Highway Loss Data Institute Bulletin Vol. 35, No. 40 : December 2018 Noncrash fire claim frequencies for Kia and Hyundai midsize fourdoor cars and SUVs an update Summary In September 2018, the Highway Loss Data Institute (HLDI) conducted a study comparing noncrash fire insurance losses for five specific Kia/Hyundai vehicles with other vehicles of the same size and class (HLDI, 2018a). The study results indicated increased noncrash fire risk for several model years of the Kia Optima, Kia Sorento, Hyundai Sonata, Hyundai Santa Fe, and Santa Fe Sport. Based on discussions with the National Highway Traffic Safety Administration s (NHTSA) Office of Defects Investigation, the engine type was considered as a possible contributing factor for the elevated noncrash fire risk. This study expands on the original study by analyzing noncrash fire claim frequency by engine variant. The model years and vehicles analyzed in this report include the 2011 15 Kia Optima (midsize four-door car), 2011 15 Kia Sorento (midsize SUV), 2011 14 Hyundai Sonata (midsize four-door car), 2011 12 Hyundai Santa Fe (midsize SUV), and 2013 14 Hyundai Santa Fe Sport (midsize SUV). As shown in the following figure, the 2.0-liter turbocharged engine had the highest average noncrash fire claim frequency (4.2 claims per 10,000 insured vehicle years) followed by the 2.4-liter engine (2.7). Comparatively, the average noncrash fire claim frequency for the control vehicles was only 1.7 per 10,000 insured vehicle years. 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Introduction Observed noncrash fire claim frequencies for Kia and Hyundai vehicles versus control vehicles by engine type 0.0 2.0-liter turbocharged engine 2.4-liter engine 3.3-liter engine 3.5-liter engine In September 2018, HLDI conducted a study investigating noncrash fire insurance losses for several Kia and Hyundai vehicles. That study was based on a petition by the Center for Auto Safety, which indicated there was a high volume of complaints for the 2011 14 Kia Optima, Kia Sorento, Hyundai Sonata, and Hyundai Santa Fe. HLDI s own analysis found significantly higher noncrash fire claim frequency for several model years of these vehicles. HLDI shared these results with NHTSA s Office of Defects Investigation. Based on these discussions, it was believed that problems with the engine may be a contributing factor to the elevated noncrash fire risk. This study expands on the prior study by examining noncrash fire claim frequencies by engine type for the affected vehicles. control group
Method Insurance data Automobile insurance covers damage to vehicles and property in crashes plus injuries to people involved in crashes. Different insurance coverages pay for vehicle damages versus injuries, and different coverages may apply depending on who is at fault. This study is based on comprehensive coverage. Comprehensive coverage insures against theft or physical damage to insured people s own vehicles that occurs for reasons other than crashes. Losses due to noncrash fires are covered under comprehensive coverage. Insurance measures Noncrash fire claim frequency is the number of noncrash fire claims divided by comprehensive exposure, expressed as claims per 10,000 insured vehicle years. Typically, under comprehensive insurance coverage, claim frequency is expressed as claims per 1,000 insured vehicle years. This distinction is made because noncrash fire claims are filed infrequently. Exposure is the length of time a vehicle is insured