To better grasp the interconnected range of socioeconomic impacts from the implementation of rural roads in northern Ethiopia, we have experimented with Causal Loop Diagrams (CLDs), a tool commonly used in systems dynamics, but generally under-used in development research. The expansion of the rural road network in Africa is praised for reducing spatial isolation, lowering transport cost, increasing access to markets, and bringing services closer to home. However, different segments of society will benefit differently from the establishment of a rural road. This difference may lead to dynamics that either exacerbate or reduce existing inequi- ties, which forms the central question for this paper. As part of a broader study on the multiple (in)direct effects of rural roads on pro- ductive employment, we undertook oral testimonies in four municipalities to explore how people perceive road impacts on livelihoods, mobility, and work. CLDs were then used to assemble those seemingly loose observations into a systematized view of the whole. The exercise reveals conflicting feedback processes that may dominate system at different times and drive the inequities between surplus food producers and laborer households up or down. The method used can be particularly useful for studying similar infrastructures that seem- ingly bring benefits to all, but may cause subtle, concealed or delayed effects, and ultimately surprising system behavior. Temporary free access online:,6yxCxJ6S