Keep a record before agreement becomes contagious

Cross-Domain EchoesShared pattern · Information Cascade

A forecasting platform can look like a hundred opinions while mostly repeating its first visible estimate. A deliberative group can likewise show near-unanimity after people see one another’s choices, despite a more divided private starting point. The two illustrated procedures capture an initial judgment before public choices can dominate it. Forecasting then exposes a distribution and its reasons; a sealed-then-public vote preserves a comparison that can prompt renewed discussion. The extra route in each diagram carries the first record forward rather than letting it disappear into the public result. A discrepancy is a diagnostic cue, not proof of an information cascade: persuasion by new evidence and social pressure are different explanations that must be checked.

Written comparison

Pre-exposure capture

Collaborative forecasting

Blind initial probabilities

Deliberative voting illustration

Sealed initial leanings

The baseline must precede the relevant public influence.

The social information step

Collaborative forecasting

Distribution and reasons

Deliberative voting illustration

Visible vote sequence

Others’ choices can inform, anchor or suppress private contributions.

Keep the baseline available

Collaborative forecasting

Compare independent reads before revision

Deliberative voting illustration

Reopen doubts hidden by public agreement

The dashed route preserves evidence rather than asserting the private record is true.

What carries across

Preserve independent evidence before showing social choices. Later agreement is easier to interpret when its private starting point has not been erased.

Where the comparison stops

Conformity from social pressure and updating on genuine new evidence are not the information-cascade mechanism. Diagnose rather than label every convergence a cascade.

• A sealed first judgment may already be influenced by prior discussion; it is not pristine ground truth.
• No numerical jury result or forecasting accuracy improvement is claimed here; the source examples are illustrations.

Conditions for this comparison

• There is a specific sequential exposure channel through which others’ choices can override private information.
• Private capture actually precedes that exposure and remains available for review.

Source entries

Shared pattern

Information Cascade

Prime

Core Idea

A sequential decision pattern in which actors observe earlier actors' choices and copy them, even when their own private information suggests otherwise, as Bikhchandani, Hirshleifer, and Welch (1992) originally characterized. Each decision reinforces the apparent correctness of the path, cascading forward through a chain of independent actors.

Collaborative forecasting

Anti-Herding Interventions

Mechanism

Example

later entrants anchor on the visible mean and submit near-copies, so the "crowd" is really one loud voice echoed ninety times.

When it helps, and when it misleads

actors rationally inferring from others' visible actions until private information stops entering the aggregate at all.

Deliberative voting illustration

Sealed Ballot Before Voice Vote

Mechanism

Example

The sealed slips come back 7–5. The show of hands, taken moments later, comes in 11–1

When it helps, and when it misleads

the sequence effect in which each person, seeing others' choices, rationally discounts their own private signal until the public tally stops reflecting anyone's real information

How it works

Take the sealed individual choice before any public declaration on the same question.