The Market for Lemons and Liars

How to cite

Burdea, Valeria; Li, Xueheng. 2018. "The Market for Lemons and Liars." AEA RCT Registry. AEARCTR-0003598. https://doi.org/10.1257/rct.3598-2.0

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

Valeria Burdea* , Xueheng Li

* Primary investigator

Registry ID
AEARCTR-0003598
DOI
https://doi.org/10.1257/rct.3598-2.0
First registered
Last updated
Status
Completed
Trial dates
2019-04-08 – 2019-04-30
Countries
United States of America (Pittsburgh)
Registry record
http://www.socialscienceregistry.org/trials/3598

Abstract

In this study we investigate the existence of an adverse selection of liars in the labor market. We then explore its consequences on the effectiveness of voluntary truth-telling as a solution for the problem of adverse selection in a lemons market. To do so, we design an experiment where individuals choose whether to sell a product in a cheap-talk market in which they can send an arbitrary message to the buyer, or in a verifiable-message market where they can only send truthful but possibly vague messages (Selection treatment). We then compare this treatment with one where the sellers are randomly allocated between the two markets (Random treatment). Using theoretical analysis we predict that the proportion of honest sellers in the cheap-talk market in the Selection treatment is lower than that in the Random treatment. Second, there are fewer trades in the Selection treatment than in the Random treatment, implying more inefficiencies in the Selection treatment.

Experimental design

This is a randomized control trial. The experiment involves two conditions: Selection (treatment) and Random (control). Each condition consists of two parts. In part 1 we measure each participant’s lying preference using the Gneezy’s (2005) lying task. In part 2, players are randomly assigned to groups of 3: one is the seller, one buyer A and one buyer B. Buyer A represents buyers in the cheap-talk market while buyer B those in the verifiable-message market. Then, the seller and one of the buyers play a market game based on Forsythe et al. (1999). In the Selection condition, the seller chooses which buyer, A or B, to play with, whereas in the Random condition, the seller is randomly matched with a buyer. This is repeated 20 times.

Design details

Randomization method
Randomization done in office by a computer.
Randomization unit
The unit of randomization is the individual.
Clusters
360 university students
Observations
360 university students
Arms
180 participants per treatment (60 sellers, each matched with one buyer A and one buyer B)
Primary outcomes
The primary study endpoint is the proportion of lying averse individuals across the Selection and Random treatments.
Secondary outcomes
The secondary study endpoint is the frequency of trade across the Selection and Random treatments.

Cited by

No citation data available for this registration yet.

Keywords