forthcoming ijrm volume 31 #2 (2014)

51
2 How Much to Give? - The Effect of Donation Size on Tactical and Strategic Success in Cause-related Marketing Sarah S. Müller a,* , Anne J. Fries b , and Karen Gedenk c ========================================================== ARTICLE INFO Article history: First received in February 7, 2011and was under review for 6½ months. Area Editor: Zeynep Gurhan-Canli ============================================================ a* Corresponding author. Sarah S. Müller is a postdoc at the University of Hamburg, Max-Brauer- Allee 60, 22765 Hamburg, Germany (Phone: +49-40-42838-6433, Fax: +49-40-42838-8607, Email: [email protected]). b Anne J. Fries is a postdoc at the University of Hamburg (Email: [email protected]). c Karen Gedenk is professor of Marketing at the University of Hamburg, Max-Brauer-Allee 60, 22765 Hamburg, Germany (Phone: +49-40-42838-3748, Fax: +49-40-42838-8607, Email: [email protected]). Forthcoming IJRM Volume 31 #2 (2014)

Upload: others

Post on 11-Feb-2022

6 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Forthcoming IJRM Volume 31 #2 (2014)

2

How Much to Give? -

The Effect of Donation Size on Tactical and Strategic Success in Cause-related Marketing

Sarah S. Müllera,*

, Anne J. Friesb, and Karen Gedenk

c

==========================================================

ARTICLE INFO

Article history:

First received in February 7, 2011and was under review for 6½ months.

Area Editor: Zeynep Gurhan-Canli

============================================================

a* Corresponding author. Sarah S. Müller is a postdoc at the University of Hamburg, Max-Brauer-

Allee 60, 22765 Hamburg, Germany (Phone: +49-40-42838-6433, Fax: +49-40-42838-8607,

Email: [email protected]).

b Anne J. Fries is a postdoc at the University of Hamburg (Email: [email protected]).

c Karen Gedenk is professor of Marketing at the University of Hamburg, Max-Brauer-Allee 60,

22765 Hamburg, Germany (Phone: +49-40-42838-3748, Fax: +49-40-42838-8607, Email:

[email protected]).

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 2: Forthcoming IJRM Volume 31 #2 (2014)

3

How Much to Give? -

The Effect of Donation Size on Tactical and Strategic Success in Cause-related Marketing

Abstract

In cause-related marketing (CM), companies promise a donation to a cause every time a

consumer makes a purchase. We analyze the impact of the size of this donation on brand choice

(tactical success) and brand image (strategic success). Our results reveal different effects of

donation size on these success measures. For brand choice, the effect of donation size is

moderated by a financial trade-off for consumers, whereas the effect on brand image is

moderated by donation framing. Specifically, we show that donation size has a positive effect on

brand choice if consumers face no financial trade-off; i.e., if they do not have to choose between

triggering a donation or saving money. The effect is negative if a trade-off exists such that higher

donations come at higher costs. Brand image is enhanced by larger donations if the framing is

nonmonetary (e.g., the campaign promises the provision of vaccinations), whereas donation size

has a negative effect if donation framing is monetary (e.g., the campaign states the Euro

amount). If campaigns use a combination of both frames, the effect of donation size on brand

image has an inverted U shape. Our results suggest that CM enhances tactical and strategic

success only if firms select the right donation size, taking into account donation framing and

financial trade-offs.

Keywords: Cause-related marketing, donation size, donation framing, promotion, choice

experiment

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 3: Forthcoming IJRM Volume 31 #2 (2014)

4

1. Introduction

In a cause-related marketing (CM) campaign, Tommy Hilfiger featured a promotion in

which 50% of the price of a specific bag would be donated to Breast Health International. In

another CM promotion, Starbucks donated $1 to the Global Fund to support people living with

AIDS in Africa for every pound of East Africa Blend coffee sold. Volvic promoted its “Drink 1,

Give 10” campaign in cooperation with UNICEF, stating that for every liter of water sold, the

company would provide 10 liters of drinking water in Africa. Procter & Gamble (P&G)

promised “1 pack = 1 vaccine” in its CM promotion, in which for every promotional package

sold, the company would donate .054€ to UNICEF, equal to the cost of one vaccination against

tetanus.

In CM campaigns such as these, the firm contributes a specific amount to a cause if a

customer buys the firm’s product (Varadarajan & Menon, 1988). This transactional element is

the main characteristic of CM: The customer must make a purchase to trigger the donation.

Corporate sponsorship of social causes has become very frequent, with spendings in North

America reaching $1.86 billion in 2011 (IEG, 2011).

CM is both a tactical tool that firms employ to increase their sales and a strategic activity

aimed at improving brand image (Ross, Stutts, & Patterson, 1991). However, whether the

investment in CM always pays off is unclear. On the one hand, by triggering a donation through

their purchases, consumers might derive utility from giving, which is known as “warm glow”

(Andreoni, 1989), and thus exhibit favorable purchase behaviors. On the other hand, CM might

raise consumer skepticism about the company’s motivation because the donation is conditional

on sales and ensures the company’s own benefit (Barone, Miyazaki, & Taylor, 2000). These

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 4: Forthcoming IJRM Volume 31 #2 (2014)

5

consumer considerations can negatively impact brand image. Whether positive or negative

effects prevail depends on several success factors (Fries, 2010).

We study one key success factor, donation size, which is particularly interesting because

it is a design element that is directly controlled by managers; i.e., they can decide how much to

give when implementing CM. Campaigns vary in their donation sizes as indicated by the

introductory examples, in which donations range from 1% of the product’s price in the P&G

example to 50% of the price in the Tommy Hilfiger campaign. The effect of investing in a larger

donation is unclear. On the one hand, consumers may derive more warm glow when donation

size increases, which should make them more likely to make a purchase. A larger donation could

also produce more favorable evaluations of the brand. On the other hand, consumers who face a

CM offer with a substantial donation may prefer to receive this money for themselves or may not

believe that the company will really donate as much as promised. Thus, donation size could also

have a negative effect on sales and brand image.

Previous research has studied the influence of donation size on CM success, but the

results are equivocal. Some studies find a positive effect (e.g., Olsen, Pracejus, & Brown, 2003),

others a negative one (e.g., Strahilevitz, 1999), and others no effect at all (e.g., Human, &

Terblanche, 2012). We therefore analyze the effect of donation size on CM success in more

depth and extend previous research by focusing on the following three aspects:

First, we acknowledge that firms use CM for both tactical and strategic purposes and

therefore study two success measures: brand choice and brand image. Previous research has

rarely compared these success measures. We expect the effects of donation size on brand choice

and brand image to differ because of different underlying drivers.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 5: Forthcoming IJRM Volume 31 #2 (2014)

6

Second, we study two potential moderators of the effect of donation size on CM success

that have not been analyzed before: the presence of a financial trade-off and donation framing. A

financial trade-off occurs when consumer choose between one brand with a CM campaign and

another brand with a price promotion. We expect that such a trade-off moderates the effect of

donation size on brand choice. The framing of a CM campaign can be monetary (e.g., 5 cents),

nonmonetary (e.g., one vaccination), or a combination of both (e.g., one vaccination, worth 5

cents). We expect framing to moderate the impact of donation size on brand image.

Third, we vary our independent variable – donation size – over a wide range and in small

intervals, which allows us to test for nonlinear effects.

In a large-scale experimental survey, we systematically vary donation size and the

potential moderators, and ask respondents to make a brand choice decision and evaluate the

image of the focal brand. In an additional exploratory study, we also measure prospective drivers

underlying CM success to shed light on the differences between tactical and strategic success.

We find that the effect of donation size is different for brand choice (tactical success)

versus brand image (strategic success). The effect on brand choice is moderated by the presence

of a financial trade-off, and the effect on brand image is moderated by donation framing.

Furthermore, we find a nonlinear effect of donation size on brand image for a combined

monetary and nonmonetary framing. Finally, our exploratory analysis suggests that brand choice

is driven by warm glow, whereas brand image mostly depends on what consumers infer about

the company’s altruism and about the effectiveness of the campaign.

Our results have important implications for managers. We show that spending more

money on a larger donation does not always produce more favorable effects, but rather donation

size has to be chosen carefully, taking into account financial trade-offs and donation framing.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 6: Forthcoming IJRM Volume 31 #2 (2014)

7

Our research contributes to the CM literature by clarifying the effects of donation size:

We explain why the effect can be positive, negative, or null. In particular, we detect differences

in tactical versus strategic success. Furthermore, we investigate the moderating effects of

financial trade-offs and donation framing for the first time and provide new insights into

nonlinear effects.

We proceed as follows. In section 2, we review existing research on donation size before

presenting our conceptual framework and deriving hypotheses about the effects of donation size

on CM success in section 3. We present the research design of our experimental survey

investigating the different effects of donation size in section 4, and its results in section 5. To

gain insights into the drivers underlying tactical and strategic CM success, we report the data and

results of an additional study in section 6. We conclude by summarizing our work and discussing

its implications for both managers and researchers in section 7.

2. Literature review

Much previous research has studied the characteristics of successful CM campaigns (for

an overview, see Fries, 2010) and has identified a broad range of success factors, including the

characteristics of the cause (e.g., Ross et al., 1991), the company (e.g., Strahilevitz, 2003), the

consumer (e.g., Wymer & Samu, 2009), the non-profit organization (NPO) (e.g., Barnes, 1992),

the product (e.g., Strahilevitz & Myers, 1998), and the fit among these factors (Zdravkovic,

Magnusson, & Stanley, 2010).

A success factor that has been analyzed in several past studies is donation size. As

indicated in Table 1, the results of these studies are equivocal, spanning positive (e.g., Dahl &

Lavack, 1995; Pracejus, Olsen, & Brown, 2003/04), negative (e.g., Arora & Henderson, 2007,

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 7: Forthcoming IJRM Volume 31 #2 (2014)

8

study 3; Strahilevitz, 1999), and insignificant effects of donation size (e.g., Arora & Henderson,

2007, study 1; Vaidyanathan & Aggarwal, 2005). Although several studies incorporate

moderating effects (Table 1), these cannot fully explain the conflicting findings. The (potential)

moderators either do not influence the effect of donation size (e.g., promotion size, donation

recipient), or they merely affect the strength of a positive or negative effect (e.g., cause

involvement, price, product type), but do not change its direction.

