31 Making public policy

It is commonly believed that Australians are uninterested in politics.^[1] Whatever the truth of this proposition, voters are generally interested in government policies that they believe will affect them, although the manner in which policy is made remains opaque for many.

We argue that public indifference to how policy is made is problematic. Public policy affects almost every aspect of our lives: policy affects our birth, the manner in which we are raised and educated, our access to health care, the quality of our physical environment, how we conduct ourselves, whom we might marry, our access to employment, our rights at work, our access to housing, how we raise our children and even the quality of our deaths and what we are able to pass on to the generations succeeding us. The policy process itself can also have repercussions for society and communities, particularly when community opinion about policy options is divided.

Policy making is, in part, an exercise in rational problem solving. It is also an intensely political process and requires the judicious consideration and balancing of complex issues, including public opinion, competing interests, social relations and the distribution of power within a society.

For the most part, policies begin as statements of values and intent. Policies often have an ideological foundation and are frequently portrayed as occupying part of a spectrum ranging from left to right. For example, governments or parties of the right or centre-right might be characterised as favouring market forces over government intervention, individual rights over collective rights and unilateralism over multilateralism. Governments of the left are typically portrayed as favouring government intervention in social and economic affairs, emphasising collective rights and preferring multilateral approaches to problem solving.

Such characterisations are, of course, simplistic. Governments of the right sometimes resort to intrusive uses of state power despite the value placed on individual sovereignty, and governments of the left sometimes resort to market mechanisms to address distributional inefficiencies.

What is ‘policy’?

This chapter is primarily concerned with formal expressions of government – or public – policy:

• as a set of values and convictions
• as operational rules designed to comply with legal requirements
• as embodied in law in the form of primary legislation or regulation.

In each case, the formal expression of policy gives form and consequence to policy intent.

In broad terms, policy can be said to represent preferred responses to problems. For any given problem there might be a number of available responses. For example, the statement ‘anyone who attempts to travel illegally by boat to Australia will be turned back to their country of departure’ is a declaration of policy. It sets out a preferred response under defined circumstances. To the extent that such a statement sets out a preferred response, it also precludes other potential responses.

Policy provides a framework for what can and ought to occur in prescribed situations. However, policy is also malleable and is subject to interpretation and adjustment as circumstances change. Changing expectations, attitudes, beliefs, values and behaviours often lead, eventually, to changes in government policy – one example being amendments made in 2017 to the Marriage Act 1961 to recognise same-sex marriage.

Similarly, technological change and environmental changes – for example, climate change – have fuelled demands for adaptive policy responses. Likewise, structural changes in the economy have driven adaptive changes in policies pertaining to industry, consumer law, employment, education and finance (among others) while global economic and political instability leads government to focus on policy responses aimed at improving productivity, enhancing economic resilience, and strengthening budget sustainability.^[2]

How does policy happen?

Public policy can be a messy business. The 19th-century American poet, John Godfrey Saxe, is reported to have written ‘Laws, like sausages, cease to inspire respect in proportion as we know how they are made’.^[3]

Public policy is an expression of political intent and a framework for action. Political parties or groupings, in and out of government, will have a set of policies – a policy platform – covering a broad and diverse range of matters. Ideally, policy platforms are internally consistent and represent a coherent narrative for governance. This is not always the case, and the highly contested nature of public policy sometimes means that governing parties bring contradictory positions to the business of government.

For a problem to be considered deserving of a policy response – having what the influential political scientist John Kingdon^[4] refers to as ‘policy salience’ – there first needs to be:

• broad agreement that a problem exists
• a broadly shared understanding about the nature of the problem
• a broad acceptance of available solutions.

Moreover, propositions about the existence and nature of problems, not to mention the nature of possible solutions, need to be tested in a variety of forums: for example, within the broader community and the electorate; within communities of interest, including geographical regions, industry sectors and civil society; within professional ‘epistemic’ communities of subject area specialists; and within political parties themselves.

The existence and importance of ‘problems’ is often highly contested, both in the community at large and within political parties – consider the unsuccessful 2023 referendum to establish a voice to Parliament for Australia’s First Peoples.^[5] Even where there is broad agreement about problems, ‘solutions’ are often controversial. There are many reasons why it is difficult to reach a majority view about the nature of policy problems and preferred solutions. Different actors and stakeholders bring different things to the table and their perspectives are shaped by their lived experience, education, qualifications, attachment to particular interests, attachment to community, ideology, religious beliefs and personal convictions.

