November 17 - Insurance Institute for Highway Safety
Transcription
November 17 - Insurance Institute for Highway Safety
November 17, 2008 David Kelly Acting Administrator National Highway Traffic Safety Administration 1200 New Jersey Avenue, SE, West Building Washington, DC 20590 Notice of Proposed Rulemaking; 49 CFR Part 571 Federal Motor Vehicle Safety Standards, Motorcycle Brake Systems; Docket No. NHTSA-2008-0150 Dear Mr. Kelly: On September 17, 2008, the National Highway Traffic Safety Administration (NHTSA) announced a proposal to amend Federal Motor Vehicle Safety Standard (FMVSS) No. 122, Motorcycle Brake Systems. The amendments would strengthen the requirements and test procedures that cover many aspects of motorcycle brake systems, including antilock braking systems (ABS) voluntarily installed by manufacturers. The Insurance Institute for Highway Safety (IIHS) supports these changes and submits with this comment our recent research showing that ABS is improving motorcycle safety. The proposed changes to FMVSS 122 will create minimum performance standards that have been tested successfully elsewhere in the world and will help keep motorcycles with unsafe brakes from being sold in the United States. The proposed ABS tests in particular will help ensure that if ABS is installed on motorcycles, the system will provide operators with adequate stopping distances and stability. IIHS and its affiliated Highway Loss Data Institute (HLDI) are studying the safety effects of motorcycle ABS as its availability increases. Results from our initial analysis of a variety of models show that ABS is having a positive effect (IIHS, 2008). Both the rate of fatal motorcycle crashes and the frequency of crashes for which insurance collision claims are filed are lower among motorcycles with ABS compared with the same motorcycles without ABS. The attached IIHS study (Teoh, 2008) shows 6.7 fatal crashes per 10,000 registered vehicle years among motorcycles not equipped with ABS during 2005-06. The corresponding rate for the same models equipped with optional ABS was 4.1 fatal crashes per 10,000 registered vehicle years — 38 percent lower. The attached HLDI (2008) study shows that the estimated effect of ABS was a 21 percent decrease in overall collision losses, primarily because collision claim frequencies for motorcycles with ABS were 19 percent lower than for motorcycles without ABS. The importance of equipping motorcycles with ABS increases as motorcycling continues to grow in popularity. Motorcycle sales more than tripled between 1997 and 2005, and motorcyclist deaths have more than doubled since 1997. Due largely to safety features that are increasingly available as standard equipment on passenger vehicles, there were fewer passenger vehicle occupant deaths in 2007 than in any year since NHTSA began collecting these data in 1975. On the other hand, more motorcyclists died in crashes in 2007 than in any other year during the same time period. David Kelly November 17, 2008 Page 2 The Institute supports the proposed strengthening of FMVSS 122 to establish stronger minimum performance standards for motorcycle brake systems. Given the beneficial effects of ABS, as documented in our research, we urge NHTSA to consider further changes to FMVSS 122 to encourage or require ABS on all motorcycles. Sincerely, Joseph M. Nolan, M.S. Senior Vice President, VRC Operations cc: Docket Clerk, Docket No. NHTSA-2008-0150 References and Attachments Highway Loss Data Institute. 2008. Motorcycle antilock braking system (ABS). HLDI Bulletin 25(1). Arlington, VA. Insurance Institute for Highway Safety. 2008. Antilock brakes on motorcycles reduce both crash frequencies and deaths. Status Report 42(9):1-3. Arlington, VA. Teoh, E.R. 2008. Effectiveness of antilock braking systems in reducing fatal motorcycle crashes. Arlington, VA: Insurance Institute for Highway Safety. SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 1 Vol. 43, No. 9, Oct. 22, 2008 IF THEY NEED TO STOP on a dime, these riders will enjoy an advantage most others don’t because of the antilock brakes on their motorcycle. Two new studies indicate crash reductions associated with anti- SR43-9 TO PDF:SR 43-9 2 10/16/08 11:37 AM Page 2 Status Report, Vol. 43, No. 9, Oct. 22, 2008 locks. Both the frequency of crashes for which insurance claims are filed and the rate of fatal motorcycle crashes go down among bikes with antilock brakes. The importance of equipping bikes with antilocks increases as motorcycling proliferates. Motorcycle sales more than tripled from 1997 to 2005. Deaths of motorcyclists have more than doubled since 1997, with some kinds of bikes having much higher death rates than others (see Status Report, Sept. 11, 2007; on the web at iihs.org). About 5,000 motorcyclists died in crashes last year. The new study of fatal motorcycle crashes was conducted by Institute researchers, while the analysis of insurance claims is by researchers at the affiliated Highway Loss Data Institute (HLDI). Adrian Lund is president of both organizations. “Even though adding antilocks won’t make motorcycling as safe as going by car, it’s something manufacturers can do to reduce the risk of traveling on 2 wheels instead of 4,” Lund says. “It’s a way to reduce the chances of overturning a bike and crashing, so it can save lives among people who choose motorcycles for their basic transportation, to save on gasoline, or just for fun.” When antilocks are needed: Stopping a motorcycle is trickier than stopping a car. For one thing, front and rear wheels typically have separate brake controls. Both underbraking and overbraking the front and rear wheels contribute to crashes (see Status Report, June 21, 1979). In an emergency, a rider faces a split-second choice to brake hard, which can lock the wheels and cause a motorcycle to overturn, or to hold back on the brakes and risk running headlong into the emergency. This is when antilocks can help. They reduce brake pressure when they detect impending lockup and increase the pressure again when traction is restored. Brake pressure is evaluated multiple times per second, so riders may fully brake without fear of locking the wheels. Antilocks won’t prevent every motorcycle crash. They won’t help a rider who’s Regression analysis revealed 21 percent lower insurance losses for motorcycles with antilocks, primarily because the claim frequency was 19 percent lower than for bikes without antilocks. These findings are based on a dataset of 72,000 insured years of 2003- FATAL CRASHES INSURANCE CLAIMS PER 10,000 MOTORCYCLE REGISTRATIONS, 2001-06 MODELS DURING 2005-06 PERCENT CHANGE IN COLLISION LOSSES FOR 2003-07 MODEL MOTORCYCLES WITH ANTILOCKS 6 0 4 -10 2 without antilocks with antilocks about to be struck from behind, for example. But the new studies indicate that antilocks reduce crashes overall and save lives. Crash reduction benefit: The HLDI study compares insurance losses under collision coverage for 12 motorcycle models with optional antilock brakes versus the same models without this option. The researchers evaluated the effects of antilock brakes on both the frequency of insurance claims that are filed for crash damage and the average cost of the damage, after accounting for rider age and gender, motorcycle age, and other factors that influence the likelihood of a crash. claim frequency claim severity SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 3 Status Report, Vol. 43, No. 9, Oct. 22, 2008 07 model Honda, Suzuki, Triumph, and Yamaha bikes (an insured year is 1 motorcycle insured for 1 year or 2 insured for 6 months each, etc.). BMW models aren’t included because it’s impossible to determine from vehicle identification numbers which ones have optional antilocks and which don’t. Harley-Davidsons aren’t included because antilocks were added after the study years. Antilock brakes “appear to reduce collision claims,” says Matthew Moore, HLDI vice president and lead author of