Method Causality

Published on April 24th, 2013


Determining Causality

By Tor G. Jakobsen

An essential question for students of the social sciences is: «How do we know that X is the cause of Y, and not vice versa?” One way to solve this question is by the use of sound theory that tells us something about the direction of causality. However, sometimes this is not enough. This article will help you demonstrate that X causes Y

Figure from Kellstedt & Whitten (2009)

According to Professors Paul M. Kellstedt and Guy D. Whitten at the Texas A&M University there are four hurdles of causality that the social science student needs to take into consideration.

  1. Is there a credible causal mechanism that connects X to Y? In other words, what is it specifically about having more or less of X that will lead to more or less of Y.
  2. Could it be that Y cause X? This will often be the case, as some explanatory power often goes both ways.
  3. Is there a covariation between X and Y? Correlation is an essential component of causality, there must be some form of measurable association between these two variables. However, be aware that correlation alone does not prove causality. Sometimes the correlation may be hidden (it could be a interaction effect, or it does not show until we have controlled for another important variable).
  4. Could there be some confounding variable Z that is related to both X and Y, giving us a spurious observed link between X and Y? A historic example of this is the link between number of storks and babies in the Danish countryside, where actually both the number of storks and babies could be explained by industrialization.


The correlation between institutions and economic performance

It is difficult to determine the causal link between good institutions and a country’s economic performance. Countries with better institutions and more secure property rights, will invest more in physical and human capital, and will use these factors more efficiently to achieve a greater level of income. This view receives some support from cross-national correlations between measures of property rights and economic development.

Yet, it is also plausible that rich economies are better able to afford good institutions. That is, the direction of causality could go the other way.

In a study by Daron Acemoglu, Simon Johnson, and James A. Robinson the authors employed an instrumental variable to answer this riddle. More specifically, they used differences in colonial European settler mortality rates to estimate the effect of institutions on economic performance. Settler mortality rates were seen as a source of exogenous variation in institutions.

There were different types of European policies of colonization, and these gave birth to either good or bad institutions. At one extreme we had the Extractive States (Belgian Congo being one example), and the other one was Neo-Europes (Australia, New Zealand, South Africa, and the United States) where the settlers tried to replicate the European institutions, focusing on private property rights.

So, the colonization strategy was influenced by the feasibility of settlements. In places where the disease environment was not favorable to European settlement, the cards were stacked against the creation of Neo-Europes, and the formation of the extractive state was more likely. The institutions persisted even after independent.


Mortality rates of first settlers as an instrument for current institutions

Acemoglu, Johnson, and Robinson used data on the mortality rates of soldiers, bishops, and sailors stationed in the colonies between the seventeenth and nineteenth centuries. The authors exploited these differences as a source of exogenous variation to estimate the impact of institutions on economic performance.

The death rate of settlers is an instrument. One cannot say whether institutions influence the economy, or if the economy influence institutions. However, we can be sure that today’s economic performance does not influence the settler mortality rates on colonial times. So, if the mortality rates are correlated with economic performance, then we can assume that the causality goes from institutions to economic performance.

Figure from Acemoglu, Johnson, & Robinson (2001)

There is a high correlation between mortality rates faced by soldiers, bishops, and sailors in the colonies and European settlements; between European settlements and early measures of institutions; and between early institutions and institutions today.


Further reading:

Acemoglu, Daron, Simon Johnson, & James A. Robinson (2001) “The Colonial Origins of Comparative Development: An Empirical Investigation” American Economic Review, 91(5): 1369–1401.

Kellstedt, Paul M. & Guy D. Whitten (2009) “Evaluating Causal Relationships” in Paul M. Kellstedt & Guy D. Whitten The Fundamentals of Political Science research. New York: Cambridge University Press: 45–66.


*Cover photo by the Field Museum Library


Tags: ,

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Back to Top ↑