European Journal of Marketing
European Journal of Marketing,
Nutrition labelling:issues and policies George Baltas
Athens University of Economics and Business, Athens, Greece

Keywords Consumer behaviour, Labelling, Food, Nutrition

Abstract Nutrition labelling of food products has received considerable attention in themarketing literature due to increasing consumer interest in health and diet issues. Nutrition
labelling of food products is intended to enable informed consumer choices and stimulate theconsumption and production of healthful products. Reviews nutrition labelling research, organisesthe literature, 留學生MBAdissertation定制and discusses implications for policy makers, managers and researchers. Analysisof the literature suggests heterogeneous utilisation and comprehension of on-pack nutritioninformation by consumers. The effectiveness of nutrition labelling depends also on theorganisation and presentation of the information, implying the importance of regulatoryissues. The topic is rich in public and marketing policy implications and provides severalopportunities for further research.
Introduction
Recognition that dietary habits can reduce the risk of chronic diseases andimprove health has generated considerable interest in nutrition labelling offood products, that is information appearing on food package labels andrelating to the nutritional composition of the products (e.g. EC Directive 90/496). Nutrition labelling is intended to enable informed consumer choice, raise
the demand for healthier food products, encourage competition on nutritionalquality, and stimulate development and production of goods with improvednutritional properties.
This paper surveys nutrition labelling research, develops a synthesis of thevaried contributions, and identifies policy and research implications. It is alsointended to help managers, researchers and policy makers keep informed of anincreasingly important area, which may not directly fall within their ownspecific professional or research interests.
The organisation of the paper is as follows. Before addressing specificissues, we motivate our discussion by briefly reviewing the main researchapproaches to nutrition labelling. The third section discusses determinants ofnutrition labelling utilisation. Within this broad theme, we consider utilisationheterogeneity, prevalence of information on negative ingredients, and
acquisition of nutrition information from other sources. Subsequently, we turnto discuss information comprehension, which is another important determinantof nutrition labelling effectiveness, besides information acquisition. The fifthsection is concerned with organisation and presentation of on-pack nutritioninformation. The role of nutrition labelling in the determination of consumption
The research register for this journal is available athttp://www.mcbup.com/research_registers
The current issue and full text archive of this journal is available at#p#分頁標題#e#
http://www.emerald-library.com/ft
The author wishes to thank two anonymous EJM reviewers, whose detailed and constructivecomments improved the paper.
Nutrition
labelling: issues
and policies
709
patterns is then discussed. Section seven provides a brief exposition to currentregulations. A concluding section discusses implications for policy makers,
marketing managers and researchers.Overview of research approaches
The interdisciplinary literature draws from various theories. Although the roleof theory is not always explicit in structuring empirical research, the literature
demonstrates how different intellectual traditions can shed light on the samephenomenon.
More specifically, within the economics of information approach proposedby Stigler (1961), the consumer searches for product information, such as
nutrition labelling, as long as the additional costs of searching do not outweighhe additional benefits of searching (Gurthie et al., 1995; Nagya et al., 1998;
Russo, 1987). In formal terms, consumers will maximise the utility of theirpurchase decisions by looking for information until the marginal cost of thesearch exceeds the marginal value (Stigler, 1961; Nelson, 1970; Avery, 1996).
The costs of utilising nutrition labels lie in the effort and time needed to gatherand process information (Russo et al., 1986). The benefits arise from healthfulfood choices through information about the nutritional quality of foodproducts. Information search increases where the perceived decision risk issignificant in terms of money or in terms of the effect being visible (Beattie andSmith, 1987; Sadler, 1999). For example, a consumer may look at labelling tofind the amount of fat, i.e. a nutrient with visible effects on the human body, butdisregard the amounts of sodium, i.e. an attribute with less visible
consequences (e.g. Keller et al., 1997).The characteristics theory (e.g. Lancaster, 1991; Anderson et al., 1992) viewsnutrition attributes as utility-generating characteristics. Product demand is
derived from the demand for product characteristics, including nutritionattributes. Hedonic price analysis (e.g. Baltas and Freeman, 2000) postulatesthat price reflects the bundle of product characteristics valued by some implicitor shadow prices and belongs to this broad economic framework (Stanley,1991).
http://www.400190.com/dissertation_writing/MBA/Psychological models of information processing activities are also dominantframeworks. In particular, important cognitive outcomes such as attention,acquisition and encoding of nutrition information have been considered in theliterature (Cole and Gaeth, 1990; Cole and Balasubramian, 1993; Moorman,1990, 1996).
Of equal interest is prospect theory, which attempts to explain how peoplevalue the goodness or badness of an option (i.e. a prospect). A hypotheticalvalue function is invoked to describe the relationship between thepsychological and actual valuation of gains and losses resulting from an action.#p#分頁標題#e#
The hypothetical value function is steeper for losses than for gains, indicatingthat, all else being equal, people weight losses more heavily than gains(Kahneman and Tversky, 1979). The principal implication for nutritionlabelling is that people may pay more attention to diet-related losses than to
EuropeanJournal ofMarketingdiet-related gains. As will be seen below, the applicability of prospect theory tonutrition labelling issues is apparent (e.g. Burton and Andrews, 1996).
The various theoretical approaches reflect the diversity in the issuesaddressed by nutrition labelling research. More specifically, informationprocessing
studies have been mainly concerned with inter-personal differencesin cognitive outcomes and differential effectiveness of alternative information
formats. Prospect theory is particularly useful in explaining the emphasis onnegative nutrients such as cholesterol, sodium and fat and the greater attention
paid to labels by ill or dieting individuals. Similarly, the economics of
information provides a structural approach to the determination of
information-processing activities and applies, in essence, a cost-benefit
analysis to nutrition information search. Differences in costs and benefits from
nutrition information then explain the heterogeneous label use by consumers
and the different effectiveness of alternative formats. Finally, the
characteristics theory and, in particular, the hedonic price model provide
monetary valuations of nutrition characteristics and reveal differences between
time periods, areas and consumer segments.
