Fair Trade and donations: Two possibilities to contribute to poverty alleviation in daily purchase decisions - Do consumers care? Nina Langen, Carola Grebitus and Monika Hartmann

Fair Trade and donations: Two possibilities to contribute to poverty alleviation in daily purchase decisions - Do consumers care? Nina Langen, Carola Grebitus and Monika Hartmann

Nina Langen1, Carola Grebitus2, Monika Hartmann2

Center for Development Research (ZEF B)

nina.langen@uni-bonn.de

Internet: www.zef.de
2 University of Bonn

Institute of Food and Resource Economics

Department of Agricultural and Food Market Research

Nussallee 21 *   53115 Bonn, Germany

Phone +49 (0) 228 73-3582

carola.grebitus@ilr.uni-bonn.de

monika.hartmann@ilr.uni-bonn.de

Internet: www.ilr.uni-bonn.de

In 2005 1.4 billion people in developing countries were living below the poverty line which means they had less than $1.25 available per day (Worldbank 2008). At the same time, the “ethical consumer” is increasingly discussed (e.g. Schulz 2008, Harrison et al. 2005).

The trend, that consumers account for ethical concerns in their purchase decision, seems to be confirmed by increasing sales volumes of Fair Trade (FT) products (Transfair 2008) as well as the increased visibility of promotion products of which a part of the sales revenue is a donation to the respectively promoted project (the so called cause related marketing (CRM)). The market for certified coffee, e.g. FT and organic, grows at double digit rates since 2000. FT labelled coffee shows a growth rate of 46% worldwide between 2004 and 2006 and 14% in 2008 in Germany (Byers et al. 2008; Transfair 2008). Organic coffee has a market share of 3.5% in Germany and shows as well double digit growth (BLE 2008). Besides single labelled coffee double certified coffee is a new trend. Worldwide nearly 50% of the FT coffee is double certified carrying also an organic certification (Byers et al. 2008). FT and to a certain extent also organic produced products are considered as ethical products (De Pelsmacker et al. 2005). The reason is that the production of FT and organic products follows specific rules which restrict the output of environmental damage, constitutes specific working conditions as well as animal well being and have the goal to reduce poverty by empowerment of marginalised producers etc.

These developments are important because purchasing FT products or donations to e.g. development aid organisations for instance in form of CRM are two of the most common options for people in industrialised countries to make an individual contribution to poverty alleviation. But what if consumers face the possibility to support producers in developing countries through a product purchase? Which (product) features play a major role? Are the concerns about working conditions, fair wages etc. just lip services? Does the responsible and conscious FT consumer take care about the differences between FT and donations?

The aim of our study is to investigate these questions by use of coffee purchase as an example. We choose coffee as a study object because it is the favourite beverage of German consumers (Deutscher Kaffeeverband 2009). Therefore we can assume that consumers really care about their coffee choice and that the results will benefit from consumers high involvement. Also, coffee is a typical product from developing countries. Thus, it is well suited for a CRM campaign, especially a donation to a charity organisation working in a coffee producing country. The market place confirms our assumption: in May 2008 Dallmayr launched a coffee called “Ethiopia” of which per sold package five trees are planted in Ethiopia. In this regard, the campaign is close to a regular donation and comparable to FT.

Consumers might contribute to charity or buy FT products for various reasons. Therefore, preferences are expected to vary across individuals. Without understanding the form and extent of preference heterogeneity it will be difficult to make assumptions regarding the relationship between donations, FT and CRM activities in combination with a product purchase. Against this background, we conducted a choice experiment with n=481 in Germany to investigate consumers’ preferences for differently labelled coffee. Our objectives are: 1. to compare attitudes towards important coffee and food features with the choices resulting from the choice experiment; 2. to perform a market segmentation in order to distinguish and compare consumers with preferences for FT and those with preferences for donations.

