Survey and transcript analyses show turnout patterns and majority desire for more information, committee says

At the May 6 meeting, the committee reviewed survey findings and preliminary transcript‑based analyses of how long warrant articles take at town meetings in Concord and other communities.

The committee’s survey analysis showed that regular Town Meeting attendees (the “regulars”) are generally more satisfied with Town Meeting as currently conducted, while those who attend less often expressed more openness to change. One committee member summarized: “The people who never attend and the people who attend most or all are actually about the same number…this underscores that the survey overrepresents people who are involved in town government.” Committee members emphasized that the survey was not a random sample of all registered voters and therefore overrepresents engaged participants.

A parallel analysis used publicly available meeting videos and automated transcripts to estimate article durations in other towns (Longmeadow, North Andover, Andover, Bedford, Acton, Carlisle and Lincoln). Eric Moore described the method: he downloaded YouTube transcripts and wrote code to identify when articles began and ended. The cross‑town results showed wide differences: in some communities a large share of articles are disposed of in under five minutes, while Concord spends more time on typical articles. The committee agreed that those comparative timing charts are interesting reference material but should be presented as supplemental context, not as used sole justification for the committee’s recommendations.

Committee members asked for clearer visuals and a short narrative describing what the data do and do not show. The chair said he would produce more accessible charts and a concise writeup tying the survey and timing analyses to the committee’s recommendations. “Everyone wants more information about what is happening with local government,” one member observed; the committee agreed to publish the survey narrative and dataset as an appendix.

Limitations noted in the meeting: the survey sample is self‑selecting and not representative of all registered voters; transcript timing comparisons are a work in progress and do not capture local customs and historic expectations that also shape meeting pace.

Next steps: refine visuals and narrative, publish the October 24 survey report and dataset as appendices, and label comparative timing analyses as supplemental reference material on the committee web page.