J Consum Policy (2013) 36:369–387

DOI 10.1007/s10603-013-9234-0

ORIGINAL PAPER

The Importance of the Indirect Transfer Mechanism

for Consumer Willingness to Pay for Fair Trade
Products—Evidence from a Natural Field Experiment

Hannes Koppel & Günther G. Schulze

Received: 15 November 2012 / Accepted: 9 July 2013 /

Published online: 3 August 2013
# Springer Science+Business Media New York 2013

Abstract A natural field experiment is conducted on the determinants of consumer’s

willingness to pay for “fair trade” (FT) products. In four treatments, subjects are offered
different choices in connection with a coffee purchase, such as buying regular coffee or FT
coffee at a premium, regular coffee with or without a premium, or donating directly or not
donating. Depending on the treatment, the premium or direct donations are either given to
FT or a well-known charity. A large part of the willingness to pay a premium for FT coffee
over regular coffee is shown to be not related to the specific attributes of FT coffee but rather
caused by the indirect transfer mechanism that FT uses, i.e., selling products at a premium
which goes to the cause of FT.

Keywords Fair Trade . Donation . Altruism . Transfer Mechanism

Introduction

The fair trade (FT) model offers an alternative to traditional trade channels intending to
support participating producers in the global South. Consumers in the North are willing to
pay a price premium over a non-FT product of otherwise equal quality. The premium is used
to pay farmers and workers better prices and wages, respectively, to guarantee stable trade
relations and to finance communal funds for social, environmental, and economic projects.
To participate in FT, the production of an entire plantation has to fulfil the FT standards.
However, considering a sales quota, producers can sell only some of their goods through the
FT channel while the rest has to be sold at world market prices (cf. Steinrücken and
Jaenichen 2007). This quota guarantees that non-FT products of otherwise equal quality

H. Koppel (*)
Alfred-Weber-Institute, University of Heidelberg,
Bergheimer Strasse 58, 69115 Heidelberg, Germany
e-mail: hannes.koppel@awi.uni-heidelberg.de

G. G. Schulze
Department of Economics, Albert-Ludwigs-University Freiburg,
Paltz der Alten Synagoge 1, 79085 Freiburg, Germany
e-mail: guenther.schulze@vwl.uni-freiburg.de
370 H. Koppel, G.G. Schulze

are sold in the regular market. FT products range from coffee, cocoa, tea, fruits, nuts, sugar,
flowers, and seed cotton to sports balls. Since its beginning in 1988, the FT movement has
experienced rapidly increasing sales; in 2009, total FT retail sales amounted to 3.4 bn Euros,
an increase of 15% over 2008 (FLO 2010). FT coffee, which generates the largest FT sales,
has mainstreamed in particular from being sold only in “world shops” to being a regular item
in supermarkets and discount outlets (Raynolds 2009). This development has led to a
growing body of research on FT. A first strand analyses the efficiency properties of FT
(Hayes 2006; Kadow 2011; Maseland and de Vaal 2002, 2008; Reinstein and Song 2012;
Weber 2007). In a more general context, the theoretical literature on the provision of impure
public goods analyses the effects of the emergence of goods that include a private consump-
tion element with a public goods element, such as environmentally friendly produced goods
or goods with a charity component. It shows that welfare effects can go both up and down
(Cornes and Sandler 1984, 1994; Kotchen 2006; Vicary 1997). A second, related strand
investigates empirically whether the FT model has actually improved the living conditions of
farmers in the South (Barham et al. 2011; Beuchelt and Zeller 2011; Valkila 2009).
A third strand enquires into the motivation of Western consumers to pay a price premium
for FT products (for a review, see Andorfer and Liebe 2012). This issue is particularly
interesting because of the nature of the FT model: It uses the purchase of a product with FT
attributes (e.g., social, economic, and environmental development issues as well as labour
standards; cf. FLO 2010) as a donation vehicle and charges a premium over the correspond-
ing non-FT product. While previous studies have analysed consumer preferences for ethical
product attributes as well as consumer attitudes and motivations to purchase products with
ethical attributes (see, for example, Cranfield et al. 2010; de Ferran and Grunert 2007;
Zander and Hamm 2010), the empirical analysis of the implicit transfer mechanism has been
neglected.1 This analysis begs the question whether the willingness to pay (WTP) such a
premium reflects only the product attributes that FT products have in addition to regular
products of the otherwise same quality or whether the donation vehicle as such mobilizes
resources. Would it be preferable to unbundle the purchase of the product and the donation
element or, stated differently, does the notion that tied transfers are welfare inferior com-
pared with untied transfers carry over to the case of FT?
Most empirical studies on FT have focused on the size of the premium and the socio-
economic determinants of the WTP the premium. Loureiro and Lotade (2005) estimate in a
contingent valuation approach that respondents are willing to pay a premium of 0.22 US$
per pound on top of the regular price of 6.50 US$. In a survey of 808 Belgians, de
Pelsmacker et al. (2005) find that consumers are on average willing to pay a premium of
10 % for FT coffee. Rotaris and Danielis (2011) show in a conjoint analysis using Italian
survey data that coffee consumers are primarily influenced by brand (aroma and flavour),
price and habit, but that the FT label also plays a significant role especially among the young
and well educated. In a cross national conjoint analysis with German and US students, Basu
and Hicks (2008) demonstrate that the WTP for FT coffee is positively related to the benefits
farmers receive up to a certain point, after which it declines.
A common feature of these studies is that they rely on surveys or hypothetical experi-
ments. However, there is also some evidence on revealed preferences which is arguably