under a given coverage type and is measured in insured vehicle years. An insured vehicle year is one vehicle insured for 1 year, two vehicles insured for 6 months, etc. Subject vehicles Table 1 shows comprehensive exposure, noncrash fire claims, and noncrash fire claim frequency for the five target vehicles by engine type. The examined vehicles include the 2011 15 Kia Optima, 2011 14 Hyundai Sonata, 2011 15 Kia Sorento, 2011 12 Hyundai Santa Fe, and 2013 14 Hyundai Santa Fe Sport. The hybrid versions for both the Kia Optima and Hyundai Sonata were excluded from the statistical analysis due to limited data. Note that the 2011 Kia Sorento has an existing noncrash fire recall (#10V388000). According to the recall, approximately 35,000 Kia Sorento vehicles were potentially affected. Table 1: Kia and Hyundai comprehensive exposure, noncrash fire claims count, and noncrash fire claim frequency by engine type Engine type Exposure Claims Claim frequency 2011 15 Kia Optima 2.0-liter turbocharged engine 411,730 204 5.0 2.4-liter engine 1,892,499 537 2.8 Total 2,304,229 741 3.2 2011 14 Hyundai Sonata 2.0-liter turbocharged engine 436,398 182 4.2 2.4-liter engine 3,617,587 982 2.7 Total 4,053,985 1,164 2.9 2011 15 Kia Sorento 2.4-liter engine 1,686,811 435 2.6 3.3-liter engine 244,852 39 1.6 3.5-liter engine 717,173 140 2.0 Total 2,648,836 614 2.3 2011 12 Hyundai Santa Fe 2.4-liter engine 347,296 73 2.1 3.5-liter engine 324,533 40 1.2 Total 671,829 113 1.7 2013 14 Hyundai Santa Fe Sport 2.0-liter turbocharged engine 241,025 69 2.9 2.4-liter engine 394,948 86 2.2 Total 635,973 155 2.4 HLDI Bulletin Vol 35, No. 40 : December 2018 2
Two control groups were constructed to accurately measure insurance losses relative to the Kia/Hyundai midsize fourdoor cars and SUVs (Table 2 and Table 3). The control groups consisted of all midsize four-door cars and SUVs manufactured in model years 2011 15, except for vehicles with a documented noncrash fire-related recall. Other gaps in model year availability were due to production either being terminated on certain vehicles or not yet started. Table 2: Noncrash fire midsize four-door cars control vehicle population Make Series Model years Exposure Noncrash fire claims Claim frequency Acura TSX 4dr 2011 14 344,320 37 1.1 Buick Verano 4dr 2012 15 535,166 83 1.6 Chevrolet Malibu 4dr (NEW) 2011 12,2014 15 2,617,380 600 2.3 Chrysler 200 4dr 2WD 2011 12,2014 799,692 193 2.4 Dodge Avenger 4dr 2WD 2011 12,2014 636,535 219 3.4 Ford Fusion 4dr 2WD 2012, 2015 1,502,125 228 1.5 Ford Fusion 4dr 4WD 2012, 2015 54,271 7 1.3 Ford Fusion hybrid 4dr 2WD 2012, 2015 43,147 5 1.2 Honda Accord 4dr 2011 12 2,472,506 293 1.2 Mazda 6 4dr 2WD 2013, 2015 210,071 18 0.9 Mitsubishi Galant 4dr 2WD 2011 12 124,217 24 1.9 Nissan 810/Maxima sedan 2011 14 986,975 243 2.5 Nissan Altima 4dr 2011 15 5,116,372 997 1.9 Nissan Altima hybrid 4dr 2011 15,884 2 1.3 Toyota Camry 4dr 2WD 2011 13, 2015 6,018,456 768 1.3 Toyota Camry hybrid 4dr 2011 13, 2015 471,175 46 1.0 Volkswagen CC 4dr 2WD 2011 70,783 18 2.5 Volkswagen CC 4dr 4WD 2011 1,774 0.0 Volkswagen Jetta hybrid 4dr 2013 23,616 3 1.3 Volkswagen New Jetta 4dr 2013 537,632 76 1.4 Volvo S40 4dr 2WD (NEW) 2011 23,112 5 2.2 Total 22,605,211 3,865 1.7 Table 3: Noncrash fire midsize SUV control vehicle population Make Series Model years Exposure Noncrash fire claims Claim frequency Chevrolet Truck Captiva sport 4dr 2WD 2013 15 239,701 59 2.5 Chevrolet Truck Equinox 4dr 2WD 2011 15 2,808,220 350 1.2 Chevrolet Truck Equinox 