Dr Phil Larkin philip.larkinanu.edu.au

1
Dr Phil Larkinphilip.larkin_at_anu.edu.au

• Comparative Politics What? Why? How?

2
Outline

• What is comparative politics
• Why bother?
• How to do it?

3
What is it?

• Subfield of political science
• Defined by its method rather than it's
  substanceMethod the act of comparison rather
  than statistics
• 'The term 'comparative politics' indicates the
  how but does not specify the what of the
  analysis' Arend Lijphart, 'Comparative Politics
  and Comparative Method', APSR, Vol.65(3)
• No agreement on how to compare

4
What is it? II

• Numeous attempts to define the field more
  narrowly. Minimalist definition'The
  presentation of empirical evidence of some kind
  to compare systematically and explictly empirical
  phenomena' Richard Rose, 'Comparing Forms of
  Comparative Analysis', Political Studies, Vol.39

5
What is it? III

• Empirical evidence
• Logical positivism discovering scientific laws?
• Ordering to aid comprehension tendencies or
  regularities
• Comparison 'systematic'
• theory guiding case selection and mode of
  comparison
• Single country comparison diachronic vs
  synchronic comparison

6
Why compare?

• It's hard so why bother?
• Escaping 'ethnocentrism'
• Replicating the laboratory in social science?
• Interrogating theory

7
Escaping 'ethnocentrism'

• 'An age-old idea of philosophers is that
  knowledge of thte self is gained through
  knowledge of others....What is true for the
  individual is even more so for societies. There
  is no nation wthout other nations. This
  diversity...is the only element that permits the
  perception of people and systems' Dogan and
  Pelassy, How to Compare Nations, p.5
• Exceptionalism or uniqueness of individual cases
  as explanation

8
Escaping 'ethnocentrism' II

• 'The tradition of writing about British (or, more
  properly, English) politics is to assert
  uniqueness through false particularisation. Every
  institution, individual, and event is described
  with nominal differences implying the absence of
  generic qualities' Richard Rose, 'Comparing Forms
  of Comparative Analysis', Political Studies,
  Vol.39
• Starting point determines explanation
• Danger or ethnocentric comparison

9
Replicating the laboratory in social sciences?

• Natural sciences laboratory conditions allow
  control of variables to eliminate them from
  explanation
• Comparison as a means of controlling variables?
  Narrow the range of indpendent variables that can
  account for the dependent variable. (ie. the
  factors that account for the phenomenon we're
  trying to explain).

10
Interrogating theory

• Can't repeat experiments
• Comparison increases number of cases improves
  liklihood of being right
• Explanation refined as cases number of increases

11
How to compare?

• Comparing like with like
• Most Similar Systems Design
• Comparing contrasting cases
• Most Different Systems Design
• Comparing lots of cases
• Large n comparisons
• Comparing a few cases
• Small n comparisons

12
Comparing like with like

• Most Similar Systems Design (MSSD) comparing
  similar cases that differ in respect of the
  dependent variable
• Replicating the laboratory similar independent
  variables 'controlled for'.
• Most common form of comparison

13
Comparing contrasting cases

• Most Different Systems Design (MDSD) comparing
  cases that differ except in relation to the
  dependent variable
• Again, attempt to control independent variables
  see relationship between independent and
  indpendent variables under very different
  circumstances
• Theda Skocpol, States Social Revolutions
  (1979) Revolutions in France, Russia and China
  compared. Different cases wht did they share?
  What can we infer about revolutions as a result
  of the comparison?

14
Comparing a lot

• Large n
• Large number of cases, more robust a the
  relationship between independent and dependent
  variables
• Scale tends to lead to quantitative research
• MDSD lends itself to large n due to focus on
  variables rather than context (though not always
  Skocpol)
• Too abstract to be useful? Doesn't tell us about
  cause stops where the politics starts?
• Too many variables, too few cases? (log
  positivism)

15
Comparing a few

• Small n allows more detailed study
• Lends itself to MSSD
• limit of available cases
• necessary detail means limits on cases
• Scale for detail danger of 'small n, big
  conclusions'?

16
Conclusion

• Differences in ways of doing it
  compare/contrast large/small does it matter?
• Single cases can be comparative
• Mutually beneficial!