Enhance Project: who pays for disaster risk in Europe, and through which partnerships

Who pays for disaster risk in Europe

A flood leaves behind a bill long after the water has gone. Somebody pays to dry out the walls, rebuild the bridge, replace the delivery van and reopen the school, and in Europe that somebody is rarely an insurer alone. According to the European Environment Agency, weather- and climate-related extremes cost the EU-27 EUR 822bn between 1980 and 2024, and less than a fifth of that was privately insured. The rest landed on households, small firms, municipalities, national treasuries and, for the largest events, the EU budget.

The losses are also bunching. The EEA puts average annual losses at EUR 8.6bn in the 1980s and EUR 44.9bn in 2020–2024. Over EUR 208bn, a quarter of the 45-year total, came in just four years, 2021 to 2024, and 2021 on its own cost EUR 65.2bn. Floods account for 47% of all losses in the EEA series, storms for 27% and heatwaves for 18%.

Insurance does not follow that pattern evenly. Storm damage is the best insured: more than 35% of meteorological losses were covered, against about 15% for hydrological losses such as floods and about 10% for climatological ones such as drought and heat. Geography matters even more than hazard. The EEA finds that under 3% of losses were insured in Bulgaria, Croatia, Cyprus, Lithuania, Malta, Romania and Iceland, while Belgium, Denmark, France, Luxembourg, the Netherlands, Liechtenstein, Norway and Switzerland all passed 35%.

Who carries the uninsured part depends on the country's rules. In Germany after the July 2021 floods, the Bundestag set up the federal reconstruction fund "Aufbauhilfe 2021" with up to EUR 30bn, half of it repaid by the Länder through VAT shares until 2050. Five years on, money is still sitting in it: by 30 June 2026 only EUR 6.2bn, about 20%, had been drawn, according to dpa. In Austria the federal Katastrophenfonds refunds 60% of the aid that the Länder pay to private victims, yet the Court of Audit found that aid covered 31% of private damage in Carinthia and 63% in Tyrol. Two neighbours with the same flood can be treated very differently a valley apart.

Reinsurers see the same split from the other side. Munich Re counted about USD 11bn of overall natural-disaster losses in Europe in 2025, "around half" of them insured, against a ten-year average of USD 35bn overall and USD 12bn insured. A quiet year flatters the ratio; a year like 2021 shows how much of the burden sits outside the market.

This reference follows those flows of money and responsibility, country by country and event by event. It uses the term disaster risk partnerships for the arrangements that decide the split: insurers and the state sharing a layer of loss, pools that spread risk across a whole market, EU funds that step in when national budgets are stretched, and the less visible deals between farmers, cities, rail operators and water users that the ENHANCE research project studied between 2012 and 2016.

National schemes for natural catastrophe insurance

Europe has no single model of natural catastrophe insurance. The six national arrangements compared on the schemes hub fall roughly into four families, and the differences between them explain much of the variation in insured shares.

Compulsory cover with a state-backed reinsurer. France and Spain attach catastrophe cover to ordinary property policies by law. Under the French CatNat regime, created in 1982, every property policy carries a surcharge that rose from 12% to 20% of the premium on 1 January 2025. Insurers can cede risk to the Caisse Centrale de Réassurance, which holds an unlimited State guarantee; that guarantee has been called once, in 2000, after storms Lothar and Martin. Payment depends on an inter-ministerial order recognising a natural catastrophe in the commune, and the Sénat reported in 2024 that only about half of the communes applying for drought recognition receive it. Spain's Consorcio de Compensación de Seguros, provisional from 1941 and permanent since 1954, works without any declaration: since 1986 it has paid automatically for listed perils on any policy in the prescribed lines, funded by a flat surcharge of 0.07‰ of the sum insured for homes. Its stabilisation reserve stood at EUR 7,518.7m at the end of 2025, which is why a 612.0% loss ratio in 2024 did not break it.

Pools bundled with fire insurance. Norway and Switzerland reach near-universal take-up by tying natural perils to fire cover. The Norwegian Natural Perils Pool, running since 1 January 1980, charges everyone the same 0.08‰ of the fire sum insured, wherever they live, and shares claims among its member insurers by market share. In 2025 claims came to about NOK 2.948bn, of which Storm Amy accounted for about NOK 1.8bn. In Switzerland, 19 cantonal building insurers hold monopolies covering about 85% of the country's buildings, and private insurers in the other cantons share claims through the Elementarschadenpool. More than 95% of buildings and contents are insured against the nine listed perils, but earthquake is not one of them. Only 21% of building values carry earthquake cover, and on 30 September 2026 the National Council rejected the federal plan to close that gap.

