Confirmation Bias

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 9.9

The tendency towards certainty and lack of openness to alternative interpretations of events, which is especially prevalent among elected representatives, builds on a psychological propensity that we all have — confirmation bias.  This refers to the well-documented and powerful inclination people have to notice, and believe casually observed evidence that tends to confirm their current beliefs, and to discount or completely ignore information that would tend to disprove their current beliefs. Accepting contradictory information risks creating the unpleasant feeling of what psychologists call cognitive dissonance. 

The claim is not that this always occurs — certainly we are all aware of times we have heard new information that prompted us to change our opinion about some issue — but confirmation bias is far more common than we realize, simply because by its very nature we can’t notice it when we are doing it. This tendency is thought to be exacerbated by what is sometimes called the “silo effect,” where people tend to select news sources that corroborate pre-existing opinions (whether by choice or Internet algorithm). In the US, conservatives may prefer to get their news from Fox News and liberals from MSNBC. But even impartial, neutral information is often interpreted differently depending on one’s prior beliefs, associated with what is known as cultural cognition. For example, depending on the prior stance of you and your friends or political leaders on gun regulation, a mass shooting in an elementary school may reinforce a belief for one person in the need to restrict gun access, but also reinforce a belief of a different person in the need for more guns in schools in the hands of “good guys” to thwart such shootings.

Breaking down the information silos is widely advocated, but in a combative electoral environment, it may not reduce polarization. Indeed, once an individual has taken on a partisan identity, or come to identify more with one side than another, exposure to opposing points of view may actually intensify polarization — what is referred to as a “backfire” or  “backlash” effect. A 2022 analysis by Philipp Lorenz-Spreen, and others that looked at many separate studies found that

“increased polarization was also linked to exposure to opposing viewpoints in one's social media feeds. In other words, being exposed to the words of political opponents did not bridge the political divide. Rather it seemed to amplify it.” [emphasis added]

In competitive situations with winners and losers, as when officials selected through partisan elections “debate” each other, there is a strong tendency towards what psychologists call “motivated reasoning.” Winning a debate requires seeking out convincing arguments for why your side is correct, and the other side is wrong. The participants are generally unaware of this distortion of their own thought process. In these partisan situations, winning feels more important than finding the truth, so even the most brilliant minds will direct their abilities to this imperative. Partisans frequently feel that their side winning is ipso facto synonymous with the truth winning. Rummaging through a vast amount of information and potential lines of reasoning to astutely come up with the most convincing case for your side may simply further bury the truth. As a result, smart people (as well as stupid people) make stupid decisions all the time.

However, researchers who have carefully reexamined and attempted to replicate earlier experiments that showed this backlash effect, have called it into question for ordinary people — at least in certain situations. When representative samples of people were given evidence on such things as whether capital punishment deters crime, or whether the minimum wage should be raised, rather than doubling down on prior beliefs (backlash), their beliefs tended to moderate in the direction of the newly presented evidence. This is hopeful for the prospects of citizens’ assemblies. This may reflect the trust in the impartiality of the messenger. The researchers noted that 

“A reasonable objection to these findings is that while individuals may not exhibit backlash when reading relatively sterile descriptions of academic studies, they may do so when arguing about a particular proposition with an opponent…. Perhaps in such antagonistic contexts, individuals become distrustful of counter-attitudinal arguments.”  

The backlash effect may be an artifact of the competitive partisan election environment. A 2023 research paper found that truthful fact-checking was more likely to backfire when the source was a political out-group member, and such fact-checking often left people with more entrenched beliefs in misinformation.

In a journal article entitled, “Identity Concerns Drive Belief: The Impact of Partisan Identity on the Belief and Dissemination of True and False News,” other researchers have published evidence that rather than standard confirmation bias, in the partisan political realm, belief in false information may be a matter of upholding group identity. It is gratifying to believe positive things about one’s own political in-group members (and their beliefs) and negative things about political out-group members. Rejecting one’s identity group’s false beliefs is psychologically painful. Politicians seeking to drive voter turnout benefit from promoting partisan loyalty as a central element of personal identity. Electoral politics undercuts efforts to base public policies on facts. Removing deliberators from this partisan in-group/out-group environment by placing them together on the same team (a citizens’ assembly) where they can come to know each other on an individual basis can sidestep this dynamic. This adds credence to the notion that citizens’ assemblies made up of everyday people, rather than politicians or extreme partisans, using good process, could avoid many of the common political psychological pitfalls.

Elected representatives also have an additional distorting incentive, which often locks them into preconceived positions. They won election by asserting a particular set of priorities and view of reality. Any change, perhaps based on new information, leaves them open to embarrassing charges of selling out, hypocrisy or flip-flopping. Bargaining and horse-trading with political opponents in an elected legislative body may be acceptable, but actually deliberating and coming to new understandings is not. Unelected citizens randomly selected don’t have this dilemma. It is a bad idea to have those going into a deliberative situation having their conclusions firmly locked in place beforehand. This is a common scenario for bargaining, but genuine deliberation requires an openness to changing one’s mind. 

As discussed in previous a post (Voter Decisions, Chapter 8.5), studies have shown that people who follow the news, and have clear political opinions on some matter are less able to absorb and use new information if it seems to contradict their current beliefs. In other words, the goblin of confirmation bias hobbles the “well-informed” more than the less informed, who remain more open to learning from a variety of sources when placed in a deliberative setting. Though now scientifically supported, keen observers have long noted that those who believe themselves to be well-informed are generally not. While serving as president, Thomas Jefferson wrote in private correspondence, 

“The man who never looks into a newspaper is better informed than he who reads them; inasmuch as he who knows nothing is nearer to truth than he whose mind is filled with falsehoods & errors.”

Since a randomly selected body would have far fewer die-hard partisans deeply invested in their team having the right answers, sortitional bodies would allow for more competent learning and weighing of competing arguments than elected bodies. If partisan elections are also common in a society for other positions, this will tend to poison the pool to some extent, so an optimal democracy would dispense with mass elections altogether (though this would only be possible after sortition had proved its superiority).