Our Programmes



Sign up to our newsletter.

Follow ODI

Putting politics into practice? Political economy analysis and the practice of development

Time (GMT +01) 16:30 18:30

Alina Rocha Menocal - Research Fellow, Politics and Governance, ODI
Stefan Kossoff - Political Economy Analysis and Governance Lead, DfID

Alex Duncan - Director, The Policy Practice

Closing remarks:
Julian Lob-Levyt - Senior Vice President and DAI Europe Managing Director

Andrew Norton - Director of Research, ODI

This event was held by ODI and DAI to launch a special issue of DAI’s Developing Alternatives journal, devoted to political-economy analysis, to which ODI contributed three articles.

Alina Rocha-Menocal’s presentation focused on:

·         Where we have come from

·         Where we are at

·         Where we need to head

Where we have come from

Donors have been grappling with how to take politics more seriously over the last decade. From this process, two realities have emerged:

1.       Politics matter

2.       Increasingly, there is recognition that the good governance agenda is not good enough as an answer to development challenges

Political economy analysis (PEA) has been an important tool in helping crystallise this. PEA is now increasingly used by a variety of actors/institutions in the international development community, including not only bilaterals (DFID, the Netherlands, etc), but also multilaterals (World Bank, the UN, the EC), and even by some in the NGO community. However there remain outstanding questions about how political economy can make a real difference.

Where we are at

Alina points to five key areas in which political economy analysis can be seen to be making a difference:

1.       PEA offers a different perspective that challenges pre-conceived notions (often normatively based) and provides a pragmatic approach (for example, by acknowledging the importance of non-state actors alongside the formal state).

2.       PEA provides an opportunity to inform more realistic policy making which can assist policies in adapting to local realities. For example, the Dutch government’s Governance and Corruption Assessments was useful in demonstrating the importance of moving from general to sector budget support in Uganda.

3.       PEA is helpful in crystallising country-specific knowledge, going beyond empty claims that ‘context matters’. Rather, PEA brings a lot of information together in one framework that can allow for a better dialogue between governments and development partners based on shared understandings of local realities.

4.       PEA can help to identify a broader range of stakeholders that may be critical to engage with but are not always obvious at first glance or through conventional analysis.

5.       PEA is critical in identifying opportunities for reform entry points by highlighting blockages to reform. This in turn can help donors to facilitate processes of change.

Where we need to head

There is a need to understand that PEA is a tool, not a magic bullet. It provides an angle to address challenges but does not solve them itself.

PEA research conducted by ODI and others, such as The Policy Practice, has found that PEA tends to be most effective, useful and amenable to practical follow-up and uptake when it is problem-focused, i.e. targeted to a particular unresolved issue in development (level as such can be macro or more micro/sectoral).

For PEA to be most useful, it must be seen as part of a multi-stage process that includes more than the production of a PEA itself. The process needs to start with clearly defining the operational purpose/objective of analysis, and there also needs to be important follow up, in terms of uptake and evaluation, once a study has been produced.

It is also important to include donor country office staff in the process from the start to tap their knowledge, ensure the usefulness of the analysis being carried out, and to promote buy-in among those who will need to translate insights into practical operations. This also points to the kinds of skills that donor staff need to have in order to fully engage with the PEA agenda and find it useful in their work. This is especially true if donors are to play a meaningful role as facilitators of change processes, rather than simply as providers of funding. ODI (in conjunction with The Policy Practice, among others), has been doing a lot of work over the past few years in providing PE training to development institutions/agencies, and there has been growing demand for it.

Alina also highlighted that it may be worth capitalising more fully on the opportunities that a PEA approach can offer at a time when donors are pressed to show results and value for money. PEA can help build more realistic theories of change focus on incremental change. Lastly, Alina also pointed out that it will be important for donors to educate their home publics about the realities of development as a political process, in order to create more realistic expectations about how quickly results can be delivered and the very real challenges that are often faced in promoting change.

Stefan Kossof welcomed the latest issue of DAI’s Developing Alternatives as a valuable addition to the PEA literature. Importantly, Stefan pointed out that PEA is not just a dismal science of pointing to the impossible, but that the case studies set out in Development Alternatives highlight how PEA can be used to understanding a variety of issues in different and innovative ways.

DFID’s story of PEA

DFID has been investing in PEA for around 10 years. Stefan’s presentation covered:

1.       What drove DFID’s interest in PEA

2.       How has this interest evolved

3.       What has been learned and what are the outstanding challenges

What drove DFID’s interest in PEA?

This began with a recognition that technical programmes were not delivering results across a range of sectors and that there was a need to understand why and how things change. This recognition occurred alongside the evolution of the governance agenda – focused on accountability, citizen-state relations and democracy – which advocated a move away from the blueprint model of development to, instead, innovating new models that are context-specific, more practical and based on ‘best-fit’ rather than ‘best practice’. Finally, DFID’s increasing involvement in fragile states exposed a number of political challenges (related to inclusive political settlements, for instance). All of this meant that DFID realised that it ‘ignored political at its peril’.

How has the interest evolved?

DFID’s ‘first generation’ PEA was through Drivers of Change analysis. This was focused very much at the national level, using the state as the level of analysis. Drivers of Change analysis strongly influenced DFID language, strategy and policy, however it was not so successful at linking findings to operational challenges on the ground. The ‘so what?’ question remained.

These limitations led to the development of a ‘second generation’ PEA, centred on the questions: what are the constraints?; and how is the politics influencing things?. On this basis, PEA has been used and now informs programme design. DFID is interested in working at a more political level in:

·         Strategy: building coalitions for change

·         Programmes: grappling with risks and assumptions of programming

·         Operations: trying to engage with a broader range of actors

Other donors are also increasingly using PEA.

