Measuring Information Burden: From Coalition-Based Reasoning to the Price System
summary
The gist
The price system is often said to economize on information <ref:1802.04595#pg2> This paper asks how much information it saves <ref:1802.04595#pg4>.
In short
The paper develops a formal framework to measure how much information is saved by a price system compared to individual reasoning over coalitions. It defines 'load-bearing coalitions' whose feasibility information must be distributed, showing that the average per-agent informational burden grows as Theta(4k/k^(3/2)), formally quantifying the economic benefit of pricing.
Key concepts
- Informational Burden
- This is the cognitive load agents face when they must reason individually over all possible coalition alternatives to make a decision. The paper formalizes this burden using logic, showing how much information must be held for individual acceptance decisions to recover the core.
- Core Boundary
- The core is reinterpreted as the exact boundary where unanimous acceptability matches the core. The authors find that this boundary is characterized by requiring every coalition to be known to at least one member, establishing a minimal informational structure needed for unanimous justification.
- Load-Bearing Coalitions
- These are specific coalitions whose feasibility information must be distributed across the population in a k-fold replica economy. Counting these coalitions provides a quantitative measure of the extra information that agents must bear due to market structures.
- Gentzen’s Sequent Calculus
- This is a formal system used to model agent reasoning as mathematical proofs. By using this calculus, the authors can rigorously verify whether an agent can derive a decision criterion from their specific set of held sentences, linking information structure directly to logical conclusions.
Terminology used across episodes
This episode discusses
The paper
Measuring Information Burden: From Coalition-Based Reasoning to the Price System · Read on arXiv
Department of Decision Sciences, Bocconi University
The price system is often said to economize on what individuals need to know. This paper measures a particular informational burden associated with decentralized reasoning in its absence: the coalition-feasibility information required for individual rejection decisions to recover the core. We develop a proof-theoretic framework in which information is represented by economically meaningful sentences and reasoning by derivations in Gentzen's sequent calculus. An agent rejects a proposed payoff vector only when she can establish the existence of a profitable coalitional deviation; failure to reject may instead reflect ignorance. For each transferable-utility game, the main theorem characterizes the informational distributions under which the set of unopposed payoff vectors coincides with the core: every noncore payoff vector must have a blocking coalition known to at least one of its members. This yields a game-specific measure of minimum informational burden. We extend the characterization to pure-exchange economies through a first-order framework that preserves finite informational premises. Applying the result to Shapley's (1975) replica economy, within the core-convergence tradition of Debreu and Scarf (1963), we show that the minimum average coalition-level informational burden grows as Θ(4 k/k 3/2), exponentially with replication. The result quantifies an informational requirement of direct decentralized coalitional reasoning that price-mediated coordination can spare agents, without claiming to measure the total informational savings of prices.
Transcript
Introduction to the show: ident: Economics. The day in economics: theory, mechanisms and the evidence beneath them.
Ruth: I'm Ruth.
Warren: And I'm Warren, here with Iris.
Ruth: We're looking at the paper: "Measuring Information Burden".
Warren: Here's the gist of the paper. The price system is often said to economize on information <ref:1802.04595#pg2> This paper asks how much information it saves <ref:1802.04595#pg4>.
Ruth: Formal Framework for Measuring Burden The authors develop a formal framework for measuring the informational burden that agents would have to bear if they had to reason individually over feasible coalitional alternatives <ref:1802.04595#pg5>.
Warren: Taking the convergence theorem of Debreu and Scarf (1963) as the benchmark, they ask how much coalition-feasibility information must be held, and how it must be distributed, for individual acceptance decisions to recover the core <ref:1802.04595#pg5>.
Ruth: Agents’ information is represented as finite sets of meaning-bearing sentences in a formal language, and reasoning is modeled as proof in Gentzen’s sequent calculus <ref:1802.04595#pg8>.
Title and authors: Ruth: So we're talking about the paper "Measuring Information Burden: From Coalition-Based Reasoning to the Price System," and it’s about quantifying the informational savings of using prices instead of individual reasoning over coalitions.
