With growing consensus on the scale of climate change and its direct impacts on societies and economies, the debate on how direct climate damages cascade through interconnected economic systems eventually leading to profound indirect effects remains open amid limited quantitative evidence. The localised nature of climate risk means that regional economies may face stark differences in how they are impacted by the direct and indirect physical risks. In Europe, the world’s fastest warming continent, concentrated local damages spill over asymmetrically into tightly-interconnected regional markets, potentially leading to increased inter-regional inequalities despite the continent’s prioritisation of economic unity through its regional cohesion policy. Solid regionalized economy-wide analysis could identify if physical climate risks serve as structural drivers of future regional inequality, offering timely insights for policy interventions to curb exacerbating socio-economic adversities. Here, using an empirical dynamic computable general equilibrium model disaggregated to NUTS2 European regions, we explore a range of regional economic projections from two particularly costly climate-driven hazards: river flooding and sea-level rise. Our methodology captures complex economic feedbacks across granular regions and sectors to estimate both direct and indirect economic repercussions of combined sea-level rise and river flood events by 2100. We find that these climate-induced hazards represent an economically divergent force for the European regions. In contrast to aggregated studies, the resulting regional economic projections reveal wide heterogeneity in combined (direct and indirect) physical risks, with the most affected regions experiencing devastating declines in GDP up to 51% by 2100. Our disaggregated projections capture the effects of two climate-induced hazards with distinct geographical hotspots - coastal and inland - which occur simultaneously though unfold at different rates, enabling a detailed assessment of regional economic inequalities. Low income regions experience by far the highest proportional losses, leading to increases in both between and within-country regional inequality. ... Read more

Why is the distribution of government spending between cash and in-kind transfers so heterogeneous across European countries? In this paper, we present a theoretical model showing that this difference depends on citizens’ trust in their own political institutions. When citizens do not trust their political institutions, they are more likely to believe that those institutions will waste economic resources while providing public services, reducing their quality. As a response, citizens demand more cash transfers at the expense of fewer in-kind transfers to minimise the quantity of low-quality services they receive. The model also shows that: (i) higher expected income increases the support for in-kind transfers and reduces support for cash transfers; (ii) paternalism increases the support for in-kind transfers over cash transfers; (iii) lower levels of civic capital reduce the demand for cash transfer and, under specific conditions, increase the demand for in-kind transfers. We conclude our analysis by testing our model’s predictions using cross-sectional data for 20 European countries. Our descriptive analysis shows that institutional trust is associated with a higher probability of supporting public investments in training programmes for the unemployed (in-kind transfers) over unemployment benefits (cash transfers). This relationship is larger in countries with higher-quality public services. ... Read more

https://doi.org/10.1007/s10797-025-09908-6. Equation (15) in the original version of the article contains a typo. In the main manuscript, Equation (15) currently reads: It should read: The same typo appears in the Online Appendix in Eq. (E.9) and in the sentence on page 16: “The optimal demand for g∗ […] is derived directly in the main body of the paper as g∗ = tp − T∗”, which should read “The optimal demand for g∗ […] is derived directly in the main body of the paper as g∗ = t(p − µ) − T∗.” The Corresponding Author takes full responsibility for the typos and apologises. The original article has been corrected. ... Read more