Fans of ranked choice voting, take note: the South Dakota Legislature uses ranked choice voting!
As I mentioned earlier, the Legislature’s Executive Board will meet Monday after session to choose interim study topics. The Executive Board will base its choices on a survey, completed by 68 of its 105 members, asking legislators to pick and rank up to five topics, out of fifteen submitted, that they’d like to see studied this summer. The Legislative Research Council has tallied the results by a ranked choice voting system, giving each topic five points for each “1” ranking, four points for each “2”, three for each “3” two for each “4”, and “1” for each 5.
Compare the rankings that would result from allowing each member to vote solely for their single favorite topic and the rankings calculated by LRC from ranked choice voting:
| Topic | rank by single vote | rank by ranked choice voting |
| A: Non-English Driver License Exams | 10 | 12 |
| B: Econ. Impact Illegal Immigration | 5 | 10 |
| C: Wind Energy Permits, Laws, Subsidies | 10 | 9 |
| D: CAFO Zoning | 8 | 13 |
| E: K-12 Mental Health & Safety | 1 | 1 |
| F: Brand Board | 3 | 5 |
| G: Trust Land County Budget Impact | 13 | 10 |
| H: 4-H | 15 | 14 |
| I: Drug Court/Treatment for Homeless | 8 | 6 |
| J: Special Ed Cost | 5 | 3 |
| K: Mental Health Services | 3 | 2 |
| L: State Health Insurance Expansion | 7 | 7 |
| M: Budget Cuts | 2 | 4 |
| N: School Capital Outlay | 10 | 7 |
| O: REAL ID Data | 13 | 15 |
In both cases, Topic E, K-12 Mental Health and Safety, comes out on top. It received seventeen first-place votes, the most for any proposed topic. Under ranked choice voting, Topic E wins 85 points for those seventeen #1 votes, plus 24 for six 2s, 21 for seven 3s, and six for three 4s, for 136 total RCV points.
But look at the differences among runners-up. Suppose the Legislature only picks four topics, as it did last summer. By straight single vote, second place would go to Topic M, Budget Cuts, with eleven first-place votes. Tied for third and thus rounding out the top four would be Topic K (Mental Health Services) and Topic F (Brand Board), each of which won six first-place votes. Just missing the four-topic cut would be Topic B (Economic Impact of Illegal Immigrants) and Topic J (Special Education Cost), each of which won five first-place votes.
Ranked choice voting produces different results that better reflect the sense of the legislators. Topic K comes in second with 105 points, thanks to being the second choice of eleven legislators, the highest number of 2s among all topics. Topic J ranks third, thanks to a similar boost from eight second-favorite votes. Topic M drops from second on straight vote to fourth under ranked choice voting: only four legislators marked M as their second choice and only one marked it third. Topic F misses the cut at fifth: it won the most third-choice votes (nine!), but that wasn’t enough to make up for lagging behind the qualifiers in first-, second-, and fourth-choice votes.
Notice also that ranked choice voting helps break ties. Ranking these fifteen topics, single-vote produces five ties, including two ties affecting top five and a three-way tie for tenth. On the same ballot, ranked choice voting produces only two ties, for seventh and tenth.
