Can foreign aid be used to spur anticorruption reforms? Many donor agencies have tried. The typical approach is to make aid to a recipient country conditional on the adoption of a series of substantive anticorruption or good governance reforms. Unfortunately, there is little data to suggest conditional aid buys reform. To the contrary, grants of conditional aid have been associated with increases in corruption, slower policy reform, and the deterioration of governance generally. While one might expect that, all else equal, conditional aid would result in relatively more aid flowing to more honest governments, it seems the opposite is true: after controlling for a country’s poverty level, regime type, and other factors, it appears that more aid goes to more corrupt countries.

Twenty years ago, a small U.S. federal agency, the Millennium Challenge Corporation (MCC), took a different approach to spurring anticorruption reform through foreign aid. The MCC, which provides large ($100M+) grants to low-income countries, embraced a strategy that differed from traditional aid conditionality in two ways. First, rather than selecting aid recipients on an ad hoc, case-by-case basis, the MCC determines eligibility using a uniform scorecard. As relevant here, the MCC requires that, to become eligible for MCC grants, a country must score above an absolute level on the World Bank Institute’s “control of corruption” index. (Countries must also score above the median for their income class on ten of twenty additional indicators.) The MCC provides grants to most countries that do meet those criteria. (Of the 80 countries are eligible under this scheme, at least 50 have received funding.) Second, and relatedly, once countries are deemed eligible, no further conditions are attached to MCC funding, which can be directed towards any purpose and is rarely withdrawn. On average, countries receive $160M in unconditional funding, though grants have been as large as $698M.

At the time the MCC was created, this approach was labelled “crude and dogmatic.” Critics complained that the MCC approach would divert aid away from the countries in greatest need of both aid and reform, and towards countries that already outperformed their peers. But the evidence strongly suggests the MCC’s approach has spurred meaningful anticorruption reforms, at least among countries near its eligibility threshold. Researchers have compared countries are right above the threshold to others right below the threshold, and found that up to 38% of countries just below the threshold have implemented substantive anticorruption reforms as a result of MCC’s creation (see here and here). Analysis of statements and correspondence with officials from MCC candidate countries (from, for example, leaked embassy cables, meeting transcripts, and the like) provides corroborating evidence that countries near the threshold utilized the scorecard to galvanize reform.

Why has the MCC’s performance-based approach been more successful in catalyzing anticorruption reform than traditional conditional aid? It’s impossible to say for sure, but the research to date suggests a few intriguing hypotheses:

• First, the rigidity of the MCC’s approach, which might seem like a disadvantage from the perspective of tailoring aid to the specific circumstances of each country, may be desirable precisely because it curtails the agency’s discretion. Even strong commitments to anticorruption reform are often drowned out by foreign policy interests, budget maximization goals, lobbying by domestic interest groups, and other factors. Though crude, MCC’s approach is harder to manipulate. The rigidity of MCC’s criteria means that more aid flows to more effective governments: while the top ten recipients of all US development aid are, on average, in 13^th percentile for control of corruption, the top ten recipients of MCC grants score in the 41^st percentile.
• Second, the rigidity of the MCC’s approach makes the conditions for aid more credible. A lack of credibility has traditionally been one of the main flaws of conditional aid. Once a grant of conditional aid has been made, the expectation becomes that aid will be disbursed—even if the conditions aren’t met. Once a grant has been announced, groups that may benefit from aid begin to lobby for modifications to the conditions, since it’s often easier to convince officials at a donor agency to move the goalposts than it is to get a recipient country to actually enact painful and controversial reforms. For example, to secure conditional aid packages, Kenya agreed to implement anticorruption reforms in its agricultural sector four times over a 15 year period, but backtracked each time after receiving aid. And Kenya is not unusual: A study of 200 conditional aid programs found “no link between a country’s reform effort, or fulfillment of ‘conditionality,’ and the disbursement rate” of aid. Because the MCC requires countries to improve their (perceived) anticorruption performance before they are eligible for grants, the condition is more meaningful. And there is evidence that this matters. Sierra Leone, for example, consulted with MCC and committed to an anticorruption plan in 2008, but didn’t receive any funding commitments from the agency until 2015, when, perhaps as a result of its plan, the country saw significant improvements its control-of-corruption index score.
• Third, unlike conditional aid programs, MCC’s approach doesn’t impose specific anticorruption reforms, but rather leaves it up to the would-be recipients to achieve a certain minimum score on the control-of-corruption index, through whatever means they choose. This has empowered local politicians and bureaucrats to implement the reforms they see as most promising. Countries near the threshold for MCC eligibility have formed Presidential taskforces responsible for designing, implementing, and tracking the reforms necessary to meet threshold indicators, and these teams have had an outsized impact on the implementation of reforms. It’s not hard to see why: Local leaders are best positioned to know which reforms will be effective and, with greater ownership, become invested in seeing them through to completion.

The MCC’s inflexibility does have costs. The MCC can’t always reach the countries most in need of reform, it can’t respond quickly to emergent crises, and it can’t tailor its programs to the concerns of domestic constituencies. But appealing as flexibility may seem, creating a credible incentive for reform requires a commitment to principles that outlast political exigencies. Only then can aid agencies escape the trap of empty promises and catalyze real performance.