Presentence Reports in Canada 2005-03

PDF Version (223 KB)

Acknowledgements

This project was conducted with the encouragement of the Heads of Community Corrections. Their support and the assistance of probation managers at the various sites were instrumental in obtaining the cooperation of judges, probation officers and counsel in each province and territory. We would also like to thank the research assistants who conducted the interviews and gathered the data in each site. They were Patricia Braiden, Dr. Don Clairmont, Catherine Foo, Claire Goggin, Tara Jackson, Ken MacDonald, Lana Maloney, Jeff Pfeifer, Don Saunders, Charles Stuart and Marcel Vachon.

Our appreciation is also extended to the Heads of Community Corrections who read earlier drafts of this report and provided comments. In particular, we would like to thank Justice David Cole for his perspective from the judiciary. Finally, we would like to thank David Daubney, Justice Canada , for their financial contribution to the project.

The views expressed are those of the authors and are not necessarily those of Public Safety and Emergency Preparedness Canada.

James Bonta, Guy Bourgon, Rebecca Jesseman and Annie K. Yessine
Public Safety and Emergency Preparedness Canada

Executive Summary

The Presentence Report (PSR) has been rarely studied in Canada although it demands a significant proportion of the resources of community corrections. It was estimated in this study that two-days of the probation officer's time were required to prepare a PSR. And yet, we know little about the usefulness of the PSR to the judiciary and how the PSR can be improved to make the reports more beneficial to the key actors in the sentencing process.

The present study obtained the views of judges, probation officers, Crown attorneys and defence counsel on the use and value of PSRs. Participants were chosen from a site in each Canadian province and territory. Over 100 judges and 198 probation officers answered questions about the PSR and nearly 100 Crown attorneys and defence counsel answered a similar, but smaller, set of questions. In addition, specific PSR cases were analyzed in detail and compared to non-PSR cases.

The majority of judges were quite satisfied with the format of PSRs and the information provided to them in the reports. On the other hand, the majority of probation officers were dissatisfied with the PSR structure and process. Probation officers complained of the insufficient training given to them, the lack of time to complete a PSR and sometimes, that the requests from judges to prepare a PSR were for minor cases.

Most of judges would like offender risk information and sentencing recommendations included in the PSR. However, judges tended to favour narrative, subjective assessments of risk and their views on sentencing recommendations in the PSR were not as strongly held as with risk assessment information. Everyone, from the judges to the defence counsel, was highly supportive of PSRs including information on offender treatment needs and recommendations regarding treatment.

In general, judges and probation officers showed relatively high agreement rates on the relevance of many treatment factors (e.g., substance abuse, current intimate relationship). However, agreement on the importance of other promising treatment factors such as social circle and family support was quite low. Although both judges and probation officers saw treatment information as very important, they failed to base certain decisions on the differing risk levels of offenders. Treatment recommendations and treatment orders were unrelated to offender risk.

PSRs tended to be used with the more serious case and appeared to influence disposition. When a recommendation for a community disposition was made in the PSR, a community sentence was imposed in 71% of the cases. Although community recommendations and sentences were more likely with low risk cases, 48% of high-risk cases where a PSR was completed received a community sentence. Thus, it appears that PSRs, with their information on treatment and risk management, provide judges with a willingness to take a risk in placing higher risk offenders into the community.

Presentence Reports in Canada

Probation officers prepare thousands of presentence reports (PSRs) for the courts each year in order to assist judges in sentencing. Although presentence investigations and the preparation of the reports represent a significant investment of probation resources there is surprisingly little research on the topic. We know little about how they are perceived by the courts, if they are valued by judges, what content is particularly important and whether they influence sentencing. The project described in this report attempts to address this gap in knowledge.

The Legal Underpinnings of the PSR
Section 721 of the Canadian Criminal Code (CCC) provides the legislative authority for the PSR. The probation officer prepares a report when requested by the court “for the purpose of assisting the court in imposing a sentence or in determining whether the accused should be discharged pursuant to section 730”^{Footnote 1}. The court can only request a PSR after the offender is found guilty or enters a plea of guilt.

Considerable latitude is given to the content of the PSR. The CCC specifies the inclusion of the following information:

“(3)(a). offender's age, maturity, character, behaviour, attitude and willingness to make amends;

(3)(b) history of previous dispositions under the Young Offenders Act and of previous findings of guilt under this Act and any other Act of Parliament;

(3)(c) history of any alternative measures used to deal with the offender, and the offender's response to those measures; and

(3)(d) any matter required by any regulation made under subsection (2), to be included in the report.”

Subsection 2 refers to the type of offence that may require a report and the content of the form. What is important in subsection 2 is that it is left to each province and territory, through the authority of the lieutenant governor in council, to set the regulations for the type of offence that may require a PSR and what additional content the PSR may include. Although there are some mandatory content requirements for the PSR there is sufficient latitude in the CCC to permit the inclusion of a wide range of other information.

Not surprisingly, a review of provincial correctional and territorial policy manuals finds varying direction to probation officers as to what should be included in the PSR, particularly with respect to sentencing recommendations, risk information and the accused's version of the events surrounding the offence (see Table 1). Cole (2001) argues that Canadian case law, in general, frowns upon the probation officer making a sentencing recommendation in the PSR although he does acknowledge that there are no absolute restrictions in making such recommendations. Similarly, the courts neither encourage nor prohibit the inclusion of information on the accused's views and attitudes toward the crime. However, recommendations around sentencing may be highly relevant when dealing with an Aboriginal offender (R v. Gladue, 1999) and risk information must be considered when contemplating a conditional sentence (R v. Proulx, 2000). Two provinces ( Manitoba and Saskatchewan ) actually include a subsection in the PSR that reports on the findings of an evidence-based, actuarial risk assessment while other jurisdictions may use such an assessment to guide the writing of the PSR without formally referencing the assessment. Nova Scotia did have a separate section on risk assessment but deleted this section after a court challenge (R. v. Elliott, 2004). The province now uses the risk assessment to structure the PSR report without directly referencing the risk instrument.

Table 1 - Provincial Guidelines for the Preparation of Presentence Reports

Notes. Where risk assessment is permitted, it is at the discretion of the probation officer in New Brunswick, Newfoundland, Nunavut and the Yukon. Sections left blank indicate no clear policy/guideline. Information is based on policy guidelines at the time of the study. Prince Edward Island and the Northwest Territories were developing guidelines at the time of the study.

Research on PSRs
Considering the wide use of PSRs in Western countries, the paucity of empirical research on the topic raises concerns regarding the value of PSRs to the courts. For some scholars, our poor understanding of the PSR and its worth reduces the PSR to “inappropriate social gossip” (Gelsthorpe & Raynor, 1995).

At a very general level, if PSRs “assist in sentencing” then there should be high agreement between the recommendations made in PSRs and the court's disposition of the case. Indeed, most studies find high agreement rates particularly when a community sentence is recommended. In the U.S. , Norman and Wadman (2000a) reported an agreement rate of 92%. Carter and his colleague (Carter, 1966, 1969; Carter & Wilkins, 1967) found that judges followed a recommendation for probation between 72% and 97% of the time compared to 27% to 67% of the time when incarceration was recommended. High agreement rates have also been reported in New Zealand (80%, Deane, 2000; 77%, Gibson, 1973) and in the United Kingdom (78%, Thorpe & Pease, 1976).

Canadian research is particularly sparse with all of the studies conducted in the 1970s and 1980s. Hagan (1975) and Boldt, Hursh, Jonson and Taylor (1983) both found that Canadian judges followed the recommendations in nearly 80% of cases. However, Gabor and Jayewardene (1978) found a concordance rate of only 42.9% for a community recommendation. High agreement rates may suggest that judges find PSRs useful but what exactly they find helpful in the reports is unclear.

A number of investigators have tried to define what makes a high quality PSR. At a basic level, quality has been defined by avoidance of discriminatory remarks, spelling and grammatical errors, and covering relevant information (Gelsthorpe & Raynor, 1995; Norman & Waldman, 2000b; Raynor, Gelsthorpe & Tisi, 1995). What exactly is relevant information to the court is an important question that has rarely been answered. Norman and Wadman (2000b) asked 48 judges to rate as “somewhat important” or “not important at all” the sections of presentence reports written in the state of Utah . All the judges rated “somewhat important” the offenders' probation history and 83.3% also rated recommendations as important. Risk information in Utah is provided only through a factual description of the criminal record but all of the judges rated this as important information in a PSR.