The same diversity is observed in data sources and data-analysis methods.
Survey data on consumer attitudes and cognitive outcomes are most common
(e.g. Nagya, 1997) but some studies do employ demand data (e.g. Mathios, 1998)
or experimental data (e.g. Cole and Gaeth, 1990) on behavioural outcomes.
Similarly, several analytical methods are used, ranging from descriptive
statistics (e.g. Abbott, 1997) to econometric models of qualitative response
(Wang et al., 1995). The analytical method is largely dictated by research
objectives, underlying theory and nature of the data. Finally, qualitative
research methods and, in particular, focus groups are also employed in a small
number of studies (Zarkin and Anderson, 1992; Shannon, 1997).
Determinants of label utilisation
Several studies report that consumers want on-pack nutrition information
(Lenahan et al., 1973; Daly, 1976; Bender and Derby, 1992; Abbott, 1997) and
make unfavourable inferences about the nutritional properties of products
lacking such information (Frieden, 1981; Huber and McCann, 1982; Dick et al.,
1990; Zarkin and Anderson, 1992; Burke, 1996). However, an important issue is
the extent to which consumers actually gather on-pack nutrition information, in
other words, use the label.
Interpersonal differences in label use#p#分頁標題#e#
The amount of search should, in principle, vary across individuals due to
heterogeneous costs and benefits of information acquisition. The literature
provides empirical evidence to suggest that this inter-individual variation may
be associated with consumer characteristics, in particular, education, gender
and age.
More specifically, education is consistently found to be positively correlated
with label use (Klopp and McDonald, 1981; Moorman, 1990; Mueller, 1991;
Nutrition
labelling: issues
and policies
711
Guthrie et al., 1995; Variyam et al., 1996; Nagya, 1997; Nagya et al., 1998;
Variyam et al., 1998). Research also indicates that label use is greater among
female consumers (Mueller, 1991; Guthrie et al., 1995; Mangleburg et al., 1997;
Nagya, 1997). Other researchers look at the effects of age and suggest that the
elderly limit the amount of information they obtain about products prior to a
purchase decision and look at nutrition labels less (Mueller, 1991; Cole and
Balasubramian, 1993).
Apart from sociodemographic variables, a number of other factors have
been identified as determinants of label usage, such as nutrition and health
awareness (Klopp and McDonald, 1981; Guthrie et al., 1995; Wang et al., 1995;
Szykman et al., 1997). Label usage may also depend on time spent on shopping
and the importance to the consumer of price and taste (Nagya et al., 1998).
Utilisation is also greater for consumers who value nutrition and product safety
over taste (Guthrie et al., 1995). A further factor that reduces nutrition
information search is the time pressure on the individual at the place of
purchase (Moore and Lehmann, 1980; Beattie and Smith, 1987).
Emphasis on negative nutrients
As mentioned above, a fundamental hypothesis in prospect theory is that, all
else being equal, losses are given more weight than gains (e.g. Kahnemann and
Tversky, 1979). In the present context, information on negative nutrients, such
as fat, may be more salient than information on positive nutrients, such as
vitamins (Burton and Biswas, 1993; Shannon, 1997; Sadler, 1999). Motivation to
search for information is greater when consumers are provided with highly
arousing negative consequences and specific ways to minimise them
(Moorman, 1990).
Nutrition information may be more useful and relevant to consumers for
product categories that are perceived to be less nutritious (Burton and
Andrews, 1996). Similarly, consumers may pay considerable attention to
labelling of fat-modified food products. Nutrition information and producer
claims have particular relevance for such products. First, consumers may use
the nutrition information in order to verify the reduced- fat characterisation of
the product (Keller et al., 1997). Second, some consumers may search for low-fat#p#分頁標題#e#
variants of food products and tend to prefer products identified as reduced-fat
or fat-free versions (Bushman, 1998). The positive impact of such ``light'' claims
on consumer demand may increase producers' incentives to innovate and
create variants with improved nutritional profiles (Ippolito and Mathios, 1993).
Additional sources of nutrition information
Nutrition labelling is one of several information channels available to
consumers, including mass media, books, advertising, education programmes
and, of course, medical advice (Feick et al., 1986; Mueller, 1991; Caswell, 1992;
Abbott, 1997).
Nutrition-related advertisements may have a positive effect on awareness of
diet-health relationships and purchase behaviour (Ippolito and Mathios, 1990;
European
Journal of
Marketing
35,5/6
712
Padberg, 1992; Jensen and Kesevan, 1993). A further source of information is
the producer claim that often appears on packaging, for example ``fat free''.
There are three types of claims. Absolute nutrient claims indicate the level of a
nutrient, for example ``99 percent fat-free''. Comparative nutrient claims
compare the level of a nutrient with that in another food, for example ``40
percent less fat than the regular version''. Health claims link the food or nutrient
with a disease or health-related condition, for example ``rich in vitamins with
antioxidant action''. Claims create expectations regarding the healthfulness of
the food, which may be confirmed or refuted by the actual nutritional
composition (Friestad and Wright, 1994; Ford et al., 1996; Keller et al., 1997).
Truthful claims may be helpful in alerting consumers to products they might
consider in order to improve their diets (Ippolito and Mathios, 1993). For
example, Ippolito and Mathios (1991) show that most consumers increased
their fiber cereal consumption once manufacturers were allowed to make health
claims.
The question as to whether claims and nutrition information have
interdependent effects is important but related research remains inconclusive.
