Hating on the MCC’s Corruption Hard Hurdle

March 17, 2014

This is a joint post with Charles Kenny.

Imagine your employer mandated you wear a watch that operated with a ten-minute window of error. You would always know the time reasonably well but would have to count on the fact that you might be 10 minutes early, 10 minutes late, or right on time.

Now imagine if you got fired for being a minute late. You’d probably ask for a different watch.  Or at a minimum, you’d request that your company not hold you to such a rigid target when they’d given you such a faulty piece of equipment.

That imprecise watch and its associated strict deadline offer an idea of what’s happening when the MCC says that, to be eligible for a compact, candidate countries must score in the top half of their income group on a measure of corruption. The MCC uses the Worldwide Governance Indicator’s control of corruption measure to make that calculation, but the people who construct that measure point out that it includes a degree of error. 

In the best of circumstances, the indicator is a dim reflection of true levels of corruption.  According to the control of corruption measure, countries could appear to be failing (a minute late in our analogy) when a more accurate measure would suggest they should have a passing grade (they’re actually five minutes early).

Our new analysis Hating on the Hurdle explores the MCC’s use of a hard hurdle for its control of corruption indicator and finds that this strict interpretation – a country must be above the median on the corruption indicator to be considered for eligibility – is doing a disservice to the MCC and its partner countries.

The report finds that the control of corruption indicator is, at best, a fuzzy measure of actual instances of corruption in given country; good performance on the indicator does not necessarily lead to development success, or even good use of aid resources; and it is incredibly difficult for countries to improve their score on the indicator despite changes to governance and policies.

Given these uncertainties with the control of corruption indicator, we recommend that the MCC take the following steps:

  1. Drop the hard hurdle. The MCC should, of course, have a corruption indicator as a part of its scorecard, but it should not be subjected to a hard hurdle.
  2. To the extent possible, base the MCC eligibility exercise on indicators that both respond to action and measure what they purport to measure. Potential measures could include sector-specific indicators related to a reduced impact of corruption, such as percentage of electricity generated that is paid for or surveyed bribes for health, police, and local government services.
  3. Demonstrate the seriousness with which the MCC takes corruption by greater use of country-specific, actionable, general-governance indicators based on factors such as membership in the Extractive Industries Transparency Initiative (where appropriate), meeting Open Government Partnership commitments, or publishing budget details.
  4. Make the prevalence of corruption measurement across multiple current indicators explicit. All of the MCC’s scorecard indicators across the three categories are linked to corruption—some explicitly so—and countries are likely to score worse on them if corruption is acting as a serious constraint on development. In other words, the MCC is capturing levels of corruption in a country through more than just the control of corruption indicator.

And lest you think this is just a fun thought exercise on the limits of data, getting MCC use of the corruption indicator right has serious real-world consequences. The MCC board of directors recently elected to revoke compact eligibility for Sierra Leone and Benin, two countries in compact development, for failing to pass the control of corruption indicator. Benin was the median score and Sierra Leone was just behind it – in effect the best “failers.”

A country’s MCC eligibility status should not hinge solely upon an indicator that does not fully measure what it purports to measure and expressly contains levels of measurement error. It may take some time for the MCC to uncover a better corruption measure with wide country coverage and regular updates. But in the meantime, the MCC should consider doing what it can do with the stroke of a pen: dropping the corruption hard hurdle.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.