Data and Code for: Random Utility with Unobservable Alternatives

The random utility model (RUM), a cornerstone in economics, is typically<br>studied under the assumption that choice frequencies of all alternatives<br>are observable. In practice, however, some alternatives have unobservable<br>choice frequencies and are commonly aggregated into a single category called<br>an outside option. We study RUM in such environments and derive a finite,<br>nonredundant system of inequality constraints on observed choice frequencies<br>that characterizes RU-rationalizability. We show that the conventional<br>practice of aggregating unobserved alternatives can miss key information<br>leading to incorrect conclusions such as that observed choices are rationalizable,<br>even when no RUM is consistent with them.<br>

随机效用模型（Random Utility Model，RUM）作为经济学的基石性理论，通常在所有备选方案的选择频率均可被观测的假设下开展研究。然而在实际场景中，部分备选方案的选择频率无法被观测，这类方案通常会被归集为单一类别，即外部选项（outside option）。本文针对此类场景下的RUM展开研究，推导出一组有限且无冗余的观测选择频率不等式约束系统，该系统可用于刻画RUM可理性化。研究表明，将未观测备选方案进行归总这一常规做法可能会遗漏关键信息，进而得出错误结论——例如，即便不存在与观测选择相符的RUM，仍可能判定观测选择具备可理性化性。