Cover Image

Not for Sale



View/Hide Left Panel
Click for next page ( 61


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 60
60 rates, perhaps subdivided by severity. The data on which 16 guide concerning alcohol-related (AR) crashes, since ef- problem driver identification and effectiveness measure- fectiveness levels (AMFs) are either given or can be estimated ments are based (traffic convictions and crashes) usually re- for some of the AR strategies. However, this procedure could side in DMV files. These data may sometimes be added to the also be used for other operator-related strategies if effective- state crash file. ness is known or becomes known in the future. The third It will also be noted that many of the strategies proposed in procedure is specific to AR strategies, and is based on advice the guide could require the enactment of legislation, depend- given in the Volume 16 guide. ing on the state in question. Selection of a strategy requiring Although most of the proposed treatment strategies do not legislation entails an assessment of the likelihood that the leg- have precise AMFs, a substantial number of the strategies in islation could be enacted in the required time-frame. Volumes 2 and 16 are supported by compelling evidence Another consideration is cost. An assessment of cost for concerning their efficacy and effectiveness. This is not true for many of the proposed strategies will require subjective approx- aggressive driving countermeasures (Volume 1). Those who imations. Some general information on treatment cost is implement strategies for reducing aggressive driving must presented for each treatment in each of the three guides (Vol- also contend with definitional ambiguity and the absence of umes 1, 2, and 16), and that information should be reviewed by a database for identifying such drivers and initiating appro- the user. Very costly strategies should be avoided unless sup- priate sanctions. These limitations will necessitate use of sub- ported by proven effectiveness data and an estimated effect size jective judgment and indirect methods. that is sufficient in economic terms (dollar benefits) to be cost- beneficial or cost-effective. Strategies that are judged to have Procedure 3 Choosing Treatments negligible or moderate operational costs (excluding start-up) and Target Subgroups Related To Illegal will usually be cost-beneficial if they produce statistically signif- Driving Actions When Treatment icant annual crash reductions as small as 510 percent over Effectiveness in Terms of Crash/Injury baseline. It is also possible to make statutory sanctions such as Reduction Is Unknown ignition interlock, vehicle impoundment, and license suspen- sion self-supporting through administrative fees and fines. The assumption here is that there is no known level of ef- The guides classify strategies into three categories: fectiveness for the treatment strategies of interest no defined CRFs or AMFs. Thus, economic analyses like those that are 1. Proven the basis for Procedures 1, 2A and 2B, and 4 are not possible 2. Tried but not proven for these treatments. Procedure 3 is aimed at helping the 3. Experimental not tried, effectiveness unknown. analyst make an educated choice of which treatments will be most effective in his or her jurisdiction, and to help the In selecting treatment strategies, priority should be given analyst develop a targeting strategy for the treatment in cases to strategies rated as proven. However, the safety planning where it is not to be applied jurisdiction-wide or to the "total team is encouraged to use their own judgment and to inde- problem" (e.g., where specific illegal-driver subpopulations pendently review the evidence cited in the guides in selecting or jurisdictions are to be targeted). treatments. The tried category includes those treatments that However, unlike road user populations covered in other have been used by agencies (in some cases used often), where guides (e.g., older drivers, pedestrians), the choice between there is little possibility of negative impacts on crash/injury alternative treatment strategies found in each of these three frequency, and where there is an expectation (but not scien- illegal-driving guides is much less oriented to specific crash tific proof) that the effect of the treatment on safety is likely circumstances (e.g., different crash types, times of crash, to be a positive one. The evidence could include poorly de- crash location types, etc.). Instead, most of the strategies are signed or executed crash/injury evaluations and indirect or related to improvements in programs, such as impounding surrogate measures that may be related to safety (e.g., behav- vehicles of repeat offenders. In addition, for both the guide ioral changes that may be related to crash/injury reduction). related to aggressive driving and the guide related to unli- The following sections present three methods for choosing censed, suspended, or revoked drivers, the strategies are and targeting strategies. The first is a modified version of Pro- related to the full group of such illegal drivers. Thus, while cedure 3 described in Section III, which is usable when treat- targeting of a chosen strategy can occur based on jurisdiction ment effectiveness (AMF) is unknown. This procedure could and selected areas within a jurisdiction, it is not apparent how be used for any of the illegal-driving strategies found in any crash data could be used to further target the strategies. AR of the three guides. The second procedure is a modification crash strategies presented are essentially oriented to three dif- of the economic analysis procedures found in Procedures 1, ferent groups of drivers young drivers in AR crashes, all 2A and 2B. It is primarily directed to strategies in the Volume drivers in AR crashes, and repeat DUI offenders in AR

