Consuming content from Fox News is associated with decreased knowledge of science and society
2020/7/31 ©PsyPost


People who visit FoxNews.com are no more or less likely to understand how the U.S. political system works compared to those who use other online media sources, according to a new study published in American Politics Research. But the study provides some initial evidence that using Fox’s website is associated with reduced knowledge about several politically-relevant topics, such as the state of the economy.

“A few years back, there was a poll that suggested Fox News viewers were less knowledgeable on important issues than people who abstained from watching the news at all,” said study author Peter R. Licari, a PhD candidate at the University of Florida.

“I kept seeing people referencing it on social media but noticed that: 1) It was only about TV news when a lot of news nowadays is consumed over the Internet; and 2) The analysis didn’t control for factors that might influence both Fox News viewership and lower political knowledge.”

“I was originally pretty skeptical of the claim — and I came into the research with that skepticism in hand. For one, there are many different kinds of ‘political knowledge,’ so an all-encompassing effect suggested by headlines and people’s conversational references to the study seemed too over-the-top — and news consumption tends to be positively correlated with political knowledge in general,” Licari explained.

“I realized I knew of a dataset where I could address these issues. I wanted to ‘productively procrastinate’ on my dissertation looking at how video games can increase civic attitudes and political participation, so I dug in.”

For the study, Licari analyzed nationally representative data from 2016 wave of the American National Election Studies. The surveys asked respondents to report their use of online news media sources, including ABC News, BuzzFeed, CNN, Fox News, The Huffington Post, NBC News, The Washington Post, and other outlets.

After controlling for age, gender, race, education, political ideology and other factors, Licari found that using Fox’s website as a news source was associated with lower levels of some types of political knowledge, but not others.

“For some kinds of knowledge, there’s little reason to suspect an association between getting one’s news from FoxNews.com and less knowledge (what I term ‘process-oriented knowledge’ in the paper — or facts about the political system and how things are run),” Licari told PsyPost.

Process-oriented knowledge included things like knowing which party currently had the most members in the U.S. House of Representatives, how many years a United States Senator is elected for, and the political office held by John Roberts.

“But, for other kinds of knowledge (what I call ‘society-oriented knowledge’ comprising important facts about society and things affecting society), the evidence does point to a possible association between visiting the site and lower levels of knowledge,” Licari explained.

Examples of society-oriented knowledge included whether unemployment had gotten better or worse over the past year, whether Barack Obama was a Muslim, and whether the world’s temperature had been slowly increasing over the past 100 years.

“Taking-in additional online sources other than FoxNew.com diminished the negative effect, but the models suggest that, for the vast majority of people, the society-oriented knowledge gap between visitors and non-visitors is pretty durable. That association is statistically and — I feel — substantively significant,” Licari said.

But as with all research, the study includes some limitations.

“First and foremost, this study could only measure statistical associations; it’s not designed to make a causal argument and people should be cautious in interpreting it as such. I included as many controls as I could sensibly do to try and limit the possibility of there being a spurious factor lurking there (such as Party ID, ideology, age, education, 2012 presidential vote, the number of conservative media sources respondents took-in overall, to name a few), but there could always be other unmeasured factors out there in the ether driving both variables,” Licari said.

“It could also be that the causal arrow is reversed. I don’t personally believe it is, but we have to be open to that possibility until other work better-positioned from a causal inference standpoint comes along and settles the issue.”

“I also wish that there were items in the dataset that could make the society-oriented knowledge scale less partisan. There are some things there (like climate change) with a definite partisan lean when it comes to getting the ‘right answer.’ That lean will undoubtedly inject some bias into the measure and into the conclusions. Although, the effects are robust even when those items are nixed from the scale and it just becomes about economic knowledge — so I think that there’s something interesting there that’s still worth following-up on and pursuing,” Licari added.

The researcher hopes his work will inspire additional investigations into the relationship between news media consumption and political knowledge.

“I think that looking at possible informational biases associated with a mainstream news outlet is important work. (And regardless of what Fox’s pundits and uber-fans say — they’re mainstream. You can either be the nation’s most viewed news network or be not-mainstream; can’t have it both ways!) I’d like to see future work come and build off the limitations I know to be there as well as those that undoubtedly are there that I’m not aware of,” he told PsyPost.

“I think, given how many pundits on the network were (and are) skeptical of the COVID-19 outbreak, there might be an understandable temptation to link this paper with the skepticism that is more common among the US right. I’d personally suggest not doing so,” Licari added.

“Instead, I’d recommend Matt Motta, Dominik Stecula, and Christina Farhart’s interesting paper “How Right-Leaning Media Coverage of COVID-19 Facilitated the Spread of Misinformation in the Early Stages of the Pandemic in the U.S.” They get at the topic directly and in a way with far better causal leverage than I’m able to get at here.”TrendMD v2.