Food Policy 44 (2014) 167176

Contents lists available at ScienceDirect

Food Policy
journal homepage: www.elsevier.com/locate/foodpol

European consumer preferences for beef with nutrition and health

claims: A multi-country investigation using discrete choice experiments
Lynn Van Wezemael a, Vincenzina Caputo b,, Rodolfo M. Nayga Jr. b,c,d, George Chryssochoidis e,
Wim Verbeke a
a

Department of Agricultural Economics, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium
Department of Food and Resource Economics, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul, Republic of Korea
Department of Agricultural Economics and Agribusiness, University of Arkansas, 217 Agriculture Building, Fayetteville, AR 72701, USA
d
Norwegian Agricultural Economics Research Institute, PO Box 8024, Dep Storgata 2/4/6, 0030 Oslo, Norway
e
University of East Anglia, Chancellors Drive, Norwich NR4 7TJ, UK1
b
c

a r t i c l e

i n f o

Article history:
Received 15 January 2013
Received in revised form 24 October 2013
Accepted 24 November 2013

Keywords:
Beef
Choice experiment
Claims
Consumer
Health
Logit
Nutrition
Preference

a b s t r a c t
The European Union (EU) nutrition labelling policy aims to facilitate consumers food choice, stimulate
innovation and facilitate the circulation of foods bearing claims across countries. However, the beef
industry has not fully taken advantage of utilizing nutrition and health claims based on the EU nutrition
labelling policy to differentiate beef products in the market. This study investigates consumer preferences
for nutrition and health claims on lean beef steak. Two choice experiments were conducted among 2400
beef consumers in four EU countries (Belgium, France, the Netherlands, United and United Kingdom).
Multinomial logit and error component models were estimated. Our results generally suggest that
consumer valuation of nutritional and health claims varies across countries. In Belgium, the Netherlands
and France, nutrition and health claims on saturated fat yielded higher utilities than claims on protein
and/or iron, while the opposite was found among consumers in the UK. The results imply that marketing
opportunities related to nutrition and health claims on beef are promising, but that different nutritional
marketing strategies are necessary within different countries.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Consumer perceptions of the nutritional value of beef are not
unequivocally positive (Van Wezemael et al., 2010). Personal
health concerns have arisen, among others about the fat content
in beef and the possible negative effect on consumers cholesterol
levels. Although there is no conclusive evidence that moderate
consumption of beef, as part of a healthy diet, has negative health
effects (Wyness et al., 2011), recent research and consumer
concerns about the association between red meat and cancer
(Santarelli et al., 2008) have negatively inuenced consumer
perception of the healthiness of beef (Van Wezemael et al., 2010;
Perez-Cueto and Verbeke, 2012). However, beef contains important
nutrients such as high quality protein, iron, zinc, vitamin D, B3,
B12, selenium and long-chain omega-3 fatty acids, all potentially

Corresponding author. Tel.: +82 2 3790 4621.

E-mail addresses: Lynn.VanWezemael@UGent.be (L. Van Wezemael), vincenzina.caputo@gmail.com (V. Caputo), rnayga@uark.edu (R.M. Nayga Jr.), G.Chryssochoidis@uea.ac.uk (G. Chryssochoidis), Wim.Verbeke@UGent.be (W. Verbeke).
1
Project work completed under the auspices of the Agricultural University of
Athens.
0306-9192/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodpol.2013.11.006

contributing to good health throughout life (McAfee et al., 2010).

The present study investigates the possible appeal of these nutritional benets when signalled to consumers by means of nutrition
and health claims.
One of the ways to inform and enable consumers to evaluate
the nutritional value and the healthiness of beef is through nutrition and health claims. Since nutrition and health considerations
can play a role in food choices (da Fonseca and Salay, 2008; Nayga,
2008; Rimal, 2005), nutrition labelling might inuence consumer
decision making processes at the point of purchase. Furthermore,
it might positively inuence consumer perceptions of the healthiness of beef products (Ares et al., 2009). Regulations for nutrition
labelling are diverse across the world. Within the European Union
(EU), EC Regulation 1924/2006 stipulates the rules for the use of
nutrition and health claims. This regulation aims to help consumers make healthy and informed food choices, stimulate and protect
innovation in the food market, and facilitate the circulation of
foods bearing claims across EU member states. In comparison to
the EU, claims in the United States (US) are very similar in nature,
but the regulatory system is more liberal, and procedures for the
use and approval of claims differ substantially (Verhagen et al.,
2010).

168

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Beef is one of the most consumed meat types in the European

and American cuisine. The EU and the US have dominant global
positions in terms of beef production and consumption, with recent numbers indicating consumers eating 17.2 and 27.7 kg per
head per year on average in the EU-27 and the US, respectively
(FAOSTAT, 2011; USDA, 2012). Over the last decades, the beef
industry and butchers have diversied their market offerings from
the traditional beef steak and roast to an increasing number of
beef-based processed products, including ready-meals. However,
up until now, the beef industry has not fully taken advantage of
the nutrition labelling regulations. Only a limited number of meat
products available in the market use the extant nutrition labelling
regulation for product differentiation (Steinhauserova et al., 2011;
Barreiro-Hurl et al., 2009). In contrast, nutrition labelling has been
an inspiring source for new product development in other food
sectors such as the dairy and snacks industries (Vyth et al.,
2010). Nutrition labelling could be a source of market opportunities for the beef sector if consumers are found to value nutrition
and health claims on beef. However, up till now only one study
has investigated the application of approved nutrition and health
claims in the European red meat sector. Barreiro-Hurl et al.
(2009) investigated consumer preferences for nutrition and health
claims regarding the fat content in pork Frankfurter sausages. Their
results indicated that consumers value approved nutrition and
health claims, and therefore conrm market opportunities that
could arise with the use of the existing nutrition labelling regulation, especially with respect to health claims. Our study complements these ndings by focussing on beef instead of pork, and by
investigating unprocessed steak compared to further-processed
sausages. Our study, however, differs from the Barreiro-Hurl
et al. (2009) study in a number of ways. First, the scope of our
study is broader with the focus on claims with respect to a variety
of nutrients. Second, our multi-country study provides an overview
of consumer preferences in a European context instead of focussing
on only one country.
The aim of this study is to investigate consumer preferences for
nutrition labelling on beef steaks. In particular, we used the choice
experiment method to achieve two specic objectives. First, we
wish to investigate consumer preferences and willingness-to-pay
(WTP) for nutrition and health claims for different nutrients on
beef steak. Second, we examine if there are differences in the results across European countries (i.e. Belgium, France, the Netherlands, and the United Kingdom). In particular, the key question
investigated in this study is whether nutrition and health claims
are appealing to European consumers, and if so, whether they are
equally or differently appealing for consumers in different European countries. No other known published study has investigated
this question in the past. Furthermore, since very few multi-country studies exist in choice experiment literature (Aristides et al.,
2004; Blaauw et al., 2010; Lusk et al., 2003), this study also contributes to the growing body of literature on multi-country choice
experiments.

