Editing Vote counting (section)

===Errors in manual counts===
Hand counting has been found to be slower and more prone to error than other counting methods.<ref>{{Cite AV media |url=https://apnews.com/video/donald-trump-georgia-district-of-columbia-national-cb651acf448449f687e12d25c0239994 |title=Explaining Election Day: How hand counting votes carries risks |language=en |access-date=2024-10-10 |via=apnews.com}}</ref>

Repeated tests have found that the tedious and repetitive nature of hand counting leads to a loss of focus and accuracy over time. A 2023 test in [[Mohave County, Arizona]] used 850 ballots, averaging 36 contests each, that had been machine-counted many times. The hand count used seven experienced poll workers: one reader with two watchers, and two talliers with two watchers.

The results included 46 errors not noticed by the counting team, including:
*Caller called the wrong candidate, and both watchers failed to notice the incorrect call
*Tally markers tried to work out inconsistencies while tallying
*Tally markers marked a vote for an incorrect candidate and the watchers failed to notice the error
*Caller calling too fast resulted in double marking a candidate or missed marking a candidate
*Caller missed calling a vote for a candidate and both watchers failed to notice the omission
*Watchers not watching the process due to boredom or fatigue
*Illegible tally marking caused incorrect tally totaling
*Enunciation of names caused incorrect candidate tally
*Using incorrect precinct tally sheets to tally ballots resulted in incorrect precinct level results.<ref name="mohave" />

Similar tallying errors were reported in Indiana and Texas election hand counts. Errors were 3% to 27% for various candidates in a 2016 Indiana race, because the tally sheet labels misled officials into over-counting groups of five tally marks, and officials sometimes omitted absentee ballots or double-counted ballots.<ref name="indiana">{{Cite news |last=BEILMAN |first=ELIZABETH |url=https://www.newsandtribune.com/news/jeffersonville-city-council-at-large-recount-tally-sheets-show-vote/article_75f432ce-cf7e-11e5-8c1a-5365ef7d3540.html |title=Jeffersonville City Council At-large recount tally sheets show vote differences|work=News and Tribune (Jeffersonville, IN)|access-date=2020-02-14 |language=en}}</ref> 
12 of 13 precincts in the 2024 Republican primary in Gillespie County, TX, were added or written down wrong after a hand count, including two precincts with seven contests wrong and one with six contests wrong.<ref name="vbg">{{Cite news |last=Contreras |first=Natalia |date=2024-03-18 |title=Texas county's GOP officials declared hand count a success, but kept finding errors |url=https://www.votebeat.org/texas/2024/03/18/gillespie-county-texas-republican-primary-hand-count-election-errors-discrepancies/ |access-date=2024-03-21 |work=Votebeat |language=en}}</ref>
While the Texas errors were caught and corrected before results were finalized, the Indiana errors were not.
 
Average errors in hand-counted candidate tallies in New Hampshire towns were 2.5% in 2002, including one town with errors up to 20%. Omitting that town cut the average error to 0.87%. Only the net result for each candidate in each town could be measured, by assuming the careful manual recount was fully accurate. Total error can be higher if there were countervailing errors hidden in the net result, but net error in the overall electorate is what determines winners.<ref name="nh">{{Cite web |url=https://dspace.mit.edu/bitstream/handle/1721.1/96548/vtp_wp11.pdf |title=Using Recounts to Measure the Accuracy of Vote Tabulations: Evidence from New Hampshire Elections 1946-2002. |last=Ansolabehere |first=Stephen |author2=Andrew Reeves |date=January 2004 |website=CALTECH/MIT Voting Technology Project |access-date=2020-02-14}}</ref> Connecticut towns in 2007 to 2013 had similar errors up to 2%.<ref name="tigran" />

In candidate tallies for precincts in Wisconsin recounted by hand in 2011 and 2016, the average net discrepancy was 0.28% in 2011 and 0.18% in 2016.<ref name="wi-an">{{Cite journal |last1=Ansolabehere |first1=Stephen |last2=Burden |first2=Barry C. |last3=Mayer |first3=Kenneth R. |last4=Stewart |first4=Charles |date=2018-03-20 |title=Learning from Recounts |journal=Election Law Journal: Rules, Politics, and Policy |volume=17 |issue=2 |pages=100–116 |doi=10.1089/elj.2017.0440 |issn=1533-1296|doi-access=free }}</ref>

