Who Are More Biased: Liberals or Conservatives?

Courtesy of Flickr

Recently Jane Mayer (2019) of The New Yorker wrote about the very close relationship between Fox News and President Donald Trump, outlining in great detail how the Murdoch family–owned news outlet functions as a tacit communication operation of the Trump White House and its political agenda. There is a revolving door of former Fox News staff working in the White House and former White House staff working at Fox News, and according to many sources quoted in the article, Fox News has erased the lines of journalistic integrity and is now functioning as a Trump propaganda machine. Mayer argued that with respect to programming, it was often unclear whether Trump controlled Fox News or Fox News controlled Trump.

Meanwhile, the president himself continues to decry the “Fake News Networks,” by which he seems to mean any outlet to the left of Fox News. According to anonymous sources cited in Mayer’s article, President Trump had such animus toward CNN that he attempted to block AT&T’s purchase of Time-Warner, the parent company of CNN. The Trump Administration’s Department of Justice sued to block the merger, but in June a district court judge approved the deal (Ivanova 2018).

Conservatives who, like the President, believe the mainstream media (MSM) have a liberal bias have been supported by evidence that journalists are more likely to identify as Democrats than Republicans. A 1990 poll of forty-nine editors and writers at The Washington Post found only one person, Tony Kornheiser, a sports columnist, was a registered Republican, and he suggested he was a RINO (Republican in name only): “I don’t think the Republican Party would claim me” (Wemple 2017). A 2014 study conducted by the University of Indiana School of Journalism found that 28 percent of journalists surveyed identified as Democrat and 7 percent as Republicans. Fifty percent said they were Independents (Willnat and Weaver 2014).

In recent years, Ad Fontes Media has produced a media bias chart for many of the most popular news sources (see below). The horizontal axis of the chart plots the left or right political bias of the source, and the vertical axis plots the factual nature of the reporting, ranging from “Contains inaccurate/fabricated info” at the bottom to “original fact reporting” at the top. The result is an inverted-U–shaped function with the most politically biased sources on both ends of the spectrum rating low on factual material (e.g., Occupy Democrats, Breitbart News), and the most politically neutral sources tending to be the highest in factual material. For example, the Associated Press, Reuters, and Bloomberg were rated both among the highest in original fact reporting and the lowest in political slant. There are a few media sources that occupy the bottom middle of the chart. For example, the National Enquirer is given only a modest right-ward skew but is also very low on factual information.

Figure 1. The 2018 version of the media bias chart by Ad Fontes Media, Inc. The vertical axis indicates level of fact-based reporting, and the horizontal axis plots liberal (left) versus conservative (right) bias. (Used with permission.)

Biased Information Processing: Conservatives vs. Liberals

As the foregoing suggests, the media landscape is filled with politically biased information that spans a wide range of viewpoints, but how biased are the people who receive this information? When you ask Republicans or Democrats how biased people of each party are, you get predictable results: Democrats say Republicans are more biased and Republicans say the reverse (see Figure 2). But psychological science cannot rely on the opinions of interested parties for answers. What happens when you put people in an experimental setting where they have a chance to act in a biased or unbiased way?

Figure 2. Mean level of agreement that the term biased accurately describes the average Republican or Democrat. Based on the responses of 951 Republicans and Democrats from Ditto et al. (2019a)¹.

The study of political ideology has a long history in the field of psychology, and the division of U.S. politics into more hardened red and blue positions—which is also happening elsewhere in the world—has made this line of study more relevant. As a result, psychologists have learned a lot about what separates the two sides. John Jost and colleagues (2008), writing in Perspectives on Psychological Science, characterize the two sides this way: The right-wing label has come to represent political views that are conservative, supportive of the status quo, and hierarchical in nature, whereas left-wing views connote progressive social change and egalitarian ideals (127).

Jost and colleagues also found that conservatives were more likely to show “system-justifying attitudes,” including support for the idea that economic systems are fair and that equality is undesirable. In addition, research has shown that conservatives and liberals can be separated by their attitudes toward various groups and activities. Table 1 shows a selection of these drawn from Jost et al. (2008).

Table 1. Groups and activities that conservatives and liberals have more positive attitudes toward (from Jost et al. 2008).

Finally, recent psychological research suggests that these ideological differences are reflected in both explicit and implicit attitudes and are related to core personality dimensions (Jost et al. 2008). Currently, the most popular psychological theory of personality assumes that we possess stable traits that affect our behavior across time and environments. The most popular trait theory at the moment is The Big Five, which attempts to characterize people by measuring their levels of five traits: openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism (McCrae and Costa 1987). Research on political ideology has shown that liberals and conservatives can be distinguished by the degree to which they show two of the Big Five traits: openness to experience (higher among liberals) and conscientiousness (higher among conservatives) (Jost et al. 2008).

