On the willingness to pay for food sustainability labelling: A meta-analysis

Giovanna Piracci1, Emilia Lamonaca2, Fabio G. Santeramo2, Fabio Boncinelli1, Leonardo Casini1
1University of Florence (Italy)
2University of Foggia (Italy)

Abstract

Sustainability labelling is an extremely complex, multifaceted and debated topic. Through a

systematic and meta-analytical approach, we disentangled the informative contents of environmental

and social labels and investigated their effect on the consumer willingness to pay for food products.

Premium prices for sustainability labels are largely heterogeneous depending on the information

disclosed. Generic and specific messages seem not to differ in terms of consumer acceptability. Not

all facets are equally important as social issues tend to be less considered. Policy interventions should

combine hard and soft measures to holistically achieve sustainability in the food system.

Keywords: Systematic review, Meta-regression, Food Policy, Sustainable food choices, Ethical

consumption, Animal welfare.

JEL Codes: D12, Q13, Q18.

1
 On the willingness to pay for food sustainability labelling: A meta-analysis

Introduction

The plethora of existing Sustainability Labels (SLs)1 in the food system signals a growing interest

from both the supply and demand sides towards sustainability issues (Asioli et al., 2020). The

Ecolabel Index (2022) tracks 147 food SLs worldwide. A sustainable food system “delivers food

security and nutrition for all in such a way that the economic, social and environmental bases to

generate food security and nutrition for future generations are not compromised” (FAO, 2018).

Therefore, labelling can address sustainability as either a holistic concept or by separately considering

both the environmental and the social dimensions 2. Environmental sustainability covers any action

aimed at protecting and preserving the environment, whereas social sustainability relates to the needs

and well-being of individuals (Retail Forum for Sustainability, 2011).

As for the labels which incorporate the environmental dimension of sustainability, the global organic

market was estimated at €106 billion in 2019 (Willer et al., 2021). The world sales of seafood

products, as certified by the Marine Stewardship Council (MSC), reached more than US$10 billion

during the biennial period 2019–2020 (MSC, 2020). Furthermore, between 2010 and 2018, the sales

revenue for foods labelled Fairtrade, the most popular certification among the social ones, grew by

124%, while the premium prices increased by 267% (Bhavsar et al., 2021).

Investigating the availability of a premium price for SLs can be of paramount importance for several

reasons. Reaching a premium price for sustainability-labelled products is a prerequisite for producers

1 The current analysis is focused on labels associated with environmental and social standards not legally imposed but adopted on a
voluntary basis, e.g., Organic, Fairtrade, Rainforest Alliance or Animal Welfare certifications. Compliance with the standard is certified
upon verification by an independent third party.
2 The economic, environmental, and social pillars are the three recognized essential dimensions of the sustainability concept (Elkington,

1997). However, our study acknowledges that sustainability attributes of products can be expressed only through the environmental
and social axes, in line with previous works (Grunert et al., 2014; Janssen and Langen, 2017; Asioli et al., 2020; Bangsa and
Schlegelmilch, 2020). This choice is due to the fact that what can be included in the economic dimension of sustainability labels is still
controversial. We assume that the economic sense of the labels relies on the profitability of the instrument and, thus, on the capability
of the attribute to generate a premium price to compensate for the sustainability practices implemented on the producer side.

2
and retailers to maintain profitability in that switching from conventional to sustainable alternatives

tend to raise production costs. The premium for SLs can potentially cover the cost of the provision of

environmental or social benefits related to the food production process, remunerating the supply of

public goods (Bougherara and Combris, 2009). On the other hand, assessing the WTP for food

attributes is an informative indicator of the acceptability of citizens for different sustainability

practices (Clark et al., 2017). Furthermore, from a policy perspective, sustainability labelling is one

of the instruments of the policy mix targeting the transition towards a more sustainable food system.

For instance, the European Farm to Fork (F2F) Strategy, one of the pillars of the European Green

Deal policy, proposes the implementation of a harmonised sustainable food labelling framework that

embraces the nutritional, climatic, environmental, and social aspects of food production (European

Commission, 2020a). To this end, understanding how the different types of information conveyed

through labelling affect the consumer WTP is expected to outline remarkable insights concerning the

effectiveness of a sustainability labelling scheme.

A large body of research has been focusing on estimating the premium prices for SLs, which are very

heterogeneous. The variability in estimated premium prices is likely due, among other factors, to

diversity of applied methodologies as well as to the food category under investigation. Based on these

premises, we conducted a meta-analysis to: (i) systematize evidence on consumer preferences for

SLs; (ii) evaluate consumer premium prices for sustainability labelling; and (iii) disentangle the

sources of heterogeneity in WTP estimates due to different types of information conveyed by the SLs.

Previous reviews and meta-analyses of consumer WTP provide insightful overviews of sustainability-

related issues, such as animal welfare (Clark et al., 2017), organic production (Katt and Meixner,

2020), carbon footprint (Rondoni and Grasso, 2021) and environmental impacts (Bastounis et al.,

2021), or explored sustainability labelling of specific food categories, such as wine (Schäufele and

Hamm, 2017). Although providing exhaustive analyses on the single SL or product chosen, these

studies fail to provide an overview on the whole phenomenon. Moreover, there is still a need to

deepen on how consumers value different sustainability labels.

3
A remarkable study is the paper by Li and Kallas (2021), who conducted the first meta-analysis on

the WTP for food SLs. Their results indicate that, on average, SLs correspond to a 29.5% premium

price, and conclude that sustainability labelling is a profitable strategy for retailers and there is a great

market potential for sustainable products. Our research expands their findings from different

perspectives. First, by considering the heterogeneity from multiple angles: the label broadness (i.e.

specific and generic information), the sustainability dimension underlying the label (i.e.

environmental and social issues), the specific category of the label (i.e. environmental impact,

production method, animal welfare, social fairness or sustainable production-related labels), and the

economic value of the carrier food products (i.e. high-priced and low-priced foods). Second, our set

of scientific works (110) provides more than six hundred observations, resulting in the largest meta-

analysis on the WTP for sustainability attributes, and allowing us to derive policy implications.

The multidimensionality and highly fragmented nature of sustainability imply that consumers may

potentially trade-off between different and specific sustainability facets. We assessed the relative

impact of the different sustainability concerns providing relevant evidence to policymakers on how

sustainability issues can be proficiently addressed through food labelling and which ones, being less

considered, need to be tackled through other types of information-provision actions.

We derive conclusions on the effectiveness of holistic SLs as a policy tool aimed to guide the

transition towards a sustainable food system.

Background

As a communication tool, sustainability labelling carries out numerous functions. First, it empowers

consumers to distinguish foods produced in compliance with certain standards from conventional

products. Consequently, consumers can make informed decisions, considering the environmental and

social impacts of their food choices (Grunert et al., 2014; Van Loo et al., 2017). Our argument draws

upon the assumption that consumers are better-off when full information is available. Nonetheless,

4
it is worth pointing out that in our increasingly intricate and information-saturated world, consumers

do not always seek exhaustive information and they can ultimately exhibit a WTP for not receiving

it (Reisch et al., 2021).

Second, SLs represent one of the soft-approach policy instruments aimed at encouraging voluntary

changes towards more sustainable consumption patterns through information provision (Noblet and

Teisl, 2015; Van Loo et al., 2017; European Commission, 2020b). In addition, communication

through labelling can mitigate the market barrier caused by information asymmetries between

producers and consumers (Rousseau and Vranken, 2013). Finally, on the supply side, labels represent

a distinctive symbol that signals to consumers certain product features or distinguishable

characteristics of production methods (de Boer, 2003). Therefore, SLs enable producers to establish

product differentiation, both across and within the food categories, and consequently allow for the

creation of a potential premium price. However, such premium, as well as the market shares, for SLs

can dramatically change depending on the food product (Ardeshiri and Rose, 2018; Dahlhausen et

al., 2018), country of sale (Akaichi et al., 2020; Menozzi et al., 2020), origin (Lim et al., 2018),

retailer (Asche et al., 2015) and, most importantly, differences in SLs (Van Loo et al., 2014; Janssen

and Langen, 2017).

Existing SLs can be extremely heterogeneous and, at the same time, strictly intertwined. This mirrors

the underlying concept of sustainability, which is primarily a complex and multidimensional issue.

Janssen and Langen (2017) coined the phrase ‘the bunch of sustainability labels,’ whereas Torma and

Thøgersen (2021) refer to ‘the sustainability labelling landscape’ to depict the current situation. So

far, no common definition or official classification has been proposed for SLs, resulting in ambiguity

and confusion about what can or cannot be recognised as sustainability labelling. Therefore, we

developed the following conceptual framework to classify the labels as the basis for our analysis.

5
Conceptual framework: A classification proposal

Applying the most basic criterion and drawing upon the study of Torma and Thøgersen (2021), labels

can be distinguished according to their broadness, i.e. whether they are specifically formulated or

generically formulated. Specific labelling addresses single and precise issues of sustainability (e.g.

the organic or carbon footprint labels); generic labelling refers to the overarching concept of

sustainability or, at least, to the overall environmental or social performance of the product (e.g. the

sustainable product claim or the eco-friendly label).

