With Donald Trump now facing an impeachment inquiry, cable news is, understandably, all-in on the Ukraine story.
Trump’s July 25 phone call with Ukraine President Volodymyr Zelensky, and the whistleblower complaint that stemmed from it, has meant that a national press corps already obsessed with Trump and Russia has also become obsessed with Trump and Ukraine. And while there are substantive (and, perhaps, politically fatal) differences between the two scandals, a look back at cable news coverage of the Trump-Russia collusion scandal offers lessons in how the press should proceed on Ukraine.
Twenty-four-hour cable news, by its very nature, is repetitive. Reporting on a top story at 11am will likely sound very similar to reporting on the same story at 7pm. Even with no significant developments, coverage of major stories often persists for days, and sometimes weeks, as a series of sustained narratives.
Yet cable news’ treatment of the Trump-Russia collusion scandal was exceptionally overwhelming. Our organization, Harmony Labs, worked with dMetrics, a natural language processing firm, to examine closed-captioning transcripts of cable news programs catalogued by Internet Archive between election day and April 19, 2019.
During that time, MSNBC devoted 32 percent of all coverage to the story, while CNN devoted 26 percent and Fox News 19 percent (Figure 1).
Furthermore, MSNBC covered the scandal for ten minutes or more on 82 percent of weekday evenings during the same time period. Even Fox News viewers, a group one might assume would be systematically sheltered from a controversy detrimental to the Republican president, could expect at least 10 minutes of coverage across 62 percent of evenings.
The opportunity cost
Why might it be problematic for cable news to cycle one top story over and over for years? One possible answer is that such emphasis on a single story might displace coverage of other issues that matter to voters. In a Gallup poll conducted before last year’s midterms, voters collectively ranked the “investigation into Russian involvement in the 2016 election” below 11 other issues in importance. Even when split by party identification, Democrats ranked healthcare, gun policy, climate change, immigration, and four other issues above the investigation.
Cable news channels systematically favored the Trump-Russia collusion scandal for nearly three years, clogging the information pipeline citizens depend on to ground their civic participation. While it is hard to grasp the breadth of important under-reported stories at a macro scale, zooming in on any given day or week reveals some of the valuable stories missed by cable news (Figure 2).
For instance, on September 21, 2017, Hurricane Maria had just devastated Puerto Rico with catastrophic flooding and power loss; the Trump Administration had just finalized its decision to withdraw from the Iran nuclear agreement; the Senate faced an 11th-hour bid to repeal Obamacare; and the death toll from a massive earthquake in Mexico rose to 250 people. And yet, that night, CNN spent 20 percent of its airtime on the Mueller scandal, while Fox News spent 29 percent, and MSNBC spent 45 percent. Sean Hannity , who hosts the most-watched cable news show in the country , never once mentioned the humanitarian crisis facing millions of Americans in Puerto Rico.
Do people care?
Together with PredictWise, a public opinion modeling firm, we conducted two surveys this past March and April to test Americans’ attitudes toward and knowledge of the Trump-Russia collusion scandal. Based on survey responses, we estimate that 43 percent of likely voters in the United States viewed the Trump-Russia story as either an “extremely” or “very” important issue, with a significant partisan skew: Democrats were almost twice as likely to call the topic an “extremely” or “very” important issue, in comparison with the whole population, and Republicans were twice as likely to report that the scandal was “not at all an important issue.”
In the two weeks leading up to, and between, our two surveys, cable networks dedicated 36 percent of primetime news airtime to Trump-Russia coverage, which portrayed the release of the Mueller report as the grand conclusion to years of reporting. But when asked how much their likelihood to vote for Trump in 2020 is dependent on the ongoing scandal, 53 percent of respondents said “not at all.” When we asked the same question two weeks later — following the end of the investigation, and the release of both Attorney General Barr’s controversial summary and the redacted report — the percentage increased to 54 percent. Other surveys echoed our findings.
Are people informed?
