{"title":"Moderate Utility","authors":"Junnan He, Paulo Natenzon","doi":"10.1257/aeri.20230085","DOIUrl":null,"url":null,"abstract":"Hotelling’s and Salop’s spatial competition models, as well as nested logit, covariance probit, elimination-by-aspects, and several other well-known discrete choice models, belong to the class of moderate utility models, where binary choices are a function of the ratio between utility difference and a product differentiation index satisfying the properties of a distance metric. We provide a behavioral foundation for this class of models. Our main result establishes that moderate utility has a single, directly testable implication: choice probabilities are moderately transitive. We use our characterization to show how the model achieves a useful compromise between explanatory power and predictive power. (JEL C25, D11, D91)","PeriodicalId":504102,"journal":{"name":"American Economic Review: Insights","volume":"3 12","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"American Economic Review: Insights","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1257/aeri.20230085","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
Hotelling’s and Salop’s spatial competition models, as well as nested logit, covariance probit, elimination-by-aspects, and several other well-known discrete choice models, belong to the class of moderate utility models, where binary choices are a function of the ratio between utility difference and a product differentiation index satisfying the properties of a distance metric. We provide a behavioral foundation for this class of models. Our main result establishes that moderate utility has a single, directly testable implication: choice probabilities are moderately transitive. We use our characterization to show how the model achieves a useful compromise between explanatory power and predictive power. (JEL C25, D11, D91)