Quantitative Criminology: Bayesian Statistics for Measuring the 'Dark Figure' of Crime

Abstract

This paper proposes a new approach to measure the “dark figure” of crime. Accurate probabilities to a variety of unrecorded or unreported crimes, including those that are difficult to measure, are needed to understand the level and type of crime occur in society, test and assess criminal activities and criminal behavior. Bayesian statistics is proposed to measure the “dark figure” of crime. The advantage of using Bayesian statistics is that a degree of belief or a state of knowledge that a criminologist or social scientist holds in a proposition plays an important role in those circumstances when very few observations or data are available. Bayesian statistics assigns probabilities to the degree of belief, or, more specifically, the degree of rational belief, and modifies or changes it mathematically based on relevant evidence relating to it. This view of Bayesian statistics is where it becomes particularly helpful in measuring the “dark figure” of crime and the two elements that constitute this figure: unreported crimes and crimes that are reported to the police but not recorded by them. The application of how we can integrate this methodology into crimes that are never reported to, or never recorded by, the police or any other legal or law enforcement agencies is explained in a simplistic manner with examples, along with the strengths and limitations associated with it. The proposed methodology is expected to provide an insight into the amount of unknown (unrecorded/unreported) crimes that take place in any given place at a given time and ensure that crimes outside the boundaries of all official/independent crime statistics can be accounted for quantitatively.