The Comparative Method: Concepts, Theory, and Cases
January 19th, 2023
Comparative Politics
● Howard Wiarda: comparative politics is “the systematic, comparative study of the world’s
political systems”
Systematic
● Eckstein: comparison in the social sciences is the substitute for experimentation in the natural
sciences
● Lijphart:
○ Experimental Method
■ Unit homogeneity
■ Random assignment (no systematic differences = maximize similarity in
selection or do random assignment)
■ Control of the treatment
○ Statistical Method
■ Partial correlation (breaking the broad group into subgroups, e.g. age is a factor
in a drug group, you breakdown the broad group and subgroup ages)
■ Random sampling (key to randomizing, maximizing unit homogeneity)
○ Comparative Method
■ Structured study of similarities and differences actress careful selected cases
■ Careful case selection
○ Case Study
■ Every case study is implicitly comparative
■ Detailed case investigation, at least implicitly comparative
■ E.g. studying a case of democracy in Canada and North Korea, and comparing
them
Comparative
● Two ways of structuring comparisons: (each way has two names)
○ Most similar systems designs/method of difference
■ Different outcome: Y and Not-Y (dependant variable)
■ Common systematic features permit the isolation of that which separates the
cases (independent variable)
■ Covariation of independent and dependent variables
■ Distinguishing feature must be explanatory — must come with a compelling
causal explanation (correlation is not causation)
MSSD: Illustration
E.g. Stathis Kalyvas, Christian Democracy in Europe
● France does not develop a durable Christian Democratic Party — other similar countries do
● France does not differ in catholicism, religiously
● But in its level of political instability, which discouraged the Catholic church from forming its
own competitive democratic party
● Most Different Systems Design (Method of Agreement)
○ Common outcome: Y and Y (same thing happens in both cases)
○ Different systematic features permit the isolation of what unites the cases
○ Establishing covariation between independent variable
○ This factor must be explanatory, it must come with a compelling casual story
MDSD: Illustration
● Levitsky and Way’s article
● Successful implementation of neoliberal adjustments politics in Argentina and Poland which are
two very different countries in the 1990s
● Factor that unites the two cases: left parties sharing social linkages, mutual trust, and identities
with labor movements
Comparative – Criticism
● Deterministic in a probabilistic age:
○ If X then Y
● Don’t entirely abandon rejected explanatory factors
● Operationalization: measurement error of a problem
● Yes, but may be a bigger problem for statistical research
● Degrees of freedom — not enough cases to settle matters
● Lijphart: try to increase the number of cases
● Cases are insufficiently
● This justifies increasing attention to interdependence
○ E.g. Diffusion and contagion
Comparative
● Comparative research can also combine statistical and single case study approach: mix and match
● E.g. Carles Boix, Democracy and Redistribution
○ Boix pursue broad statistical relationship between democracy and inequality
○ Then, Boix tests it in detailed comparative case studies of US states and Swiss Cantons
Comparative — The Right Method
● Choice of method depends on:
○ Research question
○ Concept Formation
○ Theory
Research Questions
● Good questions save time, narrowing the research focus and providing direction
● E.g. Why do political parties develop at all? Why do Christian Democratic Parties develop? What
form do they take? Why did France not develop a durable Christian Democreactive Problem?
Concept Formation
● We create concepts
● They should be clear and coherent, and helpful in addressing our research question
● Use the Ladder of Abstraction: the more abstract, the less grounded in detail
● Concept stretching: concept is too abstract
● Concept influence the number of available cases
Meaningful Study — Concepts are Crucial
○ Skocpol examines: What is a social revolution?
○ Not merely a political revolution — large set
○ But violent political overturn combines with transformation of economic and social
power — small set
○ Skocpal aims to generalize, but only based on her limited set of cases: France, Russia,
China
Theory
● Meaningful statements about the relationship between phenomena — e.g. change in X effects a
change in Y
● Captures the way the world works at some level of generality (e.g. this is what happened in this
country at this time because that’s how the world works)
● Generates predictions on how the world should be, when the world is not what it should be it
becomes an anomaly that is worth studying
● Role of theory:
○ Source of questions and puzzles, predictions and abnormality (Generates predictions on
how the world should be, when the world is not what it should be it becomes an anomaly
that is worth studying)
○ Basis for framing concepts — as in Skocpol
○ Guidance in pursuing research — where do we start?
○ Assistance in interpreting findings, drawing implications
Meaningful Study — Theoretical Approach
● Broad theoretical perspectives, operative at variations levels (individuals, group, region, nation-
state, international)
● Rationality — self-interest, maximize ranked and preferences (being re-elected in a position)
● Culture — values, beliefs, ideas, and identities
● Socio-Economic Structuralism — material conditions shape interests, constrain actors, which an
emphasis on class
● Institutionalism — human beings are influenced and shaped by the institutions that surround them
(the rules). We are surrounded by formal and informal rules that shape us and our actions.
Comparative Cases
What is a case?
● Hall: A unit is something we assign value to
● Key Question: what is a case of?
○ Cases must be constructed: states, movements, episodes in history (e.g. Germany is a
case of post-world war II democracy)
● Single-case study:
○ Great detail possible
○ More than just variables
○ Powerful if well selected, as Lijphert says (e.g. crucial case is a case that stand out as an
abnormality)
Systematic Process Analysis
● AKA process tracking
● Alternation to correlation-based approaches, including comparative method
○ Emphasize key explanatory theories (what is the debate?)
○ Draw out their predictions in full
○ Identify observations relevant to predictions — each case is a rich source
○ Test: does the entire casual chain manifest
○ Set rival theories at odds in this fashion
○ A case is filled with variety of observations in line or not in line with your theory