Another quirk of Oregon’s mail-in voting system is the way exit polls are being conducted. Normally, polling firms set up stations at polling places across the state and selectively interview voters as they emerge. The difference in Oregon is that polling firms are calling voters at home.
It’s hard to pinpoint what difference this makes in the numbers, especially since exit polls are unreliable to begin with. But consider these factors:
Young people are less likely to own a home phone, which means they could be underrepresented in polls. Few people actually fill out the contact information section of their mail-in ballots. Normal exit polls are self-selective to an extent, since you’re going to get people who are less shy or hurried; your average voter is more likely to pick up the phone than talk to a stranger in person. Normal exit polls catch voters fresh out of the voting booth; phone surveys rely on many voters who cast their ballots weeks before.
Oregon relied on phone surveys in 2004 and the results weren’t disputed. But they were also less scrutinized, since John Kerry had the nomination locked up by then. When looking at tonight’s demographic breakdowns, it’s worth keeping methodology in mind.