Tuesday, June 09, 2009


VA predictions

For the heck of it,

Creigh Deeds- 45%
Terry McAuliffe- 31%
Brian Moran- 24%

God knows, I'll be off by a ton. But I basically went with Pollster.com aggregate, and I added a few points to McAuliffe due to his ground gain and to Deeds because of the bandwagon effect.  

Monday, June 08, 2009


The big IVR test: VA Gov

Depending on whom you ask, Creigh Deeds is either surging to victory or all three candidates are lock in a tight three way primary. The traditional media such as the Washington Post are claiming that tomorrow's primary is all about "turnout". Those in the blogosphere or even the newer but still somewhat traditional Politico says that Deeds is a near shoo-in. Why the disagreement? 

Interactive Voice Response (IRV) polls done over the weekend show Deeds up by 10+ points and in the 40's. These surveys produced by Public Polling Policy and Survey USA are not viewed as reputable by mainstream outfits due to the fact that humans are not conducting the interviews and voters respond using a touchtone key pad. IRV polling causes according to some lower interview response rate, the ability for anyone to claim that they are a voter (even a dog), etc. Mainstream press make their views well known. Even Politico's Andy Barr seemed apprehensive writing the story he did (note the word "suggest" in the title).  

The benefits of IRV polling is that it is cheap and can be conducted easily over a short time span. Ergo, pollsters can poll the weekend just before the primary. They can pick up on potential trends that more traditional pollsters who use live interviewers and stopped polling before the weekend cannot. Thus, those who use this polling are seeing a potential explosion in support for Deeds. 

The traditional media who uses only traditional polls are left with data that shows Deeds in the lead, but in a very tightly fought three way race. In other words, a race in which turnout would make the difference. 

I'd put my bet on the IVR polls to be accurate as they were during the 2008 campaign. I'd be floored if tomorrow's primary is close. 

If the IVR polls are right, can the mainstream media finally stop ignoring them? If they are wrong, maybe I'll just keep quiet for a little bit. 

Saturday, June 06, 2009


Absentees are ACTUALLY UP in VA

You might be surprised to learn that despite proclamations by such politicos as Kos that "Turnout will be bleak" based on absentee ballots so far returned, absentee voting is actually up in Virginia. The article that Kos and many have quoted from the Richmond Times-Dispatch notes that "[the secretary state of the board of elections] Nancy Rodrigues said 11,549 absentee applications were sent to Democratic primary voters and 5,331 were received."

The last two times Virginia had competitive statewide primaries of a non-national scale (05 for LT. Gov and 06 for US Sen.) less than 5,000 absentee ballots were cast in the Democratic primary. In 2005, 4,525 absentees were cast and made up 3.57% of total ballots cast. In 2006, 4,721 absentees were cast and made up 3.00% of total ballots cast.

Absentees are still coming in until the day of the election, but the ones already return alone suggest that turnout should be UP in the 09 primary vs. recent years. 

While we can guess that absentees will make up somewhere around 3% of the percentage of primary ballots cast, we really have no idea whether it will be 3.00% as it was in 06, 3.57% as it was in 05, or maybe it will be 2.43% (.57% below 3.00%). 

We do not know the eventual total amount of absentee ballots, but we'll stay conservative and go with only the number that have been delivered back. With that in mind, here are simple total vote projection models based on absentee ballot percentage. 

3.57% - 149,328
3.00%- 177,770 
2.43%- 219,383

Now you would think the number of the absentee ballots returned will at least increase somewhat. If let's say another 1,000 absentee ballots are returned (a completely arbitrary number),

3.57%- 177,339
3.00%- 211,033
2.43%- 260,535

Unfortunately, I could not find in either 05 or 06, the percentage of ballots returned. My estimate is that more ballots will be returned than usual. Reason is that with all the talk of a very close election and turnout deciding the winner, voters might actually think their vote can make a difference. 

I would expect the same thinking of election day voters, "my vote can make a difference". As Kos put it, "what had been a boring race is proving loaded with last-minute drama." 

My guess (based on this completely anecdotal evidence) is that turnout is going to be above 200,000... between 210,000 and 220,000 if I were making a bet. 

Thursday, June 04, 2009


Creigh Deeds lost by 360 votes not by 323

It may be a small point, but it is one I believe needs to be made. I have seen it made in 9, count em 9, recent news stories. 

The Washington Post in their surprising endorsement of Deeds for the 2009 Democratic gubernatorial primary said in reference to the 2005 Attorney General race "The answer: Mr. Deeds lost by a scant 323 votes out of roughly 2 million cast despite being outspent 2 to 1."

The problem is that Creigh Deeds lost his 2005 Attorney General race to Bob McDonnell by 360 votes, not 323 votes. Why the error? 

The problem is that Deeds petitioned for a recount (who could blame him with the small margin?). 

The Washington Post itself reported that after the recount "this remained the closest statewide election in modern Virginia history, with 360 votes separating the two men out of 1.94 million cast."

Can we get it right? 

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]