Advertising and attachment: exploiting loss aversion through prepurchase information

• Date: 01 December 2017
• Author: Heiko Karle,Heiner Schumacher
• Published date: 01 December 2017
• DOI: http://doi.org/10.1111/1756-2171.12208

RAND Journal of Economics

Vol.48, No. 4, Winter 2017

pp. 927–948

Advertising and attachment: exploiting loss

aversion through prepurchase information

Heiko Karle∗

and

Heiner Schumacher∗∗

We analyze a monopolist’s optimal advertising strategy when consumers are expectation-based

loss-averse and uncertain about their individual match value with the product. Advertising

provides verifiablematch value information. It modifies the consumers’ reference point and hence

their willingness to pay for the product. We show that the optimal advertising strategy pools

different consumer types so that some consumers engagein ex ante unfavorable trade.Incomplete

informative advertising thus has a persuasive effect. This provides a rationale for policies that

force the monopolist to disclose important product characteristics, not only at the point of sale,

but also in all promotional materials.

1. Introduction

The economic literature on advertising usually distinguishes between “informative” and

“persuasive” advertising.1Informative advertising increases consumers’ knowledge about a

product’s existence, characteristics, price, or quality. In contrast, persuasive advertising modi-

fies consumers’ preferences regarding the product in a way that increases their willingness to pay.

Naturally, these two views on advertising have quite different policy implications. Informative

∗Frankfurt School of Finance and Management; h.karle@fs.de.

∗∗KU Leuven; heiner.schumacher@kuleuven.be.

A previous versionof this ar ticle wascirculated under the title “Creating attachment through advertising: loss aversion and

prepurchase information.” We are grateful to Kathryn Spier (the Editor), two anonymous referees, Malin Arve, Stefan

Bechtold, Estelle Cantillon, Micael Castanheira, Dirk Engelmann, Jana Friedrichsen, Michael Grubb, Paul Heidhues,

Martin Hellwig, Fabian Herweg, Claudio Karl, Georg Kirchsteiger, Sebastian K¨

ohne, Botond K˝

oszegi, Patrick Legros,

Vardges Levonyan, Luca Merlino, Wanda Mimra, Daniel M¨

uller, Martin Peitz, R´

egis Renault, David Sauer, Nicolas

Sahuguet, Martin Schonger, Christian Schulz, Michelle Sovinsky, Johannes Spinnewijn, Roberto Weber, and Jidong

Zhou as well as various seminar audiences for their valuable comments and suggestions. Karle gratefully acknowledges

financial support from the National Bank of Belgium (Research Grant, “The Impact of Consumer Loss Aversion on the

Price Elasticity of Demand”) and the ARC Grant, “Market Evolution, Competition and Policy: Theory and Evidence.”

The usual disclaimer applies.

1Bagwell(2007) uses this classification in h is extensiveoverview of the economics of advertising. He also discusses

a third type (advertising as a complement to the product), which is not relevant for this article.

C2017, The RAND Corporation. 927

928 / THE RAND JOURNAL OF ECONOMICS

advertising can be interpreted as the markets’ welfare-enhancing response to asymmetric infor-

mation problems between consumers and firms. In contrast, persuasiveadver tising is usuallyseen

as welfare decreasing, as it drives a wedge between consumers’ preferences and their intrinsic

valuation of the product.

In this article, we show that the distinction between informative and persuasive advertising

may, in practice, be blurred. Incomplete product information may have a persuasive effect and

thus increase consumers’ willingness to pay for a product. To illustrate this idea, consider the

following example.2Imagine a potential car buyer who observes an advertisement that informs

her about the design of a particular model. She likes the design and plans to buy the car. However,

just before the purchase, she learns that the model’sfuel efficiency is not great. Nevertheless, she

follows her plan and buys the car, although she wouldnot have made such a plan had she seen all

the product details—including design and fuel efficiency—in the advertisement. The informative

advertisement has a persuasive effect, as it shows the buyer something that she likes (design),

thereby creating expectations of ownership3that alter her willingness to pay, while hiding other

relevant aspects of the product (fuel efficiency).

To demonstrate that the intuition in this example is consistent with rational expectations,

we examine a simple model of trade between a monopolist and expectation-based loss-averse

consumers (K˝

oszegi and Rabin 2006, 2007).4When a consumer decides between buying or not

buying, she compares the outcome of each decision to a reference point. This reference point is

given by her expectations regarding product ownership and monetary expenses formed after ob-

serving the monopolist’sadvertisement. For example, if the consumer expects to buy the product,

not buying creates a loss in the product dimension and a gain in the money dimension. Because

losses loom larger than gains, the expectation of ownership renders buying more attractive, and

therefore increases the consumer’s willingness to pay. Conversely, if the consumer does not ex-

pect to buy the product, buying it creates a net loss through unexpected monetary expense, which

in turn decreases her willingness to pay. Thus, the role of advertising in our model is to shift

consumer expectations in favor of trade.

Our basic model of informative advertising is taken from Anderson and Renault (2006).

Consumers differ in their match value with the monopolist’s product (i.e., their utility from

consuming the good). Ex ante, they do not know their individual match value, but they learn

it by inspecting the product at the monopolist’s store. Before inspecting the good, they see the

monopolist’s advertisement, which (potentially) provides information about individual match

values. For example, the monopolist may provide full match information so that each consumer

precisely knows her type, or incomplete information by sending the same signal to consumers

with different match values. In contrast to Anderson and Renault (2006), we assume that there

are no inspection costs. If consumers were not loss-averse, the monopolist’s advertising content

would not affect their decisions in our model: consumers would purchase the product if and only

if their match value is (weakly) above the price.

2This example is borrowedfrom Ericson and Fuster’s (2011) work on expectation-basedloss aversion and extended

to the advertising context.

3In this article, we use “expectations of ownership” and “attachment” synonymously.

4By now, there exists significant empirical evidence from the laboratory and the field that supports expectation-

based reference points. Ericson and Fuster (2011) conduct a laboratory experiment in which they randomly vary the

probability of ownership of an item. Subjects who anticipate a high probability of ownership have a higher valuation

for the item than subjects exposed to a low probability of ownership. Abeler et al. (2011) and Gill and Prowse (2012)

find similar evidence in real-effort experiments. There is also evidencethat expectation-based reference points affect golf

players’ performance (Pope and Schweitzer,2011) and New York cab drivers’labor supply (Crawford and Meng, 2011).

Karle, Kirchsteiger, and Peitz (2015) show in a field-lab experiment that a reference point formed under uncertainty

affects the consumption choice at a moment when uncertainty is fully resolved.In contrast, Smith (2012) and Heffetz and

List (2014) find no significant relationship between expectations and the valuation for a good. One explanation for their

negative results could be that the wayhow subjects form expectations varies with details of the experimental design (see

Ericson and Fuster, 2013).

C

The RAND Corporation 2017.