Stuart Soroka is professor of communication at UCLA. He’s been collecting data on TV news coverage of the 2024 presidential campaign and has made his findings available on his website. We spoke this week about what he’s found. This interview has been lightly edited for clarity.
Danny Hayes: Since Joe Biden dropped out of the presidential race, your data show that the Harris campaign seems to be getting better TV coverage than the Trump campaign. Is that an accurate interpretation of your findings?
Stuart Soroka: Yes. These data capture the “sentiment” of sentences that mention either Trump or Harris. And the sentiment of Harris’ coverage has been more positive than the sentiment of Trump’s coverage for nearly every day since she entered the race. In fact, the only real exception is for the two days following the second attempt to shoot Trump. In this moment, coverage of Trump being “safe” (which is positive language) pushes the sentiment of coverage in his direction. I’m not sure we should really interpret that as a Trump advantage in media sentiment, however – it does not really reflect an evaluation of the candidate. And for the rest of the campaign, coverage of Harris has been more positive than coverage of Trump.
Can you briefly explain where these data come from and how you measure the favorability of media coverage for Harris and Trump?
I analyze television news transcripts from ABC, CBS, CNN, and Fox News, extracting every sentence that mentions either Trump or Harris. I don’t include NBC and MSNBC because their transcripts aren’t available.
I then use a content-analytic dictionary to identify the number of positive and negative words in every sentence. “Sentiment” for a sentence is the difference between the number of positive words and negative words. And candidates’ advantages in sentiment in recent months, shown on the figure below, reflect differences in the sentiment of all Trump sentences versus all Harris sentences.
Why has coverage been more favorable for Harris than Trump? I suppose the Trump campaign might say this is evidence that the media want Harris to win.
I think there are two primary reasons. The first is that there have been moments in the campaign that have been good for Harris. Harris got more positive coverage during the Democratic National Convention (DNC), for instance, when many Americans were clearly very excited about her candidacy. Harris also got very positive coverage surrounding the Sept. 10 debate, which most commentators believe she won. In these instances, Harris is clearly generating more positive media commentary than Trump.
I would not say that these results confirm that the media want Harris to win, however. Keep in mind that Fox News is one of the four networks captured in this measure. While ABC and CBS have just a few news programs, the two cable networks – Fox News and CNN – broadcast the news 24/7. Given how much content there is from the two cable networks, Fox News accounts for more than a quarter of the data.
If we break down the measure by network, Fox News is much less positive about Harris than the other networks. But even for Fox News coverage in August and September, Harris has more positive sentiment than Trump, by a small margin. There are always biases in media coverage, of course, but the changes over time in media coverage also reflect the reality of the campaign.
The second reason coverage has been more positive for Harris has to do with the things the candidates themselves are saying. Sentiment is based on the words in sentences that mention Harris or Trump. So this measure is affected by positive and negative words newscasters say about the candidates, but also by positive and negative words the candidates themselves say. And Harris’ messaging has been relatively positive, while Trump’s messaging has been more negative. As a consequence, we should expect Harris to have slightly more positive sentiment much of the time.
So if Trump says “America is a country in decline” and Harris says “I see a nation ready to move forward,” that might tilt the sentiment data more favorably for Harris that day?
That’s right. But that is not to say that the difference in sentiment is just about what the candidates are saying. There is variation over time, even as the candidates’ messaging is not changing fundamentally. And that variation is clearly linked to moments of the campaign that were good for Harris. So the tone of the Harris campaign is more positive, and she is getting more positive coverage.
Can you characterize the size of Harris’ coverage advantage? Would you say that it’s large, small, etc.?
Here is one way that I’m thinking about the 2024 data: When Biden dropped out, and especially in the midst of the DNC, coverage of Harris was very positive. These moments produced as much of a partisan difference in sentiment as we could reasonably expect in a campaign. And so, moving forward, I’ll be watching to see whether Harris can repeat that huge advantage, or if Trump can get his own similarly large advantage. That seems unlikely on either side. But any advantage that is even half of what we saw when Harris entered the race reflects a big difference in coverage. This is what we saw surrounding the debate, for instance.
A big question here is whether getting more favorable TV coverage might be responsible for Harris’ rise in the polls. What’s your take on that?
I think good coverage doesn’t hurt! Whether that coverage changes people’s opinions of the candidates, or just reflects what the public thinks about the candidates, is unclear. Media outlets are both affecting and reflecting attitudes most of the time. So we should take this measure of sentiment as an indication of the “state of the campaign,” in both media coverage and in public attitudes. And that state of the campaign has quite clearly been advantageous for Harris thus far.
