Around the world, governments and businesses are talking more and more about the need to move from today’s “take, make, waste” economy to a circular one, where products are designed to last, materials stay in use, and waste is dramatically reduced. On paper, the case is compelling: recent assessments show that shifting to a circular economy offers both a major climate opportunity and a significant economic one. A study from the European Commission’s Joint Research Centre finds that “reduction, reuse and recovery” measures could cut Europe’s heavy industrial emissions by up to 231 million tonnes of CO₂ each year, and global analyses estimate that circular models could generate around $4.5 trillion in value by 2030).

Yet progress may be stalling. The latest Circularity Gap Report shows that the share of secondary materials in the global economy fell from 9.1% in 2015 to 6.9% in 2021. Instead of becoming more circular, the world, in a recent timeframe, became less so.

What would actually help stakeholders such as consumers, companies and governments embrace circular models? In our recent research project, we reviewed more than 130 studies on circular business models to understand this very question. What we found is simple but often overlooked: circularity is not just a design or engineering challenge, it is also an engagement one. If consumers hesitate, or companies delay investments, or policymakers fail to create the right conditions, the circular shift stalls.

Our work identifies 6 practices that can boost circular economy engagement. They fall into 3 areas: helping stakeholders feel motivated, giving them opportunities, and ensuring they are able to act. Understanding these levers is key to accelerating the transition to a circular economy.

Motivation: making the case for going circular

For circular behaviour to emerge, stakeholders first need a clear reason to care. Motivation is about creating the desire to act by explaining why circular options matter, how they are beneficial and why they are worth choosing over familiar linear habits.

A first part of this involves strategic signalling: making the benefits of circular models visible, concrete and easy to grasp. Many companies now deliberately make a point of doing this. Mud Jeans, for example, communicates the exact water and CO₂ savings associated with its “Lease A Jeans” model, helping customers immediately see the environmental value of extending product life. Fairphone similarly signals the impact of modular design by showing how repairable smartphones reduce e-waste and keep devices in use for longer.

But motivation also depends on convincing stakeholders that circular options are safe, reliable and worthwhile. Even when people like the idea of circularity, they may still worry about the quality or convenience of second-hand or refurbished products. Companies are responding by offering guarantees, services and financial incentives that lower perceived risks. Decathlon, for instance, promotes its repair services and spare-part availability, reassuring customers that products can be kept in good condition for longer.

Opportunity: making circularity possible and socially acceptable

Even highly motivated stakeholders cannot engage in circular behaviour if the environment around them makes it difficult or uncommon. Opportunity is about creating the partnerships, norms and systems that make circular actions feasible, convenient and socially accepted.

“Matching” is a key part of this, in other words connecting the right stakeholders so that circular solutions can function. Few organisations can operate reuse, repair or recycling systems single-handedly; they need logistics partners, refurbishment specialists and intermediaries that help keep materials in circulation. We are witnessing more and more of these typically well-thought out matches. The fashion platform Vestiaire Collective, for example, collaborates with brands to authenticate and resell pre-owned items, creating a trusted ecosystem that individual firms would struggle to build alone. Cities such as Amsterdam foster circular procurement networks that bring together suppliers, waste operators, innovators, and citizens to jointly develop reuse and refurbishment pathways.

Opportunity also depends on legitimising circular practices, which makes them appear commonplace, expected and in line with broader societal rules. Governments play a central role here through standards and regulations. The European Union’s Right to Repair legislation, for instance, requires manufacturers to make spare parts and repair information available for many household products. This reinforces the idea that repairing rather than replacing is the default. Companies contribute to legitimising as well. When global brands like Apple promote refurbished devices as high-quality options and expand their certified repair networks, they help shift expectations about what counts as new or desirable.

Ability: giving stakeholders the capacity to act

“Closing the loop” also requires skills, knowledge and resources. Ability is about ensuring that stakeholders are equipped with the funding, infrastructure, education or practical support that make circular actions realistic in everyday life.

