pith. sign in

arxiv: 2407.02948 · v3 · submitted 2024-07-03 · 💰 econ.TH

Information Greenhouse: Optimal Persuasion for Medical Test-Avoiders

Pith reviewed 2026-05-23 23:28 UTC · model grok-4.3

classification 💰 econ.TH
keywords information avoidancepersuasionmedical testingcommitmentoptimal communicationpatient behaviordoctor-patient interaction
0
0 comments X

The pith

When upfront warnings would deter testing, doctors commit to post-test reassurance policies that preserve hope about going untreated.

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

Patients avoid medical tests because the information can produce painful beliefs about their untreated prospects. The doctor designs disclosures about how bad non-treatment would be if the patient is sick, balancing the need to make treatment compelling after a bad result against the need to keep testing attractive beforehand. The optimal strategy is warning-in-advance whenever that warning remains compatible with the patient choosing to test; otherwise the doctor commits in advance to a post-test information policy that reassures the patient about the untreated state. With voluntary consultation the doctor must sometimes provide precautionary reassurance before the test occurs.

Core claim

When the warning that supports treatment is compatible with testing, the doctor uses warning-in-advance; when such warning would deter testing, the doctor constructs an information greenhouse: a committed post-test information environment that reassures the patient about the untreated prospect. With voluntary consultation, reassurance must sometimes be moved before the test as precautionary comfort.

What carries the argument

information greenhouse: a committed post-test information environment that reassures the patient about the untreated prospect

If this is right

  • The doctor trades off making treatment more attractive after diagnosis against keeping the patient willing to test in the first place.
  • Reassurance is sometimes supplied before the test when consultation itself is voluntary.
  • The information policy is chosen knowing the patient will update beliefs about the untreated state and decide whether to test and whether to treat.

Where Pith is reading between the lines

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

  • Similar commitment devices could appear in other settings where agents avoid information that would force costly actions, such as financial or environmental risk disclosure.
  • The model implies that regulatory rules limiting post-test communication could reduce testing rates if they prevent credible commitment to reassurance.
  • Testing the boundary case where the warning exactly equals the deterrence threshold would reveal whether doctors switch abruptly to greenhouse policies.

Load-bearing premise

The doctor can credibly commit to a post-test information policy that the patient anticipates and believes.

What would settle it

A field study or experiment in which doctors are observed committing (or failing to commit) to post-test information policies and patient testing rates are measured against the model's predicted threshold for when warnings become incompatible with testing.

read the original abstract

Patients often avoid medical tests because testing produces not only useful information but also painful beliefs. This paper studies optimal communication between a doctor and an information-avoidant patient who first decides whether to take a test and, after an unfavorable result, whether to accept treatment. The doctor can disclose information about how severe non-treatment would be if the patient is sick. The main tension is between warning and reassurance. A warning can make treatment compelling after diagnosis, but reassurance can make testing acceptable by preserving hope about the untreated prospect. I characterize the optimal policy. When the warning that supports treatment is compatible with testing, the doctor uses warning-in-advance. When such warning would deter testing, the doctor constructs an information greenhouse: a committed post-test information environment that reassures the patient about the untreated prospect. With voluntary consultation, reassurance must sometimes be moved before the test as precautionary comfort.

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.

Referee Report

2 major / 1 minor

Summary. The paper studies optimal doctor-patient communication when patients avoid medical tests due to anticipated painful beliefs about the untreated state. The doctor designs disclosure about non-treatment severity to balance encouraging testing and inducing treatment after an unfavorable result. The central claim is a characterization of the optimal policy: the doctor uses warning-in-advance when it is compatible with testing, but constructs an 'information greenhouse' (a committed post-test information environment that reassures about the untreated prospect) when a warning would deter testing. With voluntary consultation, reassurance must sometimes be provided before the test as precautionary comfort.