One emerging area that is altering the structure and content of PSRs is the increasing focus on criminal behaviour and the management of offender risk (Cavidino, 1997; Stinchcomb & Hippensteel, 2001). Within these domains, offender risk assessment is assuming a major role. As already noted (Table 1), risk assessment is used in the majority of Canadian provinces and territories either as a guide in the investigation supporting the report or to speak directly on the offender's level of risk. Some observers have raised concerns over the use of risk information in PSRs. Criticisms have been voiced that offender risk information in the reports may trigger overly restrictive interventions (Bazelon, 2005), that assessors are poorly trained producing an unreliable result (R. v. Elliott, 2004), and that such information goes far beyond a description of the “character and circumstances of the offender” (Cole & Angus, 2003).

Although there is some resistance from different quarters, risk assessment is an inescapable fact of daily operations in probation and information on offender risk factors influences many aspects of sentencing. In Canada , risk information must be considered in cases of conditional sentences (R. v. Proulx, 2000), Dangerous Offender applications (Bonta, Harris, Zinger & Carrière, 1998; R. v. Moser, 2002) and Long Term Offender applications (Bonta & Yessine, 2005). Furthermore, most jurisdictions use evidence-based, actuarial risk assessment instruments to conduct their evaluations (Table 1).

Actuarial risk instruments are more accurate than the traditional, subjective assessments that were prevalent before 1990 (Bonta, 1996, 2002; Grove, Zald, Lebow, Snitz & Nelson, 2000). Today, most of these instruments assess both static risk factors (e.g., criminal history, past substance abuse) and dynamic risk factors (e.g., current employment, current substance use). Dynamic risk factors, often referred to as criminogenic needs, are especially important for correctional staff because they identify the targets for intervention (Andrews, Bonta & Hoge, 1990). Thus, it is little surprise that probation departments direct staff to structure the PSR around a “risk-need” assessment as the assessment provides clues as to what needs to be done to manage the risk of the offender.

The Present Study
The research described in this report had two main parts. The first part was a general survey undertaken with the major consumers of the PSR. Judges, Crown attorneys, defence counsel, and probation officers were asked what they considered to be important information that should be included in PSRs and what they saw as the role of PSRs in sentencing. The second part of the research involved the collection of information on specific cases. It is one thing to ask people what they consider to be important in general but how they act when faced with a real case could be quite different.

In addition to these two major components of the project, we conducted a small review of the resources required to conduct a full PSR. Very few jurisdictions have a precise knowledge of the time devoted to the preparation of a PSR. In five sites, probation officers were asked to log all activities related to the preparation of a PSR. This permitted us to calculate the amount of time probation officers spend preparing a PSR and it also provided a description of the types of activities that contribute to the preparation of a PSR.

Method

Site Selection
Prior to data collection, preliminary steps were taken to introduce the project to key stakeholders (i.e., probation managers, Judges and Crowns). This involved individual and/or group meetings, answering questions, and gauging interest in the research. With the assistance of the Heads of Community Corrections, a specific site was chosen in each jurisdiction where it was estimated that probation and the courts would provide good cooperation.

Most of the sites selected were in cities ( Whitehorse , Yellowknife , Victoria , Edmonton , Regina , Winnipeg , Ottawa , Gatineau , Saint John , Halifax and St. John's ). The remaining two sites ( Prince Edward Island and Nunavut ), because of a small geography or population base, covered the complete jurisdiction.

Procedure
Phase I. General Interviews. Once the sites were chosen, additional meetings were held with judges and the probation officers responsible for preparing PSRs. For each site, a senior judge (if he/she showed an interest in participating in the project) was asked to present the project to all the judges of that court and request their participation. In early meetings, some judges were reluctant to complete information on specific cases. Therefore, all of the judges were given the option of responding only to the survey or to participate in both the survey and the completion of case-specific information. A list of participating judges was then provided to the research team to arrange for the completion of the survey and the data gathering of case specific information.

The Northwest Territories agreed to participate in the project after the data collection had begun in the other sites. For Yellowknife , we were unable to gather case-specific information due to time constraints. The Judge's Survey was mailed to the judges in Yellowknife (the survey instruments are described in the Measures section).

The fact that the information gathered was dependent upon judges who agreed to participate in the project opens the possibility of bias. That is, the views regarding PSRs may represent a group of judges who may be unrepresentative of most judges. In an attempt to minimize this source of bias, the Judge's Survey was mailed to non-participating judges in the sites selected, a sample of judges in Quebec from courts outside of Gatineau , and a few judges who had heard about the project and were simply interested in completing the survey.

A member of the research team interviewed the judge using the structured interview (Judge's Survey). The Judge's Survey asked about the general opinions on PSRs, the role of sentencing, and the goals of the Criminal Justice System. At the end of the interview, each judge was briefed on the procedures regarding the gathering of data on individual PSR cases and encouraged to participate in the second, optional phase of the research.

At each site, probation officers who prepared PSRs were interviewed using a similar format to that used with the judges. The probation officer was briefed on the procedures regarding the collection of data on individual PSR cases. Although this second phase was optional for judges, it was mandatory for probation officers. The interviews with the judges and probation officers lasted 30 to 45 minutes.

A briefer interview was conducted with 98 Crowns and 93 defence counsel (approximately 20 minutes in length). Most of these interviews were conducted over the telephone owing to the difficulty in scheduling face-to-face appointments.

Phase II. Case-specific Information. Information was collected on individual cases in which a PSR was ordered, prepared and provided to the court for sentencing. Probation officers and judges were consulted in order to determine appropriate procedures for tracking the reports at a particular site. In all jurisdictions, with the exception of Halifax and Victoria, the first 10 cases in which a PSR was requested by a participating judge were identified for inclusion in the study.

In Halifax and Victoria, the procedures varied slightly because the judge that requested the PSR was not necessarily the sentencing judge. A PSR case was identified when the participating judge received the completed PSR. After the judge sentenced the offender, he/she completed what was required for the project and returned the data to the researchers. Once these cases were identified as project cases, the probation officer who completed the PSR was notified and requested to complete the tasks assigned for this phase of the study. In other words, the cases were identified when the judge received the PSR and the data were gathered from the probation officer after the judge received the PSR.

Judges who agreed to gather case-specific information were asked to complete three tasks after sentencing the offender. The first task involved completing a Case-Specific Questionnaire. The Case-Specific Questionnaire asked the judge to rank five sentencing principles (general deterrence, rehabilitation, punishment, specific deterrence, and incapacitation) in order to determine their importance in the sentencing of the particular case. The ranking ranged from 1 (main reason) to 5 (least important reason). The judge was also provided a list of factors about which the PSR may provide information. From this list, the judge was asked to identify those factors about which the PSR provided helpful information as well as to rank the top three factors that were most useful to the particular sentencing decision. The questionnaire also contained a few general questions about the PSR request, what was helpful and not helpful, and the PSR's overall value to the court.

The second task for the judges was to identify specific information on the PSR that they believed was important/critical in the case. This was done by directly underlining the portions of the PSR that provided this information.

Finally, the judges were asked to complete a short form gathering specific details regarding sentencing. These details included a description of the offence(s) for which the offender was being sentenced, date and disposition of the sentence, the presence of a joint submission for sentencing, and, if there was a joint submission, whether it was in accordance with the sentence imposed.

For probation officers, the Case-Specific Questionnaire contained instructions on two tasks required to gather information specifically relating to the individual case. This questionnaire was completed once the PSR had been submitted to the court. The first task required the probation officer to complete a short questionnaire that included a list of potential factors about which a PSR may provide information. From this list, the probation officer was asked to identify all factors on which this particular PSR provided helpful information to the court and, of these, to number the three most important factors for assisting the court in sentencing. The probation officer was also asked to identify all factors about which the PSR provided helpful information for the probation officer, and to number the three most important factors for assisting probation supervision. In this way, specific information in the PSR that the probation officer deemed important was identified for a) sentencing purposes, and b) supervision purposes. The questionnaire also contained open-ended questions about the specific PSR request, difficulties in preparing the PSR, and the overall value of the PSR.

The second task for the probation officer involved indicating the information deemed important directly on a copy of the PSR. Probation officers were asked to underline information that they considered particularly important to convey to the court and circle information that they believed would be useful for a probation officer supervising the offender.