the study, “but they don’t affect the severity of the crashes for which claims are filed. The cost of these claims doesn’t go down.” Lives are being saved: In a complementary study, Institute researchers examined rates of fatal crashes of motorcycle models with and without antilocks. Eight models were studied, a subset of the 12 included in the HLDI analysis. The other 4 models were excluded because of sample size limitations. A main finding is that there were 6.6 fatal crashes per 10,000 registered motorcycles without antilocks during 2005-06. The corresponding rate for the same bike models equipped with optional antilocks is 4.1, or 38 percent lower. Institute statistician Eric Teoh, author of the study, says the findings are statistically significant at the 90 percent confidence level. Antilocks on cars versus motorcycles: Passenger cars began to be equipped with antilock brakes during the 1970s, after studies conducted on the test track indicated they reduce stopping distances. However, this promise didn’t pan out in real-world crashes (see Status Report, Jan. 29, 1994). Antilocks didn’t reduce relevant collisions. “It isn’t surprising that antilock brakes are more beneficial on motorcycles than they are on cars because the 2-wheelers are so much less stable, and it’s this instability that contributes to so many crashes,” Lund points out. “By reducing wheel lockup during braking, antilocks keep a lot of motorcycles from overturning.” Antilock brakes are recent additions to motorcycles. They’re available almost exclusively as optional equipment (see list), which means shoppers have to find models on which the option is offered and then pay extra for it. Antilocks were on only 18 percent of the motorcycles included in the new studies of effectiveness. For a copy of “Antilock braking systems for motorcycles and insurance collision losses” by M. Moore and Y. Yan or “Effectiveness of antilock braking systems in reducing fatal motorcycle crashes” by E. Teoh, write: Publications, Insurance Institute for Highway Safety, 1005 N. Glebe Rd., Arlington, VA 22201, or email [email protected]. 2008 MODELS WITH ANTILOCKS Antilocks are optional except as noted; models in bold are included in one or both of the new studies of the effectiveness of antilock brakes BMW BMW BMW BMW BMW BMW BMW BMW BMW BMW BMW BMW Can-Am Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Harley-Davidson Honda Honda Honda Honda Honda Kawasaki Moto Guzzi Suzuki Suzuki Suzuki Suzuki Suzuki Triumph Triumph Yamaha Yamaha K1200GT (std) K1200LT (std) R1200RT (std) F800S/F800ST G650 Xchallenge G650 XCountry G650 XMoto HP2 Megamoto/HP2 K1200R/K1200R Sport K1200S R1200R/R1200S R900RT Spyder Electra Glide Classic Electra Glide Standard Night Rod Night Rod Special Road Glide/Road King Road King Classic Screaming Eagle Electra Glide Screaming Eagle Road King Street Glide Ultra Classic Electra Glide V-Rod Gold Wing Interceptor 800 Reflex Silver Wing ST1300 Concours 14 Norge 1200 Bandit 1250S B-King Burgman 650 Executive SV650/SV650S/SV650SAF V-Strom 650 Sprint ST Tiger FJR1300 (std) FJR1300 Electric Shift (std) 3 SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 4 4 Status Report, Vol. 43, No. 9, Oct. 22, 2008 MORE STATES BAN DRIVERS’ TEXTING California and Alaska are the latest US states to ban text messaging by drivers of all ages, not just teenagers. The two states join Connecticut, the District of Columbia, Louisiana, Minnesota, New Jersey, and Washington in banning texting by all drivers amid concern that such distractions increase crash risk. Alaska’s law took effect Sept. 1. California’s ban begins Jan. 1, 2009. Both states make the use of an electronic device to write, send, or read text messages a primary offense, meaning that police officers can pull over drivers solely for violating the bans. Texting while driving also is a primary offense in Connecticut, the District of Columbia, Minnesota, and New Jersey. Bans in Louisiana and Washington are secondary, so motorists must be violating another traffic law in order to be stopped by police for texting. Nine states have text messaging bans that apply only to novice drivers. For details on state bans on text messaging by drivers, go to www.iihs.org/laws/cellphonelaws.aspx. There’s lots of anecdotal evidence tying texting while driving to crashes, but not much data from real-world collisions. Studies have linked cellphone use with crash risk (see Status Report, July 16, 2005, and March 22, 1997; on the web at iihs.org). In one of the first published studies on texting and driving, the Transport Research Laboratory in the United Kingdom (on the web at trl.co.uk) found that texting degrades performance in a driving simulator. Researchers found that composing a text message affected driving more than reading one. The 17 drivers in the study — all were 17-24 years old — had slower reaction times, were more likely to drift out of their virtual lanes, and were more likely to reduce their speeds while they were texting. CHILD SEAT USE AMONG KIDS IN CRASHES GOES UP Use of child safety seats has surged since 1999 among restrained children younger than 9 riding in insured vehicles. Restraint types also have changed. These are the main findings of new research from the decadelong Partners for Child Passenger Safety study of the Children’s Hospital of Philadelphia (CHOP) and State Farm, with support from the Association of International Automobile Manufacturers. It’s based on 19982007 insurance claims and phone survey data on more than 875,000 kids in crashes. Overall safety seat use among restrained children 8 and younger rose to 80 percent in 2007 from 51 percent 8 years earlier. Virtually 100 percent of restrained children 3 and younger in crashes have been in safety seats since 1999. Safety seat use is much lower among older children. Progress has been made, but there’s room to improve. In 1999 only 15 percent of restrained 4-8 year-olds in the CHOP study were in an appropriate restraint — a harness restraint or booster. By 2007 appropriate restraint use in this group had quadrupled to 63 percent. The rest of restrained 4-8 year-olds rode in adult belts alone. Typically, such belts don’t begin to fit properly until kids grow to about 4 feet, 9 inches tall. “Along with the increase in the number of kids riding in child safety seats, we can also see changes in the types of restraints they are using now versus 10 years ago,” says Kristy Arbogast, director of engineering at CHOP’s Center for Injury Research and Prevention, where the study was conducted. She says more restrained 4 and 5 year-olds ride in boosters now instead of harness restraints. Only 31 percent of appropriately restrained 4-5 year-olds rode in harness restraints during 2007. Highback boosters are slightly more popular now than backless ones among restrained 4-5 year-olds. But backless boosters are used nearly 3 times as often as highbacks for 6-8 year-olds. SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 5 Status Report, Vol. 43, No. 9, Oct. 22, 2008 5 Previous CHOP research shows boosters lower crash injury risk by 59 percent for 4-7 year-olds compared with belts alone. Boosters elevate children so lap and shoulder belts are properly positioned. Earlier this month the Institute released evaluations of 41 booster models, finding that several Of the states in the study, booster seat use among 4-8 year-olds was lowest in Ohio (18 percent) and Texas (20 percent). Not surprisingly, these states don’t have booster seat laws. On the other hand, 2 of the 5 states with the highest use of boosters, Pennsylvania (72 percent) and Illinois (62 Among the study’s other findings are that 60 percent of crashes involving children occur within 10 minutes of home, and 84 percent take place within 20 minutes of home. Only 14 percent of crashes are on roads where posted speed limits are 55 mph or higher, but these crashes result in the don’t improve belt fit (see Status Report, Oct. 1, 