2014 US Congressional Election Results

[jimrtex]

Ann Kirkpatrick up 53.4%-46.8% in AZ-1.  Wow.  And Barber down 36 votes in AZ-2. 

AZ-2

The early voting counting machine broke in Cochise County, and the ballots are being moved to Graham County to be counted.

Barber carried Pima County 52%-48%
McSalley is up better than 2:1 in Cochise.

Pima turnout was 60% of 2012.
Coshise turnout (counted) is 25% of 2012.

[jimrtex]

Torie, what makes you think Barber will pull AZ-2 out?

Late counted votes, provisionals, and so forth, tend to lean Dem, typically lower SES voters. But my statement was made before Jimtex said a bunch of votes had not been counted in Cochise, which is the Pub stronghold of the CD. So Barber is probably going down, making the Pub gain 19 seats, and if the gay guy somehow manages to win in CA-52 (unlikely but possible), that will make for an even score - the result I characterized would be a big surprise, but not totally shocking to me.  I perceived that the Dem turnout would suck, and in most places it did.

A majority of votes in Arizona in 2012 were by mail, and is likely to be even higher by the passage of time.

In 2012, Barber won the early voting, and is doing so in 2014.  McSally received 60% of the election day votes yesterday.   So she might not match her 2/3 support in Cochise on election day, but should still roll up a few thousand vote margin.   Arizona requires receipt of mail ballots by election day, so voters as early as last Wednesday were urged to walk their ballots to a polling place.

The news on Cochise County was from their elections web site.

[jimrtex]

With 100% in AZ-2, McSally leads 50.6%-49.4%. 2,078 votes -- difficult to see that being overcome with absentees. McSally will probably pull this one out.

That is with absentee ballots added in.  Arizona requires mail ballots to be received by election day.  As early as last Wednesday, voters were being encouraged to walk their ballots in to either the county election officials, or the election-day polling place.

McSally received 46.9% of early votes, and 58.7% of the election day votes (71% of votes were early).

In this case, the early votes in Cochise County could not be counted because of machine failure.   Election day results were about 8K:4K in the county, and it was so heavily Republican that it was around 11K:9K.  The overall percentage for McSally in Cochise dropped from 68% to 60% with the inclusion of the early votes, but the margin increased from around 4K to 6K.

It is more probable that there will be stories about a stolen election, than Barber will win.

Factoids to be stressed:

5/6 of the vote is in Pima County, which Barber won easily.

The Douglas area on the border was carried by Barber, while McSally carried the military-dominated Sierra Vista area.  But the county seat is in Bisbee.

The early ballots in heavily Republican Cochise County were trucked out of the district to Graham County to be counted.

Even though the percentage of Republican votes dropped, the margin increased.  Republicans have to resort to algebra and the use of percentages to explain why.

Tombstone, site of the Gunfight at the OK Corral is in Cochise County.

Fort Huachuca is a base for military intelligence operations, and was a based for the 1916 invasion of Mexico.

[jimrtex]

Sacramento County has about 102,000 VBM to process, and 9,000 provisionals.

CA-3 is entirely within the county and has about 57% of the votes counted so far, which would indicate around 64,000 more votes to be counted, which is more than either Ose or Bera have received.

Ordinarily VBM would skew Republican, because it requires greater advance planning and perhaps a more stable mailing address.  But in this election, election-day voting may have gone more Republican, because turnout would be more based on self-motivation.  Folks aren't going to get enthused about voting Brown.

The next vote dump is 4 pm PST on Thursday.

[jimrtex]

Fresno County has 32.5K absentee, and 10.1K provisional votes to count.  22% of the Fresno County vote was in CA-16, so this would indicate about 10K votes yet to count in Fresno/CA-16.  The next vote dump in Fresno will be on Friday.   Incidentally, only about 25% of absentee ballots in CA-16 were returned, compared to 38%% in CA-22.

I couldn't find any information about uncounted votes in Madera and Merced counties, which together have about 3/5 of the vote.   Tacherra did slightly better (56.1% vs 54.2%) in early voting v. election day voting in Merced County.   But the votes to be counted would be Tuesday-arriving absentee ballots which might not be the same proportions.

