Millions of people owe their lives to the work that INGOs do â a gigantic heavyweight of often-hidden goodness in the world. Through these vital organisations, total strangers on all sides of the world help each other. They are a breathtaking expression of the best of what it is to be human.
The Scale of Fraud and Corruption in the Sector
While the impact of fraud and corruption goes far beyond financial loss, it would provide a helpful proxy for the scale of the problem. Problematically, however, empirical research is relatively scarce and despite efforts to improve transparency amongst INGOs, most still do not publish their annual detected losses to fraud and corruption. Of the data available, some of the most comprehensive was research by the Thomson Reuters Foundation, which found that 12 of the largest humanitarian organisations reported losses between 2009 and 2014 averaging 0.03 per cent of turnover (Esslemont, 2015). Official Development Assistance (ODA) figures might be in a similar range; the UKâs Department for International Development (DFID) reported in 2011 that it had lost 0.016 per cent (BBC, 2011), while the Australian Agency for International Development (2013) reported âpotentialâ losses of 0.02 per cent for 2011â2012.
At first glance, these figures seem encouragingly low. But all may not be as it seems. A comparison of these losses against those experienced by private sector bodies operating in similar contexts yields some interesting results. A survey by the security firm Kroll (2013) found that companies reported an average detected loss of 1.4 per cent (rising to 2.4 per cent in Africa), nearly 50 times greater than the INGO figure given above! Meanwhile, World Bank surveys (2005â2014) report estimated organisational losses to crime (as a proportion of annual sales, and including vandalism and arson with theft and robbery) of 9 per cent in the Middle East and North Africa, 7.8 per cent in Sub-Saharan Africa and 7.6 per cent in South Asia.
It is perhaps, then, no surprise that when the DFIDâs figure was put before the British Parliamentâs Public Accounts Committee, its members described it as âunbelievablyâ low (BBC, 2011). The sharp difference between the available humanitarian and global development sector statistics and those from the private sector begs the following question: does this mean that INGOs are better at preventing fraud and corruption or worse at detecting it? When we consider that a 2009 Fraud Advisory Panel (FAP) survey of UK charities found that a quarter of respondents thought their sector was at higher risk than other sectors (a proportion that rose amongst victims), that a US study found similarities in losses per incident between non-profit organisations and private companies (Greenlee et al., 2007), and our contention in Chapter 2 that INGO staff are not more honest than staff in other sectors, the answer seems to lean towards under-detection.
There is indeed a difference between detected loss and actual loss. Detected loss represents the total loss that an INGO identifies and attributes to fraud and corruption, while actual loss represents the true total of such losses, whether they were identified or not. This figure is notoriously difficult to reliably quantify â after all, fraud and corruption are designed to hide, so we should expect quantification to be difficult. Research into the difference between actual and detected loss has not been reassuring â some preliminary findings have even suggested that as little as a thirtieth of fraud is detected at best (University of Portsmouth/BDO LLP, 2013).
In Chapter 7 we explore some of the challenges INGOs face in detecting fraud and corruption. We consider how loud âbackground noiseâ obfuscates the indicators of fraud and corruption, how powerful data-driven proactive methods used by governments and private companies may be slow to make the leap to the sector, and how a tendency to operate at a level of stretch reduces the resources devoted to detective activity. Crucially, and perhaps most importantly, we describe several factors that could reduce the likelihood of âtip-offsâ from staff, volunteers or third parties. These challenges, we would suggest, support the view that the sector is under-detecting.
So, in the absence of sufficient research, widespread non-transparency about aggregate losses and probable under-detection, can we estimate the extent of fraud in INGOs? Three particular estimates may help us to do so. The first was an annual analysis of the British third sector by the UKâs now-defunct National Fraud Authority (NFA), which included INGOs. The NFAâs 2011 findings seemed to reflect the Kroll study â that these organisations experienced an average loss of 2.4 per cent of their income (revised to 1.7 per cent the following year). The second is the two-yearly survey conducted by the Association of Certified Fraud Examiners (ACFE), which in 2014 estimated a loss of 5 per cent of revenue for the average organisation. The third is a joint report from the University of Portsmouth and the professional services firm PKF Littlejohn in 2015, which used a representative sampling methodology to estimate that fraud and error losses were usually in the range of 3â10 per cent, probably around the average of 5.6 per cent.
The definitions and methodologies used in these sources vary, of course, and they are not perfect for our purposes. But they all seem broadly in line with Michael Comerâs suggestion decades ago that reliable estimates of âaverage companyâ losses to fraud were about 2â5 per cent of turnover (1977). The experience of my colleagues and I is that INGO sector finance staff, auditors and counter-fraud professionals find this range to be readily believable.
Enablers: The Likelihood of Fraud and Corruption in the Sector
An enabler allows a threat to become a reality. It does not necessarily drive it to do so, nor does its existence guarantee that it will, but it is the archetypal âopen goalâ. A range of these affect INGOs operating in the developing world and they are dynamic â consequences of the shifting landscape around them. In this section, we explore some of the most significant enablers, and perhaps suggest that with such factors in place, the risk must be high.
Cultures of Trust
Trust is important for INGOs, which often work in locations where intense monitoring is difficult. Trust helps to facilitate their operations. INGOs share this tendency with the wider non-profit sector, but it comes at a price. Some have argued that trust-based cultures in non-profits are amongst the factors that make fraud easier to conduct (Douglas and Mills, 2000), while a key theme in the FAPâs 2009 survey was that charitiesâ presumption of trust made them vulnerable to fraud. But if, as we argue in Chapter 2, INGO workers are unlikely to be any more honest than workers in other sectors, should we be so trusting?
