| The objective of this paper is to understand what drives consumers to buy extended warranties and pay high premia for them. We primarily focus on the role of risk preferences and disentangle and study their relative importance. Empirical and behavioral research on insurance is at odds with whether diminishing returns (curvature of the utility function), or loss aversion and non-linear probability weighting lead to the observed consumer behavior. This is primarily due to the inability of standard choice data to separate curvature of the utility function, loss aversion and non-linear probability weights, and the need to rely on strong parametric assumptions. We design two conjoint studies (consistent with simultaneous and sequential decision making) with choices over washing machines (with and without extended warranty) where failure probabilities and repair costs are given to subjects. Using stated choice data from the survey, consumer preferences, degree of curvature, loss aversion and probability weights can be non-parametrically identified. We find that loss aversion is significantly more important than curvature and probability weights in explaining extended warranty choices. These findings are robust to different specifications of the utility function and risk preferences. Importantly, failure to decompose risk averse behavior into that arising from curvature, loss aversion and probability weighting leads to lower washer and warranty prices, and under predicts the rate of change of warranty prices with varying repair costs. We test theory on complementary good pricing and find that estimates from the sequential choice survey rationalize the high premium consumers pay for extended warranties. Finally, forcing different retailers to sell washers and extended warranties increases (decreases) the washer (warranty) price, and makes consumers worse-off. A 5% discount on the washer price makes consumers indifferent to the proposed policy intervention. |
