1
ANALYSIS
Determinants of Demand for Green Products:
An application to eco-label demand for fish in Europe
Dorothée Brécarda*, Boubaker Hlaimib, Sterenn Lucasa, Yves Perraudeaua, Frédéric Salladarréa
a
Université de Nantes, LEMNA, Institut d’Économie et de Management de Nantes – IAE, Chemin de la
Censive du Tertre, BP 52231, 44322 Nantes Cedex 3, France.
b
Université de Rennes, LEST, 1UT Saint Malo, Chemin de la Croix Désilles, 35400 Saint Malo, France.
ABSTRACT
In this paper, we confront the theoretical motivations of the consumption of eco-friendly products and
the factors influencing the Europeans perceptions regarding the fact that “the fishes caught with an
environmental friendly technique may carry a special label”. We take advantage of the recent integration
of non-economic elements in the microeconomic analysis of consumers’ behavior in order to highlight
the factors leading to their demand for green products. Thanks to an original European survey on seafood
product carried out on more than 5 000 consumers, we test the influence of intrinsic motivation,
information, localization and socio-economic factors on the demand for an eco-label for fish.
Our results show a significant connection between the willingness of eco-labeling and seafood
features, especially the fish freshness, the geographical origin of the fish and the wild versus farmed
origin of the fish. Moreover, we prove the major role played by the fish price. We also demonstrate that
the ecological issue regarding fisheries is well connected with consumer information, intrinsic motivation
and socio-economic status: the “green fish consumer” is a young woman, well educated, well informed
on the state of marine resource and not very trustful in the well regulation of the fisheries. Moreover,
consumers who are aware of the marine resource preservation have the same profile.
Keywords: Environmental preferences, eco-label, seafood.
*
Corresponding author. Tel: 33 2 40 14 17 35. Fax: 33 2 40 17 49. E-mail address: [email protected]
2
1. Introduction
The rise in consumers’ ecological consciousness over the past years has increased their willingness to
pay for green products (OECD, 2002b). OECD (2002a) points out that 27 % of consumers in OECD
countries can be labeled “green consumers” thanks to their high willingness-to-pay and high
environmental activism. 10 % of them are “green activists” with high environmental activism but lower
willingness-to-pay. The others are “latent greens” (40 %) or “inactive” (23 %). In its 2005 paper on the
effects of eco-labeling schemes, OECD compiles several studies revealing greater consumers willingnessto-pay for eco-labeled products. The 2008 Eurobarometer shows that 75 % of the Europeans are “ready to
buy environmentally friendly products even if they cost a little bit more.” However, only 17 % of them
declare having recently bought such “products marked with an environmental label.” A reason may be the
inability of 42 % of them to discriminate environmentally friendly products from other products even
with an eco-label. Another reason can arise from the fact that some of them think that a responsible
consumption is synonymous with a lower consumption, like 75% of French questioned by
Ademe/Ethicity (2008). The question of the determinants of demand for “green products” is then
particularly prevalent.
In a standard microeconomic approach, the willingness to pay more for a green product than for a
“brown” one reflects a higher consumer’s marginal utility when he buys a green product rather than a
brown one. It also reveals consumer’s environmental preferences. However, there are several factors to
take into account for depicting such preferences. First of all, individual decisional process can be
influenced by psychological, moral and cultural factors. Frey and Stutzer (2006) associate economic and
psychological approaches in order to study “environmental morale and motivation.” They argue that
individuals are driven by altruism, social norms and reciprocal fairness, internalized norms (related to
high principles inducing self-evaluations) and intrinsic motivation (i.e. the willingness to pursue an
activity for the welfare it induces in itself). Berglund and Matti (2006) add that individual decisions
depend on ethical values and beliefs, customs, culture and several kinds of social, political and moral
values, but also on institutional settings which are likely to shape such attitudes by encouraging or
discouraging some behaviors and attitudes. In the same way, Torgler and García-Vilañas (2007) show
that political interest and political awareness are major determinants of the Spanish’ attitude towards
preventing environmental damage. Individuals’ economic behavior regarding environmental issues is also
3
justified by their citizen values. The representation of consumers’ environmental preferences through
their utility functions should also reflect more or less finely these multiple non-economic determinants.
Other important factors explaining pro-environmental attitudes are individuals’ socio-economic
characteristics. Several studies, reviewed by Torgler and García-Vilañas (2007), point out differences in
preferences according to age, gender, marital status, occupational status, localization and especially
income and education. The consumer is indeed confronted to his budget constraint that may limit his
expenses, particularly in green products. The education level may also impact consumers’ attitude
through their knowledge of environmental issues and their eco-information treatment.
The role of information diffusion and absorption has been clearly underscored in studies dealing with
eco-labels. The eco-labels are indeed an instrument used by firms and governments in order to put
forward the better ecological quality of a given product with respect to the unlabeled goods. Since the
environmental consequences of the production and the consumption of a product are generally
unobservable, the eco-label is the only way for consumers to collect such information. Firms draw on
eco-labels to win market shares thanks to a differentiation strategy surfing the wave of consumers’
ecological awareness. The government’s goal is the improvement of the environment through a
substitution between green and brown products. The success of such a policy depends on the consumers’
reaction facing this environmental information. OECD (2005) argues that consumers often accept to pay
more for eco-labeled products and that the premium they accept to pay depends on their confidence in the
certifying organization, their levels of education, their environmental involvement, and the type of
additional information available. These results have been recently confirmed, for “greener” vehicle, by
Teisl et al. (2008) and, for “green” electricity, by Ek and Söderholm (2008) and Salmela and Varho
(2006). Moral and social norms play also an important role in the electricity choices (Kotchen and Moore,
2007, Wiser, 2007, Ek and Söderholm, 2008).
The impact of seafood products eco-labeling on consumers’ behavior is an important issue. The state
of the world fisheries and aquaculture is indeed dramatically worrying despite international regulation,
such as the conservation measures adopted for the European Common fisheries policies and by the
regional fisheries organizations or the FAO’s code of conduct for responsible fisheries. According to the
2007 FAO report, about half of the world fisheries are fully exploited, 17 % are over-exploited and 7 %
are depleted. Furthermore, fishing has the particularity to be a harvest economics, since the marine
4
resources remain renewable and they are harvested at a lower rate than they are naturally replenished. The
fish eco-labeling may contribute to reach a sustainable fish exploitation if producers change their fisheries
management and consumers turn towards the eco-friendly seafood. Several seafood eco-labels, recently
analyzed by Washington (2008) for the FAO, were developed by non-governmental organizations and
few retailers and seafood industry bodies. The Marine Stewardship Council (MSC) ecol-label for
fisheries, created in 1997 by WWF, is the most well-known and the largest one. While the MSC has
adjusted its criteria and procedures in light of the FAO guidelines, it remains criticized for not
incorporating the particular situation of developing countries (Washington, 2008) and for not including
the overall environmental impact of the life cycle of seafood products (Thrane et al., 2009). Thrane et al.
(2009) emphasize the wild-caught seafood products have not only a direct effect on the targeted fish
stock, but also on the overall marine eco-system (on other species, birds, seafloor, etc.) and the external
environment (particularly on global warming). They claim an expansion of criteria used by the MSC for
eco-labeling and recommend the inclusion of energy use and chemicals, such as the Swedish KRAV
label. Since 2005, the European Commission debates in order to adopt its own label guidelines and to
take into consideration other criteria than the mere ecological sustainability (Guillotreau et al., 2008).
Thus, neither sustainability criteria, nor even certification procedures make consensus yet. Furthermore,
the most popular label, the MSC, only covers less than 1% of global fish trade. This raises the question of
the influence of such a label on seafood consumption. Several papers deal with the issue of consumers’
reactions to seafood eco-labeling. Wessels et al. (1999) analyze individuals’ preferences between labeled
and unlabeled seafood products and underscore the importance of education, knowledge of and sensitivity
to environmental and marine resource issue that favor pro-label preferences. Teisl et al. (2002) estimate
the effects of dolphin-safe labels on the canned tuna consumption and show the beneficial impact of the
label introduction on the long-term demand. Once again, the consumers’ information seems to have
strongly influenced their attitudes.
Our paper intends to provide new insights into the consumers’ eco-label demand for seafood products.
