OH-12 Special Election: Balderson vs. O’Connor on August 7

[Chancellor Tanterterg]

Fitrakis is a lot of things, and I hate him, but a Republican plant is a bit far. He's on the Green Party's national committee.

--

Interesting piece in the WSJ about the race.

TL;DR: "Balderson has been an 'unsteady campaigner' who has "struggled to raise money and is failing to motivate GOP voters." Stivers, the National Republican Congressional Committee chair who represents the neighboring 15th District, urged Balderson last month to spend more time fundraising and less time on his Ohio Senate duties. Balderson “said he isn’t aware of Washington angst about his fundraising."

Just reporting what I’ve heard Re: Fitrakis

[Virginiá]

The question is will Trump discover like most other Presidents that visits in midterms usually drive up the opposition more than supporters.

I astually not sure these trials will not aggravate Trump supporters.  My wife is getting really mad.  The trials have nothing to do with Russiagate.   They Involve old facts on which the Obama Justice Department declined to prosecute.  In some cases they failed to prosecute because Obama supporters were also involved.

Like who?

Manafort and Rick Gates, for instance? Their crimes did not get discovered until they thrusted themselves into the spotlight and into a situation where they got caught up in the investigation of someone else.

There is a huge difference between declining to prosecute and not even knowing about in the first place. The fact that you used these words is so typical of a partisan.

[Ebsy]

Foucaulf's analysis is that, assuming past party composition relationships remain consistent, Democrats only need a mid single digit lead in partisanship in the early vote, even less if "Other" voters break towards O'Connor. Whoever it was that said that O'Connor needed to win the EV by 20 points was not accurate at all.

[NOVA Green]

Question for NOVA Green: is there any way to look at the results for say 2016 by vote type (VBM, In person absentee, election day) so that it can be compared against those groups of voters partisanship?

Gass asked a similar question a few pages back, and this was my response:

56-32 is not exactly a comforting margin for O'Connor given expected E-Day turnout.

Given the number of relocated and, especially, first time voters the other category constitutes, I wouldn't be too worried if I was O'Connor

I think this hits the nail on the head. Probably tonight and throughout the week I will have some more analysis of how many of early voters are new registers and also what the return rates are for the various parties across the 7 counties. Expect to see the numbers for the latter tonight or tomorrow at the latest.

Do we know what the final early vote numbers looked like here in 2016? I thought I saw on Twitter that it was a pretty decent Republican lead.

All I have been able to find are the Franklin County numbers that I posted a few pages back on the thread....

Relevant numbers for the OH-CD 12 election in 2016 in Franklin County by Party and "Vote Type":



Franklin County-2014 numbers for OH-CD 12 by Party and Vote Type:



Franklin County: 2012 numbers of OH-CD-12 by Party and Vote Type:




Unfortunately it doesn't appear that this data is easily and freely available on the Internet, although I don't doubt that for the other Six Counties within CD-12 it is likely obtainable for a fee from the local election offices....

I guess we could at least look at the overall Franklin County OH CD-12 numbers and try to come up with a % swing, but really I doubt that the '12 and '16 GE numbers will be as representative of the EV/ED voters, compared to a lower turnout election such as '14....

Basically all we have are Vote Type (EV-MAIL, EV- In Person, ED, and Provisional) for Franklin County, but only by race........

So for example I could tell you how many people voted in ALL of Franklin County for the past X elections by "Vote Type" by election race, but unfortunately it doesn't break down to a precinct level.

One of the things I was interested at looking at were the US-PRES numbers for Precincts located within OH-CD-12 in Franklin County by Vote Type , but I couldn't find the data out there on any of the Government websites.

I couldn't find any data at all for the other Counties in the district (Might have neglected to check a few of the really small vote share ones, but likely not).

I would imagine this data is likely available through the local county election offices for a fee, or possibly as inexpensive as the cost of a few photocopies if you go in person, which is how I obtained much of my precinct data for the Oregon 1988 Presidential Election, until I discovered the joys of the Oregon SoS Election office/archives, where I could eat precinct data to my hearts content as a young Teenager until my Mom would show up 5 hours later to drive me back to my home-town, where I would spend Weeks trying to analyze the data!

