The May 2023 Emilia-Romagna floods: losses, cover and aid

Romagna, the eastern half of the region of Emilia-Romagna, is a flat and fertile plain crossed by rivers that come down from the Apennines to the Adriatic. In May 2023 extreme rain fell across the region for weeks, flooding the plain and setting off landslides in the hills. Parts of Marche and Tuscany were also hit, but the worst damage was in Romagna, in the area around Faenza, Ravenna and Forlì-Cesena.

What happened between 2 and 22 May

PERILS, which compiles industry loss figures for European catastrophes, dates the event from 2 to 22 May 2023 and treats it as one loss event: riverine flooding combined with landslides, spread over three weeks. That choice is not a technicality. Reinsurance contracts usually define an event by a time window, and a flood that runs for three weeks raises the question of whether it is one loss or several, with real money attached to the answer.

Unlike the July 2021 floods in Germany, which struck narrow valleys within a few days, the Romagna floods covered a wide, flat and densely farmed plain. Flooding on a plain also reaches far more buildings, farms and businesses than flooding confined to a valley floor.

Deaths

The two official counts differ. PERILS reports 17 deaths in its final loss announcement of 22 May 2024. The European Commission’s proposal to mobilise the EU Solidarity Fund, COM(2024) 325 of 27 August 2024, says 14 people lost their lives. Neither figure should be quoted without its source.

Total and insured losses

The chart above shows how little of the damage was insured. Both bars come from the same source, PERILS, so the comparison is consistent.

According to PERILS (22 May 2024):

• the overall economic cost was about EUR 9bn;
• the property insurance market loss was EUR 495m, the final estimate;
• that gives an insured share of about 5.5%.

PERILS gives two reasons for the gap. Flood insurance take-up among households and small and medium-sized firms was low. And around 50% of the losses hit public infrastructure (roads, embankments, bridges, schools), which is typically uninsured.

The figure declared to the EU Solidarity Fund offers a second, independent measure of the damage. Italy declared EUR 8,533m of direct damage, close to PERILS’ economic estimate although the two are compiled on different bases.

How Emilia-Romagna compares

An insured share of 5.5% is strikingly low even by European standards. Using each source’s own pair of total and insured figures, the July 2021 floods in Germany and Belgium came out at about 24% insured according to Munich Re, and the October 2024 floods in Valencia at about 38% according to Munich Re’s early estimate. Emilia-Romagna’s share was a fraction of either.

The difference comes down to institutions. In Spain, the Consorcio de Compensación de Seguros pays flood claims on every property policy through a compulsory surcharge. In Germany, cover is voluntary but widely held. In Italy, flood cover for homes was, and remains, rare. EIOPA’s data on historical insured shares, taken from the CATDAT database, put the insured share of flood losses in Italy over 1980–2024 at 3%.

Four numbers for one disaster

The Emilia-Romagna floods are a useful lesson in how different figures for the same disaster measure different things.

• EUR 9bn is PERILS’ estimate of the overall economic cost. It is the broadest figure, covering public and private losses alike.
• EUR 8,533m is the direct damage Italy declared to the EU Solidarity Fund. It follows the fund’s own rules on what counts and was compiled by the Italian authorities for the application.
• EUR 495m is the insured property loss, the amount the insurance market expects to pay. It counts only what was insured.
• EUR 378.8m is the EU Solidarity Fund grant for Emilia-Romagna, calculated from the declared damage with the fund’s formula.

Each number is right for its purpose. Problems start when they are mixed. Dividing the insured loss by the declared direct damage, for example, gives a share of about 5.8%, close to PERILS’ own 5.5% but built from two sources compiled for different purposes. The 5.5% figure, with both halves from PERILS, is the one to quote.

The grant is the number most often misunderstood. A grant of nearly EUR 380m sounds large in a headline. Set against EUR 8.5bn of declared damage it is about 4%, and it is paid to the Italian state for public operations, not to the people whose homes and businesses flooded.

Who paid

With insurers covering EUR 495m, most of the cost fell on public budgets, on households and firms themselves, and on the EU.

The EU Solidarity Fund

Italy’s application to the EU Solidarity Fund cleared the “major disaster” bar easily. Its EUR 8,533m of declared direct damage was more than double the country’s threshold of EUR 3.8bn, which under the fund’s rules is set at EUR 3bn in 2011 prices or 0.6% of gross national income, whichever is lower.