Given their relative openness to experience, it would be reasonable to think that liberals would be less biased than conservatives. When confronted with data that challenges their beliefs, liberals might be more willing to accept contrary information, whereas conservatives’ propensity to defend the status quo would make then resistant to new information. A recent meta-analysis by Ditto and colleagues (2019a) suggests otherwise. The researchers summarized studies in which participants were presented with information that contradicted their beliefs and found that liberals and conservatives were equally biased in their acceptance of this discrepant data. For example, Dan Kahan of Yale University conducted a study using a large sample of people who were scientifically matched to the demographics of the U.S. Participants reported their political orientations and were asked to take the Cognitive Reflection Test (CRT; Frederick 2005), which consists of the three questions below:

1. A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? _____ cents
2. If it takes 5 machines 5 minutes to make 5 widgets, how long would it take 100 machines to make 100 widgets? _____ minutes
3. In a lake, there is a patch of lily pads. Every day, the patch doubles in size. If it takes 48 days for the patch to cover the entire lake, how long would it take for the patch to cover half of the lake? _____ days

The CRT is a mentally challenging task that is thought to require thoughtful, deliberate reasoning if you hope answer the questions correctly. In the language of dual processing theory (Kahneman 2011), the CRT can only be mastered by engaging the slower but more powerful machinery of System 2 and cannot be solved using the simple intuitive heuristics of System 1. (By the way, the answers to the CRT are five cents, five minutes, and forty-seven days.)

In Kahan’s study, the discrepant information given to participants was about climate change and scores on the CRT. Some participants were told that research showed people who “accept evidence of climate change” tended to get more answers correct on the CRT (discrepant for conservatives). Others were told that people who “reject evidence of climate change” got more answers correct (discrepant for liberals). After receiving this information, participants were asked to rate their agreement with the view that the CRT “supplies good evidence of how reflective and open-minded someone is” (Kahan 2013, 412). Kahan found that both liberals and conservatives were biased in their evaluation of the CRT.When told it correlated with an opposing point of view—climate change deniers for liberals and climate change accepters for conservatives—each downgraded their rating of the validity of the CRT. Kahan’s (2013) study was one of fifty-one similar experiments summarized in the meta-analysis by Ditto and colleagues. Writing in the March 2019 issue of Perspectives on Psychological Science, Ditto et al. (2019a) suggest that Kahan’s findings are representative of similar studies that looked at the biases of liberals and conservatives. When confronted with information that challenges their beliefs, both sides show equivalent levels of bias.

Well, that should be a comforting thought. If we can’t get rid of bias, perhaps it is best that it be evenly distributed across both ends of the political spectrum. Not so fast.

The Ditto et al. (2019a) meta-analysis was challenged by a paper published in the same issue of Perspectives in Psychological Science by Jonathan Baron of the University of Pennsylvania and John Jost of New York University. On a general level, Baron and Jost (2019) pointed to a body of research showing that conservatives exhibit thinking styles that are substantially higher in dogmatism, rigidity, intolerance for ambiguity, and need for order; whereas, liberals show thinking styles higher in integrative complexity, cognitive reflection, and need for cognition, which involves having a positive attitude toward tasks that require reasoning. This pattern of thinking styles is inconsistent with symmetrical bias for conservatives and liberals.

Baron and Jost (2019) had some more technical criticisms of the Ditto et al. (2019a) article. For example, in deciding which studies to include in their meta-analysis, Ditto et al. started with 1,500 studies, which, after applying various inclusion criteria, they winnowed down to just fifty-one studies, and Baron and Jost suggested that the experimenters may have unconsciously chosen studies that were more likely to produce symmetrical bias. In other words, Baron and Jost accused Ditto and colleagues of being biased in the construction of their meta-analysis of bias. In a response, Ditto et al. (2019b) were quick to defend their methods of choosing studies for their analysis. But Baron and Jost made another criticism that—to me, at least—is worthy of serious consideration.

Baron and Jost (2019) point out that many of the issues used to assess partisan bias in the Ditto et al. study don’t really have two sides. For example, the Kahan study described above was about climate change. The scientific community overwhelmingly supports the idea that human activity contributes to global warming, and as a result, it is completely rational to be biased against information that contradicts that view. Baron and Jost argue that Ditto and colleagues have created a false equivalency. In many of the issues addressed in the meta-analysis, it would have been objectively appropriate to be skeptical. Baron and Jost made this point in a quote they used as an epigram for their article:

There can be no real quarrel with a willingness to infer that studies supporting one’s theory-based expectations are more probative than … studies that contradict one’s expectations. … Hence, [a] physicist would be “biased,” but appropriately so, if a new procedure for evaluating the speed of light were accepted if it gave the “right answer” but rejected if it gave the “wrong answer.” The same bias leads most of us to be skeptical about reports of miraculous virgin births or herbal cures for cancer. (Lord et al. 1979, 2106)