2. Background
For the food industry, the successful application of nutrition and
health claims on food products remains a complex issue. For instance, some of the nutrient or ingredient claims may be perceived
positively or negatively by consumers irrespective of their actual
need or function in a healthy balanced diet (Biesalski et al.,
2011). Hence, not all combinations of carrier product, nutrient or
functional ingredient, and claims are equally attractive to consumers (Verbeke et al., 2009). Not only claim-related factors (such as
named ingredient, its function, and type of benet it provides)
but also product-related factors (such as the image of the carrier

product, or previous experience with claim labelling) and personal

characteristics inuence how consumers perceive different claims
(Lhteenmki, 2012). These differences in consumer reactions to
claims complicate the strategic choice for the large-scale application of a specic claim on a specic food product.
Framing (i.e. the context within which information is presented) is known to play an important role in consumer perceptions of claims on food. Levin (1998) showed that positive
framing of a meat product attribute (e.g. 75% lean meat) resulted
in more positive product evaluations than its equivalent negative
framed description (only 25% fat). According to Prospect Theory
(Kahneman and Tversky, 1979) decisions are inuenced more by
expected losses than expected gains, implying that consumers
have a preference for avoiding a possible loss compared to achieving a possible gain. This is in line with health framing literature,
conrming that negative information tends to attract more attention than positive information (Baglione et al., 2012; Hoefkens
et al., 2011; van Kleef et al., 2005) and also has a disproportionate
stronger impact on consumer decisions (Verbeke and Ward, 2001).
The higher value attached to the avoidance of possible losses implies greater preferences for nutrition and health claims when outcomes are expressed as possible losses than as possible gains, as
shown by Levin et al. (1998) and conrmed by a meta-analysis
by Pion and Gambara (2005).
Consumer preferences differ between nutrition claims and
health claims. Nutrition claims state that a food has particular benecial nutritional properties due to the caloric value it provides (or
does not provide) and/or the nutrients it contains (or does not contain). Health claims are statements about a relationship between
food and health, which can be related to functions of the body or
disease risk reduction (EC Regulation 1924/2006). Although both
nutrition and health claims are based on nutritional factors and
share associative knowledge networks that consumers use when
processing information (Lhteenmki, 2012; Lawson, 2002), several research ndings have illustrated that health claims outperform nutrition claims. Barreiro-Hurl et al. (2009) showed that
health claims were valued signicantly higher than claims about
nutritional attributes in pork Frankfurter sausages. Also in enriched fruit juice, spread and cereals, consumers valued health
claims higher than nutrition claims (Verbeke et al., 2009). Furthermore, health claims are accepted more easily on products with a
healthy image (Bech-Larsen and Grunert, 2003) and on products
with previous health claim labelling, as was illustrated by
Lhteenmki et al. (2010) in a study investigating claims on yoghurt and bread versus raw pork chops. However, claiming health
in food products might also induce a negative expectation barrier,
lower hedonic value to consumers (Verbeke, 2006; Lhteenmki
et al., 2010), and trigger scepticism owing to perceived associations
with marketing scams by the food industry (Verbeke, 2010). Furthermore, studies have revealed that the simultaneous presence
of more than one piece of nutrition-related information on a product label can lead to lower total utility levels (Barreiro-Hurl et al.,
2008, 2010). A more comprehensive overview of the literature on
nutrition labelling can be found in Hieke and Taylor (2012).
Several studies have investigated consumer preferences for
nutrition and health claims in different countries. In a study comparing physiological function claims and disease-risk reduction
claims, Bech-Larsen and Grunert (2003) found a common pattern
between claim perceptions in Denmark, Finland and the US. However, in comparison to Denmark and Finland, the US consumers
perceived the disease-risk reduction claim as more benecial to
health, which was possibly related to the greater familiarity with
this claim type among US consumers. Saba et al. (2010) found geographical differences in consumer perceptions of the healthiness of
cereal-based products with benecial compounds between
Finland, Germany, Italy and the UK. In a large-scale study in Italy,

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Germany, the UK, and the US, van Trijp and van der Lans (2007)
found signicant differences in consumer perceptions of nutrition
and health claims. Consumers from different countries differed in
their receptiveness due to variations in perceptions of newness
and difculty to understand the benets expressed in the claims.
Despite years of globalisation, strong national eating cultures and
habits persist and result into heterogeneous consumer preferences.
Besides differences in eating cultures and food-related beliefs, also
other country-specic factors might contribute to preference
inequalities, such as previous market exposure, familiarity or attitudes towards products bearing claims (Lhteenmki et al., 2010;
Grunert et al., 2009).
The key objective of this study is to assess consumer preferences and willingness-to-pay (WTP) for nutrition and health claims
for different nutrients on beef steak across Belgium, France, the
Netherlands, and the UK. In line with the aim of the EC Regulation
to facilitate the circulation of foods bearing claims across EU member states, it is important to know to what extent nutrition and
health claims appeal to consumers in different countries. Free circulation of foods might not simply imply optimal circulation of
foods bearing claims since consumer preferences can act as a kind
of natural barrier in the circulation of a food product with a specic
claim across different European countries. Food companies should
therefore investigate whether consumer preferences for a product
bearing a specic claim are uniform across EU member states. This
study is the rst to provide insights into consumer preferences
regarding the implementation of EC Regulation 1924/2006 in the
European beef sector. Given the more liberal regulation in the
US, results will also be useful for the US (and perhaps also the Latin
American and Canadian) beef industry envisaging beef exports to
the EU.
3. Materials and methods
This study uses choice experiments to investigate consumer
preferences for nutrition and health claims in lean beef steaks. This
methodological section introduces the chosen product, attributes
and levels. The choice experimental design and data collection procedures as well as the econometric approach to choice analysis are
then explained.
3.1. Selection of choice attributes and levels
Beef steak is universally offered in supermarkets and butcher
shops throughout Europe, and European consumers are familiar
with this high quality meat cut (cf. Loureiro and Umberger,
2007). Steak, as a non-processed beef product, is perceived as
healthier compared to other, more processed beef products (Van
Wezemael et al., 2010). Furthermore, because of the focus on
health, lean beef steak (with less than 10% fat) was selected for this
study. Nutritional values of prepared (cooked, grilled or baked) instead of raw products were used in this study to better represent
actual nutrient intake.
Four criteria were used to select relevant nutrients and related
nutrition and health claims: (1) the prevalence of nutrients in lean
beef steak, (2) the nutritional value of lean beef steak in relation to
thresholds according to labelling regulations, (3) the availability of
claims for lean beef steak, and (4) consumer familiarity with nutrients. The nutritional value of prepared lean beef was taken from
the Dutch food composition databank NEVO online version 2011/
3.0 (RIVM, 2011), containing nutritional values for a wide diversity
of food products in the Netherlands. The high quality of this food
composition databank is assured by the quality framework developed within the EuroFIR project (Westenbrink et al., 2009). The
nutritional value of prepared lean beef was used as input for the

169

tool Indicator Voedingsclaims - Productschappen Vee, Vlees en Eieren

[Indicator Food Claims from the Dutch Commodity Board Cattle,
Meat, and Eggs] (IV-PVE), a tool indicating which nutrition claims
are legally allowed for use in specic food products within the EU.
The IV-PVE indicated that prepared lean beef is a good source of
iron, vitamin B6, potassium, and selenium; rich in protein, niacin
(vitamin B3), vitamin B12, zinc, and phosphorus; and poor in saturated fat, and sodium and salt. From these possible claims, those on
nutrients assumed to be most familiar to consumers were selected
for this study: iron, protein and saturated fat (i.e. a qualifying micro-nutrient, a qualifying macro-nutrient, and a disqualifying
nutrient). Claims were presented to consumers by their legally approved formulations: source of iron, rich in protein, and poor
in saturated fat. No information about the actual amounts of the
respective nutrients was provided to the participants. Our assumptions and the selection of claims were based on the results from
qualitative focus group research suggesting that European consumers acknowledge the nutritional value of beef especially
related to proteins and iron, and that they are concerned about
the fat content of beef (Van Wezemael et al., 2010). Table 1 shows
the selected attributes and corresponding levels used in the choice
experiments. Four price levels are included, reecting the range of
current market prices of (lean) beef steak in the participating
countries: EUR/kg 12, 15, 18, 21 (or UK equivalents 10, 13, 16,
19 GBP/kg).
Health claims are only authorised for use in the EU after a scientic assessment by the European Food Safety Authority (EFSA).
The use of approved statements ensures the practical relevance
of the research ndings for food policy, marketing purposes and
commercial practices. With respect to iron, protein and saturated
fat health claims, the following statements reecting scientic
evidence to date were selected based on consumers expected
positive health effects of beef consumption (Van Wezemael et al.,
2010): Iron contributes to the normal cognitive function (EFSA,
2009); Protein contributes to the growth or maintenance of muscle
mass (EFSA, 2010); and Consumption of saturated fat increases
blood cholesterol concentrations. Consumption of foods with reduced
amounts of saturated fat may help to maintain normal blood
cholesterol concentrations (EFSA, 2011).
Since the health claims state a health benet by explicitly linking it to the benecial nutrient in the food product, the presence
(or absence) of this nutrient is implied in the health claim. Therefore, to investigate the difference between consumer preferences
for nutrition claims compared to health claims, two series of choice
experiments were conducted. The rst only included nutrition
claims (further referred to as the N experiment) and the second
included both nutrition and health claims (further referred to as
the NH experiment). In the NH experiment, the nutrition claim
was explicitly presented together with the health claim, to ensure
that the results reect differences related to the health claim on
top of the nutrition claim. By presenting a health claim only, consumers may react only to the benecial nutrient that is implied in
the health claim. By presenting the health claim together with the
nutrition claim, it is assured that differences in consumer preferences refer to the claimed health benet and not merely to the
nutrient mentioned in the health claim.
3.2. The experimental design and choice tasks
The SAS-based software JMP (R) was used to produce the experimental choice design for this study. A Bayesian sequential design
was developed, as recommended by the current state of practice
(Sndor and Wedel, 2001; Scarpa et al., 2007b). In an initial phase,
a D-optimal choice design was created to collect preliminary data
among a limited number of beef consumers not selected for the nal study. This pilot study was performed during September 2011

