Faculty of Bio-Science Engineering Academic year 2011 – 2012 The impact of country-of-origin labelling (COO) on food product preference The case of Belgian versus Brazilian chicken meat Lopez Elliot Promotor: Prof. dr. ir. Wim Verbeke Co-promotor: dr. ir. Filiep Vanhonacker Thesis submitted in fulfillment of the requirements for the degree of Master Science in Human Nutrition and Rural Development
77
Embed
The impact of country-of-origin labelling (COO) on food product preference The case of Belgian versus Brazilian chicken meat
Related to food marketing and food legislation. Describes the lack of information and the debate about the importance of Country of Origin as cue from a food product and its influence on consumer's behaviour
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Faculty of Bio-Science Engineering
Academic year 2011 – 2012
The impact of country-of-origin labelling (COO) on food product preference
The case of Belgian versus Brazilian chicken meat
Lopez Elliot Promotor: Prof. dr. ir. Wim Verbeke Co-promotor: dr. ir. Filiep Vanhonacker
Thesis submitted in fulfillment of the requirements for the degree of Master Science in Human Nutrition and Rural
Development
Abstract
The purpose of this master thesis is to provide an in-depth examination of importance of
country-of-origin (COO) theory in perceptions of consumers in a national setting. It shows how
explanatory factors like socio-demographics, familiarity with a country’s products, given
importance to food attributes, and willingness to pay for a foreign country product jointly work
to explain consumers’ COO perceptions. Cross sectional data were collected from a web based
survey through the panel “thesistools” (n = 542). This is a quantitative study using a
questionnaire with 27 questions among Belgian consumers in the Region of Flanders and
Brussels.
Findings: Country-of-origin (COO) was the least important product attribute in a list of 13
attributes on average. Yet, findings showed that it‘s importance differed between consumers. A
profile is drawn of consumers with a different level of interest in origin, in terms of socio-
demographics, and different attitudinal measurements. Further, the impact of country-of-origin
was investigated for chicken meat with Belgian versus Brazilian origin, and the association
between preference for Belgian chicken meat and country image (significantly different),
ethnocentrism (significantly different), interest in foreign cultures (not significant), familiarity
with the country (not significant) was investigated, according to country-of-origin theories.
Research limitations/implications: The study used only respondents from two of the
three regions of Belgium. Future research should seek to develop a multi-dimensional scale for
chicken meat of different countries of origin.
Practical implications: It seems important to increase consumers’ familiarity with a COO
and its products to improve its overall perception. Products imported from developing countries
have the lowest level of familiarity in general. Thus, increasing familiarity with their products is
particularly important to achieve export success.
AKNOWLEDGEMENT
After working during months, I want to say that this work would not have been possible
without the collaboration of many persons that contributed to get to the final book.
First of all, I thank to God, spirituality and faith have drive my life the last years.
Then I would like to thanks to my family that in the distance have supported me and give
me the encouragement needed to pursue my objectives and dreams. To my mom who tought
me the importance of education and being an agent of change, Lupina, you have been my
example all my life, and I will always thank you for all your efforts and help and inconditional
love, mami te quiero mucho!
To my brother Emir Lopez and my sister Erika Lopez who are my examples of excellence, I
love you guys more than you think, and through all this process I thought a lot about you, is
going to take a while to get where you are but I will meet you in your careers dreams.
Very special thanks to my promoter, Prof. Dr. ir Wim Verbeke, who allowed me to be part
of an incredible dynamic and organized team or researchers of the Faculty of Bio-science
Engineering at Ugent. To Dr. ir. Filiep Vanhonacker, who helped me and supported me through
the whole procedure with his quick emails and accurate revisions, without your guidance,
patience and kindness this work would not have been possible, I will always thank you for that.
To all collaborators that facilitated me information in a very efficient and friendly way:
Bastin Valérie (Attaché), Direction générale Potentiel économique. Direction des
Source: CLE and NIS (VLAM, 1999-2009) *Refers to real slaughtered pieces without counting the losses
Flanders represents the strongest contributor for livestock production in Belgium,
therefore its selection as sample region for our research (Figure 2). For 2010, 84% of the poultry
meat in Belgium was produced in the Flemish region, while 16% in the Wallonia region,
increasing for 2011, being 85/15% respectively for both regions, denoting a decrease of the
production in the Wallonia region.
Figure 2. Production of broilers chicken in Belgium by region*, 2000-2011 (pieces)
Source: http://statbel.fgov.be FOD Economie, K.M.O., Middenstand en Energie Algemene Directie Statistiek en Economische Informatie. a. Preliminary results *Includes loses
1 For poultry it is the weight of the cold body of the slaughtered farmyard poultry after being bled, plucked and
eviscerated. The weight includes poultry offal, with the exception of foie gras. For other species, 'carcass weight' is considered to be the weight of the slaughtered animal’s cold body (EC).
Chicken Consumption frequency (%) Never 5,0 Less than Monthly 4,6 Monthly 14,4 Weekly 59,8
pg. 32
2-4 times per week 15,3 Daily 0,9
*Source: FOD Economie. Algemene Directie Statistiek en Economische Informatie, Structuur van de bevolking.
