Using payment by results to distribute international development aid risks being more about looking good than doing good, according to an academic at the University of East Anglia (UEA).
Payment by Results (PbR) - where results already achieved directly determine the amount of aid paid - is becoming increasingly important for various donors and recipients, from small NGOs to national governments. For example, governments have been paid by donors for each additional child completing primary education or for each vaccination given, while NGOs have been paid for the number of functioning wells or the additional number of students meeting a set learning outcome.
In the UK the Department for International Development (DFID) stated in 2014 that it was using PbR by default wherever it offers best value for money, while other organisations to adopt similar systems include The World Bank, global vaccine alliance GAVI and the European Commission.
However, there has been little in depth analysis into whether these relatively new agreements are likely to be more effective than traditional forms of aid that allocate money for future activities.
Dr Paul Clist, a lecturer in development economics at UEA's School of International Development, examined cases where the success of PbR was measured and led to a disbursal of aid, even when the success appeared to have been a "temporary illusion", false or at the expense of the true goal of the donor.
The findings are published in The World Bank Research Observer journal and come at a time of intense debate about how aid should be allocated, with the UK's new international development secretary calling for its budget to be overhauled.
Dr Clist said: "There are great hopes that this innovative mechanism will focus attention on ultimate outcomes and lead to greater aid effectiveness by passing the delivery risk on to recipients. Traditional systems do not focus on what is achieved, so payment by results seems quite attractive. The problem is that projects might look successful when you pay out but in reality they haven't been, or you don't give aid when you should.
"While there will, of course, be cases where PbR has a real advantage over traditional forms of aid, bad measures will lead to a large amount of 'fool's gold' that may be indistinguishable from real success. Positive evidence of PbR should therefore be treated with scepticism, and PbR contracts should only be used in the right circumstances.
"Payment by results will most often not be the right way of delivering aid, it is a way of looking good without necessarily doing good, and it definitely shouldn't be used as the default method of distributing aid. If it is used too broadly, many of the results it claims are likely to be misleading and we risk distributing money when nothing has been done or nothing has been done well. I would like to see aid done well rather than not at all."
Dr Clist argues that because the level of information available to the donor and recipient about whether a PbR contract would be favourable may be different, in some cases agreements are made where recipients are rewarded for improvements that would have happened anyway. One example he cites is a PbR contract with the Rwandan government, in which a large portion of the money distributed for improvements in the number of children sitting exams appeared simply to be related to more children being born.
In his paper Dr Clist suggests that in order to be most effective PbR needs data that is accurate, verifiable, of genuine interest and timely, for example on specific numbers of school enrolments, vaccinations, wells built, or children reading to a certain standard. At the moment the relevant data does not exist.
"The inherent tension between a desire to reward performance and alleviate desperate situations is apparent," said Dr Clist. "If countries with poor performance receive aid anyway, the incentive is lost. Equally, if these countries receive no aid, then they are disregarded.
"One of the reasons for this tension is the difficulty of separating effort from conditions. Payment by results should be used in the right circumstances and if it is to be properly judged, evidence of its success will need to be robustly proven from non-incentivized data sources, which at the moment are rare."
'Payment by Results in Development Aid: All that Glitters is not Gold', Paul Clist, is published in The World Bank Research Observer.