We describe a new class of mechanisms—referred to as booster draft mechanisms—for allocating \(n\times m\) indivisible objects among n individuals with responsive preferences. Specifically, n individuals are assigned m objects each, from m different sets of size n—called boosters—and simply receive their highest ranked object, among the remaining objects in each booster, when it is their turn to choose. We contrast booster draft mechanisms with two well-known alternatives: serial dictatorship and draft mechanisms. Our results suggest that booster draft mechanisms pose a viable alternative for resolving multi-object assignment problems.
