EC'24 Tutorial on Transaction Fee Mechanism Design
Blockchains are becoming increasingly important for the modern digital economy, with Ethereum alone holding assets valued at over $300 billion, and processing a daily transaction volume of more than $1 billion. A blockchain is a decentralized computer that users can interact with via transactions that modify the computer’s state. The blockchain is operated by miners, who collect transactions into batches called blocks, and who are responsible for reaching a consensus on the computer’s state. Due to the limits of the computer network (e.g., latency and bandwidth), block space is finite and scarce. Thus, users bid for block space in a transaction fee mechanism (TFM). The allocation of this mechanism determines the state transitions of the blockchain computer, with payments given to miners as an incentive to prioritize the allocation of transactions. A key challenge of the blockchain setting is that miners cannot be trusted: they may deviate from the protocol if profitable for them, possibly alone
In this tutorial, we present the basic TFM framework considered by the literature, discuss different aspects of TFM design, summarize previous major results, and highlight promising avenues for future work. Our goal is to introduce researchers to TFMs and equip them with the tools to conduct state-of-the-art research.
Keywords: Transaction Fee Mechanisms, Blockchains, Auctions, Mechanism Design.
Blockchains are complex multi-agent systems. The simplest model for a TFM assumes that (1) the miners are selling a single block and miners are myopic (i.e., miner incentives in future auctions are ignored), (2) all slots in a block are identical, and (3) there is a single blockchain. Our tutorial starts with this simple model and gradually challenges its assumptions by introducing more complex considerations that are also closer to reality, where, for each one, we present the corresponding major results. With this format, participants will appreciate the intricacies of a real-world TFM and learn about the great opportunity for future research.
Our panel’s focus will be on the major mechanism design challenges currently open in the TFM domain. Our speakers are thought leaders involved in both academia and industry:
Our goal is to introduce interested researchers to TFMs and equip them with the tools to conduct state-of-the-art research in this domain. Our target audience encompasses researchers, graduate students, and advanced undergraduates in EconCS. Some familiarity with auctions, mechanism design, or blockchains is helpful but not required.
Blockchains are becoming increasingly important for the modern digital economy, with Ethereum alone holding assets valued at over $300 billion, and processing a daily transaction volume of more than $1 billion. TFMs govern the resource allocation process on blockchains and are therefore inherently important for the secure, efficient, and fair operation of blockchains. The proper design of TFMs becomes even more crucial when considering that miners are profit-maximizing agents who have been observed to manipulate blockchain mechanisms for profit
In recent years, TFMs have emerged as a distinct research area of particular interest to mechanism designers, yet so far no one has done a tutorial focusing on the topic. With respect to tutorials on related topics, a blockchain tutorial was given in EC’22 (“Economics of Distributed Systems”, by Jacob Leshno and Matt Weinberg), and another in EC’18 (“Emerging Research Directions Regarding Incentives and Cryptocurrencies”, by Jacob Leshno, Arvind Narayanan, Georgios Piliouras, Alex Psomas and Matt Weinberg). Although not a tutorial, Tim Roughgarden gave a keynote talk in EC’22 on “Economics and Computation in Blockchains/Web3”, with a notable portion of the talk dedicated to TFMs.
Hao Chung is a Ph.D. student at Carnegie Mellon University, where he is advised by Elaine Shi. He is broadly interested in cryptography and mechanism design, especially in the intersection between two topics.
Email: haochung@andrew.cmu.edu
Matheus is a Postdoctoral Fellow in Computer Science at Harvard University and an incoming Assistant Professor of Computer Science at the University of Virginia. He holds a Ph.D. in Computer Science from Princeton University. Matheus’s research interests include security, applied cryptography, optimization, and their applications to blockchain systems. He is particularly interested in translating research into practice.
Email: matheus@seas.harvard.edu
Yotam Gafni is a postdoc student at Weizmann Institute, where he is hosted by Uri Feige. Yotam was a research member at SLMath Berkeley for the 2023 Fall semester, and has recently completed his Ph.D. at Technion, where he was advised by Ron Lavi and Moshe Tennenholtz. He is broadly interested in auctions, mechanism design, and the fairness and security of collaborative voting / machine-learning.
Email: yotam.gafni@gmail.com
Aviv researches the intricate relationship between the economics and security of cryptocurrencies, with a special interest in breaking cryptocurrency mechanisms. He is a Ph.D. candidate at The Hebrew University, a research consultant at Matter Labs, and a visiting researcher at the University of Innsbruck.
Email: aviv.yaish@mail.huji.ac.il