3.3 Questionnaire and Scales
First, variables representing the socio-demographic characteristics were included.
Respondents were asked about their gender, age, number of household members and zipcode.
Within demographics variables living environment was assessed by a seven-point interval scale
ranging from “rural” to “urban”. Education level was assessed by a 4 categories scale that
included: “Diploma from primary school”, “Diploma from High School”, “Diploma College (no
University)”, and “University Diploma”. Nationality of parents was assessed by three open
questions that participants had to fill in related to own, father and mother nationalities. The
variables exposed in the previous three parts contributed to profiling segments for the further
analysis. Self-perceived financial situation was probed by a 7 categories scale, with value 1 as
“difficult”, value 4 “average” and value 7 “wealthy”.
Second, meat consumption was assessed of seven categories starting from “vegetarian”
(those who eat no animal products, those who eat no meat but eggs, dairy and milk, and those
who ate no meat but fish, dairy and milk), followed by those who are “mostly vegetarian” (but
eat meat sometimes), who “mostly eat meat”, and those who “eat meat almost always”.
Frequency of chicken meat consumption was assessed by a frequency scale with six categories,
as following: “never”, “less than monthly”, “monthly”, “weekly”, “2 to 4 times per week”, “daily
or almost daily”. Responsibility for the food purchase within the family was assessed by three
categories possibilities: “I am the main responsible”, “I share responsibility”, and “Other people
in my family are responsible”.
Third, 10 product attributes were probed for their perceived importance (PI) in the food
purchasing decision process of chicken on a seven-point interval scale ranging from “totally
unimportant” to “very important”. The product attributes were: quality, taste, origin (domestic
or imported), price, appearance, freshness, environmental friendliness, availability, animal
welfare, and easy to prepare. PI reflects the individuals’ reaction from a consumer perspective,
i.e. someone who has to weigh and evaluate different product attributes before coming to a
pg. 33
purchase decision. The PI contributed to assess the importance given by consumers to intrinsic
(quality, taste, freshness, appearance) and extrinsic (origin, price, environmental friendliness,
availability, animal welfare, and easy to prepare) food attributes.
Fourth, a measurement for preference between Belgian vs. Brazilian chicken meat was
assessed by a seven categories scale, starting from “Strong preference for Brazilian chicken
meat”, “preference for Brazilian chicken meat”, “Light preference for Brazilian chicken meat” to
“No preference”.
Fifth, comparison in terms of product attributes was measured through the following
statement: “Compared to Belgium, how would you evaluate the chicken meat imported from
Brazil in respect to the following attributes: nutrition value, safety, quality, taste, price, value for
money, freshness, availability” was used. This item was measured on a seven-point interval scale
anchored at the left pole by “much worse” and at the right pole by “much better”, with “equal”
as the mid-point of the scale. This third measure is much more a public opinion, which is
presumed to be held rather independent of the consumption decisions. These statements
provided information relevant to assess the importance of country of origin in relation to
perception of extrinsic and intrinsic food attributes from a domestic vs. an imported product.
Sixth, comparison in terms of production characteristics was evaluated through the
following statement: “Compared to chicken production in Belgium, how would you evaluate the
chicken meat production in Brazil in respect to the following conditions: animal welfare, working
conditions, hygiene, quality standards, environment friendliness, use from hormones and
antibiotics” was used. This was scored on a seven-point interval scale anchored at the left pole
by “much worst in Brazil” and at the right pole by “much better in Brazil”, with “equal” as the
mid-point of the scale. Fourth, a scale that assessed to evaluate the perception of Belgian
consumers about the scale of operation from Brazilian chicken compared to Belgian, this was
scored on a seven-point interval scale anchored at the left pole by “much smaller” and at the
right pole by “much bigger”. This evaluation provided information about perception of
similarities in production practices.
Seventh, similarities in consumer’s willingness to pay for chicken meat of the following
countries was assessed: Brazil, Belgium, France, Sweden, China and USA, using a seven-point
pg. 34
interval scale anchored at the left pole by “totally not prepared” and at the right pole by
“strongly prepared”. This question aimed to compare consumer´s perception
similarities/disparities among countries. Similarities in WTP between BR vs. BE chicken meat was
assessed by the question ‘how much would you be prepared to pay for the following products’:
“Chicken from Brazil instead chicken from Belgium”, “Processed chicken products from Brazil
instead processed products from Belgium (ex. grounded)”, “Chicken meat with a label ‘produced
in Belgium’ instead a not labeled chicken meat”; this was scored on a seven-point interval scale
anchored at the left pole by “Much Less” and at the right pole by “Much more”. This question
contributed to evaluate social similarities and ethnocentric preferences.
Eight, political, economic and cultural similarities among both countries, BE vs. BR was
assessed on a 7 levels two-pole scale, going from “No similarity” to “Strong similarity”. Economic
development perception was assessed using two relevant items on a 7 levels two-pole scale,
going from “Low grade of economic development” to “High grade of economic development”,
for both countries, Belgium and Brazil. It has been found that when there is higher similarity
between countries of origin there is less difference in preference of products from the same
group, these questions attempt to assess the relationship between COO similarities and
consumer’s perception.
