Sign the Resolution
Contents | Feedback | Search
DRCNet Home | Join DRCNet
DRCNet Library | Schaffer Library | Drugs and Driving
Traffic Injury Research Foundation of Canada, 171 Nepean Street, Suite 200, Ottawa, Ontario, Canada K2P 0B4
During the past twenty years, there have been significant reductions in the magnitude of the alcohol-crash problem in Canada. To a large extent, evidence of such reductions is derived from the declining number and proportion of fatally injured drivers with positive BACs. While the precise reasons for this decline are as yet unclear, it is often presumed the reductions in alcohol-related driver fatalities are a reflection of a corresponding decrease in the overall incidence of impaired driving. This hypothesis, however, has never been fully investigated.
This paper will examine the results of roadside surveys conducted in various provinces in Canada over the past twenty years to determine the extent of changes in the incidence of drinking and driving. In addition, the paper will examine the extent to which changes in the incidence of drinking and driving correspond to changes in the magnitude of the alcohol-crash problem. The data will also be used to examine the characteristics of drinking drivers and changes in these characteristics over time. The implications of the findings for policies and programs will also be discussed.
Since 1980, there have been significant reductions in the magnitude of the alcohol-crash problem, not only in Canada (Beirness et al., 1994) but in most industrialized countries around the world (Transportation Research Board, 1994). To a large extent, evidence of such reductions has been derived from data on the number and percent of fatally injured drivers with positive BACs. While the precise reasons for this worldwide decline in alcohol-related fatalities are as yet unclear, it is often presumed that the vast array of countermeasure activities during the 1980s produced a significant decrease in the incidence of impaired driving which in turn led to a reduction in alcohol-related fatalities. To date, there has been little evidence presented that would support this hypotheses. The purpose of this paper is to test this hypothesis by examining BAC data on nighttime drivers and drives fatally injured in road crashes. More specifically, it examines the results of roadside surveys conducted in Canada over the past twenty years to determine the extent of changes in the incidence of drinking and driving and to determine the extent to which changes in drinking and driving correspond to changes in the incidence of alcohol-related driver fatalities.
Roadside breathtesting surveys of nighttime drivers provide the most valid estimates of the prevalence of drinking and driving. In Canada, roadside surveys have been conducted periodically since 1974. The first survey was a national survey in which every province participated. Since then, ten additional roadside surveys have been conducted at various times in individual provinces, the most recent having been conducted in 1993. The methodology used for each of these surveys has been comparable to that used in the original national survey of 1974, which facilitates comparisons over time.
Trends in the prevalence of driving after drinking were examined using data from these roadside surveys. Because the only national survey, was conducted in 1974 subsequent surveys conducted in various provinces at similar points in time were examined for comparability of results and, where appropriate, combined. For 1981, results from roadside surveys conducted in Saskatchewan and Quebec were combined. The results of three surveys -- Quebec in 1986, Saskatchewan in 1987 and Nova Scotia in 1988 -- were combined. Finally, the results from the two most recent surveys -- Saskatchewan and Nova Scotia -- were pooled to provide an estimate for 1993. Thus, trends in the prevalence of drinking and driving were examined from estimates available at four points in time -- 1974, 1981, 1986/88, and 1993.
Data on driver fatalities were obtained from the Traffic Injury Research Foundation's Fatality Database. This database contains information on the BAC of fatally injured drivers and is historically complete for seven provinces since 1973.
Figure 1 shows the percent of drivers in three BAC categories as determined by roadside surveys conducted in Canada at four points in time. In 1974, 16.4% of all drivers surveyed had a BAC in excess of 20 mg%. In 1981, the proportion of drinking drivers increased to 19.2%. In both of the subsequent surveys, the proportion of drinking drivers has decreased, reaching its lowest point at 11.7% in 1993. Overall, since 1981, the prevalence of drinking driving has declined by 39%.
Trends in Alcohol Levels among Drinking Drivers in Canada
Since 1981, when drinking and driving was at its highest point, the greatest decreases have been among drivers with low and moderate BACs (i.e., 21 to 80 mg% and 81 to 150 mg%). The proportion of drivers with high BACs (i.e., over 150 mg%) has shown the smallest decrease. This finding supports the contention that the gains in the drinking-driving problem achieved during the 1980s were largely among the more responsible social drinkers and that countermeasure efforts have had relatively little impact on heavier drinkers who comprise the largest proportion of fatally injured drinking drivers (e.g., Simpson and Mayhew, 1991).
