Thursday, December 17, 2020

Johnson County Number Crunching, Part 4: Who Voted

Six weeks post-election I'm well on the road to recovery. I went non-stop for about a year and a half - from the time I started booking caucus rooms in April 2019, through the primary and into the presidential. That's three whole Taylor Swift albums if that's how you keep score.

It takes some time post-election to process voter history records - a recount slows that down a bit - but we wrapped up a couple weeks back and now I've had some time to figure out who voted by party, how they voted by method, and I've done some guessing about how the no party voters went here in the People's Republic.

Voter history records are a different thing than canvassed vote totals, and the two different sets of numbers are rarely a perfect match. The main problem is that 99 counties are all balancing the books at once and that registration activity is constant. So sometimes another county takes a voter away from us before we can give them credit for voting - which is NOT the same thing as counting the vote. The statistical report for our county is within 20 of the canvassed number of voters, which is about as close as it ever gets.

Here's the party breakdown of voters by party, with a grand total and split into by election day and early voting. 

Voters Early Eday Total
total 60915   23262   84177  
Democratic 39034 64.08% 7150 30.74% 46184 54.87%
Republican 8714 14.31% 7114 30.58% 15828 18.80%
Libertarian 250 0.41% 271 1.16% 521 0.62%
Green 100 0.16% 51 0.22% 151 0.18%
No Party 12817 21.04% 8676 37.30% 21493 25.53%
The post-election registration percentages were 52.2% Democratic, 18.4% Republican, and 28.5% no party.  Overall turnout is very, very close to that. It's just a little more Democratic and a little less no party.

While the early voters were overwhelmingly Democratic, the Democrats and Republicans were nearly tied on Election Day. More no party people voted on Election Day than either major party.  Libertarians were the only voters more likely to vote on Election Day than early. I suspect that's student-age election day registrations; statistically Libertarians lean very young.

The 72.4% of the total vote that was cast early is down a little bit from the record 76.7% from the primary, probably due to the mail scare and due to COVID fatigue. The primary voting window in May was during the peak of what little semi-shutdown down Iowa had.(One stat we didn't keep: how many mailed-out ballots were returned through the postal system vs. our drop boxes.)

There's four voting plans available to most voters: the polls, mail, satellite sites, and in person at our office - which this year was actually drive-thru voting in our parking ramp. There's also "UOCAVA" voting, which stands for Uniformed and Overseas Citizens Absentee Voting Act. People outside auditor's offices often call these "military" ballots, but a lot of overseas civilians get them too. In most years there are also "absentee team" ballots, sometimes called "nursing home" ballots. We had a tiny number of these, to hospital in-patients; I lumped them into in-person below.
Here's how voters in each party chose their voting plan. 

Party Overseas Mail In-Person Satellite total early Polls Total
Democratic 424 0.92% 23619 51.14% 10696 23.16% 4282 9.27% 39021 84.49% 7163 15.51% 46184
Republican 58 0.37% 4504 28.46% 2974 18.79% 1174 7.42% 8710 55.03% 7118 44.97% 15828
Libertarian 4 0.77% 87 16.70% 112 21.50% 47 9.02% 250 47.98% 271 52.02% 521
Green 1 0.66% 44 29.14% 33 21.85% 22 14.57% 100 66.23% 51 33.77% 151
No Party 180 0.84% 6515 30.31% 4192 19.50% 1917 8.92% 12804 59.57% 8689 40.43% 21493
Total 667 0.79% 34679 41.30% 18007 21.39% 7442 8.84% 60885 72.33% 23292 27.67% 84177

Democrats were much much more likely to vote by mail and much, much less likely to go to the polls.  Everyone was about equally likely to vote at a satellite or at our drive-thru. No partys (I really hate the word "independent") and Republicans were similar in behavior - less likely mail, more likely polls. Libertarians were the only people MORE likely to go to the polls than vote early,

Here's the vote totals, which we're more familiar with. This is also lower than the total turnout of 84,198. There were 347 presidential under-votes, which may seem hard to believe but is in the normal range. Sorry, Kanye, but I've combined the lower tier and the write-ins into an "other" category.

