INSURANCE SPECIAL REPORT
Motorcycle Antilock Braking System (ABS)

December 2009 A-81

COPYRIGHTED DOCUMENT, DISTRIBUTION RESTRICTED
2009 by the Highway Loss Data Institute. 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 report in any manner without the written permission of the copyright owner.

COPYRIGHT NOTICE

2009 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 members 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.

DIRECTORS

T S. Lin, Chairman, Chubb & Son T B. Reddington, Vice Chairman, Kentucky Farm Bureau Insurance Companies T B. Anderson, National Association of Mutual Insurance Companies T P. Baum, Nationwide T H. Cohen, GEICO Corporation T E. Connell, Erie Insurance Group T F. Yager, Allstate Insurance Company T M. Deede, MetLife Auto & Home T M. Doerfler, Progressive Insurance T T. Ellefson, American Family Insurance Group T J. Feldmeier, Auto Club Group T P. Foley, American Insurance Association T J. Gillette, American National Property and Casualty T D. Griffin, Property Casualty Insurers Association of America T S. Hallworth, The Travelers Companies T S. Lough, Rockingham Group T S. McAnena, Liberty Mutual Insurance Company T S. Murphy, GMAC Insurance T T. Myers, High Point Safety and Insurance Management Corporation T J. Nutting, Farmers Insurance Group of Companies T S. Oakley, The Hartford T D. Porfilio, Kemper, A Unitrin Business T L. Stiles, State Farm Mutual Automobile Insurance Company T J. Xu, AAA of Northern California, Nevada and Utah T A. Lund, Highway Loss Data Institute The membership of the Highway Loss Data Institute Board of Directors represents insurance companies that supply data to HLDI. Financial support for HLDI is provided through the Insurance Institute for Highway Safety, which in turn is supported by automobile insurers.

CONTENTS

Introduction. 1 Methods. 2 Results. 4 Table 1 Summary Results of Linear Regression Analysis. 5 of Collision Claim Frequencies Table 2 Detailed Results of Linear Regression Analysis. 5 of Collision Claim Frequencies Table 3 Summary Results of Linear Regression Analysis. 7 of Collision Claim Severities Table 4 Detailed Results of Linear Regression Analysis. 7 of Collision Claim Severities Table 5 Results for Collision Overall Losses Derived. 9 from Claim Frequency and Severity Models Table 6 Summary Results of Linear Regression Analysis. 11 of Medical Payment Claim Frequencies Table 7 Detailed Results of Linear Regression Analysis. 11 of Medical Payment Claim Frequencies Table 8 Summary Results of Linear Regression Analysis. 12 of Medical Payment Claim Severities Table 9 Detailed Results of Linear Regression Analysis. 12 of Medical Payment Claim Severities Table 10 Results for Medical Payment Overall Losses Derived from. 13 Claim Frequency and Severity Models Table 11 Summary Results of Linear Regression Analysis. 15 of Bodily Injury Liability Claim Frequencies Table 12 Detailed Results of Linear Regression Analysis. 15 of Bodily Injury Liability Claim Frequencies Table 13 Summary Results of Linear Regression Analysis. 16 of Bodily Injury Liability Claim Severities Table 14 Detailed Results of Linear Regression Analysis. 16 of Bodily Injury Liability Claim Severities Table 15 Results for Bodily Injury Liability Overall Losses Derived. 17 from Claim Frequency and Severity Models Appendix A Distribution of Exposure for Independent Variables,. 18 Collision Coverage Appendix B Distribution of Exposure for Independent Variable,. 19 Medical Payment Coverage Appendix C Distribution of Exposure for Independent Variable,. 20 Bodily Injury Liability Coverage

