Much attention has been paid to the role of increased food production in the development of social complexity. However, increased food production is only one kind of agricultural process, and some changes in agronomic practices were geared toward stabilizing production or counteracting periodic shortfalls. The intersection between these latter strategies and sociopolitical development are poorly understood, while the long-term value of risk management strategies is often hypothesized but empirically not well demonstrated. We address these issues using recent archaeological data from the Samoan Archipelago, Polynesia. We investigate variability in, and the development of, one type of agricultural infrastructure: ditch- and-parcel complexes. In the context of Samoa’s high-volume rainfall, recurrent cyclones, and steep topography, these novel risk management facilities offered production stability and, by extension, long-term selective benefits to both emergent elites and the general populace. Their effectiveness against known hazards is demonstrated by hydrologicai modeling, while their long-term success is indicated by increased distribution and size over time. Additionally, based on their morphologies, funetional properties, chronology, and spatial patterning, we argue that this infrastructure could have been effectively used by emergent elites to gain political advantage, particularly in conjunction with environmental perturbations that created production bottlenecks or shortfalls.