How do Consumers Think About Usage-Based Auto Insurance? –A Survey Analysis from Taiwan
-
Jin-Lung Peng
Abstract
Despite the inherent benefits of usage-based auto insurance (UBI), such as loss prevention and reduced premiums, the UBI market has encountered limitations hindering its expansion in practical terms. This study aims to explore the attitudes of consumers towards UBI and the pivotal factors influencing their decision to adopt UBI. According to the survey, a better understanding of the rationale behind pay-as-you-drive auto insurance (PAYD) and pay-how-you-drive auto insurance (PHYD) plays an important role in policyholders’ acceptance. In addition, consumers with higher auto insurance premiums and individuals willing to share their data are more likely to buy UBI. Our research suggests that the promotion of UBI should be aligned with increasing individuals’ willingness to share personal information and reducing privacy concerns. This could not only increase the willingness to purchase UBI but also help balance discount expectations between policyholder and insurer.
Funding source: National Science and Technology Council
Award Identifier / Grant number: 109-2410-H-004 -027
Award Identifier / Grant number: 111-2410-H-035 -046 -MY2
-
Research ethics: Not applicable.
-
Informed consent: Not applicable.
-
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Use of Large Language Models, AI and Machine Learning Tools: LLM and AI tools are only used to improve language.
-
Conflict of interest: The authors state no conflict of interest.
-
Research funding: The authors would like to express sincere gratitude for the financial support provided by the National Science and Technology Council (109-2410-H-004 -027 and 111-2410-H-035 -046 -MY2).
-
Data availability: The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.
Appendix A: Variable Definition
Table A provides definitions for all variables.
Summary of variables and corresponding questionnaires.
| Variable | Variable definition | Question |
|---|---|---|
| Year_Driven | Years of driving experience. | How many years have you been driving? |
| Insured | A dummy variable equals one if the driver has ever insured auto insurance. | Do you have any experience with auto insurance coverage? |
| Premium | Auto insurance premium for a recent year. | How much did you approximately spend on auto insurance premiums in recent years? |
| Mileage | Average mileage per year. | What is your approximate annual mileage when driving? |
| Behavior * | A variable that enables drivers to assess the quality of their driving behavior. (1–5, 5 is the best) | How was my driving behavior? |
| UBI_Heard | A dummy variable equals one if the driver has heard of UBI auto insurance. | Have you heard of Usage-Based Insurance (UBI) policies? |
| Acceptance_PAYD * | A variable that enables drivers to assess the acceptance of pay-as-you-drive insurance policies. (1–5, 5 is the most reasonable) | I think it is reasonable to “calculate insurance premium based on driving mileage.” |
| Acceptance_PHYD * | A variable that enables drivers to assess the acceptance of pay-how-you-drive insurance policies. (1–5, 5 is the most acceptable) | I think it is reasonable to calculate insurance premiums based on “driving behavior.” |
| Agreement_information * | A variable that enables drivers to assess the agreement to provide driving data to insurers. (1–5, 5 agree the most) | I am willing to provide my driving behavior data to the insurer in exchange for purchasing UBI insurance. |
| Trust_insurer * | A variable that enables drivers to assess the level of trust in insurers on driving data collection. (1–5, 5 trust the most) | I trust that the insurance company will adequately safeguard my personal information. |
| Willingness_Change_mileage * | A variable that enables drivers to assess the willingness to reduce mileage since purchasing pay-as-you-drive insurance policies. (1–5, 5 willing the most) | I Would drive less if I had insurance that calculates premium based on driving mileage." |
| Willingness_Change_behavior * | A variable that enables drivers to assess their willingness to change driving behavior after purchasing pay-how-you-drive insurance policies. (1–5, 5 willing the most) | I would change my driving habits, such as reducing hard braking and speeding, if I purchased insurance that calculates premiums based on “driving behavior.” |
| Willingness_Buy_UBI | A dummy variable equals one if the driver is interested in purchasing UBI auto insurance. | Are you interested in purchasing UBI insurance? |
-
This table reports the summary of the variables and the survey questionnaires. For questionnaires related to personal perception, our survey responses were collected using a questionnaire designed with a five-point scale, where a higher value indicates a greater level of understanding or agreement, with five being the highest and one being the lowest. These questionnaires use a five-point scale including self-assessment of driving behavior, acceptance of PAYD, acceptance of PHYD, agreement to provide driving information, level of trust in insurers, willingness to change mileage, and willingness to change behavior.
Appendix B: Data on Perception of Behavior, Risk, and Premium
B.1. Perception of driving behavior and risks
The study focuses on consumer attitudes toward UBI through a survey, including aspects such as identifying risky driving behaviors, offering additional services, and identifying effective incentive mechanisms to increase the likelihood of purchase. First, respondents are surveyed about the factors affecting driving risk. The results are reported in Table B1, indicating that out of the 405 total respondents, 361 individuals (89 %) opted for speeding, 292 participants (72 %) favored frequent lane changes, 176 contributors (43 %) selected emergency braking, 126 respondents (31 %) chose emergency acceleration, 45 participants (11 %) singled out night driving, 35 individuals (9 %) pointed out rush hour driving, 34 respondents (8 %) highlighted mountain road driving, and 13 participants (3 %) indicated driving in downtown areas. This study demonstrates that the majority of respondents consider speeding and frequent lane changes to be the most dangerous driving behaviors and should be charged a higher premium.
Riskfactors for higher insurance premium.
| Question: I think it is reasonable to have a higher insurance premium for the following driving behaviors. (Up to three choices) | Number | Percentage |
|---|---|---|
| Speeding | 361 | 89 % |
| Frequent lane changes | 292 | 72 % |
| Emergency acceleration | 126 | 31 % |
| Emergency braking | 176 | 43 % |
| Driving at night | 45 | 11 % |
| Driving in downtown areas | 13 | 3 % |
| Driving at peak hours | 35 | 9 % |
| Driving on mountain roads | 34 | 8 % |
-
This table represents respondents’ attitudes toward behaviors that would result in higher premiums. The first column provides respondents with several options for behaviors that could potentially warrant a higher insurance premium. The second and last columns show the number and percentage of respondents who selected each behavior.
In addition, respondents are asked what additional services they think UBI can provide. This question is used to assess individuals’ preferences for UBI product design. The results are shown in Table B2. Notably, dangerous driving alerts are selected by 252 respondents (62 %), followed closely by drowsy driving prevention (61 %). Accident alerts were the third highest selected service at 221 (55 %), while safety alerts and emergency reporting received 211 (52 %) and 202 (50 %) selections, respectively. Obviously, consumers request dangerous driving alerts and prevention of drowsy driving as the primary additional services.
Additional service requested for UBI.
| Question: What additional services do I want from UBI auto insurance? (Up to three choices) | Number | Percentage |
|---|---|---|
| Dangerous driving warning | 252 | 62 % |
| Prevention of fatigue driving | 247 | 61 % |
| Car accident warning | 221 | 55 % |
| Safety alerts | 211 | 52 % |
| Emergency reporting | 202 | 50 % |
| Others | 5 | 1 % |
-
This table represents respondents’ attitudes toward desired additional services for UBI. The first column provides several options for respondents to choose from that UBI can provide. The second and last columns show the number and percentage of respondents choosing each service.
The results suggest that individuals expect rewards for mitigating anticipated risks or perceive the service as capable of reducing driving risk. Additionally, the findings indicate that insured individuals maintain a significant interest in the risk management aspects of UBI. Consequently, leveraging drivers’ data to diminish driving risks remains crucial (Gao, Meng, and Wüthrich 2019; Guillen et al. 2020). Moreover, it can potentially curtail subsidies allocated to high-risk drivers, thereby benefiting low-risk drivers.
