Case Studies in Political Science Research Paper

II. The Debate Within the Discipline

III. Examples of the Case Study Approach

A. U.S. Politics

B. Comparative Politics

C. International Relations

IV. Conclusion

I. Introduction

The case study method has always been an integral tool in the investigation of social science phenomena, being of particular value when the number of observations, or cases studied, is limited in number, restricting the utility of statistical approaches. However, for some time the individual case study approach had been supplanted by large-N, data-intensive quantitative methods as the preferred technique for empirical studies. More recently, the case study has seen a revival of interest by social scientists as part of a multimethod, holistic approach that includes formal, qualitative, and quantitative methods. Indeed, each major methodological approach plays an important role in the research cycle, with the qualitative application of the case study enlightening the inductive aspect of theory development through the identification of alternate causal explanations, new variables, or complex interactions of variables. Fundamentally, case studies allow one to go beyond often simplistic quantitative analysis and develop contextually rich and in-depth pictures of the phenomena being observed.

By itself, a case study is the history of an event, be it of short or long duration—a civil protest movement, for example, or the evolutionary process from colonial rule toward stable democracy. As such, a case study identifies the expected, predictable aspects of an event, while ideally it also captures additional but less quantifiable detail, such as the cultural context, that potentially asserts a causal role as well. Individual or comparative case studies of specific, individual events, actors, or systems allow the researcher to obtain a depth of knowledge and understanding about the object being studied that large-N quantitative studies fail to provide.

A carefully crafted case study serves several purposes within the research cycle. First, while quantitative studies identify outlying or deviant cases, those well beyond the expected normal distribution, quantitative methods are generally not able to explain the specific reasons for a particular case’s extreme variation from its population mean. The case study, however, not only provides the opportunity to identify likely reasons for these individual deviations but may illuminate previously unidentified causal variables and possible alternate explanations as well. This information potentially leads to the extension of existing theory, if not its revision, and may suggest new theoretical explanations altogether.

Additionally, the case study may be the best, or only, way to study certain phenomena because of the relatively small number of identified cases and a resulting scarcity of data, which restricts the use of quantitative methods. And while much of the earliest criticism of case studies (by social scientists) centered on their application as a mainly historical narrative, the substantive purpose of case study is to understand that history but to do so in a way that allows for the identification of critical actions, structures, or other aspects that contribute to the end result. Being able to examine with scientific rigor phenomena that either do not lend themselves well to quantitative study, or for which only a limited set of objective measures is available, makes such an approach valuable. The role case studies can play in identifying and understanding previously unknown variables and in establishing causal paths and the interdependency of variables, as well as being critical tests of existing theory, makes them not just a complement to quantitative methods but potentially of equal value (Geddes, 1990; Gerring, 2004).

Case studies are by definition qualitative, meaning that the focus of the study is not primarily the systematic manipulation of aggregated points of data, an objective exercise, but rather a study that focuses on the quality of the potential data observed, a much more subjective work. This is not to say that case studies are not objective as well: In reality, for a case study to have any influence, it must identify and measure variables to allow for reliable comparison and to build theory that is testable, replicable, and generalizable. Case study is ultimately a method that falls into two forms: the individual, within case study and the comparative across case study, usually limited to a small number of cases. Both types work to identify causal relationships and enlighten theoretical explanations. Good case study work can be either accumulating (building on previous knowledge) or original (establishing entirely new avenues of research).

II. The Debate Within the Discipline

Political scientists have had an ongoing discussion about the role of the case study approach in their field. This discussion has focused on the relative value of case study compared with other methods for evaluating and advancing theoretical understanding. Of central concern is the perceived methodological limitation of single and small-N case work within a discipline that favors quantitative methodologies. A tension results between the benefits accrued from this method and its limitations. What value can a unique examination contribute? Are hypotheses and theory valid only if they are testable and generalizable? Within these debates over the fundamental usefulness of deliberative case study work are questions that address both the inherent strengths and weaknesses of such an approach. Scholars have generally fallen into two camps, those who argue for its usefulness and those who contend it has limited utility in a discipline with a strong quantitative emphasis and reliance on scientific method.

