Guidance

Understanding political economy analysis and thinking and working politically

Published 27 February 2023

Foreign, Commonwealth & Development Office (FCDO) and the Thinking and Working Politically Community of Practice, 2023

Authors: Alan Whaites, Laure-Hélène Piron, Alina Rocha Menocal and Graham Teskey

1. Introduction

This guide ^{[footnote 1]} is adapted from work by the UK Foreign, Commonwealth and Development Office (FCDO) with inputs from members of the Thinking and Working Politically Community of Practice (TWP CoP).

It outlines how to understand and use a set of analytical tools that are collectively known as Political Economy Analysis (PEA). The guide aims to equip practitioners to act in an informed manner, given that development objectives are invariably politically complex, and entail engaging with counterparts’ political incentives and preferences.

The guide summarises different types of tools – from very light-touch to more in-depth approaches – and provides advice on how development professionals can decide what is most appropriate in a given context, with illustrations based on the experiences of teams working on these issues.

This guide will help development professionals and others to make use of PEA and to apply it to their own specific needs. The first part of the guide offers a general picture of the approach. The second part provides more specific guidance for those who are tasked with deploying a PEA.

Contents	Main audience
What is PEA, its role and purpose (Section 2)	General information for all readers
The main elements of PEA (Section 3), and Thinking and Working Politically (Section 4)	Core information for teams planning and using PEA
How to ensure quality (Section 5)	Essential reading for those directly responsible for a PEA
Important concepts and terminology (Annex)	General information for all readers

2. What is the purpose of a political economy analysis?

Applied PEA consists of a set of concepts, questions and tools that provide a clearer picture of the political context in which development assistance is provided. It does this by systematically exploring the underlying reality of policy, contestation and decisions, including those elements that are counter-intuitive. At a strategic level, these insights can help development professionals and local actors to better understand why things work in the way that they do, and to draw practical implications. Applied to specific problems or objectives, they can help to unlock constraints and identify actions, indicating the most realistic space in which to engage. The 3 main questions that a PEA will usually address are:

• what is the underlying issue or problem that a team wishes to address (which may lie beneath the more obvious consequences of those problems)?
• why does this situation persist in this form (such as insecurity or gender inequality)?
• how can change come about?

In exploring these questions, it is important to consider the nature of exclusion from processes and decisions, the external interests that might be involved (eg actors outside the context), the space and opportunities for change, and the potential role of donor agencies and other development partners.

PEA tools equip teams to explore these questions and to analyse the relevant context in ways that can be easily shared, discussed and used – enhancing internal collaboration. PEA tools enable them to on insights from economics, political science, history, and sociology, among other disciplines.

As a result, the findings of PEA should ensure that a team will benefit from a deeper understanding of the following issues:

• who are the influential actors and interest groups (see Box 1), and what are the interests (economic or otherwise), preferences and beliefs that they bring to bear?
• how power is distributed among these actors and how is that distribution of power contested, who is excluded from power and resources, why, and how?
• what practical constraints limit and shape the use of that power?
• what are the formal and informal norms (eg understanding of rules) that shape behaviour and influence decisions (eg discrimination against women, religious or ethnic minorities, or people with disabilities)?
• how do these factors affect incentives and the ability to achieve change?
• as well as the role of ideas, capacity, identities and loyalties, which priorities are likely to prevail when there are multiple objectives?
• how and why does change really happen? What are the combinations of actors, incentives, processes and capacity that lead to shifts in outcomes?

On the basis of this analysis (which may be one-off, but is better repeated on a regular basis), it is possible to identify approaches that ‘best fit’ the context, with a good awareness of both risks and opportunities. This should inform theories of change on what is feasible, while also challenging assumptions related to purely technical approaches. It will provide a picture of why some parts of partner governments work better than others, and why managing power involves prioritising certain types of action or programme.

PEA was initially regarded as being concerned only with problems – such as identifying constraints. It is now increasingly used to identify opportunities for leveraging policy change and supporting reform, particularly using action-oriented approaches termed Thinking and Working Politically (see Section 4 below). PEA helps us to better understand the political realities that partner governments face, so that we can work more effectively with them on reforms in sectors such as health, education and infrastructure. A good PEA will debunk simplistic or linear assumptions about why policies are adopted or rejected, questioning conventional default explanations, such as ‘lack of political will’. Insights from a PEA can raise awareness of unintended consequences, which is essential to avoid doing harm, and will also help in navigating challenging contexts, such as Small Island Developing States (SIDS), where small populations can limit available sources of analysis.

In summary, the objective of a PEA approach is to identify a more realistic response which is both politically feasible and technically sound.

Box 1: Data-driven diplomacy using network mapping: ‘Are we talking to the right people?’

Strategic engagement plans at the country level often identify the most significant politicians, but understanding of their networks and spheres of influence is opaque. Often individuals outside the group of top-level politicians and ministers are overlooked, leading to the tendency to talk with the same individuals on our policy priorities. A current and comprehensive picture of the political landscape of actors does not always exist.

The FCDO’s Centre for Data and Analysis (CfDA) partnered with a British embassy team to pilot a new in-house network-mapping tool, which has been developed to visualise relationships within a political landscape.

The tool takes a data-driven approach to create network maps that capture relationships by analysing news articles about politics in a given country. The network maps are created using an algorithm which searches for, and conducts analysis on, names within a string of text. The data is then exported to create maps of individuals and shows who has prominence and the strength of connections within a network.

The CfDA worked closely with the embassy team on collecting data sources, and to develop, apply and iterate this methodology to their political context. Over 11,000 news articles were analysed, spanning political alignments, and multiple languages, between January 2021 and February 2022; and nearly 500 politicians and influencers in the relevant country were identified and analysed. The network maps were then examined in a workshop to draw out key insights and implications for engagement.

This process helped to create a fuller picture of the main political actors and identified some unexpectedly strong connections between individuals in second-tier clusters. It enabled the embassy team to review and refresh their engagement plans, identifying blind-spots and new opportunities for strengthening their diplomatic networks, and to think differently about who is prominent and shaping public discourse.

3. What are the main elements of a PEA?

This guide describes approaches to PEA that range from rapid, light-touch and low-cost, through to more research-based and time-intensive work. Teams will want to consider what approaches will be most appropriate in their context and for their particular issues. This can sometimes mean using a mix of tools.

This section summarises the main elements of undertaking a PEA, looking at two different approaches. First, PEAs that are undertaken as a planned, focused and in-depth activity; and second, a type of approach that uses PEA as a continuous, real-time and ongoing part of a team’s work. This section then discusses the importance of gender-informed PEA, which is critical to either type of analysis.

For practitioners and teams with a general interest in PEA, Sections 3 and 4 provide a basic overview.

For teams and other individuals who are embarking on a targeted in-depth PEA, it is essential to read the entire guide. All readers should also consider the key terms included in the Annex.

Although many tools are focused on in-depth PEA, they usually share a common thread in the way that they try to understand how things do (or don’t) work in a given context:

1. What is the issue or problem with which a team wants to engage? What do they want to understand?

2. Why are things this way? In other words, why does the problem/issue happen? Unpacking the issue by looking at the dynamic interaction between:

• foundations/structures
• formal and informal processes and rules
• actors/stakeholders

3. How can this be addressed through development activities?

• what are realistic pathways of change?
• what are technically sound and politically feasible policy or programme interventions?
• how can related work embed this analysis, including through thinking and working politically?