A first part of this involves supporting stakeholders with the resources they need. Many organisations and individuals want to engage in circular behaviour but simply lack the means. Companies may need finance to redesign products or set up reverse logistics. Households may need convenient places to return used goods. Cities may require infrastructure that enables citizens and organisations to share goods and materials. Increasingly, these needs are being addressed. The European Investment Bank, for example, has issued dedicated circular economy loans that help firms invest in recycling capacities, keep goods and materials in use, and design out waste. Startups such as Too Good To Go provide digital infrastructure that connects retailers with customers to reduce food waste, making it easier for small businesses to participate in circular practices without building new systems from scratch.

Ability also depends on empowering stakeholders with knowledge and skills to navigate circular models. Circularity requires understanding how products can be repaired, how materials flow through a system and how to be renters, sharers or repairers rather than one-time users. Education and training help build this understanding. Repair cafés, which have grown across Europe, offer hands-on opportunities for people to learn how to fix household items alongside volunteers. Many universities now provide open-access courses on circular design principles, giving students and professionals the tools to rethink production and consumption. These initiatives can help shift circularity from a niche practice to an accessible, everyday one.

A systemic shift requires all 6 practices – not just one

What becomes clear from our research is that isolated efforts rarely work. The studies we reviewed suggest that clear communication about the benefits of circular options may have little impact if people are not reassured that these options are reliable and worthwhile. Incentives or guarantees alone may fall short when companies lack the partners needed to run repair, reuse or return systems. Even well-designed collaborations may struggle to gain traction when circular behaviour is not supported by social norms or policy signals that make it feel like the normal thing to do. And investments in new infrastructure or funding may have limited effect if stakeholders do not have the knowledge or confidence to use circular services in practice. Progress is most likely to occur when all of these elements reinforce one another.

For all stakeholders, the key is to reflect on what is needed to make circular practices part of everyday life. This includes asking questions such as:

• How do we help those we interact with understand the value of circular choices?

• How can we collaborate to create systems that make it workable to share, repair and reuse?

• Do we, and those around us, have the infrastructure, resources and knowledge to participate in the circular transition with confidence?

Recognising these shared responsibilities and needs helps ensure that consumers, companies and governments move forward together, rather than in isolation, which is essential for a successful circular transition.

A weekly e-mail in English featuring expertise from scholars and researchers. It provides an introduction to the diversity of research coming out of the continent and considers some of the key issues facing European countries. Get the newsletter!

Les auteurs ne travaillent pas, ne conseillent pas, ne possèdent pas de parts, ne reçoivent pas de fonds d'une organisation qui pourrait tirer profit de cet article, et n'ont déclaré aucune autre affiliation que leur organisme de recherche.

Mixed anxiety-depressive disorders (MADD) and suicidal thoughts, online bullying, poorer self-esteem, alcohol, cannabis and psychoactive substance use… social networks exploit young people’s vulnerability and actually help boost certain disorders that they are prone to.

This is the conclusion of a large-scale report by the French Agency for Food, Environmental and Occupational Health & Safety (Anses) which dissects the mechanisms behind digital marketing tools designed to target the specific vulnerabilities and emotional weak spots associated with adolescence.

Olivia Roth-Delgado and Thomas Bayeux are part of the team at Anses coordinating the research project. They offered to present the main teachings from this latest report.

The Conversation: What makes Anses’ “The Effects of Social media use on teenagers’ mental health” report unprecedented?

Olivia Roth-Delgado: This expert appraisal is the result of five years of research and over 1 000 sought-after articles. It is unprecedented in its originality and the extensive nature of the work that are, in our knowledge, unequalled as far as public authorities such as Anses are concerned.

For the first time, certain mechanisms pointing to the ways in which social networks operate are being linked to effects impacting the health of adolescents. These mechanisms are known as dark patterns (editor’s note: they are designed to capture users’ attention and monetise it, and they come in several forms which aren’t necessarily limited to social media. Some are also used by online shopping sites, for example).

Adolescence is a vulnerable time because the brain is still maturing. During this phase, teenage boys and girls experience changes in the way they process and handle their emotions in the reward-related circuits of the brain. They are also more sensitive to social context, which can favour risk-taking behaviour when around their peers. It is also a time of heightened vulnerability for mental health disorders.

