As political observers know, the party not in control of the White House usually does well in midterm U.S. House and state elections. (The Senate is a bit iffier because the landscape can vary enormously based on which “class” of one-third of the chamber is up for reelection in any one year.) There have, however, been two aberrations in the recent past, in 1998 and 2002, wherein the White House party gained House seats. As I discussed in a recent post, Democrats hope 2022 could be an aberration as well, thanks to positive feelings about the subsiding pandemic and a strong economy, and perhaps the continued presence in the public eye of Donald J. Trump — the man nearly all Democratic and many swing voters love to hate.
But there’s another midterm variable that should be considered: a traditional “midterm fall-off” in voting by demographic groups that have recently become Democratic bastions. This was exhibited most forcefully in the very bad (for Democrats) midterm elections of 2010 and 2014.
In case you have forgotten or were too young to follow the politics of the Obama era, an enormous wave of Democratic optimism accompanied the 44th president’s comfortable 2008 victory. There was much talk of an “Obama Coalition” built on high vote shares and high turnout among young and minority voters, who would presumably become the dominant part of a more diverse electorate in the years just ahead. Instead, Republicans won a big wave election in 2010 and had a solid win in 2014 (even after Obama’s narrow reelection victory). And many analysts (myself
included) noted that the Obama Coalition largely coincided with a portion of the electorate that had never voted strongly in midterms, while the GOP was now solidly established among older and whiter voters more prone to midterm participation.
Even after Trump’s election in 2016 gave Democratic-leaning demographic groups plenty of reasons to turn out disproportionately, there were fears the falloff would reduce or even wipe out Democratic gains in 2018. That didn’t happen, of course, as Ron Brownstein recalls:
“In 2018, more than 118 million Americans voted, exactly half of the eligible population, according to [Michael] McDonald’s calculations. That was the highest midterm turnout, as a share of eligible voters, since 1914, before women won the right to vote.
“And while the 2018 electorate was still somewhat older than in 2016, the gray shift wasn’t nearly as powerful as in the past, because young adults turned out at twice the level they did in the last midterm, of 2014. Turnout among Blacks and Hispanics also declined much less than in previous midterms, with the result that the White share of the vote actually fell from 2016 to 2018, according to McDonald’s calculations, an unprecedented pattern in recent years … The turnout wave continued into 2020, with nearly 160 million people voting and turnout among young people and people of color again rising dramatically.”
One key question for 2022 is whether the fall-off will resemble what we saw in 2010 or 2014 or instead the smaller version that appeared in 2018. The experts Brownstein consulted expect something in between, which, if combined with the kind of gains Democrats made in 2018 and 2020 among college-educated white suburbanites and older voters, could make a midterm upset possible under the right circumstances.
Democrats aren’t the only ones trying to get new or marginal voters to turn out again in 2022: Trump managed to turn out a surprising number of them in 2020 himself. Keeping them energized is job one for the GOP in 2022, and Republicans may have the advantage of the kind of enraged opposition to a “socialist” president that was so visible in the tea-party movement of 2010 — though thus far, Biden is not inspiring the same levels of hostility.
That leads to the 2022 variable that no one can entirely foresee: How visible will Donald Trump be in the midterm campaign? The threat of a Trump comeback is the easiest way for Democrats to mobilize their new 2018 and 2020 voters and for Republicans to mobilize their own. Trump could help Democrats turn the midterm from a referendum on the incumbent president (a referendum incumbent presidents typically lose) to at least partially a referendum on the once and possibly future President Trump.
All of this seems far in the future to those who are focused on Democratic efforts to deliver popular legislation in a closely divided Congress. But if we know anything about the current political environment, it’s that partisan polarization will make big swings in public opinion difficult or even impossible barring equally big changes in the quality of real life. So it may well be the small underlying currents in electoral politics, including the demographics of midterm turnout, that will determine whether Biden has at least four years or just two to implement his agenda.
With or without some help from Trump, though, Democrats really need to find ways to keep young and minority voters engaged.
going to http://www.pollingreport.com and taking all 13 polls since the debate (excluding the quickies) the results are as follows:
Bush – 47.38
Kerry – 46.31
a 1% difference well within the average 3% margin of error.
