Stéphan Marette, Hugo Molina, Guillaume Hollard, Olivier Allais, Lucie Charbonnieras, Ghewa El Dora, Kristin Kiesel, Marie-Eve Laporte, Lydiane Nabec, Emmanuel Paroissien, Jérôme Pivard, Anne Saint-Eve, Andrea Vargas *
Table of contents
- Introduction
- The main results of the contributions to the workshop
- Conclusion and open questions
- Reference by order of presentations
Introduction
While claims and front-of-pack (FOP) labels are abundant in many countries, the effectiveness of these nutritional signals in promoting healthier and/or greener food choices remains a topic of debate. Although traffic-light systems and synthetic nutrition scores may provide valuable information to consumers, they also tend to multiply the sources of information available at the point of purchase, potentially increasing inattention and information overload.
To discuss the economic impacts of food labels and scores, the Workshop on Food Labels and Score-conomics was held on April 22-23, 2026 in Palaiseau: at the Campus-Agro Paris Saclay for PSAE on April 22 and at the Bâtiment d’Enseignement Mutualisé (BEM) for the CREST on April 23. France is particularly interesting for studying scores and labels, as the Nutri-score was introduced in 2017 and recent policy debates have focused on the possible introduction of a mandatory environmental score displayed on all on foods (Soler et al., 2023). Despite the growing use of such labels, many questions remain regarding their effects on consumers and firms.
The main results of the contributions to the workshop
Based on the order of presentations, results can be summarized as Following.
In the introduction, several questions related to the impact of food labels and scores were highlighted. First, assessing ex ante the marginal effect of a new label or score remains difficult, as its effectiveness strongly depends on the informational environment in which it is introduced. Second, although the Nutri-Score was introduced in 2017, its overall economic impact on food markets and consumer choices has yet to be fully assessed. A particularly important issue concerns the reformulation of processed food products following firms’ adoption of the Nutri-Score. Andrea Vargas (for Allais et al.) presented the impact of Nutri-Score labeling on nutritional reformulation studied by Oqali. Across the six product categories examined, they find that the use of the Nutri-Score has no significant effect on reformulation at the category level in most cases. At the subcategory level, however, the effects are mixed and sometimes even unfavorable, limiting the scope for broad generalizations.
In her keynote speech, Kristin Kiesel examined whether or not food labels and claims can correct biased consumer beliefs. She highlighted that several empirical studies (i) that utilize quasi-natural and lab-in-the-field experiments strongly suggest that purchase decisions are reference-dependent and experience-driven; (ii) that consumers strategically engage with newly available information; and that (iii) memory formation is endogenous. Heterogeneity in experiences, prior beliefs, and perceptions results in subjective information processing. Detected confirmation and disconfirmation biases ultimately lead to suboptimal decision-making at individual and societal levels, giving an active role to some regulatory interventions in a context of abundance of claims and labels.
Marie-Eve Laporte (et al.) presented a paper studying when and under what conditions food scanner apps encourage healthier eating and outperform front-of-pack nutrition labels. Their analysis combines four complementary studies: a quantitative content analysis of over 16,000 online reviews and a Survey of 86 respondents in an exploratory phase, along with two experiments that test the conceptual model. The findings identify three conditions where food scanner apps outperform front-of-pack labels: when the nutritonal quality score of a product is poor, when consumers distrust the dominant market players, and when mid-range brands exhibit average brand equity. These findings suggest that food scanner apps and front-of-pack labels are more complementary than substitutable, providing useful insights for policymakers and food manufacturers.
Ghewa El Dora (et al.) presented a study analyzing whether digital food-quality disclosure affects firms’incentives to reformulate products. Using Yuka panel data from 2018–2021, they show that firms do not strategically reformulate products to cross Nutri-Score grade thresholds. They develop a theoretical model demonstrating that information disclosure does not systematically improve quality, as belief updating can either strengthen incentives to improve products or intensify price competition. To capture the additional information provided by Yuka, they construct a “credence dimension,” defined as the gap between Nutri-Score and Yuka ratings. Empirically, moderately negative information from Yuka increases reformulation incentives, whereas very negative signals have weaker or even adverse effects. Together, these findings provide a possible explanation for the limited quality improvements induced by food labels.
Hugo Molina (et al.) presented an empirical study that leverages a field experiment in which the Nutri-Score label was introduced in three out of six canteens to assess its effect on diet quality. Using individual-level panel data on meal choices, they estimate a random coefficient logit model of demand for bundles and identify compensatory behavior (e.g., complementarity between healthy main dishes and unhealthy starters). Counterfactual experiments highlight that while the Nutri-Score increased demand for healthy products in the main dish category, it also led to higher consumption of unhealthy products in the starter category. This compensatory behavior substantially reduces the overall effect of the Nutri-Score on diet quality and highlights the importance of accounting for cross-category effects when evaluating food labeling policies.
