The study of comparative politics has been primarily concerned thus far with the formal institutions of foreign governments, particularly of Western Europe. In this sense it has been not only limited but also primarily descriptive and formalistic. Its place in the field of political science has been ill-defined. Is the student of comparative politics properly concerned primarily with description of the formal institutions of various polities, or with undertaking comparisons? If with the latter, what is the meaning of comparison? Is it confined to the description of differences in various institutional arrangements? If comparison is to be something more than description of formal institutional differences, what are its aims, scope, and methods? Should the student of comparative politics attempt to compare total configurations? If not, then he has to develop a precise notion of what can be isolated from the total configuration of a system or systems and compared.
The above questions illustrate the difficulties and the challenge confronting the student of comparative politics. The problem of comparative method revolves around the discovery of uniformities, and the examination of variables in the context of uniformities between various systems. But even so, what are the particular clusters of states that have a degree of uniformity that makes the comparison and understanding of variations possible? Does the concept of “area,” as used, provide us with such a hunting ground for the study of “difference,” against a background of “uniformity”? Is the concept of culture a more acceptable one? Or does the similarity of social and economic contextual elements provide a better opportunity to compare and understand variations? What degree of homogeneity is required for comparison, and are comparisons between systems not showing the desired degree of homogeneity impossible?
The members of the seminar agreed that it would be extremely ambitious to attempt to answer these methodological questions in detail and at the same time indulge in empirical investigation. It was thought that if we could do some spadework on the methodological issues, we would invite comments and suggestions from the members of the profession and increase awareness of the need for making one’s methodological position explicit prior to undertaking comparative empirical investigation.
Comparability and uniqueness
Questions of the purpose and nature of comparison arose repeatedly, and the members of the seminar devoted considerable time to discussing them. It was agreed that the general problem of comparison is extraordinarily difficult for political science since it is unlikely that we will ever find two or more societies that are identical in all respects except for a single, variable factor. Consequently, the possibility of comparing one variable or even a set of variables against identical conditions is illusory. The alternative would seem to be comparison—at different levels of abstraction and complexity—of wider or narrower segments of the political process.
The following tentative classification of levels of comparative analysis was suggested: (1) comparison of a single problem limited to political systems that are homogeneous in character and operation; (2) comparison of several elements or clusters of elements in relation to political systems that are fairly homogeneous; (3) comparison of institutions or segments of the political process irrespective of “homogeneity”; (4) comparison of political systems as such. These four levels of comparison require increasingly higher levels of abstraction. At the fourth level, some such approach as that of “ideal types” would seem to be called for.
In general, two points of view were expressed throughout the seminar discussions. The first saw the need of a conceptual scheme that not only precisely defines the categories under which data may be collected, but also indicates the criteria of relevance to be adopted and the variables that are to be related hypothetically for the purpose of comparative study. According to the other view, given the present state of comparative studies, comparability ought to be derived primarily from the formulation of problems with limited and manageable proportions. This disagreement should not obscure the area of agreement reached by the members of the seminar.“The connecting link between general hypothetical series and particular social relations should be provided by specifying the conditions under which any or all the possibilities enumerated in the series are expected to take place.”
For example, it was agreed: (1) Comparison involves abstraction, and concrete situations or processes can never be compared as such. To compare means to select certain types of concepts, and in selection we have to “distort” the unique and the concrete. (2) Prior to any comparison it is necessary not only to establish categories and concepts, but also to determine criteria of relevance of the particular components of a social and political situation to the problem under analysis, e.g., relevance of social stratification to family system, or of sun spots to political instability. (3) It is necessary to establish criteria for the adequate representation of the particular components that enter into a general analysis of a problem. (4) The formulation of hypothetical relations and their investigation with empirical data can never lead to proof. A hypothesis or a series of hypothetical relations would be considered verified only as long as not falsified. (5) Hypothetical relations rather than single hypotheses should be formulated. The connecting link between general hypothetical series and particular social relations should be provided by specifying the conditions under which any or all the possibilities enumerated in the series are expected to take place. (6) Finally, one of the greatest dangers in hypothesizing is the projection of possible relationships ad infinitum. This can be avoided by the orderly collection of data prior to hypothesizing. Such collection may in itself lead us to recognize irrelevant relations (climate and the electoral system, language and industrial technology, etc.).
