Making an "Impact" at the 7th Annual Political Networks Conference in Montreal
An important goal at the Center for Environmental Policy and Behavior is using basic research on environmental policy to make an impact on real-world decisions. We took a bit of an unusual strategy for this at the 7th Annual Political Networks conference held last week in Montreal at McGill university. The picture features graduate student Matthew (Mateo) Robbins and me holding the poster he made for his work on spiny lobster management in Honduras. We are standing in the supporter section for the Montreal Impact, a Major League Soccer team, which that night beat the New England Revolution 3-1. Needless to say, among many French soccer chants and curse words (we don't speak French...), there were some Canadiens with some very strange looks on their faces when they saw Mateo's awesome poster. There is no doubt we made an "Impact" here--surely for the first time ever, an academic poster was unveiled in the middle of a soccer supporter section.
But Mateo's work is breaking new ground in other ways too. In the research on collaborative governance, it is rare to have three waves of network data from the same site. The Honduras Spiny Lobster Initiative collected network data at the beginning, the middle, and the end of a conservation and development project that aimed to increase the sustainability of the the local fishery. The project was implemented by the Global Fish Alliance, and they use a process called "Whole System in the Room" to bring together multiple stakeholders and build new partnerships and collaborations around fishery issues. The leaders of the Global Fish Alliance have been interested in network analysis for a long time, and I have worked with them for several years to think about how network analysis could be used to evaluate the effectiveness of their projects.
While Mateo is still working on a more detailed analysis for this project, there are some intriguing ideas at play. One of the standing assumptions of much of the literature, and also a stereotype of networks held by many practitioners, is that effective networks of this type would feature increased structural indicators of cooperation. More specifically, this would predict increased density, transitivity, and reciprocity over time. However, in the Spiny Lobster case, these network structures stayed about the same over time or even declined. Instead, it appears that the existing network ties were re-allocated to new partners, and began to span organizational boundaries in new ways. For example, some of the marginal subsistence and indigenous fishing communities seemed to have more partnerships with some of the other stakeholders. This was one goal of the Global Fish Alliance, because the lobster divers in these communities face serious health threats from poor diving practices. The re-allocation of network ties to span boundaries is an alternative indicator of success of this project, and an interesting contrast to the typical hypothesis. We will see if this finding holds under the scrutiny of more sophisticated statistical analysis of the networks.
Another interesting aspect of the data is what we are calling "network churn". Network churn occurs when different stakeholders are entering and leaving the collaborative partnership over time. Some of the early stakeholders leave because they do not find it beneficial, while new stakeholders enter the system after hearing it might be useful. Network churn presents a practical challenge to implementing conservation and development projects because it is difficult to maintain or predict communications. Network churn is also an analytical challenge, because it is more difficult to directly compare how the structural properties of the network are changing over time. The next step in this analysis is to estimate two types of statistical network models, temporal exponential random graphs (TERGM) and stochastic actor-oriented models (SAOM, sometimes called by the software name of rSIENA). Both of these models analyze the evolution of network structure over time.
Anyway, good times in Montreal and thanks to all PolNet 2014 participants who appreciated seeing lobsters and invasive sea grass along side networks of legislators and international conflict.