We propose a novel dynamic mixture vector autoregressive (VAR) model in which timevarying mixture weights are driven by the predictive likelihood score. Intuitively, the state weight of the k-th component VAR model in the subsequent period is increased if the current observation is more likely...

In the home-care services industry, caregivers drive to visit patients scattered in a district and deliver various care services at the patient's home. We use a unique data set, with a standard panel structure, recording the hours of service and the exact number of miles traveled by each...

We extend the literature structurally estimating social preferences by accounting for the desire to adhere to social norms. Our representative agent is strongly motivated by norms and failing to account for this causes us to overestimate how much agents care about helping those who are worse...