Habitual and Rational Drivers of Policy Network Structure in Seven Tropical Forest Countries: How Bounded is Our Rationality?

Many studies of environmental policy networks follow what Leifeld and Schneider (2012) call the transaction-cost perspective, which argues that policy networks’ structure reflects boundedly rational efforts to optimize the costs and benefits of network connections. In this model, agents are interdependent in that they respond to the network, but their choices of partners are made individually. This approach potentially underestimates how long-term historical practices shape networks. Here we export the power of a relational, constructivist understanding of agency for explaining environmental policy networks. Our approach, based on the version of social field theory most closely associated with Pierre Bourdieu, has been a source of inspiration for network theories of social capital (Lin, 1999), though Bourdieu himself was often dismissive of network analysis (Emirbayer & Johnson, 2008; de Nooy, 2003). We test the two approaches in the context on climate policy networks around Reducing Emissions from Deforestation and Forest Degradation (REDD+). We rely on frailty exponential random-graph models (FERGMs) to test our hypotheses (Box-Steffensmeier, et al., 2018). Assessing model estimations for networks of information sharing and collaboration across seven countries, we find that short-term network patterns tend to be better explained by the field-theoretic, while longer-term patterns are better explained by the transaction-cost perspective.