The Cost of Pain Suffering from Crime in Canada

4. Methodology

4. Methodology

To derive the total cost of pain and suffering from crime in Canada (Cj), estimates for three components are required: the number of incidents for a certain type of crime (Nj), the proportion of the population feeling worried about safety (Pwj), and the value of perceived and actual mental distress as a result of the crime experience (Vj). Given information on these three components, the total cost of pain and suffering from crime can be calculated using the following formula:

Cj = Nj * Pwj * Vj (1)

In this formula, j = any number from 1 to n and indicates each type of crime assessed.

The purpose of this section on methodology is to explain the derivation of the three components Nj, Pwj, and Vj and to discuss the limitations and boundaries related to the gathering and processing of the required information.

4.1 Number of Crime Incidents

The total number of incidents for each type of crime assessed in this study can be obtained from two sources:

In the GSS on victimization conducted in 1999, survey participants were asked to indicate whether they were victims of certain crimes, including assault, sexual offences, robbery, breaking and entering, motor vehicle theft, theft of household property, and vandalism, in the 12 months prior to the survey. The present analysis uses this information to estimate the total number of incidents for each type of crime (Nj) based on the total number of survey participants (25,876) and the total population of Canada in 1999 (30,491,294). Nj can then be derived as follows:

Nj  = Number of victims as reported in GSS * Total Population (2)
Number of survey participants in GSS

For other crimes, such as homicides, drug offences, and Criminal Code traffic offences, where the GSS provides no information, the number of incidents reported by police statistics are used. For homicides, police records are considered to provide accurate information since evidence from homicides, such as corpses, is likely to be discovered at some point after the crime has been committed. For drug offences and Criminal Code traffic offences, police statistics are likely to underestimate the actual number of offences due to underreporting and limitations imposed on policing efforts, such as budgetary constraints.

4.2 Proportion of Population Feeling Worried about Safety

The GSS on victimization contains different measures of individuals' feelings about safety. For the purpose of the present analysis, the following four questions relating to safety issues are used to construct a measure of safety perception by the public and by crime victims : [5]

  1. "safety when walking alone after dark";
  2. "safety when using public transportation after dark";
  3. "safety when alone at home in the evening/at night"; and
  4. "degree of satisfaction with safety in general."

Survey responses are summarized in Table 2. From the answers given to any or all of these four questions, the proportion of survey participants feeling very worried about safety can be derived. For example, if an individual is considered to express distress about safety when "very unsafe" or "very worried" is chosen as a response to any of the four questions, then 2,349 of the 25,876 survey participants expressed worries about their safety. If the sample of survey participants is representative of the Canadian population, then 9.1% of Canadians express general concerns about safety, whether they were crime victims in the previous 12 months or not.

The GSS on victimization further implies that victims of various crimes are more likely to express worries about safety. The proportion of crime victims expressing worries for each type of crime (Pwj) can be calculated as follows:

Pwj = number of victims expressing worries about safety (3)
total number of victims for the crime

Table 2: Survey Responses to "Safety" Questions in the 1999 GSS
Question Response Number of Individuals (total = 25,876)
Safety when walking alone in own area after dark Very safe 8919
Reasonably safe 9728
Somewhat unsafe 2737
Very unsafe 1215
Doesn't walk alone 3208
Don't know 63
When using public transportation after dark, you feel. Very worried 807
Somewhat worried 3123
Not at all worried 4196
Do not use transit at night 11093
No public transit available 6610
Don't know 42
Not stated 5
Feelings when alone at home in the evening/at night Very worried 536
Somewhat worried 4698
Not at all worried 20390
Never alone 197
Don't know 48
Not stated 7
Degree of satisfaction with safety Very satisfied 11715
Somewhat satisfied 101646
Satisfied 195
Somewhat dissatisfied 1116
Very dissatisfied 513
Dissatisfied 13
No opinion 300
Not stated 378

The measure Pwj is consistently higher for different types of crime than concerns about safety by the general public. This implies that crime victims are more likely to be concerned about their general safety than individuals who have not been crime victims. For example, 53% of individuals who were victims of assault in the 12 months prior to the survey expressed worries about safety (compared to 36% of the general population). See Table 3 for the proportion of crime victims feeling worried about safety based on information from the 1999 GSS on victimization.

Table 3: Proportion of Victims Feeling Very Worried about Safety
Type of crime Number of Victims Feeling Very Worried about Safety Number of Incidents Proportion of Victims Feeling Very Worried about Safety (Pwj)
Assault 91 549 16.6%
Sexual Assault 70 300 23.3%
Robbery 54 245 22.0%
Breaking and Entering 161 906 17.8%
Motor Vehicle Theft 165 1,252 13.2%
Theft (Not Vehicle) 249 2,194 11.4%
Mischief /Vandalism 132 947 14.0%

In estimating the total cost of pain and suffering from crime, two different measures for the proportion of the population feeling worried about safety are to be considered. For every crime that occurs, there exists the (direct) cost of pain and suffering on crime victims. The existence of crime, however, can further impose stress and (indirect) cost on the general population, including on individuals who are not crime victims. For the purpose of the present analysis, the focus is on the cost of pain and suffering experienced by crime victims.