Perhaps the most surprising result from the surveys was the inability of respondents to correctly answer questions about the facts of the scandal. In our first PredictWise survey, we posed the question, “Has the Mueller Probe led to anyone from Trump’s campaign being found guilty of ANY federal crime?” At the time of the survey, six members of Trump’s team had been convicted as a result of the Mueller investigation, but 72 percent of those who identified Fox News as their primary news source either answered incorrectly or marked “I don’t know,” while 52 percent of CNN viewers and 35 percent of MSNBC viewers failed to correctly answer that “yes, more than one person” had been found guilty.
We also asked respondents if there were any formal agreements between the Trump campaign and Russia to coordinate efforts in the 2016 election. To date, no evidence has been presented that any such agreements were made. Yet 14 percent of Fox News viewers, 35 percent of CNN viewers, and 43 percent of MSNBC viewers were under the impression that Donald Trump had a formal agreement in place with Russian actors to sway election results in his favor. On these fundamental aspects of this widely told story, significant portions of the national audience were confused. Those numbers seem to reflect a paradox: more coverage does not lead to heightened public understanding.
What did they talk about?
Our second layer of analysis examined the actual content of the coverage. Enlisting another dMetrics machine-learning model, we sorted each of the news segments identified as being about the Trump-Russia collusion scandal into narrative clusters. Plotting out these narratives over time reveals how they are born, die, and are resurrected—for example, in the shadow of the Mueller Report release (Figure 3).
The three pro-Trump counter-narratives (labeled red in Figure 3) stand out from the others. Each frames the Special Counsel’s investigation as a fraudulent, politically motivated scandal; they were aired almost exclusively on Fox News (see Figure 4, below). The consistency over time of the “No Collusion!” narrative, in particular—an alleged conspiracy, espoused by anchors, involving Hillary Clinton and a Russian Oil Company—suggests a messaging strategy that rejects the entire scandal, as opposed to rehashing or contesting the latest dramatic twist.
The other narratives, mostly aired on MSNBC and CNN (Figure 4), range from numerous theories of Trump’s Russia connection to analyses of the Mueller investigation itself. A close reading of their coverage reveals a style of reporting frequently steeped in drama and speculation. For example, the “Will Mueller subpoena President Trump?” narrative first arose in December 2017, was sustained for months on both MSNBC and CNN, and peaked in April 2018, when the New York Times obtained a copy of the questions Team Mueller wanted to ask President Trump. Primetime shows Anderson Cooper 360° and CNN Tonight with Don Lemon devoted a combined 65 percent of airtime to the leak; Rachel Maddow suggested she’d scrapped plans for one show in light of a Washington Post report that Mueller floated the possibility of a presidential subpoena in a meeting with Trump’s team three months earlier. Mueller never did subpoena Trump. One media narrative was superseded by another, and coverage of the great Trump-Russia collusion scandal churned on.
This project was undertaken as a part of Harmony Labs’ ongoing narratives project to examine how news media frame events of national interest in the U.S. The analysis was done in conjunction with dMetrics, an NLP insights firm, and PredictWise, a public opinion research firm. The cable news closed captioning transcripts were sourced from Internet Archive. For details of the methodology used by Harmony Labs to analyze cable news coverage and to survey public opinion, as well as the modeled results of the public opinion survey, please see this methods supplement.Baird Howland, Andrew Pasquier, and Brian Waniewski are the authors. Howland is a Data Scientist at Harmony Labs focused on the influence of news and entertainment media. He also supports Harmony’s Project Ratio research group, which is providing a cross-platform mapping of news content and exposure. Pasquier is a New York-based freelance writer, researcher, and global program lead at the Urban Design Forum, a non-profit gathering designers, developers, and civic leaders to debate issues facing the contemporary city. Waniewski is Executive Director at Harmony Labs. Their work focuses on understanding media influence at scale, and experimenting with using media systems to support healthy democratic culture and healthy happy people.