ࡱ> 24-./01789:;L }zbjbj jj:a:lnnn^D 4hb0LR: t ($ jL ߲{"߲߲L ca߲: 8 ߲]  ( PwC&(\w0a.a(4    Estimating the Presence of Alcohol and Drug Impairment in Traffic Crashes and their Costs to Canadians: A Discussion Paper Submitted to MADD Canada by: Applied Research and Evaluation Services (ARES) 2125 Main Mall University of British Columbia Vancouver BC V6T 1Z4 www.ares.ubc.ca December 2002 All information, opinions, and conclusions found in this report are the result of research by the Applied Research and Evaluation unit at the University of British Columbia (ARES) and do not necessarily represent the positions of any other individuals, agencies or groups. Copyright ARES, 2002. Dr. Michael Marshall and Dr. Bill Mercer of ARES would like to thank the Traffic Injury Research Foundation (TIRF) and the Insurance Corporation of British Columbia (ICBC) for their assistance in the production of this report. Table of Contents  TOC \o "1-3" \h \z Table of Contents  PAGEREF _Toc26676407 \h i Table of Tables  PAGEREF _Toc26676408 \h ii Executive Summary  PAGEREF _Toc26676409 \h 1 Argument:  PAGEREF _Toc26676410 \h 1 Calculation of Frequencies:  PAGEREF _Toc26676411 \h 2 Calculation of Costs:  PAGEREF _Toc26676412 \h 4 Conclusion:  PAGEREF _Toc26676413 \h 5 Introduction  PAGEREF _Toc26676414 \h 6 An Overview of Some Common Crash Data Sources and their Limitations  PAGEREF _Toc26676415 \h 7 Crash Severity & Frequency Data Sources  PAGEREF _Toc26676416 \h 7 Police Data:  PAGEREF _Toc26676417 \h 8 Insurance Data:  PAGEREF _Toc26676418 \h 9 Coroner Data:  PAGEREF _Toc26676419 \h 10 Alcohol and Drug Involvement Data Sources  PAGEREF _Toc26676420 \h 11 Police Data  PAGEREF _Toc26676421 \h 11 Coroner Data  PAGEREF _Toc26676422 \h 12 Filling in the Gaps  PAGEREF _Toc26676423 \h 14 Crash and Victim Counts  PAGEREF _Toc26676424 \h 14 Fatal Crashes and Fatally injured Victims  PAGEREF _Toc26676425 \h 14 Blood Alcohol Content (BAC) Count Estimations:  PAGEREF _Toc26676426 \h 17 Alcohol-only, Alcohol-and-Drug, and Drug-only Impairment Estimations  PAGEREF _Toc26676427 \h 20 An Order-of-Magnitude Calculation  PAGEREF _Toc26676428 \h 24 Facts and Arguments  PAGEREF _Toc26676429 \h 24 Assumptions:  PAGEREF _Toc26676430 \h 24 Estimation:  PAGEREF _Toc26676431 \h 25 Three Crash Costing Models  PAGEREF _Toc26676432 \h 29 Three Perspectives:  PAGEREF _Toc26676433 \h 29 Real Dollar Estimates:  PAGEREF _Toc26676434 \h 30 Discounted Future Earnings:  PAGEREF _Toc26676435 \h 30 Willingness to Pay:  PAGEREF _Toc26676436 \h 31 Impaired Driving Crash Cost Estimations  PAGEREF _Toc26676437 \h 32 Conclusion  PAGEREF _Toc26676438 \h 35 References  PAGEREF _Toc26676439 \h 37 Appendix A: Insurance Corporation of British Columbia and Manitoba Public Insurance Data & Transformations  PAGEREF _Toc26676440 \h 44 ICBC  PAGEREF _Toc26676441 \h 44 MPI  PAGEREF _Toc26676442 \h 45  Table of Tables  TOC \h \z \t "Heading 4" \c Estimated Fatalities, Injuries & PDO Vehicles, Canada, 1999  PAGEREF _Toc20193817 \h 2 Estimated % Alcohol-Involved - Fatalities, Injuries & PDO Vehicles, Canada, 1999  PAGEREF _Toc20193818 \h 3 Estimated Impairment Source - Fatalities, Injuries & PDO Vehicles, Canada, 1999  PAGEREF _Toc20193819 \h 3 Estimated # Impaired-related Fatalities, Injuries & PDO Vehicles, Canada, 1999  PAGEREF _Toc20193820 \h 3 Crash costs by Costing Model in 1999$  PAGEREF _Toc20193821 \h 4 Estimated Cost of Impaired Driving Crashes by Costing Models  PAGEREF _Toc20193822 \h 4 Table 1: Motor Vehicle-Related Fatalities, BC, 1998  PAGEREF _Toc20193823 \h 7 Table 2: # Motor Vehicle Crash Types, BC & Manitoba, 1998  PAGEREF _Toc20193824 \h 7 Table 3: Injury Severity - Miller & Blincoe 1994  PAGEREF _Toc20193825 \h 15 Table 4: ICBC Victim & PDO Vehicle to Fatality Ratios  PAGEREF _Toc20193826 \h 16 Table 5: MPI Victim & PDO Vehicle to Fatality Ratios  PAGEREF _Toc20193827 \h 16 Table 6: Injury Severity & BAC - Miller & Blincoe  PAGEREF _Toc20193828 \h 18 Table 7: Estimated Fatalities, Injuries & PDO Vehicles, Canada, 1999  PAGEREF _Toc20193829 \h 26 Table 8: % Impairment Source by Crash Type  PAGEREF _Toc20193830 \h 27 Table 9: # Victim/ PDO Vehicle - Impairment Source by Crash Type  PAGEREF _Toc20193831 \h 28 Table 10: Crash costs by Costing Model in 1999$  PAGEREF _Toc20193832 \h 32 Table 11: Estimated Cost of Impaired Driving Crashes by Costing Models  PAGEREF _Toc20193833 \h 33 Table 12: Impairment by Alcohol only Crash Costs by Costing Model  PAGEREF _Toc20193834 \h 34 Table 13: Impairment by Alcohol & Drugs Crash Costs by Costing Model  PAGEREF _Toc20193835 \h 35 Table 14: Impairment by Drugs only Crash Costs by Costing Model  PAGEREF _Toc20193836 \h 35 Table A1: ICBC Crash Types N, 1996 - 2001  PAGEREF _Toc20193837 \h 44 Table A2: ICBC Crash Fatality Ratios  PAGEREF _Toc20193838 \h 44 Table A3: ICBC N victims/vehicles  PAGEREF _Toc20193839 \h 45 Table A4: ICBC Victim & PDO Vehicle to Fatality Ratios  PAGEREF _Toc20193840 \h 45 Table A5: MPI N victims/vehicles  PAGEREF _Toc20193841 \h 45 Table A6: MPI Victim & PDO Vehicle to Fatality Ratios  PAGEREF _Toc20193842 \h 45  Executive Summary Knowledge of the extent of harm caused by traffic crashes, and by the sub-set of crashes caused by impairment, is important in the development of public policy and the allocation of countermeasure resources. This discussion paper argues that the extent of injury and property damage only (PDO) crashes is generally underestimated and hence the magnitude of the subset of impaired crashes within those categories is also underestimated. It also argues that the extent of drug-impaired crashes is underestimated. A model is presented that attempts to address these sources of underestimation by working from essentially complete data on fatalities and projecting estimations to injury and PDO events. Cost estimations are then calculated using three different costing models. Argument: The more serious a crash, the more likely it will be reported to or otherwise become known to various authorities such as the police, motor vehicle branches, insurance companies, and the coroner, and the more likely it will be investigated by one or more of those bodies. Consequently, Canadian data on motor vehicle fatalities, and whether or not the fatally injured person(s) had a measurable blood alcohol content (BAC), is largely complete, valid and reliable. That is, we have very good information the number of persons killed in crashes, and whether or not they were possibly impaired by alcohol. On the other hand, the data on whether or not a fatally injured person might have been impaired by drugs is incomplete, primarily due to a lack of testing and testing sensitivity. As crashes become less serious, there is less likelihood that they will be reported, recorded, or investigated. In order to assess the magnitude of the traffic crash and impairment-caused traffic crash problem, there is a need to find a way to estimate the number of less severe crashes, and whether or not they might have been caused by impairment by alcohol and/ or drugs. Historically, crashes reported to the police have been used as a measure of crash frequencies and types, with the police forwarding crash reports to provincial Motor Vehicle Branches for compilation and statistical analyses. However, a comparison of the frequencies of these reports with data from insurance company crash counts shows an underreporting of less serious crashes in the police-generated data. This could be because of a lack of policing resources, a reluctance on the part of drivers to report crashes to the police (but a desire for financial compensation from insurance companies) or both. Certainly, some proportion of crashes will never be reported to anyone and will just be settled privately, but insurance-based counts seem to gather many more crash instances than do police data counts. An examination of insurance-based and other data sources suggests that there may be a roughly stable relationships among the number of motor-vehicle related fatalities to the number of injuries to the number of property damage only (PDO) events. For this exercise, for every fatality, there were assumed to be 118 injuries, and 650 PDO events. Using these multipliers, one can move from the very good information on the frequencies of fatalities, to an estimation of the frequencies of less serious crashes. Similarly, an examination of BAC levels associated with different levels of crash-related injury severity (from no injury to fatality) can produce a rough estimation of the proportion likely impaired by alcohol in less severe crashes for every one percent impaired by alcohol in fatal crashes. An examination of these relationships showed that as crash severity lessened, the likelihood of impairment being a cause lessened. For the purpose of this estimation exercise, the examination of the BAC data suggested that for every one percent of fatal injuries associated with an impaired crash, about half of one percent of injury-only crashes were likely to be associated with alcohol-impairment, and about three-tenths of one percent of PDO events were likely to be associated with alcohol-impairment. To put this another way, if the percent of alcohol-impaired crashes went up by 10%, the percent of alcohol-impaired injury crashes would go up by 5% and the percent of PDO crashes would go up by 3%. Again, using these multipliers, one can move from very good information on the frequencies of impairment-related fatalities to an estimation of the frequency of impairment in less serious crashes. Finally, an examination of studies of the impairing role of drugs as well as alcohol in crashes suggested that where there is a positive BAC, about 75% of the instances involve alcohol alone, about 25% of the instances where alcohol was involved there were likely also drugs involved, and that there was an additional 10% of persons likely impaired by drugs over and above those impaired by alcohol alone or alcohol and drugs. Calculation of Frequencies: In order to estimate the extent of fatal, injury and PDO events, it was assumed that there was 118 injuries and 650 PDO events for every fatal event, and those multipliers were applied to the known number of motor vehicle-related fatal events in Canada in 1999, as reported by the Traffic Injury Research Foundation (Mayhew et al., 2001). From that it was estimated that there were: Estimated Fatalities, Injuries & PDO Vehicles, Canada, 1999 FatalitiesInjuries @118PDO veh. @ 650Total N 3,315391,1702,154,750 Again, working from Mayhew et al. (2002) the percent of persons killed in motor vehicle-related crashes, on-road or off-road, where alcohol was involved was used as a starting point to estimate the numbers injured and PDO events, using the notion that for every 1% fatal there would be 0.5% injured and 0.3% in PDO events. That resulted in: Estimated % Alcohol-Involved - Fatalities, Injuries & PDO Vehicles, Canada, 1999 FatalitiesInjuries @0.5%PDO veh. @ 0.3%% alcohol involved34.20%17.10%10.26% The next step involved applying the estimations of alcohol-only, alcohol-and-drug, and drug-only percentages to the alcohol-involved percentages. Estimated Impairment Source - Fatalities, Injuries & PDO Vehicles, Canada, 1999 FatalitiesInjuries PDO veh. % alcohol involved34.20%17.10%10.26%% alcohol only @.7525.65%12.83%7.70%%alcohol+drug @.258.55%4.28%2.57%%drug only @.103.42%1.71%1.03%% Impaired37.62%18.81%11.29% Finally, the estimated percent impaired was applied to the estimated number of fatalities, injuries, and PDO vehicles to give an estimated number of victims and PDO vehicles. Estimated # Impaired-related Fatalities, Injuries & PDO Vehicles, Canada, 1999 FatalitiesInjuries PDO veh. Total N 3,315391,1702,154,750% Impaired37.62%18.81%11.29%N Impaired1,24773,579243,185 The insurance company-generated ratios of 1.2 fatalities per fatal crash, 1.11 injuries per fatal crash, 1.44 injuries per injury crash and 1.52 vehicles per PDO crash were used to move to the crash as the units of analysis. This resulted in an estimation of 1,039 fatal crashes, 50,295 injury crashes and 159,990 PDO crashes associated with impairment by alcohol, alcohol and drugs, or drugs only for 1999. Using these crash frequency estimations, three costing models were applied (in 1999 dollars) to estimate the order-of-magnitude of impaired-related crashes in Canada in 1999. Calculation of Costs: Broadly, there are three kinds of questions that are asked about the result of a traffic crash: How much will this cost me in real dollars spent? (Real Dollar Estimate -- RDE) How much will this cost me in terms of lost goods, opportunity, or productivity? (Discounted Future Earnings --DFE) How much would I pay for this not to have happened? (Willingness to Pay -- WTP) Each model approaches the question of crash costs differently, especially in the calculation of the value of a fatal crash. The RDE figures are based on estimates from the Insurance Corporation of British Columbia (ICBC), while the DFE and WTP estimates came from an Ontario study by Vodden et al. (1994). Crash costs by Costing Model in 1999$ fatalinjury-onlyPDOReal Dollar Estimate$280,340$25,215$1,581Deferred Future Earnings$984,412$23,779$7,265Willingness to Pay$7,473,138$32,101$7,265 Using these models, in 1999 dollars, impaired-related crashes based on the above frequency estimations cost: Estimated Cost of Impaired Driving Crashes by Costing Models fatalinjury-onlyPDOSumReal Dollar Estimate$291,344,243$1,268,195,972$252,872,971$1,812,413,186Deferred Future Earnings$1,023,052,566$1,195,999,410$1,162,332,552$3,381,384,528Willingness to Pay$7,766,477,399$1,614,515,834$1,162,332,552$10,543,325,785 Conclusion: A model to give an order-of-magnitude estimation of the extent and cost of alcohol, alcohol-and-drug, and drug-impaired driving crashes in Canada was developed using a range of data sources. Overall, it was estimated that in 1999 there were over 200,000 impaired driving crashes in Canada and that impaired driving cost between $1.8 billion and $10.5 billion, depending upon the costing model used. However, these figures can only be considered as a rough approximation based on a number of assumptions, any one of which could be in error to a substantial degree. Nevertheless, the issue of the underestimation of less serious crash frequencies and the resulting underestimation of the magnitude of impaired driving crashes and attendant costs is real and worthy of consideration by researchers and policy makers alike. Finally, it must be emphasized that data on trauma and impairment by alcohol and / or drugs are strongest in the area of traffic crashes, possibly because of the high frequency and financial consequences of these crashes. However, if should be recognized that these kinds of impairment are also likely to contribute substantially to causing trauma in other activities, including boating, snowmobiling, skiing, falls and so on. This paper is restricted to the area of traffic crashes primarily because of access to data of sufficient quality and quantity, not because of an assumption that the frequency and costs of impairment-related trauma from other sources are insignificant. Introduction In a climate of limited financial and human resources, proponents of different social issues must compete for support for their issues. Policy makers at all levels of government, the public and private corporate sectors (e.g., the insurance and hospitality industries), and the general public all contribute to some extent to how these limited resources are distributed. Clearly, in this process, it is critical that resource-allocation decisions be made in an informed manner based on