It is reasonable to be skeptical of information that contradicts your background knowledge on the subject—particularly if you know what you’re talking about.According to Baron and Jost, because the Ditto et al. meta-analysis did not measure or take into account the participants’ knowledge of the topics addressed, it is impossible to know how much of their bias was justifiable and rational and how much was shear politically based prejudice. On some topics, Republicans may have greater background knowledge than Democrats (e.g., economics), but as a general trend, current political history contradicts the proposition of symmetrical amounts of knowledge among the political parties. Research shows that “low-information voters” were an important component of Trump’s support. For example, a 2017 study found that voters who were low on need for cognition and low on political knowledge showed a greater preference for Trump over Clinton in the context of the 2016 presidential campaign (Fording and Schram 2017). If liberals in the Ditto et al. meta-analysis had more knowledge of the issues used to assess bias, then they may have been behaving more reasonably and justifiably than conservatives. But, at this point, we are not in a position to answer that question.

So, Who Is More Biased?

Perhaps the most honest answer is: We don’t know. The debate between Ditto and colleagues (2019a, 2019b) and Baron and Jost (2019) has revealed some sticky problems that need to be solved before we can make a clear judgment on the matter. Furthermore, the entire enterprise may be a bit misguided—just another opportunity for finger-pointing. Rather than focusing on the contentious and fraught issue of political bias, it might be more useful to highlight people’s relative allegiance to facts. The word bias sounds uniformly bad, but having a preference for certain kinds of information over others—based on the source and on the quality of one’s personal knowledge—is entirely appropriate. If not essential. In a world awash in misinformation (see Figure 1), it is important that we learn as much as possible about the issues of the day and think intelligently and critically about the claims that come our way. There is ignorance on all sides, and intelligence and critical thinking should not be a partisan matter.

References

• Baron, Jonathan, and John T. Jost. 2019. False equivalence: Are liberals and conservatives in the United States equally biased? Perspectives on Psychological Science 14(2): 292–303.
• Ditto, Peter H., Brittany S. Liu, Cory J. Clark, et al. 2019a. At least bias is bipartisan: A meta-analytic comparison of partisan bias in liberals and conservatives. Perspectives on Psychological Science 14(2): 273–291.
• Ditto, Peter H., Cory J. Clark, Brittany S. Liu, et al. 2019b. Partisan bias and its discontents. Perspectives on Psychological Science 14(2): 304–16.
• Fording, Richard C., and Sanford F. Schram. 2017. The cognitive and emotional sources of Trump support: The case of low-information voters. New Political Science 39(4): 670–86.
• Frederick, Shane. 2005. Cognitive reflection and decision making. Journal of Economic Perspectives 19(4): 25–42.
• Ivanova, Irina. 2018. Judge approves AT&T-Time Warner merger without conditions. CBS News (June 12). Available online at https://www.cbsnews.com/news/att-time-warner-merger-approved-without-conditions-judge-rules-today-time-warner-stock-price-rises-in-after-hours-trading/; accessed March 11, 2019.
• Jost, John T., Brian A. Nosek, and Samuel D. Gosling. 2008. Ideology: Its resurgence in social, personality, and political psychology. Perspectives on Psychological Science 3(2): 126–136.
• Kahan, Dan M. 2013. Ideology, motivated reasoning, and cognitive reflection. Judgment and Decision Making 8(4): 407–424. Available online at http://journal.sjdm.org/13/13313/jdm13313.pdf.
• Kahneman, Daniel. 2011. Thinking, Fast and Slow. New York: Farrar, Straus, & Giroux.
• Lord, Charles G., Lee Ross, and Mark R. Lepper. 1979. Biased assimilation and attitude polarization: The effects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology 37(11): 2098–2109.
• Mayer, Jane. 2019. Trump TV. The New Yorker (March 11). Available online at https://www.newyorker.com/magazine/2019/03/11/the-making-of-the-fox-news-white-house.
• McCrae, Robert R., and Paul T. Costa. 1987. Validation of the Five-Factor Model of personality across instruments and observers. Journal of Personality and Social Psychology 52(1): 81–90. Available online at http://www.psychometric-assessment.com/wp-content/uploads/2013/01/5FMPeerValidationCostaMcCrea.pdf.
• Wemple, Erik. 2017. Dear mainstream media: Why so liberal? The Washington Post (January 27). Available online at https://www.washingtonpost.com/blogs/erik-wemple/wp/2017/01/27/dear-mainstream-media-why-so-liberal; accessed March 9, 2019.
• Willnat, Lars, and David H. Weaver. 2014. The American Journalist in the Digital Age: Key Findings. Bloomington, IN: School of Journalism, Indiana University.

Note

1. Thanks to Peter Ditto for providing the data required to produce Figure 2.