This started out to be a review of how ranked-choice voting (RCV) performed in its big Minneapolis test last month and what that portends about RCV’s future in Minnesota. It’s turned into a more interesting (at least to me) exploration of why the DFL Party is (somewhat secretly) divided over RCV, for reasons that are hard to nail down and are not discussed publicly.

Minnesota’s two biggest cities (both DFL-dominated) are now using RCV. A prominent DFL legislator (Steve Simon of St. Louis Park, currently running for secretary of state) is sponsor of a bill that would make it easier for more cities to adopt RCV. The DFL platform embraces RCV and recommends that it be expanded all the way to statewide races.

And yet — relying on a number of not-for-attribution conversations — I know that important DFLers dislike RCV, would not like to expand its use, would like to get the endorsement of RCV out of the permanent party platform and are interested in getting rid of RCV for Minneapolis and St. Paul elections (although that would be a major lift since it is now embedded in those cities’ charters and would have to be voted out by the public).

Minnesota Republicans, by the way, don’t seem to be much divided at all. The MNGOP platform states: “We oppose the implementation of any voting schemes that violate the principle of one man, one vote including Instant Runoff Voting.” Yes, nowadays we’re supposed to call it one person, one vote, and we’re supposed to use “ranked choice” instead of “instant runoff,” and no, it doesn’t really make sense to oppose RCV on one-person-one-vote (the Minnesota Supreme Court upheld the constitutionality of RCV in 2009). Nonetheless, the MNGOP is unambiguously opposed to RCV.

There are principled arguments on both sides of the long-running question of whether RCV is a step forward for voting and democracy. We’ve heard the arguments in recent years. But when political parties and factions within political parties are battling over something like this, it likely means there are partisan and factional stakes.

What is at the root of the intra-party DFL tension over RCV? I’m still not completely sure, but if you’ll tolerate a bunch of unattributed thoughts and arguments, I’ll tell you what I’ve heard and tentatively figured out. The one-paragraph overview is this:

The deep permanent DFL establishment crowd dislikes RCV. Some think it’s bad for the DFL as a party, some think it’s bad for the power of the establishment within the DFL. The factional split played out in the big mayoral race in which Mark Andrew, who was the establishment candidate, lost to Betsy Hodges, whose campaign was much more RCV-friendly and who seemed to end up with most of the support of the relatively anti-establishment elements of the party.

This establishment/anti-establishment analysis is also too simple. Outgoing Mayor R.T. Rybak, who certainly must be considered a big cheese in the Mpls DFL, seemed to be part of the anti-Andrew coalition, although he never openly endorsed Hodges. TakeAction Minnesota —  which has become a big part of what you might call the left wing of the DFL establishment — did endorse Hodges. Hodges’ campaign manager, Andy O’Leary, was a former executive director of the Minnesota DFL Party. So I wouldn’t want to overstate this whole idea that there is an establishment element and an insurgent element at play here.

But my impression is that the party professionals and the most important long-standing members of the Dem establishment mostly favored Andrew for mayor. Walter Mondale endorsed Andrew. Sam and Sylvia Kaplan fund-raised for him. The Rices (brothers Brian and Devin) worked for him. The billionaire Opperman family made the biggest single reported donation ($25,000) to help Andrew (technically it was to a third-party group supporting Andrew). These are very considerable assets to any DFLer. And Andrew had — by far — the best-funded campaign.

So, cutting to the chase, the election to some extent pitted a somewhat insurgent Hodges campaign that thrived on the RCV angle vs. Andrew supporters, including some influential ones whom I would characterize as members of the DFL establishment who don’t like RCV and would like to get rid of it.

How it played out

If you look for it, the Minneapolis mayoral race of 2013 exposed the factional divide within the DFL. You could identify the factions with the Andrew and the Hodges campaigns, although at times it seemed closer to being a divide that pitted all the other serious candidates against Andrew. But, of course, it ultimately accrued to the benefit of Mayor-elect Betsy Hodges.

As a city council member, Hodges was among the leaders of the decision to switch the city to RCV. She campaigned comfortably and effectively in the new system. (She had already experienced RCV in running for council in 2009.)

She seems to have formed alliances with other candidates (especially Don Samuels and Cam Winton) that helped her build and maintain her lead through the successive rounds of counting. (She won her final, convincing-but-not-majority victory on the round after Samuels and Winton were eliminated from contention, as the lion’s share of their votes were assigned to her as the second choice on those ballots.)