Measures
The general survey consisted of four versions that were tailored to suit the following groups: judges, probation officers, Crown and defence counsel. The survey gathered general information on the opinions, values, and experiences regarding the use and content of PSRs^{Footnote 2}. The survey consisted of up to four distinct sections (depending on the participant group). The first section consisted of open-ended questions that asked about the participant's views on such issues as how PSRs help the criminal justice system, the role of risk assessment in PSRs, and overall satisfaction with PSR format and content. This section of the interview was generally consistent across the four types of respondent (judge, probation officer, Crown and defence counsel), with some questions tailored according to their role within the justice system. For example, judges were asked their opinions on the potential use of forms to request specific information, probation officers were asked about the potential impact of preparing a PSR on supervision and Crowns and defence counsel were asked about the frequency of, and reasons for, PSR requests.

The remaining three sections were more quantitative in style. The second section, included in all interviews, asked participants to rate the importance of a number of different pieces of information that may be included in a PSR such as age, past response to supervision, and treatment recommendations. Ratings were based on a scale of 1 (not at all important) to 10 (most important).

The third section, included only for probation officer and judge interviews, related to general views of the criminal justice system (Crown and defence counsel were not included in order to minimize the resource demands associated with the research). Participants were asked to rate the importance of five principles of sentencing (punishment, rehabilitation, incapacitation, general deterrence, and individual deterrence) on a scale of 1 (not important at all) to 5 (very important). Using the same scale, participants were also asked to rate the importance of four different criminal justice goals (punishment, rehabilitation, protecting society, deterrence) for prisons and for community corrections.

Finally, the Probation Officer Survey included a section on the attitudes and opinions about the reasons for criminal behaviour and appropriate responses by the justice system. In this section, the probation officers were asked to rate their agreement with 11 statements expressing different views on these issues on a scale of 1 (strongly disagree) to 5 (strongly agree).

over the telephone). Participants were given the choice on how to complete the survey. In some cases where a judge agreed to a face-to-face interview, the interviews were recorded on audiotape in order to facilitate inter-rater reliability measurement.

In order to obtain an objective assessment of offender risk for the PSR cases, the Level of Service Inventory-Revised: Screening Version (LSI-R:SV; Andrews & Bonta, 1998) was applied to each submitted PSR. The LSI-R:SV is an eight-item, actuarial risk scale derived from the 54-item Level of Service Inventory-Revised (Andrews & Bonta, 1995). Total scores were calculated if no more than one item was missing; the absence of two or more items for scoring renders the scale invalid.

Non-PSR Cases
For comparison purposes, information was collected on cases where the court did not request a PSR. Each judge participating in the second phase of the study was asked to provide information on a sample of 10 non-PSR cases in which they imposed a sentence. Judges were provided with a form for providing information on offence(s) and sentence, as well as a list of 15 different areas (e.g., financial information, criminal history, and substance abuse) in which information about the offender might be available during sentencing. For each site, the judge was asked to evaluate the quantity and quality of information they knew prior to determining sentence. For example, a judge filling out the form would indicate whether anything about the offender's substance use/abuse prior to sentencing was known, and if so, whether or not the substance use/abuse was known to be a) not a problem, b) a moderate problem, or c) a severe problem.

The non-PSR cases for each judge were identified in one of two ways, depending on the judge's preference. One method required the judge to identify the first non-PSR case following a sentencing hearing for a PSR case. This method was chosen in every jurisdiction except Victoria, Ottawa , and Halifax. In these latter three jurisdictions, the judges identified a series of 10 consecutive cases for which they sentenced the offender without a PSR. Once the non-PSR case was identified, the judge recorded the information requested and forwarded it to the researchers.

Resource Monitoring of PSRs
In five jurisdictions (PEI, Ottawa, Winnipeg, Edmonton, and Whitehorse) the time and resources required to prepare PSRs was estimated. Probation officers were asked to complete a form listing the various tasks relating to the file (e.g., interviews with offender and collateral contacts, reviewing information, writing report) and the time taken to perform each task.

Coding of the General Survey Instruments
Coding procedures were developed for the open-ended questions in the survey by generating a series of response categories for each question. The raters coded the responses as present or absent, positive or negative^{Footnote 3}.

the mean Kappa for 116 variables was .90 (range of 0.46 to 1.0) and for the Counsel interviews the mean Kappa for 118 variables was .90 (range of 0.60 to 1.0).

Coding of Case-specific Information
For the case-specific questionnaires, a series of categories/issues were identified under which responses could be given for each open-ended question. The written responses of the judges and the probation officers were examined and coded according to the presence or absence of each category/issue identified.

Inter-rater agreement was conducted on two raters who coded 33 judge case-specific questionnaires and 61 probation officer case-specific questionnaires. For the judge case-specific questionnaires, inter-rater reliability was examined on 32 variables. The mean Kappa was 0.97 with a range of 0.63 to 1.0. For the case-specific questionnaires completed by the probation officers, Kappas were computed on 56 variables (mean Kappa of 0.95 with a range of 0.54 to 1.0).

Coding of the PSR
Fifty-one different variables from seven domains were coded from the PSR. The seven domains (and the number of information variables in each) were: Social Background (10), Relationships (6), Current Social Context (7), Education/Employment (6), Health (4), Substance Abuse (4), and Criminal Justice System Information (14).

For each PSR, the presence or absence of information for each of the 51 variables was coded (Information Content). In this manner, data was gathered which would allow us to describe the content of PSRs. In addition to coding the presence or absence of this information, 33 of the variables were rated by the coders as negative/problematic, neutral/not specified, or positive/strength (Evaluative Component).

For each case, we also requested that the judge underline information deemed important or useful in that case directly on the copy of the PSR. The probation officer underlined key information that was intended for the court and circled information deemed important for supervision purposes on a separate copy of the PSR. Both of these marked PSRs were coded. PSRs submitted by judges were coded according to whether or not information for each of the 52 variables was underlined when that information was present. PSRs submitted by probation officers were coded according to whether information was underlined, circled, both, or not identified when that piece of information was present. This methodology yielded information as to what both stakeholders held as important and allowed an analysis of the concordance between the two sets of views.

Inter-rater reliability was examined for a) information present/absent, b) the evaluative ratings of the information, c) the information underlined by the judge, and d) the information underlined, circled, or both by probation officer. Two raters coded a total of 63 cases. In 21 cases, we had information from both the judge and the probation officer. In 41 cases we had a PSR marked by the probation officer but not the judge, and in one case, we had a PSR marked by the judge but not the probation officer.

Information categories coded as present/absent. Inter-rater reliability was calculated on 63 PSRs. Kappas could not be computed on three of the 51 variables (Age, Educational Background, and Criminal History) as these variables were coded as present on all the PSRs examined for inter-rater reliability. The mean Kappa of the remaining 48 variables was 0.96 ranging from 0.66 to 1.0.

Information underlined by judges. Two raters coded 22 PSRs that were underlined by a judge. Kappas could not be computed on eight variables because either there was no information on these variables in the PSR (e.g., “History of sexual abuse”) or no variability (i.e., always underlined or always not underlined) on that variable (e.g., “Emotional abuse”, “Psychiatric diagnosis”, “Custodial recommendation”). For the remaining 43 variables, the mean Kappa was 0.96 with a range of 0.66 to 1.0.

Information underlined/circled/both by probation officers. Two raters coded 62 PSRs marked by probation officers. Kappas could not be computed for one variable (“Cultural background”), as there was no variability. The mean Kappa for the remaining 50 variables was 0.98 with a range of 0.88 to 1.0.

The Study Period
Data collection for the project began in Ottawa on December 1, 2002 . The other sites began their participation in the spring and summer of 2003. Data collection ended with the Gatineau site in March, 2004.

Results

The Participants
Participation in the study by the judges varied across sites and between phases of the project. In one site, only two of a possible 11 judges participated whereas in some other settings, all available judges participated. In total, 61.6% (n = 77) of the potential pool of 125 judges from the selected sites completed the Judge's Survey. Another 27 judges outside of the selected sites who learned about the project also completed the survey. Thus, there were 104 Judge's Surveys completed. Thirty of the judges at the sites (39%) agreed to gather the case-specific information required for the second phase of the project with 28 submitting underlined PSRs. A review of personal-demographic characteristics and attitudes towards sentencing and corrections revealed only one difference between the judges who only participated in the survey and those who also completed the case-specific information. Those who only completed the survey rated higher the importance of punishment as a purpose of sentencing than judges who participated in both phases of the project (4.06 v. 3.43, t = 2.8, df = 97, p < .01).

From a potential pool of 258 probation officers who were approached to complete the Probation Officer Survey, 187 (72.5%) agreed to the interview. There were an additional 11 contract probation officers who participated in the study. A total of 198 probation officers (although not the same 198 who answered the survey) completed the case-specific questionnaires and the underlining/circling component of the research. Finally, 98 Crowns and 93 defence counsel agreed to the general interview.

