Corruption and Global Health: Summary of a Policy Roundtable

Roundtable Summary

Question 1: What is the Magnitude of the Problem Corruption Represents for Global Health?

A key question for roundtable participants was: what is the best estimate for how much loss is incurred through corruption in the health sector? Participants felt this question is exceedingly difficult to answer, primarily because measuring losses to corruption is a challenge. First, there is no standard definition of corruption. Corrupt activities are also hard to identify, track, and verify in many cases. In addition, current measurement tools and approaches are often applied in only a limited way, and no single tool or approach can give a comprehensive assessment of monetary losses from corruption or the ultimate impact that corruption has on population health.  Therefore, participants agreed it is a nearly impossible task to accurately estimate losses from corruption in the global health sector overall.

Difficulties in Defining and Identifying Corruption

According to participants, there is no widely accepted standard definition of corruption, but many found Transparency International’s description helpful, albeit very general: “the abuse of entrusted power for private gain.”¹  The United Nations Convention against Corruption (UNCAC), the most prominent international treaty and policy statement on corruption and combatting it, does not define corruption per se, but instead describes various types of corruption including: bribery, embezzlement, theft, fraud, and others.²  Box 1 provides a list and short definitions for commonly used categories of corruption.

Box 1: Types of Corruption

Bribery: offering, promising, giving, accepting or soliciting of an advantage as an inducement for action that is illegal, unethical or a breach of trust.Collusion: A secret agreement between parties to conspire to commit actions aimed to deceive or commit fraud with the objective of illicit financial gain. Extortion: utilizing, directly or indirectly, one’s access to a position of power to demand unmerited cooperation or compensation as a result of coercive threats. Embezzlement: Act of dishonestly and illegally appropriating, using or trafficking the funds and goods that office holders have been entrusted with, for personal enrichment or gain. Fraud: Act of an office holder intentionally deceiving someone in order to gain an unfair or illegal advantage (financial, political, or otherwise). Favoritism/nepotism: favorable treatment of friends and associates in the distribution of resources and positions, regardless of their objective qualifications and merit. “Petty”/administrative: lower-level corruption often involving bureaucrats who control access to public services, who demand bribes before performing public duties. “Grand”/political: major embezzlement or exchange of resources such as bribes for advantages among elites at high levels of government and private industry, often associated with budgeting, position buying, investments, and large infrastructure/construction projects. State capture: when laws, policies or state institutions meant to benefit the public good have been “captured” through bribes or other means by individuals in order to foster political or personal economic interests.

SOURCE: Adapted from Hussmann K (2011). Addressing corruption in the health sector. U4 Anti-Corruption Resource Centre. http://www.u4.no/publications/addressing-corruption-in-the-health-sector-securing-equitable-access-to-health-care-for-everyone/.

Participants stated that corrupt acts are often hidden from view, making them difficult to track, study, and verify.  They pointed out that the line between corruption and inefficiency is often blurry and it can sometimes be difficult to tell the difference between corrupt acts and poor management.  Another difficulty is that there are cultural and sociological differences regarding what is corrupt vs. acceptable practice. What is considered “corruption” can vary from country to country and sector to sector. While some acts – such as outright fraudulent misdirection of funds intended for health services – are likely to be almost universally considered corrupt, in other cases – such as when gifts and money are given to health providers – it can be difficult to agree on whether a specific act constitutes corruption or not.

Therefore, participants agreed that because it is often so difficult to find and prove corruption, it remains a serious challenge to measure and quantify it.  In fact, many felt it would not be worth the effort, and might not even be possible, to comprehensively evaluate the overall impact of corruption on global health due to these limitations.

Measurement and Evaluation Tools Can Shed Light in Specific Areas

While recognizing the difficulty of estimating the total impact of corruption on the health sector as a whole, participants did state that measurement and evaluation of corruption within specific components of health systems or focused on specific health programs or activities can be helpful to understand where and how corruption occurs. In this context, participants mentioned a number of tools and approaches.

Participants stated that some of the most commonly referenced corruption measures are based on surveys asking people about their opinions and experiences with corruption, such as Transparency International’s Corruption Perception Index or AfroBarometer polls.^3,4 For example, opinion surveys of the public in 7 countries performed between 2001 and 2006 found the percent of the public who believed the health system in their country was “corrupt” or “very corrupt” ranged from 20% in Indonesia to 56% in Sierra Leone.⁵ While estimates such as these can point to general levels of concern about corruption in the health sector, participants agreed that such measures are problematic because perceptions and opinions about corruption can differ from the true prevalence of corrupt practices.