3.1. In-depth PEA

Identifying the question

Before starting an in-depth PEA process, a team will first need to agree on what is the top-level question or issue that would benefit from deeper analysis. It will want to set out the information gaps, and identify which areas they hope to understand better.

An in-depth political economy question will be framed to unpack the issues involved, including long-term drivers and short-term variables. It will have at its core an exploration of a set of interests, incentives, organisational cultures, rules and/or ideas, which may prevent a solution from being found to specific problems. The question should be objective and avoid embedding assumptions about why things are the way that they are.

For example, a reform-minded leader can face a ‘collective action problem’: a group of individuals or organisations could all benefit from the reform the leader is pursuing, but the task of gaining alignment of the actors is too complex, particularly where some actors prefer to stand aside and let others press for change (potentially letting them rock the boat, or use up their political capital). From an external ‘surface-level’ view, the inaction may seem like deliberate inaction (blocking change), or even that the actors gain from dysfunction (corruption).

Collective action problems can be significant issues that are masked by assumptions we make about levels of commitment. In reality, the number of processes and actors involved may lead to small obstacles becoming insurmountable in the absence of wider energy and ownership. PEA can help to identify which approach(es) will be needed to navigate these challenges (eg by brokering shared strategies).

Collective action problems are an example of why it is important not to prejudge a PEA question with assumptions about potential answers. For example, the question: ‘why is there no political will to improve business regulations’ is less helpful than either ‘why is improving business regulations proving so difficult? Or ‘what are the main barriers to business regulatory reform?’ Box 2 describes how one team used a PEA approach to help a country remove fake seeds from the supply chain.

Box 2: PEA as a driver of shared strategies

In country A, a seed-certification process which aimed to remove counterfeit seeds from the seed chain was riddled with systematic failures and inefficiencies. Very high margins could be achieved by selling grain as certified seed. It was even more lucrative to buy non-certified grain and mix it with some certified seed, while retaining the same selling price.

The development partner team supported a coalition to address the counterfeit-seeds issue, made up of seed companies, the Seed Trade Association, and the government. They wanted to understand why it was do difficult to reduce the incidence of counterfeit seed. The financial incentives were clear, but how did those abuses work in practice? And how did they persist for so long?

The coalition commissioned a PEA that looked both at (a) how the formal regulatory and market systems were meant to work (formal rules), and (b) everyday practice, the informal approaches that were actually followed. The intention was to use these insights to develop an action plan. They followed a 5-step process that mapped the power dynamics of various actors, their roles and how they operated in practice. As a result, they understood why some stakeholders had strong reasons to block change and found policy and technical solutions to overcome this collective action problem.

For example, Scratch Card Technology enabled farmers to verify and trace genuine seed, by loading the cards with information about the seed in question, and then inserting them in the seed packages that are distributed to various markets. Farmers were then able to phone a free number to verify the authenticity of the seed. This allowed feedback loops with private seed companies on how successful the crop is during the growing season.

An in-depth study can be applied to issues at multiple levels. Teams need to consider if the problem they want to address is situated at a global macro, sectoral, local, organisational or other level. It may involve considering the role of individual actors (eg we don’t understand why X won’t agree to Y), or complex constituencies.

It is crucial to define the focus of analysis. If the issue is too narrowly defined, broader influences arising from relevant national, regional and international political and economic processes may not be captured. To determine the most appropriate level of analysis it is useful to start with an overview of the country context and international dynamics (macro analysis). This may have been done as part of an organisation’s country planning process, and will help to decide how deeply to address specific sectors, problems and issues. Focusing the analysis can increase the likelihood of operationally relevant findings and insights. Box 3 provides an example of an economic sector macro-analysis, which was complemented by sector-specific studies.

Box 3: In-depth sector-level assessment

In country B, a team was undertaking a planning process to inform a new bilateral partnership and decided that a strategic overview of the context would be useful in advance of a meeting on economic growth in the country. The team wanted a deeper understanding of the economic reform process and its impact, as well as associated risks and opportunities, particularly to identify where the UK could offer support.

The PEA findings suggested that space was opening up for broader and deeper economic reform, with the potential for strong demand for partnership with donor agencies. It also outlined substantial challenges to this, such as lack of capacity for reform management, that were later realised and reported in programme reviews. The macro analysis provided an evidence base for a multi-year engagement that enabled the aid agency to be partners on reform issues. It then needed to be supplemented with other more specific PEAs on technical areas.

The main elements of in-depth analysis

The core of an in-depth PEA is to understand why a country situation, or a policy issue, is the way it is by exploring less obvious factors in a systematic manner.

This approach explores the dynamic interaction between 3 level of analysis:

1. Foundations or structures: long-term factors that do not change rapidly but are important for teams to consider in a rigorous analysis, such as history, natural endowments, demographic change, geopolitics, social mores, etc.

2. Formal and informal processes and rules: laws, norms, values, expected patterns of behaviour in society. These can explain, for example, whether citizens expect elections to be free and fair, or find it acceptable to pay bribes to access public services such health or to obtain building permission. These can differ markedly across issues, regions and organisations. While one government ministry may have a tradition of ‘job pricing’ (seniors asking for payments for transfers or promotions), the ministry next door may not.

3. Stakeholders or actors: individuals, organisations or groups that development practitioners want to understand better (because they are influential or excluded). Their behaviour, and their appetite to engage in a domestic reform or international negotiations, are often complex, shaped by a mix of beliefs, preferences, interests, and may involve a trade-off between these. Box 4 provides an example of network mapping. The analysis should be inclusive, recognising that gender issues are essential to building up a picture of local politics (see section 3.3).

This approach allows practitioners to set out factors systematically – by asking ‘why’ repeatedly – to go deeper than in routine discussions. This approach is also comparative – why is this the case here (eg this ministry)? but not there (the ministry across the street)? Comparative approaches are important to in-depth analysis to avoid over-extrapolating from one set of issues to another (an in-depth study helps us to understand X but the same conclusions may not be valid for Y).

Box 4: Finding the real counterpart

In country C, a team was working to promote economic growth. A series of economic reform measures were repeatedly vetoed as they worked their way through the political process. The team had assessed the reform proposal as technically sound and likely to improve a heavily aid-dependent economy.

The team could not determine where in the process the problem lay, or why, and therefore commissioned a local think tank to undertake a network-mapping exercise to determine who was involved and their positions on the reform issues. This process identified a powerful actor outside the formal ministerial structures, and this information enabled the team to rethink its influencing strategy. The approach the team supported became more politically feasible as a result.

In-depth PEA can be done quickly, particularly if there is a body of previous analysis on which to draw. However, the timeframes for in-depth PEA tend to be from two weeks through to several months, particularly if external support is used.

Done periodically this analysis can provide core background to the use of ongoing PEA processes, setting the scene for the political dynamics that are observed through more agile tools.

3.2. Ongoing (real-time) PEA

The second approach treats PEA as a continuous background process that informs work on an ongoing basis (i.e. in real time), rather than through a sequence of separate, in-depth assessments.

