Measuring fiscal disparities and equalization performance across Czech regions

Aim/purpose – This article evaluates the proposed reform of the Czech regional tax-sharing system by examining how replacing the current system based on historical shares (S0) with the proposed criteria-based formula (S1) would affect the distribution of shared-tax revenues across Czech regions.

Design/methodology/approach – The study uses a static ex ante simulation based on the 2024 shared-tax pool for Czech regions. The same aggregate tax pool is applied under both S0 and S1 to isolate distributive effects. The analysis combines per capita revenue comparisons, standard measures of horizontal fiscal dispersion, a composite Need Index summarizing the priorities embedded in S1, and a basic sensitivity analysis using alternative weighting structures.

Findings – The proposed criteria-based formula (S1) is materially redistributive and changes the relative fiscal position of several regions. Some regions gain substantially, whereas Moravskoslezský and Ústecký record significant losses, indicating that demographic, transport-related, and education-related pressures are only partly reflected in the baseline formula. The reform is not equalizing in terms of standard dispersion measures, and the sensitivity analysis confirms that the results depend materially on the formula’s weighting structure.

Research implications/limitations – The analysis is based on a single-year static ex ante simulation. It does not capture behavioral responses or the full interaction between shared taxes and broader grants and transfers. The Need Index is an internal summary of the priorities embedded in S1 rather than an external measure of comprehensive regional need.

Originality/value/contribution – The article provides an ex ante empirical assessment of a concrete reform of Czech regional tax sharing. It shows that moving from historical shares to a criteria-based formula improves transparency more clearly than it guarantees equalization, and that distributive outcomes depend substantially on how the formula is specified.