서브메뉴
검색
Essays in Microeconomic Theory, Industrial Organization, and Experimental Economics
Essays in Microeconomic Theory, Industrial Organization, and Experimental Economics
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260209102841
- ISBN
- 9798293887897
- DDC
- 152
- 서명/저자
- Essays in Microeconomic Theory, Industrial Organization, and Experimental Economics
- 발행사항
- [Sl] : New York University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 265 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-03, Section: B.
- 주기사항
- Advisor: Madsen, Erik.
- 학위논문주기
- Thesis (Ph.D.)--New York University, 2025.
- 초록/해제
- 요약This dissertation consists of three chapters, the first in the fields of contract theory and mechanism design, the second in theoretical industrial organization and competition policy, and the third in behavioral and experimental economics.Chapter 1. The first chapter proposes a novel model of multidimensional screening, where a candidate (she) with two attributes-training and talent-chooses how much hard evidence of training to present. Then, the evaluator (he) possibly verifies at a cost the value of a composite measure of the candidate's training and talent before deciding whether to accept or reject the candidate. The candidate cannot unilaterally provide evidence of talent. The composite measure is increasing in both training and talent, and the evaluator (weakly) values both training and talent in the candidate. If the evaluator is going to verify the value of the composite measure, the candidate may have incentives to withhold evidence of training-although the evaluator values training-to influence how the evaluator interprets the composite measure. Particularly, she may want to withhold evidence of training to make the evaluator attribute the composite measure to talent instead, thereby overestimating her talent.This problem arises when the composite measure is less sensitive to talent than talent is valuable to the evaluator. In that case, the optimal evaluation scheme never combines evidence and verification in the evaluation of a certain candidate. Rather, it asks for evidence of training only to accept a high-training candidate without verification. The optimal mechanism favors high- over low-training candidates: (i) It accepts some high-training candidates-including unworthy ones-without verifying their composite measure but rather only by asking them for a certain level of evidence of training; and (ii) among candidates who do not meet that level of evidence, (iia) it accepts (after verification) some unworthy candidates with high training but low talent while (iib) rejecting some worthy candidates with high talent but low training.Remarkably, this is the structure of the optimal mechanism even when the evaluator only values talent. The evaluator still optimally favors high-training candidates even though training is worthless to him. He does so because of two forces: (i) to save on verification costs by accepting high-training candidates without verifying their composite measure and (ii) due to the strategic incentives of candidates to withhold evidence of training when the evaluator verifies the value of a composite measure that is under-sensitive to talent. The two forces are complements in inducing errors in favor of high- and against low-evidence candidates. Namely, the second force exacerbates the errors due to the verification cost by decreasing the effectiveness of verification, thereby pushing the evaluator to accept high-training candidates without verifying their composite measure to save on verification costs.Chapter 2. In the second chapter, "Free entry in a Cournot market with overlapping ownership'' (co-authored with Xavier Vives), we examine the effects of overlapping ownership in a Cournot oligopoly where existing firms with overlapping ownership decide whether to enter a new market. Overlapping ownership, be it in the form of common or cross ownership, significantly differs from collusion. In infinitely repeated oligopoly games, (the prospect of) entry poses a constraint on incumbents. It limits the attainable collusive outcomes, since by colluding incumbents increase prices thereby enhancing the incentives for entry by new firms. On the other hand, (pre-entry) overlapping ownership entry acts as an additional channel through which the competitive effects of overlapping ownership can materialize.Namely, we show that in most cases-and especially when overlapping ownership is already widespread, an increase in the extent of overlapping ownership will harm welfare by (i) softening product market competition, (ii) reducing entry, thereby (in contrast to standard results) inducing or exacerbating insufficient entry, and (iii) magnifying the negative impact of an increase of entry costs on entry. Overlapping ownership can mostly be beneficial only under increasing returns to scale, in which case industry consolidation (induced by overlapping ownership) leads to sizable cost efficiencies.Chapter 3. The third chapter, "Regret, blame, and division of responsibility in games,'' studies regret in games, which has so far been analyzed as if in a single-agent context with the other players' actions treated as the state of the world. I instead propose the strategic regret approach, which accounts for the division of responsibility in games. It postulates that player i's regret (for not playing a best-response) is mitigated through blame put on player j for not playing a Pareto-improving (compared to j's actual action) best-response to player i's action.I provide experimental evidence lending direct support to both the assumptions and predictions of strategic regret. Survey questions that elicit participants' feelings in certain hypothetical scenarios show that the subjects' regret is indeed mitigated through blame assigned to others for not playing a Pareto-improving best-response. Notably, participants' anticipated regret and blame elicited in certain games have predictive power-consistent with strategic regret predictions-over their behavior in vastly different games, consistent with theoretical predictions.We conclude that, when modified to account for blame and the division of responsibility in games, regret offers novel insights on strategic behavior. More generally, the results emphasize that decision-theoretic models may benefit from modifications when applied in games. (Implicit) assumptions that are plausible (or even hardly qualify as assumptions) in single-agent settings (e.g., that the agent does not blame the random state of the world) should be reconsidered in strategic environments.
- 일반주제명
- Experimental psychology
- 키워드
- Blame
- 키워드
- Common ownership
- 키워드
- Evidence games
- 키워드
- Regret theory
- 기타저자
- New York University Economics
- 기본자료저록
- Dissertations Abstracts International. 87-03B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260203s2025 us c eng d■001000017365862
■00520260209102841
■006m o d
■007cr#unu||||||||
■020 ▼a9798293887897
■035 ▼a(MiAaPQ)AAI32115875
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a152
■1001 ▼aVravosinos, Orestis.
