What Status Should Case Studies Be Given in the Study of Comparative Politics?


The importance of case selection for comparative politics could hardly be exaggerated (Hancké, 2010: 239-240; Keman, 2011: 54). As principle units of observation and carriers of information, cases constitute the cornerstone of research design and conduct (Keman, 2011: 55). The nature and number of the cases selected for examination and comparison, therefore, to a considerable extent determine the quality and reliability of the process and outcome of the comparative analysis of politics. Such importance of case selection is amply demonstrated by the continuous and voluminous debates and discussions within the discipline of political science (Brady and Collier, in Keman, 2011:54; Keman, 2011: 54-55). Practitioners of political studies, especially comparative politics, disagree over how to make the inevitable experimental trade-off between large-scale, few-variable cases (large-N design) and small-scale, many-variable cases (small-N design) and hence what status they should be granted respectively (Hancké, 2010: 237-241). In particular, the role and status of case studies, by which is generally meant studies of one single or limited number of cases, is often contested. Concerning this issue, it is my understanding that the discipline and profession of comparative politics should not risk overlooking the utility and indispensability of case studies. I will thus argue in this essay that case studies should be utilised more frequently and widely, particularly in testing theories and generating hypotheses. In the following sections I will first demonstrate how case studies can play a central role in the confirmation or falsification of extant theories, by analysing such studies’ comparative advantage in this enterprise over its large-scale counterpart and employing representative precedents of the application of case studies in theory-testing. I shall then explain why such studies are important to the generation and formulation of new hypotheses, by showing their superior capacity for inducing and creating hypotheses relative to large-N studies, through both logical-theoretical induction and empirical examples citation. It is my hope, as aforementioned, to generate increased awareness, acknowledgement and application of the method and approach of case studies, the value of which to the comparative analysis of politics should not be underused.

Case Studies and Theory Testing

As abovementioned case studies could play a central role in the testing of extant theories. In this section I elaborate on this point. A key difference between case studies and large-N analysis, as briefly mentioned above, is the number of cases selected and thus the variables involved in these cases. In large-N studies, then, the variety of variables, the diversity of factors and hence the specificity and contingency of certain conditions and circumstances are sacrificed for such research design purposes as the minimisation of error variance and the control of extraneous variance (Peters, in Keman, 2011:57). In other words, there exists approximately a negative correlation between the number of cases studied and the possibility of studying such cases’ specific details (Caramani, 2011: 12-13). Small-N, case-oriented studies thus allows comparativists to focus more on the details and specificities of particular cases, the complex and contingent interaction and constellation of multiple factors, which hence enables researchers to conduct in-depth analyses and notice any phenomenon or behaviour unique to certain cases under particular conditions. This methodological emphasis on specificity and subtlety, as a result, makes case studies extremely suitable for testing theories. For the basic logic of theory testing is but as follows: if in the process of theory testing the researcher’s empirical observation is consistent with the theory tested, then this theory is confirmed as valid (at least at this stage); if, however, his observation contradicts with what the theory claims, then the tested theory is rendered invalid, i. e. falsified (Hay, 2002: 80-81). Therefore the more detailed and in-depth one’s empirical observation is, the more possible it is for him to spot any inconsistency between the case(s) studied and the theory under testing. Since as established case studies allow more in-depth and detailed analysis, such studies are much more suitable for examining the validity of extant theories. I now illustrate this point with examples of its empirical application.

One prominent example of the application of case studies in theory testing is Herrigel’s case study of the German political economy. In his Industrial Constructions (1996) Herrigel carried out a single country study on Germany in order to test the empirical validity of an explanatory theory proposed by Hall and Soskice (2001), namely that Germany’s coordinated market economy (CME) model is caused by its decentralised institutional structure of capitalism. This theory, as Herrigel demonstrated, proved false since the allged decentralised industrial order only concentrated in the southwestern region of the said country; whilst other industrialised regions predominantly possess an autarkic industrial order (Culpepper, 2005: 2-3). Therefore as the researcher’s empirical observation (absence of the decentralised industrial structure all across German industrial regions) contradicts with the tested theory (decentralised industrial order as the prevalent German economic structure), the theory has been falsified. Here it is of little difficulty to discern that without the detailed and in-depth analysis allowed by case study, such analytically microscopic feature as the regional variance of industrial order within a single nation-state might not be noticed at all. Similarly, Saxenian’s case study of American sub-regions falsified the theory that America’s liberal market economy (LME) model was inter-regionally homogenous; Snyder’s case study of Mexico’s economic policies falsified the theory that neoliberal market reform resulted in free market economy; Moore’s case studies of six major countries falsified the theory that democratization was mainly a political result of economic development; not to mention classical works like Tocqueville’s Democracy in America or Dahl’s Who Governs?, both of which exemplify the strength of case studies in testing the reliability of extant theories (Culpepper, 2005: 2-4; Hopkin, 2010: 301; Snyder, 2001: 3-4). Thus, as these selected examples demonstrate, case studies can and do play a central role in testing theories, a key reason for its deserved increase in usage and status.