Policy makers also need to be attuned to perceptions of ‘winners and losers’. In other words, who benefits from the policy and who perceives themselves to be adversely affected by the policy? They also need to be aware of the potential for ‘interests’ (e.g. civil society organisations, industry groupings, interest groups, communities) to mobilise for or against policy proposals. For example, opposition to the establishment of offshore windfarms became a political flashpoint in the 2024 federal election.^[6] Taking each of these factors into account, it is easy to see why it can be so difficult to reach agreement about problems and solutions.

Theoretical perspectives

In his book Analysing public policy, Peter John^[7] outlines the principal approaches for understanding the policy-making process:

• Institutional approaches, which take the view that that political organisations – such as parliaments, legal systems and bureaucracies – shape public decisions and policy outcomes.^[8]
• Groups and network approaches, which claim that associations and informal relationships, both within and outside political institutions, shape decisions and outcomes. These approaches not only consider the effects on policy of unique relationships between groups and entities, they also embrace the idea that networks of relationships affect policy outputs and outcomes.^[9]
• Exogenous approaches, which assert that factors external to the political system determine the decisions of public actors and affect policy outputs and outcomes.^[10]
• Rational actor approaches, which claim that the preferences and bargaining of actors explain decisions and outcomes. This approach is often called ‘rational choice’.^[11]
• Ideas-based approaches, which hinge on a view that ideas about solutions to policy problems have a life of their own, and that ideas circulate and gain influence independently in the policy process.^[12]

Theoretical perspectives such as these are useful in helping us to understand the policy-making process as a social, cultural, historical and political phenomenon. Each allows us to consider some facet of policy making and to understand the nature of the environment in which policy occurs.

While each of these approaches serves a particular intellectual purpose and reflects particular ‘truths’ about how policy comes to be, none tells the whole story. Nor are they necessarily mutually exclusive (e.g. many institutional accounts also rely heavily on rational actor thinking).

The reality is that making public policy is a complex social behaviour and any given policy exhibits the influences of multiple institutions, groups and networks, exogenous factors, preferences and ideas generated within epistemic communities.

A marketplace of ideas

Public policy might best be described as a marketplace of ideas and prescriptions for the broad and diverse array of matters that need to be actively governed in order for human society to function. It involves making difficult choices and negotiating multiple trade-offs between competing options. Moreover, policy exists in a highly contestable marketplace of ideas, especially in liberal democratic societies like Australia’s.

Policy practitioners need to be mindful of the ideological leanings and philosophical underpinnings of all stakeholders – both within parliament and without – in order to anticipate possible resistance to policy proposals and advise government about policy compromises that might be broadly acceptable to legislatures.

Although Australia’s polity is often portrayed as a ‘two-party system’, our parliaments are generally made up of representatives from multiple parties as well as independents who have no formal party affiliation. And although electoral contests in all Australian jurisdictions usually involve competition between two major parties – in most cases, the Australian Labor Party and the Liberal Party of Australia (an exception being Queensland, where the Liberal and National parties merged in 2008) – Australian parliaments are usually dominated by three, and sometimes four, established political parties.

Even non-governing parties and members of parliament – including minor parties, ‘micro parties’ and independents – can exert influence on policy, especially when governments do not enjoy a numerical majority in both the upper and lower parliamentary chambers (the exceptions being Queensland, the Australian Capital Territory and the Northern Territory, which are ‘unicameral’, meaning they have only one legislative chamber). The emergence in 2022 of so-called ‘Teal Independents’ signalled a potentially significant disruption of traditional party politics.^[13] It has been argued that Teal successes in metropolitan electorates contributed to Labor achieving a landslide win in the 2025 federal election.^[14]

It must be admitted, however, that when in power governments do not always adhere faithfully to the ideological positions they espoused when in opposition. Governments are usually obliged to take a pragmatic view and work within constraints imposed by the political, social and economic environment in which they are situated.

A contest of interests

Public policy is also the concern of particular interests in society, and it can be said that some policy settings can become captive to those interests.

