The preferred turnout rate numerator is the Total Ballots Counted, which is all ballots election officials counted. Not all jurisdictions report total ballots counted, so a second-best numerator is Highest Office, which in a presidential election year is the total votes for all presidential candidates including write-ins (where reported), and in a non-presidential election may be the vote for the highest turnout statewide office (typically governor) or if no statewide election was held, the sum of the congressional elections. Total ballots counted includes blank and otherwise invalid votes for the highest office. Some states may report a statistic they call total ballots cast, but is actually vote for highest office.
The preferred turnout rates are those calculated with the voting-eligible population as the denominator. The voting-eligible population (VEP) represents an estimate of persons eligible to vote regardless of voter registration status in an election and is constructed by modifying the voting-age population (VAP), by components reported in the right-most columns (scroll right in the spreadsheet).
Please see the FAQ for information on the construction of these statistics.
The denominator reflects the July 1, 2010 and July 1, 2011 voting-age population estimates extrapolated to Nov. 2012, non-citizen estimates from the 2011 American Community Survey, the year-end 2010 DOJ prison report and the year-end 2010 DOJ probation and parole report. Starting in 2010, I report the actual number of felons on probation with no estimated adjustment. Note, a '0' indicates that either a state does not disfranchise a class of felons or the state does not incarcerate felons within their borders. See The Sentencing Project for more information.
The overseas eligible population estimate is calculated from 2011 overseas civilian estimate from the State Department -- which reports 6.3 million overseas citizens in 2011 and is deflated by 75.2%, the percentage of the domestic citizen population that is of voting age according to the 2011 American Community Survey. This estimate includes military personnel. (Note a previous lower estimate of 2.1 million overseas civilians was derived from a Federal Voting Assistance Program report of the overseas population and from Department of Defense Manpower Reports; this revised estimate is more consistent with historical trends.)The national total number of ballots counted is an estimate derived from the national vote for president multiplied by the ratio of total ballots counted to vote for president in states reporting both numbers. The total ballots counted estimate is updated as additional states report.
Kansas votes for highest office adjusted for a double-count of Wilson County votes for president. Total ballots counted is unaffected.
These turnout rate estimates are updated when new data are available. Such data may be revised components of the voting-age and voting-eligible population, and any revisions to vote totals, as warranted.
Again, please see the FAQ for more information.