Our recent report on next generation financing models looks at how global health donors, specifically the Global Fund to Fight AIDS, Tuberculosis and Malaria, can enhance the health impact of grants for health service delivery by tying grant payments to achieved and verified results. Yet there are several ways to condition payments on performance and, as my colleagues and I have previously pointed out, (here, here, here and here), some ways would likely work better than others in any given setting.  Can economic theory suggest specific features of contract designs which would generate more health for the money?  

To answer this question, the Next Generation Financing Models in Global Health working group convened an expert meeting of economists, hosted in April, 2015 by the School of Public Health at the University of California at Berkeley. Han Ye at Boston University and Liam Wren-Lewis at the Paris School of Economics, wrote two path-breaking working papers for this technical meeting that apply the economic theories of regulation and incentives to the contracts between global health donors and service delivery organizations in recipient countries.

The two papers begin with the same basic assumptions about global health donors and their grant recipients. First, both papers view global health donors, such as the Global Fund and the US President’s Emergency Plan for AIDS Relief (PEPFAR), as purchasing agencies with the mandate to purchase health care services on behalf of poor beneficiaries in recipient countries. Second, they both adopt the notion that the contractors from whom donors purchase health care services have monopoly power in the delivery of the services to their beneficiaries and monopsony power in their bargaining relationship with the donor. 

Under these assumptions, the donor’s objective of ensuring the delivery of quality services to beneficiaries at the lowest sustainable cost mirrors the problem of a regulator of a natural monopoly—a problem that has received decades of attention in the economics literature within the fields of mechanism design and the theory of regulation. As we lay out in the final working group report and as I discussed in an earlier blog post, from this perspective, the global health donor is a “principal” and the contractor in the recipient country is the donor’s “agent.” The relationship between the two is characterized by asymmetric information, a situation in which the agent knows more than the principal about the cost of quality service delivery and opportunities for cost reduction.

Han-Ye’s paper provides a broad survey of the incentive mechanisms proposed in the theory of regulation and asks how each mechanism could be adapted by a global health donor to make its contracting procedures more efficient. The paper weighs the pros and cons of each mechanism, and highlights the tradeoff between the power of each mechanism’s incentives and the cost of its information requirements. Some of the mechanisms reviewed are appropriate for a single contract between a donor and a recipient; others envisage a sequence of contracts over several years, each of which is more efficient than the last. The engagement of the Global Fund and other health donors in some of the world’s poorest countries that are home to the highest disease burden is likely to endure for a decade or more into the future. It is therefore in the best interest of both health donors and recipients to adopt contract designs that will enhance efficiencies over the long term.  

Liam Wren-Lewis’s paper starts from the same assumptions, but goes deep rather than wide.  As currently practiced by the Global Fund and other health donors, contracting is essentially input financing with cost-reimbursement. From the agent’s perspective, a cost-reimbursement contract has the advantage of minimizing its financial risk. Given the bargaining power of agents, the challenge is to find a contract design that applies incentives towards improved value for the donor’s money, while also being acceptable to the agent. Wren-Lewis’s suggested solution is for the health donor to offer a contract that would give its agent the option to choose, at the end of the contract period, whether to receive a conventional reimbursement of costs from the previous period or to receive a payment proportional to the number of verified units of output it had produced (up to a target output). Of course, a contract with only two payment options is the simplest version of an elaborate multi-choice menu contract in the mechanism design literature. Despite its simplicity, however, Wren-Lewis cites literature showing that this design can achieve up to 80 percent of the efficiency gains seen with more complex contracts. By using familiar elements like a cost-reimbursement contract to develop a contract design that enhances efficiency, Wren-Lewis offers the global health donor an incremental path toward more powerful contracting. 

While Ye and Wren-Lewis’ papers were written under the auspices of our working group, their findings have wider implications for improving the efficiency of donor and government spending. For example, the contract designs they discuss could be used to structure the contractual relationship between a national government and its subnational entities such as states, provinces, or nongovernmental organizations that deliver health services. The lessons could also extend to other sectors, such as education or water and sanitation, where foreign aid funders are searching for ways to improve the efficiency of the spending they channel through government and non-government recipients.

I see these two papers as contributions to the broader literature on results-based financing, including cash on delivery models. They offer new ideas about how to shift from cost-reimbursement, or input financing, to payments for verified outcomes to improve value for money. In particular, I hope they spark the growth of a literature on the application of efficiency-enhancing mechanism design to the global health sphere.