2014 Predictions, Part I

Byline: | Category: Government | Posted at: Thursday, 30 October 2014

This is a long post, so recharge your glass before you start.

The order of states might seem a bit odd, but when you read the second post in this series (with updated numbers) it might make more sense.

Let’s start with Mississippi.  With the largest black population of any state in the Union, you would think that Mississippi is ripe for a Democratic takeover.  You would think wrong.  The white vote in the Magnolia State is so red that even when nearly half of those Republican voters are left fuming after the dirty tricks that the incumbent Senator Thad Cochran used to win a primary runoff, the Democrat Travis Childers still will be swamped by a red tide.

West Virginia.  You know how ex-smokers become even more sadistically anti-smoking than anti-smokers who never smoked?  That’s what happens when a long-blue state turns red. Shelley Moore Capito will crush Natalie Tennant and the GOP will take the state house and all three US House seats by the end of the night.

South Dakota.  Have you ever locked yourself out of the house?  Once you realize what you’ve done, you panic and start scrambling around the house in the hope that some window somewhere might be left unsecured.  That’s what happened a couple weeks ago in South Dakota.  Democrats saw that all the doors were locked–doors, being those purplish states that they might be able to walk through.  South Dakota was a window, that by some magic set of circumstances might yield entry to their goal of not being locked out of the Senate’s leadership. Sorry, but the window is locked.

In Arkansas the name Pryor won’t be enough to save the seat.  Mark Pryor has been in a steady descent since Independence Day when he last led Tom Cotton by a single point.  Falling from 45.5 his RCP average of 1 July, Pryor is now  four points lower.  In retrospect, the result will have been obvious and smart political followers will wonder why they ever thought that Arkansas was really in play.

Kentucky was never quite as bright blue as some as its neighbors.  Hence it is not quite so red now.  Like Mississippi, Kentucky is far more conservative than is its senior senator. However, unlike Mississippi, it is not so red that it could bleed away Tea Party votes and still cruise to a GOP victory.  A better Democratic candidate might have been able to exploit a divide within the GOP to win a Kentucky Senate Seat. Allison Lundergan Grimes is not that candidate.  Her campaign has been so horrendously bad that were it not for Wendy Davis, Grimes would be a bigger national punch line.  Since the beginning of September, Grimes has flat-lined at 42 percent. Mitch McConnell’s support is anemically low for a likely Senate Majority Leader, but he will still pull out a win.  I’ll explain more in a later post, but Kentucky will likely give us our first real glimpse in how the rest of the night is going to unfold.  Grimes is going to lose.  But if she keeps it close–ie, under five points–the GOP may come away disappointed for the third election in a row.  If McConnell manages to come close to Rand Paul’s 2010 margin of 55-44, then even MSNBC correspondents will bemoan the President curse on his party before midnight.

Alaska is a tricky place to predict.  Polling is usually off by more than the national average.  But it is also off in the same direction every time.  Historically, polls favor Democrats by between 3 and 12 points. Mark Begich has stayed between 42 and 43 points since the beginning of October.  The only variation has been the result of a single, unreplicated poll by a highly-partisan analyst who released the results on Facebook because he has no website, included no cross-tabs, and also found that Dan Young is ahead by only a single point.  There are so many red-flags about this poll that it’s worth betting against.  That said, Nate Silver posits that there is a 15-point spread in the margin of error in Alaska, and who is to say that Mr. Moore’s poll isn’t the correct one and everyone else is off by 15 points.  Meanwhile in the Frontier State’s gubernatorial race, Independent (D) Bill Walker is leading Republican incumbent Sean Parnell by less than two points.  But . . . the spread between the four most recent polls that make up that prediction are Parnell +5, Parnell +3, Walker +6 and Walker +9.  In order to have a meaningful of average of polls, you must have a normally distributed sample of polls.  It’s only four data points, but these polls do not appear to be normally distributed.  Which means that the average Walker lead of 1.8 is bunk.   Two of them are right and the other two are very wrong. He’s either leading by five or trailing by just as much. History tells us that Parnell and Sullivan win their races by more than the average would predict. There will come a year when this history changes.  Just not this year in America’s ultimate libertarian state when America’s ultimate anti-libertarian President is underwater by a solid double-digit amount.

