Testimony to the US House Oversight and Accountability Subcommittee Hearing

On Thursday, March 21, 2024, CGD Senior Fellow Charles Kenny appeared before the Subcommittee on National Security, the Border, and Foreign Affairs of the House Committee on Oversight and Accountability at a hearing titled “Accountable Assistance: Reviewing Controls to Prevent Mismanagement of Foreign Aid.”



Chairman Grothman, Ranking Member Garcia, members of the subcommittee, thank you for the opportunity to appear today. I want to make two points: first, corruption and financial mismanagement is a real problem globally—with negative implications for both development and foreign assistance, but it does not stop that assistance having a major impact. Second, reducing upstream bureaucracy and focusing on impact through transparency and tracking results is the best way to limit malfeasance and mismanagement while delivering more effective aid.

Every year, as many as 1.6 billion people worldwide report they pay bribes for government services. The global total of bribe payments may be as high as two percent of GDP.[1] Beyond corruption, considerable government finance is wasted on projects that don’t deliver what they promised due to factors ranging from incompetence to fraud. And US government spending, including US foreign assistance, is not immune from these problems.

That said, US aid is some of the most closely monitored of all government spending, and that monitoring suggests that while malfeasance can occur, it is comparatively rare. Let me provide some examples:

  • Between January 1, 2012, and December 31, 2020, the Special Inspector General for Afghan Reconstruction (SIGAR) issued 176 financial audit reports, covering $8.5 billion in costs incurred on awards funded by DOD, State, USAID, and USDA for Afghanistan reconstruction projects. Of that $8.5 billion, SIGAR questioned $494 million—or about six percent—based on apparent insufficient supporting documentation or noncompliance with laws and regulations. As of December 31, 2020, agency contracting officers or authorizing officials had made a determination on $417 million of the $494 million in questioned costs and disallowed $28 million, or about seven percent of them. Put the two together and you get an audit process which suggests about 0.4 percent of audited costs are eventually disallowed. It is not nothing, and it is probably less than actual misused funds. But it is also not a percentage large enough to suggest diversion destined aid to fail. That $28 million over nine years—about $3 million a year—is perhaps also worth comparing to SIGAR’s annual budget in 2019 of $55 million.[2] Even where there were substantial resources dedicated to uncovering malfeasance it just didn’t reveal all that much aid finance going places it shouldn’t in Afghanistan.

  • The World Bank has its own equivalent of the Inspector General model: the Integrity Vice Presidency. Between 2007 and 2012, the World Bank’s internal investigations division found cases of sanctionable fraud or corruption in 157 contracts worth $245 million. The World Bank’s lending volume at the time was about $40 billion a year, so contracts collectively worth about 0.1 percent of volumes over the period involved discovered and sanctionable fraud or corruption.[3]

  • The Ukraine Special Inspectors Generals report of Feb 15th 2024 notes direct budget support to Ukraine funded by USAID is primarily managed by the World Bank and only disbursed to reimburse pre-approved Ukrainian expenditures following verification of expenses. USAID has also engaged an international audit firm to conduct financial audits of direct budget support funds. In addition, KMPG will be performing a full audit of U.S. direct budget support to the Government of Ukraine. And USAID carried out its own spot checks that trace expenditures reported by the Ukrainian government to verify that direct budget support was received by intended beneficiaries: 475 spot checks were carried out without any major issues identified.[4]

  • Gavi, the vaccine alliance, is dedicated to expanding vaccination coverage in developing countries and does that in large part by delivering vaccine doses to them. The US provides annual financing of about $290 million. With colleagues, I estimated that the gap between vaccines delivered to recipient countries by Gavi and vaccines actually used on patients at between 0.5 percent and 15 percent of vaccines delivered, depending on the vaccine.[5] (That compares, for example to an 11 percent vaccine wastage rate for Covid vaccines in the US).[6]

Of course, special inspectors general and integrity departments don’t and never will be able to audit every penny to the last recipient. Whatever the level of control we will never be able to ensure zero waste or fraud. But what we can say is that leakage rates are low –and again, we can say it with more confidence about civilian foreign assistance flows than we can with a lot of other government finance.

