The impact evaluation is based on household surveys of program participants and non-participants in three geographic locations: Koussanar, Tambacounda, and Kolda. For data analysis, we employ the "double difference" or "difference in difference" impact assessment method, which involves comparing the performance of both participants and non-participants across time horizons on a range of indicators and assessing the difference between changes in both groups. In Koussanar, where the program was first implemented in Senegal, we compare the performance of program participants and non-participants at three periods: March 2013 (T1), March 2015 (T2) and March 2016 (T3). In Tambacounda and Kolda, where the program began in 2014, we compare performance of participants and non-participants from March 2015 (T2) to March 2016 (T3). The household surveys undertaken in March 2016 reached a total of 1,618 households, all of which had been part of the baseline surveys undertaken in 2013 in Koussanar and 2015 in Tambacounda and Kolda. The geographic breakdown of households surveyed in 2016 is as follows: Tambacounda – 785 households, of which 616 are participants and 167 are non-participants; Koussanar – 382 households, of which 205 are participants and 177 are non-participants; Kolda – 451 households, of which 316 are participants and 135 are non-participants