Europe is staring at a dual crisis it hasn’t managed to solve. House prices across Europe have risen 60 percent and rents 30 percent over the past 15 years, while the number of building permits has fallen 20 percent. The European Investment Bank estimates the EU currently needs 2.25 million additional housing units, roughly 50 percent more than is actually being built. And yet the buildings that do get built remain among the largest sources of greenhouse gas emissions on the continent.

Between 2010 and 2024, construction costs in the European Union rose by 56 percent, and the European Commission expects housing demand to grow by more than two million units per year.

The housing affordability crisis and the climate crisis are not two separate problems. They are one interlocked systemic failure, and Europe’s construction and real estate sector sits at the centre of both.

The challenge: three tensions, one industry

The Architecture, Engineering and Construction (AEC) sector has suffered four decades of productivity stagnation. Complex permitting regimes, fragmented governance, and an industry structure built around one-off projects have prevented it from delivering affordable, liveable, and sustainable homes at scale.

By late 2025, the supply of new housing units in the EU met only 50 percent of actual demand, compounded by soaring costs for labour and materials and a construction sector that has historically struggled with low innovation and productivity.

At the same time, buildings account for roughly 40 percent of Europe’s energy consumption and 36 percent of its CO₂ emissions.

The EU Green Deal, the Circular Economy Action Plan, and the EU taxonomy for sustainable activities are demanding deep decarbonisation – but as theWorld Economic Forum’s Reimagining Real Estate framework (2024) makes clear, technology and sustainability commitments alone are insufficient without a reconfiguration of who builds, who owns, and who governs the built environment. The WEF’s earlier Framework for the Future of Real Estate (2021) similarly warned that affordability and decarbonisation would only align if the industry fundamentally changed its business models and governance structures. Neither framework, however, mapped the concrete alternative pathways by which this transformation might actually unfold.

France is an example which shows how quickly Europe’s housing and climate goals can collide. On April 23 2026, the government announced a housing stimulus bill to accelerate construction, decentralise some decisions, and launch a third urban-renewal programme. Its most controversial proposal would allow F- and G-rated energy-inefficient homes back onto the rental market if owners commit to renovation within three years for houses and five years for apartment buildings. Under current rules, G-rated homes have been barred from new or renewed leases since 2025, with F-rated homes due to follow in 2028. The question is whether enforcement and finance will make renovation real.

Across Europe, governments are trying to expand supply without weakening climate targets. Spain has turned to industrialised construction, using EU funds to build social housing faster and cheaper, while also confronting tourist rentals and a small social-housing stock. Germany faces the opposite pressure: housing completions fell to a 13-year low in 2025, while earlier estimates put annual need at 320,000 apartments until 2030.

At EU level, the Affordable Housing Plan now links faster permitting, renovation and cost-efficient construction. Supply measures increasingly depend on whether governments can integrate affordability with decarbonisation targets.

Four plausible futures for 2040

To address this gap, we conducted a multi-year strategic foresight study with over 30 senior industry experts from across Europe, architects, developers, material suppliers, energy companies, and real estate services firms. Published in the journal Futures our study combines a horizon scan, impact-uncertainty analysis, and three rounds of expert workshops to construct four consistent scenarios for the European AEC industry by 2040.

The scenarios are not predictions. They are structured explorations of four plausible development pathways, each with a distinct logic for how decarbonisation, circularity, and housing affordability might interact under different governance arrangements.

In the first scenario, Giants rule the AEC industry, Big Tech firms and OEM-like construction companies dominate through data-driven, off-site, industrialised building. Homes become subscription services; platforms set the standards and productivity rises sharply. But affordability and tenant agency remain contested, and small firms struggle to survive.

In the second, the Circular Future: a coalition of regulators, financial institutions, and pioneering firms embeds circular principles into planning law, procurement, and finance. Buildings become documented material banks; biomaterials replace concrete; renovation dominates. Progress on carbon and resource targets is strong, but urban affordability challenges persist without deliberate policy attention to housing typologies and ownership models.

In the third, public sector leadership: governments take direct control after market mechanisms fail to deliver at scale. Binding targets, standardised typologies, and public investment programmes drive rapid decarbonisation and housing supply, but at the cost of private innovation and creative experimentation.

In the fourth, the Green energy revolution whereby the rapid decarbonisation of the electricity grid reshapes the entire housing question. Buildings become active nodes in bidirectional smart grids, and operational carbon largely disappears. But attention shifts to embodied carbon, energy poverty, and the distributional effects of a transition that benefits some households far more than others.