Marcus
I believe the point is that the methodology used before did not reflect what are historic trends, and that these numbers coming out reflect where most people thought the election is: namely a tie between Bush and Kerry. Most of the points about weighing I think these new polls are suggesting at least anecdotally maybe true- ie, they as the id’ers return to historic norms- the race is indeed tied.
It’s funny how much someone’s opinion can change only after a few days. On September 26th, a frequent poster to this site advised readers of EDM to disregard the “spin” suggesting that the CBS/NYT poll was biased in favor of Bush. Of course, this was at a time when the CBS/NYT poll indicated that Bush was leading Kerry by 8 points. In fact, his words of caution to us “libs” on September 26th were as follows,
“This CBS poll appears to use a weighting that includes MORE Dems than Reps (See the end of the file), so I’m confused by all the Lib spin that suggests this poll is biased in favor of GWB because it oversamples Reps.
Of course, if the poll does not toe the Lib line that the race is a dead heat, then the poll is to be discarded – Sheeeeesh!!”
Of course, now that the polls (including Gallup and CBS) are reflecting the effects of Bush’s undeniably disasterous debate performance, we are now advised by this same poster above,
“This (CBS) poll has about the same credibility as the forged document story IMO.”
Message: we “libs” should take the CBS poll seriously only if it shows Bush ahead, but if it shows Kerry is tied or leading it must be total crap.
Sheeeesh, indeed.
The recent swing in the polls lends credence to the idea that party ID is indeed fungible. People felt good about Kerry after the debates and so they identified themselves as Democrats (came out of the closet, if you will) to pollsters. Hence we see party ID breakdowns that now more closely reflect 2000 exit polling.
> …their Abu Gharib ambush on GWB…
>
> Posted by Smooth azz at October 5, 2004 05:02 AM
Really? So you think that ABC is more responsible for Abu Grahib than GWB is? Fascinating!
It’s probably more important to watch Friday morning job’s report than Friday night’s debate.
That’s because, in addition to the monthly number, the feds. will issues benchmark revisions for the entire year. History shows they can be very large. The average upward revisions since 1979 has been 308,000.
Robust job growth would speak for itself, which is definitely good news for the president.
On the other hand, a weak report and I think we’re looking at another 2000 squeeker.
How can you or anybody possibly know the actual number of GOP/Dem voters will turn out in November?
technically your right, we can’t know for sure the actual number of gop/dem voters. But you could make the same comment about any age group, any religion, a city vs. suburb, racial groups, ect…… by your logic you can’t do polling.
American Research Group has Kerry up by 3 in new poll today, 47-44.
http://www.americanresearchgroup.com/
The polls won’t matter – GASP – is Kerry and Edwards play it right.
I think Edwards AND ESPECIALLY KERRY have to attack attack attack in these next two debates. They can’t cede an inch of ground. If they get Cheney to come off looking half as angry, or mean as Bush that is a win. Kerry should try to drive Bush right over the edge on Friday. I think comparisons to his father, George H.W. Bush may do it. If Bush cracks again on Friday you can stick a fork in him, he will be done.
It has occurred to me, however, that maybe these polls just miss Democrats. In 2000, for example, I thought most had Bush with a consistent lead in the month before the election. I don’t remember a one showing a Gore lead. Since the election turned out to be a tie, one would have expected a distribution of polls on either side — some for Gore, some for Bush. I don’t remember it that way — am I wrong? And does it suggest some systematic, if perhaps small, bias in national polls?
American Research Group has a poll with Kerry up by 3 points in RV’s, whether Nader is in or not. It has showed Kerry consistently doing well, without the fluctuations in other polls. Wonder why?
I would really be interested in some attempt to reconcile the polls, which seem to come in two groups. My impression is that some have shown more volatility and gave a wider advantage to Bush post-RNC. Others have had less volatility all along, but more favorability to Kerry generally. On the other hand, I sort of thought Pew was one of the latter group and they’ve suddenly had Bush in the lead in their last two.
Help!
Has anybody seen the new ARG poll showing a slight Kerry lead? Ruy, how credible is this new poll?
Oops, I forgot the URL for the latest NH poll. Here it is, my apologies:
http://www.unh.edu/survey-center/elect100404.pdf
Marcus,
That’s my point exactly: We don’t really know if the EARLIER or CURRENT ID mix is correct, do we??
