Tuesday, August 04, 2015

Pop The Vote vs. Mock The Vote

One of my jobs as Caucus Coordinator for the Johnson County Democrats is keeping in touch with the other team on rooms and arrangements and logistics.

So with that in mind I spent last evening at the Johnson County Republican Central Committee meeting. I was wearing my Democrat hat - literally - not my beret. But in between some discussion of caucus sites, I did find a few items worth noting.

The bulk of the meeting was a series of updates from seven presidential campaigns - namely Bush, Carson, Fiorina, Paul, Perry, Rubio, and Santorum.

But most interesting to me was some followup from the county fair. I'm still mourning Batman's fall into second place in the auditor's office mock election at the hands of the Hillary Clinton-Wonder Woman coalition, so to ease my pain I crunched another series of meaningless numbers.

(Carly Fiorina supporters declined to state whether they, too, had backed Wonder Woman.)

The office had competition from a second mock election one building over, as Republicans offered a Popcorn Poll: drop a kernel in the jar to vote. And they actually more than doubled the auditor's GOP turnout, 1099 to 467. (TOTAL auditor turnout was 1303, counting Democrats and kids who voted just for superheroes and skipped president.)

There were some differences, of course: the Republican poll may have drawn a more partisan crowd being located at the party booth instead of in the county government exhibit. The Republicans had a simpler voting process and I'm not sure what controls they had if any on multiple voting. The auditor's office had a scout's honor sign in sheet saying "I promise to vote only once." And seriously, anyone who wants to game a mock election at a county fair really needs to get a life.

The two polls produced the same two winners: Donald Trump in first and Scott Walker in second. And Chris Christie and George Pataki were at the bottom of both.

But in between, there were some interesting differences.

Below, we have listed first the Republican results, followed by the auditor's totals. Then I list the rank difference and percentage difference between the two polls for each candidate. Note that "Uncommitted" was not a Republican popcorn poll option, so I recalculated the percentages in the auditor's mock election to factor out Uncommitted.

Rank GOP Popcorn Poll Votes % Rank vs Auditor %
1 Donald Trump 189 17.2% same -2.8%
2 Scott Walker 131 11.9% same 0.4%
3 Bobby Jindal 125 11.4% +1 2.6%
4 Ben Carson 100 9.1% +1 1.0%
5 Mike Huckabee 90 8.2% +3 3.4%
6 Rand Paul 80 7.3% +3 2.4%
7 Ted Cruz 71 6.5% -1 -0.2%
8 Carly Fiorina 60 5.5% -1 0.6%
9 Jeb Bush 54 4.9% -6 -5.5%
10 John Kasich 43 3.9% +2 0.7%
11 Rick Perry 40 3.6% +2 0.4%
12 Marco Rubio 37 3.4% -2 -0.8%
13 Rick Santorum 27 2.5% +1 -0.3%
14 Lindsey Graham 25 2.3% -3 -1.2%
15 George Pataki 14 1.3% -1 0.6%
16 Chris Christie 13 1.2% +1 -1.4%

Total GOP Vote 1099

Rank Auditor's Mock Election Votes % (of candidate votes only) Rank vs GOP %
1 Donald Trump 87 20.0% same 2.8%
2 Scott Walker 50 11.5% same -0.4%
3 Jeb Bush 45 10.4% +6 5.5%
4 Bobby Jindal 38 8.8% -1 -2.6%
5 Ben Carson 35 8.1% -1 -1.0%
6 Ted Cruz 29 6.7% +1 0.2%
7 Carly Fiorina 21 4.8% +1 -0.6%
8 Mike Huckabee 21 4.8% -3 -3.4%
9 Rand Paul 21 4.8% -3 -2.4%
10 Marco Rubio 18 4.1% +2 0.8%
11 Lindsey Graham 15 3.5% +3 1.2%
12 John Kasich 14 3.2% -2 -0.7%
13 Rick Perry 14 3.2% -2 -0.4%
14 Rick Santorum 12 2.8% -1 0.3%
15 Chris Christie 11 2.5% +1 -0.6%
16 George Pataki 3 0.7% -1 1.4%


Uncommitted Rep. 33 7.1% of total auditor vote

Total Auditor Vote 467

From this it looks like Trump is noticeably weaker in the presumably more GOP-identified Popcorn Poll - though still strong enough for first. Huckabee, Jindal, and Paul are running noticably stronger...

But I'm burying the lede here. Jeb Bush is getting less than half the support at the Republican Party booth than he did at the auditor's booth. In fact, I almost didn't post this because I though I messed up my notes and left off a "1" in front of Bush's 54 votes - but that would have put him ahead of Walker and we specifically discussed that the two polls had the same two leaders.

None of these polls are scientific of course. Everyone knows that the only vote that REALLY matters is the Hamburg Inn Coffee Bean Caucus.