A pool for one peril and one type of property. The UK's Flood Re, set up under the Water Act 2014 and launched in April 2016, lets insurers pass the flood element of eligible home policies into a reinsurance pool at prices set by Council Tax band, not by flood risk. It covers homes built before 1 January 2009 and nothing else: no businesses, no larger blocks, no newer houses. Every UK home insurer pays a levy of £160m a year from April 2025, and there is no taxpayer backstop above the scheme's limit. Flood Re estimated about 353,000 policies at 31 March 2026, and the scheme is designed to close in 2039.

Voluntary cover plus public aid. Germany and Austria leave flood insurance to the market and fill the gaps with state money. German take-up of natural-hazard ("Elementar") cover rose to 57% of homes in 2024 from 41% in 2017, according to the insurers' association GDV, with Baden-Württemberg at 94% because cover was compulsory there until 1993. On 28 September 2026 Justice Minister Hubig announced a framework under which new building policies must include elemental cover, with an opt-out and a draft bill expected in the first quarter of 2027. Austria's Katastrophenfonds, permanent since 1966, is funded from income and corporate tax, and its aid is discretionary, so victims have no legal claim to it. The Germany and Austria page follows both reform debates.

None of these families is free of cost. Compulsory and bundled forms of natural catastrophe insurance buy high coverage at the price of a weak risk signal: a Norges Bank staff memo published in January 2026 argued that the uniform Norwegian premium dulls incentives for prevention, and the same complaint is made of Spain's flat tariff. Voluntary markets keep the price signal but leave the poorest-protected households to rely on aid that arrives late and unevenly. On the evidence of the events below, the bundled models deliver more money to more people faster, and the debate in Berlin and Vienna has been moving towards them. The side-by-side comparison sets out founding dates, perils, funding and backstops for all six.

What recent European floods show

Three floods between 2021 and 2024 make the differences concrete. The insured shares in the dashboard at the top each come from one source per event, because mixing an insurer's figure with a government damage estimate produces numbers that look precise and are not.

July 2021, Germany and Belgium ("Bernd"). Munich Re put the overall loss at EUR 46bn and the insured loss at EUR 11bn, with Germany alone at EUR 33bn and EUR 8.2bn. The Commission recorded 196 deaths in Germany and 42 in Belgium. German insurers settled about EUR 8.75bn, but most of the damage was left to public reconstruction money and to households without elemental cover. The flood changed behaviour where it hit hardest: by the end of 2025 take-up had risen from 47% to 65% in North Rhine-Westphalia and from 37% to 59% in Rhineland-Palatinate, GDV reported in July 2026. In Belgium, whose practice of coupling natural-hazard cover to fire insurance Austrian insurers now hold up as a model, insurers handled about 74,000 claims and bore EUR 2.3bn, with the Walloon Region adding EUR 1.03bn. The Ahr valley case study traces how each country's rules shaped the recovery.

May 2023, Emilia-Romagna. PERILS estimated EUR 9bn of economic loss and EUR 495m of insured property loss, an insured share of 5.5%. About half of the damage hit public infrastructure such as roads, embankments and drainage, which no private policy would have covered anyway. Italy responded with Law 213/2023, which makes catastrophe cover compulsory for businesses against earthquake, flood, landslide, inundation and overflow, with the state-owned SACE reinsuring up to half of the claims. The start date slipped several times, and in the end it was staggered by company size: 31 March 2025 for large firms, 30 September 2025 for medium firms and 31 December 2025 for small and micro firms. The obligation does not reach households. The Emilia-Romagna case study covers the law and the EU grant in detail.

October–November 2024, Valencia. The DANA storm killed 232 people. Munich Re's early estimate gave USD 11bn overall and USD 4.2bn insured, a share of 38.2%, but that figure has aged quickly: by 27 March 2026 the Consorcio had paid EUR 4,378,657,847 on 212,237 claims, more than Munich Re's entire insured estimate, and expected a final cost of about EUR 4.8bn. The payments were spread across vehicles, homes, shops and industry in roughly equal parts. On the EU side, the Solidarity Fund granted EUR 946,153,691, the second-largest grant in its history, and EUR 645m of reallocated cohesion funds brought total EU support to nearly EUR 1.6bn. The Valencia case study also shows an alternative, indicative share built from two sources, to show how much the answer moves.

