Trying to get student votes out in Iowa City is like trying to make fetch happen.
This year it was - I use "was" in the past tense because if fetch is going to happen, it's going to happen before the ten day pre-registration deadline - it was not going to happen, at least not a lot.
At work yesterday, I had a long and fun chat with an out of state reporter who'd been tasked with checking in on early vote numbers in the big counties. And I had to explain, with lots of local color, that the Johnson County 2010 numbers, early and by ripple effect total, were skewed beyond recognition (we're hearing that word "skewed" a lot today) by the local 21
(tl;dr version of that: the actual measure was to keep people under 21 out of bars after 10 PM. But it played out as a Whose Town Is This Anyway culture war.)
In four solid weeks of petitioned campus voting sites in 2010, we saw 5269 voters: 1668 Democrats, 1151 Republicans, and 2374 no party or third party.
This year, without petitions, campus voting returned to a more normal schedule: three half-days at Burge and Hillcrest, and a week at the Old Capitol Mall. That netted 1577 votes: 837 Democrats, 170 Republicans, and 570 no party or third party. Not all of those were young people, either year, but most were.
The five days at Old Capitol Mall drew 1161 voters, compared to 2544 in the same five days in 2008.
Note the drop, sure, but also note the proportions. Republicans achieved near parity in 2010, but this year student voters look heavily Democratic.
Note also some unusual final numbers from 2010.
Total voters: 53,855You just don't see a 4% countywide under-vote at the top of the ballot. Actually, more like 7% in Iowa City proper because that's the only place the issue was on the ballot. And in the student precincts, that figure was even higher. Anecdotally, from being at those campus sites, the question of the day was "do I have to vote on everything or can I just vote for one thing if I want?"
Total votes for governor: 51,725
Under vote for governor: 2130
My best estimate - and that's one of my JOB jobs, making my best estimate - is that the bar issue drew in about 5000 people who would not have otherwise have voted, and about half of them voted ONLY on the bar issue. (They were offset slightly by 500ish straight ticket voters who skipped the referendum.) The other 2500 were drawn in by the bar issue but voted on at least some other items.
If you're thinking frat bros from Schaumburg, you're pretty close. This year, they're not voting at all. But in 2010 the people drawn in by the bar issue skewed more Republican than the typical student satellite voter.
That played some, not all, role in Chet Culver's slip from 68% in 2006 to 62% in 2010. (The rest was the national mood and, well, Chet.)
Johnson County's job in statewide races is to run up the score and make up for the northwest corner of the state, which is way redder than we are blue. Let's look at the People's Republic's percentages in the last few top of the ticket competitive statewide contests:
2002 Harkin 65%, Vilsack 63% - statewide wins(Set aside Vilsack 2002 because we had an unusual Green spike to 3%. At the time a vote for governor was the only way for them to keep party status, and the lieutenant governor candidate was from here. Iowa now has a third party "organization" status so people can register as Green or Libertarian, and that makes a vote for governor so much less important that the Greens aren't trying this year.)
2004 Kerry 64% - loss
2006 Culver 68% - win
2008 Obama 70% - win
2010 Culver 62% - loss
2012 Obama 67% - win
That puts the sweet spot to look for at 64 to 65%. If Bruce Braley is above that, he can pull this out. If not, it'll be really hard.
I expect that, or better, for Braley here. Joni Ernst's schtick plays well in about 90 counties. Johnson is one of the places it doesn't. Anecdotally: The non-political people I know know that I'm a political person. I don't push it on them, but sometimes they initiate conversation. And a lot of non-political people have initiated conversations with me about distate for Ernst and fear that she will win. Granted, that's in Johnson County, but that's on the street in Johnson County, not at a political event in Johnson County.
But it's not just the percentage, it's the raw number. Without the bar issue on the ballot, I expect overall turnout to slip. (Anecdote: we got almost none of the usual voter registration spike we normally see before general elections. Opening the mail I saw loads of absentee requests, very few registration forms.)
So factor out the 5000 extra student voters, and restore the student turnout to a more normal off-year level. (Actually, the campus satellite numbers were significantly better than 2006.) Factor in growth elsewhere in the county, and a couple months ago I predicted turnout of 51,903, and the first governor year general election with more votes early than election day. (We had more early than election day in both 2008 and 2012). And having watched it play out, I'm sticking with that number: 51,903. Approximately.
Note that's almost identical to the 51,725 votes for governor in 2010. So Johnson County overall turnout is likely to technically drop from 2010, but in terms of the partisan races it will be just about the same.