The second partitioning rule is based on the sustainability dimensions (as per Grunert et al., 2014;

Van Loo et al., 2014; Asioli et al., 2020). To this end, labels can be grouped as either environmental

or social. The former pertains to the environmental dimension of sustainability and signals ecological

issues such as pesticide level reduction (e.g. pesticide-free), GHG emissions (e.g. carbon footprint

labelling), biodiversity protection (e.g. biodiversity friend) or the adoption of environmentally

friendly production methods (e.g. the organic or biodynamic farming certifications). On the other

hand, social labels address ethical or moral aspects, such as concern for animal welfare (e.g. the

‘Certified Humane’ or the free-range label), the improvement of profit distribution in marginalised

countries (e.g. the fair-trade label), concern for workers’ conditions (e.g. the Social Accountability

SA8000) or the endorsement of the food chain based on direct marketing (e.g. farmers’ products or

community-supported agriculture labels).

Additionally, SLs can simultaneously involve both environmental and social issues (Janssen and

Langen, 2017; Bangsa and Schlegelmilch, 2020). The most vivid example in this group is the organic

label when referring to animal-based foods, as it jointly addresses agriculture practices with lower

environmental impacts and respect for animal welfare conditions (Janssen and Langen, 2017).

Likewise, the Rainforest Alliance certification programme guarantees both the environmental

protection of farms and forests and the promotion of workers’ rights and communities’ well-being

(Ecolabel Index, 2022). The generic label sustainable product also pertains to this partitioning since

the term sustainability may be interpreted in both directions.

6
Adding to the complexity, SLs can be further classified according to their basic scope of applicability

and, thus, with respect to their informative content. As a result, it is possible to identify the following

five macro-categories: (i) environmental impact; (ii) production method; (iii) animal welfare; (iv)

social fairness; and (v) sustainable production. Labels in the first group aim to signal the specific

environmental outcomes associated with food production (e.g. the reduced water usage, the pesticide-

free or the eco-friendly labels), and inform consumers on the ecologically positive impact of the

food. The second group of labels, i.e. the method of production, signals that the food has been

produced in compliance with specific sustainable production processes. The labels inform on the

food’s production method and the associated regulatory framework. The well-established organic

label is a remarkable, and very important, example (Lernoud et al., 2018). Other relevant examples

are the MSC, the biodynamic and the Aquaculture Stewardship Council (ASC) certifications. The

third class refers to the animal welfare labelling, which addresses information on livestock and poultry

conditions (e.g. animal-friendly, free-range or grass-fed). On the other hand, labels in the social

fairness class inform on specific facets of social sustainability directly connected to the enhancement

of human conditions (e.g. fairtrade, workers’ welfare or cause-related marketing). Finally, the

sustainable production labelling contains information targeting the overarching concept of

sustainability, such as the use of the terms ‘sustainable product’ or ‘made from sustainable

agriculture’.

The classification systems proposed are summarised in Table 1. Annex A provides further definitions

for the SLs included in the study.

[Table 1]

We only considered food information related to the production of positive externalities for the whole

society, such as the environmental benefits or ethical and social outcomes of the production process,

for sustainability labelling. Food labelling pertaining or confounding to other consumption trends,
7
most notably health or safety attributes (e.g. hormone-free or genetically-modified-organism-free

indications) and origin information (e.g. region-of-origin, country-of-origin, protected designation of

origin, protected geographical indication or food-miles indication), were not treated as SLs and hence,

not included in our analysis. Such quality attributes incorporate private and self-gains for the

individual, whereas SLs pertain to the public good dimension (Lusk et al., 2007; Asche et al., 2015) 3.

This is consistent with our research question as we aimed to determine if private market instruments,

such as labelling, can help in fostering the provision of public goods in the agri-food chain. In

addition, sustainable attributes (i.e. public attributes) differentiate food products according to non-

direct-use quality dimensions, whilst healthy, origin or nutritional attributes (i.e. private attributes)

relate to the individual’s direct use of the food (Teisl and Roe, 1998).

Methods

A meta-analytical approach, performed in accordance with the ‘Reporting guidelines for meta-

analysis in economics’ (Havránek et al., 2020), allowed us to combine and quantitatively synthesise

empirical evidence from different studies on consumer WTP for SLs while seeking to identify

heterogeneity sources among results (e.g. Stanley, 2001; Stanley et al., 2013).

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA)

protocol (Page et al., 2021) presented in Figure 1, we carried out a systematic review of consumer

studies focusing on SLs, published in the Scopus database. The Scopus database was chosen as the

repository for the systematic review since provides a complete coverage of solid academic journals

(Fogt Jacobsen et al., 2022) and constitutes one of the few scientific databases suitable for systematic

3 The organic label was added to the analysis, even though several studies suggest that organic food purchase is also driven by consumer
health (i.e., private concerns) (Magnusson et al., 2003; Schleenbecker and Hamm, 2013), besides the primary motivations related to
environmental and animal welfare issues. However, it is the most extended sustainability standard in terms of area coverage and the
one embracing the broadest range of agricultural products (Lernoud et al., 2018). For these reasons, organic production is deemed
pivotal in the transition towards more sustainable dietary patterns (Aschemann-Witzel and Zielke, 2017), and we included the organic
label in the study for a comprehensive analysis of the topic.

8
reviews (Gusenbauer and Haddaway, 2020). The relevant studies were extracted in November 2020;

without constraints in terms of time horizon. The search strategy, including the keywords used, is

detailed in Annex B. The eligibility of each article was assessed according to three general inclusion

criteria. First, the work must be focused on consumer preferences for SLs on food products. Second,

the aim of the work must be to elicit consumer WTP by applying the choice experiment method.

Third, the results must report the complete set of estimated coefficients from the model. The first

criterion restricted the analysis to food products and allowed us to focus on consumer acceptance of

sustainability attributes 4 . The second criterion permitted the selection of homogeneous studies

according to the elicitation method applied for the WTP estimates. Previous meta-analyses (e.g. Clark

et al., 2017; Printezis et al., 2019) conclude that different experimental techniques can provide

significantly different WTP measures. Therefore, we narrowed it down to the most used technique.

Finally, thanks to the third criterion, we were able to construct a measure of consumers’ WTP that

was consistent across studies, which ultimately enabled a more robust comparison between the

literature findings. From 833 articles, we selected 131 eligible studies satisfying all three inclusion

criteria (Figure 1).

[Figure 1]

From the selected studies w e collected the parameters of the SL attribute and the price attribute with

related standard errors (or t-statistics) estimated through discrete choice models. The former measures

the consumer’s marginal utility for the SL presence on the food, while the latter measures the

consumer’s marginal utility for the price of the food. In addition, we recorded each food product

reference price, which denotes the actual market price for the conventional food under investigation

4We did not restrict eligibility to empirical work. Studies focusing on methodological issues of discrete choice modelling (such as
Bello and Abdulai, 2016; Olsen and Meyerhoff, 2017), which satisfied all three inclusion criteria, were equally considered.

9
reported in each article5. We retrieved 1,287 observations due to the different types of SLs included

per study. Moreover, in case they were multiple, we collected all the estimates for the SL parameter

available in the article to account for the complete literature evidence rather than selecting a privileged

model specification (Stanley, 2001; Santeramo and Lamonaca, 2019). In this phase, we preliminarily

checked for the presence of publication selection bias using the funnel plot technique. The results are

reported in Annex C. The analysis revealed that publication selection bias was not an issue in our

sample.

Following common wisdom (e.g. Lusk et al., 2005; Deselnicu et al., 2013; Santeramo and Lamonaca,

2021), we computed a WTP index (IndexWTP) as follows:

𝛽𝑙𝑎𝑏𝑒𝑙
−𝛼𝑝𝑟𝑖𝑐𝑒
𝐼𝑛𝑑𝑒𝑥𝑊𝑇𝑃 = (1)
𝑃𝑟𝑒𝑓

As models for discrete choices are linear in the utility functions, we took the negative ratio between

the SL attribute parameter, βlabel, and the price attribute parameter, αprice, to calculate the negative

marginal rate of substitution between the two respective attributes. The marginal rate of substitution

indicates the extent to which the consumer is willing to trade one attribute for another, keeping the

utility constant. When at least one attribute is measured in monetary units, the (negative) ratio of the

two parameters will provide the financial measure of the consumer’s WTP for the non-monetary

attribute 6 (Louviere et al., 2000; Hensher et al., 2015). The WTP obtained is robust under the

assumption that the marginal utility of the price attribute of the model is negative and significant, in

accordance with the economic theory. For this reason, we excluded from the sample all the

5 In case a reference price was not given in the article, we averaged the price levels applied in the choice experiment, as in previous
meta-analysis (see, for instance, Lusk et al., 2005; Santeramo and Lamonaca, 2021).
6
In case the estimates belonged to a model specified in WTP space, we treated the estimated coefficient of the SL as the WTP itself:
WTPlabel = βk. Models in WTP space are reparametrised such that the coefficients enter the model already scaled by the price/scale
parameter. Thus, they can directly be interpreted as the marginal WTP values for the non-monetary attributes (Train and Weeks, 2005;
Scarpa et al., 2008).

