Lecture 6 - Rational Choice

1960s popular in the US discipline of economics

behavior is a rational self-interested response to incentives (homo economicus) core idea of how humans behave

• rational actor theory (individual) • public choice theory (governments) • social choice theory (aggregating individuals)

• foundationalist ontology • positivist epistemology • privileges quantitative methodology deductive approach as it starts with theory (theory of homo economicus) *inductive approaches such as behaviouralism start with observation

work out how people might behave given their expectations of how other rational self-interested people are going to behave aim: accurate prediction of how actors behave in the aggregate, not a detailed reconstruction of how they actually reach decisions !so RC models are based on idealization

Invisible hand: idea that macrolevel economic organization functions as if there is a hand organizing behaviour, it seems as if it is planned, but there is not a master-planner: micro-level interactions have the effect of giving this appearance of an organizing hand in the macrolevel the macrolevel economy functions because of the self-interest of humans on the microlevel

we also see this in behaviouralism idea that: - behaviour is predictable - scientific explanations are based on the interests, preferences and behaviours of individual actors (methodological individualism) similar logic is applicable to collective unit, with an unitary actor assumption

behaviouralism asks: why do people behave the way they do? rational choice answers: self-interest and rationality

• people act to achieve their goals, these goals reflect self-interest (e.g. goal not te be hungry, response to eat on time) -** instrumental rationality:** people are able to achieve the course of action most likely to achieve their goals • behaviour is instrumental: through cost-benefit analysis actors try to maximize their self-interest • groups of actors have the same self-interested goals which they pursue rationally (-> macroeconomic explanations) *this doesn’t mean that everyone always acts in self-interest and is always acting rational: this is merely the best way to analyse and predict behaviour, it is an aggregate assumption that self-interest and rationality are stable

• they are assumed to be complete (involve all viable options • they are assumed to be stable/fixed (do not change easily) • they are assumed to be transitive: if i prefer A over B, and B over C -> i prefer A over C !at the individual level this doesn’t always work, but at the macroeconomic level we can assume this is how people behave

logic of collective action (important in rational choice research, also mentions in the text) • given that individuals are rational groups, being composed of individuals, are also rational • if a group shares a common goal which everyone agrees is worthwhile, we can expect everyone to contribute to the realization of this goal e.g. realist theory holds the unitary actor assumption

when the economy slows down, there is not a lot of economic growth. what should be the rational response of an individual? • individual: worry about income + employment + family financial situation -> individually rational to save some money (to thrifth money) this leads to a problem on the macroeconomic level paradox of thrift: wat is individually rational to do has a collective outcome that the economy goes even worse what is better on the collective level is not better on the individual level

expectation: introduction obligation of wearing seatbelts would lead to decrease in deaths in car accidents actually: people felt more protected -> drove less carefully -> collectively more deaths and inguries

• Keynes paradox of thrift • Peltzman effect • political questions: to what extent can and should governmetns manipulate these behaviours (individual behaviour leading to poor collective questions)

collective rationality where the best outcome for the group is different than the best outcome for the individuals police needs a confession to have enough proof: if one confesses and the other doesn’t, one gets of free if both confess, both go in for e.g. 2 year if both stay quiet, they both go in for e.g. 1 year no matter how anyone models the behaviour of the other, it is the best for the individual to confess Nash equilibrium: both confess, what is individually rational for both of them to do the Nash equilibrium is not the best outcome: if they both defect they will both only have to serve 1 year Pareto optimality : best outcome, both shut up (defect) still, they will choose the Nash equilibrium, as that is in both’s self interest

• stag hunt (Rousseau): two hunters need to decide what to hunt, one is easy but doesn’t gain much meat, the other is hard to catch (has to be together) and gains much meat. how to decide? depends on what you expect the other to do: it would be better for you to hunt the easy target if you expect that the other will not choose to collaborate • game of chicken: driving cars super close to each other, the person that swerves first loses (worst outcome is that both don’t swerve, they will both die, what is individually rational to do depends on what we think the other will do = communication problem) • restaurant dilemma: if you split prices an individual has higher incentives to pick something that is more pricy, collective it is the best to all choose something cheap, otherwise everyone will pay a lot more (intermediare model) these are often communicational problems that can be resolved by communication. communication can push nash equilibria towards more pareto efficiently outcomes a lot of work in game theory consists of trying to push Nash equilibria towards more Pareto efficient outcomes

answers within the rational choice perspective: - changing the incentive structure via rewards, sanctions or nudges - government regulation: e.g. obligations (seatbeld) or taxation (making things more expensive) answers outside the rational choice perspective - socialization - norms

challenges assumption that group members tend to contribute to a collective goal (when they all see the rational interest in this goal) whether it is the case depends on what kind of group it is. Olsen distinguishes: • privileged groups: at least one member values the good enough to carry the entire cost of securing that good (including enforcing the outcome unilaterally) -> collective action problems are avoided • intermediate groups: individuals action has a perceptible effect on other individuals (means that you can have coordination: you see what everyone does to collect -> coordinating, there selective incentives to solve collective action (e.g. rewards and/or punishments) -> collective action problems are possible • latent groups: effect of individual action is not perceptible, difficult to have selective incentives (you need to see what people do in order to create enforcement mechanisms and punishments) -> widespread collective action problems

rationality assumption is flawed: people can be stupid, easy to manipulate, not enough time to calculate decisions or consequences, may not have enough information nevertheless: RC says that rationality is best to explain human behaviour, at the aggregate level these assumptions hold RC response: bounded rationality - people satisfice (good enough option) rather than maximise their self-interest) - complete absolute best self-interest is really hard to determine + costs a lot of time

• people can also act on other reasons: e.g. altruism, envy, emotion, spite, expressive motives (e.g. in parliament: how they see you sometimes more important than you self-interest) RC response: these ‘other regarding preferences’ can always be explained in terms of individual benefit (e.g. helping others can always be seen as self-interested to gain image or voldoening)

• not all idnividuals make the same choice, even if the udnerlying condiditions are the same • individuals are influenced not only by structures and institutions, but by ideals, ideology and worldview RC response: the value of parsimonious model is explaining macro-effects: the models do not need to work on the individual level, it works on the aggregate level, not every individual is the same

• rational choice explains stability, but not change • one-shot versus repeat cooperation games are different more importantly: political scientists are often more interested in what causes changes in the status quo that in what explains the status quo