Corrupt diversion of development aid in recipient countries affects both the efficacy of the intended development programs and the willingness to supply aid in donor countries. Mismanagement of development funds has spurred debate over the ability of our current aid models to achieve development goals (improved healthcare, poverty alleviation, etc.). Many possible solutions for reducing corruption’s effect on development have been tested over the years with varying degrees of success. Various approaches have been tried, including conditioning aid or loans on “good governance” policy reforms, allocating development aid to local governments or local NGOs rather than national institutions, improving oversight and tracking of aid money, and supplying loans exclusively to countries that already have relatively favorable corruption scores (called performance-based lending). Each of these models has its own limitations: Conditionality is often viewed as an affront to sovereignty and has not been terribly effective. The local approach does not address governance issues, and local actors have not always proved to be less corrupt. Oversight of funds is important but costly and imperfect. Performance-based lending seems to leave behind many poor countries that cannot jump the corruption “hurdle.”

In searching for alternative models for distributing aid in light of the aid-corruption paradox, some donors have turned to yet another approach: payments by results (PbR). PbR has been supported by the Center for Global Development (see here and here) and has gained significant traction in the past two years by bilateral donors, such as the UK and Norway, and multilateral donors, such as the World Bank. The basic premise of PbR is that payment to the recipient depends on achieved results. The donor and recipient first define the desired outcomes (e.g., increased TB vaccinations, construction of an infrastructure project, etc.) and determine the amount that the donor will give once the desired outcome is met. The donor may provide some money up front to implement the program, but the rest of the payment is contingent upon performance: The recipient carries out the project independently, the donor measures the results, and, if the results meet the agreed-upon objective, the donor releases the remaining funds. This approach stands in contrast to the traditional input model, in which a donor gives the recipient money for inputs and provides a detailed action plan along with significant oversight for achieving results.

Proponents of the PbR approach argue that it reduces opportunities for corruption in the use of development aid. PbR also boasts other advantages, such as allowing recipients greater flexibility regarding the methods they use to create outcomes. Since most of the project funds are not provided until the project has been successfully completed, recipients cannot embezzle the funds without first making a positive development contribution, reducing the fear that aid money will be used to sustain corrupt governments. PbR can also overcome the issue with performance-based lending (ignoring poor countries that are corrupt) by still allowing those in corrupt countries the opportunity to access aid funds.

Despite these benefits, PbR faces a number of important challenges. Some of these have little to do with corruption (for example, PbR allocates the risk of project failure almost entirely to the recipient, and works only for development outcomes that can be accurately and objectively measured). But even putting those issues aside, PbR’s potential to reduce corruption in development aid faces a number of important limitations:

• First, recipients can still take advantage of donors by misrepresenting the results. Donors are especially vulnerable to misrepresentation when the outcome is difficult to measure in general, or when the outcome is difficult to verify after the fact. Verification can be especially difficult in large-scale projects; for example, consider a project which mandates an outcome of an increase in country-wide TB vaccinations by X percent. Unless the donor tracks the implementation of the project closely, it can be nearly impossible for the donor to know whether the reported number of vaccinations occurred.
• Second, recipients can also take dangerous shortcuts to reduce costs, such as using cheap materials for building a bridge, and pocket the excess money from the payment. Preventing these shortcuts would require very close (and costly) donor monitoring of the projects, thus nullifying one of PbR’s main advantages.
• Third, in many PbR agreements payments are made at certain intervals (once they have achieved measurement X, the first payment is made; once they have reached measurement Y, the second payment is made), which allows the recipient to embezzle the funds after the first payment in the same way that they could with an ex ante payment.
• Fourth, and most importantly, PbR only ensures lower corruption with regards to the handling of aid money; it does not address widespread corruption within countries. Whereas other innovations such as performance-based lending or aid conditionality sought to combat overall corruption in recipient countries, PbR largely ignores this, and perhaps even acquiesces in its existence. Under PbR, recipients control how they achieve the desired outcome, allowing them to use corrupt means, such as bribery or other corrupt practices, to achieve an end that they will then be paid for.

Ultimately, PbR can safeguard the mismanagement of development aid to a certain extent, but it does little to reduce corruption and promote accountability within the country itself. That is not to say that PbR is a failed innovation. On the contrary, it is a useful advancement for those development projects with measurable and verifiable outcomes and recipients willing to assume the risk. However, donors should be wary about overusing the PbR method and recognize that this hands-off approach to projects may perpetuate corrupt practices on the ground and is unlikely to contribute to systemic change.