But take Afghanistan again: when assistance failed—and surely a lot did, not least when it came to helping the Afghan National Army become an effective fighting force—it was because of problems that don’t appear in financial audit reports. Conversely, a lot of aid that the Special Inspector General singled out for inadequate paperwork managed to have a meaningful positive impact.

Starting in 2002, USAID supported the Afghan Ministry of Health in delivering a basic backage of healthcare to 90 percent of the country at the cost of about $4.50 per person per year. USAID focused on results, and it used independent evaluations to ensure vaccinations happened and services were provided. At one point, SIGAR called for the program to be suspended because the Ministry was keeping inadequate financial documentation, but the results speak for themselves. The proportion of kids who died before their fifth birthday in Afghanistan fell from about 12.5 percent in 2001 to 5.8 percent in 2020. That declining mortality rate is saving more kids each year than the total number born in Nebraska, Nevada, and Kansas combined. Every year in Afghanistan, somewhere in the range of 180,000 parents are spared the pain of burying their young child thanks to progress against child mortality since 2001, and US assistance can take a big part of the credit.[7] That progress is at real risk under the Taliban regime, though a resumption of vaccination programs could help. But even if the fight against premature mortality weakens, Americans should surely be immensely proud of what US assistance achieved in this case.

Repeated evaluations suggest that the President's Emergency Plan for AIDS Relief started under President George W. Bush has saved millions of lives—perhaps as many as 25 million—and that knock-on effects include more affordable HIV treatment for people worldwide and more rapid economic growth in Africa.[8] Those results simply wouldn’t have happened if PEPFAR wasn’t effectively turning dollars into treatment.

Or, turning to multilateral aid, 82 percent of projects financed by the World Bank’s soft lending arm, the International Development Association (IDA) were given a satisfactory or better rating by the Bank Group’s Independent Evaluation Group. The median economic rate of return on IDA projects is 24 percent.[9] And when countries get less foreign aid because their income crosses the threshold to be eligible for support from IDA, their economic growth slows down as a result—less aid equals less growth.[10]

It is these results that matter: that aid is saving lives and promoting economic growth. Getting those results is why we provide assistance in the first place. And while good stewardship of US taxpayer dollars is vital, an excess focus on monitoring receipts and processes can get in the way of achieving those results. Back in 2010, Andrew Natsios, USAID Administrator under President George W. Bush, complained of the ‘counter-bureaucracy’ fighting against aid effectiveness: the Offices of the Inspectors General; the Office of Management and Budget; the Government Accountability Office; the Office of the Director of Foreign Assistance in the State Department; the Special Inspector Generals; the 1,977 pages of the Federal Acquisition Regulations; the Department of Defense‘s regulatory control over all overhead rates for all federal contractors and grantees; and the 450-page Foreign Assistance Act of 1961.

Natsios suggested a third or more of USAID staff were hired purely to deal with compliance issues. And he asked: “are we creating a system where every taxpayer dollar is accounted for that is incapable of carrying out its national security tasks?”[11]

Natsios also blamed the counter-bureaucracy for leaving USAID dependent on Washington contractors to implement programs because they were the only ones who understood all the requirements involved. Repeated evaluations have supported the international consensus that the most effective assistance backs projects that are chosen locally and managed locally.[12] Sadly a lot of US assistance fails both tests of locally prioritized and locally delivered because of the weight of the counter-bureaucracy.

Despite the fact that increasing local delivery of USAID-funded projects has been an explicit goal of the last three administrations, we still have a long way to go. Of US assistance contracted to firms and non-profits worldwide in 2020, four fifths still went to US firms. [13] Only about one in ten dollars went to institutions in recipient countries.[14] And the Washington Metropolitan Area Transit Authority was responsible for implementing more US foreign assistance than the government of Uganda was in 2020—and Uganda was near the top of the list of recipient government implementors in that year.[15]

We focus heavily on monitoring cash flows and processes rather than impact, and what gets monitored gets done. But we provide assistance in order to have an impact: we should be focusing our monitoring on results, not receipts. We should make sure we got what we paid for by ensuring transparency on who got paid to deliver what, and—most importantly—that it was delivered.

