Questions: Comparative Historical Analysis Across Cases and Scales

Question 1

A historian wants to understand why two neighboring countries — similar in economy, state structure, and social tensions — had different political outcomes: Country A had a revolution while Country B did not. Which research design is most appropriate?

• A. Most-different systems design — the historian should find countries that are as different as possible to maximize contrast
• B. Most-similar systems design — the similarity controls for many background variables, making the differences between the cases candidates for explaining the divergent outcome
• C. Statistical analysis of hundreds of countries, because two cases cannot support any causal claim
• D. Single-case study of Country A, because the goal is to explain the revolution, not compare

Answer: (B)

Most-similar systems design examines cases that resemble each other in most respects but differ in outcome. The similarity 'controls for' background variables: if France and England shared similar economies, state structures, and social tensions but only France had revolution in 1789, then the differences between them become candidates for explaining the divergent outcome. A most-different design would ask the opposite question: what do very different cases have in common that might explain a shared outcome.

Question 2

A historian studying peasant revolutions finds that very different societies — with different cultures, economies, and political systems — all had revolutions when they shared one structural feature: weakened landlord authority combined with a fiscal-military crisis. This research design is:

• A. Most-similar systems design, because the revolution is the common outcome across cases
• B. Flawed, because comparing across such different societies produces meaningless results
• C. Most-different systems design — the differences between cases are controlled for by their variety; the shared structural feature despite those differences becomes a powerful causal candidate
• D. Experimental, because the fiscal-military crisis variable was deliberately manipulated

Answer: (C)

Most-different systems design examines cases that vary widely in most respects but share the same outcome. If very different societies all experienced revolutions and all shared a specific structural feature, that shared feature — despite the surrounding differences — becomes a strong causal candidate. The logic: if outcome Y appears across very different contexts whenever condition X is present, then X is doing causal work. This is the design Theda Skocpol used in States and Social Revolutions.

Question 3

The primary goal of comparative historical analysis is to identify universal laws about society that apply across most contexts and time periods.

• T. True
• F. False

Answer: False

Comparative historians aim to identify mechanisms that operated across specific cases — not universal laws. Works like Barrington Moore's Social Origins of Dictatorship and Democracy or Skocpol's States and Social Revolutions make claims about structural conditions that shaped outcomes in the cases examined, while acknowledging that context matters and that their findings are not timeless generalizations. Comparison is the closest history gets to experimental control, but it produces middle-range explanations, not universal laws.

Question 4

In most-similar systems design, the key analytical move is to examine what the similar cases have in common — since shared features are the strongest causal candidates for the shared outcome.

• T. True
• F. False

Answer: False

This gets the logic backwards. In most-similar systems design, you are studying cases with a similar background but different outcomes. The shared features are 'controlled for' — they cannot explain the divergent outcomes because they are present in both cases. It is the differences between otherwise similar cases that become the causal candidates. The logic is: similar in most respects but different in outcome → the differences must explain why outcomes diverged. (Most-different systems design works the reverse way: different cases, same outcome → shared features are the causal candidates.)

Question 5

Why is comparative historical analysis described as the discipline's 'closest approximation to experimental method'? What can comparison control for that single-case study cannot?

Model answer: Historians cannot rerun events with variables changed — history happens once. A single case study can describe what occurred but cannot isolate which factors caused the outcome, because everything is changing simultaneously. Comparison introduces a form of quasi-experimental control: by choosing cases that are similar in many respects, the researcher holds background variables roughly constant and focuses attention on what differs. This is analogous to a controlled experiment where the researcher varies one factor while holding others fixed. What comparison controls for — through design rather than literal manipulation — is the alternative explanations that background similarity rules out.

The limits of this comparison-as-control are real: social cases are never perfectly matched, hidden confounding variables persist, and context always matters. This is why comparative historians speak of 'mechanisms' and 'structural conditions' rather than proofs. But even imperfect comparison generates stronger causal inference than single-case description alone, which is why the comparative method has been central to historical sociology since Weber.