DOT Logo
U.S. Department
Of Transportation





FMVSS no. 111
NPRM to Require a Rear Detection System
for Single-Unit Trucks




Office of Regulatory Analysis and Evaluation
Planning, Evaluation and Budget
August 2005





    Executive Summary
  1. Introduction.
  2. Background.
  3. Target Population.
  4. Benefits.
  5. Costs.
  6. Lead Time.
  7. Cost Effectiveness and Sensitivity Analyses
  8. Alternatives
  9. Unfunded Mandates Reform Act Analysis
  10. Small Business Impacts
  11. Appendices


    Executive Summary
    Table of Contents

    The agency estimates that approximately 79 fatalities per year (13 on-road and 66 off-road) and 148 injuries per year are attributable to straight trucks backing up. The agency believes that requiring a rear detection system will reduce the number of fatalities, injuries, and property damage crashes by giving truck operators the ability to detect objects behind the truck. In this analysis, we examine two possible counter-measures: a cross-view mirror system and a camera system.


    The agency estimates that requiring a visual rear detection system would result in a net reduction of 23 fatalities and 43 injuries annually once all single-unit trucks are equipped with a rear detection system. The present discounted value of property damage savings is estimated to be $32 million annually at a three percent discount rate and $25 million annually at a seven percent discount rate.


    The agency estimates about 18 percent of the 365,000 new single unit trucks sold annually have cross-view mirrors or video cameras, leaving the remaining 299,300 new trucks affected by this rulemaking. In addition, based on an agency-sponsored study, we determined that the maximum distance between a cross-view mirror and an outside rearview mirror to provide a meaningful image should be 5 meters (16.40 feet). Trucks with a mirror separation of more than 5 meters (16.40 feet) would be required to use a camera system.

    Therefore, of the 299,300 trucks, we estimate a counter-measure distribution of about 25 percent with mirrors and 75 percent with a camera system as follows:

      Fleet Distribution Percent of Total
    Mirrors 8-inch 14,180 4.74%
    10-inch 61,271 20.47%
    Camera Systems 223,849 74.79%
    Newly Registered Trucks 299,300 100.00%

    The estimated consumer cost per vehicle for a 20.32 centimeter (8 inch) diameter mirror and hardware is $54.64, a 25.40 centimeter (10 inch) diameter mirror and hardware is $56.85, and for a camera system, monitor, and mounting hardware is $325.10.

    The total consumer cost is estimated to be $77 million ($2004) annually.

    The net cost per equivalent life saved at a three and seven percent discount level is:

    Discount Rate Cost per Equivalent Life Saved
    3.00% $2.3 million
    7.00% $3.5 million


    I.  Introduction
    Table of Contents

    The current Federal Motor Vehicle Safety Standard (FMVSS) No. 111 specifies performance requirements and location of rearview mirrors. Currently, the requirements of this standard apply to passenger cars, multipurpose passenger vehicles, trucks, buses, school buses and motorcycles.

    We are proposing to upgrade the current standard to require a rear object detection system in straight trucks [1] with a GVWR between 4,536 kg (10,000 pounds) and 11,793 kg (26,000 pounds). The purpose of the proposed amendment is to alert drivers to persons and objects directly behind the truck, thereby reducing backing-related deaths and injuries and also lower incidences of property damage.

    This notice proposes two compliance options. Vehicle manufacturers could satisfy the proposed requirement either by installing a mirror system or rear video system that would make the area to the rear of the vehicle visible to the driver. This analysis examines the costs and benefits of a mix of cross-view mirrors and camera systems.


    II.  Background
    Table of Contents

    In March 1995, Mr. Dee Norton submitted a petition for rulemaking to the agency seeking to amend FMVSS No. 111 to require convex, cross-view mirrors on the rear of the cargo box of step vans and walk-in style delivery and service trucks. The petition was intended to prevent future tragedies similar to one that befell Mr. Norton's grandson, who was killed when he was struck and backed over by a delivery truck in an apartment complex parking lot. The driver was unable to see the area directly behind the vehicle in its side-mounted rearview mirrors.

    In determining whether to grant the petition and deciding how to substantively respond, NHTSA decided to solicit comment from the public. To this end, NHTSA issued a request for additional information, which was later followed by an Advance Notice of Proposed Rulemaking (ANPRM).

    Request for Additional Information

    NHTSA published a notice in the Federal Register on June 17, 1996, seeking information on cross-view mirrors and other alternative rear object detection systems (61 FR 30586). We received six comments in response to that notice.

    Commenters described a variety of available rear object detection devices, including both visual and non-visual systems. Visual systems include not only cross-view mirrors but also video cameras mounted on the rear of the vehicle that are connected to a monitor in the occupant compartment. Existing non-visual systems include ultrasound, radar, microwave, and infrared sensor mechanisms, which detect an object and provide an auditory signal to the driver that an obstruction is behind the vehicle, as well as audible alarms that sound whenever a vehicle is backing. These comments provided initial insights that helped NHTSA to direct the course of the rulemaking process.

    Advance Notice of Proposed Rulemaking

    NHTSA issued an ANPRM on November 27, 2000, to gather further data on key issues related to rear object detection (65 FR 70681). In addition to a request for general comments, the ANPRM posed twenty specific questions for public input, which were broken down into four main categories: (1) questions concerning rear cross-view mirrors; (2) questions concerning rear video systems; (3) questions concerning other rear object detection systems, and (4) other questions. Generally, the cross-view mirror questions concerned the size, design, and placement of mirrors, the size of the detection area behind the vehicle, the use and capabilities of exterior, audible back-up alarms, and test procedures. The rear video systems questions sought input regarding image size, display color, screen size and location, need for a system failure alert, possible conflicts with state laws against video screens/monitors in view of the driver, and test procedures.

    The questions pertaining to other rear object detection systems asked about the capabilities and limitations of these non-visual systems, including efforts to increase the range of sensors so that they are effective at higher backing speeds. These questions also raised the issue of how to develop test procedures that would ensure the accuracy and reliability of non-visual systems under a variety of environmental conditions. The "other" category of questions asked whether manufacturers who have installed rear visibility systems have experienced significant property damage prevention benefits, whether this area should be regulated by the federal government or the States, whether and how subcategories of vehicles should be defined, and whether existing commercial trucks in the applicable weight range should be required to be retrofitted with rear object detection systems.

    Comments on the ANPRM

    NHTSA received fourteen comments in response to the ANPRM, including submissions from trade associations, automobile and rear object detection system manufacturers, fleet operators, organized labor, a state agency, and individuals. [2] In addition to responding to the questions posed in the ANPRM, commenters also raised a variety of issues, including scope of the regulatory requirement, potential exclusions, alternatives to regulation, maintenance and training requirements, and preemption. The following discussion summarizes the comments received on the ANPRM.

    Scope and Exclusions

    NHTSA received a range of views regarding the scope of a regulation for rear object detection systems. Several commenters advocated narrowing coverage due to purported unsuitability of or lack of necessity for such systems on certain vehicles. For example, ATA stated that a "one-size-fits-all" approach would not be successful, because there is too much diversity in equipment and operations. NTEA stated that the rear object detection standard should only apply to "standard type vehicles."

    Commenters also offered numerous suggestions for vehicles which they believe should be excluded from the requirements of an amended standard, including tow trucks, car carriers, flat beds, stake trucks, dump trucks, tradesmen's and mechanic's bodies, platform bodies, tank trucks, any vehicle equipped with a crane or aerial device operating in a rotational manner, and other special units.

    Other commenters, such as NYDOT, urged NHTSA to expand coverage of the standard to include lighter vehicles commonly used in residential deliveries (e.g., trucks with a GVWR of 2,948 kg (6,500 lbs) to 7,258 kg (16,000 lbs)). NATC suggested that other vehicles with large blind spots (e.g., windowless vans and light trucks with campers or canopy shells) may also be suitable for coverage under a revised FMVSS No. 111. These recommended changes could bring some passenger vehicles within the scope of the rule. NYDOT suggested consideration of a phase-in period to permit earlier implementation of requirements for trucks, which can readily be equipped with existing technology.

    In response to the questions about retrofitting, commenters expressed divergent views. NYDOT urged NHTSA to take the lead on retrofitting of existing vehicles so that there would not be a patchwork of remedial activities by the 50 States. Others, such as ABC, opposed retrofitting, stating that retrofitting its entire fleet would be a "very lengthy and costly operation."

    Rearview Mirrors

    Regarding rearview mirrors, the commenters generally agreed with NHTSA's tentative determination that cross-view mirrors should be placed no more than 5 meters (16.40 feet) from the driver's side rear view mirror, as the image size arguably becomes too small beyond this distance to be useful to the driver. However, ATA urged greater clarity in how NHTSA would measure the distance between the two mirrors.

    Commenters also discussed the issue of trucks that are particularly long or high, thereby posing greater challenges in terms of rear object detection. For example, FedEx expressed concerns about situations where the height of a truck is so great that a top-mounted cross-view mirror is not visible in the side mirror. NYDOT stated that some trucks approaching 11,793 kg (26,000 pounds) may exceed the length where it would be feasible to use a cross-view mirror system, but it urged the agency to maintain some alternative rear object detection requirement for such vehicles.

    In the ANPRM, we requested comments on whether a 3 m by 3 m (9.843 feet by 9.843 feet) detection area behind a vehicle would be adequate. Some commenters suggested alternative detection zones that would be either larger or asymmetrical, but they did not provide a strong rationale or data to support their position. However, ATA and Ford suggested that NHTSA's estimation of the backing speed used to calculate the detection zone (i.e., 3 mph) underestimates actual backing speeds. These organizations stated that a reasonable estimate of backing speeds could be in the 5 mph to 8 mph range.

    Rear Video Systems

    Commenters likewise expressed a range of views on rear video systems. Some commenters, such as ABC, expressed concern about the expense of this technology. Other commenters, such as the Teamsters, specifically requested that NHTSA adopt a performance standard that would permit use of video systems.

    Reliant argued that the presence of a video camera may encourage theft (presumably of the camera), but NATC made the argument that video cameras and rear mirrors may deter theft of items from the back of the vehicle when stopped.

    In terms of the image presented by a rear video system, commenters suggested that a screen may be as small as 3.8 cm (1.5 inches) on the diagonal and as large as 25.4 cm (10 inches) on the diagonal. NATC stated that the size of the screen needed will depend upon the placement of the monitor relative to the driver's seating position. Reliant expressed concern about placing a video monitor in a truck's cab that is "already full".

    Varying views were expressed regarding screen color for rear video monitors. Mr. Silc stated that military-green monitors are more efficient than black-and-white monitors and that they provide three-times better contrast to the human eye and greater visibility. However, NATC reasoned that a black-and-white screen would be sufficient, because color would be lost in strong daylight and a black-and-white screen's contrast would be helpful in distinguishing objects and movement.

    Regarding placement for the video screen, one suggestion was to have the monitor in a location similar to a car's rearview mirror, where the driver's eyes can constantly be glancing at it. NATC urged NHTSA to conduct human factor analysis to determine the optimal placement of the monitor in the truck cab.

    NHTSA received conflicting viewpoints regarding the need for a system failure alert for the rear video system. Mr. Silc stated that it is unnecessary, arguing that if the screen is black, the system is either turned off or malfunctioning, and that either situation would be easily detectable by the driver. In contrast, the Teamsters supported use of a failure alert, expressing concern that the image of the monitor must reflect in real time the area behind the truck.

    In response to the ANPRM's questions about State laws regulating the existence and use of video monitors/screens in the occupant compartment that are in view of the driver, ATA stated that such restrictions are similar to those contained in Federal Motor Carrier Safety Regulation (FMCSR) 393.88. (That provision specifically prohibits monitors that are in view of the driver that can receive a television signal or can be used to view videotapes.)

    Audible Backup Alarms

    The ANPRM asked a number of questions regarding the efficacy of audible backup alarms and whether trucks equipped with OSHA-specified alarms should be excluded from the standard's new performance requirements. Some commenters such as NATC and ATA favored exclusion of such vehicles, arguing that audible backup alarms provide an effective warning for most pedestrians. As an added benefit, commenters stated that those systems are relatively easy to maintain.

    However, other commenters pointed out significant limitations associated with backup alarm systems. NYDOT stated that young children, who account for a disproportionate number of the fatalities and injuries related to backing crashes, might not understand or be able to properly respond to such alarms. Rostra stated that auditory backup alarms do not work adequately with the hearing impaired, and the Teamsters added that the elderly might also experience problems with such systems (e.g., due to decreased mobility and/or hearing impairment). Reliant added that these alarms can be turned off and that drivers may forget to turn them on again, and it also stated that residential customers frequently complain about loud backing alarms on trucks used at night.

