Publications

We examine the effects of unilateral structural reforms within a currency union. Focusing on the surge of German competitiveness following the introduction of the Euro, we first provide reduced-form causal evidence supporting the notion that German structural labor-market reforms in the early 2000s led to a crowding-out of manufacturing employment in other Eurozone economies. To assess the impact of this German competitiveness shock, we build a quantitative multi-sector trade model that features downward nominal wage rigidities, endogenous labor supply, unemployment-insurance benefits and international savings. The fixed nominal exchange rate can create binding nominal rigidities in response to a foreign real supply shock -- like the one prompted by the German reforms -- resulting in significant contraction of manufacturing sectors and increased involuntary unemployment across other Eurozone countries. We consider a number of counterfactual scenarios, such as the impact of German labor-market reforms in the absence of a fixed exchange-rate regime, the role of coordinated reforms within the Eurozone and a higher average inflation rate.

We use French data over the 1994-2013 period to study how imports of industrial robots affect firm-level outcomes. Guided by a simple model, we develop a novel empirical strategy to identify the causal effects of robot adoption. Our results suggest that, while demand shocks generate a positive correlation between robot imports and employment at the firm level, exogenous exposure to automation leads to job losses. We also find that robot exposure increases labor productivity and some evidence that it may raise the relative demand for high-skill professions.

We jointly study firm boundaries and the allocation of decision rights within them by confronting an incomplete-contracts model with data on vertical integration and delegation for thousands of firms around the world. Integration has an option value: it confers authority to delegate or centralize decision rights, depending on who can best solve problems that arise in the course of an uncertain production process. In line with the model's predictions, we find that firms are more likely to integrate suppliers that produce more valuable inputs and operate in industries with more dispersed productivity, and that firms delegate more decisions to integrated suppliers that produce more valuable inputs and operate in more productive industries. 

We build a dynamic heterogeneous-firm model in which real depreciations raise export demand and the cost of importing intermediates, and also affect borrowing-constraints and the profitability of engaging in innovation (R&D). We decompose the effects of real exchange rate (RER) changes on firm-level productivity growth into these channels. A number of stylised facts on manufacturing firms for a large set of countries discipline our structural model estimation: firms in emerging East Asia are very export oriented and rely little on imported intermediates compared to firms from Latin America and Eastern Europe, whereas firms from industrialized countries export as much as they import. Exporters experience an increase in cash flow, R&D, and productivity growth in response to RER depreciations, while importers experience a reduction in these outcomes. We evaluate the model's mechanisms by providing counterfactual simulations of temporary RER movements. The effects of RER swings on innovation and productivity growth are heterogeneous across regions, sizeable and very persistent. In export-oriented emerging Asia, real depreciations are as sociated with higher probabilities to engage in R&D, faster growth of average firm-level productivity and cash-flow, and higher export entry rates; we find negative average effects on these outcomes for firms in other emerging economies, which are relatively more import dependent, and no significant average effects for firms in industrialized economies. 

We study the importance of input-output (IO) linkages and sectoral productivity (TFP) in determining cross-country income differences. We find that while highly connected sectors aremore productive than the typical sector in poor countries, the opposite is true in rich ones. To assess the quantitative role of linkages and sectoral TFP differences in cross-country income differences, we decompose cross-country income variation using a multi-sector general equilibrium model. We find that (i) IO linkages substantially amplify fundamental sectoral TFP variation but (ii) this amplification is significantly weaker than the one suggested by a simple IO model with an aggregate intermediate good.  

This paper studies the impact of working from home (WFH) on work relations and public health during the COVID-19 pandemic in Germany. Combining administrative data on SARS-CoV-2 infections and short-time work registrations, firm- and worker-level surveys and cell phone tracking data on mobility patterns, we find that working from home effectively shields workers from short-time work, firms from COVID-19 distress and substantially reduces infection risks. Counties whose occupation structure allows for a larger fraction of work to be done from home experience (i) fewer short-time work registrations and (ii) less SARS-CoV-2 cases. At the firm level, an exogenous increase in the take-up of WFH reduces the probability to file for short-time work by up to 71 p.p. and the probability of being very negatively affected by the crisis by up to 77 p.p. Much of the changes in the organization of work relations are likely to be permanent and to have effects well beyond the crisis. Health benefits of WFH appeared mostly in the early stage of the pandemic and became smaller once tight confinement rules were implemented. Before confinement, mobility levels were lower in counties with more WFH jobs and counties experienced a convergence in traffic levels once confinement was in place.  

We study the impact of working from home on (i) infection risk in German regions and (ii) output using an input-output (IO) model of the German economy. We find that working from home is very effective in reducing infection risk: regions whose industry structure allows for a larger fraction of work to be done from home experienced much fewer Covid-19 cases and fatalities. Moreover, confinement is significantly more costly in terms of induced output loss in regions where the share of workers who can work from home is lower. When phasing out confinement, home office should be maintained as long as possible, to allow those workers who cannot work from home to go back to work, while keeping infection risk minimal. Finally, systemic industries (with high multipliers and/or high value added per worker) should be given priority, especially those where home office is not possible.

