MacArthur Fellow Esther Duflo recently presented to the UN MDG Summit a set of seven interventions, vetted by rigorous action research, that hold the potential to help poor countries make significant advances towards achieving the MDGs.
What I find fascinating is the emphasis on cost-effectiveness of each idea, the simplicity of technology involved and the ease of delivery of the actual intervention itself…that is, once the policy makers embrace some or all of these ideas and the respective bureaucracies start putting in place, the mechanics for implementation of these interventions. Having established that these interventions work (at least in contexts for which they have been tested and approximately similar settings), the next big task is to find global and domestic policy alliances that can push these ideas. However, that is where another huge challenge faces development professionals who work (or at least aspire to work) at the interface of evidence-based policy making and the subsequent actual implementation.
It is probably fair to assume that in many countries, policymakers can be persuaded to adopt these ideas, given the glare of the constant spotlight on country-level MDG achievement by the media and the global aid architecture. Sure, countries where foreign aid is not of significant proportions are different – but most countries seriously under-performing with respect to MDG targets are heavy foreign aid recipients. Also, countries in the above list also face serious problems of state capacity. But even in states with reasonable financial, human and physical resources (and of course, for the purpose of this discussion, overall sovereign political legitimacy), can policymakers fight their own bureaucracies? Not to imply that bureaucrats don’t make policies – but there is experience that suggests that getting middle and lower-level bureaucrats to ‘buy-in’ is a significant challenge. And as research by Lipsky and Tendler (and many others writing on policy implementation) show, often what becomes of government programmes is what the street-level bureaucrat (the frontline government officer – the primary school teacher, medical officer, village tax inspector et al)makes of it and also that not only is it incredibly difficult for their superiors to have effective control over them, but also that such control may not be in the interests of those superiors themselves. Also, in poor countries with low levels of MDG achievement, what are the incentives facing the bureaucrats to work harder than before with development interventions? What are the rewards they can expect for stellar performance and are there any deterrents at all to non-performance?
This scenario calls for pressure from both sides of the spectrum – top-down from the policymakers as well as bottom-up from the citizens. The absence of one of these elements could nullify the best efforts of the other and that is exactly why for public programmes to have a chance on the ground, sustained pressure from both ends (and in synchronised harmony) is absolutely necessary (but possible, by no means sufficient). A much-publicised audit of the NREGA in the state of Rajasthan, India (the state incidentally yields some of J-PAL’s most popular findings in the area of primary education and immunisation) that combined months of background work by the government officials (aided by a enthusiastic District Collector – the chief executive of the district administration) and civil society activists and an amalgamation of social and official audit methods is a great example of these dual forces at work.
I am sure there are numerous other examples where implementation of public social welfare programmes have been monitored and reviewed in this fashion. Of course what works in India will not work in Ghana or Peru – but the lessons remain important. And these lessons will be learnt from the ground, from observing how political mobilisation works in different settings and gaining clues to the factors that determine which ideas get a groundswell of support, not just among citizens, but amongst bureaucracies so that implementation is supported and monitored from both above and below.