Poverty is not a static condition but a dynamic process, and measuring it as such reveals challenges that annual snapshots cannot capture (Giarda and Moroni 2018, Bosco and Poggi 2020, Polin and Raitano 2014). Research has consistently found evidence of ‘state dependence’ – being poor today significantly raises the likelihood of being poor tomorrow – linked to structural factors (Bosco and Poggi 2020, Mussida and Sciulli 2022). While most households escape poverty relatively quickly in its early stages, the probability of exit declines sharply with duration, and poverty becomes increasingly entrenched over time, particularly among households with dependent children, young or elderly heads, and those exposed to employment shocks (Andriopoulou and Tsakloglou 2011). 

Significant cross-country variation exists in the duration and persistence of poverty despite similar headline rates, with welfare regimes playing a decisive role in shaping both incidence and persistence (Vaalavuo 2015). Evidence also suggests that the Great Recession amplified the scarring effects of poverty, increasing state dependence between the pre- and post-crisis periods (Mussida and Sciulli 2022, while more recent shocks such as COVID-19 risk entrenching these dynamics further, given evidence of substantial and uneven rises in poverty across European countries in the wake of pandemic-induced income losses (Palomino et al. 2020).

Most of previous research focuses on persistent poverty and is confined to specific time periods or countries. Our recent working paper (Sandor et al. 2026) applies several dynamic poverty measures to longitudinal EU-SILC data covering 2011–2022 across 25 Member States (Germany and Luxembourg excluded due to data limitations). Rather than relying on a single year’s income snapshot, we track individuals over four years and construct three complementary measures. The first classifies individuals into six poverty trajectories: transient poverty (poor for one year out of four), almost always poor (three out of four years), chronic poverty (all four years), and three cases covering patterns of two in four years (exiting poverty, in-and-out poverty, entering poverty). The second captures the incidence of poverty over time: how many people experience poverty across the observation window, and for how long (Foster 2009). The third incorporates both incidence and depth, capturing how far below the poverty line people fall and for how long (Duclos et al. 2010, Bresson et al. 2019). 

For all three measures, we first determine whether an individual is at risk of poverty in each of the four years, using the Eurostat poverty line: equivalised disposable income below 60% of the national median. The dynamic measures are then built from this yearly at-risk-of-poverty status. Because our indicator for at risk of poverty uses the longitudinal EU-SILC sample, it will not exactly match Eurostat’s official cross-sectional at-risk-of-poverty rate for the same country-year. The gap comes from differences in sample composition (the longitudinal panel is a rotation- and attrition-affected subset of the cross-sectional sample) and weighting.

A much larger share of Europeans experience poverty than headline rates capture

The standard at-risk-of-poverty rate for the EU stood at 16.2% in 2022. Yet when poverty is measured dynamically – asking not ‘who is poor today’ but ‘who has been poor at any point over the past four years’ – the picture changes substantially. Figure 1 shows that nearly a quarter (24%) of the EU-25 population (excluding Germany and Luxembourg) experienced monetary poverty at least once over 2019–2022, well above the headline figure. Several member states exceed 30%, including Bulgaria, Estonia, Greece, and Romania. 

Figure 1 Headline at-risk-of-poverty rate (2022) vs at risk of poverty at least once in the last four years, EU member states, 2019–2022

Notes: Each dot represents an EU member state. The dashed line is the 45-degree line; all countries lying above it have a higher share of people experiencing poverty at least once over 2019–2022 than their 2022 headline at-risk-of-poverty rate would suggest. The vertical axis shows the share of the population that was poor at least once over the 2019–2022 period; the horizontal axis shows the standard at-risk-of-poverty rate in 2022, defined as living below 60% of the national median equivalised disposable income. 
Source: Authors’ calculations based on 2023 EU-SILC longitudinal data and Eurostat at-risk-of-poverty rates.

This gap between the static and dynamic picture is not a measurement artefact; it reflects the reality that different people can cycle in and out of poverty at different times, which remains largely invisible to annual snapshots. Across the EU, around 18% of Europeans were poor for at least two of the four years, 12% for at least three, and 7% for all four.