Other approaches to identifying and measuring corruption referenced by participants included: studies of personal experiences with corruption by individuals, households, health workers, and government officials (e.g. how much/how often have they have been party to paying or requesting bribes or other corrupt acts), and public expenditure tracking survey (PETS) for health, which can be used to follow how public funding flows from the point of central disbursement to how it is used to pay for health services in states, counties, and other localities. One approach used by donors is the portfolio review, in which a sample of grants is investigated by an Inspector General or other neutral office.  The World Bank, the Global Fund to Fight AIDS, TB and Malaria, and USAID have all used portfolio reviews as a way to identify fraud and other instances of waste and/or corruption, and measure losses.⁶

While each of these approaches can shed some light on specific areas and instances of corruption, participants felt that none in isolation can fully determine the extent of and losses due to corruption.

Need for Continued, and in Some Cases, Expanded Measurement by Donors

Though participants felt pursuing a broad, overall estimate of losses to corruption in global health presented many difficulties they did agree that corruption in the health sector can be a serious problem, and that global health programs could do more to try to measure and understand it. Participants stated that just because corruption cannot be accurately measured at a global level does not mean such acts do not occur or that they have little impact on global health programs. To give some indication of how corruption manifests itself in health systems, participants pointed to a number of published studies that highlight the kinds of corrupt activities that can take place. Table 1 (next page) provides a selection of data points from published studies on corruption in different components of the health systems of low- and middle-income countries.

Participants felt current donor approaches to assessing the risks and impacts of corruption are often inadequate.  Typically, donors such as the U.S. will audit programs in reaction to a complaint or some prior evidence of fraud or abuse.  They said it is much rarer to see donors engage in proactive, preventative investigations focusing on a randomly selected sample of grants even though this approach could provide a more robust estimate of corruption and waste.  Such a randomized approach has been used for some U.S. domestic health programs such as Medicare to track so-called “improper payments” (includes fraud and abuse as well as other categories of misuse of funds),* but has not been widely utilized for assessing foreign assistance or global health portfolios.

Table 1: Selected Estimates of Corruption in Health from Published Studies
Year(s)	Corruption Estimate
1999	68-77% of health system user fees misappropriated or pocketed in Uganda7
2000	21% of hospital procurements estimated diverted due to corruption in Ghana8
2000-2006	In a study of 6 countries, the percent of patients surveyed reporting having to make informal payments for service ranged from 4% in Benin to 35% in Sierra Leone9
2001-2006	Opinion surveys of the public in 7 countries found the percent of the public who believed the health system was “corrupt” or “very corrupt” ranged from 20% in Indonesia to 56% in Sierra Leone9
2001-2006	Surveys in 6 countries found the percent of health worker absenteeism ranged from 6% in Cameroon to 25% in Peru to 40% in India9
2003	In Thailand, a reported 8.5% of medicines were considered sub-standard10
2005	20% of public officials in Guinea reported job purchasing in the health system is “common” or “very common”9
2011	After reviewing grant portfolios in 25 of 145 recipient countries, the Global Fund’s Office of the Inspector General reports 0.8% of disbursed grants to those countries ($39m of $4.8b total) was lost to fraud11
2012	Across seven countries in Asia, 36% of antimalarial drugs being sold were falsified; in 21 countries in sub-Saharan Africa, 20% were falsified12
2013	An internal review conducted by  Gavi, the Vaccine Alliance (a multilateral financing mechanism for immunizations in developing countries) found that 31% of one grant  ($523m out of $1.685B) to Sierra Leone had been misappropriated13
2013	The percent of patients who reported paying a bribe to obtain health services in 8 countries in central and eastern Europe ranged from 4% in Croatia to 32% in Bosnia-Herzegovina14

Further, participants felt that more studies linking corruption in the health sector and the resulting negative impacts on population health would be helpful. A few studies on this topic have been published,¹⁵ but participants felt that more evidence linking corrupt practices with real human health impacts could be an important way to understand and highlight the human costs of corrupt acts.

Question 2: Are Current Anti-Corruption Efforts Adequate?

