Current Research 

(With Joseph Francois, Bernard Hoekman and Miriam Manchin) [Submitted, Under Review]

Abstract: The literature on preferential trade agreements (PTAs) has mainly focused on the trade effects of non-trade provisions (NTPs) addressing environmental or labor policies. Using a dataset covering more than 120 countries and several decades, we employ a synthetic differencein-difference estimator to study whether such provisions are accompanied with changes in associated sustainability-related performance indicators in developing countries. Given the importance the EU places on using trade to pursue sustainability goals, we differentiate between EU and other PTAs. The analysis is motivated by two arguments that have influenced the design of deep PTAs: (i)

legally enforceable NTPs are more effective in improving nontrade outcomes in developing countries; and (ii) acceptance of NTPs will be accompanied by (more) aid from high-income partners. We find limited support for these arguments. Enforceable provisions have no effect on performance indicators, whereas non-enforceable provisions in non-EU PTAs are accompanied by deterioration in several outcome measures. Enforceable provisions are associated with less aid; we only find a positive relationship between EU aid and nonenforceable environmental provisions.

Already circulated as: Non-Trade Provisions in Deep Trade Agreements and Non-Trade Outcomes (WB WP series); Pursuing Environmental and Social Objectives through Trade Agreements (EUI Cadmus, CEPR, CITP)

CEPR ebook Paper OSF repo

(With Bernard Hoekman, Marco Sanfilippo and Rohit Ticku) [Accepted, Journal of Development Economics]

Abstract: Public procurement in developing countries accounts for approximately 13% of GDP, offering a significant market opportunity for domestic firms, especially those facing demand constraints. This paper explores the impact of selling to government entities on firm performance in Uganda, leveraging detailed administrative data and an event study methodology that corrects for firm self-selection and heterogeneity in the timing of the treatment. The results reveal that while firms increase their total sales after entering government contracts, they do not see improvements in productivity (measured by value added per worker). Moreover, sales to non-government

customers drop sharply. This reallocation effect is persistent. The effect is not observed when firms sell to large private-sector buyers. Drawing on an original survey of firms engaged in public procurement, we highlight potential drivers of these dynamics, including capacity constraints and profitability differences between public and private sector sales

Paper IGC Policy Brief Data

(With With S. Bolatto and M. Tremuli) [Submitted, Under Review]

Abstract: International trade in waste and scrap has expanded rapidly, yet it remains under-explored in quantitative trade economics. This paper analyzes the determinants of waste flows across five major material categories (plastics, paper, glass, iron & steel, aluminum) using a gravity model enriched with bilateral tariff and non-tariff measures. Leveraging HS6 bilateral customs data for 2001–2022, we compare trade elasticities between waste and non-waste products within the same HS2 sectors to assess whether differences in trade patterns are material-specific. The results —robust to dynamic lead-lag specifications assessing systematic anticipation, and to heterogeneity analyses by income level— are consistent with three distinct archetypes: information-sensitive materials (plastics, aluminum), where technical NTMs display a pattern suggestive of certification mechanisms; complementarity-driven materials (paper), which exhibit inverted tariff elasticities reflecting technological lock-in in specialized recycling infrastructure; and commodity-like materials (glass, iron/steel), where trade responds conventionally to policy and geographic frictions. Our findings point to the value of tailoring trade and circular-economy measures to material characteristics, given the systematic differences observed across archetypes.

Paper OSF repo