B.2. Consumer perception of the premium discounts in UBI
Another important objective of this study is to assess the degree to which the amount or ratio of incentive discounts demanded by consumers can induce them to change their driving behavior and mileage when considering the purchase of UBI. Table B3 illustrates the percentage of UBI premium discounts for incentivizing drivers to alter their driving behavior or mileage.
Required premium discount for behavior and mileage change (percent).
| Question: How much of a percentage of the UBI auto insurance premium discount will make me willing to change my driving behavior or mileage? | Number | Percentage | Cumulative percentage |
|---|---|---|---|
| Within 10 % discount | 14 | 3.7 % | 3.7 % |
| 20 % discount | 99 | 25.8 % | 29.5 % |
| 30 % discount | 100 | 26.1 % | 55.6 % |
| 40 % discount | 23 | 6.0 % | 61.6 % |
| Over 50 % discount | 49 | 12.8 % | 74.4 % |
| I Am not willing to change my driving behavior or mileage no matter how much percentage discount UBI auto insurance offers. | 22 | 5.7 % | 80.2 % |
| I Have no idea how much percentage discount on UBI auto insurance premiums will make me willing to change my behavior. | 76 | 19.8 % | 100.0 % |
| 383 | 100 % |
-
This table presents respondents’ attitudes toward the premium discount, which makes them willing to change their driving behavior and mileage for UBI. The first column provides several options for the premium discount ratio. The second and third columns show the number and percentage of respondents choosing each option. Moreover, the last column shows the cumulative percentage of these options.
As shown in Table B3. If the discount is less than 10 %, only 3.7 % of people are willing to make changes. However, with an increase in the discount range from 10 % to 20 %, there is a significant rise, with 29.5 % of individuals indicating their willingness to alter. Furthermore, as the discount grows to a substantial 30 %, an impressive 55.6 % of individuals express their willingness to make changes. However, 5.7 % of the respondents are unwilling to modify their driving habits or distance traveled despite the percentage of the discount offered. In addition, coefficient estimates from the Ologit and OLS models are reported in Appendix Table B4. Univariate tests for different subsamples are conducted to assess differences in other characteristics (see Tables B5 and B6).
The results of characteristics on the willingness to change mileage or behavior.
| Model | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| Dependent variable | Ologit | Ologit | OLS | OLS |
| Willingness_Change_mileage | Willingness_ Change_behavior | Willingness_Change_mileage | Willingness_ Change_behavior | |
| Independent variable | ||||
| Year_Driven | −0.019 (0.015) |
−0.005 (0.015) |
−0.009 (0.008) |
−0.004 (0.008) |
| Premium | −0.193*** (0.076) |
−0.068 (0.078) |
−0.113*** (0.043) |
−0.012 (0.039) |
| Mileage | −0.178*** (0.061) |
−0.036 (0.057) |
−0.092*** (0.032) |
−0.026 (0.029) |
| Behavior | −0.014 (0.143) |
0.017 (0.146) |
0.002 (0.079) |
−0.021 (0.072) |
| UBI_Heard | −0.079 (0.199) |
−0.110 (0.208) |
−0.057 (0.112) |
−0.000 (0.102) |
| Acceptance_PAYD | −0.055 (0.097) |
−0.057 (0.095) |
−0.041 (0.052) |
−0.072 (0.047) |
| Acceptance_PHYD | −0.048 (0.137) |
1.184*** (0.148) |
0.026 (0.071) |
0.485*** (0.065) |
| Agreement_information | 0.335** (−0.117) |
0.274** (0.122) |
0.143*** (0.063) |
0.125** (0.057) |
| Trust_insurer | 0.262** (0.106) |
0.281*** (0.107) |
0.110** (0.056) |
0.141*** (0.051) |
| Constant | – | – | 2.467*** (0.697) |
1.755*** (0.634) |
| R-squared/Pseudo R-squared | 0.0456 | 0.117 | 0.110 | 0.235 |
| Log likelihood | −521.8 | −462.3 | – | – |
| Observations | 383 | 383 | 383 | 383 |
-
This table reports coefficients from the Ologit and OLS models of willingness to change driving mileage and behavior. The dependent variables of columns (1) and (3), (2) and (4) are willingness to change driving mileage and behavior, respectively. We provided the results of Pseudo R-squared for the Ologit model and R-squared for the OLS model. In addition, the unreported results of the Tobit model about significance are consistent with the basic models, and the results are robust using different models. All variables are defined in Appendix Table A. *** indicates significance at the 1 % level, **at the 5 % level, and *at the 10 % level.
