Ideas to Action:

Independent research for global prosperity


US Development Policy


While recent aid transparency buzz has largely revolved around the latest donor rankings, MCC (always a top ranked donor) has been busy quietly raising the transparency bar yet again. The latest display of commendable openness? A concise report on closeout economic rate of return (ERR) for 94 projects in 10 compacts, as well as compiled data on original ERRs for 45 projects in 11 open compacts. It’s a short, accessible report, jam-packed with information, and well worth a read. It’s also one-of-a-kind. You just don’t find other donors systematically tracking (not to mention publishing) rates of return of closed projects in a comprehensive way like this. 

One of MCC’s distinguishing features is its commitment to financing only those projects that are expected to raise local incomes by more than the cost of project implementation. To identify such projects, MCC undertakes cost-benefit analysis, comparing total project costs with expected increases in income, and then calculates an ERR (the interest rate at which the net benefits would equal zero). Higher ERRs mean a higher ratio of benefits to costs. For MCC to consider a project economically justified, the ERR must exceed a threshold of 10%. MCC calculates “original” ERRs at the beginning of the compact with best available estimates of cost and benefit data. “Closeout” ERRs are calculated once the compact ends; at this point, data on benefits—which are analyzed over a 20 year horizon—are still projections, but cost data are known.

Here are what I think are the key takeaways of the report:

MCC’s closeout ERR analysis is fairly comprehensive.

At this point closeout ERRs are available for 61% of closed (non-cancelled) projects and will eventually be released for 83% of all completed (non-cancelled) projects. Sixteen closed projects (representing about 18% of disbursed funds or around $50 million each on average), however, will not have a closeout ERR, and it’s unclear which projects are excluded and why.

For the most part, projects are yielding positive returns…

Most closed projects’ closeout ERRs exceeded MCC’s 10% threshold, and the agency can rightly tout the 16% (weighted) average rate of return. 

…though the agency’s focus on value-for-money results has waned at certain junctures…

For fourteen projects, however (a quarter of the projects reviewed), closeout ERRs fell below the 10% threshold, coming in at a weighted average of 5%. The benefits generated by these projects are expected to be worth less than the $605 million MCC put into them. Only one of these projects (a water/sanitation project in Lesotho) had an original ERR below the 10% threshold; the others ended up with a lower rate of return than originally expected. Why? A number of factors were at play, but the most important of these were cost increases (unsurprising, perhaps, for this cohort of compacts which were hit hard by the 2008 global commodity price spikes) and diminished project scope. Now, even well-planned projects can be subject to things like implementation delays and/or higher than expected costs. But for MCC compacts, whose budget-at-signing and five-year timeline are immutable, this often means having to do less than planned. Indeed, MCC found itself needing to rescope many of its compacts in 2008-2009. Unfortunately, the use of economic analysis to guide these rescoping decisions was uneven, at best. Some decisions to continue or cut projects were based on revised ERRs, but others were not. And even when revised ERRs reflecting new costs and/or new scope were known to be below 10%, MCC sometimes decided to proceed anyway (e.g., Mozambique’s roads). This suggests, as Franck Wiebe and I have argued, that greater and more timely transparency around rescoping decisions—including publishing the data that inform the decision—would boost MCC’s accountability for pursuing cost-efficient investments throughout the lifespan of the project.

…and for certain compacts.

For the most part, projects with low ERRs were fairly small (<$50 million) and distributed across compacts (all but 2 of the 10 compacts reviewed had at least one project with a low ERR). This suggests that, in general, weaker performance is a project-specific problem, not a compact-wide phenomenon. However, three compacts stand out. Large low ERR projects dominated the Lesotho and Mozambique compacts. In Lesotho, two projects worth 44% of the total compact value had below-threshold closeout ERRs. In Mozambique three projects worth 38% of the total compact value had low ERRs. For both of these compacts, MCC proceeded with the investments in question even though mid-course revisions to the ERRs suggested early on that they would lose value. The other standout compact is Indonesia. It’s still an open compact, but it is noteworthy in that it doesn’t have an original ERR for programming worth $383 million (that’s over 10% of the total value of all the open projects reviewed). Most of this comes from the compact’s Green Prosperity Project ($300 million) which was designed as a facility that would identify and fund a series of separate qualified projects. While MCC has said that each selected project must have an acceptable ERR, there was always the question of whether it would be possible, within the five-year timeline, to set up such a sizeable facility, evaluate, and identify enough investments that meet this criteria that then can be implemented within the designated period. The fact that very little money has moved on this project three years into the five-year compact suggests that this has indeed been a challenge. 

Will we see more low closeout ERRs in the future?

It’s a possibility. First of all, original ERRs are below 10% for nine of the 45 open projects (compared to only two of 94 closed projects); they didn’t meet the standard from the outset. Not only that, among closed projects, closeout ERRs tend to be lower than original ERRs almost across the board (for 80% of projects). If this pattern holds for the open projects—whose original ERRs are already lower, on average, than the closed projects’ original ERRs—there is the potential that several investments may ultimately fall short of MCC’s value for money standard. On the other hand, there are a number of reasons to believe the open compacts may not follow the patterns of the closed compacts. The latter were hit particularly hard by global commodity price spikes; MCC policy now requires the consideration of revised ERRs during rescoping (even though decisions are not required to align with the results); and MCC—with years of experience and more and better data under its belt—may have gotten better at estimating original ERRs.

MCC has a new opportunity for learning.

Hopefully MCC views this ERR analysis as a valuable input into future project identification and design. Understanding some of the characteristics (sector, size, country, etc.) associated with the projects with weaker rates of return can help the agency think through how to approach similar projects in the future. 


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.