Skip to content Skip to main navigation Report an accessibility issue

Last Man Standing: Sentiment Analysis of Joe Biden, Presumptive Democratic Nominee

Well it finally happened. And ahead of schedule. Both major parties have presumptive nominees (i.e. candidates who have obtained or are on track to get the delegates necessary to secure their party’s nomination on the first nominating ballot). The Republicans will renominate President Donald Trump of Florida. Democrats, meanwhile, with the exit of Vermont Senator Bernie Sanders, are on track to nominate former Vice President Joe Biden of Delaware.


Images of Joe Biden and Donald Trump.

Joe Biden (left, Democrat) and Donald Trump (right, Republican) are the presumptive nominees of their respective parties and will likely face each other in the 2020 US Presidential election. Source: The Hill

Biden’s emergence as the presumptive Democratic nominee and most likely Trump general election opponent is the unlikely end to a months long saga of poorly attended rallies, anyone but Biden punditry, and the Trump led attacks, conspiracy theories, and dirt digging that ultimately ended in Trump becoming the third president in history to be impeached by the U.S. House of Representatives. While this saga is carefully inscribed here, here, and here, the purpose of this post is to examine Twitter sentiments of Biden’s emergence as the presumptive Democratic nominee.


To do this, we create a topic profile in Social Studio using the term “Biden”. Looking specifically at Twitter results, we see there are 4.8 million posts on the topic as of this article’s publication. Looking specifically, at the sentiment analysis component of the Biden social listening summary, we see that there is a mostly negative response with 61.6% of results shown as “negative” and the remaining 38.4% shown as “positive.” 


It is important to note though the limitations of the algorithm that is used to conduct the sentiment analysis. If you were to spend some time going through the results, you might not agree with each decision to mark a tweet as having a positive or negative sentiment. However, this automatic sorting does provide you a helpful place to start.


Based on the sentiment results, you might download all of the negative tweets and do an analysis to identify reasons why people have a negative opinion. This might be helpful for identifying groups such as Trump supporters, Sanders supporters, candidates who wanted a younger nominee, etc., to help see what Biden might do to reach out to these anti-Biden users. These efforts might include a progressive vice presidential nominee, a unity cabinet with members stretching from the moderate wing of the Republican Party to individuals supported by Sanders and Warren (if not one of them directly), or conceptualizing strategies that could be used to reach younger voters who have been cool to Biden’s campaign so far.


This post shows you one way in which sentiment analysis might be helpful for identifying trends useful for research or practice. If you have not looked closer at the sentiment analysis feature in Social Studio, do so today!