Data analysis shows that his negative tweets help his approval rating, but his attacks on “fake news” fall flat
Most Presidents communicate with the public through regular press briefings and press conferences. But not Donald Trump. His primary means of communication is Twitter. That medium has the advantage of immediacy, but it also results in statements at odds with the policy the rest of the administration is following. In short, he can and frequently does tweet in a fit of pique in the middle of the night without the benefit of calm, thoughtful, consideration or the advice of his advisors. As a result, some have argued that his tweeting is a political net negative to his administration.
On the other hand, his tweeting is prolific. As I write this article, Trump has been in office 681 days. Of those, according to my calculations, he has tweeted at least once 659 days. In total, he has tweeted 4898 times as President for a daily average of nearly 7.5 tweets. That doesn’t include his many retweets or his thousands of tweets from before his inauguration. Clearly, he thinks he is getting something from all this tweeting.
So which is it? Do his tweets hurt him or help him? It’s hard to characterize anything he does as strategic, but if there is anything he has done well as President it is his relentless tweeting. Paul Krugman of The New York Times even calls him the “tweeter in Chief.”
Things only get worse for Trump from here
A skilled politician might be able to turn this situation to his advantage, as Bill Clinton did. Trump is not that…
So I decided to dig into the data to come up with an answer to that question. For each day of his Presidency, I added up all his tweets for the day, and how many of those included the words “fake news,” “dumb,” “dummy,” “terrible,” “stupid,” “weak,” “dishonest,” “lightweight,” “incompetent,” “boring,” “fool,” “pathetic,” “clown,” and “disgusting,” assuming that these represented the bulk of his negative, attack tweets. Finally, I specifically added up the tweets each day that included “fake news.” The result is a database of all the days of his Presidency with the total number of tweets, the approximate number of negative tweets, and the actual number of tweets screaming “fake news.”
To that data, I attached Trump’s favorability and unfavorability ratings from three days after each tweetstorm, as downloaded from Nate Silver’s FiveThirtyEight. I figured it takes about three days for one day’s activity to impact the polls, so thus the delay. I made the data available here for you to do your own analysis.
What did I find? Well, it turns out that Trump’s angry tweeting seems to help him. I found a statistically significant positive relationship between the number of negative tweets in a day when he tweets more than once (thus his tweetstorm days) and the net positive approval rating among all adults (approval minus disapproval) for the President three days later (β=0.44, p=.049). Interestingly, the same equation also found a negative, statistically significant relationship between his “fake news” tweets and his net approval rating (β=-0.61, p=.088). Thus, it seems that his negative attack tweets help his approval rating, while his accusations of fake news don’t.
So what does this mean? As the saying goes, there are lies, damn lies and statistics, so people can get statistics to say anything they want. But what this data seems to say is that Trump’s negative tweets earn him points among much of our population. Much as we want to believe that people don’t like negative campaigning, it does work. Just ask Robert Mueller, who has seen his approval ratings plummet in the wake of relentless attacks from the Republicans. Indeed, psychologically, we are hard-wired with a “negativity bias” which makes negative information more powerful to us than positive information. Therefore, Trump seems to be tapping into his supporters when he engages in his attack tweeting.
On the other hand, his administration is an unmitigated disaster of incompetence and corruption. I believe people still do care about whether our country is on the right track or the wrong track. Therefore, on days when Trump criticism in the media for his actions — the days he is likely to tweet that everything reported is fake news — his approval rating takes a hit.
Believe me when I say that this is not the result I was hoping to find. I would have loved to report that Trump’s tweeting hurt him as I am deeply offended by it. My wife was concerned that revealing this information would be helpful to him. But I wonder, had I found a different result, would Trump’s behavior have changed?
The one lesson that does seem to come from these results is that engaging him in his negativity is a no-win proposition. Focusing on his incompetence and cruelty seems to be a better strategy, as was shown in the recent midterm election campaigns.
If you liked this post, you might also like: