skip to content

Keynes Fund

 

Summary of Project Plan


 

The objective of this project is to study the impact of automation on labour market sorting, that is, on the allocation of high- and low-skilled workers in high- and low-productivity firms.

This research question is important in light of the evidence provided so far in the economic literature that measures the contribution of labour market sorting to earnings inequality in a number of different countries (see Card et al., 2018 for a review). However, little is known to date on the drivers behind rising sorting.

One under-explored explanation is the adoption of automated production processes. In a simple theoretical framework, the decision to automate by a firm can be seen as the choice between producing a given task with human labour or with machines. Assuming that routine tasks, carried out by low-skilled workers in the absence of automation, can be performed by machines, the automation adoption will affect labour demand by displacing low-skilled workers and by increasing the value of high-skilled workers who are complementary to automation, as they can perform non-routine and non-automatable tasks. Rising assortative matching is therefore an expected consequence of increased automation, given that firms which automate become more productive and increase demand for high-skill workers.

This project aims at testing this simple theoretical framework in the data, combining detailed matched employer-employee data for Italy (one of the European countries with the highest levels of automation in manufacturing) and automation data from the Industrial Federation of Robotics (IFR).

This project contributes to two strands of the literature. First, it contributes to the literature that investigates the drivers of rising assortativeness in the labour market, which so far has focused on trade and outsorcing as possible explanations (Goldschmidt and Schmieder, 2017; Smith, 2018). Understanding the drivers of labour market sorting would also benefit the understanding of earnings inequality, given the evidence proving that sorting is one key driver behind its rise (Card et al., 2013; Song et al., 2018). Second, it contributes to the literature that investigates the mechanisms through which automation affects labour demand (Ace- moglu and Restrepo, 2020; Dauth et al., 2018). One specific and under-explored channel is precisely the reallocation of workers of different skills across firms of different productivity. At the same time, if automation increases sorting we would expect efficiency to be positively affected, given that better workers are reallocated towards better firms.

Funding for this project is needed for two main reasons. First, I plan to use a matched employer-employee dataset, that can be accessed only through the VisitInps Scholars program, i.e. a research program at the Italian Social Security Administration (Inps), that allows researchers to use the Inps archives, but only from specified locations in Italy (mainly, Rome and Milan). Hence, I need funding to cover travel costs for research visits. Second, automation data are available through IFR at a cost of 2000 Euro. The positive externality of buying this dataset is that, once a researcher of an institution is granted access, all members of the same institution are given access, too. Hence, buying this data would benefit the whole Cambridge economics community.

 

 

Salvatore Lattanzio

 

Salvatore Lattanzio is a PhD student at the Faculty of Economics, University of Cambridge. His research interests are Labour economics, Inequality, Gender wage gap.

 

Share this Project