Polls Are Not Predictions
Posted on | September 13, 2010 | 28 Comments
Paul Mirengoff of Powerline — a nice guy, whom I’ve met a few times — defends himself against Mark Levin’s criticism over the Delaware GOP Senate primary. Rather than to get into the middle of that, let me pull out a few phrases from Mirengoff’s post as illustrative of a more general dispute:
I then argued that, by supporting Christine O’Donnell in the Delaware Senate primary despite the likelihood (demonstrated by polls) that she would lose the general election, whereas her centrist opponent in the primary would likely win . . .
O’Donnell was routed when she ran for office in the past and currently is well behind the presumptive Democratic nominee in the polls. . . .
. . . polls, coupled with the state’s recent electoral history and O’Donnell’s prior poor showings, provide strong evidence that the O’Donnell is unlikely to win in November . . .
. . .my posts about the Delaware primary didn’t take a position adverse to O’Donnell until polls began to show her trailing the Democrat by double digits . . .
Other examples abound, but I think you see the point here: Mirengoff has allowed poll numbers to dominate his view of the Delaware primary and he is certainly not alone in doing so.
Yesterday afternoon, I returned to the Crystal City Sheraton after the 9/12 March and met an O’Donnell supporter who was almost frantic with distress. Someone on the Internet was trumpeting an outdated poll showing Christine trailing Mike Castle by 30 points. “We need to push back against this,” said the O’Donnell supporter.
Well, I had no time to blog at that point and thus couldn’t join the pushback effort. But a couple hours later, while I was working on my American Spectator column about the 9/12 March, the PPP poll went online:
It looks like there’s a real possibility of a major upset in the Delaware Senate primary on Tuesday night, with insurgent conservative Christine O’Donnell leading longtime Congressman and Governor Mike Castle 47-44. That 3 point lead is well within the poll’s margin of error. . . .
It’s clear that Castle’s popularity has taken a sharp turn in the wrong direction over the last month. . . .
Anti-O’Donnell readers must ponder this question: If the poll numbers in the GOP primary could shift so rapidly — if Castle could lose this much ground in the span of a month — why should we put faith in what polls tell us about the general election match-up against Democrat Chris Coons?
Paul Mirengoff has a lot of company among Republicans who have allowed this polls-as-predictions thinking to cloud their judgment about Delaware. This is an example of how the bandwagon mentality sometimes dominates political discourse. I’ve seen this type of argument used in policy discussions as well: “X percent of Americans believe global warming is a serious problem, therefore conservatives must propose policies to deal with global warming.”
Public opinion is a moving target, and the purpose of politics is not merely to react to public opinion, but to change public opinion. You will never win a fight if you walk away from the fight, and that’s what happens when you say to yourself, “Well, Delaware is a liberal state; there’s no point even trying to elect a real conservative there.”
Conservatives ought not succumb to the defeatist argument that we should retreat, ceding political terrain to our anatagonists, simply because they are strong and we are weak. When will we be stronger?
If Republicans can’t elect a conservative in Delaware this year, at what future point will such a victory be possible? In 2012? 2014? Ever?
Am I the only one who sees that this kind of helpless passivity toward the presumed liberalism of “Blue States” is a sure-fire formula for slow-motion political suicide?
Stand up and fight! WOLVERINES!
CHRISTINE O’DONNELL for U.S. SENATE

Pingback: Spin of the day « Da Techguy's Blog
Pingback: The Battle of Delaware | Parcbench
Pingback: Fishersville Mike: Maybe "unelectable" works this year
Pingback: Media wrong; Republicans wrong; O’Donnell, just ‘right’