India hand tallies paper records from a 1.5% sample of election machines before releasing results. For each voter, the machine prints the selected candidate on a slip of paper, displays it to the voter, then drops the slip into a box. In the April–May 2019 elections for the lower house of Parliament, the Lok Sabha, the Election Commission hand-tallied the slips of paper from 20,675 voting machines (out of 1,350,000 machines)<ref name="maha" /> and found discrepancies for 8 machines, usually of four votes or less.<ref name="jain" /> Most machines tally over 16 candidates,<ref name="ques">{{Cite news |date=2019-03-09 |title=What are EVMs, VVPAT and how safe they are |language=en |work=Times of India |url=https://timesofindia.indiatimes.com/india/what-are-evms-vvpat-and-how-safe-they-are/articleshow/66969640.cms |access-date=2021-11-08}}</ref> and they did not report how many of these candidate tallies were discrepant. They formed investigation teams to report within ten days, were still investigating in November 2019, with no report as of June 2021.<ref name="jain" /><ref name="nath">{{Cite news |last=Nath |first=Damini |date=2019-07-25 |title=ECI sets up teams to probe VVPAT mismatch in Lok Sabha election |language=en-IN |work=The Hindu |url=https://www.thehindu.com/news/national/ec-sets-up-teams-to-probe-vvpat-mismatch-in-lok-sabha-election/article28713704.ece |access-date=2021-11-08 |issn=0971-751X}}</ref> Hand tallies before and after 2019 had a perfect match with machine counts.<ref name="jain" />

An experiment with multiple types of ballots counted by multiple teams found average errors of 0.5% in candidate tallies when one person, watched by another, read to two people tallying independently. Almost all these errors were overcounts. The same ballots had errors of 2.1% in candidate tallies from sort and stack. These errors were equally divided between undercounts and overcounts of the candidates. Optical scan ballots, which were tallied by both methods, averaged 1.87% errors, equally divided between undercounts and overcounts. Since it was an experiment, the true numbers were known. Participants thought that having the candidate names printed in larger type and bolder than the office and party would make hand tallies faster and more accurate.<ref name="goggin"/>

Intentional errors hand tallying election results are fraud. Close review by observers, if allowed, may detect fraud, and the observers may or may not be believed.<ref name="pickles">{{Cite web |url=https://www.gov.uk/government/publications/securing-the-ballot-review-into-electoral-fraud |title=Securing the ballot: review into electoral fraud |last=Pickles |first=Eric |date=2016-12-27 |website=Cabinet Office, UK |language=en |access-date=2020-02-14}}</ref> 
If only one person sees each ballot and reads off its choice, there is no check on that person's mistakes. In the US only Massachusetts and the District of Columbia give anyone but officials a legal right to see ballot marks during hand counting.<ref name="vvstates"/>
If fraud is detected and proven, penalties may be light or delayed. US prosecution policy since the 1980s has been to let fraudulent winners take office and keep office, usually for years, until convicted,<ref name="doj8">{{Cite web |url=https://www.justice.gov/criminal/file/1029066/download |title=Federal Prosecution of Election Offenses Eighth Edition |date=December 2017 |website=US Department of Justice |access-date=2020-02-14}}</ref>
<ref name="doj-compare">{{Cite web |url=http://site.votewell.net/a/doj-elections.htm |title=Federal Prosecution of Election Offenses |website=votewell.net |access-date=2020-02-14}}</ref>
and to impose sentencing level 8–14,<ref name="level">{{Cite web |url=https://www.ussc.gov/guidelines/2018-guidelines-manual/2018-chapter-2-c |title=2018 Chapter 2 PART C - OFFENSES INVOLVING PUBLIC OFFICIALS AND VIOLATIONS OF FEDERAL ELECTION CAMPAIGN LAWS |date=2018-06-27 |website=United States Sentencing Commission |language=en |access-date=2020-02-14}}</ref> which earns less than two years of prison.<ref name="prison">{{Cite web |url=https://www.ussc.gov/sites/default/files/pdf/guidelines-manual/2011/manual-pdf/Sentencing_Table.pdf |title=Sentencing Table |date=2011-10-26 |website=US Sentencing Commission |access-date=2020-02-14}}</ref>

