Blog Roll
Who's My Legislator?

Virginia Blogs
All Things Education
Augusta Free Press
Crew of 42
Dixie Pig
Greater Greater Washington
Loudoun Progress
Ox Road South Blog
Pilot on Politics" blog
Power for the People
Richmond Sunlight
RVA Politics
Shad Plank
Vivian Paige

Find more about Weather in Arlington, VA
Click for weather forecast

Advanced Search

Another Herring Victory Party Comes Later This Week

by: Dan Sullivan

Mon Dec 16, 2013 at 07:34:07 AM EST

Obenshain photo ObenshainatSHD_zps1ce62d10.jpgThe margin might be the closest in US election history and is certainly the closest in modern history. The State Board of Elections certified the results when they convened the fourth Monday in November. The margin is 165 votes out of 2.2 million cast. It's over this Thursday (or Friday at the latest).

Obenshain, because of the margin, is within his rights to petition for the recount. The Recount Court has been empaneled and convened twice. Motions by both camps have been heard. The court has established procedural orders and a timeline for the recount.

Chesapeake, Fairfax, and Alexandria, because of their size, are going to begin today, a day earlier than the remaining jurisdictions in Virginia. Tomorrow, the remaining jurisdictions will begin and those three will continue as required. The State Board of Elections will convene on Wednesday, December 18, as will the Recount Board (to recertify the local results). All the jurisdictions will recount and then the SBE will recanvass the recounts. If there are any ballots in question from the local recounts, it will be the SBE that adjudicates.  

Anyone expecting fireworks during this procedure will be sorely disappointed. This is a very narrow process. The kinds of games that are available in other states are not in Virginia. Barring malfeasance by election officials, any drama would have to be manufactured. By design, the Virginia process is streamlined. In this case, this is the retabulation of the votes that were cast for Attorney General only on election day. Provisional ballots that were rejected will not be revisited nor will absentee ballots that were not counted. By court order, the election officials may not draw down the number of ballots that were cast (they may not reconcile the numbers; yes that is allowed in very narrow circumstances on election day). Machine errors and voter eligibility are not at issue.  

Dan Sullivan :: Another Herring Victory Party Comes Later This Week
With rare exception (and none that has been evinced in this election), the recount officials are hard-working election officials. They are not partisans during the execution of their responsibilities. They just want to get it right.

Each and every ballot will not be examined.

    There are three types of ballots in Virginia:
  • Direct Recording Electronic (DRE) - the touch screen machines
  • Optical Scan Ballots
  • Paper ballots

DREs are really easy. The recount officials will examine the machine printouts. They will look at the tapes that are from election night. If the tape is not clear, they can be reprinted (Note: these tapes have already been inspected and none were found to be illegible).

Optical scan ballots will all be re-run through the tabulator. This is a change from the past. The machines are programmed to count only those ballots cast in the election for Attorney General. The machines are also programmed to not count over-votes, under-votes, or write-ins. These are all set aside and hand-counted. There could be instances when the number of ballots tabulated does not match the number of ballots. Reconciling those number differences is not a reason for hand-counting. Only those ballots not counted will be examined.

    Terminology for the purposes of the recount:
  • Over-votes - ballots on which a voter selected more than one candidate
  • Under-votes - ballots on which, apparently, the voter marked no candidate
  • Write-in votes - what they sound like

However, if a machine is not properly setting aside the over, under, and/or write-in votes, and it cannot be fixed, the election officials do have the opportunity to hand-count all the ballots.

Of course, paper ballots are hand counted. Hand counting of paper and uncounted optical scan ballots is conducted in accordance with SBE standards (posted on the SBE website).

Only recount officials can challenge ballots. It only takes one of them.

If it is clear that there is an indication of support, an affirmation of support for only one candidate for Attorney General, then the ballot will be counted for that candidate. That does not include write-in votes for candidates who are on the ballot. Virginia law precludes write-ins for candidates on the ballot.

A lot of voters are not familiar with standardized testing procedures or have other experiences that make them familiar with properly completing a ballot. For that reason ballots get marked in a variety of ways.

If there are identical marks for each candidate, but then there is an additional clarifying mark, that is a positive affirmation, it will count for that candidate.

If only one candidate receives a mark of support in the target area of a candidate, that will count for that candidate (unless the mark is clearly negative or unsupportive). The target area is the area near the oval. The candidate area is the area around a candidate's name.

You cannot cast a vote in Virginia by negative implication. So scratching out candidates does not mean the remaining candidate received the vote.

