Left-Digit Bias in Household Inflation Expectations

How to cite

Garz, Marcel; Larin, Benjamin. 2025. "Left-Digit Bias in Household Inflation Expectations." AEA RCT Registry. AEARCTR-0016243. https://doi.org/10.1257/rct.16243-1.1

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

Marcel Garz* , Benjamin Larin

* Primary investigator

Registry ID
AEARCTR-0016243
DOI
https://doi.org/10.1257/rct.16243-1.1
First registered
Last updated
Status
In development
Trial dates
2025-09-01 – 2025-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/16243

Abstract

We study a neglected mechanism in the formation of inflation expectations: a psychological phenomenon called “left-digit bias.” This bias leads some but not all households and news media to react strongly when inflation crosses salient round numbers, such as 5% and 10%. To explore this argument, we (i) computationally analyze media content, (ii) use causal designs for observational data, (iii) conduct a randomized controlled survey experiment, and (iv) introduce the bias in New-Keynesian models with heterogeneous agents. The project addresses critical gaps in the monetary economics literature and develops guidelines for policymakers regarding communication strategies and monetary policy.

Experimental design

We conduct a randomized controlled survey experiment to evaluate the effects of exposing participants to information about the inflation rate on their inflation expectations, following Fuster and Zafar’s (2023) guidelines for information provision experiments on economic expectations. The survey experiment comes in the form of an inflation forecasting exercise (Salle et al., 2023) consisting of several rounds of forecasting. In each round, participants see real information about the past inflation rate, based on which they are asked to forecast the future inflation rate. The information treatments are designed to test for left-digit bias of participants when they process numerical information about the inflation rate, as well as the impact of framing an increase in the inflation rate as inflation breaking a historical record.

Design details

Randomization method
coin flip
Randomization unit
individual
Clusters
1,500 survey participants
Observations
1,500 survey participants
Arms
500 participants control, 500 participants treatment arm 1, 500 participants treatment arm 2
Primary outcomes
Qualitative and quantitative inflation expectations in the final round of the forecasting exercise.

Cited by

No citation data available for this registration yet.

Keywords