Measuring Democracy: Analyzing Public Sentiment in Swing States through Twitter

• In Event: Poster Session: Political Communication 1 & Politics, Literature, and the Arts
  In Poster Session: Poster Session: Political Communication 1

Abstract

The 1990s saw a wave of democratization and market liberalization develop across much of the world, notably in Latin America and much of the former Soviet Bloc. Many in the West interpreted this as proof of the inherent superiority and staying power of liberal democracy and market economics (Fukuyama 1992). Since around 2006, the trend of democratization has slowed and democratic backsliding has taken hold, including in the United States (Repucci and Slipowitz 2022). For the past few years, democracy has been an increasing topic of focus as the outlook of the future of democracy has been changing constantly. Global geopolitical and economic forces have been altering the trajectory of democracy in several places.

The 2020 U.S. Presidential election saw an increase in discourse of democratic health in the United States, due to rising questions over the legitimacy of the electoral process and outcomes, particularly from the right. With the increasing use of social media, it has become easy for individuals to sow distrust over the electoral process and call into question the legitimacy of democratic institutions. Misinformation has spread over several aspects of the democratic process of elections, such as method of voting or legitimacy of the results. This, in turn, has shifted the perception of democracy and furthered polarization among the electorate. Public sentiment can provide crucial insights to how the public feels about the direction of democracy and confidence in their elected leaders.
To understand the power of public sentiment over democracy, Twitter data from the 2020 election is analyzed to understand the difference between tweets that mention Trump and Biden and their perception of democracy, specifically in battleground states. Battleground states were chosen because of their electoral importance and nearly equal political representation. Twitter data was analyzed between 10/15/2020 to 11/8/2020 between the two datasets. The data was then filtered for swing states, and then keyword mined for the word ‘democracy’. The sentiment scores were then calculated for the tweets that mention democracy and were scored either positive, negative, or neutral. The top ten keywords of both positive and negative sentiment were looked over in the Trump and Biden datasets. There were several similar positive words between the Biden and Trump dataset, including “save”, “win”, “like”, and “united”. Conversely, the top 10 negative sentiment words for both datasets highlighted differences between the groups. The Biden dataset included “desperately”, “evil”, and “racist”, while the Trump dataset included words “attack”, “fraud”, and “threat”.

To further understand if there was a difference between the positive and negative top ten keywords between each of the two datasets, two sample t-tests were run, showing that there is a significant difference between the positive and negative words in each of the datasets. In the Biden tweet dataset, the positive keyword sentiment (n= 429) held a mean of 2.282051 whereas the negative keyword sentiment (n= 249) was at -2.240964. The Trump dataset has a positive keyword sentiment (n=413) mean of 2.033898 with the negative keyword sentiment mean of -2.140203 (n= 592). The results show that the Biden tweet dataset tends to be more optimistic and take on a positive perception of democracy whereas the Trump tweet dataset tends to have a slightly more pessimistic and destructive view of democracy. The negative words of both candidates suggest that both sides view the other as an existential threat to democracy and the United States. These findings illustrate the concerns and attitudes of individuals in swing states over their perception of democracy during the 2020 election. The Trump tweet dataset shows that individuals living in these battleground states tend to have a more pessimistic view on democracy whereas the Biden dataset carries a more cheerful and optimistic tone towards democracy. The results show the stark division between the two candidates, which have resulted in further polarization. Understanding the perception of democracy is vital as a democratic government consists of individuals who respect the rule of law and are engaged in their communities in several different ways. This brings about the future of civic engagement and what challenges it will come in motivating individuals to come and vote. This is in line with previous literature that suggests that increasing political polarization is the result of an increase in negative partisanship. Democracy functions when people have confidence and trust in their institutions and processes, and if there is a gradual decrease in that confidence, democracy will slowly wither away.

Authors

• Justin Dritschel, Columbia University

• Anusha Natarajan, Arizona State University