4dr 4WD 2011 15 1,309,862 171 1.3 Dodge Truck Nitro 4dr 2WD 2011 80,987 25 3.1 Dodge Truck Nitro 4dr 4WD 2011 84,615 21 2.5 Ford Truck Edge 4dr 2WD 2014 15 277,271 24 0.9 Ford Truck Edge 4dr 4WD 2014 15 274,193 21 0.8 Ford Truck Explorer 4dr 4X2 2011 12, 2014 15 818,475 110 1.3 Ford Truck Explorer 4dr 4X4 2011 12, 2014 15 996,453 141 1.4 Ford Truck Flex 4dr 2WD 2011 12, 2014 15 264,790 34 1.3 Ford Truck Flex 4dr 4WD 2011 12, 2014 15 115,921 16 1.4 GMC Truck Terrain 4dr 2WD 2011 15 1,205,882 160 1.3 GMC Truck Terrain 4dr 4WD 2011 15 634,803 82 1.3 Honda Accord Crosstour 4dr 2WD 2011 12 89,233 14 1.6 HLDI Bulletin Vol 35, No. 40 : December 2018 3
Table 3: Noncrash fire midsize SUV control vehicle population Make Series Model years Exposure Noncrash fire claims Claim frequency Honda Accord Crosstour 4dr 4WD 2011 12 104,685 13 1.2 Honda Pilot 4dr 2WD 2011, 2013 15 689,972 98 1.4 Honda Pilot 4dr 4WD 2011, 2013 15 1,420,406 151 1.1 Hyundai Veracruz 4dr 2WD 2011 12 61,820 6 1.0 Hyundai Veracruz 4dr 4WD 2011 12 38,073 10 2.6 Jeep Grand Cherokee SRT 4dr 4WD 2012 15 45,410 10 2.2 Jeep Grand Cherokee 4dr 2WD 2015 124,719 12 1.0 Jeep Grand Cherokee 4dr 4WD 2015 518,946 60 1.2 Jeep Liberty 4dr 2WD 2011 12 152,623 27 1.8 Jeep Liberty 4dr 4WD 2011 12 588,575 131 2.2 Jeep Wrangler 4dr 4X4 2014 15 847,080 206 2.4 Mazda CX-7 4dr 2WD 2011 12 213,618 20 0.9 Mazda CX-7 4dr 4WD 2011 12 19,775 9 4.6 Mazda CX-9 4dr 2WD 2011 15 230,437 17 0.7 Mazda CX-9 4dr 4WD 2011 15 279,761 34 1.2 Mitsubishi Endeavor 4dr 2WD 2011 10,616 1 0.9 Mitsubishi Endeavor 4dr 4WD 2011 40,833 5 1.2 Nissan Murano 4dr 2WD 2011 12 210,412 20 1.0 Nissan Murano 4dr 4WD 2011 12 352,905 44 1.2 Nissan Murano Crosscabrio 4WD 2011 12 17,186 2 1.2 Nissan Pathfinder 4dr 4X2 2011 15 495,646 57 1.2 Nissan Pathfinder 4dr 4X4 2011 15 650,689 94 1.4 Nissan Pathfinder hybrid 4dr 2WD 2014 3,105-0.0 Nissan Pathfinder hybrid 4dr 4WD 2014 5,118 2 3.9 Nissan Xterra 4dr 4X2 2011 15 77,907 5 0.6 Nissan Xterra 4dr 4X4 2011 15 268,761 22 0.8 Subaru B9 Tribeca 4dr 4WD 2013 14 9,080-0.0 Toyota 4Runner 4dr 4X2 2011 15 335,003 32 1.0 Toyota 4Runner 4dr 4X4 2011 15 792,926 69 0.9 Toyota FJ Cruiser 4dr 4X2 2011 14 11,576 2 1.7 Toyota FJ Cruiser 4dr 4X4 2011 14 217,419 30 1.4 Toyota Highlander 4dr 2WD 2012 13, 2015 662,648 52 0.8 Toyota Highlander 4dr 4WD 2012 13, 2015 1,069,878 83 0.8 Toyota Highlander hybrid 4dr 4WD 2012 13, 2015 64,732 6 0.9 Toyota Venza 4dr 2WD 2011 15 333,389 30 0.9 Toyota Venza 4dr 4WD 2011 15 315,633 23 0.7 Total 20,481,767 2,611 1.3 HLDI Bulletin Vol 35, No. 40 : December 2018 4
Total exposure and noncrash fire claims by model year for the Kia Optima, Hyundai Sonata, and the control vehicles are presented in Table 4. The studied midsize four-door cars and control vehicles totaled 28,963,426 years of exposure and 5,770 noncrash fire claims. Table 4: Exposure and noncrash fire claims by model year for the Kia and Hyundai midsize four-door cars and control vehicles Exposure Noncrash fire claims Model year Kia Optima Hyundai Sonata Control vehicles Kia Optima Hyundai Sonata Control vehicles 2011 318,562 1,625,601 6,349,026 139 603 1,086 2012 450,407 868,606 8,220,170 182 260 1,335 2013 742,114 1,153,804 3,314,967 253 239 514 2014 304,686 405,975 1,924,003 72 62 453 2015 488,460 2,797,046 95 477 Total 2,304,229 4,053,985 22,605,211 741 1,164 3,865 Similarly, total exposure and noncrash fire claims by model year for the Kia Sorento, Hyundai Santa Fe, Santa Fe Sport, and the control vehicles are presented in Table 5. The studied midsize SUVs and control vehicles totaled 24,438,404 years of exposure and 3,493 noncrash fire claims. Model year Table 5: Exposure and noncrash fire claims by model year for the Kia and Hyundai midsize SUVs and control vehicles Hyundai Santa Fe Hyundai Santa Fe Sport Exposure Kia Sorento Control vehicles Hyundai Santa Fe Noncrash fire claims Hyundai Santa Fe Sport Kia Sorento Control vehicles 2011 382,350 964,563 5,081,062 56 194 733 2012 289,479 546,725 4,233,961 57 162 502 2013 378,528 447,323 3,568,914 99 118 382 2014 257,444 409,333 3,526,848 56 88 490 2015 280,891 4,070,982 52 504 Total 671,829 635,973 2,648,836 20,481,767 113 155 614 2,611 Analysis methods Regression analysis was used to quantify the noncrash fire claim frequency differences in a given model year between the engines of study vehicles and the same size and class control vehicles with no records of a noncrash fire-related recall, while controlling for other covariates. The other covariates included calendar year, garaging state, vehicle density (number of registered vehicles per square mile), rated driver age group, rated driver gender, rated driver marital status, deductible range, and risk. Model year was not included as a covariate, as all regressions performed focused on one model year at a time. Claim frequency was modeled using a Poisson distribution, with a logarithmic link function. For space reasons, only the estimates for the differences in insurance losses for the study vehicles compared with the control vehicle population are shown on the following pages. To illustrate the analyses, the Appendix contains full model results for the noncrash fire claim frequency for the 2012 Kia Optima. To further simplify the presentation here, the exponent of the parameter was calculated, 1 was subtracted, and the result multiplied by 100. The resulting number corresponds to the difference between the study vehicle and the control vehicles of the same model years. For example, the estimate of the regression parameter for the 2012 Kia Optima with the 2.0-liter turbocharged engine on noncrash fire claim frequency was 1.2738; thus, these vehicles had a claim frequency that was 257 percent higher than the 2012 control vehicles ((exp(1.2738)-1)*100=257). HLDI Bulletin Vol 35, No. 40 : December 2018 5
Results Midsize four-door cars Figure 1 shows the estimated differences in noncrash fire claim frequency by engine variant for the Kia Optima versus the control vehicles for model years 2011 15. Here, and in subsequent figures, the vertical I-bars represent the 95 percent confidence limits for the estimates. For vehicles with 2.0-liter turbocharged engine, claim frequencies were more than triple that of the control vehicles in model years 2011 13. The difference dropped substantially for the 2014 model year but was still statistically significant. The 2015 model year showed no significant differences. Noncrash fire claim frequencies for vehicles with the 2.4-liter engine showed a similar trend, although the magnitude of the differences was less than those for the 2.0-liter turbocharged engine. 