A Measurable Decline, With Important Caveats
Poverty across the European Union fell meaningfully between 2015 and 2023, according to a peer-reviewed study published on 30 September 2026 in the open-access journal PLOS One. The research, conducted by Silvia Megyesiova and Cecilia Olexova of the Bratislava University of Economics and Business in the Slovak Republic, offers one of the more rigorous recent assessments of social deprivation across EU member states. It draws on ten distinct metrics derived from the United Nations’ 2030 Agenda for Sustainable Development, moving well beyond the single-income proxies that have traditionally dominated poverty analysis.
The headline figures are encouraging. The share of persons at risk of monetary poverty after social transfers declined from 17.4 per cent in 2015 to 16.2 per cent in 2023. The average severe material and social deprivation rate fell more sharply still, dropping from 9.7 per cent to 6.8 per cent over the same period. These are not trivial improvements, and they reflect sustained policy effort across much of the bloc during a period that also absorbed the considerable economic disruption of the COVID-19 pandemic.
Yet the aggregate trend conceals a geography of persistent vulnerability that deserves serious attention from policymakers and financiers alike. The study’s authors are careful to note that multidimensional poverty profiles evolve over time and do not necessarily follow the traditional north-south or east-west divisions that analysts often reach for as shorthand. The picture is more granular — and more instructive — than broad regional generalisations suggest.
How the Research Was Structured
The methodological approach distinguishes this study from cruder assessments. Rather than relying solely on income thresholds, Megyesiova and Olexova employed ten indicators spanning monetary poverty risk, housing cost overburden, severe material and social deprivation, digital skills, and exposure to environmental threats, among others. This multidimensional framework reflects a growing consensus among serious researchers that poverty is not reducible to a single economic variable, particularly after the pandemic exposed the extent to which physical infrastructure, information literacy, and housing conditions shape social vulnerability in ways that income data alone cannot capture.
The analysis was structured across three distinct time spans: the pre-pandemic period from 2015 to 2019, the year 2020 treated in isolation given the extraordinary disruption of the pandemic, and the recovery and post-recovery period from 2021 to 2023. This periodisation allows for a more honest reading of which improvements preceded the pandemic shock and which emerged from — or survived — it. It is a sensible design choice that strengthens the study’s policy relevance.
Country-Level Divergence
The country-level data reveal both genuine progress and stubborn divergence. In 2015, Spain, Lithuania, Latvia, and Romania all recorded poverty rates exceeding 22 per cent. By 2023, only Estonia and Latvia remained in that category, suggesting that several of the bloc’s historically weaker performers have made real structural gains. Romania’s improvement stands out: its monetary poverty rate fell from 25.4 per cent to 21.1 per cent over the period, representing a substantial reduction even if the absolute level remains high by EU standards.
The clustering analysis for the 2021 to 2023 period identifies Bulgaria, Greece, Romania, and Latvia as forming the most vulnerable grouping within the union. Several countries in Northern and Western Europe, by contrast, display considerably more stable trends, with lower baseline rates and less volatility across the three periods studied. The persistence of this vulnerable cluster — despite years of EU cohesion funding and social policy coordination — raises legitimate questions about the efficacy of top-down redistributive mechanisms and whether more targeted, country-specific frameworks might deliver better outcomes.
The Broader Policy Context
The study situates its findings within a global challenge of considerable scale. As of 2021, approximately 1.3 billion people worldwide lived in poverty, with a further 240 million classified as extreme working poor — individuals in employment but still unable to meet basic material needs. The EU’s internal record, while imperfect, compares reasonably well against this backdrop, and the downward trend in deprivation rates across the period studied reflects the relative stability of European labour markets and welfare institutions.
The authors draw a pointed conclusion from their data: poverty and social exclusion across EU countries cannot be understood through economic prosperity alone. Housing conditions, digital skills, and material deprivation each contribute independently to social vulnerability, and any credible policy response must account for this complexity. The findings also challenge the assumption that economic convergence within the bloc will automatically translate into social convergence — the most vulnerable cluster in 2021 to 2023 includes countries at very different stages of economic development, suggesting that the relationship between GDP growth and poverty reduction is neither linear nor guaranteed.
What the Data Implies for Policy
The study’s authors argue explicitly for sustainable policy responses tailored to each member state’s individual circumstances, rather than uniform EU-wide prescriptions. This is a position that sits comfortably with an empirical reading of the evidence. The divergence between member states in both poverty levels and trajectories suggests that the conditions driving vulnerability in Bulgaria differ materially from those at work in Latvia or Greece, and that policy instruments calibrated for one context may perform poorly in another.
For policymakers and institutional investors assessing social risk within the EU, the study offers a useful corrective to overly optimistic readings of aggregate progress. The bloc has made genuine headway since 2015, and that should be acknowledged. But the persistence of a clearly identifiable vulnerable cluster, the uneven absorption of pandemic-era shocks, and the limitations of income-only metrics all point toward the need for more sophisticated, evidence-driven approaches to social policy — ones that take seriously the multidimensional nature of deprivation rather than reaching for convenient single-variable proxies.