The Official Absentee & Early Voting Reports Thread

[Dan the Roman]

So far, OH absentee numbers are showing an increase in R percentage of the vote, though D are still leading.

2008

R:  19%

D:  32%

2012:

R:  24%

D:  30%

https://docs.google.com/spreadsheet/ccc?key=0AvEOdIaw0fPNdHVOZnFENDdDYVFTRi1UMlgxQ0F4OVE#gid=0

Because of how the count party affiliation, both these numbers are weak in terms of party ID.

Hard to make heads or tails of this. From what we have, I would say Democratic turnout is keeping pace with 2008(which is also supported by the Iowa numbers) while Republican turnout is up a bit. Some Republican voters in Ohio may well be Democrats who voted for Santorum in the Primary, especially in Cleveland where he did very well comparatively, but I suspect even with that they are seeing an increase. 30%(and rising due to the start of in-person) v. 32% for Democrats is probably all due to falling identification due to no primaries in 2010 or 2012.

[Dan the Roman]

NC voter registration, showed a net gain of D's over the last week of 7858.

There is now a net loss for the D's since 2008 of 80,224 in NC.

A couple points

1. Democrats are actually up in registration at this point compared to 2008. They made massive gains with in-person registration.

2. Those numbers are deeply misleading because there are almost 350,000 new voters registered in the state since 2008, only 20% of whom are white. That means a substantial number of African Americans, almost 40%, are registering as independents. Overall registration now is

71.7% White
22.3% Black

Compared with

74.9% White
21.6% Black
In 2008.

Furthermore, Democratic registration was at one point down more than 150,000, which implies Democrats have gained 90,000 votes on top of a net 250K non-white Independents. These registrants likely replaced older white Democrats who would be most likely to defect.

North Carolina is a very hard state for Obama to reach 50% in, but he will easily get 47% and almost certainly hit 48%. The demographics of the electorate are much more favorable than in 2008, which is not true in a lot of other states, and his normative margin that Romney has to overturn, to use the UK parlance, is probably 1.6% rather than the .3% he won by in 2008.

[Dan the Roman]

Absentees accounted for less than 9% of early votes cast in 2008. They also were more Republican than this year.

[Dan the Roman]

Absentee figures by Party in 2008

Republican: 122,412 (53.74%)
Democratic: 63,701 (27.96%)
Unaffiliated: 41,569 (18.25%)

Look familiar?

[Dan the Roman]

According to the latest Reuters/Ipsos poll 10% of Dems have already voted, only 4% of Republicans and 5% of Independents.
http://www.ipsos-na.com/download/pr.aspx?id=12073

Then OH is going Romney.  The votes turned in are running about 7 points for the Democrats.

Again party registration numbers in Ohio are deeply flawed (bangs head against wall).

And again, while not a perfect measurement, still telling, because the gap has closed by more than half. 

Voter registration has gone from about a 60-40 D-R ratio to about a 65-35 R-D ratio. Early voting/absentee voting has moved far less. Literally there are probably around almost twice as many "Republicans" as "Democrats" right now.

Also early voting in Cuyahoga is running ahead of 2008 numbers:

http://boe.cuyahogacounty.us/pdf_boe/en-US/2012/2008_2012InHouseVotingDailyComparison.pdf

[Dan the Roman]

2010 in Iowa was

43.7% D
38.0% R

With  349,216 Votes, 100K less than total requests so far.

[Dan the Roman]

Iowa absentees

Yesterday:

49.5% D
30.3% R
20.2% I


Today:

48.8% D
30.5% R
20.5% I

A relative drop is not an absolute one. More Democratic ballots were received yesterday than Republicans, as has been the case all week. Democrats gained about a net 1500 votes this week.

[Dan the Roman]

A relative drop is not an absolute one. More Democratic ballots were received yesterday than Republicans, as has been the case all week. Democrats gained about a net 1500 votes this week.

Actually, that is wrong.  The gap has been closing all week.  Branson initially posted how great the D's were doing.  They are not anymore. 

What part of that is wrong?

That the D's are gaining.  The gap has been closing, long term.

Not this week. Nor  in absolute terms.

10/12

Democrats 110,053
Republicans 57,341
Indies          38,801

Gap: 52,712

Today

Democrats 147,234
Republicans 92,072
Indies          61,833

Gap: 55,162

The Democratic absolute lead is growing and has been for a week and a half. But because that lead is less than the 17% margin overall from 2008, and turnout is on track to exceed 2008, the relative gap is falling.

[Dan the Roman]

Actually, you should note that there was a massive influx of D requests at the start (noted by Tender Branson on page 2 or 3).  It is like the D's entered into battle and fired all their ammunition with the first shot. 