Ninth, ethnocentric attitude to Belgian products were assessed by 4 statements, “It is
always better to buy Belgian products”, “Even if it is more expensive, I still choose to buy Belgian
products”, We should buy products from other countries only if we cannot get them in our
country”, “The quality of local products is equal to the imported products”, “Local food products
are safer”. This was assessed by a on a 7 levels two-pole scale, going from “Totally disagree” to
“Totally agree”. Several studies have found that consumer’s preference might be influenced by
their cognitive, normative and affective mechanism towards products, giving as result
preference to domestic above foreigner products (known as ethnocentric behaviour). This
question contributed to assess ethnocentric preferences.
Tenth, consumer interest toward foreign cultures was assessed through the following
affirmations: “I read a lot about other cultures”, “I have a lot of friends from other countries”, “I
travel a lot”, “I follow the international news”, this was done on a 7 levels two-pole scale, going
pg. 35
from “Totally disagree” to “Strongly agree”. Previous research have found that there is a
relationship between interest and contact with foreign cultures and consumer’s evaluation
towards foreign products. This information attempts to relate ethnocentric preferences and
COO theory.
Eleventh, awareness of presence of Brazilian chicken meat on the Belgian market was
assessed by a three possibilities category, as following: “no idea”, “no” and “yes”, followed by
an open question; “percentage if answered ‘yes’”. Familiarity Brazil was assessed by a 7 levels
two-pole scale, going from “Totally not confident” to “A lot of confidence”. This information
contributed to assess awareness of consumers.
Twelfth, willingness to pay for food labeling from BR and BE chicken products was
assessed by a 7 categories scale with the following question “How much would you be ready to
pay for the next chicken products”: with the following options, “chicken from BR instead chicken
of BE”, “processed chicken products from BR instead processed chicken products from BE” and
“chicken filet with a label ‘produced in BE’ instead a non-labeled chicken filet.
3.4 Analyses procedures
Data were analyzed using SPSS 19.0. Bivariate analyses including cross-tabulation with
Chi2-statisitics, Independent Samples T-test and One-Way ANOVA comparison of means with
Tukey post hoc tests at 95% confidence were used to profile the segments in terms of socio-
demographics, and interval-scaled attitudinal questions. Exploratory factor analysis was applied
to find underlying dimensions in the data. Cronbach’s alpha was used to measure the reliability
of the factor analysis outcome.
Given the large sample size and very low numbers of missing responses, pairwise deletion
was used as the method for treating missing values in some variables.
pg. 36
CHAPTER 4
4. RESULTS
4.1 Data Editing
Before starting the analysis, different computations were performed in order to make the
data appropriate for testing the hypotheses. Data set variables were grouped by adding values
to dichotomal variables, such as gender, and for nationality, different nationality than Belgian,
was characterized as foreigner. In order to analyse importance of COO a relative value was
computed. Missing values were not taken into account in the statistical analysis of general
importance given to food attributes, but they were substituted by mean values in the evaluation
of the relative value for COO.
Exploratory factor analysis and reliability analysis was performed on the pooled sample to
group similar variables in constructs. Factor 1 includes the items that refer to the importance
attached to the consumption of domestic products over imported products, and will further be
referred as “Ethnocentrism”. Factor 2 includes to the evaluation of political, economical and
cultural similarities between BE and BR, will be further referred as “Similarity index” (Table 5).
Factor 3, further referred as “interest in foreign cultures” grouped self reported evaluation of
respondents about affirmation such as ‘I read over other cultures’, ‘I have foreign friends’, ‘I
travel a lot’ and ‘I follow international news’ (Table 6). The reliabilities of the ethnocentrism,
similarity index and interest in foreign cultures were assessed using Cronbach’s α. The three of
them, ethnocentrism (α = 0.79), similarity index (α = 0.80) and interest in foreign cultures (α =
0.79) had sufficient internal reliability consistency. Consequently, respondent’s aggregate scores
on the three factors were calculated to be used as classification (segmentation) variables in
subsequent quartiles analysis (see 4.3 and 4.4)
In order to compute the relative importance of COO an equation was calculated and the
sample was split in four quartiles.
pg. 37
A new variable that reflects the perceived difference in economic development between
Belgium and Brazil was computed by substracting the perceived score for the level of economic
development in Brazil from the percevied score for the level of economic development in
Belgium.
