I wonder if this isn't based in some part on mobilization efforts. A good GOTV effort can increase turnout quite substantially, and I think that Dems are far more likely to run campaigns of this kind (proposition one is sustained by many studies, I'm not sure about 2)
I’m not sure about evidence re: #2 either, but I’m also very cautious (and frankly bearish) about the impacts of GOTV campaigns. On the one hand, campaigns spend lots of $$$ on them, so they obviously have nonzero effects. But on the other hand, meta-analyses of field experiments have show some small effects.
I have been thinking for a while about what to include in an online bulletin of polling resources, methods etc. This will be something I include. More to come!
As presented in the graph, the major turnout difference among congressional elections is between off-year and presidential years. They don't belong on the same graph, because the higher turnout of presidential years goes with massive short-term forces and partisan swings. Analyzed separately, especially in the 21st century, both midterm and quadrennial elections show a moderate relationship between turnout and Democratic share of the vote.
I wonder if this isn't based in some part on mobilization efforts. A good GOTV effort can increase turnout quite substantially, and I think that Dems are far more likely to run campaigns of this kind (proposition one is sustained by many studies, I'm not sure about 2)
I’m not sure about evidence re: #2 either, but I’m also very cautious (and frankly bearish) about the impacts of GOTV campaigns. On the one hand, campaigns spend lots of $$$ on them, so they obviously have nonzero effects. But on the other hand, meta-analyses of field experiments have show some small effects.
Thanks for this article. This has been a question of mine.
Aslo are you familar with a poll aggregator for live caller, education weighted polls?
If not is there a list of polls that are live caller and weighted by education?
I have been thinking for a while about what to include in an online bulletin of polling resources, methods etc. This will be something I include. More to come!
Thank you Elliot. I appreciate your reply .
As presented in the graph, the major turnout difference among congressional elections is between off-year and presidential years. They don't belong on the same graph, because the higher turnout of presidential years goes with massive short-term forces and partisan swings. Analyzed separately, especially in the 21st century, both midterm and quadrennial elections show a moderate relationship between turnout and Democratic share of the vote.
Good eye! If I can get a hold of the data I will produce a graph with two separate trends and discuss further.