    Other Non-Visual Rear Object Detection Systems

    Commenters expressed a range of views about the efficacy of a variety of non-visual rear object detection systems, such as those utilizing sonar and infrared technology. The Teamsters stated that manufacturers should be permitted to use non-visual systems as well as visual systems for rear object detection. NPTC argued that additional data are required on the effectiveness of devices other than mirrors, before such non-visual systems would be suitable as compliance options. Offering yet another possible approach, Federal Express confirmed that its vehicles are equipped with sonar backing systems used in concert with cross-view mirrors.

    Comments also were received regarding the timing of the alert and detection capabilities provided by non-visual systems. Rostra stated that detection time should be derived from the distance of a calibrated test object; the speed of the alert would depend upon the distance from the sensor and the vehicle's closing speed vis-à-vis the object. Ford stated that typical latency times for radar and ultrasonic systems are approximately 250-400 milliseconds (ms), but it added that a system's alert time could be increased by relying on multiple sensors to validate that the system is detecting a "true" target. In this context, commenters again raised concerns that NHTSA's assumption of a 3 mph backing speed may be an underestimation.

    Ford also stated that surface characteristics are very complex in the real world and that the reflective characteristics of irregular surfaces are infinite. Because of this inability of non-visual systems to detect all objects, Ford argued that NHTSA must specify a limited number of objectively defined obstacles for any certification test.

    Equipment Damage

    The ANPRM also asked questions about potential damage to various rear object detection systems. Some commenters, such as Reliant, argued that mirrors are high maintenance items due to breakage and theft. Others suggested that damage inflicted by dirt, mud, rocks, brush, and limbs could limit the mirrors' effectiveness. ATA stated that while rear detection systems could be damaged by vibration and shock, it believes that these systems could be designed to withstand most of these conditions.


    In response to questions about test procedures for the potential new rear object detection provisions, commenters generally urged NHTSA to conduct testing under as many different conditions as possible under which objects would be difficult to detect. Regarding mirrors, NATC stated that test procedures should utilize objects of various sizes, colors, heights, and positions, and the organization urged NHTSA to conduct testing under rugged conditions (e.g., vibration, humidity, and extreme high and low temperatures).

    For non-visual rear object detection systems, commenters stated that a well-defined and objective standard and test methodology are even more important, including specification of the size and shape of objects to be detected in such tests. Ford suggested use of the standard pole target developed by the International Organization for Standardization (ISO), which the company has used since 1996 for testing both its ultrasonic and radar systems. Furthermore, both Rostra and NATC stated their belief that environmental conditions should be specified as part of any performance test for non-visual systems under the standard. Factors such as temperature, rain, snow, humidity, dirt, driving surfaces, submersion, and mounting surfaces were specifically mentioned as potentially affecting such systems' detection capabilities.


    III.  Target Population
    Table of Contents

    Applicable Vehicles

    The agency is proposing to limit the application of potential Federal requirements in this area to straight trucks with a GVWR between 4,536 kg (10,000 pounds) and 11,793 kg (26,000 pounds), and also only for new trucks (i.e., older trucks would not need to be retrofitted). The lower bound of this weight range is based on FARS data (see Table III-4), which shows that the rate of fatal backing crashes for these vehicles is substantially greater than that of vehicles with lower GVWR's. Although buses are somewhat similar to straight trucks, FARS data show only one fatal backing crash for buses over a seven-year period. This is most likely due to the planned routes for buses, which are designed to minimize or eliminate backing. For this reason, buses are not included in the proposal.

    The upper bound of 11,793 kg is based on the agency's belief that it represents the maximum weight of a typical straight truck. FMVSS No. 111, however, defines requirements for a narrower weight class of between 4,536 kg and 11,340 kg (25000 lbs). The agency is requesting comments on the proposed upper bound. Specifically, are there straight trucks greater than 11,340 kg that might benefit from a rear object detection system?

    The agency raised the issue in the ANPRM about whether something should be done to improve the situation for drivers of passenger vehicles, which are involved in a larger number of fatal backing crashes. Cross-view mirrors present practical problems for cars and light trucks because of the size of the mirror needed relative to the size of the vehicle. It would be difficult to mount the mirrors high enough on cars and most light trucks so that the mirrors would not be a hazard to pedestrians and cyclists. It is unlikely that the public would accept a cross-view mirror due to the aesthetic and operational problems it would create. Lastly, light vehicles are not over represented in backing crashes. For these reasons, the agency is not pursuing a mirror solution for passenger vehicles. The more expensive video-based system is more likely to be an acceptable solution.

    In the ANPRM [3], the agency asked for comment on the appropriateness of applying such a standard to straight trucks with a GVWR of between 4,536 and 11,793 kg. We received a number of comments on this issue. The majority of these comments asked for exemption of certain types of trucks such as flat beds, stake bodies, dump trucks, tow trucks, the common light duty pick-up truck bed, and other high-cube or full-size van applications like tradesmen or mechanics bodies. Many of these vehicles either have body styles which do not allow them to have cross-view mirrors installed in an effective position or they are used in a rugged environment which would cause damage to the mirrors or other systems and thus would require frequent replacement or repair.

    ABC stated that only trucks with a cube-style cargo box should be covered, thus exempting flat bed trucks from the requirements. The National Private Trucking Council (NPTC) stated that the vast disparity in body types of straight trucks between 4,536 and 11,793 kg GVWR makes it difficult, if not impossible, to promulgate a standard that is appropriate to all vehicles. Federal Express, ATA, and TRAA all stated that cross-view mirrors or other devices may not be operationally feasible on flat beds, dump trucks, stake bodies, pickup trucks, and other special units. TRAA also included tow trucks and car carriers in its list of vehicles that it believes should not be covered. Ford stated that if the aforementioned types of vehicles are not exempted from the requirements, they should be allowed to utilize means other than cross-view mirrors to meet the standard. NTEA stated that dump trucks, platform bodies and any vehicle equipped with a crane or aerial device should be exempted.

    Conversely, NYDOT believes that there are a number of delivery-type vehicles that are lighter than 4,536 kg GVWR. As a result, it believes that these vehicles should be covered by the proposed requirements. It stated that many vehicles that enter residential areas range in weight from about 2,948 kg (6,500 lbs) to 7,257 kg (16,000 lbs) GVWR. It believes that by limiting the applicability to vehicles between 4,536 and 11,793 kg GVWR, the agency would be moving away from the intent of the original petition from Mr. Norton.

    The agency agrees with these commenters that some vehicles may not be suited for the proposed systems. For instance, in rugged environments that most dump trucks are used, any system that was installed would probably be damaged rather quickly. Dump trucks, as well as other work vehicles used off-road for a portion of the time, can experience more vibration than those used solely on the road. This not only could damage the system, but it could render it ineffective by misaligning it. Cross-view mirrors and video cameras could be vibrated out of their aimed positions. Other vehicles, such as stake bodies, tow trucks, and flat beds, may have no viable place to mount such a system. To help to better define the applicability of the proposed standard, as well as to eliminate certain vehicles on which such systems would definitely not be feasible, the agency offers the following proposals.

    Vehicles that we believe should be covered by the standard are those with cargo boxes mounted on their chasses. These vehicles can be used to deliver packages and other items, often to private residences. An example of this type of vehicle, a diaper delivery truck, was originally suggested by Mr. Norton in his petition. Further, the States of New York and Washington have applied their regulations specifically to these vehicles. Because of their constant presence in residential areas, we believe it is important that any final rule include them in the applicability. Comments are requested on better defining this group of vehicles.

    Vehicles that we believe have the potential of being covered are dump trucks and tank trucks. Commenters to the ANPRM made it clear that the rugged environment in which dump trucks are sometimes used would likely damage any system installed on the back of the vehicle. If a damaged rear vision system must be replaced on a regular basis, it would make the cost of the regulation too high. However, we are aware, through OSHA's website, that there are a number of fatalities caused by backing dump trucks. Because of this, a more rugged system might need to be developed to withstand abuse and give the drivers of these vehicles the necessary rear vision. Tank trucks used for delivery of home heating fuel, water for pools, and other liquids, while usually not used in a rugged environment, have a different problem. The design of these vehicles leaves little possibility that a mirror system could be used. The curvature of the tank may not allow the mirror to be mounted in a place where it will be effective. However, it would seem that a video system could be substituted. Comments are requested on the feasibility of mounting a mirror or video system on these vehicles.

    The final category is those vehicles that probably should not be covered. This group includes flat beds and stake bodies. These vehicles have no place to mount a mirror and have a limited number of places where a camera could be mounted. In addition, because these vehicles usually have an unobstructed view to the rear depending on their cargo, we are requesting comments on whether they should be covered. However, in the interim, these vehicles will not be excluded pending comments.

    Regarding NYDOT's comment, the agency agrees that vehicles that enter residential areas for purposes of delivery or some other type of commerce may weigh less than 4,536 kg GVWR. However, the data that we have do not clearly indicate that these vehicles are over represented, even though many do make deliveries. Many of these vehicles are also configured as passenger carrying vehicles that may not have the same rear visibility problems as larger vehicles. Consequently, the data do not identify these vehicles as being of the same risk as the larger vehicles. However, comments are requested on this issue. Therefore the proposed applicability of the rule will be for all straight trucks between 4,536 kg (10,000 pounds) and 11,793 kg (26,000 pounds) with no specific body types being excluded.

    In estimating the applicable vehicle population, Polk data from 2000 to 2002 were employed to determine the number of new registrations of straight trucks between 4,536 kg and 11,793 kg. The population data is presented below in Table III-1.

    Table III-1
    Polk registration data from calendar year 2000 to 2002 used to estimate new truck sales
      Model Year 10,000 to 14,000 lbs 14,001 to 16,000 lbs 16,001 to 19,500 lbs 19,501 to 26,000 lbs Total
    Polk 2002 2001 179,109 59,735 26,996 57,897 323,737
    Polk 2001 2000 156,072 79,512 37,496 92,560 365,640
    Polk 2000 1999 190,138 84,650 44,259 87,091 406,138
              Average = 365,172

    We used the model year (MY) of the year previous to that of the current Polk data year (e.g., for Polk 2002, we used MY 2001) because the previous year represented registrations for an entire calendar year, whereas the data that was the same year as the Polk data (e.g., Polk 2002, MY 2002) only included vehicles up to July 1st, and therefore was not a complete calendar year. Between 2000 and 2002, we found that an annual average of 365,172 straight trucks between 4,536 kg and 11,793 kg were registered. However, these 365,000 straight trucks represent an upper bound since some trucks will already have a rear detection device.

    The NPTC reported in their comments to the ANPRM that private carriers comprise the largest segment of the trucking industry accounting for 82 percent of the medium and heavy-duty trucks registered in the United States. Since data are not available regarding installation rates for privately owned trucks, we assume that they do not have any type of rear detection system. The remaining 18 percent of vehicles are comprised of fleet vehicles, which are owned and operated by large companies, such as FedEx, UPS, or Waste Management. We assumed that, being corporate vehicles, all their trucks have a rear detection system. Therefore, 100 percent of 18 percent or 65,700 (1.00 ´ 0.18 ´ 365,000) vehicles have a rear detection system. The remaining 299,300 vehicles are estimated to not be currently equipped with these systems. NHTSA requests comments on this estimate.


    For the ANPRM, the agency gathered data on the annual number of incidents of people being backed over by a motor vehicle of any type or size. The Fatality Analysis Reporting System (FARS) contains data on all fatal traffic crashes within the 50 states, the District of Columbia, and Puerto Rico. The data system was designed to assist the traffic safety community in identifying traffic safety problems, developing and implementing vehicle and driver countermeasures, and evaluating motor vehicle safety standards and highway safety initiatives. For the ANPRM, FARS data were searched and an average of 85 backing fatalities for all vehicle types was found for the years 1992 and 1993. However, by design, a fatality is included in the FARS database only if a motor vehicle is involved in a crash while traveling on a roadway customarily open to the public. This excludes other likely scenarios for backing fatalities, such as events where someone is backed over in a driveway, parking lot, or in a workplace such as a warehouse or a construction site.

    To ascertain the number of off-road backing fatalities, the agency worked with the National Center for Health Statistics (NCHS) to gather data on these incidents. NCHS obtained information on the cause of death, as recorded on individual death certificates, from each of the 50 states, the District of Columbia, and the five boroughs of New York City. NCHS and FARS data for 1992 and 1993 were used in this study to obtain average annual estimates of the number of fatalities associated with off-road and on-road fatal backing incidents. Since death certificates contain only general information as to the mode of death, it was not possible to determine the exact number of off-road fatalities because the death certificate did not explicitly state that the cause of death was by a backing vehicle. This work is described in a February 1997 Research Note prepared by the agency entitled "Nonoccupant Fatalities Associated with Backing Crashes," which is found in the public docket under NHTSA-98-4308-3. It identified an annual average of 475 backing fatalities (85 on-road and 390 off-road) for 1992 and 1993 for all vehicle types. The 1997 Research Note placed particular focus on backing crashes involving children aged one to four. This was done due to their smaller physical stature and comprehension level that places them at greater risk in situations involving backing motor vehicles. These children were found to be over represented, relative to population, in backing crashes, particularly those occurring off-road. That is, children aged one to four represent six percent of the total U.S. population, but accounted for 30 percent of all off-road backing fatalities and 16 percent of all on-road backing fatalities.