Votes Early Election Day Total
total 60672   23179   83851  
Biden 48664 80.21% 10513 45.36% 59177 70.57%
Trump 11024 18.10% 11901 51.16% 22925 27.23%
Libertarian 522 0.86% 442 1.90% 964 1.15%
Green 148 0.24% 65 0.28% 213 0.25%
Other/write in 314 0.52% 258 1.11% 572 0.68%
In my years here, Democrats have been more likely to vote early than Republicans, but this year the trend was especially exaggerated. There seemed to be intent and purpose to Republicans wanting to vote on Election Day. Even though about 3/4 of the vote was early, Trump got more votes on Election Day than early. He actually won Election Day. 
As I noted earlier, the big voting shift in Johnson County this cycle was voters moving from third parties to Biden. Of those who didn't, Election Day voters were twice as likely to vote Libertarian or Other Third Party than early voters.

Now I indulge in sheer speculation and try to figure out what the no partys did. There's no way to tell, of course, but I'll play with numbers and make some false assumptions: that all members of a party voted for their party's candidate, and that only No Party people did write ins or voted for miscellaneous candidates. Then I just subtracted the difference. 

Estimated No Party Votes Early Election Day Total
total 12574   8593   21167  
Biden 9630 76.59% 3363 39.14% 12993 61.38%
Trump 2310 18.37% 4787 55.71% 7097 33.53%
Libertarian 272 2.16% 171 1.99% 443 2.09%
Green 48 0.38% 14 0.16% 62 0.29%
Other/write in 314 2.50% 258 3.00% 572 2.70%

No party early voters voted much like the rest of the early voters, nearly four to one Biden, and just a little bit more third party. But the Election Day no party voters leaned about 5 points more Trump than the rest of election day voters. As we saw, the Election Day voters by party affiliation leaned disproportionately R as well. 

The Election Day no partys likely had a lot of weak voters or people who though COVID was Fake News, both groups that would lean Trump. End result was that no party overall was very blue by the standards of a normal Iowa place but a little redder than the standards of the People's Republic.

This really nails down my long time theory: No party voters are just like partisan voters, they just don't like checking the box for whatever reason, and once you get them in the booth they pretty much behave like partisans.

Sunday, December 13, 2020

The Caucuses Greatest Hits Version 5.0 - 2020 Edition

In a post I first wrote back in 2006 and have updated periodically. I've looked at and ranked all the caucus cycles back to 1976.  As for history and the caucuses themselves, a mixed bag.  Irrelevant nearly half the time, critical a little less often.

Not Worth The Airfare To Waterloo

21. 1984 and 2004 Republican. The Republican tradition was to hold no presidential vote at all in incumbent re-elect years.That tradition ended in 2020, not because of Trump's minor opponents but because state party leadership seemed to think not having a vote in 2020 would hurt the case for First in 2024. Ironically, this happened as other state Republican parties canceled their contests.

20. 1996 Democratic. The word went down from Des Moines to the Democratic county chairs: “The President would like a unanimous re-nomination and this WILL happen.” Self-starters in a couple lefty college precincts elected a very small handful of Nader and Uncommitted protest delegates, but those results got swept under the rug. Clinton came out and campaigned the final weekend, largely to step on the GOP story (Actually Being President trumps winning the caucus), but it was in basketball arenas, not chat n’ chews.

18 (tie). 2012 Democratic.  As close to an unopposed caucus as possible short of “The President would like a unanimous re-nomination and this WILL happen.” The state party went to bat for actually having an alignment, which Chicago didn't want. But without a live person as an Obama opponent (despite Bernie Sanders' suggestion), the dissenters were split between Uncommitted and crossing over for Ron Paul. In the end the Uncommitteds, mostly made up of folks allied with the simultaneous Occupy movement, made a lot of noise out of proportion to their 1.5% of the delegates. Rated up one notch because that 1.5% actually got honestly reported, not suppressed as in 1996.