The requests have been soaring and the percentages have been closing.

Not this week. Nor  in absolute terms.

10/12

Democrats 110,053
Republicans 57,341
Indies          38,801

Gap: 52,712

Today

Democrats 147,234
Republicans 92,072
Indies          61,833

Gap: 55,162

The Democratic absolute lead is growing and has been for a week and a half. But because that lead is less than the 17% margin overall from 2008, and turnout is on track to exceed 2008, the relative gap is falling.

Iowa's absentee requests are interesting:

Dem    49.1%
Rep    29.0%
None/Oth    21.9%
   
in 1998, they were:

Dem 47
Rep 28.2
Other 24.8

Both the R and D have improved, the the D's still hold the edge.  It has closed a bit.

Today, in terms of absentee ballot requests, Iowa Republicans have closed the gap below the 2008 levels, but not by a lot.

Party Reg    
Dem    48.1%
Rep    29.7%
None/Oth    22.1%

If in the remaining early voting period the Democrats were to gain 85,000 votes, the Republicans 75,000, and Independents 53,000, the raw numbers would be about

232,000 D
167,000 R
115,000 I

Or in percentages

45% D
32.5% R
22.5% I

In effect, the GOP would have closed the gap to 12.5%.

But the Democratic net advantage would have increased to 65,000 votes in absolute terms compared to about 43,000 in 2004. Romney would need to outperform Bush by 12,000 votes on election day with a lower election day turnout(since Iowa's electorate is not increasing) in order to win.

[Dan the Roman]

I read that if the GOP keeps the diff at 60k or less in Iowa they should win.  Thoughts?

Depends on total turnout. Ie. if 50% of voters vote early there are less election day voters left. So a 55,000 vote gap looks very different on a 600,000 turnout and a 450,000 one.

In 2004 the GOP trailed by about 53,000, 194K-141K. In 2008 they trailed by 93K, 250K to 157K.

Compared to 2008, Republicans are down 65K and Democrats are down 103K. Compared to 2004 Democrats are down 47K and Republicans are down 49K.

[Dan the Roman]

Thanks,

Will be interesting as my guess right now is about a 63,000 Democratic lead in the end.

[Dan the Roman]

Ohio updated this morning, 11.1% early voting in.  Looks like PPP, Marist and SUSA are still not credible been plenty of time now for the results to catch up...... 

Keep lauding those polls

http://elections.gmu.edu/early_vote_2012.html

Last week Obama had 74% of the early vote. Now it's 66%. People must be changing their early votes, lol.

I hope you don't actually think that's how polls work.

Are all counties reporting totals?

Look at the bottom of that page.

Also:

Counties update at different speeds.

Some counties do not include in-person numbers.

Some counties wait days to open returned absentees.

Worth noting that the one major county that both seems to count in-person and absentee numbers is both higher than that(15%) and has "Democrats" performing about 10 points better with returned absentees than with the requests everyone has been discussing re: party shifts.

[Dan the Roman]

Michael McDonald's released another update on early voting (link).

You must be logged in to read this quote.

So far there's very little evidence from the early voting numbers that Republicans have some huge enthusiasm advantage.

He's using the Marist crap, not the actual ballots. 

The "one stop" registration and mail registrations not processed was just under a 40,000 vote net gain in 2008.  The R's had a 70,000 + gain.  On election day, there will be 30-40 K more Republicans than there were in 2008.  Obama's MOV was just under 14,200.

NC is gone for Obama.

Eh, you have the same problem with those stats as you do with Early voting. The current electorate is less white than 2008, but also the gap between Ds and Rs is less. The same is true of North Carolina. Republicans may have closed the gap by 70,000, but Democrats closed the gap by about 350K in terms of the White-Black/Hispanic/Other numbers. On election day in 2008, NC registration was 73.1% white. Today it is 71.4% white. Nonwhites are registering independent.

Obama on a simple demographic measurement has probably picked up abut a net 75,000 or so voters since November 2008. He has probably lost far more than that because of his poor performance as President and generally poor poll ratings, but I wish hacks would shut up about the supposed D-R % gap as if it meant anything at all. It does not. In NC the Racial demographics, and age breakdown of the white vote matter far more.

[Dan the Roman]

Not all, no, but the big ones are. A cursory glance of the big counties shows that they aren't even close to 19%, let alone 21% as of 7 days ago.

Dan noted that counties don't always update their numbers daily and report them.

In fairness, both Cuyahoga and Franklin appear to do so, at least for in-person. Whats unclear is whether anyone is counting absentees as they come in. Counties clearly did so prior to the start of in-person voting, but they did not do so in an organized manner. Numbers were released, but it was unclear if these ballots were all of the returned ballots, or some selected at random.