    In updating the injury and fatality data for this proposal, we attempted to verify the on- and off-road information presented in the 1997 Research Note. However, we were unable to determine the exact methodology and assumptions used by the original researchers. Thus, while the underlying data set still exists, it was not possible to reproduce and confirm the results presented in that study. For this reason, we have decided not to rely on the results reported in the 1997 Research Note as part of this rulemaking.

    New Supporting Fatality Data:  In late 2002, the agency obtained death certificate data from 1998 highlighting non-crash event backing incidents and compiled this data. The death certificate information was individually examined for related information to ascertain the magnitude of the off-road fatalities discussed later in this section. As of May 2004, the death certificate study is complete. This study indicates that 91 fatalities occurred in 1998 due to backing vehicles.

    For the ANPRM, the agency found that from 1991 to 1997 an average of approximately 13 fatalities per year were attributable to straight trucks backing up. To verify the accuracy of the 1991 to 1997 data, a query of 1999 FARS data revealed 58 fatalities caused by backing vehicles, with 12 identified as being attributable to straight trucks. Again, the older data from the ANPRM is comparable with the 1999 data and will be used for the analysis.

    Workplace Fatalities/Injuries:  The agency is also concerned with fatalities and injuries that occur in the workplace. A casual observation of trucks used for deliveries and other jobs in the Washington D.C. area shows that many are already equipped with rear object detection systems. The United Parcel Service (UPS) said publicly that it intended to install video monitoring systems in all 65,000 of its delivery trucks by October 2001. The United States Postal Service (USPS) and the Potomac Electric Power Company (PEPCO) have equipped their vehicles with cross-view mirrors. Federal Express stated in its comments that it has installed sonar-based rear object detection systems on its vehicles in addition to cross-view mirrors. While their concern with backing crashes that occur on public and private roads was undoubtedly a significant factor that led the companies to install these systems, we understand that injuries and fatalities that occur in loading and parking areas at their headquarters were also a factor. For example, we are aware of workers in these areas that have been crushed into loading docks by backing vehicles. On the Occupational Safety and Health Administration's (OSHA) website, there are 15 fatalities with the cause listed as being crushed between a backing vehicle and the dock. In general, there are over 50 reports of workers being killed by backing vehicles.

    Another area of concern is construction sites. OSHA has requirements for backup alarms on vehicles and equipment used in construction to address this issue. A vehicle must be equipped with such an alarm unless the vehicle is backed up only when an observer signals that it is safe to do so. These requirements are contained in 29 CFR Part 1926 - "Safety and Health Regulations for Construction".  In the NPRM we will request information on the effectiveness of these alarms. We contacted OSHA and found that it has done no studies to determine the effectiveness of these alarm systems.

    Some backing crashes that occur in the workplace may go unreported to police as they may be handled privately by the businesses involved. These businesses may be reluctant, from a public relations standpoint, to report these injuries and fatalities. They would, therefore, not be found in FARS or NCHS data. In the NPRM, the agency will request comments on the prevalence of such workplace incidents. While these companies may not be inclined to report these data, other entities such as insurance companies, unions, and rescue personnel may do so.

    Vehicle Type Involvement in Backing Crashes: The ANPRM also attempted to determine the involvement of specific types of vehicles in pedestrian and pedalcyclist backing fatalities for both on and off-road. However, unlike the FARS data, the NCHS data collected from death certificates does not record the vehicle type involved.

    For the on-road vehicle involvement, the FARS data show the following for 1991 to 1997 pedestrian and pedalcyclist backing fatalities:

    Table III-2
    Cumulative Number of Pedestrian and Pedalcyclist Fatalities in On-Road Backing Crashes (FARS Data from 1991 to 1997)
    Vehicle Type Number of Fatalities
    Passenger car 129
    Light truck/van 139
    Bus 1
    Straight truck over 4,536 kg GVWR 81
    Unknown truck over 4,536 kg GVWR 12
    Combination truck 15
    Other 2
    Unknown 2
    Total 381

    The agency estimated from these data that, from 1991 to 1997, straight trucks were involved in 91 of the 381 backing fatalities. To get the total of 91 fatalities, 10 of the fatalities from the two unknown categories were added to the 81 fatalities attributed to "Straight trucks over 4,536 kg GVWR". This is based on a number of assumptions. We distributed the "Unknown truck over 4,536 kg GVWR" category proportionally. Therefore, straight trucks would be responsible for 10 (81/[15 + 81]) ´ 12) of the 12 fatalities in that category, leaving two fatalities that can be attributed to combination trucks. Next, we distribute the two fatalities in the "Unknown" category. Since straight trucks caused 24.01 percent (91/379) of the total number of fatalities, 0.48 (0.2401 ´ 2) of the two fatalities in the "Unknown" category was attributed to straight trucks, for a total of 91.48 (81 + 10 + 0.48), which was rounded down to 91 fatalities. Thus, straight trucks over 4,536 kg are responsible for 24 percent (91/381) of backing fatalities and we will be referring to this figure throughout the analysis.

    As stated previously, the 475 fatalities from the 1997 Research Note include off-road fatalities figures that we were unable to reproduce in a later attempt. The Research Note stated that an annual average of 85 on-road and 390 off-road fatalities occurred in 1992 and 1993 for all types of motor vehicle types. Even though the agency was unable to reproduce the off-road figure from the Research Note in a more recent attempt, we believe that the off-road fatalities should not be disregarded as suggested in the ANPRM.

    In addition, after reviewing the available literature regarding backing fatalities, it became apparent that a single national value was not available. Review of the death certificate study showed that the study could not be extrapolated to a national level, since it did not carry any weighting factors nor was it meant to be a statistically significant study. Also, abstracts of research articles reviewed in the death certificate study were also reviewed but did not make any mention of a national estimate of backing fatalities and were primarily concerned with children. Therefore, to better represent the magnitude of total fatalities, we will derive a value for the number of off-road fatalities based on available data that we feel are representative and present this methodology below.

    Determination of Off-road Fatalities: The number of on-road fatalities was gathered utilizing the FARS data found in Table III-2. The 381 fatalities represent the total number of on-road fatalities occurring over the 1991 to 1997 time span by all types of motor vehicles. To get an annual average we divide 381 by 7 (years) to arrive at about 54.43 on-road backing fatalities per year.

    Although we were unable to reproduce the off-road fatalities from the Research Note, we believe that it is a significant portion of the total fatalities and should not be excluded. Therefore, we will be using the fatalities from the 1991 to 1997 FARS data and the 2004 death certificate study to make estimates as to the number of off-road fatalities.

    The death certificate study did not explicitly state what was on- and off-road, therefore, we made reasonable assignments based on the reported location as to whether the location was public (on-road) or private (off-road) and presented it in Table III-3.

    Table III-3
    Pertinent data elements from the 2004 death certificate study
    Location Number of Victims On- or Off-Road Sub-Total Percentage
    Of Total
    Driveway 21 Off 76 83.52%
    Home 21 Off
    Parking Lot 21 Off
    Other Off-Road 13 Off
    Road/Street 13 On 15 16.48%
    Sidewalk 2 On
    Total 91   91 100.00%

    Given that 54.43 fatalities occur on-road, we now attempt to estimate the number of off-road fatalities. Using the "On-road" sub-total from Table III-3 in which 16.48 percent of fatalities occurred on-road combined with the approximately 54 fatalities from FARS, we reasoned that a total of about 330 (= 54 ¸ 0.1648) backing fatalities occur each year. Since 330 represents the total number of backing fatalities, we multiply by 83.52 percent (the percentage of off-road fatalities calculated from the death certificate study) to get 276 off-road fatalities. Lastly, since 24 percent of backing fatalities are caused by straight trucks, we multiply 54 and 276 by this value to arrive at the on- and off-road fatalities, respectively, caused by backing straight trucks, 13 and 66. Summing these values we estimate that about 79 fatalities occur annually due to backing straight trucks on public and private roadways.

    To solidify the position that the problem was most acute for straight trucks, the number of registered vehicles in the national fleet and the annual miles driven were considered to calculate the rate of backing fatalities for different vehicle types. The following breakdown of on-road fatality rate information is based on cumulative 1991 to 1997 data:

    Table III-4
    Rate of on-road fatal backing crashes (cumulative FARS data from 1991 to 1997)
    Vehicle Type Pedestrians and Pedalcyclists Killed by a Backing Vehicle per Million Registered Vehicles Pedestrians and Pedalcyclists Killed by a Backing Vehicle per 100 Billion Vehicle Miles Traveled
    Passenger cars 1.05 1.26
    Light trucks/vans 2.32 2.80
    Combination trucks 9.94 2.21
    Straight trucks over
    4,536 kg GVWR
    29.68 21.89

    It should be noted that two errors were found in the values reported in this column in the ANPRM. In the ANPRM, the number of registered vehicles used for each calculation was mistakenly summed over the 1991 to 1997 time span. Instead, an average of the number of registered vehicles should have been calculated for the seven-year span. As a result the numbers originally reported were one-seventh of their correct value. Also, the number of registered straight trucks with a GVWR of over 4,536 kg was overstated. Instead of 4.9 million vehicles, a value of 3.1 million vehicles should have been used. Correcting this error significantly strengthens the argument that straight trucks are over represented in pedestrian and pedalcyclist related fatalities.


    The agency also analyzed injury data. While these data provide a good picture of the general magnitude of backing injuries, the data in the NEISS and NASS-GES databases overlap and, therefore, cannot be added to determine the annual total number of these injuries.

    The first source of injury data was from NEISS. The U.S. Consumer Product Safety Commission (CPSC) has operated this statistically valid injury surveillance and follow-back system for almost thirty years. The primary purpose of NEISS has been to provide timely data on consumer product-related injuries occurring in the U.S. NEISS injury data are gathered from the emergency departments of 100 hospitals selected as a probability sample of the more than 5,300 U.S. hospitals with emergency departments. Surveillance data enable CPSC analysts to make timely national estimates of the number of injuries.

    The July 2000 to June 2001 NEISS file showed 139 cases in which a pedestrian or a pedalcyclist was injured by a backing vehicle. These are the first data since NEISS expanded to include other injuries such as those sustained in motor vehicle crashes. This twelve-month data sample translates into a national estimate of 7,419 injuries, of which, 148 of these injuries were attributed to straight trucks.

    Finally, we examined data from GES. GES data include only injuries incurred in police-reported incidents and, therefore, overlaps the NEISS data, which records both police-reported incidents and those that are not reported. In addition, the GES data on backing related injury crashes is probably not representative of all backing related injury crashes, because it does not include information about injuries from backing maneuvers on private areas such as driveways, parking lots, and work sites. However, we assume that the proportion of backing injuries attributable to straight trucks in the GES data to be the same in the broader NEISS data, it would translate into straight trucks being responsible for approximately 2 percent (148 ¸ 7,419) of the total injuries.

    Although NEISS provides the number of injuries due to backing straight trucks, to get a better estimate of the injury distribution of the non-fatal backing crash problem, the police-reported injury levels (A, B, C, and Unknown) in the 1996-2000 GES data were utilized and presented in Table III-5. Table III-5 represents the total number of injuries from the 1996 to 2000 time span for all types vehicles.

    Table III-5
    1996-2000 NASS-GES backing crash injury data
    Injury Type Number Percent of Total
    A-injury 285 13.25%
    B-injury 675 31.38%
    C-injury 1,132 52.63%
    Unknown 59 2.74%
    Total 2,151 100.00%
    A-injury = Incapacitating injury
    B-injury = Nonincapacitating injury
    C-injury = Possible injury
    Unknown = Unknown if injured

    Since the NEISS injury data does not directly state the MAIS level, we performed two distributions to arrive at the MAIS injury distributions. (AIS is an anatomically based system that classifies individual injuries based on anatomical injury on a six point ordinal scale of risk to life, and MAIS is simply the maximum injury level(s) an individual receives).

    First, we assumed that the NEISS injury distribution was similar to GES and combined the 148 injuries from NEISS with the percentage each injury level contributes to the total from GES (Table III-5, "Percent of Total" column) to distribute the injuries to the KABCO scale. To calculate the approximate number of A-injuries in NEISS, we multiplied the total number of NEISS injuries by the percent of A-injuries in GES or 148 ´ 13.25%; this is repeated for B-, C-injuries, and Unknowns. Table III-6 illustrates the redistributions. It is important to point out that GES only accepts incidences that occurred on public ("on-road") roadways such as highways, sidewalks, thoroughfares, etc. whereas a private ("off-road") roadway would be considered a farm road, driveway, cargo dock, and so on. Since NEISS is hospital based, their data collection is not restricted to just public roadways but also includes private ones. Thus, we assumed that the speed and severity of backing crashes were similar in both on-and off-road.

