Dissertation Title:  "Evaluating Strategies for Achieving Global Collective Action on Transnational Health Threats and Social Inequalities"

This dissertation presents three studies that evaluate different strategies for addressing transnational health threats and social inequalities that depend upon or would benefit from global collective action. Each draws upon different academic disciplines, methods and epistemological traditions.

Chapter 1 assesses the role of international law in addressing global health challenges, specifically examining when, how and why global health treaties may be helpful. Evidence from 90 quantitative impact evaluations of past treaties was synthesized to uncover what impact can be expected from global health treaties, and based on these results, an analytic framework was developed to help determine when proposals for new global health treaties have reasonable prospects for yielding net positive effects. Findings from the evidence synthesis suggest that treaties consistently succeed in shaping economic matters and consistently fail in achieving social progress.

Chapter 2 builds on this work by evaluating a broad range of opportunities for working towards global collective action on antimicrobial resistance. Access to antimicrobials and the sustainability of their effectiveness are undermined by deep-seated failures in both global governance and global markets. These failures can be conceptualized as political economy challenges unique to each antimicrobial policy goal, including global commons dilemmas, negative externalities, unrealized positive externalities, coordination issues and free-rider problems. Many actors, instruments and initiatives that form part of the global antimicrobial regime are addressing these challenges, yet they are insufficiently coordinated, compliant, led or financed. Taking an evidence-based approach to global strategy reveals at least ten options for promoting collective action on antimicrobial access, conservation and innovation, including those that involve building institutions, crafting incentives and mobilizing interests.

Chapter 3 takes this dissertation beyond traditional Westphalian notions of collective action by exploring whether new disruptive technologies like cheap supercomputers, open-access statistical software, and canned packages for machine learning can theoretically provide the same global regulatory effects on health matters as state-negotiated international agreements. As a first move, this third chapter presents a relatively simple maximum entropy machine-learning model that automatically quantifies the relevance, scientific quality and sensationalism of news media records, and validates the model on a corpus of 163,433 news records mentioning the recent SARS and H1N1 pandemics. This involved optimizing retrieval of relevant news records, using specially tailored tools for scoring these qualities on a randomly sampled training set of 500 news records, processing the training set into a document-term matrix, utilizing a maximum entropy model for inductive machine learning to identify relationships that distinguish differently scored news records, computationally applying these relationships to classify other news records, and validating the model using a test set that compares computer and human judgments. The chapter concludes by arguing that these findings demonstrate how automated methods can evaluate news records faster, cheaper and possibly better than humans – suggesting that techno-regulating health news coverage is feasible – and that the specific procedure implemented in this study can at the very least identify subsets of news records that are far more likely to have particular scientific and discursive qualities.