This paper investigates several channels through which automation affects an emerging economy. Building on a Ricardian model of trade with sectoral linkages and a two-stage production technology, in which robots replace labor in certain tasks, it is shown that domestic and foreign automation have differential effects on labor markets. Based on this model, the impact of automation on local labor markets in Brazil are estimated using a shift-share approach. Local labor market exposures to industry-level stocks of robots are derived from their initial industry-employment composition. Foreign automation is found to decrease manufacturing employment through the channel of final goods exports, while it increases employment in the mining sector through the channel of input exports. This may stimulate what has been called "premature deindustrialization" in emerging economies. To account for possible endogeneity in adopting robots domestically, robot uptake in other emerging economies is used as an instrumental variable. Domestic automation is found to directly decrease the ratio of unskilled industry workers and increase the ratio of skilled workers. Also, the wage gap between the two groups widens as a consequence of domestic automation, reinforcing income inequality.

No abstract

This paper analyzes how a specific differentiation by governments throughout the world -whether a sector was deemed "essential" or "key" -affected firm performance. During the COVID-19 pandemic, governments designated specific services as "essential," which allowed firms operating in those sectors to remain (partially) open as well as being granted other preferential treatment. This paper analyses the effects of the key-status, by mapping the countries' lists to the sectoral level, and matching these sectors with firm-level Covid-19 survey data from 27 countries. The findings reveal that, controlling for a rich set of firm-level and sectoral characteristics, firms deemed key less often reported declining sales and demand for their goods or services, and had a smaller number of furloughed workers. Nonetheless, non-key firms were more likely to employ online business activities and to change the main product or service they offered, reflecting the necessity to otherwise adjust to the economic downturn and changes in demand.

No abstract

First, I thank my principal supervisor, Holger Strulik, who always provided swift and excellent feedback, and the freedom to pursue the various topics that I found interesting. I especially appreciate discussing various academic and non-academic issues over long forest-walks in times of social distancing. I am very grateful to my second supervisor, Krisztina Kis-Katos, for the continuous support and countless talks with academic and professional advice. Her dedication to academic research but also to the well-being of the PhD students of the GlaD program have been truly inspiring. I thank Axel Dreher, the third member of my thesis committee, for his greatly valued feedback on research projects of different stages and academic insights. I would also like to express my gratitude to my co-authors