The members of the seminar, therefore, substantially rejected the arguments in favor of uniqueness, and argued that comparison between institutions not only is possible but may eventually provide, through a multiple approach, a general theory of politics and a general theory of political change. Before this development can be realized, the following research approaches should be emphasized and undertaken in as orderly a way as possible: (1) elaboration of a tentative and even rough classificatory scheme or schemes; (2) conceptualization at various levels of abstraction, preferably at the manageable level of the problem-oriented approach; (3) formulation of single hypotheses or hypothetical series that may be suggested by the formulation of either a classificatory scheme or sets of problems; (4) constant reference of hypotheses to empirical data for the purpose of falsification and the formulation of new hypotheses.
Approaches to the comparative study of political systems
Conceptual schemes. The seminar members believed that formulation of a conceptual scheme for the comparative study of politics would help provide a classificatory table and permit the elaboration of hypotheses. Comparison, it was agreed, must proceed from a definition of politics as a universally discoverable social activity. The function of politics is to provide society with social decisions having the force and status of legitimacy. A social decision has the “force of legitimacy” if the collective regularized power of the society is brought to bear against deviations, and if there is a predominant disposition among those subject to the decision to comply. As for the means of enforcing decisions, every society, generally speaking, has a determinate organization that enjoys a monopoly of legitimate authority (or political ultimacy). Moreover, the characteristic distinguishing between political relationships and other relationships is the existence of this framework of legitimacy. Conceptions of legitimacy or “legitimacy myths” are highly varied ways in which people justify coercion and conformity, as well as the ways by which a society rationalizes its ascription of political ultimacy, and the beliefs that account for a predisposition to compliance with social decisions.
But the legitimacy myth only defines the conditions of obedience. Within its framework there is the political process itself, through which numerous groups having political aspirations (policy-aspiration groups and power-aspiration groups) strive for recognition and elevation to the position of legitimacy. The factors that determine which power-aspiration group is to be invested with legitimacy, to the exclusion of all others, are the effective power factors in the system.
It was suggested, then, that the general modes of politics, for the purpose of analysis, would be as follows: Political processes are the struggle among power-aspiration and policy-aspiration groups competing for the status of legitimacy; the outcome is determined by the society’s structure of effective power, and the end state, legitimacy, is the political reflection of its general value system.
The major components of the political process, which should provide a fairly coherent classification scheme as well as the possibility of formulating hypothetical relations, are the following: the “elective” process of the system, its “formal” deliberative process, its “informal” deliberative process, its structure of “influence,” and its structure of “power.” The major tasks, envisioned under this scheme, in the analysis of political systems are: (1) to analyze the legitimacy myth of the society in terms of specific content and relationship to the society’s general myth structure; (2) to inquire into the system’s political aspirations, political processes, and effective power factors; (3) to analyze both the complexity and ultimacy of decision-making systems in the society, specifically the conditions under which political ultimacy is either diffused or concentrated, and the relationships between subsidiary and ultimate decision-making systems; (4) to provide for a theory of change through the study of “formal” and “informal” processes.
Alternative approaches. The general agreement on the usefulness of a conceptual scheme was coupled with an equally strong emphasis on the need for alternative approaches. It was thought that the present state of comparative politics calls for a “pluralistic” rather than a unitary approach, and that for each of the alternative approaches suggested, the same degree of methodological rigor should be followed as in the development of a conceptual scheme. The alternative approaches agreed upon were: the problem approach, the elaboration of a classificatory scheme or checklist to aid in more coherent and more systematic compilation of data, and the area approach.
The problem approach“This approach is flexible enough so that it can lead the research worker to examination of questions that have a varying degree of comprehensiveness in terms of both theory and empirical orientation and investigation.”
The study of comparative politics cannot wait for the development of a comprehensive conceptual scheme. Instead of aiming toward universality it may be advisable to adopt a more modest approach. The members of the seminar agreed that the “problem approach” is a step in this direction. It was pointed out that the formulation of a problem in itself has some of the characteristics of a conceptual scheme. It directs research toward various aspects of the political process and at the same time calls for an ordering of empirical data and the formulation of hypotheses or series of hypothetical relations. Furthermore, this approach is flexible enough so that it can lead the research worker to examination of questions that have a varying degree of comprehensiveness in terms of both theory and empirical orientation and investigation.