Lydiane Nabec (et al.) talked about a study aiming at understanding when Eco-Score labelling (shown on the left of figure 1) could help consumers to choose greener food products. Across two complementary studies, they examine how Eco-Score labeling affects consumers’ Willingness-To-Pay (WTP) for packets of breakfast cereals, and how this is influenced by Nutri-Score labeling and market brand position. Their results show that Eco-Score labeling has a stronger effect on consumers’ WTP when the score is unfavorable rather than favorable. However, Eco-Score Labeling effect is weaker than the Nutri-Score effect, which is more familiar to French consumers. Overall, the effectiveness of Eco-Score labeling appears to be significantly moderated by brand effects.
in this workshop
Lucie Charbonnieras (et al.) showed a study examining consumer sensitivity to local, pesticide use and technological characteristics in food products, and whether these dimensions are perceived as complementary or conflicting. Local origin and reduced pesticide use exert significant and positive effects on willingness to pay (WTP), confirming consumers’ responsiveness to territorial and environmental production characteristics. In contrast, information related to genome editing and vertical farming générâtes mixed reactions, though without strong opposition. Local production emerges as especially valued: 71% of respondents consider important to have products grown in Île-de-France, and almost all believe this production should be expanded. Finally, although 80% of participants are unfamiliar with the “Product from Île-de-France” label (shown on Figure 1), 84% support reinforcing its promotion and development.
Jerôme Pivard (et al.) detailed a lab-in-the-field experiment in a university canteen where a green tray makes the organic vegetarian dish publicly visible to fellow students, manipulating the salience of social image concerns without restricting choice. Preliminary results show that the nudge’s effectiveness is heterogeneous: it is more pronounced among individuals with stronger image-oriented behaviors, while intrinsic environmental concern and baseline meat consumption are the dominant predictors of vegetarian choice regardless of treatment. Perceived descriptive norms of meat-eating have an unexpected positive effect on vegetarian choices, and post-experiment survey data suggest no compensatory rebound in meat consumption among vegetarian choosers, which remains to be investigated using administrative data on choices in student canteens.
Kristin Kiesel (and Hyunjung Lee) presented a contribution examining how consumers respond to multiple, and potentially redundant, labels or claims (e.g., non-GMO statements added to packages that already display a USDA organic seal). Using a unique scanner dataset combined with detailed labeling information, they show that consumers prefer products featuring multiple labels and claims over products with fewer or no such messages. However, they find no evidence that redundant claims increase sales. Their results also suggest that price promotions remain the most effective tool for capturing consumers’attention at the point of sale, as immediate and direct benefits tend to dominate information processing. They conclude by arguing that labels and claims cannot be analyzed in isolation and that the effectiveness of labeling policies and disclosure laws depends critically on how manufacturers and retailers strategically position and promote their products in response to consumers’ context-specific associations.
Emmanuel Paroissien (and Olivier Allais) investigated the influence of a point-of-purchase environmental label on food choices. Using a pre-registered field experiment in French workplace canteens, they combine survey data and revealed-preference purchase data to assess whether ecolabels affect environmental knowledge, stated intentions, and actual food choices. The experiment involved a two-month introduction of the label in three of six canteens, covering more than 10,000 treated individuals. By linking survey responses to observed purchases, the study examines the channels through which ecolabels may influence dietary behavior and contribute to more sustainable consumption. Their results also highlight the difficulty of changing food behaviors through information disclosure alone. Preliminary results suggest precise null effects on both knowledge and food choices.
Stéphan Marette (for Munier-Jolain et al.) focused on a controlled lab experiment, participants first blind-tasted four types of apples under similar conditions and indicated liking scores. Stepwise information on apples was then disclosed. Willingness to pay (WTP) was elicited after each step of the successive information disclosure on varieties, organic/conventional conditions and Pesti-Score signaling the number of treatments (see the right of figure 1). Information disclosure had a significant and relatively strong impact on WTP for apples. The green color of the Pesti-Score increased WTP similarly to the organic label, while the yellow and red colors negatively impacted WTP. These findings highlight the potenal of a mandatory Pesti-Score to change the behavior of consumers and therefore to support changes toward a more sustainable agriculture.
Conclusion and open questions
The workshop highlighted the diversity of methodologies used to evaluate the economic impacts of food labels and scores, ranging from laboratory and field experiments to scanner data analysis, structural demand estimation, and theoretical modeling. Taken together, the contributions emphasized that the effects of labels strongly depend on the informational and market environment in which they are introduced. Several papers also showed that consumers exhibit genuine interest in nutritional and environmental information, although the effectiveness of labels varies substanally across contexts, product categories, and consumer characteristics.
