The Compact entails: (1) rehabilitating irrigation infrastructure; (2) improving rural roads; and (3) training 51,700 farmers to more efficiently manage their irrigation water and to cultivate high-value crops. Mathematica’s primary role has been designing and implementing the evaluation of these initiatives to determine how effective they were in improving the living conditions of rural households.

The evaluation of the training component used a group random assignment design, in which village clusters were randomly assigned to the treatment or control group based on when the interventions will be delivered. The evaluation of the roads intervention will be conducted in 2013 and will employ regression analysis. The irrigation evaluation is split into two components. The first uses a comparison group design in which the villages where tertiary canals are rehabilitated are matched to otherwise-similar villages in which tertiary canals are not rehabilitated. A separate analysis will examine the impact of larger-scale irrigation projects such as main canals.

The farmer training evaluation was released in October 2012. The roads evaluation is in progress. The irrigation evaluation will be conducted in late 2014. Mathematica also will conduct limited analyses of some smaller components of the Compact, mainly using qualitative methods.

Mathematica helped design the sampling strategy for the national household survey, which is the key source of data for the roads evaluations. Mathematica also designed a new survey and sampling strategy for 5,000 farming households that was used to evaluate the training initiative and an analogous survey of 3,000 farming households that will be used to evaluate the irrigation activities. To ensure that high quality data is collected, Mathematica provided input on the terms of reference for collecting baseline data and guidance to the local data collection firm. Qualitative data and information from community leaders are also being collected.