INTRODUCTION

According to the National Highway Traffic Safety Administration (NHTSA, 2008) motorcycle registrations increased by 75 percent during 1997-2006. Analysis by the Insurance Institute for Highway Safety of data from the Fatality Analysis Reporting System shows that, during the same time period, fatalities in motorcycle crashes increased by 128 percent. Unlike automobiles, motorcycles offer little if any occupant protection. Only 20 percent of automobile crashes result in injury or death, whereas 80 percent of motorcycle crashes have this outcome (NHTSA, 2005). Therefore any countermeasure aimed at reducing the likelihood of motorcycle crashes should significantly reduce the risk of injury or death. One technology designed to reduce the likelihood of motorcycle crashes is antilock braking systems (ABS). While in motion, motorcycles are kept stable by the gyroscopic effect of the wheels and lateral grip of the tires. If a wheel is braked too hard, so that it locks, both lateral grip and gyroscopic effect are lost. When this occurs, the motorcycle is immediately destabilized, and any remaining tire grip is engaged in uncontrolled skidding, leaving no grip for maneuvering. ABS has independent braking sensors for each wheel. If the system detects a difference in the rotation speeds of the wheels, it partially releases brake pressure to allow the locked wheel to spin and the tire to retain grip before reapplying the brake. ABS then modulates braking pressure to achieve optimum braking. The Highway Loss Data Institute (HLDI) initially reported on motorcycle ABS in April 2008, in which the 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 severities. This report updates and expands the initial analysis by adding the 2008 model year, increasing the number of make/series from 12 to 18, and doubling the collision exposure. This study also includes an analysis of medical payment coverage, which typically pays for operator injuries, and bodily injury liability coverage, which typically pays for passenger injuries.

2009 Highway Loss Data Institute

METHODS

COVERAGES
Motorcycle insurance covers damage to vehicles and property as well as injuries to people involved in crashes. Different insurance coverages pay for physical damage versus injuries. Also, different coverages may apply depending on who is at fault. In the present study, three different insurance coverage types were examined: collision, bodily injury liability, and medical payment. Collision insures against physical damage to a motorcycle sustained in a crash when the driver is at fault. Medical payment covers injuries sustained by motorcycle operators, whereas bodily injury liability typically insures against injuries to motorcycle passengers.

RATED DRIVERS (RIDERS)

For insurance purposes, a rated driver is assigned to each motorcycle on a policy. The rated driver is the one who typically is considered to represent the greatest loss potential for the insured vehicle. In a multiple-vehicle/driver household, the driver assigned to a vehicle can vary by insurance company and state. Information on the actual driver at the time of a loss is not available in the HLDI database. HLDI collects a limited number of factors about rated drivers. For the present study, data were stratified by rated driver age group (<25, 25-39, 40-64, 65+, or unknown) and gender (male, female, or unknown).

SUBJECT MOTORCYCLES

For motorcycles to be included in the present study, their vehicle identification numbers (VINs) had to have an ABS indicator. This allowed for very tight control over the study population. Twenty motorcycles met this criterion, but two of them did not have claims and therefore were excluded. There were motorcycles available with ABS that were not included because their VINs did not have an ABS indicator. All of the Honda motorcycles (both ABS and non-ABS) were equipped with combined braking systems (CBS). CBS applies braking force to both wheels when either the rear or front brake control is engaged. Even with CBS, wheel lock still is possible. With or without ABS, CBS may affect collision losses. Due to the small sample of non-CBS motorcycles in the study, the effect of CBS could not be evaluated. This is not expected to bias the results because the motorcycles in the study differed only by whether or not they were equipped with ABS. Each ABS/non-ABS 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.

ANALYSIS METHODS

Data were collected by vehicle make and series, rated driver age and gender, and vehicle age and density. Vehicle density was defined as the number of registered vehicles (<100, 100-499, and 500+) per square mile. Vehicle age was defined as the difference between the calendar year and model year, measured in years. As previously mentioned, rated driver age group and gender were included in the analysis. 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, equipped with ABS, 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 distributions of motorcycle exposure by coverage type for the six independent variables are listed in the Appendices. Rated driver factors and vehicle density were included to control for their potential impact on losses and not to produce estimates for these variables. The estimated parameters for these variables may not generalize from this subset to the much larger motorcycle population.
Highway Loss Data Institute
Regression analysis was used to quantify the effect of ABS on motorcycle 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 overall losses were derived from the claim frequency and claim 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 = 40-64, 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 presence or absence of ABS. None of these interactions were found to be significant.

