Longitudinal Analysis of Time-Use Behaviour in the United States (2003–2023): Insights from Mdcev Modelling

Abstract

This study investigates longitudinal changes in time-use behaviours in the United States from 2003 to 2023, using data from the American Time-use Survey (ATUS). Employing the Multiple Discrete Continuous Extreme Value (MDCEV) model, we examine how Americans allocate their time across key activity categories, including work, study, shopping, leisure, and travel, distinguishing between in-home (IH) and out-of-home (OoH) activities. The analysis captures significant shifts in activity patterns driven by societal, economic, and technological changes, with particular emphasis on the COVID-19 pandemic's disruptive effects in 2019-21 and its lingering influence in 2023. Key findings reveal the persistent prioritisation of IH leisure and home care activities, contrasting with declining trends in out-of-home work and study. The study also highlights demographic variations, with distinct time-use patterns observed among students, working individuals, and different genders. Utility and [[EQUATION]]-rankings illustrate changes in activity prioritisation and satiation effects, while the analysis of marginal utilities provides insights into the evolving trends in time allocation to various activities. This examination offers critical implications for urban planning, labour policies, and digital infrastructure development, emphasising the importance of understanding time-use dynamics to inform future societal and policy decisions.

Keywords: Time-use analysis, American Time-use Survey (ATUS), MDCEV, Longitudinal Behaviour Trends, In-Home Activities