✅🐀 Snidal -- Rational Choice

⭐Summary

Argument 💫

  • Challenges to rational choice theory are valid but should be used to improve RC rather than dismiss it entirely.

  • RC is a broad framework that can be applied to multiple paradigms and used either normatively or positively.

  • RC benefits from being general/broad and simple because it’s so widely applicable and flexible

  • Rational choice should be proud of how far it’s come but also acknowledge its own distortions and constraints (intellectual humility).


Perspective 👁️

Rationalist

  • obviously a fan of RC

  • a very “our approach should depend on the situation” kind of guy

    • tries to bridge gap between normative and positive uses of RC

    • argues RC is a broad framework that can be used within all IR paradigms

    • argues math and verbal (anecdotal, deductive) analyses are equally important, supporting each other and compensating for each other’s flaws

  • Very Wohlforth in the claim that rationalism is not a monolith

  • Very Snyder in the claim that “we can blend multiple approaches to get better analysis”


Key Questions ❔

  • how can one address criticisms of RC from both within RC (internal) and outside of it (external)?


  • What is RC?

    • BASIC PREMISE: RC is a theory that describes how goal-oriented actors make decisions under constraints

    • simple and general — an umbrella, not a specific theory

    • analyzes the effect of various constraints (technology, domestic interest groups, other actors’ decisions)

    • it can be normative and/or positive, empirical and/or deductive, mathematical and/or verbal


  • how empirical (see “Key Vocab”) is RC?

    • sometimes it’s empirical (testable), sometimes it isn’t

    • it’s hard to test “rational choice” as a theory because it is so broad, so you can only test applications of it


  • is it necessary to formalize how people apply RC?

    • no, many people way back in the when understood RC without mathematical modeling (ex. reciprocal strategies were expressed as “tit for tat” or “eye for an eye”

    • however, formalization (using math) and verbal theory (anecdotal, pragmatic analyses) complement each other and are best used together

    • Using math reveals things that verbal theory misses, and vice versa


  • how does RC address change over time?

    • accounts for how nonstate actors (from companies to nonprofits to even terrorists!), domestic interest groups (electorate/selectorate the leader is accountable to/derives power from), historical traumas, shifting national goals can change a nation’s preferences over time

    • however, it doesn’t always need to do this because national struggles (war, shortages, economic turmoil, wealth inequality, immigration) recur over time because states are “like units” (structured differently depending on time and place, but structured to deal with many of the same problems)


  • how does RC explain who actors are and what they want/like? does it do this effectively?

    • see above

    • it can’t fully explain what actors want/like because it can’t see inside their heads, so it will always be limited and ineffective in some ways

    • with RC one has to deduce what actors want/like from the decisions they make


  • how does RC address normative concerns, or the idea of “what should be done” in international politics?

    • RC is already kind of normative: “based on the actor’s goals, what should they do to achieve those goals given the circumstances”

    • but the author argues that RC should be more normative

      • RC shouldn’t just be used to describe how interests lead to goals, it should describe how interests should be harnessed to achieve the best possible goals (which is itself a normative statement)


Key terms 🔤

  • rationalism / rational choice: a broad theoretical framework that mainly focuses on what decisions goal-seeking actors make under constraints (limits)

  • substantive: involving important, serious issues

  • methodology: rules and methods people use to do something

  • formalization: using math models to represent and validate theory

  • “hard” methodological approaches: math (quantitative), see “formalization”

  • “soft” methodological approaches: not math (qualitative)

  • empirical: provable by observation or experiments/testing

  • dynamism: change over time

  • normativism: belief in “what should be done”

  • exogenous: outside the model (taken as given, held constant)

  • endogenous: within the model (needs to be explained)

  • equilibria / equilibrium / equilibrium analysis: assumes that each variable within a model has no effect on the others and that they are all in balance (equilibrium)

  • “aggregate as actor”: simplifying the aggregate (total) of decisions among many individuals into a single decision by one actor

    • e.g nation as actor, despite nations being made up of many decision-making groups


📝Notes

I. Challenges to Rational Choice (internal summary)

RC key framework => drawn criticism

   

Key Vocab

  • substantive: involving important, serious issues

  • methodology: rules and methods people use to do something

  • formalization: laying out set, formal rules and frameworks for a thing. making something “official.”