2008; on the web at iihs.org). Arbogast attributes the increase in booster use among older kids to education of parents and caregivers plus state laws requiring older kids to ride in safety seats. Laws in 43 states and the District of Columbia include booster provisions (on the web at iihs.org/ laws/restraintoverview.aspx). “More parents than ever now realize that kids need the help of a booster seat to make sure the belt fits properly across the bony parts of their lap and shoulder rather than across the soft belly or the neck, which are more prone to injury,” Arbogast says. percent), do require child restraints or boosters for children through age 7. CHOP researchers found that parents aren’t widely using lower anchors and tethers for children, or LATCH, which are supposed to make it easier to attach infant and child restraints securely to vehicle seats (see Status Report, Jan. 16, 1999; on the web at iihs.org). LATCH has been required in new vehicles and on child restraints since 2002. However, only 43 percent of all children buckled into restraints in vehicles equipped with LATCH in 2007 were riding in seats attached to the lower anchors, the CHOP study reports. highest rates of injury. Nearly half of all crashes involving children occur on roads with posted speed limits of 25 to 44 mph. Although the American Academy of Pediatrics recommends that children younger than 13 ride in the back seats of vehicles, about 30 percent of all 8-12 year-olds ride in the front. Positioning children in back seats reduces the risk of fatal injuries in crashes by about one-third among kids 12 and younger (see Status Report, June 27, 1997; on the web at iihs.org). For a copy of the September 2008 “Partners for Child Passenger Safety: fact and trend report” go to www.chop.edu/carseat. SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 6 6 Status Report, Vol. 43, No. 9, Oct. 22, 2008 FASTER, HEAVIER GOLF CARTS GET THUMBS DOWN FROM FEDERAL REGULATORS The federal government says it isn’t willing to trade highway safety for fuel economy in denying 4 petitions that sought to increase the maximum gross vehicle weight for low-speed vehicles and also launch a class of medium-speed vehicles. The September decisions by the National Highway Traffic Safety Administration (NHTSA) mean that the weight and top speed of these golf cart-like vehicles, which don’t have to meet all the safety rules that apply to cars, remain capped at 3,000 pounds and 25 mph. “While NHTSA agrees with the importance of environmental issues, the agency believes that it is neither necessary nor appropriate to significantly increase the risk of deaths and serious injuries to save fuel,” the agency said in denying petitions from Environmental Motors, Porteon Electric Vehicles Inc., and Mirox Corporation. The companies asked the agency to create a class of medium-speed vehicles with speed capabilities of up to 35 mph, arguing they’d fill a need for fuel-efficient vehicles for use in fast urban traffic. They’re known as neighborhood electric vehicles, street-legal golf carts, and minitrucks, among other names (see Status Report, April 6, 2002; on the web at iihs.org). These electric or gasoline-powered low-speed vehicles are designed to haul people and cargo on private land, such as retirement communities, farms, amusement parks, and construction sites, but they’re often driven on public streets. Forty-six states regulate their use, with most limiting their speed to no greater than 25 mph (on the web at iihs.org/ laws) on public roadways with speed limits of no more than 35 mph. They’re exempt from most federal safety standards that apply to cars, and they aren’t required to meet any criteria for vehicle crashworthiness, so they’d be out of their league in crashes with other vehicles going 35 mph. Electronic Transportation Applications had sought to increase the maximum allowable gross vehicle weight for electric-powered low-speed vehicles to 4,000 pounds. In denying Electronic Transportation Applications’ petition, NHTSA said, “We believe that vehicles over 3,000 pounds are capable of complying with the full requirements” of the federal motor vehicle safety standards. Increasing the allowable gross vehicle weight, NHTSA said, “would encourage the use of [low-speed vehicles] in circumstances where it could pose an unreasonable risk to safety.” The agency noted that some of the smallest passenger cars — Honda Insight and Toyota Echo, for example — have gross vehicle weights of about 3,000 pounds or less and still comply with safety standards. (Read both decisions at http://edocket.access.gpo.gov/2008/pdf/E8-22736.pdf and 22737.pdf.) Federal crash databases don’t include a specific category for low-speed vehicles so it’s hard to track their crashes. News reports frequently chronicle deaths and injuries that result when these vehicles collide with larger passenger vehicles. SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 7 NEW ‘UNDERSTANDING CRASHES’ VIDEO Why do some car crashes produce only minor injuries? How can a single crash of a car into a wall involve 3 separate collisions? Award-winning science educator Griff Jones visits the Institute’s Vehicle Research Center to answer these and other questions in a 24-minute video that’s a follow-up to a previous Institute production, “Understanding car crashes: it’s basic physics” (2000). In the new video, Jones examines the laws of nature that determine what happens to the human body in a crash. Order “Understanding car crashes: when physics meets biology” ($35) online at iihs.org/videos. SR43-9 TO PDF:SR 43-9 10/16/08 11:37 AM Page 8 NON-PROFIT ORG. U.S. POSTAGE PAID PERMIT NO. 252 ARLINGTON, VA 1005 N. Glebe Rd., Arlington, VA 22201 Phone 703/247-1500 Fax 247-1588 Internet: www.iihs.org Vol. 43, No. 9, Oct. 22, 2008 Antilock brakes on motorcycles reduce both crash frequencies and deaths ...........1 2008 motorcycles with antilocks .....................3 Text messaging bans now cover all drivers in 7 states and the District of Columbia .............4 Child safety seat use in crashes has increased during the past decade ....................................4 Souped-up golf carts will remain subject to limitations on weight and speed, the federal government decides ........................................6 New Institute video explores what happens to the human body in a car crash ...................7 Contents may be republished with attribution. This publication is printed on recycled paper. The Insurance Institute for Highway Safety is a nonprofit scientific and educational organization dedicated to reducing deaths, injuries, and property damage from crashes on the nation’s highways. The Institute is wholly supported by auto insurers: 21st Century Insurance AAA Mid-Atlantic Insurance Group Affirmative Insurance AIG Agency Auto AIG Direct Alfa Insurance Alfa Alliance Insurance Corporation Allstate Insurance Group American Family Mutual Insurance American National Property and Casualty Ameriprise Auto & Home Amerisure Insurance Amica Mutual Insurance Company Auto Club Group Auto Club South Insurance Company Bituminous Insurance Companies Bristol West Insurance Brotherhood Mutual California State Automobile Association Capital Insurance Group Chubb Group of Insurance Companies Concord Group Insurance Companies Cotton States Insurance COUNTRY Financial Countrywide Insurance Group Erie Insurance Group Esurance Farm Bureau Financial Services Farm Bureau Mutual Insurance Company of Idaho Farmers Insurance Group of Companies Farmers Mutual of Nebraska First Acceptance Corporation Florida Farm Bureau Insurance Companies Frankenmuth Insurance The GEICO Group Gainsco General Casualty Insurance Companies Georgia Farm Bureau Insurance GMAC Insurance Grange Insurance The Hartford Hanover Insurance Group High Point Insurance Group Homeowners of America ICW Group Indiana Farm Bureau Insurance Kemper, a Unitrin Business Kentucky Farm Bureau Insurance Liberty Mutual The