Thoughts about the MN Independence Party

Our friends at Humphrey Institute released the first of many Independent polls we will see between now and Election Day.

As many bloggers are going to be vigorously discussing over the next few days this first poll has a whooping Margin of error of 5.8%. Considering that most polls have a margin of error under 4% that 5.8% is nearly 40% greater than what pollsters are typically looking for. I am sure there are a number of factors involved in getting such a large margin of error, for example it is early in the campaign season with no national or federal elections in the state. While all that is interesting (hey six quarters of statistics are worth something) it not really the point.

In looking at the details Mark Dayton is leading the DFL primary race, comfortable with 10 percentage point lead. I can hear the calls now "you just said the Margin of Error was a whopping 5.8%, isn't that 10 point lead within the margin of error?" Yes, it is, however you have to understand how margin of error works and levels of certainty. Long story short they further you deviate from the result the less likely it is accurate. So while Dayton's is leading at 38% the likelihood of his polling numbers being 32% is virtually nil (statistically speaking about .001 percent chance, I haven't dealt with a poll with such a large sampling error in a while so my math might be a bit rusty). The flip side of that is true also for Margaret Anderson Kelliher in getting to 32%. So there is a .00001 percent chance that Margaret Anderson Kelliher leads. If we look at the lead from the poll it is easy to say 6 to 14 percentage point lead, which is to say a comfortable lead.

Was Professor correct in saying the results were "a slap in the face" to the DFL, yes and no. First the reality is that fresh of the endorsement, just out of the gate and the fact that Margaret Anderson Kelliher is on the campaign trail and in the news it's really a slap in the face to the democratic party because a swing as much of seven points or bounce is usually anticipated, so things are not looking good for the establishment. However it is really not a nice thing to say, especially when you have such a large margin of error.

The real key however is Matt Entenza's extremely low polling number. This is the number people are going to have to watch. Under current circumstances Margaret Anderson Kelliher needs a three person race if she is expecting to compete the more Matt Entenza becomes a non-factor the more difficult it will be for Margaret Anderson Kelliher to win the primary. Too many people are ousting the sitting politicians right now, and how poorly she performed (there is no way to sugar coat that one) in office leading up to the end of the session she has painted a huge target on herself. The way to nullify that is to split the anti-establishment vote, yes I realize that neither Matt Entenza or Mark Dayton are unknown in politics, but they were sitting at the desk when Governor Pawlenty drove all over them. If Entenza is going to make a run he has to start drawing votes, it is where he draws the votes from that will be important, if he pulls votes from Dayton than it is good for Kelliher, if he pulls from the Undecided (which is 25%) than it is good for Dayton. Entenza has the money, does he decide to use it will be the question.

The second question is where does everyone sit in terms of the general election. Dayton leads Emmer and Horner, Emmer leads Kelliher and Horner. Again more bad signs for the DFL. Not necessarily the numbers but where the vote is coming from; in this case Emmer loses about 17% of the Republican vote to Horner (6 percent of the overall vote) but right now only 10% of the Democratic vote is going to Horner, that number is historically higher and taking historical precedent into play with Horner could end up with 15% of the total vote without spending a dime. If Horner runs a well crafted campaign he looking to pull in the 20% range easily, especially if it is Emmer v. Kelliher.

As a final point I found it either disheartening or concerning the percentage of people who identified themselves with both political parties, identical 42%. The first concern is that 42% is an awfully big number, considering a year ago those numbers were in the low to mid twenties. That an impressive doubling in a year. That in itself calls more question into this poll than any other factor, however it could be summed up on whether or not the individual was going to vote in the primary, the more you ally yourself with a party the more likely you will vote in the primary, so that could be it, but we will have to watch and see.