An Estimation of the Economic Impact of Violent Victimization in Canada, 2009

• Previous Page
• Table of Contents
• Next Page

Methodology

Sources of Economic Impacts

Victimization of all five violent crimes previously mentioned (assault, criminal harassment, homicide, robbery, and sexual assault and other sexual offences), where the victims are adults (18 and over) and are not in a spousal relationship with the offender, are the sources of economic impacts that are captured in this report. The non-spousal relationship condition is imposed to avoid overlap with a previous costing report (Zhang et al. 2012), while the victim age condition is imposed to avoid overlap with any future costing research on the criminal victimization of children and youth. Offenders of all ages are included. If an impact is deemed to be mainly the result of a victimization then it is included. Any impact that does not derive from one of the five crimes is not included. Therefore, any costs or impacts incurred by the offender are not counted, as these do not result from the victimization of the offence, but from the perpetration of the offence. The decision to exclude offender costs is also supported by some economic theorists (Trumbull 1990) and, more importantly, by the lack of data that would be required to estimate offender costs.

One partial exception to this condition of not counting offenders costs is the inclusion of fines in the corrections sub-category of the criminal justice system category. Fines are theoretically a cost to the offender and a benefit to the criminal justice system (i.e., the rest of society), and because of this fines are included in this report as a partial exception only. They are included solely for illustrative purposes and are not actually included in the rest of the analysis or in the summations of total costs. Also, in practice, fine sentences can generate a net cost to the justice system due to high costs of enforcement.

To illustrate the sources of economic impacts, here are some concrete examples of crime scenarios that would be included in this report: an incident involving a 15 year old youth assaulting a 25 year old adult and an incident involving a 25 year old adult assaulting a 25 year old adult. However, given the definitions of victim and offender in Definitions, impacts associated with an assault by a 25 year old adult on a 15 year old youth would not be counted in this report, nor would an assault by a 15 year old youth on a 15 year old youth.

Geography

Theoretically, the scope of this study encompasses all Canadian provinces and territories. However, microdata for the 2009 GSS survey of the territories are not available for analysis, so all cost items based on the GSS encompass only the ten provinces. All cost items related to the criminal justice system are based on the Uniform Crime Reporting Survey 2 (UCR2) and court data and so encompass both the provinces and territories.

Time Period

The timing of the incident is the factor that determines what is included in the analysis; the timing of the impact or cost is irrelevant. Therefore, all incidents that occurred in 2009 are included, and, theoretically, all costs derived from these incidents are measured. For example, if an assault took place in 2009, all costs resulting from that assault would be included, regardless of when those costs were incurred. If the subsequent criminal court case took place in 2010, the costs associated with that court case are included; if the offender was sentenced to one year of provincial custody, to be served in 2011, those corrections costs are included as well.

Costs that are incurred long after 2009 do present some obstacles for accurate estimation. There are numerous interrelated factors influencing costs (e.g., prices, living costs, salaries, currencies, inflation) and these can change greatly over time. As the future levels of these factors cannot be known in the present, estimates of costs incurred in the future must be made based on current information of economic characteristics. Therefore, estimates of future cost items (e.g., lost future income due to mental health disability) are very tentative, as the true costs are tied to the future variation of economic factors.

Gender

All results throughout the report are calculated separately for male and female victims, to provide those in policy and programming with greater detail and context and allow them to target these issues more effectively. Some analyses in the corrections cost groups also take into consideration the gender of the offender.

Crime Categories

Five major crime categories are included in this report: assault, criminal harassment, homicide, robbery, and sexual assault and other sexual offences. These major crime categories contain specific crimes, which differ between the major data sources used. Table G.1 lists the specific crimes under each crime category according to each data source. Note that data from self-reported sources, court-reported sources, and police-reported sources are not necessarily comparable, but for the purposes of this report the specific crimes in each data source have been matched up to allow for a more comprehensive costing exercise. Each major crime category is analyzed individually in a separate section of the report: Assault, Criminal Harassment, Homicide, Robbery, Sexual Assault and Other Sexual Offences. For complete definitions of the crimes and for full analyses of cost estimations, see the appropriate sections.

Assault

Assault is identified in the self-reported GSS by setting the most serious crime variable to “assault” (MSCRIME = 304). Assault in the court-reported Adult Criminal Court Survey (ACCS) and Youth Court Survey (YCS) is separated into two offences: “major assault” and “common assault”. Police-reported data in the UCR2 includes many different types of assaults, which, for this report, are placed into groups designed to be reasonably equivalent to the court-reported aggregations. “Assault level 3”, “assault level 2”, and “assault against a police officer” in the police-reported data are deemed to be equivalent to “major assault” in the court-reported data; “assault level 1” and “other assaults” in the police-reported data are deemed to be equivalent to “common assault” in the court-reported data.

Criminal harassment

Criminal harassment is more commonly known as stalking, and is defined by the Department of Justice as behaviour that “consists of repeated conduct that is carried out over a period of time and that causes victims to reasonably fear for their safety but does not necessarily result in physical injury. It may be a precursor to subsequent violent acts” (Department of Justice 2004). For purposes of this report, three specific offences are included: criminal harassment (stalking), uttering threats, and threatening or harassing phone calls.

