個體經濟學三篇論文

Abstract

本論文涵蓋三項個體經濟學議題：不完全資訊中國餐館賽局、獨裁者的競租資源分配決策，以及公平排隊問題中瓦拉斯分配規則的公設刻畫。

第一篇文章題目為「附加排隊規則的中國餐館賽局中的阻塞現象和資訊品質探討」，本篇提出一個不完全資訊非合作賽局，並探討玩家的貝氏學習如何影響其決定。本篇文章探討中國餐館排隊問題中，資訊壅塞與訊號品質的抵換關係。餐館排隊的玩家擁有不同的事前資訊，且玩家可以選擇立刻進餐廳或等待並觀察其他玩家的訊號後才進餐廳。當多數玩家同時選擇進餐廳或者多數選擇延遲都會導致壅塞。理想上，我們預期玩家們會釋出餐廳的資訊以增進社會決策效率，但我們證明：事前擁有高品質資訊的玩家沒有動機釋出其資訊給事前擁有低品質資訊的玩家，而社會效率的排隊結果則不是均衡結果。

第二篇文章題目為「生產、衝突與獨裁者的最適偏袒行為」，本篇提出一個包括一名政策制定者及一群玩家的競賽架構，探討政策制定者的偏袒如何影響玩家投入的努力水準。本篇文章建立一個兩階段、多項獎勵競賽，其中玩家於第二階段選擇其生產性及浪費性資源投入數量，在第一階段，領導人透過調整相對權重以扭曲玩家的浪費性投入贏得獎勵的機率。我們發現當獎勵數量較多時，領導人有較強誘因鼓勵玩家投入浪費性資源，若玩家數量很多，領導人也會鼓勵玩家投入浪費性資源。我們發現在多數情況，領導人都會鼓勵玩家多投入浪費性資源，這解釋了常見的「錫鍋式」專制統治下的貪腐行為，以及公司治理中的馬屁文化。

第三篇文章題目為「具備初始稟賦之排隊問題的瓦拉斯分配規則」，本篇以合作賽局途徑定義一個解規則，並進行公設刻畫。本篇文章探討在具有初始稟賦設定的公平排隊問題下之瓦拉斯均衡概念。在本文模型中，每位決策者均持有初始入場排隊位置，並可透過金錢補償，與其他決策者交換位置到排序較前的位置。本文定義了一種新的分配規則，其結果與瓦拉斯均衡對應的分配結果相同，且此分配規則可以被無嫉妒公理唯一刻畫。

This dissertation consists of three microeconomics topics: the Chinese restaurant game with incomplete information, the dictator's rent allocation decision, and an axiomatic characterization of Walras allocation rules in the fair queuing problems.

Chapter 2 is titled ``Congestion and Information Quality in Chinese Restaurant Games with Priority Rules.'' This chapter proposes a non-cooperative game with incomplete information and explores how players' Bayesian learning affects their decisions and final social distribution. We explore the trade-off between congestion and signal quality in a Chinese restaurant queuing situation. Players are heterogeneous on the information of the true distribution; players choose entry immediately or wait and see how information goes. Too many players choosing to delay decisions will cause congestion. Ideally, we expect players holding critical information to reveal it to the public to improve social efficiency. We show that players with high-quality signals have no incentive to disclose their information to players with low-quality signals, and the efficient queuing outcome is not achieved.

Chapter 3 is titled ``Production, Conflict, and Dictator's Optimal Favoritism.'' This chapter proposes a competition structure including a policy maker and a group of players to explore how the policy maker's favoritism affects the level of effort invested by the players. We consider a two-stage multi-prize contest in which players choose a productive-wasteful effort allocation in Stage 2; in Stage 1, a ruler chooses a weight to enlarge the players' winning probability of investing unproductive effort. We find that the productive effort investment level is almost zero whenever the number of prizes is small. When the number of prizes is high, the ruler has a stronger incentive to encourage the players to invest in wasteful efforts. The ruler also encourages the players to invest in wasteful efforts as the player group size increases. We find that in most situations the ruler encourages the contestants to invest in wasteful efforts, which explains the tinpot-like autocrats' bribery behavior and company managers' favoritism.

Chapter 4 is ``The Walrasian Rule in Fair Queuing Problems With an Initial Order.'' This chapter defines a new allocation rule and characterizes the allocation rule by an axiomatic approach. We focus on analyzing the Walrasian equilibrium in the context of fair queuing problems with initial order. Each agent, occupying a particular queuing position, seeks to obtain a former position and is permitted to exchange positions with other agents by paying a monetary compensation. We define an allocation rule in which the outcome coincides with the Walrasian equilibrium and show that the rule is characterized by the envy-free axiom.