Social Spending Works Differently Across CEE and Poverty Gains Don’t Always Reduce Inequality

Two countries can devote substantial resources to social protection and still achieve very different outcomes. This is becoming an increasingly important policy problem as European governments face tighter fiscal space, ageing populations and pressure to protect households from repeated economic shocks. The challenge is no longer simply to spend more, but to make welfare systems work harder for every euro committed.

A new study, "Assessing the Efficiency and Total Factor Productivity of Social Protection Expenditure: A Study of CEE Countries," published in the MDPI journal Economies by Maya Tsoklinova, tests that proposition across 11 Central and Eastern European EU economies. Its results reveal a clear efficiency divide: Czechia and Slovakia perform strongly on poverty-oriented measures, while other countries lag, and even the leaders do not necessarily perform equally well when the focus shifts to broader income distribution.

High Spending Is Not the Same as High Performance

The study examines Bulgaria, Croatia, Czechia, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia. It uses social protection expenditure as a share of GDP as the input and measures outcomes through three models covering income inequality, standard poverty risk, severe poverty and the income share of the poorest 40%.

Across the region, average social protection expenditure rose from 18.25% of GDP in 2016 to 20.49% in 2020, reflecting the pandemic-era fiscal response, before easing to 19.21% in 2023. Over the same period, the average Gini coefficient declined from 30.75 to 29.49, while the at-risk-of-poverty rate fell from 18.29% to 17.38%. Severe poverty also declined, and the income share of the poorest 40% increased from 20.75% to 21.59%. The figures suggest gradual improvement, but the efficiency analysis reveals large differences in how effectively countries convert spending into outcomes.

In the model combining inequality and standard poverty reduction, Czechia and Slovakia remained on the efficiency frontier throughout 2016–2023. Romania and Hungary also performed at or close to the frontier in most years. By contrast, Slovenia, Croatia and Poland generally recorded the weakest efficiency scores.

The divide is substantial. Slovakia, Romania and Czechia achieved an average technical-efficiency score of 1.00 in the study's first poverty-oriented model, compared with 0.81 for Poland, 0.80 for Croatia and 0.78 for Slovenia. The gap suggests that similar commitments to social protection can produce markedly different outcomes depending on how welfare systems are designed and administered.

This does not mean lower-scoring countries should simply cut expenditure. Data Envelopment Analysis measures relative efficiency, not whether benefits are socially adequate. A system can be efficient relative to its peers while still leaving some households vulnerable. The more useful policy insight is that expenditure levels alone cannot explain the results.

Poverty Reduction and Redistribution Are Not the Same Policy Victory

Under the first two models, which focus on standard and severe poverty, Czechia and Slovakia perform exceptionally well. But when the analysis replaces poverty measures with the income share of the bottom 40%, Czechia falls from the efficiency frontier. Slovakia remains fully efficient, while Romania, Estonia and Hungary continue to perform strongly.

The shift exposes an important weakness in the way welfare systems are often assessed. A country can be effective at keeping people above a poverty threshold without substantially increasing the proportion of national income flowing to poorer households. Poverty alleviation may prevent deprivation; redistribution concerns the relative position of lower-income groups within the wider economy. They overlap, but they are not identical objectives.

For policymakers, this distinction matters enormously. A programme designed to reduce extreme hardship may look highly successful under poverty indicators but less impressive when judged by whether poorer households gain a stronger share of economic resources.

The study warns against treating one headline social indicator as a proxy for overall welfare-system performance. It also strengthens the case for better targeting. Previous evidence discussed in the paper links weaker performance in some countries to broad or insufficiently targeted programmes. Poland, for example, ranks relatively poorly despite significant social programmes, while Croatia's performance may partly reflect weaker targeting of transfers. The study itself does not prove these mechanisms, but the pattern points policymakers toward programme design rather than budget size alone.

COVID-19 Revealed How Quickly Welfare Productivity Can Break and Recover

The 2016–2023 period allows the study to capture an unusual stress test: the sudden economic and social disruption of COVID-19. Before the pandemic, productivity in the poverty-oriented models remained broadly stable. Then came a sharp break. During 2019–2020, every country recorded a decline in productivity in the first poverty model, with Lithuania, Slovenia, Estonia and Czechia among the hardest hit.

Crucially, the decline was driven mainly by what the study calls "technological change", a movement in the empirical efficiency frontier, rather than widespread deterioration in day-to-day technical efficiency.

Translated into policy language, the region's welfare systems were suddenly operating in a much more difficult environment. Governments were forced to expand support rapidly while household incomes, employment and poverty risks deteriorated simultaneously. Existing systems were being asked to do more under conditions for which they had not been designed.

The recovery is equally instructive. Between 2020 and 2023, productivity improved in most countries. Croatia, Lithuania, Slovenia and Romania recorded particularly strong growth under the first poverty model, with shifts in the productivity frontier providing much of the improvement.

The study is careful not to claim that particular pandemic reforms caused this recovery. Its methodology cannot identify individual policies. But the broader pattern raises a significant research question: did emergency digitization, administrative reform, new benefit mechanisms or EU-supported interventions leave behind a more adaptable welfare architecture?

This deserves closer investigation because resilience is becoming a core social-policy requirement. Future crises, whether financial, geopolitical, demographic or climate-related, will test welfare systems not only on generosity but also on speed, targeting and administrative flexibility.

The Real Reform Agenda Is Institutional, Not Merely Fiscal

Perhaps the most consequential result is that scale efficiency is already high across the three models. Most countries appear to operate relatively close to their optimal scale, suggesting that the largest gains may not come simply from expanding or shrinking the volume of social protection expenditure.

It shifts attention toward institutional quality. How benefits are designed, who qualifies, how accurately governments identify vulnerable households, how programmes interact with labour markets and how effectively administrations deliver payments may be more important than marginal changes in expenditure alone.

For EU institutions, the finding supports moving beyond comparisons based primarily on social expenditure as a percentage of GDP. Outcome-based benchmarking can identify cases where similar resources produce markedly different results. For governments, raising welfare budgets cannot substitute for improving welfare architecture.

The study also carries wider relevance for emerging and developing economies. Many Global South governments face the simultaneous challenge of expanding social protection coverage while operating under severe fiscal constraints. The CEE evidence suggests that building administrative capacity, targeting systems and effective institutions should accompany efforts to mobilize additional spending.

There are important limitations. The study measures contemporaneous relative efficiency and does not establish that social protection expenditure caused the observed poverty or inequality outcomes. Macroeconomic conditions, employment, demographics, tax structures and other policies can influence results.

Regardless, the study sharpens a crucial debate in social policy. A welfare state cannot be judged by the size of its budget alone; it must also be judged by what that budget achieves, for whom, and under what conditions.