A few personal-demographics of the interviewees are presented in Table 2. It is noteworthy that the probation officers were a well-educated group with 86.8% having a university degree. Two sites (Winnipeg and Halifax) used private contractors to prepare PSRs. Private contractors represented 5.5% of our sample of 198 probation officers who completed the general survey. There were no differences between private contractors and professional probation officers in terms of age. However, contractors had fewer years of experience on average than the professional probation officers (4.4 years v. 10.0 years, t = 2.04, df = 195, p < .05) and almost all were women (90.9%). With respect to defence counsel, 68.8% were private counsel and 31.2% were legal aid. Most of the defence counsel was male (79.6%) and they had slightly more experience practicing law than Crowns (17.4 years v. 12.5 years, t = 4.01, df = 179, p < .01).

Table 2 - Personal-Demographic Characteristics of the Key Stakeholders (N)

Note. Years of Experiences are for that particular role.

Table 3 summarizes the ratings that judges and probation officers assigned to the objectives of sentencing and the goals of probation and prison. In general, both judges and probation officers expressed pro-rehabilitation sentiments. The most highly rated sentencing objective was rehabilitation for both groups with 93.6% of judges and 94.9% of probation officers rating rehabilitation as “quite important” or “very important”. These ratings were in marked contrast to the ratings for general deterrence where only 33.7% of judges rated this objective of sentencing as “quite important” or “very important” compared to 48% of probation officers. Although general deterrence was not viewed as very important, there was a stronger belief in the value of specific or individual deterrence. Approximately two-thirds (68.4%) of judges and probation officers (64.7%) rated specific deterrence as “quite important” or “very important”.

When judges and probation officers were asked about the goals of probation and prisons, a few differences emerged. Probation officers assigned higher ratings than the judges to community supervision as having a protective function (t = -2.64, df = 292, p < .01) and deterrent effect (t = -3.22, df = 292, p < .01). When it came to prisons, probation officers gave higher ratings than judges to the goals of rehabilitation (t = -4.11, df = 281, p < .01), protection to society (t = -3.11, df = 282, p < .01) and deterrence (t = -2.20, df = 282, p < .05).

What the Judges Said
The Judge's Survey was a semi-structured interview consisting of 12 open-ended questions and a series of statements regarding the content of the PSR that are rated on a scale of 1 to 10. The open-ended questions, by their very nature, elicit responses of a spontaneous nature. Thus, our coding of response categories was dependent upon what the judges said with minimal prompting. The frequency of responses would have been different if the questions involved forced-choice responding or if the interviewer gave more guidance and prompts. A preliminary review of the responses from judges in the selected sites with those external to the sites found very few differences in their responses to the Judge's Survey. Thus, the data from the Judge's Survey was combined for both groups.

Table 3 - Views of Sentencing and the Goals of Corrections

Notes: M = Mean rating on a scale of 1 (not at all important) to 5 (very important) 99 Judges and 198 Probation Officers (sample size varies due to missing observations) * p < .05; ** p < .01

At a general level, 87.4% of judges expressed satisfaction with PSRs. When judges were asked how the PSR helps the court in sentencing, the most frequent answer was that it provided general background information (68%). Exactly how general background information is helpful to the court was not answered in the first question other than the comment voiced by 35% of the judges that the PSR provides independent, objective information.

Answers to some of the other questions in the Judge's Survey provided a clearer understanding of the information in a PSR that may be particularly helpful, or unhelpful, to judges. First of all, 37% of judges said that PSRs were most useful for sentencing offenders who have committed a violent offence (general violence, domestic assault and sexual assault). Judges were also asked to rate on a scale of 1 to 10 (not important to most important) 34 factors that could be present in a PSR. Three of these factors dealt with issues of violence (domestic violence, risk to re-offend violently and risk to re-offend sexually). All three were in the top 10 rated factors with ratings of 8.6 and higher (See Table 4 and Appendix A for more details). It is interesting that while concerns over violence and the risk for violence were rated highly by the judges, the risk for general recidivism was much lower (mean rating of 8.1; ranked 16 th of 34 factors).

Table 4 - The Top 10 Most Highly Rated Factors in a PSR

With a concern over the seriousness of the offence comes a need to know something about the offender's risk to commit further harm and what can be done to manage or prevent this risk. Two open-ended questions explored the views of judges concerning the role of risk assessment information in the PSR. The first general finding was that 68.3% of the judges thought that risk assessment information should be included in the PSR. Another 10.6% had no strong opinion about the inclusion of risk information. In other words, only a minority of judges (21.2%) were explicitly opposed to the inclusion of risk information (these judges expressed more concerns about the validity and reliability of risk assessment than judges who were in favour of risk information).

The second general finding concerning the role of risk assessment was with respect to the way risk information is reported in the PSR. Judges were asked to comment on whether they preferred a narrative description of risk factors, the results from a standardized risk instrument, or both. A narrative description refers to the presentation of offender risk factors without the probation officers assigning a level of risk, therefore, leaving it to the judge to make the judgement concerning risk level. The narrative approach was the most preferred option for 40% of the judges although another 20.6% would like to see a mix of both narrative and standardized risk information. Only 21 judges (20.6%) specified that they would like to have standardized risk assessment results in the PSR (18.6% had no preference as to how the information was presented).

Two factors relating to the personal characteristics of offenders were rated highly by the judges: substance abuse and psychiatric problems (ranked four and five). Both factors may be viewed as offender risk factors and/or mitigating and aggravating factors that need to be considered in sentencing. Unfortunately, we did not probe further and the judges did not explain why these two factors are important. However, both factors indicate the need for treatment.

Almost all of the judges (95.2%) were in favour of including recommendations concerning treatment. Another 1.9% had no opinion and 2.9% did not want treatment recommendations in the PSR. The judges were particularly interested in having information on the appropriateness and availability of specific programs in their local communities (voiced by almost 78% of the judges in favour of treatment recommendations). Although treatment recommendations were ranked ninth when asked specifically to rate a range of variables, offender motivation was ranked first and treatment availability was ranked third. Somewhat related to treatment amenability was past response to supervision which was ranked second.

One potential controversial area in the preparation of PSRs is the inclusion of recommendations regarding community and custodial placement (either a placement recommendation or the length of sentence). Many jurisdictions have a correctional policy or guidelines that such recommendations should not be in the PSR. A small majority of judges thought that these recommendations should be provided (59.6%) and only 11.5% of the judges said that recommendations should not be included. A minority (28.8%) had no opinion. A typical response was, “I would like to see what the probation officer thinks on this matter and it is up to me how I consider the information”. The views on placement recommendation may be “soft” as indicated by the relatively low importance rating (6.8; ranked 24 of 34 variables) given to providing this information in a PSR and that nearly one-third of the judges had no opinion on the matter.

The Gladue decision requires the court to address issues concerning the social-cultural circumstances of Aboriginal offenders. Two provinces ( British Columbia and Saskatchewan ) and two Territories ( Nunavut and Northwest Territories ) have policies on the inclusion of information relevant to Gladue. As most Aboriginal offenders are found in the western provinces and northern territories , our analysis of the question concerning Gladue was limited to these jurisdictions. In Nunavut we found all three judges agreeing that Gladue considerations did not really apply in their jurisdiction because almost all offenders were Aboriginal. PSRs in Nunavut typically discussed issues specific to native offenders such as relationships within their community and family, experience and residence on the land versus in town, and participation in traditional ceremonial, hunting, and fishing practices. Selecting the western provinces and northern territories (except for Nunavut ), 66.7% of judges said that the PSR provides useful information about the Aboriginal offender.

Finally, we asked judges if they would agree to complete PSR request forms that specified to the probation officer the kind of information that was of interest to the judge. Slightly more than three-quarters (76.9%) of the judges were agreeable to completing such a form with an additional 8.7% expressing no opinion (14.4% would not like to see such a form).

What the Probation Officers Said
In general, probation officers were evenly split on their level of satisfaction with the current use and form of PSRs. Slightly more probation officers expressed dissatisfaction with the PSR than satisfaction (42.6% v. 39.6%; 18.8% of probation officers were neutral). The majority of probation officers (77.2%) would change the PSR if they could and the most frequently voiced suggestion (16.8%) was to limit requests for PSRs to specific types of offences (e.g., violent, sexual). Other suggestions were to improve communications with the courts (15.7%), decrease repetition in the reports (13.2%), provide more details on the offence and supervision history (5.6%) and provide victim information (5.1%). One-quarter of probation officers (25.9%) felt that their training in the preparation of PSRs was inadequate.

