Nate Silver's first electoral prediction at the new 538.com was that Democrats are likely to lose the Senate. The Democratic Senatorial Campaign Committee, which exists solely to keep that from happening, quickly issued a memo hoping to undermine the charge, because Silver threatened to move perceptions of November from hard-but-doable to ugh-why-bother territory.
Campaigns spend a lot of time and energy keeping donors, activists, and supporters in a very precisely determined zone, which we'll call the "motivation" zone. All of those groups need to think that the candidate they're supporting is on the verge of losing the race, at all times, but must also think that winning is a possibility. We made the chart below to show what that looks like.
That yellow section is where campaigns want people, thinking that their candidate is just about to lose, just under 50 percent of support. Donors need to think that because they need to think that a bit more money can push the candidate over into the winning side. Activists need to think that so that they're inspired to make a few more calls to voters. And supporters need to think that because they need to feel like it's important to go to the polls on Election Day.
On either side of the motivation zone is dangerous territory for a campaign. To the right — in the space where the candidate will clearly win — is apathy. If donors and voters are pretty sure the candidate will win, they won't bother writing checks or taking off work early to go vote. To the left is despair, the fear that nothing they can do will make any difference, so why do anything? The percentages on our chart aren't set in stone, but they're a rough guide. If a candidate is polling at 40 percent against his opponent's 60 percent, it's hard to get motivated.
This is why Silver's prediction, a carefully tempered assessment of the moment that figured Republicans were likely to pick up six seats in November if the polls hold, was so dangerous for the DSCC. Silver earned a reputation in 2012 as a savant on poll numbers, calling each of the 50 states in the presidential election. As National Journal pointed out earlier this month, Democratic groups have taken to using Silver as a critical way of keeping people in that yellow motivation zone. Silver "cited North Carolina as 'the closest thing to the tipping-point state in the Senate battle,' and called Democratic Sen. Mary Landrieu's seat in Louisiana 'a true toss-up.'" Those assessments, from a deeply trusted voice on the left, opened checkbooks and phone lines.
Silver's new prediction puts the odds in Landrieu's race at 45/55 to her Republican opponent's benefit. But worse, Silver points out that the overall environment works against Democrats (as we've noted). If Silver says that the Democrats are likely to lose the Senate, Democrats believe that they're likely to lose the Senate.
Which is why, first thing Monday morning, the DSCC released that memo that did the previously unthinkable: it sought to undermine Silver's predictive abilities. In 2012, executive director Guy Cecil wrote, "Nate Silver predicted that Heidi Heitkamp had only an 8% chance of victory and Jon Tester had just a 34% chance. In 2010, he predicted that Majority Leader Reid had just a 16% chance and Michael Bennet had only a 34% chance in Colorado. All four are senators today…" In other words: Silver isn't always right.
It's safe to assume that had Silver said the Democrats would win the Senate by a healthy margin, Cecil might have isolated examples in which Silver's analysis — which is culled from existing polls from other firms — was over-optimistic. Too much fear is bad, as is too much optimism.
Cecil's point is accurate, though. There's a lot of time between now and November, and Silver's temperature-taking, particularly with a small number of polls, isn't written in stone. The thing is, that was never mentioned in those fundraising emails.
This article is from the archive of our partner The Wire.
We want to hear what you think about this article. Submit a letter to the editor or write to email@example.com.