According to participants, anti-corruption efforts can take many forms that can range from broad governance and rule of law programs to more sector-specific approaches. Global health-specific anti-corruption programs are not common, especially within the U.S. government’s global health portfolio, but participants stated that successful examples of such efforts do exist and can be instructive. Overall, participants felt more could be done by donors on health sector-specific anti-corruption policies and programs, especially in light of a number of factors that could potentially contribute to greater corruption risks within global health, such as an increasing emphasis on “local ownership” of health programs and the growing need for health programs to work in fragile, conflict and post-conflict states where disease burden is concentrated but governance can be weak.

Approaches to Anti-Corruption in Health

Participants discussed a variety of anti-corruption efforts. Some are broad-based programs that focus on improving the judicial system and supporting the rule of law or building general governance and leadership capacity of a country. Other efforts target certain areas or sectors, such as programs that address specific components of the health system. Some of the types of policies and programs in the health sector participants discussed are listed in Box 2. Participants reported that the two types of efforts – broad and sector-specific – remain isolated for the most part. They recommended fostering a combined approach that utilizes both broad and targeted anti-corruption efforts together in a coordinated fashion.

Box 2: Examples of Anti-Corruption Policies and Programs

Robust internal audit /portfolio review Collecting baseline data (absenteeism rates, informal payments, stockouts) to show that policy changes and interventions work Designing and implementing complaint mechanisms Supporting administrative law Civil society watchdog organizations & social audit; health boards Insurance fraud control (data mining for detection, transparency on entitlements, sanctions) Innovative financing approaches such as results-based aid and performance-based financing Quality improvement/clinical audit Health management systems strengthening & supervision Coordination with efforts to promote better governance in other sectors

SOURCES: adapted from Vian T, Savedoff W, and Mathisen H. (2010). Anticorruption in the Health Sector: Strategies for Transparency and Accountability. Kumarian Press; and Hussmann K (2011). Addressing corruption in the health sector: Securing equitable access to health care for everyone. U4 Anti-corruption Resource Centre Issue Brief, January.

Participants also discussed ways donors could do more to reduce the risks of corruption through changes in their practices. For example, donors can diagnose and track potential waste and corruption through more regular use of portfolio reviews (described above) to examine recipients of global health grants and the trail of funding. Findings from such reviews in the past have led to further investigations and prosecution of corruption. Losses could be recouped and these efforts could have a preventative effect against further infractions. Participants also discussed utilizing innovative financing approaches such as results-based aid and/or performance-based financing.¹⁶ Participants felt that by using such innovative approaches – delivering aid according to outcomes rather than inputs – donors can reduce the risk for abuse and fraud.

Other approaches discussed by participants involve working with country partners to reduce corruption risks in country health systems. For instance, donors could support targeted anti-corruption efforts to improve governance and oversight of health programs and better identify and minimize loss from corruption.  Examples of such approaches include: collecting and sharing baseline data on corruption, instituting quality improvement, automated monitoring and complaint mechanisms, and involving civil society through watchdog groups or community health boards to promote better oversight and accountability.  Participants pointed to several examples of successful health system anti-corruption efforts, some which are listed in Table 2 (next page).

Participants identified some common characteristics of successful anti-corruption programs for health. For example, programs can be more successful if they consist of more than one intervention because singular approaches may only shift risk; a complementary set of policy changes is typically a more effective approach.¹⁷ Participants also stated that it is often invaluable for anti-corruption efforts to take place where there is committed leadership among the implementers and partners because it is difficult to impose successful anti-corruption interventions externally.  Participants felt that successful programs reflect input and participation from patients, clients, and other affected members of the health system and communities, as this increases the transparency of health programs and generates more accountability among leaders and policymakers.

Table 2: Selected Examples of Successful Anti-Corruption Interventions in the Health Sector

Description of Anti-Corruption Program and Impact

User fee revenue theft in provincial referral hospitals in Kenya was virtually eliminated through installation of networked electronic cash registers18

A multi-pronged strategy including overhauling the government drug regulator, stepped up enforcement and a public information campaign led to an 80% reduction in fake drugs in the Nigerian market in 20049

Health Public Expenditure Tracking Surveys (PETS) in Chad and Ghana identified funding flows from central to regional/district level represented the largest risk of leakages, allowing for stricter oversight in areas of weakness and subsequent reduced losses9

Citizen participation in health boards in Bolivia led to significant decreases in illegal overpayment for drugs and supplies, for example a 40% reduction in price for intravenous solution19

Pay for performance mechanisms linking health worker bonuses to facility performance increased quality, boosted utilization, and reduced the incidence of informal payments in Cambodia20

Revised rules and transparency about hospital fees, sharing earned revenue with staff, and other rewards discouraged informal payments in Georgia21

Overall, participants felt anti-corruption efforts such as these could be effective for global health programs given additional, sustained support from donors.