Over the past twenty years, women have become increasingly more active participants in the driving population. For example, in 1974, 15% of all drivers interviewed in roadside surveys were women; in 1993, 33% were women. Women have also become more active as drinking drivers as well. In 1974, less than one in ten drinking drivers was female; in 1993, women accounted for about one-quarter of all drinking drivers.
Figure 2 shows trends in the percent of male and female drivers in Canada with BACs in excess of 80 mg% -- the legal limit in Canada. Since 1981, the percent of male drivers with illegal BACs has decreased by 47% -- from 6.2% in 1981 to 3.3% in 1993.
Percent of Male and Female Nighttime Drivers with BACs Over 80 mg%
Over the same period of time, the percent of female drivers with a BAC in excess of the statutory limit actually increased. In 1981, 1.4% of women drivers had a BAC over 80 mg%; in 1993, 1.8% were legally impaired.
Figure 3 shows the proportion of drinking drivers in each of four age groups. Since 1981 there have been reductions in all four age groups. The largest reductions, however, have been among the youngest group of drivers. The percent of drinking drivers aged 16-19 has declined by 57% -- from 16.7% in 1981 to 7.1% in 1993. Over the same period of time, drivers aged 20 to 39 experienced a 38% reduction in drinking and driving. The percent of drinking drivers aged 40 to 59 decreased from 1981 to 1986/88, and then increased again in 1993.
Trends in the Percent of Drinking Drivers According to Driver Age
Figure 4 shows the percent of fatally injured drinking drivers in seven provinces from 1973 through 1993 (represented by the line) and the percent of nighttime drivers in roadside surveys who had positive BACs (represented by the bars). The percent of driver fatalities who had been drinking increased from 58% in 1973 to a peak of 62% in 1981. Thereafter, the percent of fatally injured drinking drivers decreased dramatically, reaching its lowest point (i.e., 43%) in 1990. Since then, drinking driver fatalities have increased to 48% in 1992 and then decreased slightly to 46% in 1993.
Percent Drinking Driver Fatalities and Drinking Drivers in Roadside Surveys
The results of roadside surveys conducted over the same period of time show a similar pattern -- an increase in drinking drivers from 1974 to 1981, followed by decreases through the 1980s. On closer examination, the one group that did show an increase in roadside surveys during 1993 was drivers with very high BACs (i.e., over 150 mg%). This group is at greatest risk of becoming involved in a fatal crash and, in fact, has comprised a larger proportion of drinking driver fatalities in Canada recent years.
The increasing proportion of women among drinking drivers identified during roadside surveys is consistent with increases in the proportion of females among fatally injured drinking drivers. In 1974, 6.4% of all fatally injured drinking drivers were female; by 1993, women accounted for 12% of drinking driver fatalities.
The greatest decreases in alcohol-related driver fatalities in Canada since 1981 have been among younger drivers (TIRF, 1991). Similar decreases were noted in the prevalence of drinking among young drivers in roadside surveys.
Roadside breathtesting surveys provide the best source of information on the extent of drinking among the nighttime driving population. The expense of such surveys precludes them being conducted annually. The data presented in this paper indicate that periodic roadside surveys can be used to track trends in the magnitude and characteristics of the drinking-driving problem. Moreover, the correspondence between trends in the prevalence of drinking and driving found during roadside breathtesting surveys of nighttime drivers and trends in the incidence of alcohol involvement among fatally injured drivers in Canada suggests that alcohol-related fatalities reflect the overall prevalence of drinking and driving in the population.
This paper is based on a study of trends in roadside surveys funded by Transport Canada. The TIRF Fatality Database is jointly sponsored by the Canadian Council of Motor Transport Administrators and Transport Canada.
Beirness DJ, Simpson HM, Mayhew DR, and Wilson RJ (1994). Trends in drinking driver fatalities in Canada. Canadian Journal of Public Health 85: 19-22.
Simpson HM and Mayhew DR (1991). The Hard Core Drinking Driver. Ottawa: Traffic Injury Research Foundation.
Traffic Injury Research Foundation (1991). Canadian Trends in Drinking and Driving. Ottawa: Transport Canada.
Transportation Research Board (1994). The Nature of and the Reasons for the Worldwide Decline in Drinking and Driving. Transportation Research Circular, Number 422. Washington: Transportation Research Board.