data that are as accurate as possible. There is no question that alcohol, alcohol in combination with drugs, and drugs alone can impair drivers, and that this impairment can lead directly or indirectly to traffic crashes and other sources of trauma and mortality. There is also no question that this situation takes a serious and substantial toll in both human and financial terms. To illustrate, in 1998 in Canada, there were 555 homicides reported to the police (Canadian Center for Justice Statistics, 1999), while in the same year 2,909 persons were killed in traffic crashes and 986 of those involved a drinking driver (Mayhew et al., 2001). The precise measurement of the contribution of this impairment to the frequency and severity of crashes, as well as their cost to the social system, are more difficult questions to address. Data sources, data definitions, units of analysis, analytic techniques, and sociopolitical perspectives and interpretations differ across agencies, jurisdictions, and analysts. This is not to imply that any particular statistical analysis is in error, or that advocates of one policy or another intentionally bias their reports, but, rather, that different approaches naturally produce different results. Having said that, there is a need to explore whether a model for estimating a global order-of-magnitude of the extent and cost of impaired driving can be developed that takes into account the strengths and weaknesses of the various data sources and that could be generally used in this area. Consequently, it is the purpose of this report to begin a dialogue on what sort of model might be appropriate for such a global estimation of the extent and cost of impaired driving in Canada. An Overview of Some Common Crash Data Sources and their Limitations Crash Severity & Frequency Data Sources It is impossible to know with absolute certainty the number of traffic crashes that occur in a jurisdiction. The reason for this is that the data sources vary in how they define crashes and how they compile their information. For example, consider three common sources of crash data: data reported to Motor Vehicle Branch by the police, motor vehicle insurance record data, and coroner data. Table 1 represents the number of persons reported killed in British Columbia in motor vehicle crashes in 1998 drawn from these three sources, while Table 2 compares police reports to insurance company counts on the number of fatal crashes, injury crashes, and property damage only (PDO) crashes in BC and in Manitoba in 1998. There are a number of reasons why these numbers differ but the point is that the same question can result in very different responses, depending upon the data source queried. Table 1: Motor Vehicle-Related Fatalities, BC, 1998 # and Data SourcePoliceInsurance (ICBC)CoronerCrash Fatalities418473455 Table 2: # Motor Vehicle Crash Types, BC & Manitoba, 1998 Crash # and Data SourceBritish ColumbiaPoliceInsurance (ICBC)Fatal365418Injury17,98446,554Property Damage Only (PDO)17,418203,245Total =SUM(ABOVE) 35,767 =SUM(ABOVE) 250,217ManitobaPoliceInsurance (MPI)Fatal109138Injury6,87927,124Property Damage Only (PDO)20,13660,239Total27,12469,003 Police Data: Most jurisdictions collect data on traffic crash frequencies from police reports, but these reports probably underestimate the actual number of crashes, especially those that are less severe. For a crash to be included in police crash counts, at least three things must happen: the crash must be reported to the police; the police must attend the crash; and a report must be written and submitted to the compiling authorities (usually Motor Vehicle Branch in a Ministry). Admittedly, some jurisdictions have allowed motorists to submit a report to the police, eliminating the second step of police attendance, but the reliability and validity of self-reports can be questionable. Considering the first step of reporting a crash to the police, there is the disincentive for some drivers to report a crash for fear of precipitating Motor Vehicle Act or even Criminal Code charges, as well as avoiding civil liability. In a meta-analysis of crash reporting, Elvik & Mysen (1999) found that about 95% of the fatal crashes, 70% of the serious injury crashes, 25% of the slight-injury crashes, and 10% of the very slight injury crashes were reported to the police. Once a crash has been reported to the police, the pressures of other calls and shortage of personnel can make it impossible for the police to attend. Further, the police are probably more likely to attend more serious crashes, crashes that require traffic control (e.g., intersection), crashes that occur during times of day and days of the week that have lighter call loads, and / or those where charges might be more likely to be laid, resulting in one or more biases in the data. Finally, even when a crash has been attended, there is no guarantee that a crash report will be completed and filed appropriately. One way of getting some idea of the magnitude of difference between police counts and actual crash numbers is to compare police counts to crashes as recorded by motor vehicle insurance agencies (although the problem of unreported crashes remains for both sources). In British Columbia all vehicles must be insured through the Insurance Corporation of British Columbia (ICBC), while in Manitoba they must be insured through Manitoba Public Insurance (MPI). Using data from these Crown Corporations and the respective police data, in BC in 1999 police attended and reported on 375 of the 405 fatal crashes known to ICBC (92.6%), 19,947 of the 45,077 injury crashes known to ICBC (44.3%), and 20,987 of the 206,567 property-damage-only (PDO) crashes known to ICBC (10.2%), while in Manitoba the 1999 proportions were 75% of the 132 MIP fatal crash count, 76% of the 9,175 injury crash count and MPI and 35% of the 62,528 MPI PDO crash count. It follows that year-to-year and jurisdiction-to-jurisdiction changes in crash counts from police reports could potentially be the result of changes in police resourcing, policy and priorities as actual changes in crash rates. It is also likely that police crash counts are an underestimation of actual crash counts, especially in the case of less serious crashes. Insurance Data: Of course, insurance data only covers those crashes where one or more of those involved are covered and where claims are made. On the one hand, there is a disincentive for drivers to report crashes to their insurer because it might increase insurance rates, and this would likely be more often the case where the damage is slight and the cost of repair could be borne by the insured person. On the other hand, there is an incentive to report the crash to obtain financial relief from the crash cost (especially there is severe loss and /or if a driver believes he/she is not at fault, possibly increasing the proportion of multiple-vehicle crashes reported). In jurisdictions where one insurer covers every insured vehicle, there is less difficulty in accumulating total crash claims frequencies than in those jurisdictions with multiple insurers. Nonetheless, because the unit of analysis and activity for insurance corporations is the claim, not the crash, deriving crash counts can pose serious technical difficulties. That is, a crash can produce a large number of individual claims, so to determine crash counts, every claim, open or closed, must be compared to every other claim and then those sharing a crash aggregated into a crash count a process that can take millions of comparisons and significant expense and time in terms of computer CPU. In turn, these might be divided in terms of level of most severe injury, but that information may not be on the file, or may not be on the file in an electronically retrievable manner. Further, in terms of classification even if there are injury data, are individuals injured because they claim to be or because there is an indication of payment for an injury? When in some jurisdictions over 70% of injury claims are for soft-tissue injuries (e.g., whiplash) this becomes a serious consideration. In jurisdictions where there are many insurers, compiling crash counts becomes a very difficult task indeed due to things like differing data definitions, units of analysis, data base operating systems and the like. Last, it is the business and responsibility of insurance companies to settle claims, not report crash counts, so their motivation to produce such data may not be high in some cases. Coroner Data: While the coroner only deals with the sub-set of crashes that involve fatalities, it is the legislated duty of the coroner to determine the cause of a persons death. A difficulty that arises is that the circumstances around the death (a crash) may not be as well documented or tracked as the actual physical cause (e.g., exsanguination). Beyond the issue of classification, coroner data also can suffer from the time lag between the actual death and the time the cause of death is determined and entered into a central data set, resulting in a lag in fatality count. This lag can be significant if the time taken to determine the cause of death is substantial, as it may take several years before the number of crash-related deaths in a given year can be ascertained. Nonetheless, as the coroner has the legislated mandate to determine the cause of death, coroner numbers on fatalities due to crashes can at least be viewed as an officially sanctioned source. That is, some sort of gold standard on crash fatality frequencies should be established, and likely the coroners numbers are the best place to start. Alcohol and Drug Involvement Data Sources It is similarly impossible to know with absolute certainty the involvement of impairment by alcohol or drugs in traffic crashes that occur in a jurisdiction. Turning to the three crash-count data sources of the police, insurance claims and coroner data, only the police reports and coroner data address issues of alcohol or drugs and crashes. Police Data When the police report on a traffic crash, they often indicate whether or not they felt alcohol or drugs were contributory to the crash. In other instances they may report on whether or not they believed the driver had been drinking, whether or not they considered that to be contributory. Of course, both of these are judgment calls, and to a great extent will depend on the extent of impairment and the experience and training of the police officer. The crash report form may also allow blood alcohol content (BAC) measurements to be recorded, and may indicate whether or not alcohol or drug impaired driving charges were laid. A significant problem with these data is that for the most part, police in Canada are not trained in recognizing driver intoxication by either alcohol or drugs (especially considering the effects of alcohol impairment can resemble those caused by a head trauma), so their estimations probably under-represent actual frequencies of impaired driving in the crashes that they attend and report upon. For example Terhune (1982) found that driver BACs had to reach 0.20 or over before 58% of the police recognized impairment. Further, because the police generally arrive after the fact of the crash, they may not be in the position to even question drivers who may have gone to hospital or otherwise left the scene. On the other hand, because they are more likely to attend more serious crashes, and more serious crashes are more likely to involve driver intoxication, the sample of crashes that they attend and report upon will have a higher proportion of impaired crashes than the overall population of crashes. It would be very strange if these two sources of inaccuracy (underestimation of impairment within the sample attended and overrepresentation of impairment within the sample attended) cancelled each other out. In addition, police might be reluctant to report alcohol as a crash cause as that might lead to pressure to recommend changes, which in turn can be very time consuming. For example, in a survey of middle and senior police traffic enforcement managers, it was estimated that the laying of an impaired driving charge takes about 3.4 hours, with an additional 5.6 hours associated with the resultant court time (Mercer & Taylor, 2000). In their review of the responses of 1,545 Canadian police officers Jonah et al. (1999) concluded that the results suggest that many officers want to enforce DWI laws but that the numerous procedural and legal barriers that they confront often force them to exercise discretion in the laying of DWI charges (p.421). Nonetheless, because of their primarily subjective nature regarding the presence and role of impairment in a crash, police reports must be used with caution (see Mayhew, Bierness & Simpson, 1997 for a review focusing on Canadian data). Coroner Data While it is impossible for some drivers to be tested for alcohol or drugs because they died considerably after the crash and did not have blood or fluid samples taken soon after the crash, were incinerated, had been given IV fluids so their blood had been diluted and so on, the rate of testing fatally injured drivers for alcohol is relatively high in Canada. For example, in 1999, 81.9% of fatally injured drivers were tested for the presence of alcohol (Mayhew, Brown & Simpson, 2001). Further, a fatality may have been caused by an impaired driver who survived the crash and was never tested, resulting in an underestimation of fatalities associated with impaired driving using coroner data. On the other hand, the testing for drugs is more problematic. Testing for a wide range of drugs and their metabolites is difficult and expensive, as is the additional process of quantifying those drugs. There is no general consensus on what drugs and levels of drugs by themselves or in combination with other drugs or alcohol will result in impairment. In addition, unlike for alcohol, there is no agreed-upon or legislated level of any given drug that allows one to assume probable impairment. In addition, as noted above, while a driver may have had drugs on board, it is an open question as to whether or not those drugs had a negative influence on driving behaviours. Consequently, testing for the presence of drugs, and quantification of the amount of drugs present occurs much less often than similar testing for alcohol. For example, in British Columbia in 1996, coroner data show of those tested, 61.3% had neither alcohol nor drugs considered as contributing to the crash, 35.9% had alcohol-only contributing, 1.8% had a combination of alcohol and drugs contributing, and 1% had drugs alone contributing. Blood samples of all fatally injured drivers in BC in 1991 that were examined by the RCMP toxicology laboratory showed 37% alcohol-only; 11% alcohol-and-drugs; and 9% drugs-only. The most frequently found drugs were: 48% alcohol; 13% tetrahydrocannabinol or its metabolites (THC/THCCOOH); 4% cocaine; and 5% diazepam (Mercer & Jeffery, 1995). Arguably, coroner data, at least in Canada, are the most valid and reliable source for information on crash fatality frequencies and the presence of alcohol, but a less reliable source on the presence of drugs in fatal crashes. Finally, combining coroner data with other sources of information (e.g., police reports) to determine alcohol involvement in a crash can produce a high level of certainty as to whether or not alcohol may have been a factor in a crash. For example, of the 3315 persons killed in crashes in Canada in 1999, Mayhew et al. (2001) could identify whether or not alcohol was involved in 91.6% of the instances. Filling in the Gaps Crash and