Criminal harassment is a unique crime in the self-reported GSS. It is not included as a reportable crime in the Incident File (where the crime incident reports are found for assault, robbery, and sexual assault), but there is a question asking “have you been stalked … ?” in the Main File of the 2009 GSS. Unfortunately, no further details regarding the nature of the victimization are collected in the 2009 GSS, and so it is of limited use in calculating the costs of criminal harassment. However, the 2004 GSS contains an entire module on stalking, much of which is useful for costing. For the most complete costing results, the 2004 GSS module on stalking is used and the data are adjusted using the 2009 GSS question on stalking.

In the court data ACCS and YCS: only “criminal harassment” and “uttering threats” are categorized. On the other hand, the police-reported UCR2 has all three relevant offences – “criminal harassment”, “uttering threats” and “threatening or harassing phone calls” where the latter two offences will be grouped into “uttering threats” when estimating the costs at the court stage.

Table G.1: Crimes as specified in major data sources

Category	Self-reported data Notes de table A	Court-reported data Notes de table B	Police-reported data Notes de table C
Assault	Assault (MCRIME=304)	Major assault	Assault — level 3 — aggravated Assault — level 2 — weapon or bodily harm Assault against a police officer
		Common assault	Assault – level 1 Other assaults
Criminal harassment	Stalking (OCE_Q180=1, GSS 2009) (STALKING=1, GSS 2004)	Criminal harassment	Criminal harassment
		Uttering threats	Uttering threats Threatening or harassing phone calls
Homicide	Not available	Homicide	Murder – 1st degree Murder – 2nd degree Manslaughter
Robbery	Robbery (MSCRIME=202)	Robbery	Robbery
Sexual assault and other sexual offences	Sexual assault (MSCRIME=101)	Sexual assault	Sexual assault — level 3 — aggravated Sexual assault — level 2 — weapon or bodily harm Sexual assault – level 1
		Other sexual offences	Incest Anal intercourse Voyeurism Other sexual violations

Homicide

Homicide is one of the most difficult crimes to calculate economic impacts for because details surrounding the nature of the offence and information about the victim are often not available. Self-reported data being obviously non-existent, the only sources available for homicide are court-reported and police-reported. The court-reported ACCS and YCS list only one “homicide” offence, while three offences from the police-reported UCR2 are included here: “1st degree murder”, “2nd degree murder”, and “manslaughter”.

Robbery

Robbery should not be confused with theft or any of its related offences. Robbery is classified as a violent violation in the police-reported UCR2, while theft and other related crimes are classified as non-violent violations. Robbery can essentially be thought of as theft with violence or the threat of violence where both the victim and offender are physically present. In the court-reported ACCS and YCS robbery is classified as a crime against the person, while theft and other related offences are classified as crimes against property. Therefore, all theft and break and enter violations have been excluded.

Identifying and defining robbery is straightforward in the three main data types. Robbery is identified in the GSS by setting the most serious crime variable to “robbery” (MSCRIME = 202). The court-reported data sources contain only one “robbery” offence, while only the “robbery” offence is included here from the police-reported data (“robbery to steal a firearm”, while included theoretically, has been dropped as there were zero relevant incidents in 2009).

Sexual assault and other sexual offences

The sexual assault and other sexual offences category presents many difficulties for costing. One data issue is the recent reconfiguring of the crime formerly called “other sexual violations” in the police-reported data. According to the Uniform Crime Reporting Survey’s 2011 and Historical Canada/Province/CMA Note document:

“Other sexual violations” officially expired in 2008 and divided into the following violations: sexual interference, invitation to sexual touching, sexual exploitation, incest, corrupting children, luring a child via a computer, anal intercourse, bestiality/commit/compel/incite a person, voyeurism. Police services have been able to utilise these new codes over the past few years as their Records Management Systems have been updated to allow it. Comparison with previous years’ data should be done with caution. (Statistics Canada 2012, p. 6)

Though officially expired, there were still incidents recorded as “other sexual violations” in 2009, which causes challenges in the calculations of some cost items.

In the self-reported GSS there is one sexual offence, called “sexual assault”. In this report, victims were classified as sexual assault victims if the most serious crime variable was set to “sexual assault” (MSCRIME = 101). Sexual assaults in the GSS can be separated into two types (“sexual attacks” and “unwanted sexual touching”) based on victim responses about the nature of the crime, and these two types will be utilized when estimating pain and suffering. The court-reported ACCS and YCS simply record two types of sexual crimes: “sexual assault” and “other sexual offences”. As mentioned earlier, the police-reported UCR2 disaggregated the crime “other sexual violations” into many offences in 2008. Thus, apart from the three levels of sexual assault, this report is concerned with the police-reported crimes of “incest”, “anal intercourse”, “voyeurism”, and any residual counts of “other sexual violations”. These offences may involve adult or youth victims (though in this report only adult victims are included), but other crimes disaggregated from “other sexual violations” can only involve youth victims, such as “invitation to sexual touching” and “luring a child via a computer”. “Sexual exploitation of a person with a disability” would also be included in this report but there were no relevant counts in 2009.

Cost Categories

The costs of each crime are separated into three cost categories: justice system costs, victim costs, and third-party costs. Each of these cost categories contains sub-categories which might be further broken down into cost groups. For example, under the category “justice system costs” is the “criminal justice system costs” sub-category which contains “corrections costs” cost group. Finally, both sub-categories and cost groups may contain cost items which are the specific costs that are estimated For example, both “court costs” and “correction costs” directly belong to the “criminal justice system costs” sub-category – while “court costs” is one cost item, many costs items such as “federal custody costs” are included under the cost group “corrections cost”.

