Laelth's Election Projection

[Laelth]

I have crunched the numbers, and I am now ready to make my official prediction. I maintain that the polls that come out on the Friday before the election are sufficient to determine who will win. Little, if anything, is likely to change the minds of a significant number of voters at this point. First, here's what the national electoral college map ought to look like based upon current polling trends (measured solely by likely voters):



Since we know, however, that polls that include only likely voters must ignore newly-registered voters and voters who lack a home land-line telephone, a further analysis is necessary to account for the massive Democratic advantage in registering new voters for this election (as well as orchestrated suppression efforts on the other side). Based upon all the factors at my disposal, this is my prediction:





Methodology:

This projection assumes Ralph Nader and other candidates will get roughly 1% of the popular vote nationwide. The remaining 99% of the votes are distributed among the two major candidates, George W. Bush and John Kerry. Individual state figures were calculated by linear projection of current polling trends, as predicted in state polls, assuming that 70% of undecided voters will vote for Kerry while 30% will vote for Bush. These figures are further adjusted by four factors: 1) the state's popular vote for President in the 2000 election, assuming that states that went for the Democrat last time will be more likely to vote for a Democrat again, and vise versa; 2) the party in control of the governor's office of the state, presuming that the state's bureaucracy will give a slight advantage to that party's candidate; 3) known efforts to suppress the popular votes in selected states; and 4) voter outrage over suppression efforts in the 2000 election expressed in specific states. My calculations assume a 68% national turn-out of registered voters for this election.

Analysis:

Quite simply, Kerry should win. While both Ohio and Florida will be close, Republican efforts to suppress the vote in those states will likely not be enough to overcome the Democrats' huge gains in registered voters across the nation. In my final projection, John Kerry can lose both of these states and still win the election. Perhaps the real surprise of this election is that while Republicans have been concentrating their voter suppression efforts in key battleground states, a number of Southern and border states presumed to be solidly in Bush's camp (AR, LA, NC, VA, and WV) have been quietly trending blue, and these states, I predict, will go for Kerry based upon significant gains in newly registered voters that are not accounted for in the polls. I consider my 68% national turn-out estimate to be conservative. If 70% or more of all registered voters turn-out to vote, Kerry can expect to pick up Arizona, Indiana, and Mississippi. If 72% or more of all registered voters turn out to vote, Kerry has a chance to pick up Alabama, Georgia, and Nevada as well.

Of course, this prediction depends on our active participation in the process. Now is no time to relax. GOTV!

-Laelth

[NMDemDist2]

1. not sure we can take LA but AZ is a bigger prize and trending

blue hard..... hubby got polled tonite by the local TV station


If we stay close here in Phoenix, the rest of the state will turn us Blue. Check out the numbers I worked up from the Sec of State's registration numbers

STATE TOTALS as of 10/25/2004

Ind 655,554

Dems 914,264

Libertarian 18,261

Repubs 1,055,252

total 2,643,331

http://www.azsos.gov/election/Active_Voter_Count.pdf

According to Zogby's poll on Oct. 21:

Mr. Bush holds a small lead among men, while Mr. Kerry holds an unusually small edge among women. A bright spot for Mr. Kerry here – he leads among independents by a 62% to 31% margin.

so let's do the math shall we?

Bush gets Repubs + Libertarians + 1/3 of Indies = 1,292,031 48.87%

Kerry gets Dems + 2/3 of Indies = 1,351,300 51.12%

we win by a comfortable 2.25% margin

[Laelth]

3. That's good news.

The Arizona linear data I used from electoralvote.com shows a definite Kerry trend, but not a steep one. That, coupled with known voter suppression activity in AZ kept the state in the red until turnout hit 70%. But it's close, no doubt. It can be done!

:toast:

-Laelth

[Was_Immer]

19. Gov. Bill Richardson was on Larry King Live on CNN tonight

And he said he didnt think Kerry would take Arizona, but he predicted New Mexico would go Kerry, but he added, under the tiniest of margins.

[GainesT1958]

2. I trade Nevada with Louisiana on my prediction...

And Mo. is "iffy", but...otherwise, I'm with you on every state!:D

People are going to be surprised on Election Night..and some are going to be, as Atrios would put it: "Shocked, I say, just shocked"!:eyes:

:kick:

B-)

[abumbyanyothername]

4. He needs to factor in

The BUM rush factor in Nevada.

