Anyone following the results from the razor's-edge Arizona Senate race over the last week could hardly miss the regular line of complaints:
They're stealing the election! Martha McSally was ahead on election night! She won, dammit! Now it's being stolen! Where did all these votes come from days after the election? Looks mighty suspicious ...
A lot of people were saying such things over the last week; comment sections on websites reporting on the race were loaded with them.
These people were mostly betraying their lack of understanding of how elections in many states, Arizona recently among them, work.
If they'd lived in Oregon, they would have found nothing unusual at all. They would instead have found a familiar pattern.
It goes like this.
In states where many votes are cast outside polling places, part of the verification regimen - to make sure the votes are legitimate - involves checking the ballots one by one and often verifying that the signatures are as they should be.
Oregon is a good example of the way this works. In this state, voters have cast their ballots by mail - or by posting them in a ballot box that looks like a street-corner mail box - in the couple of weeks between receiving the ballot in the mail, and the election day deadline. Before sending it in, the voter has to sign an outside envelope. That signature is compared with one on file, as a security device. It sounds a little funky, but it has worked.
On election night in Oregon, results from many of the votes cast are released in a small flood shortly after 8 pm. These are the votes from earlier days in the return process, and early on election day. But the voters still pouring in will still be tabulated on Tuesday evening, and if there are enough of them the process of verification and counting may continue into and through Tuesday and beyond. And because a ballot submitted to a ballot box anywhere in the state by 8 p.m. on Tuesday may take a day or two to make its way from one end of the state to the other, the last of a county's votes may not come in until Thursday or Friday of election week.
Consider one more aspect to this process: Not all counties handle all this the same. Specifically, small counties generally are able to work through their ballots fairly quickly, and nearly all usually are done well before midnight in Oregon. On the other hand, the biggest counties - especially Multnomah (Portland) - take longer, and usually as a normal matter don't finish reporting until well into Wednesday.
The larger counties in Oregon, as is the case with most of the larger counties around the western states, tend Democratic. (Multnomah is an almost extreme example.) The smaller and more rural counties tend to be more Republican. What that means in practice, for people watching the vote come in on election week, is that Tuesday night generally shows a relatively strong Republican vote, but it edges down as the count goes on over the coming day or two. It's not especially unusual for a Republican to seem to be winning on Tuesday and losing on Wednesday. The 2008 Senate race between incumbent Republican Gordon Smith and Democratic challenger Jeff Merkley was a precise example of this; close election watchers around the state cautioned not to draw too many conclusions from Smith's modest lead on Tuesday night, because it would likely diminish the day after. As it did.
The same patterns are developing in many western states. Washington has seen them for some years; so have Colorado, Nevada, Utah ... and Arizona, all states where one or two large metro areas, which are trending Democratic, deliver many of their votes after many of the rural counties have already checked in.
And so it was this week in Arizona, albeit a little more drawn out than usual, in part because the race was so close.
But the pattern should not surprise anyone. It probably surprised no one in Oregon.