Therefore, we assume that determinants of consumption of labeled seafood products are the same as those
of eco-labeling wishes. Accordingly, we analyze all the determinants of demand for green products set
out by microeconomics and economic psychology. We also infer socio-economic and psychological
5
factors explaining the eco-label request.1 Moreover, we attempt to compare the socio-economic
characteristics of consumers who declare to be pro-label and those who assert to pay attention to the
marine resources level, in order to determine if those consumers are the same.
In order to assess this theoretical and empirical statement, we undertake an econometric analysis of
the European consumer willingness for eco-label in the seafood sector. More precisely, we define the
green demand as the demand of “ fishes caught with an environmental friendly technique, and which may
carry a special label”. This study rests on an original European survey on seafood product carried out
with more than 5 000 consumers in Belgium, Denmark, France, Italy and the Netherlands (Perreaudeau et
al., 2008). Although the main focus of this survey is the image of the European fishing industry, several
questions deal with environmental information and concern as well as purchase criteria for seafood.
Combining the responses to these questions with the consumers’ socio-economic characteristics allow us
to carry out an analysis of the determinants of green seafood demand.
Our results show a significant connection between the acceptability of eco-labeling and other
parameters such as the product form, the geographical origin of fish and the wild versus farmed origin of
the fish. Moreover, the consumers who are in favor of an eco-labeling policy pay more attention to prices
when buying fish, consider rather that fishing is likely to reduce fish stock overtime and believe
somewhat that fisheries are not sufficiently regulated. Furthermore, the sociological profile of a green fish
consumer is a young and well-educated woman.
The remainder of the paper is structured as follows. In section 2, we analyze theoretical determinants
of demand for green products. In section 3, we introduce the database and the econometric method. In
section 4, we analyze our empirical results and compare them with theoretical predictions. Section 5
brings this paper to a conclusion.
2. Determinants of green demand
In this section, we provide a complete view of theoretical factors encouraging and discouraging green
demand. Our analysis rests on the assumption that consumption of green products and eco-labeling
request are two sorts of so-called ‘green demand’ and have thus the same determinants. These factors
may be classed into three categories: the consumers’ intrinsic motivation, preferences and constraints. We
1
But we are not able to provide an estimation of the willingness-to-pay for the eco-label with our qualitative data.
6
do not seek here to develop an original theoretical model encompassing all these factors, but rather take
advantage of the existing literature for an analysis as exhaustive as possible of green demand.
2.1. Consumers’ intrinsic motivation
Consuming a given product because of its perceived positive environmental impact (e.g buy a linecaught sea bass or a tuna with a Dolphin-Safe label) can be described as an ecological behavior. The
individual ecological behavior might be influenced either by morale motivation (e.g some psychological
and moral determinants) or by external motivation (e.g. ecological political constraints). As Berglund and
Matti (2006) have underlined, in general, economists assume that individuals’ actions are mainly driven
by external rewards. However, there is a growing body of evidence indicating that for civic oriented
behaviors, such as environmental responsible consumption, other motivations than the external ones may
play a comparable role. Although this idea is not consensual (Arrow, 1972), this premise of rethinking the
interaction between different types of motivation has recently gained acceptance within the field of
economic psychology (Brekke et al., 2003).
Frey and Stutzer (2006) draw on four models approaching the linkages between environmental moral
and motivation. The first framework is based on the pure altruism model and can lead to crowding
effects. This model implies that an individual takes into account two dimensions in his green demand: the
private preference for the green product and the benefit that his consumption brings to others. In this
context, if the other’s benefit is obtained by an alternative way than individual’s contributions, the
consumer can be led to decrease his own contribution: it’s the so-called crowding effect. In order to
soften this effect, which may seem not very realistic, Andreoni (1990) proposes a model of impure
altruism. In this framework, individual valorizes the satisfaction resulting from the other’s benefit: he gets
“warm glow” from her contribution to others’ welfare. The altruism is all the purer as the link between
consumption and pollution is weak. This relation may be spatial (for local pollution due to consumption
of polluting products) or temporal (for future pollution resulting from consumption). When pollution and
consumption are non-separable, the purchase of green product can be motivated by the concern about
health (as a substitution or complementariness of concern about the environment). This fact is especially
true for fish because this kind of products is generally associated with a healthy image (Lambert et al.,
1996).
7
The second and third Frey and Stutzer’s models deal with social and internalized norms or, in other
words, the socially shared beliefs concerning individual behavior. With social norms, the sanction comes
from the other members of the society, whereas with internalized norms, individual sanctions himself.
Benabou and Tirole (2006) show also the theoretical importance of the social reputation in the altruism
degree. The last model is the intrinsic motivation scheme, which concerns the achievement of an activity
for the welfare it induces in itself. For example, in the case of tuna purchase with the Dolphin-Safe label,
the intrinsic motivation corresponds to the satisfaction derived from the safety of dolphins. Therefore, the
green demand increases with the altruism degree or the warm-glow of consumers and with the
relationship between consumption and pollution.
Obviously, the internalized norms and the intrinsic motivation are connected. There also are
interactions between intrinsic and extrinsic motivations. The extrinsic motivation may indeed have a
negative impact on the intrinsic motivation (the crowding out effect). A monetary reward, for instance,
may reduce intrinsic motivation because it depreciates the action in itself. This mechanism is referred as
“the hidden cost of reward” (Deci and Ryan, 1985). Frey and Oberholzer-Gee (1997) propose an
economic analysis of the hidden cost with an application to the ‘Not In My Back Yard’ (NIMBY) theory.
This effect can also be observed in other extrinsic motivations than monetary ones. For example, in the
case of Dolphin-Safe label, if the government decides to subsidize the Dolphin Safety, an individual is
likely to decrease his own participation, since the dolphin safety is supported by another “channel”.
Likewise, we may expect that fisheries regulation leads to a crowding out effect on the individual
intrinsic motivation for eco-label on seafood products. Indeed, if fisheries regulation is perceived as
sufficient, individual intrinsic motivation could be lower. Consequently, the willingness for eco-labeling
could decrease. Nevertheless, if the regulation is not perceived as sufficient enough, the effect could be
inversed leading to an increase in intrinsic motivation. In general, we observe a motivational crowding
out effect only if the extrinsic incentive is perceived as either a restriction to self-determination or as
controlling. Conversely, if the extrinsic incentives are perceived as supportive, we may observe a
motivational crowding in effect.
Even though all these frameworks are not mutually exclusive, we focus our attention on the intrinsic
model. Indeed, the link between intrinsic and extrinsic motivation will allow us to study the former
8
through the impacts caused by the latter. In this sense, we study the role played by the perception of
government’s action over consumer’s motivation.
Some microeconomic models introduced such motivations in consumers’ preferences. They provide
an adequate framework for the analysis of the green demand and, beyond that, for the actors’ strategies
within a green market. We shall study these models in the next section.
2.2. Consumers’ preferences
The common microeconomic analysis of consumers’ choices rests on a utility function that translates
consumers’ preferences among different baskets of goods. An especially appropriate approach for our
purpose is the one of Lancaster (1971) who defines goods by their intrinsic characteristics. The utility
level of a consumer then depends on the level of each good’s characteristic and not on the consumption
level itself. The impact of a product on the environmental quality may also be considered as a particular
feature, namely the green characteristic. The consumer can view this attribute as a vertical, a horizontal or
a public characteristic.
In the vertical product differentiation models, the green quality of a good is all the more high since its
negative environmental impact is low. Extending the Mussa and Rosen’s model (1978), Cremer and
Thisse (1999) assume that consumers differ by their marginal willingness-to-pay for the green quality. In
such a model, all consumers prefer the less polluting product (i.e. the best quality) but differ in their
willingness to pay for it. Thus, the consumer’s ecological consciousness is partly related to his income.2
Nevertheless, income can limit the consumer's ability to purchase a green product, and, therefore we
should distinguish between revealed and stated preferences. If income limits the revelation of preferences
for lower-income consumers, it cannot be considered as a determinant of preferences in itself. Thereby,
the demand for green product rises with the number of consumers who are sensitive to environment
matter and especially their sensitivity degree.3
2
Tirole (1988) underlines that the marginal willingness-to-pay (or taste for quality) may be interpreted as the marginal
substitution rate between income and quality: the higher it is, the more the marginal utility of income is low and thus the income is
high.
3
For other applications of such a framework to environmental policies, see Brécard (1998), Lombardini-Riipinen (2005),
Mahenc (2008), Moraga-González and Padròn-Fumero (2002) and Poyogo-Theotoky and Teesuwannajac (2002).
