With more than 50 years of experience to draw on, it is dispiriting that so much current debate about international development is still narrowly focused on the effectiveness of aid. This is perhaps understandable in the case of recent policy statements by Britain's three main political parties, all bent on persuading a financially squeezed electorate that aid spending remains worthwhile. But an aid- and donor-centric view of development also pervades more nuanced debates between development practitioners - including recent exchanges in openDemocracy, stimulated by Phil Vernon’s question (in an article on November 9th, 2009): "Is aid working?"

Phil finds aid wanting when measured against the ambitious objective of providing "sufficient impetus to overcome the strong forces that keep people poor". He is right to be concerned that donors too readily offer "northern tools for southern problems”, and use the facile language of partnership to avoid confronting the fundamental mismatch between their own "technical" goals and the political priorities of recipient governments. He is also right to question donors’ preoccupation with strengthening the normative role of the state, and their tendency to shy away from the politics of development. But while aspiring to achieve "social transformation" he fails to investigate the processes involved in making that happen, and so comes up with what is ultimately a very conventional agenda for aid policymakers, including the highly questionable one of using aid to negotiate "context-specific conditions about improvements in governance, citizenship, peace and development".

Owen Barder (14 December 2009) is right to point out that the answer to the question "Is aid working?" depends on what one believes aid should be trying to achieve, and also right in saying that money from outsiders is not a good way of trying to promote social and political transformation in poor countries. And he is right to value the direct benefits of aid in improving the quality of life for poor people. But he too readily consigns ambitions to support economic and political change (at least so far as donors are concerned) to the “too difficult” box. He therefore, also, fails to explore the local political processes with which donors inevitably engage and which they influence - sometimes in quite subtle and unexpected ways - whether or not they are trying to do so.

These openDemocracy debates - and policy statements from political parties - are symptomatic of a wider problem within the so-called "development community". Despite growing rhetoric about the importance of politics, and increasing interest by leading donors including the World Bank in using political economy analysis to illuminate underlying development problems, practitioners have not fundamentally questioned their mental models of how development happens, or their role in the process. A recent publication from the Institute of Development Studies at the University of Sussex draws a parallel between donors and new art students. One of the first things aspiring artists are taught is to close off their pre-existing knowledge of the objects they are trying to draw. They know that a chair has a back and a seat of roughly equal proportions, and four legs of identical length. But this knowledge is distinctly unhelpful in trying to capture a three-dimensional image of a chair in a two-dimensional medium, and to deal with the challenges of perspective and foreshortening. Instead, they are instructed to look at what is in front of their eyes, to forget their mental model of a chair, and to draw exactly what they see.

Development practitioners similarly need to close off their existing assumptions and mental models about governance and development. Their default position is to look at the world from the perspective of an OECD state, and to analyse problems in developing countries in terms of deficiencies when compared to an (often idealised) model of impersonal, rules-based governance. So they have focussed, with limited success, on trying to improve the investment climate, reform public services and fight corruption by strengthening formal, rules-based institutions. Moreover, implicit in much Western policy relating to developing countries is still that West is best, and that “developed” countries need to take the lead in finding solutions to the problems of “developing” countries. But power is shifting away from OECD states, while China and other BRIC countries have growing global influence, and offer alternative development models and sources of capital. Traditional approaches to managing aid relationships need re-thinking.

Instead of viewing development challenges through the lens of OECD experience, the IDS research explores in an open-minded way the complex political processes of bargaining between public and private actors, and interaction between formal and informal institutions through which common interests can be identified and elements of public authority and public goods can be created. For example, a study in Egypt shows how informal relations between politicians and investors can compensate for weak formal property rights and contract enforcement, and boost investment in the short to medium term. Research in South Asia shows that informal village level councils remain very influential, and can complement or undermine formal institutions. Studies in Ghana, Kenya and Ethiopia show how governments’ need for tax revenue has driven implicit or explicit bargaining with citizens, with the potential to enhance accountability. (For details, see the aforementioned Institute of Development Studies publication.)

Focusing in this way on local political dynamics suggests a different role for external actors: instead of trying to use aid to buy good behaviour, or to orchestrate and support rules-based reform, donors might do better to think about more indirect ways of influencing the interests and incentives of local actors. This might include using their market power to increase incentives for sustainable management of natural resources, or prioritising international action to improve financial regulation and constrain criminal activity, thus counteracting some of the perverse incentives that help to perpetuate fragile states. Donors could do more to facilitate dialogue between politicians and investors, and a common understanding of obstacles to productive investment. They could be much more alert to ways in which aid financed projects and programmes can either undermine or support incentives for effective collective action by service users or civil society groups. They could support tax reform that prioritises equity, transparency and improved collection, not just increased revenue.

Some of these ideas are starting to influence donor practice, not least in so-called fragile states. But donors need more than an alternative menu of policy options. They need a different way of thinking about the messy, complex reality with which they are engaging. A good starting point would be to ask a different set of questions - not how well does a particular country measure up against an OECD inspired model of good governance, but what is shaping the interests of political elites? (Sources of revenue are likely to be critical). What is shaping relations between politicians and investors, and might they have common interests in supporting productive investment? What might stimulate and sustain collective action by social groups to demand better services? What informal local institutions are at work, and how are they shaping development outcomes? Where does government revenue come from, and how is that shaping its relations with citizens?

Donors have much less direct influence on local development processes than they often like to think. But they could do more to help shift incentives of local actors in progressive ways - and do less harm - if they made a serious effort to understand the informal arrangements, interests and relationships that shape opportunities and drive behaviour in individual poor countries. The implications of this are quite radical: it would mean fundamental changes to the culture, mindset, staffing and organisation of aid agencies, not just changes to aid “architecture”; and much more honesty in explaining and justifying to taxpayers what aid can be expected to achieve. None of this will happen unless donors stop viewing development through the lens of OECD experience, and start looking in a much more open-minded way for local capacity and incentives that might support shared development objectives.