Table 5. Factor loading from principal components analysis for evaluation of knowledge and
ethnocentrism
Ethnocentrism Similarity index
Political knowledge -0.11 0.85 Economical knowledge -0.07 0.85 Cultural knowledge -0.03 0.81 Is always better to buy BE products 0.89 0.00 Even if more expensive, 0.90 0.03 I rather to buy BE products We should buy only imported products 0.79 -0.12 when we cannot produce them The quality of the local products is 0.42 -0.07 the same as the imported products Local food products are more safe 0.68 -0.10 % Variance explained 37.72 24.97 Cronbach’s α internal reliability 0.79 0.80
Table 6. Factor loading from principal components analysis for evaluation of interest in foreign
cultures
Interest in foreign culture
I read over other cultures 0.86 I have foreigners friends 0.81 I travel a lot 0.78 I follow international news 0.70 % Variance explained 62.05 Cronbach’s α internal reliability 0.79
pg. 38
4.2 Profiling Variables
After respondents were asked about the importance of a list of chicken meat attributes
means and standard deviation were classified and presented in descendent order, finding as the
three most important freshness, quality and taste. Country of origin on the other hand ranked
last (see Table 7). According to literature, it is often found that different cues are considered
relatively unimportant to the consumer while others are extremely important (Verbeke & Ward,
2003). Those cues that directly address to the quality consistently receive the highest scores
(such as freshness, quality, taste, appearance). The findings of the importance given to intrinsic
attributes (e.g. appearance, colour, leanless, shape, presentation), suggest that consumers will
use these attributes to infer quality of a product among the most important, while extrinsic
attributes (e.g. price, brand name, stamp of quality, country of origin, store, production
information, nutritional information) will only be determinant for final purchase when extra
information is sought. (Issanchau, 1996; Becker, 1999; Becker, 2000; Dransfield, et al., 2005;
Grunert & Wills, 2007). In the general evaluation of quality’s food attributes, freshness
constituted the most important. This is correlated with findings that show that food safety
represents the most important food attribute for meat products (Verbeke & Viaene, 1999),
being freshness related to safety. COO has been found to play a determinant role in consumer’s
preferences when studied as single product attribute (Roosen, Lusk, & Fox, 2003; Bolliger &
Responsible for food Sample 0.09 Purchase (%) Main responsible for 44.3 34.6 46.3 44.6 51.1 food purchase Shared responsibility 42.8 46.9 40.3 43.2 41.0 In food purchase Other person does 12.9 18.5 13.4 12.2 7.9 the food purchase
BR vs. BE preference Mean
pg. 43
Preference BR vs. BE Chicken meat 5.64(+1.23) 4.97
a(+1.09) 5.33
b(+1.15) 6.05
c(+1.13) 6.17
c(+1.15) 0.00
a-b-c Scores in a row with different superscripts are significantly different at P < 0.05 (1-way ANOVA and post hoc Tukey multiple comparison test). 1 Source: FOD Economie. Algemene Directie Statistiek en Economische Informatie, Structuur van de bevolking.
As summary from socio-demographic profiling we found that Q4 correspond the group
that give the highest score when evaluating RICOO, presenting a bigger percentage in the 2
older age categories, as well as main responsibility for food purchase and higher education,
which has been found to be relevant to COO preferences.
While evaluating consumer’s preferences several researchers have found that in order to
establish a more specific link is important to identify groups with similar characteristics, because
these similarities may explain favoritism towards products. For example, in their review Bilkey
and Nes (1982) found that there were differences between groups (eg, students versus non-
students, end-consumers versus industrial buyers) towards products of a specific country, as
well as differences among groups when methodological settings vary (eg, single vs. multiple cue
studies). Baughn and Yaprak (1993) reported that age is often associated with foreign product
acceptance, with younger consumers demonstrating more positive attitudes towards foreign
products. In our findings the group that score higher for preference of BE over BR chicken meat
was composed mostly for people older than 41 years old. This could be explained by the fact
that people from these ages categories are mostly in control of the type of food they purchase,
they may have bigger concerns about food safety and might be less world minded than younger
consumers who display a lower level of prejudice towards foreign products (Rawwas, Rajendran,
& Wuehrer, 2002). It has been also discussed that consumers with a better financial situation
and level of education seek for further information when selecting food products, and that COO
could be a positive or negative determinant when purchasing products (Han & Terpstra, 1998;
Ahmed & d’Astous, 2002). The same quartile also correlate to these findings, the level of their
preference for BE vs. BR chicken meat could be explained by the fact that these consumers
rather to consume a local product already known (assuming quality standards of BE) than an
imported product that they may consider unknown, and from which is difficult to evaluate due
the lack of information about the product, relaying in made in BE. The previous findings support
pg. 44
our Hypothesis 1: Segments with a different level of importance attached to COO differ in terms
of socio-demographics characteristics, meat consumption type and frequency of meat
consumption
Attributes were evaluated according to their absolute scores and listed in descendent
way. Means were compared among quartiles and differences of perception of importance for
each attribute were obtained. We noticed that there is a higher similarity in the perception of
the importance of these attributes among Q3 and Q4, which may be also correlated to the
similarity in their socio-demographics (Table 10). Our findings are a reflection of previous
results that confirm that attitudes toward COO are influenced by cultural and areal proximity
Table 10. Comparison among quartiles of RI scores of food attributes of BE vs. BR chicken
meat
Food product attributes Q1 Q2 Q3 Q4 p-value
Quality 5,85a 6,12a 6,23b 6,32b 0.00
Taste 5,88a 6,10a 6,16b 6,25b 0.03 Price 4,12a 4,63b 4,72b 5,11c 0.00 Appearance 5,35a 5,56a 5,81b 5,91b 0.00 Environmental friendliness 4,10a 5,09b 5,56c 5,90c 0.00 Animal Welfare 4,35a 5,20b 5,48b 5,58b 0.00 Facility to make 4,41a 4,93b 4,94b 4,98b 0.00 a-b-c-d Scores in a row with different superscripts are significantly different at P < 0.05 (1-way ANOVA and post hoc Tukey multiple comparison test).