Comparison between low and high required premium discount groups.
| Low premium discount requirement group (a) | High premium discount requirement group (b) | Mean difference (a-b) | |
|---|---|---|---|
| Year_Driven | 14.35 | 14.93 | −0.58 |
| Insured | 0.93 | 0.94 | −0.01 |
| Premium | 1.68 | 16.63 | −14.95 |
| Mileage | 1.45 | 1.83 | −0.38** |
| Behavior | 3.75 | 3.78 | −0.03 |
| UBI_Heard | 0.42 | 0.25 | 0.17*** |
| Acceptance_PAYD | 4.02 | 3.42 | 0.6*** |
| Acceptance_PHYD | 4.25 | 3.97 | 0.28*** |
| Agreement_information | 3.82 | 3.35 | 0.47*** |
| Trust_insurer | 3.45 | 3.03 | 0.42*** |
| Willingness_Change_mileage | 2.77 | 2.43 | 0.34*** |
| Willingness_ Change_behavior | 4.01 | 3.46 | 0.55*** |
| Willingness_Buy UBI | 0.82 | 0.65 | 0.17*** |
| N | 227 | 178 |
-
This table reports the results of the comparison between the low-discount requirement group and the high-discount requirement group. The low discount requirement group includes individuals willing to change their mileage and driving behavior for a discount equal to or less than 30 %. The high discount requirement group includes individuals willing to change their mileage and driving behavior for a discount above 30 %. We consider a 30 % discount to be a significant threshold for subjects to change their mileage and driving behavior based on the results in Appendix Table B3. Student’s t-test is used to examine whether the differences in other characteristics between these two groups are significant. *** indicates significance at the 1 % level; **, at the 5 % level; and *, at the 10 % level.
Comparison between disagree and agree groups on providing driving information.
| Disagree to provide driving information (a) | Agree to provide driving information (b) | Mean difference (a-b) | |
|---|---|---|---|
| Year_Driven | 14.08 | 14.70 | −0.62 |
| Insured | 0.92 | 0.94 | −0.02 |
| Premium | 1.81 | 1.63 | 0.18 |
| Mileage | 2.28 | 1.51 | 0.77*** |
| Behavior | 3.83 | 3.75 | 0.08 |
| UBI_Heard | 0.35 | 0.34 | 0.01 |
| Acceptance_PAYD | 3.20 | 3.85 | −0.65 |
| Acceptance_PHYD | 3.62 | 4.22 | −0.6*** |
| Trust_insurer | 2.28 | 3.44 | −1.16*** |
| Willingness_Change_mileage | 2.10 | 2.71 | −0.61*** |
| Willingness_ Change_behavior | 3.03 | 3.89 | −0.86*** |
| Willingness_Buy_UBI | 0.50 | 0.79 | −0.29*** |
| N | 60 | 345 |
-
This table reports the comparison results between the disagree and agree observations on providing driving information. This table reports the comparison results between observations that disagree and agree to provide driving information. The subsamples are divided into two groups by the agreement to provide a driving information questionnaire (Agreement_information). The questionnaire has a scale of 1–5, where 1 and 2 represent disagreement in providing driving information. Therefore, those who answered 1–2 were grouped as disagree, while those who answered 3–5 were grouped as agree. Student’s t-test is used to examine whether the differences in other characteristics between these two groups are significant. *** indicates significance at the 1 % level; **, at the 5 % level; and *, at the 10 % level.
