Digitizing Work: Driving and Measuring Changes in Information Worker Time Use and Performance Via a Longitudinal Quasi-Experiment

We study the causal effects of digitizing work on information workers’ time-use and performance at a large insurance firm. We make causal inferences and obtain unbiased estimates by exploiting a quasi-experiment: the phased introduction of Electronic Document Management (EDM) across multiple offices at different dates. We apply a difference-in-differences methodology to econometrically measure changes in a suite of performance metrics. We further triangulate on the effects of digitizing work via three complementary research techniques: extensive onsite interviews before, during and after implementation; detailed time use diaries and observation; and a series of surveys. In addition to large changes in time-use and performance, we find that digitization leads to a decline in the substitutable routine labor input and an increase in complementary nonroutine cognitive labor input. We uncover a new micro-level mechanism, “IT-enabled slack”, that explains how exactly IT can lead to payoff in terms of information worker performance.