9
In the horizontal differentiation models (d’Aspremont et al., 1979), each consumer has an ideal
variety that does not necessarily correspond to the most environmentally friendly one. The consumer’s
utility decreases with the distance between his ideal variety and the variety he really consumes. The
consumers’ taste heterogeneity leads to ideal variety heterogeneity. The green demand thus depends on
the disutility due to this distance. Without other assumption, nothing distinguishes thus a green variety
from another. In order to deal with the environment issue, the consumer’s intrinsic motivation assumption
should be implemented into the differentiation model. Eriksson (2004) and Conrad (2005) introduce the
warm glow induced by a contribution to a better environmental quality in the consumer’s utility function:
the less the product is polluting, the more the utility level is high. Hence, consumers partially internalize
the environmental externality when they buy products.
Another way to introduce environmental consciousness in consumers’ preferences is to consider the
green product as an impure public good that provides not only a common private characteristic but also a
public attribute: the environmental quality (Kotchen, 2005, 2006). For instance, the protection of
maritime resources can be assimilated to a public characteristic. Since the 20th century, the maritime
resources status is indeed gone from ‘res nullius’, belonging to nobody, to ‘res communis’, belonging to
everybody (Jagot and Perraudeau, 2006). The consumer’s utility is an increasing function of the private
characteristics, arising from the consumption of both conventional and green goods, and the public
characteristics, arising only from the consumption of green good. A rise in the consumers’ ecological
awareness, namely the degree of pollution internalization, stimulates demand for the green characteristic
when there is no substitute to the green product. However, if consumers have the possibility to make a
donation to an ecological association, which directly contributes to put forward the environmental quality,
an increase in the consumers’ ecological awareness may reduce the demand for environmental quality:
Whereas a higher environmental consciousness increases the green product demand, it reduces the
implicit price of its private characteristic that, in turn, decreases the demand for environmental quality
(that is a substitute to the private characteristic). We find here another form of the Frey’s crowding-out
effect (generated by consumer behavior herself rather than an external intervention). However, if the
green good is a complement to private consumption, the second effect is reversed, and an improvement of
ecological awareness is likely to increase the demand for environmental quality. The green demand is
10
also related not only to the environmental awareness but also to the existence of ecological organizations
that provide an alternative to the purchase of green product to act in favor of the environmental quality.
Nevertheless, some constraints can limit the purchase of green products. Let us summarize them in the
following sub-section.
2.3. Consumers’ constraints
Two main constraints carry weight in consumer purchase decisions: his income and his information
on products’ environmental characteristics.
The budget constraint plays an especially important role in the consumers’ choice between green
products and standard ones since green goods are often more expensive than their standard substitutes.
The higher prices of green products may be explained by the fact that they are generally more work
intensive, produced at a lesser scale and/or fashioned from more environmental friendly technologies. In
particular, the green fishes are more expensive than the ‘standard’ ones that are themselves relatively
expensive. For example, the average price of “standard” trawling caught sea bass sold by auction is 7.9
euros per kilogram in 2006 although the ‘green’ line caught sea bass price is 15.1 euros (OFIMER,
2008).4
Moreover, Mahenc (2008) shows that “when consumers cannot ascertain the environmental
performance of products, the price must be distorted upward to signal a clean product.” In the same field,
Teisl (2003) in a survey regarding the performance of labeling Environmentally certified forest Product,
finds that if two product shows similar environmental seals, consumers “assume that the environmental
characteristic of the higher priced product are better”. Consumers may thus prefer green products but
purchase less expensive standard ones because of their low income. This phenomenon is reinforced with
the growing competition of very low price substitutes. Conversely, the wealthiest consumers can more
easily buy their favorite products, which may be green or not.
Consumers are also confronted to incomplete information on environmental consequences of a
product from cradle to grave. First at all, environmental information of product life cycle is rarely put on
products. Thereby, consumers have to search, find and understand such an information. This may be a
4
The line caught sea bass can be considered as a green product because the technique used for its capture implies less
environmental damage than the ones used in other fishing.
11
long, costly and uncertain process. Even if consumers find some environmental information on a product,
they are not always able to interpret it. When they read the carbon footprint on products distributed by
Tesco in United Kingdom or Casino in France, for instance, do they know exactly the consequences of
these emissions for them and the whole society? When they see “responsible fishing” on seafood
products distributed by Carrefour or “selected products for a preserved ocean” on those distributed by
Casino in France do they know really the effects of these purchases on the marine resources? By
ignorance of environmental stakes or suspicion toward green label, consumers may then turn away from
green products. However, thanks to an official eco-label, as those of the Marine Stewardship Council or
the Flower in Europe, consumers know that the labeled product is kinder for the environment than the
others during all its life cycle. The eco-label may also help to reveal consumers’ environmental
preference. Conversely, the demand for green products may induce demand for eco-label by consumers
who are sensitive to ecological questions and anxious to buy more environmentally friendly products.
The previous analysis draws on different theoretical fields that underscore the central determinant of
the green demand: the ecological consciousness which, on one hand, is explained by a certain degree of
altruism and, on another hand, results in a willingness to pay more for a green product than for a standard
one, and both economic and informational constraints. In the following sections, we take advantage of the
European seafood survey to test the theoretical determinant of a particular green demand: the demand for
a “special label” for “fish caught with techniques that are respectful of the environment.”
3. The database and the econometric models
3.1. Data
The data used for this model come from a survey carried out in 2007 for Europêche ETF (Perraudeau
et al., 2008). The main purpose of the survey is to deal with the public image of the European fishing
industry. This survey was conducted in five European countries by Europêche-ETF (Belgium), CESVIP
(Italy), Fisheries Circle (Denmark), SNV-PVIS (the Netherlands) and the University of Nantes (France).
The database included 5 000 questionnaires completed by face-to-face interviews, out of which 4 748
were finally usable (847 in Denmark, 849 in Belgium, 1 110 in Italy, 1 030 in France and 912 in the
Netherlands, randomly selected with an error margin of 3-4%). The socio-characteristics of questioned
Europeans are given is Table A1 in appendix.
12
The attitudes and perceptions of European citizens towards the fishery sector are measured. The
questionnaire offers more than 50 questions about how the fishing industry is perceived by the European
citizens as well as the degree of knowledge and awareness about the seafood industry and the fisheries in
general. Environmental issues are one of the main modules of this survey.
The European fishing includes various trades according to used fishing gears and ship sizes5. The five
countries considered well cover this diversity: Denmark, Holland and Belgium have essentially industrial
or semi-industrial flotilla, which operate in the Baltic Sea and North Sea. France has an intermediate
position, with three maritime frontiers (the English Channel-North Sea, Atlantic Ocean and the
Mediterranean Sea), a diversified flotilla dominated by artisanal fleet (12-25 meters). Italy illustrates
Mediterranean countries, with a little fishing fleet, composed by vessels of less than 12-15 meters.
This survey is especially well suited to perform a comparative analysis of individual situations with
respect to green fish demand. The survey itself does not deal specifically with this issue, but with the
more general public perception of the European fishing industry. Among many others, one of the
questions was addressed in the following terms: “The fish caught with techniques that are environmentfriendly should be stamped with a specific label (do you strongly agree, agree, don’t agree nor disagree,
disagree or strongly disagree with the proposal?)”. 82 % of respondents agree upon labeling fish caught
in a sustainable way. The Table A2 in appendix highlights socioeconomic and purchase criteria of
interviewed consumers. Regarding the sociological features of pro-eco-label, they are rather young,
female, living in urban areas, well educated and earn less than 2000 euros per month. People who think
that fish caught with environment-friendly techniques should be stamped with a specific label recurrently
consider that the quantity of fish is not stable. Similarly, they rather consider that fisheries are not well
regulated. The consumers in favor of an eco-labeling policy declare to pay more attention to the fish
resources level, the product visual aspects, the fish origin, prices, the product form and to the freshness
when they buy fish. According to the citizens’ points of view regarding eco-label, countries can be
ranked: whereas Belgium and France appear as the countries where people declare to be more favorable
5
We can compare neither a 12-metre trawler which sails out in the day with a 65-metre trawler which leaves for two month
campaign, nor a 22-25-metre trawler with a same size fileyeur: periods at sea, researched species, working conditions,… are
different. Various typologies may thus be proposed in order to assess fishing activities.
