Self-Interest

game theory

an analysis of strategic interactions it is an attempt at predicting and understanding what we might think of as social decision making or interdependent choice behavior

lab games

- allows us to use motivated choices - allows us to control for ulterior motives for fairness such as pro quo, reputation building, fear of reprisal - makes it easier to measure more directly how people trade-off self-interest with concerns for others

typical results of the dictator game

- bimodal offers (largest mode at 0%, 2nd largest at 50%) - a majority of offers are positive - offers just under 50% are rare

homo economicus view on rationality

- completely selfish - self-interested - profit-maximizing - money-maximizing

repugnance

- distaste for certain kinds of transactions - barrier/constraint to markets - destroys welfare b/c it does not allow trades that would make both parties better off

goal of self-interest section

- focus on decision making when the outcomes of a choice have consequences for ourselves and others the main question "how do social preferences or concerns for fairness change our maximization problem?"

fehr & schmidt's "inequity avesion" model

- gives us a simple way to model preferences for fairness - it is a model of rational choice

three ways to save lives

- increase supply - remove technical barriers - efficient allocation

ultimatum game

- lab game - proposer offers a division of an endowment between herself and a responder - reponder accepts or rejects the offer

trust

- lowers transaction costs - makes life more convenient - having trust in society is a public good

social preferences are as rationalizable as other preferences

- perfect selfishness - perfect complements - perfect substitutes

fehr & schmidt's "inequity avesion" model, a person i's utility is determined by three things

- person i's payoff, the utility in the payoff matrix (xi) - the average disadvantage of i's payoff to others, multiplied by alpha (ENVY) - the average advantage of i's payoff over others, multiplied by beta (GUILT)

typical results for the trust/investment game

- positive amounts are usually passed - if something is passed, a positive amount is almost always returned - amount returned on average is less than amount sent

subgame perfect nash equilibrium

- refinement for nash - b/c the proposer moves first, she should choose her favorite among all equilibria

factors that impact behavior in social dilemmas

- repeated play - altruistic punishment - communication - magical beliefs - fear

ethos of economics

- respect for transactors, people are capable of choosing for themselves - respect for people's preferences - treat all market demands equally

in the ultimatum game, if it is common knowledge that both players are narrowly selfish:

- responders will always accept - proposers will choose their favorite equilibrium = subgame perfection

biases in the perception of fairness

- settlement zone - bias of self-enhancement - self serving bias

cash payments and repugnance

- some are not repugnant as gifts, but are when money is added - ie. interest for loans, paying birth mother, prostitution, donating organs

tit for tat

- start w/cooperating - continue mimicking their last move cooperating with cooperators and defecting with defectors - the most fit strategy - limits loss but allows to take advantage of gains

self-interest and normativity

- this is a problem and a reason self-interest is less focused on - economics has no default ethical position from which to deviate (if not selfishness) - there are no normative ethical principles, BUT normatively requires not contradicting one's self -- but this happens when we fall to framing effects regarding unfairness

ultimatum game WEIRD results

- typical offers 40-50% - offers > 50 (hyper fair) are RARE - offers <20% are rejected half the time (they are rarely made)

when do dictators give?

- when beta is equal to or more than 1/2 - when less than 1/2, dictators keep everything (and give nothing)

self-serving bias + perceptions of fairness

-> I are biased toward what benefits me

3 concerns regarding repugnance

1. objectification 2. coercion and exploitation 3. slippery slope

dictators never give more than...

1/2

result of exiting dictator games

1/3- 1/2 of people exit - this suggests that people feel some guilt or social pressure

trust/investment game

2 players receive the same endowment sender can pass part of her endowment to the responder (anything passed is multiplied by 3) - responder can give any amount back to sender (he is a dictator) trust is involved - sender is vulnerable, but opens up to gains

seemingly repugnant things

3rd parties block these things between two people - dwarf tossing - prostitution - paying for organs - paying birth mothers of adopted children - sam sex marriage

perfect selfishness

KEEP EVERYTHING always the corner solution

crankcase oil problem

This problem was an example of the dangers of expanding what defines utility beyond the assumption of narrow self-interest. In the extreme, if we allow for utility to be increasing due to any "present goals," we can find ourselves in a situation where we have to say that even something such as drinking oil, writhing in pain, and dying was a rational, utility maximizing, decision when it was made. This example highlights the need to be very specific about the factors influencing "social preferences." It wasn't enough to simply say that people appear to have a preference for "fair" outcomes, we needed to specify/describe precisely how people cared about fairness. This allows us to make falsifiable predictions that can be compared to the predictions from other models of self-regarding or other-regarding preferences.

