perform these "members' welfare" activities. Their omission escaped comment within the literature, however, thus eliciting no code on the basis of "no information." This interpretation is bolstered by the high correlations between the BV and AC codes for the "members' welfare" variables, although they are not as strong as those found for the "propagandizing ideas and programs" variables where the argument is elaborated.
A comparison of the mean BV scores in Tables 7.8 through 7.12 with the means in Tables 7.4 through 7.7 tells us that parties are substantially more likely to engage in propagandizing ideas and programs than in activities which provide for members' welfare. According to our scoring, the most common members' welfare activity of political parties is "providing recreational activities" (BV655), whose mean value is just barely higher that that for "operating party schools" (BV632). All the other propagandizing activities seem to be more common than any of the members' welfare activities. Nevertheless, some parties in our study did employ as tactics all or nearly all of the members' welfare activi-
TABLE 7.9b: Early 1960s: BV6.52 Running Employment Services
TABLE 7.10a: Mid 1950s: BV6.53 Interceding with Government for Members
TABLE 7.10b: Early 1960s: BV6.53 Interceding with Government for Members
TABLE 7.11a: Mid 1950s: BV6.54 Providing Basic Education
TABLE 7.11b: Early 1960s: BV6.54 Providing Basic Education