10
observations with positive and non-significant estimates for the price parameter (126 observations).

Afterwards, we normalised the measures of the WTP for SLs across articles using the reference price

for the product7, Pref, in line with previous meta-regression analyses on food attributes (Lusk et al.,

2005; Dolgopolova and Teuber, 2018). The normalisation cancels out the differences due to type of

currencies and related exchange rates, as well as measurement errors connected to diverse timing,

and units of measure across studies. The WTP index represents the effect size for the meta-regression

analysis since it allows us to compare diverse studies on the same dimensionless scale (Stanley, 2001).

A detailed analysis of the WTP index is reported in Annex D, including the complete list of articles

for the quantitative analysis (Table D3).

We investigate possible drivers of heterogeneity in consumer WTP for SLs by estimating the

following model:

𝐼𝑛𝑑𝑒𝑥𝑊𝑇𝑃 = 𝜆 + 𝜑𝑋 + 𝜌𝐾 + 𝜔𝑍 + 𝜓𝑇 + 𝜀 (2)

where IndexWTP is the vector of the observations for the dependent variable; λ is a vector of constant

terms, accounting for the average premium price for SLs; 𝜑, 𝜌, 𝜔, and 𝜓 are further parameters to be

estimated; ε stands for the vector of i.i.d. distributed error terms.

The matrix X contains variables related to the labels. The model in equation (2) is estimated in three

specifications. In the first specification (i), we tested for the broadness of formulation (i.e. specific

versus generic) of SLs. In this specification, a single variable is included in X: a dummy equal to 1

for specific SLs and 0 otherwise. In the second specification (ii), we controlled for the sustainability

dimension of the label (i.e. environmental, social, and environmental and social): the matrix X

includes a dummy for environmental SLs and a dummy for SLs addressing both environmental and

7
At this stage, we removed one article (Xu et al., 2018) because no reference price for the food was available in the study, and, thus,
it was not possible to compute the WTP index.

11
social issues (the social dimension of sustainability is treated as the baseline). In specification (iii),

we disentangled the contribution of the scope and principal information provided to consumers

through SLs: we include dummies for labels on environmental impacts, production methods, animal

welfare and sustainable production (labels on social fairness are the baseline). In addition, we

incorporated control factors in each specification to distinguish between existent vis-a-vis potential

labels, i.e. fictitiously implemented in experiments of hypothetical choices.

The matrix of moderator variables related to the structural characteristics of the studies in the sample,

K, includes the food category of the product carrying the label, the quantity of food, the region of the

study (i.e. continents) and the sampled population (i.e. consumers, purchasers, or both). Foods were

grouped according to the Harmonised System nomenclature. The quantity of food was specified as

high or normal relative to the mean quantities per food retrieved from the studies (e.g. 200 g was

classified as normal, whereas 1 kg was considered as high in relation to meat products; likewise, 1 L

was classified as normal, whilst 1 gallon was considered as high in relation to milk). This regressor

allowed us to capture the variability within food categories.

The matrix Z includes moderator variables related to the technical and methodological issues of the

choice experiments, i.e. the experiment set-up (e.g. face-to-face interview, online survey and field

and lab experiment), the experimental design applied (e.g. orthogonal, optimal orthogonal in the

difference, efficient and Bayesian efficient), the number of presented food alternatives in each choice

set and the number of total attributes describing the alternatives. We added to the empirical

specification the set of variables related to the publication process (T) to discriminate between articles

published before and after the years 2012 and 20188, the journal ranking according to the Scimago

Journal & Country Rank (SJR) at the date of publication and the subject areas of the journal.

The dependent and independent variables are described in Table 2.

8The years 2012 and 2018 have been used as threshold years of publication because they represent the 20th and 80th percentiles of the
distribution of the articles in the sample (Annex D, Figure D2).

12
[Table 2]

Adding up to this, the third specification of the model (iii) allows us to disentangle the price level

effects by splitting the sample in high-priced and low-priced food products. In particular, we

compared the single reference price of the product, i.e. the average market price for the conventional

food reported in each study, with the currency-specific mean reference price, i.e. the value of the

reference price averaged across the observations within the same currency.

The empirical relationship reported in (2) was estimated through a robust least square regression. To

correct for heteroskedasticity, we assumed independence across studies and correlation among the

observations within each study by applying a clustered structure of the error terms, as per Santeramo

and Lamonaca (2021).

In a sensitivity analysis, the specifications (i), (ii) and (iii) were estimated by including, progressively,

different combinations of control factors (cfr. Annex E, Tables E2 to E4). We estimated the base

model with only the explanatory variables related to the SLs (matrix X); we added the structural

characteristics (matrix K) and then the methodological and technical issues (matrix Z);lastly, we

estimated a full model with publication process variables (matrix T). Due to space limitations, we

will present in the next section the full model results focusing on the effect of the label-related

variables while omitting the control factors’ estimates for brevity.

Meta-regression Results

The estimates reported in Table 3 show the effect of SLs on the WTP index, highlighting the

contribution of specific SLs as compared to generic SLs (specification i); of labels covering the

environmental dimension of sustainability with respect to the social dimension (specification ii); and

of four categories of SLs, namely labels providing information on environmental impacts, production

13
methods, animal welfare and sustainability of products, with respect to labels on social fairness

(specification iii) 9 . When statistically significant, coefficients estimated for SLs were positive,

indicating that information on sustainability positively influences the premium price of products with

SLs. Results are robust to several sensitivity analyses (cfr. Annex E, Tables E2 to E4).

[Table 3]

Through specification (i), we did not detect any significant effect of specific SLs as compared to

generic SLs. Specific SLs are more dispersed than generic SLs, with a standard deviation of 0.45 for

specific SLs and 0.33 for generic SLs, and a coefficient of variation10 of 1.67 for specific SLs and

1.06 for generic SLs (Table 4). This indicates a greater heterogeneity in the percent variation in the

WTP associated with the presence of a specific SL. The specificity of sustainability aspects (both

dimensions and categories) involved in each SL may result in very different and sometimes opposite

premium prices for SLs (the WTP index for specific SLs ranges between -0.94 and 5.47, Table 4).

These specific effects may clear out each other impeding the identification of a common tendency.

Overall, the type of SLs (specific versus generic) is not enough to disentangle the effects of the

sustainability attribute on premium prices.

[Table 4]

Regarding the dimensions of sustainability, we found that consumers are willing to pay 19% more

for a product carrying an environmental SL as compared to social SL. The results corroborate with

previous studies, such as Van Loo et al. (2014), where the authors concluded that consumers are

9 In a sensitivity analysis (Annex E, Table E1), specification (ii) and (iii) are estimated by alternatively switching the base category of
the SL-related variables.
10 The coefficient of variation, obtained as the ratio of the standard deviation to the mean, measures the extent of variability in the WTP

index for each type, dimension and category of SLs.

14
willing to pay a premium for chicken breast products with carbon footprint labels ranging between

18% and 30%. We also found that the premium price is 15% greater for products with labels that

include both the environmental and social dimensions. For instance, consider a meat-based product,

beef, priced at 10.00 EUR/kg. An organic label certifying the adoption of both agriculture practices

with lower environmental impacts (i.e. environmental dimension) and improved standards of animal

welfare (i.e. social dimension) would increase the price of 1 kg of beef by 1.5 EUR with respect to a

social-labelled beef product. Similarly, Burnier et al. (2021), evaluating the influence of socio-

environmental attributes on the WTP for beef, find that consumers would be willing to pay between

1.83 and 3.03 USD/kg (equivalent to a range of 1.56 and 2.58 EUR/kg 11) more for the sustainability

certification.

The positive correlation between the WTP index and SLs covering issues related to the environmental

dimension or a mix of dimensions of sustainability is strengthened when specific categories of SLs

are considered. The premium price increases by 29% on average for labels signalling the

environmental outcomes associated with the production process (e.g. carbon footprint label or eco-

friendly label) and by 30% for labels communicating production methods (e.g. organic label) as

compared to labels concerning social fairness issues (e.g. fairtrade). Indeed, on average, larger

premium prices are associated with environmental SLs, whereas the premium price are lower for

social SLs, 0.33 and 0.20 respectively (Figure 2). The index is 0.21 and 0.16, respectively for animal

welfare and social fairness, as compared to 0.34 and 0.29 for environmental impact and production

method.

[Figure 2]

11The conversion is based on the exchange rate USD/EUR on the publication date of the study by Burnier et al. (2021), i.e., 31 March
2021, available at the European Central Bank.