Consider the case of homicide. Although the deceased victims do not experience any pain and suffering after the crime, the expected cost of each homicide is rather high. Part of the cost consists of the cost of each human life lost. Another important component is the cost of distress and fear instilled in the general population. For drug and Criminal Code traffic offences, no information is made available from the GSS unfortunately. From an economic perspective, drug and Criminal Code traffic offences are sometimes considered victimless crimes as the offenders of these crimes are often their own victims. The cost of pain and suffering from these crimes is therefore mostly borne by the distress and fear imposed on the general population. Hence, the proportion of the general population feeling worried about safety will apply in these cases.

Another consideration that can affect the cost of pain and suffering is the proportion of worries that are attributed directly to the crime experience of victims. For example, an experience with violent crime (which by definition targets the person) is much more likely to have a deeper impact on the victims feeling worried than an experience with property crime (which by definition targets non-humans). Hence, it is reasonable to assume that the worries of victims of violent crime are more likely to be caused by their crime experiences than would be the case with victims of property crime. To accommodate for such differences, the estimates provided here assume that most of the worries victims of violent crime have about safety are a direct result of their violent crime experiences, while not all the worries victims of property crime have about safety stem directly from their property crime experiences. The worries about safety of property crime victims will, therefore, be discounted using a discount factor.

4.3 Expected Value of Distress and Worry

Information on the expected value of distress and worry as a result of crimes is most difficult to assess. Ideally, victimization surveys can be designed to contain questions concerning individual willingness to pay for reduced fear and worry about safety such as the amount an individual is willing to pay for additional police services to lessen the risk of being a crime victim. In the existing literature, the estimates of victim costs, mostly for the United States, are dominated by two data sources: (1) contingent valuation placed on safety by the public and (2) the amounts awarded by juries to crime victims for the suffering and loss of productivity resulting from crime. Neither piece of information is readily available in Canada.

To minimize the health risks, individuals generally aim to avoid activities that present such risks. In the economic literature, the dominant approach to estimating the risk-dollar tradeoff is to use labour market data on worker wages for risky jobs (Viscusi, 1993). When a worker accepts a job with potential exposure to work injuries that can cause pain and suffering and with associated health risks, it is an indication of that individual's contingent valuation on the pain and suffering which might result from exposure to health risks in exchange for monetary compensation. [6] Anderson (1999) used the average value of labour market estimates provided by Viscusi (1993) as a proxy for the value of life and injury from crime. Anderson pointed out that such estimates reflected "the amounts individuals are willing to accept to enter a work environment in which their health state might change" and that "(such) values do capture perceived risks of pain, suffering, and mental distress associated with the health losses."

For the purpose of the present analysis, the values reported by Viscusi are used and converted to 1999 Canadian dollar values. Viscusi provided estimates for both fatal and non-fatal risks, where the risk measures are labour market estimates as described above. Death risks associated with a certain job are referred to as fatal risks. However, jobs that are risky also tend to be unpleasant in other respects which are referred to as non-fatal risks by Viscusi. In the context of the pain and suffering experienced by crime victims, risks of fatal injuries obviously relate to worries about death; risks of non-fatal injuries can be related to worries about decline in individual well-being as a result of crime, such as emotional distress from property loss. Viscusi implied that the implicit average value of human life from non-fatal injuries was US$52,637 in December 1990. After adjusting for inflation, that amount was equivalent to approximately CAD$72,000 in 1999. For fatal injuries, Viscusi reported that the implicit value of human life was estimated by major studies to be mostly between US$3 million and US$ 7 million in 1990, which, in 1999, was equivalent to CAD$4.1 million to CAD$9.6 million. The estimates in this report are based on non-fatal risks.

Based on the assumption that an average individual generally values life, critics pointed out that human life is therefore priceless, and it appears to be immoral to attach a monetary value to a human life. It was not the intention of Viscusi, however, to set a price on an average human life. Furthermore, there is no assumption that any worker who accepts work with certain health risks is ready to give up his or her life. What is implied with the statistical value of a human life is the loss of value to society whenever a human life is lost. From an economic perspective, there is no market for the trading of human lives, but a human life does have a value.

4.4 Sensitivity Analysis

The final estimation of the cost for various types of crime will depend on the assumptions on which the analysis is based. This is referred to as "sensitivity analysis" in the literature. For example, the cost is likely to vary when different data sources are used for the estimation. The purpose of a sensitivity analysis is to take into account uncertainty and different possible circumstances. A sensitivity analysis can help, in the case of project evaluation for instance, to determine whether the implications from the outcome of a program remain the same when different assumptions are made. The results are generally considered more "robust" or stronger if the outcome of a program does not alter under different scenarios. In estimating the cost of the pain and suffering of crime victims, sensitivity analysis provides a range of estimates for comparison purposes. For example, the implicit value of a human life was estimated by major studies to fall between a relatively wide range of $4.1 million and $9.6 in 1999. The estimated cost of pain and suffering for the different categories of crime are thus to be quite different depending on the figure used for the value of a human life.