4.3 COO and Country image
4.3.1 Product perception BE vs. BR
In order to evaluate perception of the product, eight (8) different food attributes from BR
vs. BE chicken meat were evaluated, finding that the attributes that received the highest scores
were price, nutrition value, value for money and taste. This means that respondents perceived
Belgian chicken meat to be better on these attributes as compared to Brazilian chicken meat,
being evaluated to be worse in terms of food safety and freshness (Fig 5). These results are
pg. 45
supported by previous findings that affirm that as an extrinsic attribute, COO has an influence
on consumers’ perceptions of a product’s quality and of its attributes, providing a heuristic basis
for inferring the quality of the product without (sometimes) considering other attributes
information. It appears that subjects transfer the product's COO to its specific attributes
increasing the influence of COO on product evaluations (Schooler, 1971; Bilkey & Nes, 1982;
Hong & Wyer Jr., 1989; Wall, Liefeld, & Heslop, 1991; Ahmed & d’Astous, 2002).
Figure 5. Evaluation of product attributes of Brazilian versus Belgian meat
4.3.2 Production methods perception BE vs. BR
Respondents were asked about their perception of chicken meat production conditions
in Brazil compared to Belgium. Results show that all attributes evaluated scored lower for
Brazilian than the average for chicken meat production within this scale, being the lowest
(perceived as worse) working conditions (mean = 2.4 + 1.17) (table 4).
3.68
2.96
3.37
3.64
4.37
3.66
3
3.38
1 2 3 4 5 6 7
Nutrition
Safety
Quality
Taste
Price
Value/money
Freshness
Availability
Values
Qu
alit
y at
trib
ute
s
BE BR
much worse bit equal bit better much worse worse better better
pg. 46
Figure 6. Evaluation of Brazilian Chicken Meat Production (means)
Appears to be that consumers evaluate attributes based on the country image, but this
evaluation is also related to the country image perception depending of product categories.
Pouta et al. (2010) studied COO and production methods for meat, finding that although
production methods were significant, compared to COO was minor since consumers prefered
domestic or neighbour countries products, denoting a bigger influence from COO, which
correspond to same findings that we have (Pouta, Heikkilä, Forsman-Hugg, Isoniemi, & Mäkelä,
2010). Respect to product categories, for example, Nagashima (1970) and Narayana (1981)
found differences in country image perception between Japanese and USA consumers, which
positioned USA products always in the first place. It is possible that Brazilian meat does not
belong to a category well known by Belgian consumers, and that stereotyping categories could
explain how consumers react to COO information and to the evaluation of products from
foreign countries affecting the cognitive processing of other product-related cues. (Nagashima,
attached to COO is positively related to consumer preference for domestic chicken and
Hypothesis 4: There is a positive relationship between importance attached to COO and level of
ethnocentrism.
Table 11. Difference among quartiles for preference of BR vs. BE chicken meat
Variable Q1 Q2 Q3 Q4 p-value
Preference for BE vs. BR 4.97a 5.33b 6.05c 6.17c 0.00 Chicken meat
a-b-c Scores in a row with different superscripts are significantly different at P < 0.05
(1-way ANOVA and post hoc Tukey multiple comparison test).
In order to verify which variables influenced most strongly the overall preference for
chicken meat of BR vs. BE origin, variables regarding to attributes (nutrition value, safety,
quality, taste, price, value for money, freshness, availability) and to production methods were
taken into account (animal welfare, working conditions, hygiene, quality standard,
environmental friendly production, use of hormones/antibiotics), perception of scale of broiler
and perception of chicken farms size in BR vs. BE were also used to perform a linear regression.
A linear regression analysis was applied, as result the following equation:
Y = cst -0.282*X1 -0.194*X2 -0.132*X3
Y = Preference for chicken meat BR vs. BE X1= Freshness X2= use of hormones/antibiotics X3= working conditions This results mean that if the value of freshness increase in 1 unit, then preference for BR
chicken meat versus Belgian chicken meat decrease 0.282 units, under the assumption that all
pg. 50
other variables remain the same, in other words, as higher the importance given to freshness as
less preferred BR chicken meat would be, followed in the same way by use of
hormones/antibiotics and working conditions.