Policy is often vigorously contested within political parties, a prime example being the internal debate within the federal Liberal Party around the question of climate change and strategies to reduce carbon emissions; within the federal Labor Party one finds sometimes rancorous debate about the treatment of asylum seekers.

Policy proposals from government might also be challenged by a variety of interests, including industry sectors (e.g. the Minerals Council of Australia), professional groupings (e.g. the Australian Medical Association), trade unions (e.g. the Australian Council of Trade Unions or ACTU), civil society organisations (e.g. the Australian Council of Social Service) or consumer lobbies (e.g. CHOICE). These interests represent stakeholders that stand to be affected in some way by government policy. In general, policy makers seek to consult with affected interest groups (usually through their representative organisations) in the formulation and implementation of policy. Politically powerful interests can wield significant – and sometimes disproportionate – influence. Australian examples include the first Rudd Labor government’s attempt to introduce a Minerals Resource Rent Tax and the Gillard Labor government’s national gambling reforms – these measures were staunchly resisted by industry interests and subsequently wound back.

An evidentiary basis?

In an ideal world, policy responses would have some kind of evidential basis. This might mean a combination of empirical research, statistical analysis, comparative policy studies, public consultation, evaluation studies or other forms of evidence that can be made available for independent scrutiny.^[15] However, ours is not an ideal world, and the evidence base for many public policy choices is often selective, sometimes even to the point where decision makers find themselves accused of ‘policy-based evidence making’ – a pejorative converse of the term evidence-based policy making.

‘Policy-based evidence making’ means working backwards from a predefined policy position with the aim of finding evidence that supports decisions that have already been made.^[16] It is possible that the growing trend of governments engaging private consultancy firms to produce commissioned research as an input into policy development has contributed to the perception that evidence is often crafted to fit policy preferences (although the Albanese Labor Government has taken a strong stand on reducing consultancy use and expenditure).^[17] It is also not unknown for special interests or lobby groups to produce commissioned research (of varying quality) in support of their advocacy for policy change.

Policy making is subject to bounded rationality – meaning that the decisions of policy makers are constrained by a variety of factors, such as the tractability of the problem at hand, the availability of information and the time frame within which decisions must be made. There will be times when the ‘evidence’ either fails to support, or directly contradicts, the preferred policy positions of governments, and it is not unknown for contradictory evidence to be suppressed in order to ‘protect’ policy settings that are based more in ideology or moral conviction than in any objective appraisal of the circumstances.

Finally, if evidence is to have an impact on policy governance and management, systems that are capable of incorporating new information into decision making are required. This is a perennial problem for public sector organisations, which often fail to use evaluative data generated in the course of delivering public policy to make adjustments to policy settings and/or to the service delivery architecture.^[18]

Practical policy formulation

It is government’s role to set policy objectives, and it is the duty of the public service to advise government about the technical, political and economic feasibility of those policy objectives, including any risks that might arise in their implementation. Having ‘advised’ government, the public service is obliged to give effect to government policy by developing an implementation strategy (in consultation with the government), including formulating a budget, identifying relevant internal and external capability and undertaking appropriate consultations with affected stakeholders.

It is also the responsibility of the public service to manage any risks arising in the implementation and operational phases. Policy implementation can be subject to a wide variety of operational constraints, such as short time frames, availability of resources and technical practicability. In the political realm, policy can be undone by:

• a lack of appropriate legal authority – a prime example is the automated debt assessment and recovery scheme – ‘Robodebt’ – which operated from 2016–2020 and was subsequently found to have been ‘unlawful’;^[19]
• an inability to pass enabling legislation in parliaments –  an examples is the Turnbull Liberal–National Coalition government’s withdrawal of proposed corporate tax cuts legislation in 2018^[20],and
• community/ stakeholder resistance – an example being plans by the Morrison Liberal–National Coalition to implement a proposed NDIS Independent Assessment Framework in 2021.^[21]

Typically, the public service often bears the brunt of any fallout associated with ineffectual or misguided policy formulations.

In Australia, public servants usually acquire their policy skills ‘on the job’ in the form of ‘craft knowledge’.^[22] Indeed, it is unusual for Australian public servants – unlike their North American counterparts – to enter the public service with formal training in public administration, public policy or political science. Although increasing numbers of public servants now undertake postgraduate qualifications in disciplines related to public policy, there remains a degree of scepticism among public servants about the relevance of academic learning to the ‘craft’ of public policy making.^[23]

Policy practitioners who seek to learn more about the policy process will discover an extensive theoretical literature, aimed primarily at academics, that is not easily translatable to the real-world situations confronting them.^[24] This literature is also characterised by vigorous – and often acrimonious – debate about the limitations of certain models.

