Results-based aid (RBA) has tended to be seen as too focused on simple measures of impact to work for governance. With signs of progress on indicators, and a little juggling of the concept, could there be a place for RBA in institutional reform?
Given the growing demand for evidence of development impact, some organisations are looking seriously at results-based aid. Traditional aid supports developing countries mainly by providing inputs (money, expertise etc.) With RBA, a set of targets are agreed by the partners and then the donor steps aside, leaving the partner country to manage its own development. Money only changes hands when the agreed targets are achieved.
Intuitively, this seems attractive: with RBA, ownership rests with the developing country and donor taxpayers are reassured that their money is well spent. The US took a step in this direction with the prequalification criteria for the Millennium Challenge Corporation, launched by President George W. Bush. In 2012 the World Bank launched its Programme for Results (PforR), which is now backing schemes – e.g. for clean water – that would previously have received up-front loans.
Pitfalls of results-based aid
There are, of course, potential pitfalls. An obvious one is the assumption that the partner country can cover the initial cost of the programme, when in reality this may not be the case. A proposed way around this is to create Development Impact Bonds, a concept promoted by the think tank the Centre for Global Development (CGD). With DIBs private investors assume the initial costs of a programme, also contracting whatever additional support is needed to make it happen. In return for taking on the risks and initial costs, the investor eventually receives a payment – based on the results – that includes a profit.
Even if you solve the initial cash-flow issues, however, there is a second problem with RBA: the outputs need to be measurable in order to determine when acceptable results have been achieved. This is easy enough with things like water supply – where the quality of the water, the pressure of supply and the number of working outlets can all be recorded. It is less straightforward in relation to ‘governance’, the catch-all term for everything from capacity development to democratisation.
Stephan Klingebiel of the German Development Institute (DIE) argues in a paper (PDF) on the issue that in governance, the devil – the potential ‘mis-incentives’ created by RBA – is in the detail. This is particularly true when measurement is complicated by at least three theoretically potential areas of ‘gaming’, or playing the system. In the first of these, the agreement on results is skewed to the easiest areas of reform (e.g. drafting a new strategy which may never be implemented). The second involves achieving only at the minimum level (paving the way for further programmes). While both donor and partner governments are vulnerable to these risks, they could mitigate them through careful selection and tracking of final targets.
The third problem is capacity ‘displacement’ – risks associated with the advisers employed by the investor to help implement the programme. Advisers might deliver all the agreed outputs and outcomes in order to guarantee payment – but without improving the systems involved. Displacement can be a serious problem for the partner country involved, running contrary to the need for capacity development and providing a quick fix that may not last (e.g. the programme ends, the consultants disappear and the process collapses).
Turning risks into opportunities?
A new aid instrument such as development impact bonds would, therefore, need to offset the problem of initial resourcing, while also mitigating the potential risks of displacement. One way of doing this would be to develop measures of genuine capacity development and then link those measures to the risks borne by investors. The World Bank is developing new indicators for public-sector reform (modelled on existing Public Expenditure and Financial Accountability indicators) that look at the qualitative performance of an institution, rather than just final tangible programme outputs.
The way these measures are linked to investor payments could also help to address displacement – which as a risk is nothing new. Today large sums are spent by donors on `technical co-operation’, including dedicated technical advisors within partner institutions who work to mentor and advise local colleagues (sometimes in directly risky situations). These advisers may already have to contend with potential perverse incentives. Not least to bypass counterparts in order to deliver results or to extend their role by going-slow on skills transfer.
Development impact bonds could turn these capacity risks around by asking the large consultancy companies that provide advisers to act as the initial investor (raising new capital if needed). Then, instead of measuring just the final output of the Ministries, the programme would also measure the skills transferred (using the new World Bank indicators). Only when the partner government is happy that its institution has acquired greater capacity (verified by the indicators) would the consultancy company get paid. For consultants this would help finally to lay to rest any doubts about their role, and offer clear validation of their achievements.
So, if partner governments are to be assured that advisers are trying to work themselves out of a job, then maybe financial risks should shift from donors and local institutions to the companies who stand to make the profit. A case, perhaps, for pilot programmes to test this out.
*Alan Whaites is the Team Leader of the Governance for Development and Peace Team (G4DP), Global Partnerships and Policy Division, OECD-DCD.
Interested in writing a guest blog for Governance and Development?
Contact Caroline Martin or Emilie Wilson for more information.