Skip to main content

Behavioral bias and the demand for bicycle and flood insurance

Abstract

With data from an insurer that provides coverage for both a low probability, high consequence (LPHC) risk (the flood peril) and a high probability, low consequence (HPLC) risk (bicycle theft), we investigate behavioral bias in the demand for insurance. Our analysis provides evidence which is consistent with a preference for insurance for HPLC risks over LPHC risks: we find that many more policyholders purchase add-on coverage to their homeowner’s insurance to cover the risk of bicycle theft than to cover the risk of loss due to flooding. In addition, we find mixed evidence on whether policyholders’ insurance coverage decisions are responsive to changes in their risk exposure. We find a strong relationship between wealth and the demand for both types of coverage.

This is a preview of subscription content, access via your institution.

Notes

  1. The “Zoning-System for Floods, Tailback and Heavy Rain” (ZÜRS) was developed by the Association of German Insurance Companies (GDV). The system enables insurance companies to assign almost every building in Germany to one of four flood risk classes. Please refer to Section 2 for further details.

  2. For a general discussion on the interactive role of experiments and theory in analyzing insurance demand from a behavioral economics perspective, please refer to Richter et al. (2014).

  3. Insurance agents are typically compensated with commissions based on the amount of insurance sold. The agent thus has two objectives: to provide sound professional advice and to maximize compensation payments. The companies and products with the highest commissions are not necessarily the ones that best fit the needs of the policyholder. This might create an incentive for the agent not to recommend the optimal contract and insurance company (see Schiller 2011). Inderst (2011) discusses the importance of advice in the market for retail financial services. He states that advisors who are paid by commissions might not de-bias investors but rather “increase revenues through churning when customers already have a bias towards excessive trading.” However, for two reasons this should not be a problem in the case of natural hazards insurance. Premiums, at least in Germany, are not very high and thus premium based commissions will not be very high either. Consequently, agents might focus on enhancing their relationship with their customers by providing advice to customers free from consideration of their potential commission income. Secondly, commissions only give the agent an additional incentive to inform policyholders about their flood risk exposure and the option to purchase natural hazards insurance.

  4. Since we are primarily interested in individuals’ decisions, it is advantageous that our data are from a single insurer. Our analyses are free of confounding effects that would be introduced were our data from multiple insurers.

  5. Earthquakes hardly ever occur in the part of Germany where the insurer writes coverage.

  6. In 2009, 80 % of the households in Germany had at least one bicycle (see Federal Statistical Office 2009).

  7. Typically, retail property and casualty insurance contracts in Germany are automatically renewed annually if the policyholder does not lapse. As a consequence, the contracts behave similar to multi-year contracts with annual (or semi-annual) premium payment. Through direct debit, the policyholder allows the insurer to withdraw the premium from the policyholder’s bank account every time it is due without any further action required. If the policyholder opts to pay via wire transfer, she has to transfer money from her bank account to the insurer’s bank account after receiving an invoice from the insurer.

  8. We also ran our analysis including outliers and apartments. Results did not significantly change.

  9. Policies are one-year in length and renewable.

  10. P = a*(sum insured). The factor “a” in the equation is a fixed constant.

  11. The insurer’s portfolio of in force contracts in June 2010 consists of contracts from different generations. For contracts sold before 1997, policyholders could not purchase additional coverage for natural hazard risks. The insurance company completely changed the contract design for the base contracts in 2009. In doing so they also changed the pricing scheme for the natural hazards coverage. For these policies, the flood risk classification and the building type are used for risk classification (i.e., premium is higher for policyholders living in a house than for those living in an apartment and higher for those living in flood risk class 3 than for those living in flood risk class 1). This completely changes the expected effects of risk exposure on the demand for natural hazards coverage. Thus, we restrict our analysis to contracts initially sold between 1997 and 2008. When the contract design was modified in 2009, customers holding an old contract were offered the possibility to switch to the new conditions, but they did not have to. Only a few of them did. Excluding from our analysis the policyholders that switched to the new conditions might induce a sample selection bias. Our analysis suggests this potential bias does not remarkably change our results.

  12. We do not have information on the claims for the different flood risk areas, so we cannot calculate whether premiums are adequate. But the insurance company affirmed that the premium per unit of coverage was lower for the contracts in flood risk classes 2 and 3 and they were losing money with these contracts. Hence, there seems to be some cross-subsidization (i.e., those in low risk areas overpay relative to their risk and those in high risk areas underpay relative to their risk).

  13. A premium increase due to having more insured is not the same as a price increase, which is a greater cost per unit of insured value. While we do not have the true price of coverage, but rather just the premium that was paid, we expect the premium to proxy price. A flat cost per dollar of coverage coupled with a decreasing probability of loss at higher monetary amounts suggests that the insurance loading, which is the true cost of insurance, increases with the total value of premium paid.