    • the author specifically uses formalization to mean “using math to represent theory by making models of real things”

  • “hard” methodological approaches: math (quantitative)

  • “soft” methodological approaches: not math (qualitative)

  • empirical: provable by observation or experiments/testing

  • dynamism: change over time

  • normativism: belief in “what should be done”


Key Questions

  • how can one address criticisms of RC from both within RC (internal) and outside of it (external)?

  • What is RC?

  • how empirical (see “Key Vocab”) is RC?

  • is it necessary to formalize how people apply RC?

  • how does RC address change over time?

  • how does RC explain who actors are and what they want/like? does it do this effectively?

  • how does RC address normative concerns, or the idea of “what should be done” in international politics?

  • internal criticisms

    • RC too technical

    • RC not empirical enough

  • external criticisms

    • debate between rationalists and constructivists

    • emphasizes some problems while deemphasizing other equally important ones

    • cannot explain actors, interests, institutions, or change

    • rebuttal - rationalism is rooted in realism but widely applicable


II. What Is Rational Choice

  • “individual goal-seeking under constraints”

    • actors pursue what they want/like despite challenges

  • focuses on:

    • actors = their goals, their capabilities

    • constraints= technological constraints, institutional constraints, constraints imposed by others’ choices

  • It is a methodology (a way of doing things) rather than a theory (a way of explaining or predicting things)

  • It is an “umbrella” describing many things, not one fixed framework

    • similar to Wohlforth’s analysis of realism!


Goals

  • Under rational choice, actors’ goals could be self-interested OR normative / ideological

  • Usually used positively (in ways that describe “what is”) but can also be used normatively (in ways that describe “what should be”)

  • “goal seeking” isn’t the only thing explaining behavior, but it is one of the key drivers of/ foundations of behavior


Equilibrium

  • rational choice doesn’t specify what things need to be held constant (exogenous) and what things need to be explained (endogenous)


Commitments

  • Simplicity

    • the fewer assumptions (inferences), the better

  • Generalization

    • abstract and broad, not specific and concrete

    • provides widely applicable insights

  • a lot of “rational choice” debates are actually debates about things that aren’t rational choice, even though they started with a rational choice framework


  • key terms:


    • goal-seeking = pursuing goals / desires

    • constraints = limits on behavior

    • self-regarding = self-interested

    • ideational = based on ideas

    • hyper-rational = very rational, more rational than an actual human probably would be

    • positive = based on “what is”

    • assumption = to suppose something is true

    • anomic = having no norms or values, “gone rogue”

    • exogenous = outside the model (taken as given, held constant)

    • endogenous = within the model (needs to be explained)

    • equilibria / equilibrium / equilibrium analysis = claims that each variable within a model has no effect on the others and that they are all in balance

    • constitutive = having to do with the parts that make up something

    • generalization = inferring broad trends from specific scenarios

    • tautologies =


  • broad, flexible

    • not substantial or causal, but explanatory


III. Methodological Debates and Challenges for Rational Choice

  • internal critiques

  • about methodology = “how to do rational choice”

    Rational Choice and Formalization

What is the role of math models in rational choice theory?

  • “Rational choice has an elective affinity for formalization” = people who use rational choice like to do math and use math to represent the world\

  • implicit question: “do you need to use math to use rational choice?”

    • no, you don’t need lots of math modeling [formalization] to do rational choice, many people have used other methods (game theory, anecdotal) as long as rational choice has been around and even before then. however, if correctly applied, mathematical modeling can make rational-choice analyses better ✌

    • some people knew things before math could explain it (ex. tit-for-tat), but explaining things with math can create a better understanding of those things and change people’s minds about what an effective course of action really looks like

    • just because things are “counterintuitive” doesn’t make them less true

    • using math helps better analyze arguments & assumptions

    • Schelling — math isn’t sufficient on its own to explain IR and verbal theories of military strategy are dead

      • use math to build foundations for verbal arguments

  • How to balance math (hard) and verbal (soft) approaches in RC by creating a successful model

    • Simple and elegant: model shouldn’t be so overcomplicated it overshoots reality

    • Synthesizes math and verbal logic: basic logic of the model should be in verbal, not numerical, terms; but that verbal logic should be supported by mathematical evidence

    • Testable: model should succeed under real-world testing/observation (it should be empirical)


    Empirical evaluations of RC arguments

Deductive: conclusions from abstract reasoning over existing evidence

Empirical: conclusions drawn from observation and testing

  • “RC is not empirical enough!”

    • it’s nearly impossible to observe things like actor motivations, preferences, and beliefs

    • it’s hard to test something so general and insubstantial, testing rational choice is like testing “statistics”


  • But RC is empirically useful

    • gives examples of empirically useful IR research


  • RC can’t explain everything. No theory can. So what can RC explain?