If we have any intrepid OH Atlas posters in the Metro Columbus Area that want to check in with the Ohio State Elections Department in person we *might* be able to at least be able to take a look at Franklin County US-PRES data by precinct by Vote Type, and possibly even Delaware County as well if someone wants to take a road trip out beyond the 'Burbs and Exurbs to the County seat....

Alternatively, if we have someone with "Atlas Deep Pockets", they might be able to purchase an electronic version of some type of customized report that pulls the data that would be really helpful to provide historical context to the EV Vote numbers we are seeing in this election.

Unfortunately my pockets aren't that deep at present, despite just recently getting a really nice promotion, so I haven't attempted to make phone inquires as to if the information I desire is available, and if so what will the cost be to received such data.

[NOVA Green]

Question for NOVA Green: is there any way to look at the results for say 2016 by vote type (VBM, In person absentee, election day) so that it can be compared against those groups of voters partisanship?

@ Ebsy---  Apologies.... I believe I responded to your post without actually really understanding your question.

My general thought is that since there isn't a means of looking a Vote Type anywhere other than Franklin, which itself has fundamental limitations, we won't be able to match against Partisan Affiliation (As measured by which Partisan Primary voters last voted in?).

That being said, there *MIGHT* be data available not posted on County Election websites that are available for purchase from local county election offices, even for previous elections.

Back when I was involved in political campaigns in Oregon in the late '80s, we could purchase detailed information about all voters within a County by Party Registration, Voting Address, that also included other incidental details...

Although we didn't have VbM way back then, what election campaigns would typically do was send someone over to the County Election offices every few hours on Election Day and pay to pull the report of which voters were already registered as having voted....

The purpose was to make sure we could contact voters that had not yet voted, and do whatever last minute means of trying to call, get someone to knock on doors in neighborhoods, etc to try to maximize the turnout of last-minute-voters that we knew were either supportive of our candidate, or Registered Independents located within Precincts where we knew simply jacking up Turnout would maximize our vote margins from that precinct.

In the age of Early Voting (VbM and In-Person) modern political campaigns have reached a much higher level of sophistication that anything we had in place back then, and I would imagine that local County election offices are only too happy to accept the extra Revenue to provide whatever data they have available and can legally provide to private citizens or political campaigns for a Fee.

The main question being is what data do these various Counties maintain and archive for historical elections, and in an extremely decentralized election system like we have in the US, which Counties actually even track the data that we need to properly examine these types of questions from a Political Science perspective?

[Arkansas Yankee]

The question is will Trump discover like most other Presidents that visits in midterms usually drive up the opposition more than supporters.

I astually not sure these trials will not aggravate Trump supporters.  My wife is getting really mad.  The trials have nothing to do with Russiagate.   They Involve old facts on which the Obama Justice Department declined to prosecute.  In some cases they failed to prosecute because Obama supporters were also involved.

Is your wife a conservative like you?

She has always more moderate.  But sh was the first in the household to swing to Trump after Rubio dropped out.  She thoroughly dislikes Hillary.  She got an electrical engineering degree from Vanderbilt in 1968.  She was in the buiness world till she retired in 2009.  So she knows what it is like to be in a man’s world. She believes Hillary is a phony feminist, always getting ahead on the basis of her marriage to Bill.

[NOVA Green]

So, we have been operating under the assumption that O'Connor needs to open up a massive lead in the early vote in order to overcome a large GOP edge on election day, which is the conventional wisdom in Ohio. There has been dithering about just how large, but the prior assumption here is what is important. However, when you go back and look at the partisanship of the 2016 electorate and compare it to now, the numbers paint a different picture.

In OH-12* in 2016, Democrats only enjoyed a small partisanship** advantage from the early vote compared to the election day vote, getting 22.6% in early ballots to 20.6% on election day. The GOP comprised 36.6 of the early vote electorate and 38% of the election day vote, with Other voters outnumbering both parties in both voting periods. The actual margin difference between early voters and election day voters was only 3.4 points in favor the Democrats, not exactly a gigantic margin. Additionally, if you assume Other voters split roughly 50/50, with a slight edge to the GOP, you end up very near the expected result for a Republican in this district.

What does this tell us? Democrats' performance in the early vote thus far is extremely anamalous, and while it is expected that Republicans will gain a couple more points, Democrats are going to end up at least twice their normal vote share, with Republicans under where they usually are. There is also the expectation that Other voters are going to buck trends and break towards O'Connor. We also now have considerable reason to doubt that the election day electorate is going to be nearly as Republican as has been suggested, as past election results indicate that it was not in 2016. Of course, Democrats could be abnormally banking their votes in this district, but we have no evidence of that one way or the other.