RESULTS

COLLISION COVERAGE
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 than 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 0.0000687 claims per day, or 2.5 claims per 100 insured vehicle years. The intercept outlines losses for the reference (baseline) categories: the estimate corresponds to the claim frequency for a Honda Gold Wing without ABS, with vehicle age 0, garaged in a medium vehicle density area, and driven by a male age 40-64. 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 0.87, so female rated drivers had estimated claim frequencies 13 percent lower than those for male rated drivers. The estimate corresponding to motorcycle ABS (-0.25) was highly significant (p<0.0001). The estimate corresponded to a 22 percent reduction in claim frequencies for motorcycles equipped with ABS. Individual make/series motorcycles were included in the model, and estimates of their effect on collision claim frequencies were reported in Table 2. As previously mentioned, the reference category for the make/series variable was the Honda Gold Wing. Significant predictions for make/series ranged from 1.37 for the Triumph Tiger to 5.4 for the Honda CBR1000RR. All make/series estimates were significant at the p=0.05 level except for the Aprilia Caponord and Suzuki V-Strom 650. Vehicle age significantly affected collision claim frequency. Claim frequencies were estimated to decrease 19 percent (p<0.0001) for each 1-year increase in vehicle age. Driver age was highly significant in predicting motorcycle collision claim frequency. Compared with losses for rated drivers ages 40-64 (reference category), estimated claim frequencies were 145 percent higher (p<0.0001) for rated drivers 24 and younger, 23 percent higher (p<0.0001) for rated drivers ages 25-39 and 18 percent higher (p=0.003) for rated drivers 65 and older. Rated driver gender also significantly predicted collision claim frequencies. Compared with losses for male rated riders (reference category), estimated claim frequencies were 8 percent lower (p=0.02) for drivers with unknown gender and 13 percent lower, nearly significant (p=0.06), for female rated drivers. Motorcycle collision claim frequencies increased with vehicle density. Compared with losses in medium vehicle density areas (reference category), estimated claim frequencies were 9 percent higher (p=0.04) in high vehicle density areas and 13 percent lower (p=0.002) in low vehicle density areas.

TABLE 1 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM FREQUENCIES

DEGREES

OF FREEDOM

CHI-SQUARE

31.920 432.810 289.610 87.180 7.350 23.230

P-VALUE

ABS Vehicle Make/Series Vehicle Age Rated Driver Age Rated Driver Gender Vehicle Density
<0.0001 <0.0001 <0.0001 <0.0001 0.025 <0.0001
TABLE 2 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM FREQUENCIES
PARAMETER INTERCEPT ABS ABS Model Non-ABS Model VEHICLE MAKE/SERIES Aprilia Caponord Aprilia Scarabeo 500 Harley Davidson V-Rod Honda CBR1000RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki B-King Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 VEHICLE AGE RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ RATED DRIVER GENDER Female Male Unknown VEHICLE DENSITY 0-99 100-499 500+
ESTIMATE -9.586 -0.0.100 0.871 0.662 1.0.882 0.570 0.716 0.241 0.941 0.941 1.432 0.660 1.093 0.104 1.065 0.314 0.449 -0.214 0.362 0.897 0.0.167 -0.-0.087 -0.0.081
EXPONENT ESTIMATE 6.87E-05 0.782 1.000 1.105 2.390 1.938 5.400 1.000 2.417 1.767 2.047 1.273 2.561 2.563 4.187 1.935 2.983 1.110 2.901 1.368 1.567 0.807 1.436 2.452 1.232 1.000 1.181 0.872 1.000 0.917 0.873 1.000 1.085
STANDARD ERROR 0.046 0.1.001 0.270 0.097 0.0.078 0.081 0.076 0.080 0.098 0.136 0.222 0.067 0.084 0.127 0.104 0.152 0.062 0.013 0.068 0.108 0.0.057 0.0.038 0.0.040

CHISQUARE 44,115.80 30.8

<0.0001 <0.0001
0.01 10.44 46.78 11.27 128.03 49.27 89.71 9.16 91.47 48.13 41.55 98.1 169.3 0.68 104.47 4.276.61 28.04 69.66 17.24 8.66 3.46 5.16 9.95 4.18
0.920 0.001 <0.0001 0.001 <0.0001 <0.0001 <0.0001 0.003 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.411 <0.0001 0.040 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.003 0.063 0.023 0.002 0.041
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 make/series and vehicle age had p-values less than 0.05. Neither the rated driver nor the driving environment significantly affected the claim size. 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 $8,829. The intercept outlines losses for the reference (baseline) categories: the estimate corresponds to the claim severity for a Honda Gold Wing without ABS, with vehicle age of 0, garaged in a medium vehicle density area, and driven by a male age 40-64. The estimate corresponding to the ABS effect was a 4 percent increase in claim severity. However, the estimate was not significant (p=0.3), indicating ABS does not affect claim severity. As previously mentioned, vehicle make/series and vehicle age were significant predictors of claim severity. Significant estimates of claim severities for the 18 make/series motorcycles, compared with those for the Honda Gold Wing (reference category), ranged from 23 percent lower for the Honda ST1300 to 74 percent lower for the Honda Reflex. As motorcycles age, their claim severities decrease. The model estimated a 4 percent decrease (p<0.0001) in claim severity per 1-year increase in vehicle age.

