EM3_09 - Graduate Institute of International and Development

Download Report

Transcript EM3_09 - Graduate Institute of International and Development

Epistemology and Methods
Model Building:
Concepts, Arguments, and Hypotheses
April 28 2009
„Model Building“
Deductive approach
• Theory
• Construct an explanation
• Suggest hypotheses / predictions
• “Testing”
Inductive approach
• Observation
• Theory/dominant explanation fails to account for outcomes
• In-depth analysis of explanatory variables
• Potentially new explanation / theory-building
Theories (in the positivist tradition) consist of…
•
•
•
•
•
Laws
Hypotheses
Explanations
Antecedent conditions
Variables
Law
• Observed regular relationship between two phenomena
(statements of regularity)
• These can deterministic (“if A then always B”) or probabilistic
(“if A then sometimes B, with probability X”)
• Examples:
• Democratic states don’t go to war against each other
(deterministic)
• The more democratic a state the greater the support for
environmental multilateralism (probabilistic)
Hypotheses
• A conjectured relationship between two phenomena
• Example: Legalization literature / What is the effect of an
increased legalized dispute settlement system on contracting
parties willingness to further liberalize (or commit to
additional rules)?
• Competing (working) hypotheses:
• H1: With increasing legalization of the WTO, contracting
parties are more likely to liberalize
• H2: With increasing legalization of the WTO, contracting
parties are less likely to liberalize
Explanation
• Explanation connects the cause to the phenomenon being
caused, showing how causation occurs
• What could be the explanation for H1?
• What could be the explanation for H2?
Antecedent Condition
preconditions, initial conditions
• A phenomenon whose presence activates or magnifies the
action of a causal law or hypothesis
• “A causes B if C is present, otherwise not or only weakly”
• The more democratic a state is organized, the more it
(economically) benefits from an open trading system
• C: if the level of corruption is less than…
• C: the causal effect increases with the number of existing
bilateral trade agreements the state has concluded…
Variables
• Variable: A concept that has various values, e.g. the “degree of
democracy” or “power” or “conflict”.
• Dependent Variable/outcome variable:
– War (yes, no) (nominal)
– GDP/capita (interval)
– Degree of legalization (low, medium, high) (ordinal)
• Independent variable/explanatory variable:
– the existence of a democracy (yes, no)
– Share of imports/GDP
– Distribution of power (unipolar, bipolar, multipolar)
• Intervening Variable:
– “Democracy” leads to X and X leads to “great reluctance to go to war
against another democracy”
Concepts
• Define concepts you work with!
• Good Governance, Sustainable Development, Globalization,
Power, War, Conflict, Integration, Human Rights
• Big concepts, how to disaggregate
• Think early on about variance and measurement
Concepts
• Classification/Typologies
• Typologies are theoretical constructs used when variables are
measured nominally...
• Political Systems: Presidential vs. Parliamentary systems
• Varieties of Capitalism (liberal vs. coordinated market
economies)
• Developing Countries (e.g. OECD classification - ordinal)…
Constructing Arguments / Explanation
• Ask yourself questions to locate variables (pre-condition
knowing the literature and theories)
• Example Geddes: Regime change
• Intuition!
• In order to explain regime change (DV), we try to understand
why groups concluded that the old regime had become
intolerable and how they developed the organizational strength
and popular support needed to overthrow it...(Puzzle…)
Constructing Arguments / Explanation
• The breakdown of an authoritarian regime need not lead to
democratization, but when it does, the transition involves
bargaining and negotiation (induction through observation)
• Bargaining over institutions is a central feature of regime
change
Existing explanations
•
•
•
•
•
•
•
Addressing important factors/variables
Political rivalries (internal/opposition)
Upper-class support
Risks of mass expressions of discontent
Ideology
Economic shocks
Geopolitical shocks
Theorization of one process
• Politics in Authoritarian Regimes
• „Few authors have considered how characteristics of
dictatorship affect transitions“ (omitted variable)
• Theory of politics within authoritarian regimes
• Assumptions from democratic theory (survival strategies):
officials want to stay in office, best way to do so is to give
constituents what they want
• Modifications
– who are the constituents
– what performance is necessary
– But different interests of leaders in different regimes (e.g. personalist
regime)
Theorization of one process
• The military regime
• Drawing on research of attitudes/preferences of military
officers
• Logic of seizing power vs. returning to the barracks
• Coordination game between military fractions (Battle of
Sexes)
• Ultimate goal is survival of military
• Solution: negotiations or credible first mover
Theorization of one process
Looking for implications
• As officers find themselves in battle-of-sexes game, military
regimes break down more readily in response to internal splits
(in comparison to other regimes)
• The costs for the military varies according to regime type after
regime change...
• Military regimes last less long than other authoritarian regimes
• Economic crises having stronger disintegrating effect, etc...
Hypotheses from Implication
• Key argument: regime type affects the way transition occurs!
• Geddes looks at the causal mechanism, certain observations
follow (that can be tested):
• Military regimes survive less long
• Military regimes are more quickly destabilized by poor
economic performance
• Military regimes are more likely to end in negotiations …
•
•
•
•
Various form of hypotheses
Relational (longer, less long, more likely, etc…)
As A increases also B increases
A is a necessary / sufficient condition for B
Test questions to assess arguments / hypotheses
• Falsifiable?
• Do hypotheses that form part of a „theory“ contradict each
other?
• Is there a problem of endogeneity?
KKV: e.g. degree of residential segregation on conflict
between Israelis and Palestinians in communities…
• Potentially omitted variables?
Test questions to assess arguments / hypotheses
• Case-selection
– Has the outcome (DV) already occurred...Transition to
Democracy?
– Large n necessary? Case-selection (next session)
• Operationalization
– How well can we operationalize and measure the variables
(next session)