The best synthesis so far of where we’ve got to on ‘Doing Development Differently’

March 17, 2015 8 By Duncan Green

Finally got round to reading the ‘Adapting Development’ the ODI’s latest 54 page synthesis of the theory and practice underpinning the ‘Doing adapting devt coverDevelopment Differently’ approach. It’s very good – a good lit review, laced with lots of case studies and good insights – and definitely worth a careful read. Weirdly the bit that jumped out for me was on results and monitoring (see below):

The starting point is that ‘Change is almost always driven by domestic forces, and often occurs incrementally, as a result of marginal shifts in the ways interests are perceived, especially by elites.’

ODI argues that ‘the best approach for domestic reformers and their supporters combines three key ingredients.

• Working in problem-driven and politically informed ways. This might seem obvious but is rarely the norm. Such an approach tracks down problems, avoids ready-made solutions and is robust in its assessment of possible remedies. Too often, diagnosis only gets as far as uncovering a serious underlying challenge – often linked to the character of local politics. For example, studies of medicine stock outs in Malawi and Tanzania and of human resources for health in Nepal reveal how power, incentives and institutions lead to chronic gaps in supply. It is difficult to identify workable solutions to such problems, and attempts to do so often focus on the wrong things. Doing things differently means understanding what is politically feasible and discovering smart ways to make headway on specific service delivery issues, often against the odds.

• Being adaptive and entrepreneurial. Much development work fails because, having identified a problem, it does not have a method to generate a viable solution. Because development problems are typically complex and processes of change are highly uncertain, it is essential to allow for cycles of doing, failing, adapting, learning and (eventually) getting better results. This requires strong feedback loops that test initial hypotheses and allow changes in the light of the result of those tests. Some of the greatest success stories in international development – the South Korean industrial policy being only one example – are the result of a willingness to take risks and learn from failure.

• Supporting change that reflects local realities and is locally led. Change is best led by people who are close to the problem and who have the greatest stake in its solution, whether central or local government officials, civil-society groups, private-sector groups or communities. While local ‘ownership’ and ‘participation’ are repeatedly name-checked in development, this has rarely resulted in change that is genuinely driven by individuals and groups with the power to influence the problem and find solutions.’

So much for domestic change agents, what about the aid agencies that pay ODI’s wages?:

‘Donors can help reform processes to adopt a problem-driven and adaptive approach, but if they are to be effective they must act as facilitators and brokers of locally led processes of change, not as managers. This means big changes in the way aid agencies work. And agencies will not change without new guidelines from the highest level: from ministers and other politicians who, in turn, respond to the perceptions and interests of voters and taxpayers. We propose, therefore, some major changes in how aid works and in the way aid is treated in public policy debates.

• Aiding development that is politically smart and locally led: Aid should do more to support initiatives that are problem-driven, adaptive and locally led. These initiatives need financial and other support that is fit for that purpose.

• An explicit refocusing of the debate on how aid works, not the total volume spent. There are many areas where spending that benefits poor countries could be increased, but the current debate about targets for aid spending is too focused on the ability of the donor country to pay, rather than whether those funds are used effectively. Looking at how aid works is more important than how much to spend.

• A new and more honest dialogue about development and aid with the public. According to recent evidence, ordinary citizens in donor countries are often irritated by simple ‘heart strings’ appeals. Many would welcome a frank discussion on how development happens, why it is often difficult and how aid can best support development that is both genuine and lasting. Efforts to support such a debate should be scaled up.’

dilbert auditingAnd the interesting and new (for me at least) section on results and monitoring (and how to ensure it doesn’t go the way of Dilbert, above)

‘Doing things differently should actually be more oriented towards results in two respects. First, over reasonable periods of time (which will vary according to the objective), programmes should be able to make plausible claims of having made a contribution to positive development gains, or else they should not be supported. Second, much greater efforts should be made to build up and document experience on the intermediate change processes.

Process measures of this sort could include the following.

• Measures of the extent to which issues have local salience or relevance, and whether processes give priority to local leadership and capacity. This could be based on asking simple questions about the extent to which users and local networks and organisations are involved in issue selection, design and implementation, or through perception or survey data to track how this changes over time.

• Evidence of adaptation to context. This means taking into account sub-national variance, local (formal and informal) institutions, the strength of networks, power relationships and more. This might include evidence of the use of the best knowledge available about the local political economy and its dynamics.

• Evidence of learning in action. This would measure the use of feedback loops, of evidence on past experience, and adaptation to changing conditions on the ground.

• Measures of innovation and entrepreneurial action. Sources of inspiration here may include recent attempts to monitor and measure innovation processes and impacts. Another type of test could assess the extent to which initiatives rest on a series of ‘small bets’ – i.e. spread their risk across activities – and specify the ways in which they will test and measure the effectiveness of different approaches.’

As for measuring state capability we need to ‘shift the focus of governance indicators to measures of state capacity that capture performance in core functions, rather than the adoption of particular institutional forms. Examples include the ability of a state to register all children at birth or to reduce road-traffic deaths; both of these calls for a type of state capability that could be applied to a range of development challenges. However, there is also a broader need for a better approach to identifying and measuring practical steps that can be shown to lead to improved outcomes.’

All in all a really useful contribution to building up the body of evidence and analysis underpinning the Doing Development Differently movement. But allow me one cheap shot, in relation to previous beefs about the gender blindness of much of the DDD work. The word ‘gender’ appears once in a 54 page report (in a footnote), while ‘women’ appears 5 times. In some respects, this is more same old, same old, rather than doing things differently.

And of course, as it’s ODI there are accompanying infographics etc, and this 10m video on the Philippines case study (land rights reform).

Update: and a nice riposte from Matthew Greenall who thinks this is not going to rock the boat, and is far too dominated by donors