PowerPoint Presentation - Is It Rational to Vote?

advertisement
Is It Rational to Vote?
• Political scientists study all
aspects of voting behavior. The
most interesting question, of
course, is who votes for whom
and why. But they have gotten
bogged down on what might
sound like a trivial question:
why do people even bother to
vote!
• What does Rational Choice
Theory have to say on this
topic?
Simple RCT Analysis of
the Decision to Vote
• Let us assume that our prospective
voter foresees a benefit B if her
favorite wins.
• But how likely is it that her vote will
make the difference that leads to
victory? Call that probability P. (Her
estimate of the probability that the
candidate will win regardless is R.)
• There is also a cost C incurred by
going to the polls (loss in time, the
inconvenience).
• Let us now compute the expected
utility of going to the polls.
Matrix for the Voting
Decision
Outcomes: Favorite
Wins
Favorite
Loses
Actions:
Vote
Prob = R + P
Util = B - C
Prob = 1 (R + P)
Util = - C
Not Vote
Prob = R
Util = B
Prob = 1 - R
When we compare the two actions, we see that
the expected utility of the voting option will be
less than that of staying home unless P x B > C.
Since any realistic estimate of P will be very small,
on this analysis, no individual should ever vote!
Alternative Matrix for
the Voting Decision
Decisive for Made No
Victory
Difference
Outcomes:
to outcome
Actions:
Vote
Not Vote
Prob = P
Util = B - C
Prob = 1 - P
Util = - C
This is the status quo - no
costs, no impact on the
election.
Realistically, the value of P is always going to
be very small. So unless C is set at practically
zero (which isn’t very realistic), the expected
utility of voting will always be less than that
of not voting. So from an individual perspective,
no one should ever vote!
Can Decision Theory Be
Modified to Explain
Actual Voting Behavior?
• So far, RCT can not explain why
people vote. But it does make some
plausible sounding comparative
predictions.
• According to RCT, overall voter
turnout should increase when
people’s perception of the value of B
increases and that of C decreases,
and when their estimate of P
increases.
• To explain why people vote at all,
however, we need to add a factor.
Let’s call it D for duty. People vote
when D + (P x B) > C.
Avoid Making RCT into
a Toothless Tautology
• We could collect data to check the
comparative claims of the RCT
analysis of voting. (Does increased
C in fact lower voter turnout?)
• But to say that those people who
vote are the ones who think they
have a duty to vote is getting pretty
close to saying that the people who
choose to vote are the ones who
want to vote.
• Is there an independent way to
predict who is going to feel a duty to
vote? Or must we simply wait and
see!
Another Reason not to
Just Add In Duty
• RCT analyzes decisions in terms of
the consequences of actions.
However, the D value does not
attach to outcomes but to the act
itself.
• In parallel language we can say that
most of the values that appear in the
matrix are “investment” values while
D is a “consumption” value.
• However, one could replace the duty
factor with a “guilt” factor which
only kicks in after friends say, “You
didn’t vote?”
Other Attempts to
Explain the Decision to
Vote
• Prospect Theory tells us that
people often overestimate the
numerical value of small
probabilities. Could people have
inflated ideas about how likely
it is that their individual vote
would be decisive?
• Regret Theory (see Plous, pp.
101-102) suggests another
possible explanation.
Regret Matrix for the
Voting Decision
Outcomes: So Close That Not that Close
My Vote Was
Decisive
Actions:
Vote
My candidate
won so no
regrets!
My vote didn’t
matter so
regret of C
Not Vote
Because I
didn’t vote, my
candidate lost.
Regret of B
My vote
wouldn’t have
mattered
anyway - no
regrets
The rule of Minimizing Maximum Regret
predicts people should choose to vote since
B will typically be much larger than C.
Other Psychological and
Sociological Accounts
• Other approaches talk about how
interested people are in politics, the
extent to which they are identified
with a party that mobilizes voter
turnout, the degree to which they
feel alienated from society, the
attitudes of their friends, etc.
• Many of these considerations can be
incorporated into RCT, either as
additional costs or as duty-like
considerations. None of them speak
to the core issue of whether an
individual thinks their single vote
could make a difference.
Brief summary of
Empirical Studies
• For about half of the population the
factors that appear in RCT have no
effect on their behavior - they are
moved simply by Duty.
• For the other half, considerations of
B, P and C are influential but these
factors are not combined in the
manner dictated by RCT. These
people are more apt to vote in close
elections, but they seem to be
thinking that if they vote maybe lots
of other folks like them will also
turnout and together they will make
a difference.
• Blais believes people may simply
add B and P instead of taking their
product. C has the least effect.
Download