Wednesday, August 18, 2010

Primary Post Mortem

The body has been lying in the streets for a week and the decomposing has begun, undaunted I am going to rip apart the stinking carcass of the DFL primary and try to determine exactly what happened and why. Sorry for the imagery, that’s where the title took me.

87 Counties

I posted this last week, but I still think it’s the most illustrative reason of exactly why Mark Dayton won the DFL primary.

Joe pointed something out in the comments of that post, which is that you don’t get a sense of population density from this map and that’s true, but in a way it sort of misses the point. The reason Margaret Anderson Kelliher made it close was because she racked up big vote totals in the Twin Cities, Minneapolis in particular, but the reason Mark Dayton won is that he racked up small vote totals all across the state and that’s what this map helps to illustrate.

Kelliher won Hennepin County by 16,255 votes. She won Ramsey County by 5,527 votes. Dayton won Anoka County by 2,670 votes, Itasca County by 1,129 votes and 65 counties by vote margins of less than 1,000. Kelliher won a handful of counties with less than a thousand votes as well, but most of these were won with less than one hundred votes.

Dayton’s campaign was a take on Howard Dean’s 50 state strategy, you could call it the 87 county strategy.

The Decider

Here are the final numbers:

Mark Dayton won the election by 6,972 votes. One county I didn’t mention above, St. Louis County, he won by 7,332 votes. That’s the election right there.

Losing St. Louis County by that much, coupled with losing the vast majority of counties by smaller margins, made Kelliher’s task in the metro all but impossible. I said this prior to the election:

If MAK is going to win though, she has to do better in the Twin Cities than the polls have shown. If she can’t get to parity in her base areas she doesn’t really have a chance. In order for her to win her grassroots focused campaign has to turn out the Twin Cities and St. Louis County, otherwise it’s Dayton in the general.

Kelliher did what she needed to do in the Twin Cities; it was a failure to gain any traction in St. Louis County that cost her the primary.

The X-Factor

It’s hard to say how much of this was due to Yvonne Prettner Solon’s presence as Mark Dayton’s running mate, but clearly she had an impact on the results, much more so than Kelliher’s running mate, John Gunyu did.

It’s impossible to know now what would have been, but if Kelliher had picked Tom Rukavina as her running mate would he have been able to bring home some iron range votes for her? Would it have been enough matter? These are fun questions, but in the end, the choice was made and the election is now in the books and we’ll never know what could have been.

This may have been the critical decision of the campaign when all is said and done though, Margaret Anderson Kelliher picked someone who could help her govern, Mark Dayton picked someone who could help him win the primary. Who made the right choice?

Below is a breakdown of Hennepin County into Minneapolis and not Minneapolis.

What you can see looking at this is that Kelliher dominated in the city of Minneapolis, but in the rest of Hennepin County Dayton kept it close.

Now the same breakdown of St. Louis County.

In contrast Dayton won Duluth handily and won greater St. Louis County by even more. This is where a Rukavina selection could have helped Kelliher the most since his Virginia base is in the heart of St. Louis County. I don’t mean to keep harping on this point and I honestly don’t know if a Rukavina selection gets her over the hump, but many DFLers, including myself, were a little skeptical of the Gunyu selection and the outcome of this election just amplifies that skepticism, right or wrong.

The Third Wheel

The conventional wisdom prior to the election was that Matt Entenza would hurt Kelliher more than Dayton, but there really isn’t any evidence that happened. There isn’t a lot of correlation between the candidates when you compare their number of votes per county, meaning there isn’t any evidence that Entenza affected Kelleher’s numbers more than Dayton’s. In fact the highest correlation was between Dayton and Kelliher, which is what you would expect.

As I said the correlations are not strong, but it would appear that Matt Entenza actually took votes from Mark Dayton just a little bit more than from Margaret Anderson Kelliher, it’s not to a high enough degree to be significant though. Of course, that’s just looking at the final numbers, if Entenza hadn’t been in the race the entire texture would have been different so who knows what would have happened.


To evaluate turnout I compared this year to the average turnout in 2002 and 2006. As we all know by now DFL turnout for the primary election was much more robust than most had expected, approaching almost 15% of eligible voters when most, including SOS Mark Ritchie, thought 10% would be more likely. This was more than a 50% increase over the average turnout in 2002 and 2006.

The below table of the counties with the highest turnout increase contains two numbers, “% change” is the percentage that voter turnout increased in that county versus the average of the previous two gubernatorial primaries, “vs state” is that change minus the increase of the state as a whole. So the “vs state” column is the percentage of increase above and beyond the increase of the state as a whole.

Kelliher won Carver County, Entenza won Lyon County and Dayton won the rest. These counties do not have a lot of voters though, so let’s look at some counties that do.

Above are the six counties that had the most voters on primary day, seventh place, Olmsted County had almost 10,000 less voters than Washington County, so these are the six big counties. What you see here is that Kelliher did very well in the most populous places and the counties she won had a higher turnout than the rest of the state. But this trend was confined to the most populous counties; her campaign didn’t display any particular ability to turnout the less populous counties.

Looking at the turnout from all the counties there was little correlation between turnout and the performance of any of the candidates, so while in the most populous counties Kelliher won, she seems to have done so by turning out her base, this didn’t happen in greater Minnesota and it especially didn’t happen in St. Louis County.


Unlike my colleague, Two Putt Tommy, I don’t view this election as a systematic failure on the part of the DFL party. Mark Dayton is known statewide and liked by longtime DFLers, no matter who got the endorsement and what the DFL did, this was going to be a tough race to win. Add to that the fact that Dayton didn’t make any forced errors and had the money to run lots of ads made their task even more daunting and yet, they almost pulled it off.

No comments:

Post a Comment