Persistent poverty dominates, but trajectories vary widely

Figure 2 shows that around 45% of those who experienced poverty at any point over 2019–2022 were chronically or almost always poor. And only around a third experienced purely transient poverty. Country-level heterogeneity is striking: Bulgaria, Romania, and Slovenia have high shares of chronic poverty. By contrast, Hungary, Ireland, and Slovakia have relatively high shares of transient poverty, suggesting that for many of their poor, poverty is a temporary condition rather than an entrenched state. 

Figure 2 Poverty trajectories across EU member states, 2019–2022 (share of total population)

Notes: Stacked bars show the share of the total population in each poverty-trajectory category over the 2019–2022 period. ‘Chronic poverty’ refers to being poor in all four years; ‘almost always in poverty’ to three out of four years; ‘entering poverty’ to being poor only in the final two years; ‘exiting poverty’ to being poor only in the first two years; ‘in-and-out poverty’ to alternating between years in poverty and out of poverty; and ‘transient poverty’ to being poor in exactly one year. Poverty is defined using the standard 60% of national median equivalised disposable income threshold. Countries are ordered by the share of chronic poverty. 
Source: Authors’ calculations based on 2023 EU-SILC longitudinal data.

Country comparisons reveal important nuances beyond simple incidence rates. For example, Belgium and Ireland, despite similar overall poverty incidence, display very different trajectory compositions: Belgium’s poor skews toward chronic and persistent poverty; Ireland’s toward transient. This heterogeneity has direct implications for policy design. A country dominated by transient poverty calls for different interventions than one with entrenched chronic disadvantage.

Structural factors shape poverty trajectories

To identify which characteristics are most strongly associated with different poverty trajectories, we estimate a multinomial logistic regression across seven trajectory outcomes (the six poverty trajectories defined above, plus a ‘never poor’ reference category) using a sample of 71,573 individuals from the 2019–2022 EU-SILC wave, controlling for country fixed effects and measuring all covariates at baseline.

Labour market status and education are strong predictors of poverty trajectories (Figure 3). Being unemployed is associated with a 30 percentage point lower probability of never experiencing poverty, and a 13 percentage point higher probability of chronic poverty. Being out of the labour force carries substantial but lower risks, at roughly half the magnitude of unemployment. Higher education is the strongest negative correlate, associated with a 17 percentage point higher probability of never being poor and a 6 percentage point lower probability of chronic poverty. 

Figure 3 Average marginal effects of labour market status and educational attainment on poverty trajectories, EU25, 2019–2022

Notes: Points show average marginal effects (percentage-point changes in the probability of belonging to each poverty-trajectory category) estimated from a multinomial logit. Blue markers refer to individuals who were never in poverty during the observation period, while orange markers refer to those experiencing chronic poverty. Error bars indicate 95% confidence intervals. Positive values indicate a higher probability of belonging to the respective poverty category relative to the reference group, while negative values indicate a lower probability. Reference categories are employed and low education. N = 71,573.
Source: Authors’ calculations based on 2023 EU-SILC longitudinal data.

Beyond labour market and education, housing tenure matters considerably. Renters have a 5 percentage point higher probability of chronic poverty relative to outright owners. Rural residents have a 3 percentage point higher probability of chronic poverty than urban dwellers. EU citizenship is linked to a 16 percentage point higher probability of never being in poverty and a 4 percentage point lower probability of chronic poverty. These associations are most pronounced at the extremes – never poor and chronic poor – but intermediate trajectories follow similar patterns, with effects going in the same direction though smaller in magnitude. 

Social protection matters, but the degree varies enormously

Comparing poverty rates before and after social transfers reveals substantial reductions in dynamic poverty across the EU. The proportion of the population experiencing poverty in at least two out of four years drops from around 25% before social transfers to approximately 18% after. Importantly, social transfers are relatively more effective at reducing long-duration poverty (four years in four) than at simply keeping people above the poverty line in a single year. 

Figure 4 shows that country-level variation is striking in this aspect too, with Ireland standing out again: its dynamic poverty rate is nearly halved after transfers. Czechia, Denmark, Finland, France, and Sweden also show large reductions. In contrast, Bulgaria, Greece, Italy, and Romania see comparatively smaller impacts – gaps that are especially visible in the most vulnerable/fragile group (poor in all four years). 

