2.3: Case Selection

Learning Objectives

By the end of this section, you will be able to:

• Discuss the importance of case selection in case studies.
• Consider the implications of poor case selection.

Introduction

Case selection is an important part of any research design. Deciding how many cases to include--and which cases to include--determines the appropriate research analysis methods. For instance, if a researcher decides to select a large number of cases (e.g., a survey of 1500 individuals), they will use statistical methods to conduct large-N analysis, as described in the previous section. In many facets of comparative politics, particularly if using the country as the unit of analysis, researchers rarely reach the number of cases needed for large-N analysis. If the focus is on one case, or a few cases, what strategies should researchers use to choose those cases?

A number of factors can drive case selection in comparative politics. First, case selection can derive from the interests of the researcher(s). For example, if a researcher lives in Germany, they may want to research a topic within that country, possibly using a subnational approach to compare an outcome among German states. Second, case selection can derive from an area studies-focus (e.g., a desire to understand a topic within the member-states of the European Union). Third, researchers may choose cases to facilitate a particular type of case study. Below, we discuss case selection among various types of case studies.

Types of Case Studies

As John Gerring (2017) categorizes, case studies can be descriptive or causal. Descriptive case studies are "not organized around a central, overarching causal hypothesis or theory" (p. 56). In a descriptive case study, a researcher simply seeks to describe what they observe. Suppose, for instance, that a researcher wants to describe how political institutions function in Switzerland. The researcher may discuss Switzerland's federalist system, how the Swiss use elements of direct democracy, or how the presidency rotates annually. The goal in such a study is not to make a claim about why Switzerland has a federalist system that uses direct democracy and a rotating presidency; the goal is simply to describe what exists. 

Depending on the research question, a researcher may choose cases that are typical or cases that offer diversity. Suppose, for instance, that a researcher is investigating the effects of the pandemic on medium-sized cities within the United States. If choosing a typical case, the researcher will want to find a medium-size city that exhibits key characteristics of medium-sized cities throughout the entire country. This will require the researcher to define "medium-size city" and establish the important characteristics of medium-sized cities. With those definitions, the researcher can work to identify a typical case. If choosing cases that offer diversity, the researcher may decide to examine a range of medium-size cities that vary on some dimension, such as geographical location.

Unlike descriptive case studies, causal case studies are "organized around a central hypothesis about how X affects Y" (Gerring, 2017). In a causal case study, the researcher identifies a causal mechanism, which "explain[s] how and why a hypothesized cause, in a given context, contributes to a particular outcome" (Falleti and Lynch, 2009). Case selection in causal case studies may follow an exploratory, estimating, or diagnostic approach, depending on how the researcher aims to develop a hypothesis.

Researchers use exploratory case studies to identify potential causal hypotheses. They single out the independent variables that seem to affect the outcome, or dependent variable, the most. The case study provides context that allows the researcher to assert a potential causal mechanism. If the scholar is looking to develop an "ideal-type" through their case study, they might seek out an extreme case. For example, if the researcher aims to understand the ideal-type capitalist system, they may want to investigate a country that practices a pure, or extreme, form of the economic system.

Researchers use estimating case studies when they already have a hypothesis in place. They want to use the case study to gather evidence and test the hypothesis. The appropriate case to use therefore depends on the existing hypothesis.

Researchers use diagnostic case studies to "confirm, disconfirm, or refine a hypothesis" (Gerring 2017). In a diagnostic case study, a researcher may choose a least-likely case, or a case that provides surprising evidence against the hypothesis. For example, India's high level of ethno-linguistic diversity, its relatively underdeveloped economy, and its low level of modernization through large swaths of the country suggest that it should not have democratized, or it should have failed to remain a democracy. In reality, India has been a democracy for over 70 years, making it at "least-likely" case that can provide evidence against a hypothesis about a necessary relationship between economic development and democracy.

Methods of Agreement and Difference

The discussion above tends to focus on case selection when it comes to a single case. When conducting comparative case studies, as we do in comparative politics, we must select more than one case. This presents a set of challenges. First, how many cases do we pick? Second, how do we apply the previously mentioned case selection techniques? Do we choose typical, diverse, extreme, or least-likely cases?

John Stuart Mill, an English scholar (1806-1873), developed several approaches to comparison with the explicit goal of isolating a cause within a complex environment. Two of these methods, the "method of agreement" and the "method of difference," are particularly important in comparative politics. In the "method of agreement," a researcher compares two or more cases for their commonalities. The researcher aims to isolate the characteristic, or variable, that the cases have in common (the reason for their similarities). In the "method of difference," a researcher compares two or more cases for their differences. The researcher aims to isolate the characteristic, or variable, that the cases do not have in common (the reason for their differences). From these two methods, comparativists developed two approaches.

Figure \(\PageIndex{1}\): Book cover of A System of Logic, Ratiocinative and Inductive. John Stuart Mill developed several approaches to comparison. (Source: Mill, J.S. (1843). A System of Logic, Ratiocinative and Inductive. University of Toronto Press.)

Most Similar Systems Design

The Most Similar Systems Design (MSSD) stems from Mill's "method of difference." In a MSSD, researchers select cases for comparison that are similar to each other, but in which the outcomes differ. In this approach, a researcher aims to ensure that as many of the independent variables as possible are similar across cases. While the researcher does not have randomization to rely on, as in an experiment, this case selection method attempts to "control" for as many observable confounding variables as possible. A researcher asks: If these cases look similar on variables A, B, C, D, and E, why do they have different values on the dependent variable? The researcher then aims to identity the potential cause for the difference in outcomes. 

For example, if a researcher wants to study the existence of publicly-accessible national health systems as the dependent variable, they may choose to compare Canada and the United States. Both of these countries have English heritage, English language use, liberal market economies, strong democratic institutions, high levels of wealth, and high levels of education. Yet, despite these similarities, the outcome differs: Canada has a robust publicly-accessible national health system; the U.S. does not. A MSSD design would aim to describe why these two seemingly similar cases differ on the outcome (e.g., one explanation may relate to varying political culture within Canada and the U.S.).

Most Different Systems Design

The Most Different Systems Design (MDSD) stems from Mill's "method of agreement." In a MDSD, researchers select cases for comparison that are different from each other, but in which the outcome is the same. In this approach, a researcher aims to select cases with the same value on the dependent variable, but that are quite different from one another in terms of other variables.

As an example, a researcher may be interested in learning how different countries reach the same classification of "economically liberal." In its Index of Economic Freedom, The Heritage Foundation lists countries as varied as Australia, Chile, Estonia, Malaysia, New Zealand, Singapore, Switzerland, and Taiwan as either free or mostly free. Yet these countries differ greatly from one another. Singapore is a non-democratic regime, whereas Estonia is a democratic regime. Australia and New Zealand are wealthy, whereas Malaysia is not. Chile and Taiwan became economically free countries under authoritarian military regimes, whereas Switzerland did not. A researcher will ask: Why do these different systems produce the same outcome?