The call for action

What our scenario analysis makes clear is that there is no automatic alignment between building more homes, decarbonising the stock, and making housing affordable. The same headline instruments, green finance, circular procurement, digitalisation, lead to very different outcomes, depending on who orchestrates the system and which governance logic dominates. This has direct implications for policymakers, investors, and industry leaders right now.

Three no-regret priorities emerge across all four futures:

• Deep renovation of the existing building stock is non-negotiable in every pathway; the question is only who pays and who profits.

• Digital infrastructure for monitoring energy and material performance is needed regardless of which actor is in charge.

• And new skills and organisational capabilities for industrialised construction and lifecycle thinking must be built now, not after the transition has arrived.

The EU’s first Affordable Housing Plan, launched in late 2025, and the upcoming first-ever EU Housing Summit in 2026 offer a rare political window. The question is whether policymakers will use it to address the structural governance failures our scenarios reveal, or simply add more instruments to a system whose fundamental tensions remain unresolved.

The building industry has a decade and a half to get this right. The futures exist; the choices are ours.

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.

During the entire course of World War I, approximately 25,000 American women crossed the Atlantic Ocean to attend to the needs of wounded soldiers and civilian communities in Europe. Women traditionally operated in medical units and helped care for wounded soldiers. Following US entry in the conflict, the newly established Women’s Overseas Hospitals and the American Women’s Hospitals in France drew hundreds of trained nurses to get involved in the war effort. Women’s participation, however, was not limited to the medical field. Female physicians and stenographers brought valuable skills to the front and helped the US military in a variety of domains. In 1918, for instance, the US Army Signal Corps sent 223 trained telephone operators to France to take over from inexperienced soldiers who were struggling to keep general headquarters connected with the troops who were under fire.

At a time when women experienced domestic confinement within their homes, taking part in relief organisations and being actively involved on the Western Front gradually reinforced their quest for equal rights, furthered their political agenda, and strengthened their claim for full citizenship.

Many American women seeking meaningful wartime jobs in France came from a very specific background, and many “hoped that the war would prove the forcing house in which long-standing feminine aspirations for the vote and economic equality would finally mature”.

Considerations for telling the story of the mothers to ‘America’s French orphans’

Any course focusing on American women in World War I should acknowledge the social backgrounds of the American wealthy expatriates, businessmen’s daughters, leisured wives of diplomats, and middle-class professionals who served as doctors, nurses, ambulance drivers, stenographers, and radio operators.

When teaching World War I in relation to 20th century American history to high school pupils and undergraduate students, educators traditionally focus on the neutrality of the United States and then expand on the reasons why Woodrow Wilson gradually dragged his country into the global conflict (Editorial note – For further reference: The Path to War: How the First World War Created Modern America by Michael S. Neiberg, Oxford, 2016; Neutrals, Belligerents and the Transformation of the First World War by Abbenhuis Maartje and Ismee Tames, London: Bloomsbury Academic, 2022).

Military historians linger on battles, strategies, and the decision-making process; cultural history gravitates around cultural encounters, war atrocities, and public reaction to the outbreak of the conflict; and scholars specialised in diplomacy dig into government archives, private papers, and conference proceedings to determine the responsibility of each country. But historians of women, childhood, and philanthropy have much to add to the understanding of WWI.

Presenting the big picture fatally necessitates omitting important details, but in the case of World War I studies, some entire facets of the conflict have been overlooked.

Out of interest in humanitarian organisations that operated in my home country, France, between 1914 and 1921, I have recently shifted the focus of my teaching to the plight of children during World War I. Cultural historians have long demonstrated that the French school system mobilised its youth to perpetuate a sense of national belonging in wartime and how state propaganda shaped children’s worldview. Yet I find that the various pictures of the conflict remain ethnocentric and neglect the silent but vital action of American women in rescuing France’s children.

Between 1914 and 1921, thousands of American women acted as mothers for French children displaced by the war.

In 1915, a group of American philanthropists envisioned the creation of Franco-American colonies to rescue youngest war victims from starvation and misery.

Twenty-eight colonies were established by the Committee Franco-American for the Protection of the Children of the Frontier (CFAPCF) to shelter displaced orphans from France and Belgium. All the colonies were managed and staffed by French nuns, but heavily depended on American donations and volunteers – American women. Among them were Alma A. Clarke, a former student at Bryn Mawr College, and Erica Thorp de Berry, the granddaughter of Henry Wadsworth Longfellow, a Harvard University professor and a towering figure in 19th century American literature.

