6.44. See the discussion of the empirical results.)

7.05 Relations with Foreign Organizations. The more independent the party is of international party organizations or of influence from foreign countries, the greater the party autonomy.

The expectation that these types of dependence would be positively interrelated was overwhelmingly disconfirmed by the data. The strongest correlation between these variables was in the magnitude of .23 and that was the negative relationship between 7.01 (sources of funds) and 7.05 (relations with foreign organizations). The largest positive correlation was only .09, and the mean was -.02.

Under the original conceptualization, these five attributes were thought to constitute alternative indicators of party autonomy and thus were expected to be interrelated. But it appears that autonomy, like virtue, can be compromised in a variety of settings. Any significant dependency on an outside institution may constitute a sufficient violation of autonomy. Such violations do not come in bunches and thus are not interrelated. This leads to the new argument: to the extent that parties- for whatever reason-are not structurally independent of outside institutions, they lack autonomy. Presumably, low autonomy is related to special regard for the interests and wishes of the institution on which the party is dependent in one form or another. Whether this is true, and whether any type of dependence produces more responsiveness than other types, is a matter for empirical analysis outside the scope of this book. Given the original conceptualization and the empirical data in hand, the expectations were not fulfilled, and there is precious little evidence for the validation of our measures of autonomy.

There is some evidence, however, for at least one of the indicators. Variable 7.04, relations with domestic parties, was scored independently of the "alliance" tactics mentioned in the goal orientation cluster earlier. One would expect to find high negative correlations between the coders' overall judgment of the autonomy of domestic parties (as variable 7.04 was scored) and the parties' practice of such tactics as engaging electoral agreements, participating in legislative blocs, entering cabinet coalitions, and supporting other parties' candidates for presidential elections. These correlations were precisely as expected, ranging from -.52 to - .84, with a mean of -.71. Again, the numbers of parties involved in these correlations are small, but the limited evidence is strongly in line with expectations.

INTERNAL ORGANIZATION

Whereas all the previous variables applied to the parties' external relations with society, the remaining four variables in the ICPP conceptual framework apply instead to the "internal organization" of political parties.

Degree of Organization. We define degree of organization as the complexity of regularized procedures for mobilizing and coordinating the efforts of party supporters in executing the party's strategy and tactics. Note that this definition, unlike the idea of "strong" party organization in the literature, refers to structural differentiation rather than to the location of authority, which is the subject of the next concept, "centralization of power" Seven indicators were proposed as measures of the degree of organization:

8.01 Structural Articulation. The more numerous the national organs, the clearer their functions, and the more universal the selection of members--the greater the structural articulation and thus the greater the degree of organization.

8.02 Intensiveness of Organization. The smaller the basic organizational unit of the party, the more intensive its organization and thus the greater the degree of organization.

8.03 Extensiveness of Organization. The broader the geographical coverage of the party's basic organizational units, the more extensive its organization and thus the greater the degree of organization.

8.04 Frequency of Local Meetings. The more frequent the meetings of local party units, the greater the degree of organization.

8.05 Frequency of National Meetings. The more frequent the meetings of the national committee, the greater the degree of organizations.

8.06 Maintaining Records. The more effort devoted to publishing party propaganda, maintaining a research capacity, and maintaining membership lists, the greater the degree of organization.

8.07 Pervasiveness of Organization. The more numerous the socioeconomic groups are penetrated by the party and the greater the extent of party control of these groups, the greater the pervasiveness and thus the degree of organization.

All these indicators were expected to be positively intercorrelated, and they were with an average correlation of .38. The lowest correlations tended to be with item 8.05, frequency of national meetings, which correlated with the rest at only the .26 level. A factor analysis indicated that this item headed an independent factor in the cluster. With item 8.05 removed from the set, a single-factor solution resulted, explaining 52 percent of the variance among the remaining six variables. The degree of organization scale that they produced had a reliability of .82. Excepting item 8.05, whose relationship with the rest

go to page 153