1. Was turnout abnormally low in Alaska?
Shannyn Moore wonders why, in the first election when an Alaskan was on the ticket of a major party, turnout in Alaska crashed. She writes:
Four years ago, 313,592 out of 474,740 registered voters in Alaska participated in the election-a 66% turnout. Taking into account 49,000 outstanding ballots, on Tuesday 272,633 out of 495,731 registered Alaskans showed up at the polls; a turnout of 54.9%.
That certainly seems low. In the article linked above, Nate Silver suggests that the drop is turnout in Alaska may have been due to being in a later time zone. However, Hawaii the population of Hawaii lives, on average, two times zones later than Alaska, and turnout there seems to be down only 3% (see current and 2004 results). Why did turnout in Alaska drop so much more than Hawaii?
The answer is it didn't. As tietack notes in quick hits, the correct number of ballots yet to be counted in Alaska is, according to the state website, 74,000, not 47,000. In addition to the 47,000 absentee ballots, there are also 9,500 early votes uncounted, and 16,000 "questioned" ballots. Add these into the overall turnout figures, and that would put Alaska in exactly the same 3-4% turnout drop as Hawaii. This lends credence to Silver's thesis, and raises a lot of doubt about the notion that turnout was abnormally down in Alaska.
2. Why were the polls so wrong in Alaska?
According to current results, in all three of its federal campaigns this year, Alaska broke all records for polling average error. The most inaccurately polled elections this decade are now, in order, Alaska Senate, Alaska President and Alaska House. In fact, the Alaska House campaign has more than doubled the post-2002 previous record for polling error. Here is a comparison of the final Alaskan election polls, according to Republican site Real Clear Politics, to the current results in Alaska for all three federal campaigns:
President
Final polling average: McCain +14.5%
Current results: McCain +25.3%
Shift: Republican +10.8%
Senate
Final polling average: Begich +10.3%
Current result: Stevens +1.5%
Shift: Republican +11.8%
House
Final polling average: Berkowitz +8.5%
Current Results: Young +8%
Shift: Republican +16.5%
How unusual are these shifts? Consider that in all other 2004, 2006 and 2008 campaigns, no polling average was wrong by more than 7% (see 2004 and 2006 data here).
One explanation is that the polls were not wrong, and the remaining 74,000 votes will shift the campaign back toward Democrats and within the polling average margins of error. This might be true, but we won't know until all the votes are counted.
A second explanation is that the polls were all simply modeled wrong. This is possible, and worth looking into, but we won't be able to start doing that until we know what the final results are. Again, we need to wait for the remaining votes to be counted.
A third explanation is that the campaign was tampered with.
Shannyn Moore notes that there is a history of election oddities in Alaska:
Alaska has certainly had our share of election hanky panky. Check out this link to our 2004 election results. There are 40 districts in Alaska. The Anchorage area districts run from District 17-District 32. Scroll down to the bottom of the page and pick any district from 17-32. Pay particular attention to the 3rd column labeled % turnout. Hit the back arrow and select another district. There are more precincts with voter turnout over 100% than under 100%. In other words, many more people voted in Anchorage area precincts than there were registered voters. Clearly, this is not possible. In 2006, the Democrats filed a lawsuit against the Alaska Division of Elections to release public records needed to verify the 2004 election results. The Democrats ALSO sought to have the Alaska Division of Elections release the raw election data for the 2006 election. With that history, and the bizarre anomalies in polling and voting and reports from the field of ballots not being scanned on-site due to broken machines, could this election have been stolen?
That certainly isn't an encouraging past. Also, given that Ted Stevens was recently convicted of multiple penalties, and Sarah Palin was recently embroiled in multiple ethics scandals, the people running the state aren't the cleanest folks around. This is certainly possible, but I think we need to wait for the final results, and look into the other explanations, before jumping to this conclusion.
There are other explanations, but they fail to account for the shift across all three campaigns. For example, saying it was the Bradley effect doesn't explain the Senate or House campaigns, nor does it explain why this effect only happened in Alaska. Also, arguing that Steven's closed the gap at the end due to his last minute commercial buy doesn't explain it, since that only focused on the Senate campaign. Any explanation needs to account for all three campaigns, and only the three above explanations make that cut.
This is all very fishy, but I suppose we should wait to see what the final vote tallies are before jumping to any conclusions. |
