Handbook on Sexual Violence (27 page)

Table 4.13
Total offenders proceeded against and total found guilty, all courts, England and Wales

Violence against the person

2008

Total proceeded against 59,943

Total found guilty 41,519

% of proceeded against found guilty 69%

Sexual offences

Total proceeded against 8,440

Total found guilty 5,135

% of proceeded against found guilty 61%

Rape of a female

Total proceeded against 2,233

Total found guilty 855

% of proceeded against found guilty 38%

Source
: Ministry of Justice (2010b)

Developing the measurement of attrition rates

There is controversy over the reasons for this greater attrition. The CPS (2009a:

31) notes the high number of unsuccessful outcomes in rape prosecutions due to jury acquittals. However, a report commissioned by the Ministry of Justice (Thomas 2010) found that juries convicted more often than they acquitted in rape cases and concluded that juries are not the primary source of the low conviction rate on rape. There are a number of studies that look at the various steps, including Kelly
et al
. (2005) and Feist
et al
. (2007), which identify several points rather than a single point at which attrition takes place. The scale of the attrition has become subject to some controversy, following the Stern Review’s (2010) comments on the use of the 6 per cent figure, which were in turn met with critical comments (Baird 2010; Fawcett Society 2010). So the identification of the best procedure by which to measure it is of some importance.

It may be argued that the 6 per cent figure for rape convictions is actually a high estimate; a fully comprehensive attrition rate would use as its starting point the number of crimes committed as opposed to those recorded by the police, resulting in a much lower rate of conviction. This is particularly the case for rape, and indeed for cases of intimate partner and domestic violence, for which the rates of reporting are low relative to other crimes.

Conviction rates are measures of the extent to which the perpetrators of crimes are held to account by the CJS through criminal convictions. One method, developed for rape (Kelly
et al
. 2005), includes attrition across the whole of the CJS process in one statistic. Another method, developed by the CPS (2009b), addresses only the attrition that occurs after the point at which cases are prosecuted. There are further issues: the CPS regularly includes

convictions for lesser offences than the one charged as if they constituted convictions for the offence, doubling the success rate; a practice that some have called into question and consider inappropriate (Baird 2010; Fawcett Society 2010; House of Commons Home Affairs Select Committee 2008).

In consultations on indicators for legal security, the measure for attrition proposed by Alkire
et al
. (2009) for domestic violence, rape and hate crime was based on successful prosecutions of cases as a proportion of the total number of victims (as estimated by survey data). This received widespread support except from the Home Office, which considered the proposed measures to be statistically unsound because the use of data from more than one source introduced methodological inconsistencies. The revised proposal by Alkire
et al
., following the consultation, is to report three sets of figures in raw form: number of cases (estimated from surveys); cases reported and recorded by police; cases successfully prosecuted. While this provides the raw data needed to calculate an attrition rate (expressed as proportion or ratio), such raw data is not itself an attrition rate, thus it would be difficult to use this indicator in estimating the direction of shifts over time.

A narrower way of calculating the attrition is that used by the CPS. Here, the conviction rate is the proportion of total prosecutions that lead to convictions. However, this omits the attrition of cases in all the CJS procedures prior to the decision to prosecute (reporting, recording, detecting, arresting and charging). In relation to domestic violence, the House of Commons Home Affairs Committee (2008: 89) have criticised this method of calculation:

Although some progress has been made by the Crown Prosecution Service over the last few years in increasing conviction rates for domestic violence offences, it is sobering to note that, in areas in which the attrition process has been tracked, the conviction rate for domestic violence, at around 5%, is even lower than that for rape, which is 5.7%. Without linking CPS data on successful prosecutions to data on incidence, arrest, charge and caution, the increase in successful prosecutions tells us little about the criminal justice response to domestic violence.

There is a range of possible solutions here. One may be to calculate the attrition rate for different parts of the CJS separately, for example providing specific rates for the police (from the cases recorded to the number of cases referred to the CPS) and the CPS (from prosecution to conviction). Another solution might be to bring the statistical systems into sufficient alignment such that concerns about different methodology become insignificant and a figure for the system as a whole can be produced. A further approach would be to track individual cases throughout the criminal justice system to monitor attrition more accurately at the different stages in the process and subsequently develop measures to prevent these cases being lost. But whether the procedure is to split the attrition into parts or to produce it as one for the whole system, the concept of attrition is best understood as the proportion of rape cases that do not lead to conviction.

