Title: REPETITION OF CRIMES AS A FERTILE AREA FOR FORENSIC PROFILING

Repetition of crimes as a fertile area for forensic profiling

At this point, the treatment of specific cases has been discussed. However, some forms of criminality, such as high volume crimes, are likely to show a substantial serial component. A series of studies in criminology led to the conclusion that both as small number of persons or group of offenders are responsible for a great quantity of cases , and certain victims and types of victims show different patterns in the way they are repeatedly victimised . Understanding patterns of victimisations and detecting the activity of groups of serial offenders are of crucial interest both from a repressive and preventive perspective, and also from a more strategic perspective, when the structure of some forms of criminality are scrutinised (mobility of offenders, specialisation, evolution of criminal careers, amplitude of the activity, association between criminals, classification of victims, vulnerabilities, crime phenomena, etc.). This is why modern policing strategies necessitate the implementation of analytical capabilities dedicated to the understanding of the dynamic of crime and the provision of intelligence on which informed decision may be taken.  

Forensic profiling has thus a great role to play in this perspective, providing at least that physical material and traces have a demonstrated potential for connecting the dots, or linking cases . 

Repetition of crimes – crime series

Crime series occur when the same offender operates several times. There are no definitive definitions of crime series as group of criminals may operate in different compositions, show different forms of organisation and perpetrate different types of crimes. One may accept that when the same offender operates repetitively, there is a crime series, and that individual crimes of that series are linked. Several crime series may be connected and form a graph structure which represents the activity of a group of offenders.

Thus, three forms of forensic profiling that rely on police recorded data may occur in this context:  

1. Detection of a series of crimes and reconstruction of graph structures that represent the activity of single offenders or group of offenders (group profiling)  

2. From a series of crime, extract a profile that best describes the series and its or their author(s) (individual profiling) 

3. From the graph structure, extract a profile that best describes the forms of criminality and the group of offenders (group profiling) 

Crime series detected through police recorded data are generally incomplete and largely uncertain. Not all crimes are reported by victims (level of reportability may greatly vary according to the type of crime) and the collection of traces at each crime scene may be only partial. Moreover, a series of obstacles that ranges from legal constraints to technical difficulties causes impossibilities to detect links, while they actually exist. This is the well known incapacity of connecting the dots or linkage blindness law enforcement agencies systematically suffer from .  

From this perspective, basic contribution of forensic science concentrates on crime linking through physical evidence. Detection of links is particularly efficient when the type of crime causes the exchange of physical traces. For instance, DNA databases automatically detect a great number of series, when traces are collected systematically and there is a realisation within the organisation that this detection is useful . A series of arsons, linking through the analysis of illicit drugs seized, or even series of bombing have largely benefited from this approach  

Other types of traces, such as toolmarks , earmarks, shoemarks , may be used in combination, thus augmenting the chance to detect series of crimes through at least one of these dimensions. Electronic traces may also be used for detecting the use of the same mobile phone or smart-card for organising the operation, linking cases when surveillance cameras show withdrawal with stolen bank cards perpetrated by the same person, or for linking different withdrawals perpetrated with the same stolen cards, and even linking unsolicited e-mails through technical information found for instance in the header of the message .  

All detected links present some degree of uncertainty: they may exist or may not exist and we are more or less certain of that. However, links inferred from physical material present the advantage of being the result of a direct comparison of measurable data collected at scenes of crime, while other approaches necessitate as a precondition some form of interpretation. For instance, a traditional method for linking crimes  relies upon the comparison of modus operandi which are already the result of a reconstruction process and thus necessitate uncertain reasoning to be performed as a precondition.

Other approaches consist of focusing on behavioural aspects of cases which add other levels of inferences and uncertain reasoning. This psychological based methodology has been considered as a breakthrough, as it brought systematic and structure for linking violent crimes. However, a series of drawbacks have now been detected, ranging from difficulties of filling long forms to feed the database to the lack of assessments of the global efficiency .  

Thus, by using a holistic perspective, the detection of links is envisaged by monitoring and combining all the possible dimensions. However, pragmatic constraints (cost of input and analyse data, time, knowledge, etc.) and other types of constraints limit the practical possibilities to aggregate data and multiply all those perspectives. This inevitably contributes to linkage blindness. This problem ought to be attenuated by choosing the more solid and efficient dimensions through systematic assessment and comparisons of available methods.  

Beyond the type of data used, the central mechanism for detecting series rests on proximity measurements: spatio/temporal, modus operandi or similarities between traces. There is also space here for developments, seeing that each serial offender shows an individual profile, while automatic methods for comparing cases generally rely on one single metric (see also the example about drug profiling below).  

Repetition of crimes – problems and phenomena

Forms of repetition of crimes other than the serial activity of offenders may be relevant to decide which measures (preventive, repressive or a combination) may be appropriately taken. For instance, hot-spots or crime clusters may incite to concentrate police surveillance on specific area during a certain period, independently of the serial activity of single offenders. Concentration of crimes may be also due to the attractiveness of a potential victim, a residence or a specific environment, independently of how many authors are operating. In this perspective, profiling techniques used refer to spatio/temporal crime cluster analysis that aims to alert based on relevant patterns extracted from crime data.  

More generally, the concept of the problem analysis triangle helps analysis of situations in terms of a combination of a likely offender who meets a suitable target within a specific environment without the protection of a capable guardian. These approaches belong to a set of very rich theories in criminology, called opportunities theories . In this perspective, crimes (and more generally problems that regroup recurring incidents and suggest to focus on their causes ) are considered through the lens of situations which may also be the object of profiling (what are the types of situations recurrently observed, how may I interpret this cluster of crimes?).  

Repetition of crimes – tactical, operational and strategic analysis

Intelligence-led approaches of policing proposed to estimate the usefulness of a piece of intelligence (knowledge) through its capacity to help in decision at different levels of the organisations. In this perspective, at a tactical level, a profile that results from an investigation or a detected series of crime orients investigation (who is the perpetrator, where can I find him). At an operational level, phenomena, target profiles, repeated vulnerabilities and gaps in the security system are scrutinised in order to organise some specific responses. Crime clusters are good example of profiles that may be used at an operational level. Finally, at a strategic level, understanding of the criminality structure along with its tendencies support the devise of global strategies, and decision to be taken by the management. However, all those levels rest upon a solid detection of series and an understanding of repetitive situations.  

For instance, the systematic analysis of e-mails massively sent by scammers (advance fee fraud, 419 scams, phishing, all sorts of counterfeits materials, etc.) in order to find a potential gullible victim provide accurate indications on the criminal mechanisms. This is done from a tactical to a strategic level, and concerns the infrastructures used, the localisation, or the organisation of the fraud . This example blurs the line with the wider concept of intelligence and/or surveillance.