Conventional calculations of homicide rates lead to an inaccurate reflection of Canadian trends.

• Author: Andresen, Martin A.

Introduction

Homicide rates are considered to be an indicator of general trends in violence across many societies. Often referred to as a "social barometer" (Bottomley and Pease 1986; Waller 1982), the prevalence of homicide is generally thought of as a reflection of the quality of life in a society (Gartner 1995). Traditionally, homicide rates have been calculated by dividing the total number of homicides by the total population and multiplying by 100,000 to give the homicide rate per 100,000 people:

Homicide Rate per 100,000 = (Number of Homicides/Population) * 100,000

The numerator (total number of homicides) is usually based on police reports. The denominator is simply the total population, based on census data. In Canada the conventional calculation of homicide rates using total population as the denominator produces a picture of homicide rates that rise dramatically in the 1960s, peak in 1976, and slowly decline to the late 1990s. This decline has been attributed to legislative changes, specifically, the broadening of divorce laws (1968), the abolition of capital punishment (1976), and the changing of firearms laws (1977, 1991, and 1996). (2) It is important to evaluate critically the calculation of homicide rates because of the impact such rates have on social policy.

The Canadian population

As in many other countries around the world, Canada experienced a post-World War II population explosion. The baby boom generation, generally described as those born between the years 1947 and 1961, (3) comprises approximately one-third of the population and has, arguably, shaped society for the past 50 years (see Foot 1996). With such a large portion of the population concentrated in a relatively small age range within the total population, any cohort influences of this group will have disproportionate effects on rates based on the total population. Table 1 shows how the proportions of age categories have evolved over the period under study. The age categories 15-19, 20-24, 25-29, and 30-34 rise, reach a maximum, and fall off, while all the age categories above tend only to increase.

This phenomenon is merely a reflection of the aging "baby boomers." The proportion of people within any given age category rises and falls, or simply rises, as the baby boomers age and move through that category. And most importantly for the thesis of this paper, the proportion of the population that we believe to be the prime homicidal population, those aged 15-30 years, started to fall around 1977, when the baby boomers started reaching the age of 30 years. Figure 2 illustrates this graphically.

[FIGURE 2 OMITTED]

The appropriate denominator for homicide rates

The appropriateness of the denominator is crucial, due to the effect it has on the overall homicide rate. In Canada since 1975 the number of homicides has remained relatively constant; therefore, the overall homicide rate is more reflective of the denominator than of the actual number of homicides. This could give an artificial representation of the actual state of violence in Canada.

Perhaps out of convenience or convention, the denominator for calculating the homicide rate has been the total population. But is the total population an appropriate measure for calculating the homicide rate? Are all age categories equally likely to commit homicide, or is there a sub-population that is responsible for a majority of homicides? If so, has this sub-population, from a demographic point of view, behaved differently than the population as a whole? Due to the demographic characteristics outlined above, we can see that there may be problems with the traditional method of calculating homicide rates.

Many researchers have expressed concern with the denominator used in the calculation of crime rates:

* The denominator should be generally thought of as "a measure of population that is exposed to the events, or is at risk of being involved in events, such as offenders or victims" (Reiss 1986);

* [S]ince most of the crimes counted in crime and victimization rates involve young males, it may make more sense to standardize the level of risk in terms of the proportion of young males in the population rather than in terms of the total population (Johnson and Lazarus 1989);

* [Total population] provides only a general assessment and does not take account of demographic characteristics of the population or the empirical fact that different segments of the population are involved in criminal events with different levels of frequency (Sacco and Kennedy 1994).

* Using total population as the denominator is a "crude" way of calculating crime rates (Silverman, Teevan, and Sacco 1996).

There is a strong indication that young males are the driving force behind homicide in Canada, or any country, for that matter. Hence, only the young, male sub-population should be used when calculating homicide rates to serve as a social barometer. (4)

In a cross-country study over time, Pampel and Gartner (1995) found that increases in the proportion of the 15-29 age category within the total population has a significant positive effect on national homicide rates. Using Pampel and Gartner's data, we expanded and contracted the age-range category, in order to measure which age category was the best fit for Canadian homicide rates and found that the selection of their age-range category (15-29 years) is quite robust. Our tests revealed that the appropriate age range for Canada was 15-30 years, a range used in the remainder of this paper. We propose the use of the Most Frequent Age Range of Male Offender (MFARMO) as the denominator in the calculation of homicide rates: males, 15-30 years. Due to concerns over the choice of this sub-population, the 15-30 years age range was also expanded and contracted to investigate the sensitivity of our results.

Silverman, Teevan...