18 (tie). 2020 Republican. Iowa Republicans broke with tradition in 2020, in part because they were concerned that with no actual vote they would put First at risk (this was before the 2020 Democrats removed all doubt). But even though they counted the votes against Trump, Bill Weld and Joe Walsh proved to be woefully ineffective even as protest vote placeholders.

17. 1992 Republican. The Pat Buchanan Brigade was looking like a serious threat to win New Hampshire - he ended up at 37.5% there - but the inside the Des Moines Beltway crowd stuck with the tradition of not having a vote in an incumbent year. That decision was a small win for George HW, and that no-vote decision was more important than the 2020 decision to vote. So this gets the highest rank of the de facto uncontested caucuses.

Ultimately Irrelevant

16. 1992 Democratic. Hometown boy Tom Harkin runs and wins big, though not as big as it looked because of some skilled realignment work at viability time. That 76% Harkin delegate count included a lot of stealth supporters of other candidates.

Paul Tsongas was already on the ground in Iowa when Harkin announced, but he quickly bailed. There were a couple feints from Bob Kerrey and Jerry Brown but nothing serious. Everyone showed up for the cycle's lone cattle call, the then-Jefferson Jackson dinner, but between low interest and a blizzard the hall was half empty.

In the end, Iowa kept first place after `92 only because Tom Harkin was the only Bill Clinton rival who enthusiastically jumped on the bandwagon. The other also-rans could barely hide their obvious contempt for Clinton. (Kerrey and Brown probably wrote themselves in that November.)

The long term importance of the 1992 caucuses may be that Bill, and by extension Hillary, Clinton did not have to retail campaign in Iowa, and that had a ripple effect into 2008 and 2016.

15 (tie). 2020 Democratic. In the end, the importance of the 2020 Democratic caucuses will be that they were the last, but what about the cycle itself, in terms of the traditional momentum and winnowing role? 

The results problem killed the momentum but even early on it was clear there were two leaders. And in the end two winners.

The delegate count winner, Pete Buttigeig, was a flavor of the week fad who had little traction after Iowa and who at best will end up with a make-work administration role rather than the high level cabinet post he craved. He got his breakout from obscurity not from the caucus “win”, but from pre-Iowa polling and fundraising.

I don’t have an answer for the Buttigeig problem: someone who, despite being absolutely not ready for the White House, is a rising star stuck in an area where winning a higher race is difficult. But if our rhetoric is Fight Everywhere, he should try. Indiana Senator Todd Young is up in 2022.

The popular vote winner, Bernie Sanders, lost ground from his 2016 tie, as the Not Hillary voters he had to himself that year dispersed to the rest of the field.  Despite his strong core of support, he was most people's Anyone But choice and started losing contests as soon as voters consolidated behind the eventual winner.

(As much as I loathe Sanders, who as a non-Democrat should never have been allowed to participate in Democratic nominating contests, I consider him the real winner. We kept score two ways and I consider popular vote the more fair way to call the winner. He probably had more raw votes in 2016, too, but we didn’t keep score that way that year.)

The eventual president-elect finished a poor fourth, which in prior years would have a drop-out finish. In the end, Joe Biden was the real winner of the caucuses, because he got a mulligan on his poor results. By the time we figured it all out, South Carolina and Super Tuesday had happened and the train had left the station.

Not a single candidate quit based on the Iowa results. The winnowing happened before Iowans even caucused, on factors like debate qualification, fundraising and national polls. One of those winnowed-out before Iowa candidates, Kamala Harris, became the running mate.

So the caucus results had no real impact on the nomination outcome at all. It's like we never happened. And for all the “organizing” we did – this is NOT an attack on the organizers and volunteers - in the fall we wound up losing the state to Trump, losing a Senate race we had a shot at, losing two US House seats, and losing ground in the legislature.

