Recent Posts

Changing how the DFL endorses gubernatorial candidates

by Eric Ferguson on August 9, 2018 · 4 comments

The convention hall as seen from visitor and alternate seating.

The convention hall as seen from visitor and alternate seating.

I’m writing this prior to the August 14th primary, and you might wonder why I’m writing this now, in the heat of the primary campaign when DFLer-on-DFLer campaigning is at it’s thickest (though just how negative depends a great deal on which specific race is the subject). There are two answers: one, passions about whether the endorsements made this cycle and regarding the process actually spikes right after the primary; two, this is in my mind because of recent conversations with DFLers in the last week or two with a couple connected points: the DFL has not had an endorsed non-incumbent win the gubernatorial election since Wendell Anderson, and a consensus is forming that Erin Murphy is toast. That latter opinion is based on a couple polls that are at least two weeks old by now and have other issues — not to go into a tangent, but I refer for example to the huge number of undecideds and the polling of registered voters instead of likely voters — so that opinion is premature. Not wrong, but premature, and many Murphy supporters seem in denial about the big trouble the Murphy campaign is in. By no means all, but plenty haven’t come to terms with Murphy’s situation yet.

Erin Murphy is the DFL endorsee, and if she doesn’t pull it out, we’re going to have our usual, and usually heated, discussions/arguments about how we endorse and who we endorse and whether to endorse. So I suppose I’m getting a jump on that.


When our non-incumbent gubernatorial endorsees keep losing, that begs several questions:

  • Are we talking about losing the general election, or losing the primary, and do those raise different issues?
  • Is the process OK and the choices wrong? And it’s possible to have a bad process that still gets us to good choices, just by luck rather than deliberation.
  • Is the process why we lose, or does it just not adequately address the problem?
  • We don’t have this same problem with other statewide offices, so is there something different about governor, or just statistical noise that loses are focused on governor?
  • How many instances are we actually talking about, and are they enough to say there’s a systemic problem?

Let’s start with answering the part about how many instances we’re talking about, because that’s not likely to be a matter of opinion. We’re talking about 13 gubernatorial elections, though not all include a non-incumbent endorsee. Let’s take them one at a time:

