What really works in preventing fraud against organisations and do decision-makers really need to know?

Abstract An evidence base of what works using high-quality evaluations in tackling societal problems has become the norm in many spheres, including tackling traditional crime. Yet, as we show in the example of fraud faced by organisations, high-quality evaluations are not always possible, or even necessary for tackling problems effectively. Drawing on a review of over 400 research studies exploring the prevention of fraud, this paper finds a paucity of studies meeting the highest quality of standards of evaluation using the Maryland Scale. This is largely because of the barriers to implementing the Maryland Scale, given the challenges of measuring fraud, rather than because of a low quality of research per se. In the absence of high-quality evaluations, this paper uses a novel alternative to the Maryland scale to identify a range of effective tools that organisations can use to prevent fraud. Finally, the paper provides practical and theoretical reflections upon a broader problem of how and to what extent scientific evaluations of high-quality evidence are necessary in combating fraud effectively.