Cost items are placed into cost categories on the basis of which party bears the actual impact of that cost item, not on the basis of who bears the financial burden of that cost item. For example, the sub-category “medical costs” and all of its cost items are placed in the “victim costs” cost category because the victim bears the impact of any medical costs (i.e., the injury or any physical or emotional pain resulting from the victimization that requires medical treatment) even though much of the actual expenditure is paid for by third-parties (the state or businesses through insurance).

Justice system costs

The justice system is comprised of the criminal justice system and the civil justice system. The criminal justice system is of more interest than the civil justice system for estimating the costs of crime. Although the civil justice system can play a major part in redressing some criminal acts (e.g., suing for damages in civil sexual assault cases), estimating the costs of many civil justice components is not possible at this time due to the lack of data in Canada.

Criminal justice system

The three major components of the criminal justice system are all included in this report: police (“police costs”), courts (including “court costs”, “prosecution costs”, and “legal aid costs”), and corrections. Corrections is further categorized into “federal custody costs”, “provincial custody costs”, “conditional sentence costs”, “probation costs”, and “fines”. “Fines” are unique in that they are theoretically an economic benefit to the justice system, and are therefore not included in any cost summations (see Sources of Economic Impacts). All cost items under the criminal justice system are included for every crime.

In addition to the potential corrections outcomes already listed, there are certain “other” sentences that may be given to the offender, including restitution, absolute discharge, conditional discharge, suspended sentence, payment of legal costs, and suspension of drivers license for adults, and absolute discharge, restitution, prohibition, seizure, forfeiture, compensation, pay purchaser, essays, apologies, counselling programs, and conditional discharge, among others, for youth. ^{Footnote 6} The costs of all “other” sentences are not analyzed due to a lack of available data.

Civil justice system

The civil justice system differs from the criminal justice system in that the impetus for justice comes from the victim, not the state. For many victimization incidents covered in the scope of this report, the civil justice system will not play a role, while many victims who may find some aspect of the civil system useful also find barriers in the time, the cost, and the emotional toll. One specific scenario of the civil justice system that is relevant to this report is the case in which a victim of sexual assault sues the alleged offender for damages. However, due to lack of data and a lack of contextual clarity, ^{Footnote 7} no cost estimates are made for this scenario, but a brief discussion on the subject is provided in Text Box S.1.

The only cost item included in this report for the civil justice system is “restraining or protection order costs” for criminal harassment.

Victim costs

Victims bear the greatest and most direct impacts of violent crime. The traumatic experience of being a victim may itself cause severe pain and suffering, and in many cases this is actually the greatest impact of the crime. Pain and suffering is considered an intangible cost because no financial transaction takes place and therefore its “value” is not determined by a market. The loss of life of homicide victims is the other significant intangible cost in the victim cost category.

However, there are also many tangible costs of crime for victims. Medical treatment may be required for either a physical injury or a deterioration of mental health resulting from the victimization. The victim may be forced to take time off of work and consequently lose wages due to either physical incapacitation or emotional trauma, while longer term physical and mental effects may significantly impact future earnings. Victims may also incur other miscellaneous costs mostly related to property and possessions.

Medical costs

Though much of the medical payments are made by the state or businesses, victims bear the impacts (i.e., injuries or emotional effects) of those medical costs. Assault, robbery, and sexual assault victims may seek medical attention from doctors or nurses (“doctor or nurse service costs”), or in serious cases they may visit the emergency department (“emergency department costs”) or have to stay overnight in the hospital (“overnight hospitalization costs”). Homicide victims may also visit the emergency department in cases where the victim is not yet deceased and attempts are made to save his or her life in the hospital.

As for long-term medical impacts, victims of all crimes (except homicide) may be emotionally traumatized and may seek help in counselling (“counselling costs”); data from the GSS show that many victims do indeed utilize counselling services. Data for long-term “medication costs” and “physical therapy costs” are severely limited, but what little information is available is included for relevant crimes as they are important components of medical costs. The last cost item discussed is the “costs of suicide attempts” in cases of sexual assault victimization. Sexual assault is a crime that, depending on the specific circumstances, potentially involves elements beyond the physical, including sexual, emotional, and spiritual (Frampton 1998). With such severe impacts, victims of sexual assault may fall into emotional turmoil, depression, and suicidal thoughts and behaviour. Victims who attempt suicide may cause harm to themselves, and emergency medical treatment may be required.

Text Box G.1 gives a brief explanation of the different types of counselling services available, how the costs of these services can vary greatly, and the use of counselling within and beyond victims services.

Lost productivity

Victims of assault, robbery, and sexual assault may all experience lost wages for taking time off work to recover in the short term (“lost current income”), but they may also experience long-term work loss or decreased earnings if they are permanently impacted either physically (“long-term physical disability costs”) or mentally (“mental health disability costs”). Victims of these three crimes who are not employed may also suffer from decreased productivity, whether in school (“lost education”) or in the home (“lost household services” and “lost child care services”). The lost productivity of criminal harassment victims is not estimated due to insufficient data.