13
to fish eco-labeling, Denmark and the Netherlands are in the opposite case. Italy has an intermediate
position.
Face-to-face interviews may induce several biases.6 Firstly, as our analysis is based on face-to-face
interviews, this type of data collection is prone to social desirability biases, which describe the tendency
of respondents to reply in a manner that will be viewed favorably by others. Although the respondents
were not invited in the present survey to sort out the different criteria used for their fish purchasing
behaviour, the proportion of respondents that have chosen a given item along the ordered scale of
possible answers shows that the quality characteristics at the top of their priorities whereas the
environmental criterion is cited by the lowest proportion of the population. Surprisingly, this result rather
contradicts the tendency of respondents to reply in a manner that will be viewed favorably by others. At a
global level, we consider that the SDB is limited. Secondly, a potential ‘laziness’ in the answers to the
‘agreement’ question when similar questions are presented sequentially may also induce a bias.
Accordingly the potential bias induced by the laziness in the response to the following variables (pay
attention to the product form, origin, product form, price, resource level, and wild/farmed) has been tested
with a correlation matrix. Five correlation matrixes have been established for each modality (strongly
disagree, disagree, don’t agree nor disagree, agree, strongly agree) of the variables. For each correlation
matrix, the correlation coefficients greatly differ: this supposes that respondents who have scored one
modality for one question does not report the same modality for another one. Then the laziness bias
appears to be low.
3.2. Theoretical Models
The issue of eco-labeling is analyzed through an ordered Probit model. The comparison between the
determinants of the eco-label willingness and those of the concern about the marine resources level is
carried using a bivariate ordered Probit model. In this section, only this second model will be presented in
detail, the first one being standard.
On the first hand, the issue of eco-labeling is analyzed through an ordered Probit model. The ecolabel issue is linked to other explanatory variables related to the consumption criteria commonly used in
the literature (freshness, visual quality, price, origin, product form…) and other socio-economic features
6
We thank an anonymous referee for insightful remarks on these biases.
14
underlined in the second section of this paper. In the European seafood survey, the fact that questioned
people plead for an eco-label for fish may also be related to the following explicative variables: the
demand for environmental quality (carefulness regarding the species exploitation), their degree of
personal relationship with the marine resource problem that may be inversely proportional to their
altruism degree (habitation, proximity to the coast and coast frequentation), their carefulness of the fish
characteristics or quality (origin, freshness, nature, visual quality, wild/farmed) and the fish price.
Similarly to label, these variables are ordered through a Likert scale, which has been reduced into three
categories.7 Several control variables are used such as their information on marine resource problems
(fishing impact, pollution consequences, climate change effects …) and their socio-economic
characteristic (gender, age, family status, professional situation, localization of habitation and sea
frequentation). Unfortunately, we have no information about respondents’ education level, whereas most
studies underline that it is an important variable. However, the respondents’ professional situation is a
good proxy of their education level (Desrosières and Thévenot, 1988).8 Furthermore, country dummies
are added in the specification.
On the second hand, the probability of accepting a fish eco-label and that of paying attention to the
resources level are analyzed conjointly through a bivariate ordered Probit model. We seek to estimate the
joint probability distribution of these two ordered categorical variables. A bivariate ordered Probit
models is an extension of a standard bivariate Probit model when the number of dependant categories is
more than two.
Similarly to the univariate ordered probability models, bivariate ordered probability models can be
drawn from a latent variable model. Let us assume that two latent variables
and
are determined
by:
(1)
(2)
7
This aggregation has no effect on the results, but offers a more reduced presentation.
8
The correlation between the education level and the occupation is demonstrated in several economic studies (Solon, 2002;
Fershtman and al., 1996). This allows us to take into account the dynamic evolution of the educational level through the career
progress.
15
where
and
are vectors of unknown parameters,
and
are the error terms, and subscript i
denotes an individual observation. The explanatory variables in the model satisfy the conditions of
exogeneity such that
and
. The threshold levels can be expressed as:9
(3)
with
and
. Considering
cases separately
and
, in order to avoid handling the boundary
are defined. If
standard normal with correlation ρ, then the probability
and
and
are distributed as bivariate
can be written:
(4)
where
is the bivariate standard normal cumulative distribution function. The model is estimated with
a maximum likelihood function (for further details, see Sajaia, 2008).
Several variables are used in this bivariate model: the subjective informational level on marine
resource problems (the perceptions regarding fishing activities and fishing regulation strictness) and their
socio-economic characteristics (gender, age, family status, professional situation, localization of
habitation, and sea frequentation). Thanks to country-specific dummies, we attempt to measure the
country specific effect once individual characteristics are controlled for.
4. Econometric results
The ordered Probit results presented in Table 1 show how consumers’ purchase criteria for fish
interact with their willingness to see an eco-label stamped on “green fish”. The results of the bivariate
ordered Probit model presented in Table 2 allow us to compare similitude and importance of the
determinants of probabilities to demand a fish eco-label and to pay attention to the marine resource level.
9
y1i has j modalities and
y 2i
has k modalities. y ni is equal to 0 when the individual strongly disagrees with the proposal, 1
when she disagrees, etc.
€
€
€
16
4.1. Eco-label demand
— Please insert Table 1 —
Table 1 displays the influence of fish characteristics on the eco-label demand.10 The most important
attribute affecting the probability of an eco-label demand is “to pay attention to the product form”.11 This
relation is probably due to the particular feature of seafood product for which the freshness is a key
consumption criterion, synonymous with quality. Moreover, frozen and fresh fishes can be considered as
two distinct products with different characteristics: whereas frozen fish does not substantially affect the
gross composition and nutritional properties of fish12, the fresh fish consumption is more frequently
associated with healthful eating habits and may have health benefits. Indeed, seafood products contain
protein and other nutrients such as vitamin D and selenium, omega-3 fatty acid, few saturated fat and few
calories. Despite frozen fish eco-labeling is justified for traceability reasons (fishing place, fishing
technique, type of boat, date of fishing), consumers tend to prioritize eco-certification on fresh fish.
Taking into account the fish origin is the second most important feature. Individuals who agree with
this statement have a higher willingness to discriminate the environmental content of fish products. This
can be attributed to certain preference given to local production because of well-known fishing areas and
legal framework. As illustrated by the case of the Nile Perch in the Darwin’s nightmare film (Sauper,
2004), European consumers may be reluctant to consume products originated from countries where
regulations are perceived as insufficient. An eco-label can then enhance the confidence of citizens on the
consumed products. Furthermore, environmental values are likely to play a role in the importance
attached to the fish origin since the latter is associated with transportation costs and potential detrimental
environmental consequences.
10
Guillotreau et al. (2008) also use the survey of Perraudeau et al. (2008) in order to study the fish eco-labeling policy in
Europe in the light of its past evolution and the fish consumers’ criteria.
11
Several Wald tests of equality of coefficients have been performed in order to allow us to do comparative statements about
the parameters. Furthermore, the comparative statements are contingent upon the average changes in probabilities presented in
Table 1: as our variables are qualitative, the discrete change measured by the average change in probabilities is the difference in the
predicted value as one independent variable changes values while all other are held at their means. By this way, we can interpret and
compare probabilities (see Long and Freese, 2006).
12
The amount of change depends on temperature and time of storage and method of packaging.
17
Paying attention to the visual aspect of seafood products13 is another factor, which contributes to
increase the probability of an eco-label demand. The visual aspect indeed offers the possibility to be
ensured that the safety of seafood is properly preserved. This component is consistent when the product is
fresh but not for frozen seafood.
Among the purchase criteria, wild versus farmed origin of the fish is the last analyzed factor linked to
consumers’ willingness for a seafood eco-label. Farmed fish provides between one fifth and nearly a third
of all consumed seafood (the proportion is lower in northern countries and particularly higher for France).
This is likely to ease the strain on over-fished species, but also to cause environmental damage according
to the type of farm, especially because of the use of antibiotics in aquaculture.14 Then, the low impact of
this aspect on the eco-labeling probability may arise from the fact that consumers have insufficient
information on the fish production process.
In short, the more a consumer is attentive of purchase criteria, the more he will be favorable to the
eco-label. These results are consistent with previous analysis dealing with seafood consumption (Wessels
et al. 1999, Jaffry et al., 2004) and with other food labeling (Bernués and al., 2003).
