Peace Bonds and Violence Against Women: A Three-Site Study of the Effect of Bill C-42 on Process, Application and Enforcement

5. Methods and Success Measures

5. Methods and success measures

This research project employs multiple methods for understanding the routine practices of criminal justice personnel in dealing with peace bonds and the effects of Bill C-42 on both the procedural practices and on overall issuance and breach rates.  It was decided at the outset that the study would employ both an examination of official statistical sources as well as interviews with key criminal justice personnel.  The primary statistical source for national peace bond trends is the Adult Criminal Court Survey (ACCS) data produced by Statistics Canada’s CCJS (Canadian Centre for Justice Statistics). Other sources in this report include Hamilton Regional Police statistics provided by their Family Violence Resource Unit as well as police databases in Winnipeg and Halifax. In the case of Winnipeg, we also relied on the Winnipeg Family Violence Court Database managed by Dr. Jane Ursel at the University of Manitoba.

Given the goals of the amendments to section 810 peace bonds listed in the previous section, we may construct certain measures to gauge the success of Bill C-42.

The first goal of the amendments was to increase accessibility to peace bonds, and to make it possible for third parties to apply for recognizances on behalf of applicants who may be too fearful or incapable for other reasons to commit to the process.  A prime example is a battered woman who does not wish to confront her abuser.  In this case a shelter worker, advocate, friend or police officer could apply and get the peace bond on her behalf.  We may measure the success of this first goal in multiple ways.   One approach is to examine whether peace bond issuances on a per capita basis have gone up since Bill C-42 by looking at longitudinal ACCS data.  Interviews of key informants will also tell us whether lawyers, police officers and others are aware of the option to apply on behalf of an applicant and whether or not, and to what degree, this option is ever exercised.  Of course, if third party applications are recorded on police information systems, this can also tell us whether this option is now being utilised.

The second goal of the amendments was to provide a list of possible conditions that may be imposed by judges on respondents in cases of peace bond issuances.  Police data and interviews with key criminal justice personnel, especially judges, can tell us about whether these amendments are known and whether they have had any effect on the conditions imposed.

The final goal of Bill C-42 was to increase the maximum penalty for the violation of a peace bond from six months on summary conviction to two years on indictment. We will interpret the presumption here to be that stiffer sanctions act as a deterrent and may result in fewer breaches of peace bonds. ACCS longitudinal breach data may be examined to see whether any changes to the breach rate have occurred since 1995.  Moreover, individual offender history data for Halifax and Winnipeg can tell us more precisely whether a respondent was charged during and after the duration of the peace bond.

Figure 5.0.1 summarises the effectiveness measures for this project by outlining the goals of the amendments and the data sources that will inform the analysis. It is important to keep in mind, however, that other data and discussions are also included in this report.  As this is the first national peace bond study in Canada, it was felt essential that some additional qualitative information be collected to help explain the study findings and standard practices to determine the impact of the legislation.

Figure 5.0.1: Purpose and Source of Information
Goal Effectiveness Measure Data Source
1a. Increase accessibility to peace bonds.
  • More peace bond issuances per capita.
  • Key informant reports of easier access and more use.
  • ACCS
  • Interviews
1b. Make it possible for a third party to apply for a peace bond on behalf of a person at risk of harm.
  • More third party applications.
  • Practitioner knowledge of this option.
  • Police data
  • Interviews
2. To provide concrete examples to judges of the types of conditions that may be applied to peace bonds.
  • Use of variable conditions by judges.
  • Police data
  • Interviews
3. To increase the maximum penalty for the violation of a peace bond from six months on summary conviction to two years on indictment.
  • Fewer breaches due to deterrence.
  • Stiffer penalties.
  • ACCS
  • Police data
  • Interviews

Moreover, the research had a specific focus on how these amendments affected women in violent relationships.  For the interviews, women’s shelter workers were contacted and other court and police personnel were specifically queried about peace bonds in the context of spousal and intimate violence. In the case of the police data, we also targeted cases of intimate violence or threats.

The following subsections discuss how each data source contributes to both assessing the efficacy of Bill C-42 and provides a general indication of the routine practices of criminal justice personnel in the different Canadian jurisdictions we examine.

5.1 Adult Criminal Court Survey Data

The ACCS was organised to collect data on court dispositions in criminal cases in Canada. However, not all Canadian jurisdictions report data to the ACCS: British Columbia, Manitoba, and New Brunswick are not included in the survey. It is estimated that since 1994/95, the ACCS has about 80 per cent coverage of criminal cases in Canada.  Prior to 1994/95 the survey captured only 30 per cent of court activity with only four provinces and one territory originally participating.  For this reason, this report begins with data from the 1994/95 report year.  In addition, the data are re-calculated per 100,000 population on both a provincial and national basis, eliminating non-participating provinces.

Because the ACCS survey was designed to collect data on the processing and outcome of criminal charges in court, some jurisdictions do not report the issuance of section 810 peace bonds at all because it does not constitute an offence.  However, consistent with the design of the survey, for those that do, there are "case, charge and disposition records" for both section 810 issuances and 811 breaches.  Given that an 810 does not constitute a charge, it is unclear what, for example, a reported 810 "charge record" with a "guilty" disposition means. Does this mean that there was a section 810 hearing and the applicant was successful or that the original section 810 was subsequently breached and the accused was convicted of a section 811? There is cause for confusion in other areas too, such as the inordinate number of dispositions reported as ‘other’ [1] (see Appendix A). In order to minimise the effect of these reporting irregularities, we include only charges and focus on two of the more transparent measures:  total charges (for 810) and total guilty dispositions (for 811).  We take the former to give us the overall issuance rate and the latter the breach rate. [2]

5.2 Police Data

As with most research, each data collection site posed its own unique problems.  In the case of Hamilton, due to policy and procedural factors, the police service after lengthy consideration determined that access could not be granted to their database. For Halifax, digging up police data on peace bonds proved a much longer process than anticipated because their report systems are set up for administrative purposes and are not conducive to social science research.  Winnipeg’s police database could only be accessed by research personnel in Manitoba who previously collected information from the police and entered it into a functional research database known as the Winnipeg Family Violence Court (WFVC) database.  This latter database was then accessed to provide the source of names to be searched from the police information system.  However, in the end, these police services were each able to provide the research team with valuable information.