■24510▼aEssays in Microeconomic Theory, Industrial Organization, and Experimental Economics
■260 ▼a[Sl]▼bNew York University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a265 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-03, Section: B.
■500 ▼aAdvisor: Madsen, Erik.
■5021 ▼aThesis (Ph.D.)--New York University, 2025.
■520 ▼aThis dissertation consists of three chapters, the first in the fields of contract theory and mechanism design, the second in theoretical industrial organization and competition policy, and the third in behavioral and experimental economics.Chapter 1. The first chapter proposes a novel model of multidimensional screening, where a candidate (she) with two attributes-training and talent-chooses how much hard evidence of training to present. Then, the evaluator (he) possibly verifies at a cost the value of a composite measure of the candidate's training and talent before deciding whether to accept or reject the candidate. The candidate cannot unilaterally provide evidence of talent. The composite measure is increasing in both training and talent, and the evaluator (weakly) values both training and talent in the candidate. If the evaluator is going to verify the value of the composite measure, the candidate may have incentives to withhold evidence of training-although the evaluator values training-to influence how the evaluator interprets the composite measure. Particularly, she may want to withhold evidence of training to make the evaluator attribute the composite measure to talent instead, thereby overestimating her talent.This problem arises when the composite measure is less sensitive to talent than talent is valuable to the evaluator. In that case, the optimal evaluation scheme never combines evidence and verification in the evaluation of a certain candidate. Rather, it asks for evidence of training only to accept a high-training candidate without verification. The optimal mechanism favors high- over low-training candidates: (i) It accepts some high-training candidates-including unworthy ones-without verifying their composite measure but rather only by asking them for a certain level of evidence of training; and (ii) among candidates who do not meet that level of evidence, (iia) it accepts (after verification) some unworthy candidates with high training but low talent while (iib) rejecting some worthy candidates with high talent but low training.Remarkably, this is the structure of the optimal mechanism even when the evaluator only values talent. The evaluator still optimally favors high-training candidates even though training is worthless to him. He does so because of two forces: (i) to save on verification costs by accepting high-training candidates without verifying their composite measure and (ii) due to the strategic incentives of candidates to withhold evidence of training when the evaluator verifies the value of a composite measure that is under-sensitive to talent. The two forces are complements in inducing errors in favor of high- and against low-evidence candidates. Namely, the second force exacerbates the errors due to the verification cost by decreasing the effectiveness of verification, thereby pushing the evaluator to accept high-training candidates without verifying their composite measure to save on verification costs.Chapter 2. In the second chapter, "Free entry in a Cournot market with overlapping ownership'' (co-authored with Xavier Vives), we examine the effects of overlapping ownership in a Cournot oligopoly where existing firms with overlapping ownership decide whether to enter a new market. Overlapping ownership, be it in the form of common or cross ownership, significantly differs from collusion. In infinitely repeated oligopoly games, (the prospect of) entry poses a constraint on incumbents. It limits the attainable collusive outcomes, since by colluding incumbents increase prices thereby enhancing the incentives for entry by new firms. On the other hand, (pre-entry) overlapping ownership entry acts as an additional channel through which the competitive effects of overlapping ownership can materialize.Namely, we show that in most cases-and especially when overlapping ownership is already widespread, an increase in the extent of overlapping ownership will harm welfare by (i) softening product market competition, (ii) reducing entry, thereby (in contrast to standard results) inducing or exacerbating insufficient entry, and (iii) magnifying the negative impact of an increase of entry costs on entry. Overlapping ownership can mostly be beneficial only under increasing returns to scale, in which case industry consolidation (induced by overlapping ownership) leads to sizable cost efficiencies.Chapter 3. The third chapter, "Regret, blame, and division of responsibility in games,'' studies regret in games, which has so far been analyzed as if in a single-agent context with the other players' actions treated as the state of the world. I instead propose the strategic regret approach, which accounts for the division of responsibility in games. It postulates that player i's regret (for not playing a best-response) is mitigated through blame put on player j for not playing a Pareto-improving (compared to j's actual action) best-response to player i's action.I provide experimental evidence lending direct support to both the assumptions and predictions of strategic regret. Survey questions that elicit participants' feelings in certain hypothetical scenarios show that the subjects' regret is indeed mitigated through blame assigned to others for not playing a Pareto-improving best-response. Notably, participants' anticipated regret and blame elicited in certain games have predictive power-consistent with strategic regret predictions-over their behavior in vastly different games, consistent with theoretical predictions.We conclude that, when modified to account for blame and the division of responsibility in games, regret offers novel insights on strategic behavior. More generally, the results emphasize that decision-theoretic models may benefit from modifications when applied in games. (Implicit) assumptions that are plausible (or even hardly qualify as assumptions) in single-agent settings (e.g., that the agent does not blame the random state of the world) should be reconsidered in strategic environments.
■590 ▼aSchool code: 0146.
■650 4▼aExperimental psychology
■653 ▼aBlame
■653 ▼aCommon ownership
■653 ▼aCostly verification
■653 ▼aEvidence games
■653 ▼aRegret theory
■690 ▼a0501
■690 ▼a0511
■690 ▼a0623
■71020▼aNew York University▼bEconomics.
■7730 ▼tDissertations Abstracts International▼g87-03B.
■790 ▼a0146
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17365862▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


