On September 3 the Democratic district attorney from Shawnee County, Kansas, Chad Taylor, officially announced his withdrawal from the election for U.S. Senate. This caused quite a stir among political observers, who then had to contend with the possibility of a unified challenge to incumbent Sen. Pat Roberts by independent Greg Orman. Drama ensued as Secretary of State Kris Kobach refused to remove Taylor's name from the ballot and there was a legal battle over it and now the Democrats are--kinda?--still on the ballot, but without Taylor's name next to the D. The Kansas Supreme Court hasn't made quite clear whether the Kansas Democratic Party has to nominate a replacement candidate.
It's that Senate race that's received a lot of attention in the last three weeks, but what's been neglected is that Alaska Democrats beat Kansas Democrats to the punch by one day. Gov. Sean Parnell was supposed to be a lock for re-election: after ascending to the office following Gov. Sarah Palin's resignation in 2009, Parnell was first elected to the position in 2010 by a margin of over 20 points, with the largest ever share of the vote for any Alaska gubernatorial candidate in history (59%). No one in the spring and even right up until the primary on August 19 questioned what seemed to be a fact of Alaskan political life.
Of course, it's because no one counted on two earth-shakers.
Earthquake number 1 was the release of polls including longtime Republican-turned independent and former Valdez mayor Bill Walker. Walker had considered running against Gov. Parnell in the Republican primary but stayed out. In retrospect it looks like a pretty wise move, as Gov. Parnell won over 75% of the Republican vote, and a poor primary performance could very well have damaged Walker's chances. A few days after the August 19 Republican and Democratic/Libertarian/Independence primaries (and the independent filing deadline), Hays Research released a poll it conducted for the Walker campaign: Walker led Gov. Parnell by 3 points, 43-40.
There are a few reasons we didn't see this poll as an earthquake at the time (and maybe it shouldn't be considered an earthquake, since usually you can feel earthquakes happening in real-time). First, we shouldn't be quick to consider that a lead for Walker--it was well within the margin of error. Second, polling in Alaska is usually sparse, and what few polls do get released are erratic and inconsistent with each other. Third, the poll was conducted for the Walker campaign, and partisan polls should be taken with a grain of salt.
But that grain of salt shouldn't be because the individual poll is going to be tweaked in whichever way favors the candidate for whom it's being conducted. Rather, it's mostly because internal polls are released selectively--all the polls are used as data by the campaign, but only the good-looking ones get released to the press. Think of it this way--a challenger to an incumbent might commission a pollster to conduct five surveys in a month. The first four polls show the challenger down by 8, 9, 7, and 10 points, but due to some quirks of sampling error the fifth one shows him down by 2--within the margin of error of the poll. Only that fifth poll gets released and the only people who even know about the first four are the campaign, the polling firm, and maybe the people surveyed, if they remember it. This is why, although individual polls by a campaign may not be biased, the ones that do get released sure do skew the average toward the campaign's candidate.
I say all this to explain why the Hays poll still mattered, even though it was an internal poll. Sure, it was an internal poll, and even though it showed Walker up by 3 points, it could well have been only the best of many polls conducted for the campaign--and the others could well have shown Walker down. But even if we do consider the possibility that the poll's been selectively released, the poll still suggested that Gov. Parnell was not as safe has he once seemed.
So there was that. But then earthquake number 2 was a real earthquake. On September 2--a day before Chad Taylor announced his withdrawal from the Senate race in Kansas--the former mayor of Juneau, Alaska, Byron Mallott, announced that he would no longer be running as the Democratic candidate for Governor of Alaska. Instead, he would be running as the "Alaska First Unity" candidate for Lieutenant Governor of Alaska. His running mate, running for governor under the Alaska First Unity banner? A former mayor of Valdez named Bill Walker.
 |
| The Alaska First Unity ticket--Walker (right) and Mallott. |
That's where we've come in both Alaska and Kansas. A year ago both offices--governor in Alaska and Senate in Kansas--were Republican locks; the two would be the first to be called by the media on election night; and if they were lucky they'd both get a sentence in the papers the following morning. Two things have really sold us on the possibility of a Walker-Mallott victory in Alaska, though. The first is just a qualitative assessment of the ticket. Ideally, a Walker-Mallott ticket would bring over a not-insignificant number of Democrats to Walker, and Alaska's independent streak is well documented. PPP crosstabs corroborate: independents break for Walker 44-33, and the only reason Parnell is still this close is because of his huge advantage among Republicans, who make up the overwhelming majority of registered party members in Alaska.
The second thing (which is probably more convincing to most people) is that a better-respected firm, Public Policy Polling, has finally polled Alaska and asked about a Parnell/Sullivan vs. Walker/Mallott matchup. And the result--a 42-40 lead for Walker--is not as lopsided as another Hays Research poll conducted for the AFL-CIO (a decidedly Democratic-leaning interest group in heavily unionized Alaska) would suggest (37-30, with a full third of voters undecided). But it does suggest that, as always, our knowledge of the nation's political geography blurs just outside the Last Frontier. TOSS-UP
No comments:
Post a Comment