15
Social SLs tend to be more impactful for high-priced products (Table 5). High-priced animal-based

products obtained with higher animal welfare standards increase the premium price by 86%. It

emerges that, differently from the overall tendency, the social dimension of sustainability, limiting to

the group of labels related to animal welfare, is worthwhile for a niche market, i.e. high-priced

products.

[Table 5]

Additionally, the environmental dimension of sustainability matters the most for high-value products.

The premium price is 74% higher for labels conveying information on environmental impacts and

103% greater when production methods are certified as compared to the sustainable production label

category. Consumers are willing to pay more for low-priced products with labels with information on

environmental impacts. However, the premium price for labels on environmental impacts is lower for

low-priced than for high-priced products (+22% as compared to +74%). For instance, our reference

product, beef, priced at 10.00 EUR/kg, is a high-priced product compared to other meat-based low-

priced products, such as a small whole chicken (about 1 kg) priced at 4.00 EUR/kg. An environmental

impact label would provide a premium price of 0.88 EUR/kg for the small whole chicken and 7.40

EUR/kg for the beef.

The results of Table 5 are robust to different classifications of products along their price dimensions

(cfr. Annex E, Tables E5 to E7). Furthermore, we obtained consistent findings analysing the effect of

label category on the WTP for products with different elasticities of demand (Annex E, Table E8).

Discussion

To gather generalizable evidence on consumers’ WTP for SLs, we reviewed 131 scientific

publications and meta-analysed 110 studies. To date, this is the most extensive meta-analysis on WTP

16
for sustainability attributes and the only one that looks at the WTP across labels according to their

type, sustainability dimension and specific category of information.

Overall, we found that consumers are willing to pay on average 27% more for SLs on foods, in line

with Li and Kallas (2021), which reports a premium of 29.5%. Nonetheless, our analyses point out

that looking at the overall mean across all SLs can lead to a misleading and wrong conclusion since

the field of SLs is extremely complex and multifaceted.

As for specific versus generic labelling, we did not detect any significant difference in the premium

price. This was probably because the specific label group consists of such a diverse range of labels

that any opposing impacts on the WTP index were cancelled out. On the other hand, the qualitative

analysis of the literature indicates that generic SLs are less preferred by consumers as opposed to

specific labelling (Viegas et al., 2014; Menozzi et al., 2020; Wu et al., 2020; Burnier et al., 2021).

This may be due to the fact that they sound less salient and are not directly linked to the positive

outcome they bring.

However, the establishment of an overarching (i.e. generic) SL has been advocated as a policy tool

to spread sustainability information in the food domain and encourage sustainable food choices (for

an extensive review on this topic, see Torma and Thøgersen, 2021). The future development of the

European sustainability labelling announced in the F2F points in this direction (European

Commission, 2020a) in the sense that it will be formulated to holistically convey all aspects of

sustainability. The presence of a heterogeneous plethora of specific sustainability labelling schemes

in the food market is recognised as a problem (Asioli et al., 2020). The reason is that they bring a

broad range of diverse and overlapping information (Dendler, 2014) and potentially risk favouring

consumers’ overload and confusion while jeopardising labelling reliability and consistency (Van Loo

et al., 2014; Grebitus et al., 2018). We mapped 65 different sustainability facets addressed by specific

labelling across 131 studies. Many works also corroborate the hypothesis that the simultaneous

inclusion of multiple single-issue information may negatively impact consumers’ product evaluation

(Meas et al., 2015; Hinkes and Schulze-Ehlers, 2018; Charry et al., 2019). Consequently, different
17
specific SLs may compete with each other by providing consumers with interrelated and overlapping

utilities. Negative synergy effects, often recalled as substitution effects, may arise when combining

several specific SLs on the same product. The non-proportionality between the WTP and the number

of specific SLs could be explained as an embedding effect, meaning that consumers obtain utility

from the concept of sustainability per se, whilst no additional value is gained by raising the degree of

sustainability (Tebbe and von Blanckenburg, 2018). In this regard, it is worth to provide more

scientific evidence on whether adopting the generic labelling approach – which aims to provide

consumers with a piece of holistic and unique information – instead of several specific labels is the

key to achieving a more proficient spread of the sustainability communication in food products.

Concerning the dualism of sustainability dimensions involved in food labelling, most articles

included in our review (118 out of 131) deal with labels addressing the environmental dimension of

sustainability, frequently mentioned as ecolabels. Conversely, we found that social labelling is less

considered in the literature, in line with Bangsa and Schlegelmilch (2020). The research interest in

the two sustainability pillars of food labelling seems to mirror the public opinion evaluation. Results

from the meta-analysis corroborate the idea that people value the net collective gains from eco-

friendly food-production practices and being informed through labelling of the environmental burden

associated with the production of their food. Therefore, they are willing to pay more for the perceived

improvement in the environmental quality, in line with Koistinen et al. (2013), Van Loo et al. (2014).

Specifically, label information concerning production methods and environmental outcomes is the

most impactful on the premium price for SLs. Our findings are consistent with previous works

demonstrating that the production method certification commands stronger effects than other SLs,

such as the animal welfare certification (Gross et al., 2021; Meyerding et al., 2018) or the social

fairness certifications (Fitzsimmons and Cicia, 2018; Hinkes and Schulze-Ehlers, 2018), and that

consumers highly appreciate labels attesting environmentally friendly practices, such as the carbon

footprint logo and the ‘reduced water usage’ claim (Caputo et al., 2013; Staples et al., 2020). On the

other hand, the meta-analysis stressed that social SLs are less valued among consumers compared to
18
their environmental counterparts. There is an evident inconsistency between the accepted definition

of sustainability, embracing both the environmental and social pillars, and what consumers value as

sustainable in food and, thus, what they are willing to pay for it. Qualitative findings also confirm

that preferences towards social attributes are controversial because, in some situations, consumers

were found to be positively inclined towards the presence of ethical claims on food products (Van

Loo et al., 2014; Lusk, 2019; Piracci et al., 2022), while in others, they disregarded or not paid

attention to these claims for purchasing decisions (Rigby et al., 2016; Printezis and Grebitus, 2018;

Soley et al., 2019). One of the reasons might be rooted in the fact that environmental concerns are

somehow linked to health and safety consequences, for instance, those connected to environmental

degradation. On the other hand, the social dimension of sustainability entirely focuses on improving

the conditions of others, be it in the form of human beings or animals (Capitello and Sirieix, 2020).

This may create distance as consumers can feel that these ethical purchasing decisions are not directly

relevant to themselves and never will. In addition, sustainability has always been primarily interpreted

under an ecological lens (Bangsa and Schlegelmilch, 2020). More recently, the environmental

challenges have become more familiar and better known among consumers. Publications on SLs

report that WTP can be strongly influenced by previous knowledge (Hinkes and Schulze-Ehlers,

2018; Ochs et al., 2019) and awareness (Janssen and Hamm, 2012) of the standard or of the underlying

issue. Information provision and education-oriented tools have been proven effective in increasing

the WTP and, hence, the acceptance of sustainability labelled food products (Rousseau and Vranken,

2013; Klaiman et al., 2016; Aoki et al., 2019).

In addition, the results of this study indicate that consumers’ interest in food sustainability varies

depending on the price range of the products. In the case of low-priced foods, people are willing to

pay more for the overarching labels recalling the sustainability of the product (e.g. the ‘sustainable

product’ claim), regardless of any further specific information. On the other hand, when it comes to

high-priced foods, consumers are not willing to trade off the detailed nature of the SLs for the more

generic ‘sustainable product’ claim. In this case, schemes involving specific aspects of sustainability
19
(i.e. the environmental impact labels, the production method labels or the animal welfare labels) are

the worthiest. Similar results were obtained in Contini et al. (2019) and Mazzocchi et al. (2019). Thus,

we can infer that, in low-priced products, the cost of processing more accurate information on

sustainability is higher than the expected utility gained from the product. Therefore, consumers

disregard the specific sustainability information. For instance, Naald and Cameron (2011) detected

that individuals were unwilling to pay more for the ‘humanely raised’ animal welfare certification on

chicken breast. On the other hand, for high-priced goods within the same product category, such as

ground beef or steak, consumers were willing to pay a considerable premium for the enhanced animal

welfare label (Li et al., 2018; Burnier et al., 2021).

The premium for the animal welfare category of labels was found to be positive and significant only

for the high-priced products, denoting that foods obtained through improved animal welfare standards

can be regarded as a niche market. Our findings corroborate those of Clark et al. (2017) confirming

that, although people exhibit much concern, the WTP for the animal welfare attribute is low. While

certain consumers want to receive food information, others do not and contingent upon this

information avoidance their WTP for certain labels may be small (Reisch et al., 2021).

The present work is not exempt from limitations. As it focuses on the WTP estimates stemming from

choice experiment studies, it relies on both non-hypothetical and hypothetical measures. Hypothetical

choice experiments are prone to hypothetical bias from their being not incentive compatible, thus

associated WTP estimates are upward biased as compared to estimates derived by real experiments.