A more detailed analysis of the satisfaction ratings across sites found considerable variability. The findings ranged from complete dissatisfaction at one site to 82.4% of officers in another site saying that they were satisfied with the PSR process and format.

On many other issues, the probation officers expressed views similar to the judges. Probation officers felt that PSRs would be more useful with offenders who have committed violent offences (37.9% said that PSRs are particularly useful for offenders who have committed sexual, domestic violence and general assaultive crimes). Probation officers also shared a similar perspective with judges on the role of risk assessment. The majority of probation officers thought that risk assessment information should be included in the PSR (64.6%). Another 11.2% had no opinion on the matter and 24.2% thought that risk information should not be included. When asked how the risk information should be presented, probation officers tended to be more supportive of a narrative style for presenting risk information than judges (51.3% v. 40.2%).

Pursuing the matter of offender risk information in PSRs, we attempted to score the PSRs on the LSI-R:SV, a standardized eight-item risk scale. We were able to calculate LSI-R:SV scores for 72.9% of the 247 PSRs submitted (if more than two pieces of information were unavailable the LSI-R:SV was not scored). The average LSI-R:SV score was 3.75 with 32.6% of PSR cases scoring in the low risk range, 42.7% in the medium risk category and 24.7% were high risk. Compared to a normative sample of probationers (Andrews & Bonta, 1998), the offenders for whom a PSR was requested tended to be higher risk. For example, in the normative probation sample, 68.1% of probationers were low risk compared to 32.7% in the PSR cases. This distribution of risk categories for the PSR cases was consistent across the different sites. The Gatineau and Whitehorse sites were the most likely to have items omitted whereas Ottawa and Saint John had the least number of items missing for scoring the LSI-R: SV. The most common pieces of information missing from the PSRs that was needed for scoring the LSI-R: SV was “some criminal friends” (missing in 40.2% of PSRs) followed by “arrested under age 16” (missing in 23.5% of cases).

The ratings by probation officers on our list of 34 factors were also comparable to the judges. The assessment of risk appeared particularly important when dealing with a violent offender. The assessment of risk to re-offend sexually received a mean rating of 8.9 and 8.6 for general violent re-offending. General risk to re-offend was rated lower (8.1). The concern over violent behaviour and the risk to re-offend violently was rated in the top ten, as it was for the judges (Table 4).

Probation officers ranked substance abuse and psychiatric problems as the 5 th and 6 th most important pieces of information that could be provided in a PSR. These rankings were almost identical to those of the judges where they were ranked 4 th and 5 th. The large majority of probation officers (88.4%) said that PSRs should include recommendations on the treatment needs of offenders. Interestingly, fewer probation officers (59.1%) than judges (74.0%) would like to see PSRs include specific information on programs available in their jurisdiction. Two-thirds of probation officers (66.2%) would like to make recommendations in the PSR concerning custody or community placement. Only 10.1% thought that such recommendations should not be included (23.7% expressed no opinion). When asked to rate the importance of making a placement recommendation, the probation officers assigned a higher rating than the judges (7.5 v. 6.8, t = -2.24, df = 295, p < .05).

In addition to the questions asked to judges, probation officers were asked for their views on how PSRs can be helpful in community supervision. In an earlier pilot study, the Directors of Probation thought that PSRs allow probation officers to become familiar with the client earlier in the supervision process and perhaps even start planning intervention before imposition of sentence. Most probation officers (90.4%) agreed that there were benefits to conducting a PSR assessment prior to supervision although fewer of them (66.7%) said that they end up “almost always” supervising the individual. The probation officers presented a number of advantages to completing a PSR prior to supervision. The advantages were 1) accelerating the intake process (64.6% gave this advantage), 2) facilitates rapport building and continuity of care (43.9%), and 3) provides more information than the standard intake assessment (27.8%).

Views of the Lawyers
Crown attorneys were more likely to express satisfaction with the PSR than defence counsel (83.7% v. 64.5%). Perhaps because of the adversarial nature of the roles of Crown and defence attorneys, it is not surprising that we found a few very divergent views between these two groups. For the defence, a prime motive for having a PSR is to influence the sentence, no doubt toward a community sentence. Sixty percent of defence admitted that their reason to petition the court for a PSR is to influence the sentence (only 9.2% of Crowns gave this reason). In addition, 15.1% of defence said the reason was to “buy time”.

Views on risk assessment were also quite different. Three-quarters (75.5%) of Crown attorneys expressed positive views about the inclusion of risk information in the PSR whereas only 22.6% of defence counsel thought that risk information should be included. Defence counsel was clearly negative in their views about risk assessment; 59.1% thought that it should not be in the PSR (18.3% were neutral on the matter). When asked to rate 32 variables in terms of their importance, the three items dealing with the assessment of risk received some of the lowest ratings from defence counsel (ranged from 5.5 to 5.9 on a scale of 1 to 10). For Crown attorneys the ratings assigned for the three items were higher and ranged from 8.5 to 9.2. The use of standardized risk assessment was the least preferred option with only 16.1% of defence and 30.6% of Crowns selecting this method to present risk information.

There were a few areas where the views of Crowns and defence counsel converged. First of all, they had similar views about the inclusion of custody or community recommendations with 45.9% of Crowns and 50.5% of Defence counsel saying that these recommendations should not be in the PSR. Second, the majority of Crowns and defence counsel (96.9% and 88.2%) thought that treatment needs and programming recommendations should be in the PSR. Third, both groups thought that the PSR is helpful in serious and violent cases (44.9% of Crowns and 30.1% of defence).

Specific Case Analysis
There were 162 PSRs for which information was underlined by both the judges and by the probation officers. There were 247 PSRs submitted but for 85 reports either the probation officer or the judge failed to complete the underlining task. Not all information was equally present in the PSRs with some occurring more frequently than other information. Therefore, the Ns vary for the specific piece of information. Table 5 presents the ten pieces of information that were most frequently underlined by the judges. Along with the judge's underlining, the table also shows the frequency with which probation officers thought that the information would be important to the judges.

Although judges and probation officers may underline for example, current substance abuse, with high frequency (83.3% and 76.4% respectively), it does not necessarily mean that they agree on the importance of the information in the same PSR. That is, a judge may underline information on substance abuse in one PSR case that is not underlined by the probation officer in this same PSR (and vice versa). The Jaccard coefficient shown in the far right column is a measure used to establish similarity between cases described by binary variables, in this case ratings of ''Important v. Not Important''. The coefficient, which ranges from 0 (i.e., no agreement) to 1 (i.e., perfect agreement), is based on positive co-occurrences (i.e., both judges and probation officers underline the information). Thus, Jaccard's coefficient avoids the use of joint absences of a piece of information in the calculation of agreement (i.e., both judges and probation officers do not underline the information).

Table 5 - Ten Most Frequently Underlined Variables by Judges and Their Concordance with Underlining by Probation Officers (N = 162)

BR = Base rate (the proportion of cases where the information appears).

As seen previously in Table 4, information on the offender's substance abuse (current and past) is viewed by the judges as one of the most important variables in PSRs (ranked four). Table 5 further shows that both judges and probation officers largely agreed on this information as being important for sentencing (for current and past patterns of substance use, Jaccard = .76 and .75, respectively). These results suggest that judges and probation officers agreed on the importance of information on the offender's current and/or past substance abuse in approximately three-quarters of the PSRs in which such information was provided. The relatively high values of Jaccard's coefficient found in Table 5 for attitudes toward the offence, relationship problems, and supervision history, and to a lesser but still above chance extent, for the remaining variables indicated that judges and probation officers tended to agree on the relative importance of these variables in sentencing.

Table 6 - Ten Least Frequently Underlined Variables by Judges and Their Concordance with Underlining by Probation Officers (N = 162)

BR = Base rate (the proportion of cases where the information appears).

Taken as a whole, the findings presented in Tables 5 and 6 would suggest that judges and probation officers view the information in PSRs similarly. The discrepancies could be due to the fact that probation officers write about what policy directs them to write and what their own professional experience would suggest may be relevant. Judges on the other hand, read the PSR and take out what they consider relevant and pay little attention to other parts of the report. For example, 73% of PSRs had information regarding the offender's physical health, a common bit of information included in PSRs. However, neither judges nor probation officers underlined this information very frequently (37.8% by judges and 28.6% by probation officers).

