Roger Henke, Chairman of the Board of the Southeast Asia Development Program (SADP), a development grantmaker based in Cambodia, contributes the following guest post (the first in a three-part series):

Compared with media attention to corruption among public officials and corporate interests, corruption in the non-profit sector is virtually ignored (though a recent GAB post on NGO corruption in India is a notable exception). This lack of interest is matched by the absence of sustained substantive debates about the sources of NGO sector corruption and the effectiveness of remedial interventions. My own experience with these issues derives from my involvement with the NGO sector in Cambodia. Corruption within our own house is a regular topic of informal conversation, and also makes it into our periodic sectoral assessments (though often through oblique references to concerns like “weak financial systems” and the “lack of checks and balances”). However, there are no efforts at all to go beyond these anecdotes and self-reported “weaknesses” to gather systematic, externally validated evidence about levels of corruption, let alone about issues like costs of corruption or the way it correlates with characteristics of the NGO sector that would offer entry points for positive change.

Given the comparative importance of development aid channeled through the NGO sector in countries like Cambodia, this lack of attention to NGO corruption is unfortunate. Admittedly, gathering information on local NGO (LNGO) corruption is challenging. Yet there are potentially useful sources of information that have not been exploited. For example, LNGOs are funded by grantmakers, and these grantmakers (often criticized by LNGOs for their cumbersome administrative requirements and time-consuming monitoring visits) are a possible source of data about LNGO fraud and its correlates. Additionally, the audit firms with an LNGO client base are another possible source of information.

In 2014, to test the willingness of grantmakers and audit firms to share information on their LNGO partners and NGO client base, we at SADP piloted a grantmaker and audit firm survey. The results were promising enough to repeat and expand the exercise in 2015. In this second grantmaker survey, 18 out of 26 grantmakers approached agreed to participate, and 13 of those 18 shared LNGO partner-level information (for a sample of 93 LNGOs). The grantmaker survey queried incidence and seriousness of (1) financial management problems, (2) governance problems, and (3) fraud. (In order to maximize participation, the survey prioritized brevity and simplicity over depth of information.) The audit firm survey (in which four of the five firms approached agreed to participate) asked only for some aggregate data (total number of LNGO audited, number of audits that identified fraud, number of audits that flagged serious financial system issues, etc.). Admittedly, neither the sample of grantmakers nor the sample of LNGOs is statistically representative of Cambodia’s NGO sector, but the surveys provide more valid information about corruption in development NGOs in Cambodia than has previously been available. And the quantitative picture emerging from the combination of these two data-sources about the organizational quality of Cambodian LNGOs is both revealing and disheartening. Interested readers should check out the full report; the most important findings are as follows:

• First, the survey provides strong evidence that fraud and misgovernance are serious problems. At least 25% of LNGOs in the sample had serious financial management problems, and at least 25% had serious governance problems. Approximately 15-20% on LNGOs had incidents of fraud.
• Second, the survey provides evidence that financial management and poor governance are indeed conducive to serious fraud. Of the LNGOs that had governance problems, 50% had incidents of fraud. Of the LNGOs that had serious financial management deficiencies, 80% had incidents of fraud. And for those LNGOs that suffered both from poor financial management and bad governance, the probability of fraud was a shocking 90%.
• Third, there are deficiencies in oversight both by grantmakers and by external auditors. In the case of grantmakers, approximately one-third of the fraud incidents at LNGOs went undetected by at least one of the grantmakers providing funding, and collectively grantmakers underreported serious problems at LNGO funding recipients by 35%. As for the external auditors, in 80% of the fraud cases, previous external audits failed to raise adequate red flags. And for 60% of the LNGOs with serious financial management problems, that issue was not raised in any previous external audits.

The overarching conclusion drawn in the survey report is that grantmakers need to apply a much improved and ongoing form of due diligence to their LNGO partners. Of course, in development circles the concept of “upward” accountability opens up a can of worms too big to discuss here; I will leave those larger issues to future discussion.

Given that the 2015 grantmaker survey has shown it is possible both to increase participation and to overcome the “confidentiality” hurdle, this year’s third round is going to attempt to gather additional information about both grantmakers and their LNGO partners. To the extent that we’ll manage to remain on the safe side of what time grantmakers are willing to devote to answer the survey, the resulting dataset is going to allow for a way “thicker” description, thus more nuanced understanding, and hopefully more entry-points for action. We are also exploring possibilities to replicate the survey and its approach in another ‘comparable’ country context. I hope that GAB readers will help us improve our research in this area both by offering criticism of the study summarized above, and by pointing me toward other relevant research, literature, and initiatives.