Weighted Vote Hews Closely to SDGOP Delegate Headcount; Natvig Needed to Flip 7 Jackley Voters to Win

While still apparently so shell-shocked at the mugwump convention revolt against his mainstream GOP sponsors that he hasn’t offered any substantive commentary or analysis on Saturday’s surprising nomination votes at the South Dakota Republican Party Convention in Watertown, SDGOP spin blogger Pat Powers has done us the favor of publishing the actual delegate vote counts, raw and weighted, from the contested nominations.

While the binding weighted votes are calculated by an arcane statutory formula, it turns out the weighted votes did not produce results substantially different from the raw delegate votes:

The raw votes listed above are the actual head counts. On Saturday, 366 Republican delegates voted to nominate Marty Jackley for Attorney General; 316 delegates voted to nominate David Natvig for that position. Jackley thus won 53.7% of the votes cast.

However, state law does not allow parties to simply count votes on nominations. SDCL 12-5-18 says that each delegate’s vote is to be counted as “the number of votes equal to his proportionate representation as to all delegates present from that county bears to the number of votes cast in his county at the last gubernatorial election for his party candidate for Governor.” The party must do this math to figure out the legal worth of each delegate’s vote:

1. Divide 1 by the number of delegates present from that delegate’s county. 
   1. Example: 103 delegates from Minnehaha County were present for the Jackley/Natvig AG vote. Thus, each Minnehaha delegate was worth 1/103 = 0.97% of Minnehaha’s vote.
   2. Example: 2 delegates from Faulk County were present for the AG vote. Thus, each Faulk delegate was worth 1/2 = 50% of Faulk’s vote.
2. Multiply that number by the total number of votes cast for Kristi Noem in the 2018 general election in the delegate’s county.
   1. 1. Example: Minnehaha County cast 32,355 votes for Noem in the 2018 general election. Each Minnehaha delegate’s vote is thus worth 0.97% × 32,355 = 314.13 nomination votes.
      2. Example: Faulk County cast 674 votes for Noem in the 2018 general election. Each Faulk delegate’s vote is thus worth 50% × 674 = 337.00 nomination votes.

This weighted voting, proportional to delegation size and 2018 gubernatorial vote, creates some notable disparities in the strength of each delegate’s vote. The average strength of an individual delegate’s vote, based on the 684 delegates from 58 counties counted present for the Jackley/Natvig vote, was about 247.9 votes. But delegate Jake Schoenbeck’s vote from Minnehaha County was worth 314.1 nomination votes, while his mom Donna’s vote from Codington County was only worth 177.7 nomination votes. If Jake’s sister Erin hadn’t been creamed in her precinct committeewoman primary, her vote from Pennington County would have been worth 376.1 nomination votes (so thank goodness Rapid City Precinct 3-5 voters and Jill Mills spared Jake that humiliation!). The strongest delegate in the room Saturday was the sole delegate who rolled in from Gregory County, who was packing 1,107.0 nomination votes. The weakest were each of the 5 Mellete County Republican delegates, each of whom could wield only 59.8 votes. The strongest delegate in the room was thus 18.5 more powerful than the weakest.

To put that power differential in practical perspective, consider that Jackley beat Natvig by 50 actual delegate votes. But if Natvig had persuaded just seven specific delegates to change their minds—that one super-powerful delegate from Gregory, plus the two from McCook, the two from Day, one from Roberts, and one from Deuel—and vote for Natvig instead of Jackley, the weighted vote would have tipped Natvig’s way, 85,112 to 84,114 (50.3% to 49.7%), even though Jackley still would have won the headcount vote 359 to 323 (52.6% to 47.4%).

But the weighted vote did not reverse the headcount vote in any of the three contested SDGOP nominations Saturday. In all three contests, the weighted vote did slightly exaggerate the strength of the radical right-wing insurgent candidates. The largest weighted advantage arose in Steven Haugaard’s surprise and negatively quixotic challenge to incumbent Lieutenant Governor Larry Rhoden. The weighted vote gave Haugaard 2.0 percentage points more than the delegate headcount did. Haugaard got that weighted advantage by outpolling Rhoden in his home county of Minnehaha, the only count he won where delegates had more strength than the statewide average.

Haugaard’s shortest path to reversing the weighted result would have required that he convince the 20 most powerful delegates who voted for Rhoden to come his way—again, our power broker from Gregory, then one from Roberts, two from McCook, one from Deuel, two from Tripp, one from Hanson, all ten Rhoden voters from Lawrence, and two from Butte. Haugaard would then have won the weighted vote 83,626 to 83326 (50.2% to 49.8%) while still losing the headcount 369 to 302 (55.0% to 45.0%).

Lazy incumbent Steve Barnett’s thumpage at the hands of receptionist-cum-election rigger and mass disenfranchiser Monae Johnson was clear by headcount and weighted vote. The weighting only exaggerated Johnson’s vote advantage by 0.8 percentage points. Barnett already won the small but mighty delegates from Gregory, McCook, and Day; to boost himself past Johnson in the weighted count, Barnett would have needed to wheedle at 44 flips from 14 counties, including 11 of Johnson’s 35 voters in Pennington. That effort would have tilted the weighted vote for Barnett 84,542 to 84,424 (50.03% to 49.97%) while leaving the headcount vote in Johnson’s favor 366 to 315 (53.7% to 46.3%).

44 votes out of 681, 20 votes out of 671, 7 votes out of 682—any of those margins could have been overcome by an effective, organized campaign. That slimmest margin, Natvig’s loss to Jackley, was just a well-targeted six pack and a shot away from a flipped result.

But in all three cases, the weird weighted vote, mandated by a statute created by real party animals in the Legislature, still managed to reflect, within two percentage points, the will of the actual people in the room casting ballots.