Last updated on 2018-11-17
My relative lack of enthusiasm for 2018’s five ballot measures was reflected in one odd statistic from our general election results: a larger percentage of voters chose not to vote on most of this year’s ballot measures than did 2014 and 2016 voters.
As the last items on the ballot, listed after all the statewide, Legislative, and local candidates, ballot measures tend to draw fewer total votes. Some voters rush, mark the names they recognize at the top, and don’t bother with all the additional reading at the bottom or page 2 or 3. This year, using the top participation of 339,154 voters in the gubernatorial race as our baseline, we can estimate that 8,000 to 31,000 voters left at least one ballot measure blank. The average drop-off on the ballot measures was 6.55%:
| Ballot Measure | Total Votes | Drop off |
| W | 316,782 | 6.60% |
| X | 308,024 | 9.18% |
| Z | 313,670 | 7.51% |
| 24 | 315,061 | 7.10% |
| 25 | 331,109 | 2.37% |
| Average | 316,929 | 6.55% |
Notice that the drop-off does not vary directly with ballot order: the last item on the ballot, Initiated Measure 25, the tobacco tax for vo-techs, drew the highest participation, while the second item on the ballot, Amendment X, the 55% vote threshold, drew the least interest.
Drop-off this year also appears not to correlate with complexity. Amendment W, the constitutional reboot of 2016’s IM 22, was easily the longest and most complex measure, but three far simpler measures—X, Z (single-subject rule for amendments), and 24 (banning out-of-state money in ballot question campaigns)—had worse voter drop-off. IM 25 was arguably more complicated than X, Z, and 24, but it drew more interest than any of the second-tier statewide candidate races (Secretary of State, Attorney General, etc.), where drop-off averaged 6.52%.
It appears that ballot order and complexity are overwhelmed by advertising and awareness. Every gas station in this state had posters and decals shouting NO on IM 25. Amendment W might have come in second for advertising. Look, then which two measures got the highest participation.
IM 25 had the second-lowest drop-off in the last three elections. It tied with 2016 Amendment S, the crime victims’ bill of rights, on which California billionaire Henry T. Nicholas spent millions of dollars to scrawl his poor dead sister’s name into our constitution. The recent champ for low drop-off was 2014 Initiated Measure 18, the successful minimum wage increase, which maybe didn’t have the same money behind it on either side but certainly got lots of press.
But 2018’s average ballot measure voter drop-off of 6.55% was higher than in the two preceding elections. Against the highest-turnout races in each election, average ballot measure drop-off in 2016 was 4.56%. Average ballot measure drop-off in 2014 was 3.16%.
| 2016 | Presidential vote: | 370,093 |
| Ballot Measure | Total Votes | Drop-off |
| Amendment R: Regents/vo-tech authority split | 352,111 | 4.86% |
| Amendment S: Glodt’s crime victims bill of rights | 361,335 | 2.37% |
| Amendment T: Independent Redistricting Commission | 348,881 | 5.73% |
| Amendment U: Usury! Payday Lender Protection Clause | 355,456 | 3.95% |
| Amendment V: Open Nonpartisan Primary | 354,611 | 4.18% |
| Initiated Measure 21: 36% Payday Loan Rate Cap | 357,625 | 3.37% |
| Initiated Measure 22: Anti-Corruption Act | 349,800 | 5.48% |
| Initiated Measure 23: “Fair Share” Union Dues | 350,693 | 5.24% |
| Referred Law 19: Incumbent Protection Act | 340,756 | 7.93% |
| Referred Law 20: Youth Minimum Wage Cut | 360,830 | 2.50% |
| Average | 353,210 | 4.56% |
| 2014 | US Senate Vote: | 279,327 |
| Ballot Measure | Total Votes | Drop-off |
| Amendment Q: keno/craps in Deadwood | 268,515 | 3.87% |
| Initiated Measure 17: Any Willing Provider insurance measure | 269,118 | 3.65% |
| Initiated Measure 18: increase minimum wage | 273,907 | 1.94% |
| Average | 270,513 | 3.16% |
The differences from 2014 and 2016 to 2018 are small. Slightly less interest in this year’s ballot measures could have less to do with the measures themselves and more to do with the fact that the Sutton/Noem race sucked up huge quantities of media attention and campaign energy, leaving the ballot measures undercovered compared to 2014 and 2016, when the candidate contests did not capture the same level of attention as the cowboy versus the beauty queen.
But participation in all three years support the notion that ballot order and complexity matter less than the attention that advocates and opponents alike can draw to the ballot measures.
Excellent analysis! Thanks!