Until the late 1950s, policy making was predominantly portrayed as a process of rational analysis culminating in a value-maximising decision. However, American political scientist Charles Lindblom (1917–2018) regarded the rational policy process as an unattainable ideal and proposed an alternative model, incrementalism, which focused less on abstract policy ideals and placed greater emphasis on solving concrete problems.^[25] Often described as ‘muddling through’, incrementalism describes an iterative process of building on past policies and reaching broadly agreed positions among diverse stakeholders.^[26] Lindblom’s emphasis on ‘trial and error’ would resonate with many contemporary public servants.^[27] In policy spaces characterised by complexity, incremementalism is virtually unavoidable. Describing policy measures to lift Australia’s lagging productivity as ‘a game of inches’, Productivity Commissioner, Danielle Wood observed, ‘you actually need governments to move across a range of different policy fronts at once. There’s no single lever that you pull that makes all the difference.’^[28]

The ‘Australian policy cycle’

The ‘Australian policy cycle’, formulated over 20 years ago specifically for an Australian practitioner audience, is an enduring – if somewhat idealistic – model of the policy development process. The model is a signature feature of The Australian Policy Handbook, first published in 1998, republished in its 7th edition in 2023 and billed as an ‘essential guide for students and practitioners of policy making in Australia’.^[29]

Whereas theoretical models of the policy process seek explanations through investigations of institutional, political, organisational and cultural factors that shape the policy environment, the ‘Australian policy cycle’ is more of a ‘how to’ guide and presents policy making as a sequence of practical actions. The policy cycle largely follows cabinet processes in a Westminster system, and the Handbook’s authors assert that ‘good policy should include the basic elements of the cycle’. The strength of the model is its practical approach, which captures the entirety of policy development and implementation, although it does not supply causal explanations of policy.

A policy cycle approach can help public servants develop a policy and guide it through the institutions of government. The policy cycle starts with a problem, seeks evidence, tests proposals and puts recommendations before Cabinet. Its outcomes are subject to evaluation and the cycle begins again. The policy cycle offers a modest and flexible framework for policy makers.^[30]

The policy cycle model has been criticised for suggesting a far more linear and logical progression of activities than would ever be observed in practice.^[31] Critics also point out that the model does not accurately capture the lived experience of policy professionals.^[32] The Australian policy handbook’s authors have engaged openly with such critics and have responded to their criticisms in the following terms: ‘The policy cycle does not assert that policy making is rational, occurs outside politics, or proceeds as a logical sequence rather than as a contest of ideas and interests’.^[33]

In simple terms, the policy cycle entails eight logically sequenced steps:

1. identify issues
2. analyse policy options
3. select policy instruments
4. consult affected parties
5. co-ordinate with stakeholders
6. decide preferred strategy
7. implement policy
8. evaluate success/failure.

Cross-portfolio policy co-ordination

Policy generated in one ministry or portfolio can have impacts on policy in other ministries, portfolios and agencies. Similarly, policies originating in one jurisdiction can have consequential implications for intergovernmental relations, including between national (federal) and subnational (state, territory and local) governments (one example being the impact of changes in revenue or spending decisions by the Commonwealth government upon state/territory governments) and between nations in respect of multilateral or bilateral agreements (examples being trade agreements or United Nations conventions).

In Australia, central agencies of government –  that is, the Commonwealth Departments of the Prime Minister and Cabinet, Treasury, and Finance (and their State and territory counterparts) – perform an essential policy co-ordination role. It falls to these agencies to review policy and budget proposals emanating from ministers and their departments and to seek comment from other ministries and agencies to identify any unintended consequences that might arise. Once comments have been compiled from affected agencies – including agencies responsible for government revenue – a briefing, together with recommendations, will be prepared for the consideration of Cabinet. Vetting of this nature often requires specialist knowledge of particular policy domains and of the statutory basis for government programs and services. It also depends less on political theory and more on an appreciation of the practical and pragmatic dimensions of public policy.