Victim Counts It seems fair to conclude that there is a range of counts of even fatalities and fatal crashes, and that the range of non-fatal crashes and associated injuries is even greater. Of course, there are reasonable explanations as to why these numbers vary, but when trying to estimate the extent of a problem like impaired driving, there is a profound difference in the absolute magnitude of the issue depending upon which base numbers one starts from. Consequently, it is essential that a rationale be developed that can be generally agreed upon for the purpose of order-or-magnitude estimation of crash and /or victim counts. In particular, the model should address the issues of injury and PDO crashes and drug as well as alcohol impairment. Fatal Crashes and Fatally injured Victims Of all of the sources of traffic-related fatally injured persons counts, the coroners numbers have the advantage of being both legally official and at least roughly comparable across provincial jurisdictions, as opposed to police figures and insurance figures which may vary in terms of local protocols and definitions. If coroner counts can be accepted as benchmark numbers, then the question of moving from those numbers to an estimate of the frequency of less serious crashes arises. One approach would be to make the assumption that there is a relatively constant ratio between different levels of crashes and different levels of injuries resulting from those crashes, and then using those ratios project the estimated number of non-fatal incidents from the count of fatal incidents. Certainly, everyday observation reveals that there are many more low severity fender benders than more serious crashes, and that as the seriousness of crashes increases, their frequencies decrease of the 250,217 crashes reported to ICBC in 1998, 203,245 involved property damage only (81.23%), 46,554 (18.61%) involved injury and 418 (0.17%) involved a fatality, while for the same year the MPI proportions were 87.3%, 12.5% and 0.2%. In a study of 30,871,392 persons involved in crashes, Miller and Blincoe (1994) presented frequencies divided in terms of the Maximum Abbreviated Injury Scale (MAIS), where a MAIS of 0 involved no injury, and a MAIS of 6 involved a fatal injury. Table 3 shows an extract of their published figures and manipulates the Ns into ratios of MAIS levels to fatality frequencies. That is, from the Ns, for every fatality there would have been 121.20 injuries. Similarly, for every fatality there would be at least 571.05 persons uninjured in crashes involving one or more casualty. The numbers used to arrive at 571.05 are: the sum of known persons in crashes that were uninjured (MAIS 0), plus a count of the number of vehicles in crashes were there were no casualties and where the number of persons was unknown. Table 3: Injury Severity - Miller & Blincoe 1994 Severity LevelN% of TotalFatality ratiosFatality ratiosPDO # Veh. (MAIS 0)24,035,743 77.86539.75MAIS 01,393,788 4.5131.30PDO +MAIS 0 = 571.05MAIS 14,617,228 14.96103.69MAIS 2566,849 1.8412.73MAIS 3180,111 0.584.04MAIS 421,756 0.070.49MAIS 511,386 0.040.26MAIS 1 5 = 121.20MAIS 6 (Fatal)44,531 0.141.00Fatal = 1.00TOTAL30,871,392 100.00 Turning to Canadian insurance data, on fatalities, injuries and PDO vehicle counts, British Columbian drivers are all insured by ICBC under a tort-based system, while drivers in Manitoba are insured by MPI under a no-fault system. The ratios of injuries-to-fatalities and PDO vehicles-to-fatalities are calculated in Appendix A and shown in Tables 4 and 5. The ICBC figures are based on a total sample of 2,296,374 and the MPI figures are based on a total sample of 614,644. Table 4: ICBC Victim & PDO Vehicle to Fatality Ratios 199619971998199920002001AllPDO Vehicles611.34647.17653.13705.58714.27767.62680.40Injuries158.26155.92144.68148.23144.69148.97150.32Fatalities1.001.001.001.001.001.001.00 Table 5: MPI Victim & PDO Vehicle to Fatality Ratios 199619971998199920002001AllPDO Vehicles803.72646.76508.38556.23643.56596.33618.12Injuries95.3075.5374.7182.9689.3093.2584.94Fatalities1.001.001.001.001.001.001.00 Comparing the Miller and Blincoe proportions to the ICBC proportions, a high injury claim rate for relatively low level injuries (Mercer & Halabisky, 1999) might in part explain the higher proportion of injury-level observations in the ICBC-related numbers. The relatively high ratio of fatalities to reported injuries (or low injury report levels) in the MPI data warrants further investigation. From the above, very roughly, one might speculate that, for every single fatality there are between 75 and 158 injuries, and between 508 and 767 vehicles damaged in non-injury events. This is a very wide margin for error in estimation, but arguably it would be better to work toward this sort of proportional balance than, for example, to accept the extreme kinds of injury and PDO frequency underestimations that would be derived from police reports. Clearly, one challenge for future research will be to refine these parameters. For the purposes of this discussion paper, the approximate mid-point between the BC and Manitoba samples at 118 injuries and 650 non-injury events (vehicles damaged) for every one fatality will be used. There are no particularly compelling reasons to use the mid point between these two data sets other than they are both recent, Canadian, and represent both tort-based and no-fault province-wide insurance company figures. Blood Alcohol Content (BAC) Count Estimations: Typically, policy makers, the news media and others turn to the local Motor Vehicle Branch-compiled police data for these estimates. Occasionally they will be used with some qualifications, but the qualifications are often lost in the translation to headlines and briefing note bullet points: In 1999, 2,139 (11.8%) of all police-attended injury collisions and 89 (24.5%) of all reported fatal crashes involved alcohol. (BC Motor Vehicle Branch, 2000, p. 43.) The above quote sounds authoritative and would take a page of text to properly qualify, and that qualification would have to take into account changes in police traffic personnel, changes in policing policies, training and so on, because all of those factors can in one manner or another influence counts and judgments. It has already been demonstrated that the number of police-attended crashes is substantially less than the number of crashes reported, say, to an insurance corporation. Compounding the underreporting of crash frequencies tied to police data, it has also been argued that, due to a number of factors such as a lack of training in the ability to detect impairment, the number of crashes associated with impairment is probably also underestimated in police-reported data. On the other hand, as noted earlier, coroner data in Canada are probably the best source of information on BAC and fatal crashes and police data are, at best, suspect. The problem is how to get BAC/ impairment estimations in crashes at the non-fatal levels of crash severity. It has long been established that higher levels of BAC are associated with more severe crashes (e.g., House et al., 1982; Warren et al., 1981). If it could be established (or at least reasonably assumed) that there is a relatively stable relationship between BAC levels and injury levels, then the extent of the alcohol-impaired driving problem could be projected from a knowledge of BACs in fatal crashes. In their examination of persons in crashes, Miller and Blincoe (1994) included the distribution of BAC levels by MAIS injury severity. Working from their reported frequencies, Table 6 illustrates the proportion of each MAIS level where the victim had a BAC of 0.10% or greater, as well as the ratios of injury and non-injury events to fatalities where the BAC was at that level. From these data, 39.68% of the fatally injured persons, 17.58% of the injured persons, and 13.75% of the uninjured persons had a BAC of 0.10% or greater, which in turn can be expressed as for every one fatality with a BAC of 0.10% or greater there were 53.59 injured persons and 197.91 uninjured persons at or above that BAC level. Table 6: Injury Severity & BAC - Miller & Blincoe NN BAC >=.10% MAIS >=.10% MAIS <=.10Fatality ratios <=.10PDO # Veh. (MAIS 0)24,035,743 3,349,149 13.93MAIS 0 1,393,788  148,162  10.63PDO +MAIS 0 = 13.75PDO+MAIS 0 = 197.91MAIS 14,617,228 712,198 15.42MAIS 2566,849 169,422 29.89MAIS 3180,111 55,957 31.07MAIS 421,756 6,636 30.50MAIS 5 11,386  4,510  39.61MAIS 1 5 = 17.58MAIS 1 5 = 53.69MAIS 6 (Fatal) 44,531  17,671  39.68Fatal = 39.68Fatal = 1.00TOTAL30,871,392 4,463,705 14.46 In addition, Mayhew et al. (2001) in their examination of 18,787 drivers estimated that 18.5% of serious injury crashes involved alcohol in a year where, of the 2,968 traffic-related fatalities drawn from coroner records, 30.3% of the crashes were alcohol-related (presumably the 18.5% figure would be reduced had the sample included crashes involving lesser levels of injury), while Bailey (1993), in a study of crash-involved drivers taken to hospital, found that 21% of those who were seriously injured and 14% of those who had minor injuries had a BAC in excess of 0.08%. One observation from the Miller & Blincoe (1994) data is that the proportion of positive BACs in the non-fatal injury group is roughly half of that in the fatally injured group - 39.68 to 17.58 (Mayhews injury-to-fatal ratio for alcohol-related crashes is higher than that, but it does not include lesser levels of injury). While using such a ratio is certainly not the way one would ideally wish to estimate a parameter of the extent of the impaired driving problem, it may be preferable than to abandon any attempt at order-of-magnitude estimation. Consequently, for the purposes of this paper, an assumption of a 1: 0.5 ratio of fatal-to-injury alcohol-involvement proportions will be used. To put this another way, for every 1% increase in the proportion of alcohol-impaired fatal crashes, there would be a 0.5% increase in the proportion of injury-only alcohol-impaired crashes. Similarly, for the purposes of order-of-magnitude estimation, an assumption of a 1:0.3 ratio of fatal-to-PDO alcohol-involvement proportions will be used (39.68 to 13.75). Again, to put this another way, for every 1% increase in the proportion of alcohol-related fatalities there would be a 0.3% increase in the proportion of alcohol-related PDO events. Alcohol-only, Alcohol-and-Drug, and Drug-only Impairment Estimations Governments assume that drugs other than alcohol can and do impair drivers. Section 253(a) of the Canadian Criminal Code states that it is an offense to operate a motor vehicle while impaired by alcohol or a drug, although there are no provisions for requiring a driver to be tested for drug impairment (Greenspan, M., 1989), while Royal Canadian Mounted Police breathalyzer training includes a lecture on the effects of drugs (Janzen & Walter, 1992). Researchers working in the area of traffic safety have established that drivers drink alcohol, take drugs, and have traffic crashes as a result of doing so. Generally, data are drawn from samples of drivers charged with being impaired, drivers in injury crashes, or drivers in fatal crashes. Regarding drivers suspected of being impaired (e.g., Asbjorg et al., 1990; Bjorneboe et al., 1986; Christopherson and Gjerde, 1989; Finkle, 1969; Poklis et al., 1987; Sutton and Childs, 1986; Valentour et al., 1980; White, 1981), in most instances, drug testing was only done in the absence of a given level of alcohol (usually the level needed to lay a per se charge), resulting in an incomplete drug epidemiology of these suspected drivers. On the other hand, from the drug-only results of these studies, one could argue that the drug effects may have led a police officer to suspect impairment, and that, consequently, drugs other than alcohol have an impairing role in driving. For example, Poklis et al. (1987) examined 137 drug-positive drivers who had been arrested for driving while impaired in Missouri, and reported 47% tested positive for marijuana (tetrahydrocannabinol), 22% benzodiazepines, 15% barbiturates, 11% opiates and 9% cocaine. It was estimated that, in 81% of the drug positive cases, the drivers were using the drugs for self-intoxication, as opposed to for legitimate medical reasons. Sutton and Childs (1986) examined 33 males and 6 females arrested for driving while impaired in Pittsburgh and found all had blood alcohol (average BAC .21%), with 39% positive for THC, and 5% positive for cocaine. Turning to studies drawn from injury crashes (e.g., Bailey, 1987; Dischinger and Birschbach, 1990; Honkanen et al., 1980; Kapur et al., 1990; Lucke, 1990), Bailey (1987) examined 901 drivers admitted to a trauma unit in New Zealand and found 20% had been drinking and 6.5% had used cannabis (THC), although of the fatally injured drivers, 66.7% had been drinking. He noted that THC use was related to younger drivers, while prescription drugs tended to be used by older drivers. Lucke (1990) reported that, of his sample of 200 drivers admitted to a hospital in Illinois, 54% had consumed alcohol, drugs, or both, 42% had consumed alcohol, 32% had consumed drugs, and 42% of those consuming drugs had also consumed alcohol. Drug and/or alcohol use was associated with nighttime, single vehicle crashes, and none of the police reports on the crashes referred to the use of drugs. Finally, Kapur et al. (1990) examined 277 drivers admitted to a Toronto hospital and found that 66% had used alcohol, drugs, or both, with alcohol (34%) being the most common drug, followed by cannabinoids and benzodiazepines. While hospital admission studies of injured drivers can give a clearer picture of the epidemiology of alcohol and drug use among drivers than can samples of suspected impaired drivers, they seldom can speak to the role of drugs and alcohol in traffic crashes. Similarly, research dealing with samples of fatally injured drivers (e.g., Budd et al., 1989; Caplan et al., 1990; Donelson, 1990; Everest and Tunbridge, 1990; Peel and Jeffery, 1990; Penttila et al., 1989; Root, 1990; Sweedler and Quinlan, 1990; Williams et al., 1985) can give a reasonably good epidemiological picture of the incidence of the use of alcohol and drugs, but seldom can supply information on the role of these substances in traffic crashes. Considering use levels, Budd et al. (1989) examined blood or urine samples from 600 fatally injured drivers in Los Angeles County and found a 41.5% alcohol use rate, 19% marijuana use, 8% cocaine use, and 2% barbiturate use, while Caplan et al. (1990) reported 42% alcohol use and 17% drug use, with cannabinoids at 7.4% and cocaine at 7.8% in their sample of 269 fatally injured drivers in the state of Maryland. Root (1990), reported 36% alcohol-only, 9% alcohol-and-drugs and 13% drugs-only findings in his sample of 193 fatally injured drivers in California, and Williams et al. (1985) noted that there was no statistically significant difference in average BAC between their alcohol-only and alcohol-and-drugs groupings of 440 fatally injured male California drivers. A Canadian sample of 1,169 fatally injured drivers yielded positive tests of 57% for alcohol and 11% for cannibinoids (Donelson, 1990). (See also Compton & Anderson, 1985, Simpson, 1985, and Simpson, 1990, for more complete reviews and Simpson and Vingilis, 1991, for a methodological critique of the area.) However, there is little clear evidence that drugs other than alcohol, or drugs with alcohol, play a significant role in the causation of traffic crashes. One might infer that, if the frequency of drugs found in the general