Tuesday, November 4, 2014

2014 Election Math

I've done very little modeling of the 2014 election. I did vote, but something about there not being a race for president sucks out half the appeal. Also my congressman will get re-elected in a landslide.

I did model a pair of Washington ballot Initiatives though: I-594 and I-591.

  • I-594 would expand the requirement of a background check to all gun purchases (with some exceptions)
  • I-591 would get rid of background checks (best I can tell - it's a vaguely worded bill) unless until there's a uniform national standard on background checks
My personal opinion is that we should pass 594 (even though it doesn't go nearly far enough), reject 591, then get serious about real gun control; having 32,000 gun deaths a year is a blight on our society.

But my strongest personal opinion is that I should do the best, most accurate, math I can. For example, I'm a Democrat but I think the most likely outcome nationally is the Republicans taking back the Senate and expanding their lead in the house. It's possible the Democrats pull off an upset (fingers crossed) but that's not where I'd be betting.

I'm not, however, super interested in the national race, so let's return to background checks in Washington. I took my WA specific prediction model from 2012 and re-purposed it for this year's gun vote.

Will I-591 (anti-background checks) pass?

Likely not.

I-591 has polled between 39% and 46% in 4 polls over the last 5 months. Four polls isn't a lot, but when they're consistent like that they should be taken very seriously. All in all I project I-591 has an 5% chance of passing. I made a graph of my predicted vote % of Yes votes for I-591 (42%) and confidence interval below.

Note the vertical line at the 50% mark - the mark needed for this initiative to pass. 

Will I-594 (pro-background checks) pass?

Very likely yes.

The polling on I-594 tells a similar story as 591, just on the other side of 50%. In the 4 polls done since July, I-594 has been over 60% in all polls but the most recent (it was at 59% in the poll released today). The No side for I-594 has been equally consistently bad, which adds up to good news for I-594. Really good news. I think this measure has around a 98% chance to pass. I project it passes with 61% of the vote. Graph below.

It's not impossible we see a different result tomorrow, I've played poker long enough to know that longshots come through. But it's likely we see 594 approved, and 591 rejected tomorrow. We might not even have to wait for all the mail-in ballots.

I do want to tackle one last question, since it was advanced to me by a pro-gun friend as an explanation for why he thought his side was losing:

Did the shooting at Marysville-Pilchuk High School affect either of these measures?

Probably not.

In the one poll since the shooting, support for both the pro-gun and anti-gun measures has declined slightly (and by almost the exact same amount for each measure). This is likely due to undecided voters breaking more toward No than Yes, rather than anything to do with the school shooting. The poll actually asked voters if the shooting impact their vote. 2/3 said no effect. The remainder were split close to evenly between more likely and less likely to vote.

To sum up: I-591 is probably losing, I-594 is probably winning, and it's unlikely either of those results are connected to the most recent school shooting.

If you haven't voted I encourage you to vote Yes on 594, No on 591, and Yes on reducing class sizes. If you live in Seattle I encourage you to also vote for Prop 1B (the city's universal preschool plan), against the monorail, and for the buses. 

No comments:

Post a Comment