1
The donation element incorporated in the mechanism relates FT to the empirical literature on giving and
altruism, which mainly analyses the socio-economic determinants of individual donations (see Schokkaert
2006, for an overview). While the determinants of altruism are well known (see, Bekkers and Wiepking 2011,
and Wiepking and Bekkers 2012, for surveys), the specific mechanism that fair trade uses is still an under-
researched issue. This mechanism makes the consumer a potential donor and conversely allows the individual
to make a donation by becoming a consumer.
The Importance of Indirect Transfers for Consumer Willingness to Pay 371

more reliable (Harrison and List 2004; Levitt and List 2007). Schollenberg (2012) using a
hedonic pricing approach estimates a price premium of 38 % for FT coffee of average grade
with scanner sales panel data for Sweden. Arnot et al. (2006) conduct an experiment in a
campus coffee shop, varying price differentials between regular and FT coffee and show that
consumers of FT coffee have a very low price elasticity of demand. Hainmueller et al. (2011)
show in a field experiment conducted in a US grocery store chain that sales of bulk coffee
increased by 10 % when these coffees were labelled as FT. Demand for higher priced coffee
was price inelastic while demand for low price coffee was more price sensitive. Considering
that these studies demonstrate the extent to which (different types of) individuals are willing
to pay extra for FT products, they do not shed light on the role of the specific mechanism that
the FT approach uses.
Related to the mechanism is the literature on cause-related marketing (CRM), which studies
how linking product sales to a charitable cause may serve as a marketing tool and thus as a
competitive device (Varadarajan and Menon 1988). The focus is on the characteristics that make
products suitable for CRM and the way the cause relation is established. Strahilevitz (1999) and
Strahilevitz and Myers (1998) show in a hypothetical choice experiment that hedonic products
lend themselves more easily to charity incentives than utilitarian ones. Moreover, the relative
amount of the donation (Olsen et al. 2003), the declaration of donation in absolute or relative
terms (Chang 2008), and the product-cause fit (Gupta and Pirsch 2006) matter for the decision
to purchase the good. Yechaim et al. (2003) show in a repeated choice experiment, that CRM
helps inferior products in that the attractiveness of superior alternatives is lowered. While this
literature focuses on the influence of CRM on product sales, its impact on the causes is rarely
studied (see Polonsky and Macdonald 2000, for an exception).
As the effectiveness of the particular mechanism for raising donations that the FT
approach (and potentially other causes) uses is analysed, the present research is comple-
mentary to the CRM approach. Is the purchase at a price premium a particularly useful
donation vehicle in general or is it the specific FT product attributes that allow for selling
those products at a premium?
To answer this question, a natural field experiment (for a classification of experiments see
Harrison and List 2004) is conducted in cafés at German Universities that compares the
willingness to donate indirectly through a purchase at a premium with the willingness to donate
directly. The experiment involves four treatments. In one, subjects choose between normal
coffee and FT coffee sold at a premium; this is the standard FT situation. In another, subjects can
buy normal coffee only, but are given the opportunity to make direct donations to producers of
FT coffee. The willingness to donate to FT producers either directly or indirectly can thus be
compared. However, the WTP a premium for FT coffee in the first treatment could be due to the
additional attributes of the FT coffee, e.g., higher perceived quality or better production
standards for FT coffee, rather than the motive to donate money to the producers. In addition,
in the case of FT products, individuals may have a preference to donate to poor producers
through an increased price so as to give them what they regard to be fair compensation for their
work. They may regard direct donations inferior to indirect (tied) donations, considering the
former to be charitable giving and the latter to be fair remuneration. Economically, however,
such a tied transfer is still a donation if cheaper coffee of the same quality is available.
To control for such possibilities, a third and fourth treatment was added: In the third
treatment, subjects can choose between the same two prices as in the first treatment but for
the same coffee; the price difference goes to a charitable project that benefits producers in
the third world (but not necessarily the producers of the consumed coffee). The fourth
treatment is the same as the second with the exception that the donations now go to the same
charitable project as in the third treatment. The comparison of the two sets of treatments
372 H. Koppel, G.G. Schulze

allows for the assessment of the extent to which the special attributes of FT products explain
a higher WTP for these products and the extent to which the special donation vehicle
contributes to the WTP for FT products.
In both setups, it turns out that the willingness to donate is substantially higher when
subjects are offered the indirect donation mechanism (price premium) rather than the direct
donation mechanism. Thus, the means whereby money is raised is crucial for the mobiliza-
tion of resources. Moreover, the donation through a price premium mobilizes more resources
in the case of FT coffee than in the case of regular coffee and a charity donation. Thus,
consumers are, in fact, willing to pay for the special attributes of FT products, but that
explains only the smaller part of their extra WTP for FT products.

Experimental Design

The present experiment enquires into the determinants of giving for the specific case of FT.
A natural field experiment in coffee shops was conducted, which the subjects visit as part of
their daily routine and where they spend their own money not knowing that they are part of
an experiment (for a detailed discussion on field experiments, see Harrison and List 2004).
The only difference is that they are now given a choice (to donate or not, to buy a different
coffee, and/or to pay a higher price) whereas previously they had none. Thus, the experiment
can provide some realistic insights into the different inclination for (small) donations under
direct and indirect transfer mechanisms.

General Setup

The present study aims to determine the factors that explain the WTP for FT products at a
premium in a typical, realistic setting. Since coffee is the most common FT product, the
purchase of coffee was chosen as the donation vehicle for the experiment. University coffee
shops that are self-administered by student unions were chosen as locations, but only those
coffee shops that did not sell FT or organic coffee prior to the experiment nor collected any
donations; all shops sold a cup of normal “average” coffee for 50 Eurocent (equal to
approximately 67 US cent at the time of the experiment). The selection of coffee shops
avoids a potential bias stemming from an endowment effect. This effect describes a
behavioural pattern in which individuals value a good more highly when they are endowed
with it and have to forego it than if they want to acquire it. Thus, WTP for a good is regularly
lower than the compensation that they require to forego the product (Kahnemann et al.
1990). Such an effect could occur in the context of this study: the number of people opting
for FT coffee could be higher if normal coffee was introduced at 50 Eurocent when they
previously were offered only FT coffee at 60 Eurocent compared with a situation in which
people initially could buy only normal coffee at 50 Eurocent and FT coffee at 60 Eurocent
was introduced. To avoid such an effect, the authors thus chose in all treatments the same
baseline scenario of normal “average” coffee being sold at 50 Eurocent. In addition to
normal coffee, all shops sold other products, such as non-coffee beverages, but only one type
of coffee. Other suitable venues were also found, in which the physical settings were also
very similar, in Darmstadt, Hamburg, and in two different universities in Berlin so that
communication between the patrons of different treatments was unlikely.
The following protocol was considered: Each coffee shop was sent a letter with instruc-
tions (see Appendix 1), which all sales persons were requested to read. Prior to the actual
experiment, the authors visited all participating coffee shops and gave detailed instructions
The Importance of Indirect Transfers for Consumer Willingness to Pay 373

