Spatial characteristics of serial sexual assault in New Zealand.

• Jurisdiction: Australia
• Author: Lundrigan, Samantha
• Date: 01 August 2006

This study examines the crime patterns of New Zealand serial sexual assaulters in order to determine the extent to which such offenders display spatial behaviour in line with their overseas counterparts. Hypotheses concerning journey to crime, criminal range and spatial pattern were tested. Geographical information on 76 serial offenders was extracted from police files and mapped. A series of spatial analyses were conducted in order to examine the spatial mobility of the offenders. It was found that, in line with much overseas research, offenders typically did not travel very far to offend (a median of 3 km) and operated over identifiable criminal ranges. However, in contrast to much overseas research into serial spatial patterns, New Zealand offenders were characterised by transience and commuting patterns of behaviour rather than a predominant marauder style. The implications of these findings for geographic profiling are discussed.

**********

Research shows that an offender's spatial behaviour is influenced by four interrelated groups of factors; offender background characteristics, offender knowledge/experience, target selection considerations and the environmental backcloth (Brantingham & Brantingham, 1981). These factors influence the locational decisions offenders make about an area to search and a microenvironment to exploit within that area. What research has shown is that offenders often select familiar environments, typically not too far from home, to locate a suitable target (Rhodes & Conley, 1981). Studies of serial, sexual offenders in particular have shown that their spatial behaviour is influenced by the distribution of their noncriminal, routine activities (Canter & Larkin, 1993). However, much of this research has been carried out from an Anglo-American perspective and it is an open question as to whether the same picture exists in other countries. In particular, there has been no examination of the spatial behaviour of serial sexual offenders from New Zealand. The need for such an examination is especially clear when one considers that recent advances in the field of geographic profiling are based predominantly on Anglo-American research and that geographic profiling is now a tool being used within the New Zealand context.

Spatial Studies of Serial Sexual Offenders

Studies of the spatial behaviour of serial sexual offenders have typically focused on two main areas. On the one hand, examination has been made of the average distances and range over which they operate (Warren et al., 1998; Le Beau, 1987b) and on the other hand, attempts have been made to categorise them into different spatial groups such as marauders and commuters (Canter & Larkin, 1993; Kocsis & Irwin, 1997).

Within the field of environmental criminology, there is a strong research tradition that examines the distances travelled by offenders to commit their crimes. Such studies have typically found that offenders, and in particular personal offenders such as rapists, do not travel very far from their base to offend. For example, Amir (1971) found that 82% of the rape offenders in his study lived in the same neighbourhood as their victim, while Le Beau (1987a) found that serial sexual offenders travelled an average of one mile to offend. In a similar vein, Davies and Dale (1995) found that 76% of the rapes of 79 British serial rapists were within 5 miles of home and in their study of 83 American serial offenders, Warren et al. (1998) found that they travelled an average of 3 miles to offend.

In addition to the journey to crime research, there have been a number of attempts to model or classify serial offenders' spatial patterns. Canter and Larkin (1993) proposed two conceptual models of serial offender spatial behaviour, the 'commuter' hypothesis and the 'marauder' hypothesis. With the marauder hypothesis the offender's home range, or his noncriminal environment, overlaps with his criminal range or the area where he commits his offences. In contrast, the commuter hypothesis describes an offender who moves out of his home range to a separate criminal range to commit his crimes. Canter and Larkin found that 86% of their United Kingdom offender sample fit the marauder model, while the remaining 13% were commuters.

Subsequent studies have lent support to the existence of this spatial distinction, although with variations in the proportion of offenders falling in each group (Canter & Gregory, 1994; Kocsis & Irwin, 1997; Tamura & Suzuki, 1997). Interestingly, while there has certainly been some theoretical discussion in the literature of what might differentiate between the two patterns, there has been very little actual empirical examination of the differences between the two and the validity of this method of separation has rarely been tested. Added to this is the lack of consideration given to those offenders who display both types of spatial pattern over their criminal career.

Recently though, research has begun to focus on trying to determine if there are any measurable factors that differentiate these two types of offenders. Meaney (2004) investigated the influence of demographic characteristics (age, gender, ethnicity), crime type (arson, rape, burglary), and environment type (rural or urban) on offender spatial patterns. Of the 32 Australian sexual offenders she examined, 93% were found to be marauders and compared with burglars and arsonists, she found that sexual offenders were more likely to be younger marauders living in metropolitan areas. In other words, Meaney found that the two groups could be differentiated on certain offender characteristics, offence types and environmental features of offence patterns.

The suggestion that the environmental context of offending may contribute to whether an offender displays a marauding or commuting pattern is an interesting idea that may help explain why the proportions of offenders conforming to each pattern differs depending on the country of study. Similarly, the suggestion that offenders vary according to their demographic characteristics may also be a contributory factor, although the profile of the typical serial rapist is not necessarily different in New Zealand from elsewhere. Rather, it may be more a reflection of a cultural feature of the population under study.

Explanations for Offender Spatial Behaviour

There are two principle explanations that have been put forward to account for the close relationship between an offender's spatial behaviour and his home location. The distribution of crime is closely related to the distribution of opportunities within the environment. However, opportunities only become actual targets when filtered by an offender's own knowledge and perceptions. In other words, the distribution of crime is closely related to the offender's own knowledge and experience of the environment. Brantingham and Brantingham (1993) articulated this relationship in their crime pattern theory. They proposed that the routine, noncriminal activities of an offender provide experience and familiarity with an area (called an activity space) and thus shapes where crimes are committed. In proposing an activity space for criminals, Brantingham and Brantingham (1981) saw crime as a byproduct of other activities in which the criminal engages.

A second and complimentary explanation relates to the notion of a rational offender. The rational choice explanation of spatial behaviour involves the making of decisions and choices which exhibit a trade-off between increased opportunity and greater reward the further an offender travels from home, as well as the costs of time, effort and risk (Cornish & Clarke, 1986). Journey to crime research would suggest that offenders typically travel the shortest distance possible in order to find a suitable target to exploit. This is also known as the least effort principle (Zipf, 1950) where, given a number of identical alternatives for action, an offender selects the one closest to him in order to minimise the effort involved.

Both these explanations describe processes that would be expected to be evident regardless of the cultural or environmental context concerned. Therefore, while recognising that New Zealand is a distinct environment with its own...