Louisiana.  Mary Landrieu has survived close races before.  But this year there has been only a single public poll since Independence Day showing her with over 46% in a race between her and Bill Cassidy.  And that one poll still had her losing by three points.  If the race were November 4th, she would lose a head-to-head race.  But the race isn’t November 4th.  The primary is.  The runoff is on December 6th.  The trick for her next week is to keep it close.  If she can exceed 45% in the initial contest she has a chance to get within the margin of fraud of a win.  (In Louisiana, the margin of fraud is larger than in most other states.)  If she is barely able to break 40% in November, then absent a cataclysmic change in the political environment, the only thing she gains is a one-month reprieve before she is voted out.  She won’t reach the amount she needs.

If you’re keeping score, along with Montana, this brings us to a Republican pickup of 5 Senate seats and a 50-50 split.  The loss of another Democratic seat would flip the upper chamber to the GOP.  Except that, there are two GOP seats that shouldn’t be in play, but are.

Kansas is our second example of the Locked Door Syndrome.  Only this window might be open. When one party completely dominates, it ultimately divides into two competing wings, one of which combines with the party out of power in a new coalition.  Governor Sam Brownback leads the tax-cutting government-slashing wing, while Kansas state senate Republicans are big business crony capitalists who do quite well under big government.  Senator Pat Roberts is representative of the latter wing.  These two wings cannot co-exist long in one party and thus, the logical coalition is for the party of big business to join with the party of big government under one banner.  In this sense, Kansas is on the cutting edge of what is to come nationwide.  The division is stark enough that in a more neutral political environment, it would be likely that neither wing would achieve a majority and both Brownback and Roberts would lose.  In fact, it was probably going to be the case that, had national Democrats not tried to force open a window in Kansas by supporting Greg Orman, that Brownback was going to lose this year.  The only thing now saving Brownback is that some small government Republicans who otherwise would have stayed home, will instead hold their noses to vote for Roberts in order to ensure that Republicans take the Senate. Talk about blowback.

Georgia is interesting.  You have to read Sean Trende’s analysis of eight decades of voting patterns in the Peach State in order to understand its idiosyncracies.  One of the most puzzling enigmas is its racial characterization of voters.  Trende points out that “unknown” is the fastest growing race in the state.  And nobody knows who they are.  So when you look at early vote counts and see that the zip codes with the highest proportion of early votes comes from areas with larger non-white populations, you have to remember that some of those non-whites are unknowns.  Strange.  Also strange is that, aside from Alaska, Georgia is the only state in the nation where the Democratic Senate candidate has a demonstrable trend of higher numbers since October 1st.  In fact, among the closest governor’s races in the nation, Pat Quinn in Illinois and Dannel Malloy are the only Democrats to show significant improvement over the last month. Jimmy Carter’s grandson is mired at the same 44% where he was at the end of September and where he peaked at the end of July.  So whatever is happening in the Perdue-Nunn race is not showing up in the other Georgia statewide race.

Let me digress about this point because it bears repeating:  There are only four Democrats in closely contested state-wide races who have shown an improved level of support during the month of October.  This is an enormously important point–especially this year.

All of the major poll aggregators focus on the head-to-head spread.  But not all spreads are the same.  The Democratic candidate with steady 2-point lead who is ahead 49-47 (Shaheen) is in a substantially better position going into the last few weeks of campaigning than is the Democratic candidate leading by two points but who has less than 45% support (Hagan and almost everybody else with a “D” after their name).  Barack Obama in 2012 struggled to hit the 50% mark.  But he was close enough to it that all he had to do was to demonize his opponent enough so that enough of the undecideds stayed home and 49% was enough to win when it came time to count those who actually showed up to vote.  You aren’t likely to succeed by employing that same demonization strategy when there are 14 percent undecideds in the last few weeks of the race. That’s like trying to run out the clock in football with a three-point lead and fifteen minutes left. All summer long we have seen a substantially greater number of undecideds than in years past.   By the end of October one of two things were going to happen:  undecideds were going to break for one party or the other, or they was going to stay home.  We now have a month’s worth of data showing no evidence of a break toward Democrats in 16 of the 20 most hotly contested statewide races in the country.  However, in over half of those races (11 out of 20) the GOP candidate is up over the last four weeks and/or the Democrat has noticeably declined.  (AR-Gov, AR-Sen, CO-SenGA-Gov, IA-SenKS-Gov, MA-GovMI-Gov, NC-SenNH-Sen, and NH-Gov).

