Some parties are difficult to trace through time because of a succession of splits and mergers that complicate the establishment of party identity. In addition to obfuscating the rightful ownership of the party label, the occurrence of splits and mergers inhibits the process of institutionalization by narrowing the party's focus (in the case of splits) and broadening it (in the case of mergers). Interaction patterns tend to be sharply altered through both processes. The magnitude of these changes depends on the relative size of the splinter group and the relative size of the group with which the merger takes place. If the "splinter group" constitutes a majority of the party activists, one cannot even say that the remaining group represents the same party. Similarly, if the merger occurs with another larger party, one cannot say that the party retains its identity after the merger. Within these limits (see Figure 3.1), the consequences of the splits and mergers vary according to the relative sizes of groups breaking away from the party and joining the party. Purges are also to be considered as party splits.
The time of the split, purge, or merger must be considered in assessing this variable as an indicator of institutionalization. For example, it seems that splits or mergers before 1941 (World War II) would be a poor indicator of institutionalization studied in 1950-1962. Therefore, the period during which the change occurred needs to be introduced into the concept. No splits or mergers before 1941 are counted. Splits or mergers between 1941 and 1949 are weighted less than those between 1950 and 1956, for their impact on behavior will have been dulled by then. Similarly, those that occurred between 1957 and 1962 are weighted the most.
Operational Definition. A party is assigned a value on the discontinuity variable according to the sum of its scores obtained by application of Table 3.4, which incorporates the type of change and its recency. A party that had no organizational discontinuity since 1941 would be scored 0. One that terminated in our time period would automatically be scored 19, which is one point higher than the maximum of 18 points that might be assigned by summing the bottom row in the table.
No discontinuity Experienced minor split or merger Experienced two minor splits or mergers or one
major Experienced more than one major split or merger
or one major and one minor
Experienced minor split or merger
Experienced two minor splits or mergers or one major
Experienced more than one major split or merger or one major and one minor
A "minor" split or merger is defined as a loss or gain of no more than 25 percent of the party's activists--calculated in the case of a split on the number of activists before the split or calculated in the case of a merger on the number of activists after the merger. A "major" one is a change in party activists greater than 25 percent but less than 50 percent. As recalled from our previous set-theoretic discussion of establishing party identity (Figure 3.1), a party split involving 50 percent or more of the activists terminates the existence of a party and the same for a merger which finds the original party with less than 50 percent of the set of activists in the new party. Undoubtedly, coders encountered difficulty in applying these percentages to determine the magnitude of party splits and mergers, but the percentage figures served as guidelines for gauging different magnitudes. In the absence of specific references to party "activists," they used available references to party "leaders"