A central lesson emerging from the workshop is that labels and scores cannot be analyzed in isolation. Their effects interact with brand positioning, price promotions, social norms, digital applications, and the coexistence of multiple claims and informational devices. Moreover, several contributions pointed to limited or heterogeneous effects of labels on actual behavior and firms’ reformulation incentives. Evidence of compensatory consumption behaviors, strategic information processing, and information overload suggests that providing more information does not necessarily translate into healthier or greener choices.
These findings raise several open questions for future research. First, despite the widespread adoption of the Nutri-Score since 2017, its overall economic impact on food markets, consumer welfare, and firms’ product strategies remains insufficiently documented. Second, more evidence is needed on the long-term effects of labels, especially in contexts characterized by multiple and potentially redundant sources of information. Third, understanding how consumers process increasingly complex environmental and nutritional signals remains a major challenge, particularly when attention is limited and preferences are shaped by prior beliefs and experiences.
More broadly, the workshop highlighted the need to compare informational policies with alternative regulatory instruments. In contexts where information overload limits the effectieness of labels and scores, Pigouvian taxes or subsidies targeting nutritional or environmental externalities may sometimes constitute more effective tools. Other regulatory instruments, such as advertising restrictions or bans on the promotion of unhealthy products, may also play an important role. Determining the optimal combinaon of informational, price-based, and advertising policies therefore remains an important avenue for future research and public policy design.
Reference by order of presentations (speaker in bold)
Olivier Allais, Shakila Boyer, Johanna Calvarin, Blandine Giraudeau, Marine Landreau, Virginie Molina, Odeline Molle, Ghislaine Narayanane, Fanny Sadier, Cloé Tedeschi, Marianne Trichard, Andrea Vargas.
Exploring the impact of Nutri-Score labelling on nutritional reformulation.
Kristin Kiesel. Can Food Labels Correct Biased Consumer Beliefs? Analyzing Reference-Dependent Purchases and Experience-driven Memory Formation in Retail Experiments.
Cornudet C., Marie-Eve Laporte, Berger‐Remy F., Parguel B., Richet J. When Food Scanner Apps Outperform Front‐of‐Pack Nutrition Labels: A Conditional Process Model to Foster Healthier Food Choices in Times of Growing Distrust.
Guillaume Hollard, Ghewa El Dora, Olivier Bonroy. Rate My Product: The Effects of Digital Disclosure on Food Quality.
Olivier Allais, Emmanuel Paroissien, Hugo Molina. The Cross-Category Effects of Food Labels: A Canteen Experiment.
Lydiane Nabec, Eva Delacroix-Bastien, Stéphan Marette. When Front-of-Pack Eco-Labeling Steer consumers Towards Greener Food Choices: The Branding Effect.
Lucie Charbonnieras, Lydiane Nabec, Louis-Georges Soler, Anne Saint-Eve, Stéphan Marette. Are local and technological foods antinomic for consumers?
Jérôme Pivard, Olivier Allais and Stéphan Marette. Green Trays, Greener Meals? Social Image and the Choice of Organic and Vegetarian Food in a Student Experiment.
Kristin Kiesel, Hyunjung Lee. Strategic Redundancies in Sustainability Labels and Claims: Do They Inform or Mislead Consumers?
Olivier Allais, Emmanuel Paroissien. Does Environmental Labelling Change What Consumers Know, Want, and Choose? Evidence from a Workplace Canteen Experiment
Nicolas Munier-Jolain, Christophe Martin, Fabrice Martin, Stéphan Marette, Information about pesticide treatments and consumers behaviors: an experiment with apples and a new label prototype named Pesti-Score
Find out more
Soler et al. (2023). RAPPORT DU CONSEIL SCIENTIFIQUE DE L’AFFICHAGE ENVIRONNEMENTAL
* Notes de fin
Ghewa El Dora, Université Paris 8, 93200, Saint Denis, France.
Kristin Kiesel, Department of Agricultural and Resource Economics, University of California, Davis, California, USA.
Marie-Eve Laporte, Université Paris-Saclay, Réseaux Innovation Territoires et Mondialisation, 92330, Sceaux, France
Guillaume Hollard, Center for Research in Economics and Statistics, École Polytechnique, Institut Polytechnique de Paris, Palaiseau 91120, France
Stéphan Marette, Hugo Molina, Olivier Allais, Lucie Charbonnieras, Lydiane Nabec, Emmanuel Paroissien, Jérôme Pivard, Andrea Vargas, Université Paris-Saclay, INRAE, AgroParisTech, Paris-Saclay Applied Economics, 91120 Palaiseau, France.
Anne Saint-Eve, Université Paris-Saclay, AgroParisTech, INRAE, SayFood, 91120 Palaiseau, France.
