The impact of working from home on modal commuting choice response during COVID-19: Implications for two metropolitan areas in Australia

Abstract

The need to recognise and account for the influence of working from home on commuting activity has never been so real as a result of the COVID-19 pandemic. Not only does this change the performance of the transport network, it also means that the way in which transport modellers and planners use models estimated on a typical weekday of travel and expand it up to the week and the year must be questioned and appropriately revised to adjust for the quantum of working from home. Although teleworking is not a new phenomenon, what is new is the ferocity by which it has been imposed on individuals throughout the world, and the expectation that working from home is no longer a temporary phenomenon but one that is likely to continue to some non-marginal extent given its acceptance and revealed preferences from both many employees and employ where working from home makes good sense. This paper formalises the relationship between working from home and commuting by day of the week and time of day for two large metropolitan areas in Australia, Brisbane and Sydney, using a mixed logit choice model, identifying the influences on such choices together with a mapping model between the probability of working from home and socioeconomic and other contextual influences that are commonly used in strategic transport models to predict demand for various modes by location. The findings, based on Wave 3 (approximately 6 months from the initial outbreak of the pandemic) of an ongoing data collection exercise, provide the first formal evidence for Australia in enabling transport planners to adjust their predicted modal shares and overall modal travel activity for the presence of working from home.

Keywords: Australian experience; COVID-19; Commuter mode choice; Elasticities; Mixed logit model; Segment mapping for WFH propensity; Value of time; Working from home.

Fig. 1

Days Worked in the Last Week and Before COVID-19.

Fig. 2

Days Worked from Home in the Last Week and Before COVID-19.

Fig. 3

Current Workplace Policy toward Working from Home.

Fig. 4

Comparison of Workplace Policy toward Working from Home before and during COVID-19.

Fig. 5

Working from Home by Day of Week.

Fig. 6

Days Like to Work from Home in the Future.

Fig. 7

Average Days Worked from Home by Occupation.

Fig. 8

Average Days Worked from Home by Industry.

Fig. 9

Modes Chosen for the Commuting Trip and Before and During COVID-19.

Fig. 10

Concern about Using Public Transport.

Fig. 11

Model structure.

Fig. 12

The relationship between # days WFH and weekly commuting and non-commuting trip activity and distance.

Fig. 13

Probability distributions for WFH (WFHPrev) and commute (PCommute) SEQ model.

Fig. 14

Probability distributions for WFH (WFHPrev) and commute (PCommute) GSMA model.

Fig. 15

WFH probability changes by location/socioeconomic changes in SEQ model.

Fig. 16

WFH probability changes by location/socioeconomic changes in GSMA model.