Conventional models predict that a rational farmer, given information of higher profits from a new technology relative to the old, would shift to the new technology. However, evidence abounds to the contrary. In the Philippines, the periods of sustained increase in rice yield have always coincided with government programs, mostly in the form of input subsidies, to encourage adoption of the new technology. This paper utilizes a social capital framework to inform the design of agricultural productivity programs. It emphasizes the need to examine the characteristics of farmer networks, still within the framework of optimizing the risk-adjusted return from the new technology.