Table 1 also highlights the role played by the fish price and by environmental awareness, through the
attention paid to the resource level. Both variables have a positive impact on the willingness to stamp an
eco-label on “green fish”. However, this effect is smaller than that of the fish form. The weak effect of
the fish price can be explained by the fish dearness, its healthy image, the consumers’ perception of the
scarcity of some species and their belief that a higher price signals a higher environmental quality. As
noted by Guillotreau et al. (2008), the limited effect of the heed paid to marine resource preservation
confirms the peculiar informational status of fishery products as credence goods: Even free, the
information is not spontaneously collected by consumers who pay little attention to this criterion when
they buy fish.
13
Among the main visual characteristics traditionally evoked by consumers, whole fish and filets should have firm, bright red
gills free from slime and shiny flesh. Beyond visual aspects, fish should smell fresh, not sour and fishy.
14
The purpose is to reduce disease and promote faster growth.
18
4.2. Comparison between the determinants of the eco-label demand and the
attention paid to the marine resource level.
The results of the bivariate ordered Probit model, presented in Table 2, illustrate more specifically the
real role of theoretical determinants of “green” demand and environmental concern in the seafood sector.
— Please insert Table 2 —
Individuals’ feeling about fisheries regulation enables us to understand their intrinsic motivation. An
efficient fishery regulation may indeed reduce the intrinsic motivation of consumers because of the
“motivational crowding effect” underscored by Frey and Stutzer (2006). Our results show that individuals
thinking that the fisheries are at least acceptably regulated are less likely to claim a seafood eco-label and
to pay attention to the resources level. Such individuals alleviate their own responsibility in the marine
resource issue and thus judge less necessary to stamp an eco-label on green fish and to be more attentive
to the biomass protection. These results may however be carefully interpreted because they depend on the
consumers’ perception of the regulation efficiency and not on the real regulation efficiency. Our survey
reveals that near 80 % of European respondents ignore the existence of the Common Fisheries Policy
(CFP)15. Nonetheless, a large majority reports that fisheries activities are regulated at the national (74%),
European (67%) and international (88%) levels.
The information on the state of the fish stocks plays an important role: Individuals considering marine
resources as stable are less reluctant to demand an eco-label and to be attentive to resource level. The
extent of these effects is quite similar when individuals neither agree nor disagree with this issue, but
appreciably different when they agree: the effect of the beliefs about fish stock stability is about twice as
large for the eco-label willingness as for the resource level sensitivity. These results confirm that the lack
of information may lead consumers to reject green products because of their underestimation of the
environmental consequences of their purchases.
The more an individual frequents the seaside, the more likely he claims a green fish label but,
surprisingly, the less he pays attention to the resource level. It is worth noting that the altruism degree of
individuals frequenting the seaside only few days per year may be higher to feel concerned with the
15
Adopted in 1983 with the objective of ensuring that declining fish stocks are exploited responsibly, protecting the
environment and the interest of the fishing industry and consumers, the CFP notably sets quotas for which member states are
allowed to catch what amounts of each type of fish. The CFP also includes a market organisation for the control of prices, marketing
arrangement and external trade policy (Perraudeau, 2008).
19
maritime environmental damage. However, environmental issues affecting fisheries are tricky to perceive,
even if seaside frequentation is regular. This can explain the divergent results for the effect of “seaside
frequentation”: only frequent seaside stays positively affect the probability to demand a seafood ecolabel, whereas only rare seaside stays negatively affect the likelihood of paying attention to the marine
biomass. In addition, we find no significance for the type of habitation (urban or rural), probably because
this distinction is becoming irrelevant today with the process of “rurbanization”.
The socio-economic characteristics carry for individuals’ opinion of fish eco-labeling and resource
preservation. In line with previous studies (Wessels et al., 1999, Whitmarsh et al., 2006), Table 2 shows
that they are rather young, female, and well educated. More precisely, the oldest individuals are less
likely to be pro-label than the youth. However, the age has no significantly impact on attention paid to the
resource level. Our results also support the idea of higher sensitivity of women to environmental issues,
the eco-labeling as well as the resource preservation. In addition, estimates confirm findings on other
goods (Blend and Ravenswaay, 1999, Zelezny et al., 2000, Arora and Cason, 1999 and Stern et al.,
1993). Finally, the marital status does not significantly affect the eco-label demand nor the attention paid
to the resource level. This result is confirmed by Torgler and Garcia-Valiñas (2007).
The professional situation affects individuals’ preferences for an eco-label and attention paid to the
fish stock level. Let us recall that the professional situation is the mere available variable in the survey to
measure the respondents’ education level. We also expect that this variable play a significant role in
individuals’ tastes. Individuals with intellectual profession are more inclined to want an eco-label than
farmers and workers. Students are the least attentive to marine resources’ safeguard, whereas workers and
unemployed consider more attentively this issue. Surprisingly, our estimation provides no support for
significance of individual’s income in both issues.16 This can be explained by the fact that the
questionnaire doesn’t beg the question of the willingness to pay for the environment. Because there is no
monetary engagement, individuals don’t take their income into account when they respond to the
interview.
— Please insert Figure 1 —
16
The income is not included in the estimations presented because it is correlated with the professional classification used in the
estimation (as this classification takes into account the hierarchical situation in work paid).
20
Finally, the analysis of the country fixed-effects is delicate to carry out.17 These fixed effects may
reflect the institutional, cultural and maritime practice differences across European countries. In order to
look more closely into these effects, a score ranging between 1 and 5 has been assigned to each country:
accordingly, countries with higher scores are those where the willingness for an eco-label are higher, or
those where the issue of marine resources is considered with more attention. The results are illustrated in
Figure 1. Except for France, national rankings regarding both issues are different: for instance, Belgium is
the most favorable to a seafood eco-label, but the second for the attention paid to the fish stock level;
Conversely, Denmark is the most attentive to the marine resource but the least favorable to the seafood
eco-label. Our survey reveals that subjective views of respondents about the fishing practices are quite in
line with the objective way of fishing18. This North/South dichotomy may explain the importance
attached by Northern respondents (the Netherlands and Denmark) to the level of resources. Due to the
Northern industrial fishing practices, the fish is often frozen and transformed on board, the eco-labeling
demand is then lower (see Table 1). In the Southern countries, the traditional fishing practices may reduce
the importance attached to the resource level issues but increase the eco-labeling demand. Furthermore,
the variety of fish and seafood products consumed is lower in Northern countries compared to Southern
countries. The greater consumption of fresh fish in Southern European countries ones can partly explain
these results19, but a more thorough analysis may involve a specific study of the fish sector in each
country. This research is however beyond the scope of our article.
17
Several studies emphasize that people may differ in the answers they give on a subjective question and especially on its scale
(see for instance Groot (2000) on subjective health measures and Kristensen and Johansson (2008) on job satisfaction). In our case,
these judgement effects may result from social status or national context and from habituation to objective fish consumption. This
problem of comparability can be understood in terms of response category cut-point shifts across populations. To overcome this
potential problem of comparability, we have estimate a generalized ordered Probit, which permits us to allow some regression
coefficients to vary across values of the dependent variable. The results, dressed in Table A3, show that the coefficients and their
significance levels are very similar to those found in the ordered Probit model .
18
To the question “when we speak of fishing boats, what comes into your mind?”, about 55 % of Italian respondents answer
‘small boat’ or ‘fishing launch’ (less than 12 meters), while this proportion falls to 8 % in the Danish case.
19
Let us note that the average consumption is quite similar among the studied countries except for France (29.9 kilograms per
inhabitants and per year for 2000): it is comprised between 20.4 for Belgium and 24.6 kilograms per inhabitants and per year for
Italy.
21
5. Conclusion
What are the main determinants of the demand for green products? The answer this question is
particularly important since we want to change our modes of consumption and production in order to
inflict less pressure on our natural environment. Recent development in behavioral economics and
microeconomics emphasize the theoretical determinants of green product consumption: intrinsic
motivation due to altruism, social norms, the desire for environmental public good, education, economic
constraints as income and relatively higher price of green products compared to brown products.
Could we identify these determinants in the European demand for a seafood eco-label? In order to
assess our theoretical framework, we undertook an econometric analysis of the European consumers’
willingness for an eco-label in the seafood sector, thanks to an original data set carried out for five
European countries.
