Counties to watch on primary night

IMG_0367The 2018 Republican primary is fast approaching, and the gubernatorial matchup between Congresswoman Kristi Noem and Attorney General Marty Jackley is close – a dead heat according to public polling. It could be a late night Tuesday as the results roll in, in what could could be a historically close primary.

A couple of weeks ago, this blog posted county-by-county maps, showing past GOP gubernatorial primary results from every open-seat primary since 1978. As this blog seeks to use past to understand the present, one may ask: What counties have been the “bellwethers” that most closely predict statewide results in GOP gubernatorial primaries?

Beadle County – home of the State Fair and the World’s Largest Pheasant – is the top choice. Coming in #2 is Davison County – home of the World’s Only Corn Palace. (Is kitsch predictive?)

The good news for enthusiastic Election Night watchers is that all ten bellwether counties are East River and in the Central Time Zone, meaning results can start rolling in at 8 PM CT. Here are the Top 10:

County 1978 1986 1994 2002 2010 Modified Average
1 Beadle 2.0% 1.1% 0.1% 12.0% 2.0% 1.3%
2 Davison 1.3% 3.3% 10.4% 1.7% 2.4% 2.2%
3 McCook 3.0% 3.6% 3.4% 2.6% 0.7% 2.4%
4 Grant 0.7% 4.2% 3.2% 3.0% 3.1% 2.5%
5 Day 3.2% 0.3% 15.1% 3.2% 4.3% 2.7%
6 Lincoln 2.1% 4.7% 0.8% 3.8% 4.4% 2.8%
7 Hutchinson 1.8% 3.8% 2.4% 4.1% 3.2% 2.8%
8 Clay 3.8% 2.7% 14.9% 1.1% 3.5% 2.8%
9 Yankton 2.6% 1.5% 11.2% 3.5% 3.7% 2.8%
10 Turner 3.1% 6.4% 3.9% 2.7% 2.1% 2.9%

For each county, the % for each election year is the average amount that the county’s result for each candidate differed from the statewide result. For example, if Candidate A over-performs in a county by 3%, Candidate B under-performs by 12%, and Candidate C over-performs by 9%, the county’s score for that election is the average of those three percentages: 8% in that example. The overall “modified average” is based on the five primary election scores: I gave each county one “mulligan” and threw out the worst (highest) score for each county, averaging the other four. That removes scores that are outliers, usually due geographic factors relating to the candidates. (For example, Beadle County’s high score in 2002 is because it is the ancestral home of the Rounds family.)

Here is a map that shows, graphically, the modified averages for all 66 counties (lighter color designates a lower score, and therefore a better bellwether):

Screen Shot 2018-06-03 at 10.36.11 PM

I also tried a simpler methodology: In these five primaries, how many times did each county vote for the winner? Beadle and Davison counties are among those that are a perfect 5/5 (along with Day, Grant, Kingsbury, Lincoln, Marshall, Minnehaha, Moody and Turner):

Screen Shot 2018-06-03 at 11.20.09 PM

Both maps show that West River counties tend to be poor bellwethers. They are sparsely populated and unlikely to tilt an election, and often vote for candidates who appeal specifically to ranch country, such as Walter Dale Miller or Clint Roberts. Pennington County, with a larger and more urban population, is an exception; by the first methodology, it is actually a better bellwether than Minnehaha County.

As a reminder, here are the candidates who ran in each of these five primaries:

  • 1978 – Bill Janklow (Moody County) 50.9%; LeRoy Hoffman (McPherson) 32.9%; Clint Roberts (Lyman) 16.2%
  • 1986 – George S. Mickelson (Brookings) 35.3%; Clint Roberts (Lyman) 32.1%; Lowell Hansen (Minnehaha) 18.8%; Alice Kundert (Campbell) 13.8%.
  • 1994 – Bill Janklow (Moody/Minnehaha) 54.0%; Walter Dale Miller (Meade) 46.0%
  • 2002 – Mike Rounds (Hughes) 44.3%; Mark Barnett (Hughes) 29.5%; Steve Kirby (Minnehaha) 26.1%
  • 2010 – Dennis Daugaard (Minnehaha) 50.4%; Scott Munsterman (Brookings) 17.6%; Dave Knudson (Minnehaha) 15.8%; Gordon Howie (Pennington) 12.4%; Ken Knuppe (Oglala Lakota) 3.8%

And finally, here is the ranked chart for all 66 counties, including the Top 10 stated above:

