Polling Deja Vu – Or Is It?
Many of us recall the collective media firestorm back in 2014 when the conventional wisdom was that the race to replace retiring Senator Saxby Chambliss would undoubtedly be headed into a runoff. Only one poll since September had David Perdue running at or above the 50% mark heading into Election Day. Indeed, the final RCP Average had Perdue barely ahead of Nunn, 48.2% – 45.4%.
We all know the final result. Riding a national Republican wave, David Perdue ended up trouncing Nunn by an 8-point margin 53% to 45%, avoiding a December runoff.
Alarm bells seemed to be sounding as we all woke up this morning to an AJC poll that showed Hillary Clinton leading Donald Trump by 3 points, 41% to 38%, in a 4-way race. You can see the cross tabs for yourself, here. This is the second poll in a row showing a tight race – WSB-TV/Landmark polling had the race tied at 45%.
No sooner did the hot takes begin, did many Georgia Republicans start to roll their eyes. We’ve seen this before. The polls will show a close race mired in the mid to low 40’s, but the inherent Republican advantage in turnout and party ID will deliver Trump the Peach State for the 6th straight presidential contest.
Or will it?
Looking back at the 2014 and 2012 numbers, the AJC’s record seems to be pretty good at picking the winner this close to Election Day. Their October 2012 poll nearly hit the mark – Romney won by 7.8% three weeks after the AJC pegged him up by 8%. In 2014, their September and October polls both had Perdue leading, although he was short of 50% in both of them.
Providing a warning shot to Hillary Clinton’s chances in Georgia is the infamous leaked strategy memo from the Michelle Nunn 2014 Senate campaign. In it, Nunn’s strategists assert that she would need 30% of the white vote in Georgia to win. According to AJC’s poll out today, Clinton is registering at 20% in a 4-way race.
Yet, helping Clinton are Donald Trump’s historically low numbers…among Republicans. Taking a look at the AJC cross tabs (here), Trump is only raking in 79% of self-identified Republicans. Many strategists believe that Trump needs to be north of 90% of Republicans nationally to have a chance at emerging victorious in November.
Clinton’s lead is coming from Independents, where she ties Trump at 33%, a unified Democratic party (89%), and is helped by Libertarian Gary Johnson siphoning off 18% of Independents and 9% of Republicans. Both groups are ones that Trump desperately needs on his side.
While things are still close, Trump has lost his lead of 4 points that the AJC gave him in May. This is also after the RNC in Cleveland where Republicans were urged to unite behind their nominee. Trump getting 79% of Georgia Republicans may say a lot, but “united” isn’t what comes to mind.
We should all keep in mind that Georgia has no ‘50% plus one’ requirement for presidential races. While David Perdue might have been aided by a runoff electorate, Donald Trump has no such failsafe. Hillary Clinton could conceivably win Georgia’s 16 electoral votes with numbers similar to what we have seen in recent polling.
All of this data should also be taken with a grain of salt. Singular polls are a snapshot in time and things can certainly change before Election Day. However, there’s not a lot of data that points to things going Donald Trump’s way – in Georgia, or beyond.
About Author
Cody
Add a Comment
You must be logged in to post a comment.
Hillary has very little chance of winning Georgia. She would need minorities and young voters to turn out like they did for Obama (not likely) AND she would need Gary Johnson to get in the mid-double digits (even less likely).
That doesn’t mean that Georgia can’t be an important barometer of the overall race. It would be a safe bet that Hillary plans on an electorial blowout if either of their ads start showing up on local tv. Money spent here would be better spent in actual battleground states.
No, I don’t think Johnson would need to get mid double-digits for Clinton to win Georgia, but he certainly would need more than the standard 1 percent Libertarian showing in statewide races. The Green Party candidate may collect 2 or so percent if she gets on the ballot. A Clinton win could look like something like 48 Clinton, 47 Trump, 4 Johnson and 1 Stein (Green Party). For Clinton to win, she would need a big win in metro Atlanta, perhaps 10 points or more. Granted it would be a leap for Georgia to go from Romney+8 last time to Clinton +1, but not impossible; Virginia swung from +8 Bush in 2004 to +7 Obama in 2008, and North Carolina during that time went from +12 Bush to Obama winning by less than half a point.