    Table III-6
    Average annual NEISS injury data distributed according to NASS-GES data
    Injury Type Number
    A-injury 19.61
    B-injury 46.44
    C-injury 77.89
    Unknown 4.06
    Total 148.00

    Even though the "Percent of Total" in Table III-5 is over a five-year span, an implicit assumption is that the average annual number of backing injuries and their respective distributions remain constant. Also, keep in mind that 148 injuries is an annual number from NEISS, since it comes from 2 percent of the annual number of backing injuries or 7,419.

    Given the KABCO distribution, we redistribute it once more to the MAIS scale using a distribution table from the report entitled, "Ejection Mitigation Using Advanced Glazing: Status Report II" (Table 7.2 from the report). The data set encompasses the entire NASS-CDS data pool from 1982 to 1986 with injuries sustained by motor vehicle passengers and pedestrians. The final KABCO to MAIS distributions are presented below in Table III-7a, whereas the calculations and methodology can be found in the appendix.

    Table III-7a
    KABCO to MAIS injuries distributions
      KABCO Scale Total
    A B C UNK
    MAIS-0 0.30 2.29 15.51 3.31 21.42
    MAIS-1 9.64 36.80 55.87 0.65 102.96
    MAIS-2 5.47 5.80 5.27 0.07 16.61
    MAIS-3 3.28 1.40 1.18 0.03 5.88
    MAIS-4 0.57 0.12 0.05 0.00 0.74
    MAIS-5 0.35 0.03 0.01 0.00 0.39
    Total 19.61 46.44 77.89 4.06 148.00

    Even though this calculation procedure estimates 21.42 MAIS-0 injuries (i.e., no injuries), since the data came from a hospital-based survey, we assume all of the patients brought to the hospital were injured. Thus, we redistributed MAIS-0 injuries into the MAIS-1 to -5 categories. Table III-7b summarizes the injury distributions that will be used in this analysis.

    Table III-7b
    Annual MAIS injuries due to backing crashes caused by straight trucks
    MAIS Level Number
    1 120.38
    2 19.42
    3 6.88
    4 0.87
    5 0.46

    Target Population

    Table III-8 summarizes the vehicle target population, injuries, and fatalities that will be used in this analysis.

    Table III-8
    Target population
    Applicable Vehicles 299,300
                MAIS-1 120
                MAIS-2 19
                MAIS-3 7
                MAIS-4 1
                MAIS-5 0
    Fatalities 79

    Cross-view Mirror and Video Camera Test Procedures

    The requirements of this rule will utilize cylinders of a color that provides a high contrast with the surface on which the vehicle is parked, and will be 305 mm (12.00 inches) high and 305 mm in diameter. The cylinders shall be placed in a 3,000 mm (118.11 inches) by 3,000 mm square configuration in the rear of the truck. Refer to Figure 1 for a diagram of the cylinder placement. The driver's eye location is the eye location of a 25th percentile adult female, when seated in the driver's seat.

    If cross-view mirrors are used, the geometric center of the cross-view mirror must be no more than 5,000 mm (196.85 inches) from the geometric center of the outside rearview mirror on the driver's side. In addition, it must be adjustable in both the horizontal and vertical directions and have an average radius of curvature of no less than 203 mm (8.0 inches). If a video system is utilized, the system must provide an image size of not less 90 cm2 (13.95 in2) and not more than 160 cm2 (24.8 in2), and the image must be reversed to show the scene as if it were viewed through a rearview mirror. The monitor must be mounted as close to the centerline of the vehicle as practicable near the top of the windshield, but located such that the distance from the center point of the eye location of a 25th percentile adult female seated in the driver's seat to the center of the monitor is no more than 1,000 mm (39.37 inches).

    For either system to be considered compliant, the entire top surface of all test cylinders must be visible.

    Figure 1. Location of test cylinders.


    IV.  Benefits
    Table of Contents


    This section estimates the potential injuries mitigated and lives saved if a cross-view mirror or camera vision system is installed. In order to calculate the benefits, the effectiveness of a device is needed. Two commenters [4] to the Request for Additional Information stated that in 1984 FedEx performed a pilot study in four hub cities and found that backing incidents were reduced by 33 percent when cross-view mirrors were installed. Thus, 33 percent will be the effectiveness rate that we will be using for cross-view mirrors.

    As for camera systems, neither commenters nor the agency had a quantitative value for its effectiveness. Even though we believe that a camera's effectiveness would be higher than that of a mirror system, we will be using the effectiveness of cross-view mirrors as a proxy for the camera system. However, the USPS provided the agency with anecdotal evidence supporting the claim that cameras are more effective than mirrors. In fiscal year 2003 (FY2003), they equipped two different truck models (between 4,536 kg, and 11,793 kg) for a total of 2,041 trucks with a rear-view camera system to test their efficacy. In FY2003, the trucks traveled a combined 3,432,925 miles with no reported backing incidences. As a comparison, in FY2003 USPS had 7,136 medium to heavy trucks with cross-view mirrors, which traveled a combined distance of 46,343,010 miles and experienced a total of 116 backing instances (2 involved pedestrians whereas the remaining 114 were non-pedestrian contact). This results in an overall average backing incidence rate of about 0.250 per 100,000 miles traveled. It is important to reiterate that these "incidences" encompass hitting pedestrians and also other vehicles, buildings, and so on. Therefore, the 0.250 incidence rate would be slightly lower if pedestrian contact was removed.

    Given the above data, this would imply a 100 percent effectiveness for cameras. However, in the real world, this is highly unlikely for any counter-measure. In addition, the camera equipped trucks represent a fraction of the medium to heavy trucks in the USPS fleet that contributed a small portion of the total miles traveled. Nevertheless, the potential of camera systems to be more effective than cross-view mirrors is suggested by this small sample.

    Lives Saved and Injuries Prevented

    Given the effectiveness of cross-view mirrors and camera systems are both assumed to be 33 percent, we now estimate the number of lives saved and injuries prevented if a rear detection system is installed. The number of fatalities and injuries mitigated is simply the number of fatalities/injuries times the effectiveness of the counter-measure. In addition, we assumed that if the vehicle does not back into a person, then no fatality/injury occurs, rather than the fatality/injury being mitigated. For example, suppose a person is struck by a backing straight truck and suffers an MAIS-4. Now, suppose that the truck was equipped with a mirror system and avoids the individual. The injury is not reduced to an MAIS-3 or -2 but rather no injury occurs since no contact occurs. This assumption is also carried over into fatalities. However, before we determine the net lives saved and net injuries prevented, we must first determine the injury and fatality levels when there is no counter-measure present in the fleet in order to establish a baseline.

    Equation 1 allows us to back-calculate in order to estimate the number of injuries and fatalities that may result when there is no camera or mirror system in the fleet:

    Equation 1 allows us to back-calculate in order to estimate the number of injuries and fatalities that may result when there is no camera or mirror system in the fleet                                     Equation 1


    Recall from earlier that we assumed a combined installation rate of 18% for both mirror and camera systems, and, thus, u is 18% in this case. The effectiveness, e, of either a mirror or camera system is 33%. To determine the number of, say, MAIS-1 injuries that occur when there are no backing counter-measures in the fleet, we take the current number of injuries (from Table III-8) and divide by one minus the product of the usage rate and effectiveness or 120/[1-(0.18 ´ 0.33)] and arrive at about 128 MAIS-1 injuries. In other words, if the current fleet of straight trucks with a GVWR between 4,536 kg and 11,793 kg had no backing counter-measure installed, then we would expect about 128 MAIS-1 injuries to occur. A similar calculation is applied to the remaining MAIS-2, -3,  -4, and -5 injuries and fatalities. We present the results below in Table IV-1a in the column labeled "Baseline Population."

    Table IV-1a
    Baseline population and current benefits
      Baseline Population Target Population = Current Benefits
    MAIS-1 128   120   8
    MAIS-2 21   19   1*
    MAIS-3 7   7   0
    MAIS-4 1   1   0
    MAIS-5 0   0   0
    Fatals 84   79   5
    * Inconsistencies are due to rounding error.

    In addition, taking the difference between the values in the "Baseline Population" column and the current number of injuries and fatalities ("Target Population" column) results in the number of lives saved and injuries prevented that is currently occurring at an installation rate of 18 percent ("Current Benefits" column).

    The product of the effectiveness (33 percent) and the number of injuries/fatalities in the "Baseline Population" results in the lives saved and injuries prevented if a mirror or camera system is installed on every straight truck (100 percent installation rate) with a GVWR between 4,536 kg and 11,793 kg. The results are presented in Table IV-1b in the column labeled "Total Lives Saved & Injuries Prevented".

    Table IV-1b
    Baseline population and total benefits
      Baseline Population ´ Effectiveness = Total Lives Saved & Injuries Prevented
    MAIS-1 128   0.33   42
    MAIS-2 21   0.33   7
    MAIS-3 7   0.33   2
    MAIS-4 1   0.33   0
    MAIS-5 0   0.33   0
    Fatals 84   0.33   28

    Lastly, because approximately 18 percent of the current fleet has a counter-measure installed, there currently exists some benefit attributable to these systems (i.e. the current benefits). Since we are only interested in the potential benefits if the remaining 82 percent of trucks were required to install a counter-measure, we subtract the "Current Benefit" figures from the "Total Lives Saved & Injuries Prevented" to arrive at the number of injuries and fatalities that may be mitigated if the remaining 82 percent trucks installed a rear-detection system or the net benefit.

    Table IV-1c
    Total lives saved and injuries prevented and net benefits
      Total Lives Saved & Injuries Prevented Current Benefits = Net Benefits
    MAIS-1 42   8   35*
    MAIS-2 7   1   6
    MAIS-3 2   0   2
    MAIS-4 0   0   0
    MAIS-5 0   0   0
    Fatals 28   5   23
    *Inconsistencies are due to rounding error.

    Property Damage Savings

    We believe cross-view mirrors and camera systems will also reduce property damage resulting from backing into loading docks, poles, parked vehicles, and so on. Although the agency does not have an estimate of the average annual property damage attributable to backing incidents caused by straight trucks, we attempt to make estimates based on the data in this and past economic analyses.

    We utilize the FY2003 USPS data to quantify the incidence rate of property damage. Given that 116 backing incidences occurred (with USPS straight trucks between 4,536 kg and 11,793 kg), we remove the 2 incidences that involved pedestrians to arrive at 114 incidences that involved property damage only. Since the fleet of 7,136 trucks (all equipped with a cross-view mirror) traveled a combined 46,343,010 miles in FY2003, we estimate that the incidence rate of property damage to be about 0.2460 incidences per 100,000 miles (114 ¸ 46,343,010 ´ 100,000). In other words, we estimate that even if 100 percent of the applicable vehicles had a rear-detection system, there would still be 0.246 incidences per 100,000 miles traveled (referred to as I100%).

    Based on data from the 1997 Vehicle Inventory and Use Survey (VIUS), the average truck on the road between 4,536 kg and 11,793 kg travels 13,831 miles per year (however, VIUS does not indicate the lifetime miles traveled). Data from NHTSA's analysis of the 25-year lifetime of light trucks (those less than 4,536 kg) shows a weighted average of 10,247 miles traveled per year or a lifetime mileage of 153,706 miles. Using the light truck data, we assume that trucks between 4,536 kg and 11,793 kg travel approximately 35 percent more than light trucks ([13,831 – 10,247] ¸ 10,247), or a lifetime vehicle miles traveled of 207,503 (153,706 ´ 1.35).

    In the agency's publication The Economic Impact of Motor Vehicle Crashes, 2000, the property damage caused by and to a vehicle was taken to be $1,484 per incident (Table 2, p. 9, "PDO" column). We assume that the cost of backing incidences is not as expensive as the average property damage crash, because the speed is lower. We assume the cost to be about a third of the average property damage crash and therefore contribute a third of the cost or about $495 per incident. Since the cost figure is in 2000 dollars, we apply a gross domestic product (GDP) implicit price deflator factor of 1.07 (using year 2000 as a base) to bring it up to 2004 dollars (note, at the time this report is written, 2004 was not complete and only the first quarter price deflator was available, and therefore, used). Multiplying $495 per incident by 1.07 yields a cost of $530 per incident in 2004 economics.

    Prior to determining the savings gained from a rear detection system, we must first determine the incidence rate in which there is no counter-measure to establish a baseline. This value is determined from the equation:

    Equation 2. Prior to determining the savings gained from a rear detection system, we must first determine the incidence rate in which there is no counter-measure to establish a baseline.                                     Equation 2


    Given that I0 is 0.2460 incidences per 100,000 miles, u is 1.00 and e is 0.33, I is calculated to be 0.3672 incidences per 100,000 miles. In other words, we estimate that, on average, 0.3672 incidences of property damage would be caused by a backing straight truck for every 100,000 miles it is driven if no rear-detection is installed on any truck in the fleet. However, recall that 18 percent of the current applicable fleet has a rear-detection system and therefore the incidence rate will be slightly lower than 0.3672 incidences per 100,000 miles. The backing incidence rate at an 18 percent installation rate is calculated by taking the product of 0.3672 incidences per 100,000 miles and one minus the product of the usage rate (u) and effectiveness (e) or [0.3672 ´ {1 – (0.18) ´ (0.33)}], which results in a value of 0.3453 incidences per 100,000 miles.

