Push‐me pull‐you: comparative advertising in the OTC analgesics industry

• Author: Federico Ciliberto,Simon P. Anderson,Régis Renault,Jura Liaukonyte
• DOI: http://doi.org/10.1111/1756-2171.12162
• Date: 01 November 2016
• Published date: 01 November 2016

RAND Journal of Economics

Vol.47, No. 4, Winter 2016

pp. 1029–1056

Push-me pull-you: comparative advertising

in the OTC analgesics industry

Simon P. Anderson∗

Federico Ciliberto∗

Jura Liaukonyte∗∗

and

R´

egis Renault∗∗∗

We derive equilibrium incentives to use comparative advertising that pushes up own brand

perception and pulls down the brand image of targeted rivals. Data on content and spending for

all TV advertisements in Over-The- Counter (OTC) analgesics enable us toconstruct matrices of

rival targeting expenditures and estimate the structural model. Using brands’ optimal choices,

these attack matrices identify diversion ratios, from which we derive comparative advertising

damage measures. We find that comparative advertising causes more damage to the targeted

rival than benefit to the advertiser. We simulate banning comparative advertising to find industry

profits rise.

1. Introduction

We investigate how brands strategically use comparative advertising. Comparative adver-

tising both promotes positive perception of the advertiser’s brand (the “push” effect) and detracts

from the brand image of a targeted rival (the “pull” effect).1The distinctive characteristic of

∗University of Virginia; sa9w@virginia.edu, ciliber to@virginia.edu.

∗∗Cor nell University; Jurate@cornell.edu.

∗∗∗Universit´

e de Cergy-Pontoise, ThEMA; regis.renault@eco.u-cergy.fr.

We gratefully acknowledge funding of the Marketing Science Institute under MSI Research Grant no.#4-1364 and

Bankard Fund for Political Economyat the University of Virginia. Wethank Ben Leyden for excellent research assistance.

Anderson thanks the NSF for support under grant nos. SES 0452864 (“Marketing Characteristics”) and GA10704-129937

(“AdvertisingThemes”); Liaukonyte thanks Dake Family Endowment;Renault thanks the Institut Universitaire de France.

Catherine de Fontenay, JP Dub´

e, Fiona Scott-Morton, Massimo Motta, Joshua Gans, Phil Haile, Pauline Ippolito, Ginger

Jin,Matt Shum, Andrew Sweeting, and Joel Waldfogelprovided useful comments, as did participants at numerous seminars

and at the Summer Workshopin Industrial Organization (Auckland, 2009), the NBER Summer Institute Meetings (Boston,

2009), CEPR (Toulouse, 2010), 3rd Workshop on the Economics of Advertising and Marketing (Barcelona, 2010), CES

Ifo Applied Microeconomics Conference (Munich, 2011) and Marketing Science Conference (Boston, 2012). We thank

the Melbourne Business School and the Portuguese Competition Authority for their hospitality.

1The Pushmi-Pullyu is a fictitious two-headedllama befriended by Dr. Doolittle. The heads are pointed in different

directions. When one pushes forward, it pulls the other end back from its preferred direction.

C2016, The RAND Corporation. 1029

1030 / THE RAND JOURNAL OF ECONOMICS

comparative advertising that sets it apart from purely self-promotional advertising is that specific

rivals are targeted: advertising content matters. Our fundamental objectives are (i) to develop

a novel theoretical model that explains who uses comparative advertising against whom and

to what extent, (ii) to apply this model to the US OTC analgesics industry where comparative

advertisements are extensively used, (iii) to measure the damages inflicted, and (iv) to find the

consequences of a ban on comparative advertising in a simulated counterfactual.

We describe an equilibrium model, where brands simultaneously choose prices, self-

promotion, and comparative advertising expenditures, and derive the advertising first-order con-

ditions that predict oligopoly equilibrium relationships between advertising levels (for different

types of advertising) and market shares. We use these first-order conditions to rationalize the at-

tack matrix of comparative advertising spending patterns against other brands. We show that the

attack matrix identifies diversion ratios between brand pairs, where the diversion ratios measure

the fraction of a target’s lost consumers who are diverted to a rival brand followingan attack. We

employ the diversion ratios to estimate the damages inflicted by comparative attacks.

To estimate our model, we use data constructed by Liaukonyte (2015), who coded over

4000 video files of all TV advertisements in the US OTC analgesics industry for 2001–2005

and recorded which brand(s) compared themselves to which other brand(s) or products. The US

OTC analgesics industry is particularly suitable for our analysis. First, comparativeadvertising is

prevalentand represents a large fraction of industry sales. Second, data on advertising expenditures

per ad is available for all brands for a reasonable time period. Video files are available and their

content is readily coded to determine targets.

In the empirical analysis, we deal with left-censoring of advertising (in some periods, some

brands do not engage in some types of advertising—there are corner solutions) and endogeneity

of market shares and advertising expenditures. We control for left-censoring by running Tobit

regressions. Wecontrol for endogeneity by including brand fixed effects and the prices of generic

products, which we use as instrumental variables.

Our empirical findings highlight how comparative advertising is inherently different from

self-promotion. We find that outgoing attacks are about half as powerful as direct self-promotion

ads in raising the brand’s own perceived quality. However, these attacks have a strong impact in

terms of the damage that they cause to the target. This damage is heterogeneous across attacker-

target pairs. For example, a marginal dollar of comparative advertising spent by Tylenol against

Bayer reduces Bayer’sprofit by $1.3, but a marginal dollar spent by Advil against Tylenol reduces

Tylenol’s profit by $2. These losses are much larger than for pure self-promotion advertising.

For instance, if Tylenol increased its self-promotion expenditure by $1, the decrease in its com-

petitor’s profit would range between 3¢ for Excedrin and 18¢ for Advil. Hence, much of the

harm from comparative advertising comes from its negative impact on the target’s perceived

quality. We find that higher shares are associated with higher self-promotion and comparative

advertising.

Comparative advertising also has substantial positive spillovers to rivals that are not being

attacked. For example, a marginal dollar’s comparative attack by Tylenol on Aleve increases

Advil’s profit by 10¢. This means that the benefit the third party gets from denigration of the

target’s quality is larger than the loss from an improved perception of the attacker. These results

indicate substantial “free-riding” in attacking any given target.

Despite the positive spillovers, the total damage to the industry (i.e., harm to target minus

the benefits to other industry members) remains substantial. Our measures of the damage to the

target underscore the harm inflicted by comparative advertising: outgoing attacks cause much

more damage to the target than benefit to the attacker. Spillovers are too small to make up for

the difference. For example, the positive spillover to Advil of Aleve’s marginal dollar attack

on Tylenol is 33¢, whereas the damage to Tylenol is $2. These large numbers concur with the

widespread belief among industry executives that comparative advertising potentially damages

all competitors in an industry and often results in excessive levels of advertising due to persistent

C

The RAND Corporation 2016.