In 1934, the United States had been hand-counting ballots for over 150 years, and problems were described in a report by Joseph P. Harris, who 20 years later invented a [[Voting machine#Punched card voting|punched card voting]] machine,<ref name="harber">{{Cite web |date=1983 |editor-last=Harris |editor-first=Joseph P. |editor2-last=Nathan |editor2-first=Harriet |title=Joseph P. Harris: Professor and Practitioner: Government, Election Reform, and the Votomatic |url=https://digicoll.lib.berkeley.edu/record/217520?v=pdf |access-date=2024-07-29 |website=UC Berkeley |publisher=Regional Oral History Office}}</ref>
{{blockquote|"Recounts in Chicago and Philadelphia have indicated such wide variations that apparently the precinct officers did not take the trouble to count the ballots at all... While many election boards pride themselves upon their ability to conduct the count rapidly and accurately, as a general rule the count is conducted poorly and slowly... precinct officers conduct the count with practically no supervision whatever... It is impossible to fix the responsibility for errors or frauds... Not infrequently there is a mixup with the ballots and some uncertainty as to which have been counted and which have not... The central count was used some years ago in San Francisco... experience indicated that there is considerable confusion at the central counting place... and that the results are not more accurate than those obtained from the count by the precinct officer."<ref name="harris">{{Cite web |last=Harris |first=Joseph P. |date=1934 |title=Election Administration in the United States, chapter VI, pages 236-246 |url=https://www.nist.gov/itl/election-administration-united-states-1934-joseph-p-harris-phd |access-date=2024-07-29 |website=NIST, originally published by Brookings |language=en}}</ref>}}