strategy

a complete plan of action that describes what a player will do for every possible situation throughout the game

if b = 1...

a person will never stop giving while he has more, no matter how costly giving is THUS, we assume b <1

nash equilibrium

a strategy profile such that each strategy in the profile is a best response to the other strategies in the profile - no play has the incentive to switch strategy once they are at a nash

fear

affects behavior in social dilemma - "protection" against others defecting - if a person wants to cooperate, but its afraid that others will defect, so the person defects ie. cheating person does not want to cheat, but if he thinks others will, then it is sensible to cheat

assumptions for fehr & schmidt's "inequity avesion" model

alpha > beta 0< B < 1

prisoner's dilemma

an individual's best response is at odds with what's best for society

what is equilibrium in the ultimatum game?

any accepted offer is an equilibrium (neither player will do better by changing)

the responder always accepts offers of...

at least .5

how to determine subgame perfect

backwards induction! consider what offers the responder would accept then consider which of those the proposer would most like to offer

altruistic punishment

basic emotions behind conditional cooperation can change behavior in social dilemmas - if people have positive a or b, then they may try to cooperate and punish people - altruistic BC: costs the individual and benefits everyone

beta is determined by the exchange rate...

by what giving up 1 unit to yourselves can give to the other person: 1 unit cost to me / unit cost to me + units benefit to you

communication affect on behavior in social dilemmas

can have a dramatic effect, even if the game will ultimately be played under anonymity led to large increases in cooperation

experiment w/ altruistic punishment

can keep money, or contribute to group --> 3x players could monitor whether others were cooperating and inflict losses on those who didn't

biases of self-enhancement

common in people ie. 90% of drivers rated themselves above average -- extends to judgments of what is fair -> I are biased toward what benefits me (self-serving bias)

social preferences

concern for more than our narrow self-interest can lead to several anomalies for orthodox theory examples - people reject prices that reflect supply and demand and leave them with surplus - third parties may prevent others from trading - people may cooperate when not cooperating is dominant

dual entitlement

consumers track changes in price from reference point. fairness ideas hold that consumers are entitled to the reference transaction, while firms are entitled to their reference profit

UNOS feedback

demonstrates that the process assumes that measuring attitudes is information-neutral

dictator game

dictator offers a division of endowment recipient is passive and can not do anything * it is not really a game b/c there really isn't strategy

social dilemma

each person has an incentive to do one thing, but everyone would be better off if everyone did something else *what is smart for the individual is stupid for the group --- extension so prisoner's dilemma to several players

pareto principle

efficiency principle states if every member of the society prefers one action state to another, then society must also prefer it - no one can be made better off without making someone else worse off

commensurability

everything has a price all things can be exchanged with or compensated for by other things and thus measured on a like scale.

grapefruit/avocado dilemma

example of how economics-trained individuals tend to think differently - think of problem like a psychologist's inkblot - many people quickly gravitate toward a solution that probably indicates something about where their sensitive lie when thinking about equality and maximization - the only way to guide choice is by ethics

equilibrium selection problem

how people will choose when there are multiple nash equilibria - theorists have introduced different refinements (ie. subgame perfect)

conditional cooperator

ie. fehr-schimidt player likes to cooperate with someone who is cooperating, but defect against a defector

why we assume b<1

if b>1, then people would give up wealth even if it couldn't do anything for someone else, just to get rid of your advantage

