panel data error component model. We derive moment conditions to estimate the parameters of the higher order spatial …

distribution is *entirely* known. We relax this assumption in the context of a regression error component model and develop an …

implications of discontinuous preferences on willingness to pay estimates. Using a multinomial error component logit model, we … on willingness to pay estimates. Using a multinomial error component logit model, we explore differences in scale and …-fit in discrete choice models and biased WTP estimates. Using multinomial error component model specifications we combine …

form of random e®ects and ¯rst order serial correlation in the error component model. The fact that the presence of random …

created data set. Each generic entrant is associated with an average 1% increase in the branded price. The one-way error … component model accounts for intermolecular competition, market segmentation, and endogeneity of entry and finds branded prices …