Yes Votes on Ballot Measures not Related to Length of Ballot

In my conversations about referring Senate Bill 69 and/or Senate Bill 177, numerous South Dakota neighbors have expressed the concern that crowding the ballot with lots of referenda and initiatives reduces the chances of passing any of them. Divide finite attention by increasing ballot measures, and you’re going to get more exasperated voters who just check “No” on everything. Refer bills from this Legislative session, and you’re going to make it harder for Steve Hickey and Steve Hildebrand or Rick Weiland and Drey Samuelson to sell their big initiatives in 2016.

Do more ballot measures mean a higher percentage of defeated measures? My math says… probably not.

Ballotpedia offers a list of all 333 ballot measures from 1890 to 2014. I have condensed all of those measures into a single spreadsheet here. We’ve passed 144 and defeated 189. That’s a 43.2% success rate… although “success” may not be the right word. Folks who put initiatives forward are hoping for a “Yes” vote, while folks who refer laws are generally hoping for a “No” vote. We have voted down 70% of our 44 referenda. If we count defeats of referred laws as “success” for the organizers, then the overall “success” rate for ballot measures rises to 48.7%.

But for this math moment, let’s just say Yes means Yes and No means No. How does the percentage of passed measures correspond to the number of measures on the ballot?

  • In years when South Dakotans have had just one, two, or three ballot measures, the pass rate has varied between 0% and 100%.
  • In the six years when we’ve had more than ten ballot measures, the pass rate has ranged from 9% to 92%.
  • If I get nasty and run the formal statistical correlation of each year’s ballot measure count with the pass rate, I get a small and statistically insignificant number.

More simply put, if South Dakota history means anything, folks referring laws to a public don’t affect the chances Steve and Steve and Rick and Drey have of passing their initiated measures. Ballot measure success is much more likely to be affected by the merits of each issue and the efforts of each campaign.

Bonus chart! Here’s my summary of the pass rates for each type of ballot measure:

total pass rate
Ballot measures 333 43.24%
Initiated laws 53 28.30%
Citizen-initiated constitutional amendments 12 41.67%
Legislatively referred constitutional amendments 221 50.23%
Voter referenda 44 29.55%

14 Responses to Yes Votes on Ballot Measures not Related to Length of Ballot

  1. Nick Nemec

    In other words South Dakota voters are smart enough to actually read the ballot measures they vote on rather than throw their hands up in disgust?

  2. rollin potter

    Not so NICK, It seems when visiting with them before the election most of them do not even know until they go into the box what is on the ballot!!!!!!!

  3. Bill Fleming

    I’ll argue that there’s more of a problem getting people to vote in the first place than there is in getting them to consider their vote once they’ve decided to do it. That said, I’m thinking that conservatives, being inclined to protect the status quo, will be by nature predisposed to vote for no change, unless there is a clear, compelling reason to do so. Both of the issues you are proposing, Cory, are arguments to ‘keep things the way they are.’ That should put you in relatively good stead with the traditional ‘naysayers’. ;-)

  4. Tentaively, historically, statistically, Nick, the answer appears to be “Yes.”

    There’s much more statistical analysis one could do. One could go back through the campaign finance records, find out how much was spent by ballot question committees on each side of each issue, and look for correlations with spending on each issue and total spending per question in each cycle. I could imagine the possibility that in a crowded-ballot year, if one or two issues get lots of money and press and the others go mostly unnoticed, the less-discussed measures might tend to get higher “No” votes. But that’s all speculation, not real math.

    (By the way, I’m of the impression that folks back the minimum increase last year actually outspent opponents.)

  5. Bill, I’ll take votes from whichever quarter will walk with me. But does a 43% pass rate (i.e., votes for change) fit with your hypothesis of a conservative electorate? Maybe we see that conservative split more clearly in the fact that, in initiated and referred laws, we only approve three out of ten. But when it comes to tinkering with the Constitution at the behest of our Legislature (which measures make up two thirds of our ballot measures), we’re batting .502. We’ll change our Constitution more than we’ll change our laws, which seems backwards under my assumption that conservatives would be less likely to change the constitution than change regular laws.

  6. Rollin, if what you’re saying is true, if voters are approaching the ballot with minimal information, then it’s all the more interesting that given more measures, we don’t default to “No” more often. Nick’s suggestion may still apply: even if they haven’t studied beforehand, voters may be taking as much time to read each measure whether they have two or a dozen on the ballot, and they’re about as likely to take a chance on “Yes” regardless of how many they have to read.

  7. mike from iowa

    First initiative should be to get or give voters the initiative to get out and vote. Try at gunpoint.

  8. Roger Elgersma

    I vote regularly but feel it more important to vote when I have more to vote on. Is there a correlation between number of ballot issues and voter turnout?

  9. mike from iowa

    Maybe initiatives should be printed simply and precisely-yes or no on minimum wage,yes or no make voting and running harder. Yes vote wingnuts out. :)

  10. barry freed

    Mike,
    An IM to create a State Voter’s Lottery would do the trick. You vote, you are entered to win cash. The more winners and greater payout amounts, the more people will vote.

    To the question of voter fatigue with a number of IM’s: Rather than the number of IM’s or Referals, I believe the writing of Pros and Cons affects voters decisions more. For example: the Medical Marijuana con was full of opinions, not facts. It frightened the ignorant as we still have a large population that grew up with the Hearst and Nixon lies. They think pot is heroin, or worse.
    One IM should be to establish rules of “proof” for claims made in the Pros and Cons. The Lawmakers will certainly never do that.

  11. mike from iowa

    Barry-that might work. I don’t need any incentive to vote against rethuglicans. I’m sure others do. The last farm bill was sold as an improvement and it is way worse than previous ones.And a majority of people think rethugs can fix the economy,even though they are the ones that caused the problems and tryied to block solutions. Can’t fix stoopid.

  12. Roger, that’s an interesting correlation question. I have turnout numbers handy for 1988-2014. I find a positive correlation that flirts with statistical significance for this small sample. However, higher turnout and higher ballot measure numbers could coincide because of Presidential elections: turnout is always higher in Presidential election years, so initiators and referrers may file more ballot measures in those election years.

  13. Donald Pay

    If there are lengthy ballots filled with ballot issues, isn’t the Legislature the real problem?

    Over half the ballot measures (Constitutional amendments) are put on directly by the Legislature!!! Your ballot would be much simpler if the Legislature would limit its need to fiddle with the state Constitution.

    All of the referred measures result from an action of the Legislature that many folks find objectionable enough that they spend time petitioning. If the Legislature had done a better job legislating, many of those wouldn’t gain ballot access.

    Initiated measures often result from a failure of the Legislature to address issues that many folks think need to be addressed. Isn’t that also at least partly a result of poor legislating?

    How many legislators really understand the issues they vote on? On most bills, a legislator has about a 30 minute head start on the average citizen in terms of their knowledge about the issues involved in a bill. That’s about the amount of time spent in caucus or committee on each bill. A few bills would get a bit more, most a bit less time.

  14. Indeed, Donald, the Legislature has been responsible for two thirds of our ballot measures. Far be it from me to complain about the Legislature giving us the opportunity to exercise our voice… although they are obliged to submit constitutional amendments to us, and they could check their own amending urges.

    As for referenda, we apparently are not easily provoked. By my count from the LRC’s historical records, the South Dakota Legislature has passed more than 27,000 bills since 1890. We’ve referred only 44 of those bills and overturned 31 of them. We’ve let slide 99.89% of the Legislature’s decisions.