Sen's Liberal Paradox

if people have nosy preferences, we can not respect individual rights and ensure efficient outcomes --> it is impossible to be a Paretian liberal basically, with nosiness, we can not be assured that allowing more individual freedom will increase (pareto) efficiency

shifts in repugnance

in the eye of the beholder

exiting dictator games

instead of playing the game, you can exit for some amount less than the full endowment and the recipient never finds out there is a game

pay what you want

makes people feel a social pressure

perfect complements

maximin preference - always seek to equalize payments - treat wealth as if they are complements to each other (right angle curve) -- move toward point where parties get the same amount Rawlsian

maximin rule

maximizing the minimum amount that anyone gets

fairness equilibria

nash equilibria arrived through agents best responding to other agents with social preferences

impossibility theorem

no social choice function can satisfy both pareto efficiency and minimal liberalism b/c there always exists some set of preferences that puts the two in conflict

nosy preferences

preferences over others' actions, even if they have little or no material impact on one's self - inefficient

advantageous inequity

people give to others who are behind

disadvantageous inequity

people spend money to restore equity (and destroy welfare) when the are behind

magical beliefs

play as second mover - behaved as if their anonymous behavior would influence others

row strategies

player 1

column strategies

player 2

pure coordinate game

players' interests are perfectly aligned ie. train station game, they simply want to coordinate actions ie. driving on the right side of the street

KK&T's Proposition 3

price changes will be more responsive to variations of cost than variations of demand, and more responsive to cost increases than cost decreases SO, - not allowed to raise prices due to increased demand - can raise prices due to increased costs - does not have to lower prices when costs lower

KK&T's Proposition 4

price decreases will often take the form of discounts rather than reductions in the list price - b/c canceling discounts is better than raising prices

payoffs

rewards or punishments that correspond to each possible combination of strategies that players can choose from

shortcut to determine minimal offer a responder would accept

s = alpha / (1 +2*alpha)

punishment is a...

second-order social dilemma b/c the one who does the punishment pays the cost; while everyone else can free ride on the benefits

payoff matrix

shows the payoffs resulting from each possible strategy profile

case work experiment w/ self-serving bias + perceptions of fairness

subjects read case bargain over money in role of plaintiff and defendant shows that people can not overcome self-serving biases even when they should because failed settlement will not work out as well as they think it will *key manipulation was the sequence of events, if people knew their role before reading the case when loewenstein tried to eliminate the bias by educating subjects, it only helped subjects to recognize bias in their opponents

work experiment w/ self-serving bias + perceptions of fairness

subjects took survey dictator game w/ context told that they worked 2x longer, finished 2x more tasks, both, or neither

consequentialism

the action that produces the best consequence is the best one to take

if alpha approaching infinity

the person will only accept 50/50 split... or less he will reject

if beta is less than .5...

the proposer will not be willing to split

selfish equilibrium in trust/investment game

the responder gives nothing back and the sender passes nothing in anticipation

repeated play

this impacts behavior in social dilemmas - when repeated, defection is not always the best solution - the fact that the game is repeated can change the analysis of what is equilibrium

selfish "equilibrium"

to give nothing in the dictator game

social game w/ altruistic punishment

usually, cooperation levels will decay to almost zero BUT w/ punishment, contribution rate maintains or increases

perfect substitutes

utilitarian to give away tokens to the person who values them more --- the player treats her own wealth and others' wealth as perfect substitutes (parallel diagonal lines)

strategy profile

vector of strategies, one for each player written as (player 1's strategy, player 2's strategy..)

rawlsians

want to maximize the minimum amount any person makes

utiliarians

want to maximize the total welfare across people U(x,y) = u(x) + u(y)

minimal liberalism

way of representing individual freedoms - if the difference between two action states only involves the behavior of one member, that member's preferences should be decisive over those states - no member of society can dictate what another member does in his or her private sphere

KK&T's Proposition 2

when a single supplier provides a family of goods for which there is a differential demand without corresponding variation of input costs, shortages of the most valued items will occur (b/c prices don't reflect quality differences)

KK&T's Proposition 1

when excess demand in a customer marker is unaccompanied by increases in suppliers' costs, the market will fail to clear in the short run (bc there is no excuse to raise prices to market clearing level)