Addressing this fundamental question over the potentially ambiguous nature of a case study finding, which alone can neither directly inform nor disprove a generalizable finding, Arend Lijphart (1971) states that because of its singular nature, the case study in and of itself does not directly satisfy the standards of scientific research. He does, however, credit the case study with multiple indirect benefits, making it a valuable component in establishing political science theory. He identifies six types of case studies that fall into roughly two categories: those chosen because the case itself is of interest and that are purely descriptive and those chosen to inform and build theory. The first category encompasses single case studies, which are generally detailed histories of a specific event or result and which, he argues, have value for this history alone. The thorough knowledge of a country gained by such an intensive, rich study provides critical information that others can also benefit from. Additionally, these in-depth analyses not only are a source of data for larger comparative studies but may also identify new variables of interest or suggest potentially new theoretical explanations. Lijphart’s other typologies include those case studies that are chosen specifically for theory-building purposes. They include hypothesis-generating cases in areas in which no established theory exists; theory-confirming and -informing cases, both of which test existing theories; and deviant case analysis, for cases known to have varied from the expectations predicted by theory. This third type of case often reveals additional variables previously unidentified. It may suggest a temporal ordering of variables (path dependency) or identify the sometimes critical interactions of variables. The study of deviant cases may merely suggest refinements to the way variables are operationalized within the study, still an important theoretical contribution. These last three case typologies constitute the core of comparative case study, with their usefulness coming from their deliberate selection as a test to existing theory. While Lijphart identifies certain benefits of the case study approach, his praise is still conditional, and he favors the value of large-N quantitative studies whenever possible.

Harry Eckstein (1975) addresses the utility of case studies by first noting the predominant status held by historiographic work in earlier political science research. His main contention is that this early case study work, at both the micro and the macro level, although insightful in its own right, was perceived to be severely limited in its usefulness for producing generalizable theory, because of its singular focus and the statistical consequence of an N of 1. The prevailing assumption was that what theory-building utility case study work had was inductively drawn from the events studied, and those inferences might or might not represent replicable conclusions. Eckstein questions this assumption and lays out a detailed argument supporting the utility of case study work in all stages of the theory development process, not just the nascent ones. He additionally contends that case studies may actually be most valuable at the theory testing stage. Particularly in the field of comparative politics and when studying complex, potentially unique systems, Eckstein suggests that well-designed case study methods may be the best way of testing hypotheses and cumulating generalizable theories. Indeed, he emphasizes the role of case study in its comparative application and perspective.

Before he makes his argument for the value of case study to theory building, Eckstein (1975) provides valuable definitions of case study by emphasizing the concentrated, yet flexible, aspect of an investigation into a single event or individual. This focused yet not narrowly defined approach allows the investigator to be open to unexpected observations and new conclusions. Eckstein additionally makes the important distinction that the study of one event does not necessarily mean only one measure of the results. Rather, he contends that how an event or thing is studied will dictate its number of observations. Thus, one event can be broken down into numerous observations. For example, “Astudy of six general elections in Britain may be, but need not be, an N = 1 study. It might also be an N = 6 study. It can also be an N = 120,000,000 study” (p. 85). This example illustrates his definition of a case as the single measurement of a pertinent variable observed, so that comparative study is then defined as “simply numerous cases along the same lines, with a view to reporting and interpreting numerous measures on the same variables of different ‘individuals’” (p. 85).

After he provides a useful review of the steps toward the development of theory, first the question or puzzle, followed by the formulation of a hypothesis and then a test, with the cycle likely repeating itself as refinements are made, Eckstein proceeds to describe five distinct varieties of the case study and identifies the particular uses each has. The first of these, the configurative idiographic study, is meant to be a comprehensive study of its target but one that allows for intuitive interpretation of the facts. By definition, idiographic is individualizing rather than nomographic or generalizing. Indeed, Eckstein acknowledges that this type of case study was the predominant type he first alluded to in this work. But he makes the point that the strengths of these types of case studies are their very weakness. Their rich description and often persuasive intuitive interpretations may be individually factual, but they aren’t systematic, which makes generalizable conclusions problematic and substantive theories unlikely.