    We assume that the benefits accrued by one model year of single trucks over its entire lifetime is equal to the benefits accrued by an entire fleet of trucks for a single year. In addition, we estimate that a heavy truck would travel about 207,503 miles in its lifetime. Next, we take the incidence rates at the per 100,000 miles traveled rate and adjust them up to the lifetime miles traveled of 207,503 miles. The latter represent the incidence rates that would occur over the lifetime of the trucks. Since we assumed that the benefits accrued by one model year of single trucks over its entire lifetime is equal to the benefits accrued by an entire fleet of trucks for a single year, it follows that the incidence savings calculated for one model year of single trucks over its lifetime may also serve as the incidence savings for an entire fleet in a single year.

    Given the above assumptions, we adjust the incidence rate per 100,000 miles to the incidence rate per 207,503 miles by multiplying the incidence rates by a factor of 2.07503. For instance, to determine the lifetime incidence rate given a rear detection system installation rate of 100 percent (I100%,life), we take the incidence rate per 100,000 miles (I100%) and multiply the numerator and denominator by 2.07503 to arrive at I100%,life or 0.5104 ([0.2460 ´ 2.07503] ¸ [100,000 ´ 2.07503]) incidences per 207,503 miles traveled. A similar method is applied to the zero percent and 18 percent installation rates with the results presented in Table IV-2.

    Table IV-2
    Incidence rates as a function of installation rate for 100,000 vehicle miles traveled and lifetime of applicable straight trucks
    Rear Detection System Installation Rate Incidence Rate
    (per 100,000 miles)
    Incidence Rate
    (per 207,503 miles)
    0% I0% 0.3672 I0%,life 0.7619
    18% I18% 0.3453 I18%,life 0.7166
    100% I100% 0.2460 I100%,life 0.5104

    Taking the difference between I18%,life and I100%,life results in a current net savings rate of 0.2062 incidences for the lifetime of a straight truck. In other words, we estimate that, if the remainder of the 82 percent of the applicable fleet had a rear detection system installed, we expect a savings of 0.2062 incidences per truck per year.

    Multiplying the current net savings rate of the applicable truck fleet per year, or 0.2062 incidents, by $529.29 per incident yields $119.14, the cost savings per year on a per truck basis if a cross-view mirror or camera system is installed on the entire applicable fleet of straight trucks. Next, multiplying $133.24 by the entire applicable truck fleet of 365,000 vehicles yields a current net property damage cost savings of about $39.8 million.


    V.  Costs
    Table of Contents

    Cross-View Mirrors and Camera Systems

    Various commenters submitted cost estimates to the agency in response to the ANPRM and Request for Additional Information. NATC estimates the cost of a cross-view mirror system to be less than $100; Reliant stated that the cost for a mirror and labor is $80 per mirror; and the State of New York estimated a cost of $70 for either a new or retrofit mirror; NTSB stated in their response to the Request for Additional Information that a cross-view mirror would be about $50 if purchased individually or $30 if purchased in a volume. The Norton family (original petitioners to NHTSA) also stated that rear cross-view mirrors cost about $30 when purchased in volume from one manufacturer. NATC provided a cost of $250 for "some electronic devices" in their comments to the ANPRM. They were not specific as to the nature of the electronic device, whether it is a camera system or an ultrasonic/infrared rear detection system.

    Since the data provided by commenters are somewhat outdated, the agency recently contacted aftermarket supplier and manufacturers regarding cross-view mirrors and camera systems to get more up-to-date costs and also the volume discounts, if any. The prices quoted to the agency only include parts and any associated hardware and are in 2004 dollars ($2004); they do not cover labor costs, shipping and handling costs, or any applicable taxes. We received the following:

    For cross-view mirrors, one aftermarket supplier quoted a retail price of $69.80 for an 20.32 centimeter (8-inch) model and retail price of $90.00 for 25.4 centimeter (10-inch) mirrors. The percent discounts, if ordered in volume, are found below in Table V-1a and include stainless steel mounting hardware.

    Table V-1a
    Volume discounts for 8-inch and 10-inch cross-view mirrors ($2004)
    Quantities Discount 8-inch mirrors 10-inch mirrors
    1 to 150 5 percent discount off the net total $66.31 $85.50
    151 to 500 6 percent discount off the net total $65.61 $84.60
    501 or more 7 percent discount off the net total $64.91 $83.70

    A second supplier furnished volume prices to the agency for 20.32 centimeter (8-inch) and 25.4 centimeter (10-inch) mirrors (including mounting hardware). These prices are shown in Table V-1b.

    Table V-1b
    Unit volume costs for 8- and 10-inch mirrors ($2004)
      Unit Cost
    Quantity 8-inch mirrors 10-inch mirrors
    1 to 11 $39.53 $44.19
    12 to 24 $37.53 $41.19
    25 to 100 $36.03 $39.00
    101 or more $32.55 $36.00

    The United States Postal Service reported that the wholesale cost of an 11-inch rear cross-view mirror was $43.95, with the mirrors already installed as part of the original purchasing contract and prior to calendar year 2004.

    For camera systems, prices ranged from $212.00 to upwards of $1,248.00. The more costly systems utilize a color liquid crystal display (LCD), which, we believe, are excessive for the agency's goals. Black and white displays and cameras are more than adequate for visualization purposes, and thus, will be used in our analysis.

    One supplier quoted a list price of $697.00 for a camera system, which includes a 5.5-inch black and white monitor, camera, wiring, and mounting hardware. The volume discounts they provide are found below in Table V-2.

    Table V-2
    Unit volume costs for camera systems ($2004)
    Quantities Unit Cost
    1-11 $648.00
    12-24 $548.00
    25-100 $448.00
    101 to 500 $398.00
    501 or 2500 $280.00
    2501 or more $212.00

    Another supplier quoted a retail price of $569.95 but did not mention any volume discounts. Lastly, USPS provided a wholesale cost of $242.00 but did not mention how many were purchased (however, we will assume that 2,041 were purchased since that was how many were installed in their fleet to study the efficacy of a camera system).

    Because of the varying costs garnered by the agency, we now decide upon cost figures that will most reasonably reflect what the trucking industry will incur. For mirrors, we believe that Table V-1b best represents the cost schedule whereas for camera systems, Table V-2 is most representative. Comments are welcomed as to the costs of mirrors and cameras.

    Distributions of Counter-measures

    Since the agency proposes that the furthest distance of the geometric centers of the two mirrors at which the mirror system could provide a meaningful image is 5 meters (16.4 feet), some trucks would need to be equipped with a camera system in order to be compliant. However, there are trucks that will be able to meet the performance standard with mirrors alone. We attempt to estimate the fleet distributions that will use either a mirror (both 8- and 10-inches) or a camera system.

    To estimate the distribution of the counter-measures in this analysis (i.e., 8-inch and 10-inch mirrors and camera systems), we utilize data from the 1997 Vehicle Inventory and Use Survey (VIUS), which is released by the U.S. Census Bureau of the Department of Commerce. The VIUS data do not indicate if a truck has a rear detection system installed, but does contain data regarding the lengths of affected straight trucks. Using these measurements, we make approximations as to what the distance between mirrors would be. Those that have a separation between mirrors that exceed 5 meters will require a camera system whereas those that do not might be able to use a mirror system. For those trucks that may use a mirror, we further separate them into an 8-inch or 10-inch category.

    To estimate the distance between the driver's outside rear view mirror and the corresponding cross-view mirror, we assume that the horizontal and vertical distances between the mirrors form a right triangle and that the hypotenuse of this right triangle is the separation between mirrors. Figure 2 represents a simplified sketch of a box-type straight truck with the relevant dimensions included. The variable "h" is the overall height of the truck; "L" is the overall length of the truck;  "x" is the height of the geometric center of the outside rear mirror measured from the ground; and, "d" is the distance between the geometric centers of the outside and cross-view mirrors.

    Figure 2. Sketch of an idealized box-type straight truck.
    Figure 2.  Sketch of an idealized box-type straight truck.

    In the VIUS data, lengths are grouped into 14 categories: "Less than 13.0 feet", "13.0 to 15.9 feet", "16.0 to 19.9 feet", "20.0 to 27.9 feet", "28.0 to 35.9 feet", "36.0 to 40.9 feet", "41.0 to 44.9 feet", "45.0 to 49.9 feet", "50.0 to 54.9 feet", "55.0 to 59.9 feet", "60 to 64.9 feet", "65.0 to 69.9 feet", "70.0 to 74.9 feet", and "75.0 feet or more". In addition, there are 30 categories of truck types and an unspecified "Other" category for trucks that do not fit any of the 30 categories.

    Because the VIUS data only provided the overall length of the truck with no indication of the overall height, height of the outside mirror, or distance between the mirrors, we make estimations of the measurements not available in VIUS. For the overall height of a box-type straight truck, we turn to USPS data, which gives an average height of about 2.9 meters (9.5 feet). For the height of the center of the outside rearview mirror to the ground, an internal agency document in support of the ANPRM had a height of 2.1 meters (6.9 feet) with an overall height of the truck of 2.7 meters. For some trucks, the height of the center of the outside rearview mirror is proportional to the overall height of the truck, and therefore, we adjust the 2.1 meters from the internal agency document to an amount proportional to the overall height truck height. The USPS data indicates that the overall height of the truck is about 2.9 meters, and making a slight adjustment, gives the height of the outside rearview mirror to be about 2.3 meters (2.1 ´ [2.9 ¸ 2.7]). Since both mirror heights exists in the fleet but no indication of their respective distributions is available, we take a simple average of the two numbers to arrive at 2.2 meters for the height of the center of the outside rearview.

    Given that the maximum permissible distance between mirrors is 5.0 meters (16.4 feet) and that the vertical distance between mirrors is about 0.7 meters (h – x = 2.9 m – 2.2 m), using the Pythagorean theorem, we determine the maximum horizontal distance between mirrors to be about 4.9 meters (16.2 feet). Given the maximum horizontal distance between the outside rearview and the cross-view mirrors, we now need to determine the horizontal distance between the outside rearview mirror and the front of the truck. The sum of these two measurements will provide the maximum overall length of a truck that can accept a mirror.

    For the horizontal distance between the front of the truck to the center of the outside rearview mirror, we consulted two copies of body builder manuals that contain data regarding the dimensions of a typical truck chassis. Two predominate mirror locations were identified in the manuals: the first was in which the outside rearview mirror and front bumper where in the same plane, in other words, had a horizontal separation of almost zero meters; for the second type, the rearview mirror was noticeably separated from the front of the truck.

    The data indicates that the distance ranges between zero meters (for the first type) to about 1.5 meters (5.0 feet, for the second type). We therefore take the average of these measurements to arrive at 0.76 meters (2.5 feet), but round-up to 0.91 meters (3.0 feet) because we believe that the second truck type predominates and would contribute a greater proportion to the overall population. Adding the 0.9144 meters (3.0 feet) to the 4.94 meters (16.2 feet) (estimated maximum horizontal distance between mirrors) gives us about 5.85 meters (19.2 feet). Therefore, we assumed trucks with an overall length of 5.85 meters (19.2 feet) or less may be able to use a mirror system.

    Since the available categories do not cut-off at 19.2 feet, we decide to partition the "16.0 to 19.9 feet" category into two categories—"16.0 to 19.2 feet" and "19.3 to 19.9 feet"—in order to better estimate the number of vehicles that will need a particular system. Our rationale did not split the two groups evenly (that is, 50 percent to 50 percent) but rather 82.1 percent to 17.9 percent and is as follows: since the total range of the category is 3.9 feet, we weight the 3.9 feet by 82.1 percent and subsequently the number of trucks into the "16.0 to 19.2 feet" category. The remaining 17.9 percent of trucks are placed in the "19.3 to 19.9 feet" category.

    Therefore, we estimate that trucks in the VIUS categories with an overall length of 19.3 feet or greater will need to employ a camera system. In addition, we felt that certain truck categories, regardless of length, will need to use a camera system because they probably do not have a feasible place to mount a mirror. Those trucks are "platform with added devices", "low boy or depressed center", "basic platform", "livestock truck", "winch or crane", "wrecker", "pole or logging", "oilfield truck", "tank truck (liquid or gases)", "tank truck (dry bulk)", and "concrete mixer."  Lastly, certain truck classes in VIUS are excluded, although they fall within the weight range, and include "pickup", "yard tractor", "sport utility", "station wagon", and "minivan."  The raw VIUS data is available in the appendix.

