Eliciting expectation posteriors: the role of information provision about priors

How to cite

Dovern, Jonas; Schneider, Annabell. 2025. "Eliciting expectation posteriors: the role of information provision about priors." AEA RCT Registry. AEARCTR-0016435. https://doi.org/10.1257/rct.16435-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Jonas Dovern* , Annabell Schneider

* Primary investigator

Registry ID
AEARCTR-0016435
DOI
https://doi.org/10.1257/rct.16435-1.0
First registered
Last updated
Status
In development
Trial dates
2025-08-02 – 2026-05-31
Countries
Private
JEL codes
E31
Registry record
http://www.socialscienceregistry.org/trials/16435

Abstract

Many studies of macroeconomic expectations are based on RCTs embedded in surveys that elicit priors and posteriors of expectations. Treatment effects are either estimated for the difference between posteriors and priors directly or for the slope coefficient corresponding to the weight that participants attribute to their prior when forming a posterior. A recent paper (Abberger et al., 2025) proposes to inform survey participants about their previously stated priors when eliciting posteriors. This is not commonly done due to fear of anchoring effects. This project will study the effects of following their approach on estimates of participants’ expectation updating behavior in an RCT.

Intervention

In the RCT we exogenously vary whether panelists in an online household survey in Germany (conducted with Norstat) are shown their priors when we elicit their posteriors about inflation expectations (with horizons of twelve and 24 months).

Experimental design

The information/reminder about the priors is given verbally as part of the question that asks for the posterior. It is not especially highlighted.

The control group receives only the identically formulated question (without the reminder).

Design details

Randomization method
Randomization is implemented via random number generation within the implementation of the survey (in Limesurvey).
Randomization unit
Randomization is done at the level of private households.
Clusters
n.a.
Observations
The survey will cover n=3,000 households.
Arms
Approximately n=300 households will receive the treatment with reminder about the prior. Approximately n=2700 households will serve as control group.
Primary outcomes
Posterior inflation expectations (12- and 24-months horizon, in percent)

Cited by

No citation data available for this registration yet.

Keywords