Time value estimation in sacrifice model with disaggregate data

Abstract

Two new methods to estimate time values are suggested. The methods can estimate time values in sacrifice model with disaggregate data. With the methods, it becomes possible to treat data with no aggregation in which unit prices for same origins and destinations vary according to the lot size or other individual properties of each sample, which was impossible with the existing method for sacrifice models. In addition, the methods are faster than logit model based methods, and are more flexible at probability functions assumed in sacrifice models.