of capability for organizations with policy as their primary remit.

Conclusions
This article has articulated the key dimensions of policy capability. It has argued that while all the parts described above are important, it is how they work together that matters. They need to be seen as parts of a mutually reinforcing policy infrastructure. The policy capability infrastructure is based on the views and practical experience of policy practitioners from a range of jurisdictions. It is hoped that the concept adds to the literature on policy development and the capability of public sector organizations and whole-of-government systems. Because the concept has resonated with senior officials and policy practitioners, the policy capability infrastructure could serve as a useful and generic analytical framework for describing, assessing, and improving policy capability. Policy leaders in other jurisdictions are encouraged to test it and share their insights and results, including with colleagues in academia. If it works in practice, it might also work in theory.
Working with senior officials across Australia and Aotearoa New Zealand, ANZSOG is building the notion of a policy infrastructure into its programmes and narratives. This framing of policy capability as an infrastructure broadens the definition of policy capability from a narrow focus on people and skills to a systemic approach that includes the range of systems and processes that enable and support good government decision-making. And good government decision making, that meets the needs of the public we serve, is what public policy and public administration is all about.