The Press and Peace, Examining Iraq War Coverage in Newspapers using BERT LLMs
by Jakobe Bussey
Abstract
This study utilizes state-of-the-art BERT (Bidirectional Encoder Representations from Transformers) models to perform sentiment analysis on Wall Street Journal and New York Times articles about the Iraq War published between 2002 and 2012 and further categorize them using advanced unsupervised machine learning techniques. By utilizing statistical analysis and quartic regression models, this paper concludes that the two newspapers report on the Iraq War differently, with both exhibiting a predominantly negative-neutral tone overall. Additionally, the analysis reveals significant fluctuations in negativity from both outlets over time as the war progresses. Furthermore, this study examines the objectivity of reporting between editorial and non-editorial articles, finding that non-editorials tend to report more objectively, and the neutrality of editorials remains relatively constant while the objectivity of non-editorials fluctuates in response to war events. Finally, the paper investigates variations in sentiment across different topics, uncovering substantial variations in positive, neutral, and negative sentiments across topics and their evolution over time.
Professor Peter Arcidiacono, Faculty Advisor
JEL Codes: L8, L82, H56
Illuminating the Economic Costs of Conflict: A Night Light Analysis of the Sri Lankan Civil War
by Nicholas Kiran Wijesekera
Abstract
This paper investigates the economic consequences of the Sri Lankan Civil War (1983-2009) by using event-based data on civilian and combatant fatalities in addition to night light imagery as a proxy for economic activity. By looking at regional economic activity across the island of Sri Lanka, this paper seeks to identify how violence led to declines or undershoots of economic activity in the areas in which it was most prevalent. The use of night light data gives a hyper-localized proxy measurement of this activity for each year of the war. The investigation finds that government and rebel deaths have strong, negative effects on economic activity, and that these effects spill over across time and space. Additionally, the manner in which civilian deaths occur is an important determinant of their subsequent economic impact. The paper offers new findings on the economic legacy of the Sri Lankan Civil War and extends existing work on the use of night light data to measure economic activity during conflict.
Professor Charles Becker, Faculty Advisor
Professor Michelle Connolly, Faculty Advisor
JEL Codes: H56, N45, O53
Economic Effects of the War in Donbas: Nightlights and the Ukrainian fight for freedom
Paper available to internal Duke affiliates only upon request.
Professor Charles Becker, Faculty Advisor
Professor Grace Kim, Faculty Advisor
JEL Codes: F51; H56; O52; N44
The Impact of Conflict on Economic Activity: Night Lights and the Bosnian Civil War
by Stephanie Dodd
Abstract
The tendency of violent conflict to suppress economic activity is well documented in the civil war economic literature. However, differential consequences resulting from distinct characteristics of conflicts have not been rigorously studied. Utilizing new conflict data on the 1992-1995 Bosnian civil war from Becker, Devine, Dogo, and Margolin (2018) and DSMP-OLS night light data as a proxy for economic activity, this paper investigates the disparate economic impacts that different types of conflict have on Bosnia’s municipalities.
This investigation first uses data from other Yugoslavian countries to impute pre-war night light values for conflict-affected Bosnian municipalities. Next, a spatial autocorrelation model with fixed effects is used to determine if and how the occurrence of different types of violence vary in their implications for economic activity. This analysis finds that the five types of warfare identified in the context of the Bosnian Civil war have different impacts on night lights and economic activity.
Professor Charles Becker,Faculty Advisor
Professor Grace Kim, Faculty Advisor
JEL Codes: F52, H56, O52