I'd expect Ashley Swearingen to have a bit of a coat tail effect.  Fresno should be provincial enough to turnout to vote for their mayor.

[jimrtex]

98%+ of CA-26 is in Ventura County, and 84% of Ventura votes were in CA-26.  But I could find no information on votes to be counted.

There shouldn't be any swings caused by different county reporting schedules, but it is conceivable that the ballots would be processed by precinct.   I think it would be a lot more reliable to sort ballots by precinct, before verifying singatures, and running the ballots through the vote counting machines.

[jimrtex]

CA-31 is entirely in San Bernardino County.

There are 44,436 mail ballots to be counted, 15,834 in CA-31.  There are also 3,238 election days ballots to be counted (countywide) and 14,922 provisionals.   It would take about a 55:45 breakdown for Chabot for a 2nd Miracle in the 31st.

[jimrtex]

CA-52 is entirely in San Diego County, and represents about 29% of the San Diego vote.  There are about 180K mail ballots/provisionals to be counted in the county, which would indicate around 52K in CA-52, which would represent around 1/4 of the total vote in the district.

[jimrtex]

Looks like Gardner may end up winning by just 2.5 points ...

BTW: Did anyone find the AP election results page ?

I don't think they have one. 

It's probably in their agreements with newspapers that they don't bypass them.

[jimrtex]

Barber has won in AZ, almost certainly, with stuff still out in Pima and McSally barely holding on. She might improve on the 2012 margin, but still.

Pima County had about 30K mail ballots to be counted at 11:20 AM Thursday.  Around 2/3 should be for AZ-2.  Mail ballots are verified by the Recorder and then turned over to the Elections Department for counting.  There were a lot of ballots turned over on Wednesday, with just a couple of small batches on Thursday morning.  It shouldn't take too long to count ballots.  The district had around 120,000 early ballots.  Unless they were counting before election day (is that even legal?), they were able to shove a lot of ballots through on Tuesday and Wednesday.  It is possible that all the early ballots were finished by end of day on Thursday, and are in the latest released number.

There were also about 10K provisional ballots, which at 11;20 AM were still in possession of the Recorder.  They may trickle through depending on how long they take to verify.

[jimrtex]

Sacramento County has about 102,000 VBM to process, and 9,000 provisionals.

CA-3 is entirely within the county and has about 57% of the votes counted so far, which would indicate around 64,000 more votes to be counted, which is more than either Ose or Bera have received.

Ordinarily VBM would skew Republican, because it requires greater advance planning and perhaps a more stable mailing address.  But in this election, election-day voting may have gone more Republican, because turnout would be more based on self-motivation.  Folks aren't going to get enthused about voting Brown.

The next vote dump is 4 pm PST on Thursday.

Sacramento County counted 32% of its outstanding mail ballots, but this evaporated 38% of Ose's lead.  This could end up being within a couple of hundred votes.

[jimrtex]

I did miss CA-09, but when I added it, the 216 projected lead for Ose never changed ... which was -  well - disturbing!  It turns out, that the smaller the CA-07 share of the Sacramento County pie, the higher than Bera percentage margin with respect to the last vote dump, with the two exactly offsetting each other. That was not a satisfactory state of affairs, so I hunted around, and found the CA-07 vote totals before the last vote dump, which allowed me to calculate the actual number of CA-07 votes added from the last VBM vote dump, and it turned out to be a bit smaller share of that particular county vote dump, than its share of all of the votes counted.

If I were counting ballots, I would do it on a precinct-by-precinct basis.  Recounts are done on a precinct basis, the challenger even gets to choose the order of precincts recounted.  So you will want all the ballots together.   Hand counts are also done by precinct.

Return envelopes would be easily sorted by precinct, so they would be sorted out even before signature verification is done.

Even if the ballots have the ballot type coded on them, manual auditing would be easier.  If you have a stack of 87 ballots (presumably hand counted) if they were all from one precinct, you could run them through the counting machine and check whether the ballot count for the precinct increased by 87, and the results were plausible.  Run a mix and you can't be sure whether, the single Romney vote in a 90% black precinct isn't anomalous.   But if that precinct get 75 Romney votes you know something is wrong.

But they might take all the ballots from one batch of envelopes, verify them all and run them through precinct by precinct.

But there could still be geographic biases.  If a single batch came from a particular post office branch, then it will show up in the distribution of precincts of the ballots, since most voters will mail at a nearer post office rather than a further one.

And there could also be differences connected to political leaning.  Late arriving ballots are tending to be more Democratic, than election-day and earlier returned ballots.   This is true, even though mail voters tend to be more Republican.

Republicans tend to plan better, and have more stable living conditions.   Democrats tend to procrastinate, and may have been waiting to mail their ballot until they could also make a bill payment and not have the check bounce.

CA-6 the district in the city of Sacramento (or at least more inside) is more Democratic.   So it would be reasonable to expect that a larger share of the last-arriving ballots compared to other ballots to be from CA-6 rather than CA-7.

There have been a couple of interesting cases where the late-arriving ballots didn't match this pattern.  The late arriving ballots from Madera County were stronger for Tacherra than the earlier votes.

And in San Diego, the late-arriving ballots have been much more strongly for Peters (7% difference)  This indicates that late deciders broke based on their awareness of the sexual harassment allegations.

[jimrtex]

It is extremely frustrating how in all these close western US races, the final votes always seem to favor the Democrat.  Of course it is a given that provisional favor the D candidate; just the nature of the beast that is provisional voting.  But absentees is another phenomenon entirely.  The CA Democrats must have this process down to a much better game plan.

California and Arizona have permanent vote by mail, where a ballot is automatically sent to the voter each election.  Looking at Sacramento County, it appears around 62% of voters were by mail (Sacramento County reports each election precinct as two "precincts", one for in-person voters and one for mail ballots).  This would include some voters who requested a mail ballot for this election only, and voters in mail precincts (precincts with relatively few voters, where all voting is by mail to avoid the expense of operating a polling place on election day).

The following is mixing races, and jurisdictions and is dependent on me having written down some numbers, but it should be correct as far as flavor.

Based on the Ose-Bera race in CA-7, 72% of ballots were counted on election night.  So that would make it:

72% on election night.
   38% in-person.
   34% mail ballots.
28% since election night.
   28% mail ballots.

Note the in-person/mail distribution is based on Sacramento County, while the election night/since election night is based on CA-7.

The late-counted ballots in CA-7 have been 4.32% more favorable for Bera (48.63% election night, 52.95% late counted).

In the gubernatorial race, county-wide results by mail are 1.52% more favorable for Brown than for the election day votes 62.59% vs 61.07%.

Based on distribution of ballot types and confabulating the two races and the county and district, it would suggest that the early-counted mail ballots were slightly more favorable to Kashari than the election day votes (around 1%).

Throwing in another tidbit from another county (Butte), Republicans are slightly (3%) more likely than Democrats to be PVBM voters, and around 10% more likely than minor party, and No Party Preference (NPP) voters.

In California, mail ballots must be in the hands of election officials by election night to be counted.  So you have to mail a ballot back early enough, trust the USPS to only take a couple of days to deliver a ballot in a couple of days, walk your ballot to a polling place on election day, or take it to a drive-by collection location.

Republicans are a bit more likely to fill in their ballot and return it right away.  They might live where they can trust the mail to be picked up by the USPS, or have stamps available.  Democrats are more likely to stick their ballot below a stack of unpaid bills.

Signatures have to be verified.  Ballots that are returned before the election can be handled and set aside to be counted.  The crush on the last days can not.  On election day, the staff is going to be busy.  And the election officials are going to prefer to use regular staff, or hire temporary workers for a couple of weeks, rather than trying to hire and train workers for a day or two.  If there were a law that said that all the ballots had to be counted by Wednesday or Friday, they could probably do it.  But it easier to spread it out over time.