Really there is no reason for county officials, dealing with in-person voting, a Supreme Court fight over in-person hours two weeks from now, and a whole lot of other stuff to bother going through every piece of mail they receive every day.

As a consequence I would probably use a number around 80% of the absentee requests. Which still gets you a number probably around 20% or so of 2008 turnout. Worth noting though its likely the electorate is smaller now, and that may well end up being 22-23%.

Iowa figures are almost certainly bunk though.

[Dan the Roman]

Michael McDonald's released another update on early voting (link).

You must be logged in to read this quote.

So far there's very little evidence from the early voting numbers that Republicans have some huge enthusiasm advantage.

He's using the Marist crap, not the actual ballots. 

The "one stop" registration and mail registrations not processed was just under a 40,000 vote net gain in 2008.  The R's had a 70,000 + gain.  On election day, there will be 30-40 K more Republicans than there were in 2008.  Obama's MOV was just under 14,200.

NC is gone for Obama.

Eh, you have the same problem with those stats as you do with Early voting. The current electorate is less white than 2008, but also the gap between Ds and Rs is less. The same is true of North Carolina. Republicans may have closed the gap by 70,000, but Democrats closed the gap by about 350K in terms of the White-Black/Hispanic/Other numbers. On election day in 2008, NC registration was 73.1% white. Today it is 71.4% white. Nonwhites are registering independent.

Obama on a simple demographic measurement has probably picked up abut a net 75,000 or so voters since November 2008. He has probably lost far more than that because of his poor performance as President and generally poor poll ratings, but I wish hacks would shut up about the supposed D-R % gap as if it meant anything at all. It does not. In NC the Racial demographics, and age breakdown of the white vote matter far more.

I terms of actual voters, the R/D gap is hugely important.  Most people register for a party based on their ideology.  Someone who just loves Obama doesn't say, "Hey, I think I'll register Republican."

Also, so far, the racial divide in NC is up by about 2%, but the party affiliation number is up by about the same amount.  Further, the electorate is skewing older.  

NC is gone.

The median age fell 8 years in two days. I would hold your horses on declarations about that. Secondly, black democrats vote very differently than white democrats. If Blacks are voting we know they are voting 95-5 Obama. Black Democrats will vote 99-1 Obama. White Democrats 75-25. Black Independents 85-15.

Demographics are especially important in a state like NC, far more important than Party ID given the number of DINOS.

[Dan the Roman]

Basically my point is that if Blacks are up 2 Points, but the D-R gap down, those missing Ds almost certainly were not Obama voters, and probably weren't in 2008.

Raw numbers matter here. More Democrats are turning out than 2008. More Blacks are turning out. More Republicans are also turning out. All you can do with EV/GOTV is get your voters out. Everywhere Early voting is happening, Democrats are voting. That Republicans are voting at a higher rate as well is obviously a factor that must be taken into account. But it does not indicate a failure on the part of Democrats, or any dropoff or sign of weakness. Rather it indicates Republican strength.

[Dan the Roman]

NC:

Dem    50.7%
Rep    30.5%
None/Oth    18.8%

Gap is down 0.7 points from yesterday, which should be disturbing because this should be the D's top week. 

From the same day comparison, D's are running 8.3 points behind where they were in 2008. 

Actual Changes

Democrats +54K
Republicans +58K
Indies +53K

So its only a net shift 4,000 votes Republican in a turnout of one million.

Whites +114K
Blacks +69K

[Dan the Roman]

Colorado Numbers in 2010

1,211,297

Republicans 40.7%
Democrats   34.6%
Independents 24%

[Dan the Roman]

With two weekends still to go, it'd be cool if we can call Nevada based on the early votes alone. Does anyone know if has Miller projected how much of the electorate will have voted early? Upwards of 40%?

Probably around 70%

[Dan the Roman]

No, that can't be. J.J. told me that black voters weren't very enthusiastic this year.

Actually, I've been noting that black turnout was higher in NC for a while.  It is also older there.

http://www.sos.ga.gov/elections/election_results/2008_1104/Earlyvotingstats08.htm

Early voting in Georgia in 2008 was 34.6% black, 60.7% white. This 33% is actually below 2008 pace.

Early + total election day voting in Georgia was 30.0% black, 64.1% white.


But why should anyone examine the real numbers when we can just crap them out?

It would be so wonderful if Mr. Kemp had kept up Karen Handel's updates about early voting on a day to day basis rather than replacing them with his page on zero-base budgeting.

[Dan the Roman]

VA

You must be logged in to read this quote.

http://www.vpap.org/updates/show/1094

I hope there is some sort of explanation.