*Excluding Richland and Morrow Counties
**Partisanship figures are current as of the 2018 primary

Excellent post as usual Ebsy!

I've been skeptical since the beginning about the whole "Dem needs 60% of the EV to win argument" that many on here have been basing their ballpark math on, since it is based upon overall Ohio EV numbers from Presidential Election Years.

In OH CD-12 we see a HUGE variability between DEM EV #s in "Off-Year" (Non-Pres) Years when it comes to Early-In-Person voting, which is really the main place where Dem's tend to perform extremely well.

In 2014, Dem EV-IVO share in Franklin County dropped dramatically between '12 and '14, and jumped WAY back up in '16.

Maybe I'm totally off-mark here, but what that tells me is that in PRES ELEC years, this is a method that voters (Especially Dem voters) use to avoid having to stand in long lines on election day, and not have to go through all of the hassle of going through the system to register in advance for a Mail-In-Ballot.

I would imagine that in a Non-Pres, Non-General Election Year (AKA Special Elections) that we would see an even smaller share of ABS-IVO even than in 2014....

It could be, and we have plenty of resident Ohio Election experts here, that some of these voters have shifted from ABS-In-Person to ABS-Mail in "Off-Year" and Special Elections as an alternative form of EV "Vote Cannibalization"....

Regardless ED Voting will likely comprise a significantly higher % of the OH CD-12 SE Vote Type than in the OH-2014 GE. Maybe I'm incorrect here, and just as we saw in the SE in AZ CD-08 voting type patterns are shifting in various States and Metro areas over the past few GE cycles.

One other consideration, which actually undermines my thoughts that ED voting will likely be higher as a % than the OH '14 GE, is the precinct consolidations that have been dramatically accelerating over the past few GE cycles.

Although I understand the fiscal rationale for cashed-strapped County Election Offices throughout the United States to dramatically slash the number of voting precincts to reduce the overhead costs that are essentially borne by County Taxpayers, this ultimately either:

1.) Reduces the Access to Voting for those Citizens who tend to cast Election Day ballots

(Longer Polling Station Lines for Voters within "Cities" and Metro Areas, Longer drive times for Rural/Exurban Precincts

2.) Forces Citizens into alternative forms of voting as as EV-In-Person or EV-Mail....

I'm not even going to touch the 3rd rail as to the intent of the Republican dominated State Government of Ohio, to potentially use whatever weapons in their tool-kits to disenfranchise or reduce access to voting for more Democratic leaning constituencies....

What I will say is that I have observed a dramatic decrease in the total number of voting precincts within many Counties of Ohio between '12 and '16 alone, so if you're going to look at a "Push-Pull" reason as to why many Ohio voters are shifting to Early Voting, this might well play a major factor.

Hell, most rural Townships in Licking County have only one voting precinct for Christs sake (So you gotta drive Ten Minutes to the next polling Station)! Imagine what happens if you are in Northern Franklin County and suddenly to vote in a Presidential Election after you get off work on a Tuesday, your wait time to cast your ballot goes from One Hour to Two Hours!!!

As someone who was only a Thesis shy of a Masters Degree in Political Science, it would be a fascinating study to examine changing Vote Type Patterns, Turnout, and Election Results in Ohio within just the past 8-12 Years....

Special Elections are always tricky business with tons of variables that are not easily controlled using scientific methodology.

Still, we are looking at an overwhelmingly Republican Congressional District, not just in terms of the post 2010 Redistricting Map, but the actual voting patterns of the individual precincts and municipal jurisdictions that fall within the current boundary.

I agree 100% with your fundamental point: "Democrats' performance in the early vote thus far is extremely anamalous", and "we have been operating under the assumption that O'Connor needs to open up a massive lead in the early vote in order to overcome a large GOP edge on election day, which is the conventional wisdom in Ohio"

Ebsy--- I think you're hitting the nail on the head here, especially considering the fact that O'Connor doesn't need to win any County other than Franklin, AND is still holding a lead in Delaware, AND keeping Licking County close.

I'm not even touching the "Other" Vote Share, which if we were to use a conventional 60-40 anti-incumbent Stat would even further add to the Dem Vote numbers....