[Insert Table 1 About Here]

We suggest three possible reasons why the effect of donation size on CM success can be

positive, negative, or null, which have not been studied systematically thus far: differences in

tactical versus strategic success, moderating and nonlinear effects. First, companies pursue two

main goals with CM: the tactical goal of increasing sales and the strategic goal of improving

brand image (Polonsky & Wood, 2001). Previous research on donation size mainly uses sales-

related dependent variables such as purchase intention and brand choice to measure tactical

success. Alternatively, a few studies analyze the effects on attitudes towards the brand to capture

strategic success. Only Arora and Henderson (2007), Holmes and Kilbane (1993), and Olsen et

al. (2003) investigate both types of success measures and find no differences between them. Yet,

a more in-depth analysis of the effect of donation size might reveal differences regarding its

impact on tactical and strategic success because the underlying drivers of the success measures

should be different. Specifically, purchase decisions should be driven mainly by the utility that

consumers derive from the campaign, whereas changes in brand image should result mostly from

the inferences consumers make about the brand offering the campaign. These distinct underlying

drivers should also cause the effect of donation size on brand choice versus brand image to be

moderated by different variables, as explained next.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 8: Forthcoming IJRM Volume 31 #2 (2014)

9

Second, two potentially relevant moderators have not been examined so far: the presence

of a financial trade-off and donation framing. Some studies on donation size have provided

respondents with decision tasks that involve choosing between a CM option, in which the money

is donated, and a non-CM option, which offers a price reduction of equal size (e.g., Arora &

Henderson, 2007, study 3; Strahilevitz, 1999). In this case, respondents face a financial trade-off:

they can either do something good by choosing the CM option or they can gain a financial

advantage for themselves by selecting the competitive offer. Other studies have not included

such a trade-off. Table 1 reveals that studies with a financial trade-off tend to find that larger

donations hurt sales, whereas most studies without a trade-off report that donation size has a

positive or no significant effect. These findings suggest that larger donations help only when

they come at no increased costs to the consumer. This is in line with the assertion of Burnett and

Wood (1988) that prosocial behavior depends on the cost of helping; forgoing a price discount

could be an important cost. Thus, differences in utility caused by financial trade-offs could

explain the equivocal effects of donation size on tactical CM success. To date, the moderating

effect of a financial trade-off has not been studied.

Another new potential moderator is the framing of the donation in monetary versus

nonmonetary terms. So far, two studies have examined donation size and framing. Olsen et al.

(2003) compare CM campaigns that present the donation as a percentage of the price versus a

percentage of the profit and find no differences in the effect of donation size between the two

frames. Chang (2008) shows that expressing a donation in absolute monetary value is more

favorable for small donations than a percent of price framing, whereas no difference exists for

large donations. Both of these studies compare different monetary frames. However, the

examples in our introduction illustrate that companies use not only such monetary frames (e.g.,

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 9: Forthcoming IJRM Volume 31 #2 (2014)

10

$1 in the Starbucks example) but also nonmonetary frames in which the donation is presented as

a charitable object or service (e.g., 10 liters of drinking water in the Volvic example), as well as a

combined framing that provides both types of information (e.g., 1 vaccine, worth .054€, in the

P&G example). Research on promotions has shown that the promotion’s value in relation to the

product’s price is assessed differently when it is framed in monetary versus nonmonetary terms

(e.g., Nunes & Park, 2003; Palazon & Delgado-Ballester, 2009). For CM campaigns, the effect

of donation size on brand image could also be affected by monetary versus nonmonetary framing

because these frames provide different information that might influence the inferences

consumers make about the company. The moderating effect of donation framing in monetary,

nonmonetary, or combined terms has not previously been examined.

Third, donation size could exert nonlinear effects on CM success. Most previous studies

investigate only two different levels of donation size and the range of donation sizes varies

across studies. So far, few studies have tested for nonlinear effects. Pracejus et al. (2003/04) find

an insignificant quadratic term. However, they only study a range of donation sizes from 0 - 10%

of the price, whereas in actual CM campaigns firms donate up to 50% of the price (e.g., Tommy

Hilfiger). Koschate-Fischer, Stefan, and Hoyer (2012) also use a quadratic term and report a

positive effect of donation size that is concave (i.e., weaker for larger donations). However, their

dependent variable is willingness to pay, i.e., they do not vary donation size in relation to the

product’s price. Finally, evidence for the nonlinear effects of donation size appears in the context

of charity auctions (Haruvy & Popkowski Leszczyc, 2009), which reveal a negative effect for

very large donations but a positive effect when a smaller fraction of the auction’s final price is

donated. However, whether the same mechanisms apply to both CM campaigns and charity

auctions is unclear. More importantly, all three studies analyze the nonlinear effects of donation

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 10: Forthcoming IJRM Volume 31 #2 (2014)

11

size for a monetary framing. As we explain in the next section, we expect a nonlinear effect for a

combined framing (monetary and nonmonetary), which has not been studied, yet.

3. Conceptual framework

3.1. Overview

Figure 1 depicts our conceptual framework. We analyze the impact of donation size on

both brand choice (tactical success) and brand image (strategic success). Furthermore, we

consider the presence of a financial trade-off (i.e., non-focal brand on price promotion) as a

moderator of the effect on brand choice, and donation framing (i.e., monetary, nonmonetary,

combination) as a moderator of the effect on brand image.

[Insert Figure 1 About Here]

We expect that the effects of donation size on brand choice versus brand image are

different and that different moderators are relevant because we propose that these success

measures are affected by different underlying drivers. More specifically, we assume the effect of

donation size on brand choice to be driven mostly by the utility that consumers derive for

themselves from the CM campaign, whereas the effect on brand image should be driven

primarily by what consumers infer about the company.

When consumers make brand choice decisions, they focus on themselves such that the

utility that they derive from the campaign is crucial. Consumer’s utility is determined by the

benefits and costs of the CM campaign and a key benefit of a CM campaign is warm glow.

Warm glow theory postulates that subjects derive utility from the mere act of giving, which is

known as “warm glow” or “moral satisfaction” (Kahneman & Knetsch, 1992). Consumers are

thus more likely to choose an option if it provides more warm glow (Andreoni, 1990). Triggering

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 11: Forthcoming IJRM Volume 31 #2 (2014)

12

a donation through a product purchase can offer warm glow to consumers (Strahilevitz & Myers,

1998), and the findings of Fries, Gedenk and Völckner (2010) support that warm glow is the

main underlying driver of the positive effect of CM on brand choice. Hence, we assume that the

choice of the CM product depends primarily on the campaign’s utility, which is provided by

warm glow, in relation to the costs of engaging in the campaign. The latter should be affected

when there is a financial trade-off for consumers, i.e., when selecting the CM option comes at the

cost of foregoing savings for oneself.

When consumers assess brand image, they focus on the company, so the inferences that

they derive about the brand from the CM campaign are crucial. Information integration theory

(Anderson, 1981) suggests that new information is incorporated into prior attitudes, resulting in

updated attitudes that reflect how the stimulus is evaluated. Thus, how consumers evaluate the

company’s engagement should affect its impact on brand image. We consider two aspects of this

evaluation as the main drivers of brand image: perceived altruism and perceived effectiveness.

Perceived altruism captures the degree to which consumers perceive the company to be

motivated by a genuine interest in supporting the charitable cause. Perceived effectiveness is the

degree to which consumers believe that the company will really donate as much as promised and

that this donation will actually reach the needy recipients. Perceived altruism and perceived

effectiveness have been studied as drivers of the effect of CM on brand choice and have been

found to be less influential than warm glow (Fries et al., 2010). We assume that they are more

important as drivers of the effect of CM on brand image because CM should positively affect

brand image only if consumers attribute altruistic motives to the company’s efforts and believe

the promises stated in the campaign.

In the remainder of this section, we build on these underlying drivers when we derive our

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 12: Forthcoming IJRM Volume 31 #2 (2014)

13

hypotheses about how donation size affects CM success.

3.2. Brand choice hypotheses

We expect the effect of donation size on brand choice to be moderated by the presence of

a financial trade-off because brand choice is driven by warm glow and the cost of giving. We do

not predict a moderating effect of donation framing for brand choice. Instead, the mere act of

triggering a donation through one’s purchase should induce warm glow regardless of the framing

of the donation. This is in line with the notion that when consumers contribute to a cause, they

are satisfied by the fact that something will be done without requiring detailed information

(Kahneman, Ritov, Jacowitz, & Grant, 1993).

Warm glow is an increasing function of what is given (Andreoni, 1989). Accordingly,

without a financial trade-off, i.e., when consumers do not have to choose between doing good

and saving money, a higher donation induces no increased costs to consumers and warm glow

and utility should thus increase if the donation rises. We therefore propose:

H1a. The effect of donation size on brand choice will be positive, if the consumer faces no

financial trade-off.

In contrast, when consumers choose between a brand with a CM campaign and another

brand with a lower price, they face a trade-off and buying the CM brand comes at a cost. Donors

are price-sensitive, and their likelihood of helping decreases as the cost of helping increases

(Burnett & Wood, 1988; Eckel & Grossman, 2003). We expect that this increase in costs

outweighs the increase in warm glow for larger donations. This is supported by the previous

research summarized in Table 1: almost all studies in which consumers face a financial trade-off

find a negative effect of donation size on purchase behavior. Hence, we hypothesize:

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 13: Forthcoming IJRM Volume 31 #2 (2014)

14

H1b. The effect of donation size on brand choice will be negative, if the consumer faces a

financial trade-off.

3.3. Brand image hypotheses

We expect that the effect of donation size on brand image is moderated by donation

framing because brand image is driven primarily by what consumers infer about the brand, i.e.,

by perceived altruism and perceived effectiveness. We do not predict a financial trade-off to be a

moderator in this context because a financial trade-off should affect consumers’ utility, but not

inferences about the brand.