The ongoing approach uses quicker PEA ‘tools’ to set out focused questions. One example is ‘Everyday Political Analysis’ (EPA), which was developed to help busy development professionals to build an ongoing picture of the political environment in which they work. It is light-touch, agile and can help them make decisions and offer advice that improve their agency’s impact. It illustrates the approach of simple, stripped-down frameworks. Real-time approaches help to develop an overall sense of the political context and to use this understanding to make more politically-savvy decisions – essential to influencing change.

There are 2 ‘steps’ for EPA:

• step 1: understanding interests: what makes people tick?
• step 2: understanding change: what is the space and capacity for change?

For each step a series of yes/no questions helps unpack what is going on and examples of these can be found in Hudson et al, Everyday Political Analysis, Developmental Leadership Program (2016), for further information on this approach see the full EPA paper (PDF).

Box 5 illustrates how the approach was used to understand health-sector reforms.

Box 5: Using everyday political analysis to mitigate the damage of rapid health-sector reform

In 2011 the government of country D completely restructured the health sector overnight, throwing it into disarray. These changes made no sense from a technical standpoint and would have devastating impacts on health outcomes. Local development practitioners had to work out how to respond to mitigate the damage.

The team used the Everyday Political Analysis (EPA) tool to understand where the key players stood on the change and their influence on the process. It showed that the most powerful decision-makers were not those in the most senior positions. It also identified individuals who were concerned about the restructuring but ultimately had little influence.

The tool meant that the team knew with whom it had to work to influence change, and was able to convene an informal coalition of politicians, civil society organisations (CSOs) and some international organisations, which led to the reversal of the policy.

Your organisation may not have sufficient capacity to undertake PEA on an ongoing basis. Box 6 provides an example of how to obtain real-time political economy insights through a facility set up to provide on-demand advice.

Box 6: PEA as a real-time resource for action and influence

In country E, in a context that became rapidly affected by conflict on a significant scale, many development agency staff were no longer able to travel around the country or obtain regular insights from a range of stakeholders.

It was nevertheless possible to make use of a previously established PEA facility to commission rapid political analysis on developments in regions or sectors from external national (and international) researchers. The facility was also used to obtain a better understanding of social-protection mechanisms that were now operating very differently. It produced verbal briefings, written notes, options papers, workshops, and call-down advice, in response to demand.

The team was able to draw on this analysis to inform scenario planning, business planning, policy discussions and programme design. It was also used to influence other development and diplomatic actors.

As with any form of analysis, triangulation is important – this could be done through discussion with a wider range of counterparts. Try to identify someone with whom you tend to disagree, to be sure that you are not just confirming your own biases.

The EPA approach is one of several readily available and accessible tools that complement more conventional in-depth political economy studies. Additional ongoing tools include one produced by the Effective States for Inclusive Development (ESID) research programme at the University of Manchester. ESID produced a guide to scaling approaches to PEA, including a one-hour conversational tool (Making political analysis useful: Adjusting and scaling (PDF, 132 KB), ESID Briefing Paper No.12, April 2015).

Both approaches to PEA (in-depth and ongoing) reinforce the principle of ‘triangulation’ (checking our findings and assumptions – see Section 7 below), and both share an awareness that aspects of the political context can change rapidly, meaning that an analysis may soon be out of date). Some have adopted a hybrid approach that includes the regular production of planned PEA within programme agreements made with implementing agencies – so that analysis is available at regular intervals.

3.3. Gender-inclusive PEA

PEA frameworks initially focused on who had power and on deals between elites. As a result, many PEA reports were gender-blind and failed to understand how women affect – and are affected by – their societies and politics. A gender-blind PEA is a waste of time and resources, giving a partial or distorted view of the context. Recent PEA frameworks explore both political inclusion and exclusion, and call for analysis of how women shape the political context.

An important source on gender and PEA is: Politically Informed, Gender Aware Programming: Five Lessons from Practice, Derbyshire, H. et al. (2018), Developmental Leadership Program. This paper includes an illustration of how aspects of gender and politics remain hidden under the surface of more obvious dynamics within a context (the gendered politics iceberg). It points out that there may be visible aspects of gender disparity in politics (such as less than a quarter of national Parliamentarians being women), but that these visible issues sit on top of underlying factors such as social norms and organisational constraints.

An effective PEA process will therefore take steps to explore gender issues both as they appear in the formal political sphere, and also as they are experienced by women across society, including how the agency of women can achieve political impact. Effective PEAs therefore deliver inclusive, gender-sensitive analysis by adopting participative approaches and by considering:

• how do the foundational factors affect different groups? For example, are demographic trends contributing to youth exclusion from the formal labour market?
• how do the shared understandings and norms affect different groups? Do laws give equal rights to women and men? How do gender norms and patriarchal systems exclude women from decision-making? How does this work in practice?
• how do power dynamics affect different group? Can/do women achieve political impacts and how do power inequalities influence gender dynamics?
• how do the interests and influences of key actors affect other groups? For example, which minority ethnic, religious or language groups have been able to make alliances with more powerful groups? Why might this be sustained?
• what are realistic inclusive change pathways? Under which scenarios and how might excluded groups gain more influence in the sector of interest?
• how could diplomatic dialogue or development programmes support gender-inclusive politics and power structures? How could they avoid reinforcing exclusionary processes?

Gender must never be considered an optional `add-on’ to be addressed if it seems relevant after a study is underway, it must instead be included from the inception of the analysis. PEA must consider all aspects of how differential power relationships shape political economy, and impact those involved.

3.4. Deciding how to respond: operational implications

PEA should not be seen as an end-product, but as a step in a process of action. Essential actions for applied PEA are steps to ensure that the analysis is used to inform decisions, including on risk tracking. PEA should inform decisions at all levels.

There are different ways to move from analysis to operational decisions. For example, it can be useful to develop ‘pathways of change’ (i.e. how change is likely to happen, given the analysis). These can be important building blocks within an overall Theory of Change (ToC), helping to establish how change can realistically occur, and the links, sequencing and mutual dynamics that are involved. This will prevent teams from jumping straight to a solution without making the most of the analysis. Box 7 provides an example of a developed in a politically constrained context.

Box 7: A pathway to inclusive education

A pathway of change describes how a desired objective may realistically come about, given a specific starting point and taking account of the different political economy factors that have been analysed. It makes explicit the causal factors through which change could happen (eg which combination of long-term trends, new ideas, or emerging groups of actors).

Once these pathways are developed, teams can decide how to intervene to make the desired outcome more likely, or to prevent the least desirable outcome. Sometimes, it may be best to do nothing. This way of thinking prevents teams from jumping to pre-existing solutions. It ensures they take the time to consider how development interventions will affect the different stakeholders and balance of power.

For example, in closed political contexts, the room for manoeuvre may be very limited. In country F, a team designed an issues-based programme which did not seek to change the overall political system, but instead sought to find concrete solutions towards inclusion and accountability. The analysis showed that key politicians had limited interest in improving service delivery; that ministers and members of the legislature would react only when there were major public failings; and that there was space for local initiatives. This would allow entry points for civil society groups or technical experts to offer concrete suggestions, provided these did not threaten economic interests or political power. This broad pathway of change showed how inclusion initiatives could be successful.

Within this pathway, the project team identified a counterpart with effective advocacy and networking capacity. Following a scandal about the poor performance of students in a secondary school, the project provided a civil society group with communication resources. This enabled them to make their case more effectively with the Ministry of Education, and to work in coalition with education experts and other organisations. Eventually, not only did the Ministry agree to make the necessary adjustments for the students, but the initiative led to further reforms for students.