Intangible costs

Much of the impact felt by victims is intangible. The two intangible victim costs estimated in this report are “pain and suffering costs” and “loss of life costs”. These intangible costs are very difficult to measure for a number of reasons, and even attempts to measure them can be contentious, as discussed in Valuation of Intangibles. To provide as comprehensive an estimate as possible, though, intangibles such as these must be included as they are a large component of the actual costs of crime. For details on the methods and sources used to estimate intangibles in this report, and brief discussions of the literature, see Valuation of Intangibles.

Other costs

“Stolen, damaged, or destroyed property” is one of the main costs in robbery cases and criminal harassment cases, but there may also be incidental property costs in assault and sexual assault cases. Criminal harassment victims attempting to avoid their stalker may activate special phone features such as caller ID (“special phone feature costs”), or in severe cases may actually move away from his or her current home. This latter action involves measurable costs (“moving and change of address costs”), but significant intangible costs are also involved (e.g., relocation stress, loss of contact with friends and family, loss of familiarity with home area, new schools for children, etc.) that cannot be estimated due to data limitations.

Third-party costs

Though third parties do not experience any direct impacts of someone’s victimization, they can be affected in significant ways. Employers can experience many potential costs when their employee is victimized, from lost wage expenditures to less productivity on the job. Social services, funded by governments, businesses, non-profit organizations, and individual donors, play a major role in providing support for victims. Governments are also active in providing funds for prevention, awareness, and assistance programs and campaigns. Family members may sustain costs beyond the intangible costs and funeral service costs included here.

Employer losses

Victims who are employed and who take time off of work to recover may incur a cost to employers if they have paid leave time available (“lost wage expenditure”), and for all victims who take time off of work there will be “administration costs” to the employer for the work days missed. Crimes that have major psychological impacts on victims may cause victims to be late for work and to become distracted while at work (“tardiness and distraction costs”). All employers whose employees are absent due to victimization will face “lost additional output” as their workers are not producing the expected returns on their work. As the stalking module of GSS does not contain questions on time lost and daily main activity, the available data permit estimation of employer losses only for assault, robbery, and sexual assault.

Social services operating costs

Many entities are active in providing support and care through social services. The costs of two types of social services are found in this report: “victim services costs” and “crisis centre and crisis line costs”. Usage numbers for these services are gathered from the GSS, which asks victims if they contacted or used these services, and from other government reports. Some confusion exists between the GSS questions on social services usage and the GSS question asking respondents if they had contacted or used a counsellor or psychologist because counselling services may also be available through some social services. In order to obtain estimates, it is assumed that the questions are exclusive of each other, and that counselling costs are borne by the victim, while social services operating costs are not.

Text Box G.2 provides information on criminal injuries compensation and financial benefit programs that are administered through social services. These programs are not included in the cost analysis as most of the reasons for giving compensation are already included as cost items in the report, and including compensation would therefore lead to double counting of the economic impact of victimization. For example, compensation for pain and suffering accounts for the majority of the total monetary awards, but pain and suffering is already calculated as a separate cost item in this report, so including compensation would count the value of pain and suffering twice. Similarly, compensation is often given for medical expenses or lost wages, both of which are also already included in the report.

Intangible costs

Though less contentious than the intangible costs of victims (see Intangible costs), the intangible “loss of affection and enjoyment to family” for the families of homicide victims still brings difficulties. The methods for measuring this cost are less developed than the more mature methods of valuing pain and suffering and loss of life. For details on the methods and sources used in this report see H.T.3.1 Loss of Affection and Enjoyment to Family.

Other costs

An additional cost applicable to homicide, usually borne by the family of the victim or government assistance, is “funeral service costs”.

Other government expenditures

Both the federal and provincial/territorial levels of government are active in crime prevention, education and raising awareness for victims of crime and their families, as well as the general public. In addition, governments undertake research, the rehabilitation of violent offenders and provide assistance to victims of violent crime through services and also project funding. The vast majority of both government funding for third-party service delivery and direct government expenditures on programs and policies related to violent victimization are already included elsewhere in this report, one example being the large amount of funding provided by governments for the delivery of social services, including support centres, crisis lines, and victim services.

This report examines each of five categories of violent victimization: homicide, sexual assault, assault, criminal harassment and robbery. Data on other government expenditures were not available by category of victimization or by gender. While there was information on expenditures related to sexual assault for some jurisdictions, overall the information was too vague to be of specific use. As such, this section will provide a qualitative description of federal and provincial/territorial expenditures beyond the expenditures already captured in the report.

Federal government expenditures

As noted above, it was not possible to estimate specific federal government expenditures associated with violent victimization. Focusing on one work unit, the Policy Centre for Victim Issues (PCVI) at the Department of Justice Canada, will provide an in-depth example of the additional expenditures by the federal government for victims of violence.

The PCVI was established in 2000 in response to the parliamentary report Victims Rights – A Voice Not a Veto, which recommended the development of both a victims of crime strategy and an office to respond to victims issues within federal jurisdiction. The current mandate of the PCVI is to work toward giving victims a more effective voice in the criminal justice system by:

• helping victims and their families understand their role in the criminal justice system and the laws, services and assistance available to support them;
• ensuring that the perspectives of victims will be fully considered when relevant federal laws and policies are developed.
• increasing awareness both within Canada and internationally about the needs of victims of crime and effective approaches to respond to those needs. ^{Footnote 8}

A key component of the PCVI is its collaborative approach to justice system issues for victims of crime. The PCVI works with federal department and agency partners to ensure a shared federal perspective, foster collaboration and provide opportunities for joint action. Other federal partners include: Correctional Services Canada, the Royal Canadian Mounted Police, the Parole Board of Canada and the National Office for Victims at Public Safety Canada. In addition, the PCVI works closely with its provincial and territorial counterparts as responding to the needs of victims is a shared responsibility in Canada.

