NJ-GOV: Garbage In, Poll Out

Suffolk University’s Political Research Center released its latest survey in the New Jersey Governor’s race today, showing Governor Jon Corzine holding a commanding 9-point lead over Republican challenger Christopher Christie, 42-33 percent. Independent Chris Daggett received 7 percent in the poll.

The result caused a minor stir on the Internet. A very minor stir, as even a cursory examination of the internals of this poll shows that it is deeply flawed, internally contradictory, and not at all reflective of the state of the race in the Garden State.

All polls produce outliers, and polling outfits occasionally get a bad sample. But a result this bad this late should never have been released, if Suffolk University wanted to maintain any semblance of credibility. The fact that it was released leads one to believe that the purpose of the survey was not to predict the outcome in the governor’s race, but to influence it.

Let’s start with the sample.  Suffolk surveyed 400 New Jersey residents, 360 of whom said they were likely to vote in the governor’s race.  The small sample size leads to a relatively high margin of error of 5.5 percentage points.  Worse, however, is the make-up of the sample.  Just 83 total respondents out of 400 (21%) identified themselves as Republicans versus 135 respondents (34%) that said they were Democrats.  Critically, as we’ll see later, independents made up the lion’s share of the respondents with 46 percent.

There are more independents in New Jersey’s electorate than affiliates of the two major parties.  But the 13-point split between Democrats and Republicans is far of the mark.  Democrats outnumber Republicans by a few hundred thousand.  But the split in New Jersey is much closer to 32-28-40 than this poll’s massively skewed result.  It’s no wonder, then, that an overly-Democratic sample would produce an overly-Democratic result.  But digging further into the cross tabs reveals a sample that should be full of Christie voters.

Two-thirds say that the state is off on the wrong track, and 32 percent list either taxes in general or property taxes specifically as their most important issue.  Another 26 percent list jobs or the economy as the number one issue.  Still another 15 percent list either corruption or the budget as their biggest concern.  All of these are themes that the Christie campaign has been using to hammer away at Corzine for almost two full months.  Taken together, just under three-fourths (74%) of respondents say that they are voting on the basis of one of Chrisite’s issues, yet the poll directors think this jibes with a Corzine win.

How can this be?  This sample is very confused about what it is going to do next Tuesday.  Seventy-two percent say their mind is made up about how they will vote, versus 24 percent who say they could change their mind.  But 55 percent say that they don’t know how they would vote if the election were held today.  Sixty-two percent have a favorable view of President Obama, but a third say that Obama’s endorsement of Corzine makes them less likely to vote for him.  Among the independents, Christie is winning by 5 points, while 55 percent remain undecided.  There may be a story in these results, but it is certainly not one that gives any high degree of confidence that Corzine is winning this race, let alone blowing Christie away.

That begs the question, why would Suffolk release this poll with all of its obvious problems?  It’s difficult to imagine that the polling director looked at these scatter shot results and thought this was an accurate appraisal of the state of the race.  A polling director concerned about the reputation of his center might have thought twice about releasing this data this close to an election, when there is a high degree that people will remember just how wrong it is likely to be.

On the ground in New Jersey, Daggett’s poor performance in the final two debates has stopped his campaign dead in the water.  Republicans are quietly encouraged that his campaign peaked too soon, and they expect a good percentage of his voters to come over to Christie, or stay home.

Meanwhile, Corzine’s “momentum” – which was always really a result of Daggett’s rise, not a Corzine comeback – has been completely stopped by the corruption conviction of former Bergen County Democratic Party chair Joseph Ferrirero.  Bergen is the wealthiest and most populous county in the state, and Christie has been hitting it hard, trying to reclaim the once Republican stronghold.  Ferriero’s conviction plays right into the Christie campaign’s hands.

Make no mistake, next week’s election will be close.  Statewide Republicans don’t win very often in New Jersey, and it is something of a surprise that Christie is this close this late.  But there are no moral victories.  Having been in the lead most of the summer, Republicans want a win in New Jersey to go along with an expected victory in Virginia’s governor’s race.

To get it, Republicans will need a high turnout among their base, a lower-than-normal turnout among Democrats, and late-deciders and independents to break Christie’s way.  All of this is well within reach, and much of it is likely given Corzine’s position in reputable polls.  Suffolk latest effort scratches it off that list.