Efficiency and strategy-proofness in multi-unit object allocation problems with non-quasi-linear preferences: A positive result