Despite what many
social psychologists claim, humans are mostly rational. When we are thirsty, we drink water. When're hungry, we eat food. When we're on the edge of cliff about to fall
off, we look for ways to save ourselves.
For the large majority of decisions we make everyday, humans typically
take actions that will satisfy our goals.
In certain circumstances, however, we are systematically biased. Take, for instance, the domain of
politics. It's understandable for there
to be disputes over values: privacy vs security, economic freedom vs equality,
etc. What's less understandable,
however, is the fact that people get heated up over facts. Why do liberals believe that humans
are the cause of climate change while conservatives don't? Why do conservatives believe that gun control
would increase crime while liberals believe the opposite? Liberals will say it's because conservatives
are biased, and conservatives will say it's because liberals are biased. Who's correct?
To begin, let's
catalogue the factors that may lead to ignorance of the empirical facts. There are at least three. One factor is that humans aren't always too
thoughtful or deliberate. Instead, we
use heuristics, mental shortcuts for arriving at a desired outcome. For instance, we rely on scientific experts
to tell us the truth rather than seeking it out for ourselves. Our choice of experts, however, is also often the result of nondeliberative
thinking. This unfortunate quirk of
human decision making may lead people to become uninformed or misinformed. A second factor is motivated reasoning. That is, even
if people engage their more reflective cognitive processes, they may do
so in a way that steers them away from the truth. When a person is motivated to maintain a
relationship or preserve their identity, they may selectively interpret the
evidence to suit their non-truth-attainment goals. Finally, a person may have a certain kind of
reasoning style that interferes with the attainment of truth. There is a considerable body of research, for
instance, that conservativism is associated with dogmatism, an aversion to
complexity, and a craving for closure in an argument. These cognitive traits may hinder a person in
their pursuit of truth.
It's unclear,
however, how these three factors interact to generate belief polarization. In his article, Dan Kahan outlines three
possibilities. First is what he calls
the "Bounded Rationality Position" (BRP). According to BRP, our heuristic-driven
reasoning is the most decisive factor in generating public discord over
empirical matters. On this view,
laypeople inadequately engage in effortful information processing. As a heuristic, then, these nondeliberative
folk will tend to trust the received wisdom of their particular in-group, which
in turn will lead to greater belief polarization.
A second alternative is what Kahan calls the "Ideological Asymmetry
Position" (IAP). IAP posits that
right-wing ideology matters most in distorting empirical judgments. Like BRP, IAP takes reasoning to be
heuristic-driven and inadequately engaged.
This is said to be especially true of conservatives in light of previous
correlative research on their cognitive traits.
Because liberalism is associated with, among other things,
open-mindedness, it might be thought that they would be less vulnerable to the
siren song of political bias. The final
account Kahan considers is what he calls the "Expressive Utility
Position" (EUP). This position,
unlike both BRP and IAP, posits that motivated reasoning is the most important
factor in belief polarization. A
person's primary motivation when looking at data under this view is to protect
their identity, and they will do so by means of selectively searching for and
interpreting the evidence to fit with their particular in-group. Reasoning, then, is not inadequately
engaged. Far from it. Reasoning will tend to magnify ideological differences, not mitigate
it, and this will be true across the political spectrum.
So there's the
theory; now where's the test? In the
first part of his study, Kahan presented participants with the Cognitive
Reflection Test (CRT). This test is
generally used to measure a person's disposition to engage in conscious and
effortful information processing as opposed to heuristic-driven
processing. This quick three item test
can be found here. Following the CRT,
participants were split into three conditions.
In one condition, participants were told the following:
Psychologists
believe the questions you have just answered measure how reflective and
open-minded someone is.
The second condition
tacked on an additional bit of information:
In
one recent study, a researcher found that people who accept evidence of climate change tend to
get more answers correct than those who reject evidence of climate change […] a finding that
would imply that those who believe climate change is happening are more
open-minded than those who are skeptical that climate change is happening.
The third condition
replaced the above paragraph with this one:
In
one recent study, a research found that people who reject evidence of climate change tend to
get more answers correct than those who accept evidence of climate change […] a finding that
would imply that those who are skeptical that climate change is happening are
more open-minded than those who believe climate change is happening.
Participants were
then asked how valid they personally believed the CRT was in assessing how
reflective and open-minded they were.