Thomas Bayeux: During adolescence, a culture encouraging confrontation with others, an appetite for communication and character building, consisting of testing social norms develops. All of these arguments lead us to the 11-17 age group at which these dispositions occur.

Anses’ mission as a public health authority is to assess health risks. That said, in the chapters of the study on practices and maintaining inter-generational relations, the expert review raises the potential positive effects of social media and the motivations encouraging engagement during adolescence.

The report suggests particularly concerning social media-related effects among adolescents including anxiety-depressive disorders, suicidal thoughts or self-harming. What mechanisms are at play?

O.R.-D.: Among the mechanisms we have highlighted and studied featured misleading (or mainpulative) interfaces as well as algorithms that produce personalised content. They all equate to attention-grabbing that keep social media users engaged, by offering them increasingly well-targeted or extreme content.

If a teenage boy or girl for example, searches “self-harm” once, this kind of content will be offered repeatedly and can trap them in a negative spiral.

T.B.: Capturing attention serves the business model that supports these online platforms. It gives them access to a large amount of data which it can capitalise on while equally contributing to ad space sales.

Online platforms have everything to gain from keeping people engaged using the two strategies we have outlined : on one hand, by providing personalised content using increasingly productive algorithms which ensnare users in an information loop, and on the other hand, by highlighting the most impactful content.

Dark patterns roll out familiar techniques such as likes, notifications, scrolling, reels on auto play, etc. Also known as “deceptive design patterns”, these user interfaces have been carefully crafted to trick people into doing things they wouldn’t do otherwise.

The adolescent phase greatly resonates with these “push strategies” that social media implement. At Anses, we are seeing major public health challenges as supply and demand meet, so to speak. The cocktail they produce is potentially explosive !

Where mental health-related disorders are concerned but also, bullying, and alcohol, tobacco, cannabis use along with other risk-taking behaviours that you are safeguarding against, are social networks boosting pre-existing phenomena?

O.R.-D.: Absolutely. Social networks constitute a social space. They offer a sounding board for problems that are present in society, gender stereotypes or encouraging drug use, etc.

T.B.: Social networks contribute to adolescent socialisation and social construction, they provide continuity with the world offline, encompassing both its good points and its flaws. There is no watertight barrier between what happens offline and what happens on social media.

Should the existing rules for protecting minors in society extend to social media?

O.R.-D.: This is actually the founding principle of the Digital Services Act. The European regulatory framework for digital services seeks to vet online content on very large platforms, in line with the following ethos : “What is illegal offline, is illegal online.

T.B.: This preoccupation motivates one of the key recommendations to emerge from the Anses report, which is that users under 18 can only access social networks designed and configured for protecting minors. Our intention is not for social media to be eradicated all together. But for technical solutions to be put into place to make social media a safe place for teenagers, and Anses urges platforms to become accountable in this respect.

Going forward, teenagers then discussing their social media habits with their peers, parents, teaching staff or youth workers could prove to be a very good thing. That said, it doesn’t let public authorities and online platforms ‘off the hook’ where adopting collective strategies to make social media a safe space for teenagers are concerned.

The report shows links between social media use with some disorders, without really establishing a cause-effect relationship between the two. Why is this?

O.R.-D.: The cause-effect subject remains a thorny one. It is fair to say that the expert appraisal that we are basing ourselves on is very dense and documented. Our methodology is solid, but it isn’t backed up by a "body of evidence”. That said, we can vouch for strong associations between social media use and the disorders we have mentioned for which we explicitly highlight the underlying mechanisms at work.

In relation to sleep, for example several factors are involved. When teenagers go on social media at night before bedtime, the exposure to digital blue light from screens can prolong the time it takes to fall asleep, because by stimulating our cognitive alertness, it shortens sleep duration. The long-term effects of chronic sleep deprivation on mental and physical health are well-documented. Add to that the fact that the emotional stimuli involved in going on social media can also prevent sleep. We are seeing that there is a host of proof to support this. But the concrete effects of social media on sleep in teenage boys and girls also depends on their practices.

