Asymmetric valuations in charity auctions

How to cite

Armakan, Reza. 2025. "Asymmetric valuations in charity auctions." AEA RCT Registry. AEARCTR-0016086. https://doi.org/10.1257/rct.16086-1.0

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

Reza Armakan*

* Primary investigator

Registry ID
AEARCTR-0016086
DOI
https://doi.org/10.1257/rct.16086-1.0
First registered
Last updated
Status
In development
Trial dates
2025-05-26 – 2025-11-30
Countries
Netherlands ()
Registry record
http://www.socialscienceregistry.org/trials/16086

Abstract

This project examines how auction format influences charitable giving in settings where bidders value the charitable item differently and possess common knowledge of each other’s valuations. While prior research typically finds that all-pay auctions raise more revenue than winner-pay formats like English auctions, our game-theoretic analysis predicts a reversal of this result under valuation asymmetry. Specifically, we hypothesize that English auctions may outperform all-pay auctions when bidders assign different values to the auctioned item. To test this, we designed a laboratory experiment manipulating two key factors: auction format (English vs. All-Pay) and valuation symmetry (Symmetric vs. Asymmetric). Participants engaged in bidding rounds with fixed valuations and charity framing, where a share of total revenue was split equally between bidders, regardless of outcome. Our findings aim to clarify the strategic implications of auction design in charitable contexts and provide empirical guidance for maximizing fundraising effectiveness.

Intervention

This study experimentally investigates how different auction formats—specifically English auctions and All-Pay auctions—affect charitable giving under varying valuation conditions. In each experimental session, two participants bid for an item under one of the two auction formats. The key intervention involves manipulating two factors within subjects: (1) auction format (English vs. All-Pay) and (2) valuation symmetry (Symmetric vs. Asymmetric valuations). In symmetric conditions, both bidders receive the same private valuation of the item (180 tokens); in asymmetric conditions, one bidder receives a high valuation (240 tokens) and the other a low valuation (120 tokens).

To embed the charitable context, the total revenue from each auction (either the sum of both bids in the All-Pay format or the winner’s payment in the English auction) is equally split between both participants and is explicitly framed as a contribution to a charitable cause. Participants are informed about this charitable framing prior to bidding. This intervention aims to identify whether auction format and valuation structure jointly affect the amount of money raised for charity.

Experimental design

Experimental Design (Public):

This study uses a within-subject laboratory experiment to examine how auction format and bidder valuation symmetry affect revenue in charity auctions. Each participant takes part in six auction rounds, split between two auction formats: three English auctions and three All-Pay auctions. The order of auction formats is randomized across participants.

Participants are paired in each round and assigned one of three valuation profiles: [180,180] (symmetric), [120,240], or [240,120] (asymmetric). These valuations are also randomized across rounds and are common knowledge—both bidders know each other's valuation in each round.

Auctions are framed as charity auctions: 40% of the revenue generated in each round is added to each participant’s earnings, regardless of outcome. Participants are not shown individual round results during the experiment; instead, they see their total results after all six rounds.

This design allows us to isolate the effect of auction type and valuation symmetry on charitable revenue while controlling for learning and strategic adaptation.

Design details

Randomization method
Randomization done in lab by a computer
Randomization unit
Random selection of groups (random matching) Random selection of order of auction type Random selection of order of valuations played
Clusters
100
Observations
100
Arms
100
Primary outcomes
The key outcome variables of interest in this experiment are: Revenue Generated per Auction Round Defined as: The sum of both bids in the All-Pay auction. The second-highest bid (i.e., the amount paid by the winner) in the English auction. This measures the effectiveness of each auction format in raising charitable funds. Total Revenue per Auction Type and Valuation Condition Aggregated across rounds to assess how auction format and valuation symmetry/asymmetry affect overall charitable revenue. Comparison of Revenue between Symmetric and Asymmetric Valuations To test whether symmetric valuations generate more charitable revenue than asymmetric ones, within each auction type. Bidder Behavior (Bid Amounts) Average bid amounts across auction formats and valuation types to assess how format and valuation influence strategic bidding behavior. These outcomes will be used to test the study’s main hypotheses regarding the interaction between auction type and valuation symmetry in determining the effectiveness of charity auctions.

Cited by

No citation data available for this registration yet.

Keywords