  • 1970: This is when Wendell Anderson became the last non-incumbent endorsee to win, so that’s why we’re starting here.
  • 1974: The DFL endorsed incumbent Wendell Anderson, who won reelection.
  • 1978: This is the year of the “Minnesota Massacre”. To keep the story short, because it’s something of a tangent from our main concern here, Anderson decided to appoint himself to a vacant US Senate seat. He couldn’t as governor, so he resigned, Lt. Governor Rudy Perpich stepped up to governor, and he appointed Anderson. Both won DFL endorsement for election in their new seats, but the general election voters punished them and many other DFL candidates. The DFL lost the governor’s mansion and both US Senate seats, and I think the state house flipped too. Perpich was only briefly an incumbent, sort of a partial incumbent, but it was such a weird circumstance that it doesn’t tell us anything about the endorsement.
  • 1982: The DFL endorsed non-incumbent Warren Spannaus, who lost the primary to former governor Rudy Perpich. I didn’t live in Minnesota then and have no recollection of that campaign. My admittedly un-thorough research turned up little about that race, which makes this speculation less informed than I like, so with that caveat, Spannaus might have lost, sadly, over the gun control issue he was heavily identified with. I also wonder if Perpich had the benefit of incumbency despite being a former, not incumbent, governor. Incumbents who lose the endorsement to a challenger almost always run in the primary, and they almost always win — something many DFL convention delegates seem to have forgotten the last couple years. Perpich won both the primary and general by large margins.
  • 1986: The DFL endorsed incumbent Rudy Perpich, who won reelection.
  • 1990: The DFL endorsed incumbent Rudy Perpich, who lost another weird general election. Again, this is a tangent so to keep it short, Perpich looked set to beat the scandal-plagued Jon Grunseth, when a scandal was so bad that Grunseth dropped off the ballot shortly before the election, and got replaced by the 2nd place candidate in the MNGOP primary, Arne Carlson, who won. I thought at the time that the voters were tired of Perpich after 10 years in office, and Grunseth was the only MNGOP candidate he could have beaten. Then his luck ran out.
  • 1994: The DFL endorsed non-incumbent John Marty, who won the primary but lost to Gov. Carlson. There was/is widespread opinion Marty was a bad candidate, but this was also a red wave election (though we didn’t refer to “red” and “blue” back then), and Carlson was a popular incumbent for the right party, so no DFL candidate could have won. Maybe another candidate could have gotten closer — but maybe not.
  • 1998: The DFL endorsed Mike Freeman, who lost the primary to Skip Humphrey. The reason for the loss should be obvious: name recognition. Besides being the four-term AG, Humphrey was the son of Hubert Humphrey. Humphrey was such a poor candidate that he managed to come in third in a decent year for Democrats. This was one time when the convention and primary disagreed, and the convention was right.
  • 2002: The DFL endorsed Roger Moe, who won the primary but lost the general to Tim Pawlenty. I thought Moe was a mediocre candidate, but a far bigger element was former DFL congressman Tim Penny ran with the Independence Party and split the left-leaning vote. We’ll never know for sure Moe would have won a two-way race, but more likely than not he would have.
  • 2006: The DFL endorsed Mike Hatch, who won the primary but lost the general by a whisker. Again, the IP candidate ran like a liberal and took only a small percentage this time, but it was still 5% when Pawlenty won reelection by 1%. Hatch had the infamous E-85 gaffe by his running mate, and the resulting display of his famous bad temper DFLers were concerned might cost him the election, and there was a big negative ad campaign against him in the last days that he didn’t have time to respond to, so call it over-determined why he lost. He looked formidable enough that another endorsee would likely have lost to Hatch in the primary.
  • 2010: The DFL endorsed Margaret Anderson, who lost the primary to Mark Dayton.She was outspent 3-1 by Dayton, though Matt Entenza outspent her 5-1 to come in third. Dayton skipped the endorsement altogether, whereas other successful challengers to the endorsee had gone for the endorsement and lost. I thought, at the time, still do, that she was at such a disadvantage in money and name recognition, and so much less effective at messaging than Dayton, that the endorsement was the only reason she got close. Would a different endorsee have been able to beat Dayton? Can’t tell. Dayton had as good a chance as anyone of getting the endorsement. I know that opinion is not universally shared, but I saw how much support Dayton had among caucus goers and convention delegates, yet he refused nomination. What I think gets ignored is very few candidates have the combination of name recognition, money, and message discipline Dayton had, so the idea just anyone at all could blow off the endorsement and just go to the primary is doubtful. What he showed by winning by a whisker with so many advantages is the endorsement does have potency, even though just how much is an open question.
  • 2014: The DFL endorsed incumbent Mark Dayton, who was reelected despite the red wave.
  • 2018: The DFL endorsed Erin Murphy. If she loses as looks likely, I have my theories of what’s gone wrong, but too soon to say, and those theories will look silly if she pulls it out.