Our results show a significant connection between the willingness of eco-labeling and the fish
freshness, its geographical origin and the wild versus farmed origin of the fish. Moreover, estimates
confirm the significant relationship between the eco-labeling and the price of fish: The consumers who
are in favor of an eco-labeling policy also pay more attention to prices when buying fish. This result is
surprising, as the sociological profile of the pro-eco-label corresponds to an individual from upper-social
classes, because of their much lower price sensitivity levels.
Two statements confirm that the ecological issue regarding fisheries is well connected with consumer
information, intrinsic motivation and socio-economic status. In our study, the belief that “the quantity of
fish in the sea is stable” outlines consumer misinformation. Moreover, the conviction that “the fisheries
are well regulated” is likely to crowd out the intrinsic motivation for a seafood eco-label. We show that
both assertion are negatively linked (for all the categories) with the eco-label demand. This means that
consumers who consider that “green fish” should be labeled also disagree with the idea of fish stock
stability over time, since they consider that fisheries are not sufficiently regulated. Furthermore, our
results enabled us to shape the sociological profile of a “green fish consumer”: she is young women and
well educated. In addition, this general profile is the same as that of consumers who are sensitive to the
marine resource preservation. Accordingly, the environmental aware consumers, who pay attention to the
marine resources, generally would like a seafood eco-label.
22
Surprisingly, the country effect on the probability of accepting a fish eco-label is tricky to understand.
The countries with the highest level of eco-labeling acceptability are respectively Belgium and France.
Conversely, this acceptability is lower in the Netherlands and especially in Denmark. However, northern
countries and France are the most concerned with the marine resource level. It is thus not obvious to link
the countries’ seafood eco-label willingness and their environmental consciousness.
To conclude, it seems that there is considerable potential for green consumption to develop, but that
several issues should be resolved. First, one has to study how interactions between the major actors (the
fisherman, the consumer and the policy-maker) can affect the green demand. Second, the relevant
features of green products and the ecological and responsible behaviors should be better understood by
including other controls such as the type of conservation, packaging, the level of income and the other
potential substitutable goods. Finally, the relevance of different constraints and barriers probably met by
the green consumer should be put forward. The knowledge of specific factors that encourage
environmentally friendly behaviors is essential to set up public policy to favor sustainable consumption.
Acknowledgements
We are very grateful to two anonymous referees for valuable suggestions and insightful advice.
Needless to say, we are solely responsible for any remaining errors or misinterpretations.
References
Ademe/Ethicity, 2008. Les Français et le développement durable : un nouveau palier à franchir. Available
on www.ethicity.net.
Andreoni, J., 1990. Impure Altruism and Donations to public-Goods – a theory of Warm-Glow Giving.
Economics Journal 100 (401), 464-477.
Arora, S. and Cason, T. N., 1999. Do Community Characteristics Influence Environmental Outcomes?
Evidence From The Toxics Release Inventory . Southern Economic Journal, 65 (4), 691-716.
Arrow, J. K., 1972. Problems in Resource Allocation in United States Medical Care. in Kunz and Fehr,
editors, Challenge of Life.
Benabou, R. and Tirol,e J., 2006. Incentives and Prosocial Behavior. American Economic Review, 96 (5),
1652-1678.
23
Berglung, C. and Matti, S., 2006. Citizen and consumer: the dual role of individuals in environmental
policy. Environmental Politics, 15 (4), 550-571.
Bernues, A., Olaizola, A. and Corcoran, K., 2003. Labelling Information Demanded By European
Consumers And Relationships With Purchasing Motives, Quality and Safety Meat. Meat Science 65
(3), 1095-1106.
Blend, J. R. and Van Ravenswaay, E. O., 1999. Measuring Consumer Demand For Ecolabeled Apples.
American Journal of Agricultural Economics, 81 (5), 1072-1077
Brécard, D., 1998. Incitation à innover et taxe environnementale : le rôle de la conscience écologique.
Recherches Economiques de Louvain, 64 (4). 447-466.
Brekke, K. A., Kverndokk, S. and K. Nyborg, 2003. An Economic Model of Moral Motivation. Journal
of Public Economics, 87, 1967-1983.
Conrad, K., 2005. Price competition and product differentiation when consumers care for the
environment. Environmental and Resource Economics, 31, 1-19.
Cremer, H. and Thisse, J.-F., 1999. On the taxation of polluting products in a differentiated industry.
European Economic Review, 43, 575-594.
d’Aspremont, C., Gabszewicz, J. and Thisse J. F., 1979. On Hotelling's Stability in Competition.
Econometrica, 47, 1145-1151.
Deci, E. L. and Ryan R. M., 1985. Intrinsic Motivation and Self-Determination in Human Behavior. New
York, Plenum Press, 371 p.
Desrosières, A. and Thévenot, L., 1988, Les catégories socioprofessionnelles, Paris, La découverte
Ek, K. and Söderholm, P. (2008), Norms and economic motivation in the Swedish green electricity
market. Ecological Economics, 68 (1-2), 169-182.
Eriksson, C., 2004. Can green consumerism replace environmental regulation ? A differenciated-products
exemple. Resource and Energy Economics, 26, 281-293.
European Commission, 2008. Attitudes of Europeans citizens towards the environment. Eurobarometer
295.
FAO, 2007. The state of world marine fishery resources, Rome.
Fershtman, C., Murphy, K.M. and Y. Weiss 1996. Social Status, Education and Growth. Journal of
Political Economy, 104, 108-132.
24
Frey, B. S. and Oberholzer, F., 1997. The Cost of Price Incentives: An empirical Analysis of Motivation
Crowding-Out. American Economic Review, 87 (4), 746-755.
Frey B., 1999. Morality and Rationality in Environmental Policy. Journal of Consumer Policy, 22, 395417.
Frey B. S. and Stutzer, A., 2006. Environmental Morale and Motivation. in Alan Lewis (Editors), The
Cambridge Handbook of Psychology and Economic Behavior, Cambridge, UK, Cambridge
University Press.
Groot, W., 2000. Adaptation and Scale of Reference Bias in Self-Assessment of Quality of Life. Journal
of Health Economics, 19, 403-420
Guillotreau, P., Monfort, M.-C., Perraudeau, Y., Salladarré, F., 2008. The seafood eco-labeling
experience in Europe: a new “market for lemons”? Lessons from a European survey. Paper presented
at the International Institute of Fisheries Economics and Trade conference, Vietnam, 22-25 July 2008.
Jaffry, S., Pickering, H., Ghulam, Y., Whitmarsh D. and P. Wattage, 2004. Consumer choices for quality
and sustainability labelled seafood products in the UK, Food Policy, 29, 215-28.
Jagot, L. and Perraudeau Y., 2006. De la gestion des ressources marines européennes à la gestion des
pêches maritimes. in Chaussade J. and Guillaume J. (editors), Pêche et aquaculture, pour une
exploitation durable des ressources vivantes de la mer et du littoral , Presses Universitaires de Rennes,
559 p.
Kotchen, M. J., 2005. Impure public goods and the comparative statics of environmentally friendly
consumption. Journal of Environmental Economics and Management, 49, 281-300.
Kotchen, M. J., 2006. Green Markets and Private Provision of Public Goods. Journal of Public Economy,
114 (4), 816-834.
Kotchen, M. J. and Moore M. R., 2007. Private provision of environmental public goods: Household
participation in green-electricity programs. Journal of Environmental Economics and Management,
53,1-16.
Kristensen, N. and Johansson, E., 2008. New evidence on cross-country differences in job satisfaction
using anchoring vignettes. Labour Economics, 15 (1), 96-117.
Lambert, A., Lambert, J.L. and Saives A. L., 1996. La filière des produits de la mer : innovante ou à la
traîne ? revue POUR, 149-150, 183-189.
25
Lancaster, K., 1971. Consumer Demand: A New Approach. New York, Colombia University Press.
Lombardini-Riipinen, C., 2005. Optimal tax policy under environmental quality competition.
Environmental and Resource Economics, 32, 317-336.
Long, J.S. and Freese, J., 2006, Regression Models for Categorical Dependent Variables Using Stata. 2nd
Edition. Stata Press.
Mahenc, P., 2008. Signaling the environmental performance of polluting products to green consumers .
International Journal of Industrial Organization, 26, 59-68.
Moraga-González, J. L. and Padrón-Fumero, N., 2002. Environmental Policy in green market .
Environmental and Resource Economics, 22,419-447
Mussa, M., Rosen, S., 1978. Monopoly and Product Quality. Journal of Economic Theory, 18, 301317.
