Experimental Design
We will randomly divide our target population, consisting of car drivers, into one treatment group and two control groups. The treatment group will receive two survey invitations and information about the cost of passing the toll road by email every week for 8 weeks. Control group A receives invitations to the baseline and endline survey. Control group B does not receive anything from us. The treatment group consists of approximately 6000 toll tag users, control group A consists of approximately 6000 toll tag users, while control group B consists of the rest of our study population (approximately 108,000 toll tag users).