Ford et al. (1996) report that health claims and nutrition information have
independent effects because the presence of a health claim does not influence
consumers' processing of nutrition information. Keller et al. (1997) find that
consumers rely more on nutrition information than on claims when both are
available. Nutrition information is seen as more credible and people may use it
to check the veracity of claims. In the same study, claims interact with nutrition
information in affecting perception of manufacturer credibility but claims
consistent with nutrition information do not result in more favourable product
evaluations. Mazis and Raymond (1997) find that consumers' beliefs about food
products differ when consumers are exposed to health claims on labels that#p#分頁標題#e#
include nutrition information versus when consumers are exposed to claims on
labels that do not include nutrition information. The use of claims may also be
related positively to the use of nutrition labels because some consumers tend to
use all information readily available to them (Szykman et al., 1997). One may
conclude that use of claims and use of nutrition labelling are correlated over
people in the sense that information-seekers usually look at both. More work,
however, is needed on the empirical identification of more specific interactions.
Consumer understanding of on-pack nutrition information
The comprehension, elaboration and integration of on-pack nutrition
information are cognitive outcomes of great importance for researchers and
planners. Although consumers increasingly look at labels, their understanding
of nutrition information is a major issue. For example, in a product choice test
nutrition experts make identical choices while there is significant disparity
among consumers' choices (Padberg, 1992).
It appears that while consumers claim extensive label usage and general
understanding of dietary matters, they are often confused by technical terms
(Mueller, 1991; Black and Rayner, 1992; Eves et al., 1994; Abbott, 1997) and
Nutrition
labelling: issues
and policies
713
miscalculate nutrient intake (Frazao and Cleveland, 1994). For example, some
people do not know what recommended daily values mean or how to use them
in dietary planning (Burton and Andrews, 1996). Petrucelli (1996) argues that
understanding of all available data is not necessary because their mere
existence is enough to allow comparisons between products and increase
consumer confidence in product choices.
As regards interpersonal differences, the literature suggests that age has a
significant negative impact on comprehension of nutrition information (Fusillo
and Beloian, 1977; Moorman, 1990; Cole and Gaeth, 1990; Cole and
Balasubramian, 1993; Burton and Andrews, 1996) and performance on new and
unfamiliar tasks (Salthouse, 1991).
Organisation and presentation of information
Researchers have tested various methods to improve usage and interpretation
of nutrition information (e.g. Muller, 1985; Russo and Leclerc, 1991). Particular
attention has been paid to effective label formats including summary
descriptors (Russo et al., 1986), reference information (Moorman, 1990; Burton
et al., 1994), category-specific mean nutrient values (Viswanathan, 1994),
graphical presentations (Levy et al., 1992) and verbal descriptions (Black and
Rayner, 1992).
For example, Russo et al. (1986) test a summary measure, with a number of
stars to represent overall nutritional value, but the measure appears to be too
prescriptive and ignores consumer heterogeneity in the importance of#p#分頁標題#e#
individual nutrients.
Barone et al. (1996) examine reference value and average value formats.
With the reference values, consumers form a more accurate impression of a
product's healthfulness with respect to a daily diet. The average values are
more useful for discerning between products within a product class.
Several information presentation strategies are tested by Levy et al. (1996).
Plain metric figures are easiest to use. The reference information, however,
proves better for making calculations in a wider dietary context, and
recognising high and low nutrient levels. Highlighting or grouping nutrients
makes little difference. Adjectival summary descriptions of nutrient levels and
percentages of recommended daily allowances are effective and reduce
processing effort (cf. Russo and LeClerc, 1991).
In Europe, research undertaken by the Coronary Prevention Group (Black
and Rayner, 1992) concludes that an adjectival format based on verbal banding
schemes, i.e. high, medium and low, is the most effective. A star indicator is
confusing, with consumers unable to understand that the number of stars have
an inverse relationship with how well the product performed on the fat
attribute, but a direct relationship with how well it performs on the fibre
attribute. Recommended daily intake percentages are helpful and consumers
express a desire for these percentages per serving as well as per pack.
Recently, the UK Institute of Grocery Distribution undertook a large-scale
research project to investigate formats that would supplement the labelling
European
Journal of
Marketing
35,5/6
714
dictated by EU legislation. In this study, the most useful formats are an
additional box to emphasise calories and fat per serving, guideline daily values
of calories and fat, and use of rounded numbers (Sadler, 1999).
A final issue concerns consumer preferences for alternative formats. Geiger
et al. (1991) examine consumer acceptance of multiple label formats and
conclude that consumers like labels with detailed information, bar graphs and
nutrient values in both absolute numbers and percentages. However, Levy et al.
(1992) compare actual effectiveness and consumer preference for five label
formats and demonstrate that performance and stated preference do not
coincide, suggesting that consumer acceptance may not be indicative of actual
effectiveness.
Nutritional labelling and purchase behaviour
In the UK, more than 80 percent of surveyed individuals claim that they look at
labels and that label information affects their purchase decision (Abbott, 1997).
Most consumers also use information on nutrition labels the first time they
purchase a product and this then becomes a source of new knowledge they can
draw on in subsequent purchases (Mueller, 1991; Zarkin and Anderson, 1992).#p#分頁標題#e#
Similarly, a frequently mentioned reason for switching products is to reduce
the intake of negative food components (Russo et al., 1986). As alluded to
earlier, minimising the intake of negative nutrients seems to be more important
to consumers than maximising the intake of positive ones. Nonetheless,
Variyam et al. (1996, 1998) confirm the influence of nutrition information on
intake of fibre, which is perceived as a positive attribute, and cholesterol, which
is perceived as a negative attribute.
Nutritional attributes are, of course, only a subset of objective and perceived
characteristics such as price, taste and brand name determining consumer
preferences (Muller, 1985; Zarkin and Anderson, 1992; Cole and
Balasubramian, 1993; Eves et al., 1994). Their relative importance for the
determination of consumption patterns may vary not only over people, but also
across product categories and purchase occasions (Sadler, 1999).