In 2009, federal victimization expenditures included direct and cost shared amounts related to a wide variety of activities and programming directed to prevention and awareness, services, research, policy development, consultations, symposia, education and training. The operational budget of the PCVI in 2009/10 was approximately $4M, the same as in 2008/09.

The Victims Fund, which is the funding program of the PCVI, totalled $7.75M in the same year. While not all components of the Victims Fund are relevant to adult victims of violence in Canada (e.g. Canadians Victimized Abroad), there are many components that are relevant. Descriptions of these components follow with the dollar amounts and are taken from the 2009-2010 Department of Justice Canada Departmental Performance Report which is tabled in the House of Commons each year. ^{Footnote 9}

Provincial and Territorial Funding

In 2009/2010, the Victims Fund provided $712,997 in contribution funding to provinces and territories to help meet the needs of under-served victims of crime and/or to help support victims in attending sentencing hearings and in submitting their Victim Impact Statements

In addition, the Victims Fund provided $367,941 in contribution funding to provinces and territories to support implementation of legislation to benefit victims or to advance the Canadian Statement of Basic Principles of Justice for Victims of Crime.

Finally, a total of $421,596 in financial assistance was provided to seven provinces and territories to implement victim-related Criminal Code provisions in 2009-2010.

Victims Fund Project Funding

The Victims Fund provides funding through grants and contributions to support projects and activities that encourage the development of new approaches, promote access to justice, improve the capacity of service providers, foster the establishment of referral networks, and/or increase awareness of services, legislation and assistance available to victims of crime and their families.

In 2009/10, a total of 67 projects began or were underway for a total amount of $2,198,000.

Financial Assistance for Victims to Attend Parole Board of Canada Hearings

In Canada, victims are entitled to attend hearings conducted by the Parole Board of Canada as observers or to present a victim impact statement. Attending Parole Board of Canada hearings often involves travel and accommodation away from home. The Victims Fund provides financial assistance to registered victims who wish to attend hearings for the offender who harmed them in order to help victims participate more fully in the criminal justice system.

Financial assistance is also available for a support person to accompany registered victims to Parole Board hearings or to provide child or dependent care to enable victims to attend hearings.

In 2009/10, 346 victims and 89 support persons attended parole hearings with funding from the Victims Fund for a total cost of $233,202.

Financial Assistance to Host National Victims of Crime Awareness Week Events

The Government of Canada provides funding for organizations to host National Victims of Crime Awareness Week events. In 2009, the National Victims Awareness Week took place April 26 to May 2 with the theme of Supporting, Connecting, Evolving. Through the Victims Fund, funding up to $10,000 was available for projects that promoted community awareness, recognition of victims of crime or victim services, and which supported the goals of the Week. A total of 76 organizations/communities across Canada received funding for a total of $619,072.

Text Box G.3 provides information on the federal government’s commitment to parents of murdered or missing children.

Provincial and territorial government expenditures

As with the federal government, provincial and territorial governments are committed to addressing violent victimization. In some jurisdictions, this commitment may be approached from a gendered perspective, as in addressing violence against women and girls.

As part of their work on violent victimization, the provinces and territories have also implemented numerous programs focusing on awareness, prevention, intervention, services or assistance. These programs vary across the country in order to take into account different needs in different communities. Operating costs associated with the provision of victim services, including sexual assault centres and Sexual Assault Nurse Examiners, have been included under Social Services Operating Costs and Medical Costs. Other costs, such as autopsies in the case of homicides are also included under Medical Costs. In addition to the revenues generated by the federal and provincial surcharges, some provincial and territorial governments also contribute funding to the operating costs of victim services. Most jurisdictions receive funding from the federal Victims Fund to support the implementation of federal and provincial/territorial legislation for victims of crime and to reach underserved victims. Jurisdictions use this funding for specific projects that can include the development of public legal education and information, the development of services, training, research, evaluation, and other activities.

Estimation methods

Each cost item has a unique estimation method based on the nature of the cost item, the available data sources, and resource constraints. Detailed explanations of each estimation method are available in the main sections of the report.

Data Sources

Uniform Crime Reporting Survey 2 (UCR2)

The UCR2 survey is an administrative survey that captures detailed information (incidence, accusation, and charge statistics) on all Criminal Code violations reported to and substantiated by police services across Canada. The individual-level microdata indicate the age and gender of both the victim and the offender as well as the relationship of the accused to the victim, which allows identification of non-spousal relationships. The survey is comprehensive in both detail and in geographic scope, as national (provincial and territorial) coverage of the UCR2 survey was 99% in 2009. This study uses UCR2 data to find the number of incidents requiring police resources and to estimate the number of initiated criminal court cases and corrections sentences.

