News Coverage of Tear Gas Has Increased Since The Arab Spring
During the Fall of 2014, with the protests sparked by the killing of Mike Brown in Ferguson, MO, the pro-democracy Umbrella Movement in Hong Kong, the Sunflower Revolution and others, the news was constantly reporting the use of tear gas against the citizens. Thinking back to the Arab Spring, Occupy Wall Street, and Occupy Gezi, it seemed as if the use of tear gas was on the rise. Initially, I wanted to see if I could find data that would indicate if tear gassing demonstrators was becoming a more globally stock response of the police or military instead of listening to the concerns of the citizens.
I explored the Global Database of Events, Language, and Tone (GDELT) to see if I could a reliably determine a metric for finding unique events when tear gas was used. While the dataset does tag events identified as ‘protests’, it does not include extended information about what occurred during those events. I began looking elsewhere, but finding the data representing both the time tear gas was deployed coupled uniquely identifiable locations for the events, proved difficult. Wikipedia has multiple entries about major protests, but the information is unorganized and inconsistent.
However, the data I did find was the reporting on the protest events when tear gas was used (and broadly the region where the event took place), so I decided to chart the frequency news articles mentioning ‘tear gas’ over time.
To compose the dataset, I used the New York Times Article Search API and the Guardian Content API to filter through the coverage to find the unique news events (removing anomalies in the cooking section, or opinion sections, for example).
The resulting graph functions as a hybrid between a dispersion graph showing density, and a bar chart. The vertical spikes show the number of articles per day mentioned tear gas, while the denser and darker sections shows the sustained newsworthiness of tear gas over time. The blue line running across the bottom of the graph represents the frequency of protest events throughout the world according to the GDELT data.