Nutrition labelling standards
The EU Food Law Harmonisation programme of the early 1990s was instigated
to ensure food safety, establish fair-trading measures and remove barriers to
trade. The EC Directive on Nutrition Labelling was passed in 1990 to
standardise nutrition labelling across the member states (e.g. Morris, 1993).
Under the directive, 90/496 nutrition labelling is optional but becomes
compulsory where a nutrition claim is made. Where on-pack nutrition
information is provided, it should include energy, protein, carbohydrate and fat
or all the former plus sugars, saturates, fibre and sodium. In addition, nutrition
labelling may include data on starch, polyols, mono-unsaturates, polyunsaturates,
cholesterol and listed minerals and vitamins. The directive also
dictates the order in which the nutrients should be presented, the units of
measurement and, in addition, allows information in graphical form. The
Nutrition
labelling: issues
and policies
715
nutritional composition should be expressed per 100g but may also be given
per serving. Any declared vitamins and minerals must be expressed as a
percentage of recommended daily allowance, besides absolute amounts.
The US regulations are different, mainly because nutrition labelling is
compulsory. In 1990, the National Labelling and Education Act (NLEA) was
passed, regulating health claims according to actual nutrient content and
introducing compulsory nutrition facts panels (Shank, 1992; Petrucelli, 1996).
Implementation of required changes in packaging was costly, but was
estimated to make eventual savings in the long run due to decreases in
healthcare cost (Silverglade, 1996).
Assessing the regulations
Opinions about the legislation, especially nutrition claim regulations, are
divided. Caswell (1992) argues that the legislation gives incentives to producers#p#分頁標題#e#
to improve products with less desirable nutritional profiles to avoid
unfavourable comparison by nutrition-conscious consumers. By disclosing
nutrition composition, mandatory labelling intensifies competition on
nutritional quality. Even if only a segment of the population uses nutrition
labelling, this may be sufficient to induce product changes by manufacturers
who wish to increase market share by selling to information-seeking
consumers (Burke et al., 1997). Moorman (1996) reports that the post-NLEA
labels are narrowing comprehension differences among consumers. Kristal et
al. (1998) find an increase in label usage and comprehension, although the
elderly and less educated still find them too complicated.
The NLEA regulations disqualify food products from making any health
claim, if the amounts of specific nutrients, such as cholesterol and sodium, exceed
some predetermined levels. Ippolito andMathios (1993) argue that the post-NLEA
regulated nutrition claims eliminate the potential for many types of truthful
declarations. The authors argue that consumers may improve their diets not only
by inter-category substitution (e.g. substitute fruits for desserts) but also by intracategory
substitution (e.g. substitute chicken for fat meats). It is suggested that
the NLEA disqualifies many products (e.g. over 97 percent of poultry and meat
products) from making any claims and leads to loss in consumer incentives to
switch to more healthful versions and brands of currently consumed products.
Similar conclusions are drawn by Mathios (1998), who finds that after
implementation of the NLEA, which banned health claims for product categories
exceeding a threshold value of fat, consumers shifted purchases towards
products of inferior nutrition quality. In contrast, the relaxation of the healthclaim
ban in 1985 is seen as a major contributory factor to the improvement of
consumption patterns over subsequent years (Mathios and Ippolito, 1998).
Implications and conclusions
For policy makers
In principle, according to economics of information as well as prospect theory,
increasing label utilisation requires increasing its benefits or decreasing its
European
Journal of
Marketing
35,5/6
716
costs (Russo et al., 1986). Benefit-enhancement strategies include mainly
consumer education. Cost-reduction approaches include facilitating
information acquisition and reducing comprehension effort by improving
nutrition knowledge and making nutrition information easier to use.
Variation in label utilisation and comprehension among people with different
identifiable individual characteristics suggests the potential of tailored nutrition
education programmes and materials focusing on the requirements of specific
consumer segments (e.g. Lin, 1995; Nagya et al., 1998). For example, the difficulty#p#分頁標題#e#
of older consumers in processing new complicated information highlights the
need for targeted campaigns as well as the use of additional media (e.g. Bender
and Derby, 1992). Similarly, the link between health awareness and label use
suggests that the latter can be increased by continuing to educate consumers
about important diet-disease relationships (Szykman et al., 1997).
Further, standardised labelling allows the implementation of programmes
aiming at the reduction of information-processing effort by illustrating the role
of listed nutrients and explaining the quantitative presentation of nutrient
content. In a similar vein, researchers have suggested that specific label
designs and standards may reduce consumer effort (Levy et al., 1996). For
example, a fixed spot for nutrition information on all packages might be
particularly helpful for elderly people (Cole and Gaeth, 1990).
Finally, truthful producer claims may increase consumer awareness and
interfirm competition on nutritional quality, leading to improved consumption
and production patterns (Mathios and Ippolito, 1998).
Strictly speaking, disclosure of food production methods, such as organic
farming, on package labels is not a nutrition information issue since production
methods are not directly and explicitly related to the nutritional properties of
the products. Nonetheless, it is worth noting that there is a growing movement
towards declaration of production methods, including, of course, genetic
engineering (see Bredahl et al., 1998 for a review).
For producers and marketers
Increasing consumer interest in health and nutrition places particular emphasis
on the nutritional properties of products as determinants of consumer demand.
In this respect, nutrition labelling provides a channel to demonstrate the
nutritional properties of products and an incentive to improve the nutrient
content of those that fail to meet consumer requirements in target markets.
Differential advantages can be achieved by providing overall superior
nutritional quality or exhibiting a unique nutritional profile.
The information provided in the label can also support communication
programmes based on health or, more generally, diet-related claims targeted to
certain consumer segments. For example, high values in ingredients such as
fibre and folic acid have helped cereal brands to become instantly recognised
and differentiated from the competition. Additional benefits may stem from
interactions of nutritional profile with other perceived attributes such as taste
and quality (Gallay, 1980).