General Social Survey (GSS)

The GSS is a regular annual population survey conducted by Statistics Canada; each year or cycle focuses on a particular social subject. The victimization cycle was first conducted in 1988 and now occurs every five years. The 2009 cycle of the GSS is used in this study, along with the stalking section and some sexual assault questions from the 2004 cycle. The GSS is the only national survey of self-reported victimization.

The GSS target population is non-institutionalized Canadians aged 15 and over who are able to conduct an interview in one of the official languages and who can be reached by landline telephone. GSS respondents are asked about their personal experiences of criminal victimization, regardless of whether or not the incident was reported to police. As not all crimes are brought to the attention of the police, the GSS is an important complement to police-reported crime statistics found in the UCR2 survey. Since victimizations not reported to police have no less of an effect on victims, family, and employers than those incidents that are reported to police, self-reported data recorded in the GSS give a more comprehensive and accurate reflection of the prevalence and impact of violent victimization in Canada.

The GSS victimization cycle contains two questionnaires and corresponding data files. The Main File includes all of the demographic information for respondents (such as sex, age, work status, annual income, education, and family size), details about spousal violence incidents and situations, details about criminal harassment (in the 2004 cycle but not the 2009 cycle), and screener questions that are used to determine which respondents also need to complete incident reports in the Incident File. The Incident File includes details about all reports of non-spousal victimization (excluding criminal harassment). If a respondent indicates in the Main File screener questions that they were victimized by anyone who isn’t their spouse in the past year, they are asked to complete an incident report. The Incident File contains most of the information relevant to the current study as it asks respondents about the nature and outcomes of their victimization. The Incident File contains 7,096 incident reports, each of which may contain multiple reports of “similar incidents”. These multiple reports can be taken into account with the weighting variable ADJWTVIC, which caps the number of incidents in each incident report at three. ^{Footnote 10} When weighted, the Incident File contains 9,692,321 crime incidents. The Main File has a count of 19,422 respondents, with each respondent representing an average of approximately 1,400 non-institutionalized Canadians when weighted to the total population 15 years of age and older.

Cases of criminal harassment, or stalking, do not follow the same weighting conventions as other crimes and are unique in the GSS. In the 2009 GSS, respondents were asked if they had been victims of stalking in the Main File; the only additional information provided was the number of times they were stalked in the past 12 months. This was an attempt to obtain estimates of lifetime victimization more broadly without putting a burden on respondents to provide details for each incident.

Stalking victims in the 2009 GSS did not complete a crime incident report in the Incident File. However, a stalking module was included in the 2004 GSS, and victims did provide additional details of the crime, but this was also only available in the Main File, and outcomes were given only for each victim, not for each incident. For example, if a victim activated special phone features at three separate times to avoid harassment, the 2004 GSS will only record that the victim activated special phone features at some time to avoid harassment. Therefore, all estimates for criminal harassment are based on the number of victims, not the number of incidents.

Other data sources

The Adult Criminal Court Survey (ACCS) and the Youth Court Survey (YCS) are statistical databases of the main court-related measures such as charges and convictions. They both contain demographic information of the accused (age and sex) and information on the nature and outcome of each case. However, court surveys were not designed to gather information about victims of crime, and some of the estimates based on court data also require the use of police data to approximate certain elements of the estimate. The data are collected from all jurisdictions responsible for criminal courts. The ACCS and YCS are used in many of the cost items in the criminal justice system sub-category.

Many other data sources are used in the estimates of one or a few cost items. Canadian and international government reports, academic journal articles, and independent research organizations are all represented in this study. Some of the other surveys used are:

• Adult Correctional Services Survey (ACS) and the Integrated Correctional Services Survey (ICS)
• Court Personnel and Expenditure Survey (CPES)
• Legal Aid Survey (LAS)
• Police Administration Survey (PAS)
• Prosecutions Personnel and Expenditure Survey (PPES)
• Victim Services Survey (VSS)

Limitations

Costing exercises of the scope and magnitude of the current study all face limitations related to data availability, data reliability, resource constraints, and method. As a product of these limitations, some estimates will be conservative and some estimates will not be possible at all.

The term “resource constraints” refers to the resources (time, money, data access, etc.) available to the researchers for the purposes of carrying out the costing work, and these constraints can also be causes or results of data availability and data reliability. For example, due to extremely limited data, medication costs for assault are estimated by first estimating the number of victims who may have sustained a fracture, and multiplying that number by a certain dosage and cost of pain relief medicine. Obviously, a perfect estimate would involve querying each assault victim about the medication they used because of the incident, but this method is beyond the available resources.

It is also regrettable that the current study is unable to analyze the different impacts that victimization has on different groups of the population, with the exception of the breakdown by gender. It is recognized that there is importance in estimating the costs for a variety of diverse groups (e.g., Aboriginal groups, the elderly, students, etc.) in order to examine where the impacts of victimization are the greatest, but the data sources do not permit such analyses.

Data availability

Data that are not available can influence the study in one of two ways: either the relevant cost item must be underestimated based on some other data source or a reasonable assumption, or the cost item cannot be estimated at all.

One example of the first problem is the lack of data available in situations of losses to employers where employed victims are so affected by the incident that they become unemployed, either due to resignation or layoff. No data are available in such cases, and therefore costs associated with an employee leaving, such as rehiring costs, retraining costs, and lost interim productivity while the position is unfilled, are not estimated, so only limited estimations are made for cost items concerning losses to employers.

