2023-2024
4月28日(金): 16:20-17:50 富士見校舎4階404教室
発表者: 三浦慎太郎 (神奈川大学経済学部)
タイトル: Value of Middle Managers (joint with Takakazu Honryo)
概要:
Why do modern organizations have middle managers? We address this question by analyzing the following hierarchical communication model. A subordinate has private information regarding the profitability of a new project and sends a costly signal about it (e.g., technical reports). Although the boss has to decide whether to approve the new project, she cannot freely observe the signal. Instead, the boss has two channels to learn it: (i) indirect communication with the biased (middle) manager who directly observes the signal, and (ii) direct observation of the signal by conducting costly investigations. We show that commitments to the investigation (i.e., organizations without managers) are always suboptimal for the boss even if the investigation cost is sufficiently small and the manager is sufficiently biased, providing a rationale for hierarchical organizations. By decomposing the gain from having managers, we clarify that indirect communication and costly investigation are complements if and only if the manager's bias is sufficiently large. Furthermore, we characterize the optimal direction of the manager's bias for the boss.
5月26日(金): 16:20-17:50 富士見校舎4階404教室
発表者: 濱田高彰 (東洋大学経済学部)
タイトル: Conspicuous Consumption and Excuse Device
概要:
We focus on an individual incentive to escape from a conspicuous consumption race. We use a simple signaling model of conspicuous consumption and introduce a good referred to as an “excuse good” which some consumers prefer intrinsically (called type αH), and others are not interested in at all (called type αL). Under asymmetric information between consumers and observers on the preference for the excuse good, the study shows that there exists a partial pooling equilibrium such that some consumers with type αL purchase the excuse good, even though type αL gains no direct benefit from the good. When high-income groups of type αH purchase the excuse good, observers’ perceived income for buyers of the excuse good is high. Then, if the price of the excuse good is sufficiently low, low- and middle-income groups of type αL are willing to buy the excuse good to pretend to be type αH. The study also shows a monopoly firm has an incentive to set a sufficiently low price to capture the demand from type αL. Moreover, the excuse good positively affects those who remain in the conspicuous consumption race: the competition among remaining consumers is mitigated after some consumers with type αL switch from the conspicuous good to the excuse good. The main driving mechanism for our results is the information asymmetry on the preference for the excuse good.
7月28日(金): 16:00-18:10 富士見校舎3階301教室
Historical Data Analysis Workshop
16:00-17:00
発表者1: 井上達樹 (明治大学商学部)
タイトル: The far-reaching effects of bombing on fertility in mid-20th century Japan (joint with Erika Igarashi)
概要:
Fertility changes after wars and conflicts have been observed worldwide. This study examines whether regional war damage affects postwar fertility even in areas that were not directly affected but were close to the damaged areas. In order to accomplish this, we exploit the air-raid experience in Japan during World War II. Using the municipality-level fertility data in the Kinki region in 1935 and 1947 and the data on damages from air raids in cities, we find the effects of bombing on postwar fertility in towns and villages within 15 kilometers, despite no direct damages. The estimation results suggest that severe air raids in neighboring cities increased fertility, whereas minor air raids decreased it. Moreover, the results of the quasi-experimental approach indicate that intense fears of air raids increased the fertility rate in the postwar period. Our study contributes to the literature on fertility changes in the postwar period, providing new insights into the complex relationship between war damage and fertility.
17:10-18:10
発表者2: 熊谷元宏 (一橋大学 社会科学高等研究院)
タイトル: Horses, Battles, and the State: Military Origins of Autocracy
概要:
This research explores the effect of horses on the state and battles. It exploits multiple exogenous sources of variation in the adoption of horse riding on the battlefield: (i) the regional variation in the spread of horse-riding technology, (ii) the variation in climatic suitability of horse survival, and (iii) the change in the availability of horses in the Americas during the Columbian Exchange. The research, using these exogenous variations and multiple datasets spanning several millennia, provides robust evidence of the significant impact of horses on the state and battles. Exploiting these exogenous variations in the adoption of horse riding on the battlefield, the research also establishes a short- and long-run association between horses and autocratic institutions.
8月4日(金): 15:00-18:20 富士見校舎4階404教室
Prospective Ph.D students Workshop
15:00-16:00
発表者1: 佐藤広人 (東京大学経済学研究科博士課程)
タイトル: Information Design for Sponsored Advertising (joint with Ryo Shirakawa)
概要:
How much sponsored products’ information do online platforms disclose? This study explores the platform’s information design problem of maximizing sponsored advertising revenue, wherein a consumer sequentially searches for products’ information with a deliberate order. In Weitzman’s (1979) ordered search model, the main theorem demonstrates that an optimal signal is binary. The revenue-maximizing search order is characterized by per-click advertising revenue over search cost. Furthermore, the platform’s optimal signal minimizes consumer welfare.
16:10-17:10
発表者2: 石田航 (ロチェスター大学博士課程)
タイトル: Feasibility-constrained matching market (joint with Yusuke Iwase and Taro Kumano)
概要:
Recently, it is observed that several matching markets are required to reflect social concerns or are restricted by the nature of a specific market structure. Importantly, at the stage of implementation, those are restrictions on allocation spaces. Thus we consider a many-to-one matching market in which feasible matchings are exogenously determined, as any subset of all matchings. Since the usual stability notion does not apply, we propose a new solution concept called feasibility-oriented stability. We show that it exists if and only if the empty matching is feasible. We then examine welfare analyses and comparative statics of the solution with respect to the constraints.