U.S. Government Anti-corruption Efforts and Global Health

Participants indicated that U.S. anti-corruption efforts through foreign assistance tend to concentrate on broader, cross-sectoral approaches to combating corruption. For example, most USAID assistance for anti-corruption is targeted at rule of law, democratization, and governance programs rather than sector-specific programs in health or other areas.²²

Participants did note that U.S. global health assistance is reviewed on a regular basis for fraud, waste, and abuse, typically through audits led by the Office of the Inspector General (OIG) in each of the departments and agencies responsible. For example, the legislation for PEPFAR (the President’s Emergency Plan for AIDS Relief), requires that the departments overseeing the HIV/AIDS, TB, and malaria programs authorized through that legislation submit a coordinated audit plan, and share with Congress the results of those audits. These audits are meant to determine whether funded projects meet stated goals and objectives, though in the course of performing them, auditors may encounter instances of waste, fraud, and abuse. Additional investigations of specific acts of corruption can be triggered by complaints or whistleblowers, though results from such investigations are typically not made public.

Even though participants felt audit and oversight of this kind can be helpful, some expressed concern that the current U.S. approach is ad-hoc and does not allow for a robust assessment of U.S. funding lost to corruption, waste, and/or fraud. A more informative approach, according to participants, would be to perform audits of a randomly selected subset of all grants/projects. This would provide for a more comprehensive picture of losses and point to areas of concern more effectively.²³

Participants felt there are other changes in approach and practice that the U.S. could implement to reduce corruption risks. For one, the U.S., with its annual appropriation process, can try to avoid the situation in which there is pressure to get funds “out the door” at the end of the fiscal year, a practice that can lead to less oversight and, therefore, higher risk of misappropriation or abuse by the recipients of these funds. In addition, donors can ensure country strategic and operational plans explicitly address anti-corruption goals, as they relate to the health sector specifically.²⁴ Currently, such plans only rarely incorporate or even mention anti-corruption objectives. Further, participants felt the U.S. should consider adopting results-based and performance-based approaches wherever possible, which can help reduce corruption by spending on outcomes as opposed to inputs.²⁵

Participants discussed key trends in global health assistance that could have implications for anti-corruption efforts of the U.S. and other donors. The first was the growing emphasis toward shifting USG global health financing away from U.S.-based non-governmental organizations (NGOs) and toward “local ownership” and local organizations.²⁶ Some participants argued that making such a shift increases the risk of corruption because there is potentially less accountability and fewer controls in local organizations as compared to U.S.-based NGOs. Others saw shifting resources in this manner as a way to foster sustainable capacity in countries to combat corruption because, by supporting local organizations, donors can help community stakeholders organize and hold governments accountable.²⁷ Overall, participants felt that as the emphasis on local partnerships grows, U.S. programs must ensure that sufficient safeguards are in place to minimize any risks to losses from corruption.

Another tension discussed by the roundtable participants was balancing burden of disease and governance considerations. Many of the countries where corruption risks are highest – those with weak institutions and those facing conflict and post-conflict situations – often face the greatest burdens of disease. Participants debated whether the U.S. should focus assistance where the need is greatest but governance may be weakest, or whether it was better to direct investments to countries with a track record of good governance. No simple solutions to the conundrum were identified, but participants felt it was important for the U.S. and other donors to consider this tension when developing policy. Participants also stated the U.S. needs to be sensitive to “backsliding” in governance standards as a result of governmental or societal change, which highlights the need for sustained anti-corruption efforts over time.

Question 3: How Can We Communicate About Corruption More Effectively?

Participants agreed that communicating about corruption to the public and policymakers presents a number of challenges. For one, some level of corruption is likely a part of any large-scale global health program, but can be difficult to discuss openly because the public and policymakers have proved to be extremely sensitive about the topic. Many in the public already have a strong belief that corruption represents an enormous drain on foreign assistance programs even though available evidence indicates that losses from corruption do not match perceptions, according to participants. For example, in a 2013 KFF survey, Americans reported on average they thought about half of every dollar the U.S. spends on global health was lost to corruption, and that only 23 cents of every dollar spent actually reaches the people who really need it.²⁸  Roundtable participants felt such estimates represent a tendency of the public to believe the problem is worse than it really is, and can lead to misplaced doubts about the effectiveness of global health assistance and foreign aid in general.