OCR for page 60
61 crashes. Examination of crash data can provide the analyst ("crash"), (2) variables for each vehicle in the crash ("vehi- with information concerning which of these subgroups are cle"), and (3) variables for each occupant/person in the crash producing the largest AR problem in a given jurisdiction, but ("person" or "occupant"). The variables needed to determine virtually none of the strategies within each of the three sub- whether a crash is a "target-population crash" are usually groups are susceptible to further targeting. found in the occupant/person subfile, but could also be found For these reasons, the general analysis methods presented in the general crash subfile (e.g., a "flag" for all alcohol-related under Procedure 3 in other sections of this manual are not as crashes) or "vehicle" subfile (e.g., driver information included applicable here. For that reason, only a modified Procedure 3 is with each vehicle record). In short, crash files differ from presented below one which continues to use relative estimates jurisdiction to jurisdiction. While certainly not always of the program effectiveness for different alternative treatments, the case, the following variables (or similar variables) listed in but one that does not include further targeting steps. Exhibit VII-1 will be used in this identification of "target populations." Note that while such definition is possible from both drivers in AR crashes and drivers with suspended/ Data Needs revoked licenses, there is no clear definition of "aggressive Note that Procedure 3 is a "crash-based" procedure. It drivers." Thus, the defining variables will depend on the user's assumes that the analyst wishes to choose among the alterna- definition (e.g., speeding well above average traffic speed, tive strategies and target the treatments based on crash data. multiple violations at the same time, etc.). It is noted that an alternative way of making such choices is Finally, note that identifying alcohol-related crashes through linking crash data related to problem size to an involving repeat DUI offenders will be much more difficult assessment of the existing programs in a jurisdiction, and since this group is not identified by any combination of vari- choosing to implement those strategies which are either miss- ables in the crash file. More information on defining crashes ing from the current program or have the least extensive (or for this group using a "driver history file" is included below. least effective) degree of implementation. This program- As noted above, since the strategies described in these three deficiency procedure is described more fully in a later section. guides are generally applicable to all aggressive or illegal driv- However, if the analyst wishes to choose treatments and ers, or to subsets of drivers by age, they do not lend themselves targets based on crash data, the revised Procedure 3 described to a great degree of additional targeting in many cases. AR here basically requires crash data that will allow the analyst to strategies could be chosen based on driver age (e.g., strategies (1) isolate crashes involving the specific user population of of young drivers in AR crashes vs. strategies for all AR crashes). interest (e.g., drivers involved in alcohol-related crashes) and Some of the strategies in Volume 1, the Aggressive Driving (2) define crash types or crash characteristics (e.g., AR crashes Guide (e.g., "Targeted Enforcement"), and Volume 16, the AR involving young drivers) for this user population which Guide (e.g., DUI checkpoints), could be targeted to high- would suggest strategies and target subgroups. priority locations or high-priority times of day based on crash To isolate crashes involving the population of interest, the occurrence. (They could also be targeted based on citation analyst will need to examine the data formats/coding in data that includes location of the offense, assuming that such her/his crash files to identify variables that can be used in de- enforcement is somewhat "random" across the jurisdiction.) termining whether or not a given crash is a "target-population The names of crash variables and the specific codes needed crash." Crash databases often categorize data for a given crash to conduct these targeting analyses will vary from jurisdiction into up to three subfiles (1) general accident/crash variables to jurisdiction. While not all relevant crash variables are Population Type Variable Crash Database Subfile Drivers Involved in Alcohol- Alcohol Involvement Crash Related Crashes Law Enforcement Suspect Person/Vehicle Alcohol Use Alcohol Test Person/Vehicle Violation Codes Person/Vehicle Drivers with Driver License Jurisdiction Person/Vehicle Suspended/Revoked Licenses Driver License Class Person/Vehicle Driver License Status Person/Vehicle Violation Codes Person/Vehicle Aggressive Drivers (Depends On User Definition of Aggressive Driving) Exhibit VII-1. Crash variables and subfile location by population type.

