Georgia senate seats runoff(s) megathread

[Tintrlvr]

Why was the Ossoff debate not broadcasted? I feel like it should be a bigger deal that Ossoff was literally there debating an empty podium...

It would be considered a contribution in kind under campaign finance laws. It could be considered a corporate contribution - and they could not claim that it was an uncoordinated expenditure. Non-profit sponsors could risk that status.

TV networks can only cover "legitimate" news events. A candidate speaking to the camera is an hour-long or half-hour-long infomercial, and the station would have to give the same opportunity to opponents. They can do clips, just as they could a stump speech.

When there are minor party candidates they can skirt the issue by getting a debate sponsor to set "objective standards" for inclusion in a debate. Typically, major party candidates won't debate if minor party candidates are invited.

But they *did* offer the same opportunity his opponent: Nobody stopped Perdue from showing up on the debate stage except Perdue himself. It is frankly laughable to suggest that there could have been liability for the networks if they broadcast the non-debate.

They chose not to do so because the ratings would have been poor.

[Tintrlvr]

https://twitter.com/CAMcGrady/status/1337793180180287490?s=20
what the ?

"COVID-1619" must be a reference to 1619, the year that the first African slaves arrived in what became the Thirteen Colonies and ultimately, the United States. It is still a bizarre reference, though, incorporating this year's pandemic with the injustices of the past.

and what does it all have to do with attacks on his candidacy?

He seems to be obliquely saying that the attacks are racist. I agree it's super strange (just say what you mean) but also totally irrelevant.

[Tintrlvr]

I might just be a political oddball but I do actually care about the deficit and responsible spending and I'm not convinced a blanket $2,000 stimulus check is good policy. I'd much prefer the Dems take some time to lay out all the different options for stimulating the economy (e.g., infrastructure spending, tax credits for parents that need child care for the pandemic), rather than just jump on the more checks for everyone train.

We're in a recession, the deficit doesn't matter right now (and barely matters during times of economic expansion). The best stimulus for this pandemic would be lockdowns coupled with direct aid to businesses to keep them open and direct checks to families/individuals so that they can afford to stay home and keep the virus in check.

Half the population is working from home already according to the polls and doing better financially than ever. Do they really need direct checks?

I disagree the deficit doesn't matter. We're financing all of this in the form of debt that carries interest that the Government will have to pay each year until that debt is paid off.

Who says we have to pay?

Well, as I see it, there are 3 options:

1) Government borrows more money to pay off the interest, resulting in ever higher interest payments. This just kicks the can down the road for options 2 and 3 below.

2) Government cuts spending or increases taxes to pay off the debt. I imagine both would be just as politically unpopular as the stimulus checks might be popular.

3) Government defaults on the debt or prints money to pay off the debt. See Venezuela for why this is a bad idea - hyperinflation, currency would become worthless, bank runs.

Option #4 is that government spending increases future incomes relative to the baseline significantly more than its cost, increasing tax receipts that pay off the debt incurred by spending or at least keep debt neutral compared to the no-stimulus scenario (which may be ugly regardless given the current state of things - or may not be, I remain optimistic about a rapid economic recovery). This may not be the case during a boom period, but it's clearly a very relevant factor in an economic downturn.

[Tintrlvr]

Dems lead down to +8.7 in my current model.

Again, based upon my methodology, I would lean towards this number slightly underestimating D's current lead, but that's speculation on my part.

Despite Republican's narrowing the margin, they aren't really netting any votes.



If in person votes broke how they did in the GE, the GOP would need an election day turnout of about 1.2 million to win according to my model.

Is that mathematically possible for the ED turnout to be that high?

Easily sure. The GOP should hope for about 32% of the vote to be from e-day which is very plausible.

Given that E-Day was only 19% of the vote in the recent GE, I have a hard time believing 32% could be realistic. In the same vein, of course, the JMC poll today that had E-Day at only 7% of the vote was equally unrealistic. We should expect E-Day to be somewhere in the range of 15-25% of the total vote - any more or less is extraordinarily implausible.

[Tintrlvr]

Something I am very interested in seeing:

will the two races have identical county maps?

The only county I am certain is close enough for a split result is Baldwin.

Close enough, yeah. But it seems like the sort of place where less than 1/1000 people would split their tickets.

Burke, Peach or Washington could also be split results. If the JMC poll were closer to correct, maybe Fayette could split. I agree broadly that it's unlikely any county is split, though.

[Tintrlvr]

Most of Cherokee County just dumped. Also doesn't look like enough for the Republicans. Not sure what's still out, but both Republicans are noticeably behind Trump/both Democrats noticeably ahead of Biden there now with 87% in.