Sometimes, the team or programme may conclude that there is no realistic pathway of change, and that it is better to do nothing rather than follow an approach which might have worked in other contexts. Box 8 gives the example of constitutional reform in a country where there was no fundamental agreement on the distribution of power and resources.

Box 8: When the pathway is a dead end

A multilateral body had been supporting a formal constitutional review process in fragile and conflict-affected state G. The aim was to amend and secure ratification of important outstanding provisions of the provisional constitution. Some of the issues were non-contentious – essentially requiring grammatical edits or removing erroneous or contradictory elements. Other issues had not been addressed in the provisional constitution because they touched on core conflict fault-lines of power and revenue sharing. In this context, elite relationships appeared to be frozen, with vested interest in the status quo, meaning that any progress would threaten the frail political understanding.

The constitutional support project took a technocratic approach by supporting the government department responsible for constitutional affairs and a constitutional commission. However, as relationships between central and local-level bodies deteriorated, so too did prospects for agreeing core power and resource-sharing dimensions. The project nonetheless continued to support non-contentious constitutional elements, even if wider impacts were not being achieved.

Eventually, the project team decided to undertake a consultative and self-reflective process. This was done through challenging conversations with institutional counterparts and donors, looking at the problems through a political economy lens. They understood that, without a political agreement on how power and resources were to be shared across a wider set of political actors, there was no incentive to address the contentious clauses in the constitution. In other words, the project was not based on a realistic theory of how constitutional change would occur.

The project team then realised its technical approach was ill-suited to the politicised nature of the problem. The project worked through government and could not develop a direct relationship with local actors who were distrustful of the centre. The project team opted to wrap up its work, and recommended that before any further constitutional support, core political decisions would have to be made, including dealing with the trust deficit and collective action failures between central and local bodies.

4. Thinking and working politically (TWP)

One of the driving forces behind the introduction of PEA to FCDO was Sue Unsworth, who served as the former Department for International Development (DFID)’s Chief Governance Adviser in the early 2000s. Over a decade later, she cautioned that PEA had been under-used in guiding actions (PDF, 3.8 MB), partly due to internal processes that preferred technical solutions over those rooted in the realities of the political context.^{[footnote 2]}

TWP is an umbrella term that has come to describe the need for policy-makers and practitioners to respond to the politics of their context within their strategies, rather than simply seeing this as a passive backdrop. With TWP, politics should be viewed as a positive context for active engagement, working ‘iteratively’ and ‘adaptively’ to influence those with power. Iterative means trying an approach, testing whether it works, and adjusting as we learn. Flexible and adaptive programmes have the capacity to change as the context evolves or when there is new evidence.

The ability to think and work politically may not necessarily remove obstacles to progressive change, but it does provide a compass to help navigate the political complexities faced by development actors. Many TWP programmes have identified issues or coalitions with enough traction to affect change. Box 9 provides a successful example.

Box 9: TWP: analysis into action through coalitions

An accountability programme in country H supported civic action on contentious issues where there was potential for positive reform. Combining PEA with technical expertise, it identified climate finance, climate-induced migration, fair labour migration, and food safety as policy areas where progress could be made. Adopting a TWP approach, the programme convened ‘change coalitions’ of CSOs, the private sector, and activists around each of these issues to develop solutions with policy-makers in a collaborative manner. A review of the lessons and achievements identified 4 key components of success:

• identifying a tractable issue where there was potential to bring together enough ‘winners’ from change to outweigh resistance from those who potentially lose out
• aligning interests and incentives of actors within the change coalition, and ensuring it included a broad range of stakeholders in the system
• acting as a neutral broker to facilitate alliances, balance interests, cross-pollinate ideas, and build a shared understanding within the coalition
• sponsoring insider-reformers within the state institutions (or ‘intrapreneurs’) with the power and motivation to push for results

The results of the programme demonstrate the effectiveness of TWP. More than 100 community social audits of climate finance projects such as embankments and cyclone shelters ensured that repairs were of sufficiently high quality, and led to a government commitment to engage civil society and community groups in the oversight of climate adaptation projects across the country. By working with local coalitions, including Consumer Committees in target districts, the programme contributed to the government’s introduction of a new certification system to incentivise poultry farmers to comply with safe food practices.

Development teams have a range of options to inform an iterative approach that lends itself to TWP ways of working. These include:

• using Everyday Political Analysis (see section 3.2) as a standard part of their work, adding to the insights that can inform TWP approaches
• building ongoing PEA capacity into programmes
• supporting external bodies to provide regular analysis
• undertaking more in-depth PEA studies outside the normal project cycle when the country context has changed, or new UK (or other donor country) objectives require a re-assessment of strategies

PEA can be characterised as the thinking element of TWP, while a second, and equally important, element is to use the results of PEA to work differently. So, TWP embeds PEA in programme implementation and policy influencing. Regular PEA helps test whether our assumptions about how change happens remain valid, and if not to adapt and change course accordingly.

To do this effectively, programmes need to work simultaneously on 7 functions:

1. Decision-making on approaches, including the willingness to take risks, reflecting the understanding of context and the Theory of Change.
2. Context monitoring: the regular assessment of the context (such as expected or unexpected events, inclusion–exclusion dynamics, elections, changes in programme partners), checking whether the issues on which we are focusing respond to local priorities and help to harness internally driven and locally led pressures for change, and to watch for unintended consequences. (The ‘because’ in the programme’s results framework: do the assumptions justifying our strategy and programme hold or do they need to be adjusted?)
3. Implementation: the day-to-day, week-to-week task of undertaking activities. (How are we doing on physical progress?)
4. Monitoring progress: the task of determining the kinds of change that we want to achieve, starting from the understanding that change is complex and non-linear, and that there are invariably risks and uncertainties involved. (Is the plan on track as intended? If not what adjustments are needed? How can these flexible and adaptive ways of working best be supported?)
5. Learning: our internal and reflexive questioning of progress refers first to what are we learning in terms of translating inputs and activities into outputs and outcomes (What is working and what isn’t?). And second, reflecting on how and why the programme does what it does, including testing assumptions and hypotheses about how change happens, as well as overall strategy and direction (why and how is the programme working or not working?)
6. Adapting: revising our implementation plan, adding unforeseen activities and dropping others, changing the balance of inputs, be they cash, people or events, reassessing and adjusting theories of change and related interventions, and having the space, support and mandate to do so. (How are we changing the plan and why?)
7. Influencing: thinking about both financing and engagement, teams can shape the impact of their programmes through their own engagement with counterparts on the context issues. (How are teams assisting programme objectives through their own influencing?)

This is organisationally challenging for those responsible for implementing development programmes because grants or contracts may demand predefined outputs and outcomes. Programme management systems often prefer predictability based on planning to achieve results 3 to 5 years in advance. However in the complex world in which development organisations work, ultimate objectives may not change, but the path towards them is almost never a simple process of going from A to B (PDF, 875 KB) ^{[footnote 3]}

Adaptive management inherently involves openness to learning, feedback and evidence – even when this is challenging. Box 10 lists studies which have considered experience on adaptive programming.