So having gone through the elections from Wendell Anderson to this year, the years when a non-incumbent endorsee lost the primary are 1982, 1998, 2010, and potentially this year. That’s three or four times, depending on Aug. 14th. The non-incumbent endorsee lost the general election in 1994, 2002 and 2006. Those are sample sizes of 3 or 4. Without distinguishing between primaries and general elections, that’s 6 or 7 times. The DFL has done much better at every other election, including statewide office consistently, while losing governor. Are we picking better candidates in other races? Maybe, but the defeated endorsees were an incumbent state AG, an incumbent state senator, an incumbent Hennepin county attorney, an incumbent state senate majority leader, another incumbent AG, an incumbent state house speaker, and this year an incumbent state rep. who had been assistant minority leader.
In other words, the gubernatorial endorsees were people who looked like qualified and proven candidates, even to the point of having won statewide office. That leads me to think the issue isn’t the quality of candidates, at least consistently. We’re not endorsing just any backbencher legislator or rich businessman who wants to start at the top. Were they always the best candidate? The primary voters thought that was highly arguable in 3 or 4 instances. There is only one instance where the endorsee lost the primary and the winner lost the general. There are three instances where a majority of general election voters preferred the Republican, and two where the endorsee lost by a very narrow plurality.
There are two instances where the party endorsed someone highly likely to lose when it endorsed the incumbent, 1978 and 1990. The incumbent was Rudy Perpich each time, and again, incumbents run in the primary when the party endorses a challenger, and they’re almost always going to win the primary, so there’s little point in endorsing a challenger unless the incumbent is very weak and the challenger can pull off a primary upset … which few can. So Perpich really put the party in no-win situations those years.
I wondered if the region endorsees come from might be the issue, and I count seven times the DFL endorsed someone from the central cities, five times the endorsee was from outstate, and one from the suburbs. Central city endorsees are 3-3, 3-4 if Murphy loses on the 14th, or 4-3 if she pulls it out. The only outstate endorsee to win was Perpich in 1986. The suburbs have a sample size of one. These samples are too little for a firm conclusion, but they do make me wonder if endorsing an outstate candidate is such a panacea.
Maybe what’s most interesting is I figured going through the past elections one by one, some pattern would present itself, but none is clear. To be sure, endorsing weak candidates is a formula for losing, but the weak candidates are sometimes incumbents. We can speculate about an alternative way to pick candidates, and I’m about to, but we have to do so having no way of knowing that we would get a different result, either in terms of who gets endorsed or their success rate. Really, we don’t even know that changing our system does anything to address our problem, since we haven’t even defined it.
We also have to give credence to the idea that governor is just a fundamentally different office from any other. Consider not just the DFL’s problems in electing governors, but how often governors seem to run counter to the predominant color of their state. By that, I mean that it’s not all that rare for Democrats to win in red states, and Republicans to win in blue states, even though how states vote for president has turned into a good predictor of how those states will vote for US Senate and constitutional offices (other statewide offices, like state AG or Auditor). I don’t know why governor is so different, but looking at the results, I have to lean toward thinking something is fundamentally different between governor and all other offices, and we’re just not different from other states. I had been leaning to the idea that it’s just random clustering that our endorsement losses are concentrated in the governor’s office, but looking at other states inclines me to the “governor is different” school of thought.
I know I said I was going to get into speculating about other ways of endorsing, but I have become conscious of how long this post has gotten. I’m clearly pushing into TLDR territory. So, sorry to disappoint anyone who has read this far, but I’m going to plan a separate post.
From Greg Laden: I agree that it is hard to see a pattern, or that anything that looks like a pattern is just one of those pseudopatterns that pertain only until they are broken, because they were never really there. Not finding a pattern, though, is a significant finding.
I also agree that it may simply be the case that Govs are different. For one, there are bigger stakes. Think about it. If an elected official is more powerful if they are more rare (i.e., the President — only one, vs. a member of the US house) then Governors are a bigger deal then Senators by double. That’s got to draw money, attention, power brokers, etc.
The stakes are high enough for governor that this silly little system of endorsements and primaries isn’t important.