170

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Table 1
Selected attributes and corresponding levels.
Attributes

Levels

Claim
Nutrition claim
Health claim

No claim

Price

12 /kg

Iron
source of iron
Iron contributes to the
normal cognitive
function
15 /kg

Protein
rich in protein
Protein contributes to the
growth or maintenance of muscle
mass.
18 /kg

and provided the prior parameters necessary to generate the nal

Bayesian optimal choice design in JMP. Choice sets consisting of
two alternative proles and one opt-out alternative were created,
as recommended in literature (Hensher, 2010). Each participant
answered six choice scenarios in total, so as to limit their cognitive
burden.
The choice experiment was initiated with an introductory text
informing participants about the product attributes that were included in the study. In order to lower the hypothetical bias in
the choice experiment, a cheap talk script was set up (see Appendix A). Cheap talk was rst introduced by Cummings and Taylor
(1999) and refers to the process of explaining hypothetical bias
to participants before the start of the choice experiment (Lusk,
2002). Previous research has shown that cheap talk effectively lowers hypothetical bias for consumers (Carlsson et al., 2005; Cummings and Taylor, 1999; Lusk, 2002; Silva et al., 2011; Tonsor
and Shupp, 2011). The cheap talk script used in this study (see
Appendix A) was based on the cheap talk scripts from Carlsson
et al. (2005) and Lusk (2002).
Some research has suggested that WTP might be inuenced by
the presentation of attributes in discrete choice experiments
(Umberger and Mueller, 2010). Typically, attributes are presented
to participants in a verbal context. Some studies combine verbal
with visual presentations; for instance Mennecke et al. (2007)
illustrated the type of beef steak with pictures but used verbal presentations for all other attributes. However, consumer choices are
often inuenced by product characteristics or cues that are used
less consciously, such as marbling in the case of fresh meat
(Umberger and Mueller, 2010). In such cases verbal presentation
of attributes might be less appropriate. Umberger and Mueller
(2010) suggested that the importance of intrinsic values of beef
might be underestimated if not presented visually. Since in our
study only extrinsic attributes are included, all information about
the attributes in this choice experiment was presented verbally.
However, the product itself was presented to consumers both verbally (steak) and visually. The visual presentation was a prepared
steak to avoid consumers considering raw beef steak only and to
avoid possible bias owing to perceived colour differences between
the stimuli and the steaks habitually bought by the participants.
The steak was introduced to participants as an average, normally expected, lean beef steak that can be bought in the supermarket. Although other attributes might also be important in
consumer preferences for beef, participants were asked to choose
their preferred product with the assumption that all other attribute
levels were the same for each product. Furthermore, consumers
were informed that the stated effect of the nutrients on health
had been scientically proven and that the use of the health claim
was authorized by EFSA.
3.3. Data collection
Data were collected in December 2011 in four European
countries: Belgium (BE), France (FR), The Netherlands (NL), and
the United Kingdom (UK). In each country, 600 beef consumers
(i.e., people consuming beef at least once a month) participated
in the study. Samples were representative for each of the national

Saturated fat
poor in saturated fat
Consumption of saturated fat increases blood cholesterol concentration.
Consumption of foods with reduced amounts of saturated fat may help to
maintain normal blood cholesterol concentrations.
21 /kg

populations in terms of gender and age. Participants were recruited

by a subcontracted professional market research agency (Survey
Sampling International located in Rotterdam, The Netherlands)
with a good reputation in the market research business and abiding by the ICC/ESOMAR International Code on Market and Social
Research regarding ethics in social sciences research (ICC/ESOMAR,
2008). Data were collected among a random geographic sample
from a readily available database of consumers in the different
countries willing to respond to surveys, provided by the market
agency. Half of each national sample was randomly assigned to
take part in the N experiment, while the other half of the national
sample participated in the NH experiment. In total, 2400 European
beef consumers participated in this study.
Questionnaires were distributed online to the participants in
each of the four countries. Participants were asked to indicate their
age and gender, household composition (living alone or together
with others, and whether children were present in the household),
highest obtained educational degree (primary, secondary or higher
education), and occupation (working full-time, part-time, unemployed, student, or retired). Recent beef steak consumption was reported by the number of beef steaks consumed during the last
14 days preceding the survey.
3.4. Estimation procedures: econometric models and willingness-topay
This study used choice experiments to allow participants to
evaluate different attributes simultaneously (Shen, 2005). Study
participants made a discrete choice from a set of presented alternatives which contained a number of attributes with different levels,
combined within choice sets. In our application, participants were
asked to select between three options in each choice task: two lean
beef steak proles and one no-buy option (the opt-out alternative). The opt-out alternative was included since it makes the
choice tasks more realistic. According to random utility theory,
we assume that a given alternative was selected if the perceived
utility provided by such alternative was the highest among the different choices. Thus, the nth consumers utility of choosing alternative j in choice occasion t can be represented as

U njt b0n xnjt enjt

where xnjt is a vector of observed variables relating to alternative

j and individual n; b0n is a vector of structural taste parameters;
and enj is the stochastic (unobserved and treated as random)
element.
Depending on the specications of the density of unobserved
factors and assumptions on consumer preferences, different models
can be estimated. In this study different logit specications were
used. First, a multinomial logit (MNL) model was estimated as basic
model. The MNL assumes preference homogeneity in the sample,
implying that all coefcients of the utility expression in Eq. (1) are
the same across individuals. However, preference heterogeneity
may arise in choice models because of differences in the taste
intensities (e.g. utility parameters). If heterogeneity is expected,
then MNL results will be biased, and hence employing discrete
choice models capturing random taste variations is appropriate.

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Several studies on food labelling concluded that heterogeneity is an

issue to take into account in explaining consumer preference for
labelled food products (Gracia et al., 2009; Van Loo et al., 2011). Recent applications addressed the issue of unobserved taste heterogeneity by estimating the Random Parameter Logit (RPL) (Train, 1998)
model and its variants such as the error component (EC) logit model.
In addition to the MNL model, in this application, we estimated
an EC model with panel structure proposed by Scarpa et al. (2005)
and Scarpa et al. (2007a) to capture potential heterogeneity in
tastes in each country and correlations across utilities, which
might exist since a no-buy option (the opt-out alternative) was included in the N and NH choice experiments. We used the EC model
rather than the RPL model with only random taste variation because in using this model, not only did we account for heterogeneous consumer preferences by allowing the coefcients of the
different claims to vary randomly over individuals and to deviate
from the population mean, but we also accounted for the additional variance of utility of experimentally designed alternatives
different from the opt-out alternative. In fact, the presence of the
opt-out alternative in our experimental design might cause systematic effects associated with both the opt-out and correlated
random effects across utilities between alternatives of the choice
set designs (Scarpa et al., 2007a). A Monte Carlo study comparing
the performance of different random utility models accounting
for this effect showed that the EC model was more exible and
more robust to possible misspecication than other models
(Scarpa et al., 2005). Previous studies of choice experiment applications in environmental economics (Scarpa et al., 2007a; Hess and
Rose, 2008) and food choice (Caputo et al., 2013; Scarpa et al.,
2008) have found error components of this type to consistently
improve model t in similar choice contexts.
The utility of a consumer n choosing alternative j at time t is
specied as

U njt b0n xnjt gnj enjt

where xnjt is a vector of the observed variables of alternative j, and b0

is a vector of estimable coefcients. The term gnj represents the error
components that induce heteroskedasticity and correlation over
alternatives in the unobserved portion of utility; and enjt is the
Gumbel distributed error term. For further computational details,
see Train (2003), Scarpa et al. (2005), and Scarpa et al. (2007a,b).
As in Revelt and Train (1998), in the EC, price coefcients are
assumed to be invariant across individuals, whereas the coefcients
of the claims were assumed as random parameters with a normal
distribution. This assumption assured that the WTP estimated for
the attributes and the levels of the attributes was normally
distributed.
In this study, the WTP and its corresponding condence
intervals for each claim across countries were obtained using the
Krinsky and Robb (1986) bootstrapping parametric method. By
calculating the 95% condence intervals and looking at the overlapping condence intervals, this approach allowed us to also test if
estimates of the claims are signicantly different between
countries and between the two choice experiment surveys (i.e. N
and NH). The 95% condence intervals for the claims were calculated by taking 1000 draws from a multivariate normal distribution
with means given by the estimated coefcients and covariance
given by the estimated covariance matrix of the coefcients. The
1000 draws for each coefcient were then used to calculate a
WTP at each draw (for each row) and the corresponding standard
deviations across the draws. Finally, the lower and the upper limit
of the condence interval for each nutritional claim are given by the
26th and 975th sorted estimates of WTP. Condence intervals
derived in this way are usually referred to as percentile intervals
(Hole, 2007). This approach is valuable because it not only

171

quanties WTP for different nutrition and health claims in different

countries, but it also allows investigating the added value of the
addition of a health claim on top of a nutrition claim. The method
of overlapping condence intervals was used to determine if
estimates are signicantly different between countries and between choice experiment surveys.
4. Results
4.1. Characteristics of the sample
The characteristics of the participants in the four national
samples are shown in Table 2. The distributions over household
composition, educational degree, and occupational activities show
that the sample covered a wide range of individuals. The reference
population census values between brackets show that the sample
is fairly representative for the national populations in terms of
gender and age.
4.2. European consumer preferences for nutrition and health claims
Appendix B and Table 3 report the estimated parameters for the
MNL and the EC2 models respectively segmented by country across
both the N and NH experiments.
The EC model allowing for taste heterogeneity and correlation
across utilities performed better than the MNL models for all
countries in both experiments, as indicated by the higher simulated log-likelihood and the respective lower values for BIC and
AIC. Furthermore, the statistical signicance of the estimated standard deviation parameters for the nutrition claims in each country
indicated heterogeneity in consumer preferences across countries,
undermining the assumption of homogeneous preferences of the
MNL model. Hence, we will only discuss the estimates of the EC
models below.
In all countries, the negative coefcients of the opt-out alternative (no buy) indicated that with constant prices, consumers preferred having one of the two presented beef steaks rather than
having none at all. This indicates that the selected attributes in this
study were relevant for consumers in making their choice between
beef steaks across both the N and NH experiments in all countries.
Moreover, the hypothesis of correlation across utility is corroborated since the error component for the alternative specic constants is statistically signicant. Also, as expected, the negative
price coefcients conrmed consumer preference for lower over
higher prices in all countries.
Results of the N and NH experiments suggested an increasing
probability of consumers choosing lean steak alternatives possessing claims on saturated fat in both experiments. However, the
claim on saturated fat was the least popular nutrition claim for
consumers in the UK. In the Netherlands and Belgium, the lowest
levels of utilities were attached to the nutrition claim on protein,
while in France the nutrition claim on iron was the least attractive
to consumers. In the UK, France and Belgium, the lowest levels of
utilities were attached to the nutrition and health claim on iron.
Only in the Netherlands, the nutrition and health claim on protein
was less attractive to consumers than the other nutrition and
health claims included in the study.
4.4. Willingness-to-pay for nutrition and health claims on beef
Because of the overall better performance of the EC model, the
WTP values were based on this model. To quantify the value of
2
For the estimation, 500 Halton quasi-draws rather than pseudo-random draws
are used since the former provides a more efcient simulation for this model (Train,
2003).

172

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Table 2
Socio-demographic characteristics of the participants in % of the national samples (unless stated otherwise).
Netherlands n = 600

Belgium n = 600

France n = 600

UK n = 600

Male

49.0 [49.0]

51.7 [51.6]

47.8 [47.8]

52.2 [47.86]

Age categories
1829 years
3049 years
5064 years
6575 + years

19.7
40.3
22.5
17.5

18.8
37.7
22.2
21.3

20.7
37.5
20.7
21.2

18.5
37.8
22.8
20.8

Household composition
Living alone
No children
Children 6 5 years

20.8
73.8
7.2

22.0
68.7
10.7

17.0
61.3
13.2

17.3
73.3
14.2

Education
Low (Primary education)
Medium (Secondary education)
High (Higher education)

28.0
45.5
26.5

9.3
54.2
36.5

27.2
35.8
37.0

19.3
27.2
53.5

Occupation
Full-time
Retired

28.3
20.7

34.5
29.2

38.3
28.7

39.2
24.7

Ate at least 1 beef steak in the last 14 days

Age (mean) (years)

60.2
45.9

88.7
46.9

90.2
46.4

71.0
46.5

[19.7]
[37.4]
[22.6]
[20.5]

[18.7]
[37.5]
[22.1]
[21.6]

[20.7]
[37.3]
[20.7]
[21.3]

[19.5]
[37.4]
[22.6]
[20.5]

Note: numbers between brackets are reference values for national representativeness of the sample (Source: Survey Sampling International).

nutrition and health claims regarding iron, saturated fat, and protein
for consumers in the different countries, we estimated condence
intervals for the WTP measures. The estimates in Table 4 indicated
that WTP differed between countries, nutrients (iron, fat, or protein),
and type of claim (nutrition or health claim). The WTP values are
rather high compared to the actual price of a steak. Despite the use
of a cheap talk script, it is possible that the hypothetical context of
choice caused participants to implicitly overestimate the marginal
WTP values. Experimental auctions involving real money and goods
can be used to validate the WTP estimates found in this study.
Differences in WTP values between countries existed in both
the N and NH experiments. All but British consumers were willing
to pay for a nutrition claim on saturated fat, while the British consumers are willing to pay signicantly more for a nutrition claim
on protein compared to Dutch and Belgian consumers. British
WTP for a nutrition and health claim on saturated fat is signicantly lower than in the other countries. French consumers were
willing to pay signicantly more for a nutrition and health claim
on protein than in the UK or The Netherlands.
Within countries, clear differences in WTP values for claims on
different nutrients exist. Dutch consumers were willing to pay signicantly more for a nutrition claim on iron or fat compared to a
nutrition claim on protein. In Belgium and France, WTP estimates
were not signicantly different between nutrients referred to in
nutrition claims. British consumers were willing to pay a similar
premium for nutrition claims on protein and iron. WTP values for
nutrition and health claims on saturated fat were signicantly
higher than for other nutrition and health claims in Belgium and
The Netherlands. French consumers are willing to pay on average
6.22 and 2.10 EUR/kg more for respectively a NH claim on saturated fat and protein compared to a nutrition and health claim
on iron. British WTP estimates for NH claims were not signicantly
different from each other.
Most important, the WTP values displayed in Table 4 allowed us
to quantify the value of providing a health claim on top of a nutrition claim. Comparison of the WTP values per country and per
nutrient over the two experiments showed that consumers in all
countries valued the additional health claim on iron similar to
the nutrition claim alone. This was also the case for claims on protein in the Netherlands, Belgium and the UK, while the addition of
a health claim on protein increased WTP among French consumers.
For saturated fat however, the addition of a health claim on top of

the nutrition claim signicantly increased consumers WTP in all

countries.
In summary, it can be concluded that the EC model performed
better than the MNL model, and that in Belgium, the Netherlands
and France, nutrition and health claims on saturated fat yielded
higher utilities than claims on protein and/or iron, while the opposite was found for nutrition claims among consumers in the UK.