14. 2000 Both. On the Democratic side Al Gore easily beat Bill Bradley in what was merely the first moment in the overall national dynamic; Dollar Bill made his stand on friendlier turf in New Hampshire and fell just short there, and that pretty much ended it. (Still mad: Anyone who thinks things were "rigged" against Sanders should have been a Bradley 2000 person. Basically the entire power structure of the party was behind Gore and the prevailing attitude was "how DARE Bradley 'challenge' him.")

On the Republican side it was like one of those boycott-era Olympics: W won but the toughest competitor, McCain, was a no-show playing a Screw Iowa strategy. The truly significant GOP event was the straw poll that winnowed out more candidates (E. Dole, Quayle, and Buchanan bolting to Reform) than the actual caucus (Orrin Hatch, as if that wasn’t obvious).  Comic relief: People who took Gary Bauer seriously, Alan Keyes in Michael Moore’s mosh pit.

Secondary event in nomination contest

13. 2016 Republican. Whichever contest was first would have narrowed a field that peaked at 17 candidates. The biggest event of the cycle was actually a non-event - the Ames Straw Poll that had been the dominant pre-caucus event from 1987 to 2011 was first moved out of Ames, then canceled entirely when the leading candidates refused to show up.

The field was down to a mere 12 by caucus night. Ted Cruz, Donald Trump, and Marco Rubio, in that order, each took about a quarter of the vote, with the other nine splitting the remaining 25%. The immediate two past Iowa winners, Huckabee and Santorum, fizzled, and that effectively ended their campaigns.

Rand Paul failed to keep his father's coalition, because all of the "he's alternative, dude!" vote was caucusing for Bernie Sanders. The dudebro overlap between Ron/Rand Paul and Sanders supporters, which makes little ideological sense but is clearly a Thing, is a Ph.D. dissertation for someone.

Cruz's win turned out to be an anomaly and a relatively minor event. Cruz wasn't even Trump's final opposition - that turned out to be John Kasich, merely because he refused to quit.

12. 1980 Democratic. The incumbent won the first test of Kennedy-Carter, but that battle of giants was played out on a national, even global, stage and Iowa was a bit player.

11. 2008 Republican. Important to the dynamic of the contest, but not central to the result.

Mitt Romney was looking like the guy to beat in December 2007. Which Mike Huckabee did in January 2008, after first beating Sam Brownback at the straw poll to win the mantle of THE religious conservative candidate. Had Iowa Republicans gotten behind the Mitt, they may have headed off the chaos that was the GOP field in January. Instead, we proved that there was no there there for Fred Thompson, and that the Ron Paul Яэvoutionaries were noisy in disproportion to their actual numbers (but see 2012 below). But really, we just stirred the pot, and the decisive event was in Florida between two men with Screw Iowa Lite strategies, Rudy Giuliani and John McCain.

Our next contest was very similar, but the tiebreaker is that the Iowa winner actually won the nomination:

10. 1996 Republican. What might have been: Pat Buchanan was within 3% of Bob Dole, but the social conservatives in Cedar Rapids backed Alan Keyes instead; Keyes thus won the second biggest county. One minister at one mega-church makes a different choice, and we’d have had a major upset.

Some all too obvious field winnowing (Dick Lugar???) happens. Phil Gramm gets out too, but his real stumble was in Louisiana’s jump-the-starting-gun contest a week earlier. 

Comic relief: Easily the funniest caucus! Dole, genuinely witty in his non-Satan mode, Steve Forbes the android, Alan Keyes… but they all pale next to Morrie Taylor, the tire magnate who literally tried to buy a win one vote at a time. Failed miserably but looked like he had more fun than the rest put together.

9. 1988 Democratic. Other than Tom Harkin's favorite son run which doesn't really count, this is the only time from 1976 to 2016 that Iowa Democrats did not vote for the eventual nominee. The nomination contest came down to Dukakis vs. Jackson, neither of whose fortunes were affected by Iowa.