American women helped to feed, educate, and nurse the orphaned and traumatised children who were moved to the colonies to recover and prepare for life on their own after the war. They tucked little orphans into bed, kissed them goodnight, told them stories of the gigantic country across the Atlantic Ocean, and even sang songs when they could not sleep.

Colonies operated as “humanitarian wombs” and though the survival of approximately 800 children from France and Belgium could look relatively insignificant, they carried out the first humanitarian actions toward children.

That same year, in 1915, another humanitarian organisation reached out to thousands of Americans.

Envisioned by Paris-based French industrialist, Émile Deutsch de la Meurthe, the Fatherless Children of France Society (FCFS) encouraged Americans to “adopt” France’s children who had lost their fathers to the war. Although considered orphans by virtue of being fatherless, the children were not “adopted” but rather sponsored at the rate of $36.50 per year (what would be today $900/€773). Though the tireless and skilled efforts of the FCFS staff and volunteers (mainly women), between 1915 and 1921, some 300,000 French children were spared hunger and destitution because they were sponsored by Americans.

Both organisations drew Americans’ financial support and mobilised hundreds of women across the United States. To engage donors and volunteers, they organised fairs on July 4 to remind Americans of Lafayette’s role in the American War of Independence, and spurred Americans to contribute to France’s survival. In the aftermath of the war, mourning families and those who had served were moved to support the cause of the FCFS.

The Fatherless Children of France Society more than doubled the number of sponsorships between November 1918 and January 1921, the date the organisation officially ceased to exist.

American women’s roles in reforging post-WWI communities

In the years after the war, individual Americans helped rebuild devastated France. American women set up schools and reconstructed devastated villages. For example, the American Committee for Devastated France (ACDF), co-founded by Anne Morgan, the daughter of American financier J.P. Morgan, operated on several fronts. From the Château de Blérancourt, some 350 French-speaking American women joined her task force. Among them were Mary Carson Breckinridge, the daughter of an Arkansas congressman and future founder of the Frontier Nursing Service; Lucile Atcherson Curtis, a militant suffragette who would later become the first female in the US Foreign Service; and Anna Lander West McDonnell, the niece of the first Chief Justice of the Supreme Court of the Territory of Washington.

Though the ACDF’s initial mission was to combat infant mortality, rebuild devastated villages, and finance the reconstruction of the industrial network, children’s well-being rapidly became a focus of the organisation.

The ACDF established a network of public libraries for children in the former occupied zones of Northern France. Jessie Carson became the director of a new American-style network of lending libraries for children. In April 1919, the first reading room for children was opened in the Northern French town of Vic-sur-Aisne.

The ACDF inspired American women at Wellesley, Vassar, Smith, Radcliffe, Stanford and other American colleges and universities to tackle problems related to agricultural production and devastated villagers. For example, in France’s remote and war-ravaged regions, a lack of milk contributed to infant mortality.

Funds from humanitarian organisations brought cows to the devastated regions, where underfed mothers could not breastfeed their babies. In early 1920, in Verdun (Meuse), the American-Franco Children’s League bought several cows, and Miss Butler, the president of the Vassar College unit of volunteers, organised the distribution of milk for babies; at the same time, in Reims (Marne), a “Drop of Milk Institution for Babies” opened, through the efforts of American women.

In short, when it came to humanitarian efforts to shield France’s children from destitution, hunger, and death, American women got the job done. And this is an untold story.

American women’s experiences in humanitarian missions in France during WWI are important for many reasons. First of all, they pave the way for future research on American humanitarian action during the Great War, and complement studies dealing with Franco-American relations.

Additionally, the archives of these associations are a treasure for those teaching history at the K-12, college, and graduate levels, as they contain letters from the women serving in France during and after the war. These primary sources are important first-hand accounts of the conflict.

For example, in teaching my unit on American action in France during WWI, I invited my pupils to analyse several fragments of Anne Morgan’s letters to her mother, held at the Morgan Library and Museum in New York City. Her letter, dated April 30, 1919 (Anne Tracy Morgan Papers,1888–1952, Morgan Library and Museum, New York), read as follows:

“We had proudly repaired a room to be used for the school at Camelin, when the Mayor came in and told us that in the brook, just outside the door of the school house, the head of a Boche had appeared in the water, as the brook had washed away the covering of soil that was over the body.”

