Monetary Incentives for Cyclists: evidence from a RCT in São Paulo

This randomized controlled trial evaluates the impact of monetary incentives on bicycle commuting in São Paulo, Brazil. Over a four-week period, participants are assigned weekly to receive different payments per kilometer cycled, with trip tracking conducted via a custom-developed mobile application.
Before the intervention began (August 4th), participants completed a baseline survey capturing sociodemographic information and existing cycling habits. To be eligible for the trial, individuals were required to (i) complete the survey and (ii) register in the app. As part of the app setup, participants pre-selected up to five addresses corresponding to locations related to work or education. They could also register up to four public transport hubs to allow for multimodal commuting, integrating cycling with the use of public transportation. Only participants who completed both the baseline survey and app registration were included in the randomization process.
Based on baseline behavior, participants were categorized into two groups: (1) super cyclists, who already use bicycles for all of their daily trips; and (2) casual cyclists, who cycle occasionally and also use other modes of transport in their daily commutes. All eligible casual cyclists (n = 460) were included in the intervention and randomized weekly to receive either R$ 0.30 or R$ 0.60 per kilometer traveled by bicycle, capped at 8 kilometers per trip.
Super cyclists (n = 757) were more numerous than casual cyclists. Each week, 351 super cyclists were randomly assigned to receive R$ 0.30/km, while the remaining 406 receive no payment (R$ 0.00/km). The number of paid super cyclists was determined based on the remaining weekly budget after accounting for casual cyclist payments. Both groups were subject to weekly re-randomization to allow for intra-individual variation in treatment assignment.
All payments were delivered via credits on São Paulo’s public transportation smartcard system (Bilhete Único), with payments made weekly. The trial is designed to provide causal evidence on the responsiveness of cycling behavior to monetary incentives, including the relative effectiveness of different payment amounts and the potential for incentivized mode shift. The primary hypothesis is that paying individuals for each kilometer cycled increases the number of functional trips (for work or study) made by bicycle.