On Wednesday, ODI hosted a great discussion on half-won revolutions, rivers that run deep and things that are obvious and yet hard to do. We were discussing the relationship between politics and development, what, in their excellent new book, Tom Carothers and Diane de Gramont call ‘the almost revolution’.

There was agreement that development aid is fundamentally political. However, acting and working politically, and agreeing on political goals and objectives, is easier said than done. What donors want and what those in developing countries want may not be aligned; the Bolivian government’s decision to expel USAID recently illustrates the tensions that can arise when a donor is seen to be interfering in a country’s domestic affairs. Owen Barder, from the Center for Global Development, reminded us at the event that the politics of donor countries play out here too – signs of waning public support can make the choices for development policy more constrained, and there is a limit to what aid alone can achieve. More fundamentally, the world has changed. Development is no longer dominated by the donor–recipient paradigm. Instead, there is a proliferation of new or non-traditional donors: we are witnessing an age of (greater) choice for those who receive assistance.

What does this mean for politics and development?

I came away thinking that, while we haven’t yet confronted the extent to which development practice can and should have political goals, or the question of what it takes for development practice to confront the big political issues, we are building up a much clearer picture of how to achieve politically smart development assistance. This might represent a smaller bite of the cherry but actually gives us a better sense of the nuts and bolts of how future development agencies will need to look and behave if they are to continue to remain relevant.

In our work on the politics of service delivery we’ve found that starting with the political drivers that sit beneath particular development problems – asking why there are chronic stock-outs of essential medicines, despite reform efforts, or  why health workers do not show up in rural postings – is useful for moving towards assistance that is better attuned to political realities. And we are seeing growing demand, from donor country offices, implementers, NGOs and others to apply political and governance understanding to help address these chronic bottlenecks or blockages.

It is useful to identify instances where the same types of problems persist across sectors (for example, where health, education and water services may be skewed towards particular ethnic or regional groups) and where there will be differences between sectors (such as why user demand will be weaker for rural sanitation than for primary education).

At our recent event, David Booth suggested that we know a fair amount of ‘what works’ in addressing these bottlenecks. We know that adaptive, iterative approaches for problem solving are needed (something Matt Andrews will be discussing at ODI on 21 May) and that this can be delivered through forms of arm’s-length assistance, requiring certain brokering and facilitation skills.

But we still know much less about how – and whether – donors can support large-scale political change. Experience with forms of support to political institutions (parliaments, political parties) suggests it either falls short (often opting for technocratic training) or lacks political judgement and understanding. But, as a minimum, as Diane de Gramont reminded us, there is a need to be aware of the overall effects of development assistance in different sectors on wider political and domestic accountability systems.

Crucially, the answer to this cannot and should not just be more political-economy analysis; nor should we be tempted to apply fundamentally technocratic approaches to problems that are complex and political in nature. What we need are innovative ideas for embedding politics and political thinking within the DNA of development agencies, including in hiring and promotion processes. In practice this involves turning the political-economy lens on donors themselves, as much as on the political systems in developing countries.

As Owen Barder commented last night, a UK civil servant sitting in the Treasury and looking to implement a new reform would pay close attention to the potential winners and losers of that process, and develop strategies accordingly. Why these basic principles do not consistently apply in the same way to those in development agencies is something we need to better understand. Exploring organisational incentives and pressures that may not currently support these ways of working is going to be crucial.

So how far are we from confronting the real challenges of making development aid a more political endeavour? Modesty and realism are certainly needed, but the revolution is underway.