The recount only considers votes cast (and counted) in the election. It will neither include ballots not cast in the election by including provisional or absentee ballots that were not counted nor will it exclude ballots that were cast in the election. This goes back to the earlier statement about reconciling numbers by drawing down the number of ballots cast (through a ballot pull). The procedural order from the court makes this clear.

By now you should have a sense that any dramatic change in the outcome of the Attorney General's race is highly unlikely. The votes cast and counted on election day are the votes that will be considered over the next three days. That is good for Mark Herring. It is also good for the integrity of the process in Virginia which has been under constant attack by Republicans for years now. This is in stark contrast with the position of the Herring camp, and for that matter, Democrats in general. Barring an unexpected reverse, discrediting that process is the only way that Obenshain stands a chance of stealing this election. If that becomes the strategy, we have bigger problems than Mark Obenshain.

Tags: , , , , , , , , (All Tags)
Print Friendly View Send As Email

A couple of questions (0.00 / 0)
First, thanks for an excellent summary of exactly how the recount will work.  If you only looked in the conventional media, you would seek in vain for much factual reportage at all on how the recount works.  Conventional journalists these days seem content with being stenographers, just telling us what both sides are saying.  Thanks for giving us the factual context.

The limited scope of the issues that can be considered in the recount raises a question.  The recount is only designed to provide a definitive adjudication of some issues that could have changed the outcome.  What recourse does a candidate have if he feels that the issues that the recount doesn't adjudicate might have cost him the election?  Can the losing candidate contest the election in court, or is his only recourse the legislative contest we've been hearing so much about?

The only other question I have about what you have written is on the statement that dramatic changes in the race are unlikely to come out of the recount, that only the votes cast and counted on e-day will be considered.  It may be true that only votes cast on e-day can be considered, but as you outline, there were undervotes the optical scanners did not count on e-day that will be hand-counted in the recount  for the first time, and there could be plenty enough of them to move the numbers around quite a bit.  Do all the people who tell us that there are 50,000 undervotes yet to be counted have it wrong?

My Best Answers: (0.00 / 0)
The statutory recourse is the General Assembly. That is what the court is likely to tell a candidate that seeks redress.

Regarding the undervote count: I am not certain which meaning of undervote is being used by those that claim 50,000 votes have not yet been counted. Statewide, there were 34,033 fewer votes counted for the AG race than in the race for Governor. How anyone would know whether any of these ballots are optical scan or paper and contain any marks that will qualify as an affirmation is beyond me. A true undervote on a DRE machine will remain just that. So statistically, 75% of the 34,033 are already beyond challenge.

So, yes, I do believe that those claiming there are 50,000 votes out there waiting to be counted are off the mark.

[ Parent ]
Undervotes (0.00 / 0)
For one thing, if there is no contest of an election allowed by VA law to be pursued in the courts, that raises the inherent credibility of appealing to the legislature.  Obenshain may still be left after the recount with no real justification for going to the GA, but if that's the only game in town, it sure looks less inherently wrong to go there.

As for the undervotes, some of them were certainly intentional, the voter came to vote for the top-line race, governor in this case, and didn't try to vote for either AG candidate.  I've heard this referred to as "drop-off", the tendency for fewer votes the further away you get from the top-line race.  There won't be any marks on their ballots in either the target or the candidate area, and they won't change the tally.

That 34,000 number you cite, the difference between the gov and the AG vote, is actually likely to be almost all drop-off.  Where the money is, so to speak, is in the difference between voters given a ballot, and votes cast for the top-line race.  The way I interpret the Fairfax website, is that, out of 303,000 people who voted in Fairfax, the AG race got approx. 8,000 fewer votes than that, while the gov race got approx. 5,000 fewer than that.  My interpretation is that 3,000 is probably drop-off, but 5,000 is the potential pool of undervotes that might register an intention to vote for one or the other AG candidate that could be read on a hand-count.  People who get the big numbers seem to assume that, in our Fairfax example, all 8,000 undervotes are in play, but then they also fail to pay attention to the factor that you mention, that only optically scanned undervotes are in play.  The DRE undervotes cannot be counted.

How many undervotes are actually in play would depend on how many undervotes there were on the optical scanners.  I certainly haven't been able to find any resource that tells you that, how many of which localities' votes were scanner vs DRE, so it's hard to say how many undervotes in which locality are in play.  There's a spreadsheet out there that lists what voting equipment the localities own, but not what was actually used on e-day.  The spreadsheet, for example, lists Fairfax as having way more DREs than scanners, but Fairfax seems to have left almost all those DREs in the warehouse this year, and used scanners almost exclusively.  People mention 707,000 as the number of votes cast on scanners, but I haven't seen an account of where that number comes from, and no breakdown of how many of the statewide undervotes were in that 707,000, vs the DRE votes.

