“Trade and Firm-Level Adjustments to Geopolitical Shifts: Evidence from Armenia”
Latest Version
Winner of the IMF Award for Best Paper in International or Monetary Economics at the 2026 Armenian Economic Association Annual Meetings.
Selected Presentations: Melbourne Empirical Trade Conference, Global Interactions Conference (Bari), ETSG (Milan), Université Paris-Panthéon-Assas, FIW Conference (Vienna), CITP Conference (Bristol), 24th GEP/CEPR Postgraduate Conference (Nottingham), AMIE Conference, BSE Jamboree (Barcelona), EAYE Annual Meeting (Bilbao), RIEF Conference (Paris), CEPR–PSE Policy Forum (Paris), Paris School of Economics, AEA Annual Congress (Yerevan)
This project is conducted within the framework of cooperation between the State Revenue Committee of Armenia and the Armenian Centre for Research in Economics.
Abstract
This paper examines how sanctions on a major trading partner impact the firms and trade of a neutral economy, using the case of Armenia and the Western sanctions on Russia as an example. I first develop a theoretical framework in the spirit of Ahn, JaeBin, Amit K. Khandelwal, and Shang-Jin Wei (2011 Journal of International Economics) model that features both direct and indirect trading firms and benefits/costs associated with sanctions. Using matched customs and employer–employee data covering 2018–2023, I document a sharp increase in Armenia’s intermediary role, with exports to Russia rising markedly, especially along pre-existing firm–product–partner relationships. Trade in sanctioned goods was disproportionately rerouted through Armenian firms, with higher values and unit values concentrated among incumbents. At the firm level, exporters and dual traders expanded turnover, while importers and intermediaries captured mark-ups rather than expanding production. At the employee level, wages in trading firms rose, but employment creation was limited, suggesting that gains accrued mainly to existing workers rather than through broad-based labor market expansion. In addition to the empirical findings, the paper presents a description of Armenia’s newly accessible administrative datasets as a valuable resource for future economic research.
“Jobless Industrialization and Trade Liberalization: Evidence from RTA Enactments” (R&R at Journal of Comparative Economics)
Latest Version
Abstract
This paper investigates whether trade liberalization has promoted industrialization over the past 30 years by using the enactments of Regional Trade Agreements (RTAs) as the liberalization event. Using panel data from 51 economies between 1990 and 2018 and relying on the institutional gap between the signature and the entry into force of an agreement, I estimate a two-stage least squares event-study design in which the timing of signature serves as a proxy for the time-varying conditions that lead governments to enact agreements. The findings show that RTAs have increased the share of manufacturing value added and output per worker, while decreasing the share of agriculture in production. These effects are not accompanied by a rise in manufacturing's employment share, suggesting that RTAs have promoted “jobless industrialization”. This pattern is most pronounced among RTAs enacted during 2008--2018, which generate significantly larger manufacturing-productivity gains and more negative manufacturing-employment responses than pre-2008 enactments. This productivity-employment divergence also becomes sharper as agreements include more legally enforceable provisions. Sectoral comparisons confirm that production shifts toward manufacturing relative to agriculture and other industry, without evidence of a corresponding reallocation of employment.
“Occupational Routineness and Migration Response to Mass Layoffs”
Abstract
Persistent earnings differences across local labor markets arise not only from migration costs but also from occupational switching frictions and differences in local occupational opportunity sets. Exploiting mass layoffs in France between 2003 and 2019, I show that regional reallocation is strongly shaped by occupational structure: workers move disproportionately toward destinations with similar occupational composition, and migration flows decline with occupational distance even after controlling for physical distance. To rationalize these patterns, I develop a stylized model in which displaced workers choose among staying, switching occupation, and moving, and occupational routineness shapes this choice through the return to ability, the return to occupation-specific experience, and the portability of accumulated human capital. I then show that occupational routineness does not predict selection into mass-layoff displacement but shapes adjustment among displaced workers. Workers displaced from routine-intensive occupations are less likely to move geographically and more likely to switch occupations, and occupation-specific experience pushes workers toward moving rather than switching, the less so the more routine their occupation. Non-routine workers experience larger proportional earnings losses, consistent with the destruction of higher-return occupation-specific earnings components. Geographic mobility is associated with larger earnings losses on average, and the penalty is larger for workers displaced from routine-intensive occupations, through hours rather than hourly wages. The results imply that geographic immobility is partly an occupational friction: because relocation changes which occupations a worker can access, the cost of moving is inseparable from the portability of occupation-specific human capital.
“On the Anticipation Effects of Free Trade Agreements”
with [Jose De Sousa]
Abstract
Regional Trade Agreements (RTAs) have been a key policy tool for trade liberalization since the mid-20th century. The trade literature has traditionally examined their effects using gravity models that focus on the date of enactment. This paper investigates anticipation effects by distinguishing between the signature and enactment phases of RTAs. Exploiting cross-country variation in the timing of enactment, we construct a novel bilateral dataset that separately records both dates and combines them with detailed trade and tariff information. The results show that RTA signature is associated with short-term trade destruction, which is subsequently offset by trade creation once agreements are enacted. Tariffs do not systematically decline prior to enactment, and signature alone is linked to lower applied tariffs, suggesting that pre-enactment trade reductions are not driven by tariff adjustments. We interpret the negative pre-enactment effects as anticipatory behavior, whereby firms delay transactions in expectation of lower trade costs after the agreement enters into force.