Richard S. Katz and Peter Mair, "Changing Models of Party Organization and Party Democracy: The Emergence of the Cartel Party," Party Politics, 1 (January, 1995), 5-28.

First Paragraph:
One common thread that has run through the literature on political parties, essentially since the time of Ostrogorski (1902), and that has also run through the vast variety of typologies and analyses (both normative and empirical) that have been presented in that literature, has been the view that parties are to be classified and understood on the basis of their relationship with civil society (see, for example, Duverger, 1954; Neumann, 1956; Panebianco, 1988). This has had two implications. The first has been a tendency to setup the mass-party model as the standard against which everything should be judged (Lawson, 1980, 1988; Sainsbury, 1990). The other has been to undervalue the extent to which differences between parties may also be understood by reference to their relations with the state.

Figures and Tables:
Figure 1: Parties of the cadre or caucus type.
Figure 2: Mass parties act as links between the state and civil society.
Figure 3: Parties act as brokers between the state and civil society.
Table 1: The models of party and their characteristics.

Last Paragraph:
As we noted at the beginning of this paper, much of the contemporary literature speaks of the decline or failure of parties, an emphasis which, from our perspective, is largely misconceived. In fact there is little real evidence to suggest that the age of party has waned. On the contrary, while in some respects parties are less powerful than before - enjoying, in the main, less intense partisan loyalties, lower proportions of adherents, less distinctive political identities - in other respects their position has strengthened, not least as a result of the increased resources that the state (the parties in the state) places at their disposal. To be sure, if one takes as the standard the model of the mass party, as much of this literature appears to do, then the mainstream parties are perhaps less powerful than before. That is, they are less powerful mass parties. But this, we have argued, is an inappropriate standard, which fails to take account of the ways in which parties can adapt to ensure their own survival, and which ignores the new strengths that they can acquire in short, different parties. To speak of the challenge to party rather than of its decline or failure, is perhaps to be on surer ground, albeit also fundamentally misconceived. For what we now see in western democracies is less a challenge to party in general and rather more a challenge, inevitably so, to cartel parties in particular.