17:20-18:20
発表者3: 宮下将紀 (イエール大学博士課程/香港大学)
タイトル: Unique Equilibrium of Large Quadratic Games with Incomplete Information: A Pettis Integral Approach (joint with Takashi Ui)
概要:
This paper presents a unified analysis of large quadratic games with incomplete information, where each agent's best-response strategy depends linearly on other agents' aggregated strategies and relevant payoff states. The model can accommodate a continuum of heterogeneous agents, making it applicable to diverse scenarios involving large populations of interacting agents. We provide a sufficient condition under which a unique equilibrium exists for every information structure. The condition is necessary when payoff interactions are undirected. To attain these results, we offer a solid mathematical foundation for the aggregation of strategies by appealing to the integral notion à la Pettis (1938) and investigate a linear functional equation involving Pettis integral. The solvability of this equation is linked to the eigenvalues of the corresponding integral kernel. We further explore the stability of the equilibrium by identifying a condition under which it is obtained as the limit of best-response dynamics.
10月2日(月): 16:20-17:50 富士見校舎 5階503教室
発表者: 木口憲明（一橋大学大学院経済学研究科博士課程）
タイトル: Robustly Optimal Voting Rule (joint with Shinpei Noguchi and Yuta Takahashi)
概要: We examine a strategy-proof mechanism, called robustly optimal mechanism, that maximizes the expected sum of utilities under worst-case probability distributions when there is uncertainty on the distribution of individual utility profiles. Departing from existing literature, we permit individuals to have utility functions that are neither single-crossing nor single-peaked. In this setting, we show that (1) the robustly optimal mechanism is unique, and (2) it can be implemented using the median voting rule. As a result of its robustness requirement, the optimal mechanism intrinsically possesses desirable properties such as anonymity and Pareto efficiency, both crucial for voting rules, while also being easily implementable.
11月6日(月): 16:50-18:20 富士見校舎 5階503教室
発表者: 佐藤進（一橋大学経済研究所）
タイトル: Competitive Platform Design (joint with Yusuke Zennyo)
2月16日(金): 14:45-18:20 富士見校舎 5階 503教室
14:45-15:35
発表者1：中田里志（東京理科大学経営学部）
タイトル：The Multi-Threshold Generalized Sufficientarianism and Level-Origarchy (joint with Norihito Sakamoto)
概要：This paper investigates a class of social welfare orderings that satisfy the standard and acceptable axioms in the literature: anonymity, strong Pareto, separability, and Pigou-Dalton transfer (or, convexity). Due to the lack of continuity, we show that the class of social orderings typically has some thresholds satisfying the following property, which we call level-oligarchy: individuals whose utility is less than the value are prioritized over the other individuals whose utility is greater than the value. First, we provide the novel reduced form characterization that a social welfare ordering satisfies anonymity, strong Pareto, separability, and convexity only if it is either the weak generalized utilitarian or level-oligarchy. Next, by dropping convexity and instead requiring Pigou-Dalton transfer and a mild continuity axiom, we characterize the new class of social welfare orderings, the multi-threshold generalized sufficientarian orderings, which subsumes the leximin, generalized utilitarian, and critical-level sufficientarian social welfare orderings as special cases. Therefore, we can provide a unified characterization for the important class of social orderings only by the permissible axioms. In particular, although the social judgment from both classes of orderings seems quite different, our result implies that the difference between the utilitarian and leximin orderings just comes from the degree of continuity.
15:45-16:55
発表者2：柳田翔平（一橋大学大学院博士課程）
タイトル：Overreacting to a Black Box
概要：People often receive unexpected information and tend to over-reflect it in shaping their posterior beliefs. Such reactions can also occur when people do not understand the information generation process, for example, recommendations from machine learning algorithms. In this study, we characterize a decision maker’s belief updating which exhibits overreaction to unexpected information, in the absence of knowledge about its generation process. In the resulting updating rule, overreaction occurs when the distance between prior and information is far enough. Moreover, when information is not unexpected, the updating rule coincides with the contraction rule of Ke et al.(2023).
17:10-18:20
発表者3：石川竜一郎（早稲田大学国際学術院）
タイトル: A Resolution of the Centipede Paradox (jointly with Mamoru Kaneko)
概要：The centipede paradox is the antagonism between the outcome from the BI (backward induction) theory in a centipede game and people's responses to it. We weaken the underlying postulates of the BI theory so that payo¤s may be incomparable for the players depending upon bounded cognitive abilities. In this case, a player may follow inertia when decision nodes have some distance from the start. In the CIB (conscious-choice/inertial behavior ) theory which we develop, when the players have high cognitive abilities, it exhibits the same behavior as the BI theory, but when at least one has a low ability, it induces quite opposite outcomes selecting the actions of continuation to the end area of the centipede game. To examine this result, we consider the degree of reversed causality, which is the discrepancy from hypothetical evaluations of future possibilities to the present choice. When at least one player has low cognitive ability of payo¤s, the degree of reversed causality takes a low value, meaning that the game ends near the last end node. We argue that this is compatible with the antagonism shown by people in a chain-store game reported by Reinhard Selten. The developments as a whole form a resolution of the centipede paradox.