  14. Of course, an effect of the sales channel on insurance demand could occur if policyholders who purchase through an agent differ from those who do not (for instance, regarding wealth or risk aversion). We are not aware of many studies analyzing the characteristics and behavior of policyholders between different sales channels. Bauer et al. (2002) do not find a big impact of policyholders’ socio-demographic characteristics on their propensity to purchase insurance over the internet. One exception is that individuals purchasing insurance online tend to have slightly higher incomes. Policyholders in our dataset do not remarkably differ with regard to observable characteristics between sales channels. Also, the premium policyholders have to pay does not remarkably differ between sales channels.

  15. For a more detailed description of the bivariate probit model, see Greene (1996).

  16. The preference for bicycle theft coverage over natural hazards coverage is the same if we consider only individuals living in flood risk class 2 and 3.

  17. We thank an anonymous referee for making this observation.

  18. We have no evidence that bicycle theft coverage in risk class 3 has a higher loading compared to the other two risk classes.

References

  • Agarwal, S., Driscoll, J. C., Gabaix, X., & Laibson, D. (2009). The age of reason: financial decisions over the life-cycle with implications for regulation. Brookings Papers on Economic Activity, 2009(Fall), 51–101.

    Article  Google Scholar 

  • Bauer, H. H., Sauer, N. E., & Brugger, N. (2002). Die Akzeptanz von Versicherungsdienstleistungen im Internet. Zeitschrift für die gesamte Versicherungswissenschaft, 91(3), 329–363.

  • Botzen, W., & van den Bergh, J. (2012). Risk attitudes to low-probability climate change risks: WTP for flood insurance. Journal of Economic Behavior and Organization, 82(1), 151–166.

    Article  Google Scholar 

  • Botzen, W., Aerts, J., & van den Bergh, J. (2009). Dependence of flood risk perceptions on socioeconomic and objective risk factors. Water Resources Research, 45(10), W 10440.

    Article  Google Scholar 

  • Browne, M. J., & Hoyt, R. E. (2000). The demand for flood insurance: empirical evidence. Journal of Risk and Uncertainty, 20(3), 291–306.

    Article  Google Scholar 

  • Cummins, J. D., & Doherty, N. A. (2006). The economics of insurance intermediaries. Journal of Risk and Insurance, 73(3), 359–396.

    Article  Google Scholar 

  • DellaVigna, S. (2009). Psychology and economics: evidence from the field. Journal of Economic Literature, 47(2), 315–72.

    Article  Google Scholar 

  • Federal Statistical Office. (2009). Deutschland – Land und Leute. Wiesbaden: Statistisches Bundesamt.

    Google Scholar 

  • Friedl, A., Lima de Miranda, K., & Schmidt, U. (2014). Insurance demand and social comparison: an experimental analysis. Journal of Risk and Uncertainty, 48(2), 97–109.

    Article  Google Scholar 

  • Ganderton, P. T., Brookshire, D. S., McKee, M., Stewart, S., & Thurston, H. (2000). Buying insurance for disaster-type risks: experimental evidence. Journal of Risk and Uncertainty, 20(2), 271–289.

    Article  Google Scholar 

  • GDV. (2014). ZÜRS. Available online at: http://www.gdv.de/2008/08/geo-informationssystem-zuers-geo-zonierungssystem-fuer-ueberschwemmungsrisiko-und-einschaetzung-von-umweltrisiken/.

  • Greene, W. H. (1996). Marginal effects in the bivariate probit model. SSRN Working Paper 1293106.

  • Hsee, C. K., & Kunreuther, H. (2000). The affection effect in insurance decisions. Journal of Risk and Uncertainty, 20(2), 141–159.

    Article  Google Scholar 

  • Huysentruyt, M., & Read, D. (2010). How do people value extended warranties? Evidence from two field surveys. Journal of Risk and Uncertainty, 40(3), 197–218.

    Article  Google Scholar 

  • Inderst, R. (2011). Consumer protection and the role of advice in the market for retail financial services. Journal of Institutional and Theoretical Economics, 167(1), 4–21.

    Article  Google Scholar 

  • Johnson, E. J., Hershey, J., Meszaros, J., & Kunreuther, H. (1993). Framing, probability distortions, and insurance decisions. Journal of Risk and Uncertainty, 7(1), 35–51.

    Article  Google Scholar 

  • Kriesel, W., & Landry, C. (2004). Participation in the national flood insurance program: an empirical analysis for coastal properties. Journal of Risk and Insurance, 71(3), 405–420.

    Article  Google Scholar 

  • Kunreuther, H. (1996). Mitigating disaster losses through insurance. Journal of Risk and Uncertainty, 12(2–3), 171–187.

    Article  Google Scholar 

  • Kunreuther, H., & Michel-Kerjan, E. (2009). At war with the weather. Managing large scale risks in a new era of catastrophes. Cambridge: The MIT Press.

    Book  Google Scholar 

  • Kunreuther, H., & Pauly, M. (2004). Neglecting disaster: why don’t people insure against large losses? Journal of Risk and Uncertainty, 28(1), 5–21.