Conclusions

The quantitative measurement of the extent of sexual assault has contributed to the process of making the issue one of public debate. The tension between the different measurements of sexual violence is significant. The difference between the extent of sexual violence reported to national surveys, that reported to and recorded by the police and that of convictions in court is vivid testimony to the social processes that allow only some limited numbers of cases to be made visible to the public and be brought to justice. The concepts of ‘attrition rate’ and ‘conviction rate’ are important in measuring this ‘justice gap’. Given the importance of this justice gap for both gender justice and justice in general, it is not surprising that the way it is conceptualised and measured is subject to critical commentary and public debate. While official reports have recommended a narrow focus on conviction rates after cases enter the court system, most experts have recommended that the concept encompasses the proportion of actual cases of rape that lead to criminal convictions for rape. This latter statistic captures the attrition or ‘fall out’ of cases across the criminal justice process and the points at which improvements are necessary in order to increase the proportion of cases that are brought to justice.

The source of the most reliable data on the extent of sexual violence is the British Crime Survey and in particular the self-completion module. The crime surveys are more reliable than other forms of data, for example recorded crime data, because of the notorious under-reporting of sexual violence crimes. Although survey data may still underestimate the extent of sexual violence crimes, they are nonetheless more reliable than other data. However, less common forms of sexual violence cannot be accurately identified using this method and remain invisible. This is particularly a problem for measuring sexual violence against social groups found at the intersection of two or more inequalities, here examined as forced marriage, trafficking for sexual exploitation and FGM. There are no national survey data or administrative data for these forms of violence. Indeed, there is agreement among academics, civil society actors and policy-makers that there is no reliable data or, at present, way of measuring incidents of forced marriage, trafficking or FGM, let alone agreed definitions of these forms of sexual violence. Agreeing on a definition to enable measurement is a necessary first step. Where quantitative measurements of the extent of sexual assault have contributed to making sexual assault visible and to the process of making the issue one of public debate, other forms of sexual violence remain less visible.

Kelly’s early work on the definition of sexual violence was important in broadening the understanding of sexual violence by drawing attention to issues in addition to rape and highlighting the centrality of gendered power to its analysis. Building on the subsequent 20 years of research on sexual violence, we can now develop beyond the initial concept of the continuum of sexual violence. During this period several further forms of sexual violence have been identified and named, along with the development of specific laws and policies to address them. While they are doubtless all connected, it is important to recognise the distinctions between them, the better to measure

and address them, hence we have provided data on the measurement of multiple forms of sexual violence.

The quality of the measurement of the nature and extent of sexual violence has been improving substantially over recent years, but much further improvement is still needed.

Further reading

For a further account of measurement issues in sexual assault (as well as domestic violence and stalking) see Sylvia Walby and Jonathan Allen (2004)
Domestic Violence, Sexual Assault and Stalking: Findings from the British Crime Survey
. HORS 276. London: Home Office. For an account of how rape cases fall out of the criminal justice system see Liz Kelly, Jo Lovett and Linda Regan (2005)
A Gap or a Chasm? Attrition in Reported Rape Cases
. Home Office Research Study 293. London: Home Office. For an account of the ‘complexities and pitfalls’ in measuring trafficking, see Ernesto Savona and Sonia Stefanizzi (eds)
Measuring Human Trafficking
. New York: Springer. The forthcoming book by Sylvia Walby (2012)
Gender Violence
(Cambridge: Polity) situates the identification and counting of sexual violence in a wider context.

References

Alkire, S., Bastagli, F., Burchardt, T., Clark, D., Holder, H., Ibrahim, S., Munoz, M., Tsang, P. and Vizard, P. (2009)
Developing the Equality Measurement Framework: selecting the indicators
. Manchester: EHRC. Available from: http://www.equalityhu-
manrights.com/fairer-britain/equality-measurement-framework/ (accessed 15 No- vember 2010).

Baird, V. (2010) ‘Stern review published – Rape reporting in England and Wales’,
Women’s National Commission
. Available from: http://www.thewnc.org.uk/work-of- the-wnc/violence-against-women/sexual-violence/307-stern-review-published-rape-
reporting-in-england-and-wales.html (accessed 15 November 2010).

Burman, M., Lovett, J. and Kelly, L. (2009)
Different systems, similar outcomes? Tracking attrition in reported rape cases in eleven countries. Country briefing: Scotland
. Child and Woman Abuse Studies Unit: London Metropolitan University.

CPS (2008)
Pilot on forced marriage and so-called ‘honour’ crime – findings
. Available from: http:// www.cps.gov.uk/publications/docs/findings_from_c
ps_pilot_on_forced_marriage.pdf (accessed 15th November 2010).

CPS (2009a)
Violence Against Women: Guidance
. Available from: http://www.cps.
gov.uk/ legal/v_to_z/violence_against_
women/#content (accessed 15 November 2010).

CPS (2009b)
Violence Against Women Crime Report 2008–09
. London: CPS. Available from: http://www.cps.gov.uk/publications/docs/CPS_VAW_report_2009.pd
f (accessed 15