TABLE 3 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM SEVERITIES
1.020 643.930 16.070 4.600 2.910 5.340
0.312 <0.0001 <0.0001 0.331 0.233 0.069
TABLE 4 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF COLLISION CLAIM SEVERITIES
PARAMETER INTERCEPT ABS ABS Model Non-ABS Model

VEHICLE MAKE/SERIES

ESTIMATE 9.086 0.-0.497 -1.139 -0.503 -0.-0.587 -1.355 -1.054 -0.260 -0.406 -0.826 -0.609 -0.845 -0.793 -0.850 -0.454 -0.491 -0.477 -0.042 0.089 0.122 0.0.047 0.-0.014 0.0.080
EXPONENT ESTIMATE 8,829.03 1.038 1.000 0.608 0.320 0.605 0.819 1.000 0.556 0.258 0.349 0.771 0.667 0.438 0.544 0.429 0.453 0.427 0.635 0.612 0.621 0.959 1.094 1.130 1.011 1.000 1.048 1.095 1.000 0.986 1.039 1.000 1.083
STANDARD ERROR 0.040 0.0.825 0.223 0.083 0.0.0654 0.0673 0.0635 0.067 0.0833 0.113 0.1883 0.0562 0.0714 0.1066 0.0876 0.1261 0.0513 0.010 0.060 0.088 0.0.047 0.0.032 0.0.034

CHISQUARE 52,266.80 1.02

<0.0001 0.313
Aprilia Caponord Aprilia Scarabeo 500 Harley Davidson V-Rod Honda CBR1000RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki B-King Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 VEHICLE AGE

RATED DRIVER AGE

0.36 26.16 36.59 0.17 80.56 404.93 275.2 15.04 23.7 53.43 10.46 226.35 123.27 63.67 26.83 15.18 86.47 16.25 2.2 1.95 0.2.17 0.2 1.1 5.57
0.547 <0.0001 <0.0001 0.677 <0.0001 <0.0001 <0.0001 0.0001 <0.0001 <0.0001 0.001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.139 0.163 0.794 0.317 0.140 0.656 0.295 0.018
Unknown 14-24 25-39 40-64 65+

RATED DRIVER GENDER

Female Male Unknown

VEHICLE DENSITY

0-99 100-499 500+
Table 5 summarizes the effects of the independent variables on motorcycle collision overall losses, derived from the claim frequency and claim severity models. Overall losses can be calculated by simple multiplication because the estimates for the effect of ABS on claim frequency and claim 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 corresponding two-sided p-value was derived from a standard normal distribution approximation. The estimated effect of ABS was a significant (p=0.0003) 19 percent decrease in collision overall losses. This is a strong indication that ABS is effective in reducing collision overall losses for motorcycles. Estimated overall losses for the 18 make/series motorcycles, compared with those for the Honda Gold Wing (reference category), ranged from 54 percent lower for the Honda Reflex to 342 percent higher for the Honda CBR1000RR. Ten of the make/series estimates were significantly different from the reference category, and the other seven estimates were not significant. Vehicle age also had significant effects in reducing collision overall losses. Collision overall losses were estimated to decrease 23 percent (p<0.0001) for each 1-year increase in vehicle age. Driver age was a significant predictor of motorcycle collision overall losses. Compared with losses for rated drivers ages 40-64 (reference category), estimated overall losses were 177 percent higher (p<0.0001) for rated drivers 24 and younger, 25 percent higher (p=0.0011) for rated drivers ages 25-39, and 24 percent higher (p=0.004) for rated drivers 65 or older. Estimated overall losses for drivers with unknown gender were 10 percent lower (p=0.04) than those for male rated drivers (reference category). Estimated overall losses for rated female drivers were not significant. Motorcycle collision overall losses were predicted to increase with vehicle density. Compared with losses in medium vehicle density in areas (reference category), estimated overall losses were 17 percent higher (p=0.002) in high vehicle density areas and 9 percent lower, nearly significant (p=0.08), in low vehicle density areas.

