Publications
Grossman, Matthew & Mahmood, Zuhaib & Issac, William. (2020). Political Parties, Interest Groups, and Unequal Class Influence in American Policy. Journal of Politics. Pre-print. Link.
Colaresi, Michael & Mahmood, Zuhaib. (2017). Do the robot: Lessons from machine learning to improve conflict forecasting. Journal of Peace Research. 54(2): 193-214. Link.
Colaresi, Michael & Mahmood, Zuhaib. (2017). Do the robot: Lessons from machine learning to improve conflict forecasting. Journal of Peace Research. 54(2): 193-214. Link.
- Winner: 2017 best visualization award, Journal of Peace Research
Dissertation Abstract
A range of literature has argued that countries use the United Nations (UN) to communicate their interests and coordinate their foreign policies. However, a core behavior at the UN -- public-facing speeches -- have remained understudied. While negotiation often takes place in private, the puzzle remains of why countries would then turn to a public forum to communicate--given the always-present option of a private one. In this dissertation, I expect that public forums serve a distinct purpose as compared to private negotiations: rather than being persuasive in nature, these public communications serve as being political in nature. I expect that the countries who benefit most from this broadcasting capability will be those who benefit most from communicating their position to a large audience, be those audiences international or domestic. In order to isolate the purely political nature of these public-facing speeches, I draw on the text of speeches delivered at the United Nations General Assembly (UNGA) for the years 1984-2014.
This dissertation is composed of three chapters, and makes two broad contributions. First, it provides political grounding to one of the most public, visible, and prolific behaviors in international politics: public-facing political speeches delivered at international institutions. This has implications for both the study of the UN and the study of diplomacy more broadly, and provides a framework to study other institutions in the future. Second, it casts doubt on the conventional wisdom that these types of speeches are not informative: to the contrary, I show that all three dimensions of speech--volume, content, and alignment--map on to patterns of behavior--namely, UN Peacekeeping troop deployment; intervention into civil wars; and UNGA voting similarity. This suggests that each dimension of political speech contains useful information about countries' underlying interests.
In the first chapter, I analyze why countries engage with the UN General Assembly at all. I argue that engagement with the UNGA is driven by the political benefits a country perceives from communicating opinions to politically relevant audiences. I show that countries with the capacity and interest to act in many issues, as well as countries with strong political incentives linking them to influential audiences at home, will tend to speak the most at the UNGA. I also validate that the volume of UNGA speech does in fact measure a perception of political benefit from the UN more broadly, showing that the amount of speech at the UNGA is strongly related to troop contributions to UN Peacekeeping missions
In the second chapter, I examine the scope of interests that countries speak on in their public facing speech. This provides, in part, a mechanism for the first chapter: countries with the material capacity to act in more issues, and countries with the political incentives to communicate with a more diverse audience, will tend to speak on a more diverse set of issues in their speeches. I also validate that the issues countries speak on tend to be ones of political interest: I do so by showing that when a country experiences a civil war, countries who speak similarly will be more likely to intervene in some way in the conflict.
Finally, in the third chapter, I examine the conditions under which countries use speech to reinforce or hedge against the underlying incentives and interests driving their behavior. I examine two such conditions. First, I show under conditions where coercion is more likely--such as where a large power difference exists--hedging is more likely since it is more likely there is a difference between underlying positions and observed positions. Second, I show that conditions under which friends and enemies are more clearly defined--including periods of rivalry, joint democracy, and as issues become more controversial--the cost of hedging will increase and thus countries will tend to reinforce in their speech. I use UNGA voting alignment as behavioral baselines for measuring reinforcement and hedging across issues. This chapter also relies on an ideal point measure--created in collaboration with Michael Colaresi, using a Dynamic Linear Model to cast ideal points in multiple dimensions based on speeches--to measure hedging and reinforcing speech.
This dissertation is composed of three chapters, and makes two broad contributions. First, it provides political grounding to one of the most public, visible, and prolific behaviors in international politics: public-facing political speeches delivered at international institutions. This has implications for both the study of the UN and the study of diplomacy more broadly, and provides a framework to study other institutions in the future. Second, it casts doubt on the conventional wisdom that these types of speeches are not informative: to the contrary, I show that all three dimensions of speech--volume, content, and alignment--map on to patterns of behavior--namely, UN Peacekeeping troop deployment; intervention into civil wars; and UNGA voting similarity. This suggests that each dimension of political speech contains useful information about countries' underlying interests.
In the first chapter, I analyze why countries engage with the UN General Assembly at all. I argue that engagement with the UNGA is driven by the political benefits a country perceives from communicating opinions to politically relevant audiences. I show that countries with the capacity and interest to act in many issues, as well as countries with strong political incentives linking them to influential audiences at home, will tend to speak the most at the UNGA. I also validate that the volume of UNGA speech does in fact measure a perception of political benefit from the UN more broadly, showing that the amount of speech at the UNGA is strongly related to troop contributions to UN Peacekeeping missions
In the second chapter, I examine the scope of interests that countries speak on in their public facing speech. This provides, in part, a mechanism for the first chapter: countries with the material capacity to act in more issues, and countries with the political incentives to communicate with a more diverse audience, will tend to speak on a more diverse set of issues in their speeches. I also validate that the issues countries speak on tend to be ones of political interest: I do so by showing that when a country experiences a civil war, countries who speak similarly will be more likely to intervene in some way in the conflict.
Finally, in the third chapter, I examine the conditions under which countries use speech to reinforce or hedge against the underlying incentives and interests driving their behavior. I examine two such conditions. First, I show under conditions where coercion is more likely--such as where a large power difference exists--hedging is more likely since it is more likely there is a difference between underlying positions and observed positions. Second, I show that conditions under which friends and enemies are more clearly defined--including periods of rivalry, joint democracy, and as issues become more controversial--the cost of hedging will increase and thus countries will tend to reinforce in their speech. I use UNGA voting alignment as behavioral baselines for measuring reinforcement and hedging across issues. This chapter also relies on an ideal point measure--created in collaboration with Michael Colaresi, using a Dynamic Linear Model to cast ideal points in multiple dimensions based on speeches--to measure hedging and reinforcing speech.