Policy analysis

The aphorism ‘the best is the enemy of the good’, commonly attributed to the French Enlightenment writer and philosopher Voltaire (1694–1778), neatly encapsulates a key challenge of public policy. At some level, all policy decisions represent compromises between different interests and involve considerations about political acceptability as well as economic and technical feasibility. To quote the 19th-century German statesman Otto von Bismarck, ‘politics is the art of the possible’ – likewise, policy is the art of the achievable.

Policy analysis is an important part of the ‘craft’ of policy making. The task of the analyst is to understand the implications of policy decisions in terms of their impact on the policy problems being addressed; any unintended consequences for government or the community; and their legal, economic and technical soundness. Policy analysis is essential for the provision of policy assurance and enables the analyst to provide answers to the following key questions:

• Is the policy well targeted?
• Is the policy delivery architecture well designed?
• How will performance be measured?
• How will we know if the policy is working?
• If the policy is not working, what corrective action is available?

In order to make reliable pragmatic judgements about such matters, it is important for the analyst to give close consideration to a wide range of factors. The kinds of questions the astute policy analyst might ask include:

• Is the policy framed within a particular political or philosophical perspective, and is it consistent with the values and policy platform of the governing party or parties?
• Is the policy genuinely directed towards solving a problem in public policy, or does it primarily seek to solve a ‘political’ problem by creating the impression of action while having little tangible effect?
• Have similar policies been pursued in other jurisdictions and to what effect? How might past experience inform policy implementation?
• What are the competing options to achieve the policy aims, and how do they compare? Does the policy require enabling legislation? What policy instruments or tools are available to give effect to the policy? What are the expected/hoped for impacts of the policy, and how might these be reliably measured and reported?
• Which groups or communities of interest – including classes of workers, trade unions, professional associations, advocacy organisations, industry groupings, communities and/or geographical regions and expected beneficiaries – stand to be affected by the policy and in what manner,?
• Does the infrastructure exist to give effect to the policy? Is there a functioning market framework within which the policy might be delivered? What skills base is necessary to deliver the policy? Is an appropriately skilled workforce available? What capacity exists within the public and non-state sectors to give effect to the policy?
• What will policy implementation cost? Is it affordable? Will delivery be selectively targeted, means-tested or otherwise ‘rationed’? Is it possible to offset expenditure through some form of cost recovery, such as user fees? What are the principal cost drivers in the policy space?
• How will the policy be delivered and governed? What systems or frameworks need to be established to provide assurance to government that policy implementation and delivery will occur within prescribed timeframes and budgets? What systems or frameworks are available to ensure that the policy is performing in the expected manner?
• Has provision been made for periodic evaluations of the effectiveness of the policy and/or the operational arrangements established to give effect to the policy, and is there a capacity to make necessary adjustments to the policy and/or management structures should evaluation findings so indicate?

One might also inquire as to the ‘democratic’ dimensions of policy making. Butcher (2024)  discusses important sources of democratic deficit inherent in Australian policy making and posits key criteria that might be used to conduct a ‘democratic audit’ of the policy process in a liberal democracy, namely:

• Is there an electoral mandate for the policy or, where it is not possible to seek approval from the electorate, how has the government made the case for new policy?
• Is the policy consistent with an election promise and/or party values and priorities?
• Whose interests are served by the policy?
• Has the need for the policy been established? Or is the policy ‘a solution looking for a problem’?
• Is there a legal basis for the policy?
• Is the policy process transparent and accountable?
• Does the policy require a social licence and does a social licence exist?.^[37]

Policy instruments

Policy instruments enable the application of policy decisions in practice and can be grouped into the following major categories:

1. money (spending and taxing powers)
2. law (including regulation)
3. government action (e.g. delivering services)
4. advocacy (e.g. educating, persuading)
5. networking (e.g. cultivating and using relationships to influence behaviour)
6. narratives (e.g. using storytelling and communication – including public advertising)
7. behavioural economics (e.g. using economic incentives to induce behaviour change, or ‘nudging’ as it has come to be known).^[38]

It should be noted that in the real world these categories often overlap and a mix of instruments is generally required. For instance, governments might elect to use a form of direct service delivery (government action) to achieve policy aims; the delivery of services requires statutory authority (law); is funded by government appropriations (money); and employs ‘nudge’ strategies (behavioural economics), advertising (narratives) and public education (advocacy) to achieve the government’s policy aims.