TABLE 5 RESULTS FOR COLLISION OVERALL LOSSES DERIVED FROM CLAIM FREQUENCY AND SEVERITY MODELS
FREQUENCY STANDARD ESTIMATE ERROR
-9.586 -0.0.100 0.871 0.662 1.0.882 0.570 0.716 0.241 0.941 0.941 1.432 0.660 1.093 0.104 1.065 0.314 0.449 -0.214 0.362 0.897 0.0.167 -0.-0.087 -0.0.081 0.046 0.1.001 0.270 0.097 0.0.078 0.081 0.076 0.080 0.098 0.136 0.222 0.067 0.084 0.127 0.104 0.152 0.062 0.013 0.068 0.108 0.0.057 0.0.038 0.0.040

PARAMETER

INTERCEPT ABS ABS Model Non-ABS Model VEHICLE MAKE/SERIES Aprilia Caponord Aprilia Scarabeo 500 Harley Davidson V-Rod Honda CBR1000RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki B-King Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 VEHICLE AGE RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ RATED DRIVER GENDER Female Male Unknown VEHICLE DENSITY 0-99 100-499 500+
SEVERITY STANDARD ESTIMATE ERROR ESTIMATE
9.086 0.-0.497 -1.139 -0.503 -0.-0.587 -1.355 -1.054 -0.260 -0.406 -0.826 -0.609 -0.845 -0.793 -0.850 -0.454 -0.491 -0.477 -0.042 0.089 0.122 0.0.047 0.-0.014 0.0.080 0.040 0.0.825 0.223 0.083 0.0.065 0.067 0.064 0.067 0.083 0.113 0.188 0.056 0.071 0.107 0.088 0.126 0.051 0.010 0.060 0.088 0.0.047 0.0.032 0.0.034 -0.500 -0.-0.397 -0.267 0.159 1.0.295 -0.785 -0.337 -0.019 0.535 0.115 0.823 -0.185 0.300 -0.746 0.611 -0.178 -0.028 -0.256 0.451 1.019 0.0.214 -0.-0.101 -0.0.161
OVERALL LOSSES STANDARD EXPONENT ERROR ESTIMATE
0.060 0.1.297 0.350 0.128 0.0.102 0.105 0.099 0.104 0.129 0.177 0.291 0.087 0.110 0.165 0.136 0.198 0.080 0.017 0.091 0.1387 0.0.074 0.0.050 0.0.052 0.606 0.0.673 0.765 1.173 4.1.343 0.456 0.714 0.981 1.707 1.122 2.277 0.831 1.350 0.474 1.842 0.837 0.973 0.774 1.570 2.771 1.1.238 0.0.904 0.1.174

<0.0001 0.0003

0.760 0.444 0.212 0.032 0.004 <0.0001 0.0006 0.856 <0.0001 0.513 0.005 0.033 0.006 <0.0001 <0.0001 0.370 0.731 <0.0001 <0.0001 <0.0001 0.001 0.004 0.634 0.043 0.082 0.002

MEDICAL PAYMENT COVERAGE

Summary results of the regression analysis of motorcycle medical payment claim frequencies using the Poisson distribution are listed in Table 6. Results for the following independent variables: ABS, vehicle make/series, vehicle age and rated driver gender had p-values less than 0.05, indicating their effects on claim frequencies were statistically significant. Rated driver age was marginally significant while vehicle density was not significant. Detailed results of the regression analysis using claim frequency as the dependent variable are listed in Table 6. The exponent of the intercept equals 0.000046 claims per day, or 16.8 claims per 1,000 insured vehicle years. The estimate corresponding to motorcycle ABS (-0.36) was highly significant (p=0.003). The estimate corresponded to a 30 percent reduction in medical payment claim frequencies for motorcycles equipped with ABS. The estimate corresponding to the ABS effect on medical payment claim severity was a nonsignificant (p=0.32) 13 percent increase in claim severity, indicating ABS does not affect claim severity. Rated driver age and make/series were the strongest predictors of claim severity. The predictive value of make/series is perhaps a proxy for policy limits. More expensive motorcycles are more likely to have higher policy limits than less expensive motorcycles. Higher policy limits allow higher claim severities to occur in the event of a crash. The Honda Gold Wing is the most expensive motorcycle in the study. The make/series estimates for the other motorcycles studied are less than that for the Gold Wing except for the Honda CBR1000RR, which typically is among the motorcycles with the highest collision losses primarily due to its very high claim frequency. Overall losses for medical payment coverage were calculated in the same fashion as collision coverage. ABS was estimated to reduce overall medical payment losses by 21 percent, although the estimate was not statistically significant (p=0.16).