In `88 Al Gore, dirty Prince lyrics still ringing in his ears, was the first candidate to use the Screw Iowa strategy.  It's never worked (save for the Harkin year), but nevertheless Gore wound up outlasting the two Iowa leaders.

There's a story, long told by Paul Simon loyalists, that a county chair sat on his Simon-friendly results until the Register had printed its GEPHARDT WINS headline, which mattered in the pre-internet era. Rules got changed after that so that results are reported direct from the precinct to the state without going through a county chair, but this one proved the winner-take-all-news theory that was prevalent at the time (and which was supposed to be the anchor of my aborted masters' thesis).

Comic relief: Gary Hart’s last minute return to the race, campaigning with his wife.

8. 2012 Republican. The real importance of the 2012 Republican caucuses was not its relatively small role in designating the nominee. That was always going to come down to Mitt vs. Not Mitt.  Rick Santorum never really got the bump from the win, because of the dead heat, the botched result announcement, and the recount that delayed the final result. And also because Sheldon Adelson kept Newt Gingrich on life support far too long,

No, the real importance was what happened to the Republican Party of Iowa after the presidential vote. The Romney and Santorum people both said "yay, we won," went home, and both in turn were right. The Ron Paul people stuck around, elected themselves as all the delegates and committee people, and took over the state party structure.

The consequences had a huge ripple effect through state, and even national, internal Republican politics for the next two years, until Terry Branstad, Jeff Kaufmann and the rest of the grownups took party control back in 2014 (the most important OFF-year caucus). This one may move up the charts depending on the long-term fate of the caucuses, and so may the next:

7. 2016 Democratic.  Iowa was a big deal - in the same way that the first post in an epic flame war that eventually breaks Godwin's Law is a big deal.

Had Hillary Clinton solidly beaten Bernie Sanders in Iowa, 2016 would have been over as fast as you can say "Bill Bradley" and the whole Sanders phenomenon would never have happened. Oh, he might have stayed on some ballots and accumulated a few votes. But without the dead heat in Iowa, and the money and attention that followed, he would have been an asterisk, like Dennis Kucinich playing out the string in the late states in 2004 after John Kerry had clinched and everyone else had quit.

I have long said, and the 2020 results confirmed, that half the Sanders vote was simply Not Hillary, and that alone would have gotten Martin O'Malley to 30 points in Iowa had Sanders not run. Indeed, that was probably O'Malley's whole strategy, to be the only person willing to run against Clinton.

The fact that Sanders was even allowed to run in 2016 without joining the Democratic Party was a decision by the DNC - ironically, headed at the time by the same Debbie Wasserman Schultz who supposedly "rigged" the nomination against Sanders. DWS's inability to take Sanders seriously as a threat to Clinton, and her under-estimation of Clinton's negatives, are just more signs of her ineptitude.

Sanders himself may have faded (from 49.9999% in Iowa on Caucus Night to about 25% in 2020) but his campaign mainstreamed a progressive stance that other candidates at lower levels are now seeing success with (in places very different than Iowa). The long range ranking of 2016 may move up if this turns out to be a permanent change.

Significant event in nomination contest

6. 1988 Republican. Pat Robertson pushes George HW into third place. Robertson was insignificant thereafter, but the blow made Bush go on a fight of his life attack against Bob Dole in New Hampshire. Dole took the bait and was goaded into “stop lying about my record.” This convinces HW that hard negative was the way to go. That road went through the flag factory and Willie Horton, and ended at the White House. Comic relief: Al Haig.

5. 1984 Democratic. Gary Hart barely squeaked past his old boss, George McGovern. But second, no matter how distant, was enough to make him the Not Mondale and propel him up about 40 points in eight days for a New Hampshire win, a brief but genuine shot at the nomination, and (pre-Donna Rice) 1988 front-runner status. The Right Stuff sank like Gus Grissom’s capsule, and you're an old timer if you catch that reference.