When donations are framed in monetary terms, larger donations are likely to decrease

perceived effectiveness. With larger monetary donations, consumers may become skeptical that

the company will really donate this much money, and the complete amount will reach the needy

recipients. Similar effects have been shown for price promotions, where consumers do not

believe that discounts are really as large as advertised. More specifically, consumers discount

price discounts and do so increasingly as promised savings rise (Gupta & Cooper, 1992). A

similar effect is likely to occur for CM campaigns with monetary donations such that consumers

may assume that the actual donation will be lower than advertised and will therefore increasingly

discount the advertised donation which lowers perceived effectiveness as donations become

larger. Hence, we posit:

H2a. The effect of donation size on brand image will be negative, if donation framing is

monetary.

A nonmonetary frame emphasizes the output achieved with the donation and the good the

company does. Because consumers typically cannot assess the monetary value of public goods

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 14: Forthcoming IJRM Volume 31 #2 (2014)

15

(Green, Kahneman, & Kunreuther, 1994), they most likely cannot judge the value of donations

that are expressed in nonmonetary terms, such that perceived effectiveness is not affected.

However, nonmonetary donations are perceived to require more effort from companies than

monetary donations (Ellen, Mohr, & Webb, 2000), such that this frame should indicate more

sincere company motives, which is reflected in perceived altruism. Hence, when companies

make larger donations, this should be perceived as more effort and result in stronger perceived

altruism and thus better brand image. We therefore expect the following:

H2b. The effect of donation size on brand image will be positive, if donation framing is

nonmonetary.

Finally, combined frames include both monetary and nonmonetary information and

emphasize not only how much money the company donates but also the achieved output. In this

case, opposing forces are at work: On the one hand, larger monetary donations decrease

perceived effectiveness; on the other hand, larger nonmonetary donations increase perceived

altruism. To derive an effect of donation size on brand image for a combined frame, we consider

that this frame provides maximum transparency, which should affect perceived effectiveness.

Consumers prefer transparency in donation framing (Landreth Grau, Garretson Folse, & Pirsch,

2007). Hence, in the case of small to medium donations, larger donations should not negatively

impact perceived effectiveness because the combined frame reveals not only the monetary

amount but also the output. Here, the dominant effect should be that consumers perceive a larger

donation as more charitable effort by the company, thereby enhancing perceived altruism such

that brand image becomes more favorable with rising donations. However, this only works up to

a certain donation level because in the case of a very high monetary amount, consumers might

again become skeptical about perceived effectiveness. After this point, we suppose that the effect

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 15: Forthcoming IJRM Volume 31 #2 (2014)

16

of donation size on brand image is dominated by these negative inferences and becomes

negative. In summary, we expect an inverted U-shaped effect and predict:

H2c. The effect of donation size on brand image will follow an inverted U shape, if donation

framing combines monetary and nonmonetary information.

Table 2 summarizes our hypotheses.

[Insert Table 2 About Here]

4. Research design

To test our hypotheses, we conducted a between-subjects experiment based on a large-

scale survey. Different groups of respondents considered different CM campaigns, made brand

choice decisions, and assessed the image of the CM brand.

4.1. Stimuli

We constructed choice sets with two brands per product category. Participants read the

following scenario: “You would like to buy product category X. You can choose between Brand

A and Brand B. With regard to your choice, only the brand (Brand A/Brand B) is relevant, not

the depicted flavor.” We presented photos of the two brands and information about their prices

and product sizes. In the control condition, both brands appeared without a promotion. In the

treatment conditions, one brand offered a CM campaign, and the other did not. In each product

category, the CM campaign was always tied to the same focal brand.1 The design of the CM

1 Across categories, there is variance in whether the CM brand has a larger purchase frequency than the non-focal

brand (measured as number of the last three purchases before the survey), a smaller or a similar purchase

frequency.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 16: Forthcoming IJRM Volume 31 #2 (2014)

17

campaign varied across treatment groups. For the treatments with a CM offer and a financial

trade-off, the competitive brand offered a price discount of the same size as the donation.

We investigated four product categories: chocolate bars, toothpaste, beer, and detergent.

All products were fast moving consumer goods (FMCG) that varied in their price levels and

degrees of utilitarianism and hedonism to ensure robust results. For each product category, we

chose two well-known national brands and presented products that were identical in price, size,

and flavor. The prices reflected average prices found in major German supermarkets at the time

of our study. We list the brands, sizes, and prices used in the study in Appendix A.

We employed the same NPO and cause for all product categories: all CM campaigns

promised a donation to SOS Children’s Villages to support immunization against tetanus. This

well-known charity enjoys a very good reputation (German Fundraising Association, 2009), and

immunization against tetanus represents an important and uncontroversial cause. Tetanus

remains a major risk in countries with low immunization rates (WHO & UNICEF, 2010).

We systematically varied three experimental factors, as specified in Table 3. The

donation size manipulation included eight levels: 1, 2.5, 5, 10, 20, 30, 40, and 50% of product

price.2 We used these percentages to calculate the respective donation amount in Euros and the

number of vaccinations.

[Insert Table 3 About Here]

The factor donation framing had three levels: In the monetary frame, the donation amount

was presented in Euros, such as .20€. In the nonmonetary frame, the donation was stated as the

number of vaccinations, such as four vaccinations. We used a price of 5 cents per vaccination to

translate monetary into nonmonetary donations (WHO & UNICEF, 2010). The combined frame

included both the monetary amount in Euros and the equivalent number of vaccinations.

2 A systematic research of CM campaigns in Germany revealed 50% as the maximum donation size.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 17: Forthcoming IJRM Volume 31 #2 (2014)

18

All three frames were presented without a financial trade-off. In these treatments, the CM

and the competitive brand were priced equally. We also combined the monetary frame with a

financial trade-off because this frame supports an equal framing of donation size and savings. In

these treatments, the competitive brand offered a price discount equal in size to the donation

promised by the CM brand (Arora & Henderson, 2007; Strahilevitz 1999; Vaidyanathan &

Aggarwal, 2005). We provide an example stimulus for a monetary CM campaign and equivalent

competitive price promotion in Appendix B.

In our experimental set-up, the three frames without financial trade-offs and the monetary

frame with the competitive price promotion were combined with the eight levels of donation

size. We also included a control group, such that we tested 33 conditions between-subjects.

4.2. Procedure

Subjects were randomly assigned to one of the 33 conditions. They assessed up to four

product categories, although they answered questions for a category only if they had made a

purchase in that category at least once during the previous year. This filter increases the response

quality because subjects who are familiar with a category give more valid and reliable answers

(Alba & Hutchinson, 1987). The categories appeared in the same order for all respondents. For

each participant, the experimental treatment was kept constant across all categories.

To measure brand awareness, the respondents first indicated whether they were familiar

with the two brands in each product category. They then stated their brand preferences by

specifying the number of times they had bought the two brands in their last three purchases in the

category. In line with previous research (e.g., Bouten, Snelders, & Hultink, 2011; Simonin &

Ruth, 1998), we assessed brand awareness and preferences before the experimental manipulation

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 18: Forthcoming IJRM Volume 31 #2 (2014)

19

to prevent any influence that the stimulus might have on these measures.3 After the presentation

of the stimulus, the respondents made their brand choice decision and then revealed their brand

image assessments for the CM brand.4 Finally, we asked about demographic characteristics.

4.3. Measures of CM success

We used brand choice to measure tactical success and brand image to capture strategic

success. For each choice set, respondents indicated which of the two brands they would rather

buy. Next, respondents evaluated the image of the CM brand on six seven-point semantic scales

(Völckner, Sattler, & Kaufmann, 2008; see Appendix C).

4.4. Sample

We sent the questionnaire to participants in an online access panel in Germany. As an

incentive, respondents could participate in a drawing to win Amazon gift cards. Of the 1,446

respondents who answered the questionnaire between December 2008 and January 2009, 85

were excluded from the analysis due to missing values. The final data set contains 1,361

complete observations. Of the respondents, 47% are women. On average, participants are 34

years of age and live in households with 2.3 people. The majority has a monthly net household

income ranging from 1,000€ to 2,000€. More than half (61%) are members of a church, and 25%

have children. Finally, 46% of the respondents are employed, and 35% are students.

3 These measurements might make initial preferences more salient and thus affect the measures of our dependent

variables. If this were indeed the case, it would make our hypothesis tests conservative because the experimental

treatments would have less of an effect.

4 We measured brand choice before brand image because we did not want the choice decision to be biased by

consumers’ elaboration on the focal CM brand. To test if this order causes a bias in the measurement of brand

image, we counterbalanced the order of the dependent measures for one experimental treatment in our second

study. We found no evidence that our results for brand choice and brand image were affected by the order of

these two measures.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 19: Forthcoming IJRM Volume 31 #2 (2014)

20

We find significant differences across the 33 experimental groups with respect to

occupation, church membership, and parenthood (p<.05). Therefore, we include these

demographics into our models as control variables. No significant differences emerge for other

consumer demographics (i.e., age, gender, household size, income), previous donation behavior,

or brand-related variables such as brand awareness and brand preference (p>.10). Brand

awareness rates are greater than 92% for all brands used in the study, confirming that we selected

well-known brands.

For the analyses, we pool the data across the four product categories, resulting in a total

of 4,686 observations, with a minimum of 93 observations per experimental group.

4.5. Models

For brand choice, we estimate the following binary logit model, which includes

donation size and the moderators as concomitant variables:

(1) hc

hc ( V )

1P

1 e ,

(2) C C

hc hc c c hc c h hc 1 c 1

V CAT PREPREF CAT CM ,and

(3) K S

h h k kh sh shk 1 s 1

DEMO X ,

where Phc = Probability that subject h chooses the CM brand in category c,

Vhc = Systematic utility of the CM brand for subject h in category c,

CATc = Category indicator (1 if product category c, and 0 otherwise),

PREPREFhc = Stated brand preference of subject h in category c (= number of

times the CM brand was bought during the last three purchases

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 20: Forthcoming IJRM Volume 31 #2 (2014)

21

before the survey, minus the number of times the other brand was

bought on these purchases),

CMh = CM indicator (1 if subject h sees a CM campaign, and 0

otherwise),

DEMOkh = Demographic variable k for subject h, and

Xsh = CM-related concomitant variable s for subject h.

