The independent news organization of Duke University

President Price discusses increasing partisanship of media, role of public opinion polls

President Vincent Price drew on his experience with public opinion and the media to give students some insight into modern polling and news organizations Thursday evening.

In an event sponsored by Duke Student Government, Price discussed his expertise in studying public opinion and connected this to changes in public opinion polling over the years. He also emphasized the shift in media audience over the past few decades from centrist news outlets to increasingly partisan media audiences. And no discussion on modern media is complete without the mention of "fake news."

“[Fake news is] not a new trend, nor is deep political division nor is a hyper-partisan media system,” he said. “The Founding Fathers were not exactly cool or rational actors. They were constantly accusing each other of the most heinous acts—some of it accurate, as it turns out, but a lot of it fake news.”

The modern-day problem, Price added, is centered around the lack of a trusted institution to declare certain items false and others true. Even universities have fallen victim to being written off as propagators of liberal ideas, he added. 

Some of this mistrust may come from the label of universities as "elitist," and Price said that focusing much more locally and regionally is important in regaining more public support.

"I don't think institutions like Duke have actually been as clearly invested in the interests of their they should be," he said. "It's not new—Duke has been doing this, but I think we need to step it up."

Price explained that universities and the media aren’t the only groups whose legitimacy is being questioned, but that this “loss of confidence in public institutions” is emerging as a common theme.

He was not always interested in studying public opinion—in fact, Price's senior thesis was a teleplay, and his first job was working in undergraduate admissions, where some of his duties included survey research.

Interested in improving his knowledge on the intersection of media and statistics, Price began graduate school at the University of Michigan. He began by studying how television impacted children but soon had a change of heart and “fell in love” with the study of political communication.

He recalled that in his graduate school days in the 1980s, a major critique of the U.S. media was that it was too centrist and cast more marginal views to the wayside in favor of appealing to a broad audience. Many specialists in the subject advocated for a more partisan media that would present these more extreme political viewpoints, Price explained.

This dream of a hyper-partisan media environment, he admitted, has finally been realized and is currently the landscape in which America finds itself today.

“The interesting phenomenon now is that more news is consumed than ever before in aggregate, but it’s being consumed by a community of news junkies,” he said. “If people really want to consume news, they can do this endlessly. If you want to avoid news, you can pretty much avoid it completely.”

For the latter group who chooses to shun the media, Price explained the problem lies in how these people become informed about the world around them. Rather than having their opinions anchored in a thoroughly researched argument, these citizens’ viewpoints can be easily manipulated by any opinion.

Price also discussed one of the most prevalent examples of public opinion—polling.

The percentage of people who respond to polls has been declining, he said, largely because of increasingly rampant marketing surveys.

However, Price suggested that a lower response rate might not be skewing the data too much.

“The success rate of probability sampling isn’t what it’s cracked up to be,” he said.

Price described a study in which researchers sampled a proportion of the population and received a response rate of around 20 percent. After applying the same sampling frame to subsequent attempts, these researchers drove the response rate up to higher and higher percentages. He explained that despite the large disparity in response rates, there was little difference in the data.

He also touched on the polling in the 2016 presidential election, explaining that the polls were not as big of a failure as some make them out to be. The primary errors in polling did not come from national polls but rather state polls with larger sampling error, Price explained.

“What was problematic is that [the state polls] were systematically wrong,” he said. “That is, they all got it wrong in the same direction—they all understated Trump support.”

For polling aggregators, this means that the error differences will not average out because the skew is consistently in the same direction, the president said. 

Junior Matthew Keep, who attended the event, praised the connections Price drew between his former area of expertise and current position.

"It was fascinating to hear President Price talk about his area of specialization and apply his expertise to the greatest challenges facing universities today," Keep said. "He spoke so knowledgeably about...public opinion, and I can imagine him being a terrific professor earlier in his career." 

Price cautioned against allowing polls to influence people’s decision to vote in an election.

“If we know a candidate we support in a particular election, it’s not clear what the value is of caring how far ahead or behind that candidate is,” he said. “You just go out and vote for that candidate.”


Share and discuss “President Price discusses increasing partisanship of media, role of public opinion polls” on social media.