5. Discussion
This study aimed to investigate consumer preferences and WTP
for nutrition and health claims on beef steak. Our results indicated
that in general (negatively framed) claims on saturated fat yielded
higher utilities and WTP than (positively framed) claims on protein
and/or iron, in contrast to evidence from the health framing literature suggesting that positive framing results in more positive
product evaluations. This nding can possibly be explained by
the fact that protein and iron are considered and known as benecial nutrients in beef by consumers, in contrast to fat, which is a
major issue of consumer concern also in the context of beef consumption (Van Wezemael et al., 2010). Consumer concerns about
fat levels in food have been widely documented in scientic literature (Kemp et al., 2007). Fat is among the most highly diagnostic
nutrition information that is used by consumers. The majority of
people are aware of the relationship between high levels of saturated fat consumption and cholesterol and heart disease (Garretson and Burton, 2000). Hence, the popularity of the health claim
on saturated fat in our study is consistent with the popularity of reduced cardiovascular disease risk claims in the study of van Kleef
et al. (2005). By combining nutrition and health claims in our NH
experiment, the diet-disease relationship was made clearly visible
to the participants in our study.
A notable exception is the UK where the nutrient claim on saturated fat is not the most preferred one. It is possible that this is
related to the high prevalence of obesity (among the highest across
Europe (OECD, 2010)), and thus estimated stated choice utility
scores may link to actual eating habits and preferences for foods
with a higher fat content. However, when a health claim was added
and thus the link with disease was made more explicit consumers in the UK also preferred the nutrition and health claim on saturated fat over the claims on protein and iron.

173

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Table 3
Parameter estimates of N and NH experiments from the error component (EC) model segmented by country (NL = Netherlands; BE = Belgium; FR = France; UK = United Kingdom).
Variables

NO BUY
PRICE
IRON
FAT
PROTEIN

N experiment

NH experiment

NL

BE

FR

UK

NL

BE

FR

UK

5.25 (14.35)a
0.26 (20.83)
1.45 (8.37)
1.53 (9.59)
0.71 (4.72)

4.65 (15.15)
0.25 (19.31)
1.06 (6.67)
1.38 (7.90)
0.84 (5.54)

4.47 (15.46)
0.24 (19.32)
1.00 (5.96)
1.64 (10.20)
1.21 (7.15)

4.95 (17.25)
0.28 (19.43)
1.41 (8.30)
0.33 (1.85)
1.62 (9.47)

4.73 (13.01)
0.27 (15.61)
1.51 (10.02)
2.58 (11.94)
1.12 (7.54)

3.70 (11.26)
0.18 (13.05)
1.05 (7.07)
2.07 (11.36)
1.10 (7.79)

3.42 (10.90)
0.17 (13.07)
0.94 (6.46)
1.99 (11.58)
1.65 (10.70)

4.45 (14.73)
0.25 (17.05)
1.08 (9.22)
1.17 (7.88)
1.11 (8.55)

1.22 (4.58)
0.94 (3.02)
1.58 (3.89)

0.95 (4.05)
1.38 (4.54)
1.48 (4.98)

0.98 (3.98)
1.57 (0.26)
0.79 (2.73)

1.20 (5.69)
1.62 (6.25)
0.72 (2.90)

1.22 (5.22)
1.57 (0.25)
1.18 (5.01)

0.19 (0.76)
1.24 (5.27)
0.48 (0.39)

2.45 (12.66)

2.59 (12.80)

3.56 (12.43)

2.73 (11.11)

2.44 (11.59)

2.54 (13.08)

1800
1529.43
3082.9
3148.8
0.2240

1800
1486.14
2996.3
3062.2
0.2460

1800
1404.53
2833.1
2899.0
0.2874

1800
1507.13
3038.3
3104.2
0.2353

1800
1493.30
3010.6
3076.6
0.2423

1800
1526.59
3077.2
3143.1
0.2254

Standard deviations of parameter distributions

IRON
0.84 (2.54)
1.09 (4.46)
FAT
0.64 (1.68)
1.14 (3.56)
PROTEIN
0.93 (0.42)
1.11 (4.05)
Error component
3.20 (11.95)
2.51 (12.35)
Summary statistics
N
1800
LL
1459.61
AICb
2943.2
BICb
3009.2
Adj-R2
0.2594
a
b

1800
1552.27
3128.5
3194.5
0.2124

Numbers in parenthesis are t-statistics.

AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion.

With respect to saturated fat, our results showed that consumers value the addition of a health claim to a nutrition claim. This
nding is consistent with the ndings from Verbeke et al. (2009)
and Barreiro-Hurl et al. (2009) who found that health claims outperformed nutrition claims in various food products in terms of
eliciting favourable reactions among consumers. The higher WTP
for products with a health claim might compensate the extra efforts and costs from the industry needed to scientically substantiate and obtain legal approval for the claim. However, for iron and
protein, consumer preferences based on the combined nutrition
and health claim were in most countries similar to those based
on the nutrition claim alone. This is consistent with the ndings
from van Kleef et al. (2005), suggesting that physiology-based
health benets such as cholesterol are preferred over psychology/
behavior-based benets such as cognitive functions. Another
possible explanation is that the benecial effects of familiar qualifying nutrients, like iron and protein in beef, are well known
among consumers and the mere mention of the presence of these
nutrients sufces to trigger a favorable response, in the same sense
as reported by Verbeke et al. (2009) for the case of omega-3 fatty
acids. Therefore we conclude that not all health claims consistently
add value over nutrition claims. This nding highlights the
relevance of our study, as consumer research is needed to identify
the most attractive claim for a given food product. Our results
rene and extend the ndings of past studies in agricultural
economics, food policy and related disciplines (e.g. McCluskey
et al., 2005; Beatty, 2007; van Kleef et al., 2005; van Trijp and
van der Lans, 2007; Lusk and Parker, 2009; Verhagen et al., 2010;
Lhteenmki, 2012) in that consumers do perceive and value
different claims dissimilarly.
Our results are also in line with the ndings from van Trijp and
van der Lans (2007) who found that consumer perceptions of nutrition and health claims on yoghurt differed substantially by country.
The possible inuence of country-specic factors such as market
exposure, eating cultures, familiarity or attitudes towards products
bearing claims (Grunert et al., 2009; Lhteenmki, 2012) emphasizes the usefulness of comparative multi-country research.
6. Conclusion
The results from this study have implications for the general
public, food policy makers and food companies interested in apply-

ing nutritional marketing. For the general public, a nutrition and/or

health claim that they would value could help them make better
informed decisions on whether to buy or consume the product or
not, and to make choices better in line with their actual preferences. Furthermore, opportunities offered by using nutrition and
health labels or claims can stimulate healthier product development (Vyth et al., 2010). This could in the particular case of red
meat help improve the partly unhealthy image of the sector and
its products, which is also recognised as a necessity for this sector
facing changing customer needs and wants (Grunert et al., 2011).
The insights and information from this study about consumer
reactions to nutrition and health claims on different nutrients in
beef are also relevant for food policy makers in their efforts to develop legislation, standards, common frameworks (e.g. the upcoming EU framework for the reduction of saturated fat intakes) and
information campaigns aiming at fostering healthier food choices
(Capacci et al., 2012).
The results of our study show that the existing regulations on
nutrition and health claims present a marketing opportunity for
the beef sector that is currently not being fully considered. An
increasing amount of claims is allowed by EU and US legislation,
and several nutrition and health claims apply to fresh lean beef.
Also, new product development could focus on beef products that
comply with the criteria for the use of nutrition and health claims,
since this study has shown that this can be a promising avenue.
However, for the beef industry the reduction of the level of saturated fatty acids is a challenge since fat is an important quality trait
in relation to beef tenderness, avour and juiciness (Scollan et al.,
2006) and consumers are unlikely to compromise on the taste of
their foods for the promise of future health benets (Verbeke,
2006). Furthermore, since not all claims are equally appealing to
consumers in different countries, food companies producing for
an international market may want to consider tailoring their nutritional labeling to country-specic needs to make their marketing
more effective.
This study has a number of limitations that can inspire
future research. Firstly, this study uses a hypothetical
choice-experimental approach to investigate consumer preferences. Hence, despite our use of an ex-ante calibration method
(i.e., the cheap talk script), the data are only hypothetical statements and not real market choices. Non-hypothetical experiments
would complement these ndings. Secondly, as substantial differ-