Nutrition
labelling: issues
and policies
717
Marketers should also recognise that aggregate statistics of label use may
not be representative of their target markets because the clienteles of specific
products may respond to nutrition information in a different way. As#p#分頁標題#e#
mentioned above, the consumer heterogeneity detected by several empirical
studies suggests that marketers should carefully examine the behaviour of
their target market with regard to nutrition information acquisition and usage.
It is equally true, but less often noted, that consumers differ in preferences
for particular nutritional characteristics because of different health problems,
lifestyles, tastes, etc. This heterogeneity may provide the basis for market
segmentation and development of products with optimal nutritional profiles for
identified preference segments.
Finally, at least in the EU, legislation leaves a lot of room for creative design
of labels, which may not only enhance the character of the brand but also
improve consumer understanding. For example, Kellogg's has created vitamin
and mineral logos for selected brands, which depict the function of such
nutrients in the human body.
For researchers
A first pattern that appears to transcend the character of specific data is
consumer heterogeneity in label usage. Perhaps more importantly, the
literature suggests a clear correlation between label use and individual
characteristics, in particular education and age. The heterogeneous consumer
response to nutrition labelling may be perfectly normal and not necessarily
indicative of ineffective policy. Heterogeneity characterises all elements of
consumer behaviour such as price response, brand preferences and information
processing activities. In addition, nutritional composition does not influence
purchase behaviour in a deterministic fashion and may be less important than
other factors, such as price and taste.
One may also conclude from the foregoing that label utilisation tends to be
greater for less healthful categories, negative attributes, and ill or dieting
individuals. Thus aversion of negative outcomes appears as a main
contributory factor, suggesting cost-benefit analysis and prospect theory as
well-suited analytical frameworks.
It is worth emphasising that the effectiveness of European legislation has not
been assessed empirically, despite a plethora of US studies. In view of the
disappointedly small number of European studies, there is a clear need for further
research in European markets, which we hope the present paper will encourage.
For example, consumption data before and after the implementation of current
regulations by member states may shed light on the impact of legislation, reveal
cross-country differences, and suggest ways to improve current standards.
Finally, most studies have considered cognitive outcomes, using mainly
survey data, rather than behavioural outcomes, using experimental or
consumption data. There is considerable room for research on the effects of
nutrition information on consumer food choice and demand behaviour, given#p#分頁標題#e#
the increasing availability of individual demand data.
European
Journal of
Marketing
35,5/6
718
In particular, it would be desirable to design and implement empirical
studies in which nutrition characteristics are treated as utility-generating
attributes (Baltas and Doyle, 2000). Individual optimising behaviour,
differences in information acquisition and interpretation, and variation in
budgets and preferences would then determine the observed distribution of
demand for products with different nutritional properties. The effects on
observed choice behaviour provide the ultimate yardstick for evaluating
nutrition labelling policies and standards.
[ 25 ]
Nutrition label usage of Chinese consumers
Wai-sum Siu
Associate Professor, Hong Kong Baptist University, Kowloon, Hong Kong
Tina Man-yi Tsoi
Marketing Manager, Bo Bo Tea (International) Ltd, Kowloon, Hong Kong
Though the Pacific-rim countries
like China, Hong Kong
and ASEAN countries will
soon constitute an economic
power comparable to that of
North America and Europe,
most of the empirical work on
food labelling is related to the
US and UK consumers.
Research into nutrition label
usage in a Chinese sociocultural
context is not found.
To redress this imbalance,
this study attempts to appreciate
the use of nutrition
information labels by Chinese
consumers in Hong Kong. The
research results reveal that
the frequent label users are
label-oriented, nutritionprone,
advertising lovers, in
the age group 35-50, and
they read both English and
Chinese, whereas the nonfrequent
label users are
home-oriented and can read
only Chinese.
Introduction
Previously, food manufacturers emphasised
product features and attributes, for example
taste, price, package, appearance, and technology,
to market their food products.
Inevitably, no special effort had been put into
nutrition or nutrition information. Recently,
general publics have become aware of the
important relationship between nutrition
information and pre-packaged food. The
philosophy of “the right to know” has
strongly motivated consumers to make more
informed choices. Thus, food marketers use
information labels to provide customers with
additional information, for example ingredients.
Also, in response to the advocacy of the
social marketing scholars and the specific
needs of the health-conscious consumers,
food manufacturers and processors provide
much more information, including food
nutrition information, to the general public.
One of the major research thrusts in food
marketing in recent years has been food
labelling (Feldman and Erickson, 1987;#p#分頁標題#e#
Jacoby et al., 1977; Meidan and Edris, 1990).
Though the Pacific-rim countries like China,
Hong Kong and ASEAN countries will soon
constitute an economic power comparable to
that of North America and Europe (Kaynak
and Lee, 1989), most of the empirical work on
food labelling is related to US and UK consumers.
Research into nutrition label usage
in a Chinese socio-cultural context is not
found. To redress this imbalance, this study
attempts to appreciate the use of nutrition
information labels by Chinese consumers in
Hong Kong.
Previous work
Morris (1993) predicts that, as consumers are
aware of the benefits of maintaining a healthy
diet, nutrition labelling will become increasingly
important. Feldman and Erickson
(1987) point to the importance of food
labelling in product positioning, specifically,
for the “nutrition-conscious” market segment.
This is supported by Chan’s (1985)
study on US consumers which found that the
use of nutritional information and its importance
to purchase decision are positively
related. However, Meidan and Edris (1990)
reveal that the use of nutrition information
labels by UK consumers in food choice is
notably low. Davies and Wright (1994) suggest
marketers should use a labelling strategy to
inform consumers about the wholesomeness
of the product, reduce consumers’ anxiety
and search costs, and improve consumer
decisions in evaluating preferences in order
to retain a long-term consistent brand positioning.
