Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
On subscription traps and preference reversals: The pigeonholing hypothesis
Last registered on August 05, 2020


Trial Information
General Information
On subscription traps and preference reversals: The pigeonholing hypothesis
Initial registration date
August 04, 2020
Last updated
August 05, 2020 10:10 AM EDT
Primary Investigator
University of Amsterdam, Ca' Foscari University of Venice
Other Primary Investigator(s)
Additional Trial Information
In development
Start date
End date
Secondary IDs
A first pilot study finds that when the choice menu of a service provider includes a single-use subscription, it is frequently preferred over a costlier, long-term alternative. Once single-use is replaced by an extended option for the same price, fewer subjects favor it. We conjecture that single-use in the former choice set appeals to a rather rational evaluation, while in the latter comparison the alternatives are decoded along the same category - referred to pigeonholing - with the consequence that other salient comparative criteria come to the fore and fade rational evaluation into the background. A two-dimensional framework present in most behavioral models fails to explain this type of preference reversal. Based on the conception of transaction utility (Thaler, 1985) we propose a generalization of salience theory (Bordalo, Gennaioli, and Shleifer, 2012) able to capture the effect of pigeonholing.
External Link(s)
Registration Citation
Neunhoeffer, Frieder. 2020. "On subscription traps and preference reversals: The pigeonholing hypothesis." AEA RCT Registry. August 05. https://doi.org/10.1257/rct.6262-1.0.
Experimental Details
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
share of preferences respectively for the short and the long subscription in each condition
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We ask subjects about their preferences regarding two different subscriptions, that differ in price and length, in a between-subject design.
Experimental Design Details
Problem: Imagine you are visiting New York City during the Christmas break. As
last item on your agenda before flying back home, you have saved the panorama view
from the Rockefeller Center:

Treatment condition:
L_b: d_b days free entrance for p_b $
L_l: d_l days free entrance for p_l $
Control condition:
L_b': d_b'<d_b days free entrance for p_b $
L_l: d_l days free entrance for p_l $

We conjecture that a preference reversal (PR) is limited to cases that deal
with marginal utility differences of the choice alternatives. These marginal
utility differences derive from minimal price differences of the alternatives and
more importantly from an overall low likelihood to use the service again within
the subscribed time. More specifically, based on the defining properties of salience theory (Bordalo et al., 2012) we conjecture that, depending on the perceived probability of service use under the specific context of the question, a PR can occur given the following price condition, typically not otherwise:

p_l/p_b - p_b/p_l < 1, (respectively p_l/p_b < √2)

Based on the idea of pigeonholing - the influential role of the short subscription
on the perceived utility in a subscription set evaluation - let us define the
attraction (L_b|L_l) = (d_b, d_l, p_b, p_l) for the short subscription L_b relative to
L_l, expressed in share of preferences, along four properties:

1. Diminishing relative price attraction: ceteris paribus, then for any e > 1
(d_b, d_l, p_b · e, p_l · e) < (d_b, d_l, p_b, p_l)
2. Reinforcing absolute price attraction: cet par, then for any e > 0
(d_b, d_l, p_b + e, p_l + e) > (d_b, d_l, p_b, p_l)
3. Diminishing relative duration attraction: cet par, then for any e > 1
(d_b · e, d_l · e, p_b, p_l) < (d_b, d_l, p_b, p_l)
4. Diminishing absolute duration attraction: cet par, then for any e > 0
(d_b + e, d_l + e, p_b, p_l) < (d_b, d_l, p_b, p_l)
Randomization Method
Study participants are recruited via the subject pool of the CREED
laboratory at the University of Amsterdam (randomized through the in-house experiment registration system), and the crowdfunding platform
Prolific, which is specifically targeted to randomized online surveys and experiments.
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
no cluster design
Sample size: planned number of observations
200 individuals
Sample size (or number of clusters) by treatment arms
30 individuals per treatment
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
IRB Approval Date
IRB Approval Number
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)