I am closing up shop here in Santa Monica. Leaving the wife at home to manage the kitchen reconstruction, and heading for Las Vegas to GOTV!!

Kerry will carry Nevada!

[cheshire]

5. Reno looks good too.

[Laelth]

6. Strangely enough ...

the data from electoral-vote.com shows NV trending red. That, coupled with known voter suppression activity in that state, kept it red in my model until turn-out hit 72%. Seemed strange to me, I admit, but that's what the numbers said.

And the polling data out of LA is just absurd and irreconcilable. LA went 52-45 for Bush in 2000, and I think that's a far more likely starting point for analysis than what the latest poll shows. My model also depends on targeted outrage over voter suppression in 2000, and LA is a state in which I expect to see some of that expressed. Plus, they have an interesting Senate race down there that I expect to draw people to the polls (the Republican in that race is polling only 44%). In all honestly, I think LA is a good bet.

:)

-Laelth

[flaminbats]

7. Indiana, Mississippi, and Alabama are swing states?

One Democrat told me Kerry couldn't win Indiana with Evan Bayh on the ticket, but according to your polls..he could win with just another visit.

Mississippi has among the lowest percentages of registered voters and weakest turnouts among those who are registered.

In terms of Electoral College predictions, the results on election night will surprise everyone. :thumbsup:

[Laelth]

15. Guess that includes me! LOL!

Indiana was a surprise to me too. I still have it red, even in my final prediction, but it has a Democratic governor, a substantial minority population, and it's trending blue. Those factors together make it much closer than I expected.

And Mississippi ... that will depend on African-American turnout. I'm expecting it to be very high by historical standards.

Both states still red, I think, but just barely. ;)

-Laelth

[flaminbats]

29. why my predictions are always wrong...

Time has taught me the future has already been determined. Free will is just another illusion, everything we do as individuals is nothing more than a reaction to an event which occurred in the past. The universe is nothing more than an infinite chain of actions from the past resulting in an infinite chain of reactions into the future. However..the meeting point, know as the present, gives living beings the impression that the course of time can be changed. We are part of that greater reality, which makes us just another cause and effect.

No person can leave reality and try to observe the whole from the outside, because we are all part of it! Nor could any of us move to points in which time and space does not exist without existing.

[Strawman]

8. Don't you think a 68% turnout estimate is too high?

I hope you're right, but historically, it would be quite an outlier. That's hardly a conservative estimate. 55% might be a more reasonable estimate. I think it could possibly be over 60%, but that would not be a conservative estimate. A conservative estimate would be 50%. A mid-range estimate would be around 53% and a high end, yet still somewhat reasonable, estimate would be anything between 55% and 60%.

http://www.fairvote.org/turnout/preturn.htm

I'm curious what this map looks like at 55% turnout.

[Laelth]

10. You're right, of course.

My 68% number is conservative only in the sense that I consider it on the low end of likely turnout for this year. The Secretary of State of Georgia is predicting 70% to 73% turnout in this state, and Georgia isn't even in play. Neither does it traditionally have high voter turn-out. From what I've seen, early voting turn-out is at an all time high. Admittedly, more states allow it this year than ever before, but even so, the sheer number of votes cast has been exceeding expectations. Indeed, I expect a lot of people to be surprised when (and if) all the votes get counted.

And the model looks much more bleak at 55%. ;)

-Laelth

[Laelth]

18. Oh, I should also mention that ...

my 68% figure is of registered voters ... not VAP (voting age population). 68% of VAP would, indeed, be historic. Note that 67% of registered voters voted in 2000, and Dave Leip's Atlas of U.S. Presidential Elections shows that 54.5% of the VAP voted that year. I think this year we're really looking at 70% of registered voters and 59% of VAP casting ballots, realistically--in that range, anyway. But, as I said, I chose the more conservative 68% of registered voters figure for my model.

Cool? :smoke:

-Laelth

[regnaD kciN]

21. Also...70% of undecideds going to Kerry???

55% seems reasonable, 60% something that can be hoped for, but 70% sounds to me like wishful thinking...

:shrug:

[Laelth]

26. Traditionally ...

when an incumbent sits for re-election, 66.66% (two thirds) of undecided voters vote for the challenger. I admit, the added 3.33% represents my feeling that the Chimpoleon is incompetent and a danger to the future of American democracy. But, really, an extra 3.33% for that isn't too much, is it?

:hippie:

-Laelth

[liberalpragmatist]

9. I'm skeptical

FIrst of all, I doubt undecideds will break 70-30 for Kerry. My guess is it'll be about 60-40, although the Bin Laden video might sway some of them Bush, so maybe it'll be more like 55-45. That should enable us to win a razor-thin victory.

And 68% turnout would be a miracle. Frankly, I don't think it'll happen. I think it'll be around 60%, period.

Also, don't rely too much on polls. Some states that are currently pro-Bush in the polls will be blue on election night. Some that are currently blue will be red on election night, assuming a close race.

I agree that a substantial victory is possible although at this point I'm skeptical that it'll happen. My guess is it'll be a razor-thin victory for Kerry.

[Laelth]

12. I appreciate and admire skepticism ...

but the past 70 years of our electoral history show that when an incumbent President is running for re-election, it's a landslide--one way or the other. The only real exception to this rule was the election of 1976.

Here's the data:

1996: Clinton (I) 49%, Dole 41%, Perot 8%: EC:379-159
1992: Clinton 43%, Bush (I) 37%, Perot 19%: EC:370-168
1984: Reagan (I) 59%, Mondale 41%: EC:525-13
1980: Reagan 51%, Carter (I) 41%, Anderson 7%: EC:489-49
1976: Carter 50%, Ford (I) 48%, McCarthy 1%: EC:297-240
1972: Nixon (I) 61%, McGovern 38%: EC:520-17
1964: Johnson (I) 61%, Goldwater 38%: EC:486-52
1956: Eisenhower (I) 57%, Stevenson 42%: EC:457-73
1948: Truman (I) 50%, Dewey 45%, Thurmond 2%: EC:303-189
1944: Roosevelt (I) 53%, Dewey 46%: EC:432-99
1940: Roosevelt (I) 55%, Willkie 45%: EC:449-82
1936: Roosevelt (I) 61%, Landon 37%, Lemke 2%: EC:523-8
1932: Roosevelt 57%, Hoover (I) 40%: EC:472-59

(I) = Incumbent

Perhaps this election is similar to 1976. I fear, though, that a razor-thin victory for us will turn into a razor-thin victory for the R's--as it did in 2000. The R's control the state legislatures in too many key battleground states. They're already paving the way for fraud challenges in the courts. I fear that we must win big, or we will not win at all.

This election, to me, feels more like 1992--focused on issues, energized electorate. R's championing more of the same when the electorate wanted a change and actually sought to be informed about what was going on. In short, this is a "fact-based" election as opposed to an "image" election.

If you're right and we have 60% turn-out and an undecided split of 55-45, I think we lose. But I honestly think we'll do much better than that. Here's hoping, anyway.

:toast:

-Laelth

[autorank]

14. Feels like 1980, a lot like 1980. Carter didn't get the hostages in time,

Bush didn't get Bin ShitHead in time. AMF Bush.

[Laelth]

16. I'd be happy with 1980-like results!

:headbang:

:dem:

-Laelth

[regnaD kciN]

24. It doesn't seem like either to me...

Perhaps this election is similar to 1976. I fear, though, that a razor-thin victory for us will turn into a razor-thin victory for the R's--as it did in 2000. The R's control the state legislatures in too many key battleground states. They're already paving the way for fraud challenges in the courts. I fear that we must win big, or we will not win at all.

This election, to me, feels more like 1992--focused on issues, energized electorate. R's championing more of the same when the electorate wanted a change and actually sought to be informed about what was going on. In short, this is a "fact-based" election as opposed to an "image" election.

1976 was a very unusual election because, although we had an incumbent, he had not only not been elected, he hadn't even run in a national campaign -- not even as someone's vice-presidential running mate. The largest election Ford had ever run in was for his House seat in Michigan. He was appointed to the VP slot when Agnew resigned, then took over as President when Nixon quit. That makes his status as an incumbent far different from anyone who actually ran for the job and somehow or other got 270 EV.

