“…The segmentation of electricity customers is coincidental with the original purpose of CIC surveys, which was to establish the value of reliability for customer classes in utilities.6 There are a few value of lost load studies using different elicitation strategies, including discrete choice experiments (for instance,Layton and Moeltner (2005)). Although discrete choice experiments are well-suited for estimating customers' preferences for electric service reliability, they are not well-suited for assessing the costs of power interruptions, especially for WLD power interruptions, because 1) individuals' preferences for resilient electric services are uncertain and incomplete, so it is difficult to use a single cardinal utility function to express their preferences for complex and unfamiliar alternatives; 2) studies need to abstract away significantly from what will actually happen during a power interruption because the method requires many repeated choices from respondents, taking time away from helping respondents understand the time dynamics of lost electricity-dependent services; 3) respondents' value of resilient electric services is determined by many factors, and the differences among people will be washed out by aggregating over individuals; and, 4) testing axioms of probabilistic discrete choice models, for example stochastic transitivity and quadruple condition, requires many repeated pairwise comparisons(Boxall et al, 1996;Davis-Stober, 2009;De La Maza et al, 2018;Baik et al, 2019;Baik et al, 2020).…”