Similarly, in the event of anxiety-depressive disorders or suicidal thoughts, the type of content on offer plays a major role. The two-way street factor must also come into consideration. Allow me to explain: an adolescent boy or girl who is already psychologically fragile are more likely to go on social media. Content design algorithms pick up on their emotional weaknesses and suggest emotionally-charged content. And this is precisely how teenagers get trapped in a negative spiral. Proving that there is a cause-effect relationship associated with feedback loops and bidirectional effects is however, far more complicated.

And as for social media’s impact on self-image, we also have a convincing amount of evidence demonstrating the same type of mechanisms based on repeated exposure to content that glorifies muscular men and thin women.

Girls seem more sensitive to the negative effects of social media than boys. What is this down to?

T.B.: This is one of the key takeaways of the report. Girls clearly represent a highly vulnerable segment on social media as far as health risks are concerned, and not just concerning how it impacts self-image. More girls than boys on social media are being bullied, and becoming victims of gender shaming, and social pressure… Girls pay more attention to what happens on social media, and comments that are posted.

LGBTQIA+ communities also represent a high-risk segment on social media. They are more likely to become victims of online bullying which is one of the associated health hazards, particularly mental health.

The report from Anses mentions that the amount of time spent on social media is not the only factor that should be considered.

T.B.: Time of use is helpful, but that alone isn’t enough to fully grasp the subject. Knowing how long users spend on social media allows us to study certain health factors like sedentariness, despite the growing number of digital nomad tools out there for connecting to social media. Quantifying the amount of time users engage also turns out to be precious in the case of late-night social media use, which is likely to affect sleep, for example.

However, we also know that understanding social media practices is essential for studying some of the related health side effects. It is important to know what you can do on social media : publish, like, read comments, retouch photos, for instance and the emotional attachment involved. It’s not about opposing different approaches, but aiming for complimentary.

Your report is based on a research project that fails to address, or barely addresses the impact of the very latest digital tools such as TikTok or AI chatbots. Can we assume that these new technologies increase mental health risks for teenagers as well?

O.R.-D.: The Anses’ expert appraisal draws on over a thousand articles mainly published between 2011 and 2021. Due to the time accumulated and spent researching and bringing the appraisal together, the technologies our studies focused on have naturally evolved. That said, we based ourselves on a common core of mechanisms, like deceptive user interfaces (dark patterns) and content personalisation algorithms that are related to health risks.

Therefore, our conclusions and recommendations can be applied to more recent social media. As for the question of artificial intelligence and AI chatbots, Anses recommends that the subject becomes the focus of future reports.

A weekly email in English featuring expertise from scholars and researchers. It provides an introduction to the diversity of research coming out of the continent and considers some of the key issues facing European countries. Get the newsletter!

In your recommendations, you suggest getting teenagers involved in risk-prevention programs.

O.R.-D.: Anses offers young people the opportunity to get onboard with our research, because they know best what motivates them to engage with social media as they are the ones creating and spreading new ways of using social media. This makes including them in discussions and boundary-setting with parents and teachers, all the more important. This will make them more inclined to follow the rules that they actually had a hand in making. Among the recommendations, Anses mentioned the need to promote forums in which young people can share their online experiences.

T.B.: And again, let me remind you that Anses is not recommending banning social media all together, it suggests a complete overhaul of the way networks are designed so they do not harm the health of adolescents.

Interview by Health Journalists Lionel Cavicchioli and Victoire N’Sondé, The Conversation France.

Les auteurs ne travaillent pas, ne conseillent pas, ne possèdent pas de parts, ne reçoivent pas de fonds d'une organisation qui pourrait tirer profit de cet article, et n'ont déclaré aucune autre affiliation que leur organisme de recherche.

Malgré la montée de l’extrême droite, les enquêtes montrent que les Français restent très attachés à la démocratie. En revanche, l’insatisfaction vis-à-vis de son fonctionnement est grande. Le sentiment de déni de la souveraineté populaire est au cœur du vote populiste et de l’abstention.