[Dan the Roman]

Iowa early vote turnout has now surpassed 2008's early vote turnout.

What do you think of the change in Dem proportions in the state, then? The polls certainly seem to bode well for Obama, but the swing in finalg gap isn't wondrous. Enough to hold, do you suppose, considering that Obama won by 9 in '08?

He won the early vote by 18, so a lot of McCain's vote was on election day. 

Indeed, McCain won Iowa on E-Day, but Obama's EaV (EaV = early vote, as opposed to ElV = electoral vote) margin was strong enough to overwhelm McCain's E-Day voters.

Iowa early vote turnout has now surpassed 2008's early vote turnout.

What do you think of the change in Dem proportions in the state, then? The polls certainly seem to bode well for Obama, but the swing in finalg gap isn't wondrous. Enough to hold, do you suppose, considering that Obama won by 9 in '08?

Well in 2008 the Republican campaign didn't focus on early voting at all. It makes sense that now that they do, at least to an extent, some of their election day voters are moving to early vote and thus the Democratic margin is shrinking. But if Democrats can get their margin up to 70,000 by the end of early voting in Iowa (5 more days), then I think we're in a good position.

So far, the D numbers are dropping on a daily basis in IA.  

No, the D raw vote margin has been increasing basically every day. It's now at about +62,000.

Which seems to translate into a narrow victory for Romney (though I'll admit an Obama win isn't out of the question). Keith Backer says between 60K and 120K is a tossup; that's overly optimistic (certainly, if Obama's past 90K he's got this state in the bag, I think), but I'm pretty sure the cutoff is more like 80K, not 70K. Certainly, Lief, I think you can agree at 62K Romney's more likely to win IA than Obama.

Keith Backer fails Algebra 2.

Democrats won by 96K in 2008 with a turnout of 481K Early, meaning there was more of the vote to be won on election day, and Obama won in the end by 9. Democrats currently lead by 62,000 with a turnout of 557K, and with that increasing 30K a day. If Democrats lead by 67K with a turnout of 660K, Romney has to win on election day by much more than McCain because there is a much smaller slice of the electorate.

Let's  assume the election was tomorrow and that the electorate has grown by 5% since 2008(odd since it did not grow 2004 to 2008).

Obama wins 91/8/45 among early voters. This is quite good for Romney since evidence is indies voting early are being turnout out by the Obama campaign.

Obama gets a lead of

314K
243K

Or about 61,000 votes. Total turnout on election day would be 1063. Romney would have to win 53-47.

Now lets assume the final numbers are the requests which are 291K D, 208K R, 175K I.

We get

Obama 360K
Romney 275K

Problem is, election day turnout is now only 945K. Romney now needs to win by 8, 54-46, on election day to break even.

This whole thing assumes a ten point Romney win among independents who vote early. If we assume they even break even then we get

369K Obama
265K Romney

And Romney needs to win election day by close to 12, 56-44.

So even if the absolute margin stays the same, higher turnout raises the election-day threshold.

[Dan the Roman]

I mean, if early voting ended today for some reason and we only led by 62k then that wouldn't be great. But there are five more days to go.

Obviously, I mean if EaV concluded with Democrats ahead 62K. We'll see what the final margin is. But you shouldn't count on last-minute swings towards your candidate. Pretty sure that's one of J.J.'s rules, isn't it?

Out of curiosity, why do you think the Democrats will build on their 62K? I recall it was 58K like a week ago; but that swing doesn't seem to me to be large enough to indicate some sort of sudden last-minute surge in Democratic EaV support. Or is it?

Ballots have to be postmarked by Monday.

Its been increasing from between 600 and 1200 every day for two weeks. Minimum is probably around 65K(for Tuesday, not counting late arriving absentees).

[Dan the Roman]

Iowa yesterday: 

Dem    43.0%
Rep    32.2%
None/Oth    24.8%

Iowa Today:

Dem    42.6%
Rep    32.3%
None/Oth    25.2%
   
The final gap in 2008 was 18.0.   

Absolute Gap just passed 63K. Also we are probably looking at 43% or so cast early rather than 36%. Is their weekend voting this weekend?

[Dan the Roman]

Iowa yesterday: 

Dem    43.0%
Rep    32.2%
None/Oth    24.8%

Iowa Today:

Dem    42.6%
Rep    32.3%
None/Oth    25.2%
   
The final gap in 2008 was 18.0.   

Absolute Gap just passed 63K. Also we are probably looking at 43% or so cast early rather than 36%. Is their weekend voting this weekend?

Apparently a number of counties are doing in-person today as the only weekend day of the whole period.