This post is the continuation of our media monitoring series examining coverage behavior during Election 2012. Today’s post discusses the statistical data analysis of the language used in coverage of the GOP Primary and the General Election.
In coverage of the GOP Primary, from January 1 – April 30, the most commonly used words were “Romney”, “president”, “people”, “Obama”, and “going”. Clearly, Romney was the most discussed person during the GOP primary. Romney and the other GOP candidates also frequently mentioned President Obama. Obama’s name was always wrapped up within their arguments for why he should not get a second term. These data points clearly demonstrate that all of the GOP candidates were running against the tandem of Romney and Obama – in some ways they were accurately forecasting the eventual candidates in the general election. The frequency of the term “going” is probably evidence of any political campaign, with the candidates routinely suggesting what they were “going” to do as President.
GOP Primary Coverage Word Cloud
Romney, Santorum, Gingrich Dominated GOP Primary
Many top words from GOP Primary coverage are associated with the Republican candidates themselves, showing that while they often attacked Obama, they also attacked each other. The words “conservative”, “republicans”, “republican”, “nominee” are among the most common. Also, “Santorum” and “Gingrich”, as Romney’s toughest competition in the race, made it into the top 100 words. There is also clear evidence of Election 2012’s cornerstone issue: the economy. Statements from the candidates were peppered with the words “economy, “money”, “tax”, and “jobs”.
Among newsmakers who contributed to the GOP Primary coverage, Mitt Romney unsurprisingly ranked highest. He had over 12% VoiceShare in all election coverage of the GOP Primary. He drove the frequency of many of the words in the above cloud. The other two candidates whose names were among the most common, Santorum and Gingrich, came in at second and third in our final candidate VoiceShare tally.
General Election Coverage Word Cloud
The common words in coverage of the general election, from May 1 – November 6, are slightly different than those from the GOP Primary. The word “Obama” is shown in larger font in this cloud, signifying a higher frequency. “Romney” is in larger font than “Obama”, reinforcing our 4th Estate data that Romney was more frequently discussed than Obama was. The focus on the economy is more present in this cloud with the words “jobs”, “money”, “tax”, “taxes”, and “economy” being displayed in larger font than before, indicating that “economy’ took on a greater presence within the general election . Other issues that were more prevalent within the general election are also present – such as “medicare”. The presence of the word “Ryan” is evidence of Paul Ryan’s entrance into the race as Romney’s running mate. Ryan’s entrance into the race also increased the discussion of such topics as medicare. The presence of the word “percent” is evidence of the extent of Romney’s gaffe about the “47 percent”. That comment was syndicated and amplified by many news outlets in print, television and radio.
Obama, Romney Most Heavily Influenced Narrative
Just like in the GOP Primary, Mitt Romney had the highest VoiceShare in coverage of the general election. Over one-third (36%) of the statements in coverage from May to November came from him. Obama gained significant VoiceShare during the general election (as compared to the GOP rpimary) as the race turned into a battle between him and Romney. Overall the VoiceShare slanted heavily toward the two candidates, with 60.4% of the statements from May to November coming from them. In this regard, Obama and Romney are the primary drivers of these word clouds as a majority of the most commonly word used were from Obama and Romney themselves.
Finally, it is interesting to analyze these clouds in the context of negative campaigning. It is popularly regarded that the tone of this campaign was extremely negative, even by historical standards. Even taking into account that stop words like “no” and “not” have been
excluded from the data, there is a complete absence of negative key words within these clouds. This highlights the deft touch with which the candidates undermine their opponents, and attack them when they are on the campaign trail. They do not use direct attacking language, filled with ‘negative words’. Instead, they couch their attacks in everyday language, that can’t come back to haunt them – such as a word cloud revealing their propensity for aggressive or negative terms. We have not analyzed the data, but it would be interesting to perform a word cloud of the collected commercials of each campaign to see how they compare to our news coverage clouds.