Subhajit Pramanik

Research Perspective:

My current research centres on designing and analysing dynamic matching markets, where participants, their preferences and available resources evolve over time. I use game theory and mechanism design techniques to formalize how agents interact under incomplete or costly-to-verify information. A key goal is to construct mechanisms that remain fair and efficient over multiple rounds of interaction. Another important direction is examining the price of fairness in these dynamic settings, understanding the trade-offs between equitable outcomes and overall efficiency. Recently, I have been focusing on international kidney exchange programs, exploring how to balance the cooperative gains among countries while ensuring no group is disadvantaged by participating.

My Current Projects are as follows:

A Dynamic Bargaining Framework for International Kidney Paired Exchange Programs. -- with Antonio Nicolò.
[Latest Draft]

Abstract: This paper develops a theoretical framework for International Kidney Paired Exchange (IKPE) to address efficiency and fairness concerns in multi-country kidney-paired exchange networks. Drawing on a discrete version of the Kalai–Smorodinsky (KS) bargaining solution, we propose a mechanism that first maximizes worst country’s relative gain, and then selects an allocation that achieves the largest overall transplant benefit. The model features a dynamic weighting scheme that adjusts over time to compensate countries whose participation yields lower gains in the past rounds compared to other countries, ensuring an equitable distribution of cooperative gains. We prove that the proposed mechanism guarantees Pareto optimality and individual rationality of the resulting allocation as no country is worse off by cooperating than by acting alone.

Bayesian Persuasion and Costly State Verification. -- with Raghul S Venkatesh and Souvik Roy [Under revision]
[Latest Draft]

Abstract: We study a model in which a sender tries to persuade a decision-maker by committing to a signaling strategy (Kamenica and Gentzkow, 2011). The decision-maker observes this and, in turn, commits to an interim verification strategy that reveals the true state of the world with an endogenously determined probability. The possibility of verification induces a trade-off for the sender: providing more information decreases the verification probability but also decreases the ex-ante persuasion payoff. The decision-maker trades off learning the state with more precision against the cost of verification. In equilibrium, the sender provides more information, which decreases the sender’s welfare and increases that of the decision-maker. We provide the precise condition under which social welfare is higher compared to the case without verification.

Popular Stable Matching in One-to-One Marriage Market.

Abstract: This project studies the relationship between stability and popularity in strict two-sided one-to-one matching markets. I introduce three domain axioms regulating majority comparisons on alternating paths and cycles: (DR1) odd-path endpoint-majority, (DR2) strict cycle dominance, and (DR3) uniqueness of the maximum matched set, and show that these conditions are sufficient for stability and popularity to coincide. My main theorem proves that under DR1–DR3, every popular matching is stable and, conversely, every stable matching is popular, yielding an exact equivalence between majority choice and Gale–Shapley stability. We illustrate the role of each axiom through constructed examples and show how violations of DR1 permit unstable yet weakly popular enlargements. We further analyze the master-list domain, where the stable matching is unique and forms the minimum-cardinality popular matching, and identify structural restrictions on popular matchings imposed by DR1.

Two-Sided Dynamic Information Acquisition with Sequential Interviews.

Abstract: This project studies the problem of how students and colleges (or similar parties) decide who to interview when trying to find the best match over multiple rounds. The core issue is that when a student and a college decide to interview, they both learn privately how good the match is, but each side only focuses on its own costs and benefits. They ignore the positive benefit (externality) their interview creates for the other side, like giving the other party valuable information. Because of this self-interest, the current decentralized system schedules too few interviews compared to what would be best for overall welfare (a problem called "under-interviewing"). The paper proposes a solution using a central coordination method and a mechanism called "priority tokens" These non-monetary tokens are used to ration interviews and credit each side for the expected gain they create for the other side. This mechanism effectively aligns each person's private decision to interview with the total social benefit, leading to more interviews and a better overall final matching.

Others:

Technical Reports, Online Media and Blogs:
1. Finding Optimal Matchings using Voting Rules - SRIP'22 Report (Under the supervision of Prof Neeldhara Misra). [Link]
  (Project Report - SRIP'22 at Department of Computer Science, IIT Gandhinagar.)
2. Exchange Rate and Economic Growth - a comparative analysis of the possible relationship between them (2021). [Link]
  (Based on a term paper written for the International Economics course during undergrad.)
3. An essay on labour supply decisions and reference-dependent preferences. (2022) [Link]
  (Based on a term paper written for Behavioral Economics course during undergrad.)
4. Unravelling the importance of h bonds, σ–hole and π–hole-directed intermolecular interactions in nature. (with Prof Deepak Chopra) (Journal of the Indian Institute of Science, 2020) [Link]
  (A review article in chemistry exploring the world of non-covalent bonds written during my earlier undergrad days!)
5. The Attack of Coronavirus on the Economy. (June 20, 2020) (Blog on The Qrius Rhino) - [Link]
  (Years ago, an amateur (me) dared to blog bravely, if not grammatically!)

Google Sites

Report abuse