This article analyzes the determinants of tourists’ expenditure behavior through the joint adoption of two microeconometric approaches, namely, the Stochastic Frontier (SF) and the Multiple Discrete Continuous Extreme Value (MDCEV) model. Despite the attention that analysts have dedicated to consumers’ expenditure behavior in recent years, several limitations concerning the role of budget and the phases of money allocation are still affecting the literature on the topic. In this study, the SF is employed to identify the unobserved individual maximum level of spending allotted for a trip. Once estimated, the frontier is included as a travel budget in the utility-maximizing framework of a MDCEV model. The MDCEV approach allows to simultaneously assess two moments characterizing spending decisions. That is, the decision to allocate the budget to several expenditure categories and the decision concerning the amount to allocate to each category. Data adopted for this research is collected by the Swiss Statistical Office (UST) through a representative Household Budget Survey (Haushaltsbudgeterhebung). Data related to Swiss residents’ leisure travel expenditures are investigated; more specifically, the expenditure categories considered in the analysis are Accommodation, Transportation, Shopping and Food & Beverage.

中文翻译：

---

了解游客的支出模式：多重离散-连续选择框架内的随机前沿方法

本文通过联合采用两种微观计量方法，即随机前沿（SF）和多元离散连续极值（MDCEV）模型，分析游客消费行为的决定因素。尽管分析家近年来一直关注消费者的支出行为，但关于预算作用和货币分配阶段的一些限制仍在影响有关该主题的文献。在这项研究中，SF 被用来确定未观察到的个人为一次旅行分配的最大支出水平。估计后，边界作为旅行预算包含在 MDCEV 模型的效用最大化框架中。MDCEV 方法允许同时评估表征支出决策的两个时刻。那是，将预算分配给几个支出类别的决定以及关于分配给每个类别的金额的决定。本研究采用的数据由瑞士统计局 (UST) 通过具有代表性的家庭预算调查 (Haushaltsbudgeterhebung) 收集。调查与瑞士居民休闲旅游支出相关的数据；更具体地说，分析中考虑的支出类别是住宿、交通、购物和餐饮。