Policy Cycle Analysis

The Right Answer at the Wrong Time

Consider an industry association that commissions an impeccable cost-benefit analysis of a proposed launch-liability reform, only to see the document fall flat because the relevant ministry is not yet debating costs and benefits — it is still deciding whether the issue deserves attention at all. Or a government-affairs team that invests months producing a compliance roadmap for a regulation that never emerges because the policy is stuck between formulation and adoption, with no veto player willing to sign. Or a coalition that lobbies furiously to shape a text that, once adopted, is quietly neutralised by an implementing agency with different priorities and the operational discretion to reinterpret it.

These are not failures of analysis. They are failures of timing. Each was the correct move in a different stage of the policy process. Delivered at the wrong moment, a sound argument becomes noise, a well-designed coalition becomes misaligned, and an authoritative document becomes a curiosity. The common diagnostic error is treating a policy as a single object — “the debris mitigation rule,” “the spectrum allocation reform” — rather than as an object moving through a sequence of stages, each with its own dynamics, its own actors, and its own decisive question.

Policy cycle analysis is the discipline of fixing that error. Its core claim is simple and consequential: before you decide what to do about a policy, you must decide where it is in its lifecycle, because the question that matters changes with the stage.

From Lasswell’s Seven Steps to Modern Heuristics

The idea arrived through Harold Lasswell in 1956. Writing in the early years of the post-war policy sciences, Lasswell proposed a seven-stage decision process — intelligence, promotion, prescription, invocation, application, termination, appraisal — that aimed to give the still-emerging field of public policy a shared descriptive vocabulary. The scheme was clunky, but its underlying move was durable: treat a policy not as a static intervention but as a sequence of decisions, each with distinct characteristics.

Charles Jones in 1970 and James Anderson in 1975 refined the model into something closer to the version taught today. Jones compressed the stages, brought in an explicit agenda-setting phase, and separated adoption from implementation. Anderson tightened the vocabulary further and introduced the five-stage canonical form — agenda-setting, formulation, adoption, implementation, evaluation — that became the lingua franca of comparative public administration by the 1980s. Michael Howlett and M. Ramesh, in successive editions of their policy analysis textbook through the 1990s and 2000s, gave the framework its mature expression: a heuristic, not a theory, useful precisely because it organises messy reality into questions an analyst can answer sequentially.

The model has always had critics. John Kingdon’s multiple-streams framework, developed in the 1980s, argued that problems, policies, and politics flow through institutions as separate streams and only occasionally converge — and that the stages model’s linear ordering misrepresents this coupling. Advocacy-coalition scholars, notably Paul Sabatier, objected that the stages model underweights ideology and belief systems. Punctuated-equilibrium theorists argued that policies oscillate between long stasis and abrupt change rather than moving through orderly stages.

These critiques are substantially correct, and Howlett and Ramesh themselves treat the cycle as a heuristic rather than a model of reality. But the heuristic has survived its critics because it is the best available discipline for forcing the analyst to ask a prior question: before I recommend, do I know what stage I am in? The alternative — stage-blind analysis — is worse than the stylisation.

The Diagnostic Move

What policy cycle analysis does, and what neighbouring methods do not, is force an explicit answer to the stage question before any substantive analysis begins. This discipline has consequences.

What separates policy cycle analysis from neighbouring methods is this stage-contingent vocabulary. A regulatory impact analysis assumes the policy is being formulated; a stakeholder mapping is most useful in adoption and implementation; a scenario planning exercise is typically pitched at agenda-setting. Policy cycle analysis is the method that tells you which method to use.

The Cycle at Work: Traffic Management Through the Stages

Consider an analyst asked to advise on space traffic management regulation in a major launching state. The instinct of an unguided analyst is to produce a cost-benefit analysis of candidate regulatory options. The instinct of a cycle-guided analyst is different: first, place the file in its stage.

In this jurisdiction, the issue has moved out of early agenda-setting. Multiple focusing events — a succession of close conjunctions publicly reported by monitoring services, the rapid growth of a major constellation, visible congestion in sun-synchronous orbits — have pushed orbital congestion into ministerial awareness. Several legislative proposals exist. No single authoritative option has consolidated. The issue sits in late agenda-setting bleeding into early formulation. The problem definition is still contested: some actors frame the question as a safety issue properly owned by the aviation regulator, others as a commercial competitiveness issue owned by the economic ministry, others as a national-security issue owned by the defence ministry. The framing competition, not the option design, is the decisive dynamic.

The cycle-guided analytical artefact is not a cost-benefit table but a framing map: which problem definition is gaining ground, which actors are pushing each definition, and which definition would, if it won, determine the future trajectory of the file. The analyst who produces this artefact has given the client something actionable, because coalition-building in support of a preferred framing is the action that matters during agenda-setting.

The second analytical move is the transition question. What would it take for the file to move from agenda-setting to formulation, and from formulation to adoption? In this jurisdiction, the structural blockage is jurisdictional: three agencies claim a partial interest, none has the sole authority to act, and each defends its mandate. The formulation stage cannot consolidate until one agency receives clear statutory primacy. The transition trigger, then, is either a focusing event severe enough to force legislative attention — a collision with visible debris consequences, for instance — or an executive action consolidating authority. Both are rare. The analyst’s honest finding is that this file is likely to remain in late agenda-setting for years unless a trigger arrives, and that a client’s strategy should be calibrated to that probability rather than to an imagined near-term regulation.

The analytical insight this produces — an insight unreachable from option-level analysis — is that the right first investment for a stakeholder in this file is not in compliance preparation but in framing leadership. The client who shapes the problem definition during agenda-setting has disproportionate leverage in formulation, and the framing cost is a fraction of the compliance cost.

Where It Holds, Where It Limps

Policy cycle analysis holds, better than any rival, the case where timing is the decisive variable. It is the method that disciplines an analyst to match the question to the stage, and it is the method that most reliably prevents the expensive error of producing the right answer to the wrong question. For any file spanning more than a single decision, the cycle lens is almost always worth the time it takes.

Its limits are the limits of the stages model itself. Real policy-making is not linear. Stages loop back: implementation problems reopen formulation, evaluation triggers renewed agenda-setting, adoption can reverse itself. The stages are an analytical stylisation, not a description of how files actually behave, and the analyst who treats them as a descriptive model rather than a heuristic will mis-stage every file whose trajectory is non-linear — which is most of them.

The method is also weak on the mechanisms inside each stage. It tells you which stage you are in; it does not tell you how to analyse that stage’s dynamics. Agenda-setting demands a problem-definition analysis, constructivist reading, or media framing study; adoption demands a decision-process analysis and stakeholder mapping; implementation demands principal-agent and capacity analysis. Policy cycle analysis is the wrapper; the stage-specific methods are the content.

It is also normatively silent. The cycle lens describes process; it says nothing about whether the resulting policy is good. An analyst who wants to assess desirability needs regulatory impact analysis or one of the normative evaluation frameworks. The cycle tells you whether a policy will pass, not whether it should.

A more subtle weakness is the reification risk. A stage label can substitute for analysis. “This file is in adoption” is the starting point of an argument, not its conclusion. Analysts who announce the stage and stop have done the cheapest possible version of the method.

The cycle lens pairs naturally with decision-process analysis (which unpacks the procedural mechanics of the adoption stage), stakeholder mapping (which identifies actors differently at each stage), and realist or constructivist analysis (which supplies the power and ideational dimensions the cycle model underweights).

A Note for the Practitioner