Box 10: Iterative/adaptive programming

Pett, J. (2020) Navigating adaptive approaches for development programmes: A guide for the uncertain. Working paper 589 (PDF, 232 KB). London: ODI.

Laws, E., Pett, J., Proud, E. and Rocha Menocal, A. (2021) LearnAdapt: a synthesis of our work on adaptive programming with DFID/FCDO (2017–2020). Briefing Note (PDF, 273 KB). London: ODI. See also LearnAdapt: lessons from 3 years of adaptive management, ODI: Think change.

Human Learning Systems, Adaptive Management Practice in the UK Government, Case Study (PDF, 534 KB).

Sharp, S., Riemenschneider, N. and Selvester, K. (2022) Evidence-led adaptive programming: Lessons from MUVA. Working paper. London: ODI. https://odi.org/en/publications/evidence-led-adaptive-programming-lessons-from-muva/.

The Asia Foundation, Strategy Testing: An Innovative Approach to Monitoring Highly Flexible Aid Programs, Working Politically in Practice Series.

Derbyshire, H. and Donovan, E. (2016) Adaptive programming in practice: shared lessons from the LASER and SAVI programmes. Case Study (PDF, 410 KB).

Development donors should incentivise those responsible for programme implementation to keep their political economy knowledge up to date, and to use it in practice. This can range from regular weekly check-ins during team meetings to more structured quarterly or annual workshops. These approaches can have greater impacts where programmes are designed in a less prescriptive way so they can adjust their activities and objectives. Box 11 illustrates how implementers can be encouraged to work in TWP ways.

Box 11: Twenty years of thinking and working politically

FCDO has supported governance reforms at the state level in one country for 20 years. In-depth research into what and how this work influenced governance, and also improvements in the health and education sectors, concluded that the ways in which 3 generations of programmes operated mattered significantly.

• a long-term commitment to the state partnerships which created political-level trust
• decentralised decision-making to state teams, giving implementing organisations enough space to identify priority issues and test approaches
• recruiting national staff who had both political and technical skills and had built trust with their counterparts in government, political and civic spheres. They became brokers in reform processes
• investing in understanding what motivates State Governors to support reforms (such as different political incentives, or generating healthy competition between states)
• supporting not only state-level interventions but also coalitions with civil society, traditional leaders, women’s groups, parliamentarians and media around specific reforms
• embedding PEA within the programme, by undertaking regular participatory PEAs at the state level
• Building in programme flexibility to reallocate resources, and using PEAs to propose new entry points, adjust, or end activities
• quarterly learning sessions to identify what works and share knowledge

TWP is made easier where programmes and teams have access to regularly updated PEAs. Several of the examples in this guide (eg fake seeds box and parliamentary interest featured in Box 15) are from programmes where permanent PEA capacity was built into activities; PEAs became ‘embedded’ as the programme team itself was able to undertake and use the analysis.

In other countries, civil society and democracy programmes have supported new or existing think tanks to undertake similar work (not least the original Enabling State Programme in Nepal which contributed to this new approach)^{[footnote 4]}.

Some programmes have even more ambitious objectives, such as to build the capacity of civil society and other local actors to analyse their contexts and effect change themselves. This can be supported through tailored training for target groups, as illustrated in Box 12. Such PEA approaches are often more inclusive and participatory than traditional PEAs.

Box 12: Building sustained local PEA civil society capacities

In country J with a closed political system, but where big political change was expected to come eventually, a team decided to support a civil society programme aimed at the younger generation of potential future leaders. Local academics delivered training in political economy (called ‘contextual analysis’ to be more circumspect). They adapted the content of the course to their country (eg using the official language, not English, placing emphasis on shared norms and values, and reflecting critically on the strengths and challenges of the western frameworks often used for PEAs).

The course sought to help young members of CSOs to move away from conventional ‘business as usual’ projects to the design of interventions that tried to change the way citizens and government engaged – requiring more robust analysis of the ‘the art of the possible’ and the most promising entry points.

Eighteen months after the course ended, the political transition came, and these young activists used their skills in unexpected ways – ranging from identifying the most efficient and safest locations to protest to holding the transitional government to account. They described the training as transformational, as these quotes show:

PEA helps people identify the target for change as well as identifying the possible routes – and it emphasises the importance of patience and trial and error required to get there.

The course was an important point in my life. It gave me the tools to see beyond the news and realise opportunities in seemingly desperate situations. I am using the context analysis tools to reflect with the people in my own town in different gatherings the reality of the current situation – helping to identify different scenarios.

The overall experience of knowing about and using different tools of analysis enabled me to understand the different power dynamic behind every situation and piece of news. It also triggered my passion for analysis.

(she is now enrolled to do post-graduate studies in economy and political science.)

5. Planning and undertaking an in-depth PEA process

In-depth studies often require a greater investment of funds and time, and are often linked to decision points or processes. This section will be particularly useful to those who must now undertake a planned, in-depth PEA, whether directly, or through third parties.

5.1. Clarity on the purpose and audience

Section 3 outlined the importance of identifying a clear purpose for the PEA, setting out the questions to be addressed.

When teams are planning an in-depth study, they will want to consider who are the decision-makers or partners that the organisation seeks to influence through a PEA. You should identify which stakeholders should act on PEA findings, and would benefit from being involved in the process from the beginning.

Some PEAs will remain internally focused, perhaps because they need to be undertaken quickly to influence internal processes, or contain information that is sensitive. In all forms/approaches to PEA it is essential to consider risks that may be involved for those undertaking the analysis and to take appropriate security measures.

Where possible, however, it is good practice to share the findings with those who are consulted during the PEA.

Participatory PEAs may be appropriate where the objective is to build a common understanding with external partners and/or jointly design strategies and programmes. For example, teams can organise workshops with their main partners and independent contributors to explore PEA questions. PEA produced by development programmes might be published as a contribution to more open and transparent local political discussions.

5.2. Preparing terms of reference

Following the decision to undertake an in-depth PEA, the next step is to prepare clear and realistic Terms of Reference (ToRs) – even if the exercise is only internal and does not involve any external experts.

The ToRs should set out the following:

• purpose of the PEA (the main issue/problem, and how teams will use the findings);
• the main questions to be explored – what does the team most want to understand?
• background, summarising what is already known about the issues and why they matter (based on available evidence, including previous PEAs)
• the level of analysis and the specific (how broad a picture is needed; what are the specific questions to inform action)
• methodology (a comprehensive study with desk and field research, a workshop, an everyday analysis, etc). Specify expectations on inclusion, such as ensuring that the PEA takes full account of the gender dimensions of every aspect covered (see section 3.3)
• team composition (a diverse team with the skills to fulfil the ToRs)
• risk-management and research ethics, including conflict sensitivity and consideration of safeguarding (Sexual Exploitation, Abuse and Harassment (SEAH) issues
• outputs (which do not need to be limited to a report, but could include workshops and slide deck for more accessible findings)
• budget and timeframe, which should be proportionate – if the scope is too broad and the team under-resourced, the result will be disappointing

5.3. The team

Any type of PEA can be undertaken in-house, be fully outsourced, or can take a hybrid approach with a blend of internal/external team members.

While teams have the best insights on how to influence their own organisational strategies or design programmes, they may not always have the time, networks or range of local expertise needed to complete the study. Equally PEA can be built into ongoing programmes/activities, with implementers required to produce regular analysis.