Another example of the first problem is the average number of calls made by victims to a crisis line, which is necessary information for estimating “crisis centre and crisis line costs”. The number of victims who contacted crisis lines at least once is available from the GSS. However, as crisis lines are anonymous, the number of calls made by each person who uses the service is not recorded. Therefore, some of the information needed to estimate this cost item is available, but the other information needed must be assumed based on anecdotal evidence from front-line crisis line workers. This kind of data unavailability may result in cost estimate inaccuracies.

An example of the second problem, where the cost item cannot be estimated at all, is civil sexual assault cases. Victims of sexual assault may sue the offender in a civil court in order to obtain damages for the pain and suffering experienced by the victim. This is an important element of the justice system, allowing victims a chance for financial redress. Unfortunately, the civil justice system is fragmented (as it is administered provincially) and it is difficult to both record and compile data in the civil context. Identifying those cases involving victims of sexual assault is therefore difficult, and determining the number of cases nationally and the cost to the civil justice system is not advisable. This report only provides a discussion on the topic and presents what information is available for interest only.

Data reliability

Several factors affect the reliability of the UCR2 survey. First, each police jurisdiction in Canada reports their own data to the CCJS (who oversees the UCR2), so any differences in the collection and compiling of that data between the jurisdictions will not be reflected in the UCR2. Second, differences in the way police activities are carried out between jurisdictions may cause the UCR2 to show aggregations of dissimilar data. For example, the individual responsible for deciding whether to lay a charge or not is not the same across jurisdictions. In British Columbia and Québec, the Crown decides whether or not to lay charges, while in New Brunswick the police, after consultation with the Crown, are ultimately responsible for the decision (Department of Justice 2003). Note also that the UCR2 data tables used for the calculations in this report only record the most serious offence (MSO) for each incident.

The GSS has more numerous and more serious data reliability issues than the UCR2. The main issue is the possible bias or misrepresentation of certain demographic groups (or of victimization rates and experiences) in the GSS, and this issue has many contributing factors.

The GSS is conducted by landline telephone. More and more Canadians are not using landline telephones, relying solely on cellular phones instead. The percentage of households using cellular phones only increased from 6.4% to 13.0% from 2007 to 2010. ^{Footnote 11} Any demographic groups that disproportionately avoid using landline telephones are in danger of being misrepresented in the GSS. Young people have been shown to be just such a demographic (Arcturus Solutions 2008), and if certain groups within that demographic have different victimization characteristics but cannot be reached on a landline telephone, then the GSS results may not be representative of this group. The GSS data is also weighted according to age to be more representative of the general population, but this actually exacerbates the previous issue if the results for youth are not representative of the true victimization experiences of youth.

The GSS does not cover individuals residing in institutions at the time of the survey, where an institution may be a shelter, retirement home, hospital, correctional facility, or some other non-traditional residence. If these individuals are more or less likely to have been victims of violent crime, and have different victimization experiences, the GSS will present a biased picture of victimization in Canada.

The non-response rate of the GSS is 38%. The victimization rates and victimization experiences of these non-responders is unknown. If victims or certain types of victims are more likely to avoid responding to the survey, the results will undercount victimization incidents and may possibly misrepresent the nature of victimization.

One further limitation of the GSS is the reliance on respondents to accurately recall events that may have been emotionally traumatizing. For many reasons, victims may not remember the events of their victimization or the outcomes of that victimization with perfect accuracy (Perreault and Brennan 2010). If, for example, a victim remembers that she was supported by victim services but does not remember that she visited a counsellor separately, the resulting social service operating costs and victims costs will be underestimates.

The previous limitation is closely related to another GSS reliability issue. Some victims of certain crimes (such as sexual assault or stalking) may be unwilling to disclose their experiences of criminal victimization to an interviewer due to the sensitivity of the crime and their experiences, and the incidence of these crimes is therefore likely to be underestimated ^{Footnote 12}.

Note also that the GSS weighting variable used, ADJWTVIC, caps the number of incidents in each report at three. It is possible that the true number of incidents exceeds the number of incidents obtained using the ADJWTVIC weighting variable.

One final important limitation is the quality of the data themselves. The GSS User’s Guide states the benchmark for minimum sample sizes to be used when presenting GSS data:

Users should determine the number of records on the microdata file which contribute to the calculation of a given estimate. This number should be at least 15 in the case of persons or households, and at least 40 in the case of victimization incidents. When the number of contributors to the weighted estimate is less than this, the weighted estimate should generally not be released regardless of the value of the Approximate Coefficient of Variation. If it is, it should be with great caution and the insufficient number of contributors associated with the estimate should be prominently noted. (Burns and Williams 2011, p. 25).

Many of the estimates in this report use GSS victimization results that contain less than 40 records when not weighted. The GSS estimates are simply “best estimates”. Sampling error is reduced with large samples, which is the case for many of the broader GSS variables, but as sub-sample sizes decrease (for narrower variables), sampling error increases. Therefore, many GSS results in this report should be read with extreme caution. Any reproduction of the GSS results contained in this report should only be done with full knowledge of the data quality, and should carry the same warning of caution to other readers.

Valuation of Intangibles

Intangible costs make up the largest proportion of costs in this analysis, and therefore it is important to understand the different methods used by economists to estimate the value of each intangible impact in order to choose the most appropriate one. This section contains a discussion on how pain and suffering and loss of life are valued in this report and some alternative methods that are not used.