Three types of problem approaches were suggested and discussed: (1) Narrow-range theory, which involves a relatively low degree of abstraction; i.e., it applies to homogeneous cultural contexts and deals with a limited number of variables. (2) Middle-range theory, which is conceived to include problems of fairly general importance, involving a relatively high degree of generalization, but remaining below the level of a truly general theory of politics. (3) Policy-oriented theory, which deals with the immediate practical solution of important problems and is consequently focused on problems originating in pressing conflict situations or in an overwhelming need for policy action.
Four criteria by which to select problems were suggested: the intrinsic interest of the problem to political scientists; its ability to eliminate certain key difficulties in the comparative method and the analytical utility of comparison; its capacity for advancing research beyond the current level of inquiry in the field; its probable and eventual significance for the formation of a general theory of comparative politics.
It was agreed that the formulation of the problem should be as clear and logically coherent as possible, and that it should be presented in the following form:
a. The problem must be stated precisely; it must be stated in such a form as to lead immediately to hypotheses; it must be analyzed into its component elements; its variables and the relations between them must be spelled out; and all this must be done in operationally meaningful terms.
b. Its relations to a possible general theory of politics must be described. How would the problem fit into a more general theoretical orientation and what more general questions can its solution illuminate?
c. The manner in which the problem calls for the use of comparative method must be demonstrated, and the level of abstraction that comparison would involve must be analyzed.
d. Outline a recommended research technique for dealing with the problem and justify the recommendation.
e. Enumerate possible alternative research techniques.
It was not considered the function of the seminar to state exhaustively narrow-range, middle-range, and policy-oriented problems, but a few typical ones were suggested for the purpose of illustration: An analysis of the relations between the power of dissolution and ministerial stability in parliamentary systems would fall in the realm of narrow-range theory. A study of the political consequences of rapid industrialization in underdeveloped areas of the world would be in the realm of middle-range theory. The following are policy-oriented problems: the development of constitutional government in colonial areas; how to deal with political instability in France; how to dissociate colonial nationalism from Soviet-inspired leadership and ideology; determination of policies of constitutional regimes toward totalitarian parties, e.g., the Communist Party.
A classificatory scheme
The seminar concluded that a checklist might facilitate assembling data in an orderly fashion under commonly formulated concepts. It decided to develop such a checklist for purposes of illustration. The following broad categories and subdivisions were specified:
(1) The setting of politics: an enumeration of the most significant contextual factors of all political systems, i.e., geographic patterning, economic structure, transportation and communication patterns, sociological structure and minorities, cultural patterns, values and value systems, and the record of social change.
(2) The sphere of politics: the actual and potential sphere of political decisions: conditions determining the sphere of decision making, limits on political decisions, major types of decision making, and potential changes in the sphere of decision making.
(3) Who makes decisions: Who are the “elite” supposed to be? To whom does the community impute prestige and what are the prevalent prestige images and symbols? Who actually makes the effective political decisions if they are not made by those who are supposed to make them?
(4) How decisions are made: formulation of problems, agencies and channels of decision making, some major characteristics of decision-making procedure.
(5) Why are decisions obeyed: the enforcement of decisions, compliance, consent, types of consent, ecology of compliance, measurement of compliance.
(6) Practical politics: types, purpose, organization, and techniques of policy-aspiration groups; types, goals, organization, techniques, and influence and effectiveness of power-aspiration groups.
(7) The performance of the system: stability, adjustment, and change, and their conditions, relationship between formal and informal processes, manifestations of instability and stability.
Area study and comparative politics
The third alternative approach to the study of comparative politics is the more systematic use of the area concept. However, neither geographic, historical, economic, nor cultural similarities constitute prima facie evidence of the existence of similar political characteristics. If the concept of an area is to be operationally meaningful for purposes of comparison, it should correspond to some uniform political patterns against which differences may be studied comparatively and explained.
The definition of an area on the basis of culture was considered to be worth detailed discussion. It was suggested that although primarily used by the anthropologists it might be adapted to the needs of political scientists.
To make the concept operationally more meaningful for political science, it was suggested that an attempt be made to define the area concept with reference to “political traits” or “trait complexes” or “problem configuration patterns,” in terms analogous to those used by anthropologists when they analyze the concept of culture in terms of “traits” or “trait complexes.” Such an approach to the definition of an area has not yet been undertaken despite its promise for comparative study. The very search for common political traits and patterns will call for classification and conceptualization. Once similar traits or patterns have been distinguished and have been related to geographically delimited units, the area concept will be of great value since certain political processes can be compared within the area against a common background of similar trait configuration. In this sense it was thought that future research should be directed toward developing in great detail classificatory schemes within areas.