OCR for page 60
62 presented here, Exhibit VII-2 provides some guidance con- frequency and severity of the specific types of user- cerning where example variables related to some treatment population crashes occurring in an analyst's jurisdiction. strategies might be found. Crashes specific to a given user-population were defined in Exhibit VII-1. For each user population, the analyst could begin the process by analyzing 3 to 5 years of crash Procedure data to determine the frequency of each crash popula- As described in Section 3, Procedure 3 has two basic steps. tion either total crashes or some subset (e.g., fatal and First, choose the best treatments for the user population of in- serious-injury crashes). However, since the severity distri- terest (e.g., the treatments related to AR crashes most likely bution may differ between some populations, and since to be applicable in a given jurisdiction) from among the set restricting the analysis to only fatal and serious-injury of all treatments presented in the applicable NCHRP Report crashes will severely limit the crash sample and will omit a 500 guides. Second, choose the subgroups of users (e.g., large component of the crash problem non-serious in- young AR or aggressive drivers), highway locations, or times jury and no-injury crashes a better solution is to weight of day to which the selected treatments should be applied. each crash for a given user population by an economic cost As described earlier in more detail, the choice of the "best based on its severity, and then accumulate the total crash treatments" from the listing of many potential user-population cost for each population. Information on economic cost treatments can be based on the following factors: per crash-severity level can be found in Crash Cost Esti- mates by Maximum Police-Reported Injury Severity Within a) The potential treatment judged to be the most effective, Selected Crash Geometries (22). Here, instead of using even given that effectiveness is unknown severity cost by crash type as is done in roadway-program b) The relative magnitude of the crash types and severity analyses covered in earlier sections, the analyst will use the levels that the treatment will affect basic cost per crash categorized by police-reported crash- c) The cost of the potential treatments (either jurisdiction- severity level (i.e., K,A,B,C,O). Exhibit VII-3 below wide or per-mile or per-location) presents those costs per crash. Costs for combinations of d) Other technical or policy considerations crash severity levels (e.g., K+A crashes) are presented in that report (22). This analysis of total crash cost will pro- These factors must be combined in some fashion to decide vide the analyst with overall information on which of these which treatment to choose. While there are multiple ways of three illegal driver populations is most important in making this choice, the following represents one such his/her jurisdiction. If only one of the illegal driver popu- procedure. lations is being examined, the analysis can provide useful data for public information programs concerning the total 1. Prioritize the specific user-population problem(s) to be cost of such crashes. addressed. 2. Prioritize the specific alcohol-related subpopulations to An initial issue may be whether to treat one, two or all be addressed. three of the groups covered in these guides aggressive Once one or more populations are identified, the sec- drivers, drivers involved in AR crashes, and/or unlicensed ond step involves the identification of subgroups in most and S/R drivers. This decision can be based on the need of treatment. In general, the strategies in the guides Crash Type/Issue Variable Crash Database Subfile Driver Age Driver Age Person/Occupant or Vehicle Driver Date of Birth Person/Occupant or Vehicle Time of Crash Light Condition Crash Hour of Day Crash Speed-Related Crashes (for Driver Action Prior to Crash Person (or Vehicle) Aggressive Driving) Violation Indicated Person (or Vehicle) Contributing Circumstances Person (or Vehicle) Crash Location (for Targeting County Crash Treatments) City Crash Route Crash Milepost Crash Longitude/Latitude Crash Block Address Crash Speed Limit (for Use in Speed Limit Crash Developing Cost per Crash) Exhibit VII-2. Crash variables and subfile location by crash type/issue.