Participants were concerned that entrenched public beliefs about the prevalence and impact of corruption can lead to donors to being overly sensitive about perceptions of corruption occurring in their programs. As was asked during the discussion: how much is the fear of a corruption-fueled headline limiting agencies’ willingness to confront corruption? Participants did recognize that donors sometimes find themselves in difficult situations after corrupt acts are identified, especially when the media sensationalize a story. Participants noted a well-known example of this from 2011, when an Associated Press article on corruption within some Global Fund grants led several Global Fund donors to threaten withholding support even though the corrupt acts were discovered and reported by the Global Fund itself, and the actual amount lost to corruption was much less than was insinuated in the article. As was evidenced by the Global Fund experience, there can be a disproportionate backlash for agencies and organizations when corruption is identified and publicized. Participants worried that this interplay between public perceptions and transparency about corruption has fostered an environment where leaders and program managers commit to “zero tolerance for corruption” policies even though such goals are probably unrealistic. In the current environment, participants worried that donors often wish to avoid even talking about the corruption that is an inevitable part of any health system, or even worse, may turn a blind eye or ignore the problem for fear of potential repercussions.

According to participants, shifting the pervasive negative public opinions in the U.S. and other donor countries about the extent and impact of corruption in global health assistance is a difficult and long-term task. Barring a sea change in public opinion, participants felt that a more proactive, preventative approach to communication about corruption is a better option than not discussing or ignoring it. Participants discussed preliminary research on public messaging about corruption and its impact on public opinion, which indicates that the stigma associated with discussing corruption can be ameliorated by focusing on more active, positive messages about what is being done to combat it rather than waiting until corruption is reported before acting. As an early step, participants said donors can do more to emphasize and highlight successful anti-corruption interventions, as these can lead the public and political leadership to understand that steps are being taken to address the issue.

Still, there was recognition among participants that global health programs may face a “catch-22” situation with complicated and sometimes perverse incentives. If donors and country governments do a better job at identifying corruption, the result may actually be a decline in willingness to finance global health programs in those areas or with those partners, even though the risk of loss to corruption may have been reduced. Ultimately, more needs to be understood about the relationship between implementing greater transparency and accountability interventions and the impact on corruption and public perception. Pilot studies have begun to look at such issues but results are not yet available.²⁹  Participants felt this is an area of study that donors could do more to support given the potential lessons that could be drawn.

Conclusions and Next Steps

The roundtable discussion was wide-ranging, covering many aspects of corruption and global health. Participants were in agreement that it is not really possible to know what percentage of financing for global health is lost to corruption, given data limitations and the challenge of measuring corrupt activities. Even so, participants felt corruption does occur and can dilute the impact of global health financing, and current donor approaches to assessing risks and impacts of corruption are often inadequate. Current anti-corruption programs, particularly those funded by U.S. government agencies, tend to focus on broader reform of the justice system, rule of law and governance. Expanding health sector specific anti-corruption efforts could complement these important efforts especially given that there have been some examples of successful anti-corruption programs in health. On the question about communicating about corruption, participants recognized the potential difficulties in being more transparent and direct about corruption risks and efforts to reduce them in the context of global health programs, but also felt that a more proactive approach may be a better strategy than trying to avoid talking about the problem for fear of backlash.

Although participants felt that solutions will differ by location and circumstances, and no “one-size fits all” approach can work everywhere, there are a number of concrete steps that donors, such as the U.S., could take to better address corruption in global health. Some of these steps include:

• Invest in research on further measuring and understanding corruption in the health sector;
• Dedicate more resources to investigate and understand the true risks of corruption through such tools as portfolio review applied to a broader, random sample of grants and sector-specific studies in coordination with in-country partners;
• Incorporate explicit anti-corruption policies and interventions into U.S. and partners’ global health strategic and operational plans;
• Link global health programs with broader rule of law and governance reform efforts whenever possible;
• Implement and expand anti-corruption interventions in health, building upon already proven models, which should be adapted to local circumstances;
• Promote anti-corruption efforts and highlight successes.

Ultimately, participants felt that it is important for the U.S. and other donors to dedicate more funding and time to understanding and combatting corruption in global health. While they felt the problem of corruption in health may not rise to the level of concern expressed by some policymakers and by the public at large, it is still an issue that needs to be confronted more effectively going forward.