Blackwell-Monotone Updating Rules

Mark Whitmeyer

arxiv: 2302.13956 · v6 · submitted 2023-02-27 · 💰 econ.TH

Blackwell-Monotone Updating Rules

Mark Whitmeyer This is my paper

Pith reviewed 2026-05-24 10:01 UTC · model grok-4.3

classification 💰 econ.TH

keywords updating rulesBlackwell monotonicityBayes' lawinformation economicsdecision theoryposterior distortionsnon-paternalistic evaluation

0 comments

The pith

Bayes' law is the only strictly Blackwell-monotone updating rule among those that distort posteriors in a signal-independent way.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper defines an updating rule as Blackwell monotone when more information is always at least as good for the agent in any decision problem, and strictly Blackwell monotone when more information is strictly better in some decision problem. It establishes that Bayes' law satisfies strict Blackwell monotonicity. Within the class of rules that apply signal-independent distortions to Bayesian posteriors, the paper proves Bayes' law is the unique rule with this property. When an agent's choices are assessed using her own beliefs rather than paternalistic criteria, all Blackwell-monotone rules turn out to be affine transformations of Bayesian updating.

Core claim

Bayes' law is strictly Blackwell monotone. Within the broad class of updating rules that distort the Bayesian posteriors in a signal-independent manner, it is the only strictly Blackwell-monotone updating rule. When decisions are evaluated according to the agent's own beliefs, the Blackwell-monotone updating rules are precisely the affine distortions of the Bayesian posteriors.

What carries the argument

Blackwell monotonicity, which requires that more information is always (weakly) better for the agent across decision problems, with the strict version adding that some decision problem exists where the improvement is strict.

If this is right

Any strictly Blackwell-monotone rule in the considered class must coincide with Bayes' law.
Blackwell-monotone rules under non-paternalistic evaluation reduce to affine transformations of Bayesian posteriors.
Agents using non-Bayesian rules outside this affine class will sometimes strictly prefer less information in some decision problems.
The result supplies a characterization of Bayes' law via a monotonicity property with respect to information.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The characterization may extend to settings where agents face sequences of decisions rather than single problems.
It suggests a test for whether observed updating behavior satisfies the monotonicity property by checking consistency with affine distortions.
Applications to mechanism design could use the result to restrict attention to rules that preserve the value of information.

Load-bearing premise

The uniqueness result holds only for the class of updating rules that distort Bayesian posteriors in a signal-independent manner.

What would settle it

An explicit example of a signal-independent distortion of Bayesian posteriors that is strictly Blackwell monotone but is not Bayes' law, or a decision problem in which Bayes' law fails to make more information strictly better.

read the original abstract

An updating rule specifies how an agent reacts to information. An updating rule is Blackwell monotone if more information is always better for an agent in a decision problem and strictly Blackwell monotone if, in addition, there is always a decision problem in which more information is strictly better for an agent. Bayes' law is strictly Blackwell monotone, and I show that within a broad class of updating rules--those that distort the Bayesian posteriors in a signal-independent manner--it is the only strictly Blackwell-monotone updating rule. If an agent's decisions are evaluated non-paternalistically (according to her beliefs), the Blackwell-monotone updating rules are affine distortions of the Bayesian posteriors.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Whitmeyer's paper gives a scoped uniqueness result: Bayes' law is the only strictly Blackwell-monotone rule inside the signal-independent distortion class, with monotone rules being affine distortions of Bayes.

read the letter

The main point is that Bayes' law is the only strictly Blackwell-monotone updating rule among those that distort Bayesian posteriors in a signal-independent manner. The paper also shows that Blackwell-monotone rules are affine distortions of the Bayesian posteriors when decisions are evaluated according to the agent's own beliefs. This pins down when more information stays valuable under non-Bayesian updating inside that class. The result is a clean characterization theorem that resolves a specific question in decision theory. The author states the scope explicitly from the start, which keeps the claim from overreaching, and the split between monotone and strictly monotone versions is useful because it separates cases where information is always at least as good from cases where it is sometimes strictly better. The non-paternalistic evaluation angle is a reasonable way to tie the math back to actual decisions. The main limitation is the focus on signal-independent distortions. The uniqueness holds inside that frame, and the paper flags it, but rules that twist posteriors in signal-dependent ways fall outside the result. That is a real boundary on how far the finding travels. The abstract gives no proof details, so the derivation steps for the affine property are not visible, though the statement itself shows no internal contradiction or circularity. This paper is for decision theorists and behavioral economists who work on information value or non-Bayesian updating. Someone already familiar with Blackwell's theorem and looking to extend it to other updating rules would get direct use from the characterization. It is narrow enough that not every theory group needs it, but the scoped result is the sort of thing that belongs in a theory journal. I would send it to peer review rather than desk reject.

Referee Report

0 major / 2 minor

Summary. The manuscript claims that Bayes' law is strictly Blackwell monotone and is the unique strictly Blackwell-monotone updating rule within the class of rules that distort Bayesian posteriors in a signal-independent manner. It further establishes that, under non-paternalistic evaluation of decisions according to the agent's beliefs, all Blackwell-monotone updating rules are affine distortions of the Bayesian posteriors.

Significance. If the result holds, the paper supplies a precise decision-theoretic characterization linking Blackwell monotonicity to Bayesian updating within a well-scoped class, which strengthens the foundations of information economics and updating theory. The explicit restriction to signal-independent distortions and the non-paternalistic condition are strengths that render the uniqueness claim internally consistent and testable.

minor comments (2)

[Section 2] The definition of 'signal-independent distortion' in the class of updating rules could benefit from an explicit example immediately after its introduction to improve accessibility for readers unfamiliar with the literature on non-Bayesian updating.
[Theorem 2] Notation for the affine distortion parameters (e.g., the intercept and slope terms) should be checked for consistency between the statement of the main theorem and the subsequent corollaries.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive evaluation of the manuscript, including the accurate summary of our results on Blackwell monotonicity and the recommendation to accept. No major comments were raised.

Circularity Check

0 steps flagged

No significant circularity; scoped characterization theorem

full rationale

The paper presents a characterization result: Bayes' law is the unique strictly Blackwell-monotone updating rule inside the explicitly defined class of rules that apply signal-independent distortions to Bayesian posteriors. The class definition is the scope of the uniqueness claim, and the result is framed as a theorem within that class with no reduction of the derivation to its own inputs by construction. No self-citations, fitted parameters called predictions, or ansatzes smuggled via citation are indicated in the provided material. This is a self-contained theoretical result on its own terms.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Based solely on the abstract; the paper relies on standard decision-theoretic concepts with no free parameters or invented entities mentioned.

axioms (2)

standard math Blackwell's theorem on the value of information
Definition of Blackwell monotonicity builds directly on this background result.
domain assumption Existence of decision problems in which information has positive value
Used to define the strict version of Blackwell monotonicity.

pith-pipeline@v0.9.0 · 5618 in / 1180 out tokens · 33118 ms · 2026-05-24T10:01:13.880775+00:00 · methodology

Blackwell-Monotone Updating Rules

Core claim

What carries the argument

If this is right

Where Pith is reading between the lines

Load-bearing premise

What would settle it

discussion (0)