Les Françaises et les Français seraient-ils en train de se lasser de la démocratie ? À écouter certains commentateurs, la réponse semble évidente. Montée de la défiance, abstention élevée, succès des partis de droite radicale populiste… Tout indiquerait un pays gagné par la tentation illibérale et autoritaire. Pourtant, les données disponibles livrent une image plus nuancée et, d’une certaine manière, plus rassurante.

Selon l’Enquête sociale européenne (ESS) de 2020, les citoyens restent en effet massivement attachés au régime démocratique. Une écrasante majorité juge important voire très important le fait de vivre en démocratie. Là où le bât blesse, c’est sur la pratique : près de la moitié des enquêtés appartiennent à la catégorie des démocrates insatisfaits, celles et ceux qui soutiennent la démocratie mais se disent mécontents de la façon dont elle fonctionne en France (figure 1).

Ces démocrates insatisfaits ne sont pas des antidémocrates. Ils expriment au contraire une forme d’exigence élevée : plus on attend de la démocratie, plus on est enclin à juger sévèrement son fonctionnement. Cette tension entre un attachement profond aux valeurs démocratiques et un regard critique sur la réalité institutionnelle organise de plus en plus la vie politique française.

Figure 1. Importance et satisfaction par rapport à la démocratie en France

La politisation du renouveau démocratique

Depuis une dizaine d’années, la question du renouveau démocratique est au cœur des mouvements sociaux. Nuit debout en 2016, les gilets jaunes à partir de 2018, les mobilisations contre le passe sanitaire en 2020 ou, plus récemment, la journée d’action Bloquons tout ont débordé la démocratie participative institutionnalisée. En demandant de « rendre la parole au peuple », ces mobilisations ont contribué à rouvrir le débat sur les règles du jeu démocratique : qui décide et selon quelles procédures ?

Les exécutifs ont tenté de répondre, au moins symboliquement, à cette démocratie d’interpellation. Le grand débat national et, surtout, les conventions citoyennes sur le climat, sur la fin de vie ou sur les temps de l’enfant ont consacré l’entrée des mini-publics délibératifs dans le répertoire institutionnel français. Mais ces dispositifs ont également montré les limites des innovations démocratiques lorsque les promesses de reprise intégrale des propositions citoyennes ne sont pas tenues.

La politisation du renouveau démocratique vient aussi des partis eux-mêmes. Les partis challengers se positionnent souvent en entrepreneurs de la cause démocratique. Les données Chapel Hill Expert Survey (CHES) montrent ainsi que La France insoumise (LFI) et le Rassemblement national (RN) sont les plus enclins à défendre l’idée selon laquelle « le peuple » devrait décider en dernière instance (figure 2). Bien sûr, les deux forces politiques ne mobilisent pas la souveraineté populaire de la même façon. Là où le RN privilégie l’appel direct au peuple plutôt que la médiation parlementaire pour faire pièce au prétendu laxisme des institutions et des élites, LFI promeut plutôt un approfondissement démocratique par l’effectivité et l’extension des droits.

Figure 2. Soutien à la souveraineté populaire

Il n’en reste pas moins que la démocratie libérale perd son caractère consensuel, le conflit partisan portant désormais sur le modèle lui-même, et non seulement sur sa mise en œuvre. Des visions alternatives gagnent en visibilité.

Pour corriger les dysfonctionnements d’un modèle élitiste fondé sur les élus et sur la représentation, faut-il un modèle technocratique qui donne le pouvoir aux experts ? un modèle participatif, qui s’appuie sur l’implication et la délibération des citoyens ? ou un modèle majoritariste, fondé sur l’exercice direct de la souveraineté populaire ? Surtout, comment combiner ces modèles, qui comportent chacun des avantages et des inconvénients, pour décider des fins de l’action publique ?

Comment les conceptions de la démocratie nourrissent l’insatisfaction

Tout comme les autres Européens, les Français n’investissent pas le mot-valise « démocratie » de façon univoque. L’ESS distingue trois grandes composantes : libérale (élections libres, protection des droits, indépendance des médias et de la justice), populaire (souveraineté du peuple via des référendums) et sociale (égalité et protection). L’ESS mesure à la fois l’importance accordée à chacune de ces composantes (les aspirations) et le jugement porté sur la manière dont elles sont réalisées en France (les évaluations).

