Nathan TeGrotenhius, University of Washington
Benjamin Mako Hill, Aaron Shaw
Although online petitions are an increasingly important tool that social movement organizations use to mobilize supporters to collectively pressure elites such as officials, policymakers, and corporations (Earl and Kimport, 2011; Margetts et al., 2015), the vast majority of petitions fail to garner many signatures. Existing efforts to explain the uneven distribution of signatures across petitions have looked to turbulence, social media use, recommendation engines, (Margetts et al., 2015) and to the effect of early signatures on the ultimate number of signatories (Yasseri et al., 2013; Hale et al., 2013). We borrow from organizational ecology to advance a new theory to explain petition success: that ecological forces of legitimacy and competition within topics, analogous to niches, influence a petition’s chances of obtaining signatures. These theories also suggests that focused petitions will be more successful and that concentrated topics will have more focused petitions.
The theory we advance has a basis in organizational ecology (Hannan and Freeman, 1977), resource mobilization theory (McCarthy and Zald, 1977), and resource partitioning theory (Carroll, 1985), all drawn from scholarship on organizations and, more recently, social movement organizations (e.g., Soule and King, 2008). In particular, we offer three hypotheses drawn from this previous body of theory: (1) that greater concentration of petitions within a topic will lead to more focused petitions; (2) that more focused (i.e., ‘specialized’in traditional ecological theory) petitions will attract more signatures due to the competitive nature of online petitions; and (3) that petitions will confront a trade-off between both legitimacy and competition as topic spaces become more dense or crowded. To test these hypotheses, we employ a novel computational approach by using latent Dirichlet allocation (LDA) topic models (Blei et al., 2003) to analyze the text of petitions and to group petitions into topics. This way of grouping petitions into populations affords measuring the degree to which a petition focuses on a single topic or includes multiple topics. By algorithmically assigning petitions to topics, we are able to analyze a all Change.org petitions posted within a period of one year.
We find very strong support for our first hypothesis that more focused petitions are more successful. That said, our results contradict our second hypothesis that concentrated topics have more focused petitions. Instead, we find a modest negative association between concentration and focus. Finally, we find support of our third hypothesis that the effect of density on signature count has an inverted U shape and petitions in both very relatively crowded, and relatively empty, topics tend to collect fewer signatures. This curvilinear relationship between topic density and petition success suggests support for the idea that environmental pressures on petitions include both legitimacy and competition. Our work contributes to the study of social movements by advancing a new theory of online mobilization success: that ecological forces within populations of attempts at collective action influence the chance and the extent of success at mobilizing. In particular, our results strongly support the notion that petitions focused fewer topics are more successful at attracting signatures. We argue that competition between petitions in dense niches constitutes a social failure because competing petitions may undermine their collective goals ‘a fact that is in tension with online petitioning environments that encourage the production of many focused petitions.
Our results on petition focus have clear applications for petition creators. We suggest that petition creators benefit from making their petitions focused and should attempt to offset their contribution to competition by supporting and promoting similar petitions in topic spaces that are already densely crowded.