400% 350% 300% 250% 200% 150% 100% 50% 0% -50% -100% Figure 1: Estimated differences in noncrash fire claim frequency for Kia Optima versus the control group by engine type and model year 2011 2012 2013 Figure 2 shows the estimated differences in the noncrash fire claim frequency for the Hyundai Sonata with different engines versus the control vehicles for model years 2011 14. The 2015 Hyundai Sonata was not analyzed as it was redesigned that year. For vehicles with a 2.0-liter turbocharged engine, claim frequencies were 113, 224, and 239 percent higher than the control vehicles in model years 2011, 2012, and 2013, respectively, but the difference dropped significantly in 2014. For vehicles with a 2.4-liter engine, the claim frequencies in model years 2011 14 continuously dropped from 139 percent higher to 33 percent lower than the control vehicles. 2014 2015 2.0-liter turbocharged engine 2.4 liter engine 400% 350% 300% 250% 200% 150% 100% 50% 0% -50% -100% Figure 2: Estimated differences in noncrash fire claim frequency for Hyundai Sonata versus the control group by engine type and model year 2011 2012 2013 2014 2.0-liter turbocharged engine 2.4 liter engine HLDI Bulletin Vol 35, No. 40 : December 2018 6
Midsize SUVs Figure 3 shows the estimated differences in the noncrash fire claim frequency by engine type for the Kia Sorento versus the control vehicles for model years 2011 15. The 6-cylinder, fuel-injected, 3.5-liter engine was replaced by the 6-cylinder, fuel-injected, 3.3-liter engine after model year 2013. The 2011 Kia Sorento has an existing noncrash fire recall, which affected approximately 35,000 vehicles. Noncrash fire claim frequencies were significantly higher for all vehicles with the 2.4-liter engine. The difference was largest for the 2012 and 2013 model years where claim frequencies were more than triple that of the control vehicles. The 3.5-liter engine was also associated with significantly higher claim frequencies of approximately 50 percent compared with the control group. Results for the 3.3-liter engine were not statistically significant but were associated with higher noncrash fire risk. 400% 350% 300% 250% 200% 150% 100% 50% 0% -50% -100% Figure 3: Estimated differences in noncrash fire claim frequency for Kia Sorento versus the control group by engine type and model year 2011 2012 2013 2014 2015 2.4 liter engine 3.5 liter engine 3.3 liter engine pattern indicates noncrash fire recall Results by engine type for the 2011 12 Hyundai Santa Fe are shown in Figure 4. Only the 2012 Hyundai Santa Fe with the 2.4-liter engine had a significantly higher claim frequency than the control vehicles (128 percent). 400% 350% 300% 250% 200% 150% 100% 50% 0% -50% -100% Figure 4: Estimated differences in noncrash fire claim frequency for Hyundai Santa Fe versus the control group by engine type and model year 2011 2012 2.4 liter engine 3.5 liter engine HLDI Bulletin Vol 35, No. 40 : December 2018 7
Figure 5 shows the estimated differences in the noncrash fire claim frequency by engine type for the 2013 14 Hyundai Santa Fe Sport. Vehicles with either the 2.0-liter turbocharged engine or the 2.4-liter engine had significantly higher noncrash fire claim frequencies than the control vehicles. The magnitude of the effect was larger for the 2013 model year. The 2.0-liter turbocharged engine was also associated with higher claim frequencies than the 2.4-liter engine for both model years. 