With all the archives available online and the different tools to communicate, schools and universities in France and in the United States could easily partner with each other on digital projects, along with local repository, library, museum, or university collection.

From across the Atlantic Ocean and from more than a century ago, American women’s voices bearing important witness are still waiting to be heard.

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!

Emmanuel Destenay ne travaille pas, ne conseille pas, ne possède pas de parts, ne reçoit pas de fonds d'une organisation qui pourrait tirer profit de cet article, et n'a déclaré aucune autre affiliation que son organisme de recherche.

La course à l’IA engagée à l’échelle internationale ne doit pas se traduire par un détricotage des règles préservant nos ressources naturelles.

Qui n’a pas déjà expérimenté la désagréable sensation de surchauffe de son téléphone portable ou de son ordinateur lors d’une utilisation prolongée ou lorsque vous avez ouvert trop d’onglets sur votre navigateur ?

Imaginez maintenant la chaleur dégagée par 100 000 puces de calcul de dernière génération, entassées les unes sur les autres et tournant à plein régime, et ce, dans un complexe de plus de 26 kilomètres carrés soit environ 4 stades de football. Placez ce grille-pain géant dans une région où la température est de 35 degrés en moyenne et peut atteindre les 50 °C l’été, et vous voilà devant le projet « Stargate UAE » visant à construire jusqu’à 5 gigawatts de puissance de calcul installée dans un immense centre de données à Abu Dhabi.

Ces projets de centres de données dits « hyperscale » visant à alimenter l’essor de l’intelligence artificielle (IA) se multiplient dans le monde, que ce soit aux États-Unis avec le projet Prometheus de Meta prévoyant la construction d’un centre de données de la taille de Manhattan, et même en France avec le « Campus IA ».

Au-delà de leur importante consommation énergétique, ces mastodontes soulèvent d’autres problèmes. Pour fonctionner correctement, ils ne peuvent pas atteindre des températures trop élevées, et contiennent donc des systèmes de refroidissement qui permettent aux composants électroniques de fonctionner à plein régime tout en évitant qu’ils ne se détériorent sous la chaleur qu’ils dégagent.

Comment ces centres de données sont-ils refroidis ? Quel est l’impact de leur refroidissement sur l’environnement, et comment les rendre plus sobres ?

Centres de données, refroidissement et consommation en eau

Il existe plusieurs techniques pour refroidir un centre de données. Pour le résumer simplement, les systèmes de refroidissement reposaient auparavant exclusivement sur des systèmes de ventilation (comme dans votre ordinateur) ou de climatisation (comme dans votre voiture) qui utilisent la circulation de l’air comme source de fraîcheur et rejettent l’air chaud à l’extérieur.

Une deuxième solution de refroidissement utilise l’eau, beaucoup plus efficace que l’air pour transférer la chaleur. Celle-ci permet de rafraîchir des plaques placées proches des composants électroniques, et/ou de rafraîchir l’air ventilé dans l’entrepôt de données.

Dans le premier cas (la climatisation), l’opérateur a besoin de beaucoup d’énergie pour faire tourner les pompes et systèmes de ventilation. Dans le deuxième (le refroidissement liquide), l’entreprise a besoin de moins d’énergie mais nécessitera l’accès à une source d’eau douce (l’eau salée endommagerait les tuyaux et composants) afin d’alimenter son système en eau fraîche.

Les opérateurs de centres de données sont donc face à un arbitrage complexe : doivent-ils utiliser des systèmes de climatisation énergivores ou bien du refroidissement liquide qui, cette fois, nécessite la consommation d’importantes ressources en eau ?

En effet, la consommation en eau des data centers est estimée à 560 milliards de litres chaque année dans le monde, soit l’équivalent de la consommation annuelle en eau potable de 10 millions de Français.

Cette soif insatiable se retrouve également dans les chiffres publiés par les Gafam. Ainsi, Google a vu sa consommation nette d’eau augmenter de 28 % entre 2023 et 2024, atteignant 30 milliards de litres dont environ un tiers provient de régions en stress hydrique. Microsoft, pour sa part, estime que 46 % de ses prélèvements d’eau ont lieu dans de telles zones en 2024.