BTW: I haven't thanked you yet for your A+++ updates and effort posts on this thread!!! It was a bit easier for me running numbers on AZ CD-08 '18 SE, since the EV/ED ratio skews much higher EV than OH CD-12, AND we don't have the whole funky business of "What was the last Partisan Primary You Voted In" as a means of trying to determine voter identity and partisan affiliation.

One Special Request before E-Day (If you have the data and time available) would be to try to separate the ABS-MAIL vs ABS-IVO data by County and "Party ID", as an election sweetener for us all as we move towards the grand finale!!!!

[Ebsy]

It shouldn't be too much of a problem to get a partisanship breakdown for mail in vs. in person absentees, though the data will need some cleaning since not all the counties track it the same way. I will say from simply looking at the data over the past several weeks that mail ins seem to be very Republican while Democrats definitely dominated in person absentees, but I can get you the exact breakdown.

[BuckeyeNut]

Trump switching his visit venue is going to mess up the Pelotonia charity bike race route. LOL

[NOVA Green]

Great post as always, Nova. Let me offer some trivia, being familiar with the county, albeit not a resident.

Granville is the home of Dennison University, and not-coincidentally quite Democratic. It regularly appears as a (non-Atlas) blue splotch on an otherwise largely red election map.

Reynoldsburg in the SW is split. Democrats failed in 2015 in a major push to win the mayoral race and council majority, but it votes Democratic at the presidential and federal level. It's a city where Republicans years are numbered at the local level. However, the city is actually divided between Licking, Fairfield, and Franklin Counties, and the Licking County portion is IIRC at least somewhat more Republican than the City as a whole. I'm tempted to say O'Conner needs to win the district's portion of Reynoldsburg, or at least come real close, to win the race.

Between annexation and suburban growth, Pataskala has become the second largest city in the county after Newark. It's still reliably Republican, though not quite as overwhelmingly as before.

If O'Connor wins Reynoldsburg and Newark, even narrowly, he should be alright.

Badger--- Thks as always for your insightful analysis, as well as your contributions as one of Atlas resident Subject Matter Experts (SME) when it comes to the great state of Ohio.

Here are the US-PRES '04 to '16 numbers from Reynoldsburg.



The Licking County portions of the City are moving Democratic at a PRES level faster than a Hound Dog on a Fox Hunt in the Shires of the UK....

Here are the '04 to '16 US PRES numbers from Granville Village....



Crazy comparing and contrasting '04 vs '16....

Although my PC is overheating, so I'll need to shut-down and restart right now, it's pretty cool for the Dems that they are seeing these types of massive swings among College Educated and Suburban Anglos in Ohio, still even with these types of massive swings in certain parts of Licking County, we have seen even more massive swings between '12 > '16 in areas accounting for a much larger chunk of the Vote Share....

The key to a Democratic Hypothetical Victory in OH CD-12 runs through an extremely narrow pathway that combines both the massive shifts in relatively educated and upper-middle-class Anglo "Metro" voters combined with Obama Democrats in the Cities and smaller communities of "Downstate CD-12), for whom for many Ancestral Republicans he was one of the only Democrats they had ever voted for at a Presidential Level, as well as occasionally voting for DEMs on down-ballot races.

This is one of the tricky dynamics of this race that makes it perhaps a bit more like PA-18 than AZ-08....

Honestly it doesn't completely surprise me, having had the privilege and pleasure of having lived Four Years in the Buckeye State as a Young Man in the Mid '90s, somewhere roughly between Columbus, Dayton, && Cinci.

"Ancestral Democrats" in many of these places will be key--- regardless of 1992/1996 PRES Votes, we can still look at the '00 to '16 Votes....

If Trump shows up at Orange Township in Delware County, and his 'Pub loses the OH CD-12 SE that will be monumental....

[Brittain33]

Foucaulf's analysis is that, assuming past party composition relationships remain consistent, Democrats only need a mid single digit lead in partisanship in the early vote, even less if "Other" voters break towards O'Connor. Whoever it was that said that O'Connor needed to win the EV by 20 points was not accurate at all.

Again, after the 2016 experience, I question how we can have confidence in early vote as a predicative measure. I really appreciate your work and it is an interesting data point to show engagement... but I would be cautious about trying to derive benchmarks with significance since the numbers are still so low compared to the final totals.