    Prior to excluding any vehicle categories, we first distribute the "Other" category into the 30 known truck categories and then exclude the previously mentioned truck groups. Next, those trucks we feel can pass only with a camera system, regardless of length, and those trucks that are 19.3 feet or greater are grouped together to arrive at the number of trucks that will need a camera system. We assumed that trucks 4.85 meters (15.9 feet) or less used an 8-inch mirror whereas those in the "16.0 to 19.2 feet" category used a 10-inch mirror. The population figures we arrived at are presented in Table V-3.

    Table V-3
    Counter-measure distributions based on 1997 VIUS data
    Counter-measure Number Percent of Total
    Mirrors 8-inch 166,204 4.74%
    10-inch 718,161 20.47%
    Camera Systems 2,623,725 74.79%
      Total 3,508,090 100.00%

    Because the rule will not be retroactive, the absolute figures in Table V-3 will not be used since they represent current trucks on the road and not the number of new trucks. However, we feel that the distributions are representative of new vehicles and employ the percentages to make an estimate of the total cost of this rulemaking. In other words, we believe that 4.74 percent of the fleet will use 8-inch mirrors, 18.70 percent will use 10-inch mirrors, and the balance of 76.56 percent will employ a camera system. From Table III-8, we take the 299,300 "Applicable Vehicles" and multiply it with the percentages in Table V-3 to estimate the number of trucks that will receive a particular counter-measure. The number of trucks needing an 8-inch mirror is 0.0474 times 299,300 or 14,180 trucks; for 10-inch mirrors, the number of trucks is 61,271; and for cameras, 223,849 trucks. The results are summarized in Table V-4.

    Table V-4
    Estimated breakdown of counter-measures by fleet size
      Fleet Distribution
    Mirrors 8-inch 14,180
    10-inch 61,271
    Camera Systems 223,849
    Newly Registered Trucks 299,300

    Consumer Costs

    The total cost incurred by consumers is estimated in this section. We assume that manufacturers or suppliers produce mirrors and camera systems and, in turn, sell them to vehicle manufacturers to install on new trucks. These costs are passed on to consumers. Since costs are lower when purchased in volume, manufacturers are almost certain to purchase a rear detection system in quantity rather than individually. In addition, because we estimate that cameras will comprise a significant proportion of the population of those trucks that are affected, we anticipate that manufacturers will purchase a larger quantity of cameras when compared to mirrors. We also believe that both the competition amongst mirror and camera manufacturers and as the technology becomes more commonplace will inevitably lower the prices. Therefore, from Tables V-1b and V-2, respectively, we assume that manufacturers will purchase mirrors in batches of 101 or more at a cost of $32.55 for each 8-inch mirror with mounting hardware and $36.00 for each 10-inch mirror with mounting hardware. For cameras we assume that the large increase in demand, and competition, will bring down the price for all purchasers to $212 (the current price when purchasing a quantity of 2501 or more), which includes camera, monitor, wires, and mounting hardware.

    Since the estimated newly registered trucks will not have any rear-detection system installed, to accurately estimate the cost burden, labor costs to install a mirror or camera must also be considered. We estimated that for a mirror system, once it is in production and the task becomes repetitive, it will take about 3 minutes (0.05 hours) to install a mirror and about 6 minutes (0.10 hours) to install a camera system. Comments are welcomed as to the installation times of the respective systems. An hourly wage of $32.97 for assembly line workers, which includes fringe benefits, is used for both mirrors and cameras. Lastly, because the unit costs of the counter-measures were provided to the agency as wholesale costs, we must apply a markup to the costs to bring them up to a consumer cost (i.e., what a consumer would pay if he or she purchased a single mirror system or camera system off the shelf), and also markup any other variable costs associated with the counter-measure, which, in this case, would be the hourly wage. Table V-5a presents the wholesale cost of installing a single counter-measure. The unit cost of a counter-measure is added to the product of the hourly wage and time it takes to install a system (duration).

    Table V-5a
    Wholesale unit installation cost calculations ($2004)
    Counter-measure Unit Cost   Hourly Wage   Duration   Unit Installation Cost
    8-inch mirror $32.55 + $32.97 ´ 0.05 = $34.20
    10-inch mirror $36.00 + $32.97 ´ 0.05 = $37.65
    Camera system $212.00 + $32.97 ´ 0.10 = $215.30

    Table V-5b takes the wholesale unit installation cost and marks it up by a factor of 1.51 to a consumer unit installation cost.

    Table V-5b
    Wholesale cost markup to consumer cost ($2004)
    Counter-measure Unit Installation Cost
      Markup   Unit Installation Cost
    8-inch mirror $34.20 ´ 1.51 = $51.64
    10-inch mirror $37.65 ´ 1.51 = $56.85
    Camera system $215.30 ´ 1.51 = $325.10

    Lastly, Table V-5c shows the consumer cost of each of the counter-measure and the total consumer cost.

    Table V-5c
    Total costs ($2004)
    Counter-measure Unit Installation Cost
      Quantity   Total Installation Cost
    8-inch mirror
    10-inch mirror
    Camera system
    Total Consumer Cost

    Thus, the total estimated consumer cost is about $77 million.

    Maintenance Costs

    There will be an increase in maintenance costs for either a mirror system or a video system. Mirror systems can be damaged by hitting tree branches, backing incidences, theft, vandalism, etc. Video systems can be damaged when the vehicle is struck in the rear, or by theft, vandalism, etc. The agency requests comments on the frequency of losses due to these factors combined and the average costs to repair such systems.


    VI.  Lead Time
    Table of Contents

    We are proposing that the new rear object detection requirements to prevent backing deaths and injuries would be effective for new vehicles manufactured one year after publication of a final rule. Installation of cross-view mirrors would not involve substantial engineering efforts or changes in manufacturing processes. Manufacturers might need additional time to implement more technically demanding video systems, although we believe that one year would likely provide sufficient time for manufacturers to incorporate these systems as well.

    Our intent is not to limit the solution to the backing problem to mirrors, as demonstrated by our proposed compliance options. However, we believe that the technology is readily available and vehicle manufacturers and owners could apply that technology rapidly. We do not see any reason to delay realization of the anticipated safety benefits by unnecessarily prolonging implementation of the amended standard.


    VII.  Cost Effectiveness And Sensitivity Analyseis
    Table of Contents

    This section combines costs and benefits to provide a comparison of the estimated injuries and lives saved per dollar spent. It should be noted that costs occur when the vehicle is purchased, but benefits accrue over the lifetime of the vehicle. Therefore, benefits must be discounted to reflect their present value and to put them on a common basis with costs.

    In some instances, costs may exceed economic benefits, and in these cases, it is necessary to derive a net cost per equivalent fatality prevented. An equivalent fatality is defined as the sum of fatalities and nonfatal injuries prevented converted into fatality equivalents. This conversion is accomplished using the relative values of fatalities and injuries measured using a "willingness‑to‑pay" approach. This approach measures individuals' willingness to pay to avoid the risk of death or injury based on societal behavioral measures, such as pay differentials for more risky jobs. The column labeled "Relative Value per Injury" in Table VII-1 presents the estimated relative rational investment level to prevent one injury, by maximum injury severity. The data represent average costs for crash victims of all ages. Injuries are assumed to be valued based on the relative costs of MAIS injuries.

    The relative value of an MAIS injury when compared to a fatality is found by using data from Table A-1 (page 62) of the agency's publication, The Economic Impact of Motor Vehicle Crashes, 2000. For an MAIS-1 injury, an adjusted cost is first determined by subtracting the costs due to "Travel Delay" ($777) and "Property Damage" ($3,844) from the "Comprehensive cost [5]" ($15,017), where the calculation is $15,017 – $777 – $3,844 = $10,396. An analogous cost for a fatality is determined in a similar fashion, where the "Comprehensive" cost for a fatality is $3,366,388; cost due to "Travel Delay" is $9,148; and "Property Damage" is $10,273. The adjusted cost for a fatality is therefore $3,346,967 ($3,366,388 – $9,418 – $10,273). Dividing the MAIS-1 adjusted cost by its fatality partner results in 0.0031 ($10,396 ¸ $3,346,967), which represents the relative value of an MAIS-1 injury when compared to a fatality. In other words, a single MAIS-1 injury is equal to 0.0031 of a fatality, or 322 MAIS-1 injuries would be equal to one fatality, in terms of costs. The relative values (with the adjusted costs in parentheses following the MAIS injury level) for MAIS-2 ($153,158), MAIS-3 ($306,465), MAIS-4 ($720,748), and MAIS-5 ($2,384,403) injuries are calculated in a similar fashion and are presented in Table VII-1. Keep in mind that these relative values use costs based on calendar year 2000 economics.

    Table VII-1
    Relative value per injury and equivalent fatalities
    Injury Level Relative Value per Injury   Injuries   Equivalent Fatalities
    MAIS-1 0.0031 ´ 35 = 0.1087
    MAIS-2 0.0458 ´ 6 = 0.2746
    MAIS-3 0.0916 ´ 2 = 0.1831
    MAIS-4 0.2153 ´ 0 = 0.0000
    MAIS-5 0.7124 ´ 0 = 0.0000
    Fatal 1.0000 ´ 23 = 23.0000
        Equivalent Fatalities = 23.5664

    Table VII-1 also lists the equivalent fatalities, which is simply the product of the relative value per injury and the number of injuries. Taking the MAIS-1 to -5 injuries and "converting" them into fatalities and summing reveals an equivalent fatality of 23.5664.

    Appendix V of the "Regulatory Program of the United States Government", April 1, 1990 - March 31, 1991, sets out guidance for regulatory impact analyses. One of the guidelines deals with discounting the monetary values of benefits and costs occurring in different years to their present value so that they are comparable. The agency performed a cost-effectiveness analysis resulting in an estimate of the cost per equivalent life saved, as shown on the previous pages. The guidelines state, "An attempt should be made to quantify all potential real incremental benefits to society in monetary terms of the maximum extent possible".  For the purposes of the cost-effectiveness analysis, the Office of Management and Budget (OMB) has requested that the agency compound costs or discount the benefits to account for the different points in time that they occur.

    There is general agreement within the economic community that the appropriate basis for determining discount rates is the marginal opportunity costs of lost or displaced funds. When these funds involve capital investment, the marginal, real rate of return on capital must be considered. However, when these funds represent lost consumption, the appropriate measure is the rate at which society is willing to trade-off future for current consumption. This is referred to as the "social rate of time preference", and it is generally assumed that the consumption rate of interest—that is, the real, after-tax rate of return on widely available savings instruments or investment opportunities—is the appropriate measure of its value.

    Estimates of the social rate of time preference have been made by a number of authors. Robert Lind [6] estimated that the social rate of time preference is between zero and 6 percent, reflecting the rates of return on Treasury bills and stock market portfolios. Kolb and Sheraga [7] put the rate at between one and five percent, based on returns to stocks and three-month Treasury bills. Moore and Viscusi [8] calculated a two percent real time rate of time preference for health, which they characterize as being consistent with financial market rates for the period covered by their study. Moore and Viscusi's estimate was derived by estimating the implicit discount rate for deferred health benefits exhibited by workers in their choice of job risk.

    OMB Circular A-4 recommends agencies use both three percent and seven percent as the "social rate of time preference".

    When a crash is severe enough to potentially result in an occupant fatality or injury, which could be at any time during the vehicle's lifetime, safety benefits may occur if an appropriate counter-measure is introduced. For this analysis, the agency assumes that crashes over the truck fleet's lifetime will occur in proportion to the number of miles a given year's new truck fleet will be driven from year to year as it ages, or, in other words, exposure to a particular crash event is proportional to the VMT.

    Thus, for a three percent interest rate, a discount factor of 0.8054 is determined, whereas at a seven percent interest rate, the discount factor is 0.6315. We apply these discount factors to the estimated benefits in order to bring them to a present value.

    The cost effectiveness is calculated by dividing the net cost burden (vehicle costs minus property damage savings) by the equivalent fatalities. Cost figures are presented in Tables VII-2a and VII-2b.

    Table VII-2a
    Discounted property damage cost savings
    Interest Rate Property Damage Benefits   Discount Factor   Net Property Cost Savings
    No discount $39,827,832 ´ 1.0000 = $39,827,832
    3.00% $39,827,832 ´ 0.8054 = $32,077,336
    7.00% $39,827,832 ´ 0.6315 = $25,151,276

    Table VII-2b
    Net costs after property damage cost savings
    Interest Rate Cost Burden   Net Property Cost Savings   Net Cost Burden
    No discount $76,988,310 - $39,827,832 = $37,160,478
    3.00% $76,988,310 - $32,077,336 = $44,910,974
    7.00% $76,988,310 - $25,151,276 = $51,837,034

    Lastly, taking the net cost burden and dividing by the equivalent lives saved yields the cost per equivalent life saved at a particular discount level. These calculations are presented in Tables VII-3a and VII-3b.