The Commission’s proposal COM(2024) 325 set the grant for Emilia-Romagna at EUR 378,833,540, including an advance of EUR 94,708,385 already paid. The same proposal allocated EUR 67,811,826 for Tuscany.

The grant follows the fund’s standard formula: 2.5% of damage up to the threshold plus 6% of damage above it. Put plainly, the fund paid about 4% of the declared damage. It is designed to support emergency operations and the restoration of public infrastructure, not to compensate private losses. The wider role of the fund is set out in the post on EU solidarity in disasters.

A worked example: a Faenza workshop

Consider a small metalworking firm on the edge of Faenza, flooded to a depth of a metre in May 2023. Its machinery, stock and premises are damaged. Like most small Italian firms at the time, it has fire and theft cover but no flood cover.

In 2023 that firm had no insurance claim to make for the flood. Its route to compensation ran through public aid schemes, with their application rules, ceilings and waiting times. The EU Solidarity Fund grant reached Italy as money for emergency and recovery operations, not as payments to the firm.

Run the same flood forward to 2026. Under Law 213/2023, as phased in by later decrees, a small or micro firm should have had compulsory catastrophe cover from 31 December 2025, with SACE, the state-backed export credit and insurance company, reinsuring up to 50% of claims. The firm would now claim on its own policy. The change shifts the first layer of the cost from the state’s budget to the insurance market, with the state still behind it as reinsurer.

Italy’s compulsory business cover law

The May 2023 floods came while Italy was already debating catastrophe insurance. The budget law for 2024, Law 213/2023, settled it for businesses. Article 1, paragraphs 101 to 111, requires businesses to insure against earthquake, flood, landslide, inundation and overflow. SACE reinsures up to 50% of the indemnities, under a state guarantee, according to the Ministry of Enterprises and Made in Italy (MIMIT).

The start date moved several times. As set out by MIMIT (page last updated on 3 June 2026) and by Ingenio’s report on the Milleproroghe decree:

Some secondary sources give 1 October 2025 for medium firms. MIMIT gives 30 September 2025.

Five changes of date in eighteen months tell their own story. A compulsory market of this size needs insurers to design and price products and businesses to find and budget for them, and each deferral bought time for one side or the other. The cost of deferral is that every month of delay is a month in which a flood would have found most small firms still uninsured.

EIOPA noted in its 2025 dashboard summary that Italy’s new mandatory scheme for non-agricultural firms entered into force in 2025 and is expected to narrow the protection gap for commercial buildings. The word “commercial” matters. The law does nothing for homes.

The gap the law leaves

Italy’s total protection-gap score on EIOPA’s 2025 dashboard is 12 out of a possible 20, the joint highest in the EU and EEA alongside Greece. Its flood score is 2.5 on a scale of 0 to 4, just at the level EIOPA flags for monitoring. Its earthquake score is 4, the maximum.

Those scores need reading with care. EIOPA builds each one from a matrix that sets modelled risk, on a scale of 0 to 4, against insurance penetration in bands of 0–25%, 25–50%, 50–75% and 75–100%. A score of 3 or more counts as a relevant gap, and 2.5 means the gap should be monitored. Italy’s flood score of 2.5 therefore says that its flood risk is meaningful and its flood insurance thin, but not yet at the level EIOPA treats as a relevant gap. The May 2023 floods, with 5.5% insured, suggest that the national average may flatter regions such as Romagna.

The European Environment Agency’s figures point the same way at European level. Its indicator of October 2025 found that only about 15% of losses from hydrological events, which include floods, were insured across the EU over 1980–2024, against more than 35% for meteorological events such as storms. Flood is the peril Europe insures least well, and Italy is one of the clearest cases.

Households remain outside any compulsory arrangement. A family whose house in Faenza flooded in 2023 faces the next flood in much the same position: insured only if it chose to buy cover, and otherwise dependent on public aid. That is a political choice, and a contested one. Making home cover compulsory would spread the cost across all owners, including those in low-risk areas, and would raise affordability questions for poorer households in the south, where earthquake risk is highest. Leaving it voluntary keeps the burden on public budgets after each disaster.

The comparison with Spain’s Consorcio and France’s CatNat, both of which attach cover to every property policy, is set out on the schemes comparison page. How EIOPA builds its scores is explained on the protection gap page, and the case-study atlas places Emilia-Romagna alongside the July 2021 floods in Germany and Belgium and the Valencia floods of 2024.