It must be a tough time for donors. From every side new research is showing that the development world is more complex and more political than their standard approaches can cope with.
The rise of Political Economy Analysis points to the need to think and work politically, and to ensure that foreign aid does not undermine local institutions. Meanwhile evidence of successful institutional reform points to the need for problem-driven iterative adaptation (PDIA) – not a one-size-fits-all, best-practice approach. Research on the underlying complexity of development problems suggests that a responsive, learning and networked approach is required. Recent work points to a negative effect on performance that comes from trying to monitor and manage complex projects too closely.
It can only be better that donors have a deeper understanding of the change processes that actually lead to improved development outcomes, and a clearer idea of what external actors can do to support them.
But this better understanding often leads to lists of things to avoid, rather than ideas for new approaches that may work better. When there are new approaches suggested, they are often very cheap but extremely difficult given the constraints of the donor itself; both of which can be a problem when aid money still needs to be spent.
An important question facing donors is: if all this new thinking is true, what can conditionality be used for? Isn’t this the worst possible instrument in this new world? Conditionality can be an overly simplistic method when dealing with complex, hard-to-monitor reforms. It could dangerously tilt government accountability towards donors at the expense of its accountability toward their domestic political setting. And specifying targets in advance could forestall the ‘purposive muddling’ that could be vital for a reform’s success.
But these questions miss a key point emerging from these new research agendas: the need to separate the role of the ‘reformer’ from the role of the donor. As has been said many times, donors cannot lead reform, but they can support it.
This leads me to a suggestion for the use of conditionality as a powerful tool to support reform. Rather than targeting conditionality on indicators of reform, donors could focus it on those factorswhich create an enabling environment for reformers.
There are many candidates for such factors (see Matt Andrews book), but the most obvious to me is the production of information, and public access to it.
Information can be used for two key processes: information for reformers; and information to build political will. Where there are (potential) reformers in government, their ability to identify problems and opportunities for reform, to develop potential solutions to those problems, and to ‘sell’ those solutions to political leaders, will depend on their ability to clearly identify what is actually happening. And in a complex policy environment it is critical for reformers to be able to learn and adapt using feedback on the effects of the first stages of a reform. In relation to political will, information that is publicly available can also be used by parliament, interest groups and civil society organisations to push for reform. These two effects can reinforce each other: reformers inside the government will be more successful in convincing politicians to act, if influential voices outside government are also calling for reform.
Some will argue that you can publish the information, but you can’t be sure that all these other good things will flow from it. Of course that is true. But, if there is no information, you can be sure those good things won’t happen. Information is not sufficient for successful reform – but it is necessary. Many people think this implies it is not worth doing. On the contrary, its absence is a binding constraint, which means it must be done first.
The next question might be: what is the role for donors in this world? In fact, prioritising the kind of data on which to place conditionality (whether on the public finances, service delivery, poverty and inequality, taxes or many other issues) is a key challenge. To galvanise the virtuous circle described above, the topic probably needs to have some political resonance, and some interest groups prepared to engage with it. So isn’t this choice also political? Maybe, but less so than prescribing the actual reforms. And one potential way donors could helpfully work politically is by strengthening the ability of people in developing countries to claim their rights from the state. Indeed, if knowledge wasn’t power, why would governments be so unwilling to share it?
Finally, if donors do want to use conditionality to improve the provision of information, what does the new research tell us about the best way to do this? Firstly, due to the concept of ‘isomorphic mimicry’, we know not to attach conditionality to obvious targets like establishing a new statistics agency or passing a new law on data. Secondly, while ensuring standard statistical techniques are used, we don’t want to get in the way of PDIA: reformers improving their processes in ways that work for the local context. So how about attaching conditionality to the public access to data in a usable form? This would be evolution not revolution, and also has the helpful characteristic that it uncouples donor spending from the cost of providing the data (as opposed to ‘strengthening’-type programmes). So donors are free to choose the amount they are willing to pay to see the data appear.
New research suggests that donors could improve their support for institutional reform. If creating anenabling environment for reformers is important, then using conditionality to create information would be a good place to start. There would also be a significant spillover effect from the new data, by providing opportunities for future research and new evidence.