Figure 4 The dynamic risk of poverty before and after social transfers for the chronically poor (poor in four out of four years) in member states, 2014–2022

Notes: Each panel shows the share of the population that is poor in all four years (those in ‘chronic poverty’) before and after social transfers, for each member state over the 2014–2022 period. The gap between the two areas reflects the poverty-reducing effect of social transfers. Before-transfer income excludes pensions, derived from EU-SILC variable HY022. Poverty is defined using the standard 60% of national median equivalised disposable income threshold. 
Source: Authors’ calculations based on EU-SILC longitudinal datasets from 2015 to 2023.

While these comparisons cannot fully isolate the effect of transfers from other country differences, they suggest that social protection systems differ markedly in how closely they track sustained low income. This is particularly relevant in light of evidence that welfare regimes shape not only current poverty dynamics but their intergenerational transmission, with countries where persistent poverty is less effectively addressed tending to exhibit stronger intergenerational poverty associations (Bavaro et al. 2024).

Policy implications

These findings are directly relevant as the European Commission adopted the first ever EU Anti-Poverty Strategy in May 2026. Progress has been made in reducing the at-risk-of-poverty rate, but understanding why this pace remains modest requires looking beneath the headline rate. The relevant question is not only how many are poor, but how persistently, how deeply, and which groups are most at risk of entrenched disadvantage, and ultimately, which policies have proven more effective at dislodging these disadvantages. Dynamic poverty measurement is not merely a statistical refinement; it offers a clearer picture of who is at greatest risk of being left behind. 

Standard poverty indicators remain valuable for tracking overall progress, but complementing them with dynamic measures can improve the targeting and design of anti-poverty interventions. Transient and persistent poverty are likely to call for different kinds of intervention. Short-term income shocks point toward emergency income support and improved benefit access. Persistent poverty is more closely linked to structural factors requiring longer-horizon investment such as education, childcare, healthcare, place-based activation, and minimum income schemes with adequate coverage.

A policy framework that does not distinguish between these poverty trajectories may not achieve the full impact of its resource allocation. Our analysis also underscores that the effectiveness of social protection cannot rely on a single strategy. Anti-poverty policy requires a combination of income transfers, human capital investment, and structural reforms, with the right mix depending on each country’s trajectory composition and institutional context. Crucially, dynamic poverty analysis equips policymakers to identify not only who is currently poor, but how long they have been in poverty, how deeply they are affected, and who is at risk of becoming trapped – and to intervene before persistence sets in.

Authors’ note: The views expressed are those of the authors and do not represent the official position of the European Commission.

References

Andriopoulou, E, and P Tsakloglou (2011), “The determinants of poverty transitions in Europe and the role of duration dependence”, IZA Discussion Paper No. 5692.

Bavaro, M, R Carranza, and B Nolan (2024), “Intergenerational poverty persistence in Europe and the Great Gatsby curve”, VoxEU.org, 28 October.

Bosco, B, and A Poggi (2020), “Middle class, government effectiveness and poverty in the EU: A dynamic multilevel analysis”, Review of Income and Wealth 66(1): 94–125.

Bresson, F, J-Y Duclos, and F Palmisano (2019), “Intertemporal pro-poorness”, Social Choice and Welfare 52: 65–96.

Duclos, J, A Araar, and J Giles (2010), “Chronic and transient poverty: Measurement and estimation, with evidence from China”, Journal of Development Economics 91: 266–77.

Foster, J E (2009), “A class of chronic poverty measures”, in T Addison, D Hulme, and R Kanbur (eds.), Poverty Dynamics: Interdisciplinary Perspectives, Oxford University Press.

Giarda, E, and G Moroni (2018), “The degree of poverty persistence and the role of regional disparities in Italy in comparison with France, Spain and the UK”, Social Indicators Research 136(1): 163–202.

Mussida, C, and D Sciulli (2022), “The dynamics of poverty in Europe: What has changed after the Great Recession?”, Journal of Economic Inequality 20: 915–37.

Palomino, J C, J G Rodríguez, and R Sebastian (2020), “Inequality and poverty effects of the lockdown in Europe”, VoxEU.org, 16 June.

Polin, V, and M Raitano (2014), “Poverty transitions and trigger events across EU groups of countries: Evidence from EU-SILC”, Journal of Social Policy 43(4): 745–72.

Sandor, A M, S Da Costa, and A K Pisiotis (2026), “Poverty dynamics in the EU”, European Employment and Social Outlook Working Paper 2026/02, European Commission.

Vaalavuo, M (2015), “Poverty dynamics in Europe: From what to why”, Publications Office of the European Union.