The four-week card that kept working
March 16, 2026 · 2 min read
In Perth, Australia, researchers ran a randomized trial with retail employees: one group received a transit card loaded with four weeks of fare, along with reminder messages, while a comparison group waited. During the trial, cardholders were more than twice as likely to use public transport. The more interesting result came afterward: elevated use persisted once the incentive period ended.
2x
Cardholders were more than twice as likely to use public transport, and elevated use outlasted the incentive.
Source: Behavioural Insights Team RCT, 2022
Set this beside the airport study we have written about before, where a free week of bus travel changed nothing that lasted. The two cannot be compared directly, but the design differences are visible. Perth put a working card in each person's hand rather than an offer to claim. It ran for a month rather than a week, long enough to meet the same journey twenty times and let a routine form. And the reminders arrived during the change, when they had something to attach to.
Habits decay, and that is plannable
New travel routines need maintenance, and the evidence says so with unusual clarity. Across seven Dutch peak-avoidance projects, 80 to 100 percent of participants still showed the new behavior months after incentives stopped, about half after one year, and about a third after two. A program that looks finished at launch is really at its midpoint.
What to apply
When you introduce a mobility benefit, give it a duration measured in weeks of repeated journeys, put the working version in people's hands rather than a link to request it, and plan the reinforcement before launch: the reminder rhythm, the check at six months, the refresh at one year. Measurement built into the rollout is what turns a good first month into a durable second year.
The limit
The Perth trial was modest in size, 134 participants against 101 controls, and persistence was measured over the study horizon. The Dutch decay figures come from program evaluations, mostly of departure-time changes. Both point the same direction; neither is the last word.
Sources: Behavioural Insights Team randomized controlled trial, 2022; MuConsult evaluation for Beter Benutten, 2017.