How the undervotes that a hand count can rescue and count as votes will break, seems to me to depend mainly on which localities used the scanner, vs the DRE.  I would expect the Fairfax undervotes that can be rescued would break about the same as the already counted Fairfax votes broke, and Herring to therefore gain votes, maybe hundreds of votes.  Statewide, if the scanner was used disproportionally in localities that broke for Herring, he's going to gain.  If for Obenshain, he will gain.  Even if the undervotes in play statewide are a lot less than 50,000, even if they are as few as 5,000, it wouldn't take a big skew in terms of the vote break in where in VA those 5,000 are located to change the race dramatically.  For example, if it is only 5,000 that can be read by hand-count, Obenshain only needs a 2,600/2,400 break in scanner VA to win.  That's just 52% to 48%.  The 64 thousand dollar question is -- what is the vote break in scanner VA?  

People have suggested that the break actually works in Herring's favor, that Blue VA voted disproportionally by scanner compared to Red VA.  I'ld sleep a lot better over the next week if that could be confirmed.    

[ Parent ]
All Good Points (0.00 / 0)
The relationship between the gubernatorial numbers and the AG numbers was simply an effort to find a rationale for the 50,000. I believe most commentators do not know that the undervotes in question are in the AG race only.

My understanding is that about 75% of Virginia uses the DRE.

My guess is that over 90% of the undervotes are not votes at all; that they are what they are. Walking into the world of supposition and accepting your lower number of 5,000, that would mean some 500 ballots which might add to the vote total.

Consider this: Of the persons voting who have the less extensive standardized testing experience, who might they be supporting? If the undervote is due to unscanable marks, which way will they fall?

I am of the opinion that the affirmation calls that are made will add to the Herring margin. I admittedly also want to believe this.

[ Parent ]
How the undervotes might skew (0.00 / 0)
You can posit all sorts of reasons that the tendency to have flubbed marking the ballot in ways the scanner can read would skew for one candidate or the other.  But I don't think that any factor other than how that locality voted, is at all likely to be very strong.

Sure, people unfamiliar with the bubble-sheet format are much more likely to have gone to the candidate area instead of the target area to indicate their vote, than people who have used bubble sheets for national standardized testing, or in college or HS, etc.  So we're talking about older voters, people in their 60s or 70s who came out of their academic years before the era of the bubble sheet.  We're talking about people who never took national standardized tests or went to college.  Did either of these groups break for Herring vs Obenshain nearly as dramatically as the difference in break among localities?  Older voters may actually have gone Obenshain, I haven't reviewed the exit polling, but I doubt by very much.  Non-college or non-HS may have broken somewhat in Herring's favor, but I would be surprised if by much.  

Then you have people who were just in an hurry and didn't recognize the format.  The ballot wasn't really like a bubble sheet, it didn't have long rows of bubbles, just 4-5 little unobtrusive bubbles.  People in a hurry would tend to just circle a name, not noticing the bubbles at all.  How would "people in a hurry" skew?  Beats me, but I would be surprised if it's by much.

I really can't see any predictive variable at all likely to have much weight except for locality.  If we recover x votes from Fairfax, they're going to split the same as the Fairfax votes we have already counted, strongly for Herring.  If we recover y votes from Chesterfield, they're going for Obenshain by the same margin as the Chesterfield vote we've already counted.  Every other factor is probably small, and there's so many of them moving in different directions that they will tend to cancel each other.    

[ Parent ]
Update by Herring campaign (0.00 / 0)
On the Herring campaign's conference call earlier this afternoon, they made a few points:

*They have observers (700-800 volunteers) throughout Virginia and are tracking the numbers, as precincts are counted, against the certified results.
*They continue to expect small movement, but certainly nothing that would affect the results of the election.
*As of the call at 2:30 pm, Herring had picked up 55 votes, Obenshain 35 votes, for a net gain of 20 votes for Herring (making the total margin as +185 for Herring as of 2:30 pm). They expect this trend to continue.
*The recount seems to be going along relatively smoothly, the challenge level has been "relatively low" (at most 20 challenges as of 2:30 pm).
*They feel very confident that Herring will be declared the winner for the 3rd time.
*The recount should finish up Wednesday, then the court will review Wed/Thur/Fri, hopefully wrapping this up by the end of this week.