References
Braun, A., N. Haeusle, and P. Thistle. 2023. “Risk Classification with On-demand Insurance.” Journal of Risk & Insurance. Advanced online publication. https://doi.org/10.1111/jori.12429.Search in Google Scholar
Butler, P., T. Butler, and L. L. Williams. 1988. “Sex-divided Mileage, Accident, and Insurance Cost Data Show that Auto Insurers Overcharge Most Women.” Journal of Insurance Regulation 6 (3): 243–416.Search in Google Scholar
Che, X., A. Liebenberg, and J. Xu. 2022. “Usage-based Insurance – Impact on Insurers and Potential Implications for InsurTech.” North American Actuarial Journal 26 (3): 428–55. https://doi.org/10.1080/10920277.2021.1953536.Search in Google Scholar
Cheng, J., F. Y. Feng, and X. Zeng. 2023. “Pay-as-you-Drive Insurance: Modeling and Implications.” North American Actuarial Journal 27 (2): 303–21. https://doi.org/10.1080/10920277.2022.2077220.Search in Google Scholar
Derikx, S., M. De Reuver, and M. Kroesen. 2016. “Can Privacy Concerns for Insurance of Connected Cars Be Compensated?” Electronic Markets 26 (1): 73–81. https://doi.org/10.1007/s12525-015-0211-0.Search in Google Scholar
Desyllas, P., and M. Sako. 2013. “Profiting from Business Model Innovation: Evidence from Pay-As-You-Drive Auto Insurance.” Research Policy 42 (1): 101–16. https://doi.org/10.1016/j.respol.2012.05.008.Search in Google Scholar
Dijksterhuis, C., B. Lewis-Evans, B. Jelijs, D. de Waard, K. Brookhuis, and O. Tucha. 2015. “The Impact of Immediate or Delayed Feedback on Driving Behaviour in a Simulated Pay-As-You-Drive System.” Accident Analysis & Prevention 75: 93–104. https://doi.org/10.1016/j.aap.2014.11.017.Search in Google Scholar
Ferreira, J.Jr, and E. Minikel. 2012. “Measuring Per Mile Risk for Pay-As-You-Drive Automobile Insurance.” Transportation Research Record 2297 (1): 97–103. https://doi.org/10.3141/2297-12.Search in Google Scholar
Gao, G., S. Meng, and M. V. Wüthrich. 2019. “Claims Frequency Modeling Using Telematics Car Driving Data.” Scandinavian Actuarial Journal 2019 (2): 143–62. https://doi.org/10.1080/03461238.2018.1523068.Search in Google Scholar
Geyer, A., D. Kremslehner, and A. Muermann. 2020. “Asymmetric Information in Automobile Insurance: Evidence from Driving Behavior.” Journal of Risk & Insurance 87 (4): 969–95. https://doi.org/10.1111/jori.12279.Search in Google Scholar
Greenberg, A. 2009. “Designing Pay-Per-Mile Auto Insurance Regulatory Incentives.” Transportation Research Part D: Transport and Environment 14 (6): 437–45. https://doi.org/10.1016/j.trd.2009.01.003.Search in Google Scholar
Guillen, M., J. P. Nielsen, A. M. Pérez-Marín, and V. Elpidorou. 2020. “Can Automobile Insurance Telematics Predict the Risk of Near-Miss Events?” North American Actuarial Journal 24 (1): 141–52. https://doi.org/10.1080/10920277.2019.1627221.Search in Google Scholar
Holzapfel, J., R. Peter, and A. Richter. 2023. “Mitigating Moral Hazard with Usage-based Insurance.” Journal of Risk & Insurance. Advanced online publication. https://doi.org/10.1111/jori.12433.Search in Google Scholar
Hultkrantz, L., and G. Lindberg. 2011. “Pay-as-you-speed an Economic Field Experiment.” Journal of Transport Economics and Policy (JTEP) 45 (3): 415–36.Search in Google Scholar
KPMG. 2020. Taiwan Insurance Report, Klynveld Peat Marwick Goerdeler (KPMG). Available at: https://assets.kpmg.com/content/dam/kpmg/tw/pdf/2020/08/tw-kpmg-taiwan-insurance-report.pdf.Search in Google Scholar