Main Street America Group Markel Corporation Mercury Insurance Group MetLife Auto & Home Michigan Insurance Company MiddleOak MMG Insurance Mutual of Enumclaw Insurance Company Nationwide Insurance N.C. Farm Bureau Mutual Insurance Company Nodak Mutual Insurance Norfolk & Dedham Group Ohio Casualty Group Oklahoma Farm Bureau Mutual Insurance Company Oregon Mutual Insurance Palisades Insurance Pekin Insurance PEMCO Insurance The Progressive Corporation Response Insurance Rockingham Group Safeco Insurance Samsung Fire & Marine Insurance Company S.C. Farm Bureau Mutual Insurance Company SECURA Insurance Shelter Insurance Sompo Japan Insurance Company of America State Auto Insurance Companies State Farm Tennessee Farmers Mutual Insurance Company Tokio Marine Nichido The Travelers Companies Unitrin USAA Virginia Farm Bureau Mutual Insurance West Bend Mutual Insurance Company Zurich North America FUNDING ASSOCIATIONS American Insurance Association National Association of Mutual Insurance Companies Property Casualty Insurers Association of America Highway Loss Data Institute Bulletin Motorcycle Antilock Braking System (ABS) VOL. 25, NO. 1 Antilock braking systems (ABS) on motorcycles are designed to allow riders to make urgent, yet controlled stops. Motorcycles are fundamentally unbalanced; they are kept stable at very low speeds by a rider holding the handlebar and maintaining balance. At higher speeds, stability comes almost exclusively from the gyroscopic effect of the wheels. While at speed, if one of the wheels stops rotating for a fraction of a second, the result is immediate instability. The effect is more pronounced if it happens to the front wheel, where a fall is almost inevitable — especially while cornering or leaning the motorcycle. ABS has independent braking sensors for each wheel. If the system detects a locked wheel, it releases the brake to allow that tire to retain grip before reapplying the brake. ABS then modulates braking pressure to achieve optimum braking. This Highway Loss Data Institute (HLDI) bulletin compared the collision losses of 12 motorcycle models available with optional ABS. Model years of the motorcycles studied ranged from 2003 to 2007. Significant reductions in collision claim frequencies and overall losses were found for motorcycles equipped with ABS. No significant reductions were found for claim severity. For motorcycles to be included in the study, their vehicle identification numbers (VINs) had to have an ABS indicator. This allowed for very tight control over the study population. Twelve motorcycles met this criterion. It should be noted that there were motorcycles available with ABS that were not included in the study because their VINs did not have an ABS indicator. Data were collected by make and series, rated driver age and gender, and vehicle age and density. Vehicle age was defined as the difference between calendar year and model year measured in years. Previous HLDI studies have shown that all of these factors have a significant impact on insurance losses. For insurance purposes, a rated driver is assigned to each vehicle on a policy. Information on the actual driver at the time of a loss is not available in the HLDI database. For the present study data were stratified by rated driver age group (<25, 25-39, 4064, 65+, or unknown) and gender (male, female, or unknown). The dataset also was stratified by make/series and vehicle density (<100, 100-499, and 500+ vehicles per square mile). For example, a 1-year-old Honda Gold Wing, ABS equipped, with a 40-64 year old male as the rated driver, and garaged in an area with a vehicle density of 100-499 vehicles per square mile constituted one unit of observation. The distribution of motorcycle collision exposure for the six independent variables is listed in Appendix A. It is important to note that rated driver factors and vehicle density were included to control for their potential impact on losses and not to produce estimates for APRIL 2008 those variables. The estimated parameters for those variables may not generalize from this subset to the much larger motorcycle population. Regression analysis was used to quantify the effect of ABS on motorcycle collision losses while controlling for other covariates. Claim frequency was modeled using a Poisson distribution, whereas claim severity was modeled using a Gamma distribution. Both models used a logarithmic link function. Estimates for collision overall losses were derived from the claim frequency and severity models. Reference categories for the categorical independent variables were assigned to the values with the highest exposure. The reference categories were as follows: make/series = Honda Gold Wing, ABS = without ABS, rated driver age range = 25-39, vehicle density = 100-499 vehicles per square mile, and rated driver gender = male. Losses for each unit of observation were weighted by the exposure in the linear regression. The key independent variable in the model, ABS, was treated as categorical. Models were constructed that examined the interaction of the rated driver factors and vehicle density with the presences or absence of ABS. None of those interactions were found to be significant. Summary results of the regression analysis of motorcycle collision claim frequencies using the Poisson distribution are listed in Table 1. Results for all independent variables in the model, including ABS, had p-values less then 0.05, indicating their effects on claim frequencies were statistically significant. Detailed results of the regression analysis using claim frequency as the dependent variable are listed in Table 2. The table shows estimates and significance levels for the individual values of the categorical variables. To make results more illustrative, a column was added that contains the exponents of the estimates. The exponent of the intercept equals 0.0000826 claims per day, or about 3 claims per 100 insured vehicle years. The intercept outlines losses for the reference (baseline) categories: the estimate corresponds to the claim frequency for a new Honda Gold Wing without ABS, garaged in a medium vehicle density area, and driven by a male age 25-39. The remaining estimates are in the form of multiples, or ratios relative to the reference categories. For example, the estimate corresponding to female gender equals 0.81, so female rated drivers had estimated claim frequencies 19 percent lower than those for male rated drivers. The estimate corresponding to motorcycle ABS (-0.22) was highly significant (p=0.002). The estimate corresponded to a 19 percent reduction in claim frequencies for motorcycles equipped with ABS. TABLE 1 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM FREQUENCIES DEGREES OF FREEDOM Vehicle Age 1 CHI-SQUARE P-VALUE 196.270 < 0.0001 < 0.0001 Rated Driver Age 4 57.040 Vehicle Density 2 18.150 0.0001 Vehicle Make/Series 11 292.300 < 0.0001 Rated Driver Gender 2 8.060 0.0178 ABS 1 9.580 0.0020 TABLE 2 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM FREQUENCIES PARAMETER ESTIMATE EXPONENT (ESTIMATE) STANDARD ERROR CHISQUARE INTERCEPT VEHICLE AGE RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ VEHICLE DENSITY 0-99 100-499 500+ RATED DRIVER GENDER Female Male Unknown VEHICLE MAKE/SERIES Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Suzuki Bandit 1250 Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 ABS ABS Model Non-ABS Model -9.4014 -0.2628 3.015 0.769 0.0826 0.0192 12948.8 187.54 <0.0001 <0.0001 0.1791 0.5839 0 -0.1433 -0.0747 1.196 1.793 1.000 0.866 0.928 0.0925 0.094 0 0.0649 0.1026 3.75 38.58 0.0529 <0.0001 4.88 0.53 0.0272 0.4663 -0.1332 0 0.122 0.875 1.000 1.130 0.0582 0 0.0514 5.23 0.0222 5.62 0.0177 -0.205 0 0.0431 0.815 1.000 1.044 0.0845 0 0.0527 5.88 0.0153 0.67 0.4135 0 0.8988 0.5596 0.7658 0.1745 1.4525 0.7284 1.0251 -0.084 1.077 0.3632 0.3276 1.000 2.457 1.750 2.151 1.191 4.274 2.072 2.787 0.919 2.936 1.438 1.388 0 0.1077 0.1093 0.1025 0.1128 