2. The national polls are right and the state polls are wrong. There's a tension between polls being done at the state level -- mainly in swing
states -- and the national polls. The former show Obama winning the electoral college, the latter show the popular vote being a tossup or maybe even going to
Romney. Now, in principle it's possible that this discrepancy isn't a contradiction. Maybe Romney is "wasting" more votes with huge majorities in red
states than Obama is "wasting" in less lopsided blue states -- so Obama has more votes "left over" for swing states. That's the way Obama could clearly win in
the electoral college while roughly tying in, or even losing in, the popular vote. But Nate Silver of the New York Times recently provided reason to think
this isn't what's happening. He says it looks more like either the state polls or the national polls are just plain wrong. [Update, 11/4, 8:50 p.m.: Obama supporters will be happy to hear that new data is making Silver reconsider. This evening he tweeted: "Everything gelling a bit: 1) US polls catching up to state polls 2) bad
polls for BHO in noncompetitive states help explain remaining spread."]
Silver thinks the state polls are more likely to be right than the national polls. But what if it's the other way around? What if the state polls are
systematically wrong? Maybe owing to some methodological flaw that's more common among low-budget state-level pollsters? And maybe the impact of this flaw
has been magnified by recent changes in demographics or in personal communications technology or whatever? Silver enlists a chart to
show that national polls have often been more biased (i.e., systematically overstated either Democratic or Republican performance)
than state polls. But the chart also shows that the last time national polls were markedly more biased than state polls was back in 2000 -- and in the most recent
election national polls were ever-so-slightly less biased than state polls. Maybe this is a trend, and state polls are getting less and less
reliable? I'm not saying this is the case. But as sources of anxiety go, it's not a bad one. Indeed, Silver wrote in a Saturday morning post that if his prediction of an Obama victory is wrong -- an outcome he gives a 16 percent probability -- it will almost certainly be because the state polls are wrong.
3. The Bradley Effect (named after former African-American mayor of Los Angeles Tom Bradley). Are some people reluctant to admit to pollsters
that they're voting against a black candidate? If so, then both state and national polls could be overstating Obama's popular support. The conventional
wisdom is that the Bradley Effect didn't amount to much in 2008, but Doug Wilder, former black governor of Virginia, thinks it could be a factor in 2012.
4. Benghazi. Actually, even a worrier like me can't find much to worry about here. The conspiracy theory suggested by Fox
News -- that someone somewhere in the federal government ordered American security forces not to come to the rescue of Ambassador Chris Stevens and
others in the Benghazi consulate -- hasn't held up. Romneyites have flogged the hell out of
the Benghazi story, but I think that what little mileage they can get out of it they've already gotten.
