Motivation for occupational fraud : an analysis of the 'fraud triangle' using economic logic

2013 
The 'fraud triangle' constitutes a very popular criminological model of occupational fraud that explains the dynamics underlying occupational fraud. However, given that occupational fraud is essentially an economic crime, there is little evidence in the extant literature that provides an explanation of the 'fraud triangle' in purely economic terms. In this paper, we separately consider each of the three vertexes of the "fraud triangle" and establish an underlying economic linkage that coherently connects each of the vertexes. We first consider the "incentive" vertex and hypothesize and prove that in equilibrium, the "net valence" (i.e. net incentive to an employee from an act of occupational fraud) is zero. It is when the equilibrium is disturbed that a positive "net valence" for fraud arises, identified in our framework by an increase in the probability of an employee not being caught after committing fraud. We then move to the second vertex i.e. "opportunity" and derive, using a finite-state Markov chain, a separate, closed-form expression for the equilibrium probability measure identified in our analysis of the first vertex. Finally, we show that the last vertex - "rationalization"; represented in our framework via an equilibrium penalty-to-reward ratio for an act of occupational fraud from an employee's viewpoint is very sensitively dependent on the likelihood of the organizational culture towards fraud continuing to remain in a slack state in the future given that it is currently in a slack state. We also provide a hypothetical numerical illustration of our framework for a specific setting of the Markov state transition matrix. We believe that our analysis is a first step in providing a rigorous mathematical ratification of the criminological theory of "fraud triangle" and therefore helps to open up the field for more meaningful empirical enquiry in future.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    4
    Citations
    NaN
    KQI
    []