0.2295 0.0954 0.0747 0.1352 0.1357 0.2177 0.0949 69.6 26.2 55.81 2.39 40.06 58.29 188.54 0.39 63 2.78 11.92 <0.0001 <0.0001 <0.0001 0.1218 <0.0001 <0.0001 <0.0001 0.5345 <0.0001 0.0953 0.0006 -0.2151 0 0.806 1.000 0.0709 0 9.19 0.0024 P-VALUE Motorcycle collision claim frequencies increased with increases in vehicle density. Claim frequencies in high vehicle density areas were estimated to be 13 percent higher (p=0.02) than those in medium vehicle density areas, whereas claim frequencies in low vehicle density areas were estimated to be 13 percent lower (p=0.02). Claim frequencies were estimated to decrease 23 percent (p<0.0001) for each 1-year increase in vehicle age. Individual make/series motorcycles were included in the model, and estimates of their effect on collision claim frequencies were found to be significant. As previously mentioned, the reference category for the make/series variable was the Honda Gold Wing. Significant predictions for make/series ranged from 1.39 for the Yamaha FJR 1300 to 4.27 for the Suzuki Bandit 1250. All make/series estimates were significant at the p=0.001 level except for the Honda ST1300, Suzuki V-Strom 650, and Triumph Tiger. The lack of significance for these individual values of the make/series variable indicate that estimated claim frequencies for these motorcycles were similar to the Honda Gold Wing (reference category). Driver age was highly significant in predicting motorcycle collision claim frequency. Estimated claim frequencies for rated drivers 24 and younger were 79 percent higher (p<0.0001) than those for rated drivers ages 25-39 (reference category), whereas estimated claim frequencies for rated drivers ages 40-64 were 13 percent lower (p=0.03). The estimated 7 percent decrease in claim frequency for rated drivers 65 and older was not significant. Driver gender also significantly predicted collision claim frequencies. Estimated claim frequencies for female rated riders were 19 percent lower (p=0.02) than those for male rated riders. Summary results of the regression analysis of motorcycle collision claim severities using the Gamma distribution are listed in Table 3. Of the six variables included in the analysis, only vehicle age and make/series had p-values less than 0.05. Neither the rated driver nor the driving environment affects the claim size. TABLE 3 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM SEVERITIES DEGREES OF FREEDOM CHI-SQUARE P-VALUE Vehicle Age 1 12.400 0.0004 Rated Driver Age 4 3.420 0.4907 2 2.490 0.2880 Vehicle Make/Series Vehicle Density 11 316.960 < 0.0001 Rated Driver Gender 2 0.190 0.9075 ABS 1 0.070 0.7953 Detailed results of the regression analysis using motorcycle collision claim severity as the dependent variable are listed in Table 4. The structure of the table, as well as the variables and reference categories, are the same as those used for claim frequency in Table 2. The variables and reference categories that were used for claim frequency were used for claim severity. The exponent of the intercept equals $9,089. The intercept outlines losses for the reference (baseline) categories: the estimate corresponds to the claim severity for a new Honda Gold Wing without ABS, garaged in a medium vehicle density area, and driven by a male age 25-39. The estimate corresponding to the ABS effect, a 2 percent decrease in claim severity, was highly nonsignificant (p=0.8), indicating ABS does not affect claim severity. As previously mentioned, vehicle age and make/series were significant predictors of claim severity. Not surprisingly, as motorcycles age their claim severities decrease. The model estimated a 6 percent decrease (p=0.0004) in claim severity per 1-year increase in vehicle age. Estimated claim severities for the 11 make/series motorcycles, compared with those for the Honda Gold Wing (reference category), ranged from 21 percent lower for the Honda ST1300 to 75 percent lower for the Honda Reflex. All of the make/series estimates were significant at the p<0.05 level. TABLE 4 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM SEVERITIES EXPONENT (ESTIMATE) STANDARD ERROR CHISQUARE PARAMETER ESTIMATE P-VALUE INTERCEPT VEHICLE AGE RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ VEHICLE DENSITY 0-99 100-499 500+ RATED DRIVER GENDER Female Male Unknown VEHICLE MAKE/SERIES Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Suzuki Bandit 1250 Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 ABS ABS Model Non-ABS Model 9.1148 -0.0608 9088.817 0.941 0.0849 0.0172 11515.1 12.57 <0.0001 0.0004 0.0819 0.1743 0 0.0624 0.0415 1.085 1.190 1.000 1.064 1.042 0.0838 0.1029 0 0.0586 0.0866 0.95 2.87 0.3286 0.0905 1.14 0.23 0.2866 0.6317 -0.0342 0 0.0553 0.966 1.000 1.057 0.0563 0 0.0514 0.37 0.5432 1.16 0.2822 0.0319 0 0.0078 1.032 1.000 1.008 0.0728 0 0.0501 0.19 0.6613 0.02 0.8766 0 -0.5373 -1.3808 -1.1479 -0.2339 -0.8037 -0.9265 -0.9031 -0.989 -0.5166 -0.5712 -0.4631 1.000 0.584 0.251 0.317 0.791 0.448 0.396 0.405 0.372 0.597 0.565 0.629 0 0.0928 0.0916 0.0922 0.0989 0.1774 0.0908 0.0801 0.1179 0.1154 0.1755 0.0863 33.54 227.27 154.93 5.59 20.53 104.19 127.02 70.33 20.03 10.59 28.81 <0.0001 <0.0001 <0.0001 0.018 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0011 <0.0001 -0.0155 0 0.985 1.000 0.0597 0 0.07 0.795 Table 5 summarizes the effects of the independent variables on motorcycle collision overall losses, derived from the claim frequency and severity models. Overall losses can be calculated by simple multiplication because the estimates for the effect of ABS on claim frequency and severity were in the form of ratios relative to the reference (baseline) categories. The standard error for overall losses can be calculated by taking the square root of the sum of the squared standard errors for claim frequency and severity. Based on the value of the estimate and the associated standard error, the level of statistical significance (p-value) can be obtained from a probability distribution table. The estimated effect of ABS was a 21 percent decrease in collision overall losses that was significant (p=0.01). This is a strong indication that ABS is effective in reducing collision overall losses for motorcycles. Motorcycle collision overall losses were predicted to increase with increased vehicle density. Estimated overall losses in high vehicle density areas were 19 percent higher (p=0.01) than those medium vehicle density in areas (reference category), whereas estimated overall losses in low vehicle density areas were 15 percent lower (p=0.04). Estimated overall losses for the 11 make/series motorcycles, compared with those for the Honda Gold Wing (reference category), ranged from 66 percent lower for the Suzuki V-Strom 650 to 91 percent higher for the Suzuki Bandit 1250. Only about half of the make/series estimates were statistically different from the reference category. Driver age was a significant predictor of motorcycle collision overall losses. Estimated overall losses for rated drivers 24 and younger were 113 percent higher (p<0.0001) than those for rated drivers ages 25-39 (reference category). Rated drivers ages 40-64 and 65 and older had slightly lower estimated overall losses, but these estimates did not reach statistical significance. Vehicle age also was significantly predictive of collision overall losses. Overall losses decreased by an estimated 28 percent (p<0.0001) for each 1-year increase in vehicle age. TABLE 5 RESULTS FOR COLLISION OVERALL LOSSES DERIVED FROM CLAIM FREQUENCY AND SEVERITY MODELS PARAMETER INTERCEPT VEHICLE AGE RATED DRIVER AGE Unknown 1. 