The disciplined configurative study, a term Eckstein (1975) credits to Sidney Verba, turns this relationship around somewhat; rather than building theories on interpretations, interpretations should be driven by theory. This implies that the details of a case should either confirm or disprove a theory that ought to apply to it. The problem with this approach is, as Eckstein points out, its “discipline.” The strict and usually narrow application to a case of a hypothesized theory should either confirm or deny it. In essence, Eckstein suggests that this approach may be too restrictive. It may also lack the flexibility to accommodate more intricate relationships not already identified or suggested by existing theory. He also worries that interpretation of cases on an existing theory presumes that the theory itself is correct and suggests that existing theory, however valid, may “compel particular case interpretations” (p. 104, italics added) with its emphasis on generalizability at the expense of more individualized findings.

Eckstein’s (1975) third type is heuristic case studies, which are deliberate searches for discovery, often a result of trial and error. These are meant to be creative, stimulating the imagination of the researcher toward new ways of looking at a problem, focusing on broader, more generalizable relationships. This discovery is incremental and is often developed in sequential studies as the new theory is further refined. The reason for heuristic case study is given rather succinctly by Eckstein: Theories do not arise from data alone but rather from the imagination of the researcher, after discerning puzzles and then patterns. Case studies, with their intensive analysis, increase the likelihood that these critical relationships will be found, particularly when they are carefully chosen to advance theory building. One caveat Eckstein offers on heuristic case studies is that they often produce too much—multiple explanations, too many variables, and a resulting complexity of interactions that are not only unwieldy but make generalization impossible.

Case studies are also used to probe the likelihood of proposed theories, a form that Eckstein (1975) calls plausibility probes. These are an intervening step before testing, to determine whether the expense of testing is warranted. Although the usefulness of such a study is limited to this end, and alone it cannot confirm a theory, it can, however, improve the prospects of testing, and for this reason it has value.

A more critical example of case study in theory building is the crucial case study. Eckstein (1975) confronts the dilemma of a single observation and the inability to correctly determine a statistical relationship on the basis of such limited information as a source of potential error for any theory based on it. The inductive fallacy is the error made when one derives a theory from only the observed (gathered) data, without further testing. The critical caveat is that one cannot test a theory with the same data used to originate the theory, and therefore another such example must be found. The crucial case is just such a test of a proposed theory. If all those variables deemed critical to a theory exist, then the results should be as predicted by the theory. Conversely, one can study a case similar in most respects, yet lacking in the hypothesized critical components, as a way of demonstrating that similar results did not result because the causal variable was missing. Although these most likely and least likely case study designs cannot absolutely confirm or deny theory, they are important tests of the likelihood of the theory and the correctness of the causal relationships being proposed.

Eckstein’s (1975) thorough typology and analysis of the case study method methodically crafts an argument for the benefits of case study work. These include the insight made possible by the rigorous, thorough inspection in a carefully crafted case study and its across-discipline utility in identifying new variables and new causal mechanisms leading to the generation of new theory. To accomplish this goal, Eckstein emphasizes that case study selection must be driven by theory, and not by interest or convenience.

Charles Ragin (1987), in The Comparative Method, devotes a chapter to the discussion of case-oriented comparative methods and addresses the likelihood that even the most meticulously performed case study is unlikely to produce definitive explanations. However, identifying critical contextual facts may help determine the causal relationships underlying the observed phenomena. It is important to note that Ragin emphasizes the value that an intensive case study accrues to its researcher. Deep understanding of an event or case in its entirety, rather than merely knowing pieces of information, allows for more contextually rich comparison to other events. This richness can only enhance the reliability of the causal inferences drawn. Such depth of knowledge is likely limited to a small number of cases, and indeed this complexity is a constraint on the case study researcher. Case study is, as Ragin shows, a successful strategy for analyzing complex, multicausal events and at the same time still cohesively connecting them theoretically. He concludes with a nice summation of the strengths of the case study method: Case studies make possible the discovery of patterns of relationships and difference, with all deviations requiring an explanation, necessitating a thorough knowledge of the data. Since case study work does not rely on statistical probabilities such as frequency or distribution, a single case can be critical and can potentially prove or disprove a hypothesis. Case study work is holistic and requires a thorough understanding of the entire event, not just targeted aspects of it, and finally, case study encourages creative new ways of examining behavior and events. Particularly in the identification of complex interactions and the importance of context in understanding their role, Ragin makes the point that case studies provide a methodologically distinct approach.

In Designing Social Inquiry, Gary King, Robert O. Koehane, and Verba (1994) argue that the same level of testable, scientific rigor can be applied to qualitative work that quantitative scholars are able to use in their statistically based work; qualitative work includes, of course, case studies. King et al. focus on research design with an emphasis on the logic of inference, to use the facts that are known to learn about facts as yet unknown. This is then used to identify causal relationships and construct theories that can then be tested. King et al.’s emphasis on the latter stages of research design, producing theory that is testable and thus falsifiable, challenges case study researchers to think rigorously about their work, to recognize the similarities of quantitative and qualitative work with respect to empirical rigor, and to approach their work as such. King et al. argue that the primary way to do this is to see qualitative data more quantitatively, and to accomplish this from a practical standpoint, they advise maximizing the number of observations (from which measures are taken) whenever possible. At the same time, when adding an observation is not possible, they recommend summarizing on the outcome of interest instead, in order to avoid issues of micronumerosity (having more variables than observations). Echoing Eckstein (1975), King et al. remind us that the size of a case study, its N, is often determined by the level of analysis chosen: Is it one single event, several incidents within that event, or many more individual acts? In addition to constructing a design that allows for multiple observations, the authors emphasize the requirement of designing theories that can be falsified (i.e., the null hypothesis can be tested). King et al. also address the importance of reducing the potential bias introduced through case selection. They emphasize the care with which cases must be chosen, as there must be “the possibility of at least some variation on the dependent variable” (p. 129). Other potential sources of selection bias they cite are investigator induced: choosing cases because data are available or because one has a particular interest in or understands the language, or the larger bias that often occurs when case selection is correlated with the dependent variable. In this instance, the process being studied has already been selected for over time, leaving as evidence only its most recent iteration and losing any obvious trace of what may have been critically important in the intervening stages.

Not all scholars implicitly agreed with the arguments made by King et al. (1994), and a lively review symposium in response to it appeared in the journal American Political Science Review. In it, Ronald Rogowski (1995) challenges King et al.’s concern with the testability of single-observation studies and relates three examples of just such single-case studies that do succeed under this limitation. He offers additional examples in response to their admonitions against dependent variable selection bias and comments that without deliberate selection based on a case’s anomaly (its status as a statistical outlier), one of the core benefits of case study work would be lost. Rogowski sums up by emphasizing the importance of not losing the benefits of good qualitative work at the expense of increased quantifiability. In the same symposium, David Collier (1995) also takes issue with how King et al. address selection bias. However, although Collier generally concurs with their position, he argues for a bit more nuance when one is faced with some of the realities of the comparative method. Additionally he identifies the importance of valuing the context of research findings as more important perhaps than their generalizability, and he gently suggests that King et al. could be less rigid in their appraisal of qualitative methods.

Since case study is just that, an intensive examination of at least one item, how cases are selected is a fundamental issue. In comparative case studies, this issue is particularly relevant because small-N studies suggest that there exists more than one unique example of what is being examined and therefore a larger population to choose from. As a result, concerns over potential selection bias contribute prominently in discussions of the case study method. In “How the Cases You Choose Affect the Answers You Get: Selection Bias in Comparative Politics,” Barbara Geddes (1990) addresses this issue by reexamining three prominent comparative studies. She neatly demonstrates how the potential error of case selection on the dependent variable can particularly impact results in small-N studies. Essentially a primer on selection bias, this article outlines the importance not only of identifying the most likely causal reasons some event occurred, but also of examining the counterfactual as well. Geddes makes the point that by not providing a larger sample, selected randomly (rather than on the dependent variable) for testing the proposed relationship between cause and effect, one is really comparing only “the differences among the selected cases” (p. 132). She then shows how such an error can also occur in a path-dependent argument. In both examples, misleading findings resulted from researchers’ not expanding the population from which the targeted cases were drawn. Had they done so, they would have had a larger and likely more random sample to test. Geddes’s final example involves time-series studies and the determination of the appropriate end point of a case study. In this instance, she shows how changing the dates of a study would affect its results drastically, and she also makes the point that historical case studies are especially vulnerable to selection bias based on the time frames chosen for analysis.