to all people working in the coffee shop on how to implement the treatments; the necessary
material as well as the coffee were provided. The experiment was conducted in each coffee
shop over a period of 2 weeks with observations recorded on Tuesdays, Wednesdays, and
Thursdays. In the first week, the coffee shops operated as usual and the sales personnel
recorded the number of coffees and other products sold on each particular day. These
accounts serve as reference points, in order to find out whether the treatments attracted
more customers in the second (i.e., the treatments) week, which may result in biased samples
for the different treatments. Fortunately, the Wilcoxon-signed-rank test rejects the null for
each treatment (i.e., coffee shop), thus there is no significant structural difference in the
samples. In the second week, each coffee shop implemented one treatment (see below) while
normal coffee continued to be sold at the regular price. At the counter where the coffee was
sold, the introduction of a new alternative (different coffee at higher price, donation box,
etc.,) with signs and material provided by the trade company (GEPA) and the charity
(MISEREOR) used in the experiment was clearly marked. All signs and material were of
equal size and colour, so that the level of awareness and information about the new
alternative was equal for all treatments. A standard type of donation box which is used
widely for charitable donations at counters or in street collections and is easily recognizable
as such (see Appendix 2 for pictures of the settings) was used. The authors specifically
instructed all sales personnel in all the treatments not to proactively advertise the new
alternative but to briefly mention it. Thus, customers were aware of the new alternative
but should not feel pressured into opting for it.
On the last day of every treatment, patrons were interviewed using a standardized
questionnaire. The following information were asked: Gender, number of semester, field
of study, marital status, number of siblings, visits to developing countries, a general donation
frequency, variables indicating the wealth status, and religiousness. Lastly, the authors asked
what people associated with “FT.” Interviews were carried out only on the last day and only
after the customers had bought coffee, to avoid influencing their future behaviour in the
experiment. Until this point, patrons were not aware of the experiment. This may be
important, as when people are aware of being part of an experiment they may have a
stronger self-image or sense of identity and therefore behave differently (Akerlof and
Kranton 2000; Levitt and List 2007). Studying people’s behaviour in their natural habitat
without them being aware that they are under study avoids such—potentially large—bias.
The intention of the questionnaire was to determine whether populations were comparable in
their socio-economic composition. Also, the authors wanted to analyse whether individuals
regarded FT mainly as a means to donate or whether they associated specific product
attributes with FT coffee and hence with the WTP a premium.

Treatments

Four independent treatments were conducted. The announcements for the treatments were
similarly designed, except for the differences between treatments.

Treatment 1 (T1) The consumer could choose between normal coffee at 50 Eurocent, as
prior to the experiment, and FT coffee at 60 Eurocent. This premium of 20% approximately
reflects the price difference in the market between “standard” coffee (that all coffee shops
used prior to and during the experiment) and FT coffee.

Treatment 2 (T2) People could buy normal coffee only, and they had the choice of making a
donation in a newly installed donation box. In this case, the donation was given to the most
374 H. Koppel, G.G. Schulze

prominent FT organization in Germany, GEPA, indicated by a corresponding sign “gepa—Das

Faire Handelshaus” (“gepa—The FT Company”).2 It is the same company that produced the FT
coffee used in T1.
The possibility that the WTP a price premium for FT coffee is not entirely or even not
primarily motivated by the wish to donate cannot be excluded. Instead, individuals could be
willing to pay extra for the specific production standards under which FT coffee is produced
or for a perceived higher quality of FT coffee compared with normal coffee. They could also
regard the higher price as a vehicle to achieve a fair remuneration for the producers’ work
and thus prefer to channel the donation to the producers through the indirect mechanism as
they may regard the direct donation mechanism as charitable giving rather than fair
remuneration. Some might not even be aware of the donation component in the indirect
mechanism. Likewise, subjects might not consider the FT organization GEPA to be a charity,
and therefore, would not be inclined to put money in the donation box. These effects would
lead to a larger difference between the indirect and direct donation treatment than is caused
by the different methods to donate.
Therefore, a second set of treatments were conducted: The authors used normal coffee as
in the first treatment, sold at the regular price and in the third treatment also at the higher
price with the premium being donated. This excludes different product attributes as a reason
to pay a premium. Moreover, a well-known and credible German charity for development
cooperation (MISEREOR) was chosen as recipient of the donations in both treatments to
exclude the second reason for a bias, namely that FT organizations may not be seen as
charitable organizations and therefore direct donations may be low. Since the money went to
a specific relief project supporting poor producers in general and not to the particular
producers of coffee that the subjects were consuming, the third motivation, a fair remuner-
ation for producers of the consumed product, was absent as well. Thus the study focused on
the different methods of raising donations.

Treatment 3 (T3) Participants could choose to buy the same normal coffee at 50 Eurocent or
at 60 Eurocent. If they opted for the second option, the difference was donated to
MISEREOR. The donation went to a specific relief project for small-scale producers,
including coffee producers, in a developing country.3

Treatment 4 (T4) The fourth treatment is the same as the second, in which normal coffee is
sold and a donation box installed, except that the donation goes to the same MISEREOR
relief project as in T3.
A comparison between T3 and T4 shows a possibly different inclination to donate
through indirect and direct donation mechanisms. Moreover, it shows whether it is solely
the difference in the donation vehicle that has an impact on the amount of resources

