Italian network model: Shock propagation and policy evaluation

Banca d'Italia researchers Fabio Bertolotti, Andrea Linarello, and Patrick Zoi developed a quantitative framework to analyze shock propagation within Italy's production network.

Building on Baqaee and Rubbo (2023), their model incorporates trade openness and endogenous labor supply, calibrated with detailed input-output data for 98 sectors.

The study highlights how network linkages significantly amplify shocks to upstream industries, while dampening those to downstream sectors, compared to an 'island' economy without inter-sectoral connections.

For instance, a 1% productivity shock in electricity generation, a central upstream sector, raises aggregate GDP by 0.06% in the network model, three times more than in an 'island' economy.

This amplification occurs as lower electricity prices reduce marginal costs across numerous downstream industries, propagating productivity gains.

The framework also captures non-linearities, showing that linear approximations systematically underestimate aggregate impacts, especially for large shocks.

The model was applied to four distinct policy scenarios, consistently demonstrating the substantial role of network effects in aggregate outcomes.

These included R&D tax credits, entry regulation reforms, a spike in imported oil and gas prices, and AI adoption.

In all cases, network forces accounted for a significant share of the aggregate effects.

For instance, an R&D tax credit increase led to a 0.11% GDP rise, with 35% from network spillovers.

The study highlights how policy impacts are amplified or redirected through the interconnected production structure, making network considerations crucial for effective policy design.