Teams often use ongoing approaches, such as the Everyday Political Analysis framework, without external support. Sometimes, teams will benefit from external facilitation, to draw out their own insights, ensure everyone contributes and responds to expert challenges, such as during a scenario-planning process.

Where a PEA team is formed for an in-depth, one-off study, it can similarly be entirely in-house, external or hybrid. Any PEA team will benefit from having the following characteristics, where possible:

• a multi-disciplinary team, with political, economic and social expertise to analyse a wide range of political economy factors
• cross-cutting skills, such as, gender and inclusion, are essential. Other relevant expertise, such as working in conflict-affected contexts, will add value. The development of the PEA approach should actively consider how inclusion is embedded at every stage. The team should itself reflect an awareness of diversity (gender, language, regions, etc).
• sector/technical knowledge depending on the PEA scope
• country/local knowledge. Where outside experts are being used it is best to avoid establishing an entirely international team, as the objective of a PEA is to get ‘under the surface’, which depends on having team-members with nuanced understanding of different factors, and the networks to set up interviews or obtain insights from a wide range of perspectives. Ensuring that the team has access to local voices is essential to good PEA
• understanding of the organisation that has commissioned the PEA. It is useful to have on the team someone who is familiar with the ultimate user of the analysis, how decisions are made, and even the presentational style
• a team leader who has PEA experience and can also facilitate processes and communicate with different stakeholders. This is critical when the PEA results in unexpected or challenging findings
• the team leader should be aware of and manage power dynamics within the team and with those who have commissioned the work. If national experts or younger/female team members cannot contribute fully, or if some stakeholders have already decided what the PEA findings should be, the results will be less than optimal

5.4. Ensuring quality

The person responsible for commissioning and managing the PEA should ensure that the ToRs are properly prepared and followed so that study can contribute to useful impacts.

A solid PEA needs a rigorous design that sets out the analytical framework: agreement on the main questions and sub-questions, with a clear plan for how data will be collected and analysed, and implications for future work.

If your team is producing or commissioning a PEA, you need to assess whether the analytical framework and methodology are suitable, and within both budget and timeframe.

PEA processes generally use qualitative research to examine the why and how of different issues (for example the passage of new legislation on transparency or education), not just the what, where, when, or who. Qualitative research is also used to uncover trends in opinions, networks, and other patterns which evolve over time.

The need to ‘triangulate’ emerging findings dictates that qualitative data (such as the opinions of observers) need to be checked against other respected sources of information to avoid bias, perhaps documents or interviews with a wider range of informants. For example, some actors may believe that lack of leadership commitment is behind the inability to deliver X, although other data suggest that there are neither the financial resources nor technical expertise to get this done.

Research methods that are appropriate for PEAs include:

• a desk review to ensure that the team have a sense of the existing evidence and areas of consensus. It is important to bear in mind that the views of marginalised communities, sensitive political issues, and research in local languages may not be available in academic or grey material
• local engagement (also called ‘field work’) – it is beneficial to visit a variety of locations, spending time outside the capital, to hear a wider range of views
• key informant interviews (KIIs) and focus group discussions (FGDs) are the most common qualitative data-gathering tools. Informal, off-the-record conversations can be useful (eg with journalists or politicians) but need to be triangulated (while also respecting the informants’ request for confidentiality)
• quantitative data and surveys may add value, and can make it easier to share findings externally
• some software, such as network mapping, can help visualise findings
• research ethics should be followed, including the need to ensure informed prior consent, offer confidentiality, and being aware of the power relations between external researchers and those being interviewed. It may be more appropriate for a female team member to interview women, for example

In contexts affected by conflict and fragility, it is particularly important to consider how research methods may affect potential risks to those involved – either within the team or external stakeholders, interviewees, participants in group discussions and so on.

Box 13: Making an in-depth sector-level assessment

A macro-level PEA was commissioned in country K to be undertaken by 2 experts, one national and one foreign. The process consisted of a focused, but in-depth, literature review; a set of more than 40 confidential interviews with key stakeholders; and a synthesis of findings and recommendations. These covered prospects for economic reform as well as other sectors for development support such as social protection and education – though the latter was supplemented by a subsequent sectoral PEA.

Building on an initial literature review and remote engagement with some central stakeholders, including colleagues across the organisation, the exercise focused on a 3-week mission for face-to-face meetings with stakeholders in government and civil society, and with cooperating partner and multilateral agencies. The research team triangulated the views and perspectives offered in these meetings against the preliminary desk-based research, additional data collection, and the research team’s appreciation of the field.

In addition to regular exchange and discussions with colleagues the research team also validated findings through 3 workshops with the responsible team and relevant sector advisors, as well as selected external stakeholders.

The analysis outlined the broader context of economic reform and how the political economy is changing, including the institutional landscape and individual interest structures as a means to identify feasible pathways to change, and to understand the potential impact of reform. It analysed the politics of reform, set out in 4 specific technical areas requiring more detailed research, to understand the technical and political feasibility of reform, based on an analysis of the balance of power and interest structures. Lastly, it provided insights into developing and providing specific opportunities for development partnership.

PEAs have also been done remotely, with some national experts undertaking in-person interviews in the country and other team members joining remotely or relying on notes. This can be less than ideal, but well facilitated online workshops have enabled good enough discussion. Research quality also improves with mixed methods and triangulation (see above).

Reflexivity is important in order to manage potential bias. This means being aware of one’s own power and biases or prejudices. It can be a useful exercise for teams to hear from outsiders about how they are perceived in a country or sector, as it can be very different from the team’s assumptions. It is also important to be conscious of how others perceive the PEA team, or the organisation that has commissioned its work, since this will potentially influence interviews and messaging.

Workshops/webinars are useful for the team to reach emerging findings internally. They can also be used for validation with external partners, and for drawing implications. Where possible, decision-makers and other users of the PEA findings should be given the opportunity to engage with the PEA process, to give it legitimacy.

Dissemination of findings. Where possible (see 5.1) it is good to share the findings of any study involving human subjects with those involved in the research or affected by the study. ToRs should set out how findings will be disseminated. Producing a redacted version of the report is one option.

Even if you are using ‘light-touch’ or everyday political analysis tools the principles of triangulation and reflexivity remain critical, otherwise findings will reflect bias and may simply entrench pre-existing views. For example:

• if you are undertaking the analysis on your own, take the time to speak with those you often disagree with, or read newspapers/documents from respected sources you rely less on. It will give you a fuller picture
• if you only have time to organise one workshop, make sure external experts from a range of perspectives are invited, including from diverse social backgrounds (eg paying attention to gender, ethnicity, age – see Section 3)
• if you want your programme or policy team to undertake light-touch PEA on a regular basis, agree a calendar in advance, allocate staff time and develop tools that suit your team and can up updated quickly (eg stakeholder matrix, events or evidence logs)

PEA processes are designed to fit a particular set of questions and needs, the approach will differ according to context, budget and objectives. In applying this guide and developing a PEA approach it is useful to also be aware of wider work, tools and approaches. It is recommended that teams consider the resources listed in Box 14.

Box 14: Drawing on different PEA tools and guides

Browne, E. Gender in Political Economy Analysis. hdq1071.pdf (gsdrc.org) (PDF, 720 KB).

