Structural Funds, Misallocation and Regional Disparities

Technical Summary: Structural Funds, Misallocation, and Regional Disparities

Problem and Motivation
Despite the substantial expansion of European Union (EU) structural funds over the first two decades of the 2000s, empirical evidence documents a persistent increase in GDP per capita dispersion across EU regions (σ-divergence). This counterintuitive finding challenges the conventional view that cohesion policy effectively reduces territorial disparities. Existing literature offers mixed results regarding the impact of structural funds on regional convergence, with effects often found to be heterogeneous and conditional on institutional quality or absorptive capacity.

The authors propose that this divergence may be explained by the unintended consequences of large-scale public transfers on resource allocation. Specifically, they hypothesize that while structural funds may directly reduce income gaps, they simultaneously induce misallocation—the inefficient allocation of productive resources across regions. By potentially redirecting production factors toward low marginal productivity regions, structural funds could worsen allocative efficiency, thereby generating a divergence channel that offsets their redistributive benefits. This paper aims to reconcile the evidence of increasing dispersion with the expansion of transfers by explicitly modeling the interplay between EU funds, regional misallocation, and income convergence.

Methodology
The study utilizes a harmonized panel dataset of NUTS-2 regions covering 26 EU countries from 2000 to 2022. The analysis integrates data from the EU Cohesion Policy Open Data Portal (for fund disbursements) and the European Commission's Annual Regional Database (ARDECO) for economic indicators.

• Variables:

  • Dependent Variable: The GDP per capita gap (GDPpc GAP), defined as the difference between the average real GDP per capita of the 5% richest regions (the frontier) and the specific region's GDP per capita.
  • Misallocation Measure: Adapted from Hsieh and Klenow (2009), misallocation is measured as the absolute log deviation of regional Revenue Total Factor Productivity (TFPR) from the EU-wide mean. This metric captures the dispersion of the marginal revenue product of inputs across regions; higher dispersion indicates greater allocative inefficiency.
  • Treatment: Aggregated annual modeled expenditure of all EU structural funds (including ERDF, CF, ESF, EAFRD, etc.) at the regional level.

• Econometric Strategy:
  The authors estimate a simultaneous equation model using Three-Stage Least Squares (3SLS) with region and year fixed effects. The system consists of two equations:

  1. Misallocation Equation: Explains regional misallocation as a function of lagged EU funds and lagged GDP gaps.
  2. Convergence Equation: Explains the GDP per capita gap as a function of lagged misallocation and lagged EU funds.

  Identification: To address endogeneity, the study employs a quasi-experimental identification strategy based on the institutional eligibility thresholds for Objective 1 status (regions with GDP per capita below 75% of the EU average). A binary instrument (Eligibility) is constructed based on pre-program income conditions relative to the EU average, following the framework of Becker et al. (2010). This instrument captures the exogenous variation in access to funds induced by the EU's assignment rules, rather than persistent income differences.

Key Results
The estimation yields three primary findings regarding the direct and indirect channels of structural funds:

1. Offsetting Channels: EU structural funds operate through two opposing mechanisms.

   • Direct Effect: Funds have a negative effect on the GDP gap, directly promoting convergence (reducing income disparities).
   • Indirect Effect: Funds increase regional misallocation. Since increased misallocation widens the GDP gap, this channel acts as a divergence force.
   • Net Impact: In the full sample, the indirect divergence channel absorbs approximately 85% of the direct convergence effect. While the total effect remains slightly negative (convergence-enhancing), the magnitude is drastically reduced by the allocative distortions induced by the transfers.

2. Heterogeneity by GDP Quartile: The relationship between funds, misallocation, and convergence varies significantly across the income distribution (analyzed by quartiles of the GDP gap):

   • Frontier Regions (Q1 & Q2): Here, the direct effect of funds is positive, meaning transfers are associated with widening the GDP gap. This aligns with a Solow-type mechanism where richer regions, facing less capital scarcity, utilize funds primarily for productivity (TFP) growth, moving further ahead. However, in these regions, the indirect effect of misallocation is convergence-enhancing (negative), partially offsetting the divergence.
   • Middle/Lagging Regions (Q3): This quartile exhibits the clearest convergence pattern. The direct effect is negative (convergence), driven by capital accumulation (factor deepening). However, the indirect effect is divergence-enhancing (positive), as misallocation imposes higher costs on regions further from the frontier. The indirect channel offsets about 65.6% of the direct gain.
   • Most Lagging Regions (Q4): Neither the direct nor the total effect is statistically robust, suggesting that for the most structurally weak regions, transfers alone are insufficient to generate a measurable convergence effect.

3. Explaining σ-Divergence: The aggregate increase in misallocation, particularly its concentration in lagging regions where it exacerbates income gaps, provides a mechanism for the observed σ-divergence despite increased funding.

Significance and Contributions
The paper contributes to two main strands of literature: the empirical analysis of EU structural funds and regional convergence, and the literature on misallocation and aggregate productivity.

• Reconciling Mixed Evidence: The study offers a unified explanation for the heterogeneous and often inconclusive findings in previous literature. It suggests that the net effect of cohesion policy depends on the balance between its intended redistributive impact and its unintended distortionary effects on resource allocation.
• Mechanism Identification: By explicitly modeling the misallocation channel, the authors demonstrate that the effectiveness of structural funds is not solely a function of the volume of resources transferred but critically depends on how those resources alter the efficiency of factor allocation.
• Policy Implication: The findings imply that for structural funds to effectively promote convergence, policy design must prioritize mechanisms that relax factor accumulation constraints without exacerbating allocative distortions. In regions far from the frontier, where accumulation is key, funds must be managed to avoid the "good" inefficiency that accompanies productivity growth in frontier regions, which can otherwise widen disparities.

The authors conclude that the persistence of regional disparities in the EU during the 2000–2022 period is consistent with a scenario where the direct convergence benefits of structural funds are largely neutralized by the indirect costs of increased misallocation.