For all we know, CBS/NY Times COULD HAVE padded their earlier polls with Reps, to pounce with a Kerry “comeback” story later, or it could be that Dems were more enthusiastic after the debate, and may have been more willing to respond to a pollster.
The fact is: We don’t really know the motive behind them suddenly reversing the party ID to now support Kerry. Given that BOTH entities are run by partisan Dems, CBS complicity in perpetrating a forged document fraud, and the NY Times open disdain & hostility for GWB, I don’t trust them at this point.
Finally, in response to Tony and others, I have no problem recognizing a poll that is favorable to Kerry – And will gladly acknowledge if & when he takes the lead. I’m just skeptical that the race has changed fundamentally from where it was prior to the debate because CBS/NY Times, Newsweak & CNN/Gallup decided to flips party IDs in their samples.
Cloudy,
Today the Columbus dispatch published registration information. In Ohio people do not declare a party when they register, so estimates of the impact are based on whether the location is heavily democratic or republican.
Yesterday alone, 20,000 people registered in Franklin County (Columbus) and 98,000 had previously registered, which brings the total of registered voters to 837,000. I read somewhere else that this is approaching or exceeding census estimates for the county.
Cuyahoga County (Cleveland) also received about 20,000 registrations yesterday, and they had received 235,000 voter registratin cards, 100,000 of them estimated to be new voters. The others probably name changes etc. I think if you haven’t voted in four years you have to reregister, so I think many were people who hadn’t been active who want to vote this year. Cleveland is heavily democratic, Columbus is predominately democratic.
In Hamilton County -Cinncinnati -(more Repub leaning), new registrations totaled 64,000 as of Saturday.
Even small, rural counties like mine, are getting record registrations, around 200 a day as of last week.
Typically that would be thought of as a problem in this repub leaning county, but this is no typical year. Some of these new votes include friends of my college aged kids who are voting Kerry.
Spent Saturday walking the neighborhoods with a Dem list that represented swing voters or undecided voters. Finding out if they had decided, trying to talk with undecideds. My impression, not very scientific, of course, but undecideds had already switched to Kerry at about 7-3. A few still undecided, but sometimes I think we ought to let some undecideds alone. I mean why should my well researched, passionate vote be canceled out by someone who doesn’t know who they are going to vote for as they walk in the booth?
More good news, I talked with folks who had decided to vote for Kerry after seeing the debate.
My prediction is that Ohio will go for Kerry.
Yard signs are more for Kerry here (again this is repub territory) and we can’t keep them stocked in the Dem storefront.
Those polls do not pick up on all the new registrations. Took my colleagues out to lunch last week to talk about why I’m voting for Kerry.
It will be all about GOTV now that everyone is registered.
Go Johnny go!
Smooth, are you kidding me? Clarke and Abu Gharib shook faith in Bush considerably. As I recall, Kerry was way ahead all that time. The NG story didn’t have much negative impact, not because memos were faked, but because NOBODY CARES about Bush’s NG service.
BTW – Where’s the outrage over the fake Kerry story from FOX (which was not only posted on its Web site but discussed on FNC) and the “who among us” quote made up by Maureen Dowd? Oops. Doesn’t fit the librul media paradigm. Never mind.
Smooth,
Love how you’re getting into the polling internals now that things aren’t going W’s way…
From the RealClearPolitics site, now using only polls from after the debates, Bush’s lead is down to 1.8% in a 3-way, mostly LV’s (total of 6 polls) and 0.4% in a 2-way, again mostly LV’s, total of 5 polls.
If anyone’s not noticed, SJ will criticize the polls most in favor of Kerry while not addressing the polls in favor of Bush. Good to see Ruy being evenhanded and posting all the polls.
I’ll happily take a 0.4-1.8% Bush lead at this point. If Kerry can keep it at that, the late break should go his way and be enough to put him over the top.
This will probably be my only post today. Lots of stuff on my plate.
Jazz
Several polls show that actually contrary to your propaganda the differential between men and women for Kerry was at a nadir (not only according to CBS, but also Newsweek and other polls).
