False advertising

• DOI: http://doi.org/10.1111/1756-2171.12228
• Published date: 01 June 2018
• Date: 01 June 2018
• Author: Chris M. Wilson,Andrew Rhodes

RAND Journal of Economics

Vol.49, No. 2, Summer 2018

pp. 348–369

False advertising

Andrew Rhodes∗

and

Chris M. Wilson∗∗

There is widespread evidence that some firms use false advertising to overstate the value of

their products. We consider a model in which a policy maker can punish such false claims.

We characterize an equilibrium where false advertising actively influences rational buyers and

analyze the effects of policy under different welfare objectives. We establish precise conditions

where policy optimally permits a positive levelof false advertising and show how these conditions

vary intuitively with demand and marketparameters. Wealso consider the implications for product

investment and industry self-regulation and connect our results to the literature on demand

curvature.

1. Introduction

Buyers are often reliant on firms to obtain information about product characteristics. To

exploit this, some firms deliberately engage in what we call “false advertising”—the use of

incorrect or exaggerated product claims. They do this in a range of different contexts, and

despite potential legal penalties. Recent policy cases include Dannon, which paid $21 million to

39 US states after it misled consumers about the health benefits of its Activia yogurt products;

Skechers, which paid $40 million after falsely stating that its toning shoes helped with weightloss;

and Kellogg’s, which paid $5 million for wrongly claiming that its breakfast cereals enhanced

childrens’ immune systems. Similarly, in some related examples, car manufacturers such as

Volkswagen, Hyundai, and Kia have incurred a series of multimillion dollar penalties after

∗Toulouse School of Economics, Universityof Toulouse Capitole; andrew.rhodes@tse-fr.eu.

∗∗Loughborough University; c.m.wilson@lboro.ac.uk.

We are thankful to the Editor,David Martimor t, andtwo anonymous referees. We also thank Alexei Alexandrov, Mark

Armstrong, Mike Baye, Joao Correia-da-Silva, Daniel Garcia, Andy Gavill, Matthew Gentzkow, Thomas Jeitschko,

Justin Johnson, Bruno Jullien, Martin Peitz, Michael Powell, Robert Ritz, Jean Tirole, Tianle Zhang, and various audi-

ences, including those at the North American Econometric Society (Philadelphia, 2016), CEPR (Zurich, 2015), ESSET

(Gerzensee, 2015), IIOC (Boston, 2015), EEA (Mannheim, 2015), CREST (2015), HECER (2015), THEMA (2015),

Toulouse (2016), the Berlin IO Day(2015), the 7th Workshop on the Economics of Advertising and Marketing (Vienna,

2014), EARIE (Munich, 2015), the NIE Workshopon Advertising (Manchester, 2014), and the University of East Anglia

(2015). Weare also grateful to Kamya Buch for her research assistance. Rhodes acknowledges financial support from the

European Research Council (ERC) under the European Union’sHorizon 2020 research and innovation programme (grant

agreement no. 670494).

348 C2018, The RAND Corporation.

RHODES AND WILSON / 349

cheating tests in order to make false claims about their emission levels or fuel efficiency.1

Additional evidence also comes from academic research, which carefullydocuments the existence

of false advertising and its ability to increase demand (e.g., Cawley,Eisenberg, and Avery, 2013;

Mayzlin, Dover, and Chevalier, 2014; Zinman and Zitzewitz, 2016; and also Rao and Wang,

forthcoming, for false advertising in the form of fake user reviews).

However, despite this, the theoretical literature has largely ignored false advertising. In this

article, we develop a model where false advertising arises in equilibrium and actively influences

rational buyers. Tougher legal penalties reduce the frequency of false adverts but also increase

their credibility. As a result of the latter effect, we show when and how stronger penalties can

reduce buyer and social welfare. In particular, by using some results on demand curvature, the

article derives precise conditions on demand and market parameters such that a policy maker

optimally uses a low penalty to permit a strictly positive level of false advertising. We then

consider several wider issues, including investment incentives and industry self-regulation.

In more detail, Section 2 introduces our main model, where a monopolist is privately

informed about its product quality. Although we later extend the results in a number of ways,

we initially focus on the case where quality is either “high” or “low” and where the two types

have symmetric marginal costs. The policy maker first commits to a penalty for false advertising.

Then having learned its type, the firm chooses a price and a (possibly false) claim about its

quality. Buyers subsequently update their beliefs and make their purchase decisions, before the

policy maker instigates any penalties. We believe this setup closely approximates manyimpor tant

markets where buyers are unable to verify claims, or can only do so after a long time, and where

policy plays a key role in regulating advertising.

Section 3 characterizes an equilibrium with some appealing economic properties, in which

the high type advertises truthfully, but the low type may engage in false advertising. This equi-

librium does not resemble that of a standard price signalling game, due to the assumption of

symmetric costs. Instead, it smoothly unifies several otherwise separate cases, depending on the

level of the penalty. First, when the penalty is large, there is no false advertising. Here, as in

the standard disclosure literature, quality claims are verifiable and product quality is perfectly

revealed within a full separating equilibrium. Second, when the penalty is small, the low type

always conducts false advertising within a full pooling equilibrium. In this case, advertising is

cheap talk, and so buyers simply maintain their prior beliefs. Finally, when the penalty is moder-

ate, our equilibrium involves a novel form of partially verifiable advertising. Here, the low type

mixes between (i) pooling with the high type by engaging in false advertising with a relatively

high price, and (ii) advertising truthfully with a relatively low price. Therefore, when buyers

observe a high claim, they positively update their belief that quality is high. Hence, in contrast to

full separation, false advertising does arise in equilibrium but unlike in full pooling, advertised

claims actively inflate buyers’ beliefs beyond their priors even when false.

Section 4 analyzes how marginal changes in the penalty affecta variety of welfare measures.

Wefirst consider buyersur plus. Here, a reduction in the penalty increases the probability of false

advertising and generates two opposing effects. The first “persuasion” effect harms buyers by

prompting them to buy too many units at an inflated price. The second effect derives from the

impact of false advertising on damaging the credibility of claims. This issue goes back to at least

Nelson (1974) and is well documented empirically, with studies showing how deception lowers

credibility and reduces buyers’ purchase intentions (e.g., Newell, Goldsmith, and Banzhaf, 1998;

Darke and Richie, 2007). However, rather than viewing this impact as detrimental, we document

a beneficial “price” effect, whereby false advertising counteracts monopoly power by lowering

buyers’ quality expectations and prompting lower prices.

To compare these two effects under a relatively general form of demand, we utilize some

recent results on demand curvature and cost pass-through (Anderson and Renault, 2003; Weyland

1For further details, see www.ftc.gov/enforcement/cases-proceedings, goo.gl/yw437p; goo.gl/J1yJPx,

goo.gl/7GWYsz. Accessed October 23, 2017.

C

The RAND Corporation 2018.