One thing that is bothersome in California is that the counties are independent on how they operate and even report things to the SOS.  So some counties don't even bother to report the number of uncounted ballots, or only do it every couple of days.  And counts are release in spurts.  It can really give the impression that counties are holding on to ballots (eg in CA-16 where Fresno released some updates, the Madera, the some more in Fresno, and then Merced.

[jimrtex]

^ There's a running conspiracy theory at RedRacingHorses that Colleen Mathis, one of the commissioners that drew that AZ map, was in league with Democrats; she purportedly pushed the commission to draw a D-friendly map. They call the AZ map the 'Mathismander.'

The commission  had a Republican counsel and a Democratic counsel.  The Republican counsel was chosen on a 3-2 vote by the two Democratic commissioners and the independent sleeper agent.

If Arizona keeps its commission it should expand its membership and require concurrent majorities among the Republican, Democratic, and other members.  It should be required to comply with open meeting laws, and the competitive districts provision should be jettisoned.  It is took subject to cheating, and most adjustments to make districts more competitive violate equal protection.

[jimrtex]

Crunched the number on CA-16, and the Dems hold it.  Based on the splits of the remaining ballots, the past split of Fresno County of about 63-37 would have to drop to around 55-45 for the race to be even at the end. It isn't happening. Costa will win by about 650 votes, plus or minus.

CA-07 does not look too good for the Pubs either. Ose needs a swing to him of a couple of points on what remains, from what went before. Still in the hunt, but the underdog now.

So that leaves AZ-02, where somebody else claims that the final projection is for a 40 votes McSally win. So that one remains a slight tilt GOP seat.

Addendum: Oh, wait a minute! I forgot that Tacherra had a 741 vote lead!  Yes, Costa is expected to generate a lead out of the remaining ballots of 651 (assuming the remaining ballots break like the ones counted to date (don't know if the last vote dump was more Dem than the election day counted votes, because I can't find the details of the last vote dump numbers), leaving him 90 votes short. So this race remains a tossup more or less. My bad.

There has been one vote dump in Fresno County (last Friday).  The next is scheduled for Wednesday the 12th at 3 pm PST.

Election night was 16235:9750 or 62.48% Costa.
The update was 1884:3581 or 65.53% Costa.

There are 10,300 VBM and 10,100 provisional ballots yet to be counted.

22.7% of county votes are in CA-16.  22.5% of mail ballots returned are in CA-16, but 30.4% of ballots sent out.  The return rate was much worse in CA-16, than the other districts.

Ballots Issued and Returned by District (PDF)

I don't know whether the returned includes the final 10,000 ballots or not.

But if we assume 22.7% x (20,400 outstanding) x 31.06% (Costa plurality in first dump) = 1438 Costa margin.

Provisional votes include:

(1) First federal election, but voter doesn't have ID.

Probably not too many of these.

(2) VBM person appears in person.   They either did not receive their ballot, the lost it, or they spoiled it and did not bring it with them.  If the mail ballot showed up by election day, then it is counted.  The reason the provisional ballots are counted last to make sure there wasn't a VBM ballot.

It is possible that these ballots skew differently than the late-arriving VBM ballots.  A Republican might be more likely to recognize that their ballot might not arrive in time, and instead go to the polls.  But a Democrat could be more likely to lose their ballot.  And a Democrat might be more likely to show up without their unvoted mail ballot.  And some of the provisional ballots won't count, since the mail ballot did arrive.

(3) Voter has a new address in the county, but did not re-register.  These provisional ballots will be counted if the signature at the polls matches the signature on the registration record.

Madera County has 973 provisional ballots remaining.  They don't anticipate releasing any updates prior to just before the results have to be certified (December 5).  Madera County is split between two CDs.

Merced County has not updated their results from election night - the SOS website has more recent results.  The number of votes in the update, somewhat matches the number of VBM mail ballots that were reported to the SOS.  So I suspect they are in a situation similar to Madera.

Finishing up the VBM in Fresno County should put the result into a virtual tie.

