A Non-Compensatory Random Utility Choice Model based on Choquet Integral

摘要: We present a random utility maximisation (RUM) based discrete choice model to simultaneously consider three behavioural aspects of the decision-maker - i) evaluation each attribute (e.g., constant marginal beyond thresholds), ii) aggregation attributes in systematic part indirect utility, and iii) flexible substitution patterns between alternatives. Corresponding aspect, we use fuzzy membership functions capture how evaluates range attributes, replace weighted-sum function with Choquet integral (CI) non-linear interactions adopt multinomial probit (MNP) probability kernel represent estimate proposed using constrained maximum likelihood estimator. A comprehensive Monte Carlo study is performed establish statistical properties estimator demonstrate superiority over traditional MNP weighted sum terms goodness fit recovery effects. Note that CI-based specification provides complementarity pairs coupled their individual importance ranking as by-product estimation. This information could potentially help policymakers making policies improve preference level for an alternative. The advantages considering cut-off non-compensatory behaviour are illustrated empirical application understand New Yorkers' preferences towards mobility-on-demand services.