TABLE 6 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF MEDICAL PAYMENT CLAIM FREQUENCIES
9.640 92.390 57.850 9.140 7.840 1.820
0.002 <0.0001 <0.0001 0.058 0.020 0.403
TABLE 7 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF MEDICAL PAYMENT CLAIM FREQUENCIES
PARAMETER INTERCEPT ABS ABS Model Non-ABS Model VEHICLE MAKE/SERIES Aprilia Scarabeo 500 Harley Davidson V-Rod Honda CBR1000RR Honda CBR600RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki B-King Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 VEHICLE AGE RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ RATED DRIVER GENDER Female Male Unknown VEHICLE DENSITY 0-99 100-499 500+ ESTIMATE -9.985 -0.0.041 0.206 0.759 1.0.367 0.744 0.559 0.582 0.502 0.343 0.617 0.347 1.137 0.406 1.029 0.677 0.028 -0.234 0.064 0.529 0.-0.165 -0.0.253 -0.0.104 EXPONENT ESTIMATE 4.61E-05 0.1.042 1.229 2.135 4.1.444 2.104 1.750 1.789 1.651 1.410 1.854 1.415 3.119 1.501 2.797 1.968 1.028 0.792 1.066 1.698 1.0.848 0.1.288 0.1.109 STANDARD ERROR 0.109 0.1.004 0.261 0.309 0.0.252 0.168 0.182 0.170 0.292 0.416 0.714 0.191 0.187 0.257 0.257 0.310 0.190 0.031 0.141 0.191 0.0.150 0.0.093 0.0.096 CHISQUARE 8,387.580 9.050

<0.0001 0.003

0.000 0.620 6.010 66.640 2.120 19.620 9.480 11.640 2.950 0.680 0.750 3.300 37.010 2.500 16.020 4.760 0.020 55.330 0.210 7.680 0.380 1.210 0.150 7.340 0.080 1.180
0.968 0.431 0.014 <0.0001 0.146 <0.0001 0.002 0.001 0.086 0.409 0.387 0.069 <0.0001 0.114 <0.0001 0.029 0.884 <0.0001 0.649 0.006 0.537 0.271 0.703 0.007 0.772 0.277
TABLE 8 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF MEDICAL PAYMENT CLAIM SEVERITIES
1.010 53.340 0.000 34.650 15.970 0.060
0.314 <0.0001 0.981 <0.0001 0.0003 0.970
TABLE 9 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF MEDICAL PAYMENT CLAIM SEVERITIES
ESTIMATE 8.018 0.-1.129 -0.791 0.076 -0.-0.440 -0.410 -0.570 -0.441 -0.172 -0.471 -0.049 -0.957 -0.566 -0.746 -0.973 -0.323 -0.549 -0.001 0.756 -0.422 0.0.093 0.-0.319 0.0.018
EXPONENT ESTIMATE 3,034.798 1.0.324 0.453 1.079 0.0.644 0.664 0.566 0.643 0.842 0.624 0.952 0.384 0.568 0.474 0.378 0.724 0.578 0.999 2.130 0.656 1.1.097 1.0.727 1.1.019
STANDARD ERROR 0.113 0.0.897 0.304 0.323 0.0.262 0.1716 0.1949 0.1672 0.3156 0.4172 0.905 0.1965 0.1977 0.2535 0.2758 0.3219 0.1772 0.033 0.152 0.206 0.0.154 0.0.105 0.0.103
CHISQUARE 5,032.420 1.000

<0.0001 0.318

Aprilia Scarabeo 500 Harley Davidson V-Rod Honda CBR1000RR Honda CBR600RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki B-King Suzuki Burgman 650 Suzuki SV650 Suzuki V-Strom 650 Triumph Sprint ST Triumph Tiger Yamaha FJR1300 VEHICLE AGE

