Published
Sold to China: Container Traffic in the Port of Piraeus, 2023, Review of International Economics - special issue on China. with Pamina Koenig, Sandra Poncet, Claude Duvallet, Yoann Pigné
This paper analyzes the effects of the acquisition of the Port of Piraeus by the Chinese shipping operator COSCO in July 2016 on the organization of container traffic in Europe. Using real-time container ship positions provided by vessel tracking systems between 2015 and 2019, we study the impact of the privatization of the Greek port on the traffic of Piraeus and competing ports by vessels of different operators, and specifically COSCO. We use a difference-in-differences approach. The number of calls by container ships to Piraeus increased following its privatization, but this increase in attractiveness corresponds essentially to ships operated by COSCO whose capacity exceeds 3,000 twenty-foot equivalent units, and in particular the largest of them. We do not identify any crowding out effect between operators in Piraeus: the use of Piraeus by the vessels of other operators remains relatively unchanged. The privatization of Piraeus seems to have imposed the Greek port as COSCO’s transhipment hub for the European market without this being to the detriment of ports in any other particular European area.
City Size and Risk of Being Unemployed (Regional Science and Urban Economics, 2019) (with Carl Gaigné)
We study the relationship between city size and risk of being unemployed. We introduce a new mechanism, job pooling, as a source of agglomeration economies, in a model of risk sharing with an imperfect labor market and risk-neutral agents. Despite competition across workers to get a job in biggest cities in population (job competition), they tend to be located in large cities because tight labor market yields income gains from sharing firms among workers that do not know ex-ante the negotiated wage rates. The covariance between the probability of being unemployed and firm-specific shocks is decreasing with respect to the number of workers. This explains why higher city size exhibits higher unemployment rate on average and weaker fluctuations. From French data, we show that the elasticity of unemployment with respect to employment density is positive and around 0.2. In contrast, the elasticity of unemployment fluctuation with respect to employment density is negative and around -0.02. Our estimations control for the endogeneity of city size and for the ability of workers to move across industries in a particular local labor market.
Forced Displacement and Technology Adoption: An empirical analysis based on Agricultural households in Bosnia and Herzegovinia (Journal of Development Studies, 2018) (with Lionel Védrine)
We use the Bosnian Living Standards Measurement Study (LSMS) survey to show that conflict-induced displacement of agricultural households affects dramatically the adoption of new technologies in agriculture. We exploit the heterogeneity in the level of violence in the pre-war location to account for the selection bias. This instrumental variable seems to be a source of exogenous variation in our case because violence aims to ethnic cleansing, without economic consideration. We find that displaced are less likely than stayers to adopt fertilizer. However, we find out that under-adoption may be mitigated by the access to the property of the displaced.
Trade, wage adjustment, and job reallocation (with C. Gaigné)
We use matched employer-employee data to study the impact of import shockson firms’ performance, the wage trajectories of their employees, and labor realloca-tion across firms. To guide our estimations, we first develop a trade model in whichheterogeneous firms, producing under oligopolistic competition, bargain over wageswith heterogeneous workers. Within a given skill group, the impact of imports shockson wages varies across workers, according to export intensity of their initial employerand their status (stayer vs. movers). In accordance with the model’s predictions, wedocument that import shocks do not lead to a reallocation of workers toward moreproductive firms and do not generate higher sales for most efficient firms. Workersemployed by non-exporting firms experience large income losses in response to im-port competition. Latest Version
China, the Port of Piraeus and Trade ( with Pamina Koenig and Sandra Poncet)
Within the 80% of global merchandise trade in volume transported by sea, 35 % is carried in containers. Out of the seven to ten major companies transporting these containers across the world, one of them, Cosco, is State-owned by China. It started investing in European maritime infrastructures, purchasing a controlling share in the largest port of Greece, Piraeus, in 2016. In this paper we analyze the trade and welfare consequences of the 2016 investment by Cosco. We use detailed shipping data on the time and date container vessels enter and leave terminals and ports for the whole world, during 2015-2019, which we complement with timevarying information on the operating company of the ship. We illustrate the heterogeneity of ports in terms of port-calls by operator, and show that the surge in attendance in Piraeus after the purchase is almost entirely due to Cosco adding the stop on its routes. We estimate the impact of the purchase on bilateral trade flows, and highlight that Cosco-operated trade flows are more affected than others, suggesting the decrease in trade cost disproportionately impacted those with respect to vessels operated by other carriers.
We compute the general equilibrium effect of the purchase on world real income, based on the trade model with indirect routes by Allen and Arkolakis, 2022 adapted to shipping by Heiland et al., 2023. Finally we quantify the effect of geopolitical scenarios involving the presence of Cosco in European port.