County 1978 1986 1994 2002 2010 Modified Average
1 Beadle 2.0% 1.1% 0.1% 12.0% 2.0% 1.3%
2 Davison 1.3% 3.3% 10.4% 1.7% 2.4% 2.2%
3 McCook 3.0% 3.6% 3.4% 2.6% 0.7% 2.4%
4 Grant 0.7% 4.2% 3.2% 3.0% 3.1% 2.5%
5 Day 3.2% 0.3% 15.1% 3.2% 4.3% 2.7%
6 Lincoln 2.1% 4.7% 0.8% 3.8% 4.4% 2.8%
7 Hutchinson 1.8% 3.8% 2.4% 4.1% 3.2% 2.8%
8 Clay 3.8% 2.7% 14.9% 1.1% 3.5% 2.8%
9 Yankton 2.6% 1.5% 11.2% 3.5% 3.7% 2.8%
10 Turner 3.1% 6.4% 3.9% 2.7% 2.1% 2.9%
11 Gregory 13.7% 2.6% 0.3% 4.3% 5.0% 3.0%
12 Douglas 11.6% 5.3% 0.2% 3.2% 4.2% 3.2%
13 Spink 7.9% 3.0% 1.0% 6.6% 3.1% 3.4%
14 Pennington 3.7% 2.5% 5.8% 4.0% 3.8% 3.5%
15 Union 5.7% 1.6% 5.0% 10.1% 1.8% 3.5%
16 Miner 4.6% 6.4% 0.4% 7.9% 3.6% 3.8%
17 Kingsbury 1.8% 5.2% 3.9% 12.2% 4.4% 3.8%
18 Lawrence 6.1% 3.9% 5.2% 7.3% 0.8% 4.0%
19 Lake 3.3% 2.9% 6.6% 7.7% 3.8% 4.1%
20 Deuel 6.9% 3.6% 2.5% 5.3% 5.4% 4.2%
21 Minnehaha 1.2% 8.2% 11.6% 4.1% 3.6% 4.3%
22 Butte 2.9% 5.6% 4.4% 7.9% 4.9% 4.4%
23 Hyde 1.3% 6.4% 7.2% 6.5% 3.5% 4.4%
24 Hand 5.4% 4.7% 2.1% 9.5% 5.5% 4.4%
25 Aurora 10.2% 5.4% 2.6% 6.5% 4.1% 4.6%
26 Brule 9.7% 4.4% 0.7% 6.8% 7.6% 4.8%
27 Jerauld 3.6% 6.2% 11.4% 5.3% 4.4% 4.9%
28 Codington 12.5% 4.4% 8.4% 1.7% 5.1% 4.9%
29 Sully 5.5% 6.7% 10.9% 5.0% 2.7% 5.0%
30 Faulk 8.1% 5.5% 8.4% 4.8% 2.0% 5.1%
31 Marshall 11.6% 3.9% 15.4% 4.0% 1.8% 5.3%
32 Hanson 9.1% 7.3% 5.4% 6.3% 2.4% 5.4%
33 Bon Homme 6.9% 7.1% 17.0% 3.4% 4.5% 5.5%
34 Brown 16.0% 5.2% 10.6% 1.8% 4.3% 5.5%
35 Tripp 11.9% 7.5% 10.7% 0.7% 3.5% 5.6%
36 Potter 9.4% 3.0% 14.3% 4.5% 5.8% 5.7%
37 Fall River 5.3% 5.2% 4.8% 8.0% 10.5% 5.8%
38 McPherson 38.2% 13.4% 1.8% 0.9% 7.3% 5.8%
39 Custer 5.5% 7.1% 5.5% 5.7% 7.9% 6.0%
40 Hughes 3.9% 6.7% 20.9% 9.9% 3.8% 6.1%
41 Walworth 18.0% 12.2% 6.3% 2.6% 3.3% 6.1%
42 Harding 5.9% 13.2% 4.6% 1.4% 17.1% 6.3%
43 Edmunds 19.3% 6.2% 11.3% 3.5% 4.7% 6.4%
44 Charles Mix 12.2% 4.7% 3.8% 10.9% 6.2% 6.4%
45 Dewey 3.3% 9.7% 22.6% 8.6% 4.3% 6.5%
46 Clark 11.7% 4.6% 7.4% 8.3% 5.9% 6.6%
47 Sanborn 6.2% 9.7% 8.5% 6.3% 6.3% 6.8%
48 Meade 12.3% 5.6% 11.0% 5.9% 5.8% 7.0%
49 Roberts 16.9% 6.2% 15.1% 3.9% 3.8% 7.3%
50 Hamlin 14.7% 6.9% 7.4% 10.7% 5.1% 7.5%
51 Moody 18.1% 4.6% 22.3% 2.9% 4.6% 7.6%
52 Bennett 16.4% 7.9% 4.1% 12.3% 6.8% 7.8%
53 Ziebach 8.3% 6.0% 14.9% 4.9% 11.8% 7.8%
54 Perkins 7.1% 7.1% 8.9% 8.1% 17.3% 7.8%
55 Oglala Lakota (1) 9.6% 10.5% 0.7% 18.7% 11.9% 8.2%
56 Jones 16.5% 5.0% 19.0% 4.4% 7.8% 8.4%
57 Campbell 27.9% 21.9% 3.3% 5.8% 3.6% 8.6%
58 Haakon 17.5% 9.3% 16.2% 3.2% 6.4% 8.8%
59 Buffalo 14.5% 12.3% 0.9% 21.6% 7.9% 8.9%
60 Brookings 3.4% 18.7% 10.6% 8.3% 13.3% 8.9%
61 Stanley 18.7% 6.6% 20.7% 9.3% 1.5% 9.0%
62 Jackson (2) 14.3% 8.2% 16.9% 3.9% 9.9% 9.1%
63 Mellette 14.6% 7.6% 10.8% 9.5% 9.5% 9.4%
64 Todd 10.4% 8.6% 7.5% 12.8% 11.1% 9.4%
65 Corson 6.2% 13.0% 12.9% 11.5% 11.6% 10.5%
66 Lyman 34.9% 17.2% 14.1% 8.9% 2.4% 10.6%

Oglala Lakota County was known as “Shannon County” until 2015 but is referred to by its current name throughout this post.

  1. Jackson County’s 1978 score is based on the cumulative vote in Jackson and Washabaugh counties; the two counties reported results jointly in 1978 and were merged in 1983.