14-24 25-39 40-64 65+ DENSITY 0-99 100-499 500+ RATED DRIVER GENDER Female Male Unknown VEHICLE MAKE/SERIES Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Suzuki Bandit 1250 Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 ABS ABS Model Non-ABS Model FREQUENCY STANDARD ESTIMATE ERROR SEVERITY STANDARD ESTIMATE ERROR ESTIMATE OVERALL LOSSES STANDARD EXPONENT ERROR (ESTIMATE) P-VALUE -9.4014 -0.2628 0.0826 0.0192 9.1148 -0.0608 0.0849 0.0172 -0.2866 -0.3236 0.1185 0.0258 0.7508 0.7235 0.0155 <0.0001 0.1791 0.5839 0 -0.1433 -0.0747 0.0925 0.094 0 0.0649 0.1026 0.0819 0.1743 0 0.0624 0.0415 0.0838 0.1029 0 0.0586 0.0866 0.2610 0.7582 0 -0.0809 -0.0332 0.1248 0.1394 0 0.0874 0.1343 1.2982 2.1344 1.0000 0.9223 0.9673 0.0365 <0.0001 -0.1332 0 0.122 0.0582 0 0.0514 -0.0342 0 0.0553 0.0563 0 0.0514 -0.1674 0 0.1773 0.0810 0 0.0727 0.8459 1.0000 1.1940 0.0387 -0.205 0 0.0431 0.0845 0 0.0527 0.0319 0 0.0078 0.0728 0 0.0501 -0.1731 0 0.0509 0.1115 0 0.0727 0.8411 1.0000 1.0522 0.1207 0.4839 0 0.8988 0.5596 0.7658 0.1745 1.4525 0.7284 1.0251 -0.084 1.077 0.3632 0.3276 0 0.1077 0.1093 0.1025 0.1128 0.2295 0.0954 0.0747 0.1352 0.1357 0.2177 0.0949 0 -0.5373 -1.3808 -1.1479 -0.2339 -0.8037 -0.9265 -0.9031 -0.989 -0.5166 -0.5712 -0.4631 0 0.0928 0.0916 0.0922 0.0989 0.1774 0.0908 0.0801 0.1179 0.1154 0.1755 0.0863 0 0.3615 -0.8212 -0.3821 -0.0594 0.6488 -0.1981 0.1220 -1.0730 0.5604 -0.2080 -0.1355 0 0.1422 0.1426 0.1379 0.1500 0.2901 0.1317 0.1095 0.1794 0.1781 0.2796 0.1283 1.0000 1.4355 0.4399 0.6824 0.9423 1.9132 0.8203 1.1298 0.3420 1.7514 0.8122 0.8733 0.0110 <0.0001 0.0056 0.6921 0.0253 0.1325 0.2653 <0.0001 0.0017 0.4570 0.2908 -0.2151 0 0.0709 0 -0.0155 0 0.0597 0 -0.2306 0 0.0927 0 0.7941 1.0000 0.3549 0.8047 0.0147 0.0128 APPENDIX A DISTRIBUTION OF EXPOSURE FOR INDEPENDENT VARIABLES VEHICLE AGE 0 1 2 3 4 5 RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ VEHICLE DENSITY 0-99 100-499 500+ RATED DRIVER GENDER Female Male Unknown VEHICLE MAKE/SERIES EXPOSURE PERCENT OF TOTAL 291 12,177 20,388 18,147 12,748 8,724 0% 17% 28% 25% 18% 12% 4,886 1,696 11,274 47,687 6,933 7% 2% 16% 66% 10% 21,281 29,331 21,863 29% 40% 30% 7,034 43,147 22,294 10% 59% 31% EXPOSURE ABS WITHOUT Honda Gold Wing (Touring) 28,904 Honda Interceptor 800 (Sport) 1,925 Honda Reflex (Scooter) 3,036 Honda Silver Wing (Scooter) 2,882 Honda ST1300 (Sport) 3,118 Suzuki Bandit 1250 (Standard) 136 Suzuki Burgman 650 (Scooter) 3,219 Suzuki SV650 (Unclad Sport) 8,468 Suzuki V-Strom 650 (Dual Purpose) 2,986 Triumph Sprint ST (Sport) 908 Triumph Tiger (Dual Purpose) 695 Yamaha FJR1300 (Sport) 3,610 Total 59,886 PERCENT OF SERIES 82% 76% 87% 85% 68% 80% 92% 100% 97% 82% 94% 59% 83% EXPOSURE ABS WITH 6,366 623 443 506 1,468 33 288 11 90 198 42 2,520 12,589 PERCENT OF SERIES 18% 24% 13% 15% 32% 20% 8% 0% 3% 18% 6% 41% 17% The Highway Loss Data Institute is a nonprofit public service organization that gathers, processes, and publishes insurance data on the human and economic losses associated with owning and operating motor vehicles. COPYRIGHTED DOCUMENT, DISTRIBUTION RESTRICTED © 2008 by the Highway Loss Data Institute, 1005 N. Glebe Road, Arlington, VA 22201. All rights reserved. Distribution of this report is restricted. No part of this publication may be reproduced, or stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright owner. Possession of this publication does not confer the right to print, reprint, publish, copy, sell, file, or use this material in any manner without the written permission of the copyright owner. Permission is hereby granted to companies that are supporters of the Highway Loss Data Institute to reprint, copy, or otherwise use this material for their own business purposes, provided that the copyright notice is clearly visible on the material. HIGHWAY LOSS DATA INSTITUTE 1005 North Glebe Road Arlington, VA 22201 Effectiveness of Antilock Braking Systems in Reducing Fatal Motorcycle Crashes Eric R. Teoh October 2008 ABSTRACT The effect of antilock braking systems (ABS) on motorcyclist fatal crash risk in 2005-06 was studied by comparing fatal crash rates per registrations of motorcycles with and without ABS. Study motorcycles included those for which ABS was optional equipment and could be identified as present by the model name. Fatal motorcycle crashes per 10,000 registered vehicle years were 38 percent lower for ABS models than for their non-ABS versions. INTRODUCTION Annual motorcyclist deaths in the United States have more than doubled, from 2,077 in 1997 to 4,697 in 2006 (Insurance Institute for Highway Safety, 2008), and motorcycle registrations have increased by about two-thirds, from 5,167,693 in 2000 (earliest year for which data are available) to 8,642,243 in 2006, according to data from R.L. Polk and Company. Many factors contribute to motorcycle crashes, but improper braking was identified as a major pre-impact factor in a study of motorcycle crash causation (Hurt et al., 1981) and again, 20 years later, in the Motorcycle Accident In-Depth Study (MAIDS) (Association of European Motorcycle Manufacturers, 2004). Operating the brakes on most motorcycles is much more complicated than on four-wheel vehicles. Most motorcycles have separate controls for the front and rear brakes, with the front brake usually controlled by a lever on the right handlebar and the rear brake controlled by a pedal operated by the rider’s right foot. During braking, a rider must decide how much force to apply to each control. As with other types of vehicles, much more deceleration can be obtained from braking the front wheel than from braking the rear wheel. Motorcycles are inherently less stable than four-wheel vehicles and rely on riders’ skills to remain upright during extreme maneuvers such as hard braking. Braking too hard and locking a wheel creates an unstable situation. Locking the front wheel is particularly dangerous, with loss of control being almost certain. A locked rear wheel is more controllable but still can lead to loss of control if the rider simultaneously tries to steer the motorcycle, as in an emergency avoidance maneuver. However, in an emergency requiring full stopping power, riders concerned about wheel lock may be reluctant to apply full force to the brakes, particularly to the front brake. Both Hurt et al. (1981) and MAIDS (Association of European Motorcycle Manufacturers, 2004) had examples of both loss of control due to wheel lock and failure to adequately brake. Although proper braking practices can be taught, rider training courses have not been shown to be effective in reducing motorcycle crashes (Mayhew and Simpson, 1996) or have provided mixed results at best (Billheimer, 1998). In an effort to address the issue of under-braking (especially the front wheel), manufacturers have developed braking systems that essentially link the front and rear brake controls. These systems, collectively referred to here as combined braking systems (CBS), apply braking force to 1 both wheels when either control is engaged. The degree to which braking force is applied to the front wheel, for example, when the pedal for the rear brake is pressed varies by design, but the concept is the same. CBS has been shown to reduce stopping distances of experienced riders on closed test tracks (Green, 2006) and would be expected to be beneficial in situations in which a rider under-brakes (or does not brake) the front wheel to avoid locking it or causing the motorcycle to pitch forward. Even with CBS, however, it still is possible to lock a wheel during hard braking, often with catastrophic consequences. ABS has been developed to help riders solve this dilemma. The system monitors wheel speed and reduces brake pressure when impending wheel-lock is detected. Brake pressure is increased when traction is restored, and the system evaluates and adjusts brake pressure many times per second. These systems allow riders to apply brakes fully in an emergency without fear of wheel-lock. ABS was first developed for commercial aircraft in 1929 (Maslen, 2008) and was first implemented in production automobiles with the 1971 Chrysler Imperial (Douglas and Schafer, 1971). BMW was the first manufacturer to implement ABS on a motorcycle with its K100RS Special model in 1988 (Tuttle, 2001). ABS and CBS are not necessarily related; either or both can be implemented on a