PSR and Non-PSR Cases
We thought that it would be informative to examine some of the differences that may exist between cases where a PSR was requested and cases where it was not. Information was collected for 249 PSR cases (i.e., where probation officers submitted an underlined/circled PSR) and 196 non-PSR cases. Three jurisdictions ( Saskatchewan , Quebec , and Northwest Territories ) were unable to provide data on non-PSR cases. Table 7 presents the some of the characteristics of PSR and non-PSR cases. There were no differences in the age of the offenders (33.1 years v. 33.6 years). However, there were a number of statistically reliable differences in the criminal offences committed by offenders for whom a PSR was prepared, their gender, and in most court outcomes. It appears that offenders who had committed personal injury offences and who had multiple convictions were more likely to have a PSR requested. With respect to race, we analyzed data from the jurisdictions with significant Aboriginal populations. Although it appeared that Aboriginal offenders were less likely to have a PSR (36.8% of PSR cases were for Aboriginal offenders compared to 50.8% of non-PSR cases), the differences were statistically unreliable ( c 2 = 5.77, df = 2, p = .056).

Table 7 - The Characteristics of PSR and Non-PSR Cases

* Differences are statistically significant, p < . 05
For Offender Race, data from British Columbia, Alberta, Manitoba and Yukon only.

PSRs and Sentencing
There is a belief that the PSR somehow influences the sentence of the court. As evident in Table 7, the presence of a PSR is associated with a community sentence and treatment recommendations are also more likely in cases where a PSR was completed. In view of the finding that a PSR is more likely requested for cases where serious offences were committed, it appears that the information presented in the PSR mitigates the likelihood of a custodial sentence. Another possible explanation is that judges (or counsel) were already contemplating a community sentence and looked to the PSR for support. One important factor in sentencing is the presence of a recommendation for a community sentence in the PSR. A recommendation for a community sentence (i.e., probation or a conditional sentence) was made in 140 PSRs and judges gave a community sentence in 71.4% of the cases. No such recommendations were made in 98 PSRs and judges gave a community sentence in 53.1% of these cases ( c 2 = 8.43, df = 1, p < .01).

The positive relationship between the recommendations for a community sentence made by the probation officer and the court's sentencing of the case does not mean that it is the only factor taken under consideration and that judges place offenders into the community without regard to risk. The PSRs were reviewed and coded for a range of variables that were present in the reports (e.g., poor supervision history, response to substance abuse treatment, financial problems). Nineteen variables produced correlations of .15 or higher with a community sentence. Many of these 19 variables were then grouped into general categories. For example, Lifestyle Stability consisted of the variables Steady Residence, Gainful Employment/Education, and Positive Financial Circumstances. Criminality consisted of items such as Persistent Criminal History, Poor Supervision History and Antisocial Attitudes. Other variables such as Gender stood by themselves. As shown in Table 8, low risk (i.e., lifestyle stability, and a lack of a criminal lifestyle and social service agency involvement), as expected, showed a significant association with a community sentence.

Table 8 - Variable Domains Related to a Community Sentence

Although the lack of substance abuse was related to a community sentence (Table 8), individuals with a substance abuse problem could still receive a community sentence. Substance abusing offenders were likely to receive a community sentence if there were positive indicators of responding to substance abuse treatment. The presence of treatment plans or a history of positive response to treatment increases the likelihood of a community sentence ( r = .31 and r = .27).

It was expected that offender risk would be related to a community recommendation and placement and that the assessment of offender needs would be related to treatment recommendations and orders. Table 9 summarizes the relationship between actuarial, evidence-based assessments of offender risk and needs and the recommendations in the PSR and court sentences. Some jurisdictions include a summary of an actuarial risk assessment in the PSR (e.g., Saskatchewan , Manitoba ) whereas other jurisdictions do not have this information (e.g., Alberta , Quebec ). Objective, evidenced-based assessments of risk were positively related to a recommendation for a community placement and a community sentence. That is, offenders judged low risk were more likely to have a community recommendation and a community sentence than offenders who are viewed as medium and high risk. For example, 95% of low risk offenders received a community recommendation from the probation officer and 90% were actually placed under community supervision by the courts. With respect to a recommendation for treatment and a treatment order, offender risk level was unrelated (Table 9). For example, 70% of low risk offenders, who by definition would have few criminogenic needs, received a treatment recommendation and only 68% of high-risk offenders (i.e., many criminogenic needs) received a treatment recommendation. Likewise, the courts imposed treatment conditions as part of the sentence for 73.7% of low risk offenders and 69.6% of high-risk offenders.

Table 9 - Community Placement and Treatment Recommendations by Offender Risk (%)

The Resource Demands of a PSR
Time sheets were analyzed for five sites. On average, a probation officer spent 14.1 hours preparing a PSR. However, there was significant variation in the time spent on a PSR across the sites (Table 10). The Edmonton site spent the least time on the preparation of the PSR (6.1 hours) and Whitehorse spent the most time (17.1 hours). Interestingly, staff preparing PSRs in Winnipeg and Whitehorse spent the most time directly with clients, double the time of the other sites. We were unable to ascertain whether variations in the time spent preparing a PSR were related to the quality of the reports.

Table 10 - Resource Demands in Preparing a PSR Across Sites

Discussion

This study surveyed the views and opinions of key criminal justice players on the role of PSRs in sentencing. It is also the largest survey of judges on this topic ever conducted in Canada . Hogarth (1971) examined the views of 78 judges in Ontario and since then there has been no other study that directly examined the views of judges vis-à-vis the PSR. Even the international literature yields only a few studies querying judges (Cavadino, 1997; Norman & Wadman, 2000b; Rush & Robertson, 1987). In addition to asking judges about their views of the PSR, probation officers' opinions were also sought. To our knowledge, there is only one other Canadian study that gathered information from probation officers (N = 765; Hagan, 1975) and one American study (N = 150; Norman & Wadman, 2000b). Finally, this is the only Canadian study reporting the views of Crowns and defence counsel with respect to PSRs.

Despite the low response rate from the judges (61.6%), supplementary analyses suggested that the judges who participated in the study were fairly comparable to judges who did not participate. This conclusion is provided with the caveat that the sample size is relatively small. The participation rate among probation officers was 72.5%. On the whole, the judges and probation officers shared pro-rehabilitation attitudes (over 90% rated rehabilitation as important or very important). However, a surprisingly high number of judges and probation officers (over two-thirds) thought that specific deterrence was important or very important.

The idea of specific deterrence as an important goal of sentencing is inconsistent with the evidence. A number of studies and reviews of the literature of “get tough” interventions clearly show that deterrence has either no effect on reducing offender recidivism (Andrews & Bonta, 2003; Bonta, Wallace-Capretta & Rooney, 2000a; MacKenzie, Brame, McDowall & Souryal, 1995) or it is actually associated with small increases in recidivism (Petersilia & Turner, 1993; Smith, Goggin & Gendreau, 2002). Although judges are required by law to consider deterrence when sentencing, there is a clear need to reinforce among the judiciary and probation officers the ineffectiveness of sanctions on recidivism.

Both judges and probation officers thought that PSRs were particularly useful with serious cases, however, what is meant by “serious” could vary. The most common view of the seriousness of a crime refers to a personal injury offense (e.g., a sexual or nonsexual assault, robbery, impaired driving causing death). Indeed, PSRs were more likely to be requested in cases with a personal injury offence (47.4% v. 26.5% of non-PSR cases; Table 7). However, a personal injury offence is not necessarily indicative of a risk to re-offend (Quinsey, Harris, Rice & Cormier, 1998). When a more comprehensive assessment of offender risk was undertaken, a third of all PSRs were categorized as low risk offenders on the LSI-R:SV with only 24.7% of the PSRs conducted on high risk offenders. Although judges must consider ordering a PSR when a possibility of a custodial sentence exists (Ruby, 2004) it may be helpful for judges and probation officers to be more aware of the distinction between offence seriousness and high risk to reoffend and that the usefulness of PSRs may not be only for “serious” cases.

Risk Assessment in the PSR
The majority of the judges (59.6%) would like to see offender risk information included in PSRs (21.2% were against including this information). If we include those who had no strong opinion on the matter then 78.9% of judges were agreeable to including offender risk information (the comparable figure for probation officers was 75.8%). In addition, both judges and probation officers rated risk assessment information as highly important and frequently underlined this information in PSRs. These findings confirm Cole and Angus' (2003) prediction that “as risk prediction instruments…continue to expand…such a move will likely be welcomed by the Canadian judiciary” (p. 355).

