arXiv subjects

Seungjin Han

Publications and source records attributed to Seungjin Han.

10 recordsLinked to original sources

arXiv2026

Ranking Treatment Saturations under Clustered Network Interference

In this paper, we study how to rank a finite set of treatment saturations for a target population with clustered network interference. We propose an empirical success (ES) ranking rule that, for each pair of saturations, selects the saturation level with the higher estimated welfare using data from a two-stage randomized saturation design. We adopt the statistical decision theory framework with additively separable regret loss to assess the performance of the ES ranking rule. We derive non-asymptotic upper bounds on the maximum regret of the ES ranking rule that depend on the within-cluster network only through a single combinatorial summary of its dependency structure. We exploit these bounds to characterize a quasi-optimal first-stage saturation distribution within the two-stage randomized saturation design. We further show that the ES ranking rule is asymptotically optimal among threshold ranking rules in the sense of minimizing an upper bound on the worst-case regret.

econ.EM↗

arXiv2026

Contracting with Imperfect Commitment: Minimal Canonical Contracts

Contract theory typically assumes full commitment by the principal, but many contracts fix some payoff-relevant decisions while leaving others discretionary. We ask when imperfect commitment is equivalent to full commitment. For contracts in which a committed baseline is followed by a bounded discretionary adjustment, as in commercial-insurance schedule rating or civil penalties, bounded discretion is allocation-neutral. When contractible and non-contractible decisions are distinct instruments, the equivalence fails. We characterize optimal single-principal contracts and show that simple-offer equilibria are robust under competing principals. The methodological contribution is an extended taxation principle that makes these analyses more tractable.

econ.TH↗

arXiv2024

Optimal Wage Band for Job Matching with Signaling

We study an optimal wage band problem in a competitive matching labor market where education signals worker ability. We prove uniqueness of the competitive signaling equilibrium under a general class of utility and profit functions and show that the optimal wage band problem is isomorphic to a simpler optimal ability threshold problem. Using a parametric model, we analyze how wage bands improve welfare relative to no intervention. Our results highlight novel mechanisms driven by asymmetric information, contrasting with existing literature. The framework is broadly applicable to settings where agents invest in costly signals under asymmetric information in competitive matching environments.

econ.TH↗

arXiv2024

Competing Mechanisms in Games Played Through Agents: Theory and Experiment

This paper proposes Competing Mechanism Games Played Through Agent (CMGPTA), an extension of the GPTA (Prat and Rustichini (2003)), where a Principal can offer any arbitrary mechanism that specifies a transfer schedule for each agent conditional on all Agents' messages. We identify the set of equilibrium allocations using deviator-reporting mechanisms (DRMs) on the path and single transfer schedules off the path. We design a lab experiment implementing DRMs. We observe that implemented outcomes are efficient more often than random. A majority of the time, Agents do tell the truth on the identity of a deviating Principal, despite potential gains from (tacit) collusion on false reports. As play progresses, Agents learn to play with their counterparty Agent with the average predicted probability of collusion on false reports across groups increasing from about 9% at the beginning of the experiment to just under 20% by the end. However, group heterogeneity is significant.

econ.TH↗

arXiv2024

Multidimensional Signaling with a Resource Constraint

We study multidimensional signaling (cognitive/non-cognitive) as a sender's portfolio choice with a resource constraint. We establish the existence of a unique monotone D1 equilibrium where the cognitive (non-cognitive) signal increases (decreases) in sender type and the sum of the two increases in sender type. The equilibrium is characterized by two threshold sender types. The low threshold is one where a kink occurs in signaling. The constraint is binding only for sender types above it. The high threshold is the other one, above which all types spend all the resources in cognitive signal with pooling and discontinuity on the top.

econ.TH↗

arXiv2023

Common Agency with Non-Delegation or Imperfect Commitment

In classical contract theory, we usually impose two assumptions: delegated contracts and perfect commitment. While the second assumption is demanding, the first one suffers no loss of generality. Following this tradition, current common-agency models impose delegated contracts and perfect commitment. We first show that non-delegated contracts expand the set of equilibrium outcomes under common agency. Furthermore, the powerful menu theorem for common agency (Peters (2001) and Martimort and Stole (2002)}) fails for either non-delegated contracts or imperfect commitment. We identify canonical contracts in such environments, and re-establish generalized menu theorems. Given imperfect commitment, our results for common-agency models are analogous to those in Bester and Strausz (2001) and Doval and Skreta (2012) for the classical contract theory, which re-establish the revelation principle.

econ.TH↗

arXiv2023

Robust Equilibria in General Competing Mechanism Games

This paper proposes the notion of robust PBE in a general competing mechanism game of incomplete information where a mechanism allows its designer to send a message to himself at the same time agents send messages. It identifies the utility environments where the notion of robust PBE coincides with that of strongly robust PBE (Epstein and Peters (1999), Han (2007)) and with that of robust PBE respectively. If each agent's utility function is additively separable with respect to principals' actions, it is possible to provide the full characterization of equilibrium allocations under the notion of robust PBE and its variations, in terms of Bayesian incentive compatible (BIC) direct mechanisms, without reference to the set of arbitrary general mechanisms allowed in the game. However, in the standard competing mechanism agme, the adoption of robust PBE as the solution concept does not lead to the full characterization of equilibrium allocations in terms of BIC direct mechanisms even with agents' separable utility functions.

econ.TH↗

arXiv2023

Optimal Delegation in Markets for Matching with Signaling

This paper studies a delegation problem faced by the planner who wants to regulate receivers' reaction choices in markets for matching between receivers and senders with signaling. We provide a noble insight into the planner's willingness to delegate and the design of optimal (reaction) interval delegation as a solution to the planner's general mechanism design problem. The relative heterogeneity of receiver types and the productivity of the sender' signal are crucial in deriving optimal interval delegation in the presence of the trade-off between matching efficiency and signaling costs.

econ.TH↗

arXiv2022

Statistical Treatment Rules under Social Interaction

In this paper we study treatment assignment rules in the presence of social interaction. We construct an analytical framework under the anonymous interaction assumption, where the decision problem becomes choosing a treatment fraction. We propose a multinomial empirical success (MES) rule that includes the empirical success rule of Manski (2004) as a special case. We investigate the non-asymptotic bounds of the expected utility based on the MES rule. Finally, we prove that the MES rule achieves the asymptotic optimality with the minimax regret criterion.

econ.EM↗

arXiv2021

Monotone Equilibrium in Matching Markets with Signaling

We introduce a notion of competitive signaling equilibrium (CSE) in one-to-one matching markets with a continuum of heterogeneous senders and receivers. We then study monotone CSE where equilibrium outcomes - sender actions, receiver reactions, beliefs, and matching - are all monotone in the stronger set order. We show that if the sender utility is monotone-supermodular and the receiver's utility is weakly monotone-supermodular, a CSE is stronger monotone if and only if it passes Criterion D1 (Cho and Kreps (1987), Banks and Sobel (1987)). Given any interval of feasible reactions that receivers can take, we fully characterize a unique stronger monotone CSE and establishes its existence with quasilinear utility functions.

econ.TH↗