Basic HTML Version

wonderful. The feedback loop can make you think that things are better than they are. In
economic cases, the difference between somebody being excessively positive about the
stock market and somebody being excessively negative is the difference between motivated
and fear-driven inference. Of course, this is happening not only at the individual level, but
also at the social level, with people sharing information and emotions to produce shared
Why do people engage in motivated inference? Why has natural selection not led
people to be more rational than they are? There is actually a neurological reason for this: the
brain did not evolve to do probability and utility very well. Probability theory was only
developed in the seventeenth century, and utility theory came much later. These are cultural
developments and they can often be useful tools, but they are not built into our brains.
There is a good deal of empirical evidence that in the brain there is no division
between cognition and emotion, between probability and utility. For instance, George
Loewenstein (2001) and others wrote a great paper called “Risk as Feeling” in which they
summarized the evidence for people estimating risk in terms of emotions. Another social
psychologist, Norbert Schwartz (1990), wrote a paper called “Feelings as Information,” in
which he said that we assess probabilities and utilities through our emotions. There is also
some related work in neuroscience by Sam Harris (2008), who is more famous for his
critiques of religion. He has found that the neural correlates of belief and disbelief are tied in
with different brain areas: when people agree with something, it is connected with the
positive emotion parts of the brain, whereas when people find something false, it is the
negative emotion parts of the brain that are involved. So the brain is using emotions for
these purposes and that can get us into trouble.
Unfortunately, there are not yet any psychological experiments documenting the
occurrence of fear-driven inference. Dave Nussbaum tried to do some experiments, but the
problem is that in order to do effective tests you have to make people really afraid, about
which there are obvious ethical concerns. However, anecdotally, rumination (thinking about
things over and over again) generates an amplified feedback loop in which the
representations feed on each other to become an internally reinforcing system that makes
you convinced that what you fear is actually true.
Fear-driven inference can produce a critical transition in a complex system in which
there is a bubble based on motivated inference. When something really big happens in the
economy, like the collapse of Lehman Brothers in 2008, an emotional transition takes place
toward fear-driven inference. This happens to individuals every day: you are in a good
mood and then you get some bad news that puts you into a bad mood. The emotional
transition leads to fear-driven inference, which is my attempt to explain economic booms
and busts.
Of course, sometimes this happens in reverse. Take a political revolution like the
Arab Spring, where people were driven by fear-driven inference. They thought there was
nothing they could do because the oppressive regime was too strong, but then there was an
emotional transition sparked by a particular event, like a man in Tunisia setting himself on
fire. People then went over to the other extreme and motivated inference made them think