This study proposes a new approach to utilize information from existing choice experiments to predict policy outcomes for a transfer setting. Recognizing the difficulties from pooling raw data from experiments with different designs and sub-populations we first re-estimate all underlying Random...