some type of party archive, repository, library, or research division as an institutionalized organizational resource.

In general, we limit our concern to maintenance of records as carried out by the national party organization, rather than local or regional organizations. But we do allow for national access to membership lists maintained by local or regional organizations if the national organization itself does not maintain such lists.

Operational Definition. In the absence of adequate data to make more specific coding distinctions, we were forced to be broadly judgmental in scoring parties on this variable. We employed an arbitrary progression of scores to weight our components according to their presumed importance in record maintenance. We gave parties the sum of the scores applicable from Table 9.6.

The highest possible score that a party can obtain from this table is 2 for publishing party propaganda, plus 6 for maintaining party archives, plus 8 for maintaining membership lists--for a total of 16. The lowest possible score is 0.

Coding Results. According to our scoring of BV806, political parties throughout the world vary greatly in maintaining records. The distributions in Tables 9.7a and 9.7b show parties spread rather evenly along the continuum of possible scores for BV806. Slightly more than 10 percent of the parties earn a code of 16 by rating highly on all the scale components: pursuing an energetic publishing program, maintaining an outstanding archive or research division, and keeping excellent membership lists. But a comparable percentage of parties does little more than publish occasional propaganda (code 1), while a few do nothing at all along this line (code 0). Our fortune in coding party organization turns upward with BV806 as we return once more to the 85-90 percent success level. For the purist, however, we must point out that there is a significant correlation of .40 between BV806 and AC806, indicating the return also of a tendency to downgrade parties for not maintaining records if the literature neglects discussing the activity.

"Pervasiveness of organization" refers to party penetration into mass social and economic groups representing politically significant sectors of the population. Typically organized into youth cadres, women's clubs, commercial associations, labor unions, agrarian leagues, and religious bodies, such groups articulate their aspirations while mobilizing support for the party. Duverger refers to such party-affiliated organizations as "ancillary" organizations, which he describes as "various bodies, created by the party and controlled by it constitutionally or in fact, which make possible wider or greater participation: wider participation, by grouping around the nucleus of members proper, satellite organizations made up of supporters; greater participation, through completing the political organization of the member by organization on the family, social, and cultural plains" (1963, pp. 106-107). Although Duverger sees these ancillary or "auxiliary" organizations as controlled by the party, they may, in the words of Clement Moore, lead an "equivocal existence," operating as pressure groups with "a measure of freedom to influence policy" and yet subordinating their particular interests to the party interest, sapped of their vitality by close party supervision (1965, p. 159).

"Pervasiveness" in our conceptualization is measured by the number of major sectors of society represented in organizations ancillary to the party and by the proportion of individuals in each sector involved in the organizations. We do not here assess the situation in which an organization controls the party--a subject for the "autonomy" variable cluster--but we do evaluate the extent of party control of ancillary organizations and the extent of organizational activity in the party's behalf. We really assess the consequences of party control rather than the processes of control, for we score party control or activity as "high" if the organization is

go to page 106