Green New Deal Sentiment Analysis

Count of Tweets

The chart above that shows a huge spike in tweets about the Green New Deal the day after the debate—confirming our theory that we would get more data about the GND during this time frame as a large and more diverse group of people are enageing with the topic.

Location of Tweets in LA in 2020

As you can see from the results, there are some inconsistent geo locations included despite the geo filter. However, the majority of located tweets are in Larose and New Orleans.

Location of Tweets in the US post Debate

Of these 269 tweets, 14 come from LA and 12 from Brooklyn. This indicates that the discussion of the GND is occurring in larger, progressive cities and not in areas like Louisiana that have a greater need for climate change policies.

Text Analysis of GND Tweets post Debate

Word Counts

We return to the first query of all GND tweets in the 2 weeks following the debate to have a large sample size for our text analysis. This yields 12,094 tweets. These results would indicate a generally positive opinion of the GND, with the caveat that Twitter users are a selection of people who are not representative of the general population in terms of income and education.

Sentiment Analysis

The median polarity of tweets is 0.125, the mean 0.118. This indicates a slightly positive skew to the discussion surrounding the GND.

)

The median subjectivity is .497 (on a scale of zero to one) indicating that the subjectivity skews lower.

)

We then plot subjectivity and polarity against each other, and observe the trend that as subjectivity increases, so does polarity in both directions. That is, as the tweets get more and more subjective, they are also either most positive or more negative. This trend makes sense given the charged nature of the Green New Deal.

)

We group the biased tweets by the hour of the day and generate a box and whiskers plot of polarity for each hour. Polarity skews negative around 1am and 3am.