Charitable asymmetric bidders

• DOI: http://doi.org/10.1111/jpet.12406
• Author: Olivier Bos
• Date: 01 April 2020
• Published date: 01 April 2020

© 2019 The Authors. Journal of Public Economic Theory published by Wiley Periodicals, Inc.

J Public Econ Theory. 2020;22:320–337.320

|

wileyonlinelibrary.com/journal/jpet

Received: 30 September 2017

|

Accepted: 12 October 2019

DOI: 10.1111/jpet.12406

ORIGINAL ARTICLE

Charitable asymmetric bidders

Olivier Bos

1,2,3

1

LEMMA, Panthéon‐Assas University,

Paris, France

2

Institut Universitaire de France (IUF),

Paris, France

3

ZEW ‐Leibniz Centre for European

Economic Research,Mannheim, Germany

Correspondence

Olivier Bos, LEMMA, Panthéon‐Assas

University, 4 rue Blaise Desgoffe, 75006

Paris, France.

Email: olivier.bos@u-paris2.fr

Funding information

European Science Foundation,

Grant/Award Number: PGPPE ‐2373;

FP6 People: Marie-Curie Actions,

EST-Early-Stage Training

Abstract

Recent papers show that the all‐pay auction is better at

raising money for charity than the first‐price auction

with symmetric bidders under incomplete information.

Yet, this result is lost with sufficiently asymmetric

bidders under complete information. In this paper, we

consider a framework on charity auctions with asym-

metric bidders under some incomplete information. We

find that the all‐pay auction still raises more money than

the first‐price auction. Thus, the all‐pay auction should

be seriously considered when one wants to organize a

charity auction.

1

|

INTRODUCTION

Fundraising activities for charitable purposes have become increasingly popular. One reason is

the growing number of nongovernmental organizations with humanitarian or social purposes.

Another one is the decrease of government participation in culture, education, and related

activities. The purpose of these associations is either the development and promotion of culture

or aid and humanitarian services. Even in France, a country without any fundraising tradition,

some organizations began to appear, such as the French Association of Fundraisers

1

in 2007.

Commonly used mechanisms to raise money are voluntary contributions, lotteries, and

auctions. Even though most of the fundraisers still use voluntary contributions,

2

auctions are

increasingly used. Indeed, for some special events or particular situations, auctions provide a

----------------------------------------------------------------------------

This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivatives License, which

permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no

modifications or adaptations are made.

1

http://www.fundraisers.fr/

2

There is further evidence of this phenomenon on the Internet with the emergence of sites, such as http://www.JustGive.org.

particular atmosphere. The popularity of auctions for charity purposes can also be observed by the

increase in Internet websites offering the sale of objects and donating a part of their proceeds to

charity. Well‐known examples include Yahoo! and Giving Works of eBay. Many others have been

created, such as the Pass It On Celebrity Charity Auction

3

in 2003, where celebrities donated objects

whose sale revenue contributed to a “charity of the month”. We can also cite cMarket Charitable

Auctions Online

4

created in 2002 and selected as a charity vehicle by more than 930 organizations.

Therefore, given the well‐developed and wide theoretical literature on altruism and

charitable fundraising (e.g., Andreoni, 1989, 1998, 2004), there is a growing and recent research

on charity auctions.

5

Goeree, Maasland, Onderstal, and Turner (2005) and Engers and

McManus (2007) investigate an independent private value’s model and show that all‐pay

auctions are better at raising money for charity than winner‐pay auctions. Moreover, Schram

and Onderstal (2009) lead a lab experiment and confirm these theoretical results. However,

Carpenter, Homes, and Matthews (2008) run a field experiment in four American preschools. In

their experiments the ranking of the revenues is reversed. They attribute this result to the

unfamiliarity of the participants to the mechanism and endogenous participation (see

Carpenter, Homes, & Matthews, 2010 for a theoretical investigation of the endogenous

participation). In addition, we can also investigate this question in a situation where people are

different in the sense that they do not have the same beliefs. Indeed, Goeree et al. (2005) and

Engers and McManus (2007) assume that bidders have the same altruism parameter and

valuations are drawn from the same distribution. Bos (2016) provides an answer with complete

information. He investigates a model with complete information and heterogeneity on the

bidders’values, and shows that when the asymmetry among bidders is strong enough, the

ranking of the expected revenues is affected. In particular, winner‐pay auctions outperform all‐

pay auctions. Damianov and Peeters (2018) confirm in a lab experiment the nonoptimality of

all‐pay auctions in a complete information framework.

The point of this paper is then to determine, whether all‐pay auctions are still better at

raising money for charity when bidders are asymmetric under some incomplete information

(provided by a uniform distribution). To do so, we compare the first‐price auction and the

(first‐price) all‐pay auction. The former is the most used sealed‐bid auction in practice and the

latter raises theoretically more money for charity than winner‐pay auctions in a symmetric

incomplete information framework. We do not consider the optimal auction determined by

Goeree et al. (2005), the lowest‐price all‐pay auction, because of its difficult implementation and

potential participants misunderstanding.

6

If we conclude that all‐pay auctions are still better

with asymmetric bidders and our specific incomplete information setting with uniformly

distributed values, we should seriously consider implementing all‐pay auctions to raise money

for charity in some environments. Indeed, to the best of our knowledge, all‐pay auctions

have never been implemented in real life for charity purposes. However, it seems easy to do.

For example, every bidder could buy a number of tickets simultaneously as in a raffle.

Contrary to a raffle, though, the winner will be the buyer with the highest number of tickets

in hand.

3

http://www.passitonline.org/

4

http://www.cmarket.com/

5

For a good understanding on prosocial behavior see Munoz‐Herrera and Nikiforakis (2019), who provide a thorough overview on the theoretical, lab, and field

experimental works of James Andreoni.

6

Difficulties occurred on the field to implement the first‐price all‐pay auction (Carpenter et al., 2008; Onderstal, Schram, & Soetevent, 2013) and should be worst

with the lowest‐price all‐pay auction.

BOS

|

321