The notion that development agencies will be more effective if they understand the politics of the places within which they work is pretty straightforward.  Indeed, why would this be in any way incorrect?  About a decade ago a wave of work started in development agencies based on this proposition – including DFID’s ‘drivers of change’ studies in Malawi, Zambia and Bangladesh, and Sida’s ‘power analysis’ approach.  More recently the European Commission has started to focus on politics, and the World Bank has produced a range of work raising some interesting narratives – that political economy analysis (PEA) needs a ‘problem-driven approach’ to be influential and that sector level analysis carried out alongside national institutions can be a key way forward.  From my own experience as a co-champion of the Bank’s Community of Practice on PEA, I recall well that operational staff were easily convinced of the importance of PEA – but translating this into an adequate budget to support useful analysis was a big challenge.  To get more traction on operationalising political analysis in the current climate you need to be able to demonstrate the difference it makes, and this remains a challenge.

A recent meeting at ODI revisited the question with thoughtful reflections from Alina Rocha Menocal of ODI, Stefan Kossoff from DFID, Alex Duncan from the Policy Practice and DAI’s Julian Lob-Levyt.  The meeting was co-hosted with DAI and launched a jointly produced edition of the journal Developing Alternatives – focusing on ‘Political economy analysis and the practice of development’.  I took away three main messages from the discussion:

• Development agencies have got better at understanding political context – but it’s hard to ‘prove’ this has happened in any simple way. This is largely because the most important element is a culture change – where consideration of the politics of inclusive development becomes a legitimate area for analysis and consideration in decision-making.  It is also because a huge amount of the work that is commissioned never gets released publicly (because it is sensitive and may appear out of tune with the mandate of a development agency).  And finally because the biggest impact of development actors getting better at this may simply be avoiding very big and very significant screw-ups.  For example, some have argued that development agencies played a significant role in creating the conditions for the Rwandan genocide through poorly informed engagement.  If development agencies have got better at understanding conflict dynamics and how they can make sure their actions do not promote division, this is a significant improvement.  But it’s hard to prove – being about the ‘counter-factual’ – because you cannot easily measure the impact of what did not happen.
• We need better stories on the link between political analysis and country ownership of development actions.  Sometimes the practice of PEA seems associated with an unhelpful ‘othering’ of development partners.  In this way, what should make development agencies less technocratic in their professional stance can actually work the other way – as the neutral technocrats look for the reasons why their technical solutions (which they believe to be obviously correct)  are not adopted (‘it must be the politics!’).  We need to get away from this.  There are three key points here: firstly any analysis should reflect on the political economy of donors themselves and their impacts on the wider context; secondly – where it makes sense – we need to develop a stronger practice of engaging and supporting an analysis of the political dimensions of development alongside country partners; and thirdly, we should be using political economy analysis to inform policy dialogue.  There are many examples of this – again we are perhaps hampered from learning how to do this better by the fact that many of these cannot be publicly disseminated in any simple way. Recent work conducted by ODI in the water sector suggests that it is of crucial importance to get the process right in conducting and following up the analysis if it is to work in these ways.
• Understanding the global context for national and local political change is an important part of effective practice, where we have some good examples but need more.

So where do we go from here?  Do we just say ‘understanding political dynamics is obviously useful, but it’s hard to prove it works apart from the odd well-documented case’ and leave it at that?  Not quite.  We need to look for new ways of understanding what works in pursuing this, and develop a better appreciation of the progress we have made and the progress we still need to make.  This will be a significant strand of our forward work on politics and governance within ODI and we look forward to building on this discussion....