The remainder of the paper is as follows: in the next section theoretical background information is given. In section 3 the methodological background is described. Section 4 provides estimation results from the econometric analysis. We finish with some concluding remarks.
2. Theoretical Background
2.1. The Phenomenon of Ethical Consumption

Ethical consumption as a form of market behaviour became obvious in the last three decades (Harrison et al. 2005). The definition of the ethical consumer is seldom exclusive and mostly descriptive. Harrison et al. (2005) explain ethical purchase behaviour as a traditional consumption plus a concern. The assumption that a consumer purchases the cheapest good which is fitting his needs leads to the definition of the traditional purchasing. If people deviate from the normal assumption and consider other concerns like working conditions for disadvantaged producers in developing countries or absence of child labour in their purchase, then their shopping decision can be called ethical purchase behaviour. The motives of people to buy a certain kind of product are manifold. They vary from a concern for environmental issues over political to religious, spiritual and social motives. The one important common point which is independent from the motives of consumption is that ethical consumers bear in mind the effect their purchase decision has not only on themselves but especially “on the external world around them” (Harrison et al. 2005). When we combine the definition of Tallontire et al. (2001) - an ethical consumer is a customer feeling responsible towards society - and that of De Pelsmacker et al. (2005) saying that these feelings are expressed by means of his purchase behaviour then we arrive at the relationship between FT and charitable giving. Besides this, in the context of FT it is often talked about ethical or responsible consumers (Ruwet 2007).

If we follow Priller and Sommerfeld (2005) and define donations as a form of social participation, a contribution to welfare production which is able to maintain and open up social connecting forces in modern societies, we can state that one precondition out of a bundle of motives for a donation is that the giving individual is an ethical being. And if we define, according to Nicholls and Opal (2005), FT as a kind of an alternative market mechanism which is neither donations nor non-profit but a form of political and ethical consumption, an individual buying fairly traded products is an ethical consumer. Therefore we can link ethical consumption patterns, donations, FT and CRM.
2.2. Donations in Germany
2.2.1. Donation volume in Germany

Regularly available numbers regarding donations to non-profit organisations in Germany are provided by the GfK, Deutscher Spendenmonitor, and the National Income statistic (see table 1). Their results differ strongly with respect to the donation amounts (from 2.6 to 7 billion EUR/year) and the donation purpose in Germany (Priller and Sommerfeld 2005). For instance TNS Infratest (2008) reports that development projects benefit of about 19% of the 2.8 Billion EUR which were donated in 2007 in Germany. GfK (2008a) reports different percentages on a different basis: in the first half of 2008 9.3% of humanitarian help, which is 80% of the total donation volume, was given to long-term development projects and 18.7% went to first aid.

Table 1: Charitable giving survey data in Germany

Survey	Year	Billion EUR	Method
GfK Charity Scope	2005	2,4	Respondents at least 10 years old, 10.000 interviews, diary, monthly
	2006	2,0
	2007	2,0
Deutscher Spendenmonitor	2005	3,5	Respondents at least 14 years old, 4.000 interviews, Face-to-Face, yearly
	2006	3,4
	2007	2,8
National Income Statistic	2001	2,9	Taxpayer, complete inventory count

Source: Sommerfeld (2008) for the data until 2006, data for 2007 from GFK (2008a) and TNS Infratest (2008), data for the national income statistic from Buschle (2006).

2.2.2. Socio-demographic and regional differences of donors

The willingness to spend and the amount of donations depend on age (younger people spend less), economic situation which often depends on education level (wage earner spend more than trainees or unemployed people, retirees and housewives spend most), religious denomination (raises the probability of contribution, while there is not much of a difference between Catholics or Protestants) and the number of children in the household (positive correlated with donation) (Buschle 2006). Related to the entity of taxpayers most of the donors can be found among the married couples with three or more children. No differences in terms of donation habits can be found regarding gender (GfK 2008b). In 2008 more than 50% of the monetary donations come according to GfK (2008b) from people older than 60 years (which are only 26% of the panel) and more than 50% of these are given by those people older than 70 years.

As can be seen in table 2 in recent years the number and the sales of FT products increased. The overall sales volume of FT products with the certification mark is 142 Mio. € in 2007 (Transfair 2008). 70% of all FT products are also certified organic. Coffee is the front runner of the FT products in Germany: it has more than 50% share of sales (LZ Net 2007); with a market share of 1% in 2005 (Krier 2005).