Toutefois, il faut avoir à l’esprit que les besoins en eau des data centers ne sont pas uniquement liés aux systèmes de refroidissement. Pour obtenir une vision globale de l’impact du développement de ces infrastructures sur les ressources en eau, il convient de prendre également en compte l’eau utilisée par les centrales électriques qui les alimentent, ainsi que l’eau consommée lors du processus de fabrication des composants électroniques. Des chercheurs estiment ainsi que les mégacentres de données construits spécifiquement pour les besoins de calcul de l’IA utilisent, en moyenne, jusqu’à 20 millions de litres d’eau par jour, soit autant qu’une ville de 10 000 à 50 000 habitants.

Peut-on rendre les centres de données moins gourmands en eau ?

Il existe des solutions innovantes pour limiter cette consommation et rendre les systèmes de refroidissement plus efficients. Des entreprises, comme OVH Cloud, Nvidia ou Nebius, développent et déploient de nouvelles architectures de systèmes de refroidissement liquide au plus proche des puces de calcul. Ces nouvelles techniques permettent de réduire, selon les chiffres annoncés, jusqu’à 50 % de la consommation en eau. Toutefois, elles restent encore onéreuses à mettre en œuvre et assez peu développées sur le parc existant.

De manière plus générale, la principale source de perte en eau lors du fonctionnement des centres de données vient du fait qu’ils reposent aujourd’hui pour la plupart sur des circuits ouverts, conduisant à l’évaporation d’une grande partie de l’eau utilisée. C’est pourquoi les nouveaux centres de données devraient idéalement reposer, autant que possible, sur des systèmes de refroidissement en circuit fermé, évitant ce phénomène d’évaporation. Néanmoins, ce type de refroidissement peut s’avérer plus cher, conduit souvent à une hausse du besoin en électricité, et n’est pas évident à mettre en œuvre dans tous les centres de données « historiques » qui n’ont pas été conçus pour le mettre en œuvre.

Des propositions plus farfelues sont aussi avancées, telles que l’envoi de data centers dans l’espace ou bien en immersion dans les océans. Néanmoins, l’apport réel de ces propositions reste encore largement débattu, que ce soit pour des questions de faisabilité technique (bon courage pour réaliser la maintenance de votre centre de données sous-marin !) ou de bénéfices en termes d’émission de CO₂ par rapport à un centre construit sur terre – le cabinet de conseil en décarbonation, Carbone 4, fondé par Alain Grandjean et Jean-Marc Jancovici a, à cet égard, montré que les data centers spatiaux risquaient d’avoir un impact carbone plus important que sur terre en raison des émissions liées au lancement.

À lire aussi : Pourrait-on faire fonctionner des data centers dans l’espace ?

Pour un développement raisonné des centres de données, conscient du caractère fini des ressources naturelles

Au-delà de la faisabilité technique, ces discours risquent de nous détourner du vrai problème : le développement massif de centres de données hyperscale très gourmands en eau, dont une bonne partie dans des territoires où cette ressource se fait rare et conduit à des conflits d’usage.

Ce développement ne se fait pas dans un vide juridique. Les règles du droit de l’environnement, de l’aménagement du territoire et de l’urbanisme prévoient un certain nombre de régimes d’autorisation et d’évaluation environnementale en amont de la construction de ces projets, notamment en France avec le régime des installations classées pour la protection de l’environnement (ICPE).

Néanmoins, la course à l’IA engagée à l’échelle internationale conduit les pays à rivaliser d’ingéniosité pour attirer les investisseurs quitte, parfois, à assouplir les contraintes réglementaires comme c’est le cas actuellement en France avec la loi dite de simplification de la vie économique récemment adoptée. Il est urgent de prêter attention à l’ode à la « simplification », qui provient des discours politiques au sein de l’Union européenne et transcrite dans la politique menée par la Commission européenne, mais qui ne doivent pas se traduire par un détricotage des règles préservant nos ressources naturelles.

Plus généralement, ces débats soulèvent la question de l’usage : alors que certaines économistes parlent de « bulle de l’IA », qui peut réellement prédire quels seront les véritables usages futurs de ces infrastructures ?

Dans les années 1960, il fallait un bâtiment entier pour faire tenir un ordinateur, ils tiennent aujourd’hui dans notre smartphone. Si les IA de demain tiennent aussi sur nos terminaux, doit-on réellement sacrifier nos ressources naturelles pour créer ces mastodontes ?

À lire aussi : Charles Ponzi nous permet-il de comprendre la bulle de l’IA ?

Thomas Le Goff est Research Fellow au sein du think thank Centre on Regulation in Europe (CERRE).