1. When is a crime not a crime
When is a crime not a crime? Using the victim experience to determine the principal components of crime
Iain Brennan
BSC Victimology Conference, University of Lincoln
29 Jan 2015

2. Crime labelling and crime seriousness
Long interest in how the public feel about crime
(e.g. Beccaria, 1769)
Many studies have ranked and rated the seriousness of different crime types using general population samples (Sellin & Wolfgang, 1964)
Very few have used real-life cases and victims
Too noisy for comparison? Not enough data?

3. Victim responses to crime
What is a crime to some may not be a crime to others Victim data can provide insights into seriousness and the factors that make an incident ‘a crime’ Identifying factors that lead to ‘crime discounting’ can identify a previously hidden sample of victims, improve police ascertainment of crime and inform the democratic processes of criminal law

4. Proto-typicality Crime seriousness Perceptions of crime Context
Legality
Harm
Unjust
Proto-typicality
Impact
Future
vulnerability
Crime seriousness
Ruback, Greenberg & Westcott,1984

5. Contextual factors
Vignette studies show perceptions of assault affected by consequences, offender characteristics, victim characteristics, precipitatory factors and incident- specifics (Ruback & Thompson, 2001) Studies of actual crime reporting show that reporting likelihood is affected by relational distance, incident visibility and victim intoxication Communities have an influence on crime reporting and attitudes towards the use of law (Ellison, 1991; Anderson, 1999)

6. Research questions
What factors are associated with crime labelling and discounting? How do different levels of social ecology influence decisions about crime – event, individual, community?

7. Method
Sample: BCS respondents 2008/9-2010/11 n=139,515 Inclusion criterion: “has anyone, including people you know well, deliberately hit you with their fists or with a weapon of any sort or kicked you or used force or violence in any other way?” n=3,325

8. Variables - Outcome
Crime labelling Did you think that what happened was A crime 2. wrong, but not a crime 3. or just something that happens
Not a crime

9. Predictors/covariates
Harm: injury; weapon used; emotional harm Unjustness: number of offenders; victim initiation; personal attribution Future vulnerability: fear of crime Prototypicality: relational distance; history of victimisation Level effects: neighbourhood (LSOA); neighbourhood deprivation Multilevel logistic regression analysis to identify factors that predict crime labelling clustered within neighbourhoods

10. Results
35% of people who experienced violence did not regard it as a crime
Little evidence of neighbourhood clustering
Crime labelling: Weapon use, injury, emotional damage, multiple assailants, personal attribution and relational distance predicted increased likelihood of labelling incident as a crime
Crime discounting: Victim initiation, frequency of victimisation
Stats

11. Interpretation
For victims, ‘crime’ is more than just harm x legality, it is a complex decision influenced by many other contextual factors
Ruback’s model is a valuable one for understanding victim response to crime, but it is improved by the inclusion of context
Black’s theory of relational distance still holds even when the decision has no real-world consequences

12. Interpretation
Neighbourhood effects are minimal, suggesting that non-reporting within neighbourhoods reflects compliance rather than conformity
‘Crimeworthiness’ is informed more by the characteristics of the incident than local norms

13. Learning from the victim
Some violence is acceptable to some people
Resilience
Control/ownership of law
but discounting may also be instrumental:
Neutralisation
Control of ‘victim’ identity

14. Future research directions
Determining contextual influences – e.g. alcohol, war zone
Need for further understanding of the mechanism for discounting
Continue use of large data sets to test complex theories about victims

15. Practical implications
There are a cohort of victims beyond those who do not report – violence is under-labelled as well as under-reported
Organisations who work with victims should not just focus on changing victim behaviour in response to violence but their interpretations of it
Recognise a population of victims beyond those who do not report

16. Thanks! I.Brennan@hull.ac.uk

17. Stats output Measures of association (Odds ratios)
Odds (95% CI)
Individual-level variables
Weapon
1.89 (1.41–2.54)***
Injury
1.71 (1.42–2.05)***
Emotional harm
(compared to no emotional response)
Just a little
1.09 (0.84–1.42)
Quite a lot
2.76 (2.06–3.70)***
Very much
5.73 (4.11–7.98)***
More than one assailant
0.79 (0.65–0.97)***
Relational distance
(compared to family member/intimate partner)
Friend/acquaintance or neighbour
1.11 (0.81–1.52)
Other
1.78 (1.37–2.33)***
Incident attributed to victim characteristic
1.35 (0.98–1.85)
More than one reported violent victimisation
0.79 (0.65–0.97)*
Fear of victimisation
1.07 (0.97–1.19)
Victim initiation of violence
0.39 (0.22–0.69)***
Neighbourhood-level variables
High crime and disorder
1.23 (1.00–1.53)

18. Limitations British ‘Crime’ Survey may affect labelling
Known and unknown unmeasured confounding variables
Difficulty in identifying ‘pure’ victims
Phenomenon may be specific to UK