This is also true where the process of organizing an initiative is outside thecontrol of any party, established or opposition (even though new parties, as yet
barely distinguishable from interest groups or social movements, may still beneWt
from a free-for-all). In the long run, however, such a situation approaches
unmediated direct democracy and can be expected to have some of the negative
eVects on all parties, particularly in terms of agenda setting, as spelled out above.
The form in which electors are asked to vote in both referendums and initiatives
is generally to approve or disapprove a speciWc proposal such as joining the EU
(e.g. in many of the Eastern European candidate members in 2004 ). Majority
disapproval means that the situation remains unchanged. In other words, the
alternative to the proposed change is usually the status quo. Voters tend to favor
no change as a safe alternative when they are confused or unclear about the eVects
a new proposal will have. This happens quite often. Insofar as popular voting
undermines the agenda of established parties, this tendency favors them.
The status quo, and the government position, are also reinforced in federal
systems such as Switzerland and Australia where not only is a national majority
required for a measure to pass, but also majorities in a majority of states (e.g. in
Australia in four out of six). This puts considerable blocking power in the hands of
states with small populations which may even form the majority of federal units.
Federal as well as democratic values are being protected here and territorial
minorities may be eVectively safeguarded from a steamrolling popular majority if
they predominate within some of the units. Similar eVectsXow from provisions
that a popular vote is valid only if total voting passes a stipulated threshold
(commonly, 50 percent of electors or voters in the previous general election). In
several recent Italian votes, opponents have urged abstention. This eVectively lines
up the apathetic and uninterested on your side under a stipulated turnout rule and
has been very successful in defeating new proposals.
Phrasing the policy question as a yes or no choice—an almost universal
practice—helps to simplify decisions for voters and avoids the kind of Condorcet
voting cycles which might arise from rank-ordering a set of alternatives (Arrow
1951 ). Where the maneuver is permitted, opponents of the measure may, however,
also seek to dilute its eVects by putting a series of modiWed alternative proposals on
the same ballot (often involving little change to the existing situation). Even if the
original proposal also gets a majority, the composite policy which emerges as the
Wnal outcome will dilute the eVects of change fairly eVectively. Tactics of this kind
are common in the unregulated situations typical of California, and have been used
in Switzerland. In more regulated situations, such tactics are prohibited and courts
stipulate that only one proposal may be put in each policy area (Mendelsohn and
Parkin 2001 ; Le Duc 2003 ).
All this reinforces the general point that parties are better served by extensive
regulation of direct policy votes than by an absence of them. Quebec, where a
whole codiWcation of the process has been passed into law, can be contrasted with
604 ian budge