One Percent Manual Tally

Lynda Roberts, Registrar of Voters, Elections

Election Code Requirements for One Percent Manual Tally

Definition: "One percent manual tally" is the public process of manually tallying votes in one percent of the precincts, selected at random by the elections official, and in one precinct for each race not included in the randomly selected precincts. This procedure is conducted during the official canvass to verify the accuracy of the automated count. CA Elections Code Sec. 336.5

During the official canvass of every election in which a voting system is used, the official conducting the election shall conduct a public manual tally of the ballots tabulated by those devices cast in 1 percent of the precincts chosen at random by the elections official. If one percent of the precincts should be less than one whole precinct, the tally shall be conducted in one precinct chosen at random by the elections official.

In addition to the one percent count, the elections official shall, for each race not included in the initial group of precincts, count one additional precinct. The manual tally shall apply only to the race not previously counted.

Additional precincts for the manual tally may be selected at the discretion of the elections official. CA Elections Code Sec. 15360.


The one percent manual tally is conducted by the Elections office during the twenty-eight day canvass period after every election. It is a hand count of a random sample of one percent of the precincts in the election. The Registrar of Voters posts a public notice of the date and time of the random selection of precincts and the manual tally five days before the event. The manual tally of ballots begins immediately after the random selection of precincts.

The purpose of the one percent manual tally is to verify the accuracy of the machine count of ballots. After the manual tally of each randomly selected precinct, the elections official compares the manual tally of the ballots in the precinct with the machine tally to verify that both are the same.

Selecting the Random Sample*


The one percent random sample is selected and the manual tally is conducted after all of the ballots in a particular category (precinct or central count ballots) have been counted so that every vote cast in that category has at least a one percent chance of being drawn and counted in the sample for the category.

The categories from which the one percent manual tally are selected and counted are the same as the categories reported by the elections official in the Statement of Vote. These categories are:

  • Ballots counted at the polls on precinct based optical scanners.
  • Ballots counted by central count scanners (mail and provisional ballots).

Method of Selecting the Random Sample

The precincts in the election are numbered with consecutive numbers from 001 up to the actual number of voting precincts used in the election. Precinct one is numbered 001, precinct 2 is numbered 002 and so on until all of the precincts in the election have been numbered.

The random sample of precincts is picked using three 10-sided dice with numerals from 0 to 9. Tossing the three dice generates a 3 digit number from 000 to 999. The dice are red, white and blue (or black) where red is the first number, white is the second number and blue is the third number. The dice are tossed until they show a number that is within the range of numbers of the actual precincts.

For every contest not included in the original random sample, another sample is drawn as follows: the precincts within each of the ballots types not included in the original sample are numbered consecutively and the dice are thrown again to select the precinct for the sample. This process is repeated until precincts representing all of the contests have been selected.

Procedure for the Manual Tally

After the precincts for the manual tally are selected, a board of four people counts them using tally sheets.

Preparing for the Tally

  1. Record the precinct number of the precinct to be recounted on the top of the tally sheet and on the sample ballot and initial both.
  2. Write the name of the contest and the names of the candidates in each contest to be recounted on the tally sheets.
  3. Start with the first contest and separate the ballots for each candidate into stacks of 10.
  4. Separate out overvotes and undervotes in the contest and put these ballots into separate stacks.
  5. Count all of the ballots to make sure the number of ballots to be manually tallied matches the number of ballots counted by the machine.

Counting the ballots

  1. One board member reads the name of the first candidate in the first contest on each ballot and one board member observes to make sure the reader is calling the vote correctly.
  2. Two board members mark their separate tally sheets as the candidate's name is called out. The marks are hash marks going in the same direction across the first row on the tally sheet. On the second row, the hash marks should go in the opposite direction of the hash marks on the first line (//// and then \\\\).
  3. The board members who are marking the tally sheets call out "10", "20", "30", etc. after each group of ten votes is counted for a candidate or measure. Recount if the count of ten is different between the two board members.
  4. When finished counting each contest, the supervisor checks the manual tally against the results for that contest on the summary report. When the supervisor verifies that the tallies match, he/she initials the tally. The counters write the total for each candidate in long hand on the tally sheet.
  5. After the votes for the first candidate are counted and the results have been checked, the board re-sorts the ballots for the second candidate and continues counting votes for that candidate and the rest of the candidates in the contest in the same manner.

Counting Contests Where the Number to Vote for is More Than One

When the number to vote for in a contest is more than one, sort the ballots for the first candidate and count the first candidate. Then resort the ballots for the second candidate and count the second candidate and so on until all of the candidates have been counted. Supervisor should check the tally after each candidate is counted in case the ballots have to be recounted.

Handling Discrepancies

Should a discrepancy between the manual tally and the machine count occur, the ballots should be manually recounted until two manual tallies match the machine tally.

If the machine tally cannot be reconciled with the hand count, and there is no reasonable explanation for the discrepancy, the Registrar of Voters will take additional steps as needed to confirm that the count is accurate. This may include recounting all ballots on another machine and performing a second audit of the new machine tallies.

Completing the Manual Tally

When the manual tally is completed, the Registrar of Voters confirms that the machine count and the manual tally results match and that discrepancies have been resolved.

The ballots and results of the manual tally are then placed into containers, sealed and kept for the required amount of time for that election.

Rev. 6/8/2011

* The method of selecting the random sample for the one percent manual tally described in these procedures was developed by David Wagner, UC Berkeley Professor of Computer Science, Arel Cardoso, UC Berkeley graduate student, Nancy Bickel, Berkeley League of Women Voters and Judy Bertelson, Voting Rights Task Force who worked on this project as a subcommittee of the Alameda County Registrar of Voters' Election Advisory Committee in 2005-06.