driving population were lower than the frequency found in the crash victim population, then drugs might play a role in crash causation, and that generally appears to be the case. In a roadside survey study in Quebec in 1999 (Dussault et al. 2001), toxicological analysis of 2,281 urine samples given by drivers showed 5.22% cannabis, 3.66% benzodiazepines, 1.09% cocaine 1.08% opiates, 0.35% barbiturates, 0.07% amphetamines and 0.03% PCP. Additionally, in a 1998 telephone survey of 1,422 drivers in British Columbia, 3.2% admitted to using marijuana and driving within the past month (Mercer & Linguanti, 1999). In their examination of 13,618 suspected impaired drivers who showed sufficient signs of impairment to be taken to a police station for a breathalyzer test, Jeffery et al. (2000) found that 6.4% had a BAC under 0.08% and 0.8% had a BAC of 0.0%, suggesting that they may have been impaired by drugs other than alcohol, while De Gier (1993 cited in Vingilis, 2002) estimated that at least 10% of all people injured or killed in road crashes were taking some sort of psychotropic medication. One study that argued that drugs other than alcohol contribute causally to traffic crashes examined one years data on all fatally injured drivers in British Columbia (Mercer & Jeffery, 1995). Blood samples, driver records, and crash records of these 41 female and 186 male fatally injured drivers were examined. Toxicologies showed: 37% alcohol-only; 11% alcohol-and-drugs; and 9% drugs-only. The most frequently found drugs were: 48% alcohol; 13% tetrahydrocannabinol or its metabolites (THC/THCCOOH); 4% cocaine; and 5% diazepam. With regard to causes, they noted that in the alcohol-only group, the police identified alcohol impairment as a cause in 55% of the crashes, but identified alcohol impairment in 84% in the alcohol-and-drug group, even though the alcohol levels between the two groups did not differ. That is, they argued that the increase in the police attribution of impairment as a cause was due to the doubly impairing effects of alcohol combined with drugs. They also noted that impairment was not cited as a cause in any of the drug-only group, but that the police virtually never cite impairment by drugs as a crash cause, possibly because of their inability to legally gather drug impairment evidence and possibly because of a lack of drug recognition training. Finally, as noted above, coroner data on drugs and driving will tend to underestimate the presence of drugs. While the exact role of drugs other than alcohol in traffic crashes is still to be determined, the presence of these drugs in the driving population and the crash population is undeniable. On the other hand, except for the relatively rare samples of intensively tested fatally injured drivers, the extent of the presence of these substances in traffic crashes can only be roughly estimated. Working from the data presented by Mercer & Jeffery (1995) on toxicologies of fatally injured drivers in British Columbia, of those who had a BAC of 0.08% or more, 76% had alcohol only, 24% had a combination of alcohol and drugs, comprising 100% of those presumed to be impaired at a 0.08% BAC. There was also an additional number of persons with drugs but no alcohol on board and that number was the equivalent of 19% of those with a BAC of 0.08%. If one assumed that the relative proportion of alcohol, alcohol and drugs, and drugs only remained constant within the impaired group in each crash/injury severity level, then these proportions could be roughly applied to each of these levels. That is, the percentage of BACs over 0.08% would be multiplied by 0.75 to estimate the proportion of alcohol-only incidents, by 0.25 for alcohol and drugs combined and by 0.20 for drug-only incidents, then summed to estimate the overall proportion of incidents involving alcohol impairment, alcohol and drug impairment, and drug-only impairment. However, while there is little controversy over whether or not a person is impaired at a BAC of 0.08%, the issue of drug types, levels, and combinations that may or may not impair driving is still under investigation (see, for example, Vingilis & MacDonald, 2002 and Longo et al., 2001). Nevertheless, some sort of inclusion of a drug-impaired factor should go into a model of driving and impairment. Consequently, for the purposes of this discussion paper, for the percentage of BACs over 0.08%, the multipliers of 0.75 will be used to calculate alcohol-only impairment, 0.25 for alcohol and drug impairment, and 0.10 (half of the 0.20 from Mercer & Jeffery) will be used to calculate drug-only impairment. An Order-of-Magnitude Calculation Facts and Arguments In order for politicians and other officials to make relatively informed legislative and policy decisions, there is a need to know the relative magnitude of the impaired driving crash problem. Within this context, it may be better to have an order-of-magnitude global estimation that is imprecise than to have more precise measures of bits of the problem but without an attempt at an overview. The following is an imprecise order-of-magnitude estimation of the extent of the impaired driving crash problem in Canada in 1999. Assumptions: Data from police crash reports can be used to help define the extent of impaired driving but should not be used exclusively to measure either the number of crashes or the proportion of impaired driving crashes. In particular, injury and property damage only (PDO) crashes are under-reported, as is the contribution of impairment to crashes. Coroner data are the most accurate source of the count of fatal injuries resulting from impaired driving and the proportion of drivers having positive BACs, especially those supplemented by police reports. Coroner data underestimate the involvement of drugs in traffic crashes. The relative levels of severity of crashes remain (somewhat) constant, so that, by knowing the proportional relationships among the levels of severity and the absolute frequency of one level, the frequencies of other levels of severity can be estimated. For the purposes of this estimate, 118 injuries and 650 non-injury events (vehicles damaged) for every one fatality will be used. The relative levels of impairment associated with crashes vary systematically with crash severity, so that by knowing the proportion of impaired crashes in one level of crash severity the proportions can be estimated in other levels of severity. For the purposes of this estimation, an assumption of a 1: 0.5 ratio of fatal-to-injury alcohol-involvement proportions and a 1:0.3 ratio of fatal-to-PDO alcohol-involvement proportions will be used. The relative proportions of alcohol-only, alcohol-and-drugs, and drugs-only leading to impairment-related crashes remain constant from crash severity level to crash severity level, while the proportion of impairment-related crashes decreases with a decrease in the levels of crash severity. For the purposes of this estimation, the percentage of BACs over 0.08% would be multiplied by 0.75 to estimate the proportion of alcohol-only incidents, by 0.25 for alcohol and drugs combined and by 0.10 for drug-only incidents, then summed to estimate the overall proportion of incidents involving alcohol and or drug impairment. All best guess order-of-magnitude assumptions of proportional relationships can be improved with better data and need to be reviewed. Estimation: From Mayhew et al (2001) in 1999 in Canada, 3,315 persons died in motor vehicle incidents on and off road within 12 months of the incident. In 3,037 (91.6%) of these cases, it was possible to determine if alcohol was a factor. Of these known cases, 1,038 (34.2%) involved alcohol. Extrapolating this figure to the total number of motor vehicle fatalities (3,315 x 0.324) it can be estimated that in Canada during 1999, 1,134 persons died in alcohol-related crashes. (p. 14). The number of injured persons and the number of vehicles in PDO crashes were estimated using multipliers of 118 and 650 to the number of fatalities. The % alcohol involved for these two groups (injury and PDO vehicles) was estimated using multipliers of 0.5 and 0.3 to the proportion of alcohol involved for the fatally injured group. The estimated percent and number impaired was calculated using a multiplier of 0.75 on alcohol involved to estimate alcohol-only, 0.25 on alcohol involved to estimate alcohol and drugs and 0.10 on alcohol only to estimate drugs only. Table 7: Estimated Fatalities, Injuries & PDO Vehicles, Canada, 1999 FatalitiesInjuries @118PDO veh. @ 650N 3,315391,1702,154,750% alcohol involved34.20%17.10%10.26%% alcohol only @.7525.65%12.83%7.70%%alcohol+drug @.258.55%4.28%2.57%%drug only @.103.42%1.71%1.03%% Impaired37.62%18.81%11.29%N Impaired1,24773,579243,185 In order to help dispel any notion that these estimates are precise measures, one should always qualify them with words like it can be roughly estimated that approximately and round appropriately, as in: From an examination of police and coroner data it can be roughly estimated that in Canada in 1999 impaired driving crashes resulted in approximately 1,200 persons killed, 73,000 persons injured, and 243,000 persons/vehicles involved in property-damage-only (PDO) crashes.  In turn, using the ICBC ratios noted above of 1.2 fatalities per fatal crash, 1.11 injuries per fatal crash, 1.44 injuries per injury crash and 1.52 vehicles per PDO crash one can to move to the crash as the units of analysis. This results in an estimation of 1,039 fatal crashes, 50,295 injury crashes and 159,990 PDO crashes associated with impairment by alcohol, alcohol and drugs, or drugs only. Using this model, Figure 1 and Table 8 show the proportion impaired by different sources across the three crash types. Figure 1  Table 8: % Impairment Source by Crash Type fatalinjuryPDODrug3.4%1.7%1.0%Alc+drug8.6%4.3%2.6%Alcohol25.7%12.8%7.7%Non-impaired62.4%81.2%88.7% Figure 2 and Table 9 show the frequencies of victims by impairment source. It is noteworthy that the count of fatally injured victims is overwhelmed by the counts in the other two crash categories such that they cannot register on the scale of the graph. This illustrates the importance of including a full assessment of impairment in crash types that are normally under-reported in terms of impairment as a cause. Figure 2: # Victim/ PDO Vehicle - Impairment Source by Crash Type  Table 9: # Victim/ PDO Vehicle - Impairment Source by Crash Type fatalityinjuryPDODrug 113  6,689  22,108 Alc+drug 283  16,723  55,269 Alcohol 850  50,168  165,808 Non-impaired 2,068  317,591  1,911,565  Three Crash Costing Models The assessment of damage to society that stems from traffic crashes is as much an exercise in philosophy and values as it is in costing and economics. That is, costs can be defined in a number of ways, resulting in a range of conclusions. For example: As the property damage component of a crash results in costs to the individual vehicle owner, either directly or through insurance premiums, an argument can be made that this crash cost component is not a cost to society because the vehicle repair industry benefits and thus the equation is balanced. The question of whether or not to include the costs of crash prevention in the "social costs" equation can also pose difficulty, as these prevention costs (public education, traffic enforcement etc.) can rise as crash rates fall, resulting in no net dollar gain. What time frame should be used to assess costs? If a person becomes permanently disabled as the result of a crash, should the lost opportunity cost be calculated over the person's lifetime and assessed into the year of the crash, or amortized over the years the person is expected to live? What is the ethical process that can put a cash value on something that cannot be purchased, like a life? Indeed, can money be used as a common denominator in the arena of human pain, suffering, and loss? And so on. Indeed, every point in a costing exercise involves assumptions and definitions of terms that result in the end product being at best an educated guess. However, as long as the costing model is understandable, replicable, and apparently reasonable, it can be used to assess crash costs. Three Perspectives: Broadly, there are three kinds of questions that are asked about the result of a traffic crash: How much will this cost me in real dollars spent? (Real Dollar Estimate -- RDE) How much will this cost me in terms of lost goods, opportunity, or productivity? (Discounted Future Earnings --DFE) How much would I pay for this not to have happened? (Willingness to Pay -- WTP) Real Dollar Estimates: The answer to the first question involves an assessment of expenditures directly stemming from an occurrence - a "real dollar" estimate (RDE). For a traffic crash, vehicle repairs, medical costs, insurance payouts, and the like would be considered, and the result would be a figure representing actual dollars spent. Obviously, this sort of figure also represents actual dollars that would be saved if the crash had not occurred. However, it does not take into consideration the broader harm to society resulting from a traffic crash, such as lives made less productive or the costs of enforcement and the courts. Consequently, it seriously underestimates the costs of a crash to society. On the other hand, it may be of particular interest to policy makers, as the dollars associated with the crash are connected to largely immediate and tangible fiscal consequences. Discounted Future Earnings: The second question attempts to consider the costs of a crash by calculating not only immediate dollars spent, but also the productivity that society loses from lost days at work, reduced employment opportunity, or a life shortened. It tries to build a scenario of what would have happened if there had been no crash, and then place a value on that. For example, if a 45-year old steel worker was killed, this approach would consider the value of 20 (assumed) years of lost productivity and add that to the crash cost. This "discounted future earnings" (DFE) approach results in very much larger costs than the RDE model, but is not a realistic estimate of actual dollar savings had the crash not occurred. That is, it deals with theoretical savings, not actual dollars spent. On the other hand, it is a much better way of assessing the harm to society than the RDE approach, because it is more comprehensive. However, in the DFE model, retired persons, persons in low-paying jobs, unemployable persons and the like -- who may well be making significant contributions to society -- do not carry much weight. Consequently, the DFE model, by only looking at the dollar-value of productivity in the market and ignoring the rest of a person's life, also underestimates the harm to society caused by traffic crashes. Finally, it may be of less interest to policy makers because the costs associated with the crash are, in effect, spread over many years, so that the savings resulting from crash reductions would not be fully realized for many years. Willingness to Pay: The third question tries to get at the global value of the occurrence by asking what society would pay to avoid that occurrence. This "willingness to pay" (WTP) model reasons that the true cost of a crash to society is what society would be willing to pay for it never to have occurred. While the methodologies of arriving at this global value can be as circuitous and flawed as those of the RDE or DFE approaches, the WTP model is the most conceptually clean and reasonable -- a thing is only worth what people are willing to pay for it, irrespective of attached dollar "costs". On the other hand, arguably the amount persons would be willing to pay would vary from time to time and circumstance to circumstance. This may be the best estimate of "social costs" but of least use to policy makers. Like the DFE model, the WTP model does not deal in immediate real dollar costs, and cannot be used to estimate real dollar savings of any crash reductions. Impaired Driving Crash Cost Estimations It has been estimated that all crashes cost national economies up to 2.8% of their GNP if quality of life costs are eliminated and up to 5.7% if quality of life calculations are included (Elvik, 2000), with a mean weighted value of 1.4% and 3.1% respectively. As impaired driving crashes tend to be more serious (and thus costly) than non-impaired driving crashes, one would expect them to comprise a substantial proportion of the total crash costs. Miller & Blewden (2001) estimated that alcohol-related crashes made up 38.5% of the crash costs in New Zealand, while Miller & Blincoe (1994) looking at American data stated that alcohol-involved crashes accounted for over a third of all motor vehicle crash costs. Unlike American tort settlements associated with crashes, compensation for pain and suffering and the like is limited in Canada. However, medical costs (bodily injury), lost wages, court costs, material damage costs, rehabilitation costs and crash and death benefits are all covered by crash insurance, and represent a good measure of the real dollar costs of crashes. In 1997 ICBC estimated that an average fatal crash cost $272,970, an average injury crash cost $24,552, and a PDO crash cost $1,539 (Mercer & Halabisky, 1999). Statistics Canada figures show inflation, as measured by the Consumer Price Index, increased by 2.7% between 1997 and 1999, so in 1999 dollars the costs would be $280,340, $25,215, and $1,581. Vodden et al. (1994) examined crash costs in Ontario, Canada, and developed both a deferred future income and willingness-to-pay model. Their estimations for fatal, injury-only, and PDO crashes using a DFE model were $831,429, $20,084 and $6,136, while their WTP model estimated $6,311,772, $27,112 and $6,136 respectively, all in 1990 dollars. Statistics Canada figures show inflation, as measured by the Consumer Price Index, increased by 18.4% between 1997 and 1999, so in 1999 dollars the costs within the DFE model would be $984412, $23779, and $7265, while within the WTP model they would be $7,473,138, $32,101 and $7,265. Table 10: Crash costs by Costing Model in 1999$ fatalinjury-onlyPDOReal Dollar Estimate$280,340$25,215$1,581Deferred Future Earnings$984,412$23,779$7,265Willingness to Pay$7,473,138$32,101$7,265 Table 11 and Figure 3 illustrate the outcome of using the values from the three models to estimate the cost of impaired (alcohol only, alcohol & drugs, drugs only) driving crashes in Canada in 1999. Table 11: Estimated Cost of Impaired Driving Crashes by Costing Models All Impairedfatalinjury-onlyPDOSumEstimated crash N 1,039  50,295  159,990  211,325 Real Dollar Estimate$291,344,243$1,268,195,972$252,872,971$1,812,413,186Deferred Future Earnings$1,023,052,566$1,195,999,410$1,162,332,552$3,381,384,528Willingness to Pay$7,766,477,399$1,614,515,834$1,162,332,552$10,543,325,785 Similar to the estimates involving crashes and victims, these sorts of estimations should be well qualified and appropriately rounded to reflect their approximate natures (e.g., $1.8 billion, not $1,812,413,186). Figure 3: Estimated Cost of Impaired Driving Crashes by Costing Models  As the proportion of alcohol-related fatalities were multiplied by 0.75 to obtain an estimate of alcohol-only impaired incidents, by 0.25 to obtain an estimate of alcohol-plus-drug impaired incidents, and by 0.10 to obtain an estimate of drug-only impaired incidents, the overall figures for all impaired crashes can be broken out into the three categories by multiplying the figures in Table 13 by 68.18%, 22.73% and 9.09%, as is illustrated in Tables 12 through 14. Table 12: Impairment by Alcohol only Crash Costs by Costing Model fatalinjury-onlyPDOSumEstimated crash N 709  34,292  109,084  144,085 Real Dollar Estimate $198,643,802$864,679,072$172,413,389$1,235,736,263Deferred Future Earning$697,535,840$815,454,143$792,499,467$2,305,489,451Willingness to Pay $5,295,325,499$1,100,806,250$792,499,467$7,188,631,217 Table 13: Impairment by Alcohol & Drugs Crash Costs by Costing Model fatalinjury-onlyPDOSumEstimated crash N 236  11,431  36,361  48,028 Real Dollar Estimate $66,214,601$288,226,357$57,471,130$411,912,088Deferred Future Earnings$232,511,947$271,818,048$264,166,489$768,496,484Willingness to Pay $1,765,108,500$366,935,417$264,166,489$2,396,210,406 Table 14: Impairment by Drugs only Crash Costs by Costing Model fatalinjury-onlyPDOSumEstimated crash N 94  4,572  14,545  19,211 Real Dollar Estimate $26,485,840$115,290,543$22,988,452$164,764,835Deferred Future Earnings$93,004,779$108,727,219$105,666,596$307,398,593Willingness to Pay $706,043,400$146,774,167$105,666,596$958,484,162 Conclusion In order to allocate countermeasure resources, policy makers, legislators and others need to know the overall extent of the problem of impaired driving, at least at an order-of-magnitude level. However, no single data source is acceptable for such an estimation: insurance-based crash counts are probably the best assessment of injury and PDO crash numbers (but not their alcohol involvement), police reports the best source of crash descriptions (but not epidemiology), coroner reports and coroner reports enhanced by police reports are the best source for alcohol involvement information and fatality counts, and toxicological studies the best source for drug involvement information. Finally, in terms of tracking changes in alcohol involvement in fatal crashes, coroner and enhanced coroner data are likely the most reliable information source. The approach taken in this paper has been to try to combine the strengths of different data sources to identify relationships and to develop ratios that would allow the extent of the impaired driving problem to be estimated, working from the most reliable data source enhanced coroner data. The impact of impaired driving on injury and PDO crashes was estimated, as was the influence of drug-impaired driving. While it would be irresponsible to attempt at considering this impact without including non-fatal crashes and drug-impairment, the actual ratios used to arrive at this estimation are certainly up for review. For example, it may be that about 10% of all impaired crashes involve drugs only, or it may be 5% or 15% or some other proportion that is a matter that warrants research and discussion but not to include some estimation would underestimate the extent of impaired driving crashes. Similarly, an estimation of the human and financial impact of impaired driving using only fatal crashes would seriously underestimate the problem. In terms of specific costing models, the model of choice is probably a function of the use to which it would be put. Analysts trying to measure or project program effectiveness in order to build budgets would likely be attracted to a real dollar estimate, while those trying to reflect the extent of harm that impaired driving causes to society would be better to use a model based on willingness-to-pay. Again, there are no absolutely right or wrong approaches or answers, just approximations. Having said that, in terms of large, medium and small, these analyses and estimations argue that impaired driving costs are massive. Certainly, it costs Canadians at least almost two billion dollars per year, and perhaps as much as over ten billion dollars per year, and likely involves over 200,000 crashes per year. Finally, the next steps in estimating the impact of impaired driving might be for those involved in research and advocacy in the area to work toward refining a more complete model that could be used by decision makers to understand the extent of the preventable tragedies of impaired driving crashes. References AAAM. The Abbreviated Injury Scale. Des Plaines, IL: Association for the Advancement of Medicine, 1985. Ajsbjorg, S.C., Bjorneboe, H.G. and Morland, J. Screening for drug use among drivers suspected of drunken or drugged driving. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Bailey, J.P. The role of alcohol and drugs in injury road accidents in New Zealand. In Noordzji, P.C. and Rozbach, R. (Eds.) Alcohol, Drugs and Traffic Safety - T86. Elsevier Scientific Publications, New York, 1987. Bailey, J.P. The Impact Of The Sale Of Liquor Act 1989 On Injury Drink Driving Accidents. Institute of Environmental Science and Research Limited, Poirua, New Zealand, 1993. Bierness, D.J. Alcohol involvement in recreational vehicle operator fatalities in Canada. In the Proceedings Of The 16th International Conference On Alcohol, Drugs And Traffic Safety, (C. Dussault & D. Mayhew, Eds.) August 4-9, 2002, Montreal, Quebec, 2002. (online at www.icadts.org) Bierness, D.J. Alcohol involvement in snowmobile operator fatalities in Canada. Canadian Journal of Public Health, 92 (5), 359-360, 2001 Bjorneboe, A., Bjorneboe, G.E.A., Bugge, A., Christophersen, A.S., Drevon, C.A., Gadeholdt, G. Gjerde, A. and Morland, J. Drugged driving in Norway, 1986. In Prevention and Control: Realities and Aspirations. Proceedings of the 35th International Congress on Alcoholism and Drug Dependence, Oslo, 1988. National Directorate for the Prevention of Alcohol and Drug Problems, Oslo, Norway, 1989. Budd, R.D., Muto, J.J. and Wong, J.K. Drugs of abuse found in fatally injured drivers in Los Angeles County. Drug and Alcohol Dependence, 23, 1989. Canadian Center for Justice Statistics. Canadian Crime Statistics, 1998. Statistics Canada, Ottawa, Ontario, 1999. Caplan, Y.H., Levine, B.S., Goldberger, B.A. Drugs in driver fatalities: a preliminary study in the state of Maryland. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Chesher, G.B., Franks, H., Hensley, V.R., Hensley, W.J., Jackson, D.M., Starmer, G.A. and Teo, R.K.C. The interaction of ethanol and THC in man. The effects of perceptual, cognitive and motor functions. Medical Journal of Australia, 2, 159-163, 1976. Christophersen, A.S. and Gjerde, H. Comparison of female and male drugged drivers. In Valverius, M.R. (Ed.) Women, Alcohol, Drugs and Traffic. Dalctrof, Stockholm, 1989. Compton, R.P. and Anderson, T.E. The Incidence of Driving Under the Influence of Drugs, 1985: An Update of the State of Knowledge. Office of Driver and Pedestrian Research, National Highway Traffic Safety Administration, Washington, D.C., 1985. Dalrup, T., Reudenbach, G. and Kimm, K. Cannabis und alkohol im strassenverkehr. Blutalkohol, 24 (2), 1987. De Gier, J.J. Driving licences and known use of licit and illicit drugs. Maastricht: Institute for Human Psychopharmacology, University of Limburg, 1993. Dischinger, P.C. and Birschbach, J. Alcohol and drug use among victims of vehicular crashes admitted to trauma centers. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Dussault, C., Lemire, A.M., Bouchard, J. & Brault, M. Drug use among Quebec Drivers: The 1999 Roadside Survey. Highway Safety Research and Strategy, Societe de lassurance automobile du Quebec, Quebec, Canada, 2001. Donelson, A.C. Cannabis and alcohol use among drivers and pedestrians fatally injured in traffic crashes. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Elvik, R. How much do road accidents cost the national economy? Accident Analysis & Prevention, 32, p. 849-851, 2000. Elvik, R. & Mysen, A.B., Incomplete accident reporting: meta-analysis of studies made in 13 countries. Transportation Research Record 1665, pp. 133-140, 1999. Everest, J.T. and Tunbridge, R.J. The incidence of drugs in road accident fatalities in Great Britain. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Finkle, B.S. Drugs in drinking drivers: a study of 2,500 cases. Journal of Safety Research, 1(4), 1969. Fleming, Z. & Mercer, G.W. (1998) Enhanced CounterAttack: Results of a Province-Wide Impaired Driving Roadcheck Campaign, June4 to December, 1997. Insurance Corporation of British Columbia, North Vancouver BC, January 1999. Fleming, Z. & Mercer, G.W. (2001) Enhanced CounterAttack: Results of a Province-Wide Impaired Driving Roadcheck Campaign, May 14 to December 31, 1998. Insurance Corporation of British Columbia, North Vancouver BC, January 2001. Greenspan, E.L. Martin's Annual Criminal Code, 1990. Canadian Law Book Inc., Aurora, Ontario, 1989. Honkanen, R., Ertansa, L., Linnoila, M., Alha, A., Lukkari, I., Karlsson, M., Kilviluoto, O. and Puro, M. The role of drugs in traffic accidents. British Medical Journal, 281, 1980. House, E.G., Waller, P.F. & Stewart, J.R. Blood alcohol level and injury in traffic crashes. In the Proceedings of the American Association for Automotive Medicine Conference, Ottawa Ontario, October 4-6, 1982. Janzen, K.E. and Walter, L. Drugs and the impaired driver. RCMP Gazette, 54 (11), 1992. Jeffery, W.K., Fromm, F., Go, G.V., Worsley, R., & McNeill, J.H. Data analysis of breath alcohol concentrations taken from suspected impaired drivers in British Columbia, 1992 1993. Journal of Traffic Medicine, 28 (1-2), 31-34, 2000. Jonah, B., Yuen, L., Au-Yeung, E., Paterson, D., Dawson, N., Thissen, R. & Arora, H. Front-line police officers practices, perceptions, and attitudes about the enforcement of impaired driving laws in Canada. Accident Analysis & Prevention, 311, 421 433, 1999. Kapur, B.M., Vingilis, E., McLellan, B., Nelson, W. and Sheu, W.J. Alcohol and drugs among MVA admissions to a regional trauma unit. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Longo, M.C., Lokan, R.J. & White, J.M. The relationship between benzodiazepine concentration and vehicle crash culpability. Journal of Traffic Medicine, 29 (1-2) 36-43, 2001. Los Angeles Police Department. Monrovia Police Department Drug Recognition Expert Program 1991 Staff Report. Los Angeles, California, 1992. Lucke, R.E. Alcohol and other drug involvement in serious traffic crashes - development of a research protocol. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Maki, M. and Linnoila, M. Traffic accident rates among Finnish out-patients. Accident Analysis and Prevention, 8, 1976. Manitoba Public Insurance Corporation Strategic Research Department.; Traffic Collision Statistics Report, 2002 (Manitoba Public Insurance crash counts were supplied by Mr. Oscar Mwanza of MPI) Manitoba Transportation & Government Services. Traffic Collision Statistics Report, 2000. Mayhew, D. R., Bierness, D. & Simpson, H.M. Indicators of the Alcohol-Crash Problem. Traffic Injury Research Foundation of Canada, Ottawa, Ontario Report to Trransport Canada, December 1997. Mayhew, D.R., Brown, S.W. & Simpson, H.M. The Alcohol-Crash Problem in Canada: 1999. Traffic Injury Research Foundation of Canada, Ottawa, Ontario, 2001. Mercer, G. William & Halabisky, L. (1999) Crash-type and injury-type cost calculations: the role of whiplash in a tort-based system. In: Proceedings of Traffic Safety and Auto Engineering Stream of the World Congress on Whiplash-Associated Disorders. Vancouver, British Columbia, Canada, Feb. 7 11, 1999. Mercer, G.W. and Jeffery, W.K. (1995) Alcohol, drugs and impairment in fatal traffic accidents in British Columbia. Accident Analysis and Prevention, 27(3), 335-343. Mercer G.W. & Linguanti, F. Public Attitudes and Behaviours on Alcohol and Drug Impaired Driving in British Columbia, 1998. Paper presented at the Canadian Association of Forensic Scientists conference, Edmonton, Alberta, November, 1999. Available through the Insurance Corporation of British Columbia, North Vancouver, British Columbia. Mercer G.W. & Taylor, R. (2000) Traffic policing parameters: British Columbia Association of Cheifs of Police Safety Committee Questionnaire Responses. Insurance Corporation of British Columbia and BCACP, North Vancouver, BC. Miller, T.R. and Blincoe, L.J. Incidence and cost of alcohol-involved crashes in the United States. Accident Analysis and Prevention, 26 (5), 1994. Miller, T.R. & Blewden, M. Costs of alcohol-related crashes: New Zealand estimates and suggested measures for use internationally. Accident Analysis & Prevention, 33, pp. 783-791, 2001. Milner, G. Drinking and driving in 753 general practice and psychiatric patients on psychotropic drugs. British Journal of Psychiatry, 115, 1969. Motor Vehicle Branch. 