    Table VII-3a
    Discounted equivalent lives saved
    Interest Rate Equivalent Lives Saved   Discount Factor   Discounted Equivalent Lives Saved
    No discount 23.5664 ´ 1.0000 = 23.5664
    3.00% 23.5664 ´ 0.8054 = 18.9804
    7.00% 23.5664 ´ 0.6315 = 14.8822

    Table VII-3b
    Cost per equivalent life saved
    Interest Rate Net Cost Burden   Discounted Equivalent Fatalities   Cost per Equivalent Life Saved
    No discount $37,160,478 ¸ 23.5664 = $1,576,842
    3.00% $44,910,974 ¸ 18.9804 = $2,366,176
    7.00% $51,837,034 ¸ 14.8822 = $3,483,157

    Sensitivity Analyses

    The effectiveness used in the preceding part of this analysis is based on a single study conducted by Federal Express in 1984.  Because of the age of this study, it is appropriate to examine the dependency of our results on the effectiveness rate. We performed two different sensitivity analyses to determine how sensitive the results of the analysis are to the estimated effectiveness of rear detection systems. In the main body of the analysis, we estimate that rear detection systems are 33 percent effective in reducing crashes attributable to straight trucks backing up. In this sensitivity analysis we examine what would be the net costs, benefits, and cost per equivalent life saved if rear detection systems were 20 percent, 40 percent, or 60 percent effective. Table VII-4 shows these results.

    Table VII-5 shows breakeven points, in terms of effectiveness, for costs and benefits using three assumptions. First, that a statistical life is valued at $3.5 million. Second, that a statistical life is valued at $5.5 million. Third, that property damage cost reductions equal the cost of the rear detection equipment. The results of these analyses show that if we assume that the value of a statistical life is $3.5 million, this rule will be cost effective if effectiveness is 27 percent or higher (at a 3 percent discount rate) or 33 percent or higher (at a 7 percent discount rate).   If we assume that the value of a statistical life is $5.5 million, this rule will be cost effective if effectiveness is 20 percent or higher (at a 3 percent discount rate) or 25 percent or higher (at a 7 percent discount rate).

    Property damage savings equal the cost of the rule at an effectiveness of 54 percent (at a 3 percent discount rate) or 60 percent (at a 7 percent discount rate).

    Table VII-4
    Cost per equivalent life saved
    3 Percent Discount Rate
    Effectiveness Net Cost Burden
      Discounted Equivalent Fatalities   Cost per Equivalent
    Life Saved
    20% $60.7 ¸ 10.79 = $5.6
    33% $44.9 ¸ 18.98 = $2.4
    40% $33.6 ¸ 23.11 = $1.5
    60% $-20.7* ¸ 36.40 = -$0.1
    7 Percent Discount Rate
    20% $64.2 ¸ 8.46 = $7.6
    33% $51.8 ¸ 14.88 = $3.5
    40% $42.9 ¸ 18.12 = $2.4
    60% $0.3 ¸ 28.54 = $0.01
    *  At a 60 percent effectiveness rate, property damage savings are $121.3 million undiscounted. Discounted at 3 percent, property damage savings are $97.7 million ($121.2*0.8054). Property damage savings are higher than consumer costs of $77 million resulting in a net savings to society.

    Table VII-5
    Breakeven Point Analysis Effectiveness Required
    Discount Rate At $3.5 million per
    life saved
    At $5.5 million per
    life saved
    Where Property
    Damage Savings
    Equal Costs
    3.00% 27% 20% 54%
    7.00% 33% 25% 60%


    VIII.  Alternatives
    Table of Contents

    In this section, we attempt to analyze an alternate regulatory approach for this rulemaking. In particular, we examine the consequences of an increase in backing speed from 3 mph to 6 mph. The agency believes that if the backing speed is 6 mph rather than the current 3 mph, then the test set-up (see Figure III-1) would change from a 3 meter by 3 meter area to a 3 meter by 5 meter area. If this were to happen, we believe that only a camera system would be able to have any injury/fatality mitigation potential. In addition, those trucks that currently have a cross-view mirror installed will incur an incremental cost to upgrade to a camera system.

    The agency has identified that about 365,000 newly registered straight trucks per year would fall under this rulemaking. Of those trucks, we estimate that about 18 percent (65,700 trucks) currently have a rear detection system installed whereas the remaining 82 percent (299,300 trucks) have no counter-measure.

    As mentioned before, within the 18 percent of trucks with a counter-measure, some trucks have cross-view mirrors whereas others have a camera system. The agency does not have information as to the distribution of mirrors and cameras nor did any commenters provide a figure to the agency. Therefore, we assume that 80 percent have a cross-view mirror and 20 percent have a camera system, which translates into 52,560 (65,700 ´ 0.80) trucks with a cross-view mirror. Of the 52,560 trucks with a cross-view mirror, we further refine the distribution between 8- and 10-inch mirrors based on data from Table V-3. In the table, we estimate that 166,204 trucks will need an 8-inch mirror whereas another 718,161 will need a 10-inch one for a total of 884,365 trucks. From these figures, we estimate that trucks in the current fleet with a cross-view mirror, 18.79 percent (166,204 ¸ 884,365) or 9,878 (52,560 ´ 0.1879) trucks will have an 8-inch mirror where as 81.21 percent (718,161 ¸ 884,365) or 42,682 (52,560 ´ 0.8121) trucks will have a 10-inch mirror. The agency welcomes comments as to the counter-measure distribution currently in the fleet.

    The cost of upgrading from a cross-view mirror to a camera system is the product of the total affected vehicles and the incremental cost difference between a mirror and camera. From Table V-5b, the consumer incremental cost to go from an 8-inch mirror to a camera is $273.46 ($325.10 – $51.64); from a 10-inch mirror to a camera is $268.25 ($325.10 – $56.85). Applying these cost figures to their respective vehicle populations results in a total incremental upgrade cost of $2,701,238 (9,878 ´ $273.46) for 8-inch mirrors and $11,449,447 (42,682 ´ $268.25) for 10-inch mirrors. The cost to install a camera system on those vehicles that currently have no counter-measure is $97,302,430 (299,300 ´ $325.10). Thus, the total cost burden is expected to be about $111 million ($2,701,238 + $11,449,447 + $97,302,430).

    The agency believes that a camera system is more effective than a cross-view mirror but only have anecdotal evidence to support this statement. Therefore, the effectiveness is still assumed to be that of a mirror system or 33 percent, and thus, the lives saved, injuries mitigated and property damage savings do not change. The alternate net cost burden and cost per equivalent life saved is presented below in Tables VIII-1a and VIII-1b, respectively.

    Table VIII-1a
    Alternate net costs after property damage cost savings
    Interest Rate Cost Burden   Net Property Cost Savings   Net Cost Burden
    No discount $111,453,115 - $39,827,832 = $71,625,283
    3.00% $111,453,115 - $32,077,336 = $79,375,779
    7.00% $111,453,115 - $25,151,276 = $86,301,839

    Table VIII-1b
    Alternate cost per equivalent life saved
    Interest Rate Net Cost Burden   Discounted Equivalent Fatalities   Cost per Equivalent Life Saved
    No discount $71,625,283 ¸ 23.5664 = $3,039,297
    3.00% $79,375,779 ¸ 18.9804 = $4,181,987
    7.00% $86,301,839 ¸ 14.8822 = $5,798,997


    IX.  Unfunded Mandates Reform Act Analysis
    Table of Contents

    The Unfunded Mandates Reform Act of 1995 (Public Law 104-4) requires agencies to prepare a written assessment of the costs, benefits and other effects of proposed or final rules that include a Federal mandate likely to result in the expenditure by State, local or tribal governments, in the aggregate, or by the private sector, of more than $100 million annually (assumed annually with inflation with base year of 1995). Adjusting this amount by the gross domestic product price deflator for the year 2004 results in about $118 million (115.5 ¸ 98.11 ´ $100 million).

    These effects have been discussed in detail in the "Costs" and "Benefits" sections of this Preliminary Regulatory Evaluation. To summarize, NHTSA is issuing this proposed rule to require a rear detection system be installed on straight trucks between 4,536 kg (10,000 pounds) and 11,793 kg (26,000 pounds). The proposed rule will improve the visibility of the area behind trucks allowing drivers to detect the presence of persons and objects directly them, thereby reducing backing-related deaths and injuries and also lowering incidences of property damage. The cost of the proposed rule to install either a cross-view mirror or camera system is estimated to be $77 million (see Table V-5c).


    X.  Small Business Impacts
    Table of Contents

    A. Initial Regulatory Flexibility Analysis

    The Regulatory Flexibility Act of 1980 (5 U.S.C §601 et seq.) requires agencies to evaluate the potential effects of their proposed and final rules on small business, small organizations and small Government jurisdictions in the United States.

    The 5 U.S.C §603 requires agencies to prepare and make available for public comments initial and final regulatory flexibility analysis (RFA) describing the impact of proposed and final rules on small entities. Section 603(b) of the Act specifies the content of a RFA. Each RFA must contain:

    1. A description of the reasons why action by the agency is being considered;
    2. A succinct statement of the objectives of, and legal basis for a proposal or final rule;
    3. A description of and, where feasible, an estimate of the number of small entities to which the proposal or final rule will apply;
    4. A description of the projected reporting, recording keeping and other compliance requirements of a proposal or final rule including an estimate of the classes of small entities which will be subject to the requirement and the type of professional skills necessary for preparation of the report or record;
    5. An identification, to the extent practicable, of all relevant Federal rules which may duplicate, overlap or conflict with the proposal or final rule;
    6. Each final regulatory flexibility analysis shall also contain a description of any significant alternatives to the proposal or final rule which accomplish the stated objectives of applicable status and which minimize any significant economic impact of the final on small entities.

    1. Description of the reason why action by the agency is being considered

      NHTSA is considering this action to improve the ability of single unit truck drivers to detect pedestrians of other objects behind their truck to reduce fatalities, injuries, and property damage.

    2. Objectives of, and legal basis for, the proposal or final rule

      NHTSA is proposing these changes under the Authority of 49 U.S.C. 322, 30111, 30115, 30117, and 30666; delegation of Authority at 49 CFR 1.50. The agency is authorized to issue Federal motor vehicle safety standards that meet the need for motor vehicle safety.

    3. Description and estimate of the number of small entities to which the proposal will apply

      The proposal would affect the manufacturers of single unit trucks, mirror manufacturers and video camera manufacturers.

      Business entities are defined as small businesses using the North American Industry Classification System (NAICS) code, for the purposes of receiving Small Business Administration assistance. One of the criteria for determining size, as stated in 13 CFR 121.201, is the number of employees in the firm. To qualify as a small business in the Heavy Duty Truck Manufacturing (NAICS 336120) or in the Motor Vehicle Body Manufacturing  (NAICS 336211) categories, the firm must have fewer than 1,000 employees. Mirror manufacturers might be categorized in the "All Other Motor Vehicle Parts Manufacturing" (NAICS 336399), in which case the firm must have less than 750 employees to be considered a small business. The agency does not know what NAICS video camera manufacturers would fall into, however, it is likely that video camera manufacturers are not small businesses.

      NHTSA estimates there are about 750 manufacturers of single-unit trucks in the United States, most of which are small businesses. In addition, we estimate that there are about a dozen mirror manufacturers in the United States, of which we believe about 3 are small businesses. Comments are requested on these estimates.  

    4. A description of the projected reporting, recording keeping and other compliance requirements of a proposal or final rule including an estimate of the classes of small entities which will be subject to the requirement and the type of professional skills necessary for preparation of the report or record

      There are no reporting or record keeping requirements associated with the proposal. Manufacturers will need to certify compliance with the final rule, just like with all safety standards.

    5. An identification, to the extent practicable, of all relevant Federal rules which may duplicate, overlap or conflict with the proposal or final rule;

      There are no rules that duplicate, overlap, or conflict with the proposal.

    6. Each final regulatory flexibility analysis shall also contain a description of any significant alternatives to the proposal or final rule which accomplish the stated objectives of applicable status and which minimize any significant economic impact of the final on small entities.

      Two sets of alternatives are being considered for this proposal. The first involves the test set-up and the difference between the alternatives is in the size of the test cylinder detection area behind the vehicle. The proposal is for a 3 meters by 3 meters area. The alternative is for a larger area of 3 meters by 5 meters. We believe the larger area will require the use of a video camera. This will have an impact on the mirror manufacturers.

      The second alternative deals with the applicability of the proposal and whether certain types of vehicles should be exempt from the proposal because it is either impractical to install the countermeasures on the vehicle, or for some other reason. The agency has asked for comments on vehicle applicability, which could minimize the impacts of the proposal to some extent.

    In summary, the agency believes that the proposal will have a significant economic impact on a substantial number of small businesses. There are a substantial number of single-unit truck manufacturers and the cost of video cameras is significant. We anticipate that single unit truck manufacturers will pass this cost on to consumers. The increase in price could have a small impact on the sales of single unit trucks. However, we expect this influence to be minor since we believe the price elasticity for single-unit trucks to be very inelastic. In other words, we don't believe the increase in price will have much of an influence on the level of sales. These vehicles are bought for business purposes and there are not many purchase alternatives. The impact on competition for single unit truck manufacturers is unknown at this point. If, as we anticipate, the cost of video systems comes down to a basic level for all purchasers, regardless of purchase size, there will be little impact on competition since this rule will be applicable to all manufacturers and thus all entities will be subject to the same regulatory impact.