Nonetheless, the empirical evidence suggests that the marginal WTPs are equivalent in hypothetical

and non-hypothetical contexts (see, for instance, Lusk and Schroeder, 2004; Taylor et al., 2010), and

choice experiments are the most adopted methods in studies on SLs (Lusk, 2018). Therefore, we

opted for WTP estimates from choice experiments to achieve a valid representation of the

phenomenon, while controlling for bias in results that may be induced by the use of multiple types of

experimental techniques (Clark et al., 2017; Printezis et al., 2019). In other terms, our setting is

controlled, homogenous and, thus, reliable.

20
The analysis is built upon the classification of SLs in categories that are not necessarily exclusive.

This drawback trades off the value of modelling a multifaceted topic. Being unaware of classifications

of wide use, we have synthesized the existing evidence and proposed plausible classification criteria,

which are not conclusive, but rather a starting point for future work.

Within the scope of our study’s limitations, we recognize that our study is rooted on the assumption

that consumers demand for full information. However, the modern, complex, and information-

abundant food choice environment may lead to scenarios in which consumers do not consistently

pursue comprehensive information on sustainability (Reisch et al., 2021). This insight suggests a

potential avenue for further investigation into the evolving consumer information preferences and

their economic implications.

Conclusions and Policy Implications

We analyse how consumers WTPs for SLs vary with the types of sustainability information provided

through labels, and, specifically, according to the broadness of formulation, the underlying

sustainability dimension, and the label category. Our analysis aims to synthesize the existing

knowledge and empirical evidence on the topic and provide generalisable conclusions.

Interestingly, generic informative content does not differ much as compared to specific formulation

of labels. This thought-provoking result stands out in the policy debate on the implementation of

harmonised and overarching labelling systems tackling sustainability, as is the case of the European

F2F Strategy. Developing a comprehensive SL ruled by the government could guarantee transparency

and increase consumer trust. Moreover, unambiguous communication would be effective to address

consumer criticisms, such as scepticism, lack of knowledge of SLs or information overload.

Nonetheless, further efforts are required to uncover how citizens would react to the introduction of

generic labels on sustainability. Future research should focus on how information may prompt people

to switch towards more sustainable purchasing patterns. For instance, it would be crucial to assess

21
under what circumstances the interpretability and identifiability of the generic (i.e., holistic)

information provided are guaranteed.

As for the sustainability dimensions, the social and ethical labels are less considered than the

environmental-related counterparts. The potential inconsistencies between the perception of

sustainability and the official definition, which embraces both the environmental and social pillars,

may explain the tendency we observe in the literature. Awareness campaigns may help reducing this

mismatch and foster investments with positive social and ethical impacts.

A further contribution is our focus on the price dimension. We have highlighted how the information

on sustainability signalling the environmental outcomes of the product, the production methods or

animal welfare improvements may lead to higher premiums when provided on expensive foods rather

than low-priced products. The adoption of environmental and animal welfare certifications seems

more efficient in high-ended food categories, whereas the generic claim ‘sustainable product’ is more

appropriate on low-ended products.

To conclude, our results stressed that not all sustainability facets are equally important among

consumers at the point of purchase, although there is a growing concern for sustainability issues. The

economic rationale behind this well-known attitude-behaviour gap is that individuals are not

incentivised to pay a premium since their actions result in a minimum global outcome that cannot be

instantly perceived (Bonnet et al., 2020). The main limitation in relying on SLs to provide public

goods is the free-riding behaviour (Lusk et al., 2007). Consumers will likely purchase the cheaper

non-certified foods as long as they can free-ride off the purchasing of sustainable foods by others.

Due to these reasons, market-based interventions, such as food labels, should be regarded as a part of

the policy tool belt rather than the main instrument to drive the sustainable transition. An integrated

policy approach, combining hard and soft measures, is required to achieve complete sustainability in

the food system, i.e. to develop a food system that contributes to all the different pillars of

sustainability in a balanced manner (European Commission, 2020b). Labelling can support and

encourage informed choices and strengthen the demand for more sustainable alternatives, which is
22
particularly true in the case of already established preferences, such as the environmental-labelled

products.

Lastly, while the direct impact of SLs, such as the WTP, has been proven to exhibit significant

heterogeneity, the indirect impact of these labels on the agri-food value chain cannot be

underestimated. This influence may encompass dynamic shifts within the global food system as well

as alterations in marketing, trade, and cooperation strategies (e.g. Hoekman and Sabel, 2019;

Santeramo and Lamonaca, 2022). Notably, in the modern and globalized agri-food value chains the

role of services, and (among them) the use of product standards necessitate technology innovations

and upgrading by producers and processing firms (Barrett et al., 2022; Santeramo et al,. 2023), and .

These complex effects warrant further exploration, as they could hold valuable insights into the

evolving dynamics of sustainable product markets.

23
References

Akaichi, F., Revoredo-Giha, C., Glenk, K., & Gil, J. M. (2020). How Consumers in the UK and Spain Value the
Coexistence of the Claims Low Fat, Local, Organic and Low Greenhouse Gas Emissions. Nutrients, 12(120).
Aoki, K., Akai, K., Ujiie, K., Shimmura, T., & Nishino, N. (2019). The impact of information on taste ranking and
cultivation method on rice types that protect endangered birds in Japan: Non-hypothetical choice experiment with
tasting. Food Quality and Preference, 75(April 2018), 28–38.
Ardeshiri, A., & Rose, J. M. (2018). How Australian consumers value intrinsic and extrinsic attributes of beef products.
Food Quality and Preference, 65(October 2017), 146–163.
Asche, F., Larsen, T. A., Smith, M. D., Sogn-Grundvåg, G., & Young, J. A. (2015). Pricing of eco-labels with retailer
heterogeneity. Food Policy, 53, 82–93.
Aschemann-Witzel, J., & Zielke, S. (2017). Can’t Buy Me Green? A Review of Consumer Perceptions of and Behavior
Toward the Price of Organic Food. Journal of Consumer Affairs, 51(1), 211–251.
Asioli, D., Aschemann-Witzel, J., & Nayga, R. M. (2020). Sustainability-Related Food Labels. Annual Review of
Resource Economics, 12(1), 1–15.
Bangsa, A. B., & Schlegelmilch, B. B. (2020). Linking sustainable product attributes and consumer decision-making:
Insights from a systematic review. Journal of Cleaner Production, 245, 118902.
Barrett, C. B., Reardon, T., Swinnen, J., & Zilberman, D. (2022). Agri-food Value Chain Revolutions in Low-and Middle-
Income Countries. Journal of Economic Literature, 60(4), 1316–1377.
Bastounis, A., Buckell, J., Hartmann‐boyce, J., Cook, B., King, S., Potter, C., Bianchi, F., Rayner, M., & Jebb, S. A.
(2021). The impact of environmental sustainability labels on willingness‐to‐pay for foods: A systematic review and
meta‐analysis of discrete choice experiments. Nutrients, 13(8).
Bello, M., & Abdulai, A. (2016). Impact of ex-ante hypothetical bias mitigation methods on attribute non-attendance in
choice experiments. American Journal of Agricultural Economics, 98(5), 1486–1506.
Bhavsar, A., Diallo, C., & Ülkü, M. A. (2021). Towards sustainable development : Optimal pricing and sales strategies
for retailing fair trade products. Journal of Cleaner Production, 286, 124990.
Bonnet, C., Bouamra-Mechemache, Z., Réquillart, V., & Treich, N. (2020). Viewpoint: Regulating meat consumption to
improve health, the environment and animal welfare. Food Policy, 97, 101847.
Bougherara, D., & Combris, P. (2009). Eco-labelled food products: What are consumers paying for? European Review of
Agricultural Economics, 36(3), 321–341.
Burnier, P. C., Spers, E. E., & de Barcellos, M. D. (2021). Role of sustainability attributes and occasion matters in
determining consumers’ beef choice. Food Quality and Preference, 88, 104075.
Capitello, R., & Sirieix, L. (2020). What does ‘sustainable wine’mean? An investigation of french and italian wine
consumers. In S. L. Forbes, T.-A. De Silva, and A. J. Gilinsky (Eds.), Social Sustainability in the Global Wine
Industry (pp. 137–154). Palgrave Pivot, Cham.
Caputo, V., Nayga, R. M., & Scarpa, R. (2013). Food miles or carbon emissions? Exploring labelling preference for food
transport footprint with a stated choice study. Australian Journal of Agricultural and Resource Economics, 57(4),
465–482.
Charry, A., Narjes, M., Enciso, K., Peters, M., & Burkart, S. (2019). Sustainable intensification of beef production in
Colombia—Chances for product differentiation and price premiums. Agricultural and Food Economics, 7(1).
Clark, B., Stewart, G. B., Panzone, L. A., Kyriazakis, I., & Frewer, L. J. (2017). Citizens, consumers and farm animal
welfare: A meta-analysis of willingness-to-pay studies. Food Policy, 68, 112–127.
Contini, C., Boncinelli, F., Romano, C., Scozzafava, G., & Casini, L. (2019). Price vector issue in a choice experiment:
A methodological proposal. Food Quality and Preference, 75(February), 23–27.
Dahlhausen, J. L., Rungie, C., & Roosen, J. (2018). Value of labeling credence attributes—common structures and
individual preferences. Agricultural Economics (United Kingdom), 49(6), 741–751.
de Boer, J. (2003). Sustainability labelling schemes: The logic of their claims and their functions for stakeholders.
Business Strategy and the Environment, 12(4), 254–264.

