本文認為有兩種類型的代理尋租模式。一組代理有兩種類型的租金估值。選擇一個代理商的基礎(chǔ)上競賽成功函數(shù)規(guī)模報酬不變的贏家。我們首先使純策略的納什均衡特性化。然后研究如何平衡工作比，獎品功耗比，總的努力水平響應，當代理變化的異質(zhì)性或代理時，成分變化�；�于這些均衡分析中，我們調(diào)查是否增加一個初級階段可以享受比賽的組織者誰的目的是最大限度地從各個代理商預計總工作量。最佳的競賽結(jié)構(gòu)已充分的特點，在不同程度的異質(zhì)性和代理商組成。

　　焦倩：新加坡國立大學經(jīng)濟學博士，現(xiàn)任中山大學嶺南學院講師。主要研究領(lǐng)域為應用微觀理論，競爭與拍賣，和產(chǎn)業(yè)組織理論。學術(shù)論文發(fā)表于Public Choice, Pacific Economic Review等經(jīng)濟學國際期刊上。

　　原文：This article considers a rent-seeking model with two types of agents. A group of agents has two types of valuation of the rent. One of the agents is selected as the winner based on contest success function with constant returns to scale. We first characterize the pure strategy Nash equilibrium. Then examine how the equilibrium effort ratio, the prize dissipation ratios, and the total effort level respond when the heterogeneity of agents changes or when agents composition changes. Based on these equilibrium analyses, we investigate whether adding a preliminary stage can benefit a contest organizer who aims at maximizing total expected effort from all agents. The optimal contest structure has been fully characterized at different level of heterogeneity and agents composition.