Secondary links

User links

Frameworks for looking at policymaking

Researchers have developed many different frameworks that are useful for understanding policymaking. Some frameworks may be useful starting points for strategizing on how to best influence policymaking. This guide focuses on a selection of these policy analysis frameworks and highlights the implications of each for governance assessments:

The policy stages framework

The stages model divides the policy process into a series of stages that can be useful as entry points for influencing policymaking. The model often includes the stages of:
  1. Identification and agenda setting of a policy problem
  2. Formulation of policy options
  3. Approval of policy
  4. Implementation of policy
  5. Evaluation of policy

In the first stage, governance assessments may assist in identifying new policy issues and set these on the political agenda. Civil society in particular may use governance assessments for this purpose. In the second stage, governance assessments may assist in developing policy recommendations for policy reform. In the final stage, governance indicators may be used in the evaluation of policy, such as National Development Plans.

Because this model is the most common way practitioners think about the policy process, it is important to be aware of some of its shortcomings:


  • In real life the proposed sequence of stages is often jumbled. For example, policies may be approved post implementation, or the evaluation of a policy may simultaneously lead to the formulation of new policies. Governance assessment exercises may therefore wish to target policy-making across the stages of policy-making and not only focus on agenda setting, evaluation or any other single stage.
  • Usually there is not one policy cycle, but a process of multiple, interacting cycles, involving numerous policy proposals at multiple levels of Government. Governance assessment exercise that aim to cover a broad range of issues will in particular see that inserting findings and recommendations into policy-making may occur across a wide range of policy initiatives at local and central Government, in Parliament, within sectors and within the Judiciary.
  • The policy cycle may not be effective, and it may therefore not be a good entry point. For example, if governance assessments align themselves with national monitoring and evaluation plans, this is only helpful if these plans are acted upon and followed up by Government.
  • The stages model is not a causal model. It does not identify a set of causal drivers that govern the process within and across stages. In particular, it provides no answer as to why or why not governance assessment policy recommendations are acted upon or not in the approval stage. This will require an additional understanding of actors, interests, institutions and incentives.

The multiple-streams framework

The multiple streams model views the policy process as consisting of three streams of actors and processes:
  1.  A problem stream: this stream consists of evidence about various problems and the proponents of various problem definitions.
  2. A policy stream: this stream involves policy recommendations and proponents of various solutions to policy problems.
  3. A politics stream: this stream consists of elections and decision-makers.
For governance assessments it therefore becomes important to:
  • Link the governance assessment’s definitions of governance problems with issues already raised, such as key policy issues.
  • Provide understanding of the existence, magnitude and depth of governance problems in the problem stream.
  • Link the definition of a problem with solutions in the policy stream. In most cases many policy initiatives already exist, and it is important to link with the proponents that support similar policy recommendations to that of the governance assessment. In cases where Government commission governance assessments as part of reform initiatives there may nevertheless be important to make these links: Assessment conclusions may not correspond with the preferred policy solutions of the entity that commissioned the research.
  • Link definitions and solutions with the politics stream. This may include linking to the national mood in a country. The notion of 'national mood' refers to that a fairly large number of individuals in a given country tend to think along common lines, and that the mood swings from time to time. For example, the national mood' may be the perception that the country's development is hindered by high levels of corruption. Linking to the politics stream may also include linking to pressure group campaigns, in this case, anti-corruption campaigns for example.

Institutional incentives framework

This framework is an umbrella for several frameworks that use individuals’ self-interest as their starting-point and look at how institutions create incentives that change the behaviour of people.

Sometimes what is a rational choice for an individual would be an irrational choice if a group of people were to decide what is best for the community. For example, providing better services for the poor may be a public good that would benefit all through reducing crime and in the long-term increase overall wealth, but go against the immediate interests of some groups that may have to pay more in taxes for the state to pay for poverty alleviation programs. Incentives for decision-makers may favour the immediate interests of the latter group compared to that of the long-term interests of society as a whole.

This framework points out several incentives of decision makers and bureaucratic systems that affect how governance assessment may feed into policy. In particular:

  • It may be important to keep in mind the need for immediate quick win for policy makers in order ‘to sell’ policies to their constituents, and not only focus assessment recommendations on long-term objectives.
  • It may also be more politically feasible in some circumstances to focus recommendations on ‘universal’ services instead of 'targeted' approaches: There may be more willingness in the population to for example improve education for all, than improve education only for the poorest.

Major policy change framework

In this framework policymaking is characterized by long periods of incremental change punctuated by brief periods of major policy change. Punctuations may be caused by regime change, changes in the national mood, changes in international or geopolitical circumstances, a political turn-over in politics, as well as other changes in national circumstances.

For governance assessments events that cause a shift in a country’s policy direction may constitute an opportunity for real change. For governance assessments it is important to seize these windows of opportunity. For example, the African Peer Review Mechanism ( that took place in Ghana from 2003-2006 occurred after a turn-over in politics which marked the end of 20 years of rule by President John Rawlings. The 'State of Democracy' assessment that took place in Mongolia occurred as a follow-up to the Fifth International Conference of New and Restored Democracies held in Ulaanbaatar in 2003.

However, this framework also correctly lowers expectations of what governance assessments can achieve during periods without major policy change which will be the time most governance assessments take place: Normally one can only realistically expect smaller and incremental changes.

The advocacy coalition framework

The advocacy coalition framework focuses on the interaction of advocacy coalitions, each consisting of actors from a variety of institutions and organizations who share the same policy ideas, within a policy area. This approach highlights several aspects of policy-making important to governance assessment exercises that aim to affect it:
  • Mutual learning between advocacy coalitions, or two different groups of people with different opinions on a policy area, is more likely to take place if there is a level of informed contestation. Governance assessments should seek to provide evidence that can be interpreted and analysed from different viewpoints, and facilitate evidence-based debate between opponents with different views to facilitate mutual learning.
  • Mutual learning between advocacy coalitions is more likely to occur if what is learned is uncontroversial and does not touch upon the core values of any of the groups. When building consensus in governance assessments, it may therefore be useful to start with issues that are the least problematic for stakeholders to agree to.
  • A focus on short-term decision-making will underestimate the influence of policy analysis because such research is primarily used to alter the belief system and ideas of decision-makers over time. Governance assessments should therefore aim to be in the game for the long-run.