Meanwhile, the endorsement system seems to work quite well at the state House and Senate level, and for CDs most of the time, and other offices.
This is one reason I get annoyed at people who want to throw the whole endorsement system out, adducing only the evidence that it seems to go south in the governor’s race so often.
Anyway, excellent post, looking forward to the solution that we can adopt!
My suggestions to consider:
– shorten the time between endorsement and primary, then don’t get all worried about the primary. If the time is short, the endorsement will mean more.
– examine and evaluate the ways the endorsement benefits a candidate and see if that needs to be updated or adjusted.
– update the endorsement process. Delegates should be required to represent a candidate, and the first vote should require that representation (i.e., be automatic) for governor, like a national presidential convention. That would solve a number of problems including unfaithful delegates. Of which there are potentially many, since there is no check against that at this time.
From Mac Hall: Eric,
An interesting assessment … and may I offer some thoughts ?
I am not a big fan of history … every year is different … the candidates, the number of candidates, the issues and timing of events (Grunseth / Wellstone) are different each time. Plus, can you really compare any election to today with the post-Citizens United campaigns to yesterday ? But, I get your point.
That said, Minnesota is such an odd state with the impact of the Independence Party on the general election results … and the willingness of voters to go into the voting booth and vote for someone that they know has no chance of winning — but they just cannot vote for the DFL or MNGOP. (Heck, just look at 2016 presidential election and the number of votes that McMullin, Johnson, and Stein got.) Your comment about Tim Penny is spot-on … it was his strength in Rochester that tilted the election … Pawlenty won Rochester the next time as an incumbent.
And primaries do not get the participation that a November election does. Look at how Alexandria Ocasio-Cortez took out Joe Crowley in NY14 by 4,116 votes but less than 27,000 were cast … in the 2016 general election, Crowley got over 147,000. People don’t participate in primaries … and since Minnesota is an Open Primary, crossovers and independents can vote. IMO, Dayton won in 2010 because the MNGOP thought he would be an easier candidate to beat (read my analysis). Will Republicans have to vote in their 2018 primary because of the Pawlenty-Johnson contest … or will they have fun by voting for Murphy and Ellison in hopes of being able to brand the DFL in November ?
So, that takes us to today’s political world and the caucus endorsement versus the primary. Ya gotta wonder if there would be as much hand wrangling if the 2018 DFL Governor’s race was called a no-endorsement after the convention’s third round ? And this year, there are so many candidates who just announced right before the filing deadline that even the convention delegates may look at them differently today.
I wonder if the endorsement process should be restricted to just vetting candidates … so that you could have multiple vetted candidates for the primary.
Guess we’ll just have to wait and see the results and your next post.
From Dan:
Am I having deju vu? Nope, another election cycle and another debate about the worthiness of the DFL endorement process. Ok, then. Its very simple. Caucuses disnfranchise voters. Instead of having polling places all over and open all day (with early and absentre voting as well) there are far fewer places to vote and a much shorter window of time. As a result, lots of people are disenfrnchised. People who work nights and can’t get off. People with young kids to care for. People who travel for work. Active military. Old and/or disabled people who can’t walk 8 blocks to the site in the winter when there is nowhere to park, or if they get there, unable to stand in line for an hour or two waiting to sign in. Voting should be accessible as possible. We complain about voter ID laws and voter purges done by Republicans making it difficult to vote, but caucuses make it extremely difficult, if not impossible for people to vote. The DFL choosing candidates through the caucus process is hypocritical, undemocratic and really, completely untenable. Maybe, just maybe, you could justify the large scale disenfranchisement of voters if the caucus process had good results. But it hasn’t in nearly 50 years in governor’s races. And its not hard to figure out why. When you disenfranchise large groups of people, the people who do the choosing are unrepresentative of the DFL electorate. The endorseed candidates lose in the primary or the general because they aren’t what the DFL electorate wanted. They are what a small, unrepresentative group wanted. A primary may not have high turnout, but no one is prevented from voting and the makeup of the voters is going to be more representative. We ditched caucuses for presidential elections and the DNC is working to limit and discourage caucuses with its new rule changes. You can try to polish the caucus turd all you want, but at the end of the day its still a turd.
UPDATE: Here’s part 2.

{ 3 comments… read them below or add one }

{ 1 trackback }

Previous post:

Next post: