Monday, November 19, 2018

2018 General Election Number Cruncher

Two weeks ago I left my office at a relatively early 11:20 PM and headed a few blocks to what, based on everything I had seen all day, I expected would be a victory party.

As I arrived I saw flocks of people leaving, looking like the cat had just died, and one of the first filled me in that they had just called the governor's race for Reynolds.

It's taken me a couple weeks to absorb that, to catch up on the mental pile of election cleanup work confronting me in what is now just 40 hours of work a week, and to absorb the local numbers. And as I look at the Johnson County numbers, I see a landscape with little relief, a monochrome map of dark blue and darker blue.

Johnson County could not have done more. We got Obama-level turnout and Obama percentages. I say this not to brag, but to acknowledge that our best in the People's Republic is not enough to win the state. It is a necessary condition, yes, and we have to do it again each and every time, but we have to do something else somewhere else.


But as I look at my local numbers, a landslide this deep inundates any minor differences, and the partisanship of a general election masks subtle local trends.

The biggest trend to note is turnout.


I had estimated (this is literally my job) about 58,000 total voters - 30,000 early and 28,000 on Election Day. That would have been an incremental increase from our old "real" record of 52,959 in 2014. (2010 was technically bigger at 53,855 but that was with 2500 undervotes for governor and a weird turnout spike on campus caused by a bar age admission issue.)

We had seen the increase coming during early voting, which ended up 4000 over my prediction, and ordered more ballots from the printer. Almost immediately on Election Day it became clear we would need them. By the 11 AM turnout update we decided to ship out everything we had. Campus still lagged, but they kicked in before the 3 PM update. We had to print a few extra in-house for one of the dorm precincts but otherwise we were fine.

Turnout ended up at 68,262, nearly 15,000 above the record and closer to the presidential record than to the old midterm record. Like I said: Obama turnout.

And the Democrats got Obama shares of the vote.

Better, in fact: Barack topped out at 69.9% in 2008. This year, Dave Loebsack led the Democratic ticket in the "seriously" contested races at 72.4%, a notch above Fred Hubbell's 71.6 and just below treasurer Mike Fitzgerald's 73.3 over token Republican, um, ... (I had to look) Jeremy Davis. (Tom Miller won 84% in a race with a Libertarian and no Republican.) With his TV name ID, State Auditor-elect Rob Sand ran just a little bit better than the other down-ballot challengers; there was very little gap between Tim Gannon and Diedre DeJear.

By coincidence the Johnson County vote split almost exactly between early voting (34,119) and election day (34,143), for the closest race of the election. That makes comparing numbers easy. And almost race by race, the early vote numbers were 15 to 16% better for Democrats than the Election Day tallies. So many Democrats voted early that there were a disproportionate share of Republicans left yet to vote on Election Day. Still, Hubbell's 64% of the Johnson County Election-Day-Only vote was far better than he did in any other county, and his 79% on the absentees approaches Kim Jong Un levels.


Johnson County voters had their eyes on the prize and were in no mood to protest-vote. The write in vote nearly vanished in 2018, dropping from 964 (1.25%) in the presidential to just 24 votes for governor.
Voters exercised their write-ins on an obscure back of the ballot race for a two year short term on the Soil and Water commission, where no candidate filed. 3319 voters cast write-ins with the winner, appointed incumbent Bonnie Riggan, tallying just 93.

The back story on that is Soil and Water had two separate contests: two full four year terms and the two year term. WHY state law doesn't just have everybody run in the same race and give the third place person the short term, I don't know.

The language on the paperwork is non-intuitive and asks if the candidate is seeking "to fill a vacancy." EVERY appointee gets this wrong - they think "the seat isn't vacant - I'm in it." Some township people got it wrong, too. But township officials only need to re-do an affidavit. Soil and Water commissioners need signatures.

So on the day before the deadline, all three incumbents filed for the two full terms. The next day, the deadline passed and no one had filed for the short term. Rather than run against her colleagues, Riggan withdrew to run as a write in for the short term.

But the blank line on the ballot caused much electoral amusement. Everyone and their cousin thought it would be cute to get their friends to vote for them. There were at least three social media campaigns going on. And us election workers couldn't explain any of this to voters, who usually only asked after they had their ballots, because explaining the story pretty much says, vote for Bonnie.

The overall non-two party vote fell from 7.6% for president in `16 to just 1.8% for governor this year. If you thought Gary Siegwarth of the "Clean Water Party" would be a refuge for Hubbell-hating lefties, you were wrong as he scored just 0.4%.

So the bulk of the third party vote (1.35%) went to Libertarian Jake Porter, taking a big drop from the 3.6% won by 2016 presidential candidate Gary Johnson. Porter's statewide 1.6% means the LP loses the full party status it craved for decades after just two years, the same fate that befell the Greens in 2002. And it wasn't just a top of the ticket thing - the LP was under 3% in all of the D vs R vs L contests (typically third parties do better in down ballot races that some voters see as less "important.")

The third party contender I was worried about was Daniel Clark, a former Bernie Sanders national delegate who defected to Jill Stein and who was running as an  independent against Dave Loebsack.

There was pro-Clark paid Facebook advertising claiming to be from Democratic Socialists of America (DSA says it wasn't really them), and the far left has always had a chip on its shoulder about Loebsack. My theory has always been: Since he beat the last anti-war Republican, Jim Leach, the left expected Loebsack to be at the left edge of the party - today we would say "Bernie" or "Alexandria," but in the vernacular of 2006 it was "Kucinich."