With regard to 2014 polling, yes, it seemed to be too pro-Democratic, but that was when Obama’s numbers were not all that great; this year, Obama’s numbers sure look better than Trump’s do. But in any event, if Trump even has to struggle to win Georgia say by 3 or 4 points, you get the feeling the race is over; if Georgia is even close, that means other swing states like Ohio and Florida are probably in the bag for Clinton.
And Larry Sabato’s Center for Politics website hasn’t changed its basic format in months—yes, they have changed some ratings from “lean” to “likely” but overall, as Andrew points out above with the 538 Site, Sabato has the same thing, Clinton winning the 332 electoral votes from the Obama states of 2012 plus the 15 electoral votes of North Carolina—347 to 191. Since the 1992 election of Bill Clinton, every Democratic presidential candidate, win or lose, has taken more than 250 electoral votes (the “worst” showing being John Kerry, who won 252 electoral votes in 2004 against Bush the second—though he ended up with 251 in the final count because of a faithless elector). Sabato has Clinton at 249 electoral votes in the safe/likely category and just 163 such electoral votes for Trump. At that point, if she gets Florida’s 29 electoral votes, game over, no matter what happens in Ohio and Pennsylvania. As for Georgia, Clinton almost certainly would have to win Cobb and Gwinnett to have a chance of winning statewide; it will be interesting to see how both counties vote in November, with Gwinnett especially trending Democratic these days.
Looking at recent electoral results in Georgia, it’s possible for Clinton to win, but to do so, she’d need to really win the Atlanta metro. In the longer term, I’m still confident that Georgia will be the next in the VA/NC Southern swing state change. The key is that ATL and its inner suburbs are slowly becoming the next Northern VA. In a POTUS year, Fulton, Cobb, Dekalb, and Gwinnett collectively have 1.4 million voters, about 1/3 of the state’s total. Traditionally, Republican Cobb and Gwinnett balanced out solid Democratic Dekalb and lean Democratic Fulton, so the impact on the statewide vote was marginal. That’s stopped being the case over the last two election cycles:
2000: Dem +28,731
2004: Dem +44,032
2008: Dem + 270,415
2012: Dem + 221,616
There is a back sliding from ’08-’12, but in ’08 Obama somewhat contested the state and aired ads while in ’12 they didn’t spend a dime. And his margin slid back a bit both state and nationwide. And even then, he still shrank the margin in Gwinnett by another 2,000 votes. (Clayton’s also become a large Dem vote source in the past four cycles, jumping from D+20,000 in 2000 to D+67,000 in 2012)
In the broader metro Atlanta area, which accounts for a little under 2/3 Georgia’s votes, the margin’s been cut back too:
2000: Rep + 93,094
2004: Rep + 233,972
2008: Dem + 48,835
2012: Rep + 28,863
So, the Dems have turned Metro Atlanta’s inner suburbs into a vote source, and have cut Metro Atlanta into pretty much 50-50. So why do they still lose? Well, the big shift from 2000 on is that the end of the last of the Yellow Dog Dems means they get slaughtered in the rest of the state:
2000: Rep + 119,865
2004: Rep + 314,133
2008: Rep + 253,142
2012: Rep + 277,895
If Clinton is going to win Georgia, she needs to make up something like 260,000+ votes in the Metro Atlanta area. Is that possible? Sure, if Trump is radioactive to the kind of GOP women and educated business class Republicans in those areas. Basically, she’d have to win the Atlanta Metro/Exurbs 56-44 to make up a 260,000 vote gap in the non-Atlanta parts of the state. And that region includes GOP strongholds like Cherokee (Romney +56,660 in 2012), Forsyth (R+51,309) and Hall (R+33,429). Now, the latest poll (44-40 Clinton) does have her winning Atlanta and its exurbs combined 60-40, which would be enough to put her over the top. I think you’re going to see more of a shift in those inner suburbs this fall with Trump as a catalyst; HRC may even become the second Democrat to win Cobb or Gwinnett since JFK (the other being Jimmy Carter in 1976; even as favorite son he lost both to Reagan in 80). But barring a major investment disparity, I suspect the exurbs and rural areas will carry Trump over the line by enough.
