Eleonora Brandimarti (University of Glasgow)
Virginia Minni, Raffaella Sadun and Jorge Tamayo
We study task reorganization after technological change using worker-task-level data from a large Latin American bank. Two sequential interventions, a mobile app automating routine transactions and a subsequent redefinition of performance evaluation criteria, separately shift production technology and organizational design without displacing labor. Automation raises output per worker by roughly 30 percent but generates no task upgrading, despite freeing capacity. Reallocation toward complex tasks occurs only after the firm explicitly restructures performance evaluation. Adjustment is slower among high-tenure workers and high-productivity managers accelerate upgrading primarily for low-tenure workers, amplifying within-team inequality. Automation-induced task complementarities do not occur automatically: task upgrading requires deliberate organizational redesign, and its returns depend on the distribution of adjustment costs within the firm.