24
Dendler, L. (2014). Sustainability Meta Labelling: An effective measure to facilitate more sustainable consumption and
production? Journal of Cleaner Production, 63, 74–83.
Deselnicu, O. C., Costanigro, M., Souza-Monteiro, D. M., & McFadden, D. T. (2013). A meta-analysis of geographical
indication food valuation studies: What drives the premium for origin-based labels?. Journal of Agricultural and
Resource Economics, 204-219.
Dolgopolova, I., & Teuber, R. (2018). Consumers’ willingness to pay for health benefits in food products: A meta-
analysis. Applied Economic Perspectives and Policy, 40(2), 333–352.
Ecolabel Index. (2022). The Ecolabel Index. Retrieved July 22, 2022, from https://www.ecolabelindex.com/
Elkington, J. (1997). Cannibals with forks: The triple bottom line of 21st century business. London: Capstone Press
Publication.
European Commission. (2020a). Communication from the Commission to the European Parliament, the Council, the
European Economic and Social Committee and the Committee of the Regions: A Farm to Fork Strategy for a fair,
healthy and environmentally-friendly food system. 5(1), 55.
European Commission. (2020b). Towards a Sustainable Food System. Group of Chief Scientific Advisors - Scientific
Opinion, (8).
FAO - Food and Agriculture Organisation of the United Nations. (2018). Sustainable food systems: concept and
framework. FAO, Rome.
Fitzsimmons, J., & Cicia, G. (2018). Different tubers for different consumers: Heterogeneity in human values and
willingness to pay for social outcomes of potato credence attributes. International Journal on Food System
Dynamics, 9(4), 354–374.
Fogt Jacobsen, L., Pedersen, S., & Thøgersen, J. (2022). Drivers of and barriers to consumers’ plastic packaging waste
avoidance and recycling – A systematic literature review. Waste Management, 141, 63–78.
Grebitus, C., Peschel, A. O., & Hughner, R. S. (2018). Voluntary food labeling: The additive effect of “free from” labels
and region of origin. Agribusiness, 34(4), 714–727.
Gross, S., Waldrop, M. E., & Roosen, J. (2021). How does animal welfare taste? Combining sensory and choice
experiments to evaluate willingness to pay for animal welfare pork. Food Quality and Preference, 87(March 2020),
104055.
Grunert, K. G., Hieke, S., & Wills, J. (2014). Sustainability labels on food products: Consumer motivation, understanding
and use. Food Policy, 44, 177–189.
Gusenbauer, M., & Haddaway, N. R. (2020). Which academic search systems are suitable for systematic reviews or meta-
analyses? Evaluating retrieval qualities of Google Scholar, PubMed, and 26 other resources. Research Synthesis
Methods, 11(2), 181–217.
Havránek, T., Stanley, T. D., Doucouliagos, H., Bom, P., Geyer-Klingeberg, J., Iwasaki, I., … van Aert, R. C. M. (2020).
Reporting Guidelines for Meta-Analysis in Economics. Journal of Economic Surveys, 34, 469–475.
Hensher, D. A., Rose, J. M., & Greene, W. H. (2015). Applied choice analysis. Cambridge university press.
Hinkes, C., & Schulze-Ehlers, B. (2018). Consumer attitudes and preferences towards pangasius and tilapia: The role of
sustainability certification and the country of origin. Appetite, 127(April), 171–181.
Hoekman, B., & Sabel, C. (2019). Open plurilateral agreements, international regulatory cooperation and the
WTO. Global Policy, 10(3), 297-312.Janssen, D., & Langen, N. (2017). The bunch of sustainability labels – Do
consumers differentiate? Journal of Cleaner Production, 143, 1233–1245.
Janssen, M., & Hamm, U. (2012). Product labelling in the market for organic food: Consumer preferences and
willingness-to-pay for different organic certification logos. Food Quality and Preference, 25(1), 9–22.
Katt, F., & Meixner, O. (2020). A systematic review of drivers influencing consumer willingness to pay for organic food.
Trends in Food Science and Technology, 100(April), 374–388.
Klaiman, K., Ortega, D. L., & Garnache, C. (2016). Consumer preferences and demand for packaging material and
recyclability. Resources, Conservation and Recycling, 115, 1–8.
Koistinen, L., Pouta, E., Heikkilä, J., Forsman-Hugg, S., Kotro, J., Mäkelä, J., & Niva, M. (2013). The impact of fat
content, production methods and carbon footprint information on consumer preferences for minced meat. Food
Quality and Preference, 29(2), 126–136.
Lernoud, J., Potts, J., Sampson, G., Schlatter, B., Huppe, G., Voora, V., … Dang, D. (2018). The state of sustainable
25
 markets – statistics and emerging trends 2018. Geneva: ITC. Geneva: ITC.
Li, S., & Kallas, Z. (2021). Meta-analysis of consumers ’ willingness to pay for sustainable food products. Appetite, 163,
105239.
Li, X., Jensen, K. L., Lambert, D. M., & Clark, C. D. (2018). Consequentiality beliefs and consumer valuation of extrinsic
attributes in beef. Journal of Agricultural and Applied Economics, 50(1), 1–26.
Lim, K. H., Hu, W., & Nayga, R. M. (2018). Is Marine Stewardship Council’s ecolabel a rising tide for all? Consumers’
willingness to pay for origin-differentiated ecolabeled canned tuna. Marine Policy, 96(October 2017), 18–26.
Louviere, J. J., Hensher, D. A., & Swait, J. D. (2000). Stated choice methods: analysis and applications. Cambridge
university press.
Lusk, J. L. (2018). Separating Myth from Reality: An Analysis of Socially Acceptable Credence Attributes. Annual
Review of Resource Economics, 10, 65–82.
Lusk, J. L. (2019). Consumer preferences for cage-free eggs and impacts of retailer pledges. Agribusiness, 35(2), 129–
148.
Lusk, J. L., Jamal, M., Kurlander, L., Roucan, M., & Taulman, L. (2005). A meta-analysis of genetically modified food
valuation studies. Journal of Agricultural and Resource Economics, 30(1), 28–44.
Lusk, J. L., Nilsson, T., & Foster, K. (2007). Public preferences and private choices: Effect of altruism and free riding on
demand for environmentally certified pork. Environmental and Resource Economics, 36(4), 499–521.
Lusk, J., & Schroeder, T. C. (2004). Are choice experiments incentive compatible? A test with quality differentiated beef
steaks. American Journal of Agricultural Economics, 86(2), 467–482.
Magnusson, M. K., Arvola, A., Hursti, U. K. K., Åberg, L., & Sjödén, P. O. (2003). Choice of organic foods is related to
perceived consequences for human health and to environmentally friendly behaviour. Appetite, 40(2), 109–117.
Mazzocchi, C., Ruggeri, G., & Corsi, S. (2019). Consumers’ preferences for biodiversity in vineyards: A choice
experiment on wine. Wine Economics and Policy, 8(2), 155–164. https://doi.org/10.1016/j.wep.2019.09.002
Meas, T., Hu, W., Batte, M. T., Woods, T. A., & Ernst, S. (2015). Substitutes or complements? Consumer preference for
localand organic food attributes. American Journal of Agricultural Economics, 97(4), 1044–1071.
Menozzi, D., Nguyen, T. T., Sogari, G., Taskov, D., Lucas, S., Castro-Rial, J. L. S., & Mora, C. (2020). Consumers’
preferences and willingness to pay for fish products with health and environmental labels: Evidence from five
european countries. Nutrients, 12(9), 1–22.
Meyerding, S. G. H., Gentz, M., Altmann, B., & Meier-Dinkel, L. (2018). Beef quality labels: A combination of sensory
acceptance test, stated willingness to pay, and choice-based conjoint analysis. Appetite, 127(May), 324–333.
MSC. (2020). Celebrating and supporting sustainable fisheries. The Marine Stewardship Council Annual Report 2019-
2020. Retrieved from https://www.msc.org/docs/default-source/default-document-library/about-the-msc/msc-
annual-report-2019-2020.pdf
Naald, B. Vander, & Cameron, T. A. (2011). Willingness to pay for other species’ well-being. Ecological Economics,
70(7), 1325–1335.
Noblet, C. L., & Teisl, M. F. (2015). Eco-labelling as sustainable consumption policy. In Handbook of Research on
Sustainable Consumption (pp. 300–312).
Ochs, D., Wolf, C. A., Widmar, N. O., Bir, C., & Lai, J. (2019). Hen housing system information effects on U.S. egg
demand. Food Policy, 87(May), 101743.
Olsen, S. B., & Meyerhoff, J. (2017). Will the alphabet soup of design criteria affect discrete choice experiment results?
European Review of Agricultural Economics, 44(2), 309–336.
Page, M.J., McKenzie, J.E., Bossuyt, P.M., Boutron, I., Hoffmann, T.C., Mulrow, C.D., … Moher, D. (2021). The
PRISMA 2020 statement: an updated guideline for reporting systematic reviews. International Journal of surgery,
88, 105906.
Piracci, G., Boncinelli, F., & Casini, L. (2022). Wine consumers ’ demand for social sustainability labeling : Evidence
for the fair labor claim. Applied Economic Perspectives and Policy, 1–20.
Printezis, I., & Grebitus, C. (2018). Marketing Channels for Local Food. Ecological Economics, 152(May), 161–171.
Printezis, I., Grebitus, C., & Hirsch, S. (2019). The price is right!? A meta-regression analysis on willingness to pay for
local food. PLoS ONE, 14(5), 1–23.

