This paper proposes an origin-based air travel demand model that assumes saturation at the origin level and explicitly accounts for substitutability between destinations. We simultaneously integrate demand generation and allocation by means of a multilevel aggregate nested logit formulation that covers the choices of whether or not to travel by air, where to travel (destination), and how to travel (itinerary). Two specifications are proposed to reflect systematic differences between lengths of haul and the bootstrap is applied to jointly address endogeneity issues and data missingness. The validity of the proposed approach is tested over the entire network of outbound air trips from Italy in 2018.

本文提出了一种基于原点的航空旅行需求模型，该模型在原点水平假设饱和，并明确说明了目标之间的可替代性。我们通过多层嵌套的logit公式同时集成需求生成和分配，该公式涵盖了是否乘飞机旅行，在哪里旅行（目的地）以及如何旅行（旅行）的选择。提出了两个规范以反映运输长度之间的系统差异，并且将引导程序用于共同解决内生性问题和数据丢失。该方法的有效性在2018年从意大利出发的整个出境航空网络中进行了测试。