(Page 29 of 34) - Parties Change: So What? authored by Harmel, Robert.

Finally, our last proposition deals with our two models of sequenced, “reactive”

change and their relative likelihoods of resulting in disruptions of internal party politics.
Our proposition is derived primarily from the distinction between the purposes – at least as
argued above – for engaging in one or the other type of reactive change. Cumulative
reactive change is the born of the desire to forge greater coherence in the party’s
ideology/issue profile and/or its organization, whereas compensatory reactive change is
designed specifically to “keep peace” in parties where prior change has resulted in self-
assessed “losers.” Given not only the purpose of compensatory change, but also the
observation that greater coherence is often won at the expense of one interest or another,
with accumulated losses as the reactive changes continue, it follows logically that

P18:

For situations where change causes further change, the cumulative type is more
likely than compensatory change to result in severe disruptions to internal politics.

Conclusion

It has been our purpose here to encourage more empirical research on the

consequences of party change. Our method has been the generation of a number of
potentially testable propositions, tying certain party- and change-level “mediating factors”
to the likelihood that party change will result in various intended and unintended
consequences. Ultimately, the value of the exercise should and will be judged -- even more
so than by whether the propositions withstand empirical testing -- by the extent to which
the propositions prove testable at all. And that will depend, to a very large extent, on
whether indicators can be found – or developed – for each of the important concepts. If
not, then this exercise in agenda setting could also prove to have been an exercise in futility.
We have already spoken to the measurement issues related to party change and to the
dependent variables, but until now have said nothing about the measurability of the all-
important “mediating factors.” It is incumbent upon us to do so now.

Existing data sets are at least indicative of the availability of measures of some

critical variables, though individual researchers may wish to extend their procedures to
additional parties. While Harmel and Janda’s Party Change Project (PCP) has not
collected data on party institutionalization, Janda’s earlier International Comparative
Political Parties (ICPP) project did do so for multiple indicators for the 153 parties
covering the 1950-1962 period. Short of replicating the laborious procedures for more
recent years, one may be satisfied to use “age” as a surrogate for instititutionalization.

so, then party ages for older parties (i.e. those “significant parties” born prior to 1963) may

Janda (1980; 1990: 10), for instance, uses age as one indicator of

institutionalization, along with electoral stability, legislative stability, and leadership
competition.