1990 Traffic Accident Statistics. Ministry of Attorney General, Victoria, British Columbia, 1992. Motor Vehicle Branch. 1999 Traffic Collision Statistics. Ministry of Attorney General, Victoria, British Columbia, 2001. Peel, H.W. and Jeffery, W.K. A report on the incidence of drugs and driving in Canada. Canadian Society of Forensic Sciences Journal, 23 (2,3), 1990. Penttila, A., Karhunen, P.J. and Pikkarainen, J. Women, alcohol, drugs and fatal traffic accidents in Finland. In Valverius,M.R. (Ed.) Women, Alcohol, Drugs and Traffic. Dalctrof, Stockholm, 1989. Perez-Reyes, M., Hicks, R.E., Bumberry, J., Jeffcoat, R., and Cook, C.E. Interaction between marijuana and ethanol: Effects on psychomotor performance. Alcohol Clinical Experimental Research, 12, 268-276, 1988. Polkis, A., Maginn, D. and Barr, J.L. Drug findings in "driving under the influence of drugs" cases: a problem of illicit drug use. Drug and Alcohol Dependence, 20, 1987. Root, I. Drugs of abuse and fatal automobile accidents. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Simpson, H.M. Polydrug effects and traffic safety. Alcohol, Drugs and Driving, 1, 1985. Simpson, H.M. The epidemiology of traffic injuries involving alcohol and other drugs. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. Simpson, H.M. and Vingilis, E. Special population surveys: the epidemiology of drugs and driving. In Ferrara, S.D. and Giorgetti, R. (Eds.) Methodology in Man-Machine Interaction and Epidemiology on Drugs and Traffic Safety. Addiction Research Foundation of Italy, University of Padova, Italy, 1991. Smith, G.S., Keyl, P.M., Hadley, J.A., Bartley, C.L., Foss, R.D., Tolbert, W.G. & McKnight, J. Drinking and boating: a population-based case-control study of recreational boating facilities. In the Proceedings Of The 16th International Conference On Alcohol, Drugs And Traffic Safety, (C. Dussault & D. Mayhew, Eds.) August 4-9, 2002, Montreal, Quebec, 2002. (online at www.icadts.org) Statistics Canada. Economic Indicators Canada www.statcan.ca Canadian Statistics Menu- Latest Indicators. Stein, A.C. A simulator study of the effects of alcohol and marijuana on driving behaviour. In Noordzji, P.C. and Rozbach, R. (Eds.) Alcohol, Drugs and Traffic Safety - T86. Elsevier Scientific Publications, New York, 1987. Subramanian, R. Transistioning to Multiple Imputation A new mathod to estimate missing blood alcohol concentration (BAC) values in FARS. National Highway Traffic Safety Administration (NHTSA) Technical Report DOT HS 809 403, 2002. Sutton, L.R. and Childs, P. Driving under the influence of alcohol and/or other controlled substances in a metropolitan area. In Society of Automotive Engineers (Eds.) Alcohol, Accidents and Injuries. Detroit, 1986. Sweedler, B.M. and Quinlan, K.E. Alcohol and drugs among fatally injured drivers of heavy trucks. In Perrine, M.W.B. (Ed.) Alcohol, Drugs and Traffic Safety - T89. Proceedings of the International Conference on Alcohol, Drugs and Traffic Safety, National Safety Council, Chicago, 1990. United States Department of Transportation, National Highway and Traffic Safety Administration. Drug Evaluation and Classification Program. Washington, D.C., 1989. Terhune, K.W. An evaluation of crash culpability to assess alcohol and drug impairment effects. In 26th Annual Proceedings, AAAM, Ottawa, Canada, Oct. 4-6, 1982; 329-348. Valentour, J.C., McGee, M.P., Edwards, R.P. and Goza, B.S. A survey of drug use among impaired drivers in Virginia. Medico-Legal Bulletin, 29 (5), 1980. Vingilis, E. & MacDonald, S. Review: drugs and traffic collisions. Traffic Injury Prevention 3, 1-11, 2002. Vodden, K, Smith, D., Meng, R., Miller, T., Lall, A., Beirness, D.J., Mayhew, D, Simpson, H.M., Kazakov, A., & Tasca, L. The Social Cost of Motor Vehicle Crashes in Ontario. Safety Research Office, Safety Policy Branch & Research and Development Branch, Ministry of Transportation and Highways, Ontario, March, 1994. Warren, R.A., Simpson, H.M., Buhlman, M.A., Bourgeois, L.A. & Chattaway, L.D. The relationship of driver blood alcohol to injury severity. In the Proceedings of the American Association for Automotive Medicine Conference, San Francisco, October 1-3, 1981. White, J.M., Clardy, D.O., Graves, M.H., Kuo, M.C., Macdonald, B.J., Weirsema, S.J., Fitzpatrick, G. and Pearce, D.S. Testing for sedative-hypnotic drugs in the impaired driver: a survey of 72,000 arrests. Clinical Toxicology, 18 (8), 1981. Williams, A.F., Peat, M.A., Crouch, D.J., Wells, J.K. and Finkle, B.S. Drugs in fatally injured young male drivers. Public Health Reports, 100, 1985. Appendix A: Insurance Corporation of British Columbia and Manitoba Public Insurance Data & Transformations ICBC Using the crash as the unit of analysis, Insurance Corporation of British Columbia (ICBC) figures in Table A2 show that, for every fatal crash, there were between 110 and 115 injury crashes and between 473 and 566 PDO crashes, with overall averages of 114 and 504 respectively over the 6-year period between 1996 and 2001. Table A1: ICBC Crash Types N, 1996 - 2001 Crash Types 199619971998199920002001sumPDO201,500202,666203,245206,567206,292215,6401,235,910Injury50,68248,47246,55445,07743,55443,954278,293Fatal4264264184053953812,451Sum252,608251,564250,217252,049250,241259,9751,516,654 Table A2: ICBC Crash Fatality Ratios Crash Types199619971998199920002001allPDO473.00475.74486.23510.04522.26565.98504.25Injury118.97113.78111.37111.30110.26115.36113.54Fatal1.001.001.001.001.001.001.00 In order to move from the victim as the unit of analysis to the crash (and visa versa), one would need to estimate the number of persons killed in a fatal crash, the number of persons injured in a fatal crash, and so on. ICBC supplied actual counts of injuries and fatalities (the latter coming from their fatals data base compiled from combining coroner, police, and ICBC data sources). They have also used the following: 1.2 fatalities per fatal crash; 1.11 injuries per fatal crash; 1.51 vehicles per fatal crash; 1.44 injuries per injury crash; 1.82 vehicles per injury crash; and 1.52 vehicles per PDO crash (Fleming & Mercer, 1999, p. 13). When these numbers are applied to the crash counts in Table A1, the victim and vehicle counts are: Table A3: ICBC N victims/vehicles Crash Types199619971998199920002001sumPDO vehicles306,280308,052308,932313,982313,564327,7731,878,583Injuries79,28774,21968,43465,96163,51963,610 415,030 Fatalities5014764734454394272,761 Sum386,068382,747377,839380,388377,522391,8102,296,374 From Table A3, the relative proportions of injuries and PDO vehicles to fatalities are shown in Table A4. Table A4: ICBC Victim & PDO Vehicle to Fatality Ratios 199619971998199920002001allPDO611.34647.17653.13705.58714.27767.62680.40Injury158.26155.92144.68148.23144.69148.97150.32Fatal1.001.001.001.001.001.001.00 MPI Manitoba Public Insurance PDO vehicle and victim counts for the same time period are shown in Table A5, while the ratios of injuries and PDO vehicles to fatalities are shown in Table A6. Table A5: MPI N victims/vehicles 199619971998199920002001sumPDO94,839 89,899 78,799  83,991  98,465 93,624  539,617 Injuries 11,245  10,499  11,580  12,527  13,663  14,640  74,154 Fatalities 118  139  155  151  153  157  873 Sum 106,202  100,537  90,534  96,669  112,281  108,421  614,644  Table A6: MPI Victim & PDO Vehicle to Fatality Ratios 199619971998199920002001allPDO803.72646.76508.38556.23643.56596.33618.12injuries95.3075.5374.7182.9689.3093.2584.94fatalities1.001.001.001.001.001.001.00 2. That is: (number of fatalities divided by fatalities per fatal crash) = number of fatal crashes. ((number of fatal crashes multiplied by number of injuries per fatal crash) subtracted from (number of injuries)) divided by number of injuries per injury crash = number of injury crashes number of PDO vehicles divided by number of vehicles per PDO crash = number of PDO crashes  See, for example, Bierness (2001), Bierness (2002), and Smith et al. (2002).  Things that can affect counts across different agencies include differences in the time from crash to death, whether or not collisions not on public roads are included, differences in required reporting criteria, availability and willingness of agencies and governments to participate in data collection, differences in crash type definitions and so on.  It is likely that crash count frequencies are more accurately estimated in jurisdictions with only one vehicle insurer, as multiple insurers may have different reporting and classification criteria.  Beyond being a source for statements about alcohol or drug driver impairment and associated charges, police data also contain relatively objective statements concerning driver demographics, crash time, location, circumstances, and so on. These parameters can be used to develop profiles of crashes that are impaired-related and non-impaired related that can then be applied to crashes that do not contain direct statements regarding judgments on the extent of impairment. For example, if from police data (or other data sources such as coroner data) it were determined that impaired driving crashes were highly related to a combination of characteristics such as driver age, gender, time of crash, severity of crash, number of vehicles involved and so on, then data sources that do not contain direct statements on impairment could be used to estimate impairment frequencies. The application of impaired crash profiles from police data to the more numerous insurance data on crashes has been used successfully to track the effects of enforcement programs on impaired driving crashes (in Fleming & Mercer, 2001). Some of these estimation models that access large databases can be very sophisticated - see, for example, Subramanian (2002).  It should be kept in mind that even the coroner data probably underestimate the presence of alcohol due to issues of fluid collection after alcohol had been metabolized, as well as the infusion of fluids as a medical procedure (e.g., blood transfusions). Nonetheless, they are the best data available for estimating levels of the physical presence of alcohol.  The notion relatively constant needs to be qualified. Major changes in driver behaviour, such as substantial changes in restraint device use, or in traffic safety policy such as a reduction in speed limits, could very well change the overall relationship between fatality frequency and injury frequency, lowering relative fatality frequency and increasing relative injury and property damage only (PDO) frequencies. These global trends would have to be kept in mind and examined from time to time to adjust any estimation model for their effects.  The MAIS injury category numbers are grouped together to produce a simple injury category, as that is the way many crash databases are aggregated. In turn, this will allow the application of the aggregated data ratios to available Canadian crash data sources.  The issue of the level of driver BAC at which a crash can be classified as being caused by impairment is debatable. Some people may become physically impaired at a very low BAC, while others would not be impaired. Further, a person may be impaired but not be at-fault in a crash.  For purposes of this discussion paper, no attempt will be made to reconcile the differences between ratios built upon a BAC of 0.10% and those that might be built upon 0.08% if the data were available.  The position being taken here is that, while the proportion of impaired crashes seems to reduce with reduced levels of crash severity, there is no compelling reason to believe that the relative weighting of alcohol, alcohol-and-drugs, and drugs only changes within the impaired group at each severity level.  Mayhew et al. defined a motor vehicle fatality as alcohol-related if there was at least one drinking driver or drinking pedestrian in the fatal crash, as revealed from coroner and other data sources (e.g., police reports). While this would include relatively low BACs, the majority of persons with alcohol on board tend to be above the legal limit of 0.08%. For example, in their sample of fatally injured drivers in 1999, 66.9% had a zero BAC, 3% were between 0.01% and 0.049%, 3% were between 0.05% and 0.08% and 27.1% were above 0.08%. 2. That is: (number of fatalities divided by fatalities per fatal crash) = number of fatal crashes. ((number of fatal crashes multiplied by number of injuries per fatal crash) subtracted from (number of injuries)) divided by number of injuries per injury crash = number of injury crashes number of PDO vehicles divided by number of vehicles per PDO crash = number of PDO crashes  In the first quarter of 2002 the Canadian GNP was $1,106.4 billion according to the Statistics Canada Economic Indicators Table.  ICBC no longer publishes average crash type values.  PAGE 2 PAGE 18 PAGE ii PAGE 3 EMBED MSPhotoEd.3   /1?@K`auv,֩֩֩jgCJUmHnHujCJUmHnHuOJQJmHnHujoCJUmHnHu0JmHnHu jUCJ0JjUmHnHu5 5B*phCJ, mHnHujCJUmHnHu1    01?@$a$ =&/:e*zdz|z@ghKLM_`1l[ + u  :  $  $  $ ,-./01LMNghijkl; < = V W X Y Z ߽߭ߝߍ}jOCJUmHnHujCJUmHnHujWCJUmHnHujCJUmHnHuj_CJUmHnHu0JmHnHuOJQJmHnHu mHnHujUmHnHujCJUmHnHu1Z [ % & ' ) * + T U V o p q s t u j!CJUmHnHuj?!CJUmHnHuj CJUmHnHujG CJUmHnHujCJUmHnHujUmHnHu mHnHu0JmHnHuOJQJmHnHu2      4 5 6 8 9 : c d e ~  ߽߭ߝߍ}j$CJUmHnHuj'$CJUmHnHuj#CJUmHnHuj/#CJUmHnHuj"CJUmHnHu0JmHnHuOJQJmHnHu mHnHujUmHnHuj7"CJUmHnHu1    2 3 4 6 7 8 Y Z [ t u v x y z       ! " # < j'CJUmHnHuj&CJUmHnHuj&CJUmHnHuj%CJUmHnHuj%CJUmHnHujUmHnHu mHnHu0JmHnHuOJQJmHnHu2 8 z  B v eGll & $  $  $  $ < = > @ A B U V W p q r t u v DEF_`acd߽߭ߝߍ}j)CJUmHnHuj{)CJUmHnHuj(CJUmHnHuj(CJUmHnHuj(CJUmHnHu0JmHnHuOJQJmHnHu mHnHujUmHnHuj'CJUmHnHu1deopq&'(ABCEFGKLMfghjklopq jUjc,CJUmHnHuj+CJUmHnHujk+CJUmHnHuj*CJUmHnHujs*CJUmHnHujUmHnHu mHnHu0JmHnHuOJQJmHnHu5opqLMNghijkl  }jK/CJUmHnHuj.CJUmHnHujS.CJUmHnHuj-CJUmHnHuj[-CJUmHnHu0JmHnHuOJQJmHnHuj,CJUmHnHu mHnHujUmHnHu1    ABC\]^_`a    CDE^_`bcdj1CJUmHnHuj;1CJUmHnHuj0CJUmHnHujC0CJUmHnHuj/CJUmHnHu0JmHnHuOJQJmHnHujUmHnHu mHnHu4 a d s!r=O0r&& $     RSTmnoqrs !QRSlmnpqr߽߭ߝߍ}j4CJUmHnHuj#4CJUmHnHuj3CJUmHnHuj+3CJUmHnHuj2CJUmHnHu0JmHnHuOJQJmHnHu mHnHujUmHnHuj32CJUmHnHu1789;<=./0IJKMNOtuvj 7CJUmHnHuj6CJUmHnHuj6CJUmHnHuj5CJUmHnHuj5CJUmHnHujUmHnHu mHnHu0JmHnHuOJQJmHnHu2*+,./0QRSlmnpqr--..#.6.R.W.// 00&0506011<1O1V1߽߭ߝ՘5>* 5B*ph B*ph jUj8CJUmHnHuj8CJUmHnHuj8CJUmHnHu0JmHnHuOJQJmHnHu mHnHujUmHnHuj7CJUmHnHu5   ##%%c*d*,,,,-,-----. $$Ifa$ $$Ifa$....#.+.5.6.7.//d^d\\\$If$$If.\ ~vvv04 a $$Ifa$ ////// 0 00&0WQ$If$$If.\$ 04 a $$Ifa$ $$Ifa$ &0-04050600011'1d^\\ZQ $$Ifa$ 0$$If.\$ 04 a $$Ifa$ '111;1<1O1V1]1d1e1y111111|||||| $$Ifa$$Ifj$$If.\~ , 4 a $$Ifa$111111111111111$Ifj$$If.\~ , 4 a $$Ifa$V111223"3(3}3~3]4^4444444z8888::\:l:::;;;CC{EE@MAMM$N%N&NNNOO(O)O/O0O1O2O?O@OGOHOIOJOKOLOlOOOZZ"p#pt$tnoŒ"ōW5mHnHu j5U>* j0JU 5B*ph5 B*phV1122223333"3(303:3xr$Ifj$$If.\~ 4 TbT4 a $$Ifa$ $$Ifa$ 0 :3;3F3M3T3[3\3g3m3t3|3}3~355} 0 $$Ifa$$Ifj$$If.