The one group that was less than enthusiastic about risk information in PSRs was defence counsel. More than half of defence counsel (59.1%) thought that risk information should not be included in PSRs (on the other hand, 75.5% of Crown attorneys like this information included). Thus, it is not surprising that in the few cases we could identify, it was the defence that would challenge risk assessment in the courts (R. v. Elliott, 2004).

What is Important in PSRs? The Role of Offender Rehabilitation
In addition to the judiciary's interest in offender risk assessment, issues around offender treatment played a prominent role. This was the one area where almost all the key court officials and probation agreed. At least 88% of probation officers, Crown attorneys and defence counsel wanted treatment recommendations included in the PSR. For the judges, 95.2% wanted treatment recommendations. Although we know that the judges and probation officers had pro-rehabilitative attitudes that could partly explain these findings, we do not know to what extent the results with the lawyers stemmed from their attitudes toward rehabilitation (amazingly, almost every Crown attorney (96.9%) wanted to see treatment recommendations in the PSR).

When judges and probation officers were asked to assign ratings to different variables that could be found in a PSR, treatment factors, especially around substance abuse, were ranked in the top ten (see Table 5). Furthermore, judges and probation officers showed satisfactory levels of agreement (agreeing in approximately 75% of cases when it came to the relevance of substance abuse). The interest in the treatment of offenders is in marked contrast to the trend in PSRs in the United States where the emphasis is more on just deserts and preventative incapacitation. With the widespread use of sentencing guidelines in America , the role and influence of the PSR had diminished and their content was often limited to historical descriptions of fact and social background (Bazelon, 2005; Norman & Wadman, 2000b; Stinchcomb & Hippensteel, 2001). The recent U.S. Supreme Court decision ( January 12, 2005 ; Cohen, 2005) that struck down federal sentencing guidelines may now alter how PSRs are prepared in the U.S.

In Canada , evidence-based, offender risk-need instruments are used in almost every province and territory. The value of risk-need instruments is that they provide information that goes beyond decisions about level of supervision or custody. They provide information on the needs of offenders that must be addressed in order to reduce recidivism (Andrews & Bonta, 2003; Bonta, 2002). That is, risk-needs instruments play an important role in offender rehabilitation. Consequently, using risk-needs assessments to either structure the PSR (e.g., as in Ontario and Nova Scotia ) or to directly communicate to the court what needs to be done to manage offender risk (e.g., as in Manitoba and Saskatchewan ) could serve the courts well by providing treatment information.

A question remains as to how can PSRs be improved to give the courts better treatment information. At present, many PSRs are a mix of treatment relevant information, information that is largely unrelated to treatment, and important information that can inform treatment but is missing. For example, nearly three-quarters of PSRs described the offender's health and both judges and probation officers felt that this information was of little interest (Table 6). In addition, information on the offender's social circle was evident in 55% of PSRs and furthermore, when it was present, judges and probation officers did not think that this information was of much value. However, knowledge of the offender's social circle is probably one of the best predictors of criminal behaviour and a highly relevant treatment target (Andrews & Bonta, 2003; Gendreau, Little & Goggin, 1996). In order to improve the content of PSRs consideration should be given to the information in Tables 5 and 6 in combination with the empirical evidence on the criminogenic needs of offenders.

The Influence of the PSR on Sentencing
Perhaps because many judges see probation officers as knowledgeable about offender behaviour, a majority of judges (59.6%) would like the PSR to include sentencing recommendations. However, some judges (11.5%) were against the inclusion of sentencing recommendations with another 28.8% having no opinion on this matter. The judges who were neutral on the subject could be influenced by defence objections to such recommendations and if we add them to judges who are clearly against sentencing recommendations we have a significant proportion of judges who may question probation officers who include sentencing recommendations in the PSR (40.3%). However, interest was particularly high among the judges when a recommendation for community placement was made (albeit half of the lawyers were opposed to such recommendations). When a community recommendation was made in the PSR it was followed by the courts in 94.2% of cases, it was underlined in the PSRs by the judges in 73.2% of cases (Table 5) and judges gave a community sentence more frequently in PSR cases than in non-PSR cases (Table 7).

Of course, judges did not follow blindly the community recommendations made by the probation officer. Judges paid attention to factors they considered indicative of re-offending (e.g., criminal history). Other variables that increased the likelihood of a community sentence were lifestyle stability, the lack of a substance abuse problem and minimal criminality (Table 8). In general, low risk offenders were most likely to receive a community placement especially if it was supported by an evidence-based risk assessment (when an actuarial risk assessment was conducted 95% of low risk offenders were recommended for community placement compared to 79% when the assessment was based on clinical judgment).

Forty-four per cent of high-risk cases were recommended for community placement (Table 9). In this study, we found that PSRs would more likely be ordered in serious cases and that the courts were more likely to give a community sentence, especially a conditional sentence (although strictly speaking a conditional sentence is a jail sentence). In Canada , conditional sentences permit the judge to order the offender to serve their time in the community rather than being placed in prison (Gemmell, 1999). Lundrigan (1998) also found conditional sentences more prevalent among cases where a PSR was requested. Thus, it appears that the presence of a PSR encourages judges to take a risk and place some offenders into the community who otherwise would have been sentenced to incarceration. This conclusion about a willingness to take a risk is not new and has been noted by others (Campbell et al., 1990; Hogarth, 1971; Weinrath, 1999).

Taking risks with some offenders, however, could be problematic. The offender treatment literature does provide evidence that some higher risk offenders can be successfully treated provided they are given the right programming (Andrews & Bonta, 2003). Therefore, the treatment recommendations in PSRs must be appropriate for the offender. If PSRs provide relevant treatment recommendations, and they are actually delivered, then we may expect reductions in recidivism among offenders with a PSR. However, prior research has found that the presence of a PSR is not associated with reduced recidivism (Davies, 1969; Rumney & Murphy, 1952; Weinrath, 1999).

One critical principle of effective offender treatment programs is the matching of offender risk level to the intensity of programming (Andrews, Bonta & Hoge, 1990; Bonta, Wallace-Capretta & Rooney, 2000b; Bourgon & Armstrong, 2005). In this study, probation officers showed poor adherence to the risk principle of effective rehabilitation. As seen in Table 9, more than half of low risk offenders had a treatment recommendation whereas the research literature would suggest that these offenders would require very little treatment. In addition, only two-thirds of high-risk offenders had a treatment recommendation when all of them probably had treatment needs.

Summary and Conclusions

Looking back to Hogarth's (1971) landmark study on Canadian sentencing, we found many similarities as well as important differences between Hogarth's (1971) investigation and our current findings. Judges continue to express general satisfaction with the PSR. Hogarth (1971) reported that approximately two-thirds of judges in his sample were satisfied and this level of satisfaction rose to 87.4% in the present study. Not everyone, however, expressed such high satisfaction levels and this was most evident among probation officers. The results point to a need for better training on the preparation of the PSR and a dialogue with the judiciary to ensure that non-serious and “frivolous” cases are not referred for a PSR investigation.

Perhaps the most important changes that have occurred over the past 30 years are the adoption of evidence-based, risk-needs assessment instruments and a better understanding of effective offender rehabilitation programming. In the 1970s, no probation department in this country was using an actuarial risk instrument. Assessments of risk were based upon clinical and professional judgments. Hogarth (1971) found that PSR were used mostly for serious offences (e.g., robbery, assault causing bodily harm, indecent assault, etc.) but there was no recognition that the seriousness of the offence is often unrelated to the risk to re-offend. Given that PSRs appear to have an influence on sentencing, the best available assessment methods are needed to ensure appropriate sentences. The contemporary use of evidence-based risk-needs instruments to either structure the PSR investigation or to report directly upon offender risk and needs in the reports provide the type of information that can better inform the courts. For the most part, judges welcome this information.

The judges in this study held favourable attitudes toward offender rehabilitation. Hogarth (1971) also found judges to rank “reformation” highly although in practice specific deterrence was weighted more heavily (pp. 288-289). Attitudes toward rehabilitation undoubtedly influenced the keen interest from the judges in information on treatment. Information on treatment needs were valued by all the key actors in the sentencing process and judges were particularly interested in treatment recommendations.

If we can summarize the major changes over the past three decades in what judges expect from PSRs and what is being reported in PSRs then it would be this: PSRs have become treatment oriented, case management reports. In 1971, Hogarth found that family background was cited by judges as an “essential” piece of information in 61% of cases whereas offender attitude to rehabilitation, mental condition and use of alcohol and drugs were cited as essential less than 20% of the time. In the present study, the relevance of these pieces of information has reversed with treatment information most highly valued by judges and most frequently found in PSRs. The trend is clear: PSR will become more and more a document that provides a risk-need description of offenders along with a formulation of a strategy for managing risk. More importantly, PSRs are becoming increasingly evidence-based and this transformation will translate into benefits for judges by assisting in sentencing.