Chinese consumers in Hong Kong have
become more health- and nutrition-conscious
(Hong Kong Trade Year Book, 1992). Recently,
the Hong Kong Government has amended the
Food and Drug (Composition and Labelling)
Regulations. The regulations require food
packers and manufacturers to label their
products in a prescribed, uniform and legible
manner. Specifically, the following information
should be marked in either English or
Chinese language or in both languages on the
label of the prepackaged food:
1 name of the food;
2 list of ingredients;
3 indication of “use by” or “best before” date;
4 statement of special conditions for storage
or instruction for use;
5 name and address of the manufacturer or
packer; and
6 count, weight or volume of food.
Most food manufacturers and packers only
provide the minimum information required
by the regulations, whereas the provision of
nutrition information is rare.
As of to date, there is no major study on the
use of nutrition information by Chinese consumers
in Hong Kong. The only available
work was undertaken by the Consumer Council
on the observation of the enforcement of#p#分頁標題#e#
the Food and Drug (Composition and Labeling)
Regulations (Choice, 1988). Kindle (1983)
postulates that Chinese social cultural values
influence Chinese not to switch brands,
specifically unfamiliar brands. Chan’s (1985)
empirical research into Chinese consumers
in Hong Kong suggests that the use of nutrition
information and its importance to food
British Food Journal
100/1 [1998] 25–29
© MCB University Press
[ISSN 0007-070X]
[ 26 ]
Wai-sum Siu and
Tina Man-yi Tsoi
Nutrition label usage of
Chinese consumers
British Food Journal
100/1 [1998] 25–29
choice are negatively related. However, it is
still not clear how and to what extent Chinese
consumers use nutrition information labels
in purchase decisions. Thus, this research
attempts to describe and identify the behaviour
of Chinese label users.
Research design
Research method
A multi-phase research approach was
adopted. First, two focus group interviews,
each with nine respondents, were conducted
in order to throw light on the awareness,
knowledge and use of nutrition information
labels in purchasing pre-packaged food.
Based on the focus group interview results
and insights obtained from previous studies
(including Chan, 1985; Feldman and Erickson,
1987; Meidan and Edris, 1990), a questionnaire
was designed to solicit information on
the behaviour of label usage. The questionnaire
consisted of four parts: a filter question,
40 psychographic items, nine demographic
questions and a question about the frequency
of using nutrition information labels. The
filter question was “Are you the main food
buyer for the household?” Thirty-six psychographic
items derived from Meidan and
Edris’s (1990) research and four additional
items on advertising, country of origin, word
of mouth and packaging, which were generated
from the focus group interviews, were
included in the questionnaire. A five-point
Likert scale (suggested by Yu et al. (1993) as
being the most appropriate scaling for attitudinal
research into Chinese consumer behaviour)
was used to determine the Chinese
respondents’ level of agreement and disagreement
to the items. Nine demographic questions,
including sex, age, personal income,
number of children, age of children, marital
status, education, employment and language
usually read, were included in the questionnaire.
A five-option question – “How often do
you use nutrition information labels in
selecting food items you usually buy?” – was
used to identify the user status of respondents.
Respondents who answered “Always”
or “Often” were classified as “Frequent#p#分頁標題#e#
users”, while those who responded “Sometimes”,
“Occasionally” and “Never” were
called “Non-frequent users”. The questionnaire
was translated into Chinese and then
back-translated to eliminate translation
errors and biases. The questionnaire was pretested
and minor amendments were accordingly
made.
Sampling plan
Following the sampling plan of a similar
study (Meidan and Edris, 1990), the quota
sampling method was adopted. Three districts,
namely Wan Chai (the high income
district), Tsuen Wan (the middle income
district) and North District (the low income
district), were selected. Seventy subjects aged
18 and over were chosen from each district.
This constituted a sample size of 210. Questionnaires
were administered outside retail
outlets of two supermarkets chains –
Hutchison Whampoa’s Park ‘N Shop and
Dairy Farm’s Wellcome. One retail outlet of
the two supermarket chains in each sample
district was selected randomly. These two
supermarket chains were selected because
they possessed the largest number of outlets
in Hong Kong and had attained a combined
market share of 70 per cent, among the 170
operators (South China Morning Post, 1994).
Every sixth shopper leaving the supermarket
was asked to answer the questionnaire by
personal interview method. Couples who
shopped together were counted as one unit.
Table I shows the demographic characteristics
of the respondents.
Results and findings
Factor analysis was adopted to analyse the
psychographic items (Kaiser-Meyer-Olkin
measure of sampling adequacy = 0.6534,
Bartlett test of sphericity = 2130.6784,
p < 0.01). As suggested by Aaker (1971), factors
with eigen value greater than one were
selected. Fourteen factors, which accounted
for 64.9 per cent of the total variance, were
identified. As suggested by Malhotra (1993),
the surrogate variable, or the variable with
the highest loading on each factor, was used
for further analysis. Thus, 14 surrogate variables
were identified. They were bargainhunter,
nutrition prone, label-oriented, advertising
lover, brand switcher, time saver,
tinned food user, brand loyalter, dieter, social
entertainer, weight watcher, home-oriented,
home avoider and health conscious. Details of
the surrogate factors are shown in Table II.
Stepwise Multiple Discriminant Analysis
in the SPSS for Windows 6.0 computer software
package was used to investigate the
differences between “Frequent users” or
“Non-frequent users” by 14 surrogate psychographic
items and nine demographic variables.