In the utility function (Equation (2)), we include category-specific intercepts and control

for stated preference heterogeneity with a PREPREF variable for each category (Ailawadi,

Gedenk, & Neslin, 1999; Horsky, Misra, & Nelson, 2006). The parameter h captures the effect

of a CM campaign on utility. It differs across subjects because different respondents receive

different experimental treatments, as described by the concomitant variables Xsh, and because

response is heterogeneous. We control for demographic variables that vary between the

experimental groups (Section 4.4.). Finally, we use a continuous mixture model to capture

unobserved heterogeneity in all parameters except those for PREPREF and the demographics,

which are household-specific. We assume that all heterogeneous parameters follow normal

distributions and estimate their means and standard deviations.

For brand image, we estimate the following linear regression model with the same

independent and concomitant variables:

(4) C C

hc hc c c hc c h hc 1 c 1

BIMAGE CAT PREPREF CAT CM ,

where BIMAGEhc = Image of subject h of the CM brand in category c.

The models for both brand choice and brand image are developed in consecutive steps.

Starting with demographics, we add the concomitant variables Xsh stepwise to check for model

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 21: Forthcoming IJRM Volume 31 #2 (2014)

22

improvements from adding moderating and nonlinear effects. The models are nested, as outlined

in Table 4. Appendix D summarizes the operationalization of the independent variables.

[Insert Table 4 About Here]

Model 1, with the main effects of donation size, donation framing, and a financial trade-

off, is our base model (Jaccard, 2001; Jaccard, Turrisi, & Wan, 1990). In Model 2, we add

interactions of donation size with donation framing and a financial trade-off. With Model 2 we

can test our hypotheses about the moderating effect of a financial trade-off (H1a and H1b) and

donation framing (H2a and H2b). To examine the nonlinear effects of donation size (H2c), we

incorporate quadratic terms in Models 3 and 4.5 To provide evidence for the significance of the

interaction effects (Jaccard, 2001; Jaccard et al., 1990), we first include a quadratic term for

donation size (Model 3), and then add quadratic terms for the interactions with donation framing

and a financial trade-off (Model 4). Although our hypotheses do not feature all possible

moderating and nonlinear effects for both brand choice and brand image, we estimate all four

models with both dependent variables to ensure that we do not miss any effects.

We estimate our models with simulated maximum likelihood (Train, 2009) using the

MAXLIK module in GAUSS.6 We test whether pooling across the product categories is

appropriate using likelihood ratio tests for the logit models of brand choice and Chow tests for

the regression models of brand image. In the Chow tests, the improvement in model fit when we

move from a pooled model to four separate models is not significant for any of the models

(p>.05). In the likelihood ratio tests, no fit improvement is significant at the 1% level; for Models

5 We also tested for thresholds by allowing the coefficient of donation size to be different below and above a

threshold in our model 2. We inserted thresholds at donation sizes of 10 and 30%, which are common for price

promotions (e.g., van Heerde, Leeflang, & Wittink, 2001). However, none of these thresholds improved model

fit, neither for brand choice nor for brand image (p>.10). Details are available from the authors upon request.

6 We rescaled donation size (by dividing it by 100) in the brand image models to facilitate the estimation

(Ailawadi, Gedenk, Lutzky, & Neslin, 2007).

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 22: Forthcoming IJRM Volume 31 #2 (2014)

23

2 and 3, the improvement is significant at the 5% level. However, with more than 4,000

observations, even small differences tend to be significant, and we find no substantive

differences for the effects of our experimental variables across the four categories. Thus, we

consider pooling to be appropriate.

5. Results

5.1. Brand choice

Table 5 contains the fit measures for our four brand choice models. Because the models

are nested, we use likelihood ratio tests to determine whether more comprehensive models offer

a significant improvement over simpler ones.

[Insert Table 5 About Here]

Model 1 includes the main effects of our experimental manipulations of donation size,

donation framing and a financial trade-off. In Model 2, we add the moderating effects of

donation framing and a financial trade-off to the effect of donation size. The likelihood ratio test

shows model fit is improved significantly, indicating that the effect of donation size on brand

choice is moderated. In Models 3 and 4, we add quadratic terms to capture the nonlinear effects

of donation size but find no significant improvements. This is in line with our predictions: we

expected nonlinear effects for brand image but not for brand choice. We rely on Model 2 to test

our brand choice hypotheses and present its parameter estimates in Table 6.

[Insert Table 6 About Here]

All coefficients for the control variables exhibit plausible signs. The positive PREPREF

coefficients (p<.01) indicate that consumers are more likely to choose the CM brand when they

preferred it over the competitive brand in their recent purchases. Respondents with children react

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 23: Forthcoming IJRM Volume 31 #2 (2014)

24

more favorably to CM campaigns (p<.01), in line with previous research (Ross, Patterson, &

Stutts, 1992). Respondents who did not indicate whether they were church members also react

more favorably to CM, but this effect is only weakly significant (p<.10).

Regarding our first hypothesis, we find that the presence of a financial trade-off

moderates the effect of donation size, as indicated by 7. In support of H1a, donation size has a

positive effect on brand choice when there is no financial trade-off, according to the significant

and positive 1, which captures the effect of donation size for a monetary frame without a

financial trade-off. As expected, we find no differences in the effect of donation size across the

three frames without financial trade-offs; neither 5 nor 6 is significantly different from zero. To

formally test H1a for the nonmonetary and combined frames, we test whether the sums of the

respective coefficients (1 + 5 and 1 + 6) differ from zero using a Wald test (Greene, 2008).

We find a weakly significant positive effect of donation size for the combined frame (p<.10), but

for the nonmonetary frame, the effect is only close to significance (p=.11). Thus, the results

support H1a for the two frames with a monetary component and without a financial trade-off.

The effect of donation size on brand choice is negative when consumers face a financial trade-

off: A Wald test shows that the sum of the coefficients 1 and 7 is significantly negative

(p<.01), thereby supporting H1b.

To provide a sense of the strength of the effects of donation size on brand choice, we

simulate the changes in brand choice probability for different donation sizes and frames. For the

simulation, we use the estimated parameter means from Model 2. We assume that a consumer

chooses between two brands that are equally preferred (category dummies and PREPREF equal

zero). For the demographic variables, we use the most frequent levels (i.e., no children, church

member, full-time occupation). We present the simulation results in Figure 2.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 24: Forthcoming IJRM Volume 31 #2 (2014)

25

[Insert Figure 2 About Here]

Figure 2 reveals that the positive impact of donation size for frames without financial

trade-offs is moderate. For example, with a monetary frame, a campaign with a donation of 1%

of the price increases brand choice probability by 14.3 percentage points (from 50% without a

campaign to 64.3%). Increasing the donation size to 20% of the price earns the firm another 5.6

percentage points in brand choice probability, which is unlikely to offset the loss in margin. For a

donation of 50% of the price, brand choice probability increases to 77.8%.

For a monetary frame with a financial trade-off, brand choice probability is 70.4% for a

1% donation but falls to 6.7% for a 50% donation and an equal competitive discount. Here, the

negative effect of donation size on brand choice is substantial. The simulation demonstrates that

for donations of up to 13.1% of the price, consumers prefer a CM campaign over a price

reduction of the same size, but when donations increase further, CM can no longer compete with

an equivalent competitive price promotion. From this point on, consumers are more attracted by

the competing firm’s discount than by the focal firm’s CM campaign. The finding that

consumers would rather have the money for themselves than donate it to a cause is in line with

research on willingness to pay for ethical products, which shows that consumers are willing to

pay only a limited premium for social attributes (Auger, Devinney, Louviere, & Burke, 2008).

In summary, our results suggest that the moderating effect of a financial trade-off can

explain most of the equivocal findings on the effect of donation size on tactical CM success in

previous research (Table 1). We find that the effect of donation size is positive if consumers face

no financial trade-off, but becomes negative when larger donations induce higher costs to

consumers.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 25: Forthcoming IJRM Volume 31 #2 (2014)

26

5.2. Brand image

Table 5 presents model fit for the regression models with brand image as the dependent

variable. We use likelihood ratio tests to test for improvements in fit in our hierarchy of nested

models. Model 1 includes demographics and the three experimental factors as concomitant

variables. In Model 2, we add interaction effects of donation size with donation framing and a

financial trade-off, and find that model fit improves significantly. Thus, the moderators affect the

impact of donation size on brand image. Next, we add nonlinear effects of donation size, but the

incorporation of a quadratic term for donation size in Model 3 does not improve model fit.

Hence, the effect of donation size on brand image is not nonlinear per se. Model 4 includes

quadratic terms for the interactions of donation size with donation framing and a financial trade-

off. In Model 4, though, we encounter problems with multicollinearity; thus, we exclude the

quadratic terms DONSIZ2 NONMON and DONSIZ

2 TRADE-OFF from the model.

7 The

reduced Model 4* represents a significant improvement over Model 3. We therefore use Models

2 and 4* to test our hypotheses and list their parameter estimates in Table 6. Again, the

parameters for all control variables have plausible signs: Previous preferences relate positively to

brand image (p<.01), and the effect of CM on brand image is more favorable for respondents

with children in Model 2 (p<.10).

In Model 2, we find support for H2a: The impact of donation size is negative when the

frame is purely monetary, regardless of whether the competitive brand is on promotion or not.

Specifically, the significant negative coefficient 1 shows that the effect of donation size is

negative for a monetary frame without a financial trade-off, and the interaction with a financial

trade-off (7) is not significant. A t-test further reveals that the sum of 1 and 7 is significantly

7 This modification does not limit our insights. In several alternative models, we find no significant parameters for

the terms we exclude, and the nonlinear effect of donation size for a combination frame remains stable.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 26: Forthcoming IJRM Volume 31 #2 (2014)

27

negative (p<.05), providing formal support for H2a. The sum of 1 and 5 is significantly

positive (p<.05), which supports H2b: The impact of donation size is positive when the frame is

purely nonmonetary.

