The proposed new Approach has several really useful qualities, among which I would count:
•The frank admission that although moving from ‘best practice’ to ‘best fit’ has been on the Bank’s intellectual agenda since 2000, operationalising it remains extremely hard.
•A clear recognition that the obstacles include both correctable problems on the Bank side (lack of continuity in country dialogue; the Bank’s structural divisions; inflexibility of previously available lending instruments, etc.) and understandable incentives on the client side (the legitimacy and prestige factors in institutional mimicry; tokenistic compliance with best-practice objectives to keep aid flowing, etc.).
•Proposals for addressing the Bank-side obstacles within a flat-funding context which look appropriate and practical – to the extent an outsider can judge.
•A good outline of how, in the context of more continuous, joined-up and flexible working at country and regional levels, the Bank’s public sector management diagnostic and design work might become at least as sophisticated as the emerging approaches to topics such as growth or industrial policy.
All this is terrifically important and promising. Does it add up to a breakthrough? I would be more confident that it does if two challenges could be addressed (and I don’t see why they couldn’t be).
First, communication of the new thinking, and appreciation of its implications, is hindered in a couple of important places by an old language redolent of ‘best practice’ conventions. This concerns a) risk, and b) indicators.
Although the text recognises in various places that uncertainty implies opportunities as well as risks, the headline messages are all about doing a better job of managing risks. Why not be more up-beat, saying more about not missing opportunities? The language of ‘well-accepted theories of change’ is used once (para. 107). Why not make fuller use of that language?
Relatedly, the passages on indicators still seem too little informed by Matt Andrews’ warnings against ‘indicators without theory’. Surely we all now agree that what should be tracked are indicators derived from appropriate theories of change; that is, theories that capture well the tacit knowledge of country contexts that Bank staff often possess. The point is made (para. 101) that the needed indicators are not necessarily quantitative, but that is secondary to whether they are informed by a sound, ‘best fit’ theory.
Second, the question why explicit knowledge about what works in public sector reform remains ‘surprisingly limited’ is more of a mystery than it needs to be. The principal problem is surely that what works in PSR cannot be divorced from the broader question of ‘what works in development’, and the answers to that question are largely about country politics. That means they lie somewhat outside the comfort zones of operational Bank staff. But the Approach is already pushing out the frontiers in this respect (in para. 34 and Figure 4 for example), so why not take a couple of further steps?
Stuti Khemani has already pointed out that according to North, Wallis and Weingast, ruling elites within Limited Access Orders (LAOs) respond to incentives that make them unlikely to be (genuine, not tokenistic) champions of reforms the Bank might want to support. This is a challenge that the Approach needs to pick up somehow, because otherwise it will remain the proverbial elephant in the room, effectively preventing the desired shift from tacit to explicit knowledge.
The good news is that research into the politics of what works in development is generating more fine-grained propositions than are to be found in North et al. Rather than treating LAOs as a group, Africa Power and Politics research is looking for salient differences among states of a broadly neopatrimonial type within low-income Africa. The variable of how the ruling elite manages (or not) the utilisation of major economic rents has proven significant to comparative experience in Africa, as well as previously in SE Asia. This suggests that while North et al. may not be wrong at the rather stratospheric level of generality at which their comparisons are conducted, there are some slightly less dismal stories to be told at the level of particular sub-sets of LAOs – as well as within particular countries.
Here then is a topic, and a body of ongoing research, to which Bank thinking could be turning to flesh out the Approach. A broader alliance with what is going on in policy-oriented politics research would put the pursuit of more explicit knowledge about public sector reform on firmer foundations. That would make the Approach as a whole more compelling.