Slides

The Comparative Method: Concepts, Theory, and Cases

Comparative Politics

  • Definition by Howard Wiarda:
    "the systematic, comparative study of the world's political systems"

Systematic

  • Eckstein's View:
    "Comparison in the social sciences is the substitute for experimentation in the natural sciences"

  • Importance of control:

    • Lijphart outlines key aspects:

    • a) Experimental Approach:

      • Unit homogeneity

      • Random assignment

      • Control of treatment

    • b) Statistical Approach:

      • Utilizes partial correlation and random sampling

    • c) Comparative Structured Study:

      • Focus on systematic study of similarities and differences across carefully selected cases

    • d) Case Study:

      • Detailed case investigation, often implicitly comparative

Comparative

Structuring Comparisons

  • Two primary methods to structure comparisons:

    1. Most Similar Systems Design (MSSD)

    • Focuses on cases that are similar in most aspects but differ in one important regard

    • Key Components:

      • Different outcomes: Y and Not-Y (dependent variable)

      • Common systemic features allow isolation of what separates the cases (independent variable)

      • Required covariation of independent and dependent variables

      • The distinguishing feature must be explanatory, necessitating a compelling causal story

    1. Most Different Systems Design (MDSD)

    • Focuses on cases that have different systemic features but yield similar outcomes

    • Key Components:

      • Common outcome: Y and Y

      • Ability to isolate what unites the cases through differing systemic features

      • Establishing covariation between independent variable(s) and outcomes

      • Again, the factor must be explanatory and backed with compelling causal reasoning

Illustrations of MSSD and MDSD

  • MSSD Example:
    The Rise of Christian Democracy in Europe (Stathis Kalyvas, 1996):

    • France fails to develop a durable Christian Democratic Party despite similarities in Catholicism and religiosity with other countries.

    • The differentiating factor is political instability that discouraged the Catholic Church from forming a competitive party.

  • MDSD Example:
    Between a Shock and a Hard Place: The Dynamics of Labor-Backed Adjustment in Poland and Argentina (Steven Levitsky & Lucan Way, 1998):

    • Successful neoliberal policies in both Poland and Argentina during the 1990s despite strong labor movements.

    • Contributing factors: Labor-left party alliances sustained in Argentina longer due to stronger party structure and incentives for labor leaders.

Comparative - Criticisms

  • Deterministic Bias:

    • Perceived as deterministic in a probabilistic age.

  • Operationalization Issues:

    • Measurement error as a significant challenge.

  • Degrees of Freedom:

    • Insufficient number of cases to establish causation.

  • Case Independence:

    • Cases may not be sufficiently independent.

  • Explanatory Factors:

    • Encouragement to not fully abandon rejected explanatory factors, while recognizing it may be a larger issue in statistical research.

  • Lijphart's Recommendations:

    • Increase the number of cases for better detection of patterns, particularly important for small N.

    • Focus on interdependence, including aspects of diffusion and contagion.

Comparative - Research Methodology

  • Mixing Methods:

    • Comparative research can utilize both statistical and case study approaches (i.e., Carles Boix’s Democracy and Redistribution).

  • Case studies to corroborate statistical findings:

    • Initial broad statistical relationships explored then tested through detailed case studies (e.g., US states and Swiss cantons).

Choosing the Right Method

  • Determined by:

    1. Research question

    2. Concept formation

    3. Theory

  • Good research questions refine focus, save time, and offer clear research direction.

  • Example Questions from Kalyvas:

    • Why do political parties develop?

    • Why specifically do Christian Democratic Parties develop?

    • What forms do they take?

    • Why did France not develop a durable Christian Democratic Party?

Concept Formation

  • Abstract vs. Concrete:

    • Concepts range from abstract (i.e., "fruit") to concrete (i.e., "The Braeburn apple I had for breakfast").

  • Creation of Concepts:

    • They should be clear, coherent, and helpful to the research question.

  • Ladder of Abstraction:

    • As concepts become more abstract, they become less grounded in detail.

  • Concept Stretching:

    • If too abstract, concepts can stretch beyond usefulness, affecting the number of available cases.

Study: Concepts Are Crucial

  • Reference to Theda Skocpol, States and Social Revolutions (1979):

    • Examines the definition of a "social revolution" as not merely collective actions but violent political upheavals that include transformation of economic and social power.

    • Aims to generalize based on limited cases: France, Russia, China.

Theory

  • Definition:

    • Meaningful statements about the relationships between phenomena (e.g., change in X affects a change in Y).

  • Role of Theory:

    1. Source of questions, puzzles, predictions, and anomalies.

    2. Basis for framing concepts (as exemplified by Skocpol).

    3. Guidance for starting research.

    4. Assistance in interpreting findings and drawing implications.

Theoretical Approaches

  • Broad Perspectives at Various Levels:

    1. Rationality:

    • Focuses on self-interest and maximizing highest-ranked preferences.

    1. Culture:

    • Considers values, beliefs, ideas, and identities.

    1. Socioeconomic Structuralism:

    • Material conditions shape interests and constrain actors, emphasizing class.

    1. Institutionalism:

    • Examines how both formal and informal rules shape interests and actions.

Comparative: Defining Cases

  • What is a Case?

    • Hall’s Definition:

    • “Units in which the relevant outcome takes on a specific value.”

  • Key Question:

    • What exactly is it a case of?

  • Constructing Cases:

    • Cases must be constructed, which can include states, movements, or episodes.

  • Single-Case Studies:

    • Offer great detail and more than just variable observations.

    • Are powerful when well-selected (e.g., crucial or deviant cases).

Systematic Process Analysis

  • Also referred to as process tracing:

    • This method serves as an alternative to correlation-based approaches including comparative methods:

    • Focus on key explanatory theories to draw out complete predictions.

    • Identify relevant observations per these predictions, treating each case as a rich source of information.

    • Test the manifestation of the entire causal chain.

    • Set rival theories against one another in this systematic fashion.