Technologists, interested in demonstrating the value of data analytics products to the field of public policy, argue that analysis of social media sites can provide insights about policy perspectives among the wider public. At its core are assumptions that diverse groups can provide wise answers to social queries, and that governments can utilize crowd-sourcing as a deliberative tool to improve policy decision-making. The use of data analytics for social networking sites like Twitter could inform policymaking if it could identify areas of potential agreement among a diverse group of individuals. This paper uses the De Groot learning model to test the capacity of online crowds to come to consensus about social problems. Using this model, and the Idle No More social movement as a case, this paper examines the degree that Twitter is a "wise" crowd in terms of its structure, meaning it is capable of sustaining an inclusive exchange of information over time. The paper finds that structurally, the Twitter network is the sort that can converge on a topic, however the larger the discussion, the more the network structure will be centralized, meaning small groups can dominate the messaging. However, influential actors change over time, suggesting that network structure may not matter as much as information flows across influential events.