    Article  Google Scholar 

  • Kunreuther, H., & Pauly, M. (2005). Insurance decision-making and market behavior. Foundations and Trends in Microeconomics, 1(2), 63–127.

    Article  Google Scholar 

  • Kunreuther, H., Novemsky, N., & Kahneman, D. (2001). Making low probabilities useful. Journal of Risk and Uncertainty, 23(2), 103–120.

    Article  Google Scholar 

  • Landry, C. E., & Jahan-Parvar, M. R. (2011). Flood insurance coverage in the coastal zone. Journal of Risk and Insurance, 78(2), 361–388.

    Article  Google Scholar 

  • Laury, S. K., McInnes, M. M., & Swarthout, J. T. (2009). Insurance decisions for low-probability losses. Journal of Risk and Uncertainty, 39(1), 17–44.

    Article  Google Scholar 

  • McClelland, G. H., Schulze, W. D., & Coursey, D. L. (1993). Insurance for low-probability hazards: a bimodal response to unlikely events. Journal of Risk and Uncertainty, 7(1), 95–116.

    Article  Google Scholar 

  • Michel-Kerjan, E., & Kousky, C. (2010). Come rain or shine: evidence on flood insurance purchases in Florida. Journal of Risk and Insurance, 77(2), 369–397.

    Article  Google Scholar 

  • Nichols, A., & Schaffer, M. E. (2007). Clustered standard errors in Stata. Available online at http://repec.org/usug2007/crse.pdf.

  • Pfeifer, C. (2011). Risk aversion and sorting into public sector employment. German Economic Review, 12(1), 85–99.

    Google Scholar 

  • Pohl, R. (2004). Cognitive illusions: A handbook on fallacies and biases in thinking, judgment and memory. New York: Psychology Press.

    Google Scholar 

  • Richter, A., Schiller, J., & Schlesinger, H. (2014). Behavioral insurance: theory and experiments. Journal of Risk and Uncertainty, 48(2), 85–96.

    Article  Google Scholar 

  • Schiller, J. (2011). Ökonomische aspekte der entlohnung und regulierung unabhängiger versicherungsvermittler. Zeitschrift für die gesamte Versicherungswissenschaft, 100(1), 113–130.

    Article  Google Scholar 

  • Schoemaker, P., & Kunreuther, H. (1979). An experimental study of insurance decisions. Journal of Risk and Insurance, 46(4), 603–618.

    Article  Google Scholar 

  • Schwarcz, D. (2009). Insurance demand anomalies and regulation. Journal of Consumer Affairs, 44(3), 557–577.

    Article  Google Scholar 

  • Shafran, A. P. (2011). Self-protection against repeated low probability risks. Journal of Risk and Uncertainty, 42(3), 263–285.

    Article  Google Scholar 

  • Siegrist, M., & Gutscher, H. (2006). Flooding risks: a comparison of lay people’s perceptions and experts’ assessments in Switzerland. Risk Analysis, 26(4), 971–979.

    Article  Google Scholar 

  • Slovic, P., Fischhoff, B., Lichtenstein, S., Corrigan, B., & Combs, B. (1977). Preference for insuring against probable small losses: insurance implications. Journal of Risk and Insurance, 44(2), 237–258.

    Article  Google Scholar 

  • Sutter, D., & Poitras, M. (2010). Do people respond to low probability risks? Evidence from tornado risk and manufactured homes. Journal of Risk and Uncertainty, 40(2), 181–196.

    Article  Google Scholar 

  • Sydnor, J. (2010). (Over)insuring modest risks. American Economic Journal: Applied Economics, 2(4), 177–199.

    Google Scholar 

  • Urbany, J. E., Schmit, J. T., & Butler, D. D. (1989). Insurance decisions (or the lack thereof) for low probability events. Advances in Consumer Research, 16, 535–541.

    Article  Google Scholar 

  • Viscusi, W. K., & Zeckhauser, R. J. (2006). National survey evidence on disasters and relief: risk beliefs, self-interest, and compassion. Journal of Risk and Uncertainty, 33(1), 13–36.

    Article  Google Scholar 

  • Zhou-Richter, T., Browne, M. J., & Gründl, H. (2010). Don’t they care? Or, are they just unaware? Risk perception and the demand for long-term care insurance. Journal of Risk and Insurance, 77(4), 715–747.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Mark J. Browne.

Additional information

The authors are thankful for the helpful comments from an anonymous reviewer as well as from the participants at the 2011 ARIA meeting, the Munich Behavioral Insurance Workshop 2011, the 2012 EGRIE meeting and the 2012 annual meeting of the Deutscher Verein für Versicherungswissenschaft. All remaining errors are the authors’.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Browne, M.J., Knoller, C. & Richter, A. Behavioral bias and the demand for bicycle and flood insurance. J Risk Uncertain 50, 141–160 (2015). https://doi.org/10.1007/s11166-015-9212-9

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11166-015-9212-9

Keywords

  • Decision making under risk
  • Risk assessment
  • Insurance demand

JEL Classifications

  • D81
  • D83
  • D84