Data/code files for the following project: I study the viability of Twitter-based measures for measuring public attitudes about the police. I find that Twitter-based measures track Gallup's measure of public attitudes starting around 2014, when Twitter user base stabilized, but not before 2014. Increases in Black Lives Matter protests are also associated with increases in negative sentiment measures from Twitter. The findings suggest that Twitter-based measures can be used to acquire granular evaluations of police performance, but they can be more useful in analyzing panel data of multiple agencies over time than in tracking a single geographical area over time.
Time Period(s): 1/2010 – 12/2021
Collection Date(s): 2021 – 2022
Universe: Public attitudes data of US cities in the 2010s
Data Type(s): aggregate data; observational data; survey data
The following publications are supplemented by the data in this project.