    We do not believe that the impact on mirror manufacturers will be significant. While there will be an increase in sales of about 77,000 units, this is not a huge increase in sales for the average mirror manufacturer.


    XI.  Appendices
    Table of Contents

    Since we know the number of fatalities, we exclude the "K" column and also subtract the "Fatal" figure from their corresponding "Total", which leaves the number of injuries only. Similarly, since the number of uninjured ("NO" column) and injuries of unknown severity ("ISU" column) were not reported, we also exclude those categories from our calculations. Thus, Tables XI-2a and XI-2b represents the figures used to distribute the injuries from NEISS and GES (see Table III-5) into the MAIS scale.

    Table XI-1
    KABCO/MAIS injury distribution table (1982 to 1986 NASS CDS Injury Data).
      KABCO Classifications
    MAIS 0 34,125 251,763 1,313,849 567,577 2,243 55,920,352 16,197
    MAIS 1 1,106,880 4,039,582 4,731,293 111,255 2,899 4,493,704 151,977
    MAIS 2 628,338 636,692 445,949 11,258 1,188 125,755 33,822
    MAIS 3 376,136 153,411 99,524 5,431 238 17,347 9,352
    MAIS 4 65,427 13,620 4,229 139 394 716 3,688
    MAIS 5 39,650 3,518 1,219 310 0 0 288
    Fatal 12,194 1,350 645 0 168,780 60 819
    Total 2,262,750 5,099,936 6,596,708 695,970 175,742 60,557,934 216,143
    A = Incapacitating Injury NO = No Injury MAIS 0 = No Injury MAIS 4  = Severe
    B = Non incapacitating Injury ISU = Injured, but severity unknown MAIS 1 = Minor MAIS 5 = Critical
    C = Possible Injury UNK = Unknown if Injured MAIS 2 = Moderate AIS = Abbreviated Injury Scale
    K = Killed   MAIS 3 = Serious MAIS = Maximum AIS

    Table XI-2a is a summary of the data used from Table XI-1 to calculate the decimal fractions in Table XI-2b. To calculate the decimal fractions in Table XI-2b, we take a particular KABCO injury level at a specific MAIS injury level and divide by its corresponding "TOTAL" figure. For example, to determine the fraction of MAIS 0 injuries for the A-level injury on the KABCO scale, we take 34,125 injuries and divide by 2,250,556 to arrive at 0.01516. This is repeated for the remainder of Table XI-2a to arrive at Table XI-2b.

    Table XI-2a
    Adjusted absolute values for the KABCO/MAIS injury distribution table from Table XI-1
    0 34,125 251,763 1,313,849 567,577
    1 1,106,880 4,039,582 4,731,293 111,255
    2 628,338 636,692 445,949 11,258
    3 376,136 153,411 99,524 5,431
    4 65,427 13,620 4,229 139
    5 39,650 3,518 1,219 310
    TOTAL 2,250,556 5,098,586 6,596,063 695,970

    Table XI-2b
    Relative values for the KABCO/MAIS Injury distribution table from Table XI-2a
    0 0.01516 0.04938 0.19919 0.81552
    1 0.49183 0.79229 0.71729 0.15986
    2 0.27919 0.12488 0.06761 0.01618
    3 0.16713 0.03009 0.01509 0.00780
    4 0.02907 0.00267 0.00064 0.00020
    5 0.01762 0.00069 0.00018 0.00045
    TOTAL 1.00000 1.00000 1.00000 1.00000

    To determine the number of MAIS injuries for a given KABCO injury, we take a particular number of injuries from Table III-5 and distribute it with their respective column in Table XI-2b. For example, to determine the MAIS distribution for "A" injuries, multiple 19.61 by 0.49183 to get 9.64 MAIS 1 injuries; for MAIS 2 injuries, multiply 19.61 by 0.27919 to get 5.47. The remainder of the KABCO to MAIS injury distributions are found above in Table III-7a.

    Table XI-3
    VIUS Data
    Truck Category LESS THAN 13.0 FEET 13.0 TO 15.9 FEET 16.0 TO 19.9 FEET 20.0 TO 27.9 FEET
    Pickup 1,201 42,492 168,710 13,856
    Panel or van 267 5,453 27,938 3,680
    Multistop or step van 1,024 31,718 267,974 192,950
    Platform with added devices 493 8,847 50,914 118,225
    Low boy or depressed center 0 0 1,971 2,680
    Basic platform 2,267 46,324 256,281 380,983
    Livestock truck 267 130 3,552 11,280
    Insulated nonrefrigerated van 0 500 1,133 4,086
    Insulated refrigerated van 267 2,638 9,147 33,479
    Drop-frame van 0 1,798 2,405 5,598
    Open-top van 27 310 554 2,243
    Basic enclosed van 176 7,854 67,718 137,152
    Beverage 0 362 3,202 12,809
    Public utility 554 6,268 44,059 65,898
    Winch or crane 0 311 4,409 14,223
    Wrecker 329 4,759 28,420 44,435
    Pole or logging 0 1,049 1,270 5,050
    Auto transport 0 48 769 4,001
    Service truck 420 14,473 77,543 58,449
    Yard tractor 0 1,306 674 19
    Sport utility 7,211 6,221 10,011 646
    Station wagon 0 416 4,824 614
    Minivan 282 2,561 35,545 974
    Oilfield truck 0 433 2,613 5,704
    Grain body 0 5,188 34,414 116,711
    Garbage hauler 0 881 3,714 11,316
    Dump truck 1,608 22,589 113,092 149,445
    Tank truck (liquids or gases) 27 4,272 14,600 71,468
    Tank truck (dry bulk) 0 155 3,176 4,020
    Concrete mixer 0 0 521 430
    Other 0 131 1,866 2,475
    Total 16,420 219,487 1,243,019 1,474,899

    Table XI-3
    VIUS Data (continued)
    Truck Category 28.0 TO 35.9 FEET 36.0 TO 40.9 FEET 41.0 TO 44.9 FEET 45.0 TO 49.9 FEET
    Pickup 838 1,057 374 538
    Panel or van 571 0 0 0
    Multistop or step van 14,585 1,513 1,820 22
    Platform with added devices 31,258 4,284 1,669 1,093
    Low boy or depressed center 4,796 3,979 2,905 2,766
    Basic platform 85,550 23,289 9,793 6,721
    Livestock truck 3,266 2,019 717 54
    Insulated nonrefrigerated van 1,953 0 0 138
    Insulated refrigerated van 11,845 986 180 0
    Drop-frame van 7,564 97 382 454
    Open-top van 187 320 0 0
    Basic enclosed van 85,824 6,917 1,533 1,552
    Beverage 17,995 3,054 1,744 557
    Public utility 14,578 4,239 2,424 1,308
    Winch or crane 7,236 624 517 132
    Wrecker 11,620 1,085 104 367
    Pole or logging 1,776 328 54 0
    Auto transport 2,997 98 0 172
    Service truck 2,557 1,102 422 418
    Yard tractor 0 0 48 97
    Sport utility 123 0 0 0
    Station wagon 78 0 0 0
    Minivan 0 0 0 0
    Oilfield truck 1,907 535 0 302
    Grain body 18,397 514 521 342
    Garbage hauler 8,132 681 540 370
    Dump truck 28,346 10,268 5,744 3,612
    Tank truck (liquids or gases) 11,392 1,534 236 351
    Tank truck (dry bulk) 752 420 0 0
    Concrete mixer 327 0 0 0
    Other 515 0 420 0
    Total 376,965 68,943 32,147 21,366

    Table XI-3
    VIUS Data (continued)
    Truck Category 50.0 TO 54.9 FEET 55.0 TO 59.9 FEET 60.0 TO 64.9 FEET 65.0 TO 69.9 FEET
    Pickup 0 0 0 0
    Panel or van 0 0 0 0
    Multistop or step van 50 0 841 0
    Platform with added devices 936 463 46 97
    Low boy or depressed center 1,293 1,439 696 220
    Basic platform 4,432 1,821 810 227
    Livestock truck 220 344 167 172
    Insulated nonrefrigerated van 22 46 123 0
    Insulated refrigerated van 688 300 941 29
    Drop-frame van 0 54 141 166
    Open-top van 0 0 0 0
    Basic enclosed van 2,831 7,710 6,484 1,611
    Beverage 1,225 1,530 621 0
    Public utility 790 229 616 158
    Winch or crane 172 328 50 0
    Wrecker 0 0 0 0
    Pole or logging 197 0 0 0
    Auto transport 135 135 167 22
    Service truck 123 0 0 0
    Yard tractor 0 294 0 122
    Sport utility 0 0 0 0
    Station wagon 0 0 0 0
    Minivan 232 0 0 0
    Oilfield truck 0 0 0 0
    Grain body 131 512 160 0
    Garbage hauler 132 0 0 0
    Dump truck 953 957 0 0
    Tank truck (liquids or gases) 0 611 186 97
    Tank truck (dry bulk) 197 0 266 0
    Concrete mixer 0 0 0 0
    Other 0 197 0 123
    Total 14,759 16,970 12,315 3,044

    Table XI-3
    VIUS Data (continued)
    Truck Category 70.0 TO 74.9 FEET 75.0 FEET OR MORE Total
    Pickup 0 0 229,066
    Panel or van 0 0 37,909
    Multistop or step van 0 0 512,497
    Platform with added devices 29 0 218,354
    Low boy or depressed center 48 538 23,331
    Basic platform 0 0 818,498
    Livestock truck 0 0 22,188
    Insulated nonrefrigerated van 0 0 8,001
    Insulated refrigerated van 317 0 60,817
    Drop-frame van 0 0 18,659
    Open-top van 0 0 3,641
    Basic enclosed van 2,153 964 330,479
    Beverage 97 0 43,196
    Public utility 0 0 141,121
    Winch or crane 0 0 28,002
    Wrecker 0 0 91,119
    Pole or logging 0 0 9,724
    Auto transport 123 0 8,667
    Service truck 0 0 155,507
    Yard tractor 0 0 2,560
    Sport utility 0 0 24,212
    Station wagon 0 0 5,932
    Minivan 0 0 39,594
    Oilfield truck 0 0 11,494
    Grain body 0 0 176,890
    Garbage hauler 0 0 25,766
    Dump truck 0 48 336,662
    Tank truck (liquids or gases) 210 0 104,984
    Tank truck (dry bulk) 0 0 8,986
    Concrete mixer 0 0 1,278
    Other 0 3,229 8,956
    Total 2,977 4,779 3,508,090

    [1] A "straight truck" is a single-unit truck composed of an undetachable cab and body. Body types routinely incorporated as part of straight trucks include an enclosed box, flat bed, dump bed, bulk container, or special purpose equipment".

    [2] Comments were received from: (1) the National Private Truck Council (NPTC); (2) the American Trucking Associations (ATA); (3) the Towing and Recovery Association of America (TRAA); (4) the National Truck Equipment Association (NTEA); (5) Ford Motor Company (Ford); (6) Sheffield Partners LLC (Sheffield); (7) Rostra Precision Controls, Inc. (Rostra); (8) Reliant Energy (Reliant); (9) ABC Supply Co., Inc. (ABC); (10) Federal Express Corporation (FedEx); (11) the International Brotherhood of Teamsters (Teamsters); (12) the New York Department of Transportation (NYDOT); (13) the Nevada Automotive Test Center (NATC); and (14) Ronald G. Silc.

    [3] The FMVSS 111 ANPRM can be found in the public docket under NHTSA-2000-7967-1. When using the on-line Docket Management System (, enter only the middle series of digits, 7967.

    [4] FedEx did not directly provide the agency with the details regarding the pilot study, however two commenters stated that FedEx study resulted in a reduction in backing crashes. Nevada Automotive Test Center (NHTSA-2000-7967-7) provided the 33 percent figure whereas the International Brotherhood of Teamsters (NHTSA-2000-7967-8) only qualitatively mentioned that a reduction in incidences occurred.

    [5] "Comprehensive costs" combines both economic costs and values for "intangible" consequences, such as pain and suffering, quality of life, and so on.

    [6] Lind, R.C., "A Primer on the Major Issues Relating to the Discount Rate for Evaluating National Energy Options," in Discounting for Time and Risks in Energy Policy, 1982, (Washington, D.C., Resources for the Future, Inc.).

    [7] J. Kolb and J.D. Sheraga, "A Suggested Approach for Discounting the Benefits and Costs of Environmental Regulations: unpublished working papers.

    [8] Moore, M.J. and Viscusi, W.K., "Discounting Environmental Health Risks: New Evidence and Policy Implications", Journal of Environmental Economics and Management, V. 18, No. 2, March 1990, part 2 of 2.