Something interesting we’ve seen over the past month is the online volume trends of the top candidates in the Democrat Presidential Primary. We will often see large daily variations in volume, but the top tier of candidates has generally been maintaining consistent average volumes over weekly or monthly periods.
The consistency we normally witness is why it is interesting to see the volume trend of discussions regarding Joe Biden throughout July. Biden’s average volume during the past two weeks of the month dropped 30% from his average volume during the first two weeks. In comparison, Warren and Harris saw their volumes decrease a more reasonable 11% and 15% respectively, while Sanders saw a huge increase of 49%, although it came with the largest decrease in sentiment, primarily due to the negative comments surrounding his campaign’s minimum wage issues.
The lack of online discussion about Biden starts with a lack of reasons to discuss him, and this could be due to the other candidates pursuing the media’s attention more aggressively, or simply a decision by his team, such as more debate prep over the past two weeks. Whether it is coincidental or by design, the numbers indicate his campaign team has been comfortable allowing others to take the spotlight in recent media cycles. Whether this continues post-debate is a trend we will be looking for in our online data in the days ahead.
The President's recent Baltimore tweets have caused yet more criticism in the media and online, but will it have any impact on his overall support?
The past two weeks have given us a number of examples that demonstrate the pattern of activity when the President says or tweets something controversial. Tweets about the “Squad” that garnered massive attention, a rally where the “send her back” chant took over the media narrative for days, and the Mueller testimony all made for stories that generated huge volumes of discussion.
Through our partnership with Scott Rasmussen (http://bit.ly/SRTrump), we’ve been monitoring the online sentiment expressed toward the President, which gives us real-time insights into the public reaction to his actions.
If we focus on the last two weeks, we can see the large initial drop in sentiment caused by the Squad tweets. As we discussed in a previous post, we found most of this negativity expressed by people who were previously negative towards the President, and it wasn’t impacting his base of support in a meaningful manner.
From the chart (h/t www.scottrasmussen.com), we see that POTUS’s sentiment started to recover within days. It also appears that the House condemnation of his comments and how that event played out may have provided the impetus for the recovery. The initial negativity of the rally chant the next day put the brakes on the recovery, but it then continued through to the end of the week.
The same pattern occurred when Robert Mueller testified at the end of the past week. Initial negativity drove sentiment temporarily lower, but it is already recovering back to the levels witnessed before the Squad tweets.
It is this recovery from the negative opinion that has been a constant characteristic of the President’s sentiment scores over the past months. While events fuel non-supporters comments in the short term, Trump supporters tend to rapidly come to his defense to change the negative momentum to positive.
This pattern makes sense when we look at what polls have been showing for the President’s approval ratings. Just as our average sentiment tends to stay in a range we would characterize as slightly negative, the Scott Rasmussen/Harris X polling finds the same consistency in their numbers as well. As pointed out in the July 25th update, "since January 28, the number approving of the president’s performance has stayed within two points of 46% every single daily release." The consistency found in the polling results mirrors the consistency we see in the President’s sentiment over the medium to long term.
Overall, for all of the ups and downs that appear to be playing out in the Oval Office, the reality is that public opinion of the President tends to be remarkably consistent regardless of his actions.
Understanding this is likely a key factor in the President’s communications strategy, and Eyesover can provide this information to any organization, in real-time, to help inform strategic decisions.
When a negative story breaks in politics, it is always a challenge for the politician or political party to know what their next step should be. Whether the story is breaking in traditional media or online, in those first few hours, it is next to impossible to know how big the story is, who is talking about it, or how it is impacting the politician or party.
Let's take a look at Donald Trump's recent tweets about the "Squad." There is no question the overall impression of the tweets and the related backlash took a decidedly negative tone. Trump's overall average sentiment dropped 65% over the three days after he tweeted his comments (-0.0982 to -0.1618 on a -1 to +1 range via www.scottrasmussen.com). While our analysis usually finds the President’s sentiment in a slightly negative territory, it takes an enormous number of negative comments to drive the average sentiment below the -0.15 level.
But from where is the negativity coming? Is Trump losing supporters due to his comments, or is the backlash primarily from people who would never vote for him regardless of his actions? Many feel that much of Trump’s motivation for his approach to social media is to reinforce his base, so the importance of knowing the reaction of those who have previously indicated support for him can't be overstated.
As it happens, this is exactly the kind of analysis you can perform with Eyesover. From a sample (n=126,962) of accounts commenting on the President’s tweets over the past few days, we can determine how many people have changed their opinions towards POTUS. To do this, we compare the sentiment of previous comments to the sentiment currently being expressed from the same account.
The numbers show that while some previously positive individuals have turned negative, Trump has been gaining new supporters at roughly the same pace over the past few days. The lack of a significant change in supporters indicates the negativity shown in his overall sentiment scores is largely caused by individuals who have always spoken negatively about the President and are outside of the base of support he has built and is trying to maintain.
In any campaign, it is critical to know if changing issues require a change in strategy to mitigate risk or take advantage of opportunities. Eyesover provides the data campaigns need to determine when to change strategy or when to maintain the status quo.
While recent polls have shown a tight race between the Conservatives and the Liberals for the lead in the Canadian political landscape, at the same time, there has been some interesting movement among the NDP, Greens, and PPC as well.
When we look at the share of online mentions for each party during June, we see the majority are discussing the Liberals (61%), which is understandable due to the fact they are the governing party. The Conservatives follow at 28% while the New Democrats, People's Party, and Greens are 5%, 3%, and 3% respectively.
The fact that our tracking has found a sustained downward trend in sentiment over the past few months for the Liberals aligns with their recent fall in the polls. The trend continues to indicate that much of the online activity for the Liberals has been increased criticism of their recent actions.
But while the focus tends to be on the LPC and to a lesser extent, the CPC, we can't forget that there is an increasing likelihood that the NDP, Greens, and PPC will have a say in who will be able to form a government this fall.
When we look at the third, fourth, and fifth party mentions online, what catches the eye is the volume of PPC mentions. The party was discussed more than the Green Party in June and had just over 50% the number of mentions of the NDP. For a one-MP party that is less than a year old, there is considerable online traction.
Geographically, comparing where the mentions are coming from compared to provincial populations, the PPC mentions are over-weighted in Ontario, Alberta and BC despite the fact their sole MP, Maxime Bernier, is from Quebec.
The share of online mentions we show above aggregates daily mentions from the same account so it does not skew the one person, one vote concept, but we can also dive into the raw data to get an idea of how active people are when it comes to mentioning a party. To make sure our analysis does not get skewed by fake accounts or bots, we apply a thorough vetting to ensure accounts represent real individuals or organizations, and we ignore non-original comments such as retweets.
Of particular interest with the PPC is the fact that the average number of mentions from an account is 2.46. This level of activity is far higher than any of the other parties (CPC: 2.04, LPC: 1.89, Green: 1.72, NDP: 1.56) possibly indicating the PPC has a higher level of support intensity than other parties. The low number for the NDP could be an indication of a lack of enthusiasm within their base.
Something we will be watching over the next few months is to see how concentrated the PPC support is, and if they can turn their base's enthusiasm into a larger vote count in some ridings.