Optimal advantage allocation in repeated contests
研究两阶段Tullock竞赛中，设计者如何根据历史表现分配生产力优势以最大化总努力，发现不对称程度决定最优策略：中等不对称时奖励首轮胜者，高度不对称时奖励首轮败者。
We study a two-stage Tullock contest with asymmetric players where a designer allocates a productivity-enhancing advantage based on past performance. The advantage’s magnitude is exogenously fixed, so the designer’s choice is purely allocative: who receives it. We characterize the optimal rule and show a sharp threshold in asymmetry. With moderate heterogeneity, favouring the stage-1 winner strengthens incentives and maximises total effort. With sufficiently large heterogeneity, favouring the stage-1 loser restores competitiveness in stage 2 and raises aggregate effort. The results clarify optimal contest design under limited favouritism instruments.<br/><br/>