Friday, July 31, 2015

Mock Election: Able To Leap To Conclusions In A Single Bound

I'll admit to taking the Johnson County Fair mock election way too seriously. It's an occupational hazard, since I've been involved with it since its launch in 1999.

The shortcomings are obvious: self-selection, small children voting, the non-random sample of the fair. But over the years its predictive value, while nothing to base a campaign plan on, has been not half bad.

I spent more time at the fair than usual this year and ballot secrecy is a little loose - since the real job is to show kids how to mark a ballot and put it in the machine, you see a lot. And a lot of ballots seemed to be split, with the child marking a superhero choice, and the parent guiding a presidential pick.

One thing that is clear: fun questions help turnout. Our top three years have been the three years of the Weipert Administration, when we started doing the fun questions.

2015 participation TRIPLED from the last caucus year, 2011, which saw an all-time low 408 voters. That year, the caucus vote was accompanied by no less than NINE issue questions. (Sample: "Which issues would cause you to vote against a candidate if he/she disagreed with your position? Vote for no more than EIGHT." You could see kids and parents recoiling in confusion.)

Enough background. Let's dig into the results.

Total Voters 1303
Wonder Woman 270 22.0%
Batman 194 15.8%
Captain America 163 13.3%
Spiderman 124 10.1%
Flash 116 9.4%
Superman 105 8.5%
Iron Man 104 8.5%
Batgirl  90 7.3%
Supergirl 63 5.1%
Democratic  Presidential Candidates
Hillary Clinton 289 46.5%
Bernie Sanders 168 27.0%
Uncommitted Dem. 72 11.6%
Martin O'Malley 39 6.3%
Lincoln Chafee 33 5.3%
Jim Webb 21 3.4%
Republican  Presidential Candidates
Donald Trump 87 18.6%
Scott Walker 50 10.7%
Jeb Bush 45 9.6%
Bobby Jindal 38 8.1%
Ben Carson 35 7.5%
Uncommitted Rep. 33 7.1%
Ted Cruz 29 6.2%
Carly Fiorina 21 4.5%
Mike Huckabee 21 4.5%
Rand Paul 21 4.5%
Marco Rubio 18 3.9%
Lindsey Graham 15 3.2%
John Kasich 14 3.0%
Rick Perry 14 3.0%
Rick Santorum 12 2.6%
Chris Christie 11 2.4%
George Pataki 3 0.6%

My aggressive Bat-campaigning was only enough to help the Caped Crusader to a stong second place finish. I had a good, four point, issue based argument: Bat Cave, Batmobile, Robin thrown in for free, and best theme song.

But the BatVote was split with BatGIRL, as there was definitely a trend of young female voters supporting female superheroes. Together, the BatVote would have been in first place.

Instead, Wonder Woman took an early lead she never relinquished, and with no Wonder Man to split the vote, the finished first.

There may have been another factor: a high correlation with the Democratic presidential race.
Many, many blog posts have been written in many, many basements about comic books and gender, and I can scarcely do the genre justice. But Hillary/Wonder Woman was definitely a thing.

That Clinton percentage below 50 is a bit misleading, because the Uncommitted vote was largely a kid vote, and party ballot choices (to represent the caucuses, we had separate Democratic and Republican ballots, with the same superheroes on each) were often based on preferences for elephants over donkeys.

But there is, legitimately, a good sized chunk of uncommitted, or rather undecided, Democratic vote, and 12% is not far off the mark, especially if you factor in the Waiting For Biden folks.

So did Sanders disappoint? His percentage may not be representative of Johnson County, but the county fair looks a lot more like Iowa than most of the county does.

Lincoln Chafee did a bit better than I'd expected, but frankly the first line on the alphabetical ballot helped.

The talk of the fair, though, other than superhero related smack talk, was the Trump lead. No one seemed focused on the tiny percentage; the first place rank was the word. And that's going to keep going in the larger world as well.

Yet the first place rank clearly only exists because of how badly the rest of the field is splintered. The "establishment" candidates Walker and Bush, combined, beat Trump, and if you throw in some votes from a Kasich or a Rubio or a Perry you start to build a lead. The social conservative cluster of Carson plus Cruz plus Huckabee plus Santorum beats Trump.

And these percentages aren't really all that far off any given poll of the week. The pattern of preferences among the non-Trump candidates might as well be a group of children randomly marking ballots at a county fair.

If one of the Kiddie Table Debate candidates wants to be a hero to 80% of the Republican Party, they'll drop out and endorse someone in the top four or five. Could be enough for a cabinet seat.

Also worth noting from what I overheard: People have noticed Bobby Jindal on TV.

But for me, the most disappointing performance of the fair was the poor showing by definitive superhero Superman at sixth place, just one vote over Iron Man.