\~ 4 TbT4 a55555@6A667U7V78888888 $$Ifa$$If 0 & F 888888899 9mcZZZmTZZ$If $$Ifa$  0$If$$If\ m & 04 a 9'9(9;9F9N9U9V9W99d^dXV 0$If$$If\ m & 04 a $$Ifa$ 99:: :::::3:@:O:n8eee $$Ifa$}$$Ifbrjr6$&bbb4 a $$Ifa$$If O:\:k:l::::::::::;xXxD}$$If1rjr6$&bbb4 a $$Ifa$ ;;;;;";#;<<Z>[>AAxrrpnhhhnnn^ 0}$$If1rjr6$&bbb4 a $$Ifa$ AAAAA&C'CEEFFIIII JJJ7J8JMMMMN $ 0$Ifa$$IfNNN NN%N}}} $ 0$Ifa$ 0$Ifk$$Ifl40Z<N04 la%N&N7N;N?NCNnxdWWW $ 0$Ifa$ 0$If$$Ifl4\Z<G04 laCNDNENNNNnljdW $ 0$Ifa$$If$$Ifl4\Z<G04 laNNNNN $ 0$Ifa$$Ifm$$Ifl40Zj04 laNNNNNNNNNNO OO<xkkXxkkxkk $ 0$Ifa$ 0$If|$$IflFZj 0    4 la OOO1OIOJOKO|ooo $ 0$Ifa$$If|$$IflFZj 0    4 laKOLOUO\OlO $ 0$Ifa$$IfX$$Ifl4Z04 lalOmOsOwO{O|OOOOOOOO<xkkTxkkxkk $ 0$Ifa$ 0$If|$$IflFZj 0    4 la OOOOOOOOOPPRRT|oomkmmmmm $ 0$Ifa$$If|$$IflFZj 0    4 la RmTnTRVSVVVZZ[[[\\\\^^WaXaLcMc%d&ddddddgggJiKiMiwixijjjjjllZm[m%p&p"r#ruuvvvvvxx 0x||~~qr 78 !"LM89  0QRS $$Ifa$$If 0 Œ֌ZTKKKBZ $$Ifa$ $$Ifa$$If$$Ifr # 82 04 a#$+6K$$Ifr # 82 04 a $$Ifa$$If6<CDELUZ`QtK$If$$Ifr # 82 04 a $$Ifa$`abirw|}~QpKQl$If$$Ifr # 82 04 a $$Ifa$~K$$Ifr # 82 04 a $$Ifa$$IfƍǍ֍ލQK$If$$Ifr # 82 04 a $$Ifa$ QpKB $$Ifa$$If$$Ifr # 82 04 a $$Ifa$/05ZTTTTRII $$Ifa$ 0$$Ifr # 82 04 aWa/SŐ͐4Xđʑ ! *+,#bȢۢ͢#9>Ѩר05ˮѮ@Fد߯7=,2.4X^msɹϹƻ0J j0JU B*ph5 5B*ph>*Z5:?DINR $$Ifa$RS`! $$Ifa$$$Ifhִb d4t"20    4 a`gnu| $$Ifa$! $$Ifa$$$If1ִb d4t"20    4 aŐ̐ $$Ifa$̐͐ؐ! $$Ifa$$$If1ִb d4t"20    4 aؐݐ $$Ifa$! 0$$If1ִb d4t"20    4 a45:?DINSW $$Ifa$ $$Ifa$ 0 WXelszHB9999 $$Ifa$$If$$Ifhִ "&    4 a?$$Ifִ "&    4 a $$Ifa$đʑˑ֑?$$Ifִ "&    4 a $$Ifa$֑ۑ $$Ifa$PHA;;;; 0 0$$Ifִ "&    4 aPQJKM|}NOkl "#%1> $$Ifa$$If 0>Kabv>8/ $$Ifa$$If$$Ifֈ$  B$B 804 a $$Ifa$54$$Ifֈ$  B$B 804 a $$Ifa$ $$Ifa$΢ $$Ifa$$If & $If$$Ifֈ$  B$B 8 204 a>$$Ifֈ$  B$B 804 a $$Ifa$ &'( $$Ifa$$If()09AGHGA8888 $$Ifa$$If$$Ifֈ$  B$B 804 aHIJQY`f>|8$If$$Ifֈ$  B$B 804 a $$Ifa$fghipqy>8$If$$Ifֈ$  B$B 804 a $$Ifa$yz $$Ifa$&8 $If$$Ifֈ$  B$B 8 204 aǣȣУѣף $$Ifa$& $If$$Ifֈ$  B$B 8 204 a !"#>$$Ifֈ$  B$B 804 a $$Ifa$#$%efީߩBCEOPuv"#XYDEQ 0HLsxhlNO&) $HM >CU[sz%&ijw9CJUmHnHu mHnHu5mHnHu 5B*ph B*ph65 j0JU>*MQRJK`a()*,NOPdetuv & F 0 & F 0 0tu45-./;<rsijk & F 0 0 & F 0 & F 0p|$Ifj$$Ifb\@ HPX ^ 4 a $$Ifa$ $$Ifa$  $$Ifa$$Ifj$$If.\@ HPX ^ 4 a 4;BH $$Ifa$$Ifj$$If.\@ HPX ^ 4 aHI\bhn $$Ifa$$Ifj$$If.\@ HPX ^ 4 ano $$Ifa$$Ifj$$If.\@ HPX ^ 4 a=>TUzkY & F 0h$If^h & F 08$If$If 0h$If^h 0j$$If.\@ HPX ^ 4 a UrsSToiiiiii 0V$$Ifl$h%04 la 0$If 0h$If^h & F 0h$If^h & F 08$If  oTi$Ifj$$If1\ZM {x)4 ax $$Ifa$ $$Ifa$ d& / 0 #(-.7<AFGOU[`andh$Ifj$$If"\ZM {x)4 ax $$Ifa$ntz$&iklmzx & / 0j$$If1\ZM {x)4 ax $$Ifa$ ijk?IijMR   ( ? }    # 4 5      V W X u5E,H:HH_0>@JK4!\!W"CJjK 5CJUmHnHu5 j0JU>* 5B*ph B*ph5mHnHujd5CJUmHnHuL|$Ifj$$If1\M x)4 a $$Ifa$ $$Ifa$  '>OPXm $$Ifa$$Ifj$$If"\M x)4 a*+23XY'(^ 0j$$If"\M x)4 a(STUijghi & F & F^ & F rsulmEF  $$Ifa$$If 0    ' / 6 7 P dZQQQdK$If $$Ifa$  0$If$$If\ m & 04 a $$Ifa$P Y a h i |      d^dX 0$If$$If\ m & 04 a $$Ifa$  ` a b        j}$$Ifbrjr6$&bbb4 a $$Ifa$$If 0    ( > ? T a p }    x8xX}$$If1rjr6$&bbb4 a $$Ifa$         # 3 4 5 6 xDxrr 0}$$If1rjr6$&bbb4 a $$Ifa$ 6    V X Y Z 012uv| $$Ifa$$If K&/ 0{rrrr $$Ifa$$If}$$If.r2 :B%P T4 a (5DE]jw4{{rrrr<{rrrr $$Ifa$$If}$$If1r2 :B%P T4 a ,-<{{rrrrllj{ 0 $$Ifa$$If}$$If1r2 :B%P T4 a -3?CGHZqxriiii $$Ifa$$If}$$If.r2 :B%P T4 a $$Ifa$  -:G${{rrrr8{rrrr $$Ifa$$If}$$If1r2 :B%P T4 a GH[\kx4{{rrrrpjh{ 0 $$Ifa$$If}$$If1r2 :B%P T4 a 5H^xriiii $$Ifa$$If}$$If.r2 :B%P T4 a $$Ifa$ ^_tu${{rrrr4{rrrr $$Ifa$$If}$$If1r2 :B%P T4 a  #0=>?@KL${{rrrrlljl 0 $$Ifa$$If}$$If1r2 :B%P T4 a LM! # . /   !!""d#e#$$%% 00^`0 00^`0 0W""" ####$t$$$$%%'*','.'@(h())))Q*s**+++++,@,)-Q-.<../V/020e00o11V2p2r2s22333\44G5^5566666W7r7_8}88869^9U:s:::;;<<<< =5===>>C??A@|@@A >*OJQJOJQJ0J>*H*>*]%&&6'7'''(())**++++,,--..//J0 00^`0 0 00^`0J0K00022~22b3c3J4K444k5l5A6B666778899 00^`0 0 00^`09::;;J<K<<<====>>B?C?|@}@%A&A|B}BbC 0^`0 0^`0 00^`0 0 00^`0ACAABCCCCCDDDD"E?EAEDEdEEEFFFFFgGG@HgHiHkHII!I#I~IIXJrJtJuJJK"LvLMMMM8NFN]NyNzNN=OeOO%P1QPQQRRSOTdTfThTTTTTT#VVWW X XXXXZZg[[[[+\3\d\ B*ph 5B*phCJ0J>*H*>*[bCcCCCDDLEMEEECFDFFFGG|H}H+I,I!J#J}J~JKK 00^`0 0 00^`0KLLHNINNNOOPPbQcQRR*S+SSStTuTTT#V$V 0^`0 00^`0 0 00^`0$V&W'WXXXX!Y"Y#Y(Y)YlZmZnZZZZZZZZZ $$Ifa$ $$Ifa$$If 00^`0 0 00^`0ZZZ$$IfhִK @$QRRQRR0    4 a $$Ifa$ZZZZZZZZ[ $$Ifa$[[[! $$Ifa$$$If"ִK @$QRRQRR0    4 a[[[$[+[2[9[A[ $$Ifa$A[B[H[! $$Ifa$$$If"ִK @$QRRQRR0    4 aH[L[P[T[X[\[`[f[ $$Ifa$f[g[k[! $$Ifa$$$If"ִK @$QRRQRR0    4 ak[s[{[[[[[[ $$Ifa$[[[! 0$$If1ִK @$QRRQRR0    4 a[[[[[[[[[[[ $$Ifa$ $$Ifa$$If 0 [[\! $$Ifa$$$Ifhִ V6$88888880    4 a\\\\\$\+\2\ $$Ifa$2\3\:\! $$Ifa$$$If1ִ V6$88888880    4 a:\A\H\O\V\]\d\k\ $$Ifa$k\l\r\! $$Ifa$$$If1ִ V6$88888880    4 ad\l\\\__`"`W```````oaaaaab&b,bc5ckcuccccccBd{ddddee2e7effffg glhmh5i6inn}o~oqqrrssttuuuupyqyyy*z+z-z.z4z5z6z7z8z:z0JmHnHu0J j0JUjOJQJU jU>* j0JU5 B*ph 5B*phOr\w\|\\\\\\ $$Ifa$\\\! 0$$If1ִ V6$88888880    4 a\\____________ $$Ifa$ $$Ifa$$If 0 ___!  $$Ifa$$$Ifhִ |gR=#0    4 a___````"` $$Ifa$"`#`,`! $$Ifa$$$If"ִ |gR=#0    4 a,`3`:`A`H`O`W``` $$Ifa$``a`l`! $$Ifa$$$If"ִ |gR=#0    4 al`p`t`x`|```` $$Ifa$```! $$Ifa$$$If"ִ |gR=#0    4 a```````` $$Ifa$```! 0$$If"ִ |gR=#0    4 a`6a7aoapauazaaaaaa $$Ifa$ $$Ifa$$If 0 aaa! $$Ifa$$$Ifhִb d4t"20    4 aaaaaaaaa $$Ifa$aaa! $$Ifa$$$If1ִb d4t"20    4 aaaaaaaab $$Ifa$bbb! $$Ifa$$$If1ִb d4t"20    4 ab bbbb!b&b+b $$Ifa$+b,b-b! 0$$If1ִb d4t"20    4 a-b.b2b3bbbcccc!c&c+c0c4c $$Ifa$ $$Ifa$ 0 00^`0 04c5c9cAcIcQcZcH????9$If $$Ifa$$$IfִN N"-?@@@@@@    4 aZccckcucvccc6$$$IfִN N"-?@@@@@@    4 a $$Ifa$ $$Ifa$cccccccc?$$IfִN N"-?@@@@@@    4 a $$Ifa$ccccccccc $$Ifa$cccddd#d-dH(?????? $$Ifa$$$IfִN N"-?@@@@@@    4 a-d7dAdBdCdDd{d?886 0$$IfִN N"-?@@@@@@    4 a $$Ifa${d|dddddddd $$Ifa$ $$Ifa$dddddddHB9999 $$Ifa$$If$$IfhִD "p"poooopo    4 adddddddd?$$IfִD "p"poooopo    4 a $$Ifa$ddddee ee?$$IfִD "p"poooopo    4 a $$Ifa$eee#e(e-e2e7e $$Ifa$7e8e9e:eeHA2( t^t` 00^`0 0$$IfִD "p"poooopo    4 aee]f^ffglh5inn}oqrsstuwWxXxyypyy)z*z,z & F^ t^t`,z-z8z9z:zFzGzHzTzUzVzazbzczdzzz{z|z}z 00^`0h]h&`:z;zAzBzDzEzFzHzIzOzPzRzSzTzVzWz]z^z_z`zazdzezvzwzxzyz}z jsCJUjh A UV jU0JmHnHu0J j0JU& 00 P/ =!"#$%4 0000 P/ =!"#$%. 00P/ =!"#$%oDd bX<  C Ab* Su ldf|Dn* Su ldf|PNG  IHDR=[WgAMAPLTEJJRRR!RZ!Z!Z)Z)c1c1c1k1k9kBk9sBsBsJsJ{R{R{RRZcZcckksss{{ƌƌƔƔΔƔΜΜΥΥֵֵ֭֭޵ֵ޽޽ pHYsodKIDATx\K8^ P`2Ek|MtJ4b)ǖ9YMr5Y?AdZije2nYk1ccgGXL֤uʨ3n)+ڛ_-qwז>jJKd;+zK}|_7 G_ǯxWaMϋ-Wڷc䚘pkhJ֕J{7rZ`EҶ[lV~,NO}d[i<} N֧UBjpkϢzBa'mH+:õx$_(+rFaPLF@1SH;ёuقʲ$iE/-+%tJf[ -znMY7;92JgT:OLVʭ(ZD&D䁉]b*V|UyoHlNyܟHDlE5خ45PZhNJ%codM--dJ"[WQFwP.%k E2^zDǗhZ~Ҍ78T-l"olA:Q@h4ssܷd{fc=ʱV "4YH-/`ľnKeck4Q[KvKdx`Y~F07őn-sj+B~-Gr"bjbVWʹZ57'ǫmЉ,h YdM&Z6 -KUB43:[+xXw(n%i7)b^ 3nb Agӳ'vRjsz1oFV5,ӄ5׸E=|uN : $n\X|ֹ#-NQkuZ/0X|vdNW|"c;a|t)yUt"T NMZ01VpY' <ָ_7PƂDdc[O |j 0i v9Rs^aa M?2w+[%n"vdD۬N8ZzUHGoAD"dSAVBPlioZ~kaP0Yv&[S᫛` uTaƪݪI`.y}tCSx dFj$VОO`܉SDHgӥcy/>d_ŚB8ŠQFڥ"e-1Uw[ώ]uX?c$3:l"Nds+) |h3Б&>5/t9>nSIgMg WFqbGYD8Ц|FD}p`5qEH }׹9V 4/Vy<u'0>uq|=Q֭3x^)UC%M5CyY|$+ *<%9"0dk@=1&7'ɭB`8, LbWNMs?g͹VG\vmOmA^(.lQ~ٶsLSP"̔)돚M~yȘlmAX1Tƌ}^ 1dk SUoύR"8Gc$!cP Gr[\_I Z6`'nۥ~Y-Nސ1\3詄2|nɡ>OɃ*ZqH$ƪb7#^2|HQTY|v$bqISᤶLO/e$qH t% xGeU2n&i\|2-^qk'Y`+X;X%pNڊxz8F*QY3tG̬{⊰{A{-i2.ԅ$u陕̃Zx3jI| 3DTrUv0], `э?sof־x1jfZuxzsƧ\ذpZWhSTKyi߈ݣ>CczF~jan3\64fj Y3[3uOl$k-Jt}J9\+֪O||ݮ*nM܃"{8daȀ<=$T}!c5GTݴs-`F1WrT2tm( c"6_jd#eVdKC/'{w,D;Bs"e-@=ݝ !谵Grd=sQ^굨zvYpꉵ|AO"na lqZA]RjƆ3.n&( PGm{ɍ6' k Zѝ/mY^qu  7R "<լ#Eٚl0 R6pKnw`y2_θ!+شd̗F;4r@ﴃꀛ~:5]3 Ļ(17(kh-9Ɵ8w]g JA;Gbʗlq8- /<~ +Ư^6b夽Oy964NKOzeJ ȕcynx.%nzEIg6 ᵇe$- Yy5<Չ:]4A#a/4+5A?^hmw'qdý1)rwMB#upM+]S]:I8Hh".N!eFP~Ojb iPT.!)ʐC{_I3w!86[b-J'5GO` 9mqĚ`>I);R0!p_uB&RPGh.zc Q& (ӥ&aٰY>03ClԄMiic`?7R菭@1'Ktx`F &hE| h:y]aZ bV7ʫnx*sWeU'WrZt,E d*:(Zz 4hF8AĀxR6l︘*N6EMq${0 :䨭g1! >aV.Y<P _'uoo+|>55AJǢ}RSȆ˟/&bv vM {QLr%sAԩəkkrVҦ HD?῔YY0bR{"pC ~H$6E7R_pb8FɂDlY14 #} `<#u2?aAgLV-L z٣TR#! pNt/ NGq=EbnBfIBg*xF)Ax5N x۫aYbY7Xtwͺ6 -w *M%-ZNj0KEg ك(KXF6z)YTVɂaQJ~? g|ʺ^Ǭ|^>QALl;`>`Xg\ffo"ɺYz+kFnؒ::C7ϒM6S{>O.(3Juw3K71s56z>1_giuE_\gS"/,꡷opzͥH@yN+Zr8D>>pkB"<ȼ fB+0=R6q-ߑqC5Y0<)oa @<儅P%K`)%Xi -ʳ0?" **]cx%7o\' R:3 Cb~w?DL<?" ;Ԧ4p yoV.f5|, /NڣSI#~6E!e{<4l#NQrg:7ȂcUDk`5,8b[]I?`_aO ܽX} di[1 /=aק9M;ք?bV%+ţņ:' J_Sb;?#1 Ry<{)Y)a-'iA@rRa0 R?#Y-for ;'d9Z)c+cտB wdK\ω{d9OK_?Q^r?Kv߈RKWYI\p'^{!;>5@d.ln} o5BP6ݰ罃z"YnvᝃA7C??CɅD#ώ_J2InD)IENDB`|DyK  _Toc26676407|DyK  _Toc26676408|DyK  _Toc26676409|DyK  _Toc26676410|DyK  _Toc26676411|DyK  _Toc26676412|DyK  _Toc26676413|DyK  _Toc26676414|DyK  _Toc26676415|DyK  _Toc26676416|DyK  _Toc26676417|DyK  _Toc26676418|DyK  _Toc26676419|DyK  _Toc26676420|DyK  _Toc26676421|DyK  _Toc26676422|DyK  _Toc26676423|DyK  _Toc26676424|DyK  _Toc26676425|DyK  _Toc26676426|DyK  _Toc26676427|DyK  _Toc26676428|DyK  _Toc26676429|DyK  _Toc26676430|DyK  _Toc26676431|DyK  _Toc26676432|DyK  _Toc26676433|DyK  _Toc26676434|DyK  _Toc26676435|DyK  _Toc26676436|DyK  _Toc26676437|DyK  _Toc26676438|DyK  _Toc26676439|DyK  _Toc26676440|DyK  _Toc26676441|DyK  _Toc26676442|DyK  _Toc20193817|DyK  _Toc20193818|DyK  _Toc20193819|DyK  _Toc20193820|DyK  _Toc20193821|DyK  _Toc20193822|DyK  _Toc20193823|DyK  _Toc20193824|DyK  _Toc20193825|DyK  _Toc20193826|DyK  _Toc20193827|DyK  _Toc20193828|DyK  _Toc20193829|DyK  _Toc20193830|DyK  _Toc20193831|DyK  _Toc20193832|DyK  _Toc20193833|DyK  _Toc20193834|DyK  _Toc20193835|DyK  _Toc20193836|DyK  _Toc20193837|DyK  _Toc20193838|DyK  _Toc20193839|DyK  _Toc20193840|DyK  _Toc20193841|DyK  _Toc20193842sD7J&*   9&" WMFC,$ ,LlpW- EMFL @    !" !" !   0 &% '% % % " !% %   % % " !% %   % % " v!% %   Rp Arial$0w" 1w wh"1w" !0" L  -X0x\ Araldv% ( Rp ArialL  -X0x\ Araldv\w K\D(#0 Kdv% % % " v!% %   Rp Arial       -X0x\ Aral[dv% ( Rp Arial  -X0x\ Aral[dv[awhwpw[\w L\D(#0 Ldv% % % " v!% %   RpArial    Dt'g0F -X0x\Aral[dv% ( RpArial'g0F -X0x\Aral[dv[awhwpw[\wq \Dt,(#0q dv% % % " v!% %   Rp Arial 7Tdh0 D $Yl0h0<3<3 _20 @20"0k& -X0\ Araldv% % % % % % % % % % " v!% %   Rp Ariallw ! -X0\ Araldv% % % % % % % % % % " v!% % %   RpArial <w!w!%  w! %xw 8h'g0: -X0l\Araldv% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % " !% % %   u& % ( '% (     +p% % % " !% % %   u% ( % (   V0JHJHHJJH& % % % % " !% % %   JD U9mJ   6J6J6J6J6J6J6Jn6nJ[6[JH6H% % % " !% % %   uU9m&% ( JH  6H66J6JH% % % " !% % %   u% % % " !% % %   p% % % " !% % %   JH% % % " !% % %   JF& % ( '%   V0V0YlYllYYlV0+]+]]++]F(GDIC% %   _  ^t$0T ( % % Ld__T0!??%    % ( % %  FGDIC% V0____% F(GDIC% %   S  ^t$0T &" WMFC L( % % LdYSkYSU!??%    % ( % %  FGDIC% V0YSlYSSlYlYS% F(GDIC% %   O  ^t$0T ( % % Ld+O~\+OT!??%    % ( % %  FGDIC% V0+O]+OO]+]+O% F(GDIC% %   N  ^t$0T ( 33333333% % LdN^NT!??%    % ( % %  FGDIC% V0N_NN__N% F(GDIC% %   K  ^t$0T ( % % LdYKRYKU!??%    % ( % %  FGDIC% V0YKSYKKSYSYK% F(GDIC% %   J  ^t$0T ( ffffffff% % Ld+J~N+JT!??%    % ( % %  FGDIC% V0+JO+JJO+O+J% '333% (  333V0HNHHNNHV0YHKYHHKYKYHV0+HJ+HHJ+J+H% % % " !% % %   333JH% % % " !% % %   333u&% ( JH  6J% F6JF6JF6JF6JF6JF6JF6JF6JFn6JnF[6J[FH6JHJ6% J 6J 6 6 6% % % % % % " !% % %   333 +   T%AAL|% Impaired in Crash Type      % % % % " !% % %   u% % % " !% % %   u% % % % " !% % %   u  TX-? AA-LP0% T`&?AA&LT10% T`&?AA&LT20% T`&?AA&LT30% T`&?AA&LT40% T`&?AA&LT50% T`&?AA&LT60% T`&y?AA&yLT70% T`&f?uAA&fLT80% T`&S?bAA&SLT90% Td@?OAA@LT100% % % % " !% % %   u% % % % " !% % %   u  Tl&AALXfatal  Tpo&AAoLXinjury   T`Df&AADLTPDO % % % " !% % %   u% % % % % " !% % %   U,I  TX1CAAX1 L`Crash Type   % % % % " !% % %   u% ( '% (     +U9n% % % " !% % %   U9m% % % " !% % %   U9m&% ' % (    +^g  TYHkAAY LdNon-impaired   % % % " !% % %   U9m% % % " !% % %   U9m&% ( F(GDIC% %   ^  ^t$&" WMFC L0T ( % % LdR^ZfR^ !??%    % ( % %  FGDIC%    +R^[g%   Tx`YkAA`YL\Alcohol  % % % " !% % %   U9m% % % " !% % %   U9m&% ( F(GDIC% %   ^  ^t$0T ( ffffffff% % Ld^f^ !??%    % ( % %  FGDIC%    +^g%   T|YkAAYL\Alc+drug   % % % " !% % %   U9m% % % " !% % %   U9m&% ( '333% (  333  +^g  Td Y2kAA YLTDrug  % % % " !% % %   U9m% % % " !% % %   u% % % " !% % %   u& % ( % (     +p% % ( " !  u" v!  " !    % "   v  ''  ----'-- --'-- --'--  Arial- Arial---'--  Arial- Arial---'-- Arial-Arial---'--  Arial----------'--  Arial---------"System- '--- Arial- -- -- ------------------------------ '--- ,u- -  q--- '---  ,u--  $JHHJJH- --- '---  , DJ, DJJ JJJJJJnJn[J[HJH--- '---  ,u,um9U-  HJ HJHJ--- '---  ,u--- '---  ,p--- '---  ,HJ--- '---  ,FJ-  -  $$YllYYl$+]]++]-- -B( - - @ !0T_-  - - - -$___- -- -B( - - @ !USY-  - - - -$YSSlYlYS- -- -B( - - @ !TO+-  - - - -$+OO]+]+O- -- -B( 33333333- - @ !TN-  - - - -$NN__N- -- -B( - - @ !UKY-  - - - -$YKKSYSYK- -- -B( ffffffff- - @ !TJ+-  - - - -$+JJO+O+J- 333-   333$HHNNH$YHHKYKYH$+HHJ+J+H--- '-- -  333,HJ--- '-- -  333,u-  HJ J-FJFJFJFJFJFJFJFJnFnJ[F[JHFHJJ- JJ   ------ '-- -  333,+   +2 % Impaired in Crash Type      ---- '-- -  ,u--- '-- -  ,u---- '-- -  ,u  2 -0% 2 &10% 2 &20% 2 &30% 2 &40% 2 &50% 2 &60% 2 y&70% 2 f&80% 2 S&90% 2 @100% --- '-- -  ,u---- '-- -  ,u 2 fatalr  2 oinjury   2 DPDO --- '-- -  ,u----- '-- -  ,I,U 2 1X Crash Type   ---- '-- -  ,u- -    o:U--- '-- - ,m9U--- '-- - ,m9U- -   h^ 2 Y Non-impaired   --- '-- -  ,m9U--- '-- -  ,m9U-  -- -B( - - @ ! ^R-  - - - - h\^R- 2 Y`Alcoholg  --- '-- -  ,m9U--- '-- -  ,m9U-  -- -B( ffffffff- - @ ! ^-  - - - - h^- 2 YAlc+drug   --- '-- -  ,m9U--- '-- -  ,m9U-  333-   333 h ^  2 Y Drug  --- '-- -  ,m9U--- '-- -  ,u--- '-- -  ,u-  -   q- - ' ,u' '  - '^D>K.*_ N/&" WMFCQ <<=lX6 EMF<= @    !" !" !   0 &% '% % % " !% %   % % " !% %   Rp ArialP @w0(_20 @20"0k \X) -X0\ Araldv% Rp Arial 3xwо IȬqS0PȬ4 ؿ Arial{0 2 2\w I\4PȬ#0 Idv% % % % % % Rp Arial+xwإS I , qS0+< ]  Arial.wGw}`\wS I\<+ #0S Idv% Rp Arialb N0TM0|xgrvJ0|xg<r -X0  \ Araldv% % % % % % % RpArial,=,0009,xw 7(X'g0* -X0\p\Araldv% RpArial(xwо ZqS0Z4 (ؿArial{0 2 2\w \4Z#0 dv% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % " !% % %   & % ( '% (     +% % % " !% % %   % ( % (   V0gMQgMMQgQgM& % % % % " !% % %   gIUMg   6g6g6g6gM6M% % % " !% % %   MgM  6M6Q6gQ6gM% % % " !% % %   % % % " !% % %   % % % " !% % %   gMQ% % % " !% % %   gKQ&% ( ' %   V0o0Po00QoQo0V0:P:Q:Q:F(GDIC% %   +  ^t$0T ( % % Ldo+/o+Q!??%    % ( % %  FGDIC% V0o+0o++0o0o+% F(GDIC% %   y  ^t$0T ( % % Ld:y:yQ!??%    % ( % %  FGDIC% V0:y:yy::y% F(GDIC% %   )  3f^t$0T ( f3ffffffff% % Ldo)*o)Q!??%    % ( % %  FGDIC% V0o)+o))+o+o)% F(GDIC% %   s  3f^t$0T ( f3% % Ld:sx:sQ!??%    % ( % %  FGDIC% V0:sy:ssy:y:s% '% (  V0o()o(()o)o(V0:qs:qqs:s:q% % % " !% % %   gMQ% % % " !% % %   & % ( gM  6gQ% bQ6gQb6&^WMFC<<=gb6gb6gb6gbM6gMgQ6Q% gV6gQ2V62QV6QV6Q% % % % % % " !% % %    .  T (AA L# Impaired Victims by Crash Type         % % % % " !% % %   % % % " !% % %   % % % % " !% % %     TTAIEYAAAILP-BTx"U%AA"L\500,000 TUAA L`1,000,000eTUAA L`1,500,0001TyUAAy L`2,000,000ATEUUAAE L`2,500,000l% % % " !% % %   % % % % " !% % %     Tl`rAA`LXfatal  Tp`rAA`LXinjury   T`R`trAAR`LTPDO % % % " !% % %   % % % % % " !% % %   hx  Tk}AAk} L`Crash Type   % % % % " !% % %   % ( ' % (     +M% % % " !% % %   M% % % " !% % %   M&% '% (    +  T\AA LdNon-impaired   % % % " !% % %   M% % % " !% % %   M& % ( F(GDIC% %     ^t$0T ( % % Ldfnf !??%    % ( % %  FGDIC%    +fo%   TxtAAtL\Alcohol  % % % " !% % %   M% % % " !% % %   M&% ( F(GDIC% %     3f^t $0T ( f3ffffffff% % Ld !??%   % ( % %  FGDIC%    +%   T| AAL\Alc+drug   % % % " !% % %   M% % % " !% % %   M& % ( '% (   + Td!FAA!LTDrug  % % % " !% % %   M% % % " !% % %   % % % " !% % %   &% ( % (     +% % ( " !  " !  " !    % "     ''  ----'-- --'--  Arial- Arial------ Arial- Arial-------Arial-Arial- -- ---- -------------- --- ---------"System- '--- ,- -  --- '---  ,--  $gMMQgQgM- --- '---  ,UIg,UIgg gggMgM--- '---  ,,MMg MQQgMg--- '---  ,--- '---  ,--- '---  ,QMg--- '---  ,QKg-  -  $o00QoQo0$:Q:Q:-- -B( - - @ !Q+o-  - - - -$o++0o0o+- -- -B( - - @ !Qy:-  - - - -$:yy::y- -- 3f-B( f3ffffffff- - @ !Q)o-  - - - -$o))+o+o)- -- 3f-B( f3- - @ !Qs:-  - - - -$:ssy:y:s- -   $o(()o)o($:qqs:s:q--- '-- -  ,QMg--- '-- -  ,-  Mg Qg- QbQgbgbgbgbgMbMgQgQ-VgQgV2Q2VQVQ------ '-- -  ,.  