Research

Ongoing work

• Building Up Local Productivity: Infrastructure and Firm Dynamics in Mexico,2024 (with M. Busso). What determines the aggregate and distributional effects of new transportation infrastructure? One key overlooked channel is the role that infrastructure policy plays in changing the incentives of firms to enter, exit, and grow -- in turn generating endogenous changes in local productivity. In this paper, we document and quantify the importance of this channel by using detailed Mexican microdata and a spatial general-equilibrium model that incorporates firm dynamics. Leveraging random delays in the construction of highways, we empirically show that productivity grows in places with better transportation infrastructure. Firms play a critical role in driving these results: highways increase firms' size, entry rates, survival rates, and total factor productivity. Then, by calibrating our model on census data between 1998 and 2018, we find that new highways over this period increased welfare and income by half a percent, similar to its costs in terms of GDP. Moreover, we find substantial spatial reallocation of workers and production. Nearly half of these effects are explained by endogenous changes in local productivity, which is driven by firm dynamics. 

• Making a Growth Miracle - Historical Persistence and the Dynamics of Development,2024 (with J. Gathen). What explains growth miracles? Periods of rapid economic growth are driven by aggregate changes in the economic environment and economies slowly catching up with previous changes. Building on 40 years of plant-level manufacturing panel data for Indonesia, we show that large population changes and the slow entry, exit and growth of plants drive prolonged periods of catch-up growth. Motivated by the empirical evidence, we build a model of plant dynamics, which we estimate on the micro data along the observed growth path without assuming that the economy is at a steady state. Catch-up growth starting from initial conditions in 1975 accounts for 42% of Indonesia’s subsequent industrialization. But catch-up growth does not become less important over time because further changes in worker and plant demographics induce new adjustments. Observed changes in government policy explain few changes in worker and plant demographics and thus drive at most 10% of observed economic growth.

• Dysfunctional Firm Dynamics and Mexico’s Dismal Productivity Performance, 2019 (with S.Levy).  We study a common growth paradox in developing countries, where fast industrialization might be coupled with low aggregate productivity growth. We argue that the paradox can be explained by two opposing forces. On the one hand, governments introduce policies that promote growth, such as trade liberalization, competition agencies, and regulatory bodies; and on the other hand, policies segment the economy into formal and informal sectors. We construct a 20-year establishment-level panel dataset for Mexico, a country where manufacturing exports grew from seven to 33% of GDP, but labor informality barely changed, firm informality increased, and TFP growth was negative. We find that many high-productivity formal firms exit; surviving firms' size and productivity hardly grow, and many informalize; entrants are less productive than survivors, mostly because of large informal entry. Finally, we show that while manufacturing performs better, its contribution to TFP is modest because informality persists in the sector; and despite spectacular export growth, the country is now de-industrializing.

Working papers

• The heterogeneous impacts of import competition on Mexican manufacturing plants, 2019 (with J. Blyde).  We study the impact of import competition on Mexican firms between 2003 and 2013 by exploiting variation in Chinese import penetration across industries. We find that the trade shock induced a decline in employment, sales, exports, and productivity. Importantly, the effects are heterogeneous: smaller and less efficient plants experiencing the largest adjustments, while the most efficient plants exhibited relatively minor effects and, for some outcomes, no effects at all. The productivity gap between small and large plants has been increasing over time and that the reallocation of resources has been productivity-enhancing, particularly in sectors that have experienced large-scale import penetration from China.

• The longitudinal linkage of mexico’s economic census 1999-2014, 2018 (with M.Busso and S. Levy). This technical note describes the methodology to construct a longitudinal dataset using the Economic Censuses of Mexico from 1999 to 2014. The procedure is based on an algorithm that links establishments with identical or significantly similar location, legal entity and industry. Since a set of longitudinal identifiers is already available for the 2009 and 2014 Economic Censuses, it is used to validate our results, obtaining 90% accuracy. The paper links 1.44 million establishments for the period 1999-2004, 1.52 million for 2004-2009 and 2.15 million for 2009-2014.

Policy reports

• NAFTA-USMCA and Wages in Mexico, 2022 (with S. Levy). This note focuses on some features of Mexico’s labor market that, in our view, are crucial to understand the effects of NAFTA-USMCA on wage in Mexico. In our assessment we find that (i) Despite NAFTA, average wages in Mexico did not increase from their pre-NAFTA levels, although in its absence they would have been marginally lower. (ii) So long as Mexico’s current domestic regulations remain -particularly those pertaining to labor and social insurance- it is unlikely that the U.S.-Mexico-Canada Agreement (USMCA), the trade pact that superseded NAFTA in 2019, will lead to higher average wages. (iii) If the USMCA increases labor costs significantly in the USMCA-related segment of the economy, aggregate productivity in Mexico may suffer.

Also circulates as LABOR, USMCA Forward: Building a more competitive, inclusive, and sustainable North American economy