174

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Table 4
Estimated marginal WTP values (in EUR/kg) based on the EC model (NL = Netherlands; BE = Belgium; FR = France; UK = United Kingdom).
N experiment

a
1
2

NH experiment

NL

BE

FR

UK

NL

IRON
5.44a [6.6114.252]

BE

FR

UK

4.26 [5.522.99]

4.11 [5.542.78]

5.04 [6.243.94]

5.62 [6.784.50]

5.89 [7.764.20]

5.49 [7.613.78]

4.27 [5.173.36]

FAT
5.78 [7.784.49]

5.60 [7.074.15]

6.73 [8.195.44]

1.20 [2.480.05]

8.45 [9.936.67]

11.66 [14.299.44]

11.71 [14.229.34]

4.60 [5.823.43]

PROTEIN
2.71 [3.751.62]

3.42 [4.582.27]

4.96 [6.313.61]

5.81 [6.924.65]

4.22 [5.323.15]

6.20 [7.944.56]

9.70 [11.957.52]

4.39 [5.483.30]

Marginal willingness to pay estimates in EUR/kg.

Maximum: upper 95% condence interval level (975th sorted estimate of WTP).
Minimum: lower 95% condence interval level (26th sorted estimate of WTP).

ences between countries were observed, the results cannot be generalised to other countries. Also generalisations to beef or meat
products other than steak are beyond the possibilities of our study,
as consumers health perception of beef varies across beef products
(Van Wezemael et al., 2010). However, the theoretical and analytical framework could be applied to other commodities as well in
the future. Furthermore, this study selected only three nutrients,
while many claims related to other nutrients are also applicable
to beef. Future research might investigate whether claims with respect to other nutrients besides iron, protein or saturated fat could
be appealing to meat consumers. Thirdly, the research design did
not include a nutrition facts panel, although mandatory under EU
legislation. The nutrition facts panel is often presented at the back
of the packaging of fresh meat, while claims are expected to be presented front-of-pack. Therefore, it can be expected that claims are
considered independently of the nutrition facts panel by most consumers. However, if consumers are being confronted with the actual amount of the nutrients on which the claim is based, this
might inuence their claimed preferences. Since consumers may
prefer to check claims against actual values in the nutrition information panel (Chan et al., 2005), presenting the actual amount of
the involved nutrient can possibly affect the credibility of the claim
and consumer preferences. Fourthly, consumer preferences for
nutrition and health claims are known to be dependent on individual differences such as socio-demographic factors (Verbeke, 2005),
nutritional knowledge (Baglione et al., 2012), familiarity (Wills
et al., 2012) or experienced disease state (van Kleef et al., 2005).

These factors could be used in the future to better characterize

consumer preferences for nutrition and health claims on beef
within countries.
Acknowledgement
This work has been performed within the EU FP6 Integrated Project ProSafeBeef, Contract No. FOOD-CT-2006-36241. The nancing
of the work by the European Union is gratefully acknowledged.
Appendix A. Cheap talk script
From previous similar studies we know that people often respond in one way but act differently. In studies where people do
not actually have to pay money for a product when indicating a
particular preference, people state a higher willingness to pay
than what one actually is willing to pay for the good in the store.
A possible reason for this is that people do not really consider
how big the impact of this extra cost actually is on the available
family budget. It is easy to be generous when you do not really
have to pay for it. In the store, people might think in a different
way: the amount of money spent on this good cannot be spent to
other things. We ask you to respond to each of the following preference questions just exactly as you would if you were in a real
store and had to pay for your choice. Please keep this in mind
when answering the following questions.

Appendix B. Parameter estimates of N and NH experiments from the multinomial logit (MNL) model segmented by country

Variables N experiment

a
b

NH experiment

NL

BE

FR

UK

NL

BE

FR

UK

NO BUY
PRICE
IRON
FAT
PROTEIN

3.11 (14.85)a
0.20 (16.05)
1.14 (10.34)
1.13 (9.78)
0.66 (6.49)

2.85 (13.77)
0.19 (15.03)
0.84 (7.72)
1.02 (8.98)
0.77 (7.66)

2.84 (13.69)
0.19 (15.21)
0.81 (7.24)
1.23 (10.63)
1.05 (10.31)

3.14 (15.68)
0.22 (16.40)
10.07 (9.57)
0.28 (2.44)
1.30 (12.53)

2.09 (9.19)
0.16 (12.65)
1.12 (11.31)
1.57 (14.20)
0.64 (8.84)

1.90 (8.39)
0.13 (10.22)
0.87 (9.21)
1.51 (14.14)
0.95 (9.43)

1.77 (7.70)
0.12 (9.51)
0.80 (8.32)
1.50 (13.91)
1.40 (13.65)

2.47 (11.65)
0.18 (13.69)
0.94 (10.01)
1.04 (9.76)
1.03 (10.15)

Summary
N
LL
AICb
BICb
Adj-R2

statistics
1800
1740.36
3490.7
3518.2
0.0941

1800
1780.63
3571.3
3598.7
0.0794

1800
1743.87
3497.7
3525.2
0.0962

1800
1696.36
3402.7
3430.2
0.1082

1800
1733.10
3476.2
3503.7
0.1118

1800
1736.08
3482.2
3509.6
0.0919

1800
1693.89
3397.8
3425.3
0.1075

1800
1710.83
3431.7
3459.1
0.1050

Numbers in parenthesis are t-statistics.

AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion.

