Public policy is made through a process in which political institutions identify problems or objectives, gather evidence, develop possible responses, choose among options, establish legal or administrative authority, implement decisions and evaluate their effects. It is not normally a single decision made by one person. Ministers, legislators, civil servants, analysts, regulators, local authorities, experts, interest groups, affected communities and sometimes courts or international institutions can influence different stages. The exact process varies between political systems and policy areas.
In parliamentary government, elected ministers typically establish political priorities while public officials develop advice, analyse options and manage implementation. Legislation may be required for some policies, while others can be implemented through existing statutory powers, regulations, administrative decisions, spending programmes or guidance. The OECD describes public governance as encompassing formulation, implementation, communication and evaluation, while noting that policy making does not necessarily follow a linear path.
The process usually begins with recognition of a policy problem, although governments can also begin from an opportunity, political commitment or proposed solution. Problem definition is important because different definitions of the same situation can lead to different policy responses. Rising unemployment, for example, could be framed as a skills problem, a demand problem, a regional inequality problem, a labour market regulation problem or a combination of several issues.
Policymakers therefore need evidence about causes, affected groups, existing interventions and likely future developments. Government policy systems increasingly use data, research, stakeholder engagement, behavioural analysis, economic analysis, forecasting and user research to improve diagnosis. The UK Policy System Framework identifies inputs such as time, resources and diverse evidence alongside practices, relationships, culture and context as components of the environment in which policy is made.
Once a problem has been defined, policymakers can develop objectives and possible interventions. An objective states what government is trying to achieve, while a policy instrument is a mechanism used to influence behaviour or produce a desired outcome. Instruments can include legislation, taxation, spending, regulation, information campaigns, public services, grants, procurement rules, standards or direct government operations. Several options may be developed rather than immediately selecting the first proposed solution.
Analysts can assess costs, benefits, risks, distributional effects, feasibility, implementation requirements and unintended consequences. The UK Green Book provides a framework for appraising costs, benefits and risks of options for achieving government objectives and explicitly distinguishes appraisal from the political decision about what objectives should be pursued.
Evidence is used throughout policy development, but evidence does not mechanically determine policy. Governments must make choices about priorities, values, resources and acceptable trade offs. Scientific evidence can indicate the likely effects of an intervention, economic analysis can estimate costs and benefits, administrative data can reveal patterns of service use and consultation can identify practical consequences experienced by affected groups. Political institutions then determine which objectives should be pursued within constitutional and legal constraints. Open policy making approaches can involve users and stakeholders in diagnosing problems and developing possible solutions. The UK Open Policy Making toolkit describes stages including diagnosis, discovery, development and delivery, illustrating how user needs, evidence and practical testing can feed into policy design.
A policy may require formal political approval before it can be implemented. In a parliamentary system, legislation can require approval by Parliament, while spending decisions may require appropriate budgetary authority. Ministers can sometimes use existing statutory powers to make administrative or regulatory changes without passing new primary legislation. The legal route therefore depends on the policy instrument. Once a government has chosen an option, officials may draft legislation, regulations, guidance, funding arrangements, operational procedures and performance measures.
Legal advisers examine whether the proposed action is within existing powers and compatible with applicable law. The policy decision and its legal implementation are consequently related but distinct stages. A political announcement does not necessarily mean that the administrative machinery, legal authority and resources required to deliver it already exist.
Implementation is where policy decisions are converted into practical activity. Departments may need to establish new services, change regulations, distribute funding, recruit staff, develop digital systems, communicate requirements, train professionals or establish enforcement arrangements. Implementation should be considered during policy design because a theoretically effective policy can fail if delivery organisations lack capacity, information, funding or workable procedures. UK government implementation guidance explicitly warns that policies and projects can fail when delivery planning is postponed until after policy decisions have been made. It describes implementation as a cross cutting discipline involving tools for analysing problems, planning delivery and understanding how policy interventions produce real world effects.
Evaluation determines whether a policy achieved its intended outcomes and why. Monitoring can establish what activities occurred and whether outputs were delivered, while evaluation investigates effects and causal relationships. A programme might successfully distribute a service to thousands of people but still fail to achieve its intended social outcome. Conversely, an apparently expensive intervention might produce benefits that justify its cost. Good evaluation can therefore inform continuation, modification, expansion or termination.
Current UK government guidance describes evaluation as important for understanding whether and how interventions work and for ensuring public expenditure is used effectively and efficiently. The 2026 Magenta Book emphasises evaluation as a means of supporting accountability, future spending decisions and iterative improvement of public services.
Public policy making is therefore better understood as an iterative system than as a simple sequence ending with implementation. Evaluation can reveal that the original problem was misunderstood, that an instrument did not work as expected or that external conditions changed. Governments can then revise the policy, redesign implementation, change funding, introduce another instrument or abandon the intervention. Political priorities can also change before evaluation is complete.
Public policy is consequently shaped by evidence, institutions, political authority, administrative capacity, law, resources, stakeholder interests and changing circumstances. The OECD explicitly notes that policy making does not necessarily follow a linear path and describes formulation, implementation and evaluation as interconnected stages. This means that a complete account of policy making must consider not only how governments choose policies but also how those policies are translated into administrative action and subsequently learned from.