The number of provisional between Merced and Madera (CA-16) is similar to the provisionals for Fresno (CA-16).   Results so far Fresno 63.01% Costa vs. Merced+Madera 58.57%.

Merced and Madera are exceptions to the late-counted votes being more favorable to the Democratic candidate.  I think we could still a result within 200 votes (50.1% to 49.9%).

[jimrtex]

AZ-2 is an R+3 district, as is AZ-1. AZ-9 is R+1. Yeah, those secret Democrats are really drawing the lines to protect themselves

Those GOP  PVI's were inflated by McCain being on the top of the ballot in 2008.

Inflated, but not inflated that much. A R+3 district is still a pure tossup, maybe just slightly Republican even after accounting for McCain's overperformance. So we have four strong Republican districts, two pure tossups, one slightly Democratic district and two Democratic districts. How is that not a fair map?

The way to gerrymander is to pack your opponents into some districts, while giving your districts a slight advantage.

Now imagine you have a district that is perhaps 53% R, compact and formed from natural communities, such as cities or counties.  You decide to make it more "competitive."

How do you do this, you move Democrats in, and Republicans out.  But switching areas that
are 53% Democratic and 47% Republican is tedious and takes a lot of switches, even though the political similarity suggest a greater community of interest.

So instead you grab a small 80% chunk of Democrats and expel some Republicans to another district that is already overwhelmingly Republican.

But you overlook the fact that the measurement of voters in the 80% Democratic area was in a midterm election, and perhaps against a weak candidate (think of the Ohio gubernatorial race, where a somewhat unpopular incumbent in an extremely competitive state, blew out his Democratic opponent by 30%).   Or maybe the statewide candidate happened to be from the area and ran a couple of percentage points above the partisan trend in the area.

Didn't the commission systematically underpopulate the competitive legislative districts in Arizona?

[jimrtex]

It says a lot about the mental state of Republican party leaders when obscene gerrymanders like Ohio are considered to be the norm and a non-partisan commission map, which nonetheless allowed Romney to win a higher percentage of districts than he did the popular vote, is considered a Democratic gerrymander because there were seats, which although Obama did not win in either of his comfortable victories, which a Democrat could plausibly win. For the GOP, anything that is not an absurd Republican octopus gerrymander must be a Democratic gerrymander. UnInksing believable.

Romney was not running for Congress in Arizona.

[jimrtex]

Costa takes an 88-vote lead CA-16 based on the latest dump from Fresno County.

Most of the votes outstanding are provisional ballots:

Fresno 10,700.  With about 22% of votes in the county from CA-16, this would make 2430 for CA-16.  If Costa takes 65%, that is a gain of +730.

Madera 973, with 64% of the vote in CA-16, this is around 619 in CA-16.  If Tacherra takes 67% that is -210.

Merced 1670, all in CA-16.  If Tacherra takes 56% that is -200.

So around +320 net for Costa.   This assumes provisional ballots will all be counted (after all they are call 'provisional' for a reason), and have a distribution similar to the late counted ballots.   If would take around a 9% swing towards Tacherra to tie it up.

[jimrtex]

So, Tacherra actually gained on the last update -- a net of 11 votes. Costa now leads by 75 votes.

That was a 77:66 minidump from Merced (and slightly weaker at 53.8%, than previous votes).

Then Fresno had a 1020:395 dump for Costa.  72,1% is his best result yet.  I think we are getting into the provisionals,

A curious happening in the AZ-2 race.  There was also a special school board election on a separate ballot.  The original ballot instructions said to vote for not more than 3, when they should have said not more than 2.  So new ballots were sent out, with corrected instructions, and voters were told they could vote the new ballot.

Because of the mixup, and to make sure whether voters sent in a second ballot, the school board ballots were set aside.  It turns out that a bunch of voters put both ballots in the school district ballot envelope.   Fortunately, it was in a Republican leaning area (Green Valley) and it added +23 to McSally's lead.

There are about 750 provisional ballots in Pima County which were determined to be invalid.   So that race will likely end up in court.