400% 350% 300% 250% 200% 150% 100% 50% 0% -50% -100% Figure 5: Estimated differences in noncrash fire claim frequency for Hyundai Santa Fe Sport versus the control group by engine type and model year 2013 2014 2.0-liter turbocharged engine 2.4 liter engine The 2.0-liter turbocharged, 2.4-liter and 3.5-liter engines were all associated with significantly higher noncrash fire claim frequencies across vehicles and model years. Figures 6 8 show the trends of noncrash fire claim frequencies by vehicle age for those with the 2.0-liter turbocharged engine (i.e., Kia Optima, Hyundai Sonata, and Hyundai Santa Fe Sport), the 2.4-liter engine (i.e., Kia Optima, Hyundai Sonata, Kia Sorento, Hyundai Santa Fe, and Santa Fe Sport) and the 3.5-liter engine (i.e., Kia Sorento and Hyundai Santa Fe). For all engine variants, the risk of noncrash fire increases with vehicle age, but the 3.5-liter engine had a smaller growth rate compared with the other two engines. 12 10 Figure 6: Observed noncrash fire claim frequencies for 2.0-liter turbocharged engine by vehicle age Kia Optima Hyundai Sonata Hyundai Santa Fe Sport 8 6 4 2 0-1 to 1 2 to 4 5+ HLDI Bulletin Vol 35, No. 40 : December 2018 8
12 10 8 6 Figure 7: Observed noncrash fire claim frequencies for 2.4-liter engine by vehicle age Kia Optima Hyundai Sonata Hyundai Santa Fe Hyundai Santa Fe Sport Kia Sorento 4 2 0-1 to 1 2 to 4 5+ 12 10 Figure 8: Observed noncrash fire claim frequencies for 3.5-liter engine by vehicle age Kia Sorento Hyundai Santa Fe 8 6 4 2 0-1 to 1 2 to 4 5+ Discussion For the vehicles examined, the overall trends in noncrash fire claim frequencies by model year were consistent with the results of the previous study (HLDI, 2018a). In general, the studied Kia/Hyundai vehicles of all engine types had higher claim frequencies than the control vehicles. This could be an indication of a problem unrelated to any specific engine that affects all vehicles. On the other hand, vehicles with the 2.0-liter turbocharged engine had substantially higher noncrash fire claim frequencies compared with the other engine types and the control vehicles. This could be an indication of further problems with the turbocharged engine. A concurrent study by HLDI (2018b) indicates that in general, turbocharged engines are associated with higher noncrash fire claim frequencies compared with nonturbo engines. Further investigation is needed to determine whether the turbocharged engine is a determinant or only an influencing factor for the elevated noncrash fire risk. Vehicles with the 2.4-liter engine also exhibited elevated noncrash fire risk, although not as high as the 2.0-liter turbocharged engine. Both engines are from the Theta II engine family, which are currently the subject of ongoing NHTSA investigations (RQ 17-003 and RQ 17-004). In 2017, Kia and Hyundai recalled several series of vehicles with Theta II engines, which could cause problems with oil flow or engine seizing (#17V224000 and #17V226000, respectively). The recalled vehicles included several model years of the vehicles examined in this study. Although these recalls were not noncrash fire-related, these engine problems could be related to the elevated noncrash fire risk for vehicles with the 2.0-liter turbocharged engine and 2.4-liter engine. The 3.3-liter engine and 3.5-liter engine are from the Lambda engine family, which showed smaller differences in noncrash fire claim frequency compared with the control vehicles. HLDI Bulletin Vol 35, No. 40 : December 2018 9
The results of this analysis will be shared with NHTSA s Office of Defects Investigation in support of their ongoing investigations. Regression analysis was used to quantify the difference between the study vehicles and same size and class control vehicles without noncrash fire-related recalls while controlling for other covariates. This study like most HLDI studies controls for calendar year, garaging state, vehicle density (number of registered vehicles per square mile), rated driver age group, rated driver gender, rated driver marital status, deductible range, and risk. These covariates are highly correlated with collision claim frequency. However, unlike collision claims, noncrash fire claims do not result from a crash and can occur without a driver in the vehicle. Therefore, it is not expected that all the covariates typically used would be relevant to the noncrash fire claims. In similar studies on noncrash fire losses for different vehicles, HLDI conducted several analyses with and without different covariates, but the inclusion or exclusion of certain covariates did not significantly impact the results. HLDI will continue to evaluate the effect of these covariates on noncrash fire insurance losses in future studies. For consistency with other studies, the results presented in this bulletin include the usual covariates except for the model year, because only one model year was examined at a time. References Highway Loss Data Institute. (2018a). Noncrash fire insurance losses for Kia and Hyundai midsize four-door cars and SUVs. Loss Bulletin, 35(32). Arlington, VA. Highway Loss Data Institute (2018b). Noncrash fire losses for turbo/supercharged engines. Loss Bulletin, 35(42). Arlington, VA. HLDI Bulletin Vol 35, No. 40 : December 2018 10
Parameter Appendix: Illustrative regression results noncrash fire claim frequency for 2012 Kia Optima Degrees of freedom Estimate Effect Standard error Wald 95% confidence limits Chi-square P-value Intercept 1-14.3349 0.1156-14.5615-14.1083 15377.80 <0.0001 Calendar year 2011 1-0.3133-26.9% 0.3382-0.9762 0.3496 0.86 0.3543 2012 1-0.6862-49.7% 0.1168-0.9151-0.4573 34.52 <0.0001 2013 1-0.7389-52.2% 0.0987-0.9324-0.5454 56.03 <0.0001 2014 1-0.6273-46.6% 0.0919-0.8075-0.4471 46.55 <0.0001 2015 1-0.4373-35.4% 0.0847-0.6033-0.2713 26.67 <0.0001 2016 1-0.1732-15.9% 0.0781-0.3263-0.0201 4.91 0.0266 2018 1 0.0433 4.4% 0.0886-0.1303 0.2169 0.24 0.6249 2017 0 0 0 0 0 0 Deductible range 0 1-0.1931-17.6% 0.1249-0.4379 0.0517 2.39 0.1220 Registered vehicle density 1 50 1-0.7341-52.0% 0.2542-1.2323-0.2359 8.34 0.0039 51 100 1-0.3662-30.7% 0.0906-0.5437-0.1886 16.33 <0.0001 101 200 1-0.3109-26.7% 0.2054-0.7134 0.0916 2.29 0.1300 201 250 1-0.1510-14.0% 0.0841-0.3159 0.0138 3.22 0.0726 501 1000 1 0.0336 3.4% 0.0802-0.1236 0.1908 0.18 0.6751 1001+ 1-0.2559-22.6% 0.3557-0.9531 0.4412 0.52 0.4718 251 500 0 0 0 0 0 0 <50 1 0.4800 61.6% 0.0992 0.2855 0.6745 23.40 <0.0001 50 99 1 0.0607 6.3% 0.1012-0.1376 0.2590 0.36 0.5484 100 249 1 0.0715 7.4% 0.0834-0.0920 0.2349 0.73 0.3915 250 499 1-0.2317-20.7% 0.0906-0.4093-0.0541 6.54 0.0106 500 999 1-0.0631-6.1% 0.0810-0.2219 0.0956 0.61 0.4357 1000+ 0 0 0 0 0 0 State Alabama 1 0.5252 69.1% 0.1684 0.1951 0.8553 9.72 0.0018 Alaska 1-0.0011-0.1% 0.7142-1.4008 1.3987 0.00 0.9988 Arizona 1-0.2418-21.5% 0.2196-0.6723 0.1886 1.21 0.2708 Arkansas 1 0.3305 39.2% 0.2479-0.1555 0.8164 1.78 0.1826 California 1-0.0826-7.9% 0.1234-0.3244 0.1592 0.45 0.5030 Colorado 1-1.3510-74.1% 0.4560-2.2448-0.4572 8.78 0.0031 Connecticut 1 0.1375 14.7% 0.2404-0.3337 0.6086 0.33 0.5673 Delaware 1 0.6183 85.6% 0.3045 0.0215 1.2151 4.12 0.0423 Dist of Columbia 1 0.8434 132.4% 0.4192 0.0217 1.6651 4.05 0.0443 Florida 1-0.0061-0.6% 0.1288-0.2586 0.2464 0.00 0.9620 Georgia 1 0.0901 9.4% 0.1464-0.1970 0.3771 0.38 0.5386 Hawaii 1-1.3674-74.5% 1.0040-3.3353 0.6004 1.85 0.1732 Idaho 1-1.0444-64.8% 0.7140-2.4439 0.3550 2.14 0.1435 Illinois 1-0.2332-20.8% 0.1660-0.5586 0.0921 1.97 0.1599 Indiana 1-0.1189-11.2% 0.2158-0.5419 0.3042 0.30 0.5818 Iowa 1 0.0676 7.0% 0.2533-0.4289 0.5641 0.07 0.7897 Kansas 1-0.4438-35.8% 0.3147-1.0607 0.1731 1.99 0.1585 Kentucky 1-0.3571-30.0% 0.2410-0.8294 0.1153 2.19 0.1385 Louisiana 1 0.3488 41.7% 0.1836-0.0111 0.7087 3.61 0.0575 Maine 1-0.0165-1.6% 0.4575-0.9131 0.8801 0.00 0.9712 Maryland 1 0.3053 35.7% 0.1659-0.0198 0.6305 3.39 0.0657 Massachusetts 1-0.0392-3.8% 0.1967-0.4247 0.3463 0.04 0.8420 HLDI Bulletin Vol 35, No. 40 : December 2018 11