    • motivation based on consequence

    • motivated based on “appropriateness”

  • sometimes people like to “curve fit” / “rationalize” their models (tailor their data to fit what they already believe about it) instead of actually testing them

  • but rational choice can help prevent this

    • using math helps disprove “rationalized” arguments

    • emphasis on simplicity helps reveal what is elegant, well-founded theory and what is complicated mental gymnastics

    • emphasis on generalizable conclusions/insights limits cherry-picking specific situations

  • “fixed preference” = in any given scenario, an actor will want [thing].

    • criticism that this ignores change over time

  • arises from balancing broad theory with testing

big problem to assume “preferences caused by behavior” / “norms caused by behavior”, because behavior causes preferences and those two things are not interchangeable (a tautology)


two ways to have quality qualitative testing of RC models

  •   case studies

  • “analytic narratives” - evaluating case studies within a RC framework, using RC to describe actors, preferences, choice, constraints, and other variables

    • can help describe trends over time

    • can help bridge the gap between valid logic and necessary simplicity

    • can explain things both inside the case and across cases through RC

  • Balancing analytic narratives (Elster)

    • actors within the model shouldn’t be given the ability to play 8d chess; there should be room for irrationality, miscalculation, etc. because that is how it is in reality

    • “aggregate as actor” - simplifying groups made up of a variety of individuals into a single individual decision-maker

    • within an actor’s calculations, rational choices/strategy-based decisions can share space with moral concerns


IV. Substantive Challenges for Rational Choice

    Incorporating Dynamics and Change

  • equilibrium (holding certain things constant) = big obstacle to understanding change, as in reality everything is constantly shifting

  • “equilibrium path” = status quo, prevailing general conditions

    • when is it constant, when does it shift?

    • example: “anarchy,” a core assumption of realism

  • author argues RC works best concerning change when applied to cooperation and institutions

  • why do RC analysts hesitate to analyze change? two assumptions/reasons

    • because states are “like units” facing similar problems yesterday, today, and tomorrow

      • power struggles are constant, financial issues are constant, immigration issues are constant. History repeats itself

    • because large-scale change happens very slowly

    Endogenous Actors and Preferences   

traditional assumption: actors and preferences remain constant, what changes are the things that constrain those actors/preferences

  • because you can’t see preferences, but it’s easier to see constraints

  • fixed preferences = concrete, testable arguments; dynamic preferences = untestable, inconsistent arguments

  • but preferences do change!

  • fixed-preference analysis is good sometimes, but not all the time


  • RC & Domestic-level interest groups

    • audience costs

    • two-level games

    • commitment problems

  • preference change comes from changes in underlying constituencies

    • i.e changes in policy/decision-making come from changes in opinion by the people the policymaker/decision-maker is accountable to/derives power from

      • authoritarian = “selectorate” / elites

      • democratic leader = “electorate” / general voting public


  • RC & International actors

  • explaining the role of nonstate actors (NGOs, MNCs, terrorist groups) in international politics


  • RC and Actors’ Movements to Change Each other

  • ADD MORE ACTORS

  • Changing membership within international institutions

    • conditions for membership can change based on the issue

    • firms/NGO might participate in solving one issue but not another

  • Actors trying to change other actors

    • Human rights NGOs trying to make a brutal police state less repressive


RC and Actors’ Movements to Change Their Own Preferences

  • change in information = change in preferences

  • change in institutions =change in preferences

    Normative and Policy Analysis

Instead of “based on actors’ interests, what will happen”, RC should look at “how can we align actors’ interests with what ‘should’ happen”

  • it is normative: a theory about how people “should” behave in each situation

  • but it could be more normative

    • addressing resource inequalities within cooperation & how those inequalities create “winners and losers”

    • can help incorporate interests into normative analyses, preventing idealistic miscalculation

    • can help narrow discussions about the role of international law

V. Conclusion

  • Math (“soft)” and verbal (“hard”) methods work with each other in RC, rather than being separate from / superior to each other

  • Flaws in RC are because

    • RC analysts cannot perfectly predict how the world works

  • Inability to explain change can be an asset in one situation and a liability in another

  • Limitations regarding normative/strategic preference ordering, change over time are what make RC effective, it would be too broad otherwise


  • rational choice should be proud of how much it’s revealed about IR, but also realistic about its flaws and limitations in doing so