In 1992, OTOH, Bush 41 was behind from about the time of the Democratic convention on. By the time the debates came around, Clinton was up by double digits, and the race really came down to whether or not Clinton would make a last-minute stumble that would snatch defeat from the jaws of victory.

The only election this feels like, to me, was the last one -- with the added urgency for those on the Democratic side that can no longer rationalize, as they did even in the days following the 2000 debacle, that Bush would likely have no choice but to govern from the center in a bipartisan manner because of the closeness of the election.

[KharmaTrain]

11. There's A Big Overlooked Sample Here...

No one talks about it and I'm glad, since it's gonna be our surprise on Tuesday night. It has been in the past.

This election will almost boil down to city vs. rural...in almost every major state. That's why states like New York, Illinois & California have been solid Blue and will never be bluer.

The "battlegrounds" are really the cities: Cleveland, Philadelphia, Detroit, Milwaukee vs. the smaller towns. And this game will be going on big time this year in Colorado, Arizona, Nevada...even Texas.

It boils down to the massive undersampling of black and hispanic voters. Having done my share of elections here in Cook County, there wasn't a one where the black and hispanic vote wasn't under polled and would come storming through on election day.

In 2000, at this juncture the polls had Bunnypants over 50% and he was supposed to win the popular vote but could lose the electoral (still not sure how that figured)...anyway, the larger black and hispanic vote has a lot to do with Gore getting a majority and surely made a difference in Wisconsin, Michigan and New Mexico that year (and I'll throw in Florida, since we DID win that).

Again, you're not hearing much about the black and hispanic vote and I'm glad to see it since we're going to be very, very pleasantly surprised on Tuesday.

Earlier today I posted a link to a large rally staged by Michael Moore & Tom Joyner in Miami...not a single person responded to it.

[aquart]

20. What do you mean not a single person responded?

Nobody said Done or nobody clicked on the link?

[Dogmudgeon]

23. Not Philly -- it's solid Kerry Country

Even Pittsburgh and environs -- home of Rick Santorum -- are trending Democratic right now. We'll have a comfortable, if not huge, margin of victory here in PA.

Even our "rednecks" are getting educated. They're tired of voting for the people who impoverish them.

--bkl

[autorank]

13. Laelth you're a trip!!! Good work. Welcome to DU!

[Laelth]

17. My pleasure.

:D

and thanks for the welcome.

-Laelth

[Dogmudgeon]

22. Good work! But ...

... I think it's too optimistic.

You're projecting a 68% voter turnout with relatively wide margins of state victory for either candidate; I'm thinking it will be more like 55-60%, with smaller margins of victory, and will take until Thanksgiving to complete the tallies mandated from the results of GOP lawsuits.

In Pennsylvania alone, military personnel have been given until November 10th to get absentee ballots in.

Still, I hope I'm wrong and you're right. I'd love to see an election-day landslide for JK with 75% of all eligible voters voting.

I've done my part and intend to do more. Eight staunch non-voters are now on fire with the craving to turn Bush out, and at least three Republicans have crossed over (possibly 5, but I'm only counting those who actively gripe about Bush).

My only regret is that my grandmother won't get a chance to vote against Bush and see him humiliated. She passed away a few days ago, and we don't do Daley-style voting in my bailiwick. So we're all voting in rememberance of her.

--bkl

[Laelth]

25. Sorry to hear about your grandmother.

... And I don't think PA will be even remotely close. Your grandmother would be proud.

Peace!

-Laelth

[eyepaddle]

27. Wow,

Laelth you might be the wonkiest wonk to ever wonk! Do you do this professionally? At first I thought your assumptions were a tad high, but then I realized I was reading your post wrong....RV vs Voting Age Populaiton really made a difference.

I'll defecate if we win Louisiana though, I lived there for a couple of years and thought it wsa pretty staunchly conservative. Granted I was in the oilfield (and yes, I DID feel like I was a part of the problem) ;) There are some Dem roys of light though, like in the Senate. As long as we're here, LA DU'ers, I've been gone awhile--how's Landrieu workin' out?