Hierarchies of Values
INGO workers have to make choices, often in very difficult circumstances, and this can lead to the stratification of values. Well-meaning people may not adequately prioritise deterring, preventing, detecting or responding to fraud and corruption. The starkest example is the aid worker who slips a border policeman a small payment to allow their trucks of medical equipment to pass freely on their way to beneficiaries. But another might be where organisational cultures value innovation, empowerment and delivery, and are suspicious of regulation. Internal controls might be seen as restrictive, whereas finding inventive ways to deliver outcomes in complex scenarios is praised. Other cultures might be action- or goal-orientated, especially those found in INGOs that work in humanitarian emergencies. These cultures can sometimes risk divorcing those urgent humanitarian aims from the compliance, risk management and reflective learning that actually support them. While both innovation and humanitarian aims are vital, a failure to follow and develop clear and effective policies, procedures and systems for the management of people, assets, funds and stock contributes to the ideal conditions for fraud and corruption.
Governance
INGOs vary in terms of models of governance, but a common theme can be that the necessities of working in fragile or under-developed environments may lead to substantial devolved responsibility. While this level of empowerment â particularly in the concentration of decision-making at the country level â makes an INGO programmatically nimble, ramifications can arise if a corresponding level of accountability is not embedded. Country management can become an accountability âbottleneckâ, for example, controlling the flow of information and divorcing the business unit from wider organisational culture. Business units may then develop in the image of their management rather than within the ethics, values and desired behaviours of the organisation. Similarly, if highly empowered managers are not incentivised properly, they may de-prioritise the effective holistic management of fraud and corruption in favour of the business objectives upon which they are measured.
Dynamism
The prevention of fraud and corruption thrives within stability. Where operating locations, ways of working, themes and modes of programming, and organisational priorities are stable, it can be easier to plan business, manage risk and bed down internal processes while identifying variance. In Chapter 6, however, we explore how INGOs live in a highly dynamic world. In particular, we note how variance in income (and, in the case of institutional donors, their varying funding requirements) introduces uncertainty at the outset of the planning and finance cycle, which causes ripples throughout the process. We also consider that there can be rapid change in political, economic, social, technological, environmental, legal and organisational (PESTELO) factors affecting INGOs. The consequence of this constant state of change is a reduced ability to prevent and detect fraud and corruption.
Operating Environments
The anti-corruption organisation Transparency International (TI) conducts annual research known as the Corruption Perceptions Index (CPI). This study ranks countries according to how corrupt their public sectors are perceived to be and has become a respected measure of corruption risk. In the 2014 CPI, the best-performing countries were Denmark, New Zealand and Finland, while the worst-performing countries were Sudan, North Korea and Somalia. That year, most of the DFIDâs country programmes were in countries that appeared in the bottom half of the CPI (DFID, 2015), as were almost all of the top 10 highest recipient countries of net ODA and official aid in 2013 (World Bank, 2015).
Many developing countries do seem to present an elevated risk of fraud and corruption to organisations operating within it. There are a complex set of inter-relating factors behind this, many linked to poverty-related issues that draw INGOs to the developing world in the first place. While we do not intend to infer that the âdeveloped worldâ or the âglobal northâ is corruption-free â it is certainly not â many developing countries do present challenges.
In Chapter 5, we discuss deterrence. States that do not enjoy an effective and reliable national infrastructure, such as transparent and accountable courts and police, may leave organisations operating in them at a disadvantage. Experienced aid workers recount anecdotes about bizarre judicial decisions against INGOs and in favour of local actors. Almost all of Irish Aidâs âkey partnerâ countries in 2014 (Department of Foreign Affairs and Trade, 2015) were ones in which respondents to TIâs 2013 Global Corruption Barometer (GCB) said that the police or judiciary were the most corrupt national institutions of all. A consequence of this might be that INGOs are known to stand a lower chance of obtaining criminal justice or redress, that their cases are believed to be vulnerable to interference or that they may choose not to seek it. The net effect is the same â a clear message to potential perpetrators that the risk to them may be lower than we would hope. And in Chapter 6, we consider how the internal controls so vital to the prevention of fraud and corruption could be affected in some contexts. We describe, for example, how unstable environments may aggravate compliance with controls and how informal markets may affect the provision of documentation like receipts.
In Chapter 8, we look at the management of incident response â how we handle suspected fraud and corruption. Some types â such as kickbacks â affecting INGOs can be difficult for them to prove through internal investigation. This can be compounded if the local police are ineffective, disinterested or politicised. Further, the availability of evidential opportunities such as government registries and databases (company registrations, electoral rolls and so on) may be low, while in some contexts â such as remote programme management â the opportunities to examine documentation, computer drives and conduct interviews may be limited.
Later, in Chapter 10, we consider the role of local culture more closely, a word which we tend to use in an anthropological sense (reserving âcontextâ for the broader local environment in which an INGO operates). We consider that different cultures can have very different perspectives on things like financial transactions and contracts, and describe the need for INGOs to take action to address any tensions. For example, cultures in some places may involve very strong ties and obligations to familial and social networks. Some individuals could find the need to alleviate the financial difficulties of friends or family more compelling than the contents of the INGOâs code of conduct. They may be more likely to empathise with the more readily obvious plight of lower-income households in their networks than with their relatively faceless employers. Further, differing perspectives on matters like transactions can c...