OCDE, 2002a. Report of the OECD workshop on information and consumer decision-making for
sustainable
consumption,
Working
Party
on
National
Environmental
Policy,
ENV/EPOC/WPNEP(2001)16/FINAL.
OCDE, 2002b. Towards Sustainable Household Consumption? Trends and Policies in OECD Countries.
Paris.
OECD, 2005. Effects of Eco-labelling Schemes: Compilation of Recent Studies. Joint Working Party on
Trade and Environment, COM/ENV/TD(2004)34/FINAL
OFIMER, 2008. Key figures for the fisheries and aquaculture sector in France.
Perraudeau, Y., 2008. Les politiques des pêches et de l’aquaculture en France et en Europe . in
Guillotreau P. (Editors), Mare economicum. Enjeux et avenir de la France maritime et littoral, Presses
Universitaires de Rennes, 25-67.
Perraudeau, Y, Martinez, F. and Salladarré, F., 2008. Image et valorisation sociale du secteur des pêches
maritimes en Europe. Rapport final du LEN corrail (Université de Nantes), Europêche et ETF pour la
Commission Européenne, DG Emploi (forthcoming).
Perraudeau, Y., 1996. A la recherche d’une adéquation ressources-marché. Actes du colloque des 9e
rencontres interrégionales de l’AGLIA, 25-27.
Poyago-Theotoky, J. A., Teerasuwannajac, K., 2002. The Timing of Environmental Policy: A Note on the
Role of Product Differentiation. Journal of Regulatory Economics, 21 (3), 305-316.
26
Sajaia Z., 2008. Maximum likelihood estimation of a bivariate ordered probit model: implementation and
Monte Carlo simulations. The Stata Journal, Forthcoming
Sauper H., 2004, Darwin’s nightmare, documentary, France/Austria/Belgium, 107 mn, 35mm.
Stern, P., Dietz, T. and Kalof L., 1993. Values Orientations, Gender, And Environmental Concern.
Environment and Behavior, 25 (3), 322-348.
Solon, Gary, 2002. Cross-Country Differences in Intergenerational Earnings Mobility. Journal of
Economic Perspectives, 16, 59-66.
Teisl, M. F., Roe, B., Levy, A. S., 2002. Can Eco-Labels Tune A Market? Evidence From Dolphin-Safe
Labeling. Journal of Environmental Economics and Management, 43, 339-359.
Teisl, M. F., 2003, What We May Have Is a Failure to Communicate: Labeling Environmentally Certified
Forest Products, Forest Science, 49 (5), 668-680.
Teisl, M. F., Rubin, J. and Noblet, C. L., 2008. Non-dirty-dancing? Interactions between eco-labels and
consumers. Journal of Economic Psychology, 29, 140-159.
Thrane, M., Ziegler, F. and Sonesson, U. (2009), Eco-labelling of wild-caught seafood products, Journal
of Cleaner Production, 17, 416-423.
Tirole, J., 1988. The Theory of Industrial Organisation. MIT Press, Cambridge, 479 p.
Torgler, B. and Garcia-Valinãs, M. A., 2007. The determinants of individuals Attitudes Towards
Preventing Environmental Damage, Ecological Economics, 63, 536-552.
Washington, S., 2008. Ecolabels and Marine Capture Fisheries: Current Practice and Emerging Issues.
Globefish Research Program, 91.
Wessels, C.R., Johnston, R.J. and Donath, H., 1999. Assessing consumer preferences for eco-labeled
seafood: the influence of species, certifier, and household attributes. American Journal of Agricultural
Economics, 81 (5), 1084-89.
Whitmarsh, D. and Wattage, P., 2006. Public attitude towards the environmental impact of salmon
aquaculture in Scotland. European Environment, 16, 108-121.
Zelezny, L.C., Chua, P.-P. and Aldrich C., 2000. Elaborating on Gender Differences in
Environmentalism. Journal of Social Issues, 56, 443-457.
27
Appendix
— Please insert Table A1 —
— Please insert Table A2 —
— Please insert Table A3 —
28
Table 1. Estimation results of the ordered Probit model
Eco-label
Variables
Coefficient
t-statistic
Average change
Pay attention to the product form (frozen vs fresh)
Disagree
Ref.
Don’t agree, nor disagree
0.124
0.82
0.019
Agree
0.440***
2.81
0.070
Pay attention to the origin of fish
Disagree
Ref.
Don’t agree, nor disagree
0.177***
2.91
0.028
Agree
0.292***
5.13
0.046
Pay attention to the visual aspect of fish
Disagree
Ref.
Don’t agree, nor disagree
0.087
0.90
0.014
Agree
0.252***
3.07
0.039
Pay attention to the price
Disagree
Ref.
Don’t agree, nor disagree
0.149
1.40
0.023
Agree
0.289***
2.75
0.045
Pay attention to the resources level
Disagree
Ref.
Don’t agree, nor disagree
0.048
0.90
0.007
Agree
0.152***
2.83
0.023
Pay attention to the wild or farmed origin of fish
Disagree
Ref.
Don’t agree, nor disagree
0.083
1.38
0.013
Agree
0.097*
1.75
0.015
Number of observations
3356
LR chi2(37)
674.83***
Log Likelihood
-3553.95
Pseudo R2
0.0867
The reported coefficients are estimated with an ordered Probit model.
The significance thresholds are respectively 1%(***), 5%(**) and 10%(*).
The third column corresponds to the average change in the predicted probability of eco-label demand.
The specification includes control variables: gender, age (4 categories), the family status (2 categories), the type of job (9
categories), the type of the habitation (2 categories), the general point of view regarding fishing activities (3 categories), the
perception of fishing regulation strictness (3 categories), the sea side frequentation (4 categories), and the countries effects (5
categories).
29
Table 2. Estimation results of the bivariate ordered Probit model
Pay attention to the
resource level
Eco-label
Coefficient
The fisheries are:
Insufficiently regulated
Acceptably regulated
Well regulated
The quantity of fish in the sea is stable
Disagree
Don’t agree. nor disagree
Agree
Seaside frequentation
Never
Between 1 and 10 days per annum
Between 11 and 30 days per annum
More than 30 days per annum
Type of habitat (Ref. = urban)
t-statistic
Coefficient
t-statistic
Ref.
-0.345***
-0.391***
-7.79
-7.21
Ref.
-0.107***
-0.096**
-2.63
-1.88
Ref.
-0.253***
-0.463***
-4.51
-7.90
Ref.
-0.279***
-0.257***
-5.60
-4.85
Ref.
-0.017
0.127
0.200**
-0.016
-0.21
1.50
2.32
-0.32
Ref.
-0.242***
-0.158**
-0.122
-0.015
-3.11
-1.98
-1.51
-0.34
Ref.
0.042
-0.008
-0.210*
-0.128***
-0.002
0.57
-0.09
-1.67
-3.27
-0.04
Ref.
0.040
0.020
-0.076
-0.066*
0.023
0.59
0.26
-0.65
-1.82
0.55
Ref.
-0.497**
-0.017
0.028
0.202***
-0.078
-2.48
-0.20
0.36
3.04
-0.94
Ref.
0.001
0.154**
-0.057
-0.078
-0.073
0.01
2.00
-0.78
-1.29
-0.93
0.071
0.88
-0.186**
-2.50
0.139
1.38
0.079
0.85
0.181
1.60
0.227**
2.24
10.38
7.93
5.65
2.41
Ref.
-0.210***
-0.144**
-0.416***
-0.193**
-2.73
-2.01
-5.75
-2.54
Age
15-25 years
25-45 years
45-65 years
More than 65 years
Gender (Ref. = women)
Marital status (Ref. = couple)
Professional situation
Farmer
Worker
Employed
Self-employed
Intellectual profession
Intermediary profession
Student
Retired
Without employment
Countries
Denmark
Ref.
Belgium
0.878***
France
0.610***
Italy
0.436***
The Netherlands
0.195**
Number of observations
LR chi2(25)
Correlation coefficient (t-test)
Log Likelihood
The reported coefficients are estimated with a bivariate ordered Probit model.
The significance thresholds are respectively 1%(***), 5%(**) and 10%(*).