BUT – it is important not to overstate this uptake by other donors and PEA does, in many ways, remain a niche issue, with most donors still focused on more traditional aid effectiveness concerns.

What has been learned and what are the challenges?

PEA has revealed that prospects for change might be a lot weaker and the time horizon for change might be a lot longer that previously thought. Donors therefore need to be much more realistic about how much they can impact the rules of the game and political context. Donors must also work much harder to engage with non-state actors, keeping in mind the challenges that this can pose for state sovereignty.

Yet these challenges do not mean that nothing has been achieved or that we should give up.

PEA has changed the culture within DFID but despite this, there are still many incentives within the organisation at play that may mean PEA does not always filter into practice on the ground. There is still a tendency towards technical solutions and best practice approaches.

More recent concerns about how the results agenda may negatively impact on support for PEA should not be overplayed. The challenge for researchers and PEA advocates is to demonstrate the impact of PEA work. There is also a need to manage expectations of PEA – there is no need to sell it as a radical tool that leads to solutions that wouldn’t have come about without it.

Alex Duncan provided some comments on the presentations, asking how far donors have tried to embed PEA in their work. Virtually all bilateral have engaged in PEA to greater and lesser extents and this has led to changed practice within many donor organisations in some contexts. Yet there is by no means a consensus on PEA and uptake remains uneven. Why so?

1.       The stickiness of selling PEA to senior bureaucrats (new staff skills have to be developed, which takes up resources)

2.       There are respectable models of analysis that do not involve PEA (and its added costs). While these models don’t answer all the questions – neither does PEA

3.       PEA takes agencies into difficult and sensitive areas that has traditionally been the realm of diplomats. This can, at worst, risk challenging the Paris Declaration principle of ownership.

So while people may understand the logic of PEA, there are still reasons that may lead them to shy away from its use.

What have been the effects of PEA?

1.       At the level of thinking: PEA helps to be clear about how development change can happen. It sets more realistic expectations and seeks to understand local realities (including the presence of broader stakeholders). This has had substantial, although uneven, impact.

2.       At the level of strategy: PEA looks for opportunities for change that build on existing institutions, rather than building new ones from scratch, or from other models. This has led to a broadening of the types of interventions that donors undertake (although there is no evidence yet to prove whether donor strategies are more effective because of PEA).

3.       At the level of operations: PEA is effecting programmes but this is often isolated. If donors shared PEA analysis more broadly this could be more apparent.

4.       At the level of development impact: Achieving impact here is the ultimate justification for PEA work. Currently, the jury is still out. While there is evidence of more informed project design it is not yet year whether development effectiveness has improved as a result.

A good step forward would be for DFID to undertake a review of its investment in PEA (perhaps in partnership with the Netherlands) to assess the impact of PEA to date.


Group questions and discussion centred on the results agenda, what attributes make PEA effective, the relationship between PEA and conflict analysis and country ownership. The following key points emerged:

·         PEA requires longer term research outside of the short research usually commissioned by donors. This would allow for more sustained uptake.

·         While problem-centred PEA might have been most effective to date, it is important going forward that this is linked back into sector work, where getting buy-in from sector professionals could lead to more sustained use of PEA (rather than it being a niche issue for governance teams).

·         PEA can be seen to infringe on ideas of country ownership. However, PEA is also intended to improve donors’ understandings of local context and realities. As a result, it should lead to more honest, open and informed dialogue with country partners that shows a greater understanding of the constraints that they face. In this way, PEA can be instrumental in better understanding the challenges to genuine country ownership and how to address them.

·         We shouldn’t be more demanding of PEA than we are of any other tools of analysis – there is little concrete evidence for any of them changing the world. The change that is likely to result will almost always be incremental and therefore difficult to measure.

·         Should be a research focus on what PEA has changed operationally

·         PEA is most effective where there is a clearly defined problem, where the commissioning agency is supportive of PEA approaches, where the political environment is open enough to allow discussion of the problems, and where there are some ideas on the table regarding what should be done (so that these can be tested).

To close discussions, Julian Lob-Levyt discussed DAI’s increasing involvement in PEA. He pointed to concerns about the narrow focus on quantifiable results that has occurred in fields like health and warned against such approaches in broader development. PEA offers a way of trying to understand the limitations of traditional development responses, to grapple with the diversity of actors that need to be engaged with, as well as to improve the manner in which donors engage. Yet despite these advantages, PEA does not provide a way to reliably predict how a political context will unfold. It is, rather, another tool that reveals important and often overlooked information about the contexts in which we work.


Over the past ten years, there has been a growing recognition within the donor community that development aid needs to take politics into account if it is to be effective. International agencies that long hesitated to speak openly about the political dimensions of their work are now more willing to do so. As part of this, the use of political economy analysis to better understand contextual realities has proliferated.

A more politically informed approach to development is promising: it would entail moving away from models of development based on ideal standards and blueprints (‘best practice’) towards recognising multiple paths to institutional performance grounded on contextual realities (i.e. ‘best fit’). Yet, translating the insights and lessons emerging from the politics agenda into real changes in developmental practice has proven much more challenging, and uptake among donors remains weak.

This event, co-hosted by ODI and DAI, builds on a dialogue that ODI convened in December 2010 on how development actors might go about incorporating political insights more seriously into their work. The event will also launch a special issue of DAI’s journal Developing Alternativeson political economy analysis.