motorcycle. ABS has not significantly reduced crash risk for passenger vehicles (Farmer et al., 1997; Farmer, 2001), but there is reason to expect ABS will be more helpful to motorcycles because of the instability that occurs with any wheel-lock. Studies conducted on closed test tracks have demonstrated that ABS can reduce the motorcycle stopping distances (Green, 2006; Vavryn and Winkelbauer, 2004). It is clear that reducing wheel-lock is crucial in maintaining stability during hard braking. These results suggest that ABS has the potential to reduce motorcycle crashes in real-world situations. Serious motorcycle crashes identified from insurance claims were analyzed in a small study to determine, by crash reconstruction, how certain crashes would be affected by ABS (Allianz Center for Technology, 2005). About half of the 200 crashes studied were deemed to be relevant to ABS, and the majority of those involved another vehicle violating a motorcyclist’s right-of-way. Crash reconstruction analyses showed that between 8 and 17 percent of these crashes could have been avoided had the motorcycles been equipped with ABS. No results were presented on how increased stability or stopping power provided by ABS might have decreased the severities of the crashes that were deemed inevitable. A study by the Highway Loss Data Institute (HLDI), conducted in conjunction with the present study, found that motorcycles equipped with optional ABS had 19 percent fewer insurance claims for collision damage per insured vehicle year than the same motorcycle models without ABS (Moore and Yan, in process). The goal of the present study was to evaluate the effectiveness of ABS in reducing the rate of fatal motorcycle crashes on public roads in the United States. Specifically, rates of fatal crash involvement per registered vehicle were compared for motorcycle models with and without ABS installed as optional equipment. 2 METHODS Data on fatal motorcycle crashes were extracted from the Fatality Analysis Reporting System (FARS), a national census of fatal crashes occurring on public roads that is maintained by the National Highway Traffic Safety Administration. Exposure data consisted of national motorcycle registration records obtained from R.L. Polk and Company. Each vehicle record in both databases was indexed by its vehicle identification number (VIN), which encodes vehicle information, and the first 10 digits of the VINs were used to determine make, model name, and model year according to records in a motorcycle features database created and maintained by HLDI. Vehicles with missing or invalid VINs were excluded. To be included in the study, a motorcycle model was required to have ABS as an option and the presence of that option must have been discernable directly from the model name (e.g., Honda Gold Wing vs. Honda Gold Wing ABS). This eliminated bias due to the comparison of different makes or styles of motorcycles. Although ABS has been an option on BMW models for much longer than the study period, the BMW systems cannot be identified from the model name alone. All BMW models were excluded. The final study population (Table 1) included eight make/model motorcycles, each with both ABS and non-ABS versions. Some vehicles were excluded due to zero registrations of the ABS model during the study years in the Polk records. Because none of the study vehicles with ABS were available in model year 2000 or earlier, the analysis was restricted to 2001 or later model year vehicles. Among the motorcycles included, all of the Hondas (both ABS and non-ABS) were equipped with standard CBS; CBS was not available on any of the others. Table 1 Study motorcycles Non-ABS motorcycles ABS motorcycles Model years Make/model Model years Make/model 2001-06 Honda Gold Wing 2001-06 Honda Gold Wing ABS 2001-06 Honda Interceptor 800 2002-06 Honda Interceptor 800 ABS 2001-06 Honda Reflex 2001-06 Honda Reflex ABS 2003-06 Honda ST1300 2003-06 Honda ST1300 ABS 2002-06 Honda Silver Wing 2003-06 Honda Silver Wing ABS 2003-06 Suzuki Burgman 650 2006 Suzuki Burgman 650 ABS 2001-06 Triumph Sprint ST 2006 Triumph Sprint ST ABS 2003-05 Yamaha FJR1300 2004-06 Yamaha FJR1300 ABS At the time this study was conducted, registration data were available only for 2000 and 2005-07, and FARS data were not yet available for 2007. There were no registrations of the ABS versions of these motorcycles in 2000. Therefore, data were analyzed for years 2005-06. Fatal crash rates per 10,000 registered vehicle years for each motorcycle model, both ABS and non-ABS versions, were calculated by dividing 10,000 times the number of motorcyclist fatal crash involvements in 2005-06 by the number of 3 motorcycles registered during these years. Because registration counts spanned 2 years, the denominator was interpreted as registered vehicle years instead of registrations. Fatal crash rates per vehicle registrations for ABS and non-ABS motorcycle models were compared by calculating a rate ratio (RR) equal to the crash rate for ABS models divided by the crash rate for non-ABS models. If ABS has no effect, then the rate ratio should be 1.0. A rate ratio of less than 1.0 would indicate the fatal crash rate for ABS models is lower than the rate for non-ABS models. Similarly, a rate ratio greater than 1.0 would indicate the fatal crash rate for ABS models is higher than the rate for non-ABS models. One way to calculate the rate ratio is to calculate the rate for non-ABS motorcycles as total crash involvements divided by total registered vehicle years, and analogously for ABS motorcycles. However, to reduce any bias that may have occurred from relative differences in registrations among motorcycle models, an alternative rate for non-ABS motorcycles was calculated as the weighted average of fatal crash rates for each vehicle model, where the weights were taken as the number of registered vehicle years of ABS-equipped motorcycles. Thus, for any given motorcycle model the ABS and nonABS fatal crash rates received the same weight in calculating the overall fatal crash rates for motorcycles with and without ABS. Ninety and 95 percent confidence intervals (CIs) for the rate ratios were calculated based on standard error estimates derived by assuming that the number of fatal crash involvements follows a Poisson distribution as follows: Var(ln(RR)) = VarABS + Varnon-ABS where VarABS = ∑8 1 X ,ABS and Varnon-ABS = ∑8 1 r X ,non‐ABS ∑8 1 r X with Xi,ABS as the number of fatal crash involvements of the ith model motorcycle equipped with ABS, analogously for Xi,non-ABS, and ri as the ratio of registered vehicle years of the ABS model to registered vehicle years of the non-ABS model for model i. lower 95% confidence limit = exp[ln(RR) – 1.96(Var(ln(RR)))1/2] upper 95% confidence limit = exp[ln(RR) + 1.96(Var(ln(RR)))1/2] RESULTS Table 2 presents fatal crash involvements, registered vehicle years, and the rate of fatal crash involvements per 10,000 registered vehicle years for the study motorcycles during 2005-06. Motorcycles manufactured by Honda, particularly the Gold Wing model, dominated the sample, but the pattern for all but two of the motorcycles was a lower fatal crash rate for ABS-equipped motorcycles. Across all ABSequipped motorcycles, the rate of fatal crash involvements per 10,000 registered vehicle years was 4.1, compared with 6.7 for the same motorcycles not equipped with ABS. 4 Table 2 Motorcycle fatal crash involvements and registered vehicle years, 2005-06 Non-ABS models ABS models Fatal crash Registered Rate Fatal crash Registered involvements vehicle years (x 104) involvements vehicle years Honda Gold Wing 63 93,608 6.7 6 19,547 Honda Interceptor 800 8 10,437 7.7 3 2,307 Honda Reflex 6 14,858 4.0 1 2,644 Honda ST1300 2 7,003 2.9 2 3,580 Honda Silver Wing 11 12,273 9.0 1 1,278 Suzuki Burgman 650 8 9,618 8.3 0 309 Triumph Sprint ST 1 4,476 2.2 0 135 Yamaha FJR1300 8 8,734 9.2 2 6,486 Total 107 161,007 6.7* 15 36,286 Rate (x 104) 3.1 13.0 