Fritz, V., Levy, B. and Ort, R. (2014) Problem-Driven Political Economy Analysis: The World Bank’s Experience. Washington, DC: World Bank.

Green, D. Putting gender into political economy analysis: why it matters and how to do it. https://oxfamapps.org/fp2p/putting-gender-into-political-economy-analysis-why-it-matters-and-how-to-do-it/.

Whaites, A. (2017) The Beginner’s Guide to Political Economy Analysis. London: UK National School of Government International, London.

DFAT (2022) ‘Political economy analysis and adaptive management guidance note’.

Haines, R. and O’Neil, T. Putting gender in political economy analysis: why it matters and how to do it. https://gadnetwork.org/gadn-resources/2018/5/9/putting-gender-in-political-economy-analysis-why-it-matters-and-how-to-do-it.

Rocha Menocal, A., Cassidy, M, Swift, S., Jacobstein, D., Rothblum, C., and Tservil, I. (2018) Thinking and Working Politically through Applied Political Economy: A Guide for Practitioners. Washington, DC: USAID.

DEVCO (2011) Using Political Economy Analysis to Improve EU Development Effectiveness (PDF, 668 KB). Concept Paper. Brussels: European Commission.

Hudson, D, Marquette, H., and Waldock, S. (2016) Everyday Political Analysis. Birmingham: Developmental Leadership Program.

OECD (2015) A Governance Practitioners Handbook (Section 1 – politics).

Unsworth, S. (2008) ‘Framework for Strategic Governance and Corruption Analysis (SGACA): Designing Strategic Responses Towards Good Governance’, prepared for the Netherlands Ministry of Foreign Affairs: Netherlands Institute for International Relations (Clingendael).

There are many other tools, such as sector-specific framework and those that combine PEA with conflict or gender. This online PEA library is regularly updated with new PEA/TWP tools, reports and case studies.

5.5. Are we informed users? Doing our homework on which tools and approaches to use

The wide variety of PEA tools (Box 14 lists some of these) makes it difficult to generalise about specific approaches. Different tools will emphasise different aspects of the 3 elements within in-depth PEA (these were: foundations, shared understandings, and stakeholders, section 3.1). When developing a PEA, a team should consider whether one particular approach is more likely to fulfil the objectives, either by reviewing materials on the tools or by consulting a PEA specialist.

An example of the variations in the tools would be Network Mapping (also known as Political-Stakeholder Analysis). This is one of the easiest tools to map actors’ interests regarding a policy area and their influence over the desired change process. This tool can therefore provide a good understanding of the second and third areas of interest (shared understandings and stakeholders), but only very limited insights into the first (foundations). Box 15 provides an example.

Box 15: Network mapping – politics as a people thing

In country L, the parliament plays a powerful role in political life. It has been involved in contestation over where authority really lies, and also on how to manage economic reform. A team wanted to support economic growth, while recognising that laws and regulations are not always driven by simple economic logic. The team built into its new economic reform project an ongoing PEA element that would see the implementor develop a capability to map economic issues onto political incentives and interests.

One of the tools developed was a network map that captured the economic interests of stakeholders. The network mapping tool allowed publicly available information to be entered for stakeholders including politicians. This could be used to identify potential blockages for reform measures. The tool could look at stakeholders depending on their investments and to see common interests.

Following an election process, the team was able to use this capability to map where stakeholders stood on issues such as political reform. And the same ‘network’ principles helped to produce forecasts of where different groups and organisations would position themselves on key reform issues.

PEA is also a lens that can be combined with other analytical tools, for example gender and social inclusion analysis, institutional, or conflict analysis. These different analytical tools are usually complementary and can benefit from specific political economy questions, for example about the nature of the state and the interaction between politics and economics.

5.6. Are we bringing the elite bargain (or political settlement) into focus?

In recent years 2 ‘sibling’ concepts have been used to help understand how power is managed. Among development agencies, the use of ‘Political Settlement’ thinking gained ground from 2008 onwards^{[footnote 5]}; and from 2017, the World Bank’s idea of ‘Elite Bargains’ ^{[footnote 6]} has been useful in focusing on how dysfunctional elite relationships might arise.

Elite Bargain thinking provides helpful background in considering whether a ‘development bargain’ is feasible in a context, and how it can be supported. The concept of a development bargain has an important tradition in academic and policy-oriented research, including work funded by the UK^{[footnote 7]}. As recently outlined by Stefan Dercon, it refers to an agreement among elites to pursue a shared vision or plan for development (see ‘core argument’^{[footnote 8]}. While both in-depth and ongoing approaches to PEA can help to bring an elite bargain into focus this task is often a core element in the ToRs for an in-depth study.

Elite Bargains,^{[footnote 9]} according to the World Bank, are the processes through which elite actors and their organisations coordinate to determine outcomes. Elite bargains are dynamic, constantly adapting to changes in each actor’s relative power, incentives, and preferences. The development path is bumpy: shocks (eg in terms of trade) and gradual developments (such as urbanisation) can alter elite power and preferences, often benefiting one group at the expense of another. In the face of these changes, bargains that cannot accommodate new actors or demands may collapse. At other times, elite bargains successfully adapt to changes by accommodating new demands through credible rules for elite–elite and elite–citizen interactions.

During policy bargaining processes, the unequal distribution of power – power asymmetry – can influence policy effectiveness. Power asymmetry is not necessarily harmful, but its negative manifestations can be reflected in capture and exclusion. Policies can be blocked, captured, or rendered ineffective when their design does not account for asymmetries in bargaining power.

State capacity is largely a function of power; ruling elites invest in the capacity of governing structures when incentives make it sensible to do so – and they neglect those investments when they do not. Such investments tend to improve institutional functions, but increasing the strength of bureaucratic actors is risky, creating the possibility of institutional champions that can contradict ruling elite preferences (see trade-offs below). Moreover, bureaucracies often serve the purposes of patronage and rent distribution; undermining these arrangements is politically challenging and can destabilise elite bargains.

One avenue for reform is to encourage greater contestability within the elite bargain – including through greater social and political accountability. Understanding how to promote both an accountable elite bargain, and effective institutions, is often the starting point for a macro PEA exercise.

This text is largely adapted from: 2017 World Development Report: Governance and the Law, World Bank

Annex: Key political economy concepts

This Annex lists some of the concepts often used during PEA processes. They include key issues that teams should consider as they plan their own PEA approach and draft their Terms of Reference.