Table 2: Fair Trade sales figures in Germany 2004-2007

Year	Sales volume of fair traded products [Mio EUR]	% to PY	Sales volume with Fair Trade labeled products [Mio EUR]	% to PY	coffee sales [t]	% to PY	Fair Trade and organic certified [%]
2004	99	7	n.s.	n.s.	n.s.	n.s.	n.s.
2005	121	22	73	25	3600	9	64
2006	155	30	110	51	4000	18	70
2007	193	23	142	29	4600	11	> 70

Source: Forum Fairer Handel (2004, 2006, 2007, 2008).

The German consumer initiative stated that women buy more (40% buy FT) than men (33% buyers), that higher income classes (more than 2500 EUR/month) are more likely FT buyers than people with a lower income, higher educated people (high school) are more often buyer (50%) than less educated (32%) (Verbraucher Initiative 2007).

Choice experiments are a flexible approach to record preference data from individuals in artificial but at the same time realistic situations. Realistic in the sense, that a situation is created where an individual should compare alternatives through their attributes and come to a decision between the alternatives. For more information see e.g. Adamowicz et al. (1998).

To analyze the data gathered with the choice experiments random utility models based on latent class / finite mixture modelling are applied (e.g. Scarpa and Thiene 2005). Latent class analysis assumes that within the basic population different groups or segments with varying preference structures, which result in different preference parameters β, can be distinguished. In a simultaneous process the latent class model (LCM) estimates utility parameters of the different attributes and the probability of the affiliation of the respondents to these components. This means that it simultaneously determines and describes product choice and group membership while it separates the sample in several, homogenous subgroups which map the heterogeneity in the population (Holmes and Adamowicz 2003). Every consumer is attached to that segment where he has the biggest probability of affiliation to a segment, near 1 (Gensler 2003). The respondents are grouped into the segments based on statistical information criteria (Greene 2003). The Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) are frequently used to determine the number of segments. Therefore, model parameters are estimated for increasing numbers of segments until the point where an additional segment does not improve the model fit according to the named criteria (Wedel and Kamakura 2000).

During the coffee choice experiment participants were asked to make six choices. Each choice set consisted of four coffee packages representing different attribute bundles and various attribute levels. The experimental design included four coffee attributes with different levels each. Namely, Price: 2.99 EUR, 3.99 EUR, 4.99 EUR, 5.99 EUR; Organic: no, yes; Label: no label, FT, charity organization; Donation: no donation, 0.2 EUR or 0.5 EUR or 1 EUR directly to the producer.

Data from the consumer survey are analyzed by means of LCM. The optimal number of classes in the LCM was identified by assessing the AIC, its variant AIC 3, and BIC as well as the log likelihood statistic from 1 to 5 class models. With the increasing number of classes the log likelihood statistic as well as the AIC and BIC values present remarkable changes (the values become smaller) and the R² value increases strongly. A conditional bootstrap with 500 draws showed that the 5 class model does not increase significantly model fit. Thus, we chose the 4 class model which gives overall best results. These results are reported in table 3.

Table 3: Parameter estimates of the 4 class model

The comparison of the coefficients reveals differences between the classes with regard to the highest valued attributes and allows the naming of the groups: class 1 - the price conscious coffee shoppers - are the most price sensitive and with 43% of all respondents belonging to it the biggest class, Class 2 - the donors - love the donation in combination with the product purchase but not FT and is with 28% of the respondents a bit bigger than class 3 (25%), class 3 - the FT and organic lovers - highly value organic production and FT and class 4 - the denier - is the smallest class with 2% and dislikes any kind of label on the coffee. The different parameter estimates for the classes support the existence of preference heterogeneity in the sample for coffee attributes. Results from the model of choices indicate that class 1 is a bit similar to class 3 in having positive parameter estimates for organic production, FT and the donation amount labelled on the coffee pack. The high z-values lead to the conclusion that all the estimates for class 1 are significant. Class 2 cannot be analysed with respect to organic production and FT because the z-values are not significant. Interestingly class 2 shows the lowest parameter estimates compared to the other classes for the labelled amount of donation. Class 3 is the least price sensitive class and has a strong preference for organic and FT indicated by the high magnitudes of the parameter. No clear statement can be given about the preference for a high amount of donation because the z-values for 0.2 and 1 EUR are not significant. Nevertheless, we can state that class 3 prefers a donation of 0.5 EUR over no donation and therefore CRM activities positively influence the utility of class 3 members. Class 4 members strongly dislike organic and FT labelling and “no amount” of donation influences the members’ utility positively.

