Breakdown of Minnesota's voting record on amendments

MPR reporter Sasha Aslanian, who has been covering the marriage amendment debate, asked me a question last week about how non-votes have factored into proposed amendments to the Minnesota Constitution.

Since 1898, voter ratification of a proposed amendment requires not just more "yes" votes than "no" votes, but a majority of "yes" votes from all ballots cast in the state; non-votes on a ballot measure effectively count as "no" votes.

I dug up the data and turned up some interesting facts:

• There have been 145 amendment votes since non-votes have counted as "no" votes. Seventy-one of those proposed amendments have been ratified; 74 have been rejected.

• Only 15 percent of those rejected amendments would have been defeated on the basis of "no" votes alone. Put another way, 85 percent required at least some blank votes to ensure defeat.

• The lowest percentage of non-votes was 3.19 percent in 1988 on the proposal to authorize a state lottery. The highest was 55.81 percent on a proposal in 1900 to increase debt limit of municipalities borrowing permanent school funds.

In 1898, after 40 years of requiring a simple majority of yes votes to ratify a proposed amendment, voters approved an amendment to count the non-votes as "no" votes on constitutional questions, creating a higher hurdle for ratification. It became known as the "brewer's amendment" because it was reportedly pushed by the liquor industry to make it difficult to enact prohibition measures via the constitution.

Question: What can the data tell us about how non-votes will factor into the proposed marriage and voter ID amendments?

The numbers suggest that when indifference about a proposed amendment is low it becomes much more difficult to defeat. In fact, there appears to be a marked threshold when non-votes drop below 20 percent. When non-votes amount to more than 20 percent of votes cast, the amendments have been rejected 72 percent of the time. When the blanks fall below 20 percent the rejection rate drops to 22 percent.

Since 1954, non-votes have never amounted to more than 20.19 percent of the total.

Given that history and based on the amount of activity around the marriage and voter ID amendments, I think it's safe to predict ballot indifference is going to fall below the 20 percent threshold.

It seems to me, interest in the two proposed changes is more like the establishment of a lottery (approved in 1988) or prohibition of alcohol (rejected in 1918) than authorizing investment of school and university funds in first mortgages on improved farms (rejected in 1916), or allowing legislators to seek election to other offices and to provide resignation procedure for legislators (rejected in 1966).

Bottom line: Polls show voter opinion close enough that non-votes could certainly factor in the outcome of either proposal. But strategically, opponents would be unwise to count on that boost.

You can look at the data and draw your own conclusions.

UPDATE: An earlier version of this story incorrectly stated that in 2006 there were no non-votes on the proposed amendment to dedicate the motor vehicle sales tax to highways and public transportation and that there were no other statewide measures on the ballot. Information published by the state incorrectly reports the tallies for the 2006 amendment. There was an election for a U.S. Senate seat and Minnesota constitutional offices in 2006.