Each demographic variable was classified#p#分頁標題#e#
into dummy (0,1) variable because the
option under each question could not be considered
an interval scale. Each variable was
used in the stepwise discriminant analysis
using Wilks method and entry criteria for
p < 0.01 and exit criteria for p > 0.10. After five
iterations (See Table III), five variables,
namely label-oriented, nutrition-prone,
[ 27 ]
Wai-sum Siu and
Tina Man-yi Tsoi
Nutrition label usage of
Chinese consumers
British Food Journal
100/1 [1998] 25–29
advertising lover, age, home-oriented and
read only Chinese, were identified in the
discriminant model. The statistical results of
Table IV suggest that at the 1 per cent level of
significance (Wilk’s Lambda = 0.725816, Chisquare
value = 65.694, d.f. = 6, p value < 0.01)
the function discriminates significantly
between the two types of users. The hit ratio
on 30 per cent hold-out sample is 70.83 per
cent, indicating a good model fit as suggested
by Boyd et al. (1991). The standard canonical
discriminant function coefficients of the five
variables are: label-oriented (0.63297), nutrition-
prone (0.56114), advertising lover
(0.40024), age (0.34364), home-oriented
(–0.45195) and usually read only Chinese
(–0.47381). Nutrition prone, label-oriented,
advertising lover and age are positively
Table I
Demographic characteristics of the sample
Sex Male 31 (14.8%) Female 179 (85.2%)
Age 18-34 61 (29.0%) 35-54 114 (54.3%)
55-64 21 (10.0%) 65 and over 14 (6.7%)
Monthly Under HK$2,000 11 (5.2%) HK$2,000-9,999 42 (20.0%)
personal HK$10,000-17,499 67 (31.9%) HK$17,500-37,499 59 (28.1%)
income HK$37,500 and over 31 (14.8%)
Usual language Chinese only 97 (46.1%) English only 8 (3.8%)
read Chinese and English 96 (45.7%) Others 9 (4.4%)
Marital status Never married 37 (17.6%) Married 158 (75.2%)
Widowed 12 (5.7%) Divorced/separated 3 (1.4%)
Have children Yes 155 (73.8%) No 55 (26.2%)
Age of children Below 3 16 (7.6%) 3-5 19 (9.0%)
6-10 47 (22.4%) 11-17 62 (30.0%)
18 and over 65 (31.0%)
Education No schooling/ 11 (5.2%) Primary 55 (26.2%)
kindergarten
Lower secondary 44 (21.0%) Upper secondary 54 (25.7%)
Matriculation 9 (4.3%) Tertiary (non-degree) 11 (5.2%)
Tertiary (degree) 26 (12.4%)
Working status Full-time 91 (43.3%) Part-time 28 (13.3%)
Student 6 (2.9%) Retired 17 (8.1%)
Housewife 63 (30.0%) Others 5 (2.4%)
Table II
Surrogate psychographic variables
Surrogate factors Questions
Bargain-hunter In food shopping, you try to save money whenever you can
Nutrition prone You believe that the main value of food should be in its nourishment rather
than in its taste or flavour
Label-oriented You usually spend as much time as possible reading labels before you buy#p#分頁標題#e#
Advertising lover You would like to buy a brand because of attractive advertising
Brand switcher You change brands from time to time for the sake of variety and novelty
Time saver You dislike food shopping because it takes up too much of your time
Tinned food user You cannot get along without tinned food
Brand loyalter You continue purchasing a particular food brand because you are satisfied with it
Dieter You usually use packaged diet food at least one meal a day
Social entertainer You consider yourself a very social, outgoing person
Weight watcher You wish to lose weight quickly and improve appearance
Home-oriented You usually spend a lot of time in the kitchen
Home avoider You like to be free and to escape from the drudgery of the kitchen
Health conscious You dislike chemical ingredients
[ 28 ]
Wai-sum Siu and
Tina Man-yi Tsoi
Nutrition label usage of
Chinese consumers
British Food Journal
100/1 [1998] 25–29
related to the “Frequent users”, while homeoriented
and usually read only Chinese are
related to the “Non-frequent users”. The
“Frequent users” have higher group mean
values in nutrition prone (4.22951), labeloriented
(4.01639), advertising lover (3.09836)
and age (0.55738) than that of “Non-frequent
user”. On the other hand, the “Non-frequent
users” have higher group mean values in
home-oriented (3.09396) and read only
Chinese (0.53020), compared with the “Frequent
users”.
The results appear to suggest that the “Frequent
users” are “nutrition seekers” as they
believe that nutrition is more important than
taste. They are in the age group 35-54. They
tend to buy a particular brand because of
attractive advertising and are willing to
spend much more time on reading information
labels. In contrast, the “Non-frequent
users” are “home-oriented” because they
usually spend a lot of time in the kitchen.
They are of age either above 54 or between 18
to 35. They read only Chinese.
Discussions and implications
Mueller (1991) reveals that the US label users
are mostly senior citizens. The results of this
research, however, reveal that the Chinese
frequent nutrition label users are relatively
younger, at the age of 35 to 54. Also, the Chinese
frequent users are nutrition conscious,
outgoing and are more willing to purchase a
specific brand because of its attractive advertising.