Significance. If the characterization is correct, the paper introduces the information greenhouse as a distinct persuasion instrument under information avoidance, extending the literature on optimal information design and medical decision-making. It offers a precise condition distinguishing pre-test warnings from committed post-test policies, which could inform empirical work on doctor communication strategies.

major comments (2)
  1. [Model primitives and commitment assumption] The distinction between warning-in-advance and the information greenhouse (abstract, paragraph on main tension) is load-bearing on the assumption that the doctor can credibly commit to a post-test disclosure rule that the patient anticipates and believes. The model must specify the commitment technology or provide a microfoundation; absent this, the patient anticipates ex-post adjustment and the greenhouse construction collapses to the warning-in-advance case.
  2. [Characterization result] The abstract states a characterization result but supplies no equations, parameter restrictions, or proof sketch. The full derivation of the optimal policy (including the precise condition separating the two regimes) must be presented with explicit primitives on the patient's avoidance cost, belief updating, and the doctor's objective to allow verification that the result does not rely on post-hoc restrictions.
minor comments (1)
  1. [Abstract] The abstract could briefly note the key modeling assumptions (e.g., common knowledge of the commitment technology) to help readers assess applicability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which help clarify the model's assumptions and presentation. We address each major comment below and outline planned revisions.

read point-by-point responses
  1. Referee: [Model primitives and commitment assumption] The distinction between warning-in-advance and the information greenhouse (abstract, paragraph on main tension) is load-bearing on the assumption that the doctor can credibly commit to a post-test disclosure rule that the patient anticipates and believes. The model must specify the commitment technology or provide a microfoundation; absent this, the patient anticipates ex-post adjustment and the greenhouse construction collapses to the warning-in-advance case.

    Authors: We agree that credible commitment is essential to sustain the information greenhouse as distinct from warning-in-advance. The model treats the disclosure rule as chosen and announced ex ante, with the patient forming beliefs accordingly; this follows the standard information-design framework. To address the concern, we will add a dedicated paragraph in the model section discussing institutional microfoundations (e.g., reputational incentives for doctors, regulatory requirements on disclosure, or repeated patient interactions) that can support commitment. We will also explicitly note the collapse to the warning case if commitment fails. This addition will be made without altering the core analysis. revision: partial

  2. Referee: [Characterization result] The abstract states a characterization result but supplies no equations, parameter restrictions, or proof sketch. The full derivation of the optimal policy (including the precise condition separating the two regimes) must be presented with explicit primitives on the patient's avoidance cost, belief updating, and the doctor's objective to allow verification that the result does not rely on post-hoc restrictions.

    Authors: The full manuscript (Section 3 and Appendix) derives the optimal policy explicitly. Primitives are defined as follows: patient utility includes an avoidance-cost term linear in the posterior probability of severe non-treatment; beliefs update via Bayes' rule given the chosen disclosure; the doctor maximizes expected patient welfare subject to the testing and treatment participation constraints. The separating condition is the comparison between the belief threshold that induces treatment under warning and the threshold that sustains testing. We will revise the abstract to state this condition concisely and add a short proof sketch to the main text for accessibility, while retaining the full derivation in the appendix. revision: yes

Circularity Check

0 steps flagged

Derivation self-contained from primitives; no circular reductions identified

full rationale

This is a standard mechanism-design theory paper that characterizes optimal disclosure policies by comparing the patient's testing and treatment incentives under different information structures. The distinction between warning-in-advance and the information greenhouse is obtained by checking when a treatment-inducing signal is compatible with the patient's participation constraint on testing; both cases are derived directly from the model's payoff primitives and the commitment assumption stated in the setup. No equations reduce a claimed prediction to a fitted input by construction, no uniqueness theorem is imported via self-citation, and no ansatz or renaming of known results is used. The result is therefore independent of its own outputs.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The model rests on standard domain assumptions from information economics about Bayesian belief updating and utility from beliefs; no free parameters or invented entities are identifiable from the abstract alone.

axioms (2)
  • domain assumption Patients form beliefs about non-treatment severity that generate avoidance when those beliefs are painful.
    Stated as the source of test avoidance in the abstract.
  • domain assumption The doctor can commit to and the patient anticipates an information policy after the test.
    Required for the greenhouse construction to be credible.

pith-pipeline@v0.9.0 · 5667 in / 1386 out tokens · 24700 ms · 2026-05-23T23:28:25.508507+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.