The races for elected offices are not the only campaigns being fought this fall. There are also four ballot questions for voters to decide this November, with camps on both sides trying to persuade voters of the merits of a "yes" or "no" vote on each.
Massachusetts is not as referendum-happy as other states like California, but we do have a long history of ballot initiatives. Since 1919, Massachusetts has voted on 173 ballot questions. History shows that ballot question voting was a bit of a specialized endeavor early on, with a significant percentage of voters who showed up to vote in other races not voting one way or the other on the ballot questions.
Blank votes are not a trivial matter for ballot question proponents. For the types of questions on the ballot this year to pass, a measure must be approved by not only a majority of voters voting on the question but also 30 percent of all voters who cast ballots in the election. Results have not been anywhere near that level of roll-off in recent years, so the 30 percent threshold has not arisen as an issue. Since 2000, the average ballot question has been left blank by just 7 percent of voters, meaning more than 90 percent of voters who show up for governor, senator or president are also making a choice on the ballot questions before them.
Not Everyone Votes On Ballot Qs
At first blush, this 7 percent figure looks impressive. But digging deeper into the numbers shows that that blank ballots are not evenly distributed. In towns with lower income and a higher minority populations, more voters leave ballot questions blank.
In Lawrence, for instance, an average of 16 percent of voters have not voted on each ballot question since 2002. On 2010's Question 1, to repeal the sales tax on alcohol, roll-off was north of 20 percent in 29 precincts, all in Boston and Gateway Cities like Lawrence, Chelsea, Springfield and Fall River. All of the high roll-off Boston precincts were majority-minority, according to U.S. Census figures.
The electorate in midterms is already less diverse than the state's entire population of registered voters. So with fewer minority voters casting ballots, and more leaving the ballot questions blank, the voting power of the state's minority communities is seriously diluted.
The practical implication for campaigns is that the voters deciding the ballot questions in any given year are different, demographically, than those voting for the elected offices higher up the ballot. Ballot campaigns need to persuade an electorate that is wealthier, less urban and less diverse than the electorate as a whole.
Don’t Worry About Ballot Length Or Order
Ballot campaigns seize on their question number as a way to communicate to voters. Expect to see "Vote no on 1" or "Yes on 2" bumper stickers and ads around town between now and November. But as prominent as these question numbers are in the campaign, the order of the ballot questions seems not to affect how many blank ballots are cast on the question.
One might think a ballot question at the end of the list might be overlooked by voters who tire before reaching the back page of a long ballot. But our review of nearly 100 years of elections results shows no such relationship. Wherever a question is placed on the ballot, it seems to stand an equal chance of voters finding it. Also of no obvious impact is the number of ballot questions on the ballot in a given year. The data shows that voters will work their way to the end of a long ballot at the same rate as they will finish a short ballot.
This counterintuitive finding should bring comfort to those campaigning for initiatives placed far back in a thick ballot. In 2000, when there were 8 ballot questions, Question 1 actually generated slightly more blank votes than did Question 8. In other years, the differences between early and late questions were either negligible, or not consistently ordered from least to most blank votes.
What Are The Odds?
Since 1990, 21 out of 46 ballot questions have passed, a 46 percent success rate. There are several types of ballot initiatives, each of which has a different track record in recent years. All the ballot questions this year are “Initiative Petitions for a Law” seeking to make new law or change existing law through the ballot box. This has been by far the most common type of ballot question in recent years, representing 36 of the 46 ballot questions since 1990. Of the 36 Initiative Petitions for a Law brought before voters since 1990, 18 have passed, a 50 percent success rate.
Steve Koczela is the lead writer for Poll Vault and president of The MassINC Polling Group. Rich Parr is research director of The MassINC Polling Group.
Correction: An earlier version of this story incorrectly reported the type of two of this year's ballot questions. We regret the error.