A survey conducted by Verbeke and Ward (2003) explored the importance of traceability,
COO, and several beef quality cues in Belgium. They found that survey participants expressed
more interest in labeling cues denoting quality and quality standards than in labeling cues
related to traceability and origin (Verbeke & Ward, 2003). Ehmke (2008) found that COO
information was not as important as genetically modified content information (France, USA, and
Niger) or organic production (China). Becker (2000) found that on average, COO, alongside
colour and place of purchase, were regarded as most helpful in assessing both eating quality
and food safety concerns of meat. In Germany and Sweden COO was the most important factor
determining both eating quality and safety aspects. In the UK, however, colour, leanness, or
place of purchase was regarded as most important. Roosen et al., (2003) using data from
Germany, France and UK in their analysis, found that consumers (mostly from France and
Germany) place more importance on labels of origin than any other product attribute such as
brand, price, marbling or fat content; more than 90% of surveyed consumers wanted a
mandatory labeling program for beef produced from cattle fed genetically modified crops.
Thus, appears to be that for individuals with quality and food safety information needs,
COO information is relatively less important, yet when extrinsic cues are used as extra
information by consumers, COO has an effect in their final purchase decision.
4.5 COO and Willingness to Pay
To evaluate perception of economic development a new variable was created by
substracting values given to BE-BR economic development as results the values given by
consumers mean that as negative the value as better development perceived for BR and worse
perceived development for BE (from -6 to 6 . The mean value was 1.47 (+2.02), meaning that the
perception of economic development was higher for BE, although not with the highest score
(Table 12).
pg. 51
Table 12. Difference among quartiles for perception of economic development scores BE-BR
Variable Mean(SD) Q1 Q2 Q3 Q4 p-value
Perception of economic 1.67(2.42) 1.50(2.16) 1.59(1.79) 2.01(1.90) 0.25 development between BE-BR
To evaluate willingess to pay for chicken meat comparing perceived similar countries we
found that consumers evaluate more favorably France and Sweden, than Brazil, USA and China
(Table 13).
Several findings suggest that the familiarity with products made in a country can be a
good predictor of COO perceptions (Ahmed & d’Astous, 2002). Consumers positive evaluation
for similar neighboring countries might be regarded to concerns about countries that are
geographically located far away, and that due the time of transport can become a risk for food
safety, as well as concerns for quality standards in less developed countries. Previous
researchers have affirmed the same, for example Juric and Worsley (1989) found that food from
neighbouring countries are perceived as being superior to food from more distant countries due
to similar cultural beliefs and areal proximity. Alfnes (2004) showed, that on average, Norwegian
consumers preferred domestic or Swedish beef to beef from more distant countries, in addition,
beef from developed countries was preferred to beef from less developed countries such as
Botswana (Alfnes, 2004). Roth and Romeo (1992) found that consumers were willing to buy
autos and watches from Japan, Germany, and the U.S. appearing this to be related to these
countries' high overall image. Likewise, respondents' unwillingness to buy these products from
Mexico and Hungary appears due to the poor overall image of these countries. These results are
similar to other studies that found that automobiles manufactured in Japan, West Germany, and
the U.S. were preferred over those from England, France, Brazil, Mexico, Taiwan, and South
Korea. Subcultural biases in preferences might lead consumers to favour products from
countries with a similar ethnical and economic situation, especially if there are intra-national
variations in culture (Laroche M. , Papadopoulos, Heslop, & Bergeron, 2003).
pg. 52
Table 13. Willingness to pay for chicken meat of developed and developing countries
COO (%) Not willing Willing Well prepared
BE 4.3 8.9 86.8 FR 12.7 15.0 72.3 SWE 19.5 18.8 61.7 BR 58.5 22.1 19.4 USA 65.0 19.6 15.3 CHINA 85.8 9.1 5.0
Respondents were asked about their perception of similarity from political ideas,
economic situation and culture between BE and BR. An exploratory factor analysis and reliability
test was performed on the pooled sample. A new variable was computed as average of the
political, economic and cultural knowledge. The findings concerning similarity index (political,
economical and cultural similarities between Brazil and Belgium) confirm that groups that have
similar knowledge about a country, would have a bigger influence in assessing COO than groups
that differ in these ideas. Several studies have found that when evaluating information of COO
as attribute, categories of products play an important role (Han & Terpstra, 1998; Tseng &
Balabanis, 2011). In our results we found that the majority of respondents were ready to pay
more for a European country (France, Sweden) than for a non-European country (Brazil, USA or
China) and this may have sense if we take into account that freshness, quality and taste were
the attributes with a higher scores. This supports our Hypothesis 5: Consumers are willing to pay
more for domestic products/developed country products than for foreigner/developing country
products.
4.6 Interest in foreign cultures
Respondents were asked about if they read and have friends of other cultures, travel a lot
and keep updated about international news. An exploratory factor analysis and reliability test
was performed in the pooled sample, and new grouping variable as average of the 4 considered
was created and oneway-ANOVA applied to compare means from quartiles, no differences
among quartiles were found (Table 14).
pg. 53
Different researchers have investigated the interest in foreign cultures and the
relationship with products evaluation. For example, Papadopoulos and Heslop (1986) compared
Canadian consumers who had visited a country with those who had not, and found that visiting
a country reduces the gap between the more global, prevailing public image of its products and
its actual capabilities. Balabanis et al. (2002), affirmed that a greater level of direct contact with
a country or its products lead to more objective consumer product perceptions. Our findings
show that Q4 had the highest score for interest in foreign cultures, however the higher
preference for BE over BR chicken meat, which could be explained by socio-demographic that
support that older individuals tend to score less favorable foreign products, and as well that
people with a higher level of education and financial situation tend to evaluate foreigner
products less positively than those with lower educational level and financial situation.