[Tintrlvr]

Trump cleared 70% in Cherokee in November... Loeffler/Perdue not matching that...

He did not - Trump got 68.8% in Cherokee. But Loeffler is at 67.7% and Perdue is at 68.3%, in both cases of course with all of the non-R vote going D (whereas Jorgensen got 1.7% there at the GE).

[Tintrlvr]

Damn, I didn't notice Rockdale was done. That went from Biden+30.8 at the GE to Ossoff+34.8 and Warnock+35.2! 90% of GE turnout, too.

[Tintrlvr]

At work. If reading this right Fulton and DeKalb have not been counted or very little has? Ruh roh Raggy

Lots of Fulton is already in, but DeKalb has very little counted. There are some Republican areas with a lot out, too, such as Forsyth, but overall it does seem like the Democrats are likely favored in what is left.

[Tintrlvr]

Bartow County just dropped, too, so the Republicans are really completely out of votes now. Ossoff will almost certainly take the lead with what's left out, and Warnock is guaranteed victory.

[Tintrlvr]

Ossoff is down 19k, not 3k. There was a tabulation error. Perdue still isn't gonna win

It wasn't a tabulation error. Bartow County reported right after DeKalb; it was the only area of GOP strength left with significant votes out.

[Tintrlvr]

Cobb, DeKalb, Henry are the only counties remaining with large numbers of outstanding votes, correct?

Fulton has a fair number left, too. Not percentage-wise of overall votes in Fulton, but Fulton is huge.

[Tintrlvr]

Ossoff is down 19k, not 3k. There was a tabulation error. Perdue still isn't gonna win

It wasn’t a tabulation error, the biggest tranche of remaining GOP vote came in a few minutes after the DeKalb dump.  It actually bumped Ossoff above 95% on the needle.

So is it fair to say that the NYT "Needle" model was taking it into consideration from the start?

Yes. In fact, Bartow reporting pushed Ossoff from 94% to 95% chance to win.

[Tintrlvr]

There's still a sizable chunk of Newton still out too. They haven't counted their mail in votes apparently.

Ah, that would make sense. I've been wondering about the handful of counties that say they are all done but are showing wild swings against the grain to the Republicans. Newton is by the far largest (and the only competitive one), but there are a couple of random small ones in North Georgia, too.

[Tintrlvr]

Is there a sense of when Warnock and Ossoff will be sworn in? I remember Doug Jones didn't join the Senate until several weeks after the special election in 2017. Could McConnell still be the majority leader the first week or two of the Biden Presidency?

With regard to Ossoff, probably depends on how close the regular election will be. The New York Times projects a 1.1% win, which is outside recount territory. Warnock's win should be certified very soon. I'm not familiar with exact senate procedures, but I doubt this can be delayed for very long.

Certification even in the Warnock race will take a couple of weeks because they need to wait for military absentees and count provisional ballots. Those won’t change the result of course, but certification doesn’t happen until after every vote is counted.

[Tintrlvr]

This is a big win for election pollsters, which showed Warnock having a larger lead and both candidates winning by a very thin 1-2% margin.  It's actually uncanny, especially when compared to all the election day polls which greatly overestimated Biden in several states.  Looks like they got it bang on (within 1%!)

Georgia polling was very good in the GE, too, FWIW.

[Tintrlvr]

Webster County sent a weird signal being the first county to post complete results last night, and having by a fairly large gap, the largest swing R since Nov of any county in the state.

I think earlier today it was said that Webster County hasn't counted its absentee ballots yet, which would explain the apparent large R swing. The same thing happened in a few other counties last night that originally said they were done but it turned out they hadn't counted absentees, most notably Newton County.

[Tintrlvr]

I'm not really getting the point of all these 8-year old Ossoff tweets that people have been digging up.  They all seem totally inoffensive and unremarkable to me.

I think people find them funny and relatable.

[Tintrlvr]

More from Miles:

https://twitter.com/JMilesColeman/status/1347033774069080064

Yeah, it appears that while Ossoff clearly gained way more nonwhite votes from November, Perdue gained a few college+ voters.

Turnout was lower than in November, I wonder if that had any effect. Also the 100k who didn't vote in November but voted now I think may play a role

That would imply a disproportionate fall-off in college educated white Biden voters as compared to Trump voters. Highly unlikely.

The more obvious explanation is that a notable number (though still a clear minority) of McCain-Romney-Clinton-Biden voters voted for Perdue and Loeffler. Which ultimately isn’t that shocking; some recent converts still split their ballots/prefer non-Trump Republicans.

[Tintrlvr]

More from Miles:

https://twitter.com/JMilesColeman/status/1347033774069080064

Yeah, it appears that while Ossoff clearly gained way more nonwhite votes from November, Perdue gained a few college+ voters.