2
Although “GEPA” is not a charity as such and does not raise donations directly, it is the institution which is
widely known as the leading fair trade organization in Germany. Since “GEPA” does not have a way of
accepting donations, they offered to donate the money to “MISEREOR.”
3
MISEREOR is one of the most prominent German charities for Development Cooperation. It has received
the Seal-of-Approval (a label given to credible charities) from the German Central Institute for Social Issues
(http://www.dzi.de/dzi-institut/german-central-institute-for-social-issues/, 15 Nov 2012) since the seal was
established in 1992. The donations went to the project P22302 of MISEREOR. For details of the project
see http://www.misereor.de/projekte/projektpartnerschaften/haiti-waldgaertenkleinbauer.html, 14 Sept 2011,
“Haiti: Karibikparadies—vom Winde verweht” (“Haiti: Caribbean paradise—Gone with the wind”).
The Importance of Indirect Transfers for Consumer Willingness to Pay 375

mobilized. A comparison between T1 and T2 may show this as well, but results might also
be influenced by different perceived product attributes of FT coffee, offered in T1 but not in
T2, compared with normal coffee (see above). Such a difference in the evaluation of the
product will be shown by comparing T1 and T3. A second bias may arise because people do
not regard the FT organization GEPA as a charity and therefore are less inclined to donate
directly. If that were true, it would show up by comparing T2 and T4. These comparisons
and detectable effects are summarized in Table 1.
Evidence shows that donations are more frequent and higher if others can observe them
as compared with a situation of anonymity (Andreoni and Petrie 2004; List et al. 2004;
Soetevent 2005). This could lead to a bias in the experiment, if the degree of observability
was not uniform across treatments. Thus, the authors are confident that the results are not
affected by such a bias because public observability of the act of donating was the same in all
treatments: The donation was made in public and could be observed by others only at the
time it was made, but not afterwards. People could observe the dropping of the coin in the
donation box as well as overhear the order for the FT coffee or the coffee purchased at a
premium but once the transaction was concluded there was no visible sign of the donation,
nor was there any indication of how many people had donated before. The donation box was
non-transparent, so that the amount donated thus far could not be discerned, and there was
no record of how many people had opted for the more expensive coffee. Likewise people
could not distinguish FT coffee (T1) or other coffee purchased at a premium (T3) from
regular coffee purchased at 50 Eurocents once it was sold since cups, smell, colour, etc.,
were identical.

Results

Experimental Results

For the first week, before the treatment was implemented, all sales of regular coffee and
other products were recorded by the sales personnel. Records for the second week, the
treatment week, include again all sales of regular coffee and other products, and for T1 also
the number of FT coffees sold at 60 Eurocents, for T3 the number of regular coffees sold at a
premium price of 60 Eurocents and for T2 and T4 the number of donations made in
connection with the purchase of regular coffee or other products. No donation was made
without any purchase. The individual donations were not counted, but only the total amount
donated in each box was accounted for.
For the indirect donation mechanism (T1 and T3), the amount of donation was restricted
to the price premium of 10 Eurocent. For the direct donation mechanism (donation box, T2
and T4), the individual donation was unrestricted and not observed. This is of little
consequence since only the possible differences in the total amount of resources mobilized

Table 1 Measured effects by comparison of treatments

Treamtents T2 (direct and FT) T3 (indirect and donation)

T1 (indirect and FT) Direct vs. indirect (including attributes) Attributes (of product)
T4 (direct and donation) Attributes (of organization) Direct vs. indirect (excluding attributes)
376 H. Koppel, G.G. Schulze

per capita were of interest, which were observed. Essentially, differences in per capita
donations were observed by comparing imputed distributions of standardized donations.
For the direct donation mechanism, computation focused on how many people would have
donated the same total amount if direct donations were restricted to 10 Eurocent as in the
case of indirect donations, and by using nonparametric tests to compare the two distribu-
tions. Different numbers of subjects in the four treatments were accounted. Details of
the imputation method as well as the sales statistics for all treatments are given in
Appendix 3.
A first impression of the different inclination to donate in the four treatments is depicted
in Figure 1. It shows the share of people in each treatment that have donated the actual or
imputed donation of 10 Eurocent. There is a marked difference between direct and indirect
donation mechanisms which is larger for the FT treatments (T1 and T2) than for the relief
project treatments (T3 and T4). It points to the biases described above related to the
differential perception of the FT company and the relief project.
A significant difference in donation behaviour between the two FT treatments (T1 and
T2) was found; Fisher's exact test rejects the null hypothesis of equality at the 1% level of
significance (p<.01). The direct donation mechanism raises an average donation of 0.9
Eurocent whereas the indirect donation mechanism commands a per capita donation of 6.9
Eurocent.
The same qualitative result of a significant difference between indirect and direct
donations for the relief project treatments T3 and T4 was found. Direct donations raise an
average of 2.4 Eurocent, while indirect donations are on average 5.5 Eurocent. Again
Fisher's exact test rejects the null hypothesis at a 1% level of significance (p<.01). Thus,
the donation vehicle—direct vs. indirect donations—plays an important role in mobilizing
resources, for charities and for FT (coffee).
However, as mentioned above, people may not see FT companies as charities, because
they do not collect (direct) donations, but operate as normal companies. That may reduce the
willingness to donate directly and thus donations per capita could be lower in T2 than in T4.

0.8
0.7
0.6
0.5
Share

0.4
0.3
0.2
0.1
0
Indirect Transfer Indirect Transfer Direct Transfer Direct Transfer
Fair Trade Donation Donation Fair Trade
(T1) (T3) (T4) (T2)

