The Cost of Pain Suffering from Crime in Canada

2. Literature Review

In the current literature, victim costs are typically estimated from two data sources:

In his comprehensive study, Anderson (1999) estimated the value of risks to life and health in the case of violent crime in the United States. Anderson used labour market estimates as a proxy for the value of life and injury based on individuals' contingent valuation on work that might affect their health and cause pain and suffering. He estimated that the total cost of the risk to life and health violent crime presents amounted to US$574,395 million. Although such a proxy of values does capture the perceived risks of pain, suffering, and mental distress, the level of distress may be considerably higher in the case of violent crime than the level of distress expected from a risky work environment. Furthermore, the individuals who choose to do risky jobs may place a lower value on their health. Therefore, Anderson concluded that the calculation of the value of risks to life and health based on labour market estimates possibly represents the lower bound of estimates for the risks to life and health that violent crime presents.

An example of using jury awards to evaluate the cost of the pain and suffering of crime victims was provided by Cohen (1988). Cohen's most significant contribution to the literature is his methodology for estimating the pain, suffering, and reduced quality of life endured by crime victims. Cohen and his colleagues have conducted numerous studies to estimate the cost of crime that included the cost of pain and suffering.

Cohen (1988) combined actual victim injury rates and jury awards in personal injury accident cases to obtain monetary values for pain, suffering, and fear. To obtain a measure of pain and suffering, Cohen used information from compensation awarded by courts to victims, also known as "jury awards." Jury awards are divided into two categories: "special" damages intended to compensate for medical expenses and lost wages and "general" damages intended to compensate for the pain and suffering of the victims. Cohen obtained data on the two estimates for different kinds of crimes and provided estimates for pain and suffering using the following regression:

gd = c + a *sd

In this regression, gd = general damages, sd = special damages, c is a constant and a is the estimated coefficient of sd. For example, Cohen estimated the following regression equation for gunshot victims:

gd = US$17,957 + US$5.20 * sd

This means that for every US $1,000 of special damages (medical costs + lost wages) awarded by the court, the estimated award for general damages that covers pain and suffering is US $17,957 + US$5.20 * 1,000 = US$23,157.

Cohen also combined estimates of the value of life with crime-related death rates to obtain the risk of death from crime. The probability of death is calculated by dividing the number of murders associated with each type of crime by the number of those crimes committed. The probability of death is then multiplied by the estimated value of a statistical life, that is, US $2,000,000 (this value is based on estimates from previous studies [see Viscusi, 1993]), to obtain the risk of death for different types of crime.

Cohen then combined the values for risk of death, pain, and suffering with direct losses to obtain a total cost to victims for ten different types of crime. The average cost of crime to victims was found to be the highest in the case of kidnapping (US$110,469), followed by bombing (US$77,123). For each of the cost categories, bombing was found to contribute the largest loss of value in terms of direct loss (US$24,737); the highest cost for pain and suffering was found in rape (US$43,561); and kidnapping was most costly in terms of risk of death. Bank robbery, for example, was found to have a total cost of US$18,810 (US$4,422 for direct losses, US$10,688 for pain and suffering, and US$3,700 for risk of death).

Cohen (1990) estimated the costs to victims of rape, robbery, assault, car theft, burglary, and larceny. His estimates included the cost of pain, suffering, and risk of death, and he reported that the total cost per year in the United States was US$113 billion. In another study, Miller, Cohen, and Wiersema (1995) estimated that the cost of violent crimes and property crimes to victims was US$472 billion per year, including the cost of pain, suffering, and reduced quality of life.

In 1994, Cohen, Miller and Rossman estimated the total cost of violent crimes, including rape, robbery, assault, and murder, in the United States to be $183 billion per year. For each type of violent crime identified in the study, three types of cost are measured to contribute to the total cost: (1) direct costs such as medical and insurance costs; (2) lost productivity measured by forgone market wages and fringe benefits and housework; and (3) the cost of pain and suffering. The costs of non-fatal injuries were estimated by multiplying estimates of the loss in quality-adjusted life years (QALYs) [3] by the value of a life lost per year. Quality of life losses were rated on a scale that included different kinds of actual and psychological pain and suffering. In short, Cohen made a significant contribution to the development of important guidelines for estimating the cost of the pain and suffering experienced by crime victims.

An alternative method for estimating the value of reduced quality of life as a result of crime is to compare the housing prices in high and low crime areas, sometimes referred to as the "revealed preference" approach. In a study using this approach, Rizzo (1979) suggested a possible relationship between the amount of crime and reduced housing prices. Rizzo began his analysis by stating that the social cost of crime should include the "estimated market value of goods and property stolen or destroyed, loss of earnings due to personal injury or death, a rough estimate of expenditures to avoid crime by individuals and public agencies, and so forth [such as reduced value of housing]." Rizzo proposed a theoretical framework by assuming a closed system comprised of n communities or blocks that is characterized by a competitive housing market. Housing services produced in a community are affected by the level of crime in the community, the amount of land, and the amount of capital. Furthermore, the level of crime is affected by the amount of self-protection taken by the community. A representative consumer weighs the cost of self-protection against the benefit of crime reduction.

Based on this theoretical framework, Rizzo first examined the cost of crime in terms of reduced housing prices. He used data from Chicago to examine the relationship between crime rate and rental price per unit, holding constant the size and other characteristics of apartments. Variables used to explain rent differences included the average number of rental rooms on a block, the proportion of relatively new housing built after 1960, the family income average, the distance from the University of Chicago campus, the proportion of white population on the block, and various regional dummy variables. [4] Variables used in the empirical analysis were mostly entered in logarithmic form to give percentages instead of absolute differences. Hence, according to Rizzo, "the sum of the coefficients should be interpreted as the effect of a one per cent increase in crime as we move from one sub-area to another in this community."

From the regression estimates, Rizzo attempted to derive the cost of crime for victims and potential victims. The author concluded that, on average, the cost of crime was found to be higher in the Hyde Park-Kenwood area than the entire city of Chicago. Rizzo's study has provided the methodology to estimate partial crime costs based on the estimated value of housing.

These different methods for calculating victimization costs are being criticized, however. Jury awards are being criticized for inflating the value of pain and suffering (Zimring and Hawkins, 1995) and, therefore, overestimating victimization costs. The willingness-to-pay approach requires information from the public that has to be obtained through either labour market estimates or victimization surveys. Critics pointed out that the study of victimization has received limited attention until recently (Winterdyk, 2000) and that information from victimization surveys is limited. Furthermore, the results and recommendations provided by various studies depend heavily on the sample used in a particular survey and can, therefore, be biased. There are currently no agreed upon methods among researchers for estimating the intangible costs of victimization, such as the cost of pain and suffering or the value of a shattered life. This is perhaps one of the main reasons why cost-benefit analyses have not been widely applied to evaluate different kinds of crime prevention programs and policies. The present study attempts to provide some preliminary estimates for the cost of the pain and suffering endured by crime victims in Canada by using information from a victimization survey.