Jennifer Hadden raises important questions in three main areas: case selection, conceptual clarity, and future research. First, as she indicates, my cases were selected for variation on independent variables, including regime type. Although I did not spell this out in the book, I also selected for cases in which the transnational networks engaged in advocacy were science-based, rather than only driven by norms. In response, Hadden suggests adopting a most similar or most different selection method, either of which requires selecting in part on the dependent variable (network influence). The most similar selection method, for example, selects cases with similar starting conditions, but that vary in outcomes. This narrows down which independent variable correlates with outcome variations.

Such an approach would pay dividends, but was beyond my scope at the time. As Hadden notes, my research indicates that network influence is not a binary variable. This complicates measurement, particularly when preliminary research gives inaccurate measures of actual network influence. For example, documents from the World Bank and preliminary conversations with environmental agencies in Egypt and Mexico significantly overstated the effectiveness of environmental governance and network influence in my cases. However, additional within-case studies building on the two Mexican cases in the book will be useful in future research. To that end, I have a chapter expected this year based on a new within-case study in Jamaica.

Hadden also asks an important conceptual question, namely whether it was justice framing or local connections that best explains support for conservation. First, justice framing is inseparable from local connections. The procedural element of justice in marginalized communities requires connections with those communities. Second, justice framing emerges as a response to environmental problems that exacerbate socioeconomic and political marginalization. In other words, communities and individuals appeal to notions of justice to redress actual or threatened violations of their rights. When the Mexican Center for Human Rights (CEMDA) or Maroon communities in Jamaica argue that environmental threats and proposed solutions threaten the welfare of marginalized communities, and demand local participation as part of restorative practices, they are using the language of justice. This means that not every management approach without local participation creates injustice, nor does every instance of local involvement indicate justice framing. Top-down projects are not unjust if there are no negative costs externalized to marginalized people, while projects with local participation are not justice movements if they are not carried out in response to rights violations.

These observations point to a fruitful research agenda. New case studies can also identify factors that determine how networks generate consensus. Preliminary research suggests that networks are more likely to do so when members agree to adopt a mutually constituted, if limited approach of looking at a problem. This suggests that consensus depends on social agreements about, among other things, appropriate areas of study, as well as empirical data testing. However, social relationships are an important mediating factor in shaping how science influences environmental policy. It is imperative that these social relationships be structured in ways that empower marginalized populations.