Click here to find my research statement
(Abstract). Modern economies are organised as dense production networks, yet the transmission of shocks is conventionally understood through observed Input-Output linkages and vertical cascade effects. This paper argues that propagation additionally operates through horizontal complementarities: the extent to which sectors share upstream suppliers or downstream buyers. These interdependencies shape a latent network of connections, linking also otherwise unconnected sectors, and generate simultaneous responses to idiosyncratic shocks through structural demand and supply similarities. Because it moves Domar weights, horizontal propagation is hidden in Hulten's theorem while governing the effect of sectoral volatility on aggregate output: horizontal reallocation among sectors shifts the average level of GDP. This new channel complements vertical propagation, offering a unified view of comovement rooted in the horizontal geometry of common interactions. Using Input-Output data, I show that sectoral responses vary systematically with horizontal relations, forging novel implications for the transmission of sectoral and aggregate shocks.
JEL: C67, D57, E32, J21
KEYWORDS: production networks, Input-Output economy, horizontal transmission, network distance, sectoral comovement, aggregate fluctuations
Presented at: 25th Research in International Economics and Finance (RIEF) meeting, Université Paris-Saclay, 39th Annual Irish Economic Association, Central Bank of Ireland, Catholic University of Milan, "La Sapienza" University of Rome, "Tor Vergata" University of Rome, University of Milano-Bicocca and University of Oxford.
(Abstract). This paper examines the determinants of U.S. wage inequality between industries. It distinguishes two channels of Skill-Biased Technological Change: a quantity effect, mirroring variations in the distribution of capital and labour inputs, and a structural effect, reflecting differences in substitution elasticities across factors of production. Counterfactual exercises show that evolving technological heterogeneity in substitution patterns dominates wage dispersion. A quantitative decomposition reveals that structural forces, mostly reflecting substitutability of routine with non-routine labour, account for 94% of observed wage variance, exacerbated by stronger sorting and segregation effects. Upon neutralising structural differences, the quantity channel reckons merely 6-15%. Indeed, structural heterogeneity, rather than mere changes in factor quantities, is the dominant force behind U.S. wage inequality.
JEL: E24, J31, J82, L16
KEYWORDS: wage inequality, technological change, labour force composition, elasticity of substitution
Presented at: Catholic University of Sacred Heart, Roma Tre University, University of Milan, University of Milano-Bicocca, Naples School of Economics: 4th PhD and Post-Doctoral Workshop, University of Naples "Federico II", and University of Pavia.
(Abstract). Production networks shape aggregate supply not only through whom firms buy from, but also through whom they sell to. Studying the final destination of sectoral output makes the presence of a network a general equilibrium force. We introduce final demand proximity: the extent to which a sector's output reaches final consumers directly or indirectly through other sectors. Given that firms sell both inside the network, to other firms, and outside it, to final demand, price-setting decisions reallocate demand across these two margins. This makes marginal revenues and costs jointly network dependent, endogenizing the role of the production network. We show this mechanism in a simple framework and then formalize it in a multi-sector model. Final demand proximity shapes the slope of sectoral supply curves, and therefore also of aggregate supply. We then exploit country-sector variation within the European Monetary Union to test this prediction. Consistently with the theory, we find that supply curve slopes vary systematically with both network participation and final demand proximity. While the effect of participation is monotonic, the effect of final demand proximity is non monotonic.
JEL: E31, E32, E52, F41
KEYWORDS: production networks, real rigidities, inflation dynamics, supply curve
Presented at: Banque de France, Federal Reserve Board, International Association for Applied Econometrics (IAAE) Annual Conference 2026, Nova School of Business and Economics, 2nd Macroeconometrics in Salerno Workshop, University of Salerno, and Catholic University of Sacred Heart.
Firms Compete, Networks Shape: The Relational Origins of Market Power (draft coming soon)
(Abstract). Market power is a property of the network of alternatives around a buyer, not only of the firm alone. An Input-Output system records these alternatives directly. This paper uses this structure to develop a theory in which markups arise from the interaction between a supplier's dominance within the relevant market and the buyer's ability to escape that market by reorganizing production. The theory shows that (i) markups decompose into a vertical component, determined by the number of rivals, and a horizontal component, determined by relative dominance of co-suppliers; (ii) greater dispersion in downstream expenditure shares creates market power rather than reallocating it, even when the number of suppliers is unchanged; (iii) within each market the ranking of markups coincides with the ranking of shares. These network origins of market power generate factor misallocation and policy-induced markup changes. U.S. Input-Output data confirm that horizontal dominance is quantitatively important for markup dispersion.
JEL: C67, D4, D43, D57, L11, L14
KEYWORDS: production networks, market power, Input-Output table, buyer-supplier structure, horizontal dominance, endogenous markups, misallocation, transition policy
Distributional Foundations of Input-Output Economies
Where to Sit in the Chain: Microfoundations and Micro-to-Macro Implications of Global Value Chain Positioning, with Siena D.
Job Vacancies through COVID-19 Lockdowns in the Euro Area, with Bertucci A.