OCR for page 60
63 Crash Severity Speed Limit Comprehensive repeat-offender crashes cannot be calculated, then the an- Category Cost/Crash* alyst could not use this modified Procedure 3 to choose < 45 mph $3,622,200 Fatal (K) between AR groups, but could use the program-deficiency > 50 mph $4,107,600 Serious injury (A) < 45 mph $195,700 procedure described following this procedure later in this > 50 mph $222,300 section to make AR treatment choices. < 45 mph $62,200 Moderate injury (B) Note that the number of crashes involving a repeat of- > 50 mph $91,600 Minor injury (C) < 45 mph $40,100 fender will almost always be a very small part of all AR > 50 mph $49,500 < 45 mph $7,000 crashes in any jurisdiction. Thus, if the choice of AR No injury (O) strategies is to be based primarily on the size (or economic > 50 mph $7,800 * Crash Costs in 2001 dollars harm) of the crash problem, the AR strategies related to total AR crashes or AR crashes involving young drivers Exhibit VII-3. Crash cost by crash severity level will always be the choice. However, calculating the num- and posted speed limit (22). ber of repeat-offender crashes when possible will provide for aggressive and unlicensed/suspended drivers are aimed the analyst (and the public) with solid information on the at the entire population of such drivers. However, the relative size of that part of the AR crash problem. strategies in the AR guide are essentially oriented to three 3. Identify possible treatments for use for each high-priority different groups of illegal drivers young drivers in AR illegal driver group. crashes, all drivers in AR crashes, and repeat DUI offend- The analyst will then review the pertinent NCHRP Re- ers in AR crashes. The problem size (or total crash cost) port 500 series guides and list treatments that would be for the first two groups can be calculated using the proce- most appropriate for each of the high-priority illegal dures in the above paragraph in combination with the driver groups identified in the above step. The choice "Driver Age" variable in the crash file. However, the isola- should be limited to those treatment strategies that are tion of alcohol-related crashes involving repeat DUI classified as proven or tried in the guides. offenders will be much more difficult unless this group is 4. Rate the possible treatments based on estimated identified by variables in the crash file. In a limited num- effectiveness. ber of crash files, information will be added concerning Since this procedure deals with treatment strategies whether a crash-related AR citation is the "first" or a "sub- with unknown effectiveness, this appears to be impossible. sequent" offense. Those AR crashes coded as having a However, for a given set of possible treatments for a par- "subsequent" AR citation would be the target crashes. ticular user group, it may be possible to make a judgment However, if this information is not available, then the an- concerning which treatment strategy would be expected to alyst will have to rely on other data sources. If the analyst be most effective. This judgment will be most difficult for has a usable "driver history file" (see Section II) and if that the aggressive-driver strategies, where there is essentially file includes information on crashes (in addition to con- no information on crash-related effectiveness. The judg- victions), then he/she could use those data to estimate the ment will be somewhat easier for the strategies in the other number of crashes related to repeat offenders. (Note that two guides since they contain some information on esti- the crashes counted should occur after or at the same time mated effectiveness for some of the strategies. as the second or subsequent DUI offense.) If that driver- 5. Choose best treatment(s) by considering estimated history file does not contain AR crash information, then effectiveness, cost and other technical and policy the process will be much more difficult. The analyst will considerations. have to use the driver-history file to identify the popula- The analyst will then combine the output of the steps tion of repeat offenders based on the number of past AR above with at least two other factors in making a final de- convictions. This group would then have to be matched to cision on which treatment(s) to implement the cost of the crash files for the time period under examination (say, the treatment and other technical and policy considera- by driver license number, which is usually found on both tions. Unfortunately, there are no good guidelines for how files), and specifically to crashes that occur either after or to "weight" the different factors. While problem size (total at the same time as the second or subsequent AR offense. crash cost) and assumed treatment effectiveness are key The identified crashes would then have to be further factors, there may be technical, policy, and cost consider- screened to determine which are alcohol-related before ations that will remove certain treatments from consider- the estimate could be made. Finally, if the state has a cita- ation even if they are felt to be effective. The analyst will tion tracking system that includes information on crash have to choose the final treatments based on best judg- occurrence in conjunction with an AR citation, this could ment. The procedure outlined above will at least ensure provide the needed crash-related counts. If the number of that the major factors in the decision are clearly defined.