Nonprofit, nonpartisan journalism. Supported by readers.


Why partisan bias alters our view of facts

But here’s the good news: For a tiny bribe, the partisan distortion disappears.

The U.S. unemployment rate from 2008-2015. In a recent Washington Post poll, by 76-18, Democrats said that unemployment had gone down. By 53-38, Republicans said it had gone up.
Bureau of Labor Statistics

One of the struggles within the struggle to believe in the benefits of democracy is the struggle to believe that the thinking of average voters is rooted in some kind of coherent factual or logical reality, something more rational than tribal loyalty (whether we define “tribal” according to race or just according to partisanship).

Of course, every citizen has a right to base his vote (or the decision not to vote) on any reason that satisfies him. One of the most common types is a person who always votes, and always votes for the same party, and often it is the same party for whom his parents voted. This could be rational, but it is more likely akin to a quasi-religious quality, just as most people follow their family’s religion, usually without giving serious consideration to other faiths.

Changes in the media environment over recent decades make it easier for news consumers to pay attention only to TV, radio, print and online sources that confirm one’s existing partisan/political biases. This is sometimes called the “Fox Effect,” a term certainly coined by a liberal. I’m sure it affects plenty of people on both sides of the spectrum and I’m sure unthinking partisan voting is also common among the many Americans who pay little interest to news or politics.

The syndrome that liberals sometimes call “Obama Derangement Syndrome” refers to the semi-joking theory that the presence of Barack Obama in the Oval Office has driven many conservatives so crazy that they cling to impossible beliefs, contradicted by all evidence. The idea that Obama is a secret Muslim is perhaps the craziest one but, according to this recent poll, 43 percent of Republicans said it was so.

Article continues after advertisement

More typical of such dichotomies, when asked by pollsters whether the economy has improved during the Obama years, the responses track closely along the partisan loyalty of the person being asked. Here’s a recent example from the Washington Post. If you didn’t click through, the question was whether the unemployment rate had gone up or down during the Obama presidency. By 76-18, Democrats said — accurately — that unemployment had gone down. By 53-38, Republicans said it had gone up.

“Unemployment” can be measured several ways. The official measure of the unemployment rate counts as “unemployed” someone who is looking for a job but doesn’t have one. If you’re not looking, you’re not “unemployed” by that traditional measure. The number who have stopped looking includes some who are called “discouraged workers,” meaning they have given up looking for work.

The rate of “discouraged workers” has risen during the Obama years. It’s certainly not crazy to suggest that this increase in discouraged workers is part of Obama’s economic record, and not a good part. But the idea that Republicans are more attuned and Democrats less attuned to the problem of discouraged workers makes no sense. To the degree that they are aware of it, Republicans are obviously more (and Democrats less) motivated to notice the rise of discouraged workers under Obama.

(By the way: This whole partisan pattern occurred during the Reagan years as well. Only in those years it was Democrats who were far more likely to say the economy had gone downhill, when in fact by most normal measures it had gotten significantly better.)

But I was quite struck by a scholarly study that economics reporter/analyst Neil Irwin highlighted in a recent piece for The New York Times’ Upshot blog, which suggested that at some level people know when they give these answers out of partisan bias that they are giving wrong answers. Here’s Irwin’s version of the setup, without even referencing the Obama syndrome:

“Did unemployment get better or worse during Ronald Reagan’s presidency? In a 1988 survey, some 80 percent of dedicated Republicans accurately said it had improved, compared with 30 percent of loyal Democrats. In the 1990s, the pattern reversed on a range of factual questions about economic and fiscal issues. In a 1997 survey, for example, Republicans were far less likely than Democrats to acknowledge that the budget deficit had declined during the Bill Clinton administration.”

But political scientists apparently wondered the same thing I wondered: When people give wrong, partisan answers to factual questions, do they believe what they are saying or are they somewhat knowingly engaging in partisan spin?

Small bribes

So two teams of scholars one led by John Bullock of the University of Texas, the other by lead author Markus Prior of Princeton, tried a wrinkle. Before asking the question, they told the respondents that they would receive one dollar if their answer was accurate. And, in one version of this experiment, they told people that if they weren’t sure of the correct answer, they would be paid 33 cents for responding “I don’t know” rather than guessing and getting the answer wrong.

The results were funny or sad, depending on how you look at them. When offered these tiny bribes for either giving the right answer or admitting they didn’t know, the normal partisan gap in the responses basically disappeared.

Article continues after advertisement

I asked University of Minnesota professor Howie Lavine, who specializes in political psychology, what he made of this. He was aware of the research and said that people who answer survey questions in a way that favors their preferred party aren’t lying to help their party in the realm of public opinion as much as they are “gratifying their partisan identity.”

“Think of it as an expressive act that gives them a psychological benefit,” Lavine said. “They feel good expressing a belief that puts their partisan identity in a positive light. It’s a self-esteem boost. When your group is doing well, you feel better about yourself.”

The question might be: Did unemployment go up or down during the Obama presidency? But, Lavine said, “It’s almost as if they’re answering a different question, something like: Do I like Obama or not? Am I a Republican or not?”

Lavine added: “People are increasingly shuttering themselves within partisan echo chambers. They’re told over and over again that everything the other party does is bad. This tendency has been getting stronger. It got worse from the first President Bush to Clinton, worse during the second Bush and worse again from Bush to Obama, where it also got caught up with race.”