I'd SOOOOOO love to break up the right wing hold of the South's electoral votes.:toast:

[Laelth]

28. No, just an amateur.

But I've been lurking on the RealClearPolitics blog all year and just listening and learning from the numbers geeks over there. I had to suffer through a lot of neocon bile, but it was worth it just to learn some of the basic principles and methods of poll analysis. Plus, I've been reading the actual poll reports from the pollsters and watching the state-by-state linear projection graphs at electoral-vote.com religiously. I figured that after all that self-education I should take a shot at a prediction and see how I fared. Lastly, I wanted to make a prediction that factored in the massive Democratic advantage in voter-registration this year as well as control of state government and active voter suppression efforts by the R's. These are things that the scientific-based numbers geeks don't adequately account for in their polling. I'm not sure that I've accounted for these factors correctly, either, but at least I weighted my calculations with these in mind, and that's something that the numbers geeks who get paid for it are not at liberty to do. I make no promises, but I feel like I've got good arguments in favor of this particular prediction.

And as for LA, I admit, that's my most radical call. LA has not been heavily polled this year, so it's hard to make a solid, scientifically-based prediction there. What I don't believe is that the R's have a solid lock on the South. FL, LA, AR, TN, MO, KY, and WV went for Clinton in 96. LA, AR, MO, KY, TN, WV, and GA went for Clinton in 92. I think it's unwise for the Democratic party to write off the South. The Plains states, sure. Those are gonna go R for the foreseeable future. But not the South (South Carolina excepted). If Democrats can energize the African-American electorate in the South, the R's can not take the South for granted. And that's what I'm seeing this year. My projection also takes into account the outrage of the African-American community over their targeted disenfranchisement in FL in 2000. That's why I'm willing to stick my neck out and call LA, AR, MO, WV, VA, NC, and FL for Kerry now.

In a few days we'll see whether my assumptions are valid or just the obviously biased ravings of a wanna-be wonk. ;)

:smoke:

-Laelth

[mconvente]

30. Yeah but Clinton was from the South

I think we have the best chance of getting Arkansas in the southern states. And Virginia of course if that counts as "South" - I think that's more Atlantic Coast state than true deep south.

[Laelth]

32. True.

But we've got Edwards on the ticket this time, and I think Clinton easily will pull his home state our way once again. Ultimately, though, I think the demographics of many Southern states make them winnable for Democrats in a way that the Plains states are not. Notice that even in my final projection I have NV and AZ red. That's because the African-American populations of those states are not large enough to tip them our way in a high-turnout election where anger over disenfranchisement is a major motivating factor. I still think that with 72% turn-out of registered voters we can win both AZ and NV, but if that happens, I'm predicting we win AL, GA, and MS too.

Of course, I could be entirely wrong. :dunce:

I could be crazy. :crazy:

We won't know until Nov. 3, and that's only if we're lucky.

:smoke:

-Laelth

[Bucky]

31. I'll say this for you-- you are a BRAVE predictor!

The idea that Louisiana, North Carolina, and Virginia will flip blue, but Nevada & Tennessee won't is certainly cutting against the grain. Among your reds, you have Alaska and Indiana giving Kerry 40% of the vote! Wowzers! If it's that good, however, I don't see how your model can say Kerry's only getting 52% of the popular vote. Those kind of numbers would probably mean cloer to 65% of the vote.

I'll have to take your Current Trend Indicator and still take Arkansas, Colorado, Iowa, and West Virginia from you. I hate to do it, but Bush is moving well in these last few days. It blows my mind that this is so, but somehow he's managed to keep people scared. Still, that gives us a 299-239 victory. Still wide enough to give us a win even if they steal Florida again (which would make it 272-266).

[Laelth]

33. I'd be completely happy with your prediction ...

just so long as we win.

Here's my thinking on NV--it's already been stolen. The state's actually trending red in the polls right now. They went red in 2000. They have a red governor, and we know the R's were working to suppress the vote there. These factors keep the state in the red in my model until turn-out hits 72% of RV's. That's possible, of course, but I kept my model conservative at 68% of RV's (just so I wouldn't be too disappointed). LOL!

If you take a look at electoral-vote.com, you'll see that VA and NC are both "weak" Bush, whereas TN is strong Bush. I also used the linear projection models from that site pretty faithfully ... just projecting out the lines and seeing where they fell, without adjustment (and I've got IN going to Bush by a razor-thin margin, in part because of how sharply blue the state is trending right now).

Either way, it's been fun, and I certainly hope I'm right. I don't want another razor-thin victory that we end up losing in court.

Kerry in a landslide!

-Laelth