This is a lot of great info, thanks for sharing it. I think you mentioned or insinuated it, but Clinton’s job may not be to flip votes but rather to depress Republican turnout and have strong showings in Fulton, DeKalb, Clayton, etc.
Republican turnout in Walton, Barrow, Forsyth, Cherokee, Henry will be places to watch. They aren’t as big as Cobb and Gwinnett but the margins are great for Rs and the turnout numbers need to match 2012.
It’s certainly an uphill climb, but I still think Trump has the slight edge here. But he could always become more radioactive.
Thanks for the compliment! Certainly, turnout is going to be a big issue. Overall it was slightly down in Georgia from 08-12 even as the population grew (Total vote statewide dropped about 26K out of ~3.9 million). Clinton’s job is definitely to get good turnout from her base and have traditionally GOP voters either switch or stay home. I think she’ll do a bit better with the kind of GOP suburban business people (Office Park Republicans: Romney folk) and women. But those kind of voters can also help her just by not being able to pull the lever for Trump.
First thought on your advice: Henry isn’t a great place for Republican margins anymore. If I had to make any bet this year on Georgia’s election returns, it’d be that Henry County flips this year, following Douglas, Rockdale, and Newton in the line of formerly solid GOP suburban counties that have gone persistently, but narrowly, blue in 08-12. It’s gone from being Rep +21,663 in 2004 to Rep + 6,588 in 2008 to Rep +2,925 in 2012: Even as Obama lost the state by more, Henry’s margin cut in half. And in 2014, it went narrowly for Nunn and Carter, even as they lost the state by the same margin Obama did two years prior. (It’s also been a boom county: 39,000 votes in 2000 to 64,000 in 2004 to 88,000 in 2008 to 90,000 in 2012).
In terms of the other four counties you name, you’re dead right: Cherokee and Forsyth, especially, have been the flip side of the Dem gains in the inner and South/Southeast Suburbs. They’re boom counties that have become big GOP vote engines in the state. In terms of vote totals, they’ve doubled in size since 2000, and in 2012 gave Romney 60 point margins and a 108,000 raw vote margin: Enough to nearly balance out Fulton. (Back in 2000, they only gave Bush a 47,000 vote margin.) Knock that back a bit, and that helps Clinton make up the 250,000 votes she needs to balance out the non-Atlanta part easier.
I’d definitely agree that I think Trump has the slight edge here: The vote in Georgia is very inelastic and very racially polarized: There just aren’t that many swing voters in the state, and making up that many votes is going to be hard. But if anyone can turn off the kind of suburban white voter for whom, since 2002, voting GOP is just the ‘don’t give it a second thought’ default option, it’s Trump.
I think Johnson pulling 12% is significant, even in a poll this far out. I’m failing to understand why most of the polls are only offering 2 choices when the actual ballot will have at least 3 names.
I’d suggest only two are polled because three-quarters or more of those choosing non-major party candidates in pre-election polling change their minds and choose a major party candidate in the booth.
If Trump loses Georgia, it will be because Republicans stay home. If he keeps up the pace of alienating people that might end up being quite possible. Doubly so since he seems to have no ground game to speak of in any battleground state to turn out the vote, let alone here.
I will say his overall electoral strategy is probably the right one. The only viable path for Republicans going forward appears to be flipping states in the Industrial Midwest that are still overwhelmingly white. Of course, this should be easier for him than a more traditional Republican candidate since almost all of his policy positions, except for his hard line on immigration and national security, are much more democratic than republican.
I still think its a cycle or 2 too early for Georgia to be a battleground but Trump is so erratic who knows. I do think now that Clinton has pulled ads in Virginia and Colorado, its quite possible that money gets redirected to Arizona and Georgia.
When Bush 43 says he is worried he might be the last Republican president elected he may not be far off. Presuming Clinton wins and gets re-elected (because everyone seems to get 2 terms now no matter how badly they screw up the first one), the next competitive presidential election won’t be until 2024. By that time, Georgia and Arizona are very much in play (possibly even Texas which is already a majority minority state but for whatever reason latinos there don’t vote anywhere near their percentage of the population), and Florida will be a strong lean D state. At that point, the Republican party would have to make significant changes to its social positions to appeal to more moderate Republicans in the northeast and industrial midwest to have any kind of path to 270 in states that are still overwhelmingly white.