Parameter Appendix: Illustrative regression results noncrash fire claim frequency for 2012 Kia Optima Degrees of freedom Estimate Effect Standard error Wald 95% confidence limits Chi-square P-value Michigan 1-0.0558-5.4% 0.1987-0.4453 0.3337 0.08 0.7790 Minnesota 1-0.4682-37.4% 0.2401-0.9388 0.0024 3.80 0.0512 Mississippi 1 0.7887 120.1% 0.1827 0.4307 1.1467 18.64 <0.0001 Missouri 1-0.0057-0.6% 0.1948-0.3875 0.3760 0.00 0.9766 Montana 1-0.6256-46.5% 0.7161-2.0291 0.7778 0.76 0.3823 Nebraska 1-1.5246-78.2% 0.7127-2.9214-0.1278 4.58 0.0324 Nevada 1-0.2289-20.5% 0.3052-0.8271 0.3693 0.56 0.4533 New Hampshire 1-0.3705-31.0% 0.5101-1.3701 0.6292 0.53 0.4677 New Jersey 1-0.1018-9.7% 0.1711-0.4371 0.2335 0.35 0.5517 New Mexico 1-0.8803-58.5% 0.4590-1.7799 0.0192 3.68 0.0551 New York 1-0.0462-4.5% 0.1408-0.3222 0.2297 0.11 0.7425 North Carolina 1-0.1288-12.1% 0.1760-0.4737 0.2162 0.54 0.4645 North Dakota 1-7.9150-100.0% 28.7026-64.1711 48.3411 0.08 0.7827 Ohio 1-0.0397-3.9% 0.1584-0.3502 0.2708 0.06 0.8020 Oklahoma 1 0.2119 23.6% 0.2283-0.2355 0.6594 0.86 0.3532 Oregon 1-0.5238-40.8% 0.3651-1.2395 0.1919 2.06 0.1514 Pennsylvania 1 0.0438 4.5% 0.1537-0.2575 0.3452 0.08 0.7755 Rhode Island 1-1.0417-64.7% 0.7127-2.4386 0.3553 2.14 0.1439 South Carolina 1 0.3674 44.4% 0.1743 0.0258 0.7089 4.44 0.0350 South Dakota 1-0.4402-35.6% 0.5865-1.5897 0.7092 0.56 0.4529 Tennessee 1 0.1857 20.4% 0.1779-0.1630 0.5345 1.09 0.2966 Utah 1-0.4904-38.8% 0.3881-1.2511 0.2702 1.60 0.2064 Vermont 1-7.3558-99.9% 26.3787-59.0570 44.3454 0.08 0.7804 Virginia 1-0.0635-6.2% 0.1718-0.4003 0.2733 0.14 0.7117 Washington 1-0.1843-16.8% 0.2418-0.6583 0.2896 0.58 0.4459 West Virginia 1 0.3450 41.2% 0.2938-0.2308 0.9209 1.38 0.2402 Wisconsin 1 0.1166 12.4% 0.2010-0.2774 0.5106 0.34 0.5619 Wyoming 1-8.2759-100.0% 36.5825-79.9763 63.4246 0.05 0.8210 Texas 0 0 0 0 0 0 Rated driver age <25 1-0.0410-4.0% 0.0944-0.2260 0.1439 0.19 0.6636 66 1-0.2906-25.2% 0.0815-0.4504-0.1308 12.71 0.0004 Unknown 1-0.0895-8.6% 0.1589-0.4009 0.2219 0.32 0.5732 25 65 0 0 0 0 0 0 Rated driver gender Male 1 0.0215 2.2% 0.0566-0.0895 0.1325 0.14 0.7043 Rated driver marital status Unknown 1 0.2435 27.6% 0.2385-0.2238 0.7109 1.04 0.3072 Female 0 0 0 0 0 0 Single 1 0.3134 36.8% 0.0590 0.1977 0.4291 28.18 <0.0001 Unknown 1-0.0737-7.1% 0.2328-0.5300 0.3825 0.10 0.7514 Married 0 0 0 0 0 0 Risk Nonstandard 1 0.2485 28.2% 0.0899 0.0723 0.4247 7.64 0.0057 Standard 0 0 0 0 0 0 Engine type 2.0-liter turbocharged engine 1 1.2738 257.4% 0.1264 1.0260 1.5216 101.51 <0.0001 2.4-liter engine 1 0.7491 111.5% 0.0970 0.5589 0.9393 59.59 <0.0001 Others 0 0 0 0 0 0 HLDI Bulletin Vol 35, No. 40 : December 2018 12
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. DW201812 HH Highway Loss Data Institute 1005 N. Glebe Road, Suite 700 Arlington, VA 22201 +1 703 247 1600 iihs-hldi.org COPYRIGHTED DOCUMENT, DISTRIBUTION RESTRICTED 2018 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 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.