(Page 34 of 34) - Value Cleavages and Partisan Conflict: The 2004 American Presidential Election in Comparative Perspective authored by Gunther, Richard., Beck, Paul., Kuan, H.C.. and Smidt, Corwin.

. This conclusion stands in sharp contrast with the title, the bold statements on the dust jacket, and the argument of the first five

chapters of Morris Fiorina's 2005 book, although our findings are most consistent with the data presented in the final three chapters of
Fiorina's book. This apparent inconsistency is the result of the fact that Fiorina addresses this question from two different perspectives.
His assertion that there is no deep cultural cleavage in American society is based upon an examination of marginals and simple
comparisons between the distribution of opinion between "red" and "blue" states (at the aggregate level), between the old and the young,
or between men and women. Seen from this perspective, there is no marked polarization of the American public; and this same
conclusion can be drawn from the data we presented in Table 1 and the accompanying narrative in the text. These seemingly
incompatible findings raise a fundamental question: what defines the "real" politics in a country--the distribution of public opinion at the
mass level, or the manifestation of opinion in electoral behavior. We regard public opinion as "raw material" which is then molded into
political behavior through the electoral mobilization efforts of elites, the media and partisan secondary associations. It is the mobilization
of segments of the electorate into sharply opposing camps that has the most direct and powerful impact on the nature of politics in a
democratic system. Political elites and parties structure the political agenda, frame political discourse, engage in electoral competition
with each other, and formulate policies, and the way they choose to perform these functions may be substantially out of step with modal
public preferences. Sharply divisive conflict at the elite level, with each party's respective electorate mirroring that polarization, may be
an unfortunate departure from the moderate preferences of the modal voter, but it is not a "myth." While our interpretation clashes with
the bold language used by Fiorina in many parts of his influential book, he, himself, acknowledges this undesirable reality, and attributes
"the hijacking of American democracy" to the takeover of once pragmatic catch-all parties by ideological purists and extremists. In short,
he attributes the polarization of the American electorate to changes in the intermediation roles played by parties and their allied secondary
associations and communications media.

. We would like to thank Hans-Jürgen Puhle for providing us with the analysis of Uruguay's trajectory of political development, upon

which this discussion is based.

. Indeed, when supporters of the Concertación parties are examined separately it is noteworthy that the mean-scores on this Traditional

Conservative scale more clearly divide Socialists from Christian Democrats (by .6 points on a 10-point scale) than they do the aggregate
of Concertación supporters from the right-wing opposition (.1 point).

. The next stage of this research project (CNEP III, within which several Asian and African countries have been incorporated),

therefore, will add to this Western European core a number of additional value questions that should have greater resonance and political
relevance in a wider array of countries.

. Correlations between our core measures of fundamental support for democracy (agree/disagree, "Democracy is the best form of

government for a country like ours" [DemBest]; and preferences for democratic vs. authoritarian regimes) were significant at the .001
level only in Spain with regard to Traditional Conservatism (Pearson's r = .11), Social Democracy (.14) and Postmaterialism (.12). Only
one of these value scales (Modern Social Democracy) in Greece was correlated with DemBest at a level of statistical significance of .01
(Pearson's r - .09). None of the other values scales in Spain, Greece, Uruguay, Chile and the United States was correlated with support
for democracy at a level of significance of .05 or better. Due to the incoherence of the values clusters that emerged from our analyses in
Hungary and Hong Kong, these correlations were not calculated.

. The percentages of the variance in party ID were somewhat lower, ranging from .25 in the United States to .11 in Uruguay, where

only a small minority of respondents claimed to identify with a political party.