(As an aside to my digression, the movement we’re seeing undercuts Nate Cohn’s argument that the polls are biased in the GOP’s favor.  Even if that were true, a consistent skew won’t disguise movement.  October’s polls clearly have shown movement in a Republican direction.)  However, even though we see movement now, doesn’t mean that it will continue (Nate Silver’s point). What it does mean is that in the next five days, absent a massive change, in all but three races Democrats have already persuaded all the voters that they are going to persuade, while Republicans have some potential upside still remaining. (This BTW, is how a “wave” forms causing almost every race to “tilt” in the same direction.)  In a later piece I want to explore this strange year and expand on what I discussed regarding Kansas and tell you what I think it portends for the future alignment of the parties.

End of digression.

When it comes right down to it, Michelle Nunn has to force a runoff if she expects to win.  If she does, January 6th is a long way away.  Since control of the Senate won’t be at stake by then, this could very well be the year when Georgia Democrats end their streak of losing runoff races.

Colorado.  Wow, did Mark Uterus, I mean, Udall, blow it.  And I predicted it two years ago:

” . . . pinning your party’s hopes on the most vocal advocates of a highly controversial social issue, when there is near universal agreement that other issues are more important, gives your party’s megaphone to those who are both extreme and irrelevant. Sandra Fluke is this year’s Terri Schiavo. For every already-Democrat she inspires to vote, she turns off at least one independent for the crime of insulting them by ignoring larger issues. Karl Rove’s plan to drive up Evangelical turnout in 2004, while it worked then, gave rise four years later to Mike Huckabee, who is perhaps the most demagogic and dangerous major presidential candidate to have run for office since William Jennings Bryan beclowned himself and his party in the late 19th century.”

Barack Obama won with the “war on woman” sword in 2012 and Mark Udall will die by that same sword this year.  The desperation NARAL radio ad preposterously postulates that in “Cory’s [Gardner’] world” there will be condom shortages and a ban on birth control. Again, if it weren’t for Wendy Davis,Udall and Grimes would be fighting for the “2014 I am not a witch award.”  Both sides should study the Colorado race, but especially Democrats, as Colorado was the most winnable race that they didn’t have to lose.  Oh, BTW, Udall flames out so spectacularly that he takes down Hickenlooper with him.

Iowa.  You do not say that lawyers make better legislators anywhere.  You especially do not say that lawyers make better legislators than do farmers when you’re trying to win a race in Iowa.  This election was over the moment that recorded remark hit the news.  B.C. (Before Catastrophe) Joni Ernst was five points down and never over 36% in the polls.  Since then Bruce Braley has never been higher than he was the day he made that stupid remark. Braley too is a strong contender for the “I am not a witch” award.  Aided by Braley’s collapse and Terry Branstad’s large margin in the governor’s race, three of the Hawkeye State’s four US House seats end the night red.

North Carolina and New Hampshire are worth discussing together.  Prognosticators have been lumping them that way because both female Democratic incumbents have clung to one or two-point leads even as their male GOP opponents have gained noticeably over the last month of the race.  But for reasons I discussed in my digression above, having the same spread does not mean that they are in the same place.  Kay Hagan is stuck at +/- of one point of 45% since Labor Day, which is exactly where she was on the Fourth of July.  Feminists could say that Hagan has hit a glass ceiling.  Jeanne Shaheen is bounded by a similar +/- one-point range.  The difference is that her range is +/ one point of 48%.  Another difference is that North Carolina is a shade more red than is New Hampshire.  This is especially true in midterm years.  Hagan’s supporters will point to early voting totals that show that compared with 2010, Democratic turnout is up two-percent, while Republicans are down six.  What they overlook, however, is that Republican Richard Burr beat his Democratic opponent by nearly 12%.  So Hagan’s apparent eight-point improvement over Elaine Marshall’s early numbers four years ago is good, but it is not good enough.  Kay Hagan and Jeanne Shaheen end the night on the opposite sides of  their races:  Shaheen wins hers while Hagan loses.

The what-might-have-been states:  Minnesota, New Mexico, and Virginia. All three could have been won with the right candidate and the right message.  Instead of finishing with 53 or 54 Senate seats (depending on Georgia’s January outcome), Republicans could have seized as many as 58 seats.  That they played defense while Democrats shot themselves in the foot made short-term sense, but was a lost major long-term opportunity to remake the Republican brand.

There’s still a few days left and thus time or a few more shoes to drop.  More to follow after a few more polls.

Share this post:

Comments are closed.