5.2.1 Halifax

In order to identify peace bonds in cases of domestic violence from the Halifax Regional Police, the cases identified as ‘private recognizances' on their ‘hazard screen.’ [3] had to be cross-referenced with known ‘domestics’ from the Victim Services Unit.  As it was discovered that far more cases may have been related to intimate violence than were captured by Victim Services, the Halifax police researcher had to re-enter the incident report number from the hazard screen to access the actual incident report to determine the relationship between the applicant and respondent.

Given that when peace bonds lapse they are removed from the hazard screen, only cases from 1998 onwards could be accessed.  The cases were then logged onto a data collection sheet. In all, there were 233 cases collected from the Halifax Police information system.  Of these, 84 (or 36 per cent) were determined to be spousal (or intimate) violence related.

5.2.2 Hamilton

As was previously indicated, the Hamilton Police eventually denied researchers from this study access to their database. Previous to this decision, the Hamilton Police Family Violence Resource Unit had provided us with internally collected data on peace bonds. These are monthly spousal assault summaries showing dispositions including peace bond issuances.  However, without respondent identification and without a completed data collection sheet, there was no way to build a tracking database for Hamilton.

Under the assumption that data from police files was forthcoming, a researcher had already begun interviews with Hamilton and Ontario informants.  These interviews are the primary source of information for Hamilton.

5.2.3 Winnipeg

The Winnipeg data is essentially police augmented data from the Winnipeg Family Violence Court (WFVC).  This part of the data collection was contracted to Jane Ursel at the University of Manitoba.

Since the WFVC data did not include certain information needed for completion of the standard data collection sheet for this study, it was necessary to gain access to the Winnipeg Police information system.  The WFVC data did not include the Finger Printer Serial (FPS) numbers or dates of birth for the respondents.  This information was required for the next part of the analysis, when the data collection sheet was forwarded to the RCMP for the tracking of respondents.

The WFVC database contained 340 cases of peace bond issuances from 1992 to 1997.  Since all of the cases were originally derived from the WFVC database, all peace bond data from Winnipeg is considered family violence related.

5.2.4 CPIC

The RCMP’s Canadian Police Information Centre (CPIC) collects charge and disposition records for offenders across the country.  These are supplied by police forces.  Almost all police services in Canada now have either in-car access or dispatch access terminals to CPIC to check the background of suspects or search for outstanding warrants.

After data collection sheets were completed for Halifax and Winnipeg, these were passed on to the CPIC office in Ottawa so that background checks could be run on all respondents issued peace bonds in those two cities.  The CPIC data was then added to the case file history of each respondent and both the data collection sheet data and the CPIC information was input into an SPSS data file.

The combined local and federal police data makes it possible to find out whether and when individual respondents committed offences while under conditions of a peace bond.  We can thus calculate an individual breach rate apart from court processing data gleaned through ACCS statistics. Instead of relying on aggregate charges coming through the courts with no way of linking these to particular individuals, we can now calculate breach rates on an individual basis.

5.3 Interviews

We conducted a total of 26 interviews for this report. These were conducted by telephone with key informants in Ontario (n=8), Nova Scotia (n=11), and Manitoba (n=7). Seven informants were Magistrates, Justices of the Peace or Judges, six were lawyers or Crown attorneys, seven were police officers and six were shelter workers.  A larger number of potential informants had originally been contacted but some were dropped because of their lack of contact with Criminal Code peace bonds, thus leaving a sample size of 26. We tried to focus on the particular city in which the site research centred:  Halifax, Hamilton, and Winnipeg, however, in certain cases knowledgeable informants were located outside these sites but within respective provinces.

  Thus, the interview data in this report do not reflect Canada as a whole but rather the specific sites chosen. The questions asked varied depending on the informant. While a certain core number of questions were asked of all interviewees, the interview schedule changed for judges, lawyers, police officers, shelter workers, and justices of the peace because of their particular institutional and professional experience and perspective (see Appendix B). Participants were asked to comment on the use, enforcement, and disposition of peace bonds especially in cases of intimate violence or threats.  Specific questions concerning knowledge of Bill C-42 amendments were raised, and in particular the utilisation of third party applications, and the process and utility of peace bonds in situations of intimate or spousal violence. The interviews proved to be a very valuable source of data. Institutional interpretations of law and the practical application of legal processes are the real determinants of accessibility, enforcement, and overall efficacy.

[1] The Canadian Centre for Justice Statistics (CCJS) is presently conducting an audit of how participating jurisdictions are reporting peace bond data. However, the results of this endeavor were not available at the time of the writing of this report.

[2] Based on discussions with the survey manager at CCJS, we are assuming here that all charges reported in a yearly cycle are more reliable than using cases as our unit of analysis and that the only reliable disposition are those categorised 'guilty' under 811 charges.

[3] A hazard screen pops up to alert Halifax police officers when there is an outstanding warrant or bond when a suspect’s name is queried on the information system.