Follow me on Twitter. Follow Blue Virginia on Facebook and Twitter.

Also see (0.00 / 0)
this spreadsheet, which currently has Herring's lead up to 246 votes.

Follow me on Twitter. Follow Blue Virginia on Facebook and Twitter.

[ Parent ]
I'm assuming (0.00 / 0)
that these increases in votes counted are from votes the optical scanners kicked out because of 1) overvotes; 2) undervotes; or 3) write-ins.  

I guess my question is, why aren't the optical scanners programmed to kick these votes to the side, just as they are with the write-ins, in the first place.  If the election officials can look at them now and determine if the vote should be counted, can't we assume that they could have done the same thing on election night and our election night count would have been more accurate?

[ Parent ]
Good questions. (0.00 / 0)
I'm not sure what the answers are - anyone?

Follow me on Twitter. Follow Blue Virginia on Facebook and Twitter.

[ Parent ]
My understanding (0.00 / 0)
Which could be outdated, but this was true of earlier elections in which I was an observer, is that on e-day, the scanners are programmed to reject only overvotes (the little bubbles are filled in for more than one candidate in any race), plus undervotes (no bubbles filled in) for all the races.  If you vote in at least one race, the scanner would not kick your ballot back even if you undervoted all the other races.

The machines are programmed this way because many people turn out to vote only for the top-line race, and don't care to vote in other races.  It's felt to be an invasion of their right to a secret ballot to subject them to the fuss that would be caused if their failure to vote every race were dinged by the machine and the ballot spit back telling everyone in the room that they decided not to vote every race.  People who vote more than one candidate in one race, or don't vote any candidate in any race, on the other hand, are assumed to have made a mistake that they would want to correct because if they don't, their voted doesn't count.  Overvotes count as no vote.  Undervote every race, and why did you bother to go to the polls?

[ Parent ]
It is my understanding (0.00 / 0)
that write-ins are separated from other votes in the optical scan machines because every one of those votes has to be recorded on election paperwork.  I'm not saying that under votes need to be kicked out when the voter puts the ballot into the machine, only that they should be segregated within the machine just as the write-in votes are segregated.  At the end of the evening, if election officials at the precinct level are not allowed to make determinations on those ballots, they could be forwarded to the electoral boards for their determination in the canvass.

It's been proven in this election there are a fair number of under votes that were never counted in the AG race.  From my experience yesterday, of the under votes kicked out of the machines in our County, 24% were counted by election officials as obvious votes for either Mr. Herring or Mr. Obenshain.  The other 76% were either write-ins or ballots which were not marked (true under votes).  

[ Parent ]
Herring Campaign Obviously Wrong (0.00 / 0)
They say that "they expect small movement, but certainly nothing that would affect the results of the election."  Yet win Fairfax County with 67 precincts counted Herring has gained 158 votes and Obenshain 77 with Republican leaning districts being counted tomorrow.  That is not a "small movement" in an election where the final count was a difference of 165 votes.  This is an entire new ballgame.

[ Parent ]
I presume they meant overall movement for Virginia (0.00 / 0)
as a whole. If that's the case, then we'll need to see what happens tomorrow, when more Republican areas do their recounts.

Follow me on Twitter. Follow Blue Virginia on Facebook and Twitter.

[ Parent ]

Make a New Account



Forget your username or password?



Donate to Blue Virginia

The purpose of Blue Virginia is to cover Virginia politics from a progressive and Democratic perspective. This is a group blog and a community blog. We invite everyone to comment here, but please be aware that profanity, personal attacks, bigotry, insults, rudeness, frequent unsupported or off-point statements, "trolling" (NOTE: that includes outright lies, whether about climate science, or what other people said, or whatever), and "troll ratings abuse" (e.g., "troll" rating someone simply because you disagree with their argument) are not permitted and, if continued, will lead to banning. For more on trolling, see the Daily Kos FAQs. Also note that diaries may be deleted if they do not contain at least 2 solid paragraphs of original text; if not, please use the comments section of a relevant diary. For more on writing diaries, click here. Thanks, and enjoy!

P.S. You can contact us at and you can subscribe to Lowell's Twitter feed here. If you'd like to subscribe to Miles Grant's Twitter feed, click here. For Teacherken, click here. For Kindler, click here.

P.P.S. To see the Blue Virginia archive, please click here. To see the Raising Kaine archive, please click here. To see the Blue Commonwealth archive, please click here.

RSS Feed

Subscribe to Blue Virginia - Front Page

Powered by: SoapBlox