Fig. 1 Share of pepole donating 10 eurocent per treatment

The Importance of Indirect Transfers for Consumer Willingness to Pay 377

In addition, one motivation for buying FT coffee at a premium may be perceived differences
in standards of production or in quality (T1 vs. T3). The latter effect would show if the
probability to donate was significantly higher in T1 than in T3. T3 uses normal coffee prior
to the experiment and thus there are no different standards or quality as a motivation to pay
extra. There seem to be differences in the distribution (Fig. 1), but Fisher's exact test rejects
the null only at a 12 % level of significance, which is above the normal significance levels.
This result suggests that the standards of production and perceived quality of FT coffee do
play a minor role in the mobilization of resources and highlights the role of the transfer
mechanism within the FT model.
Likewise, the authors test whether there are significant differences in the direct donations
to the FT organization GEPA and the MISEREOR relief project by comparing T2 and T4.
Indeed, direct donations to GEPA raise significantly lower average amounts than those to the
well-known charity MISEREOR. This finding indicates that FT is not seen as charity in the
traditional sense and therefore subjects are disinclined to donate directly.
For treatments T2 and T4, the actual number of people who made a donation was smaller
than the imputed one, which indicates that individual donations were larger per capita than
the imputed ones (at 10 Eurocent). This raises the question whether more resources could
have been mobilized with the indirect transfer mechanism if a different premium was
chosen. Since the authors already obtain the result that the indirect transfer mechanism
raises significantly more money than the direct one, our basic result would be unaltered if
they did. However, the authors might have underestimated the maximum difference—it may
be even larger with an optimized premium and did not consider this as a big issue because
the actual premium found in the coffee market was approximately used, which should have
already been optimized by the FT companies. Likewise, a willingness to donate an amount
smaller than 10 Eurocent cannot materialize in the indirect donation mechanism, but in the
direct one. However, this might bias results, in the same direction leading to an underesti-
mation of the differential effect of the indirect vs. direct donation vehicle.

Survey Results

To explore the motivation behind donating (differentially) through purchasing FT products,

a survey on the last day was carried out so as not to influence future consumption behaviour.
A total of 82 subjects of all treatments were asked “What do you associate with FT?”
Respondents could choose up to three answers out of six possibilities given. Results are
provided in Figure 2. It turns out that most of the subjects associate FT with development aid
and standards of production, a smaller number with donations and the warm glow, and a still
smaller, but substantial number with better quality (taste or health). In fact, approximately
50 % of FT coffee is also certified as organic (cf. FLO 2010 or www.fairtrade.net).
In other words, a significant number of respondents associate FT with special
product attributes (production standards or quality, i.e., taste and health). This suggests
that some of the WTP a premium for FT coffee may be attributed to the valuation of
these attributes, rather than only the wish to donate to a group of people in need.
Noted also was that most people were well aware that a purchase of a FT product
implies a donation or giving aid to needy people. Figure 2 contains answers from all
individuals surveyed regardless of their decision to donate (explicitly or implicitly
through a purchase at a premium).
378 H. Koppel, G.G. Schulze

Health

Taste

Standards of
Production

Good Feeling

Donation

Development Aid

0 10 20 30 40 50 60 70 80

Fig. 2 Associations with FT (in per cent)

To shed light on possible differences between those who donated to FT or bought a FT

product and those who did not regarding what they associate with FT, only the responses of
the 52 individuals in treatments T1 and T2 were displayed in Figure 3, i.e., those treatments
that involved FT, and distinguish between the two groups.4 While most response rates are
similar to the overall figures and not very different for the two groups, two stark differences
emerge: Only 21% of those who bought FT products or donated to FT regarded FT as
involving a donation, whereas that view was held by 45% of those who did not. (Better)
Standards of production were regarded by 86% of the FT supporters as a characteristic of FT,
while the figure was only 53% for non-supporters. This suggests that standards of produc-
tion and development aid are the most important motivations behind purchasing FT
products.
These survey findings are in line with our experimental results. In the present experiment,
the WTP a premium is shown to be higher for FT coffee sold at a premium than for regular
coffee sold at the same premium that benefits the charity (T1 vs. T3). The survey shows that
86% of the supporters of FT associate better product standards with FT, an attribute that the
regular coffee does not have. In addition, a much smaller share associates better taste and
more healthy products with FT, non-altruistic motives for paying a premium. This explains
the difference in the WTP between T1 and T3. People are willing to pay for product
attributes in addition to their WTP for a good cause (as shown by the attribute “development
aid” and “good feeling”) that FT shares with the charity MISEREOR, as both organizations
support a similar group of people—peasant farmers in the developing world. The survey also
shows that only a small share of FT supporters (21%) regards purchases of FT products as a
donation whereas contributions to the charity MISEREOR clearly is such a donation. This
explains why people donated much less directly to FT than to the charity (T2 vs. T4). It also
explains partly that the willingness to contribute to FT is much higher in the indirect

4
In order to increase sample size we aggregated T1 and T2, even though some of the individuals in T2 who
did not donate to FT might have purchased a FT coffee in T1.
The Importance of Indirect Transfers for Consumer Willingness to Pay 379

no donation/ purchase donates to/ purchases FT

Development Aid 0.76

0.71

Good Feeling 0.34

0.36

Health 0.05
0.14

Taste 0.18
0.21

Standards of Production 0.53

0.86

Donation 0.45
0.21

Fig. 3 Characteristics associated with FT (in per cent)

treatment than in the direct treatment (T1 vs. T2)—FT is seen as a charity by only a few and
the product attributes are present in T1 but absent in T2.
What our survey cannot explain, however, is why there is such a stark difference between
T3 and T4, involving only the charity. A difference in product attributes is absent and the
charitable cause is obviously the same in both treatments. Still people are willing to donate
significantly more if the donation is included in the product price rather than made
separately. FT uses that very mechanism—including the donation in the product price.
Without comparing the two FT treatments to the two charity treatments one might be
tempted to ascribe the entire WTP a premium for FT products (as shown in T1) exclusively
to the product attributes; our survey might have been interpreted that way if it were evaluated
independently of the experiment. Seen in the light of the experimental results, however, the
survey results are in line with previous findings: They explain the differences between T1
and T2 on the one hand and T3 and T4 on the other hand. They cannot, and do not, explain
the different effectiveness of direct and indirect donation vehicles.
The questionnaire included two further attitudinal questions on FT. First, people were
asked to what extent they agreed with the statement “I trust that the money will be used for
its stated purpose.” While the overall trust was relatively high, there was no clear difference
between supporters of FT and those that did not purchase a FT product or donate to FT. If at
all, trust was higher among the non-supporting group. The second statement was “Do you
think FT is necessary for the survival of farmers?” Interestingly, supporters and non-
supporters of FT regard its necessity quite similarly, the average values are 2.29 and 2.32
on a scale of 1 (“absolutley agree”) to 6 (“never”). Thus, neither difference in trust in the FT
mechanism nor in the belief of its necessity for the survival of farmers can explain donation
patterns.
The second purpose of the present survey was to check whether the different results were
driven by heterogeneous samples by using the information from our questionnaire on the last
day of the experiment. A number of socio-economic characteristics, such as gender, number of
semester, field of study, marital status, number of siblings, visits to developing countries,
variables indicating the wealth status, and religiousness were compiled. As is well-known,
charitable giving is affected by gender, wealth and income, age, and other socio-economic
characteristics. Mann–Whitney U test and the Kruskal–Wallis H test were used to test whether
the samples were drawn from the same population. Fortunately, samples were comparable with
380 H. Koppel, G.G. Schulze