Set side by side, the three events say something simple. Where cover comes with the ordinary policy, as in Spain and Belgium, money reaches people through the claims system within months. Where it is optional and patchy, as in Germany, Italy and Austria, the recovery runs through public programmes that are slower and less predictable, and in Germany's case still largely unspent five years later. The case-study atlas maps these events alongside older ones, from the 2010 ash cloud over Iceland to Alpine rail lines in Austria.

The protection gap in Europe

The European Insurance and Occupational Pensions Authority (EIOPA) does not reduce the shortfall to a single percentage. It uses a dashboard. Each of 30 EU and EEA countries gets a score from 0 to 4 for each of five perils (windstorm, wildfire, river and pluvial flood, coastal flood and earthquake), built from a matrix of modelled risk against insurance penetration. A score of 3 or more marks a relevant gap and 2.5 means "monitor", so the country total can reach 20. The UK is not covered.

In the 2025 update Greece and Italy share the highest total, 12, driven by earthquake scores of 4 and, in Greece's case, a wildfire score of 3. Romania follows at 11.5, with a flood score of 3, above the line EIOPA treats as a relevant gap. The EEA average fell to 5.5 from 6.5 the year before, and Spain dropped from 6 to 5. EIOPA's flood chart sets EIOPA's flood scores against the historical share of flood losses that was insured; the contrast between France, at 54%, and Italy, at 3%, is the CatNat regime made visible.

EIOPA's own headline is that "only around a quarter of the losses were insured" in Europe over 1980–2024, which is higher than the EEA's figure of under 20%. The two do not contradict each other. EIOPA's historical series covers natural catastrophes including earthquakes, while the EEA indicator is limited to weather- and climate-related extremes in the EU-27, and they draw on different loss databases. Anyone quoting either number should say which one it is.

At global level the Swiss Re Institute put the 2025 natural-catastrophe protection gap at USD 112bn, or 51% of economic losses. Europe is not the worst-placed region by that measure, but its gap is concentrated in a handful of perils and countries, and that concentration is what the dashboard shows well. The protection gap page reads the scores peril by peril and explains how the matrix works.

For households in England the practical first step is to check the long term flood risk of an address, which the official service rates in four bands from high to very low.

<iframe src="https://www.enhanceproject.eu/embed/protection-gap/" width="640" height="560" style="border:0" title="Insured share of flood losses by country (EIOPA 2025)" loading="lazy"></iframe>

Disaster risk partnerships beyond insurance

Insurance is only one layer. Above and around it sit public instruments that pay out on different rules, and a growing set of tools designed to reduce the loss before it happens. Disaster risk partnerships, in the sense used here, are the arrangements that join these layers up, or fail to.

The EU Solidarity Fund, created in 2002, has mobilised over EUR 9.6bn for 110 natural disasters and 20 health emergencies, according to the Commission's regional policy directorate. A country qualifies for a "major disaster" when direct damage exceeds EUR 3bn at 2011 prices or 0.6% of gross national income, whichever is lower, and it must apply within 12 weeks. The grant is 2.5% of damage up to the threshold plus 6% of damage above it. A worked example shows the scale. Italy declared EUR 8,533m of direct damage after the 2023 floods against a threshold of EUR 3.8bn:

• 2.5% of EUR 3,800m comes to EUR 95m;
• 6% of the remaining EUR 4,733m comes to about EUR 284m;
• together that is about EUR 379m, within a fraction of a percent of the EUR 378.8m granted for Emilia-Romagna (Tuscany received a separate EUR 67.8m).

In other words the Fund covered less than a twentieth of the declared damage. It is a gesture of solidarity and a useful cash injection for public reconstruction, not a substitute for insurance. The post on EU solidarity in disasters sets out its history and the political arguments around it.

The Union Civil Protection Mechanism, set up in October 2001, sends help in kind, not in cash. The EU and 10 participating states pool teams and equipment, and the Mechanism had been activated more than 880 times since 2002, 64 of them in 2025, according to DG ECHO. Its rescEU reserve, created in 2019 and fully EU-funded, includes firefighting planes and helicopters, medical evacuation capacity and stockpiles for chemical, biological, radiological and nuclear emergencies; current funding covers 12 new planes and 5 helicopters.

