The present research program investigates the underlying mechanisms that drive changes in the structure of communication networks among strategic decision-makers confronted with changes in their resource environment. This investigation is carried out through five experiments that engage human participants in a novel n-armed bandit task either as individuals (Studies 1, 2A, and 2B) or as members of interactive groups (Studies 3A and 3B). This research program makes several novel contributions. Primary among these is an important refinement to the “scarcity” construct employed by a number of disciplines across the social sciences. Despite a lack of consensus around the definition of scarcity, there is apparent agreement that a downward shift in resource levels is a key component. The present experiments cleanly separate the effect of downward shifts in resource levels from other components of the scarcity construct. I find that a downward shift in resource levels alone is not sufficient to produce the "scarcity effects" identified by prior research that did not control for comorbid components of scarcity that often present as confounds. However, I find that competition (Study 3B) does produce behaviors consistent with scarcity effects (e.g. reduced information-seeking and information hoarding). I situate these results in the context of the scarcity and social learning strategies literatures and discuss the implications for constructing a coherent definition of scarcity.