At both open town meetings and elected town meetings–like Brookline’s–counted votes are a longstanding custom, sometimes called “dividing the house.” Until recently, Brookline’s approach was for the moderator to appoint a few town meeting members as “tellers”–usually six–who would count the votes.

If unable to call a vote by scanning the hall–or if seven town meeting members rose to “doubt the vote”–the moderator would organize a standing, counted vote. The tellers would count those standing to vote “yes,” then those standing to vote “no,” then those standing to vote “present” (also called “abstain”). Typically, a counted vote would take at least five minutes. In our traditions, legislative voting is a public act, whether or not detailed records are kept. Tellers had an unusual privilege: their own votes were invisible to the public.

Roll-call votes: When first elected moderator in 1970, Justin Wyner began roll-call votes, at the request of 35 or more town meeting members. Roll call had been fairly common in state legislatures but not in town meetings. Mr. Wyner would call the names of town meeting members and confirm their votes by repeating them aloud. The town clerk would record the votes and make lists available to the public. A roll-call vote took at least 20 minutes; there was rarely more than one each town meeting.

Automation for recording votes began at a 2012 fall town meeting, a contribution to “open government” for Brookline. Each town meeting member uses a small wireless device. Edward “Sandy” Gadsby, the current moderator, sets low thresholds for a recorded vote: seven town meeting members who “doubt the vote” he calls after a show of hands but 35 town meeting members if a recorded vote is requested in advance. He also initiates an electronic vote when he says he is unable to call a vote by scanning the hall–doing so instead of conducting a standing, counted vote.

The current approach to electronic voting allows 40 seconds, recommended by a 2012 moderator’s committee, during which votes appear on a projection screen as town meeting members cast them. Then Mr. Gadsby reads the totals and declares whether or not a motion has passed. Most records of the electronic votes appear fairly promptly on Brookline’s municipal Web site, as computer files called “Electronic Recorded Votes.” Some electronic votes are not being shown in the “roll call vote” computer files available from the Brookline municipal Web site–a dubious practice. Publicly available video recordings of town meeting sessions show individual votes being acquired during such events. The public collection and use of those votes at town meetings strongly suggests that all the electronic votes are public records under Massachusetts laws.

Problems, uncertain votes: Some obvious and some hidden problems occurred this year. The voting system appeared to malfunction during the vote for Article 22 on May 29; Mr. Gadsby called for that vote to be repeated. It appeared the system might continue recording after a 40-second voting period had ended. Changes would show on the projection screen. Mr. Gadsby apparently tried to take totals at 40 seconds in order to declare the votes, but the electronic system seemed to be confusing him. On May 29, he called out corrections to the town clerk for an amendment to Article 10 and for the main motion on Article 23, correcting the vote on the amendment twice.

Comparing “Electronic Recorded Votes” files found the day after each of the first two sessions with the “Combined Electronic Recorded Votes” file found six days after the close of the town meeting showed uncertain votes. Four votes by Precinct 4 town meeting members shown in the file found the morning after the May 27 session differed from corresponding votes shown in the “combined” file, inspected six days after the close of the town meeting. The first two changed from no vote being recorded to a Yes vote; the third from No to Yes and the fourth from Yes to No.

Totals of the votes taken from the “Combined Electronic Recorded Votes” were discrepant with both the totals listed at the end of that file and the totals announced at town meeting–used to decide whether motions passed or failed. Differences of up to 3 votes occur between the Yes or No votes declared at town meeting and totals of Yes or No votes found in the “Combined Electronic Recorded Votes” file. None of the uncertain and discrepant votes would have been enough to change a result at this town meeting, but mistakes could cloud results of closer votes.

Absent town meeting members, missing votes: Some precincts had much larger fractions of missing votes than others–that is, votes which could have been cast by town meeting members who were absent at the time of a vote or who did not vote. Based on “Combined Electronic Recorded Votes” as found today, the at-large town meeting members and those from Precinct 6 had the lowest fractions of missing votes: 8 and 9 percent. Town meeting members from precinct 15 had the highest fraction of missing votes: 55 percent. The average among the precincts was 24 percent missing votes.

A table of recorded votes following this article comes from “Combined Electronic Recorded Votes” as found today. Votes are indicated as Y for “yes,” N for “no,” P for “present” or “abstain,” and A for absent or no vote recorded. An ampersand (&) marks an uncertain vote, differing between sources examined. From the A entries for the 4-1/2 hour second session, it looks as though town meeting members started to leave around 9:30 pm. The moderator and the Board of Selectmen might consider scheduling more sessions with shorter durations. This year’s twelve total hours of an annual town meeting, for example, might have run as four sessions of three hours each.

– Craig Bolon, Brookline, MA, June 9, 2014

Addendum, June 11, 2014. A written public records request was filed with Town Clerk Patrick Ward today for copies of the records of two electronic votes that were conducted and recorded during the third session of the 2014 annual town meeting, on June 2. Copies of the public records request have been sent to the Supervisor of Public Records in the office of the Secretary of State and to Town Administrator Mel Kleckner. Those were votes on motions to refer, under Articles 26 and 29. Records of the votes were collected by elected officials at a public meeting and used for a public purpose. However, the votes by individual town meeting members do not currently appear in the “Combined Electronic Recorded Votes” file for this town meeting, available on Brookline’s municipal Web site as of June 9, 2014.

Brookline 2014 annual town meeting, recorded votes as of June 9, 2014