Term	Description
Agency	The term used for the ability of individuals to affect the world around them. Political agency is the capacity for actions that can affect the political sphere. The many variables that arise from an individual’s peculiar personality traits (knowledge, judgement and instincts) are all bound up in their agency.
Confirmation bias	The tendency to gravitate towards evidence and views that reinforce our own reading of a situation or problem, giving less credence to those elements that might challenge this. This helps to explain the importance of triangulation, and of challenging all the assumptions and perceptions that are offered. This can be particularly true where we have default explanations to apply to problems (eg ‘the cause must be corruption’)
Elites	When the World Bank produced its 2017 World Development Report on governance it suggested that ‘What distinguishes elites from citizens is elites’ ability to directly influence the design and implementation of a certain policy. Elites can vary from one policy to another’ [footnote 10]. Elites are those with a privileged voice on how power is managed, whether this voice helps or hinders political leadership. Elites go beyond those in formal government roles and rarely exist in isolation, but rather rely on constituencies, positions, or resources to secure their profile and role. Elites usually need to be understood both in the context of their relationship with each other and in terms of their links to social groups and constituencies.
Identity-based politics	In any society individuals possess a complex mix of identities which have varying degrees of influence on their behaviour. One person’s approach may be strongly shaped by affinity with the traditional regional group of their parents/family, another’s only weakly. In any context there can be appeals to political loyalties based on identity, and identities can also affect levels of personal safety/risk anywhere. In fragmented societies, building a political constituency based on identity can be the political norm, and it may be easier to communicate messages through shared reference points that draw on a sense of social community. Identity can also be the most significant factor in how an individual is positioned in relation to reciprocal politics (is a particular group in or out of favour for patronage?)
Incentives	The factors that motivate people to act in a certain way go well beyond financial rewards, or personal interests. For example, a strongly held belief might lead somebody to promote action that would harm their own financial position. Incentives therefore take multiple forms (as can disincentives) and where people have conflicting incentives, they will often ‘trade off’ the gains/losses based on a prioritisation that may not be immediately obvious to others.
Issue-based politics	This is often seen as the antithesis of reciprocal and identity-based politics, with an approach focused on common visions, and willingness to join together around ideas. In reality, politics are never purely ‘issue-based’ or reciprocal’. There will be a mix of factors that relate to benefit for social groups, appeals to identity, and ideas/beliefs. For individuals this mix will differ greatly: some will be driven mostly by ideas, others perhaps more based on a sense of which group traditionally represents their region or set of interests.
Norms/rules and values	People operate according to numerous sets of rules, written and unwritten. These include social norms (what constitutes acceptable behaviour). Traditionally, a distinction has been drawn between ‘informal’ and ‘formal’ rules. This distinction may be meaningless for counterparts for whom nearly all rules can be ‘formal’, because they are well understood and generally accepted. The greater issue is what rules are in operation, how are these understood, and what systems exists to enforce, mediate, and navigate them. Norms may therefore differ based on location, gender, age, social background and a host of other factors that will be readily understood within organisations, but often opaque to outsiders.
Political space	A World Bank study of leaders who led reform processes found that they need 3 ‘As’ to get things done (Acceptance, Authority, Ability), but often leaders struggle to secure these in a consistent way. Others suggest that leaders may not have a ‘reform horizon’, do not know what reform is possible or how it has been achieved elsewhere. These challenges can be cited as a lack of political space – the room for change. Operating on the basis of the familiar (default tactics – such as patronage) may seem most practical.
Reciprocal politics	A number of terms refer to systems where political loyalty is rewarded by elites, perhaps through public-sector employment, preferential treatment for public services, or even direct payments. The terms ‘clientelism’, ‘patronage’, ‘neo-patrimonialism’, ‘pork-barrel politics’, and the ‘political marketplace’, all refer to aspects of these issues, including the negotiating process that can take place. Ultimately, they all point to politics that entails tangible benefits for supporters and constituencies as a ‘reward’ for being within a leader’s political constituency/support-base.
Rent-seeking	Rents are the benefits that accrue from the control of resources. Rent-seeking usually means pursuing those benefits by virtue of power (eg corruption), rather than productivity. Different political theories are focused on rent-seeking. For PEA, rent-seeking is both a useful concept in understanding power, and also a potential risk as a catch-all answer. Equally, rent-seeking can vary greatly across systems and groups, and good PEAs avoid generalisations.
Social networks	The relationships among individuals shape perceptions and decisions, and the nature of personal connections differ markedly based on cultural factors. Relational bonds can be forged (or broken) based not only on family or friendship ties, but also shared experience and interests. Common reference points can benefit some people as they seek to relate to those in positions of power, in some contexts, being in the same cohort for university or the military is particularly important. These factors go beyond providing a form of loyalty between individuals, they can also make it easier to influence through insider knowledge of what might shape the ‘agency’ of key actors, and can marginalise those without access to these networks.
Trade-offs	PEAs often point to dilemmas regarding ‘trade-offs’; the losses or gains involved in a particular course of action. For example, improving the professionalism of the military will improve security, but doing so might involve selecting leaders on merit, rather than on personal loyalty. The leader is faced with trading the gains of a more competent national defence, against the potential loss of their own hold on an important source of power. Most issues involve multiple trade-offs at different levels and the complexity of potential unintended consequences. For those on the outside, the trade-offs involved in any given situation are often hard to see, and it can be tempting to boil down such issues and to portray them as simple linear choices.

Avoiding conceptual pitfalls

Political analysis has been influenced by ideas originating with 2 schools of thought: ‘structural-functionalism’ and ‘new institutional economics’. Models and technical terms are derived from these origins and have been used to explain contexts (eg Limited Access Orders). While models can help to illustrate thinking, it is sometimes tempting to use these as doctrinaire templates. Theory should be used primarily as a stimulus to ideas, questions and discussion, (eg elite bargain ideas can help in analysing information) but not to be taken as encapsulating the whole picture. Concepts and terminology can also best be used to provide a shared language for discussion of real-world observations.

1. The lead author was Alan Whaites, with contributions from Laure-Hélène Piron, Alina Rocha Menocal and Graham Teskey from the Thinking and Working Politically Community of Practice (TWP CoP). Published by the FCDO Governance Cadre and the TWP Community of Practice. The authors are grateful for feedback from Jonathan Rose (USAID) and Verena Fritz (World Bank) and numerous FCDO staff. The content does not represent UK Government policy. The views expressed remain those of the authors alone. ↩

2. Unsworth, S. (2015) ‘It’s the politics! Can donors rise to the challenge?’, in Whaites, A. (ed.) A Governance Practitioner’s Notebook. Paris: OECD. ↩

3. Rachel Kleinfeld, Improving Development Design and Evaluation, Carnegie 2015 ↩

4. Unsworth, S. and Booth, D. (2014) ‘Politically Smart and Locally Led Development’. London: Overseas Development Institute. ↩

5. Behuria, P., Buur, L., & Gray, H. (2017). Studying political settlements in Africa. African Affairs, 116(464), 508–525. ↩

6. World Bank. (2017). World Development Report 2017: Governance and the law. World Bank. ↩

7. See for example Booth, D. (2012). Development as a collective action problem. Africa Power and Politics Programme Synthesis Report. ODI; Kelsall, K., N. Schulz, W.D. Ferguson, M. vom Hau, S. Hickey, and B. Levy (2022) Political Settlements and Development: Theory, Evidence, Implications. Oxford: Oxford University Press; Meehan, P. (2018). ‘What are the key factors that affect the securing and sustaining of an initial deal to reduce levels of armed conflict?’. Literature Review for the Elite Bargains and Political Deals Project. Stabilisation Unit.   DFID; Rocha Menocal, A. (2022) Why incorporating organised crime into analysis of elite bargains and political settlement matters: Understanding prospects for more peaceful, open and inclusive politics, SOC ACE Research Paper No. 15. Birmingham, UK: University of Birmingham. ↩

8. Interesting discussions are available at: https://www.youtube.com/watch?v=zUf7dirW7NY and https://www.youtube.com/watch?v=15TTnED0HDs ↩

9. Summary adapted from World Bank (2017) Development Report 2017: Governance and the Law. Washington, DC: World Bank. ↩

10. World Development Report 2017, p. 20. ↩