Predicting insurance demand from risk attitudes

Abstract

Can measured risk attitudes and associated structural models predict insurance demand? In an experiment (n = 1,730), we elicit measures of utility curvature, probability weighting, loss aversion, and preference for certainty and use them to parameterize seventeen common structural models (e.g., expected utility, cumulative prospect theory). Subjects also make twelve insurance choices over different loss probabilities and prices. The insurance choices show coherence and some correlation with various risk-attitude measures. Yet all the structural models predict insurance poorly, often less accurately than random predictions. Simpler prediction heuristics show more promise for predicting insurance choices across different conditions.

Insurance Demand Experiments: Comparing Crowdworking to the Lab

“…These similarities are in line with the study by Goodman et al (2013) that is largely supportive of crowdworking populations as a subject pool for behavioral experiments. Goodman et al highlight that such online participants tend to produce "reliable results" that are broadly consistent with average behavior observed in 1 Two companion papers use data from this experiment, Jaspersen et al (2020) and Jaspersen et al (2022). Both papers analyze research questions distinct from the one analyzed here, and simply combine the online and in-person subjects in their analyses.…”

“…Even when considering all available options (i.e., 0%-100% in 1% increments), a subject still might choose a corner solution as the best available option. Indeed, many established models of choice would predict these corner solutions for some of the commonly assumed combinations of preference parameters (Jaspersen et al, 2022). Thus the filter used in column (4) of Table 7 does not adequately categorize subjects based on their likely choice sets.…”

“…Insurance demand is a popular focus in experimental economics due to the relative simplicity of insurance decisions and the importance of insurance markets for economic welfare (for a review, see Jaspersen, 2016). Research evaluates many important and varied topics, including insurance demand in various situations (e.g., Bajtelsmit et al, 2015;Laury et al, 2009;Zimmer et al, 2018), behavioral patterns such as reactions to framing (e.g., Powell & Ansic, 1997), the tendency to make decisions in isolation (Gottlieb & Mitchell, 2020), and the predictive validity of entire theories of decision-making (Jaspersen et al, 2022). However, experimental research on insurance demand does not come without complications.…”

Probability weighting and insurance demand in a unified framework

“…Several alternatives have been proposed to explain underinsurance of LPHI risks. One suggestion is diminishing marginal sensitivity (Schmidt 2016), though evidence of this behavioral assumption is mixed (e.g., Chung et al 2019;Harbaugh et al 2010;Jaspersen et al 2021). Friedl et al (2014) argue that social comparison can make insurance against highly correlated risks less attractive and present evidence from a laboratory experiment.…”

“…EU has often been criticized for its prediction of partial coverage at actuarially unfair premiums, as it tends to conflict with choices observed in the laboratory and the field (e.g., Jaspersen et al 2021;Shapira and Venezia 2008;Sydnor 2010). Probability weighting allows us to explain the evidence.…”