We can see the usefulness of that tie-breaking power by comparing votes for Topics B and J:
| COUNTS | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O |
| 1 | 2 | 5 | 2 | 3 | 17 | 6 | 1 | 0 | 3 | 5 | 6 | 4 | 11 | 2 | 1 |
| 2 | 3 | 4 | 5 | 1 | 6 | 3 | 2 | 4 | 4 | 8 | 11 | 5 | 4 | 4 | 2 |
| 3 | 2 | 0 | 4 | 3 | 7 | 9 | 6 | 3 | 7 | 5 | 6 | 2 | 1 | 6 | 3 |
| 4 | 3 | 0 | 4 | 4 | 3 | 2 | 3 | 4 | 5 | 7 | 6 | 4 | 3 | 5 | 0 |
| 5 | 5 | 1 | 1 | 2 | 0 | 3 | 5 | 0 | 4 | 5 | 1 | 4 | 5 | 4 | 5 |
| rank | 10 | 5 | 10 | 8 | 1 | 3 | 13 | 15 | 8 | 5 | 3 | 7 | 2 | 10 | 13 |
| POINTS | |||||||||||||||
| 1 | 10 | 25 | 10 | 15 | 85 | 30 | 5 | 0 | 15 | 25 | 30 | 20 | 55 | 10 | 5 |
| 2 | 12 | 16 | 20 | 4 | 24 | 12 | 8 | 16 | 16 | 32 | 44 | 20 | 16 | 16 | 8 |
| 3 | 6 | 0 | 12 | 9 | 21 | 27 | 18 | 9 | 21 | 15 | 18 | 6 | 3 | 18 | 9 |
| 4 | 6 | 0 | 8 | 8 | 6 | 4 | 6 | 8 | 10 | 14 | 12 | 8 | 6 | 10 | 0 |
| 5 | 5 | 1 | 1 | 2 | 0 | 3 | 5 | 0 | 4 | 5 | 1 | 4 | 5 | 4 | 5 |
| 39 | 42 | 51 | 38 | 136 | 76 | 42 | 33 | 66 | 91 | 105 | 58 | 85 | 58 | 27 | |
| rank | 12 | 10 | 9 | 13 | 1 | 5 | 10 | 14 | 6 | 3 | 2 | 7 | 4 | 7 | 15 |
Topic J, the reasonable and practical study of special education costs, gets five first-place votes, the same as the right-wing-inspired Topic B, the study of the economic impact of illegal immigration. Topic J would produce useful, actionable data on an issue with significant and demonstrable fiscal impacts on the budget of every school district in the state. Topic B would simply foment more extremist chest-thumping rhetoric about an issue the Legislature can do little if anything to regulate (say it with me again: immigration is a federal issue). DiSanto, Frye-Mueller, Gosch, Howard, and Kevin Jensen are as passionate about yelling about illegal immigrants’ impact on our economy (though not acknowledging past research showing the impact is not negative) as Bartels, Duvall, Johns, Johnson, and Wiik are about studying special education needs, so a single-vote system ranks those topics equally. But get past first choices, and we see that support for Topic J is deeper. J and four other topics get more second-choice votes than B. Nobody marks B as a third or fourth choice, and only Livermont marks B as a fifth choice. Only ten legislators put illegal immigration in their top five; thirty legislators deem special ed funding that important. By that count of unweighted top-five interest shown, B came in dead last, while J tied with K for top-five interest.
I don’t think the Executive Board is bound by this survey. They can choose to study whichever and however many topics they want. However, ranked choice voting helps the Executive Board more accurately gauge interest in topics and differentiate topics with broader interest from measures with narrower, more single-minded and noisy support.
And if our Legislature has the brain power to use it for their summer study topic selection, regular voters can handle it on crowded ballots like the Sioux Falls mayoral race. Let’s do away with runoffs and elect leaders who better reflect their true popular support with ranked choice voting!
Cory writes:
I’d like to point out that many advocates of ranked-choice voting are referring to something a little different when we recommend using it to elect candidates to political office.
In the most commonly advocated system, if no candidate gets 50 percent of the vote, the votes for the candidate who gets the fewest votes are reallocated to those voters’ second choices. If there’s still no candidate with 50 percent of the vote, the votes for the remaining candidate with the fewest votes are reallocated to those voters’ next (second or third) choices, and so on until a candidate has 50 percent.
Imagine a four-way race. It’s possible for the candidate with the third-most votes in the initial balloting to move into second place when the fourth-place candidate’s votes are reallocated, and then into first place when the third-place (initially second-place) candidate’s votes are reallocated. In other words, every voter has direct input into the final outcome, and no candidate goes into office with less than 50 percent of the vote.
Establishment Republicans tend to dislike this system, because it would allow Libertarians to actually vote for Libertarians without the supposed risk of “wasting” their votes or “spoiling” elections. In other words, many Republicans would prefer to risk a “spoiled” election rather than to risk a strong Libertarian showing on Election Day.
That is how grudznick thought ranked voting worked, but Mr. Evans explained it to me differently. As a dimwitted old man still confused, I am still in favor of using Mr. Evans’ version during the public voting to help Libertarians and further confuse voters, most of whom are not very smart or up to speed on the issues anyway.
What Kurt describes is the way ranked choice voting is done in Minneapolis, St. Paul, Duluth and other Minnesota government entities. The start was a little rocky, until voters and candidates became acclimated to the system. While there are still some naysayers, the overall attitude toward RCV in MN seems to be positive.
It sounds like we may start using it in South Dakota soon, Ms. Geelsdottir, if Mr. Evans has his way. I’m OK with that.
There are currently five ranked-choice voting bills in the New York legislature.
https://ipview.blogspot.com/2018/03/ny-ranked-choice-voting-rcv-update.html