Questions: Comparative Politics: Method and Approach

Question 1

A researcher wants to explain why Chile and Argentina both democratized in the 1980s despite having very different economies, cultures, and geographic sizes. Which comparative design does this illustrate?

• A. Most-similar systems design (MSSD), because both countries are in South America
• B. Most-different systems design (MDSD), because the cases differ on many variables but share the outcome
• C. A natural experiment, because the timing of democratization was determined by an external shock
• D. A case study, because only two countries are examined

Answer: (B)

MDSD selects cases that differ on many background characteristics but share the outcome of interest. If Chile and Argentina democratized despite being very different in other respects, the researcher infers that something common to both — not the many differences — explains the outcome. MSSD would instead hold many variables constant to isolate one. Two-country comparisons can implement either design; it is the logic of selection that defines the design.

Question 2

Comparative politics is essentially just describing political systems in multiple countries side by side.

• T. True
• F. False

Answer: False

Comparative politics is a method for causal inference, not mere description. Describing Germany's parliament and Japan's parliament side by side is journalism or area studies. Comparative politics asks: given variation in institutional design, what explains differences in outcomes? This requires a research design that systematically controls for confounding factors — the logic of controlled comparison, whether qualitative or quantitative.

Question 3

When would a researcher prefer the most-different systems design (MDSD) over the most-similar systems design (MSSD)?

Model answer: MDSD is preferred when a researcher wants to establish that a common factor explains an outcome across very different contexts, strengthening external validity by showing the finding is not limited to a specific regional or cultural setting. MSSD is preferred when contextual similarity allows tighter control of confounders to isolate a single causal variable.

The two designs address different inferential problems. MSSD maximizes internal validity by holding context constant; MDSD maximizes external validity by showing robustness across diverse contexts. Choosing between them depends on whether the goal is to isolate a causal mechanism or to demonstrate its generalizability.