Finally, we test for a nonlinear effect of donation size for a combined frame, using Model

4*. With respect to the expected inverted U shape, we use t-tests pertaining to the sum of the

coefficients for donation size and its interaction with the combination frame (1 + 6), as well as

the sum of the two quadratic terms (8 + 9) (Jaccard et al., 1990). The sum of 1 and 6 is

positive and significantly different from zero (p<.05). The sum of 8 and 9 is weakly significant

and negative (p<.10). That is, the effect of donation size follows an inverted U shape for a

combined frame, and H2c is supported.

To illustrate the effects of donation size on brand image for the different frames and to

assess the strength of the effects, we again run a simulation. We use the estimated parameter

means from Model 4*, and the same assumptions about PREPREF and demographics as in the

brand choice simulation. Figure 3 presents the results.

[Insert Figure 3 About Here]

Figure 3 reveals the negative effect of donation size for monetary frames, the positive

effect for the nonmonetary frame, and the inverted U-shaped effect for a combined frame. When

a CM campaign presents both monetary and nonmonetary information, donation size first has a

positive effect on brand image and then a negative one. The turning point is reached at a

donation of 35.3% of the price – well within the range of realistic donation sizes.

CM with large monetary donations (> 15.9% of the price) and very small donations with

a combined frame (< 6.9% of the price) hurt brand image. The former finding is in line with our

reasoning that high monetary donations might lower consumers’ perceived effectiveness. The

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 27: Forthcoming IJRM Volume 31 #2 (2014)

28

latter indicates that with a combined frame, transparency is counterproductive in the case of very

small donations. Revealing the exchange ratio between money and the charitable object

demonstrates how few resources are necessary to achieve a considerable outcome. In turn,

consumers likely perceive very small combined donations (e.g., 2 cents equaling 0.4 vaccinations

in our study) as paltry, which results in lower perceived altruism.

Overall, we find changes between +.37 and –.12 on a seven-point scale; brand image in

the control group was 4.95. Given that the well-known brands in our study possess established

images that are unlikely to change much because of a single experimental treatment, these effects

are substantial. In summary, Figure 3 shows that donation size can have substantial effects on

brand image and highlights the importance of considering donation framing when deciding on

the size of the donation.

6. Underlying drivers

To derive our hypotheses about the effects of donation size on tactical and strategic

success, we relied on different underlying drivers. In a second study, which is exploratory in

nature, we collected data on these potential underlying drivers, and analyzed how they affect

brand choice and brand image. For this purpose, we regressed the two success measures on warm

glow, perceived altruism, and perceived effectiveness.

6.1. Data

We included a subset of the stimuli from our large-scale survey. In contrast to our main

study, we varied product category between-subjects because of the longer questionnaire, which

now included measures on potential underlying drivers. To keep the number of experimental

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 28: Forthcoming IJRM Volume 31 #2 (2014)

29

groups tractable, we used fewer donation sizes (2.5, 10, 30 and 50% of product price) and only

two product categories (chocolate bars and toothpaste). We employed the same frames as in our

large-scale study (monetary, nonmonetary, and combined without a financial trade-off and

monetary with a competitive price promotion). The three frames without financial trade-offs and

the monetary frame with the financial trade-off were combined with the four levels of donation

size. We employed all combinations for the two product categories, resulting in 32 conditions,

which we varied between-subjects.

The procedure and measurements for the success variables were the same as in our first

study. In addition, we included measures to capture the underlying drivers. After making their

brand choice decisions and brand image assessments, participants indicated their warm glow,

perceived effectiveness of the campaign and the perceived altruism of the company running the

campaign (all on seven-point multi-item scales, see Appendix C).8

We invited members of an online access panel in Germany to participate in our survey.

As an incentive, respondents could participate in a drawing to win Amazon gift cards. Between

August and October 2011, 1,402 respondents answered the questionnaire. We excluded 34

participants due to a response time of less than 2.5 minutes (the mean was 7.8 minutes) or

because they clicked-through all multi-item scales (straight line response on all scales). The final

data set contains 1,368 complete observations. Among our respondents, 62% are women. On

average, they are 33 years of age and live in households with 1.2 people. The most respondents

have a monthly net household income between 1,000€ and 2,000€. More than half (60%) are

church members, and 25% have children. 55% of the respondents are employed, and 31% are

students. Thus, the sample’s demographics are very similar to those of our main study. We do

8 We also measured self-sufficiency and ease of imagination of the donation as additional potential drivers to test

for alternative explanations. Because these did not prove to be relevant, we excluded them from the analysis. Full

results are available from the authors upon request.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 29: Forthcoming IJRM Volume 31 #2 (2014)

30

not find significant differences between the experimental groups on any consumer demographics

(i.e., age, gender, household size, income, church membership, parenthood), previous donation

behavior or brand-related variables such as brand awareness and brand preference (p>.10).

6.2. Results

Our hypotheses are based on the reasoning that brand choice is driven by warm glow, and

brand image is driven by perceived altruism of the firm and perceived effectiveness of the

campaign. Therefore, we regressed both success measures on these three potential drivers. We

also included category-specific intercepts and controlled for stated preference heterogeneity with

a PREPREF variable for each category.9 A Chow test for the regression model of brand image

and a likelihood ratio test for the logit model of brand choice indicate that pooling across the

product categories is appropriate (p>.10). All variance inflation factors are below 1.68, indicating

no problems with multicollinearity.

The estimation results are displayed in Table 7. For the brand choice model, fit is good,

as indicated by the value of .328 for Nagelkerke’s R2. In line with our reasoning, the only

significant driver of brand choice is warm glow, which is crucial for the utility consumers derive

from a CM campaign. In contrast, perceived altruism and effectiveness do not disclose

significant effects.

[Insert Table 7 About Here]

For brand image, the linear regression’s R2 value of .159 is satisfactory, given that we

study the same well-known brands as in our main study, for which images have been formed

over a long time in the consumer’s mind. In line with our reasoning, perceived altruism and

perceived effectiveness both exert a significant positive effect on brand image. Warm glow also

9 Since our dataset contains only one observation per respondent, we do not model unobserved heterogeneity.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 30: Forthcoming IJRM Volume 31 #2 (2014)

31

has a significant effect, which most likely represents a spillover of the good feeling consumers

experience through the campaign onto the brand’s image. However, the coefficient for warm

glow is the smallest. Thus, the primary drivers of brand image are perceived altruism and

perceived effectiveness, which both relate to what consumers infer from CM about the brand.

In summary, this study suggests that in CM, brand choice and brand image are indeed

affected by different underlying drivers. For brand choice, the utility the consumer derives from

the campaign is crucial, and this is determined by the warm glow the campaign triggers. In

contrast, brand image is mainly affected by the inferences consumers make about the brand

involved in the campaign. It is critical for consumers to believe in the company’s sincere motives

and that the donation will be used as promised.

7. Summary and implications

We have investigated the impact of donation size on the effect of CM on brand choice

and brand image in a large-scale experimental survey with different product categories. An

additional exploratory study provides insights into the underlying drivers of consumer behavior

in the context of CM. Our key findings are the following:

The effect of donation size on brand choice depends on the presence of a financial

trade-off. If consumers face no trade-off, larger donations increase brand choice

probability. However, if consumers have to choose between doing good and savings

for themselves, larger donations and larger trade-offs respectively will decrease brand

choice probability.

The effect of donation size on brand image depends on donation framing. With a

monetary frame, larger donations are less favorable for brand image and CM

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 31: Forthcoming IJRM Volume 31 #2 (2014)

32

campaigns with high monetary donations can even hurt brand image. For a purely

nonmonetary frame, donation size has a positive effect on brand image, and for a

combined frame, the effect follows an inverted U shape.

In CM, tactical and strategic success appear to be driven by different mechanisms.

Brand choice depends on consumers’ utility and thus warm glow is the crucial driver.

Brand image improves when consumers make positive inferences about the company

and thus is mainly affected by the perceived altruism of the company and the

perceived effectiveness of the campaign.

Our systematic analysis of donation size helps explain why donation size can have

positive, negative, or no effects on CM success. First, we show that tactical success (brand

choice) and strategic success (brand image) are affected differently. This is attributed to

differences in their underlying drivers.

Second, we identify two new moderators of the effect of donation size on CM success.

For brand choice, larger donations exert a favorable effect as long as consumers face no financial

trade-offs, whereas the effect is negative when an alternative brand offers a price promotion of

equal size. This finding goes a long way toward explaining the contradictory previous results

summarized in Table 1. Except for Chang (2008) all studies that find a negative impact include a

financial trade-off (e.g, Arora & Henderson, 2007; Strahilevitz, 1999) while in none of the

studies reporting a positive effect a larger donation comes at higher cost to consumers (e.g.,

Koschate-Fischer, Stefan, & Hoyer, 2012; Pracejus, Olsen, & Brown, 2003/04). For brand

image, the effect of donation size is moderated by donation framing. Through the introduction of

this moderator, we extend the scope of previous research on donation size, which has studied

only monetary frames.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 32: Forthcoming IJRM Volume 31 #2 (2014)

33

Third, we find that the effect of donation size is nonlinear for a frame that combines

monetary and nonmonetary information: It follows an inverted U shape. So far, nonlinear effects

have only been studied for monetary frames (Koschate-Fischer, Stefan, & Hoyer, 2012; Pracejus

et al. 2003/04). Our results support the findings of Pracejus et al. (2003/04) in that we do not find

nonlinear effects for monetary frames. We note that Koschate-Fischer et al. (2012) find a

concave effect for a monetary frame, but with a different success measure, i.e., willingness-to-

pay. Their result may simply reflect consumers’ reluctance to pay more for larger donations,

whereas in our study product price remained constant even with larger donations.

Our results suggest important implications for managers who intend to use cause-related

marketing. First, large donations are not essential for tactical success. CM can be a cost-effective

sales promotion instrument to increase brand choice because even small donations have a

substantial positive impact. In contrast, most price promotions must pass a 10 - 20% discount

threshold to significantly affect purchase intentions and behavior (Gupta & Cooper, 1992; van

Heerde, Leeflang, & Wittink, 2001). Rising CM donations increase brand choice probability only

moderately, which is most likely not sufficient to offset their additional costs.

