comparative politics
sub-field of political science studying
and analyzing them empirically by exploring their similarities and differences across political systems
e.g. why do young people participate politically differently than older generations?
implicates = comparative politics is defined by its substance (really broad: anything that takes place at a political system) and its method
WHY COMPARISON?
intuitive/implicit comparison made explicit
comparative methods
= rules and standards of comparative analysis
how should a comparison be carried out in order to enhance or improve description, prediction, theory-testing and classification?
ISSUES IN COMPARATIVE POLITICS
Methods in Comparative Politics
Lijphart (1971): 3 main methods CP
Case Study Method
= intensive examination of one particular case and the context in which it exists
comparative component - some don’t have a comparative component -> offering context (e.g. saying it is an example of…)
(e.g. electoral performance populist party France last election, if you don’t say why you study this specific case it’s not really comparative)
Deviant Case Study = identifies and examines an exception to what is generally expected from an established theory
Theory-Testing Case Study = probing a theory in a new empirical context to which it is supposed to apply
Comparative Method (Small-N)
= a systematic analysis of a small number of cases (small-N analysis) (= interpretative + not numerically measured indexes)
how many cases? where does this type of method end (when is it large-N)?
= very valuable tool to further our understanding of certain phenomena (robust understanding)
Problems:
- Too Few Cases, Too Many Variables -> Increase N (risk of conceptual stretching: using cases that don’t fit) + MSSD
(there are always other factors that may affect X (e.g. eco success small countries Europe)
- Selection on Dependent Variable -> MDSD
e.g. Katzenstein: Small States in World Markets: Why are (7) small European countries more successful economically than their larger and economically more powerful neighbors?
- they have adopted corporarist model of policy-making
Statistical Method (large-N)
= comparative method based on a large N of cases, using statistical techniques to examine relationships between variables
main differences with small-N =
advantage = if you have the data, you may reach very robust/large conclusions
e.g. relation corruption and human development -> correlation line between these levels
Conclusion
main branches polsci
classifications and typologies
clustering differences into different classes of phenomena according to common attributes that they share
CP enables this
we would we do this?
classifications are the first (descriptive) step in arriving at some sort of explanations
- without having good classifications we will not be able to explain properly certain phenomena (e.g. democratic vs non-democratic you need this distinction to e.g. answer if eco. development correlates with democratization)
unit of analysis vs level of analysis?
deviation from the book: they do
Deviant case study
identifies and examines an exception to what is generally expected from an established theory
Theory-Testing Case Study
= probing a theory in a new empirical context to which it is supposed to apply
e.g. Lijphart: Power-Sharing in South Africa post-1990
- does power-sharing in South Africa work in the same way as it did in the Netherlands pre-1960 to sustain democracy
MSSD
most similar system design
select for cases that you assume are similar on almost all variables that may influence what you want to explain
e.g. used in Katzenstein’s study of 7 small economically successful European states
in this way you account for alternative explanations
! we see this most often (e.g. often look at democratic European states, as we assume they are equal)
MDSD
most dissimilar system design
select one factor that is common in systems that are overall really different
e.g. Katzenstein did wrong: it’s not …, it’s something else (Katzenstein didn’t look at small economic successful states that don’t have …)