26
Reisch, L. A., Sunstein, C. R., & Kaiser, M. (2021). What do people want to know? Information avoidance and food
policy implications. Food Policy, 102, 102076.
Retail Forum for Sustainability. (2011). Labelling. Retrieved May 13, 2022, from Issue Paper n.7 website:
https://ec.europa.eu/environment/industry/retail/issue_papers.htm
Rigby, D., Burton, M., & Pluske, J. (2016). Preference Stability and Choice Consistency in Discrete Choice Experiments.
Environmental and Resource Economics, 65(2), 441–461.
Rondoni, A., & Grasso, S. (2021). Consumers behaviour towards carbon footprint labels on food: A review of the
literature and discussion of industry implications. Journal of Cleaner Production, 301, 127031.
Rousseau, S., & Vranken, L. (2013). Green market expansion by reducing information asymmetries: Evidence for labeled
organic food products. Food Policy, 40, 31–43.
Santeramo, F.G., Jelliffe, J. Hoekman, B. (2023) Manuscript Title: International Agri-food Value Chains: Exploring the
role of Trade and Services. RSCAS Working paper.
Santeramo, F. G., & Lamonaca, E. (2019). The Effects of Non-tariff Measures on Agri-food Trade: A Review and Meta-
analysis of Empirical Evidence. Journal of Agricultural Economics, 70(3), 595–617.
Santeramo, F. G., & Lamonaca, E. (2021). Objective risk and subjective risk: The role of information in food supply
chains. Food Research International, 139, 109962.
Santeramo, F. G., & Lamonaca, E. (2022). Standards and regulatory cooperation in regional trade agreements: What the
effects on trade?. Applied Economic Perspectives and Policy, 44(4), 1682-1701.
Scarpa, R., Thiene, M., & Train, K. (2008). Utility in willingness to pay space: A tool to address confounding random
scale effects in destination choice to the Alps. American Journal of Agricultural Economics, 90(4), 994–1010.
Schäufele, I., & Hamm, U. (2017). Consumers’ perceptions, preferences and willingness-to-pay for wine with
sustainability characteristics: A review. Journal of Cleaner Production, 147, 379–394.
Schleenbecker, R., & Hamm, U. (2013). Consumers’ perception of organic product characteristics. A review. Appetite,
71, 420–429.
Soley, G., Hu, W., & Vassalos, M. (2019). Willingness to Pay for Shrimp with Homegrown by Heroes, Community-
Supported Fishery, Best Aquaculture Practices, or Local Attributes. Journal of Agricultural and Applied
Economics, 51(4), 606–621.
Stanley, T. D. (2001). Wheat from chaff: Meta-analysis as quantitative literature review. Journal of Economic
Perspectives, 15(3), 131–150.
Stanley, T. D., Doucouliagos, H., Giles, M., Heckemeyer, J. H., Johnston, R. J., Laroche, P., … Rost, K. (2013). Meta-
analysis of economics research reporting guidelines. Journal of Economic Surveys, 27(2), 390–394.
Staples, A. J., Reeling, C. J., Widmar, N. J. O., & Lusk, J. L. (2020). Consumer willingness to pay for sustainability
attributes in beer: A choice experiment using eco-labels. Agribusiness, 36(4), 591–612.
Taylor, L. O., Morrison, M. D., & Boyle, K. J. (2010). Exchange Rules and the Incentive Compatibility of Choice
Experiments. Environmental and Resource Economics, 47(2), 197–220.
Tebbe, E., & von Blanckenburg, K. (2018). Does willingness to pay increase with the number and strictness of
sustainability labels? Agricultural Economics (United Kingdom), 49(1), 41–53.
Teisl, M. F., & Roe, B. (1998). The Economics of Labeling: An Overview of Issues for Health and Environmental
Disclosure. Agricultural and Resource Economics Review, 27, 140–150.
Torma, G., & Thøgersen, J. (2021). A systematic literature review on meta sustainability labeling – What do we (not)
know? Journal of Cleaner Production, 293, 126194.
Train, K., & Weeks, M. (2005). Discrete Choice Models in Preference Space and Willingness-to-Pay Space. In R. Scarpa
& A. Alberini (Eds.), Applications of Simulation Methods in Environmental and Resource Economics (pp. 1–16).
Van Loo, E. J., Caputo, V., Nayga, R. M., & Verbeke, W. (2014). Consumers’ valuation of sustainability labels on meat.
Food Policy, 49(P1), 137–150.
Van Loo, E. J., Hoefkens, C., & Verbeke, W. (2017). Healthy, sustainable and plant-based eating: Perceived (mis)match
and involvement-based consumer segments as targets for future policy. Food Policy, 69, 46–57.
Viegas, I., Nunes, L. C., Madureira, L., Fontes, M. A., & Santos, J. L. (2014). Beef credence attributes: Implications of
substitution effects on consumers’ WTP. Journal of Agricultural Economics, 65(3), 600–615.

27
WCED - World Commission on Environment and Development. (1987). Our common future. Oxford University Press.
Willer, H., Trávníček, J., Meier, C., & Schlatter, B. (2021). The World of Organic Agriculture. Statistics and Emerging
Trends 2021. In The World of Organic Agriculture. Retrieved from
https://shop.fibl.org/de/artikel/c/statistik/p/1663-organic-world-2015.html
Wu, X., Hu, B., & Xiong, J. (2020). Understanding Heterogeneous Consumer Preferences in Chinese Milk Markets: A
Latent Class Approach. Journal of Agricultural Economics, 71(1), 184–198.
Xu, P., Su, H., & Lone, T. (2018). Chinese consumers’ willingness to pay for rice. Journal of Agribusiness in Developing
and Emerging Economies, 8(2), 256–269.

28
 Tables

Table 1. Classifications of SLs according to different criteria.

Partitioning Criterion Classes of SLs Examples

Broadness of Formulation (type) Specific Organic, carbon footprint, fair-trade
Generic Sustainable product, eco-friendly, green food
Sustainability Dimension (dimension) Environmental Pesticide-free, biodiversity friend, biodynamic farming
Social Fair-trade, animal welfare, free-range
Environmental and Social Organic*, rainforest alliance
Scope and Information Provided (category) Environmental Impact Carbon footprint, water footprint, eco-friendly
Organic, Marine Stewardship Council, Aquaculture
Production Method Stewardship Council
Animal Welfare Animal welfare, free-range, grass-fed
Social Fairness Fairtrade, workers’ respect, cause-related marketing
Sustainable Production Sustainable product, made from sustainable agriculture
*If applied to animal-based foods.

29
Table 2. Description and summary statistics of the variables included in the model.

Variables Type Description Mean Std. Dev.