Because open-mindedness is almost universally considered a positive
trait, participants have an emotional stake in believing their group to be more
(or at least not less) open-minded than their ideological opponents. Hence, if a liberal were biased in favor of
liberalism, they would discount the CRT's validity in the third condition while
accepting the CRT's validity in the second.
The opposite would be true for conservatives; they would discount its
validity in the second condition and accept it in the third. All the theories above predict some motivated
reasoning, but they each generate different hypotheses about the form that such
reasoning would take.
- IAP predicts that
motivated reasoning should be especially pronounced among conservatives. Liberals, on the other hand, should have
roughly similar validity ratings regardless of condition. Furthermore, conservatives should score worse
than liberals on the CRT itself given that the CRT actually does measure
cognitive reflection.
- BRP predicts that
people who score poorly on the CRT will be more inclined to express polarized
sentiments. That is, low cognitive
reflection will lead to more bias. This
result, however, should hold regardless of one's political affiliation. Political affiliation should also be
unrelated to actual CRT scores.
- EUP is the reverse
of BRP. It predicts that polarization
will increase as CRT score
increases. That is, contrary to BRP, greater cognitive reflection will lead to more
bias. Similar to BRP, however, and
unlike IAP, EUP is neutral on political affiliation.
Before moving on to
the results, take a moment to guess which theory you think was best supported.
The first step of
data analysis was to compare liberals and conservatives on CRT scores. Contrary to IAP, conservatives actually
scored significantly better than
liberals on the CRT, a surprising result given the literature on conservatives'
other cognitive traits. The results of
the second step of data analysis were also contrary to IAP's predictions. Both liberals and conservatives split in
their perceptions of CRT validity depending on which condition they were
in. Conservatives believed the CRT was
valid when it favored their own side but not so when it didn't. The same pattern was evident among liberals
as well. So much for IAP then…
Next up on the
chopping block is… *drum roll* … BRP! As
you'll recall, BRP predicts that as people's CRT scores increase, they should
exhibit less ideological bias. Cognitive
reflection, on this view, should have a salutary effect on polarization. However, this was not borne out by the
data. Indeed, as people scored higher on
the CRT, they exhibited more and more partisanship. In particular, liberals who scored high on the CRT were remarkably more averse to accepting the CRT's
validity when in the third condition (in which climate change skeptics appeared
more open-minded). Additionally,
conservatives who scored highly on the CRT welcomed it with open arms when in
the third condition. In conjunction,
these findings jive more convivially with the predictions of EUP; greater
cognitive reflection leads to greater political bias for both liberals and
conservatives.
In my previous post,
I discussed Jonathan Haidt's Social Intuitionist Model, a model of moral
judgment that describes humans in a less than favorable light. Humans, it is said, arrive at their moral
judgments through automatic intuitive processes and later use their powers of
reason, not to correct potential errors, but rather to rationalize all those
errors away. Like a magician waving his
wand, the motivated reasoner can make any problem disappear. Of course, the problems are still there. They're just hiding in a hat now instead of
out in the open. Kahan's research
supports this interpretation of moral judgment.
In Kahan's research, those who were most reflective were also the best
magicians; they were more able to twist their judgments to fit the narratives
of their in-groups. This picture, though
grim and pessimistic, is the one we must look to if we hope to paint a brighter
future for humanity. Kahan's research
suggests that we need to not only nudge people's intuitive judgments towards
the truth as many others have suggested.
This will only do so much. Given
that a large portion of our bias comes from our more reflective moments, we
have to also remove the incentives people face for forming beliefs on grounds
unconnected to the truth of such beliefs.
Indeed, in a sense, it's rational for, say, a conservative to believe that
gun deregulation will decrease crime.
Not because this belief is particularly supported by evidence but rather
because he will be expressing his commitment to his group, and his group will
in turn support him. It's rational, too,
for a liberal to believe genetically modified foods are devil spawns, again,
not because it's supported by evidence, but because by believing such things,
they will be boosting their status within their group. Removing or circumventing these cultural associations will be the work of future researchers.
Of course, this polarization is
not true of every topic. In fact, most
topics don't exhibit such polarization.
(e.g. that the Moon revolves around the Earth, that earthquakes result
from shifting tectonic plates, that height is hereditary, etc.) Like I said at the outset of this post,
people are mostly rational. But for
those domains where we have some trouble, like politics, more research is
needed to figure out exactly how we get things wrong and how we can fix our
mistakes.
Frederick, S. (2005). Cognitive reflection and decision making. Journal of Economic perspectives, 25-42.