respect to all socio-economic characteristics, and only some differences in the field of study
were found, which might be due to different locations of university coffee shops within the
campus. However, Bekkers and Wiepking (2011) show in their review on philanthropy that
field of study does not have a clear effect on donation behaviour. Given the stark differences in
treatment outcomes, the authors are confident that the limited differences in programme
composition do not invalidate our results.

Discussion

In this paper, the authors have inquired into the determinants of the willingness of consumers
to pay a premium for FT products and shown that this WTP extra is driven to a large part by
the transfer mechanism that the FT model uses and only to a smaller part by the specific
attributes that FT products have. Up to now the literature on the motivation to buy FT
products has focused on these specific attributes such as better quality and better
product/labour standards. This exclusive focus has overlooked the effect of the mechanism
with which the donations were raised. Furthermore, people are shown to be more inclined to
make small donations if they are linked to the purchase of a good rather than when they are
made independently through a donation box. This phenomenon was not limited to FT
products but applied to charities in general and it contributed substantially more to the
WTP extra for the FT good than the specific attributes of FT products. In other words, the FT
model appeals to consumers not only because of its specific approach to help disadvantaged
producers in the global South, it also uses a specific marketing concept to its advantage by
selling products at a premium that goes to a good cause. By linking two elements in one
transaction, FT companies are able to mobilize significantly more resources than they
otherwise would.
The present findings are in line with earlier studies that show that framing substantially
influences the decision to give to other people. Contrary to earlier studies, however, the
present study shows that already the simple mechanism through which the money is raised is
decisive for the amount of resources mobilized for a certain cause. By making potential
donors become consumers, the FT model is able to raise more money than through direct
donations for the same cause.
The present findings could only be derived through a difference-in-difference experi-
mental approach by comparing direct and indirect donation mechanisms for FT and for a
well-known charity. Product attributes play a role for FT treatments but are absent in the
charity treatment; the comparison of these two sets of treatments only allows assessing the
relative importance of the two effects. The survey results are in line with our experimental
findings and should be interpreted in connection with our experimental results. Surveys
cannot ask convincingly for the effect of a specific transfer mechanism, which individuals
might not even be aware of. This might lead to the conclusion that the entire WTP a
premium for FT products is attributable to the product characteristics. The authors show
that those attributes do play an important role in explaining the success of the FT model, but
that the specific transfer mechanism is even more important.
Thus, the experiment raises two important issues. First, it suggests implications for the
policy of other philanthropic organizations as our findings may well be applicable to other
philanthropic causes. The research was motivated by the soaring sales of the FT organiza-
tions, which use the indirect donation mechanism as exclusive business model, but our
primary result holds for a well-known charity as well. Anecdotal evidence suggests that our
findings are not limited to our specific setup. For example, the “Bauhaus Archive” for design
The Importance of Indirect Transfers for Consumer Willingness to Pay 381

collects more than four times as much money through the sale of overpriced chocolate than
through a donation box at the entrance.5
If the present findings extend beyond FT and the charity that were used, as expected,
charitable NGOs and philanthropic foundations may, in addition to raising money through
direct donation, use the indirect donation mechanism by selling consumption products at a
premium. This may require cooperating with a firm that sells the respective goods (as not all
NGOs may engage in selling products themselves). Furthermore, the present findings
suggest that it may be in the interest of the cooperating firm to offer one variant with and
another one without a donation component of the otherwise identical product, unlike in
CRM which bundles entire sales of a product with a donation component. Such simple price
discrimination offers caring consumers a welcome and easy possibility to donate while it
does not deter price-sensitive consumers from purchasing the cheaper product variant. With
this simple heuristics, consumers can choose the variant that serves them best (cf. Thøgersen
et al. 2012). Individuals may also only occasionally purchase the variant that includes the
donation component.
Second, there were clear evidence that people are more willing to donate small amounts if
their donation is linked to a purchase, but it is not clear as to why this is so. It may turn out
that the logic of gift exchange experiments extends to the mechanism of donations. Falk
(2007) shows that the frequency of donations significantly increases if a solicitation letter
accompanies a gift. More generally, people are more inclined to make a gift if they get
something in return. In the context at hand, people may regard the purchased coffee with its
specific product attributes as something received in return (for the donation plus the ordinary
price), and thus are more inclined to make the donation. They may not have the same
association when they buy coffee and make a donation in a box because these two
transactions are not as clearly linked. It may also turn out that the transaction costs are
lower if the two transactions are linked, even though direct and indirect donations are both
made in connection with a purchase and a payment process, and in both setups the sales
person communicated in the same way the possibility of making a donation. For instance, it
could be that the mental transaction costs are lower. Whether these hypotheses stand up to
closer scrutiny is left for future research.

Acknowledgements The authors are grateful to Max Albert, Björn Frank, Nils Hesse, Sevtap Kestel,
Krisztina Kis-Katos, and two anonymous reviewers for helpful comments. Of course, all remaining errors
are that of the authors.

Appendix 1: Instructions for Sales Persons

& Do not say anything about an experiment! Be as you normally are.

& The experiment is conducted over 2 weeks, on Tuesday, Wednesday, and Thursday.
& In the first week, nothing changes, except that you fill the tally sheet.
& In the second week, the experiment is conducted and the tally sheet is extended.
& Be extremely careful when filling the tally sheet and do not mix up the columns in order
to record accurately sales of coffee, other products, donation and FT coffee.

