Quality verification in markets with reputation: Theory and experiment

Last registered on February 07, 2020

Pre-Trial

Trial Information

General Information

Title
Quality verification in markets with reputation: Theory and experiment
RCT ID
AEARCTR-0005404
Initial registration date
February 06, 2020

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
February 07, 2020, 1:39 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
BI Norwegian Business School

Other Primary Investigator(s)

Additional Trial Information

Status
In development
Start date
2019-08-18
End date
2020-11-30
Secondary IDs
Abstract
I investigate the interaction between two potential solutions to the moral hazard problem in markets for experience goods: quality verification and reputation building. I set up a dynamic quality choice model where short-lived buyers chose between a trusting and a distrusting action, and where a long-lived seller chooses between acting trustworthy or not. Through the presence of commitment behavioral types and the ability of buyers to observe the history of the seller, the seller can build and maintain a reputation. I extend this framework by letting buyers verify product quality at a cost. Contrary to what might be expected, the ability to verify product quality may hurt buyers. The reason is twofold. First, the ability to verify limits the ability of buyers to credibly commit to punishing distrusting behavior. Second, verification limits the ability of buyers to learn about the seller type from observation of previous actions. I implement a lab experiment to test the empirical content of the model.
External Link(s)

Registration Citation

Citation
Våge Knutsen, Magnus. 2020. "Quality verification in markets with reputation: Theory and experiment." AEA RCT Registry. February 07. https://doi.org/10.1257/rct.5404-1.0
Experimental Details

Interventions

Intervention(s)
I set up repeated quality choice game where buyers can observe the seller's history. I exogenous vary the opportunity of buyers to verify quality at a cost before they chose whether or not to purchase.
Intervention (Hidden)
In the experiments subjects have the rolls as buyers and sellers. A seller is matched with a sequence of buyers, one in each period. Sellers chose between high and low quality, where high quality is costly, and buyers chose whether or not to purchase. Buyers observe what the seller did in the previous period. The experiment consists of two treatments. One treatment where buyers can verify and one where they cannot verify. Sessions are run with blocks of five subjects and each session consists of 15 rounds. The 15 rounds are exactly similar. In each round four subjects have the role of buyers and one subject has the role of the seller. The seller is matched with each buyer once and in a random order. When a round is completed, a new subjects is given the role the seller, and the previous seller becomes a buyer. A given subject will be a seller in three rounds (never twice in a row), and a buyer in 12 rounds. Subjects are informed about the details of the matching procedure.
Intervention Start Date
2019-08-18
Intervention End Date
2020-11-30

Primary Outcomes

Primary Outcomes (end points)
The primary outcome measure is the role of reputation. This is measured as the share of buyers who purchase from a seller with a history of high quality less the share of buyers who purchase from a seller with a history of low quality.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Quality choice of sellers.
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
I employ a between subjects design where I vary the ability of buyers to verify product quality.

Observational unit: average behavior in blocks of 5 subjects. Subjects are kept within blocks and unique subjects are used in all treatments.

Random matching of subjects within a block in each new game.
Experimental Design Details
Randomization Method
Random matching of pairs done by computer software in the lab.
Randomization Unit
Blocks of 5 subjects.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
14 blocks
Sample size: planned number of observations
70 subjects
Sample size (or number of clusters) by treatment arms
7 blocks in each treatment.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
A pilot study has been carried out in order to calculate the sample size needed to reach a sufficient power for the tests. Data from 3 blocks for each treatment has been collected. In the treatment without verification the primary outcome measure is 0.47 (st.dev is 0.1454), and in the treatment with verification the primary outcome measure is 0.12 (st.dev is 0.1487). Based on these figures i have conducted power simulations. The simulations indicate a total sample of 70 subjects (14 blocks) yields a power above 99 percent
Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
IRB Approval Date
IRB Approval Number

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials