The Causality Compass

Lesson 4: Granger Tests and Temporal Lag Analysis

RESEARCHER:

Objective

History is a path-dependent process. In this activity, you will determine the direction of causality between two historical variables and identify appropriate lag structures based on historical context.

1. Directionality Dilemma

"In the 17th-century Dutch Republic, there is a strong correlation between **maritime trade volume** and **scientific publications in optics**."

Hypothesis A (A → B)

Trade wealth funded scientific academies and created demand for navigation lenses.

Hypothesis B (B → A)

Advancements in optics (telescopes) improved navigation, leading to safer/more trade.

If you ran a Granger Causality test and found that Science in Year \(T\) predicts Trade in Year \(T+5\), which hypothesis is supported?

2. Identifying Historical Lags

For each of the following scenarios, identify the causal lag (in years/decades) and explain the historical mechanism behind the delay.

Scenario: Public Health Investment

Variable X: Massive state investment in clean water infrastructure in London (1850).
Variable Y: Increase in average adult life expectancy.

Estimated Lag:

(e.g., 5 years, 20 years, etc.)

Mechanism for Delay:

Scenario: Colonial Taxation

Variable X: A sharp increase in tax rates in a Spanish colony (1780).
Variable Y: Recorded tax revenue in the Seville archives.

Estimated Lag:

(e.g., 1 year, 3 years, etc.)

Mechanism for Delay (Consider bureaucracy/travel):

Critique of Granger Logic

Granger causality is based on prediction, not "true" physical causation. If people begin hoarding food today because they expect a famine tomorrow, the "Hoarding" (Variable A) will Granger-cause the "Famine" (Variable B) because it happens first.

How should a historian account for "Forward-Looking Expectations" when interpreting Granger results?

Causality Compass Worksheet

Causality Compass Worksheet