5
Telephone conversation with Ms. Seydell, Bauhaus archive. The Bauhaus Archive is a non-profit organi-
zation that runs a museum and an archive on Bauhaus designs in commemoration of the group of artists,
architects and designers in Germany that was founded in 1919 and dissolved by the Nazis (http://
www.Bauhaus.de).
382 H. Koppel, G.G. Schulze

& It is very important that you do everything meticulously and with a maximum of
accuracy.
& If you have to sell different brands of coffee, it must be visible for consumers that they
get the coffee that they paid for.
& If a consumer is asking why you are doing this never say anything about an experiment.
& New coffee (T1): “It is a new coffee and we wanted to give you a choice.”
& Donation box (T2 and T4) or coffee with donation premium (T3): “We wanted to give
you the opportunity to donate for this cause.”
& Never open the donation box! If it is full you have to install another one.
& For T4: As you sell at different prices, the whole amount donated through the higher
price must be donated for the mentioned purpose. Please provide a receipt. If you do not
want to donate it, give it to us and we will donate it.

Appendix 2: Experimental Settings

Fig. 4 Treatment setting (1)

The Importance of Indirect Transfers for Consumer Willingness to Pay 383

Fig. 5 Treatment setting (2)

Fig. 6 Treatment setting (3)

384 H. Koppel, G.G. Schulze

Appendix 3: Sales Statistics and Imputation

Tables 2, 3, 4, and 5 report the sale statistics for all treatments for both subsequent weeks.

Table 2 Sales statistics of treat-

ment 1 (in units) Day Normal coffee Fair trade coffee Other products

Tuesday 38 44
Wednesday 39 47
Thursday 23 12
Tuesday 9 32 53
Wednesday 15 23 43
Thursday 8 16 19

Table 3 Sales statistics of

treatment 2 (in units) Day Normal coffee Donation Other products Donation

Tuesday 103 294

Wednesday 123 395
Thursday 98 310
Tuesday 74 1 395 1
Wednesday 106 2 311 0
Thursday 101 2 360 0
Total donation: 3.00 Euro

Table 4 Sales statistics of

treatment 3 (in units) Day Normal Coffee including Other
coffee donation products

Tuesday 14 3
Wednesday 9 3
Thursday 12 5
Tuesday 5 10 6
Wednesday 8 2 1
Thursday 5 10 2
The Importance of Indirect Transfers for Consumer Willingness to Pay 385

Table 5 Sales statistics of

treatment 4 (in units) Day Normal coffee Donation Other products Donation

Tuesday 89 9
Wednesday 63 12
Thursday 70 9
Tuesday 76 16 1 1
Wednesday 46 13 10 1
Thursday 55 14 11 0
Total donation: 5.40 Euro

For the direct donation mechanism we compute how many people would have donated
the same total amount if direct donations were restricted to 10 Eurocent as in the case of
indirect donations. In treatment T2, a total of six people made a donation, which summed up
to 3.00 Euros (Table 3). That is a per capita donation of 50 Eurocent for those donating. One
donating individual did not buy coffee, but another product, which we have to exclude for
reasons of comparison, because this person would not have had a chance to donate in the
indirect donation treatments 1 and 3. Thus, we have a total donation of 2.50 Euros for the
five people that donated and bought coffee. If each individual were restricted to a 10
Eurocent donation, the sum of 2.50 Euros would have been achieved if 25 people had
donated out of the 281 people that bought coffee in that treatment. Thus, when comparing
treatments T1 and T2 we compare the actual distribution of treatment T1 with 71 people
donating 10 Eurocent each through their FT coffee purchases at a higher price and 32 people
buying coffee at the cheaper price and donating nothing with the imputed distribution in
treatment T2, in which 25 out of 281 people donated 10 Eurocent (Table 2).
The analogous calculus is done for T4. In treatment T4, a total of 45 people made a
donation, which summed up to 5.40 Euros (Table 5). That is a per capita donation of 12
Eurocent for those donating. Two donating individuals did not buy coffee, and are therefore
excluded. The total donations for the 43 people that bought coffee are therefore 5.16 Euro. If
donations were restricted to 10 Eurocent 51.6 (≈52) individuals would have achieved a total
donation amount of 5.16 (≈ 5.20) Euro. In order to compare the distributions between
treatments a conservative rounding was made to construct the imputed distribution. The
expectation was that tied transfers raise on average more donations. The detection of a
treatment effect, assuming that 52 individuals donated 10 Eurocent, is less likely than the
alternative assumption that 51 individuals donated 10 Eurocent. Thus, the imputed distribu-
tion for T4 contains 52 out of the 177 individuals donating 10 Eurocent.

References

Akerlof, G., & Kranton, R. (2000). Economics and identity. Quarterly Journal of Economics, 115, 715–753.
Andorfer, V. A., & Liebe, U. (2012). Research on fair trade consumption—A review. Journal of Business
Ethics, 106, 415–435.
Andreoni, J., & Petrie, R. (2004). Public goods experiments without confidentiality: A glimpse into
fundraising. Journal of Public Economics, 88, 1605–1623.
Arnot, C., Boxall, P. C., & Cash, S. B. (2006). Do ethical consumers care about price? A revealed preference
analysis of fair trade coffee purchases. Canadian Journal of Agricultural Economics, 54, 555–565.
Barham, B. L., Callenes, M., Gitter, S., Lewis, J., & Weber, J. (2011). Fair trade/organic coffee, rural
livelihoods, and the “agrarian question”: Southern Mexican coffee families in transition. World Devel-
opment, 39, 134–145.
386 H. Koppel, G.G. Schulze