Case Studies and Hypothesis Generating

Having substantiated the indispensable role case studies could play in theory testing, I now turn to the other major application of such studies, namely hypothesis generating. The generating and proposing of new hypotheses, in general, depend on the capacity of the researcher to notice and problematize new, unique phenomena and outcome previously undescribed or unstudied. This in turn requires the research design and method to allow the researcher to focus more on details and depth, on specificities and contingencies of particular cases. Expressed simply, new hypotheses require new observation; new observation requires detailed analysis. This, as I have shown in the last section, is best served by case-oriented studies. Because case studies, being limited in number but (consequently) abundant in variables, allows the comparativist to ‘gain an in-depth understanding of their research subject’ which thus maximises the possibility of his discovering any formerly unnoticed behaviour or phenomena which could then be used to generate new hypotheses (Vromen, 2010: 255). In addition, to view the same matter from an alternative (though not altogether separate) perspective, as implied in the section above, the falsification of an old theory per se is but the generation of a new hypothesis (itself to be tested in future researches), albeit one that rejects the condition-outcome relations of the invalidated theory (Hancké, 2010: 238). And as the first section demonstrates, since case studies are particularly suitable for theory testing, they become simultaneously so for hypothesis generating.

To give an example of case studies’ contribution to hypotheses generation, consider Antonsich’s recent case study of four European regions to understand EU citizens’ attachment to and notions of the EU. Through detailed and in-depth interviews and analyses of selected participants from the regions of Lombardia (Italy), Pirkanmaa (Finland), North-East England (United Kingdom) and Languedoc-Roussillon (France), Antonsich found patterned variance in people’s notional attachment to the EU. For instance, whilst one Italian participant claimed ‘there is a greater belonging with those who speak your language’, an English respondent revealed ‘I don’t actually feel European, I just want the benefits from being a European citizen’ (Antonsich, in Vromen, 2010: 260-261). On such basis he proposed the new hypothesis that there exist three different notions of ‘Europe’: cultural-national, cultural-transnational, and functional-utilitarian (Vromen, 2010: 260). Therefore as seen in this case, case study offered the researcher the in-depth insight into subtle and qualitative variance hardly noticeable on a large-N scale, which greatly contributed to the generation of new hypothesis. Together with multiple other examples, empirical applications demonstrate the superior capacity of case studies in hypothesis generating.


To summarise, in this essay I have argued and demonstrated the proposition that case studies should be utilised more frequently and widely, ergo be given a more active and prominent status, especially regarding the two enterprises of theory-testing and hypothesis-generating. First, with the help of representative precedents of the application of case studies in examining the reliability of existing theories and a comparative analysis of case studies and large-N methods, I have demonstrated the centrality and suitability of case studies for the (in)validation of extant theories. Next, by analysing the relative advantage of case studies over large-N studies in the generation of hypotheses through logical, theoretical induction as well as the citation of relevant empirical cases, I have substantiated the sui generis value and utility of case studies for hypothesis-generating. Therefore as I have managed to establish in this essay, case studies possess irreplaceable value and significance for the comparative analysis of politics, thus should be given a more active and preeminent status which matches its academic anility and potential vis-à-vis comparative politics.


Caramani, D. (2011) ‘Introduction to comparative politics’, in in Caramani, D. (ed.) Comparative Politics (second edition), Oxford: Oxford University Press, pp. 1-19.

Culpepper, P. D. (2005) ‘Single Country Studies and Comparative Politics’, Italian Politics & Society, 60: 2-5.

Hancké, B. (2010) ‘The Challenge of Research Design’, in Marsh, D. and Stoker, G. (eds) Theory and Methods in Political Science (third edition), Basingstoke: Palgrave Macmillan, pp. 232-248.

Hay, C. (2002) Political Analysis: A Critical Introduction, Basingstoke: Palgrave Macmillan.

Hopkin, J. (2010) ‘The Comparative Method’, in Marsh, D. and Stoker, G. (eds) Theory and Methods in Political Science (third edition), Basingstoke: Palgrave Macmillan, pp. 285-307.

Keman, H. (2011) ‘Comparative Research Methods’, in Caramani, D. (ed.) Comparative Politics (second edition), Oxford: Oxford University Press, pp. 50-63.

Snyder, R. (2001) Politics after Neoliberalism: Reregulation in Mexico, Cambridge: Cambridge University Press.

Vromen, A. (2010) ‘Debating Methods: Rediscovering Qualitative Approaches’, in Marsh, D. and Stoker, G. (eds) Theory and Methods in Political Science (third edition), Basingstoke: Palgrave Macmillan, pp. 249-266.


Written by: Haoyu Zhai
Written at: University of Bristol
Written for: Dr Andrew Wyatt

Date written: March 2015

Please Consider Donating

Before you download your free e-book, please consider donating to support open access publishing.

E-IR is an independent non-profit publisher run by an all volunteer team. Your donations allow us to invest in new open access titles and pay our bandwidth bills to ensure we keep our existing titles free to view. Any amount, in any currency, is appreciated. Many thanks!

Donations are voluntary and not required to download the e-book - your link to download is below.