Decisive event in nomination contest

4. 2004 Democratic.

Iowa was the whole ball game in 2004. Nothing that happened after Iowa mattered nearly as much as what happened in Iowa.  The guy who won got the nomination, and the guy in second got VP.  And the guy in third?

The Dean Scream goes down as the single most memorable caucus moment, but everyone forgets The Scream was after The Much More Important Disappointing Third Place. (Which was helped by the most coordinated cross-campaign effort I ever saw, as the other candidates worked together to realign in whatever way hurt Dean the most.) After Dean had been the front-runner for months, Iowans got scared at the last minute - mostly thanks to Dick Gephardt, who went on a suicide attack that took them both out and set Gephardt up for reward in the Kerry Administration.

Made History

3. 1976 Democratic. This one made both Jimmy Carter and the caucuses themselves. Carter didn’t actually win this, you know. He was second to Uncommitted. But I know folks who still brag “Jimmy Carter slept on my couch.”

I’m torn about ranking a caucus that directly produced a president below one that didn't. But read on.

2. 1980 Republican. In the first true Iowa Republican caucus, obscure former ambassador, spy boss, and failed Senate candidate George Herbert Walker Bush shocked the ten foot tall colossus of the GOP, Ronald Reagan. This one win puts Poppy on the map and ultimately on the ticket (after the botched Ford “co-presidency” deal at the `80 convention).

So why rank this ahead of Jimmy Carter, especially since Bush Sr. lost that 1980 nomination? The ripple effect. No Iowa win = no Bush 41. And with no HW, do you REALLY think Bush 43 or 45 (please clap) would have made it on their own? 1976 made a president, but 1980 made a dynasty.

Number 1: 2008 Democratic. There's no question the 2008 Iowa Democratic caucuses created a president. Iowa was the honing ground for Barack Obama's message and appeal and ground game. We eliminated the entire second tier, and proved that voters in one of the whitest places in America would support a black candidate. Remember, a lot of African-American voters were sticking with Hillary Clinton before Iowa, because Obama "couldn't win." Iowa shattered that myth and the perception of Clinton's inevitability.

It's too soon to tell, and Trump's win blurred things, but the 2008 caucuses may have ushered in not just one president, but a whole era, a new alignment of states that ends the 1968 Nixon-Wallace southern-western coalition for good, at least at the presidential level.

2008 was a whole new map. As late as the first John Edwards campaign, people were sill seriously saying it was impossible to break the Republican "electoral college lock" without southern rural white male voters. Trump has since proven those voters are gone forever - maybe in the urban and exurban North as well.

But that old South has been replaced by the new South, driven by suburban voters who migrated from the North, women, and minorities. Florida has disappointed teh last two times, but Virginia is now solid blue, North Carolina is in play, Georgia flipped in 2020, and even Texas is on the horizon. Barack Obama fueled this alignment, which would not have been possible without that Iowa win.

The 1976 caucuses made one president, but his victory is a mere footnote to a Republican era, brought about by the intensity of Watergate and the Nixon pardon. The 1980 Republican caucuses made two presidents, but they followed the electoral footsteps of others.

How many presidents in an era? Obama wasn't able to transfer this alignment to an immediate successor, because some anchors of the old coalition fell in 2016 (Trump essentially drew three cards to an inside straight with his narrow Wisconsin-Michigan-Pennsylvania wins), but Biden seems to have partially restored and in some places expanded the Obama coalition. On the other hand, he slipped in others, most notably the rural Upper Midwest. Some of those rural Iowa Obama counties aren't ever coming back - but in the big picture Georgia and Arizona are a nice tradeoff.

If the 2008 caucuses ushered in an Obama Realignment, like the FDR Relignment or the Nixon-Wallace Realignment, they could lead to four or five presidents.It's a weaker case than it was a few years ago, but 2008 still deserves the number one spot.