Innovation Policy in Production Networks [Paper Link]
Abstract: Governments use R&D incentives to boost economic growth, and increasingly target them at particular sectors. How should this support be targeted when sectors are linked through supply chains? I develop the first tractable theory of R&D-driven growth in a production network, where sectors buy inputs from one another. Compared to a social planner, the market allocates too little R&D to sectors with upstream positions, low markups, or high productivity growth. This misallocation lowers the level and growth rate of output. The upstream result arises because the market power of firms in downstream sectors shrinks the market size for innovation in upstream sectors. A simple set of production and R&D subsidies, which does not depend on the network, restores the efficient allocation. When only R&D subsidies are available, the best policy depends on network position and favours upstream sectors. I calibrate the model to US manufacturing and find that over a quarter of R&D is misallocated. Too little goes to motor vehicle parts, semiconductors and fertilizer, which are relatively upstream, and too much to pharmaceuticals, which is relatively downstream with slow productivity growth. Reallocating R&D efficiently would raise output by 28% and growth by 0.4 percentage points a year in the long run. A model without input-output linkages understates the output gain by a quarter and gets the direction of misallocation wrong for one in six sectors.
The Clean Transition in a Network Economy [Paper Link]
Winner of the European Economic Association's Young Economist Award 2024
Abstract: Governments increasingly provide sector-specific subsidies to support the adoption of clean technologies. Given that interventions in one sector may have knock-on effects on upstream and downstream sectors, how should governments optimally target these subsidies? To address this question, I develop a static model featuring emissions, climate damages, endogenous choice between clean and dirty technologies, and intermediate inputs in production. I find that the total effect of a sector-level intervention on emissions and welfare depends on a direct (i.e., non-network) channel and three network channels. I show that the presence of these network channels may cause well-meaning interventions to backfire and increase aggregate emissions. I calibrate my model to a bloc of 23 advanced OECD economies and find that network channels significantly dampen or reverse the direct channel of clean technology subsidies. As a result, accounting for the network meaningfully changes the set of sectors to target.
Presentations: CEPR Macroeconomics and Growth Annual Symposium 2025, NBER Conference on Decarbonizing Industrial Production 2025, Society for Economic Dynamics Conference 2025, the 10th Annual Conference on Network Science in Economics (2025), the RIEF Network Doctoral Meeting 2025, EEA-ESEM Congress 2024, the Bank of England Macro Brownbag Seminar, the Vigo Workshop on Dynamic Macro 2024, the Warwick Economics PhD Conference 2024, the Society for Economic Dynamics Conference 2023, the EDGE Jamboree 2023
Green Innovation by US Public Firms: Trends and Determinants (with Zeina Hasna, Samuel Pienknagura, Jaden Kim)
Abstract: We investigate empirically the dynamic relationship between state-level policies and the innovation behavior of US publicly listed firms. To this end, we construct a novel dataset of US state-level policies that support renewable energy and promote energy efficiency and combines it with patent records and firm-level financial data. we find that changes in the count of policies are associated with a large, positive and statistically significant increase in green innovation in the medium-term (within three to five years). However, impacts vary by policy and patent type, and depend on the stringency of climate policies at the national level, on state-level variables and on firm characteristics. On average, state-level policies are more effective at stimulating overall innovation and brown efficiency technologies compared to pure green technologies. In addition, we find suggestive evidence that financial incentives stimulate energy efficiency innovation more than regulations over the medium term, especially for pure green and brown efficiency patents. Finally, state-level policies have larger effects when: (1) national environmental policy is less stringent or highly uncertain; (2) states have larger economic size or population; (3) firms have larger stocks of green innovation indicating path dependency holding at the firm-level.
Presentations: ESADE Workshop on Climate Change Macroeconomics 2024
Energy Shocks in Production Networks: Evidence from the UK (with Raphael Abiry)
Abstract: We empirically investigate the impact of gas price shocks on sectoral prices in the UK, and the role of input-output linkages in mediating this process.
Presentations: Bank of England PhD Internship Workshop, Bank of England Structural Economics Division PhD Showcase, Bank of England Brownbag Seminars
Gas2CPI - Scenario Analysis of UK Energy-Price-to-Inflation Dynamics (with Raphael Abiry, Giulia Gardin, and Dooho Shin)
Abstract: This paper introduces Gas2CPI, a bottom-up modelling framework that maps wholesale crude oil, natural gas, and carbon price movements into consumer price inflation. The model captures key institutional features of the UK energy system - including regulated retail price caps, fixed-price components, and pass through between oil and petrol, and gas and electricity prices - and combines direct effects from household energy bills with indirect effects propagated through supply chains. Gas2CPI enables transparent scenario and counterfactual analysis, illustrating how energy shocks can generate persistent inflationary pressures across the CPI basket.
We apply the framework to the recent energy price surge following the Iran war. Using energy futures as of late March and excluding any 'second round' (or general equilibrium) effects - for example operating through wages - we estimate the impact of the energy shock on UK inflation. The initial inflation response is driven by higher household energy bills, while indirect supply chain effects and price adjustments elsewhere in the CPI basket generate more persistent dynamics. Oil price shocks pass through relatively quickly, whereas inflationary impacts from higher gas and electricity prices emerge with a lag.