Andrew, her chief opponent in the race for mayor, told me during the campaign that while he was trying to warm up to the new system, it wasn’t natural to him to ask voters to consider ranking him as their second choice. “I’m being coached to do that,” he said.

Analysis of the voting data confirm that those who voted for Hodges were most likely to rank at least one more choice. Those who voted for Andrew were least likely. In the end, since Hodges and Andrew were the last two candidates standing in the rounds of counting, it doesn’t really matter whether their supporters ranked a second choice (since those ballots counted for the first choice candidate all the way through).

There was a time, in the middle of the campaign, when it seemed reasonably likely that Andrew would lead among first-rank votes but might be overtaken by second- and third-preference votes for Hodges.

In the end, Hodges led on every round of counting. I believe that prevented a bigger potential storm among Andrew supporters that might have burst if the new voting system had more clearly cost him the election. Andrew supporters to whom I spoke expressed more or less respect for the convincing nature and scope of Hodges’ win. She led on every round, and it was never really close.

It is hard and a tad far-fetched (but, I discovered, still somewhat possible) for some Andrew supporters to convince themselves in the aftermath that, if this election had been held under the old primary and two-candidate-runoff system, Andrew might have prevailed, and that scenario also provides some clues as to why some elements of the deep party establishment doesn’t like RCV.

Why don’t they like it?

When I asked people who had been close to the campaign why the establishment crowd seemed to dislike RCV they explained to me — slowly and clearly, as to a small child — that the establishment has thrived, succeeded and become the establishment under the previous system. When you are winning under a system, you don’t want to change it, they said.

When I pressed for concrete mechanisms by which RCV seems to reduce the influence of party activists, professionals and fundraisers, they came up with a couple of possible theories.

For example, uniting the party machinery behind an endorsed candidate is one of the ways the party expresses its power.

At the DFL endorsing convention in June, Andrew led on every ballot but was never able to reach the 60 percent threshold needed for endorsement. The result was a deadlocked convention and no endorsement. Some DFLers, especially Andrew supporters, believed that the logic of RCV caused several candidates to believe that their strength on second- and third-place rankings might enable them to come from behind during the counting.

Don Samuels and Jackie Cherryhomes are sometimes mentioned as candidates who, under the old system, would have come into the convention pledging to abide by the endorsement and, when it became clear that they wouldn’t be endorsed, would have dropped out of the race and perhaps made it possible for Andrew to be endorsed and gain support among those believe in the endorsement system. An endorsement also frees up the party machinery itself to get openly behind the endorsed candidate, even against other DFLers who are contesting the race.

This whole theory falls apart, of course, unless you believe that in a non-RCV race, Andrew would have been endorsed. There were instances of non-endorsement in Minneapolis mayoral races in the pre-RCV era. But if the logic of RCV cost Andrew the endorsement, it’s an example of how Andrew supporters believe RCV undermined his chances.

Another theory is one that goes straight at one of the pro-RCV arguments. RCV advocates claim that the system discourages negative campaigning. In a conventional election, the voter has two realistic choices on the final round. If you run attack ads against your opponent, you may persuade their supporters to vote for you, or you may just turn them off and prevent from voting at all. Either way, it’s a net gain for you.

But in a multi-candidate RCV situation, you have to think about getting voters — voters who are planning to vote for one of your opponents — to at least rank you second or third. In this scenario, if A attacks B, you may drive supporters to C, and you may reduce your chances of getting B’s supporters – who resent your attacks on B — to rank you second.

Most observers of the 2013 race agree that it was impressively clean and positive. Some complain that without negative campaigning, it’s hard for a candidate to draw clear distinctions. And this race, with so many candidates and so little negative campaigning, may have suffered on that score. But if you are the best-funded candidate (and Andrew was, as the establishment candidate is likely to be), perhaps your money doesn’t do you as much good as it might have under the old system, when the biggest expenditure of most campaigns is advertising, and often attack advertising.

I don’t know how far-fetched you may find factors like these. Perhaps they help you understand why disappointed Andrew supporters may have convinced themselves that — especially compared to the previous two-step system of city elections in which Andrew and Hodges might have faced off one on one — RCV undermined the power of the party establishment and the larger campaign machinery to make their advantages felt in the race.

Effective Democracy is a year-long series of occasional reports supported by the Chicago-based Joyce Foundation, as part of a grant made to MinnPost and the Wisconsin Center for Investigative Journalism.