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

References
Ares, G., Gimnez, A., Gambaro, A., 2009. Consumer perceived healthiness and
willingness to try functional milk desserts. Inuence of ingredient, ingredient
name and health claim. Food Qual. Pref. 20 (1), 5056.
Aristides, M., Weston, A.R., FitzGerald, P., Le Reun, C., Maniadakis, N., 2004. Patient
reference and willingness-to-pay for Humalog Mix25 relative to Humulin 30/
70: a multicountry application of a discrete choice experiment. Value Health 7
(4), 442454.
Baglione, S.L., Tucci, L.A., Stanton, J.L., 2012. Self-reported nutritional knowledge
and the acceptance of health-related food benet claims. Brit. Food J. 114 (4),
453468.
Barreiro-Hurl, J., Gracia, A., de-Magistris, T., 2008. Using latent classes to determine
consumer heterogeneity in nutrition label valuation. Acta Agric. Scand C 5 (3
4), 178193.
Barreiro-Hurl, J., Gracia, A., de-Magistris, T., 2009. Market implications of new
regulations: impact of health and nutrition information on consumer choice.
Span J. Agric. Res. 7 (2), 257268.
Barreiro-Hurl, J., Gracia, A., de-Magistris, T., 2010. The effects of multiple health
and nutrition labels on consumer food choices. J. Agric. Econ. 61 (2), 426443.
Beatty, T.K.M., 2007. Recovering the shadow value of nutrients. Am. J. Agric. Econ.
89 (1), 5262.
Bech-Larsen, T., Grunert, K.G., 2003. The perceived healthiness of functional foods: a
conjoint study of Danish, Finnish and American consumers perception of
functional foods. Appetite 40 (1), 914.
Biesalski, H.K., Agget, P.J., Anton, R., Bernstein, P.S., Blumberg, J., Heaney, R.P., Henry,
J., Nolan, J.M., Richardson, D.P., van Ommen, B., Witkamp, R.F., Rijkers, G.T. and
Zllner, I., 2011. In: 26th Hohenheim Consensus Conference, September 11
2010. Scientic Substantiation of Health Claims: Evidence-based Nutrition.
Publications form USDA ARS/UNL Faculty, paper 836, 22 p.
Blaauw, D., Erasmus, E., Pagaiya, N., Tangcharoensathein, V., Mullei, K., Mudhune, S.,
Goodman, C., English, M., Lagarde, M., 2010. Policy interventions that attract
nurses to rural areas: a multicountry discrete choice experiment. B World
Health Organ 88 (5), 350356.
Capacci, S., Mazzocchi, M., Shankar, B., Brambila Macias, J., Verbeke, W., PerezCueto, F.J.A., Koliol-Kozakowska, A., Piorecka, B., Niedzwiedzka, B., DAddesa, D.,
Saba, A., Turrini, A., Aschemann-Witzel, J., Bech-Larsen, T., Strand, M., Wills, J.,
Traill, B., 2012. Policies to promote healthy eating in Europe: a structured
review of policies and their effectiveness. Nutr. Rev. 70 (3), 188200.
Caputo, V., Nayga Jr, M.R., Scarpa, R., 2013. Food miles or carbon emissions?
Exploring labelling preference for food transport footprint with a stated choice
study. Aust. J. Agric. Resour. Econ. 57, 118.
Carlsson, F., Frykblom, P., Lagerkvist, C.J., 2005. Using cheap talk as a test of validity
in choice experiments. Econ. Lett. 89 (2), 147152.
Chan, C., Patch, C., Williams, P., 2005. Australian consumers are sceptical about
but inuenced by claims about fat on food labels. Eur. J. Clin. Nutr. 59 (1),
148151.
Cummings, R.G., Taylor, L.O., 1999. Unbiased value estimates for environmental
goods: a cheap talk design for the contingent valuation method. Am. Econ. Rev.
89 (3), 649665.
da Fonseca, M.D.P., Salay, E., 2008. Beef, chicken and pork consumption and
consumer safety and nutritional concerns in the city of Campinas, Brazil. Food
Contrl. 19 (11), 10511058.
EFSA Panel on Dietetic Products, Nutrition and Allergies, 2009. Scientic Opinion
on the substantiation of health claims related to iron and formation of red
blood cells and haemoglobin (ID 249, ID 1589), oxygen transport (ID250, ID
254, ID 256), energy-yielding metabolism (ID 251, ID 1589), function of the
immune system (ID 252, ID 259), cognitive function (ID 253) and cell
division (ID 368) pursuant to Article 13(1) of Regulation (EC) No 1924/2006.
EFSA J. 7 (9), 1215.
EFSA Panel on Dietetic Products, Nutrition and Allergies, 2010. Scientic Opinion on
the substantiation of health claims related to protein and increase in satiety
leading to a reduction in energy intake (ID 414, 616, 730), contribution to the
maintenance or achievement of a normal body weight (ID 414, 616, 730),
maintenance of normal bone (ID 416) and growth or maintenance of muscle
mass (ID 415, 417, 593, 594, 595, 715) pursuant to Article 13(1) of Regulation
(EC) No. 1924/2006. EFSA J. 8 (10), 1811.
EFSA Panel on Dietetic Products, Nutrition and Allergies, 2011. Scientic Opinion on
the substantiation of health claims related to foods with reduced amounts of
saturated fatty acids (SFAs) and maintenance of normal blood LDL-cholesterol
concentrations (ID 620, 671, 4332) pursuant to Article 13(1) of Regulation (EC)
No. 1924/2006. EFSA J. 9 (4), 2062.
FAOSTAT, 2011. Databank from the statistic division of the Food and Agriculture
Organisation of the United Nations. <http://faostat.fao.org/>.
Gracia, A., Loureiro, M.L., Nayga Jr., R.M., 2009. Consumers valuation of nutritional
information: a choice experiment study. Food Qual. Pref. 20 (7), 463471.
Garretson, J.A., Burton, S., 2000. Effects of nutrition facts panel values, nutrition
claims, and health claims on consumer attitudes, perceptions of disease-related
risks, and trust. J. Publ. Pol. Market 19 (2), 213227.
Grunert, K.G., Lhteenmki, L., Boztug, Y., Martinsdottir, E., Ueland, ., Astrm, A.,
Lampila, P., 2009. Perception of health claims among Nordic consumers. J.
Consum. Policy 32 (3), 269287.
Grunert, K.G., Verbeke, W., Kgler, J.O., Saeed, F., Scholderer, J., 2011. Use of
consumer insight in the new product development process in the meat sector.
Meat Sci. 89, 251258.

175

Hensher, D.A., 2010. Hypotheticla bias, choice experiments and willingness to pay.
Trans. Res. B-Meth. 44 (6), 735752.
Hess, S., Rose, J.M., 2008. Should reference alternatives in pivot design SC surveys be
treated differently? Working Paper Institute of Transport and Logistics Studies,
ITLS-WP-08-09, 24 p.
Hieke, S., Taylor, C.R., 2012. A critical review of the literature on nutritional
labelling. J. Consum. Aff. 46 (1), 120156.
Hoefkens, C., Verbeke, W., Van Camp, J., 2011. European consumers perceived
importance of qualifying and disqualifying nutrients in food choices. Food Qual.
Pref. 22 (6), 550558.
Hole, 2007. A comparison of approaches to estimating condence intervals for
willingness to pay measures. Health Econ. 16, 827840.
ICC/ESOMAR, 2008. International Code on Market and Social Research. <http://
www.esomar.org/index.php/codes-guidelines.html>.
Kahneman, D., Tversky, A., 1979. Prospect theory: an analysis of decision under risk.
Econometrica 47 (2), 263292.
Kemp, E., Burton, S., Creyer, E.H., Suter, T.A., 2007. When do nutrient content and
nutrient content claims matter? Assessing consumer tradeoffs between
carbohydrates and fat. J. Consum. Aff. 41 (1), 4773.
Krinsky, I., Robb, A.L., 1986. On approximating the statistical properties of
elasticities. Rev. Econ. Stat. 64 (4), 715719.
Lhteenmki, L., Lampila, P., Grunert, K., Boztug, Y., Ueland, ., Astrm, A.,
Martinsdottir, E., 2010. Impact of health-related claims on the perception of
other product attributes. Food Policy 35 (3), 230239.
Lhteenmki, L., 2012. Claiming health in food product. Food Qual. Pref. 27 (2), 196
201.
Lawson, R., 2002. Consumer knowledge structures: background issues and
introduction. Psychol. Market 19 (6), 447455.
Levin, I.P., Schneider, S.L., Gaeth, G.J., 1998. All frames are not created equal: a
typology and critical analysis of framing effects. Organ Behav. Hum. Dec. 76 (2),
149188.
Loureiro, M.L., Umberger, W., 2007. A choice experiment model for beef: what US
consumer responses tell us about relative preferences for food safety, countryof-origin labeling and traceability. Food Policy 32 (4), 496514.
Lusk, J.L., 2002. Effects of cheap talk on consumer willingness-to-pay for golden rice.
Paper presented at the American Agricultural Economics Association Annual
Meeting, Long Beach, California, 38 p.
Lusk, J.L., Parker, N., 2009. Consumer preferences for amount and type of fat in
ground beef. J. Agric. Appl. Econ. 41 (1), 7590.
Lusk, J.L., Roosen, J., Fox, J.A., 2003. Demand for beef from cattle administered
growth hormones or fed genetically modied corn: a comparison of consumers
in France, Germany, the United Kingdom, and the United States. Am. J. Agric.
Econ. 85 (1), 1629.
McAfee, A.J., McSorley, E.M., Cuskelly, G.J., Moss, B.W., Wallace, J.M.W., Bonham,
M.P., Fearon, A.M., 2010. Red meat consumption: an overview of the risks and
benets. Meat Sci. 84 (1), 113.
McCluskey, J.J., Wahl, T.I., Li, Q., Wandschneider, P.R., 2005. US grass-fed beef:
marketing health benets. J. Food Distr. Res. 36 (3), 18.
Mennecke, B.E., Townsend, A.M., Hayes, D.J., Lonergan, S.M., 2007. A study of the
factors that inuence consumer attitudes toward beef products using the
conjoint market analysis tool. J. Anim. Sci. 85 (10), 26392659.
Nayga Jr., R.M., 2008. Nutrition, obesity, and health: policies and economic research
challenges. Eur. Rev. Agric. Econ. 35 (3), 281302.
OECD, 2010. Health at a Glance: Europe 2010. Multilingual Summaries, Paris.
Perez-Cueto, F.J.A., Verbeke, W., 2012. Consumer implications of the WCRFs
permanent update on colorectal cancer. Meat Sci. 90 (4), 977978.
Pion, A., Gambara, H., 2005. A meta-analytic review of framing effect: risky,
attribute and goal framing. Psicotherma 17 (2), 325331.
RIVM (Rijksinstituut voor Volksgezondheid en Milieu), 2011. Nederlands
Voedingsstoffenbestand [Food Composition Database The Netherlands] NEVOonline version 2011/3.0. <http://nevo-online.rivm.nl>.
Revelt, D., Train, K., 1998. Mixed logit with repeated choices: households choices of
appliance efciency level. Rev. Econ. Stat. 80 (4), 111.
Rimal, A., 2005. Meat labels: consumer attitude and meat consumption pattern. Int.
J. Consum. Stud. 29 (1), 4754.
Saba, A., Vassallo, M., Shepherd, R., Lampila, P., Arvola, A., Dean, M., Winkelmann,
M., Claupein, E., Lhteenmki, L., 2010. Country-wise differences in perception
of health-related messages in cereal-based food products. Food Qual. Pref. 21
(4), 385393.
Sndor, Z., Wedel, M., 2001. Designing conjoint choice experiments using managers
prior beliefs. J. Market. Res. 38 (4), 430444.
Santarelli, R.L., Pierre, F., Corpet, D.E., 2008. Processed meat and colorectal cancer: a
review of epidemiologic and experimental evidence. Nutr. Canc. 60 (2), 131
144.
Scarpa, R., Ferrini, S., Willis, K.G., 2005. Performance of error component models for
status-quo effects in choice experiments. In: Scarpa, R., Alberini, A. (Eds.),
Applications of Simulation Methods in Environmental and Resource Economics.
Springer Publisher, The Netherlands, Dordrecht, pp. 247273.
Scarpa, R., Thiene, M., Marangon, F., 2008. Using exible taste distributions to value
collective reputation for environmentally-friendly production methods. Can. J.
Agr. Econ. 56 (2), 145162.
Scarpa, R., Willis, K., Acutt, M., 2007a. Valuing externalities from water supply:
status-quo, choice complexity and individual random effects in panel
kernel logit analysis of choice experiments. J. Envir. Plann. Man. 50 (4),
449466.