References

Andrews, D. A., & Bonta, J. (1995). The Level of Service Inventory–Revised: User's Manual. Toronto , ON : Multi-Health Systems.

Andrews, D. A., & Bonta, J. (1998). The Level of Service Inventory–Revised: Screening Version User's Manual. Toronto , ON : Multi-Health Systems.

Andrews, D. A., & Bonta, J. (2003). The Psychology of Criminal Conduct (3 rd. ed.). Cincinnati , OH : Anderson .

Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52.

Bazelon, E. (2005). Sentencing by the numbers. New York Times Magazine, (January 2), 18-19.

Boldt, E. D., Hursh, L. E., Johnson, S. D., & Taylor, K. W. (1983). Presentence reports and the incarceration of Natives. Canadian Journal of Criminology, 25, 269-276.

Bonta, J. (1996). Risk-needs assessment and treatment. In A.T. Harland (Ed.), Choosing Correctional Options that Work: Defining the Demand and Evaluating the Supply (pp. 18-32). Thousand Oakes , CA : Sage Publications.

Bonta, J. (2002). Offender risk assessment: Guidelines for selection and use. Criminal Justice and Behaviour, 29, 355-379.

Bonta, J., Harris, A., Zinger, I. , & Carrière, D. (1998). The dangerous offender provisions: Are they targeting the right offenders? Canadian Journal of Criminology, 40, 377-400.

Bonta, J., Wallace-Capretta, S., & Rooney, J. (2000a). Can electronic monitoring make a difference? An evaluation of three Canadian programs. Crime and Delinquency, 46, 61-75.

Bonta, J., Wallace-Capretta, S., & Rooney, J. (2000b). A quasi-experimental evaluation of an intensive rehabilitation program. Criminal Justice and Behavior, 27, 312-329.

Bonta, J., & Yessine, A. K. (2005). The National Flagging System: Identifying and responding to high-risk, violent offenders. (User Report 2005-03). Ottawa : Public Safety and Emergency Preparedness Canada .

Bourgon, G., & Armstrong, B. (2005). Transferring the principles of effective treatment into a “Real World” prison setting. Criminal Justice and Behavior, 32, 3-25.

Campbell, C., McCoy, C., & Osigweh, C. A. B. (1990), The influence of probation recommendations on sentencing decisions and their predictive accuracy. Federal Probation, 54(4), 13-21.

Carter, R. M. (1966). It is respectfully recommended. Federal Probation, 30(2), 38-42.

Carter, R. M. (1969). Variations in presentence report recommendations and court dispositions. Research Report, 2(5), 27. Washington State Department of Institutions.

Carter, R. M., & Wilkins, L. (1967). Some factors in sentencing policy. Journal of Criminology and Police Science, 38, 253-273.

Cavadino, M. (1997). Pre-sentence reports: The effects of legislation and national standards. British Journal of Criminology 37, 529-548.

Cohen, A. (2005). Putting “guide” back in the guidelines. CBS News.com ( January 12, 2005 ).

Cole, D. P. (2001). Using PSRs to evaluate and respond to risk. Paper presented to the National Judicial Institute Criminal Law Seminar, February.

Cole, D.P. & Angus, G. (2003). Using pre-sentence reports to evaluate and respond to risk.
Criminal Law Quarterly 47, 302-357

Davies, M. (1969). Probationers in their Social Environment. London : Her Majesty's Stationary Office.

Deane, H. (2000). The influence of pre-sentence reports on sentencing in a district court in New Zealand . Australian and New Zealand Journal of Criminology, 33 , 91-106.

Gabor, T., & Jayewardene, C. H. S. (1978). The pre-sentence report as a persuasive communication. Canadian Journal of Criminology, 20, 18-27.

Gelsthorpe, L. & Raynor, P. (1995). Quality and effectiveness in probation officer's reports to sentencers. British Journal of Criminology 35, 188-200.

Gemmell, J. (1999). Conditional sentences. In J. V. Roberts & D. Cole (Eds.), Making sense of sentencing. (pp. 63-76). Toronto : University of Toronto Press.

Gendreau, P., Little, T., & Goggin, C. (1996). A meta-analysis of the predictors of adult offender recidivism: What works! Criminology, 34, 575-607.

Gibson, R. E. (1973). Recommending sentence – New Zealand probation officer's presentence reports. Victoria , New Zealand : Australian and New Zealand Society of Criminology, 231-239.

Grove, W. M., & Meehl, P.E. (1996). Comparative efficiency of informal (subjective, impressionistic) and formal (mechanical, algorithmic) prediction procedures: The clinical-statistical controversy. Psychology, Public Policy, and Law 2, 293-323.

Grove, W.M., Zald, D.H., Lebow, B.S., Snitz, B.E., & Nelson, C. (2000). Clinical versus mechanical prediction: A meta-analysis. Psychological Assessment 12, 19-30.

Hagan, J. (1975). The social and legal construction of criminal justice: A study of the presentencing process. Social Problems, 22, 620-637.

Hogarth, J. (1971). Sentencing as a Human Process. Toronto , ON : University of Toronto Press.

Lundrigan, V. (1998). “It is recommended…” A Study of the Impact of the Pre-sentence Report on Sentencing. Unpublished Master's Thesis. Memorial University , St. John's , Newfoundland .

MacKenzie, D. L., Brame, R., McDowall, D., & Souryal, C. (1995). Boot camp prisons and recidivism in eight states. Criminology, 33, 327-357.

Norman, M. D., & Wadman, R. C. (2000a). Probation department sentencing recommendations in two Utah counties. Federal Probation, 64(2), 47-51.

Norman, M. D., & Wadman, R. C. (2000b). Utah presentence investigation reports: User group perceptions of quality and effectiveness. Federal Probation, 64(1), 7-12.

Petersilia, J., & Turner, S. (1993). Evaluating intensive supervision probation/parole: Research of a nationwide experiment. Washington , DC : National Institute of Justice, Research in Brief.

Quinsey, V. L., Harris, G. T., Rice, M. E., & Cormier, C. A. (1998). Violent offenders: Appraising and managing risk. Washington , DC : American Psychological Association.

R. v. Elliott [2004] Provincial Court of Nova Scotia.

R. v. Gladue [1999] S.C.R. 688.

R. v. Moser [2002] O.J. No. 1344 (QL) (S.C.J.).

R. v. Proulx [2000] 1 S.C.R. No. 61.

Raynor, P. (1996). Evaluating probation: the rehabilitation of effectiveness. In T. May & A. A. Vass (Eds.) Working with Offenders: Issues, Contexts and Outcomes. (pp. 242-258). London : Sage.

Raynor, P., Gelsthorpe, L., & Tisi, A. (1995). Quality assurance, pre-sentence reports and the probation service. British Journal of Social Work, 25, 477-488.

Rosecrance, J. (1985). The probation officers' search for credibility: Ball park recommendations. Crime & Delinquency, 31, 539-554.

Ruby, C. (2004). Sentencing (6 th ed.). Toronto , ON : LexisNexis/Butterworth.

Rush, C., & Robertson, J. (1987). The utility of information to the sentencing decision. Law and Human Behavior, 11, 147-155.

Rumney, J., & Murphy, J. (1952). Probation and social adjustment. New Brunswick , NJ : Rutgers University Press.

Smith, P., Goggin, C., & Gendreau, P. (2002). The effects of prison sentences and intermediate sanctions on recidivism: General effects and individual differences. (User Report 2002-01). Ottawa : Public Safety and Emergency Preparedness Canada .

Stinchcomb, J.B. & Hippensteel, D. (2001). Presentence investigation reports: A relevant justice model tool or a medical model relic? Criminal Justice Policy Review 12, 164-177.

Thorpe, J., & Pease, K. (1976). The relationship between recommendations made to the court and sentences passed. British Journal of Criminology, 16, 393-394.

Weinrath, M. (1999). Are new directions warranted for the presentence report? An empirical assessment of its predictive utility in the adult court system. Journal of Crime and Justice 22, 113-129.

Appendix A - Ratings of Specific Information

** Mean Differences are statistically significant, p < .01.
* Mean Differences are statistically significant, p < .05.
^a Some cases missing due to unavailability of data.

Date modified:

2016-01-19