Fig. 8

a, b Bipartite networks of funding entities connected to regions their ads dominate for the Partially-Regional group (a) and the Regional group (b). Region nodes are uniquely colored, whereas entity nodes are grey. Entity nodes are larger if they spend more in the network, and region nodes are larger if more was spent on them in the network. High-spending entity nodes are labeled with their “disclaimer” component. The weight and label of an edge from entity E to region R reflects \(E's\) total spend on ads dominated by R. Both networks were visualized in Gephi (Bastian et al. 2009) using the same layout algorithm (Hu 2005) and same parameters (colors, size, spline, etc). c Degree distribution showing, for a given x value, the number of entities in each network with degree at least x. d Estimated total spend on regionally-dominated ads per region, separated by partition group. The total is obtained by summing the lower bound of each relevant ad’s estimated spend bin. The Regional group’s map (right) is completely yellow, indicating the lowest level of spending in each region; in contrast, the Partially-Regional group’s map (left) shows very little yellow