Some of America’s partners have valuable lessons when it comes to transparency and results measurement. Ukraine has a world-leading government electronic procurement system called Prozorro that combines competition with incredible levels of transparency far in advance of US federal procurement systems.[16] When it was introduced, Prozorro led to a dramatic increase in the proportion of government procurement that was conducted competitively alongside cost savings and reduced contracting times.[17] It is great to see that the Ukraine government has extended the use of Prozorro to defense purchases this year.[18] Transparency allows third parties to see if finance providers are getting what they paid for. It is a fantastic tool to enlist citizens in recipient countries to track results, and that has been happening in Ukraine.

Meanwhile, Congress has put in place some of the underpinnings for a results-based accountability agenda: 2016’s Foreign Aid Transparency and Accountability Act was an important step in that direction, setting expectations for timeliness, quality and accessibility of data as well as extend monitoring and evaluation alongside performance metrics. The bipartisan Foundations for Evidence-Based Policymaking Act—signed into law in2019—pushed for more data and evaluation.[19] But I would urge you to go further, and ensure more US foreign assistance uses the most effective tools to deliver backed by an accountability system that helps ensure that delivery.

We should spot check if people are being paid when using budget support, as USAID is already doing in Ukraine. And we should be doing more to survey beneficiaries and physically audit infrastructure financed under all aid projects. Where possible, rather than paying for potential outcomes, we should be basing payments on achieved results. Combined with transparency in project delivery targets that allow beneficiaries themselves to act as monitors, these approaches can assure impact.

At the same time, because of the long bipartisan concern over the costs of bureaucracy to delivery in US foreign assistance, I hope Congress will actively support USAID Administrator Power’s effort to reduce red tape.[20] One small additional way that Congress might help: the USAID website contains 40 mandated reports to Congress on various topics in 2023 including the use of social and behavior change communication and an annual tuberculosis report.[21] With respect, it might be worth confirming all such reports are still read and appreciated by the intended audience.

And because of the long bipartisan consensus around the need for more local delivery, I hope Congress will also support the USAID Administrator in pushing the localization agenda, an effort that does appear to be having some impact.[22]

In particular, the experience with support to countries including Ukraine and Jordan shows that government-to-government assistance works. That’s also demonstrated by the strong record of the Millennium Challenge Corporation (MCC), which finances locally prioritized and locally managed assistance. And yet on average, about 2 percent of aid is provided as direct budget support (cash) to foreign governments.[23] USAID government-to-government pilots are demonstrating that the agency can do more.[24] And if USAID expanded the use of systems that pay on the basis of results achieved, it could guarantee impact in its work with governments.[25]

I mentioned MCC’s strong delivery model. At the same time its country selection approach is arbitrary and unfortunate, not least relying on a measure of corruption that its own creators provide good reasons for believing is not fit for purpose.[26] This is an example of where a preemptive reaction to the risk of malfeasance can do more harm than good. The MCC corruption measure is a very weak indicator of general corruption risk in a country and an even weaker indicator of corruption risks to MCC compact success. And yet that measure excludes a number of countries that would considerably benefit from MCC support. Take the MCC scorecard on Ukraine –the country passes 14 out of 20 indicators, and passes on democratic rights, but despite its control of corruption score improving in recent years, it is still just below the median.[27] For that reason, Ukraine is denied support from MCC. But this says considerably more about weaknesses in a measure based on dated and inaccurate perceptions than it does about corruption in Ukraine.[28] Congress should encourage the MCC to reconsider the hard hurdle on corruption rather than pushing for authorization for MCC to work in upper-middle-income countries where each dollar of assistance will have less impact.[29]

Moving from an approach dominated by arbitrary screening criteria and tracking receipts and processes toward one that emphasizes transparency in finance delivery with close monitoring of outputs and results will ensure the US gets what it pays for from development assistance, with minimal leakage and maximum impact. Thank you.


[1] Kenny, Charles (2017) Results Not Receipts: Counting the Right Things in Aid and Corruption Washington: CGD

Rights & Permissions

You may use and disseminate CGD’s publications under these conditions.