Lahrmann, H., N. Agerholm, N. Tradisauskas, K. K. Berthelsen, and L. Harms. 2012. “Pay as You Speed, ISA with Incentives for Not Speeding: Results and Interpretation of Speed Data.” Accident Analysis & Prevention 48: 17–28. https://doi.org/10.1016/j.aap.2011.03.015.Search in Google Scholar
Lemaire, J., S. C. Park, and K. C. Wang. 2016. “The Use of Annual Mileage as a Rating Variable.” ASTIN Bulletin: The Journal of the IAA 46 (1): 39–69. https://doi.org/10.1017/asb.2015.25.Search in Google Scholar
Litman, T. 2005. “Pay-as-you-drive Pricing and Insurance Regulatory Objectives.” Journal of Insurance Regulation 23 (3): 35–53.Search in Google Scholar
Litman, T. 2011. “Pay-As-You-Drive Pricing for Insurance Affordability.” Victoria Transport Policy Institute Report. https://vtpi.org/payd_aff.pdf (Accessed Dec 15, 2022).Search in Google Scholar
Paefgen, J., T. Staake, and F. Thiesse. 2013. “Evaluation and Aggregation of Pay-As-You-Drive Insurance Rate Factors: A Classification Analysis Approach.” Decision Support Systems 56: 192–201. https://doi.org/10.1016/j.dss.2013.06.001.Search in Google Scholar
Parry, I. W. H. 2005. “Is Pay-As-You-Drive Insurance a Better Way to Reduce Gasoline Than Gasoline Taxes?” American Economic Review 95 (2): 288–93. https://doi.org/10.1257/000282805774670482.Search in Google Scholar
Soleymanian, M., C. B. Weinberg, and T. Zhu. 2019. “Sensor Data and Behavioral Tracking: Does Usage-Based Auto Insurance Benefit Drivers?” Marketing Science 38 (1): 21–43. https://doi.org/10.1287/mksc.2018.1126.Search in Google Scholar
TIGF. 2018. “Taiwan FinTech and Insurtech Report.” Taiwan Insurance Guaranty Fund (TIGF). Available at: https://www.tigf.org.tw/xmfile?xsmsid=0M076543574431881145.Search in Google Scholar
Tselentis, D. I., A. Theofilatos, G. Yannis, and M. Konstantinopoulos. 2018. “Public opinion on usage-based motor insurance schemes: A stated preference approach.” Travel Behaviour and Society 11: 111–8, https://doi.org/10.1016/j.tbs.2018.02.003.Search in Google Scholar
Tselentis, D. I., G. Yannis, and E. I. Vlahogianni. 2017. “Innovative Motor Insurance Schemes: A Review of Current Practices and Emerging Challenges.” Accident Analysis & Prevention 98: 139–48. https://doi.org/10.1016/j.aap.2016.10.006.Search in Google Scholar
© 2025 Walter de Gruyter GmbH, Berlin/Boston
Articles in the same Issue
- Frontmatter
- Featured Articles (Research Paper)
- Net Reserve Calculation for Whole Life Insurance Under Mean-Reverting Stochastic Interest Rate Models
- Integrating Remote Sensing Data in Crop Insurance: A Solution to Data Scarcity in India
- How do Consumers Think About Usage-Based Auto Insurance? –A Survey Analysis from Taiwan
- Measurement of Risk Culture in General Insurance Companies in India – An Empirical Validation Using a Causal Model
Articles in the same Issue
- Frontmatter
- Featured Articles (Research Paper)
- Net Reserve Calculation for Whole Life Insurance Under Mean-Reverting Stochastic Interest Rate Models
- Integrating Remote Sensing Data in Crop Insurance: A Solution to Data Scarcity in India
- How do Consumers Think About Usage-Based Auto Insurance? –A Survey Analysis from Taiwan
- Measurement of Risk Culture in General Insurance Companies in India – An Empirical Validation Using a Causal Model
