This project is a pilot study of how social networks among farmers influence the adoption of water-saving agricultural technologies in India. Laser levelling is a technology that can grade an agricultural field to a flat surface by using a laser-guided scraper. The benefits of laser levelling include improved crop yields, reduced labor time spent weeding, and in particular, a reduction of up to 20-25% in irrigation water usage. This last benefit represents a positive externality. Because groundwater use is unpriced, farmers have minimal incentives to conserve on it. However, the collective benefit from conserving groundwater in India is substantial. Groundwater sustains around 60 percent of agriculture in India, while 80 percent of the 743 million people living in rural areas use groundwater for meeting their domestic water needs. However, there is evidence that groundwater stocks are rapidly depleting, and some estimates suggest that if this trend continues, food production could fall by around 25 percent by 2025. In Australia and in other countries in South Asia, the private returns to laser levelling have induced relatively high adoption by farmers (up to 50-80% for rice fields in Australia), generating commensurate water savings. In India, estimates suggest that it takes 1 to 2 years for private benefits to exceed upfront costs. Furthermore, various government initiatives make laser levelling equipment widely available to farmers at a subsidized rental rate. However, adoption of this technology in Punjab, where our experiment is set, remains low: it is estimated that only one-seventh of all cultivable land has been laser leveled. Two possible reasons for this low adoption rate are that small farmers are cash constrained before sowing (the time when levelling would be useful) or that farmers are poorly informed about the technology. In the first phase of our pilot study, we will conduct comprehensive surveys in a small number of villages that have recently received access to laser levelling through a cooperative. With these surveys, we will build a dataset of basic demographic information, current agricultural practices, and a network mapping indicating whose information or behaviour an individual cited as an influence on his own decisions. This dataset will provide preliminary evidence about how farmers might respond to price incentives for leveller adoption and about the channels – including social networks, government outreach, and mass media – through which farmers learn about new agricultural technologies. A second phase of the pilot study involves policy interventions to encourage laser leveller adoption. In some of the pilot villages, we plan to introduce either an extreme subsidy (free leveller rental) or an informational campaign to raise awareness about the technology. Two separate informational treatments will be used: an untargeted public campaign, and a targeted campaign. The latter will leverage knowledge of a village’s social network to target information to focal individuals who are expected to have disproportionate influence on their fellow farmers’ choices. Insights from this second phase will provide guidance for a full scale field experiment in a larger number of villages in Punjab.