We characterize the agglomeration patterns of industries and plants in Europe, distinguishing Eurozone countries and the U.S. Using a micro-level index, we quantify the degree of geographic concentration in industrial activities and explore how firm heterogeneity, industry attributes, and location fundamentals jointly explain the observed patterns. Our analysis shows that there is a clear hub-and-spoke structure in the geographic distribution. Larger and more productive plants, especially the superstars of each industry, are more centred than their smaller, less productive counterparts. The greater agglomeration surrounding superstars is particularly pronounced in the Eurozone but not present in the rest of Europe. Location fundamentals also play an important role and can sometimes mitigate the importance of agglomeration economies around large firms. Regions with different levels of economic development, including in education and technology, exhibit distinct agglomeration patterns. The findings suggest heterogeneity in the ability of regional policies to build superstar-centred industry clusters.  

A number of theories in organizational economics and industrial organization suggest that vertical integration, while costly, increases productivity. It follows from firms' maximizing behaviour that higher prices in the product market ought to induce more integration. Trade policy provides a source of exogenous price variation to assess this prediction: higher tariffs should lead to higher prices and, therefore, to more integration. We construct firm-level vertical integration indices for a large set of countries and industries and exploit variation in applied most-favoured-nation tariffs to examine the impact of tariffs on firm boundaries. The empirical results provide strong support for the view that higher output prices generate more vertical integration. Our estimates of the average price elasticity of vertical integration are in the range 0.4-2.  

We develop a model of directed technology adoption, frictional unemployment and migration to examine the effects of a change in skill endowments on wages, employment rates and emigration rates of skilled and unskilled workers. We find that, depending on the elasticity of substitution between skilled and unskilled workers and the elasticity of the matching function, an increase in the skill ratio can reduce the relative unemployment rate of skilled workers and decrease the relative emigration rate of skilled workers (brain drain). We provide empirical estimates and simulations to support our findings and show that effects are empirically relevant and potentially sizable. 

We study trade policy in a two-sector Krugman (1980) trade model, allowing for wage, import and export subsidies/taxes. We study non-cooperative trade policies, first for each individual instrument and then for the situation where all instruments can be set simultaneously, and contrast those with the efficient allocation. We show that in this general context there are four motives for non-cooperative trade policies: the correction of monopolistic distortions; the terms-of-trade manipulation; the delocation motive for protection (home market effect); the fiscal-burden-shifting motive. The Nash equilibrium when all instruments are available is characterized by first-best-level wage subsidies, and inefficient import subsidies and export taxes, which aim at relocating firms to the other economy and improving terms of trade. Thus, the dominating incentives for non-cooperative trade policies are the fiscal-burden-shifting motives and terms-of-trade effects. 

We consider a model in which exporting requires finding a local partner in each market. Contracts are incomplete and exporters must learn the reliability of their partners through experience. Export behavior is state-dependent due to matching frictions. Better legal institutions alleviate contracting frictions especially in sectors with large contracting problems, thereby increasing state dependence and reducing hazard rates by more in those sectors. Moreover, hazard rates decline with the age of the relationship, as unreliable partners are weeded out. We find strong evidence in favor of the model's predictions when testing them with French firm-level data. 

Cross-country differences in sectoral total factor productivity (TFP) are at the heart of Ricardian trade theory and of many models of growth and development. Our knowledge of the magnitude and the characteristics of cross-country differences in sectoral TFP is still limited however. This study fills the gap by showing how sectoral TFP differences can be backed out from bilateral trade data using a hybrid Ricardo-Heckscher-Ohlin model. This approach allows us to overcome the data problems that constrained previous studies and to provide a comparable set of sectoral productivities for twenty-four manufacturing sectors in more than sixty countries at all stages of development. Our results imply that TFP differences in manufacturing sectors between rich and poor countries are substantial and far more pronounced in skilled labor and R&D intensive sectors than in others. We also apply our productivity estimates to test development theories that have implications for cross-country industry-level productivity patterns.  

This paper studies cross-country differences in productivity from an open economy perspective by using a Helpman-Krugman-Heckscher-Ohlin model that embraces the single-cone model and a one-sector economy with factor deepening as particular cases. To estimate the model, I combine tools from development accounting and the factor content of trade literature. When simultaneously fitting data on income, factor prices and the factor content of trade, I find that the one-sector model is by far better supported by the data than the single-cone model. Rich countries have far higher productivities of human capital than poor ones, while differences in physical capital productivity are not related to income per worker. Finally, I estimate an aggregate elasticity of substitution between human and physical capital that is significantly below one. 

This paper investigates the dynamics of export relationships -- defined as shipments by a given firm to a given destination in a given year -- using a panel of almost 25,000 French exporters over the five-year period 1995-1999. We describe how these export relationships evolve over time and present a number of stylized facts, which we relate to different theories of export dynamics, such as a dynamic sunk-cost model and the recent literature on exporting and learning.

We find that export relationships are very dynamic: a large fraction of export relationships are created or destroyed every year and export values within relationships fluctuate substantially. Most of these dynamics are explained by relationship-specific shocks rather than by supply and/or demand shocks. Moreover, upon entry, export values are small but they gradually expand as relationships mature. Finally, while many export relationships are volatile, others are persistent. Having previously exported to a given destination substantially increases the probability of exporting there in the current period. We argue that, taken together, these facts are more in line with a learning model than with the sunk-cost hypothesis.