V.O. KEY CONTENDED that "the nature of parties must be sought through an appreciation of their role in the process of governance" (Key 1964, p. 200). If this is true for the study of American parties, which held Key's attention, then it is certainly true for the comparative analysis of parties across nations, where there is far more variation in the position that parties occupy in the political system. One source of variation lies in the party's legal status. Obviously, legal parties have opportunities to participate in government decision making that are denied illegal parties. Another important source of variation lies in the party's political "strength." A "strong" party tends to have more impact on public policy than a "weak" one. Still other sources of variation can be found in the party's national orientation and its conditions of origin.
Duverger has probably provided the most general conceptual discussion of a party's "governmental status" in his chapter titled "Strength and Alliances" (1963, pp. 281-351). Still, his conception was restricted to the party's share of the products of electoral competition (he favored using the percentage of seats won in the legislature), and he did not consider some of the broader aspects of governmental status. The tendency of other scholars to concentrate on the electoral process of indicators of governmental status has focused research on measuring the "competitiveness" of party systems. While the approach to interparty competition has had its roots in the study of politics in Western democracies, the concept of interparty competition has since been generalized to apply to the politics of underdeveloped countries and European party-states (see LaPalombara and Weiner 1966; and Wiatr 1964). Broadening the view of party competition beyond simple electoral competition requires that other factors be considered to provide an adequate measure of the party's governmental status.
Within the ICPP project, "governmental status" refers to the nature and extent of the party's participation in national politics. A party that is "high" on governmental status (1) enjoys government favor rather than interference, (2) claims identification with the nation's chief executive, (3) holds cabinet positions, (4) engages in party activities throughout the country, (5) holds a majority of seats in the legislature, and (6) receives strong popular support in elections. Briefly, and in today's parlance, a party that is high on governmental status would be identified with the "establishment."
Note that the concept of governmental status is logically independent of the previous concept, institutionalization. A party may rate low in governmental status and yet be institutionalized; conversely, an "establishment" party, even in a one-party state, may not be institutionalized and may suddenly fold when confronted with a major political event, for example, the Convention People's Party in Ghana upon Nkrumah's downfall.
The concept of institutionalization differs from governmental status--and all the other dimensions in our conceptual framework--in one important aspect from the standpoint of research design. With the exception of "year of origin,'' the basic variables used to indicate institutionalization all require observations over time. Given the criterion of stable interaction patterns in the conceptualization of institutionalization, this is as it should be, for measures of stability necessarily involve observations over time. Therefore, our institutionalization variables each display only one score, which is the distillation of the party's experience for that indicator during our period of interest.
For governmental status and the other concepts, however, it is meaningful to score parties on their positions at individual points in time. One might imagine that a party's governmental status, for example, could differ dramatically from one year to the next, depending the outcome of a critical election or perhaps a forceful attempt to change the government. As explained in Chapter 2, our observations lack sufficient refinement to assess parties on these concepts on a yearly basis. In fact, our information base supports little more than distinguishing differences for a party between the "first" and ''second'' halves of our period.
Independent judgments of party positions for the first and second halves were made more readily within the governmental status variable cluster than within any others. Ups and downs in party fortunes tended to be re-