you need a title.
On Sun, 11 May 2003 Jennifer_Fitzgerald(a)brown.edu wrote:
Hi, all. Comments/criticisms/questions are welcome on
our working
abstract. Thanks a lot!
Abstract: Research showing the positive effects of citizen initiated
ballot propositions on electoral turnout suffers from omitted variable
bias by ignoring the presence of items that land on ballots via state
legislatures.
this is not really omitted variable bias. the latter is when you have the
same observations and a control variable was omitted. in your case, you
are saying that a different set of observations would provide a better
test.
Further, past work mistakenly implies that citizen
initiatives are substantively different from other
types of ballot
proposals due to their origins among the people. In this paper we
contribute to this literature by examining the effects of citizen
initiatives and legislative referenda, which we argue are much more
alike than has been previously acknowledged, on turnout in midterm and
presidential elections.
you could say all of that above better. as I understand what you're
doing, here's something like what I'd say:
A large literature over the past xxx years has attepted to demonstrate
that the existence of citizen-initiated ballot propositions increases
voter turnout. This effect has long been the subject of historical,
qualitative, and journalistic analysis and has now been confirmed by a
number of extensive systematic studies. Unfortunately, this extensive
literature has been taken to have major implications for democratic theory
and the design of electoral institutions despite the fact that not a
single article has included a control group -- ballot propositions
initiated by legislators instead of citizens. We collect the first
dataset of this type of proposition and demonstrate that they increase
voter turnout in exactly the same way as citizen initiated propositions.
This result would seem to invalidate the claims in this literature that
the increase in turnout has anything to do with the fact that the
propositions studied in the literature were citizen-initiated, and calls
into question the policy and constitutional implications suggested. In
contrast to the pervasive "direct democracy" view, we argue that creative
politicians will pursue their policy interests through whatever mechanisms
are available, whether they be citizens or full time legislators. The
form of the institution is much less important....
Gary
We find that the number of citizen initiatives
and the number of legislative referenda are each positively correlated
with turnout rates in midterm elections, but not in presidential
elections. We also find that once the number of ballot propositions on
midterm ballots reaches a certain threshold, the impact on turnout
declines. This offers support for the ballot fatigue hypothesis, which
until now has not been supported empirically.