It becomes evident that half of the respondents have a positive attitude towards FT and CRM activities. As these respondents are not identical cannibalism between CRM activities and FT is not very likely. Thus, we can state that at least some German coffee drinkers care about poverty of marginalised people in developing countries in daily purchase decisions.

BLE (2008): Kaffee - der Deutschen liebstes Heißgetränk, http://www.oekolandbau.de/grossverbraucher/studien-und-info-materialien/marktdaten/wachstumsmarkt-kaffee-september-2008/, 25.09.2008.

Buschle, N. (2006): Spenden in Deutschland, Ergebnisse der Einkommensteuerstatistik 2001, in: Statistisches Bundesamt (ed.): Auszug aus Wirtschaft und Statistik (2), 151-159, Wiesbaden.

Byers et al. (2008): Value-adding standards in the North American food market - Trade opportunities in certified products for developing countries, FAO Commodities and Trade Technical Paper (11).

De Pelsmacker, P. et al (2005): Do Consumers Care about Ethics? Willingness to Pay for Fair-Trade Coffee, The Journal of Consumer Affairs, 39 (2) 363-385.

Deutscher Kaffeeverband (2009): Kaffeetrends: Zertifiziert und „light“, http://www.kaffeeverband.de/186.htm, 08.01.09.

Forum Fairer Handel (several volumes): Zahlen zum Fairen Handel in Deutschland, http://www.forum-fairer-handel.de/cms/, 20.02.2009.

Gensler, S. (2003): Heterogenität in der Präferenzanalyse: ein Vergleich von hierarchischen Bayes-Modellen und Finite-Mixture-Modellen, in: Albers et al. (eds.): Beiträge zur betriebswirtschaftlichen Forschung, 107, Deutscher Universitäts-Verlag, Diss., Frankfurt/Main.

GfK (2008a): GfK Charity Scope – Der Spendenmarkt 2007, http://www.gfk.com/imperia/md/content/ps_de/chart_der_woche/2008/kw16_08_charity.pdf, 28.04.2008.

GfK (2008b): Spendensituation und Spendenverhalten – Aktuelle Entwicklungen, Charity Scope 2008: Spenden- motivation – Eine soziostrukturelle Analyse.

LZ Net (2007): Deutschland spielt bei fair gehandelten Produkten in der Top-Liga, 29.03.2007.

Priller, E. and Sommerfeld, J. (2005): Wer spendet in Deutschland? Eine sozialstrukturelle Analyse, WZB Discussion Paper, Wissenschaftszentrum Berlin für Sozialforschung.

Schulz (2008): Megatrend Nachhaltigkeit - Marktpotenziale von LOHAS & Co., https://umho.uni-hohenheim.de/fileadmin/einrichtungen/umho/Aktuelle_Berichte/Megatrend_Nachhaltigkeit_-_Marktpotenziale_von_LOHAS___Co._-_25._April_2008.pdf, 03.06.2008.

Sommerfeld (2008) : Die Bedeutung des Spenden- und Gemeinnützigkeitsrechts in der Praxis - Eine Analyse von Experteninterviews, http://www.dzi.de/Spendenstatistik/tagung13102008/Sommerfeld.pdf, 02.02.2009.

Tallontire et al. (2001): Ethical Consumers and Ethical Trade: A Review of Current Literature, Policy Series 12, Chatham, UK: Natural Resources Institute.

TNS Infratest (2008): Bundesbürger spendeten in 2007 weniger, http://www.tns-infratest.com/presse/pdf/Presse/20080116_TNS%20Infratest_Spendenmonitor2007.pdf, 16.01.2008.

Transfair (2008): Doppelt gut - Bio im Fairen Handel, http://www.transfair.org/presse/detailseite-presse/browse/4/article/45/doppelt-gut.html, 20.02.2008.

Verbraucher Initiative e.V. (2007): Fairer Handel: Stetiges Wachstum, http://www.fair-feels-good.de/pdf/86.pdf?title=Fairer_Handel:_Stetiges_Wachstum, 30.05.2008.