The more recent discussion of case study work has focused increasingly on understanding the role of this method as part of a comprehensive research strategy. John Gerring (2004) emphasizes how, by failing to accommodate the bounded aspect of case work, most commonly used definitions for case study are inadequate. He offers the definition of case study as “an intensive study of a single unit for the purpose of understanding a larger class of (similar) units,” with units being “spatially bounded phenomena” (p. 342). This implies the study of a unique event or thing, at one point in time, with the goal of generalizability, which, he argues, provides a more theoretically useful interpretation. Gerring then provides a comprehensive discussion of the methodological ambiguities that occur in case studies and identifies six areas in which case studies are vulnerable. With these as a guide, he outlines the strengths and weaknesses of case (within-unit) study versus across-unit study. He notes that the case study method is more suited to descriptive inferences than to causal ones. It is a method that has a special affinity with intensive, focused studies rather than those that are extensive and broad. Case study is more likely to have high internal validity and weak external validity. It facilitates the defining of causal mechanisms, and not the testing of causal effects, performing better when causal mechanisms are deterministic instead of probabilistic. Finally, case studies are well suited to exploratory research but are limited in their uses for confirming hypotheses, yet they are preferred when across-case studies cannot provide adequate variance for the relationship being studied. With this enhanced clarity, and by situating case studies not apart from but as a complement to noncase methods, Gerring suggests that case study methods should be accepted as an equally worthy methodological approach by the entire discipline and that rather than favoring one method over another (often exclusively), scholars should use the method most suited to their question, their data, and their theory.

With Alexander George and Andrew Bennett’s (2005) Case Studies and Theory Development in the Social Sciences, the debate within the discipline over case studies is brought up to date. Both authors have been longtime advocates of case study methods, and this latest work is a very thorough argument for the value of case study methods as part of a research strategy that includes both quantitative and qualitative methods as well as formal theory (Bennett, 2004, offers a chapter-length article distilled from this material, as well). George and Bennett disagree with King et al.’s (1994) contention that there can be only one “logic of Inference” (p. 11). George and Bennett discuss the relationship between case studies and the systematic building of theory. They compare the strengths and weaknesses of case studies and first identify four strengths, all areas in which statistical methods tend to be weak. These include concept validity, the potential for discovering new causal variables and deriving new hypotheses, a better understanding of the relationship between causal variables and possible path dependency, and the ability to identify or model the complex interactions of these variables. Weaknesses of case studies include the potential for introducing selection bias from the cases chosen and the inability to accurately measure the relative strength of an effect. Also, because of their single or very small number, case studies are relatively unique and not necessarily representative; cases chosen from a small pool may not necessarily be independent of one another, and they do not have a rich number of observations from which to judge the strength of associations between variables. George and Bennett advocate the use of the structured, focused comparison, which allows for the collection of data that can be systematically compared with other cases as well as accumulated. In this way, scientific rigor is added, and the utility of case methods is likely increased. The authors then outline the method of case study, from designing the research to executing the study and to drawing conclusions from the findings. In all steps, the role of theory is predominant: It drives the design and motivates the findings. In addition to being the definitive authority on case methods, George and Bennett present a compelling argument for using multiple methodological approaches in a research program. Not only do they show how qualitative and quantitative methods complement each other; they integrate formal modeling as well. This approach is gaining momentum in political science today, making a qualitative skill set not merely useful but necessary.

