2016 Massachusetts State Election Audit Report

This is the post-election audit report required by Massachusetts General Laws Chapter 54, section 109A.

Overview

As required by law, 3% of all Massachusetts precincts were audited after the 2016 State Election. Of the 350,241 votes that were tallied for the required offices and the chosen ballot question during the audit, there was a change of 2,946 or 0.84% from the votes tallied on election night and the hand-count audit. With a total of 4 candidates on the ballot, and 5 certified write-ins, there was a 1.20% difference between votes tallied on election night and votes hand counted during the audit for the office of President and Vice President. For the ballot question that was audited, there was a 0.55% difference between the votes tallied on election night and the votes hand counted during the audit. Of the rest of the offices audited (Representative in Congress, Senator in the General Court, and Representative in the General Court) the differences between votes tallied on election night and votes tallied during the audit were 0.85%, 0.72%, and 0.77% respectively. Following, is a breakdown of those offices.

Office of President and Vice President

Almost half (48.50%) of the differences found in this particular office came from “All others,” “Blanks,” and the five (5) “Certified Write-In” candidates. Some votes appear to have been recorded as “All others” on election night, but were later attributed to a “Certified Write-In” candidate during the audit. The remaining difference, or a total of 0.62% of the total votes cast in the office of President and Vice President, came from differences in the votes tallied for candidates who were on the ballot in the 2016 election.

Office of Representative in Congress

40.92% of the differences found in this particular office came from “All others” and “Blanks.” A few votes appear to have been recorded as “Blanks” on election night, but were later attributed to “All others” or candidates on the ballot during the audit. As indicated in the “Clerk Notes” section of this report, clerks occasionally found names that had check marks, circled names, or other markings not read by the voting machines on election night. The remaining difference, or a total of 0.50% of the total votes cast in the office of Representative in Congress, came from differences in the votes tallied for candidates who were on the ballot in the 2016 election.

Office of Senator in the General Court

34.39% of the differences found in this particular office came from “All others” and “Blanks.” The remaining difference, or a total of 0.47% of the total votes cast in the office of Senator in the General Court, came from differences in the votes tallied for candidates who were on the ballot in the 2016 election.

Office of Representative in the General Court

53.98% of the differences found in this particular office came from “All others” and “Blanks.” The remaining difference, or a total of 0.36% of the total votes cast in the office of Representative in the General Court, came from differences in the votes tallied for candidates who were on the ballot in the 2016 election.

Ballot Question

29.07% of the differences found in the audit for the ballot question came from “Blanks.” The remaining difference, or a total of 0.39% of the total votes cast for the ballot question, came from differences in the votes tallied for “Yes” and “No” for the audited question that was on the ballot in the 2016 election.

Clerk Notes

The clerks that participated in the audit offered several theories as to what could have caused the 0.84% difference between ballots tallied on election night and ballots counted during the audit. Many of the local election officials believed that some of the difference may have come from human error during the audit. In particular, local election officials believed that at least some of the discrepancies came from different judgement calls made during the audit than on election night by poll workers. Another reason cited for the difference was the occasional mismarking of the ballot by the voter or failure to follow instructions. As an example, some clerks cited that names of candidates were sometimes circled or underlined, rather than the vote indicator being completed. Similarly, some of the vote indicators had check marks or an “X” to indicate a vote, which may not be picked up by the machines. Some clerks also commented that the increased number of absentee combined with the early ballots, which are both pre-scored, filled the ballot boxes quickly as they do not lie flat when deposited into the tabulator.

Conclusion

Of all votes audited, 99.16% of the votes recorded on Election Day were the same as during the audit. Nevertheless, we will continue to work with local election officials throughout the Commonwealth to ensure that the accuracy of votes tallied on election night is at its highest levels.

Download the 2016 Post-Election Audit Report (XLSX)

Download the Post-Election Audit Narrative Report (PDF)