Resource Allocation Under Exogenous Information Regimes in Multi-Battle Team Contests

Abstract

A central leader allocates a budget across multiple fronts before agents choose effort. We study a $T$-front team contest with heterogeneous players under three exogenous information regimes: full, partial, and no disclosure. Under full disclosure, every interior allocation-game Nash equilibrium features battle-by-battle spending parity: with equal budgets, leaders match each other's spending within each battle, although the common spending level may vary with the realized type pair. Under partial and no disclosure, we characterize neutral-opponent benchmark best responses. Partial disclosure yields a prior-only tilt direction: toward strong-player battles when strong types are in the majority and toward weak-player battles otherwise. Under no disclosure, Bayesian effort pooling can make weak-player battles have higher marginal product than strong-player battles; the benchmark first-order driver then points toward weak-link protection over a wide region and becomes the neutral-opponent benchmark best response under the stated concavity and interior-effort conditions. We call this mechanism Bayesian productivity reversal. Three-battle strategic-response checks suggest that these directions are robust away from the relevant benchmark thresholds.