Understanding State-Dependent Price Setting: Evidence from a Lab Experiment

How to cite

Isabelle, Salle. 2025. "Understanding State-Dependent Price Setting: Evidence from a Lab Experiment." AEA RCT Registry. AEARCTR-0015767. https://doi.org/10.1257/rct.15767-1.0

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

Salle Isabelle*

* Primary investigator

Registry ID
AEARCTR-0015767
DOI
https://doi.org/10.1257/rct.15767-1.0
First registered
Last updated
Status
In development
Trial dates
2025-04-15 – 2026-04-30
Countries
Private
JEL codes
D22; D43; E31; E70
Secondary IDs
EB-12945
Registry record
http://www.socialscienceregistry.org/trials/15767

Abstract

This individual decision-making experiment investigates how firms adjust their prices in response to cost-push shocks, such as tariffs or supply-chain disruptions. Recent inflation episodes have shown that firms change prices more frequently and by larger amounts when inflation is relatively high, but real-world data often lacks the availability and the clarity needed to study the pass-though between cost-push shocks and selling prices in detail. We design a lab experiment where participants act as firms: setting prices, forecasting future prices, and responding to shocks in a simulated market.
Participants face three types of shocks—sudden tariffs, large gradual supply shocks, and small ones— and the rest of the market is simulated with rational behavior. The experiment also tests how price-setting changes depending on the market structure (many firms vs. very few) and the type of price adjustment friction (either a cognitive or monetary menu costs). Subjects make decisions under time pressure and are paid based on profits and forecast accuracy. The design allows researchers to isolate how firms respond to different cost environments in a controlled, replicable way.

Intervention

This is an RCT with a 2-by-2 design where the treatment variables are the degree of market competition (two firms versus 100) and the nominal rigidities (a small cognitive cost or a monetary menu cost for changing the price). We randomly attribute individual participants to one of the four following treatments upon entry in the lab. Everything else (incentives, payoff, underlying model, interface) is kept the same across all treatments.

Experimental design

This is an RCT with a 2-by-2 design where the treatment variables are the degree of market competition (two firms versus 100) and the nominal rigidities (a small cognitive cost or a monetary menu cost for changing the price). We randomly attribute individual participants to one of the four following treatments upon entry in the lab.
Subjects then play multiple independent sequences of price-setting games where they will be successively confronted with a large and a small cost-push shock, a large demand shock and a Markov-process shock describing highly uncertain changes in their costs (the order of the four shocks is randomized for each subject). Only two sequences are drawn for payment, where their cumulative profits are paid out to the subjects, on top of a participation fee and the forecasting bonus points drawn out of sequence (participants also have to submit forecast of the market price). Most of the student-subjects are in the lab, but some will be online to replicate the conditions of the entrepreneur-subjects. Two treatments only out of the four will be reproduced with entrepreneurs (about 30 by treatment).

Design details

Randomization method
Randomization done by a computer.
Randomization unit
individual lab subject
Clusters
240 student-subjects (60 per treatment) and 60 entrepreneur-subjects, for a total of 300 subjects.
Observations
300 subjects
Arms
90 for two treatments (monetary menu costs and duopoly and another one chosen to be reproduced with the entrepreneurs) and 60 in each of the two other treatments.
Primary outcomes
Dynamics of individual mark-ups, frequency of price changes, sizes of price changes

Cited by

No citation data available for this registration yet.

Keywords