3.8 5.6 7.8 0.0 0.0 3.1 4.1 *Overall non-ABS rate is weighted by registered vehicle years of ABS-equipped motorcycles. The effect of ABS on fatal crash involvement is given by the rate ratio estimate for ABSequipped motorcycles against non-ABS motorcycles. This estimate and associated 90 and 95 percent confidence intervals are provided in Table 3. The rate ratio estimate corresponds to an approximate 38 percent reduction (computed as (RR-1)×100%) in the rate of fatal crash involvements per 10,000 registered vehicle years for the ABS models over the (weighted) non-ABS models. Table 3 Estimated rate ratios and confidence intervals for those estimates for comparing ABS and non-ABS fatal crash rates Rate ratio 0.615 95% confidence interval (0.352, 1.074) 90% confidence interval (0.385, 0.982) DISCUSSION Results of this analysis provide evidence that ABS is effective in reducing fatal motorcycle crashes. Study motorcycles with ABS had a fatal crash involvement rate 38 percent lower than that for their non-ABS versions during the study years. Although the estimated effect of 38 percent is large, it is not statistically significant at the customary 0.05 level. ABS is a relatively recent option on motorcycles, and the option was purchased on only 18 percent of registered study vehicles during 2005-06. More data are required to obtain a more precise estimate of ABS effectiveness in reducing fatal motorcycle crashes. However, as the estimate becomes more precise, it is quite likely that it will continue to indicate a benefit of ABS. If there were no effect of ABS on fatal crash involvement, an estimate as large as the 38 percent reduction in this study would be expected to occur by chance less than 10 percent of the time. Thus, there is considerable confidence that ABS is preventing fatal crashes among motorcyclists. This confidence is bolstered by the fact that a separate analysis of insurance collision coverage losses among crashes of all severities also shows a reduction in crashes of about 19 percent for motorcycles equipped with ABS (Moore and Yan, 5 2008). These results provide confirmatory evidence of the expected benefit of ABS from engineering principles, test-track trials, and a crash reconstruction analysis. The substantial effectiveness estimate observed in this study is not, however, without limitations. ABS was studied as optional equipment, so the cohort of motorcyclists who choose to purchase ABS may differ from those who decline to purchase it. In particular, motorcyclists who choose ABS may be more concerned about safety than those who decline, thus leading to lower fatal crash rates due to safer riding practices. Because of the small sample of ABS-equipped motorcycles, it was not possible to carefully examine how rider factors such as helmet use and speeding differ between the two groups. For example, 78 percent of the non-ABS riders were helmeted, compared with 60 percent (9 of 15) of the ABS riders. Also, the prevalence of these factors is not known for riders of study motorcycles that were not involved in fatal crashes. Therefore, it was not possible to study how such factors influenced the observed reduction in fatal crash rate for ABS-equipped motorcycles. Aside from differences in rider factors, it is also possible that riders who choose ABS accumulate more miles than those who decline, which would result in an upward bias in the fatal crash rate for the ABS cohort relative to the non-ABS cohort. As may have occurred in passenger vehicles (Grant and Smiley, 1993), motorcyclists may tend to drive ABS motorcycles more aggressively than non-ABS motorcycles, also resulting in a higher than expected crash rate for the ABS group. Without more extensive data, it is not possible to know the magnitude or nature of any bias of the estimated rate ratio comparing crash rates for ABS and non-ABS motorcycles. With or without ABS, CBS also may reduce the likelihood of certain types of crashes. However, due to the small sample of non-CBS motorcycles in this study, the effect of CBS could not be evaluated. Still, CBS is not expected to bias the results because the braking systems of the ABS and non-ABS motorcycles differed only by whether or not they were equipped with ABS. In other words, each ABS/nonABS pair either did or did not have CBS. ABS showed a benefit in both the CBS and non-CBS groups, suggesting the presence of CBS on some of the motorcycles did not confound the observed effect of ABS. ABS cannot be expected to prevent or mitigate all types of crashes, as demonstrated in the Allianz Center for Technology (2005) study. For example, ABS would not affect the outcome or likelihood of a crash involving a motorcycle struck from behind. The small sample of ABS motorcycles in FARS and the lack of detailed information on precrash events in FARS precluded examination of the effects of ABS on crashes that would likely be influenced by its presence. ACKNOWLEDGMENTS Motorcycle VINs were decoded by Marvin Campbell of the Highway Loss Data Institute, which also maintains the database used to decode these VINs. Adrian Lund, Anne McCartt, and Charles Farmer of the Insurance Institute for Highway Safety contributed helpful comments and suggestions. This work was supported by the Insurance Institute for Highway Safety. 6 REFERENCES Allianz Center for Technology. (2005) AZT Accident Research on Effectiveness of ABS in Motorbikes. Presented at 2005 RCAR Conference, Association of European Motorcycle Manufacturers. 2004. MAIDS in-depth investigations of accidents involving powered two wheelers. Final Report 1.2. Brussels, Belgium Billheimer, J.W. 1998. Evaluation of the California motorcyclist safety program. Transportation Research Record 1640:100-09. Douglas, J. and Schafer, T. 1971. The Chrysler ‘Sure-Brake:’ The first production four-wheel anti-skid system. SAE Technical Paper Series 710248. Warrendale, PA: Society of Automotive Engineers. Farmer, C.M. 2001. New evidence concerning fatal crashes of passenger vehicles before and after adding antilock braking systems. Accident Analysis and Prevention 33:361-69. Farmer, C.M.; Lund, A.K.; Trempel, R.E.; and Braver, E.R. 1997. Fatal crashes of passenger vehicles before and after adding antilock braking systems. Accident Analysis and Prevention 29:745-57. Grant, B. and Smiley, A. 1993. Driver response to antilock brakes: a demonstration of behavioural adaptation. Proceedings of the Canadian Multidisciplinary Road Safety Conference VIII. Ottawa, Ontario: Transport Canada. Green, D. 2006. A comparison of stopping distance performance for motorcycles equipped with ABS, CBS and conventional hydraulic brake systems. Presented at the 2006 International Motorcycle Safety Conference, Long Beach, CA. Hurt, H.H.; Ouellet, J.V.; and Thom, D.R. 1981. Motorcycle accident cause factors and identification of countermeasures; Volume 1: technical report. Report no DOT HS-805-862. Washington, DC: National Highway Traffic Safety Administration. Insurance Institute for Highway Safety. 2008. Fatality facts, 2006: motorcycles. Arlington, VA. Available: http://www.iihs.org/research/fatality_facts_2006/motorcycles.html. Accessed: July 21, 2008. Maslen, J. (Spring 2008) ABS: 30 years of life saving. Roadsafe Magazine. Available: http://www.roadsafe.com/magazine/2008spring/abs.htm. Accessed: July 23, 2008. Mayhew, D.R. and Simpson, H.M. 1996. Effectiveness and role of driver education and training in a graduated licensing system. Ottawa, Ontario: Traffic Injury Research Foundation. Moore, M. and Yan, Y. In process. Antilock braking systems for motorcycles and insurance collision losses. Arlington, VA: Highway Loss Data Institute. Tuttle, M. 2001. One track mind. Rider Magazine July;28(7), p. 8. Vavryn, K. and Winkelbauer, M. 2004. Braking performance of experienced and novice motorcycle riders – results of a field study. Presented at the 2004 International Conference on Transport and Traffic Psychology, Nottingham, United Kingdom. 7