Count'em Up - Suffolk County DA Straw Poll

On April 9, 2018, the five candidates for Suffolk County District Attorney shared the stage at the English High School in Jamaica Plain to make their case to the voters. Voter Choice Massachusetts was invited to conduct a straw poll with Ranked Choice Voting to help voters navigate this crowded field and find out which of these candidates was able to consolidate a majority of attendees' support, which "pick one and pray" plurality voting -- the way we vote now -- routinely fails to do. 

Civic and political organizations, as well as the public at large, are increasingly recognizing the critical role district attorneys play in our lives. Today, more than ever, the public is growing acutely aware of how important it is to have public oversight and involvement in our criminal justice system so that it works the very best it can. As a result, people are paying more and more attention to district attorney elections statewide and nationally.

The candidates are, in order of appearance in the photo: Rachael Rollins, Greg Henning, Shannon McAuliffe, Evandro Carvalho, and Linda Champion

The forum was sponsored by JP Progressives and the Boston NAACP and was covered by this article in Commonwealth Magazine. (See the event Facebook page for more details about the event.) No Republican candidates have announced to run in their primary for this office, making the results of this Democratic primary decisive for the final outcome.

There are several self-described progressive candidates running. On the other hand, candidate Greg Henning has received significant campaign contributions from outgoing DA Dan Conley as well as police officers. This suggests he may be capable of consolidating a plurality of support from more moderate or conservative voters.

The way our “plurality” elections work now, voters can only pick one candidate on their ballot, and the winner is the person who gets the most votes, but not necessarily a majority -- more than 50%. The four self-described progressive candidates could end up splitting the support of their shared voter base and hurting each other. This would allow a more self-described moderate or conservative candidate to win with as little as 20% of the vote. If it turns out that the broad electorate leans progressive, then there is a problem. This splintering of ballots of like-minded voters between similar candidates could mean that the majority of the electorate could be denied a winner who they support.

Let’s see what the straw poll was able to tell us. (If you're not familiar with how Ranked Choice Voting is counted, here is a 1.5 minute primer from Minneapolis.)

The Poll Results


The poll results are by no means scientific, and should never be construed as an accurate representation of the votes to be cast in the state primary election on September 11th. The poll only reflects the views of the attendees of the forum who chose to express their candidate preferences on our sample Ranked Choice Voting ballot (which can be viewed at the top of this post). (See Disclaimer Detail at the bottom of this post.)

Initial preferences

In the first round, in which only voters' 1st-choice votes were counted, Rachael Rollins is in the lead with a plurality of the vote. However, both Greg Henning and Shannon McAuliffe won solid support from the audience. Would Rollins struggle to maintain her top position as other candidates’ voters moved on to their second choices or was she really the most preferred candidate?

The field consolidates

After the two lowest vote getters are eliminated and the ballots of their supporters counted towards those voters' next choices, the picture did not change much. At this point, Greg Henning would be eliminated as having the least support in the poll of the remaining candidates.

This poll could have still gone to either Rollins or McAuliffe, depending on the preferences of Henning’s supporters. Rollins was holding a 10 point lead over McAuliffe with Henning’s supporters still counted separately, reinforcing that she may have been the clear favorite of progressives voting at the forum (57% to 43%).

And the winner is...

Rachael Rollins won the RCV straw poll. She picked up 55% of Henning’s votes, giving her a decisive double-digit victory. Thanks to the straw poll, we can be certain that the winner has true majority support -- not merely a plurality win of the "biggest faction".

Full round-by-round results

See full detail of round-by-round results with vote transfers here:

What does this all mean?

Attendees of this forum went all-in with ranking their ballots, proving how easy it is to use Ranked Choice Voting, even while sitting or standing in an auditorium. 93% of voters ranked three or more candidates. Just 4% voted for only a single candidate. Ranked Choice Voting took the pain of strategic voting out of the process, allowing voters to simply record their honest preferences: 1st choice, 2nd choice, 3rd choice.

Rachael Rollins was the decisive favorite at this forum by a strong 56.2% majority, as well as being the top choice among a crowded field, with 40.2% of the first-choice preferences. What we see from the first and second round is that Rollins had about 52-57% of the vote going towards the more progressive leaning candidates, in a poll where 74-77% of participants voted leaning progressive.

This suggests that with multiple self-described progressive candidates running and splitting votes, even if 65% of the general electorate votes progressive, the leader of this group, like Rollins in this poll, could still only come away with vote numbers in the low 30% range. This would mean a loss to a more moderate/conservative candidate like Henning if he consolidates just 35% of the vote. If Henning were to win with a majority of votes, then good for him and good for voters. However, putting candidates in office who are not supported by a majority of voters is not healthy in a democracy. Such vote-splitting is the Achilles' heel of how we elect candidates today using our "pick one and pray" plurality voting system.

There is a better way

The 5-way Suffolk County DA race is a textbook example: our voting system breaks down when similar candidates want to run. If we had Ranked Choice Voting, then similar candidates can freely run. Voters can enjoy the benefits of having more choices. No more pressuring candidates to drop out or pressuring voters to rally around "one" front-runner. No more "lesser of two evils" voting. There would be no more vote-splitting because voters can choose a cluster of similar candidates, and ranked choice voting would choose the strongest of the cluster, and every group of voters with shared interests stays unified. RCV allows every voter to choose their favorite as their 1st choice, then choose backup choices (2nd, 3rd, etc) if their first choice does not have enough support.

About Voter Choice Massachusetts

In light of Maine becoming the first state in US history to enact Ranked Choice Voting; and in light of a profoundly divisive 2016 presidential election season; perhaps you, like us, recognize how much a better voting system is needed right now. Voters should not be pressured to choose between the “lesser of two evils”. Candidates with the best ideas, not the biggest bank accounts or the insider connections, should have a fair shot at running for office. The people of Massachusetts deserve better -- we deserve ranked choice voting.

At 10,000 members and over 200 active volunteers, Voter Choice Massachusetts is the largest and most active statewide Ranked Choice Voting movement in the nation right now (besides Maine); but we still need your help to win Ranked Choice Voting statewide. That's where you come in. Volunteer. Donate. Connect with us online.

***Disclaimer Detail

  • Voter Choice Massachusetts is a strictly non-partisan organization and does not endorse any candidate, party, or partisan ideology.
  • This poll was conducted at an event sponsored by organizations whose supporters lean strongly progressive.
  • The poll was conducted using ranked paper ballots distributed amongst attendees. 82 valid ballots were collected. This is a very small sample size in the context of the race.
  • It is very early in the race. Candidates' messages and voters concerns may still be evolving.
  • Attending a forum like this may result in election participants being more informed than the average voter, making them a non-representative sample.
  • The results may be biased based on the candidates' performances on this particular evening.