Intangible impacts have no market price at which “values” are determined. Estimating these “values” is difficult since every victimization experience is different, and the intangible impacts felt by each victim are consequently different. Therefore, there are many different methods to estimate the values of intangibles. The very act of placing a monetary value on intangible costs can be considered inappropriate or insensitive, but it is vital to do so when estimating the costs of social phenomena. Great effort has been made in this report to treat these issues with proper respect and sensitivity. Intangible impacts are obviously very personal and affect victims profoundly, and in no way is this report implying that the impacts are “worth” the values assigned to them; the values assigned simply provide a somewhat standard way to compare the magnitudes of different social issues, and without including intangible costs these cost analyses would be extremely incomplete. By including intangible costs for all social issues, a more accurate comparison across issues can be made.

Pain and suffering

Pain and suffering costs are relevant to assault (A.V.3.1 Pain and Suffering Costs), robbery (R.V.3.1 Pain and Suffering Costs), and sexual assault and other sexual offences (S.V.3.1 Pain and Suffering Costs). Pain and suffering costs are also relevant to criminal harassment, but data are not available in order to include them in the estimation.

In this report, the values of pain and suffering for assault and robbery come from McCollister et al. (2010). McCollister et al. (2010) uses the method developed in Cohen (1988) and updates it by using more recent data. The method involves taking the compensation amounts awarded by juries for different crimes (by way of the injuries incurred and the likelihood of each crime resulting in each injury) and subtracting all likely tangible costs (e.g., medical, etc.) associated with those victimizations to get the juries’ valuations of pain and suffering. Though both Cohen (1988) and McCollister et al. (2010) focus on the US, it is assumed that pain and suffering in Canada is similar due to lack of Canadian information.

An alternative approach for estimating pain and suffering is to analyze Disability Adjusted Life Years (DALYs), or the converse measure Quality Adjusted Life Years (QALYs). DALYs are measures of the amount of healthy life lost due to an injury or disease (including psychological injury or disease), or the amount of life lost due to premature death caused by that injury or disease. Access Economics (2004) uses DALYs and the value of one year of life to calculate the pain and suffering costs of domestic violence victims in Australia. This is possible in the Australian case as Access Economics (2004) had access to detailed data on the injuries sustained by victims of domestic violence, but there is no such detailed information available for victims of crime in Canada. Dolan et al. (2005), however, uses the QALY method to estimate general values of pain and suffering for victims of rape and sexual assault.

The approach used by McCollister et al. (2010) is not suitable for use in the sexual assault and other sexual offences section because it only includes the value of pain and suffering for “rape/sexual assault” and does not distinguish between the two. Though Canadian law classifies three levels of sexual assault and does not distinguish between rape and sexual assault, the GSS does distinguish between “sexual attacks” and “unwanted sexual touching”. In this report, a proportion of “sexual attacks” are further estimated to be “rape”. All sexual offences not deemed “rape” are then classified as “sexual assault”. The values of pain and suffering for sexual assault and other sexual offences are therefore taken from Dolan et al. (2005), which distinguishes between the more serious offence “rape” and “sexual assault” and is based on the QALY method described above. This method is used to guarantee a conservative estimate, as classifying all GSS incidents as “rape” and assigning the pain and suffering value of rape to those incidents may cause the pain and suffering of some of the less serious incidents to be overestimated.

A third method not used in this report, also described in Dolan et al. (2005), is the stated or revealed preferences (“willingness to pay”) approach, which measures how much people are willing to pay for some intangible benefit of reducing crime based on their actual statements or behaviour.

Loss of life

Loss of life costs are relevant to homicide (H.V.2.1 Loss of Life Costs).

The value of a statistical life (VSL) in this report is taken from Viscusi (2008). Viscusi (2008) uses the common “willingness to pay” or “willingness to accept” method. In its simplest terms, the willingness to pay method measures how much money a person is willing to pay to reduce the risk of death in some event, activity, or job; by stating a monetary amount that a person would pay to reduce the risk of death the person has revealed how much he or she values life. For example, as stated in Zhang et al. (2012), “if an individual is willing to pay $500 to eliminate a 0.01% risk of death, the implicit VSL for that person is $500 / 0.01% = $5 million.” There is no standard method for calculating the VSL, and many different values for life have been proposed.

There are complications with calculating the VSL through this willingness to pay method. The most important issue is that the willingness to pay method is income constrained; that is, people with lower incomes would place lower valuations on reducing the risk of death because of their lower incomes. A related problem is that the valuation of life is not constant across the population (not only due to income levels but other characteristics as well) and time (as age is one determinant in an individual’s willingness to pay). The nature of the risk is also a factor in the valuation of life, as seen in Viscusi (2009), where deaths resulting from some events (such as natural disasters) have a much lower valuation than deaths resulting from other events (such as terrorist attacks). An example of other factors affecting individual willingness to pay may include a person’s happiness level. All of these issues mostly pertain to an individual’s assessment of willingness to pay, but in this report the VSL is derived from an average willingness to pay based on a sample of labour force participants; this method mitigates some of the problems associated with income constraints, population, age, and nature of death that arise when willingness to pay is defined on an individual basis. In the absence of alternative methods, the willingness to pay method is considered suitable for use in this report.

• Previous Page
• Table of Contents
• Next Page