En soustrayant les évaluations aux aspirations, on obtient des scores de déficit démocratique perçu : plus l’écart est grand, plus les attentes sont déçues. Les résultats sont éloquents.

Sur la démocratie libérale, les Français perçoivent un déficit, mais d’ampleur relativement modérée. C’est sur la souveraineté populaire et, surtout, sur la dimension sociale que les écarts sont les plus prononcés. Beaucoup, notamment parmi les démocrates insatisfaits, souhaitent une démocratie où les citoyens sont davantage appelés à trancher les grandes orientations politiques, et où l’État joue son rôle de protection contre les risques sociaux.

Quand les déficits démocratiques s’expriment dans les urnes

Comment ces déficits perçus se traduisent-ils dans les comportements électoraux ? Des modèles statistiques permettent de dessiner plusieurs profils.

Les personnes qui perçoivent un fort déficit en matière de démocratie libérale ont tendance à s’abstenir davantage ou à se tourner vers la gauche, qu’elle soit modérée ou radicale. En revanche, ce déficit n’alimente pas le vote pour la droite radicale populiste. Plus les attentes libérales sont déçues, plus la probabilité de voter pour le RN recule (figure 3A).

Il en va différemment pour le déficit de démocratie populaire. Les électeurs qui estiment que « le peuple » n’a pas suffisamment son mot à dire, que les décisions sont confisquées par les élites ou que les référendums manquent sont plus enclins à voter pour la gauche radicale ou pour le RN, mais aussi à s’abstenir. C’est ici que se loge le cœur du vote populiste et de l’abstention : non pas dans un rejet abstrait du système, mais dans la conviction que la souveraineté populaire est entravée (figure 3B).

Par ailleurs, le déficit de démocratie sociale ne profite pas à la droite radicale. Il tend même à réduire l’abstention et le vote RN, ainsi qu’à accroître le vote pour la gauche modérée. Autrement dit, celles et ceux qui jugent que la démocratie française ne tient pas ses promesses en matière de justice sociale sont loin de s’abstenir ou de voter mécaniquement pour les partis de droite radicale (figure 3C).

Figure 3. Probabilités de vote en fonction du déficit démocratique perçu.

Plus instructif encore, ces déficits démocratiques perçus ont un impact plus prononcé sur le vote que l’insatisfaction globale vis-à-vis de la démocratie. Être un démocrate insatisfait augmente bien les probabilités statistiques de voter pour des partis challengers ou de s’abstenir. Mais les attentes déçues vis-à-vis de la démocratie influencent encore plus fortement ces probabilités.

Au total, il est difficile de bien saisir les logiques électorales du soutien à la démocratie si l’on s’en tient au diagnostic d’une insatisfaction généralisée, sans se demander quelles conceptions de la démocratie motivent le vote protestataire et l’abstention.

C’est parce qu’ils restent profondément attachés aux composantes sociales et populaires de la démocratie que les Français sont des démocrates insatisfaits, enclins à sanctionner les partis de gouvernement ou à se détourner des urnes. L’insatisfaction vis-à-vis de la démocratie ne vient pas de citoyens hostiles à la démocratie libérale, mais de citoyens exigeants et porteurs d’autres manières de « faire démocratie ».

Cet article est tiré de l'ouvrage French Democracy in Distress. Challenges and Opportunities in French Politics (Palgrave Macmillan, 2025), sous la direction de Élodie Druez, Frédéric Gonthier, Camille Kelbel, Nonna Mayer, Felix-Christopher von Nostitz et Vincent Tiberj. Une conférence autour de cette publication est organisée à Sciences Po, le jeudi 29 janvier 2026, de 17 heures à 19 heures, 27 rue Saint-Guillaume, Paris (VIIᵉ).

Frédéric Gonthier a reçu des financements du programme Horizon Europe n°1010952237: TRUEDEM—Trust in European Democracies.