Second, large CM donations may not be able to compete in a promotion-intensive

environment in which consumers face trade-offs between doing something good and savings for

themselves. In many FMCG categories, price discounts are in the range of approximately 20% of

the product price (e.g., van Heerde, Leeflang, & Wittink, 2000). Donating that much money to a

cause would increase brand choice probability but not enough to maintain market share when the

competitor offers an equivalent price promotion. However, van Heerde, Leeflang, and Wittink

(2004) observe huge variations in price discounts, ranging from 5% to 51% of the price. Against

smaller discounts, CM is likely to prevail.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 33: Forthcoming IJRM Volume 31 #2 (2014)

34

Third, larger donations can help or hurt strategic success depending on their framing.

When CM campaigns use a monetary frame, the effect of CM on brand image becomes less

favorable with increasing donations and can even turn negative. In contrast, donation size has a

positive effect on brand image for campaigns with nonmonetary framing. Finally, managers may

combine monetary and nonmonetary information in their donation framing. In this case, a

medium donation size is optimal for brand image. Therefore, managers should carefully align

donation size and donation framing in CM to create a positive effect on strategic success. If they

succeed in this task, CM may be an attractive alternative to price promotions, which – even if

possibly more effective in the short run – typically hurt brand loyalty in the long run (Neslin &

van Heerde, 2009). CM, in contrast, may help managers to increase consumers’ brand loyalty by

adding a philanthropic component to their brand.

Finally, managers should pay attention to their communication in CM campaigns. They

should appeal to the warm glow consumers derive from participating in the campaign to enhance

tactical success. At the same time, to improve brand image, they need to make a credible claim

that the company is committed to help the cause and that the donations reach their targets.

Our study also has some limitations that provide opportunities for further research. Our

measures may suffer from a social desirability bias (e.g., Lautenschlager & Flaherty, 1990;

Nancarrow, Brace, & Wright, 2001), such that respondents might be more likely to choose the

CM brand and evaluate its brand image more favorably than they would in a real purchase

situation. Even if our data suffered from this social desirability bias, though, it would be unlikely

to affect our results regarding donation size because the bias would be the same for all

experimental groups. However, our measures of the absolute effect of CM would be biased,

which would make it challenging to derive specific implications for the optimal donation size.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 34: Forthcoming IJRM Volume 31 #2 (2014)

35

We suspect that our data are not affected by a strong social desirability bias because we find

negative effects of CM on brand image and a strong effect of a competitive promotion on brand

choice. Nevertheless, it would be worthwhile to validate our results with field data.

Furthermore, we study only fast moving consumer goods. Investigating the impact of

donation size and its two moderators for durable goods might lead to further insights. For

example, the effect of a financial trade-off might be different for higher priced products.

Previous research has indicated that CM is more successful for hedonic as compared to

utilitarian products (e.g., Strahilevitz, 1999; Strahilevitz & Myers, 1998). The proposed

underlying mechanism is that CM reduces customers’ guilt associated with indulging in hedonic

products. We find no differences for the effects of donation size between hedonic and utilitarian

products, maybe because in our scenarios we explicitly told consumers that they want to buy a

product, such that their purchase incidence decision was already made. Further research should

look more deeply into the role of category type and test whether the feeling of guilt indeed plays

a role.

Another limitation of our study is that brand image is formed over a long period of time,

such that it could be interesting to validate our results with a study that repeatedly measures

brand image over a longer time period. This would also present an opportunity to measure how

improvements in brand image translate into brand equity and affect purchase decisions in the

long run (Keller, 1993).

Finally, we have studied a specific type of financial trade-off where the competitive brand

offers a price promotion. Our results are interesting for firms who consider using CM to compete

in a promotion-intensive environment. Future research may want to study the trade-offs that

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 35: Forthcoming IJRM Volume 31 #2 (2014)

36

occur when the CM brand increases its price. The respective results might help firms decide

whether they can pass on the costs of the donations to consumers.

Despite these limitations, we think that our study yields interesting results with important

implications for managers and researchers, and we hope that further research will build on it.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 36: Forthcoming IJRM Volume 31 #2 (2014)

37

Table 1

Research on donation size

Study Donation size†

Financial

trade-off Dependent variable Result Moderating effects

Dahl & Lavack

(1995)

$0.0025 vs. $0.1 no Perceived exploitation of NPO

Product appeal

+ Promotion size: n.s.

Garretson Folse,

Niedrich, & Landreth

Grau (2010)

0.13 - 32% of price no CM participation intentions +

1.88 - 67.5% of price no CM participation intentions +

2.5 - 40% of price no CM participation intentions +

Holmes & Kilbane

(1993)

0 - 6.8% of price no Attitude toward ad + Price: n.s.

Koschate-Fischer,

Stefan, & Hoyer

(2012)

0 - 40 cents no Willingness to pay + Attitude toward helping: +

Warm glow: +

Cause involvement: +

Cause organization

affinity: +

5 vs. 40 cents no Willingness to pay + / n.s. Fit: + / n.s.

5 vs. 40 cents no Willingness to pay + / n.s. Fit: + / n.s.

5 vs. 40 cents no Willingness to pay + / n.s. Fit: + / n.s.

Olsen, Pracejus, &

Brown (2003)

1 vs. 10% of price no Attitude toward ad

Attitude toward brand

Purchase intention

+ Framing (% of price vs. %

of profit): n.s.

Pracejus, Olsen, &

Brown (2003/04)

0 - 10% of price no Brand choice +

Smith & Alcorn

(1991)

$0.1; $0.25; $0.4 no Intention to use coupon +

Arora & Henderson

(2007), study 3

1 vs. 5% of monthly

credit card charge

yes Brand choice

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 37: Forthcoming IJRM Volume 31 #2 (2014)

38

Chang (2008) 5 vs. 25% of price no Behavioral intention Price: -

Strahilevitz (1999) 5 vs. 50% of price yes Brand choice

1 vs. 25% of price yes Brand choice

1 vs. 25% of price yes Brand choice Product type: + (hedonic)

Subrahmanyan (2004) 1 - 20% of price yes Purchase likelihood

Arora & Henderson

(2007), study 1

0 - 45% of price no Brand choice

Purchase likelihood

Attitude toward brand

n.s.

Fries, Gedenk, &

Völckner (2010)

5 vs. 15% of price no Brand choice

n.s.

Holmes & Kilbane

(1993)

0 - 6.8% of price no Attitude toward store

Intention to respond

n.s. Price: n.s.

Human & Terblanche

(2012)

$0.18 vs. $1.14 no Attitude toward cause alliance

Attitude toward campaign

CM participation intentions

n.s. Donation recipient: n.s.

Vaidyanathan &

Aggarwal (2005)

6.3 vs. 12.5% of price yes/no Willingness to buy n.s.

Van den Brink,

Odekerken-Schröder,

& Pauwels (2006)

0.1 vs. 25% of price no Brand loyalty n.s.

Notes: + = positive effect; - = negative effect; n.s. = no significant effect; †

= donation sizes were transformed into % of price when possible.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 38: Forthcoming IJRM Volume 31 #2 (2014)

39

Table 2

Hypotheses

Brand choice Brand image

Effects of donation size

H1a +, if no financial trade-off H2a , if donation framing monetary

H1b , if financial trade-off H2b +, if donation framing nonmonetary

H2c ∩, if donation framing combined

Notes: + = positive effect; - = negative effect; ∩ = inverted U-shaped effect.

Table 3

Experimental factors

Factor Level Realization

Donation size 1 / 2.5 / 5 / 10 / 20 / 30 / 40 / 50%

of product price

Converted into Euro amount and/or

number of vaccinations

Donation framing Monetary Amount in Euros

Nonmonetary Number of vaccinations

Combined Amount in Euros and equivalent

number of vaccinations

Financial trade-off

(only in

combination with

monetary frame)

Present Competitive brand offers price

discount of equal size as donation

by CM brand

Not present No competitive promotion

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 39: Forthcoming IJRM Volume 31 #2 (2014)

40

Table 4

Model specification

Variables Model 1 Model 2 Model 3 Model 4

Category variables

CAT_CHOCOLATE

CAT_TOOTHPASTE

CAT_BEER

CAT_DETERGENT

PREPREF_CHOCOLATE

PREPREF_TOOTHPASTE

PREPREF_BEER

PREPREF_DETERGENT

Demographic variables

Child_YES x CM

Church membership_YES x CM

Church membership_ NO_RES x CM

Occupation_FULL x CM

Occupation_PART x CM

CM variables

CM

DONSIZ

NONMON

COMBI

TRADE-OFF

DONSIZ x NONMON

DONSIZ x COMBI

DONSIZ x TRADE-OFF

Nonlinear effects

DONSIZ2

DONSIZ2

x COMBI

DONSIZ2

x NONMON

DONSIZ2

x TRADE-OFF

Notes: x = interaction effect; = included in the model.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 40: Forthcoming IJRM Volume 31 #2 (2014)

41

Table 5

Model fit and improvement

Model Model 1 Model 2 Model 3 Model 4 Model 4*

Brand choice

Log likelihood -2,360.015 -2,318.506 -2,317.138 -2,314.786

Likelihood ratio test:

Reference model Model 1 Model 2 Model 3

Chi2 83.017 2.737 4.705

(p) (<.001) (.255) (.582)

Brand image

Log likelihood -6,329.615 -6,317.384 -6,316.166 -6,312.979

Likelihood ratio test:

Reference model Model 1 Model 2 Model 3

Chi2 28.210 2.437 6.373

(p) (<.001) (.296) (.041)

N = 4,686

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 41: Forthcoming IJRM Volume 31 #2 (2014)

42

Table 6

Parameter estimates for brand choice and brand image models

CM success measure

Independent variables

Brand choice Brand image

Parameter estimates (standard errors) Model 2 Model 2 Model 4*

Category variables† Mean SD Mean SD Mean SD

β1 PREPREF_CHOCOLAT 1 ( .752*** .317)

( .204*** .021)

( .203*** .021)

β2 PREPREF_TOOTHPASTE 1 ( .295*** .213)

( .195*** .016)

( .197*** .016)

β3 PREPREF_BEER 1 ( .791*** .270)

