In reviewing some of the comments in the Oregon thread, I realized that there were a couple of obvious things I could be doing to further improve my turnout model:
(i) Adding a time trend -- turnout has been VERY high in the past several primaries,
and
(ii) While still using the percentage of the voting-age population as the explanatory variable, include the number of Kerry voters in 2004 as an independent variable.
This increased my turnout estimate in Oregon to about 660,000 voters. Obama also gains about a point in his statewide margin against Clinton because we're now showing turnout to be especially heavy in the Portland areas.
So our "official" projections are Obama by 13.8 points, and turnout of 661,470. Full results are below. Kentucky projection to be up later this morning.
Tuesday, May 20, 2008
Revised Oregon turnout estimate
-- Nate at 5:40 AM
Labels: oregon, primaries, turnout models
Subscribe to:
Post Comments (Atom)
16 comments
Hey... love your model.
BUT... im a little confused. I read that 800'000 ballots had already been cast and that they were predicting the final turnout of Oregon to be around 1 million.
(http://www.oregonlive.com/politics/oregonian/index.ssf?/base/news/1211261111282260.xml&coll=7)
Or is that including the Republican primary as well? im confused
Turnout in OR has already surpassed 700K, according to news reports. I'm not sure your model works for mail in votes.
Last I checked, that does include GOP primary voters. Anyone want to take a stab at how many votes go to Paul?
alpaca,
I would not be surprised to see Ron Paul take 18-20% of Oregon.
Your work is fantastic. A great source for electoral math this whole cycle. Please tell me you've been approached by the Obama campaign about doing this for more than just entertainment?
I look forward to marveling at how close you are to the real results.
You are the Moneyball of American Politics.
I echo what anoodle said. Great work.
Were any projections done for Kentucky? (I've looked but didn't find any - could easily have missed them).
Could you throw us please one extra number - no need to provide analysis, just the bare number - what your OR prediction would be *without* ideology, i.e. with the NC/IN model? +3 Obama? +5? +7? Would be useful for all of us to understand whether anything dramatic did take place or not.
I keep posting methodological quibbles so let me tell you once again, this work is awe inspiring.
Could you please remark about the maps that were released by Karl Rove? Thanks.
As of 3:30pm Monday,
CD-1 had 75,899 of 143,294 Dem ballots returned, or 52.97%.
CD-2 had 66,530 of 128,324 Dem ballots returned, or 51.85%.
CD-4 had 108,092 of 195,180 Dem ballots returned, or 55.38%. (52,280 of those ballots were in Lane).
CD-3 and CD-5 share Clackamas county and I don't know what the split is. 203,648 of 399,920 Dem ballots returned, or 50.92% combined.
Overall, 454,169 of 866,718 Dem ballots had been returned by 3:30pm Monday afternoon, or 52.40%. PA was the closed primary record at 55%, for reference.
Link:
http://blog.oregonlive.com/mapesonpolitics/2008/05/ballot_returns__5_19_08a.pdf
I'll be linking your post/predictions in my Oregon Tracking Diary on Kos tonight, along with a few others.
PN,
In the last two primaries in Oregon, 29 percent of the ballots came in on the last day. So that would extrapolate outward to turnout of about 640,000, pretty darn close to by estimate.
Note from Kentucky: the linking of cities to the congressional districts caused a deep chuckle.
Covington is generally treated as a small political pariah in the booming suburban three-county region known simply as "Northern Kentucky." The city sends one of our very rare African-American representatives to the state capitol, but he's surrounded by the most aggressive conservatives in the commonwealth. No one here would use Covington to define the Fourth district.
Corbin,similarly, is a useful interstate exit for driving into Eastern Kentucky, but it's simply not a cultural or political center of the Fifth District.
And Murray, though a lovely place, has no chance of being the center of that long gerrymandered First District. (For cultural purposes, I'll note that the eastern most part of that district really is Appalachian in culture, being based in what we call "the Knobs" and racked with deep poverty and population decline.)
Mostly, we just don't do cities--and Louisville barely consents to work with the rest of us. But you can simply name the districts by region-- Far West, Central West, Louisville, Northern Kentucky, Eastern Kentucky, Central Bluegrass--and we'll all know where we fit (except for the east end of that silly First District, of course).
The turnout in Multnomah County (Portland area) is a bit lower than the statewide average as of yesterday: only 49% compared to 52% statewide.
http://blog.oregonlive.com/mapesonpolitics/2008/05/ballot_returns__5_19_08a.pdf
There is a large bloc of Non-partisan independents in Oregon that Obama would really do well with, *if* this were an open primary...sigh.
Poblano - do you know how Oregon's bloc of independent voters (24% of all registered voters) compares with other states? Seems on the large-side to me. Is there any relationship between the percentage of independents in a state and Obama's support?
As I have said in previous thread, it seems turnout is likely to be around 700K. Good revision Poblano!
How well does the candidate visit variable stand-up when the other variables that might make an area or state attractive to that candidate are well-controlled? Seems like visits might reflect more than affect. But Pobl. did the math.
I certainly hope you are right. I'd see a 19% loss in Kentucky as a big win for Obama.
I think it'll be 30+% myself. But... I'm happy that you think it'll be as low as 19% as you've been correct in the past.
I agree with your number. I took 866,718 and multiplied by .74 and that's why my turnout prediction in my post tonight is 641,371.
Post a Comment