{| class="wikitable sortable" style="font-size: 95%; text-align: center; "
|+ class="nowrap" | Errors in hand-counted tallies for candidates
|-
!Place||Year||Candidate tally errors, as % of votes counted||Reference Standard||Notes
|-
|New Hampshire towns||1946-1962||0.83%||careful hand recount||wtd avg is sum of absolute values of errors, divided by total ballots<ref name="nh"/>
|-
|New Hampshire towns||2002||2.49%||careful hand recount||20% in one town; others average 0.87%<ref name="nh"/>
|-
|Connecticut towns||2007-2013||up to 2%||investigations of differences between hand & machine counts||"routinely show up to 2% error"<ref name="tigran">{{Cite journal |last=Antonyan |first=Tigran |display-authors=etal |date=2013-06-21 |title=Computer Assisted Post Election Audits |url=https://voter.engr.uconn.edu/voter/wp-content/uploads/AS-2013.pdf |journal=State Certification Testing of Voting Systems National Conference |via=University of Connecticut}}</ref>
|-
|Experiment, optical scan style ballots||2011||1.87%||Known values in experiment||As % of all 120 ballots, not candidate's ballots<ref name="goggin"/>
|-
|Experiment, read to talliers||2011||0.48%||Known values in experiment||As % of all 120 ballots, not candidate's ballots<ref name="goggin"/>
|-
|Experiment, sort & stack||2011||2.13%||Known values in experiment||As % of all 120 ballots, not candidate's ballots<ref name="goggin"/>
|-
|Wisconsin precincts||2011||0.28%||careful hand recount||Table 6 "0.59% of the ballots" "out of 3,019" where 3,019 is total number of ballots<ref name="wi-an"/>
|-
|Wisconsin precincts||2016||0.18%||careful hand recount||Table 7a. "0.59% of the ballots" but 0.18% if exclude write-ins<ref name="wi-an"/>
|-
|Indiana, Jeffersonville||2016||3%-27%||Newspaper tally||Over-counted groups of 5 tally marks, and omitted or double-counted groups of ballots<ref name="indiana"/>
|-
|Colorado audits||2018||0.8%||Consensus between election computer & Sec of State staff||Errors by audit boards in determining voter intent on individual ballots. No manual totals done.<ref name="yrs3">{{Cite web |date=2020-04-16 |title=Colorado Risk Limiting Audits: Three Years In |url=https://www.sos.state.co.us/pubs/elections/RLA/files/OverviewThreeYearsIn.pdf |access-date=2021-10-27 |website=Colorado Sec. of State}}</ref>
|-
|Colorado audits||2019||0.2%||Consensus between election computer & Sec of State staff||Errors by audit boards in determining voter intent on individual ballots. No manual totals done.<ref name="yrs3"/>
|-
|India national election audit||2019||8<ref name="jain">{{Cite news |last1=Jain |first1=Bharti|date=2021-06-03 |title=Tallying of VVPAT slips and EVM count in constituencies that went to polls recently throw up 100% match|language=en |work=Times of India |url=https://timesofindia.indiatimes.com/india/tallying-of-vvpat-slips-and-evm-count-in-constituencies-that-went-to-polls-recently-throw-up-100-match/articleshow/83192568.cms |access-date=2021-11-08}}</ref> of 20,625 machines audited<ref name="maha">{{Cite news |last1=Mahapatra |first1=Dhananjay|date=2019-04-09 |title=Supreme Court: Count VVPAT slips of 5 booths in each assembly seat |language=en |work=Times of India |url=https://timesofindia.indiatimes.com/india/count-vvpat-slips-of-5-booths-in-each-assembly-seat-sc/articleshow/68786810.cms |access-date=2021-11-08}}</ref>||Discrepancy between hand tally of [[VVPAT]]s & election computers||They investigated and have not released analysis, so it is not clear how many of these were errors in hand tally.<ref name="jain"/>
|-
|Colorado audits||2020||0.6%||Consensus between election computer & Sec of State staff||Errors by audit boards in determining voter intent on individual ballots. No manual totals done.<ref name="co2020">{{Cite web |date=2020-11-25 |title=2020 General Election Risk-limiting Audit Discrepancy Report |url=https://www.sos.state.co.us/pubs/elections/RLA/2020/general/DiscrepancyReport.pdf |access-date=2021-10-27 |website=Colorado Sec of State}}</ref>
|-
|[[2021 Maricopa County presidential ballot audit|Maricopa County, AZ audit]]||2021||15%||Paper-counting machine||Audit & machine count were contracted by state Senate<ref name=anglen>{{Cite news |last=Anglen |first=Robert |date=2021-10-12 |title=New Arizona audit review shows Cyber Ninjas' ballot count off by 312K |language=en-US |work=Arizona Republic |url=https://roselawgroupreporter.com/2021/10/new-arizona-audit-review-shows-cyber-ninjas-ballot-count-off-by-312k/ |access-date=2022-07-29}} [https://archive.today/20211012210419/https://www.azcentral.com/story/news/politics/elections/2021/10/11/new-arizona-audit-review-shows-cyber-ninjas-ballot-count-off-312-k/6094144001/ Full article]</ref><ref name="scan-az">{{Cite web |date=2021-10-08 |title=Scan of Pullen Pallets Binders 1-45 |url=https://statecraftlaw.app.box.com/v/senateauditpublicreadingroom/file/870446599996 |access-date=2022-07-29 |website=Statecraftlaw}}</ref>
|-
|Mohave County, AZ, experiment||2023||0.15%||Logic & Accuracy Test ballots||46 errors were 0.15% of 30,600 contest totals on 850 test ballots.<ref name="mohave">{{Cite web |date=2023-07-25 |title=Ballot Hand Tally |url=https://resources.mohave.gov/Repository/Calendar/08_01_2023BOSAgenda0fe47379-660b-465f-a8b5-4eb9fc976f30.pdf |access-date=2024-03-02 |website=Mohave County Board of Supervisors}}</ref>
|}
Data in the table are comparable, because average error in candidate tallies as percent of candidate tallies, weighted by number of votes for each candidate (in NH) is mathematically the same as the sum of absolute values of errors in each candidate's tally, as percent of all ballots (in other studies).