Conventional measure
One switch, two very different days
Both days collapse into a single category.
The Question
Does telework remove weekday travel, or rearrange it?
Planning tools still assume a fixed workplace anchors the weekday, and the evidence on hybrid work has been mixed. Much of that disagreement is a measurement problem: a yes or no telework status files a short day and a full one under the same answer. This study measures telework in hours within the day instead.
This study
Hours on the day, not a label on the worker
Measured in hours, the two days separate.
Approach
Midweek person-days from a regional travel diary
-
Data
Travel Behavior Inventory · 2019, 2021, 2023
Three waves of the Twin Cities regional travel survey, the same data behind the region's travel demand model.
-
Unit of analysis
One employed worker, one weekday
Midweek person-days only. Mondays and Fridays carry their own telework rhythm.
-
Key measure
Telework in hours, allowed to bend
Hours worked from home that day, specified so travel can rise and then fall.
-
Outcomes
How many trips, what kind of day, and why
Total trips, the type of travel day, and a split between work and non-work travel.
What We Find
Weekday travel rises with partial telework before it falls
Finding 01
The middle of the range travels the most
Moderate telework days record more total trips than days with none. Full days at home record clearly fewer, and the turn comes early: between roughly 2.7 and 2.9 telework hours.
Finding 02
A shift between kinds of days, not a smaller average
Partial telework pushes weekdays up into the busiest travel categories; full-day telework pushes them back toward quiet and no-trip days. Hybrid work changes what kind of weekday it is, not just the average.
Finding 03
Commuting falls the whole way; errands rise, then fall
The chance of any work trip declines steadily as telework hours go up. Non-work travel does not: it peaks near four telework hours, about a quarter above baseline, then falls back. At moderate telework the extra errands outweigh the lost commute.
Finding 04
The turn survives matching
Telework concentrates among higher-income, office-based workers. Matching telework days to comparable non-telework days moves the turning point by a fraction of an hour, so the pattern is not simply about who teleworks.
Four specifications: total trips and travel-day categories, each on the full sample and the matched sample.
Partial and full telework are behaviorally distinct, not different doses of the same behavior.
Why It Matters
What changes around three hours of telework
Below three hours, telework adds travel to the day. Above it, telework removes it. Three consequences follow.
Ask how long, not whether
A yes or no telework question cannot tell a three-hour day from a seven-hour one. A daily duration item recovers the difference.
Peak reduction is not demand reduction
Partial telework lowers the commute peak while adding off-peak trips. Models that treat telework as a proportional cut to work trips will underpredict weekday demand.
Who benefits from that flexibility
Partial teleworkers gain flexibility and mobility, full teleworkers gain flexibility alone, and workers who cannot telework gain neither.
Scope. Outcomes are trip counts and types of travel day, not distance, mode, or trip chaining. Telework intensity is observed rather than assigned, and results describe midweek travel in one region.
Status
Under review at a peer-reviewed transportation journal.
Full results will be linked here after publication. Happy to talk through the approach in the meantime.