As promised in my first post, here's the deal with undervoting/roll-off. First, the definition is pretty simple for those not familiar. Roll-off is the phenomenon of voters not completing a ballot with a selection in each contest. It tends to occur more towards the bottom of the ballot in races like "Soil and Water Conservation District Supervisor."
Obviously, this year I was pretty interested in this phenomenon as it affected those of us down the ballot from Barack Obama. Many people were expecting landslides down the ballot for Democrats. We wanted an empirical estimate of what it would be, and more importantly, if it would be. Would there be an Obama wave to ride for the rest of the Democratic Party? Are young/AA voters more likely to roll off of the ballot?
From my analysis of previous elections, the answer to that exact question was no. What was frightening, however, was the nuance. While young and/or AA voters do not roll off at a rate greater than average voters, newly registered young and AA voters were very much more likely to roll off in the 2008 primary in NC. Simply put, there would be a downballot Obama wave, we just needed to learn how to surf it and make sure it got all the way to shore. That's why you saw all kinds of advocacy of the straight party ticket beyond standard turnout efforts. Because we knew turnout was not enough, and we knew by how much.
This study triggered my academic interest in the roll-off phenomenon. I just finished a paper studying the phenomenon in NC judicial elections and proposing an informal model of how to think of roll-off in a functional manner as it occurs within a single ballot. It's rather informal, so don't be surprised by the tone or absence of a thorough lit-review. If anyone is interested, here it is. Let me know in the comments section if you see anything you contest or have anything to say. It is, to say the least, a working paper that I hope to develop over the course of the next semester.
I don't like the fact that you can vote exactly down party lines on some ballots, although you had to apparently still select a President. Don't you think it promotes partisanship?
ReplyDeleteFirst, I don't understand why in NC the law was written that way, but it was certainly deliberate if you look at the text, so there must be some reason. (Section e right here.
ReplyDeleteOf course it promotes partisanship. But do the two parties drafting all of the legislation really want people crossing party lines? Think of it this way. On a given ballot say you've got 20 or so sets of pairwise of strategic interactions (contests). It is much less costly for each player if he can make an alliance with the other 9, making it all one big red vs. blue extravaganza. The strategy becomes simpler and the costs go down. More or less, it's like a market seeking an efficient outcome. So yes, it causes more straight ticket voting in all likelihood.
For the sake of argument, I'll put an interesting spin on it.
Given, straight ticket voting causes the more unified the parties become
More unified parties cause them to be more susceptible to median voter theory
As the median voter phenomena occur, the parties converge to the middle
Thus, straight ticket voting promotes centrist politics
Without it, you've got people all over the place and there is little order which allows for all kinds of political extremism as each individual politician seeks to excite the ends of his/her base more than the other politicians who are targeting the same base - because name recognition matters more than party id.
Just a very loosely sketched out thought, but I'd be interested to hear your response
-Will