This poll weights respondents to reflect the percentage of votes in the 2000 election that came from heavily Democratic, heavily Republican, and swing counties. News reports (the best ones in the Times, to be fair) make clear that new voter registration is much heavier in Democratic than in Republican counties. The poll results which are weighted to the 2000 voting pattern will not reflect this.
Also, the poll is weighted for education and not for income. This will fail to correct the bias of low-income Kerry voters being unrepresented in the sample, and may actually increase the bias through the weighting process. (See my comment on the last CBS-Times poll.)
again, another periodic update on the website for helping with voter reg in swing states, including the blockages of registrants:
http://acthere.com/
By the way, the Kerry camp has been allowed by the FEC to start raising money for a fund for any post-election litigation and such
About the “gap gap”. We have two dimensions and TWO kinds of gap gaps now. Before, there were differences in the amount of bounce from the debates, but all the polls agreed the race is now within two % + or -. Now there are two polls that have minimal bounces AND still show a large lead of 5-7 points remaining for Bush. Of the debate bounce, either they see one or two percent increase at most and no more, or well over 5% and nothing in between. Over now eight polls (NY Times, ABC, Pew, Zogby, Gallup, Carville, LA TImes, and NEWSWEEK, there is not only no consensus, the results don’t even spread along any sort of continuum.
If one averaged them all out you would probably get about a four percent bounce net, leaving about a two percent lead to Bush — my guess that’s where the truth is. And Cheney v. Edwards should help at least another notch, though not nearly as much.
I really wonder again if there are any surveys of the new registrations to see the party breakdown and location of the new voters, probably concentrated in swing states such as Ohio, Nevada, and (probably less likely to affect pres campaign, MO). If the news of record registrations is true this could easily decide the election.
A few substantive political points
AGAIN, IT IS IMPERATIVE THAT THE FLIPFLOP ISSUE BE CONFRONTED FORCEFULLY AS A “MERE SPIN” AND SOON, POSSIBLY BY EDWARDS IN THE DEBATE. He’ll need to back it up, eg showing the bogusness of the exploitation of the ‘voted for it before voted against it blooper’ as well as the weakness of the claim on NAFTA and the Patriot Act, challenge the media, and secondarily, note that Kerry’s campaign could easily match all the supposed Kerry flipflops at their own website and then do it. “Mere spin” can then at least mitigate the solidified flipflop cliche that has 60% of voters saying Kerry just tells people what they want to hear.
I also hear much talk about the draft, but little attention to the possibility of a draft that doesn’t require MILITARY service, although allowing it to be fulfilled by noncombat military duty (with those signing up for combat duty of course exempt). This could be the deciding issue of the election easily.
Some of the incredible spins on Kerry from the debates, from people in mainstream media and watchdogs who should know better. The ‘global test’ issue cast as a requirement for PRIOR world acceptance by Safire, the $200 billion scored for being too high as there’s some official (VAST UNDER)estimate of the money costs of the war. Those costs don’t cover long term commitments like paying for those injured or disabled over the long term, shifting costs on to state and local governments, eg from National Guard callups, hidden and delayed military costs (even wear and tear), etc etc etc. Kerry if anything UNDERestimates the costs of the wars like everyone including the media does. This is a pet peeve of mine. Pre-emptive war was read by Safire to mean pre-emptive NUCLEAR war, which Kerry never suggested, and numerous observers are completely confused about N Korea, including some media watchdogs. Biden pointed out that the Chinese WANT us to go into bilateral negotiations to SUPPLEMENT, as Kerry suggested, the multilateral approach. Yet these misstatements are not only repeated by the Bush campaign but by highly respected journalists, commentators and watchdogs that people trust. People need to watchdog these.
In the interests of honesty I’ll point out that I am much more dovish than Kerry and way to his Left, just preferring him to Bush. I also didn’t join the many of my colleagues who followed the Ralph ‘if Roe v Wade gets overturned, so what?’ Nader juggernaut in 2000 (even as they supported Hillary against Lazio.
******************************************
The flipflop issue needs to be effectively countered soon so the last weeks of the campaign can successfully focus on DEFICIT, JOBS, MEDICARE SCAM, ENVIRONMENT, and MEGATAX CUTS for the RICH in the midst of all this.
I still say the machine agenda is a better predictor of who will be the next prez than all this.