176

L. Van Wezemael et al. / Food Policy 44 (2014) 167176

Scarpa, R., Campbell, D., Hutchinson, W.G., 2007b. Benet estimates for landscape
improvements: sequential Bayesian design and respondents rationality in a
choice experiment. Land Econ. 83 (4), 617634.
Scollan, N., Hocquette, J.F., Nuernberg, K., Dannenberger, D., Richardson, I., Moloney,
A., 2006. Innovations in beef production systems that enhance the nutritional
and health value of beef lipids and their relationship with meat quality. Meat
Sci. 74 (1), 1733.
Shen, J., 2005. A review of stated choice method. Discussion Papers in Economics
and Business, vol. 0527, 22 p.
Silva, A., Nayga, R.M., Campbell, B.L., Park, J.L., 2011. Revisiting cheap talk with
new evidence from a eld experiment. J. Agric. Resour. Econ. 36 (2), 280
291.
Steinhauserova, P., Rhurkova, I., Ruprich, J., 2011. Comparison of meat products
composition bearing/not-bearing claims marketed in the Czech Republic. Acta
Vet Brno. 80 (4), 379384.
Tonsor, G.T., Shupp, R.S., 2011. Cheap talk scripts and online choice experiments:
looking beyond the mean. Am. J. Agr. Econ. 93 (4), 10151031.
Train, K., 1998. Recreation demand models with taste differences over people. Land
Econ. 74 (2), 230239.
Train, K.E., 2003. Discrete Choice Methods with Simulation. Cambridge University
Press, Cambridge, UK.
Umberger, W.J., Mueller, S.C., 2010. Is presentation everything? Using visual
presentation of attributes in discrete choice experiments to measure the
relative importance of intrinsic and extrinsic beef attributes. Paper prepared for
presentation at the Agricultural and Applied Economics Association AAEA, CAES
& WAAEA Joint Annual Meeting, Denver, Colorado, 29 p.
USDA, 2012. Statistics and Information: US Cattle and Beef Industry 20022011.
<http://www.ers.usda.gov/topics/animal-products/cattle-beef/statisticsinformation.aspx>.
van Kleef, E., van Trijp, H.C.M., Luning, P., 2005. Functional Foods: health claim-food
product compatibility and the impact of health claim framing on consumer
evaluation. Appetite 44 (3), 299308.

Van Loo, E., Caputo, V., Nayga Jr., R.M., Meullenet, J.F., Ricke, S.C., 2011. Consumers
willingness to pay for organic chicken breast: evidence from choice experiment.
Food Qual. Pref. 22 (7), 603613.
Van Trijp, H.C.M., van der Lans, I.A., 2007. Consumer perceptions of nutrition and
health claims. Appetite 48 (3), 305324.
Van Wezemael, L., Verbeke, W., de Barcellos, M.D., Scholderer, J., Perez-Cueto, F.,
2010. Consumer perceptions of beef healthiness: results from a qualitative
study in four European countries. BMC Publ. Health 10, 342.
Verbeke, W., 2005. Consumer acceptance of functional foods: socio-demographic,
cognitive and attitudinal determinants. Food Qual. Pref. 16 (1), 4557.
Verbeke, W., 2006. Functional foods: consumer willingness to compromise on taste
for health? Food Qual. Pref. 17 (12), 126131.
Verbeke, W., 2010. Consumer reactions to foods with nutrition and health claims.
Agro-Food Ind. Hi-tech. 21 (6), 58.
Verbeke, W., Ward, R.W., 2001. A fresh meat almost ideal demand system
incorporating negative TV press and advertising impact. Agric. Econ. 25 (23),
359374.
Verbeke, W., Scholderer, J., Lhteenmki, L., 2009. Consumer appeal of nutrition and
health claims in three existing product concepts. Appetite 52 (3), 684692.
Verhagen, H., Vos, E., Francl, S., Heinonen, M., van Loveren, H., 2010. Status of
nutrition and health claims in Europe. Arch. Biochem. Biophys. 501 (1), 615.
Vyth, E.L., Steenhuis, I.H.M., Roodenburg, A.J.C., Brug, J., Seidell, J.C., 2010. Front-ofpack nutrition label stimulates healthier product development: a quantitative
analysis. Int. J. Behav. Nutr. Phys. 7, 65.
Westenbrink, S., Oseredczuk, M., Castanheira, I., Roe, M., 2009. Food composition
databases: the EuroFIR approach to develop tools to assure the quality of the
data compilation process. Food Chem. 113 (3), 759767.
Wills, J.M., Bonsmann, S.S.G., Kolka, M., Grunert, K.G., 2012. European consumers
and health claims: attitudes, understanding and purchasing behaviour. Proc.
Nutr. Soc. 71 (2), 229236.
Wyness, L., Weichselbaum, E., OConnor, A., Williams, E.B., Benelam, B., Riley, H.,
Stanner, S., 2011. Red meat in the diet: an update. Nutr. B 36 (1), 3477.