Valuation of Self-Insurance and Self-Protection Under Ambiguity: Experimental Evidence

Abstract

This experimental study, first, compares the individual valuations of two risk reduction mechanisms: self-insurance and self-protection. Second, it investigates these valuations when the loss amount is ambiguous, and compares these values with valuations when loss amounts are known. Results confirm that there exists no "framing effect" due to the two risk reduction mechanisms. Ambiguity in the loss amount has a weak impact on the valuation, and using different representations of ambiguity does not change the valuation. Moreover, the mean ratios of ambiguous to risky bids are greater than one for low loss amounts indicating ambiguity aversion. These ratios are not significantly different from one for high loss amounts regardless of the probability of loss levels. Finally, 28 percent of the sample behaved consistent with the predictions of "anchoring and adjustment", while only 6 percent supported the "maximin" predictions.

Keywords: self-insurance, self-protection, risk, uncertainty

JEL Classification: C91, D81