Basics of Rational Choice Theory
• For
each option, list all the possible alternative outcomes.
• To
each outcome, assign a numerical probability and a numerical value representing
its desirability.
• Calculate
the expected value for each option.
• Choose
the option with the highest expected value.
Calculating the Expected Value of an Option
• For
that option, sum the values of each possible outcome, weighted by the
probability that that outcome will occur.
• This
sum represents the benefit that one can expect on average to obtain if one were to choose that option.
• The
term expect here is similar to what we mean when we speak of life
expectancy.
Requirements on Probability Assignments
• Rational
Choice Theory itself puts only formal constraints on the probabilities. We only
require that they not violate the probability calculus.
• Sometimes
people speak of such probabilities as being subjective or personal.
• They
need not correspond to actually observed frequencies as long as they are
internally coherent.
Example of Coherent, But Non-Objective Probability Assignments
• Suppose
Jones mistakenly believes that the probability of drawing a face card (A,K,Q,J)
from a standard bridge deck is one-third.
• RCT per
se does not criticize this assessment as
long as Jones also believes that the probability of drawing a non-face card is
two-thirds.
Requirements on Evaluations Assigned to Outcomes
• Again,
RCT puts only minimal formal constraints on the assignment of measures of
personal value.
• For
example, preferences must be transitive,
consistent and comparable.
• But
RCT per se does not care how bizarre or
immoral one’s tastes are.
Other Non-Issues in RCT
•
Rational Choice Theory
also puts only formal constraints on how one defines the decision situation.
•
The options considered
must be taken to be mutually exclusive.
•
The outcomes listed must
be both mutually exclusive and exhaustive.
•
RCT does not care about
how correctly one has identified the outcomes of the various options as long as
they are internally coherent.
Three Senses of Rationality
• RCT
attempts to lay down the formal
characteristics of what is sometimes called instrumental rationality.
• It
does not talk about whether the contents
of our beliefs are rational,
including beliefs about probabilities.
• Neither
does it attempt to answer questions about which desires or goals
are rational ones for humans to
pursue.
Uses of Rational Choice Theory
• RCT
serves as a “logic of decision”. It doesn’t tell you what
your beliefs or probabilities or values should be.
• But it
does tell you useful things about what they cannot be.
• This
permits us to critique and analyze our own decision making.
Example of the Flying Parents
• Suppose
parents of small children have to go on a trip and are wondering whether they
should fly together or fly separately.
• They
assume that the children are more apt to become orphans if they fly in the same
plane.
• But it
is more convenient and pleasant to fly together. What is the rational choice
for them to make?
Setting up the Matrix
• The
first step in any such problem is to lay out the alternative options and then
the possible outcomes of each option.
• In
this case there are two simple options: fly separately or fly together.
• (We
assume there is no other alternative - such as driving.)
Outcomes for the “Fly Together” Option with Probabilities
Outcomes for the “Fly Separately” Option with Probabilities
Outcomes for the “Fly Separately” Option with Probabilities
Outcomes for the “Fly Together” Option with Values
Exploring the Choice
• We can
first do some algebraic manipulations of the equations for the two expected
values to see if any simplifications or approximations can be made.
• We can
then insert plausible values for p, x,
y, and z to find out which option is the most reasonable one to choose.