I believe five steps would take us a long way towards a governance-for-development practice with solid grounding in evidence and experience.
1. Governance for development should be treated as a skill-set, not as an area of work.
It is time to stop working on governance, and start addressing all kinds of development (including peace and security) problems with the benefit of insights from the study of governance.
If I am even half right about the five myths, there is no justification for dedicating scarce resources to trying to improve governance comprehensively ahead of significant social and economic change.
But the importance to development practice of understanding the way political and administrative systems work remains undiluted.
2. The governance advisory resources at the disposal of major development agencies should be redeployed massively into joint ventures with other advisory groups.
There is growing support for politically smart (informed and astute) working in all areas of development, including conflict reduction, inclusive economic growth, social equity and public services. Adaptive planning or entrepreneurial reform approaches are getting greater recognition among growth and sector specialists.
There is real potential, therefore, for a pooling of wisdom among different professional cadres to identify change objectives that are simultaneously technically sound (high impact, sustainable, etc.) and politically feasible (if pursued in smart and adaptive ways).
3. In these joint ventures, the focus should be on solving specific problems.
Recombining existing comprehensive approaches to sector issues with equally overblown governance reforms will not do the trick. In the spirit of the recent ICAI review of DFID security-sector programmes, the target should be particular changes with high development pay-offs that are potentially soluble with politically smart tactics and ‘good enough’ institutional arrangements.
Any genuine opportunity to eliminate a wider governance failure should be seized, of course. As Cammack and I have argued, unresolved collective action problems underlie many development blockages. However, facilitating collective action solutions represents the upper end of the scale of ambition for donor-funded interventions.
Problem-solving initiatives should geared to achieving ‘just enough’ change in relationships and institutions. They should be prepared to work with, or around, the constraints created by current political or administrative incentives.
4. The aim should be to enable groups and individuals to pursue their interests in ways that are better for development.
While individual and collective altruism cannot be discounted, convincing people to act against their fundamental interests is usually neither possible nor necessary.
Being politically smart is partly a matter of using new ideas and better evidence, including evidence of social threat, to show powerful people that their objectives can be met in ways than are less harmful to social inclusion and development.
Along with collective action on behalf of the powerless, new ideas that change elite perceptions of self-interest are an element in almost all progressive change, as Dani Rodrik has argued.
5. Quality programming, and accountability to taxpayers, should be ensured by robust trialling of specific change initiatives.
It is commonly assumed that the flexible problem-driven development practice advocated here poses particularly difficult challenges within official agencies that have a strong ‘results orientation’ and rigorous monitoring. But this arises partly from the fact that the more adaptive programmes remain stuck in a governance ghetto and are not really problem-driven.
Where initiatives are seeking to address specific problems, making multiple ‘small bets’ on possible ways forward, it is perfectly feasible for every specific initiative to be assessed on a pre-established timetable. As recognised in MM4P and business start-up circles, this enables timely corrections, with further options trialled on the same tough-minded basis.
Monitoring does not need to be based on demonstrating inherently contestable ‘process’ gains (improved relationships, capacities, etc.). Until such time as there is robust, quantitative and attributable evidence of a specific development problem solved, quality assurance and accountability can rest on explicit period-specific performance objectives set in advance.
Some development agencies have been moving in the direction suggested here, by setting up small units that harness governance expertise to specific development challenges. I would encourage them to go further and faster.