3524
495.24***
0.174*** (8.53)
-9012.18
30
31
Table A1. Descriptive statistics of the individual socioeconomic characteristics
Variables
Gender (men)
Age
15-25 years
25-45 years
45-65 years
More than 65 years
Marital status (Single)
Professional situation
Farmer
Worker
Employed
Self-employed
Intellectual profession
Intermediary profession
Student
Retired
Without employment
Habitation (urban)
Seaside frequentation
Never
Between 1 and 10 days per annum
Between 11 and 30 days per annum
More than 30 days per annum
Income
Less than 1000€/month
1000 and 2000€/month
2000 and 3000€/month
More than 3000€/month
Countries
Denmark
Belgium
France
Italy
The Netherlands
The quantity of fish in the sea is stable
Disagree
Don’t agree. nor disagree
Agree
The fisheries are:
Not at all regulated
Insufficiently regulated
Acceptably regulated
Well regulated
Very well regulated
Pay attention to the resources level
Disagree
Don’t agree. nor disagree
Agree
Pay attention to the visual aspect of fish
Disagree
Don’t agree. nor disagree
Agree
Pay attention to the wild or farmed origin of fish
Disagree
Don’t agree. nor disagree
Agree
Pay attention to the price
Disagree
Don’t agree. nor disagree
Agree
Pay attention to the origin of fish
Disagree
Don’t agree. nor disagree
Agree
Pay attention to the product form (frozen vs fresh)
Disagree
Don’t agree. nor disagree
Agree
Source: Perraudeau et al. (2008)
Observations
4747
Mean
0.4894
Min
0
Max
1
4746
4746
4746
4746
4747
0.2533
0.3902
0.2739
0.0826
0.6210
0
0
0
0
0
1
1
1
1
1
4731
4731
4731
4731
4731
4731
4731
4731
4731
4744
0.0095
0.0710
0.2323
0.0797
0.1729
0.0664
0.2234
0.1131
0.0317
0.3942
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
4731
4731
4731
4731
0.0633
0.3418
0.3071
0.2878
0
0
0
0
1
1
1
1
4668
4668
4668
4668
0.3194
0.4023
0.1645
0.1138
0
0
0
0
1
1
1
1
4748
4748
4748
4748
4748
0.1784
0.1788
0.2169
0.2338
0.1921
0
0
0
0
0
1
1
1
1
1
4546
4546
4546
0.1744
0.4495
0.3761
0
0
0
1
1
1
4264
4264
4264
4264
4264
0.0963
0.3571
0.3462
0.1417
0.0587
0
0
0
0
0
1
1
1
1
1
4062
4062
4062
0.4289
0.2461
0.3250
0
0
0
1
1
1
4359
4359
4359
0.0624
0.1149
0.8227
0
0
0
1
1
1
4228
4228
4228
0.2895
0.2223
0.4882
0
0
0
1
1
1
4438
4438
4438
0.0657
0.3535
0.5808
0
0
0
1
1
1
4339
4339
4339
0.2688
0.2265
0.5047
0
0
0
1
1
1
4378
4378
4378
0.0466
0.6812
0.2722
0
0
0
1
1
1
32
Table A2. Average score of the Willingness to see an eco-label stamped
on “green” fish according to socioeconomic and consumption behaviour criteria
Socioeconomic variables
Index
Purchase criteria variables
Gender
The quantity of fish in the sea is stable
Women
4.25
Disagree
Men
4.13
Don’t agree. nor disagree
Age
Agree
15-25 years
4.26
The fisheries are:
25-45 years
4.23
Not at all regulated
45-65 years
4.14
Insufficiently regulated
More than 65 years
3.98
Acceptably regulated
Marital status
Well regulated
Married
4.20
Very well regulated
Single
4.19
Pay attention to the resources level
Professional situation
Disagree
Employed
4.24
Don’t agree. nor disagree
Farmer
3.78
Agree
Worker
4.10
Pay attention to the visual aspect of fish
Self-employed
4.17
Disagree
Intellectual profession
4.18
Don’t agree. nor disagree
Intermediary profession
4.13
Agree
Student
4.27
Pay attention to the wild or farmed origin of fish
Retired
4.06
Disagree
Without employment
4.35
Don’t agree. nor disagree
Income level
Agree
Less than 1000€/month
4.26
Pay attention to the price
1000 and 2000€/month
4.21
Disagree
2000 and 3000€/month
4.15
Don’t agree. nor disagree
More than 3000€/month
4.04
Agree
Habitation
Urban
4.27
Non-urban
4.07
Seaside frequentation
Pay attention to the origin of fish
Never
4.03
Disagree
Between 1 and 10 days per year
4.16
Don’t agree. nor disagree
Between 11 and 30 days per year
4.22
Agree
More than 30 days per year
4.25
Pay attention to the product form (frozen vs fresh)
Countries
Disagree
Italy
4.28
Don’t agree. nor disagree
Belgium
4.46
Agree
The Netherlands
3.99
France
4.33
Denmark
3.83
*The score associated with the answer ‘strongly disagree’ is 1, the score with the answer ‘strongly agree’ is 5.
Source: Perraudeau et al. (2008)
Index
4.50
4.27
4.01
4.41
4.37
4.07
4.03
3.66
4.10
4.19
4.35
3.90
4.05
4.25
4.03
4.15
4.33
3.88
4.12
4.28
3.96
4.12
4.37
4.04
4.06
4.25
33
Table A3. Estimation results of the ordered Probit and generalized ordered Probit model
Ordered Probit
Generalized Ordered Probit
Coefficient
t-statistic
Pay attention to the product form (frozen vs fresh)
Disagree
Ref.
Don’t agree, nor disagree
0.124
0.82
Agree
0.440***
2.81
Pay attention to the origin of fish
Disagree
Ref.
Don’t agree, nor disagree
0.177***
2.91
Agree
0.292***
5.13
Pay attention to the visual aspect of fish
Disagree
Ref.
Don’t agree, nor disagree
0.087
0.90
Agree
0.252***
3.07
Pay attention to the price
Disagree
Ref.
Don’t agree, nor disagree
0.149
1.40
Agree
0.289***
2.75
Pay attention to the resources level
Disagree
Ref.
Don’t agree, nor disagree
0.048
0.90
Agree
0.152***
2.83
Pay attention to the wild or farmed origin of fish
Disagree
Ref.
Don’t agree, nor disagree
0.083
1.38
Agree
0.097*
1.75
The fisheries are:
Insufficiently regulated
Ref.
Acceptably regulated
-0.345***
-7.79
Well regulated
-0.391***
-7.21
The quantity of fish in the sea is stable
Disagree
Ref.
Don’t agree. nor disagree
-0.253***
-4.51
Agree
-0.463***
-7.90
Seaside frequentation
Never
Ref.
Between 1 and 10 days per annum
-0.058
0.67
Between 11 and 30 days per annum
0.081
0.91
More than 30 days per annum
0.160*
1.73
Type of habitat (Ref. = urban)
-0.016
-0.32
Age
15-25 years
Ref.
25-45 years
0.001
0.01
45-65 years
-0.048
-0.55
More than 65 years
-0.288**
-2.19
Gender (Ref. = women)
-0.091**
-2.25
Marital status (Ref. = couple)
0.007
0.16
Professional situation
Farmer
Ref.
Worker
-0.351
-1.55
Employed
-0.062
-0.73
Self-employed
0.025
0.31
Intellectual profession
0.196***
2.86
Intermediary profession
-0.022
-0.26
Student
0.076
0.92
Retired
0.108
1.02
Without employment
0.113
0.97
Countries
Denmark
Ref.
Belgium
0.890***
8.86
France
0.503***
5.58
Italy
0.409***
4.48
The Netherlands
0.294***
3.19
Threshold 1
-1.430
Threshold 2
-0.921
Threshold 3
-0.172
Threshold 4
1.093
Number of observations
3356
LR chi2(37/100)
674.83***
Log Likelihood
-3553.95
Pseudo R²
0.0867
The significance thresholds are respectively 1%(***), 5%(**) and 10%(*).
Coefficient
t-statistic
Ref.
0.090
0.402**
0.59
2.52
Ref.
0.166***
0.298***
2.71
5.18
Ref.
0.118
0.293***
1.21
3.53
Ref.
0.158
0.304***
1.48
2.87
Ref.
0.031
0.142***
0.57
2.63
Ref.
0.079
0.083
1.30
1.48
Ref.
-0.309***
-0.401***
-6.70
-7.05
Ref.
-0.210***
-0.418***
-3.55
-6.79
No
No
No
No
3356
788.25***
-3463.06
0.1101