Due to limited exposure, only 12 of the 18 motorcycles used in collision coverage analysis were used in analysis of bodily injury liability coverage. Summary results of the regression analysis of motorcycle bodily injury liability claim frequencies using the Poisson distribution are listed in Table 11. Results for all of the independent variables except rated driver gender had p-values less than 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 12. The exponent of the intercept equals 0.0000085 claims per day, or 3.1 claims per 1,000 insured vehicle years. The estimate corresponding to motorcycle ABS (-0.394) was significant (p = 0.03). The estimate corresponded to a 33 percent reduction in bodily injury liability claim frequencies for motorcycles equipped with ABS. The estimated claim frequency for rated drivers 24 and younger was more than 4 times that for rated drivers ages 40-64 (reference category). Of the 12 estimates for make/series, only two were statistically different from the reference make/series. Claim frequencies were estimated to be 0.474 for the Yamaha FJR1300 and 2.614 for the Honda CBR1000RR. Claim frequencies were estimated to decrease 16 percent (p = 0.0002) for each 1-year increase in vehicle age. None of the variables in the analysis were shown to have a statistically significant impact on bodily injury liability claim severity. Although ABS was estimated to reduce overall bodily injury liability losses by more than 43 percent, the estimate was not statistically significant (p = 0.185).

REFERENCES

National Highway Traffic Safety Administration. 2008. Traffic Safety Facts, 2007. Report no. DOT HS-810-990. Washington, DC: US Department of Transportation. National Highway Traffic Safety Administration. 2005. Without Motorcycle Helmets We All Pay the Price. Washington, DC: US Department of Transportation.
TABLE 11 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF BODILY INJURY LIABILITY CLAIM FREQUENCIES
5.050 22.610 14.540 17.980 4.010 6.420
0.025 0.020 0.0001 0.001 0.135 0.040
TABLE 12 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF BODILY INJURY LIABILITY CLAIM FREQUENCIES
PARAMETER ESTIMATE EXPONENT ESTIMATE 8.47E-06 0.0.968 2.614 1.000 0.784 0.502 1.036 1.015 0.799 0.337 0.677 0.936 0.474 0.839 1.172 4.116 0.1.424 0.0.749 0.1.293 STANDARD ERROR 0.159 0.0.409 0.342 0.000 0.360 0.3926 0.284 0.260 0.517 1.006 0.330 0.363 0.316 0.047 0.288 0.353 0.0.185 0.0.147 0.0.150 CHISQUARE 5,429.750 4.690

INTERCEPT -11.679 ABS ABS Model -0.394 Non-ABS Model 0 VEHICLE MAKE/SERIES Harley Davidson V-Rod -0.033 Honda CBR1000RR 0.961 Honda Gold Wing 0.000 Honda Interceptor 800 -0.243 -0.689 Honda Reflex Honda Silver Wing 0.035 Honda ST1300 0.015 Kawasaki Concours 14 -0.225 Suzuki Bandit 1250 -1.088 Suzuki Burgman 650 -0.390 Suzuki SV650 -0.066 Yamaha FJR1300 -0.747 -0.176 VEHICLE AGE RATED DRIVER AGE Unknown 0.159 14-24 1.415 25-39 -0.007 40-65+ 0.354 RATED DRIVER GENDER Female -0.153 Male 0 Unknown -0.289 VEHICLE DENSITY 0-99 -0.163 100-500+ 0.257

<0.0001 0.030

0.010 7.900 0.460 3.080 0.020 0.000 0.190 1.170 1.390 0.030 5.600 14.110 0.300 16.110 3.660 0.210 3.860 1.000 2.930
0.936 0.005 0.500 0.079 0.901 0.955 0.664 0.280 0.238 0.855 0.018 0.0002 0.582 <0.0001 0.976 0.056 0.643 0.049 0.316 0.087
TABLE 13 SUMMARY RESULTS OF LINEAR REGRESSION ANALYSIS OF BODILY INJURY LIABILITY CLAIM SEVERITIES
0.200 6.490 0.230 4.520 0.150 0.200
0.652 0.839 0.628 0.341 0.928 0.906
TABLE 14 DETAILED RESULTS OF LINEAR REGRESSION ANALYSIS OF BODILY INJURY LIABILITY CLAIM SEVERITIES
ESTIMATE 10.150 -0.-0.289 0.0.622 -1.273 -0.856 -0.512 -0.572 -0.014 -0.862 -0.144 -0.003 -0.043 -1.347 -0.665 -0.0.238 0.0.079 0.0.162
EXPONENT ESTIMATE 25,578.310 0.0.749 1.1.0.425 0.599 0.564 0.986 0.422 0.866 0.997 0.958 0.260 0.514 0.1.269 1.1.082 1.1.176
STANDARD ERROR 0.293 0.0.994 0.0.748 0.906 0.872 0.531 0.832 1.535 0.696 0.913 0.609 0.088 0.699 0.806 0.0.390 0.0.320 0.0.367
CHISQUARE 1,201.140 0.210