Table 14. Difference among quartiles for interest in foreigner cultures
Variable Mean(SD) Q1 Q2 Q3 Q4 p-value
Interest in foreign cultures 4.60 4.59 4.50 4.85 0.10
Awareness of BR chicken in BE market was assessed by the question “do you think that BR
chicken meat is present in BE market”, a majority of the respondents of the sample with 70.3%
“had no idea”, 3.5% said “no” and 26.2% said “yes”. These results support the COO labeling
theory, that affirm that when consumers have no access to extrinsic information the evaluation
of the perceived quality is based in intrinsic attributes, and could be explained that due their
lack of knowledge about the presence of Brazilian chicken meat in the Belgian market, they may
relay in what “they know” about quality standards of domestic products and use their beliefs
when it comes to perception of quality standards of BR, evaluating its products in a less positive
way.
pg. 54
4.6.1 COO and familiarity
To evaluate the familiarity and its relationship with the aversion of the respondents with
BR among the quartiles one way-ANOVA was applied. It was expected that as greater the
exposure of consumers to BR as less aversion, no differences were found among the quartiles.
Table 15. Difference among quartiles regarding familiarity to BR
Variable Mean Q1 Q2 Q3 Q4 p-value
Familiarity with BR 2.25 2.27 2.39 2.41 0.72
It has been proposed that attitudes are influenced by both, subjective familiarity (the
subjects think they are familiar with the stimulus, COO) and objective familiarity (actual
exposure to the stimulus, COO objective knowledge), and this could partially explain why the
respondents evaluation for preference of BE over BR chicken meat was high.
It has been described that country image is the overall perception a consumer form of
products from a particular country, based on their prior perceptions of the country's production
and marketing strengths and weaknesses. This definition brings country image closer to the
means consumers use in assessing products. What consumers know (or think they know) about
a country's manufacturing ability, flair for style and design, and technological innovativeness,
seems much more congruent with their product perception formation than do other, less
production and marketing-oriented factors. There is also a possible a link between COO labeling
and preference. Consumer’s product’s evaluation is based on the information given by
governments/marketers, and by not knowing the COO of a product they may prefer this product
based in other attributes evaluation or based in the country image and belief that they have
towards that country.
pg. 55
4.7 Influence of level of processing
To evaluate correlation between willingness to pay for chicken meat products different
options were given to the consumers. Negative correlation among the willingness to pay for BR
chicken filet and for processed chicken products from BR over BE was found (p-value=0.00).
Respect to labeling of chicken filet with a brand “made in BE” was negatively associated with
willingness to pay for chicken meat products from BR instead BE, meaning this that consumers
that prefer BE labeled chicken products are less willing to pay for BR chicken meat products. A
negative correlation between readiness to pay for processed meat products from BR over BE
was found in relationship with chicken meat labeled as “made in BE”, meaning this that
consumers whom prefer “made in BE” chicken filet are less willing to pay for chicken meat
products of BR.
Table 16. Difference among quartiles for willingness to pay for BR vs. BE chicken products
Variable Mean Q1 Q2 Q3 Q4 p-value
Willingness to pay 3.25a 3.08b 2.96b 2.46b 0.00 for chicken filet from BR over BE Willingness to pay for 3.13a 2.79b 2.79b 2.39b 0.00 Processed chicken products Willingness to pay 4.32a 4.74b 4.83b 5.14c 0.00 for a ‘made in’ BE label instead none
During the review of literature we found that Belgian consumption of chicken meat have
increased along the years, due a shift of dietetic patterns, financial reasons, health concerns,
etc., and also that imports of BR chicken meat have significantly increased, most of all, for the
frozen and processed type. It could be a link between these findings and the current absence of
a mandatory labeling of COO when food is processed, meaning this that although consumers
purchase BR chicken meat through their habitual consumption, they are not aware of it, and
when performing evaluation of preference they relay in COO and domestic products, assuming
that the standards of quality are good if domestic processed although the original food COO is
other than BE. This correlate positively with our findings that suggest that Q4 which gave the
pg. 56
highest score preferring BE chicken meat over BR, is less willing to pay for chicken filet and
processed chicken products from BR, but willing to pay more for a ‘made in Belgium’ label,
instead no label. During the review of the statistics of type of chicken meat products we found
that chicken filet is the favorite cut preferred by Belgian consumers, which seems to correlate to
our findings and COO, denoting that the segment that preferred the most BE over BR chicken
meat, Q4, gave even less score when it was an evaluation between filet and processed chicken
meat products, giving more importance to chicken filet over processed chicken meat, and
therefore scoring even lower for this type of chicken cut.
pg. 57
CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS
The first objective of this study was to verify the importance of different food product
attributes in food purchasing decisions of chicken meat (and the ranking of COO in that list). The
results show that COO as food attribute is less important than other food attributes when an
overall evaluation of the product is done, yet a significant effect in consumer’s preference when
evaluating BR vs. BE chicken meat, being the last one significantly preferred among Belgian
consumers.
