Research Article: Latent influence networks in global environmental politics

Date Published: March 7, 2019

Publisher: Public Library of Science

Author(s): Benjamin W. Campbell, Frank W. Marrs, Tobias Böhmelt, Bailey K. Fosdick, Skyler J. Cranmer, Sergi Lozano.


International environmental treaties are the key means by which states overcome collective action problems and make specific commitments to address environmental issues. However, systematically assessing states’ influence in promoting global environmental protection has proven difficult. Analyzing newly compiled data with a purpose-built statistical model, we provide a novel measurement of state influence within the scope of environmental politics and find strong influences among states and treaties. Specifically, we report evidence that states are less likely to ratify when states within their region ratify, and results suggesting that countries positively influence other countries at similar levels of economic development. By examining several prominent treaties, we illustrate the complex nature of influence: a single act of ratification can dramatically reshape global environmental politics. More generally, our findings and approach provide an innovative means to understand the evolution and complexity of international environmental protection.

Partial Text

The international dynamics underlying interstate environmental politics are increasingly important to understand because effective international environmental policy, as is the case with other areas of international policy, depends upon the concerted efforts of national governments and is prone to a litany of collective action problems. The most public means by which states can pledge to take environmental action is by ratifying, joining, or acceding international environmental treaties, particularly those with detailed and enforceable provisions about actions to be taken. Joining international environmental agreements via ratification comprises benefits, particularly in the long run as more concerted efforts toward solving environmental problems become possible and environmental quality may improve due to these, but also costs as actors have to somewhat alter existing patterns of behavior in order to overcome inefficient or ineffective policies. Importantly, ratification decisions are not made in a vacuum and are often thought to be linked to the actions of other states. For instance, consider the problem of air pollution and acid deposition (acid rain) in Europe in the 1970s. While initial efforts to address this issue internationally failed, due to pressures from the US and the Eastern bloc during the Cold War, it was primarily the influence of the Nordic states and Germany that eventually led to the establishment of the Convention on Long-Range Transboundary Air Pollution in 1979 and other states joining this agreement [1, 2]. This illustration underlines that actor linkages [3, 4] and structural group leadership [5] are frequently seen as key determinants in international environmental governance. And, indeed, more recently have EU states been aspiring to leadership roles in global environmental policymaking and to influence others to join their efforts. To analyze these foundational dynamics underlying international environmental politics, we developed a novel method to estimate the international influence network and how it informs one of the most pressing issues of our time—environmental protection.

To assess the proposition that influence networks help shape the evolution of international environmental protection, we augment and analyze data on international environmental treaty ratification. We compiled data on international environmental agreements, which focus on treaties that were open for ratification globally between 1972 and 2000. We based the coding of the data on the original treaty documents and texts, which do hardly change over time. With this re-coding of the ratification data for every state-treaty pair, we not only validate the original data from [19], but add information on the exact day and month of a ratification event. Second, we updated the data on available covariates, doing much to address the problem of missing values in the original data. The outcome variable for our models is a binary indicator for whether a state ratified a given treaty in the year of observation (see Methods for details). The inclusion of a state-treaty-year triplet in the data is an indication that a state could have ratified a treaty during the given year. In other words, for a state-treaty-year to be included, the state has to have had the potential to ratify the treaty in that given year. A state-treaty pair remains in the data until 2000 if no ratification occurs, but leaves the data before if it has joined an agreement.

From patterns in environmental treaty ratification, the BLIN infers two weighted, directed networks. These networks describe either the positive or negative influence among states and treaties. Negative influence occurs when the one event makes another potential event less likely. For example, if Ireland ratifies a given treaty, France may be less likely to ratify the same treaty. Positive influence reverses this dynamic, wherein one event makes another potential event more likely. For example, if Iceland ratifies a given treaty, Norway may be more likely to ratify the same. These influence relationships are also weighted such that larger weights refer to stronger influence relationships between actors. To be clear, the BLIN model provides correlational results that are consistent with the causal process of influence. With observational network data such as ours, a direct causal test (e.g., an experiment or matching process that relies on independent observations) is not possible. For ease of interpretation, we parse the positive and negative influence networks into four different networks based upon the direction of influence and the nodes included.

We present and assess latent influence networks underlying international politics. The implications of this study matter for a variety of reasons. Most importantly, our findings constitute a systematic demonstration of an underlying latent influence network in international politics, including its geometry, structural features, and interdependencies. This complex network of influencer-influenced relations has long been considered, but now may be used to study a variety of international and domestic outcomes. The networks presented provide means of answering numerous foundational questions in the social sciences, including whether material resources imply influence, or whether power is truly the currency of international politics.




Leave a Reply

Your email address will not be published.