Natural catastrophes are becoming more frequent and more severe, but the losses they cause are increasingly uninsured. Less than a quarter of the economic damage from natural catastrophes in the EU is currently covered by insurance, and in several countries the share is below 5%. This digest brings together recent work that asks why this protection gap matters for the macroeconomy, how climate change is likely to widen it, and what that implies for growth and welfare.
A Growth Model with Catastrophes and Insurance In Giuzio, Rousová, Kapadia, Kumar, Mazzotta, Parker & Zafeiris (2026), we build a stochastic growth model in which disasters arrive as Poisson shocks to the capital stock, and insurance determines how quickly destroyed capital is rebuilt. The key mechanism is financial, not just monetary: insurance payouts relax the borrowing and liquidity constraints that otherwise force households and firms to divert savings and investment into reconstruction, prolonging the recovery. Fully insured damage is rebuilt immediately and growth is unaffected; uninsured damage depresses output for as long as reconstruction takes. We then let climate change operate through two separate channels: a gradual rise in temperatures that erodes capital productivity directly (which insurance cannot touch), and a rising catastrophe probability that pushes up insurance premiums and shrinks coverage (insurance retreat), amplifying the growth cost of the first channel.
Evidence from 47 Countries, 1996–2019 The empirical part tests this mechanism using several thousand EM-DAT disaster events matched to quarterly GDP growth across 47 developed and middle-income countries. A catastrophe causing damage equal to 1% of GDP lowers quarterly GDP growth by about 0.2–0.25 percentage points on impact if uninsured. But if a high share of the damage is covered by insurance, the initial growth hit is largely offset, and for very high insured shares we even find a small immediate rebound. Splitting large-scale disasters into high- and low-insured-share groups shows that the drag on annual growth from poorly insured catastrophes can persist for up to three quarters, while growth after well-insured disasters does not deviate significantly from trend. The results survive a battery of checks for endogeneity between insurance coverage and economic development, including splits by GDP per capita, credit depth, and government effectiveness, and by whether a country has a government-backed catastrophe insurance scheme.
Recovery and Risk-Sharing across Countries von Peter, von Dahlen & Saxena (2024) reach a strikingly consistent conclusion using annual data for over 200 countries between 1960 and 2011. They show that GDP recovery from natural catastrophes is faster and more complete the larger the insured share of damages, and that aggregate GDP losses are overwhelmingly driven by the uninsured component of damage. Their reduced-form evidence on the persistence of the macroeconomic effect of uninsured disasters across a set of mostly lower-income countries complements our quarterly-frequency sample of developed and middle-income economies, and the convergence of the two sets of results, despite very different samples, frequencies, and time periods, is reassuring evidence that the insurance–recovery link is not an artefact of any one dataset.
Financial Adaptation and Sovereign Risk Phan & Schwartzman (2024) study how disaster insurance and catastrophe bonds interact with sovereign default risk in a calibrated growth model. The comparison matters because governments and insurers typically act on different timelines. Sovereign disaster relief and post-catastrophe fiscal support are almost by construction ex-post: they arrive only once damage has already occurred, and their prospect can even discourage preparedness, since households, firms, or local governments may under-invest in resilience if they expect the state to cover the shortfall regardless. Insurance, by contrast, is priced ex-ante and can be designed to reward risk-reducing behaviour before a catastrophe hits, for instance through lower premiums for flood-proofed buildings or climate-resilient construction, an approach known in the industry as impact underwriting. This distinction is central to the moral hazard concern we raise in our own paper: when the public sector is expected to backstop losses regardless of preparedness, it can crowd out exactly the private incentives for adaptation that well-designed insurance contracts are able to provide. Phan and Schwartzman model full insurance as a device that equalises a country’s net worth regardless of whether disaster strikes, and show how the interaction between catastrophe risk and sovereign credit constraints shapes optimal financial adaptation. Where their model treats insurance largely as a binary net-worth stabiliser, our framework puts the extent of insurance coverage and the speed of post-disaster reconstruction at the centre, making the two papers natural complements: one emphasising the sovereign financing side of climate risk and its incentive effects, the other the private capital-rebuilding side.
Modeling Reconstruction Dynamics Hallegatte, Jooste & McIsaac (2024) build a detailed structural macroeconomic model of the medium-run impact of disasters, distinguishing infrastructure from non-infrastructure capital and public from private capital. Consistent with our reduced-form growth model, they find that the speed of reconstruction is a crucial determinant of the ultimate economic cost of a disaster, and that financial interventions, including insurance, are central to accelerating that reconstruction. Their richer treatment of monetary policy and capital heterogeneity offers a natural extension of the mechanism we keep deliberately stylised in order to derive clean empirical predictions.
Key Lesson Across very different modelling strategies, samples, and time horizons, these papers converge on the same point: the macroeconomic cost of a natural catastrophe is not fixed by the physical damage alone, it depends on how quickly destroyed capital can be rebuilt, and insurance is one of the fastest channels available for financing that rebuild. As climate change raises both the frequency of catastrophes and, through insurer risk aversion and rising premiums, the risk of insurance retreat, closing the protection gap, e.g. through deeper private insurance penetration, catastrophe bond markets, public–private schemes, or cross-border risk pooling, is not just a matter of household or firm-level protection. It is a first-order determinant of how costly climate change will be for growth and welfare.
The views expressed here are those of the authors and do not necessarily reflect the position of the European Central Bank.