Turnout was lower than in November, I wonder if that had any effect. Also the 100k who didn't vote in November but voted now I think may play a role

That would imply a disproportionate fall-off in college educated white Biden voters as compared to Trump voters. Highly unlikely.

The more obvious explanation is that a notable number (though still a clear minority) of McCain-Romney-Clinton Trump-Biden voters voted for Perdue and Loeffler. Which ultimately isn’t that shocking; some recent converts still split their ballots/prefer non-Trump Republicans.

Ossoff did quite a bit better than Clinton amongst white college educated voters

I thought about saying McCain-Romney-Trump-Biden voters, and probably the largest group of defectors are such. The important point is voters who have historically been Republicans but voted against Trump at least once, maybe twice, and not necessarily because they are ideologically closer to the Democrats.

[Tintrlvr]

The real question is going forward whether or not GA becomes Virginia, a blue state, or North Carolina, a swing state that’s always in reach but still hard.

The difference with North Carolina, which I think many people (including myself) did not appreciate in 2008, is that the Democrats were overperforming among rural white voters in North Carolina relative to other Southern states in the 00s (possibly because rural North Carolina is relatively industrialized compared to the rest of the South), and therefore still had considerable room to fall among rural white voters over the next decade, which counterbalanced strong Democratic trends in the cities. Much, though not quite all, of that room has now been fallen, so I do expect North Carolina to start moving leftward again soon. But, more to the point at hand, that's absolutely not the case in Georgia, where the Democrats have already hit rock-bottom in the rural white vote, and have basically been there for at least a decade.

[Tintrlvr]

GA-06 also has by far the most Warnock-Perdue voters of any district.

[Tintrlvr]

Also interesting to note that the 538 polling averages were right on the money in estimating Warnock's margin, but slightly overestimated Ossoff (they put him at +1.8 ). Probably too small to warrant serious analysis, but I do half-wonder what made the dropoff between the two candidates wider than expected. Shy Enlightened Suburban Ticket-Splitter Effect?

I think there are two main dynamics at play that contributed to this:

The first is in line with your "shy enlightened suburban ticket splitter effect" - something I've already written about last week. This effect of course was confined to white, high-income, highly-educated areas that were much more Republican just a few short years ago; even though Perdue saw his biggest gains relative to November in those places, the gap between Warnock and Ossoff was also largest there.


The second comes down to a simple yet shifting phenomenon that has been observed to varying degrees in multiple elections dating back to at least 2008: there are more black voters who are willing to vote for a black Democrat (but not a white Democrat) than there are white voters who are willing to vote for a white Democrat (but not a black Democrat) as a percentage of their respective electorates.

However and frankly, I think the turnout differential in the runoff flipped the ratio, as even November's election showed that Democrats had more to lose than gain by running, say, a black female than a white male; I can't emphasize this enough, but relative to #trends, expectations and where resources were/were not allocated in '20, Biden won GA based on his substantially better and unexpected performance with white rural voters in the northern half of the state relative to 2018's numbers (this alone netted him ~25k votes in the margin relative to/harmonized with Abrams' performance). We also saw this phenomenon in 2018: Barrow had a closer margin than Abrams, with him doing considerably better in whiter rural areas while drop-off and support was considerably worse in blacker urban areas.

I imagine it wasn't just higher black turnout in the runoff that offset this, but higher black turnout of the right type of voter; just like with other non-white groups, I'm willing to bet a lot of lower-info black Trump '20 voters simply stayed home for the runoff.


In the runoff electorate, probably 4-5% of black voters backed Warnock but not Ossoff, while 1-2% of white voters backed Ossoff but not Warnock. In a normal election, however, this discrepancy isn't as big (and as mentioned, even in 2018 and 2020, was still reversed). Combined with the low-energy white bipartisan suburbrons who wanted to seem racially progressive but didn't want unified government, and the margin difference can more or less be explained (especially when factoring in that Perdue was better-liked across racial lines than Loeffler and also had the benefit of actual incumbency; also accounts for the slight-to-non-existent gap in non-black, non-suburban parts of the state between Warnock and Ossoff).

This analysis strikes me as far too focused on the Democrats and not enough on the Republicans, which is where there was actual, meaningful difference. Warnock-Perdue voters in ex-Milton (and similar areas) were probably mostly former straight-ticket Republican voters who have voted for Perdue before and found him familiar enough to justify splitting their tickets, while Loeffler was a political unknown who had done nothing to win their sympathy. It had nothing to do with Warnock and Ossoff per se.

To the extent there were Loeffler-Ossoff voters, they seem to have been white and rural, i.e., some Dixiecrat remnants on whom anti-Warnock scare tactics worked.