Assessing the Value of Museums with a Combined Discrete Choice / Count Data Model

We develop a model for museum visits and estimate it on a large sample of holders of a museum card that provides free access to all museums. Our model distinguishes two stages, referring to the determination of the number of trips and destination choice and can deal with an effect of income on destination choice. We use a mixed logit model for destination choice and a count data model for explaining the number of trips. The model implies that welfare analyses should not only take into account the change in the logsum variables, but also changes in the number of trips. In our empirical application we find substantial local interest effects that cause a correlation between the attractiveness of a museum and its distance to the residential location of the visitor. The ranking of museums on the basis of their estimated attractiveness differs substantially from that based on the change in income that would be needed to compensate for its disappearance.