When development finance institutions (DFIs) use subsidies to support private firms in developing countries, they fundamentally change the nature of their business. They transition from using market-based tools to help markets work better to using government grants to back favored firms.That change requires a different operational model. To ensure the maximum development impact of scarce aid resources, subsidies should be competitive wherever possible, capped if not competitive, and transparent in every case.
Subsidized finance should be competitive not least because creating a competitive process can help ensure government subsidies are focused on a public policy priority (the subject of the competition). But it also helps ensure the most efficient use of subsidies. A competitive approach is based on selecting the firm which will provide the development outcome (loans to underserved customers, power generation or distribution, as it might be) at the lowest subsidy rate. It uses the least amount of aid resources to deliver the desired investment project. And the process ensures “additionality”—that the subsidized finance from the DFI is actually needed to deliver the project. (If it isn’t, the winning firm will bid for the project on the basis of zero subsidy.) Absent a subsidy competition, client firms have every incentive to suggest they need the largest possible subsidy to complete a given investment, while DFI staff lack the information to know if that is true—and face strong incentives to complete deals. That results in public aid resources being turned into private rents.
For that reason, if subsidized projects are not competitive, it is vital that subsidies are capped. That at least will limit the scale of excess subsidy payments. Preferably, subsidies should be capped at a level low enough that the value of the subsidy to the firm is smaller than the cost of compliance to DFI processes, including environmental and social conditions. That would mean firms only take DFI finance if money is not available from the market. In turn, that ensures the project is additional—it would not have taken place without the DFI. Additionality really matters. There is far too little private investment in poorer countries. If DFI support merely backs investment that was already going to happen, it has no impact on investment levels.
Any cap higher than a minimal level increases the risk of wasting aid, and should require a strong public justification around project innovation and market-making. This is especially the case because for DFI projects to have impact at scale, they must demonstrate that an activity is sustainable and profitable in a country. The more that the project is subsidized, the less likely that the demonstration effect kicks in—other firms will reasonably assume the project was only possible because of the subsidy.
Finally, DFI projects should be transparent not least because more information on projects will increase any demonstration effect. Transparency also helps inform potential clients of what they can expect from working with a DFI, which will help increase the pipeline of quality projects. And because DFIs are using public finance, taxpayers have a right to know how that finance is being used, especially in the case of noncompetitive subsidy allocation, which raises legitimate concerns over governance. Transparency should apply to on-lent resources in financial intermediary projects as well.
DFIs can deliver on this agenda, as I discuss here. But if and where they say they can’t, they should be returning subsidy finance to parts of the aid system that can meet these minimal governance conditions for the efficient use of public funds.