broad confederation of individual studies and research
emphases with a tenuous relationship to one another than
it is a highly focused, clearly demarked or well-integrated
subfield of the discipline" (p. 138). In a section on "The
Search for Theory," he characterized the American
research as "self-consciously empirical and atheoretical"
(p. 145). He saw special promise in the work of
Schlesinger and said, "The more theoretical applications
and conceptual developments that are attempted, the more
models generated and explanations advanced, the better it
will be for a field that has not been known for the quality
or variety of its theorizing" (p. 148).
I agree with Crotty that research on American
parties can benefit greatly from closer attention to
theoretical applications, and there are other American
theorists besides Schlesinger and Downs. Kamens
( 1989), for example, has proposed a theory of party
development to account for the paradox that U.S. parties
have grown stronger organizationally since the 1960s
while becoming less important as vehicles for mass
mobilization. His explanation focuses on the
nationalization of politics and shifts in culture with the
rise of higher education and the mass media.
Nevertheless, there are many more examples of
theorizing about parties in the comparative literature than
in the American literature. For instance, I have already
cited theoretical efforts by Duverger, Katz, Strom, and Budge and Keman. There is also Ware's flow-chart
model of party behavior (1987, 108), Panebianco's theory
of party transformation (1988, 262-273), the theories of
candidate selection analyzed by Gallagher and
( 1988), Hamilton's well-developed model of determinants
of socialist party radicalism (1989, 30-31), the exposition
of coalition theory in Laver and
Schofield ( 1990), and Schlesinger's theory of the multinuclear party (1991, 151-
172). Party scholars can lament the lack of party theory
no longer. Our challenge now is to assimilate, develop,
and extend existing theory rather than wait for a general
theory to descend from on high. Even if students are
primarily interested in U.S. party politics, they could
sharpen their analytical skills and theoretical insights by
paying more attention to comparative political parties and
by reading the European literature.
Virtually all those surveyed reported that they were
familiar with the American Political Science Review (99%) and the Journal of Politics (91%).
LaPonce shows that all national journals are basically
ethnocentric, with the British Journal of Political Science least so.
Nevertheless, in a comparison of British and American journals, Crewe
and Norris found that "the proportion of American political scientists
reading U.K.-based journals was two and a half times the proportion of
British political scientists who read U.S.-based journals" (1991, 526).
Neumann's formal definition of a political party was "the
articulate organization of active political agents, those who are
concerned with the control of governmental power and who compete for
popular support with another group or groups holding divergent views"
4. Ware ( 1987, 17)
failed to recognize that "government"
means "public office" in the United States in the context of this
However, Schlesinger takes the opposite position: "I
would argue that the compulsion to seek an all-inclusive definition of
parties blinds us to the great varieties and types of political
organizations that the restricted view allows us to identify in
democracies, and therefore, the crucial distinctions that should be made
between them" (1991, 203).
Intercorrelations among indicators of age, leadership
competition, legislative stability, and electoral stability produced a
single factor solution for 150 political parties and a scale with a Cronbach reliability coefficient of .79 ( Janda 1980b, 143-144, 155
Given the proliferation of parties in the former communist
countries, one needs a reference guide to party politics, and some have
already been published. Szajkowski ( 1991)
listed more than 500 parties
in 12 countries in the region, and other books by Pribylovskii ( 1992)
and Abramov and
Darchiyev ( 1992)
described hundreds of parties and
proto-parties in Russia alone.
Laver and Schofield actually listed another technique,
dimensional analysis of parliamentary roll call votes, but this method
has been primarily limited to party analyses in single countries, not in
In contrast to the traditional "proximity" theory of
voting proposed by Downs, an alternative "directional" theory is
proposed by Rabinowitz,
Listhaug ( 1991)
. This theory
assumes that political issues are bipolar, and that voters decide
according (1) to the direction of their preference and (2) to the strength
of their preference. "Similarly, parties advocate different directions of
policy and present them with different levels of intensity" (p. 149).
Given a voter slightly left of center, "proximity theory predicts a
preference for the party nearest the center, while directional theory
predicts a preference for a party farther away" (p. 150). See also Macdonald,
Rabinowitz ( 1991)
11. Charlot ( 1989, 353)
credits Seiler ( 1986)
distinguishing between two sequences of party formation. In one, the
issue orientation precedes the laying claim to power and the resulting
partisan alignment. In the other, the partisan alignment precedes the
issue orientation and the laying claim to power. Seiler's work is in
In a later study of 108 parties in 19 elections over two
widely spaced elections, Rose and
Urwin ( 1975)
found little support for
regionalism as a basis of party cohesion.
One early, and lonely, exception is Anderson ( 1968)
who worked to relate the organizational theory literature to the study of
state and local parties.
Although the study of party organization is better
developed in the American literature, even there it is a neglected topic. Epstein ( 1991)
examined 238 articles and research notes published in
the American Political Science Review
from March 1986 through December 1990 and found only one item, a research note, on extragovernmental party organization.
This is similar to Duverger's concept of "community"
The six-item scale for degree of organization
(complexity) had a reliability of .82, as measured by Cronbach's alpha.
The eight-item scale for centralization of power had an alpha of .83.