|
|||||||||||||
|
third matrix. Note that no associated AC codes are given for reflection, for they would be identical to the corresponding ACs for concentration. The party support scores calculated in Table 5.2 are based on data source 15 for West Germany. As noted earlier, we possess alternative data sources for scoring party support in some instances. In fact, an alternative source was finally chosen for scoring the German parties on education. Source 14 was preferred over source 15 because of 14's larger effective sample size. The corresponding printout for data set 14 is contained in the section for West Germany, Part Two of this volume. Data set 15 has been presented here only for purposes of discussion, providing the reader an opportunity to gain a better understanding of our use of "alternative data sets" in evaluating the reliabilities of our measures. William H. Form defines "occupations" as "relatively continuous patterns of activities that provide workers a livelihood and define their general social status" (1968, p. 245). Cross-cultural studies of industrialized and industrializing societies disclose a general applicability of occupational classifications according to the complexities of skills and show consistency in status attached to those skills. This rough cross-cultural equivalence between occupations and status suggested that we accept indicators of status and income as surrogate variables for occupation. Whenever relevant sample surveys of our countries employed these alternatives to occupation, we coded our parties on them as well. As reported previously in our discussion of reliabilities, the correlations among the occupation-status-income variables were markedly lower than those for alternative categorizations or sources for the other five cultural differentiators. This finding casts doubt on the substitutability of status and income for occupation, despite the conceptual arguments in the literature. Our general procedure in selecting data sets for scoring the bases of party support was to identify variables or categorizations that served to differentiate among the parties in a system. Because the occupation variable tended to differentiate parties better than either status or income variables, some categorization of the occupation variable was usually chosen over status and income when a choice was possible. If a finer classification of occupation produced more differentiation among the parties, then it was selected over the cruder alternative classifications. But, most typically, the three-part division of occupations into white collar, nonmanual labor, and farmers proved serviceable for our purposes. In some Third World nations, even this crude classification seemed inappropriate to their nonindustrialized economies, forcing us to substitute either rough social class judgments (despite the finding discussed above) or to omit scoring the parties entirely. The results of our coding for occupation are given in Table 5.4 along with the results for all the other cultural differentiators. From the information in Table 5.4, one can see that we were more successful in coding parties on occupation than on any of the other cultural differentiators. We were able to code 87 percent of the 135 parties in the first part of our period and 88 percent of the 147 parties in the second part. The meanings of the statistics for the parties' attraction, concentration, and reflection of occupational groupings must be determined through comparison with statistics for the other five cultural differentiators. But there are distinct limits on such comparisons, because of great variations in the numbers of parties scored on the other differentiators. This problem will be discussed in the context of the results for religion. Rose and Urwin describe "religious divisions" as "the main social basis of parties in the Western world today" (1969, p. 12). In global terms, confessional conflicts constitute a pervasive source of systemic cleavage and are frequently reflected in the structure of party systems. The distribution of the support of members of identifiable religious factions among the political parties in a country is presumed to be a valuable index of both national integration and the structure of potential disintegration. Two significant dimensions of religious conflict are discernible. In nations characterized by religious heterogeneity the parties are likely to reflect the support of distinguishable religious groupings. In societies dominated by a single religion, however, intrafaith as well as interfaith conflict is apt to be salient to the structure of the party system. Indeed in some nations where more than one institutional church flourishes, such as the Netherlands, parties exist which attract support from individuals of diverse religious backgrounds but who have in common their anticlericalism. As Converse notes, "the question of negative as well as positive reference groups is . . . vital to the study of social cleavages" (1968, p. 279). Our assessment of religious support for about one-half of the parties in our study appears to support Rose and Urwin's contention of the importance of religious divisions, at least in comparison with occupation as a cultural differentiator. Table 5.4 shows that parties are |
|||||||||||||
|
|||||||||||||