III. Examples of the Case Study Approach

A. U.S. Politics

In the field of U.S. politics, the classic example of a grounded, participant observer case study must be Alexis de Tocqueville’s (1835/2004) Democracy in America. Although most modern scholars of U.S. politics solve their N of 1 problem by focusing on the subunits of U.S. government, using states or administrations, court terms or congressional voting records as their unit of analysis, Tocqueville analyzed the United States as a single entity. He drew his conclusion, that it is citizens’ affinity for joining in and participating at all levels of civic life that strengthens democracy and enables it to flourish, from his personal observations as he extensively toured the country in the early 1800s. A more modern work in U.S. politics that is rooted in qualitative case study work is Richard Fenno’s (1978) examination of congressional members, Home Style: House Members in Their Districts, in which he used extensive interviews and considerable time observing congressmen, both in Washington, D.C., and, critically, in their districts. This self-styled “soaking and poking” enabled a comprehensive, in-depth observation that allowed Fenno to identify the paradox of individual representatives’ being very well-liked by their constituents at the same time as the institution of Congress is collectively viewed much more critically, and he explains much of the paradox by the personal relationships developed through district service. David Mayhew’s (1974) Congress: The Electoral Connection also looks at the relationship between members of Congress and their constituents and is another example of a work based on inductive reasoning rooted in extensive in-depth participant observation. Mayhew finds that it is the incentive for reelection that motivates the individual behavior of both congressmen and the Congress. Through committee assignments, leadership positions, and vote trading (among other means), congressmen ensure their reelection chances. Mayhew suggests that with Congress motivated as a whole by mutual self-interest, it is no surprise that the structural arrangements of Congress, its organization of the leadership and committee system, have evolved to facilitate this behavior.

B. Comparative Politics

The case study method is used most extensively in the subfield of comparative politics. Using primarily small-N research designs, many significant works have been produced. Included among these is Barrington Moore’s (1966) Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World. Moore examines five societies to compare their experiences with modernization and the economic revolution that ensues. He concludes that there are three likely outcomes, dependent on the country’s social structure, and these in turn predict the likelihood of a successful transition to democracy or descent into dictatorship. The most well-known example of an intensive single-country case study must be Robert Putnam’s (1993) Making Democracy Work: Civic Traditions in Modern Italy, also an excellent example of historiographic work. The subject of lively debate within the discipline, social capital, that is, the extent to which citizens are participatory and invested in their communities as a result of their civic relationships, was found by Putnam to be a necessary component of a successful democratic society. Putnam argues that the associational experience of northern and central Italy developed interpersonal trust and fostered more democratic local governments, but the lack of similar groups in the south left them with less. Another such single-case work is Robert Bates’s (1989) Beyond the Miracle of the Market: The Political Economy of Agrarian Development in Kenya. This work, which focuses on the intersection between land use, government institutions, and public policies, relies on a critical understanding of the economic, political, and cultural forces at work in Kenyan society. The complex interplay of economics and politics that Bates studies is only fully appreciated when the cultural context is included; the influence of tribal affiliations and Kenya’s British colonial legacy are just two examples. These kinds of rich, multilayered observations and intimate knowledge of a society can be accomplished only with case study methods, with which Bates combines quantitative rigor as well.

C. International Relations

International relations scholars have also extensively used the case study method to selectively examine the actions of elite actors and organizations during critical events. Case study work is used to evaluate existing theory as well as propose alternate explanations to better understand the often complex motivations of and among nation-states. In Essence of Decision: Explaining the Cuban Missile Crisis, Graham T. Allison (1971) examines the Cuban missile crisis and, primarily through interviews, reconstructs the often conflicted decision-making process of all the major participants. To do this, he approaches the same event from the perspective of three different decisional- behavior models. These competing approaches are collectively used to illustrate the author’s thesis: that despite internal pressures to the contrary, it was the actions and the decisions of the two leaders that successfully resolved the issue. Alexander George and Richard Smoke’s (1974) Deterrence in American Foreign Policy is an example of a focused-comparison case study that examines 11 instances of the failure of U.S. deterrence policy. George and Smoke use process tracing to establish the causal explanation, which would not be possible without the depth of knowledge acquired in these case histories. In doing so, they critique existing theory and are able to offer a new, more dynamic, explanation. In another example of a focused-comparison study, Stephen M. Walt (1987), in The Origins of Alliances, looks at alliance formation and contrasts two distinct types: those made for mutual support to defend against a threat and those that are more opportunistic (or perhaps pragmatic), in which one aligns with the threat itself. Walt then explores the likely causes of these choices, looking specifically at shared ideology and the influences of foreign aid. His concentrated case study of states in the Middle East during a single period allows him to develop the depth of knowledge necessary for such a study, in which data alone would be inadequate.