[Ebsy]

Foucaulf's analysis is that, assuming past party composition relationships remain consistent, Democrats only need a mid single digit lead in partisanship in the early vote, even less if "Other" voters break towards O'Connor. Whoever it was that said that O'Connor needed to win the EV by 20 points was not accurate at all.

Again, after the 2016 experience, I question how we can have confidence in early vote as a predicative measure. I really appreciate your work and it is an interesting data point to show engagement... but I would be cautious about trying to derive benchmarks with significance since the numbers are still so low compared to the final totals.

Well, the early vote indicators in Ohio showed that Democrats were in trouble in 2016, so I am not exactly sure what the "2016 experience" is.

[Brittain33]

Foucaulf's analysis is that, assuming past party composition relationships remain consistent, Democrats only need a mid single digit lead in partisanship in the early vote, even less if "Other" voters break towards O'Connor. Whoever it was that said that O'Connor needed to win the EV by 20 points was not accurate at all.

Again, after the 2016 experience, I question how we can have confidence in early vote as a predicative measure. I really appreciate your work and it is an interesting data point to show engagement... but I would be cautious about trying to derive benchmarks with significance since the numbers are still so low compared to the final totals.

Well, the early vote indicators in Ohio showed that Democrats were in trouble in 2016, so I am not exactly sure what the "2016 experience" is.

What I was thinking of was early vote totals in Nevada and Florida which seemed to indicate a strong Democratic win based on past results. However, there was a massive surge for Trump on Election Day.

I don’t think there’s a massive surge of Republican support that has gone undetected because Republican turnout in specials has been poor all year and I don’t think OH-12 will be any different. But I think predicting the outcome from early vote totals with precision is hard to do based on the high variability.

Personally I’m considering this race a tossup based on polling and PVI and obvious Republican dog sweat, and don’t expect to know more until Tuesday night.

[Antonio the Sixth]

The Nevada results actually perfectly matched Ralston's predictions based on the early vote. Florida is still a puzzle, admittedly (as is NC to a lesser extent).

[junior chįmp]

GOP in D I S A R R A Y

https://www.twitter.com/darreldrowland/status/1025763229019058176

[Gass3268]

https://twitter.com/PollsterPatrick/status/1025783143956185088

https://twitter.com/PollsterPatrick/status/1025785145360621574

[libertpaulian]

GOP in D I S A R R A Y

https://www.twitter.com/darreldrowland/status/1025763229019058176

Dems need to counter the Pelosi Strategy with the Jordan Strategy.

[Dr. Arch]

https://twitter.com/danielmarans/status/1023249788152766465

[Ebsy]

Update:



Democrats improved in their absentee performance from yesterday, and Other voters continued to increase while Republicans dropped, possibly signalling a Democratic bump this weekend. Republicans continue to lag behind where they would be in a normal general election.

[ON Progressive]

Update:



Democrats improved in their absentee performance from yesterday, and Other voters continued to increase while Republicans dropped, possibly signalling a Democratic bump this weekend. Republicans continue to lag behind where they would be in a normal general election.

Dems went from 49.44% to 48.14% and that's improving?

[Ebsy]

Update:



Democrats improved in their absentee performance from yesterday, and Other voters continued to increase while Republicans dropped, possibly signalling a Democratic bump this weekend. Republicans continue to lag behind where they would be in a normal general election.

Dems went from 49.44% to 48.14% and that's improving?

Democrats were 36% of returns on Thursday, and near 40% on Friday.

[Devout Centrist]

Franklin County vote share is only down marginally. If it's at 35% or greater on Monday night, that's a good sign for O'Connor.

[Hindsight was 2020]

Update:



Democrats improved in their absentee performance from yesterday, and Other voters continued to increase while Republicans dropped, possibly signalling a Democratic bump this weekend. Republicans continue to lag behind where they would be in a normal general election.

Dems went from 49.44% to 48.14% and that's improving?

Due to continued increase of “others” that should break heavy for O’Connor

[Co-Chair Bagel23]

Franklin County vote share is only down marginally. If it's at 35% or greater on Monday night, that's a good sign for O'Connor.

It'll definately be above 35 % in EV, question is if we can hoist it up throughout the entire fight.

[Ebsy]

For those who asked:



As expected, vote by Mail narrowly favor GOP registrees, but Democrats are dominating among In-Person returns.