IIA. Vote Motive and Rational Ignorance

Rational Agent Model

• MB > MC → do it

• MC > MC → don’t do it

Voting Equation (for a Rational Agent)

MB - MC > 0

pB - C > 0

• p= probability your vote is decisive → extremely tiny

• B= net utility difference between candidates = | V2 - V1 |

• C= cost of voting

• Equation does not hold and yet people still vote, why?

Reasons for Voting

1. derive utility from act of voting: civic duty, limit social pressure

2. expressive voting

3. minimizing maximum regret

Expressive Voting

• utility is derived from expressing your opinion

• voting has an instrumental (useful) value and expressive (what does it say about you?) value

• expressive value is independent of the probability that your vote is decisive

Min-Max Regret Strategy

Choose action that minimizes possible future regret

• regret= the difference between the best possible outcome, given that you know it, and what you actually got

Biggest possible regret= that you did not vote and your vote would have been decisive

*does not look at probabilities, only outcomes

Rational Ignorance

rational to be ignorant because there is a point where the cost of learning > benefit

• ignorance= less than 100% knowledge of something

• may be because have high opportunity cost of time

• in politics, ratioal to be ignorant because MB is so low

“Voting, Rational Absentation, and Rational Ignorance” [Johnson]

Johnson thinks that voting can be seen as rational when broader, non-monetary factors are considered

Condorcet’s Model proved that…

in a simply majority, cannot always predict the equilibrium

Condorcet Paradox

individual voters that have transitive (consistent) preferences can display (collective) intransitive preferences (irrational and cyclical) under majority rule

• problem: whoever decides order of votes can control the result— agenda control

Cycling Definition

= don’t come to a decision (no equilibrium)

Pareto Set

= set of points for which cannot help without harming someone else

Definition of Logrolling

the practice of trading votes within a legislature to ensure the passage if multiple issues

Core Assumption of MVT [Downs]

“Parties formulate policy in order to win elections rather than win elections in order to formulate policy.”

→ maximizing likelihood will win

Hotelling’s “Spatial Location Theory”

product differentiation decreases so firms converge to some specific point (point that maximizes profit is inefficient for consumers)

What does logrolling accomplish?

Allows policies to pass that might not under majority rule

Median Voter Theorem [Downs]

in a majority rule system with single-peaked preferences and two candidates, the candidate who wins the median vote will win the election

Assumptions of MVT [Downs]

1. vote maximizers

2. duopoly

3. perfect info

4. majority rule

5. single issue policy space

6. spacial mobility (candidates and parties)

7. no absentations

8. individual preferences are single-peaked

Majority

50+1

Plurality

most votes win

super majority

majority greater than simple majority

simple majority runoff

simple majority vote and candidate with lowest number of votes is dropped (can lose the Condorcet winner)

Condorcet criterion

choose the candidate who defeats all others in pairwise elections in majority rule

Hare system

get rid of the candidate with the lowest number of first place votes; keep going until majority is found

Coombs system

get rid of candidate with most number of last place votes; keep voting until majority is reached

Borda count

each candidate given a score [1,m]; most number of points wins

Which system yields Condorcet winners most often?

Borda Count

Two Types of Costs

1. external costs= costs one expects to incur as a result of the actions of others

2. decision-making costs= costs one expects to incur while participating in the whole set of collective decisions defined by a single collective action