2024 Post-Election Audit Report Narrative
In accordance with the provisions of Section 109A of Chapter 54 of the General Laws, a post-election audit was conducted of the ballots cast in 3% of all precincts in Massachusetts following the November 5, 2024 State Election.
As required by law, a public, non-computerized, random drawing was held on November 7, 2024 to select the precincts to be audited. City and town election officials for the selected precincts were notified of the post-election audit on November 7, 2024.
In total, 72 precincts and sub-precincts were drawn at random. Due to the manner in which ballots were stored in the Town of Dartmouth, resulting in the comingling of ballots from two precincts, one of the selected precincts could not be audited. As such, 71 precincts are included in this report.
Those municipalities posted their audit times and locations in accordance with law and the public post-election audits were held November 8-12.
To conduct the post-election audit, local election officials oversaw a hand count of all ballots that had been counted through November 5, 2024. As required by law, the offices of Electors for President and Vice President, Senator in Congress, Representative in Congress, Senator in General Court, and Representative in General Court were counted in each precinct, if more than one candidate was printed on the ballot. A statewide ballot question, Question #2, also randomly selected during the drawing, was counted as well.
The results of the post-election audit hand counts were reported by local election officials to the Secretary of the Commonwealth’s Elections Division, along with the original tallies from Election Night for comparison.
An analysis of the post-election audit of the November 5, 2024 State Election is detailed below.
One Ashburton Place, Room 1705, Boston, Massachusetts 02108 • (617) 727-2828
TOTAL BALLOTS CAST
Among the 71 precincts audited, 99,744 cast ballots were counted on Election Night. Local election officials reported that they counted 99,788 ballots during the post-election audit, which resulted in 44 additional ballots being counted among all 71 precincts. The additional 44 ballots account for 0.04% of ballots audited.
Of the 71 precincts audited, 54 reported no changes in the number of ballots cast, while 17 precincts reported a difference, likely due to tabulator jams and poll worker error.
Of the 17 precincts reporting a difference in the total number of ballots cast, 15 had changes of fewer than 10 votes. The two precincts where the number of additional ballots was greater than 10 were both located in the City of Boston, where officials reported that the increase was due to the failure of poll workers in those precincts to count a small number of ballots not read by the tabulator. These ballots should have been hand-counted by poll workers after the close of polls and recorded on a paper tally sheet.
ELECTORS FOR PRESIDENT AND VICE PRESIDENT
Among the 71 hand-counted precincts, the electors for Ayyadurai and Ellis received 5 fewer votes as a result of the post-election audit; electors for De La Cruz and Garcia received 10 additional votes; electors for Harris and Walz received 77 additional votes; electors for Oliver and ter Maat received 2 fewer votes; electors for Stein and Caballero-Roca received 12 additional votes; and electors for Trump and Vance received 3 fewer votes than those counted on Election Night.
The category of “All Others,” which includes write-in votes, received 47 fewer votes than those counted on Election Night, likely due to tellers in the audit interpreting the intent of the voter differently in a manner that was different than the interpretation of the poll workers on Election Night.
Among all audited precincts, there were 14 fewer blanks than reported on Election Night and 16 additional overvotes. While certain tabulators report overvotes separately from blank votes, other tabulators include overvotes among the blanks, as they are reported in official results. This indicates that the tabulators were accurately counting votes for more than one candidate as blank votes or overvotes, and that the actual increase of overvotes was only 2.
SENATOR IN CONGRESS
Among all audited precincts, Elizabeth Ann Warren received an additional 113 votes as a result of the post-election audit, while John Deaton received 5 votes fewer. The number of votes reported on Election Night as “All Others” increased by 10 and the blank votes were reduced by 79, with five additional overvotes.
The largest increase in votes for Warren occurred in the two Boston precincts that had previously failed to include the hand-counted ballots in their Election Night totals. Other changes appear to be primarily attributable to the interpretation of audit tellers of ballots that tabulators had read as blank votes. This can occur when voters mark their ballots in a manner that is contrary to instructions, such as marking the ballot outside of oval or making marks on more than one oval for the same office.
REPRESENTATIVE IN CONGRESS
Precincts in 4 of the Commonwealth’s 9 congressional districts audited the results for Representative in Congress, with other districts listing only one candidate on the ballot for this office, and therefore not subject to the audit.
In all 4 of the audited districts, candidate vote totals changed by fewer than 10 votes, as did the total number of votes cast.
SENATOR IN GENERAL COURT
In the 71 precincts audited, 10 State Senate districts were contested and therefore subject to the audit.
Four districts had no changes in the total number of votes cast, while all districts had a difference of 3 or fewer ballots cast districtwide. The largest change in the number of votes cast for a candidate was 8, in the Berkshire, Hampden, Franklin and Hampshire District, which includes 2 towns that do not use ballot tabulators.
REPRESENTATIVE IN GENERAL COURT
In the 71 precincts audited, 14 State Representative districts were contested and therefore subject to the audit.
Ten of the 14 districts reported no changes in the total number of ballots cast, while the other 4 districts reported changes of 3 ballots or fewer. The largest vote total change for any candidate was 9 votes, in the Third Berkshire District. The clerk in one town in that district reported that discrepancies were due to the clerk’s error in transferring numbers from one tally sheet to another.
QUESTION #2
Among the precincts audited, there were 76 additional “Yes” votes, 93 additional “No” votes, and 135 fewer blanks. Ten overvotes were also reported, suggesting that some voters did not mark their ballots according to instructions.
The increase in vote totals for the “Yes” and “No” options and the corresponding decrease in blanks for this ballot question suggest that audit tellers were able to interpret markings that the tabulators had interpreted as blanks.
CONCLUSION
Local election officials reported in their audit reports that nearly all discrepancies between the Election Night count and audit results could be attributed to “human error.” Most often, the human error was made by poll workers in tallying hand-counted ballots, calculating totals on tally sheets, or interpreting write-in votes. In a few precincts, where the number of ballots increased, it appears that poll workers erred in failing to tally ballots contained in the handcount compartment of the ballot box.
It appears that tabulators used in the November 5, 2024 election counted ballots accurately when voters marked ballots according to instructions, with some communities that used tabulators reporting fewer changes than the communities that originally counted all ballots by hand on Election Night.
Based on these conclusions, the Elections Division will be providing additional training to local election officials regarding the tallying and reporting of votes, so that those officials are best prepared to instruct their poll workers.