The last topic considered was the relationship between comparative politics and international politics. It was pointed out that contemporary study of international politics has entered a new stage. Since the national interest is now a central concept of international politics, what tests or criteria are to be used in identifying the interests that shape the foreign policy of any state? Is the student of international politics to accept, at face value, the definition of national interests given by statesmen? Or does the examination of power factors, geography, historical development, etc. offer the student of international politics certain rough tests for defining the “objective” interests of the state?
Assuming that the concept of national interest provides a focal point for investigation, two approaches seem possible, either separately or in combination: (1) To determine analytically what the national interest “ought to be” under certain conditions. One could then compare this evaluation of the national interest with its “actual” definition as provided by the actions and pronouncements of the particular nation. (2) To describe, in terms of the following categories, the reasons why nations define their national interest in certain terms and not in others: survival prerequisites; objective physical conditions—geography, natural resources, tradition, past decisions, value systems, etc.; institutional channels through which the national interest is defined and set—i.e., interest programs and images and how they are defined in the political process; policy interplay between independent political units and their respective interests, including the pattern of reaction within each; the subjective pluralism of the society, by which the content of the interest images is set. The first four categories are in the nature of factors that set limits on the definition of the national interest. The last category defines the possibilities in substantive terms. On the basis of this scheme, it was suggested, there might be a useful “division of labor” or cooperation between students of international politics and of comparative politics. The second and fourth categories could best be handled by students of international politics, the others by specialists in comparative government.“In the context of national interest, foreign policy is to be conceived as a dynamic interplay of the given or chosen goals, the organic elements that set limiting conditions on the selection of goals, and the selection of strategy or means for the achievement of the given or chosen goals.”
More generally, however, a cooperative effort between students of international politics and comparative politics should be centered in the following areas of mutual interest: (1) The process of decision making has become a function in international politics through existing organizations. How does this decision-making take place? Is it accompanied by any broad legitimacy ideas or myths that transcend the national states? (2) The concept of national interest provides for meaningful concepts for the study of foreign policy. The concept will have to be broken down into component parts, some of which would be studied by students of comparative politics, while others remain within the domain of international politics. Given analogous conditions, generally speaking, the definition of national interest varies by individual states. Determination of what accounts for this variation seems to be the proper task of the specialist in comparative politics. (3) Study of the focal point at which the states meet—diplomacy and negotiations through which conflicts are resolved or common objectives realized—is a cooperative task that admits no arbitrary allocation of duties, for the action of each state depends upon domestic conditions, internal images, and traditional forces. Relations between states, on the other hand, have repercussions on domestic myths, images, authority symbols, and institutions. Domestic and international politics are in this sense complementary factors. (4) The student of international politics may also join with the student of comparative politics in attempting to define and study an area. Certain uniform outlooks and behavior patterns may be due to similar experiences shared by a number of states that can be geographically identified. The study of such uniformities and differences is primarily a joint task. (5) Finally, goals of foreign policy may be jointly studied. In the context of national interest, foreign policy is to be conceived as a dynamic interplay of the given or chosen goals, the organic elements that set limiting conditions on the selection of goals, and the selection of strategy or means for the achievement of the given or chosen goals.
This article is a summary of a longer report on the proceedings of the interuniversity summer research seminar on comparative politics held at Northwestern University during July and August 1952. Plans for the seminar were described briefly in the March 1952 issue of Items, p. 7. While the author is responsible for the statements appearing here, they summarize a collective product and at times reproduce lines of thought expressed originally by the participants both during the meetings and in their respective essays. Members of the seminar were: Samuel H. Beer and Harry Eckstein, Harvard University; George I. Blanksten and Roy C. Macridis, Northwestern University; Karl W. Deutsch, Massachusetts Institute of Technology; Kenneth W. Thompson, University of Chicago; and Robert E. Ward, University of Michigan. Richard Cox of the University of Chicago acted as rapporteur.
Roy C. Macridis (1918–1991) taught political science at Brandeis University since 1965. During the Second World War, Macridis served with the Office of Strategic Services and migrated to the United States in 1944. He was a member of the Council’s Committee on Comparative Politics from 1954 to 1958.
This essay originally appeared in Items Vol. 6, No. 4 in December of 1952. Visit our archives to view the original as it first appeared in the print editions of Items.