( .395*** .030)

( .396*** .031)

β4 PREPREF_DETERGENT

( .921*** .082)

( .230*** .019)

( .231*** .019)

Concomitant variables††

γ Constant

( .423 .311)

( .832** .293)

( .124 .123)

( .101 .308)

( .156 .134)

( .126 .104)

η1 Child_YES

( .352*** .135)

( .080* .048)

( .070 .047)

η2 Church membership_YES

( .103 .122)

( .007 .044)

( .014 .043)

η3 Church membership_NO_Res

( .580* .308)

( .064 .106)

( .052 .105)

η4 Occupation_FULL

( .051 .138)

- ( .024 .049)

- ( .027 .049)

η5 Occupation_PART

( .093 .163)

- ( .043 .058)

- ( .045 .058)

1 DONSIZ

( .014** .006)

( .005 .014)

- ( .540** .218)

( .500* .258)

-1 ( .203** .581)

( .253 .377)

2 NONMON ( .273 .235) (1

.219

.224) - ( .101 .086)

( .298*** .091)

- ( .092 .090)

( .230** .113)

3 COMBI

( .148 .225)

( .267 .703)

- ( .176** .080)

( .051 .105)

- ( .301** .105)

( .011 .114)

4 TRADE-OFF

( .362 .255)

1 ( .311*** .332)

- ( .026 .078)

( .374*** .080)

- ( .032 .101)

( .332*** .096)

5 DONSIZ x NONMON - ( .002 .009)

( .001 .019)

1 ( .122*** .348)

( .162 .966)

1 ( .105** .371)

( .488 .578)

6 DONSIZ x COMBI - ( .002 .009)

( .009 .018)

1 ( .314*** .320)

( .721** .316)

3 (1

.525**

.097)

( .613 .509)

7 DONSIZ x TRADE-OFF - ( .085*** .014)

( .020 .017)

( .027 .306)

( .010 .434)

( .021 .438)

( .196 .323)

8 DONSIZ2

1 (1

.413

.096)

( .396 .763)

9 DONSIZ2

x COMBI -4 (2

.703**

.185)

(1 .957 .536)

Notes: N = 4,686; * p<.10; ** p<.05; *** p<.01 (two-sided); SD = Standard deviation; Standard errors in parentheses; † Category constants available upon request; †† Donation size rescaled (divided by 100) for brand image models.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 42: Forthcoming IJRM Volume 31 #2 (2014)

43

Table 7

Results underlying drivers

CM success measure

Independent variables

Brand choice Brand image

Parameter estimates (standard errors)

Category variables†

PREPREF_CHOCOLATE 1 .190*** (.103) .162*** (.026)

PREPREF_TOOTHPASTE .704*** (.092) .190*** (.030)

Underlying drivers

Warm glow .393*** (.053) .046** (.019)

Perceived effectiveness .091 (.062) .067*** (.023)

Perceived altruism - .027 (.064) .111*** (.023)

Nagelkerke’s R2 .328

Chi2 352 .253

(p) (< .001)

Log likelihood -631 .189

R2

(Adj. R2) .159 (.155)

F 42 .741

(p) (< .001)

Notes: N = 1,368; * p<.10; ** p<.05; *** p<.01 (two-sided);

Standard errors in parentheses; † Category constants available upon request.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 43: Forthcoming IJRM Volume 31 #2 (2014)

44

Figure 1

Conceptual framework

Donation size

Donation framing

Tactical CM success:

Brand choice

Financial trade-off

Strategic CM success:

Brand image

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 44: Forthcoming IJRM Volume 31 #2 (2014)

45

Figure 2

Change in choice probabilities through a CM campaign

Figure 3

Change in brand image through a CM campaign

-0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0 10 20 30 40 50

∆ C

hoic

e pro

bab

ilit

ies

Donation size (% of price)

Nonmonetary

Combined

Monetary

Monetary

with financial

trade-off

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0% 10% 20% 30% 40% 50%

∆ B

rand I

mag

e

Donation size (% of price)

Nonmonetary

Combined

Monetary

Monetary

with financial

trade-off

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 45: Forthcoming IJRM Volume 31 #2 (2014)

46

Appendix A

Product stimuli

Product category Brand Size Price

Chocolate bars KitKat 200g 1.99€

Duplo

Toothpaste Colgate 75ml 1.99€

Odol-med3

Beer Bitburger 24 x 0.33l 12.49€

Warsteiner

Detergent Persil 4.75kg 12.49€

Ariel

Notes: Italics = CM brand.

Appendix B

Example stimuli for monetary CM campaign and competitive price promotion

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 46: Forthcoming IJRM Volume 31 #2 (2014)

47

Appendix C

Multi-item scales

Measures Source

CM Success

Brand image

Rating the CM brand on Likert scales:

-3 = bad and +3 = good

-3 = not likeable and +3 = likeable

-3 = low quality and +3 = high quality

-3 = not trustworthy and +3 = trustworthy

-3 = unpleasant and +3 = pleasant

-3 = unattractive and +3 = attractive

Völckner,

Sattler, &

Kaufman,

2008†

.95

(S1)

.92

(S2)

Underlying drivers (Study 2)

Warm glow

Extent to which participants agreed/disagreed with the following

statements:

When I purchase [CM brand name], I feel good because I do

not only spend money for myself but also for other people.

I feel comfortable if I donate for a good cause by purchasing

[CM brand name].

I am pleased that I do not only get a product by purchasing

[CM brand name], but that I also do a good deed at the same

time.

Arora &

Henderson,

2007;

Andreoni,

1989; Fries,

Gedenk, &

Völckner,

2010; Monin,

2003

.93

Perceived effectiveness of CM campaign

Extent to which participants agreed/disagreed with the following

statements:

I believe that the donated money reaches the needy persons.

I am convinced that little of the donated money is wasted.

I assume that the donated money will be distributed in favor of

the cause.

I trust in the fact that the donated money will be used for the

cause.

I believe that the company actually donates as much as stated in

the CM campaign.

Fries, Gedenk,

& Völckner,

2010; Sargeant

& Lee, 2004;

Webb, Green,

& Brashear,

2000

.93

Perceived altruism of company

Extent to which participants agreed/disagreed with the following

statements:

The manufacturer conducts the campaign in order to do a good

deed.

The campaign is an honest effort.

The manufacturer is not truly committed to the purpose of the

donation.

Fries, Gedenk,

& Völckner,

2010; Nowak,

2004;

Strahilevitz,

2003; Webb,

Green, &

Brashear, 2000

.93

Notes: = Cronbach’s alpha in our study; †we added the item trustworthy; S = Study.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 47: Forthcoming IJRM Volume 31 #2 (2014)

48

Appendix D

Variable specifications

Variable Specification

Category variables

CAT_CHOCOLATE 1 if product category is chocolate bars, 0 otherwise

CAT_TOOTHPASTE 1 if product category is toothpaste, 0 otherwise

CAT_BEER 1 if product category is beer, 0 otherwise

CAT_DETERGENT 1 if product category is detergent, 0 otherwise

PREPREF Number of times the CM brand was bought in the last

three purchases in the category prior to the survey,

minus the number of times the other brand was

bought / Divided by 3

Demographic variables

Child_YES 1 if subject has children, 0 otherwise

Church membership_YES 1 if subject is member of a church, 0 otherwise

Church membership_NO_Res 1 if subject did not indicate church membership,

0 otherwise

Occupation_FULL 1 if subject works full time, 0 otherwise

Occupation_PART 1 if subject works part time, 0 otherwise

CM variables

DONSIZ Donation size of the CM offer

NONMON 1 if the CM campaign includes nonmonetary donation

framing, 0 otherwise

COMBI 1 if the CM campaign includes combined nonmonetary

and monetary donation framing, 0 otherwise

TRADE-OFF 1 if the competitive brand offers a price promotion,

0 otherwise

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 48: Forthcoming IJRM Volume 31 #2 (2014)

49

References

Ailawadi, K. L., Gedenk, K., Lutzky, C., & Neslin, S. A. (2007). Decomposition of the sales

impact of promotion-induced stockpiling. Journal of Marketing Research, 44(3), 450-

467.

Ailawadi, K. L., Gedenk, K., & Neslin, S. A. (1999). Heterogeneity and purchase event feedback

in choice models: An empirical analysis with implications for model building.

International Journal of Research in Marketing, 16(3), 177-198.

Alba, J. W., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal of

Consumer Research, 13(4), 411-454.

Anderson, N. H. (1981). Foundations of information integration theory. New York: Erlbaum.

Andreoni, J. (1989). Giving with impure altruism: Applications to charity and ricardian

equivalence. Journal of Political Economy, 97(6), 1447-1458.

Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow

giving. The Economic Journal, 100(401), 464-477.

Arora, N., & Henderson, T. (2007). Embedded premium promotion: Why it works and how to

make it more effective. Marketing Science, 26(4), 514-531.

Auger, P., Devinney, T. M., Louviere, J. J., & Burke, P. F. (2008). Do social product features

have value to consumers? International Journal of Research in Marketing, 25(3), 183-

191.

Barnes, N. G. (1992). Determinants of consumer participation in cause-related marketing

campaigns. American Business Review, 10(2), 21-24.

Barone, M. J., Miyazaki, A. D., & Taylor, K. A. (2000). The influence of cause-related

marketing on consumer choice: Does one good turn deserve another? Journal of the

Academy of Marketing Science, 28(2), 248-262.

Bouten, L. M., Snelders, D., & Hultink, E. J. (2011). The Impact of Fit Measures on the

Consumer Evaluation of New Co-Branded Products. Journal of Product Innovation

Management, 28(4), 455-469.

Burnett, J. J., & Wood, V. R. (1988). A proposed model of the donation decision process. In E.

Hirschman, & J. Sheth (Eds.), Research in Consumer Behavior, 3, (pp. 1-47). Greenwich,

CT: Elsevier JAI.

Chang, C.-T. (2008). To donate or not to donate? Product characteristics and framing effects of

cause-related marketing on consumer purchase behavior. Psychology & Marketing,

25(12), 1089-1110.

Dahl, D. W., & Lavack, A. M. (1995). Cause-related marketing: Impact of size of corporate

donation and size of cause-related promotion on consumer perceptions and participation.

In AMA Winter Educators' Conference Proceedings, 6, 476-481.

Eckel, C. C., & Grossman, P. J. (2003). Rebate versus matching: Does how we subsidize

charitable contributions matter? Journal of Public Economics, 87(3/4), 681-701.

Ellen, P. S., Mohr, L. A., & Webb, D. J. (2000). Charitable programs and the retailer: Do they

mix? Journal of Retailing, 76(3), 393-406.

Fries, A. J. (2010). The effects of cause-related marketing campaign characteristics – A literature

review. Marketing - Journal of Research and Management, 6(2), 145-157.

Fries, A. J., Gedenk, K., & Völckner, F. (2010). Cause-related marketing: Designing successful

campaigns. University of Cologne Working Paper.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 49: Forthcoming IJRM Volume 31 #2 (2014)

50

German Fundraising Association. (2009): Spendenbilanz ausgewählter Organisationen 2005-

2008. Retrieved March 15, 2010, from http://www.fundraisingverband.de/fileadmin/

pdf_upload/1Spendenbilanz_2005-2008.pdf.

Garretson Folse, J. A., Niedrich, R. W., & Landreth Grau, S. (2010): Cause-related marketing:

The effect of purchase quantity and firm donation amount on consumer inferences and

participation intenions. Journal of Retailing, 86(4), 295-309.

Green, D. P., Kahneman, D., & Kunreuther, H. (1994). How the scope and method of public

funding affect willingness to pay for public goods. Public Opinion Quarterly, 58(1), 49-

67.

Greene, W. H. (2008). Econometric analysis. (6th

ed.). Upper Saddle River, NJ: Pearson Prentice

Hall.

Gupta, S., & Cooper, L. G. (1992). The discounting of discounts and promotion thresholds.

Journal of Consumer Research, 19(3), 401-411.

Haruvy, E., & Popkowski Leszczyc, P. T. L. (2009). Bidder motives in cause-related auctions.

International Journal of Research in Marketing, 26(4), 324-331.

Holmes, J. H., & Kilbane, C. J. (1993). Cause-related marketing: Selected effects of price and

charitable donations. Journal of Nonprofit & Public Sector Marketing, 1(4), 67-83.

Horsky, D., Misra, S., & Nelson, P. (2006). Observed and unobserved preference heterogeneity

in brand-choice models. Marketing Science, 25(4), 322-335.

Human, D., & Terblanche, N. S. (2012). Who receives what? The influence of the donation

magnitude and donation recipient in cause-related marketing. Journal of Nonprofit and

Public Sector Marketing, 24(2), 141-160.

IEG. (2011). IEG Sponsorship Report, Retrieved June 2, 2011, from

http://www.sponsorship.com/ About-IEG/Press-Room/Economic-Uncertainty-To-Slow-

Sponsorship-Growth-In.aspx.

Jaccard, J. (2001). Interaction effects in logistic regression. Thousand Oaks, CA: Sage.

Jaccard, J., Turrisi, R., & Wan, C. K. (1990). Interaction effects in multiple regression. Thousand

Oaks, CA: Sage.

Kahneman, D., & Knetsch, J. L. (1992). Valuing public goods: The purchase of moral

satisfaction. Journal of Environmental Economics and Management, 22(1), 57-70.

Kahneman, D., Ritov, I., Jacowitz, K. E., & Grant, P. (1993). Stated willingness to pay for public

goods: A psychological perspective. Psychological Science, 4(5), 310-315.

Keller, K. L. (1993). Conceptualizing, measuring, and managing customer-based brand equity.

Journal of Marketing, 57(1), 1-22.

Koschate-Fischer, N., Stefan, I. V., & Hoyer, W. D. (2012). Willingness to pay for cause-related

marketing: The impact of donation amount and moderating effects. Journal of Marketing

Research, 49(6), 910-927.

Landreth Grau, S., Garretson Folse, J. A., & Pirsch, J. (2007). Cause-related marketing: An

exploratory study of campaign donation structures issues. Journal of Nonprofit & Public

Sector Marketing, 18(2), 69-91.

Lautenschlager, G. J., & Flaherty, V. L. (1990). Computer administration of questions: More

desirable or more social desirability? Journal of Applied Psychology, 75(3), 310-314.

Monin, B. (2003). The warm glow heuristic: when liking leads to familiarity, Journal of

Personality and Social Psychology, 85(6), 1035-1048.

Nancarrow, C., Brace, I., & Wright, L. T. (2001). Tell me lies, tell me sweet little lies: Dealing

with socially desirable responses in market research. Marketing Review, 2(1), 55-69.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 50: Forthcoming IJRM Volume 31 #2 (2014)

51

Neslin, S. A., & van Heerde, H. J. (2009). Promotion dynamics. Foundation and Trends in

Marketing, 3(4), 177-268.

Nowak, L. I. (2004). Cause marketing alliances: Corporate associations and consumer responses,

Journal of Food Products Marketing, 10(2), 33-48.

Nunes, J. C., & Park, C. W. (2003). Incommensurate resources: Not just more of the same.

Journal of Marketing Research, 40(1), 26-38.

Olsen, G. D., Pracejus, J. W., & Brown, N. R. (2003). When profit equals price: Consumer

confusion about donation amounts in cause-related marketing. Journal of Public Policy &

Marketing, 22(2), 170-180.

Palazon, M., & Delgado-Ballester, E. (2009). Effectiveness of price discounts and premium

promotions. Psychology & Marketing, 26(12), 1108-1129.

Polonsky, M. J., & Wood, G. (2001). Can the overcommercialization of cause-related marketing

harm society? Journal of Macromarketing, 21(1), 8-22.

Pracejus, J. W., Olsen, G. D., & Brown, N. R. (2003/04). On the prevalence and impact of vague

quantifiers in the advertising of cause-related marketing (CRM). Journal of Advertising,

32(4), 19-28.

Ross, J. K., Patterson, L. T., & Stutts, M. A. (1992). Consumer perceptions of organizations that

use cause-related marketing. Journal of the Academy of Marketing Science, 20(1), 93-97.

Ross, J. K., Stutts, M. A., & Patterson, L. (1991). Tactical considerations for the effective use of

cause-related marketing. The Journal of Applied Business Research, 7(2), 58-65.

Sargeant, A., & Lee, S. (2004). Trust and relationship commitment in the United Kingdom

voluntary sector: Determinants of donor behavior, Psychology & Marketing, 21(8), 613-

635.

Simonin, B. L., & Ruth, J. A. (1998). Is a company known by the company it keeps? Assessing

the spillover effects of brand alliances on consumer brand attitudes. Journal of Marketing

Research, 35(1), 30-42.

Smith, S. M., & Alcorn, D. S. (1991). Cause marketing: A new direction in the marketing of

corporate responsibility. Journal of Consumer Marketing, 8(3), 19-35.

Strahilevitz, M. (1999). The effects of product type and donation magnitude on willingness to

pay more for a charity-linked brand. Journal of Consumer Psychology, 8(3), 215-241.

Strahilevitz, M. (2003). The effects of prior impressions of a firm’s ethics on the success of a

cause-related marketing campaign: Do the good look better while the bad look worse?

Journal of Nonprofit and Public Sector Marketing, 11(1), 77-92.

Strahilevitz, M., & Myers, J. G. (1998). Donations to charity as purchase incentives: How well

they work may depend on what you are trying to sell. Journal of Consumer Research,

24(4), 434-446.

Subrahmanyan, S. (2004). Effects of price premium and product type on the choice of cause-

related brands: A singapore perspective. Journal of Product & Brand Management,

13(2), 116-124.

Train, K. E. (2009). Discrete choice methods with simulation. (2nd

ed.). Cambridge: Cambridge

University Press.

Vaidyanathan, R., & Aggarwal, P. (2005). Using commitments to drive consistency: Enhancing

the effectiveness of cause-related marketing communications. Journal of Marketing

Communications, 11(4), 231-246.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)

Page 51: Forthcoming IJRM Volume 31 #2 (2014)

52

van den Brink, D., Odekerken-Schröder, G., & Pauwels, P. (2006). The effect of strategic and

tactical cause-related marketing on consumers' brand loyalty. Journal of Consumer

Marketing, 23 (1), 15-25.

van Heerde, H. J., Leeflang, P. S. H., & Wittink, D. R. (2000). The estimation of pre- and

postpromotion dips with store-level scanner data. Journal of Marketing Research, 37(3),

383-395.

van Heerde, H. J., Leeflang, P. S. H., & Wittink, D. R. (2001). Semiparametric analysis to

estimate the deal effect curve. Journal of Marketing Research, 38(2), 197-215.

van Heerde, H. J., Leeflang, P. S. H., & Wittink, D. R. (2004). Decomposing the sales promotion

bump with store data. Marketing Science, 23(3), 317-334.

Varadarajan, P. R., & Menon, A. (1988). Cause-related marketing: A coalignment of marketing

strategy and corporate philanthropy. Journal of Marketing, 52(3), 58-74.

Völckner, F., Sattler, H., & Kaufmann, G. (2008). Image feedback effects of brand extensions:

Evidence from a longitudinal field study. Marketing Letters, 19(2), 109-124.

Webb, D. J., Green, C. L., & Brashear, T. G. (2000): Development and validation of scales to

measure attitudes influencing monetary donations to charitable organizations, Journal of

the Academy of Marketing Science, 28(2), 299-309.

WHO & UNICEF. (2010). Immunization summary: The 2010 edition. Retrieved January 22,

2010, from http://www.childinfo.org/files/Immunization_Summary_2008_r6.pdf.

Wymer, W., & Samu, S. (2009). The influence of cause marketing associations on product and

cause brand value. International Journal of Nonprofit and Voluntary Sector Marketing,

14(1), 1-20.

Zdravkovic, S., Magnusson, P., & Stanley, S. M. (2010). Dimensions of fit between a brand and

a social cause and their influence on attitudes. International Journal of Research in

Marketing, 27(2), 151-160.

Forthc

oming

IJRM V

olume 3

1 #2 (

2014

)