72  # Impaired Victims by Crash Type         ---- '-- -  ,--- '-- -  ,- --- '- - -  ,  2 IA-72 "500,000g2  1,000,0002  1,500,0002 y 2,000,0002 E 2,500,000--- '- - -  ,---- '-- -  , 2 `fatal0  2 `injury   2 `RPDO --- '-- -  ,----- '-- -  ,xh 2 }k Crash Type   ---- '-- -  ,- -    N--- '-- - ,M--- '-- - ,M- -    2  Non-impaired   --- '-- -  ,M--- '-- -  ,M-  -- -B( - - @ ! f-  - - - - pf- 2 tAlcoholg  --- '-- -  ,M--- '-- -  ,M-  -- 3f-B( f3ffffffff- - @ ! -  - - - - - 2 Alc+drug   --- '-- -  ,M--- '-- -  ,M-  -    2 !Drug  --- '-- -  ,M--- '-- -  ,--- '-- -  ,-  -   - - ' ,' '  - 'gDM.,| 3&" WMFCI #Cl[-7 EMFC @    !" !" !   0 &% '% % % " !% %   % % " !% %   Rp Arial3 x xx @) S Sj f _20 @20"00k \  -X0\ Araldv% Rp Arial.oYh0x( 0  @xdp' 0k PM`k.@T 90@Ari\w{ \p#0{ dv% % % % % % Rp Arial.oYh0x( 0  @xdp' 0k PM`k.oY5l\w ]\+p#0 ]dv% Rp Arialb N0TM0|xgrvJ0|xg<r -X0  \ Araldv% % % % % % % RpArialDt'g0F -X0x\Araldv% RpArial*kVh0x( 0  @xdp' 0k PM`k*@T90@Ari\wA e\ -p#0A edv% % % % % % % Rp  Small Fonts i 0009 xw M!,g0 Ҭ -X0<P\ Smll Fontsdv% Rp  Arial.oYh0x( 0  @xdp' 0k PM`k.88@T 90@8Ari\w \.p#0 dv% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % " !% % %   & % ( '% (     +% % % " !% % %   '% ( % (   V0KSKKSSK& % % % % " !% % %   GW6   66666666h6hK6K% % % " !% % %   &% ( K  6K6S6S6K% % % " !% % %   % % % " !% % %   % % % " !% % %   KS% % % " !% % %   IS& % ( '% (    V0JRJJSSJV0l.Rl..SlSl.V06LcR6LcLcS6S6LF(GDIC% %   5  ^t$0T ( ffffffff% % Ld5R5-!??%    % ( % %  FGDIC% V05R55SS5% F(GDIC% %   0  ^t$0T ( % % Ld0R0-#!??%    % ( % %  FGDIC% V00R00SS0% F(GDIC% %   1  ^t$0T ( ffffffff% % Ldc1Rc1-"!??%    % ( % %  FGDIC% V&" WMFC C0c1Rc11ScSc1% '% (  V0o)Ro)o)SSoV0$R$$SS$V01R11SS1% % % " !% % %   KS% % % " !% % %   &% ( K  6S% ~S6S~666~6~6~6~6~6~6~h6h~K6KS6S% X6SKX6KSX6SX6S% % % % % % " !% % %   .   T(AALImpaired Crash Cost Models      % % % % " !% % %   % % % " !% % %   % % % % " !% % %     TXmLxYAAmLLP$0T./x<AA./Lh$1,000,000,000T.xAA.Lh$2,000,000,000T.xAA.Lh$3,000,000,000T.xAA.Lh$4,000,000,000T.xAA.Lh$5,000,000,000T.xAA.Lh$6,000,000,000T.xAA.Lh$7,000,000,000T.axnAA.aLh$8,000,000,000T.DxQAA.DLh$9,000,000,000% % % " !% % %   % % % % " !% % %     TlbtAAbLXfatal  TbtAAb Ldinjury-only     T`ibtAAibLTPDO % % % " !% % %   % % % % % " !% % %   z  TAA L`Crash Type   % % % % " !% % %   % % % % % % " !% % %   ,  Rp Arial%|8 w\wn \pw#0n dv% Tx%AAL\Dollars % ( % % % % " !% % %   % % ( '% (     +% % % " !% % %   % % % " !% % %   &% ' % (     + TKAALtReal Dollar Estimate      % % % " !% % %   % % % " !% % %   &% ( F(GDIC% %     ^t$0T ( % % LdU]U !??%    % ( % %  FGDIC%    +U^%   TcAAcLxDeferred Future Income        % % % " !% % %   % % % " !% % %   &% ( '% (    + )  T.AA.&WMFCCLpWillingness to Pay   % % % " !% % %   % % % " !% % %   % % % " !% % %   & % ( % (     +% % ( " !  " !  " !    % "     ''  ----'-- --'--  Arial- Arial------ Arial- Arial-------Arial-Arial- - - ---- Small Fonts-  Arial- - - - --- -------- - - ------- - - -- --- -------- - - --"System- '- -- ,- -  --- '- --  ,--  $KKSSK- --- '- --  ,WG,WG6 6hhKK--- '- --  ,,- K KSSK--- '- -- ,--- '- -- ,--- '- -- ,SK--- '- -- ,SI- -  $JJSSJ$l..SlSl.$6LcLcS6S6L-- -B( ffffffff-- @ !-5- -- --$55SS5--- -B( -- @ !#-0- -- --$00SS0--- -B( ffffffff-- @ !"-1c- -- --$c11ScSc1-- $o)o)SSo$$$SS$$11SS1--- '- --  ,SK--- '- --  ,- K S- S~S6~6~~~~~~h~hK~KSS-XSXKSKXSXS------ '--- ,.  .2 Impaired Crash Cost Models      ---- '--- ,--- '--- ,- --- '- -- ,  2 Lm$02 /.$1,000,000,0002 .$2,000,000,0002 .$3,000,000,0002 .$4,000,000,0002 .$5,000,000,0002 .$6,000,000,0002 .$7,000,000,0002 a.$8,000,000,0002 D.$9,000,000,000--- '- -- ,---- '--- , 2 bfatalr  2 b injury-only0     2 biPDO --- '--- ,----- '--- ,z 2  Crash Type   ---- '--- ,- - - --- '- -- ,, Arial- 2 Dollarsg -  - --- '- -- ,---   --- '-- - ,--- '-- - ,--   %2 Real Dollar Estimate      --- '--- ,--- '--- ,- -- -B( -- @ ! U- -- -- _U-(2 cDeferred Future Income        --- '---  ,--- '---  ,- -   *  "2 .Willingness to Pay   --- '-- - ,--- '-- - ,--- '-- - ,--   - -' ,' '  - 'rD  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrtuvwxyz{|}~      !"#$%&'()*+,36=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~Root Entry  FC5Data sWordDocument ObjectPoolCC_1100418408@96>@)C@{ COle CompObjuObjInfo  @96>@)#Microsoft Photo Editor 3.0 PictureMSPhotoEditor MSPhotoEd.39qOh+'0 $ @ L X dpxCONTENTS <6IOlePres000 FCONTENTSV301Tablea      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~ H%Bݽ⯷ʜ޽˷ɰ­}͒ϓꦵ̘ή߷Ԥ礢ȺͲީҴ᭻颱ȕؾܵθøƦ|ӰӪƦǯ߼ثљۯֳֹٳӯկȨµ樹ӭޣ°İs۾אָvyɗαղ߬ˢ͹;άѪٻ⥰ΥƼߢӬ۬ͦ͜Ӭ⮽Ԯ۷ڢШӲ챼Ңȡ褯ţɳಽ쮹ϟǜ뭸Εţȵҭϻٴ䴿ݳ͗–ٰٝãʱܩϭίףɞĸߺԿ޹㯻կԹ䶾ѩʭ޷ܾ⯸׬ȔΪйݽҵʴɽٶҧѫٿذȱϭ·θϯȭƯȶ̽۸ոոջվۼνϿӷɵǸʺҿԿԽҸ»ҵط޼تͩӚ̳ٯڳܢɬ⶿ξٿЯǴλռ׺պۺԭǯɯɶοҢȯʮɫȥĨź޻ڬɦìɴѱί̴ѷԵҽڶӮ˾Ӹоپ۵ҫȲԾݽ;ȿ푤ęϗ˻ѝʚǝ趾ٽ׽޹Ѵݼܶнݶθһռ׼׹ԩȫȼ׳йڳвϵ᳾мԜ̙䵿⶿馴϶ٶֶسݾѻӯĻ߾ح̯˿ڵҼ᳾Эʲ߭ʵ۹ԴϷׯؼҨӯ»دʱ޸ծ˶ᱼΫȵקıطԽ߫Ѡȩ׳Ьڦ㱹̳ܿڼٱε௺̪Ǹץ®Ꜩپۦʪӭ־⶿лٻ୸ʦô贿ќ޳ιߵᇒՎʤԭʸҞº߿թįݴ쟪͢ԓݺغ۩͛ݵͲξ߹ӹ䲽ϷݾۻظնӶӵҥۓÖûԝŞ˫ݻܱթսٻؿվԶӫȧù߿߿տДݬҾΟ̽ݾ׷߷Բ۳ֺع皤˜įѴشݹӹٶ̾ᴿմ߾߹շӸ޶Ҳμܼ߷޴Ѻ֬ɶܨwóζ̴ʲƼڽߺ޴ճп߾۹ڵѴеѼؽٱͭɶ޹ջ׽ٽټػ׻׼غֻ׼ػ׺ֹպֻ۲ߞnض޽޼ټپ׳кؿ׿׾־־ֿ׽ݽڼٻؿݿܿ׵۽ݶꦦŭþ޶ܻڿڵҺݿܿݺ׹ӹ׽޿ݽۻٻٿݿݿݾܼںغػٿ޿־ٺ۴ٴ۸Чб㕘׶ڹ߮ͼӴô~ȭũ˳߳ͺܿڴӸسб߶ֹپ߿߾޽ݾ޾޽ݺڹٹٺڿݿߺܺܠĦ嫶ԯۼ߿߽馩ҳ̬¥ھݸ׼ݵҰ̽ܳոػۼܼܻۻۼܺڼܼܼܻۺڻۼܾͷԳ֭կڱзּٿֺ׽ױζ鵾ϰ꯿ֳӪնⱼصz}nqx۬|~~|x޻ڹֺػպۺڻ۽޻ۿܻڽܾݾݽܿ䤵Ѣҭ۽ѫҨ̻trua[]`Z\ڭݷģ̛̔įگϧäզߡӴܻܽھֻռ߾޽ݿ߾ܿ߰٭ѮϿгٯŀ_]`smoXRT׭ݴ߻Էԯ鸾̭鳽᤯ͫ۾ݿٷԷԹָҹӻ߿ݿ޾ݾ۴಺ϳֲosraaa_[\lfh]WY԰ߺ㱿ٺ߾ؿֺܲ߼ϻڿܺҵ϶йӿٿ׿ݿܾڼ޾ܳдӺްyz~WYVjigiefb\^b\^Ӵ᯽ػ뛡Ξ˟źھۿ־ջҶͰʯɱζӺлѼҾ־ֻշѴθܺԽꮽޭ߾۵һ[\`kjfa]Zjddlfh_Y[㧵гߴ߶᯽ױڸ۾ۨǷܻݿڿڽڿڼ׸׹ػݾ޻عع߾ؿ䫺ٯ޷۽ٷqrvbcgmjea]Zidalcdhcgݪյׯܭگ⩷Ѳ۳ܸڵطݼٽؽں׹ֶֹյֵ׸ܻ߼ߺ۹ڽ߯ٶߺ޼ߺ׺ׂܿWX\ghldaZniekcabYZ{vz簿ԱହٶՠȰۭݷݵ۸߿ܹطԲӯЭίӴط߻ܻ߽ܽԺ׽^_c^_cdeimhbgb^c[Yh_`|۷竸غ˜İ۫ൻݶṽشϯʰ˷Ҽ׽ܾܿۼۼݶټfdekjhea`d`_hc_fa]jeaXTSߡ¨Ш⫻߮߭߾پ޾پٿڿڽ׽׼ּֽ޵Һݳֹۻyyy_]^`_]a]\lhghc_lgcid`]YX~ԬЬԫۧץըܿ׿ۼٵ߾۬ɭ][\hfggcba]\iedfa^kfca\YfbaܟԸܨҭҹ|z}a_`fbckgfhbbiccidafa^`[Xuqr˘Ģͱ篾սݼ׺ɾ\Z]iefc_`iccgaae__keee`]e__{wxߴ߶߸谿֞ƥЙؾpnqhcgeabicegaae__h_`gaae__d^^zuy׭ӝ֢ɰٳ۽սfaekfja[]e_af``lcdrijc]]e__f`b鲽ӣ̧ҥηϮȨidhlgk^Y]pjle_akbch_`f]^f`biccpjlᲽݩһڳԻ虥Ƴ޹⏦ЩܔØàfa^fa^gb_hc`hc`hc`hc`hc`nfda[]tsyת޻׼ٻؼdžˬ훬~ӚƜŨ鰿onvfa^fa^gb_hc`hc`hc`hc`hc`g_]`\]޻׺ּٽڼټڸո٣ַݎᗦŗôѫիϻڨȵifmfa^gb_gb_hc`hc`hc`hc`hc`jabnim۾ھڽڼټ߹׳д߿諶ҜļЇǵ߹۹㥳ͬkjo^\agb_gb_hc`hc`hc`hc`gb_gb__YYxswܾڽ߿ܶԲϵ޺ָհͶ᰼ʹ݌|䥴՝ʻ檸ջ߷Լbbdhfigb_gb_hc`hc`hc`hc`gb_gb_]WYݹ׷ָֹ鴼ϧ«Ϣƿֳ᭸Եبӥֲӵvxw\Z[fbchc`hc`hc`hc`hc`gb_gb_fa^rnoݻټ۹ܽѱ˻ڽ׾׺ɸǹⶽٸcedb`alhihc`hc`hc`hc`hc`gb_fa^fa^߼ݺ۸ۿݹӺЭêŶѿ۾޿ҳܻጟ“¡۾z{^`]a`^gcbhc`hc`hc`hc`hc`gb_fa^fa^߿߿޾޾޾߼޻ݺ޸޸ݷݷݷݷ޸޸߸޷ݶݶݶ޷߸۷ܸݹ߻ݽܿ뙦ڽ``bfdekgfd`_idaidahdafbad`_hfg\Z[ustî۵߱߳޲޲ݱݱ޲޲߳߸޷޷ݶݶݶݶݶ߻۴ޭÚԹ栯Њ餫Ȧٖȭffhb`aiedhdcidac^[kgdfbad`_hfg][\ߵ޴ݳ޵ݴݴݴݴݴݴ޵߸߸߸߸߸޷޷ݶݹ޺߻߼޾⳽زղڔǢᥰƋڢͧԩկ̸nnpfdec_^jfemhelgdd_\jfcd`_fbaecdljk߽߽߽߽߽߽߽߽߽޼ݻܺڽڽټټټټڽڽߺߺ޹ܵܵ޷߸ް߰޵۴ٴ۶޻޽Ԧǭ蚦ѭؼ͢ǧ۫ο[[]b`agcbidahc`idajebgc`ea`hdc_]^߾ݽ߻޺޺ܸ۷۷ڶڶ۷۷ܸ޹޹ߺߺߺ޹ݸܷݶ޷߸ܵڳܴ޸ݹڶٶڷ޲޴ߵ޷ߺӾѿܰ̿⬷˳䬸Ə矪êսخ׶ߞėyy{dbcc_^mhed_\e]Ze`]idafb_lhgb^]gef߲߲ޱޱޱޱ߲߲ܸܸݹݹܸܸ۷۷ݴ޵߶߾ݻܺݻܼںܻ޽ܲݳ޴޶޶޶ҥѯ媶ά쳾ЩͬҽРϡή颰˘è߫ө[[]ljkb^]lgdf^[mebhc`gb_hdamihYUT~|}սحͱԺ߻߻߻߻߻߻߶߶޹߹ڴܶ޸޸߹۾ܿךּէլ⳽شʟϘĦꪺӤξڢϝȰ||~_]^ihfc_\gb_jb_phekfcidahdagcbZVUڶݸޫü߲߲޼߽߼߼ٿ޲سۿԷ尿ܻ驺ΟȨݭݩغ||~ZZ\ljkc_^hc`e`\ldah_Ze`]kfciebb^]c_^zxyéɢԺ޺ߺ߸ݺڸܯ׳֪з礯Ϣ֪ަhc`hc`hc`hc`hc`hc`hc`hc`idamhehc`gb_fa^`[X}⦶ͩеϧۿڽݶ߶巿٬Ҭޡїƕsnrhc`hc`hc`hc`hc`hc`hc`hc`e`]jebhc`jebmhec^[ida}߽޹߹ۧƮʫŦݷߵ߽֯ү氼ְa\`hc`hc`hc`hc`hc`hc`hc`hc`gb_jebgb_gb_mhelgdc^[_ZWӷ߸宼ٸ޽ߴӫ޻߹皧Ǐ׼پgbfhc`hc`hc`hc`hc`hc`hc`hc`hc`jebidad_\fa^mhelgde`]__g箺а׫޼߾ѣ`^ae`dhc`hc`hc`hc`hc`hc`hc`hc`e`]gb_kfchc`d_\gb_idafa^cad][^ܹ⯹غ׼߸ݻ޻߸׾۶ܾZX[kfjhc`hc`hc`hc`hc`hc`hc`hc`gb_gb_jebkfcidagb_gb_gb_hbbc_`igj޷ܺݺݹ߷߷߷޾նݽusv][^kfjhc`hc`hc`hc`hc`hc`hc`hc`idaidagb_hc`kfckfcjebmhehbbnjka_bvw|޷ۻۻܸ޶޶߷߷ܽ׳վ[Y\b`cfaehc`hc`hc`hc`hc`hc`hc`hc`fa^jebhc`fa^idahc`fa^idafbalhi^^`\_dt{ݻݵݲۻݵݼ߳৻֩ӻgef_[Zlgdhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`jb`d\Zldbkbcc^d{~޼޶޳ۻܲݹݾݼ߹Խ䴸ijܪ؆xvʃ€Ư潿޸ۭڠɴmklgcbidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`h`^ldbi`ad_ciisݫ̾޼޶ߴܼܷܻܰܽܿ?@@=AAABCDAC<>>?AB?@?A?A>@>@>@>@?A?A>>==9999@@AA::6682<6NJb`rsÎϬ߹ZV@>DBA?:897;9=;><=><=<=;<;<:;:;:;=?=?<>;=9;8:8:8:3669:=@ABDDCCBBA;<<=<=;<;<:;;<=>87GG߮ɻػƥPL4/    $&EJahͯٵ٥ɵﲶx|}jkmpqspqs{~dbcd`_c^[idahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`idakcakc`f^\hde\[atx⪴Ͳի̵߽߷޾޲ݴܸܿ'% ,' } <3~ uӬ$' %(ۼ߬ӻ̒monZZZ```fff]]]a`ecbgzz|eeedbcfbahc`lgdhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gc`idaldae][lffa_dknwٸܹ߽߿߸BCms ^_ |ާ™ķur |  աʦϿƁXXX___gefdbchdeiefiefeab^\]ecf```ighfbaidafa^hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gcbhc`ogdg_\kcahdelkqںrt| !XY pլ빺"!ihf^][dcaokjiedc]]f``icckbcaXYkeegcd\Z]mnp___ecdc_^gb_e`]hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`edbiebphekb]f^[f``a_b ek!$AG:BCACAB@A?A?A?B@B@C=?9B@:,&  >?=9=9=9=9>:>:>:>:<9>;?<>;<9;8<9=:=:=:>;A=A=D=D>D>=9<8=9>:>:=9:673u"#v۾FFz}۲Θ|\_;< jnw_[Xhdakgde`]c^[nfdogeg_]h][ndcskid^^VTU~~`acbbbecdhdcidaojghc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`dcajfcle_kb]h_Zg_]^Z[xvy()xw;: }|  rxUU  .( ('۽`t/1|sx~X]chdafa^e`]hc`ldbh`^h^]lbapbaodbdZYicclhiWWYfgk`acaaafdeied`[Xjebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`bb`kgdg`ZjaZmd_ldac_^|z{A= ikޞMT޾߻qt}|AERW noqVWRik`hc`hc`hc`hc`hc`hc`hc`hc`f``jfehdejfgcadYWZᦡpkhc^[jebgb_jebidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`{{{  &*AD#پ޾߻sv ,( glڵݨ٨ML ̲֮׆`acde`de]hc`hc`hc`hc`hc`hc`hc`hc`gaaiccd`_gcdgefa_bvtwzurb]Zgb_ojgjebjebhc`gb_hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`faguzšҲ  X^̸Ϲӹӻտ߿߼߻x{ [\sy䠦ƍǽ ~ᲽӨ噜bcg^^^kjf__Whc`hc`hc`hc`hc`hc`hc`hc`hbbicce__fbcjfggehfdgmknfa^c^[kfcjebe`]hc`gb_gb_hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gcbaenȳ⧵ϰٲ׺*' tzPSӶѼּּ־߿߻߻߶{~dh)'}t{ݪװ٬ꌎ氶а줨pqsccefecgc`fc^hc`hc`hc`hc`hc`hc`hc`hc`jb`jebjddhbbgcdiefecd`^akfckfcidad_\e`]idahc`gb_hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`mjc^`_szۢ˧ҠƥNOUX~ Z`׻ֹԷҶѶѷҶѶ߻޵߶wz  ۽ھۿyzFDﭱopr\\Zhgeiebe`]ojghc`hc`hc`hc`hc`hc`hc`hc`kcakcamecicce__fbcjfgiefhc`idafa^idanifmhejebidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gcZda\\]_෽߶᦯̩鄆"!FJغշҵдϲͰˮɱ˴ζзѸҼ߻߻߶޶߷kn }ټӻҾx{ ۽86ٸ롡rsw^^^cb^gd_gb^fa^kbchc`hc`hc`hc`hc`hc`hc`hc`lb`ia^jb`kcahbbgaahdcjfgfa^idagb_kfckfcfa^idajebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`d_[qjd`[Wx{Ͳܣ̬ϭ֠ų޻ߵ ⚢ Z[ۼٺ׹ָշԴѲϭȰ˳βͰ˱̷ҽ߼߻߻߻߶޵޴ݷܾZ] !ٽkn  վ޻  sx~jiq[Z_cb`he^gcZje_h`^e\_hc`hc`hc`hc`hc`hc`hc`hc`ka_ka_ia^jb`jebhbbgaahdcidalgdhc`idae`]b]Zkfcidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`e_aqfbjaZhji͞Şݺ޴ߴ41xrwCHtطԶӶӶӷԶӴѲϪūƭȮɯʱ̲ʹϳ˼ߵ߻߻߻޵ݴ߷޻MO}75MPٿQOяhowUZ`_]hedifec^[TgcZqjdg_]mbfhc`hc`hc`hc`hc`hc`hc`hc`i_]ndblb`ia_ldbiccgaakeee`]jebgb_jebgb_idaqlifa^hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`kbgi[ZnaY[[Ytךݻܸڼڿݹ߱HG  @@榫 [=D޷74 y~ x޹طֺٽܿ޿@?ذӽ~  MR楤trwd^`h_`mcbhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gb^lgcg_\d_\kijھܹܵڷۻܶ߱| !jnܵݶ߸85 %$#߽ܺٺٽA@ ڽHDheӔmlr_]`d^`jabkcahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`kfbkfbe`\c^[a_`ܶݯ۱۶۴ݰ޷޷߸߸߸߸߸ x ZY߻ݾݿٽոֹٲڳݶ߸9640 {|ڻغ׹ֺ׻ؼٽ0.E@>>䔗zzabg_^chcgjdficckbchc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`d`]e`\jealgdd`aopu޸ݯܲ۶۴ݰ޷޷޷޷߸߸߸߸01bckoz ۺЭ˭ܯӮγֽ̹ܺкۺܲװ׳ٲڳܵ޷߸96:6  02ٽʲɴθռٿ޴ݾۼٻػؿ0*~  z{߿ѐmmodch`_daacigjjfge__e__keehc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gfbd`]id`id`a]\[Z_ݹܹۻݹ޳߸߸߸[a12~ RVݿگٰܵݶݶ޷޷߸߸߸41 95rn  ),޿ܾܾ޽ݷ WX  KK䝝ttvggg[[[ighhfgiefhdegcbea`gaaicchc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hgeea^id`jb_hc`[Y\ijo| s ir ޶ߵߴ߶߸߸߸޷޷޷$!?Bէȴrmjd_\niffa^gb_jebidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`ifmڧԭڧԦӶ簽ݬټ޾޶޴߶߹߸ݹ޹޹޹ݶݮ޻ܱ㋋}     Թ߼޻ܹݺ߻߻޺޹޳ݲ޳ۿ gkW\  ~9:cd߹ӱͶʹϹӼطϴ̵˹ϽӺϻоݲΪƬȱͰе⚞JM۠Ϣ̺PQ  adŔxspd_\gb_jebgb_lgdjebfa^hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc``^c~⧵СʥΟȫԿ寽ص޸޹ߺݹݹܹݺ߼߻޺ݺںٹٺټټݲ۬ޯݴںڼ۸ߴݵ"w   {3+֚еұѰЮЮϯϯήέ֩ҨԬܶ׽ܿҷ̲ʳ͸ռٸ׮İʯ˯θܠimNP&)[aT\r}ŠݵظʴƲȹӸոлӻӼԻӷծģƴηϲƯ³˷ܯǤɾ߽м߸Աͮʰзދ      `gӹӽֻԶ۸ܴڻ#  |ᝥ졥^YVfa^lgdfa^hc`lgdhc`idahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`d_cXV[ܩͮҹݫϛɷݿݽ߻ߺ޺ۻۻ۽ܮߴٵܹ޺ٰ>: ܲڴٳٲ׳ֲձմմٺѳзڽ׶ѽ܏}II׳ιҼ߹ھѺʽлиѼ߿ԾԸ˩Ǽձʵοؽپ{bfǐؗԔіӝڤ᧮䧮䤭ԫ۷礭Ԛʚʖ֕ՔԓӓӓӐЌ̚ᡦꤩ훡ߗбھξؽvw LGͽߢ³߹ӡǪݴؾswzjeblgdjebhc`kfchc`d_\kfchc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`ied_Z^ebi֨DZй޺ݺֶڷڬ^Z       '"~޹޻׿ꔙ  87  +1ܲԪƮy}IJgn٣ĦDzۖó☧Ʃbcgidae`]e`]kfckfcfa^fa^fa^hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`fa^mih]X\߿ۿӷ޺ʀ  D<ůġêŭ¦JJnֽ  -+y7=ۧæ¸ݵѴ| zֵܾ⹿ߴڼ۷  12ɘ¸뢯Œݺ甕]^`e`]gb_idakfce`]fa^mhefa^hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`d_Ynif^Z[lipڷӸ  4Ǹ߱ڬܰқ!swԬ*% ''ܫǭɾңkoҹ߽Źƺtx  힩ŜöϠƻޱҠopr^_aidanifkfcidad_\fa^ojghc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gcZfa]c_`^Y_޳Q;az  ޽ۺڹݼݼݼ޽޽޽޽ݼݼԾٽٰ߮ݻH@  xjn  ߻ gi᷽߫˲դº۱͟PT ιƴEJ e_ܘʨʷּ׷ϺՓ[\^fgifa^jebd_\idalgdgb_idad_\hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`ibXc^Xkee`[arov޹ܥ^[ ``ݽܴݵ߼޻սԹ j ec඼  RTضԺ߸߻ޮ˫ȽܭǫŻݴα;<鮽ļĬ  87ĥ߰Աͭҭᨰթa]\fa^kfchc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`iccb\\hdcd`_nji^][pomtsqzwe`]fa^mhegb_gb_idajebhe``\Yiefc^bfaelhiiebeb]hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`eb]jebkbcf]`e`f齏 5/ݷݵ߷߼߼Ӽbexy[Y bfּյҷߵϯ44߿佺˿ü¹EBqv蘘۟ӣөiedd_\fa^hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`icckeenjib^]ea`ʂ}zfa^jebfa^hc`hc`e`]da\okheab|w{hde^ZWmjehc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gd_hc`jabh_bb]c÷#  {ܯ߷߼ؽ  LI܀}ڿԺں''&$ݺжۼް˺þ¾89$%י¼۸خ׮ӗ֤׻㪹ְ֢nlmjfejebc^[hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`keeicc_[Z]YX󈃀`[Xidafa^lgdkfchc`da\a]Zd`afbcc_\ifahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`he`fa^jabkbe_Z`rqwA9 vlޭ[V VW>; `]  پݯɮ31ڵѳϹޯҾϠǹۻޤƒۚٞҨ`^_c_^qlid_\hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`b\\keegcbfbavqnb]Zlgdfa^hc`d_\e`]eb]d`]okha^Yhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`ifae`]i`akbe^Y_pou޷ݶݲea@;߬ݶ޾׺} 1- ޮTK ߺںզ 42ܼԺ߳VZ 31ѡϰә٭㪹֯ݿooodbcd`_jebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`mgggaa_[Zkgf|y^YVc^[jebgb_kfchc`niffc^~{ea^jgbhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`he`fa^i`ai`c`[awv|߽޼绎  ص׮ۭ^[pX\":2#ؽ崽ίɺ;:ټ٤ϰމ  {⏚ٿӥϳ窱^^^dbclhg^YVjebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`a[[gaahdc~}fa^fa^lgdhc`e`]gb_e`]gb_ZWRc_\da\hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`he`hc`i`ai`ca\b}|Էz ~ڶ۲ڬò  /( ,⶿ж零PO׻睪ʑǨزֹ׏ަ¡Ѣʪqrt```dbclhga\Yidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`nhhhbbgcbqpne`]kfchc`hc`hc`hc`jebfa^he`vqugcdgc`c`[hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`he`hc`jabjada\bzy;0BE޹KKBG  sϭ߾ּӺ햢 dcҮښ͖ӧwԚŤƧݿܣԶۤ՜֞컴ӭWXZaaahfgd`_nifhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`_YYjdd祤dcakfclgde`]mhekfcd_\ida_ZW~upt`\]fb_lidhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`ifagb_kbclcf_Z`srx5>  ߹ܬé   ׷̿  ۴вػy{ nj洼ϾյƳޖ͛ͻӴҿ/.YXح峺ԾΥ˦ͫ՟Ӝʩxspfa^lgdc^[jebfa^kfchc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`idagb_hc`idaidaea`kgf{wvb^]hc`hc`lgdidagb_b]Zb]Zwro񬨧gcbgc`d_\jebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`fa^fa^iedcab]]]qrt~~  |^U>A۵ܹ߽ں &ݻᶿδnp }~ܶӷܘǒƫ紼򁇻 |擄șϭ߫lgdd_\idagb_lgdd_\idahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`idaidaidagb_gb_idaidagb_\XW}yxhdcea`idahc`fa^idaojgd_\e`]񋇆b^]ea^kfcidahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gb_gcbecd___vwy QMѷ ݷڷfZ E=ܴϾ_`z᳾Ҳ﫼Ќ~~acs ;=Ϙ첿ңçթqlie`]e`]fa^jebkfce`]idahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`jebhc`fa^idakfcidagb_lhg~zya]\fbaidahc`c^[e`]kfc]XUlgdrnmfbaea^mhec^[hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`idaidafbagef___{|~\T9-ܽ$ݷ>9 91⳻γLLݼݮ޷иѽݦ̨B= =<⵶ֶξz  ׻޵һஹͰϣ™`[Xidajebgb_idahc`hc`kfchc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`e`]hc`gb_e`]jeblgdidagb_^ZYd`_lhgkgfidajebjebjebjebhc`njifbad`]mhehc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`jebfbagefZZZ<< vz~߻ߺ߻߶㿲 ޵rt 5(޶  >>Զ10RRef^`dfac]_]_`bdffhfhfgfgfgghghhihihihjcefhkmjlgigidfehcfUX?A<=9:./67ݾָҽPJ  x 81eaϊީ೾ڴ챼ؕt|@F "%ޯЪʹhime`]lgdlgdkfcgb_fa^kfcjebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`fa^jebhc`fa^jebhc`e`]ida~gcbc_^fbajebhc`lgdgb_^YVgb_mihd`_hdajebgb_hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`fa^kfcfbadbcWWW߯һ_a}KQ۸UK 2"߿ֺZ[ 8+۷ݯ޸ -.ں     ܾڽվ׼ֵϼ߻rp @>/-315320.+xx@B{|WX\hc`hc`jebkfcgb_gb_jebgb_hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`idamheidahc`jebd_\d_\snk_[Zjfekgfhdcjebfa^kfckfce`]tol픒ea`a]\mifgb_b]Zhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`fa^kfcfbab`a\\\̭ۤ٫بӸ   "¥  ߺپֺҿټԾCC?1ܳױ߹zo  TUտۻӷϹѾӽ־߹жϺӿ׾ؿ -,ݸԷ{}޻𞟣`aeabfgb_gb_idafa^gb_jebhc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`jeblgdhc`idajebc^[lgd܂~}_[Zb^]hdca]\hc`fa^hc`gb_d_\lgdtrsb^]b^]jfchc`idahc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`hc`gb_idafbaa_`iii஼׹#  {}87|߼ݿ43G:߹UM ZUݾۿ <<߼  u{