Using MPRS to Estimate Outcomes and Enable Adaptive RCTs

Travis Lybbert of the University of California, Davis, in the U.S. will develop a simple, low-cost approach that uses mobile phone records to track changes in living standards and use it to evaluate new mobile money saving products and services in Haiti. They will develop prediction algorithms for mobile phone records using in-person and phone surveys collected as part of a planned randomized control trial (RCT), which will be used to evaluate a new type of savings account they are developing with a leading mobile services provider. The adaptive randomized control trial will involve 13,500 working poor in Haiti and will use their mobile phone records to measure the value of the savings account.