Basu, A. K., & Hicks, R. L. (2008). Label performance and the willingness to pay for fair trade coffee: A
cross-national perspective. International Journal of Consumer Studies, 32, 470–478.
Bekkers, R., & Wiepking, P. (2011). Who gives? A literature review of predictors of charitable giving
I—Religion, education, age and socialization. Voluntary Sector Review, 2, 337–365.
Beuchelt, T. D., & Zeller, M. (2011). Profits and poverty: Certification's troubled link for Nicaragua’s organic
and fairtrade coffee producers. Ecological Economics, 70, 1316–1324.
Chang, C.-T. (2008). To donate or not to donate? Product characteristics and framing effects of cause-related
marketing on consumer purchase behavior. Psychology and Marketing, 25, 1089–1110.
Cornes, R., & Sandler, T. (1984). Easy riders, joint production, and public goods. The Economic Journal, 94,
580–598.
Cornes, R., & Sandler, T. (1994). The comparative static properties of the impure public good model. Journal
of Public Economics, 54, 403–421.
Cranfield, J., Henson, S., Northey, J., & Masakure, O. (2010). An assessment of consumer preference for fair
trade coffee in Toronto and Vancouver. Agribusiness, 26, 307–325.
de Ferran, F., & Grunert, K. G. (2007). French fair trade coffee buyers’ purchasing motives: An exploratory
study using means-end chains analysis. Food Quality and Preference, 18, 218–229.
de Pelsmacker, P., Driesen, L., & Rayp, R. (2005). Do consumers care about ethics? Willingness to pay for
fair-trade coffee. Journal of Consumer Affairs, 39, 363–385.
Falk, A. (2007). Gift exchange in the field. Econometrica, 75, 1501–1511.
FLO. (2010). Growing stronger together, annual report 2009–10. Bonn: Fair Trade Labelling Organizations
International.
Gupta, S., & Pirsch, J. (2006). The company-cause-customer fit decision in cause-related marketing. Journal
of Consumer Marketing, 23, 314–326.
Hainmueller, J., Hiscox, M. J., & Sequeira, S. (2011). Consumer demand for the fair trade label: Evidence
from a field experiment. Available at SSRN: http://ssrn.com/abstract=1801942.
Harrison, G., & List, J. A. (2004). Field experiments. Journal of Economic Literature, 42, 1013–1059.
Hayes, M. (2006). On the efficiency of fair trade. Review of Social Economy, 64, 447–468.
Kadow, A. (2011). The fair trade movement: An economic perspective. mimeo, Department of Economics,
University of Glasgow. http://www.eea-esem.com/files/papers/eea-esem/2011/2614/Kadow_FT_EEA.pdf.
Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1990). Experimental tests of the endowment effect and the
Coase theorem. Journal of Political Economy, 98, 1325–1348.
Kotchen, M. (2006). Green markets and private provision of public goods. Journal of Political Economy, 114,
816–834.
Levitt, S., & List, J. A. (2007). What do laboratory experiments measuring social preferences reveal about the
real world? The Journal of Economic Perspectives, 21(2), 153–174.
List, J. A., Berrens, R. B., Bohara, A. K., & Kerkvliet, J. (2004). Examining the role of social isolation on
stated preferences. The American Economic Review, 94, 741–752.
Loureiro, M. L., & Lotade, J. (2005). Do fair trade and eco-labeling in coffee wake up the consumer
conscience? Ecological Economics, 53, 129–138.
Maseland, R., & de Vaal, A. (2002). How fair is fair trade? De Economist, 150, 251–272.
Maseland, R., & de Vaal, A. (2008). Looking beyond the cooperative: Fair trade and the income distribution.
In R. Ruben (Ed.), The impact of fair trade (pp. 223–237). Wageningen: Academic.
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, 170–180.
Polonsky, M. J., & Macdonald, E. K. (2000). Exploring the link between cause related marketing and brand
building. International Journal of Nonprofit and Voluntary Sector Marketing, 5, 46–57.
Raynolds, L. T. (2009). Mainstreaming fair trade coffee: From partnership to traceability. World Development,
37, 1083–1093.
Reinstein, D. A., & Song, J. (2012). Efficient consumer altruism and fair trade. Journal of Economics and
Management Strategy, 21, 213–241.
Rotaris, L., & Danielis, R. (2011). Willingness to pay for fair trade coffee: A conjoint analysis experiment with
Italian consumers. Journal of Agricultural and Food Industrial Organization, 9 (1), Article 6.
Schokkaert, E. (2006). The empirical analysis of transfer motives. In S.-C. Kolm & J. M. Ythier (Eds.), Handbook
of the economics of giving, altruism and reciprocity (pp. 128–181). Amsterdam: North-Holland.
Schollenberg, L. (2012). Estimating the hedonic price for fair trade coffee in Sweden. British Food Journal,
114, 428–446.
Soetevent, A. R. (2005). Anonymity in giving in a natural context—A field experiment in 30 churches.
Journal of Public Economics, 89, 2301–2323.
Steinrücken, T., & Jaenichen, S. (2007). The fair trade idea: Towards an economics of social labels. Journal of
Consumer Policy, 30, 201–217.
The Importance of Indirect Transfers for Consumer Willingness to Pay 387

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, 215–241.
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. The Journal of Consumer Research, 24, 434–446.
Thøgersen, J., Jørgensen, A. K., & Sandager, S. (2012). Consumer decision making regarding a “green”
everyday product. Psychology and Marketing, 29, 187–197.
Valkila, J. (2009). Fair trade organic coffee production in Nicaragua—sustainable development or a poverty
trap? Ecological Economics, 68, 3018–3025.
Varadarajan, P. R., & Menon, A. (1988). Cause-related marketing: A coalignment of marketing strategy and
corporate philanthropy. The Journal of Marketing, 52(3), 58–74.
Vicary, S. (1997). Joint production and the private provision of public goods. Journal of Public Economics,
63, 429–445.
Weber, J. (2007). Fair trade coffee enthusiasts should confront reality. Cato Journal, 27, 109–117.
Wiepking, P., & Bekkers, R. (2012). Who gives? A literature review of predictors of charitable giving
II—Gender, marital status, income and wealth. Voluntary Sector Review, 3, 217–246.
Yechaim, E., Barron, G., Erev, I., & Erez, M. (2003). On the robustness and the direction of the effect of
cause-related marketing. Journal of Consumer Behavior, 2, 320–332.
Zander, K., & Hamm, U. (2010). Consumer preferences for additional ethical attributes of organic food. Food
Quality and Preference, 21, 495–503.
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.