Thus, food marketers should allocate
more funds to research and development in
order to increase the nutrition value of their
foods. They should produce high quality
advertisements with clear and strong messages
on good nutrition value.#p#分頁標題#e#
Petty and Cacioppo (1986) reveal that,
should manufacturers and distributors fail to
provide clearer, more standardized
Table III
Group mean values and standardized coefficients
Standardized canonical
Group mean value discriminant function
Questions User Non-user coefficients
Nutrition prone
You believe that the main
value of food should be in
its nourishment rather than
in its taste or flavour 4.22951 3.69799 0.56114
Label-oriented
You usually spend as much time
as possible reading labels before
you buy 4.01639 3.18792 0.63297
Advertising lover
You would like to buy a brand
because of attractive
advertising 3.09836 2.75839 0.40024
Home-oriented
You usually spend a lot of
time in the kitchen 2.65574 3.09396 –0.45195
Age
35-54 years old 0.55738 0.53691 0.34364
Language usually read
Chinese only 0.29508 0.53020 –0.47381
Table IV
Result of discriminant analysis
Canonical discriminant functions
Fcn 1*
Eigen value 0.3778
Pct of variance 100.0
Cum pct 100.0
Canonical corr 0.5236
After fcn :
Wilk’s Lambda 0.725816
Chi-square 65.694
df 6
Sig 0.0000
留學生dissertation網Note: *the one canonical discriminant function
remaining in the analysis
[ 29 ]
Wai-sum Siu and
Tina Man-yi Tsoi
Nutrition label usage of
Chinese consumers
British Food Journal
100/1 [1998] 25–29
information labels, consumers will rely on
simple peripheral cues, for example strong
visual labels based on association with prior
beliefs or inferences, or use source effects,
such as word of mouth, to make purchase
decisions. Davies and Wright (1994) also point
to the use of inserts on nutrition to arouse
awareness. However, food marketers in Hong
Kong are not active in providing nutrition
information to consumers. The findings of
this research also reveal that frequent label
users are willing to spend much time on reading
labels and they use both Chinese and
English. Thus, new market entrants with
unfamiliar brands may consider using nutrition
information labels as a means of achieving
a competitive edge over the well-known
brands. The nutrition information labels
should be in great detail and preferably in
both Chinese and English. Also, should
supermarkets, food stores and convenience
shops wish to target the “nutrition seekers”
and position themselves as customer-care
retailers, they should provide extra services
such as placing posters on recommended
daily allowances or nutrition information
near the food shelves.
Meidan and Edris (1990) suggest educating#p#分頁標題#e#
consumers on the importance of using nutrition
information labels. French and Barksdale
(1974) also point to the use of legislative
measures to increase consumers’ awareness
of nutrition information. It would be quite
likely that in the near future policy makers
would require food marketers to ensure full
information disclosure and provide details of
the nutrition contents of their products. This
may be strategically important and useful to
food marketers since some additives may not
have been previously considered beneficial or
harmful by consumers. Proactively, food
marketers may substitute the artificial additives
with natural ingredients and also convey
the nutrition information to consumers
by the non-numerical systems, for example
graphs, as suggested by Morris (1993).
Conclusions and recommendations
As consumers become more health-conscious
and policy makers impose tighter regulations
on food labelling, marketers should recognize
consumers’ specific need for food nutrition
information so as to maintain market competitiveness.
New market entrants might
make use of nutrition information labels to
divert consumers from relying on simple
peripheral cues or word of mouth. To the
Chinese consumers, both Chinese and English
nutrition information should be provided.
Moreover, an attractive marketing
communication blend about the nutrition
value of their food products should be
planned thoroughly and executed to inform
and persuade Chinese consumers.
References
Aaker, D.A. (1971), Multivariate Analysis in Marketing:
Theory and Application, Wadsworth
Publishing, Belmont, CA.
Boyd, H.W., Westfall, R. and Stasch, S.F. (1991),
Marketing Research: Text and Cases, Richard
D. Irwin, Homewood, IL, pp. 602-3.
Chan, T.S. (1985), “A cross-cultural comparison of
information processing in consumer food
choice”, Hong Kong Journal of Business Management,
Vol. 3, pp. 25-34.
Choice (1988), November, p. 22.
Davies, M.A.P. and Wright, L.T. (1994), “The
importance of labelling examined in food
marketing”, European Journal of Marketing,
Vol. 28 No. 2, pp. 57-67.
Feldman, L.P. and Erickson, L.W. (1987),
“Labelling as an element in food marketing
strategy: the United States experience”, Food
Marketing, Vol. 3 No. 2, pp. 3-18.
Food and Drugs (Composition and Labeling) Regulations
(1993), Chapter 132 Subsidiary Legislation,
July, The Government Printer.
French, W.A. and Barksdale, H.C. (1974), “Food
labeling regulations: efforts toward full disclosure”,
Journal of Marketing, Vol. 38, July,
pp. 14-9.
Hong Kong Trade Year Book 1992 (1992), The Hong
Kong Chinese Importers’ and Exporters’#p#分頁標題#e#
Association, Part 2, pp. 52-4.
Jacoby, J., Chestnut, R. and Silberman, W. (1977),
“Consumer use and comprehension of nutrition
information”, Journal of Consumer
Research, Vol. 4, September, pp. 119-28.
Kaynak, E. and Lee, K.H. (Eds) (1989), Global
Business – Asia-Pacific Dimensions, Routledge,
London.
Kindle, T.I. (1983), “A partial theory of Chinese
consumer behavior: marketing strategy
implications”, Hong Kong Journal of Business
Management, Vol. 1, pp. 97-109.
Malhotra, N.K. (1993), Marketing Research: An
Applied Orientation, Prentice-Hall, Englewood
Cliffs, NJ.
Meidan, A. and Edris, T.A. (1990). “Nutrition label
usage: behaviour in food choice decisions”,
British Food Journal, Vol. 92 No. 8, pp. 14-21.
Morris, E. (1993), “Nutrition-labelling developments”,
British Food Journal, Vol. 95 No. 5,
pp. 12-15.
Mueller W. (1991). “Who reads the label?”, American
Demographics, Vol. 13 No. 1, January,
pp. 36-40.
Petty, R.E. and Cacioppo, J.T. (1986), “The elaboration
likelihood model of persuasion”,
Advances in Experimental Social Psychology,
Vol. 19, pp. 123-205.
South China Morning Post (1994), 17 November.
Yu, J.H., Keown, C.F. and Jacobs, L.W. (1993),
http://www.400190.com/dissertation_writing/MBA/“Attitude scale methodology: cross-cultural
implications”, Journal of International Consumer
Marketing, Vol. 6 No. 2, pp. 45-63.