Post-covid Telework and Productivity: A Large Scale Analysis

This paper investigates the causal impact of post-Covid telework adoption on firm-level productivity in France, where hybrid work—typically around two days per week—has become the dominant form of telework. Using matched data from the Dares's ACEMO-Covid employer survey and Insee's FARE administrative database, we examine whether the share of teleworkers in 2022 is associated with productivity growth between 2019 and 2022 in the non-agricultural firms excluding finance and real estate. Ordinary Least Squares (OLS) estimates reveal a positive but moderate relationship: a 10-percentage-point increase in telework share is associated with a 0.7 to 1.0-percentage point increase in labour productivity growth over the period. The relationship is non-linear, with marginal gains declining once the telework share exceeds roughly 20-30 percent. To address potential endogeneity—including reverse causality and unobserved firm characteristics—we implement an instrumental variable strategy based on the pre-pandemic configuration of firms' premises. Specifically, we use the surface per employee of rented office spaces physically separated from production or retail activities in 2019. Firms operating such separate office units were organizationally better positioned to adopt telework and potentially reduce real-estate costs. The instrument strongly predicts telework adoption and passes several validity checks. Two-stage least squares estimates yield a significantly larger causal effect with a semi-elasticity of about 0.27. The Local Average Treatment Effect (LATE) implies that a 10-percentage-point increase in the telework share raises productivity growth by about 2.7 percentage-point. The IV coefficient suggests that productivity gains are concentrated among firms whose telework adoption was facilitated by their pre-existing organizational structure. Consistent with this interpretation, firms operating separate offices before the pandemic and adopting telework subsequently reduce obsolete office space and increase office equipment investment. However, these adjustments remain quantitatively limited, indicating that the estimated productivity gains likely reflect broader organizational and managerial changes associated with hybrid work arrangements.