Agent-Based Modeling of Office Attendance Dynamics in the Post-COVID Era

Abstract

This study develops an agent-based model (ABM) to examine the mechanisms driving office attendance behaviorin post-COVID hybrid work environments, with a focus on testing the hypothesis that interaction effects―wherethe benefits of face-to-face collaboration depend on colleagues’ attendance―constitute the primary driver of attendancebehavior. As hybrid work becomes mainstream amid increasing workforce mobility and growing individualdiscretion in attendance decisions, organizations seek to optimize performance while individuals pursue well-beingand productivity. Drawing on organizational behavior theory, this model mathematically represents office attendanceas an outcome of individual decision-making based on utility functions. Employees are modeled as autonomousagents with diverse preferences and constraints, categorized into four types. The framework integrates seven keyvariables across home and office utility functions, including family environment, office grade, and team interactioneffects. Simulation results validate the hypothesis that interaction effects, along with commuting costs, primarily driveattendance patterns. Using the validated model, we explore how attendance policies that balance organizational performancewith individual well-being vary according to task composition and industry characteristics. Results suggestthat industries with higher team task ratios benefit from more frequent in-office attendance. Furthermore, groupbasedrotation attendance schemes maintain organizational performance comparable to all-employee simultaneousattendance while improving individual well-being and enabling efficient office space utilization. The frameworkprovides a quantitative tool for designing flexible attendance policies, offering practical insights for post-pandemicworkplace management across different industries and task structures.