I worried that, because of Loebsack's history of big Johnson County wins, disgruntled lefties unhappy about a Hubbell-led ticket but eager to defeat Kim Reynolds would use their protest votes against the supposedly "safe" congressman. But the mood was to punish Republicans for Trump, and that meant voting for Democrats. Clark fizzled at just 0.7% - enough to swing some races, but not enough here. What's funnier is the 1.7% won by Libertarian nominee Mark Strauss, who apparently did not get the memo that "Republican" Christopher Peters is a small l libertarian who used to be a big L Libertarian.

Loebsack and Hubbell had near-identical totals in Iowa City, Coralville, and North Liberty. They were in the 80s in many Iowa City precincts and Hubbell topped out at 90 in precinct 18.The congressman ran about 500 votes and 3% better in the rest of the county, allowing him to carry Jefferson East and Cedar Township where Hubbell didn't. Both lost the county's two historically Republican townships, Sharon and Washington.

Two places that turned around a bit were Oxford and Lone Tree. Both swung heavily to Trump, but Hubbell took 53 and 54%. Loebsack was in the upper 50s and low 60s, same as 2016, so the swing in the presidential may have been about that race and those two candidates.

Our only top tier legislative race was in Senate 39 where Democratic incumbent Kevin Kinney easily dispatched Republican Heather Hora. Kinney rolled up a 68-32 margin in fast growing North Liberty and Tiffin, and only narrowly lost Jefferson East, Sharon, and Washington. The negative TV ads that tried to paint Kinney as, in effect, a Johnson County liberal backfired due to Kinney's popularity and persona. It's hard to call an ex-deputy sheriff soft on crime.

Liberal hopes were high for Jodi Clemens, challenging Bobby Kaufmann in House 73. But Clemens only carried the Johnson County part of the district 54-46, not enough to overcome Kaufmann's margins in Cedar County and in heavily Republican Wilton in Muscatine County.

This district has tormented local Democrats for close to 20 years, since the 2001 map, as it looks so good on paper but is so loyal to Kaufmanns pere and fils.



Clemens carried the Johnson County early vote two to one but lost election day 55-45, and this mostly rural part of the county saw a higher share of election day voters than the city precincts did. She topped 60% in the small piece of Iowa City in the district and in Scott Township, a technically "rural" precinct that is trailer court dominated. But Clemens only narrowly carried Solon and lost the surrounding townships.

This race is an object lesson in how a purist stance on campaign finance reform functions as unilateral disarmament. Clemens had a firm policy of individual donations only (though she did accept money from county parties). That got her crossed off all the labor lists, the women's lists, the environmental lists, and, if she was ever in consideration, off the state party's target list.

The House 73 northeast corner shares Senate District 37 with Coralville and far west Iowa City. After a solid win in a serious primary, Zach Wahls had a near walkover in the general , defeating Libertarian Carl Krambeck 82-17% in Johnson County. Wahls was in the 70s in the rural areas and in the 80s in town.

Wahls spent much of the cycle helping other candidates. He's raised leadership-level money before even taking office and with Hubbell's loss he immediately becomes one of the most prominent new faces in the Capitol. The RAYGUN t-shirt is no doubt coming soon. The national publicity just keeps happening, too:



Libertarian Krambeck did almost as well as Joe Bolkcom's Republican challenger, Pat Wronkiewicz, in Senate 43 (most of Iowa City). Bolkcom crushed his first ever GOP opponent 79-20%. "Wronk" did best, ballpark of 30%, in the two dorm precincts, 3 and 5, where there was at least some awareness that he was a student, and in precinct 24 in Windsor Ridge which is just a smidge less blue than the rest of town. (Same is true of precinct 8 but that's in Wahls' district.) The Republican reached 35% in Hills, which counter-trends in local school elections but is reliably blue in general elections.

The one courthouse race was for the Board of Supervisors and the two Democrats beat the one Republican. Pat Heiden and Janelle Rettig took near-identical totals in Iowa City, Coralville, and North Liberty, with some variation by precinct and with Heiden slightly ahead.

I noticed a trend I have never seen before: the two Democrats rotated names as the law requires and there was a noticeable advantage across the Iowa City, Coralville, and North Liberty precincts for the person listed first. It was almost as if a lot of the new voters in this high turnout election were unaware they could vote for two. The flip-flopping roughly balanced out, though.

In the remainder of the county, though, Rettig ran 2000 votes behind Heiden and 352 votes behind Republican Phil Hemingway, who is running in the December 18 election to replace the late Kurt Friese. He'll likely face the winner (and maybe the loser?) of Tuesday night's Democratic nominating convention; the GOP meets Saturday to presumably nominate Hemingway.

No surprise that Phil's strongest showings were in Sharon and Washington in the southwest, and in Cedar Township in the red-trending northeast corner. He also narrowly carried Swisher and Sheyville, and won Cosgrove by one vote over Heiden. Rettig was not, however, shut out in the rurals, carrying Hills.

With only one Republican in a vote for two race, the undervote was high.  The average voter cast just 1.5 votes in the supervisor race. That means on average half the voters skipped an oval - probably many more when you take into account people who skipped the supes entirely. Early voters marked 1.58 ovals on average, while the more GOP election day electorate marked just 1.42 and undervoting was at its peak in the precincts Hemingway carried.

Well, that was fun. Let's do this again in four weeks.

No comments:

Post a Comment