Food origin is an important product characteristic for many consumers. It affects the
significance of how consumers distinct foods.
The second objective of our master thesis was to profile groups with different levels of
interest in COO in terms of socio-demographics, responsibility for food purchase, meat
consumer profile, and consumption frequency, in order to compare groups with similar profiles
and to establish a relationship among these profiles and their level of preference or not for
domestic vs. imported products. Different authors affirm that world-minded consumers tend to
be younger, better educated and more affluent (Hett, 1993). Consumers with more income and
education accept foreign products more readily (Niss, 1996). Several authors found that
younger, wealthier and more educated consumers evaluate foreign products more favourably.
Our findings support the previous through the segmentation of quartiles and the comparison of
evaluation of these quartiles with significant higher preference for Belgian over Brazilian chicken
meat. The groups that evaluated highest this preference was composed for older, main
responsible for food purchase and highly educated respondents. Yet these findings contradict
those obtained by Niss (1996), this could be explained as well as the relatively significant low
familiarity with Brazil (and its products) and due the country image that they have over Brazil as
producer and developing economy.
Our third, and last objective was to verify whether theory of COO is also applicable in food
product preference for the case study of Belgian versus Brazilian chicken meat in the context of:
pg. 58
- Country Image
- Ethnocentrism
- Interest for foreign cultures
- Perceived similarity
The results of the present research lead us to affirm that there is an influence of country-
of-origin on these variables, and based on our sample we could conclude that Belgian
consumers prefer BE chicken meat, and that if they would have to choose, they would be able
to pay for a product that originate from a country with similar beliefs, to which they are more
familiar with and possess a similar level of economic development.
It is possible that Belgian consumers evaluate products based in countries stereotypes
(negative or positive), therefore the management of a product’s national image is an important
element in the strategic marketing decision when related to consumer’s preference.
With the expanding global economy, aspects concerning country-of-origin (COO) and its
communication are gaining in importance especially in the agro-food sector. Our findings
corroborate that for many consumers, country-of-origin is a determining purchase criterion in
food consumption.
Consumers make decisions about the quality of products based on a systematic process of
acquisition, evaluation and integration of product information or cues. For this reason it
becomes apparent why extrinsic product cues have gained importance within food products
evaluation. However, COO has been identified as well as a credence attribute, affecting
evaluation of products differently.
Appears to be that concerns over quality standards were related as well as a variable than
influenced our results, from which we concluded that origin can be considered as an important
attribute or cue in consumers’ evaluating process for food quality and safety aspects because
some consumers considered own country products more trustable because of better food
safety and production quality standards, and that is the probable reason why some consumers
are willing to pay more for domestic or labeled own country-of-origin products.
For agribusiness and marketers these insights open up positioning potentials and are
relevant for strategic marketing and communication purposes. If subjects transfer product's
pg. 59
COO to its specific attributes increasing the influence of COO on product evaluations, how could
countries change their image, improving their chance to penetrate foreign markets and gain
confidence from consumers. This means that managers would benefit by having a better
understanding of a product's COO, as well as identifying the dimensions along which country
image should be improved. One possible strategy when an unfavorable mismatch exists was
proposed by Roth and Romeo (1992), they suggested to consider a joint venture within a
favorable match country; e.g. a Hungarian car manufacturer may benefit from manufacturing
and/or marketing its cars with a German partner. This type of strategies should be considered
by countries that differ in strengths and weaknesses across products and that could benefit
from having a partner with a better country image, or with a more trustable known experience
in the market.
Evaluation country of origin as an attribute is not an easy task, and we also got to the
conclusion that international marketers should analyse subcultural differences before expanding
into targeted countries and evaluate own country image. Then the producer may need to de-
emphasise or perhaps even disguise or hide the product’s origin in order to make it more
acceptable. Finding subcultural differences would allow managers to improve the development
and implementation of marketing strategies, and could lead them to bussiness success and to
understand better which are the possible market niches to be filled.
Little consumer research was found specifically with respect to perception of country of
origin labeling from Belgian consumers and chicken meat products from domestic versus
imported origin. From theory it can be expected that consumers would have a higher preference
and a better image from countries with a similar (higher) level of development and production
specialization, however this may be cross linked to product-specific and country-specific.
The limitations of this work was that the survey was the self reported and subjective
nature of the measure by respondents. The biggest part of the sample was composed by
respondents that live in Flanders, therefore Flemish speaking. This could represent a bias when
evaluating this sample as representative for Belgian consumers without including consumers
from Wallonia, which may present different socio-demographic profile
pg. 60
Further research is needed to strength knowledge about COO perception of Belgian
consumers among other food attributes and among countries.
pg. 61
Bibliography
ABEF. (2010). The World Eats Better witn Brazilian Chicken. Brazilian Chicken Producers and