Warren: It’s essentially building a formal way to measure that burden, looking at how much information agents need to hold if they have to reason individually over every possible combination of players >
Iris: The paper sets up this framework by using a formal language for agent information and modeling the decision-making process as a proof in Gentzen’s sequent calculus >
Ruth: And they use the convergence theorem of Debreu and Scarf from one thousand nine hundred sixty-three as their benchmark to figure out exactly how much coalition feasibility information is needed to recover the core > <ref:1802.04595#pg1,the convergence theorem of Debreu and Scarf>
Warren: The main finding there is that for unanimous acceptability to match the core across all transferable utility games, every coalition must be known by at least one member >
Iris: They also show that if you remove even a single coalition's information from everyone in it, you can break that equivalence for certain games >
Ruth: So they redefine the core as this exact boundary for what we can unanimously justify with a minimal but well-defined information structure >
Warren: Then they take this and apply it to the k-fold replica economy to count exactly which coalitions' feasibility information must be distributed across the entire population >
Iris: This leads them to show that the average per agent informational burden scales as Theta of four k over k cubed over two, which is exponential in terms of the economy's size >
The paper's summary: Ruth: To summarize what they’re doing, this paper takes the classical idea that institutions economize on knowledge and shows how much that saving actually is when we switch from individual reasoning to a price system >
Warren: They use a formal proof structure to determine the exact minimal informational structure required for unanimous acceptability to coincide with the core in these types of games >
Iris: They establish that this minimal structure requires every coalition to be known by at least one agent, and that removing just one piece of information from a coalition breaks the agreement >
Ruth: The paper then moves into a quantitative measure by applying this to replica economies to count the load-bearing coalitions and find their average per agent informational burden >
Warren: They find that this burden grows exponentially in terms of the number of players or agents, specifically as Theta of four k over k cubed over two >
Iris: This result is significant because it gives a precise, formal answer to the Hayekian question about how much information the price system actually saves >
Ruth: It moves beyond just saying prices save info; it shows exactly *how much* information, quantified by this specific scaling factor, is being saved in terms of agent cognitive load >
The paper's improvements: Warren: The authors suggest that the framework itself is a major improvement because it provides a rigorous method for measuring relevant information that isn't just some vague intuition >
Ruth: They introduce the formal language and proof-theoretic apparatus, which gives them exactly the kind of structure needed to make information measurable, not just an abstract concept >
Iris: The real methodological value is in providing this formal setting where relevant information can be measured, showing that institutions economize not only on transaction costs but also on this informational burden >
Ruth: They’ve also applied it to the pure exchange economy and found that the load-bearing coalitions are determined by the geometry of the core in replica economies >
Warren: In an exchange economy setting, they show that for a k-fold replica economy where k is greater than or equal to two, there’s a set Lk of load-bearing coalitions that includes individual coalitions and mixed pairs and one-asymmetric coalitions and the grand coalition >
Iris: And they confirm that the average per agent informational burden in this exchange economy setting grows as Theta of four k over k cubed over two, which diverges to infinity >
Conclusion: Ruth: So to wrap up, the paper "Measuring Information Burden: From Coalition-Based Reasoning to the Price System" gives us a formal answer on how much information is saved by the price system in terms of agent cognitive burden >
Warren: They show that this saving is quantified by an average per agent informational burden growing as Theta of four k over k cubed over two, which means this saving gets exponentially bigger as the economy gets larger >
Iris: The paper provides a rigorous way to measure information, using formal logic and proof theory to determine the minimal knowledge structure needed for equilibrium >
Ruth: It really frames institutional efficiency in terms of information cost, showing that mechanisms economize on cognitive simplicity alongside transaction costs >
Warren: So this is a precise answer to the Hayekian question about how much information the price system saves, quantified by this specific scaling factor and diverging as we scale up >
More episodes
- Daily Summary for 2026-11-19
- 2608.21274-Recommendation Quality and the Concentration of Consumption: Experimental Evidence from Netflix
- 2408.02757-A nonparametric test for diurnal variation in spot correlation processes
- 2306.16393-High-Dimensional Canonical Correlation Analysis
- 2605.17299-Geometric Brownian motion with intermittent entries and exits
- 2607.04753-Fooling Yourself: how narratives shape beliefs
- 2408.14671-Inference on Treatment Effects with Repeatedly Mismeasured High-Dimensional Controls
- 2201.06898-Difference-in-Differences Estimators for Treatments Continuously Distributed at Every Period
- 2501.15307-Influence Functions: A Unified Differentiation-Based Framework
- The day in conversation for 2026-11-19