Government policies aimed at reducing the harms from the use of tobacco products provide a good example because they employ all the instruments named above:

1. Money: the collection of excise tax on cigarette sales to provide a source of funds for medical research and for non-government organisations involved in anti-smoking programs.
2. Legislation/regulation: setting age restrictions on the purchase of tobacco products, banning smoking in public places and restricting the sale, advertising, distribution and packaging of tobacco products.
3. Government action: funding the delivery of preventative health services aimed at assisting smokers to quit.
4. Advocacy: there have been multiple education campaigns on the health risks associated with tobacco and how to quit.
5. Networking: successive governments have entered into partnerships with representative bodies, such as the Australian Medical Association, and non-government organisations advocating smoking reduction.
6. Narratives: anti-smoking campaigns utilising various media and featuring testimonials by former smokers and/or portraying the health and other impacts of smoking on real people.
7. Behavioural economics: levying excise taxes to increase the purchase price of tobacco products and/or offering financial incentives to quit smoking.

It is worth noting that the choice of policy instrument is all too often a function of familiarity, as opposed to optimal fitness for purpose (in other words, policy makers stick to what they know). Other factors influencing the choice of instruments include:

• the characteristics of the policy area in question (e.g. some policy areas might have a long history of recourse to particular models of implementation, and this predisposes policy actors in those areas to prefer those models)
• available resources (e.g. some policy instruments might entail significantly higher establishment and running costs than others, or they might require skills or technologies that are in short supply, leading to the selection of less optimal but more feasible options)
• ease of administration and/or administrative traditions (e.g. some policy instruments might be inherently easier to administer, while others entail greater complexity and risk; in some policy domains particular traditions – say, centralised, hierarchical management frameworks, as opposed to decentralised, distributed frameworks – might predominate, predisposing practitioners towards the selection of instruments that ‘fit’ with the existing administrative culture)
• the political dimension (e.g. recourse to particular policy instruments might be precluded because they are not considered to be acceptable to the community and/or they might be unacceptable to governments on ideological grounds).^[39]

implementation failure

The true test of any policy lies in its implementation. Implementation gives practical effect to policy. It is a complex process requiring application of a range of technical and management skills. Many seemingly ‘good’ policies fail in their implementation, resulting in a failure to achieve expected outcomes or in unintended ‘perverse’ outcomes. As observed by Hummel et al.:

Policy implementation is a complex and contested process that can involve multiple stakeholders, sectors, systems, and responsibilities. It is not a straightforward progression from policy design but rather a dynamic and political endeavour.^[40]

The so-called Robodebt scheme offers a potent example of cascading technical, administrative and ethical failure.^[41] Designed to enable Australia’s welfare payments agency, Centrelink, to recover alleged overpayments from welfare recipients, the automated debt recovery system wrongly raised debts that were not in fact owed. At the same time, an aggressive compliance regime effectively placed the onus on vulnerable recipients to prove they had not been overpaid, while senior officials and ministers downplayed growing community concerns. In 2021, the Federal Court approved a class action settlement of $1.8 billion, after finding that the scheme was unlawful.^[42]

Implementation failure can occur anywhere from the policy design phase through to ‘roll out’, and can be caused by any combination of:

• inadequate research, design and planning
• poor co-ordination and inadequate consultation with stakeholders
• insufficient resourcing and capacity constraints
• legislative and regulatory gaps
• proceeding too quickly and/or failure to ‘pilot’
• failure to anticipate and/or effectively manage risks
• ineffective governance and/or administrative architecture
• multiple and/or incompatible policy goals.

Implementation failure entails significant costs in terms of finite resources (such as money, labour and time), reputation and trust. These include a failure to realise intended policy aims; loss of public confidence; costs of restoration, rectification or redress; costs arising from bringing failed programs to a premature end; lost opportunities (opportunity costs); and, for governments, loss of political capital (with potential electoral consequences). It is important to recall that policy making is, and remains, inherently ‘political’ and that ‘policy success’ will always be a contested assessment. Indeed, it might be said that ‘failure’ has been ‘weaponised’ in Australia’s contemporary political culture.

