If I have seen further, it is by standing on the shoulders of giants.


Quantitative Explorations of Mass Movements in the Middle East

Sitting at the intersection of my methodological and substantive interests, my dissertation consists of three separate but related papers that each explore an aspect of mass mobilization with a different quantitative and data driven approach: Latent Class Analysis, causal mediation analysis, and machine learning-based natural language processing and keyword recommendation.

Together they present a holistic perspective on the nature of mass mobilization in the Middle East. Individually, each paper delves more deeply into character of mass movements—durable mobilizations of large numbers of people for a common political purpose—successively narrowing in on the subject. When large numbers of individuals join in a collective struggle against a perceived societal wrong, what form does their struggle take? How does the occurrence of these mass struggles impact the perceptions of individual well-being and grievance for the populations that experience them? What language do people use to articulate their experiences with mass movements—either as participants or observers—and how can we identify and what can we learn from their linguistic choices?


Mass Movements Project
The Forms and Implications of Mass Resistance and Empowerment
with Stephen Kosack

How do mass movements aimed at political and economic change originate and what are their effects? Systemic change in political institutions and social and economic performance often occurs when large numbers of citizens mobilize either in opposition to a government that is not serving their interests or in support of one that is. Recent mobilizations such as the Arab Spring have renewed interest in the power of such mobilizations. But mass movements are rare. Masses of citizens rarely find it rational to sacrifice for social change, and even when they do, they rarely succeed. This rarity constrains the study of mass movements. Mass mobilization is the subject of a large, interdisciplinary academic literature, including some of the most influential scholarship on political and economic change. But in contrast to studies of the causes and consequences of political institutions, which have been able to draw on rigorous, cross-national data on political institutions, systematic studies of the effects of mass mobilization have had to rely on cases and single-country studies.

The Mass Movements Project is producing the first comprehensive cross-national dataset of mass movements, covering the nature and characteristics of all movements of more than 1,000 participants over the last century in all countries worldwide. The data we gather about each movement include 216 variables, covering its purpose, scale and scope, organization, tactics, leadership, and affiliation with the government. It will permit the first rigorous examination of theories about the origin and implications of mass movements that have until now been examined only in specific cases, including:

  1. movement characteristics and how those characteristics shape movements’ capacity for action and likelihood of surviving repression;
  2. the role of political, material, and social deprivation, the availability of resources either from within the movement or from outside, and political institutions that require compromise or allow airing of many viewpoints for movement emergence and trajectories; and
  3. the implications of mass movements for economic and political change, such as democratization, economic performance, and factors of human development such as health care and education.

Currently, the project is focusing on the developing world and has completed coding mass movements in 19 countries in the Middle East and North Africa—as well as more than 10 selected countries in Sub-Saharan Africa, Central and Southeast Asia, and Latin America—from 1800-2012.


Government as a Complex System
A Network-based Approach
with Albert-László Barabási, Michele Coscia, Ricardo Hausmann, and Stephen Kosack

A government is a complex system, within which institutions and agencies play a pivotal role in shaping the economy, society, and individual behavior by making and enforcing rules and regulations and providing myriad services. Yet, fundamental questions remain about basic aspects of how governments work. How are governments organized to execute their many functions? How much does their organization vary, and what explains that variation? Does governmental organization relate to government effectiveness?

These sorts of questions have inspired voluminous literatures in political science, economics, and sociology. But exploring them has long been difficult for lack of comprehensive data about how governments are organized. The growing presence of governments on the Internet presents a unique opportunity to gather these data. This project develops novel methods for community discovery, graph mining, and analytical web crawling to transform the information in U.S. government websites and their linkages into network data on government activity and connectivity. The internet footprint of this governance structure—more than 30 million public agency webpages and their connections–provide a comprehensive picture of the organization and activities of 50 semi-sovereign state governments operating under a single governing framework atop one of the world’s most complex societies.

The picture that emerges is both familiar and illuminating. It supports a number of core expectations from across the social sciences about basic aspects of government structure, but also allows a far deeper and more nuanced view than has previously been possible of the range of functions that U.S. governments perform and how public agencies interact with each other in executing these functions. The data also permit investigation of whether differences in governmental organization across the 50 U.S. states are associated with differences in the effectiveness with which they deliver crucial government services such as health and education. In the process, the data provide an opportunity to examine predictions of longstanding theoretical traditions in the social sciences against the first holistic representation of U.S. government structure in all its variation and complexity. The results raise important challenges for prevailing institutional theories in economics, political science, and sociology about how governments are structured and the factors that explain variation in that structure. In short, the extensive footprint that U.S. governments leave on the internet provides a new window on government—a window that enables a clearer, deeper picture of how governmental institutions work as well as a method for advancing theoretical understanding of the causes and consequences of that structure.