Effect measure
WTP Index Continuous Percent variation in the willingness to pay associated with the presence of the SL. Range [-0.94; 5.47] 0.27 0.44
Label-related information (X)
Specific Dummy 1 if the label is specific, 0 if generic 0.93 0.26
Environmental Dummy 1 if the label is related to the environmental dimension of sustainability, 0 otherwise 0.50 0.50
Sociala Dummy 1 if the label is related to the social dimension of sustainability, 0 otherwise 0.28 0.45
Both Dimensions Dummy 1 if the label is related to both the environmental and social dimensions of sustainability, 0 otherwise 0.22 0.41
Environmental Impact Dummy 1 if the label informs about the environmental impacts of the product, 0 otherwise 0.17 0.38
Production Method Dummy 1 if the label informs about the production method of the product, 0 otherwise 0.51 0.50
Animal Welfare Dummy 1 if the label informs about the animal welfare respect of the product, 0 otherwise 0.24 0.43
Social Fairnessa Dummy 1 if the label informs about the positive outcome on social fairness of the product, 0 otherwise 0.04 0.20
Sustainable Production Dummy 1 if the label informs about the overall sustainability of the product, 0 otherwise 0.03 0.17
Potential Label Dummy 1 if the label was fictitiously implemented for the study purposes, 0 if existing 0.30 0.46
Structural characteristics (K)
Beverages Dummy 1 if the food belongs to the beverage category; 0 otherwise 0.02 0.15
Coffee Dummy 1 if the food belongs to the coffee category; 0 otherwise 0.03 0.17
Dairy Dummy 1 if the food belongs to the dairy category; 0 otherwise 0.04 0.19
Egg Dummy 1 if the food belongs to the egg category; 0 otherwise 0.13 0.33
Fruit Dummy 1 if the food belongs to the fruit category; 0 otherwise 0.14 0.35
Meat Dummy 1 if the food belongs to the meat category; 0 otherwise 0.33 0.47
Seafood Dummy 1 if the food belongs to the seafood category; 0 otherwise 0.06 0.24
Vegetable Dummy 1 if the food belongs to the vegetable category; 0 otherwise 0.16 0.36
Baby food Dummy 1 if the food belongs to the baby food category; 0 otherwise 0.03 0.16
Other categoriesa,b Dummy 1 if the food belongs to the other less represented food categories; 0 otherwise 0.07 0.26
Food quantity Dummy 1 if the food quantity is high; 0 if normal 0.44 0.50
Asia Dummy 1 if the region of the study is Asia; 0 otherwise 0.09 0.28
Africa Dummy 1 if the region of the study is Africa; 0 otherwise 0.02 0.13
Australia Dummy 1 if the region of the study is Australia; 0 otherwise 0.02 0.16
Europe Dummy 1 if the region of the study is Europe; 0 otherwise 0.57 0.50
30
 Variables Type Description Mean Std. Dev.
North America c Dummy 1 if the region of the study is North America; 0 otherwise 0.29 0.45
South Americaa Dummy 1 if the region of the study is South America; 0 otherwise 0.01 0.11
Consumers Dummy 1 if the sampled population in the study was consumers; 0 otherwise 0.14 0.35
Purchasers Dummy 1 if the sampled population in the study was purchasers; 0 otherwise 0.53 0.50
Consumers and Purchasers Dummy 1 if the sampled population in the study was consumers and purchasers; 0 otherwise 0.04 0.20
Residents Dummy 1 if the sampled population in the study was residents; 0 otherwise 0.22 0.41
Academicsa Dummy 1 if the sampled population in the study was academics; 0 otherwise 0.01
Methodological and technical issues (Z)
Face-to-face interview Dummy 1 if the choice experiment was administered through face-to-face interview; 0 otherwise 0.17 0.37
Online survey Dummy 1 if the choice experiment was administered through online survey; 0 otherwise 0.60 0.49
Field experiment c Dummy 1 if the choice experiment was administered through field experiment; 0 otherwise 0.20 0.40
a
Lab experiment Dummy 1 if the choice experiment was administered through laboratory experiment; 0 otherwise 0.03 0.17
Efficient design Dummy 1 if the experimental design was efficient; 0 otherwise 0.33 0.47
OOD design Dummy 1 if the experimental design was optimally orthogonal in the difference (OOD); 0 otherwise 0.15 0.35
Bayesian design Dummy 1 if the experimental design was Bayesian efficient; 0 otherwise 0.15 0.35
Orthogonal design Dummy 1 if the experimental design was orthogonal fractional factorial; 0 otherwise 0.26 0.44
a
Other designs Dummy 1 if the experimental design was fractional factorial, full factorial, random or saturated; 0 otherwise 0.08 0.28
Alternatives Continuous Total number of alternatives in the experimental design. Range [2; 8] 3.60 0.98
Attributes Continuous Total number of attributes included in the experimental design. Range [1; 12] 4.15 2.03
Publication process information (T)
Before 2012 Dummy 1 if the paper was published before 2012, 0 otherwise 0.10 0.30
Before 2018 Dummy 1 if the paper was published before 2018, 0 otherwise 0.62 0.49
th
Q1 Dummy 1 if the journal is in the 25 percentiles (Q1) of the SJR, 0 otherwise 0.47 0.50
Q2 Dummy 1 if the journal is in the 50th percentiles (Q2) of the SJR, 0 otherwise 0.47 0.50
a th
Q3 Dummy 1 if the journal is in the 75 percentiles (Q3) of the SJR, 0 otherwise 0.05 0.22
EEF Dummy 1 if the subject area of the journal is Economics, Econometrics and Finance (EEF), 0 otherwise 0.75 0.43
ABS Dummy 1 if the subject area of the journal is Agricultural and Biological Sciences (ABS), 0 otherwise 0.77 0.42
SS Dummy 1 if the subject area of the journal is Social Sciences (SS), 0 otherwise 0.19 0.39
Notes: The mean for dummy variables represents the share of observations equal to 1. Std. Dev.: Standard Deviation. a set as the baseline in the model. b food categories with less
than 10 observations, which are honey, chocolate, pizza, pasta and bread, flour and milling products, algae-based foods, oils and fats, jam, cereals and ready-to-eat meals. c dropped
in the model due to multicollinearity.

31
Table 3. Meta-regression results.

Type Dimension Category

Variables
(i) (ii) (iii)
Specific 0.04
[0.38]
Environmental Dimension 0.19*
[1.84]
Both Dimensions 0.15*
[1.75]
Environmental Impact 0.29*
[1.70]
Production Method 0.30*
[1.91]
Animal Welfare 0.15
[1.07]
Sustainable Production 0.25
[1.22]
Constant 0.77* 0.60 0.53
[1.74] [1.51] [1.24]
Observations 648 648 648
R-squared 0.19 0.20 0.21
Notes: Ordinary Least Square (OLS) estimation of equation (2). The dependent variable is the index of willingness to
pay (WTP) associated with the presence of the sustainability labels (SLs). The control factors (variables for structural
characteristics, methodological and technical issues, publication process information) are included in all specifications
and omitted for brevity. The full set of results can be found in Annex E (Tables E2 to E4). The base category is generic
SL in specification (i), social dimension in specification (ii) and social fairness in specification (iii). The t-statistics are in
brackets. * Significant at the 10% level.

32
 Table 4. Descriptive statistics of WTP Index by label characteristics.

Obs. (n) Obs. (%) Mean Std. Dev C.V. Min. Max. Median
Total 648 100 0.27 0.44 1.63 -0.94 5.47 0.19
Label Characteristics
Type Specific 601 93 0.27 0.45 1.67 -0.94 5.47 0.19
Generic 47 7 0.31 0.33 1.06 0.00 1.65 0.20
Dimension Environmental 326 50 0.33 0.53 1.61 -0.47 5.47 0.21
Social 181 28 0.20 0.34 1.70 -0.94 2.88 0.19
Both Dimensions 141 22 0.24 0.30 1.25 -0.08 1.65 0.14
Category Environmental Impact 113 17 0.34 0.57 1.68 -0.04 5.47 0.22
Production Method 333 51 0.29 0.44 1.52 -0.47 3.28 0.18
Animal Welfare 155 24 0.21 0.36 1.71 -0.94 2.88 0.18
Social Fairness 27 4 0.16 0.22 1.38 -0.33 0.50 0.22
Sustainable Production 20 3 0.27 0.40 1.48 0.03 1.65 0.10
Notes: Obs: observations; Std. Dev.: Standard Deviation; C.V.: Coefficient of variation; Min.: Minimum; Max.:
Maximum.

33
Table 5. Meta-regression results by product price range.

Average reference price by currency

Variables Low priced High priced
Environmental Impact 0.22* 0.74**
[1.82] [2.13]
Production Method 0.17 1.03***
[1.35] [2.75]
Animal Welfare 0.03 0.86**
[0.16] [2.27]
Social Fairness -0.20 0.45
[-0.84] [1.03]
Constant 0.69* -1.55
[1.75] [-1.46]
Observations 485 163
R-squared 0.32 0.69
Notes: Ordinary Least Square (OLS) estimation of equation (2). The dependent
variable is the index of willingness to pay (WTP) associated with the presence of the
sustainability labels (SLs). The control factors (variables for structural characteristics,
methodological and technical issues, publication process information) are included in
all specifications and omitted for brevity. The full set of results can be found in Annex
E (Table E5). Products are classified as low or high priced according to the average
reference price by currency. The t-statistics are in brackets. *, **, *** Significant at
the 10, 5 and 1 percent level, respectively.

34
Figures

Figure 1. The PRISMA flow diagram.

35
Figure 2. Distribution of the WTP index and average WTP by category.

Notes: The included values of the WTP index range from the 5th to the
95th percentiles. SF denotes the average WTP for labels in the social
fairness category; AW indicates the average WTP for labels in the
animal welfare category; SP signals the average WTP for labels in the
sustainable production category; PM refers to the average WTP for
labels in the production method category; EI represents the average
WTP for labels in the environmental protection category.

36