In this lesson, we will start by discussing
neoclassical economics as a foundation for thinking about behavioral economics.
Neoclassical economics has at it's core expected utility theory which
presumes to describe both how people make
decisions and how those decisions should be made.
However in recent decades,
psychologists and behavioral economists have
identified anomalies in behavior that couldn't be explained by utility theory.
This has consistently and considerably enhanced our understanding of human behavior.
Let's start by explaining expected utility theory before identifying some of
it's limitations and some key elements that are addressed by behavioral economics.
Expected utility theory assumes people act out of
self-interest that they know what's best for
themselves and that they're able to act on that understanding.
This could be described by a person being able
to assess the probability of different potential outcomes.
Having a clear sense of
how good positive utility or how bad disutility they'd feel about each outcome.
And then through a process of backwards induction
figuring out which choice path has the highest net present value.
People try to maximize their expected utility,
the sum of the probabilities times the utility for each outcome.
Here are some key assumptions of expected utility maximisation.
One is that people have
fairly stable preferences that are reasonably consistent over time.
A second is that the starting point doesn't really matter.
A third is that people have a good understanding of probabilities and can
estimate the probabilities of potential events reasonably, accurately.
Fourth, while the future may be less important than the present,
people discount the future at constant rates.
Fifth, the framing of choices doesn't matter.
Six, people are fully rational.
So, if these assumptions were all true,
which unfortunately they are not,
this would have a couple of important implications.
One is that the government should only intervene in developing
programs in cases of clear market failure.
Or for example, where people's actions harm others through negative externalities.
An example of a negative externality would be,
if I walk into a restaurant or some other close space and I started smoking cigarettes,
that would have negative benefits to my neighbors and that should be accounted for.
Another implication of this if
all these assumptions were true would be that problems like
obesity could basically be explained away as a reflection of people's preferences.
If you're a clinician or a public health professional,
you'd probably find that a bit problematic.
Over the years, Daniel Kahneman and Amos Tversky.
Did a lot of very interesting and important work in behavioral economics.
A lot of this was summarized in a paper they wrote
that was published in econometrics in 1979 that
summarized a theory called Prospect theory which provided
an overarching conceptual framework for describing
the anomalies they'd observed in a number of individual studies.
There are several key points.
One is how people feel about a set of possible outcomes depends on their starting point.
This is called reference dependence.
Decision makers evaluate outcomes as gains or losses depending on their starting point.
A second important point is the notion of
diminished sensitivity to both gains and losses depending on the starting point.
A third is that people exhibit loss aversion,
the disutility of a loss is much stronger than the utility of equivalent dollar gain.
Fourth, people overweigh small probabilities.
This is called non-linear probability weighting.
Let me explain further what's meant by each of these.
Here's an interesting example of reference dependence.
Let's imagine two people who each have four million dollars today.
Neoclassical economics would assume they have roughly equivalent utility of wealth.
However, then consider that one of them had a million dollars yesterday,
and the other had seven million.
Clearly, the person who went from one million to four million would be ecstatic.
The person who went from seven million to four million would be despondent.
Very different depending on where their starting points were.
If we go back to the graph that depicts prospect theory,
what we have in the Y axis is a measure that's
similar to utility that depicts psychological value.
People started at a reference point in the middle of the graph.
And then as the graph goes to the right and upwards we're in the realm of gains.
As the graph goes to the left and downwards,
we're in the realm of losses.
We can illustrate the point about diminished sensitivity to gains and losses
by considering the plight of a homeless person who has no money,
who somebody generously gives $100.
We can see that psychological value namely the shift
in the Y-axis goes up a lot from zero dollars to $100.
But if we contrasted that with somebody who already had
$100 or $200 or pushing this out further let's say,
$1,000 or $10,000, receiving a
$100 would no longer have such a big impact on my psychological value.
In essence, that gets smaller and smaller,
the more I start with.
And the same would be true in the realm of losses.
Loss aversion is defined as the disutility of
losing money being much greater than the utility of gaining the same amount of money.
A number of studies have shown that people have what Kahneman
calls a loss aversion ratio in a range of 1.5-2.5.
This could be illustrated by the fact that when given a 50,50 chance of
winning $150 or losing $100 most people
would not voluntarily enter into this gamble because of the potential pain
of losing $100 is much greater than the joy of winning a $150.
Under standard expected utility maximisation,
it'd be a no brainer to take this gamble since the expected utility of point five times
negative 100 plus point five times 150 would be $25 which is greater than zero.
Going back to the graph of prospects theory,
we can see how loss aversion is depicted by
looking at the fact that there's a steeper slope when we're in the realm of losses,
than we're in the realm of gains.
In other words, psychological value changes for each unit shift
in the X-axis when where in the realm of losses relative to gains.
Thinking of non-linear probability weighting is
important as well because this has a lot of important implications.
In expected utility, probabilities are linear.
In contrast, in prospect theory,
small probabilities are over-weighted.
This is because of the possibility effect.
Think about the frenzy that your neighbors might have
experienced in the last Powerball lottery.
They felt like they could actually win even
if the probabilities were infinitesimally small.
And in essence, those small probabilities near zero are commonly over-weighted.
We could think about this by contrasting
the value of a shift and probability of zero to point
01 with that of shift in probability from point 50 to point 51.
Shifting from point 50 to a point 51 doesn't excite too many people going from zero to
point to 01 when you're thinking about the possibility of
something like winning a large prize becomes quite important.
Similarly, people value certainty a lot when it comes to large probabilities.
So, a probability of point 99 is worth much less than the probability of one.
We could also see how there at that change of point 01 is
much more important than if we were to go from point 50 to point 51.
There are lots of important implications of this in terms of
the purchase of insurance to protect against rare events,
lotteries to win money or willingness to take
a less attractive deal if there is a 100 percent probability.
In this module, I've given you a flavor for
neoclassical economics and how it is enhanced by behavioral economics.
We shouldn't discard the basic notion that people are somewhat
rational and try to make decisions that maximize their expected utility.
What behavioral economics has contributed is a better understanding of
how the notion that people consistently maximize expected utility is incomplete.
This is in areas like reference dependence,
diminished sensitivity to gains or losses,
loss aversion and non-linear probability weighting.
We'll build it on this further in subsequent lectures. Thank you.
Christian TerwieschAndrew M. Heller Professor at the Wharton School, Senior Fellow Leonard Davis Institute for Health Economics Co-Director, Mack Institute of Innovation Management
Amol S. Navathe, MD, PhDAssistant Professor of Medical Ethics and Health Policy
David A. Asch, MD, MBAProfessor of Medicine and Professor of Medical Ethics and Health Policy
Roy Rosin, MBAChief Innovation Officer
Kevin Volpp, MD, PhDProfessor of Medicine, Division of Health Policy / Professor of Health Care Management