IV. Conclusion

As the previous examples illustrate, case study work is applicable to a broad range of theoretical questions. Indeed, for many situations, a case study examination is the only way to rigorously examine an event. Case study can be used in either half of the research cycle: to deductively test the hypothesized research question or to inductively explore the results of empirical observations. It is also a valuable method for developing original theoretical insight, which can often form the basis of a research design using more statistically robust methods. Case study in and of itself serves a vital informative purpose as well, allowing in-depth appreciation of often nuanced yet critical conditions of the larger phenomenon being observed. Finally, the case study is increasingly being appreciated as a necessary component of comprehensive political science research today: Together with traditional quantitative methods that provide reliable statistical probabilities for a tightly focused view, and formal theory methods that produce more soft-focused or abstract explanations, case study work provides a necessary contribution by filling in the gaps, compensating for the inevitable shortcomings when formal and quantitative methods are applied to real-life questions and problems. Most critically, a well-crafted case study gives the researcher a level of knowledge and understanding of the matter being examined that no other method allows. This benefit alone justifies the application of case study methods to social science research today and in the future.

Bibliography:

1. Allison, G. T. (1971). Essence of decision: Explaining the Cuban missile crisis. Boston: Little, Brown.
2. Bates, R. (1989). Beyond the miracle of the market: The political economy of agrarian development in Kenya. Cambridge, UK: Cambridge University Press.
3. Bennett, A. (2004). Case study methods: Design, use and comparative advantages. In D. F. Sprinz & Y. Wolinsky Nahmias (Eds.), Models, numbers and cases: Methods for studying international relations (pp. 19-55). Ann Arbor: University of Michigan Press.
4. Collier, D. (1995). Review: Translating quantitative methods for qualitative researchers: The case of selection bias. American Political Science Review, 89, 461-466.
5. Collier, D., & Mahoney, J. (1996). Insights and pitfalls: Selection bias in qualitative research. World Politics, 49, 56-91.
6. Eckstein, H. (1975). Case study and theory in political science. In F. Greenstein & N. Polsby (Eds.), Handbook of political science: Vol. 7. Strategies of inquiry (pp. 79-137). Reading, MA: Addison Wesley.
7. Fenno, R. (1978). Home style: House members in their districts. Boston: Little, Brown.
8. Geddes, B. (1990). How the cases you choose affect the answers you get: Selection bias in comparative politics. Political Analysis, 2, 131-150.
9. George, A. L., & Bennett, A. (2005). Case studies and theory development in the social sciences. Cambridge: MIT Press.
10. George, A. L., & Smoke, R. (1974). Deterrence in American foreign policy: Theory and practice. New York: Columbia University Press.
11. Gerring, J. (2004). What is a case study and what is it good for? American Political Science Review, 98, 341-354.
12. King, G., Koehane, R. O., & Verba, S. (1994). Designing social inquiry: Scientific inference in qualitative research. Princeton, NJ: Princeton University Press.
13. Lijphart, A. (1971). Comparative politics and the comparative method. American Political Science Review, 65(3), 682-693.
14. Mayhew, D. R. (1974). Congress: The electoral connection. New Haven, CT: Yale University Press.
15. Moore, B., Jr. (1966). Social origins of dictatorship and democracy: Lord and peasant in the making of the modern world. Boston: Beacon Press.
16. Putnam, R. (1993). Making democracy work: Civic traditions in modern Italy. Princeton, NJ: Princeton University Press.
17. Ragin, C. C. (1987). The comparative method: Moving beyond qualitative and quantitative strategies. Berkeley: University of California Press.
18. Rogowski, R. (1995). Review: The role of theory and anomaly in social scientific inference. American Political Science Review, 89, 467-470.
19. Tocqueville, A. de. (2004). Democracy in America. New York: Library of America. (Original work published 1835)
20. Walt, S. M. (1987). The origins of alliances. Ithaca, NY: Cornell University Press.