<0.0001 0.648

Harley Davidson V-Rod Honda CBR1000RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki Burgman 650 Suzuki SV650 Yamaha FJR1300 VEHICLE AGE
0.080 0.020 0.690 1.980 0.960 0.930 0.470 0.000 1.530 0.0.240 3.710 0.680 0.020 0.370 0.110 0.060 0.020 0.190
0.771 0.888 0.405 0.160 0.326 0.334 0.491 0.993 0.216 0.875 0.996 0.626 0.054 0.409 0.887 0.541 0.736 0.806 0.876 0.659
TABLE 15 RESULTS FOR BODILY INJURY LIABILITY OVERALL LOSSES DERIVED FROM CLAIM FREQUENCY AND SEVERITY MODELS
-11.679 -0.-0.033 0.-0.243 -0.689 0.035 0.015 -0.225 -1.088 -0.390 -0.066 -0.747 -0.176 0.159 1.415 -0.0.354 -0.-0.289 -0.0.257 0.159 0.0.409 0.0.360 0.393 0.284 0.260 0.517 1.006 0.330 0.363 0.316 0.047 0.288 0.353 0.0.185 0.0.147 0.0.150
INTERCEPT ABS ABS Model Non-ABS Model VEHICLE MAKE/SERIES Harley Davidson V-Rod Honda CBR1000RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki Burgman 650 Suzuki SV650 Yamaha FJR1300 VEHICLE AGE RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ RATED DRIVER GENDER Female Male Unknown VEHICLE DENSITY 0-99 100-499 500+

17,1,984 1,837 1,1,378 1,045 30,671

3,55 2,634 9,420

35% 83% 99% 100% 81% 77% 89% 83% 72% 54% 78% 99% 80% 96% 84% 64% 87% 28% 77%
65% 17% 1% 0% 19% 23% 11% 17% 28% 46% 22% 1% 20% 4% 16% 36% 13% 72% 23%
VEHICLE AGE -RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ RATED DRIVER GENDER Female Male Unknown DENSITY 0-99 100-499 500+ 306 5,973 9,330 8,211 6,799 5,090 3,323 1,060 EXPOSURE (YRS) 4,4,257 25,977 5,130 EXPOSURE (YRS) 2,155 25,866 12,070 EXPOSURE (YRS) 13,389 16,099 10,603
1% 15% 23% 20% 17% 13% 8% 3% % 10% 1% 11% 65% 13% % 5% 65% 30% % 33% 40% 26%
APPENDIX C DISTRIBUTION OF EXPOSURE FOR INDEPENDENT VARIABLE, BODILY INJURY LIABILITY COVERAGE
Harley Davidson V-Rod Honda CBR1000RR Honda Gold Wing Honda Interceptor 800 Honda Reflex Honda Silver Wing Honda ST1300 Kawasaki Concours 14 Suzuki Bandit 1250 Suzuki Burgman 650 Suzuki SV650 Yamaha FJR1300 Total
1,825 1,424 55,249 4,707 6,302 6,112 6,357 1,5,109 3,071 4,420 96,605
13,922 1,1,110 2,209 1,10,317 33,470
79% 100% 80% 76% 87% 85% 69% 55% 82% 82% 96% 30% 74%
21% 0% 20% 24% 13% 15% 31% 45% 18% 18% 4% 70% 26%
VEHICLE AGE -RATED DRIVER AGE Unknown 14-24 25-39 40-64 65+ RATED DRIVER GENDER Female Male Unknown VEHICLE DENSITY 0-99 100-499 500+ 1,206 20,356 29,044 27,415 23,034 17,077 10,458 1,487 EXPOSURE (YRS) 8,977 1,304 14,454 88,583 16,758 EXPOSURE (YRS) 5,835 72,444 51,797 EXPOSURE (YRS) 39,326 52,750 38,000
1% 16% 22% 21% 18% 13% 8% 1% % 7% 1% 11% 68% 13% % 4% 56% 40% % 30% 41% 29%
1005 North Glebe Road Arlington, VA 22201