Readers wishing to explore this topic further can look to Dowding et al.^[43]

Artificial intelligence and policymaking

Artificial intelligence (AI) is rapidly becoming a ubiquitous feature of our everyday lives. In the process It is transforming our media landscape, revolutionising our ability to assimilate information, and challenging our ability to discern truth from fiction. The application of AI to politics and policymaking is a subject that animates much scholarly debate – sometimes pitting AI utopians against AI deniers and doomsayers. What is clear is that AI is set to become an important tool of policy practice.

Many champion the capacity of AI to deliver ‘practical, problem-driven solutions that reduce administrative burden, streamline workflows, and improve navigation and responsiveness’^[44] while others argue that we do not presently have the guardrails in place required to manage the risks posed by AI.^[45] Hart et al. point out that AI can both augment and amplify (as opposed to replace) human capability: keeping ‘a human in the loop’ is essential, especially when dealing with the ethical and moral issues inherent in public decision-making.^[46]

While Hart et al. believe that AI ‘has prompted a reassessment of the fundamentals of policy craft’, they caution that ‘we need to not just consider the strengths and limitations of these tools but also address the ways that these features might fundamentally change how policy work is conceptualised, developed, and supported’.^[47]

The risks posed by this emergent technology in government include problems of transparency and accountability, algorithmic bias, threats to privacy and national security, and ethical risks posed by an overreliance on the technology. The risks to government also extend to its use by third party contractors. In 2025, for example, it was reported that a major consulting firm was found to have used the application ChatGPT to produce a report to government that was ‘littered with bogus citations and contained fabricated and hallucinated legal precedents that would have been part of the official evidence if not exposed by journalists’.^[48]

We do not have the space here for an exhaustive analysis of the implications of AI for the ‘craft’ of policy making. Fortunately, there is a large and growing literature to which the interested reader may turn if they wish to drill deeper.^[49]

Conclusions

In this chapter we have attempted to introduce readers to a spectrum of ideas about the nature, formulation and ‘craft’ of policy making. In so doing, we have tried to:

• acquaint readers with the major theoretical approaches to understanding the policy process
• equip readers to more effectively understand past and present policy debates
• enable readers to interrogate the processes of policy development, implementation and evaluation.

Policy design and implementation is a complex and imperfect process that is often seen as more of a ‘craft’ than a formal discipline. Policy professionals tend to ‘learn on the job’, and even those who have formal qualifications in public policy or exposure to the academic study of policy often find that the pragmatic reality of policy making aligns poorly with policy theory.

The Australian policy cycle and the public value approach offer sound practical templates for policy design and evaluation. Unfortunately, as will be attested by many policy professionals working within government, ‘policy craft’ is seldom conducted in full accordance with such orderly, rational models.

In contemporary policy spaces, effective policy craft increasingly comes down to working effectively within networks inside and outside government. Today’s policy professional needs to be acutely aware that governments have many sources of policy advice and that many of these sources have vested interests in particular outcomes. Above all, a capacity for critical reflection and an ability to anticipate the risks and consequences of policy choices provide the foundation of sound policy practice.

It is our hope that the concepts canvassed in this chapter will assist readers to make sense of scholarly and media accounts of policy histories and policy making in different domains and of the changing role, form and modus operandi of the public sector.

References

About the authors

Dr John R. Butcher is an adjunct research fellow in the John Curtin Institute of Public Policy within the Curtin Business School at Curtin University. His principal research focuses on the relationship between government and the not-for-profit sector. He is a contributing author to Australia’s Evolving Democracy: A New Democratic Audit (2024), co-author (with David Gilchrist) of Collaboration for Impact (2020), co-editor (with David Gilchrist) of The Three Sector Solution (2016) and co-author (with John Wanna and Ben Freyens) of Policy in Action (2010).

Dr Trish Mercer is a Visiting Fellow at the Centre for Social Policy Research, Research School of Social Sciences, Australian National University. Her diverse career in the Australian Public Service, with over 20 years at the senior executive level, spanned policy development and research, financial management, program implementation, community education and direct service delivery experience. Trish is now involved in public policy research and executive education training. Trish co-edited and contributed to Learning Policy, Doing Policy (2021), exploring how theories of the policy process can be transferred and taken into practice. Her social policy interests include early childhood, schools and employment services. Trish is working with former colleagues to investigate how public servants exercise their ‘agency’ to influence policy design and program implementation.