Parametric cover pays a fixed sum when a measured value passes a threshold, such as wind speed or water depth, without waiting for a loss adjuster. The Financial Stability Institute and the International Association of Insurance Supervisors estimated parametric premiums at USD 14.8–18bn in 2023, about 0.8% of global property and casualty premiums. The best-known sovereign pools are outside Europe: the Caribbean's CCRIF, founded in 2007, had made 82 payouts worth about USD 483m to its 39 members by October 2025. In the UK, FloodFlash sells a sensor-triggered flood policy to businesses. The parametric insurance guide explains triggers, basis risk and where the model fits.

Risk reduction incentives are the weakest link in most schemes, and the strongest argument for changing them. Flood Re's Build Back Better option, in place since 2022, pays up to £10,000 towards resilience measures during repairs. The French Sénat has proposed lowering the CatNat deductible for policyholders who invest in prevention. In Austria, prevention was the largest single spending line of the Katastrophenfonds in 2025, at EUR 233.7m. In Germany, the insurers' "Elementar Re" proposal would make opting out of cover mean giving up a claim to state aid, an attempt to stop public money from discouraging private cover. Whether any of these works at scale is still an open question; Flood Re's own 2025-26 report concedes that take-up of resilience measures has been inconsistent.

One idea from the research literature has not gone away. The ENHANCE project's Romania case study argued that post-disaster aid can blunt the incentive to prevent damage, and proposed turning the Solidarity Fund into a pre-disaster, risk-based instrument. The Fund's rules have not changed in that direction, but the moral-hazard argument now turns up in almost every national reform debate, from Berlin to Vienna.

Solidarity is the oldest of these arrangements and the least formal one. Why solidarity matters most when a disaster strikes follows the idea from neighbours filling sandbags to the EU funds that reimburse governments afterwards.

The Enhance Project reference and the ENHANCE research

Enhance Project is a reference on how Europe shares the cost of natural disasters: national schemes, EU instruments, the protection gap, UK flood risk and the events that test all of them. Terms used across the pages are defined in the disaster risk glossary.

Much of the background comes from the ENHANCE project, a research programme funded under the EU's Seventh Framework Programme (grant agreement 308438) that ran from 2012 to 2016. Its full name, Enhancing risk management partnerships for catastrophic natural hazards in Europe, sums up its subject. The project brought together 24 partners from academia, government, companies and international organisations in 11 countries, and built its work around 10 case studies. They ranged widely in hazard and scale:

• controlled flooding of farmland to protect towns in the Po basin after the May 2012 earthquake damaged drainage works;
• the 2003 forest fires in Chamusca, Portugal, which burnt nearly 22,000 hectares;
• Alpine hazards on Austrian rail lines, where the operator ÖBB Infra and its partners disagreed over cost sharing;
• storm-surge risk culture on the Wadden Sea coast, drought in the Júcar basin and flood risk in London;
• the 2010 Eyjafjallajökull ash cloud and its effect on European aviation.

The project's working papers, among them D5.1 on economic instruments for risk reduction and D5.3 on novel and improved insurance instruments, are listed with their authors in the documents library. Its newsletters and press material, which carried interviews with Margareta Wahlström of UNISDR and EU Commissioner Christos Stylianides, are gathered in the media corner.

ENHANCE partners took part in the Third UN World Conference on Disaster Risk Reduction in Sendai, Japan, on 14–18 March 2015, and wrote nine "Letters from Sendai" during and after it. The final letter was blunt: none of the seven targets in the adopted Sendai Framework set a quantitative degree of progress, and most were framed as global outcomes, not national ones. Another letter reported that the partners had found no appetite in Europe for harmonising insurance approaches, which still holds for the six national schemes a decade later.

The project's final book, announced on 31 October 2016, reported that ten public, private and civil-society partnerships had been set up or developed, with an emphasis on the financial sector. One of the smaller results was also one of the most concrete: the Rotterdam Port Authority, the Municipality of Rotterdam and the Dutch infrastructure ministry committed roughly EUR 200,000 to further climate-risk analysis of the port. In an interview with Jaroslav Mysiak of the Fondazione Eni Enrico Mattei, who coordinated the case studies, the yardsticks for judging whether such a partnership works were described as still missing. Readers comparing Flood Re with the Consorcio, or the Aufbauhilfe with the Solidarity Fund, will find that this is still the hardest part of the question.