Fact check: ‘133 million registered voters’ argument uses flawed logic

Posts circulating on social media allege that “only 133 million registered voters voted” to falsely claim that the number of votes cast for Joe Biden is not mathematically possible. The calculation on which this claim is based on is misleading. The nationwide voter turnout is usually calculated using the eligible-voting population as a denominator, not the number of registered voters. Around 239,247,182 people were eligible to vote in 2020. While there is no available figure yet for nationwide registered voters for 2020, Reuters calculations found there were an estimated 206,557,583 registered voters as of the publication of this article.

Reuters Fact Check. REUTERS

Some iterations feature a screenshot of a tweet by @MSMFactChecking on Dec. 8 here ( archived version ). The tweet reads: “Donald Trump got 74 million votes and (SIC) There are 133 million registered voters in the USA (SIC) If every single registered voter went out and voted there would only be 59 million votes left for Biden. How did he get 81 million votes?”

In a subsequent tweet, the account explained here how this number was “calculated” based on “registered voters” and “turnout”. The tweet reads: “just to clarify there were 213,799,485 registered voters in 2020 of which conflicting sources say turnout was between 60% - 65%. I averaged it out at 62.5% which means just over 133,500,000 registered voters voted. 74,000,000 Trump + 81,000,000 Biden =155,000,000”


Reuters could not find a source for the alleged number of registered voters mentioned (213,799,485), nor a nation-wide number of registered voters for 2020. According to the latest published information by the U.S. Census Bureau, based on their Voting and Registration Supplement to the Current Population Survey (CPS), ( here , table 4.A), in 2018 there were around 153,066,000 registered voters in the U.S.

The U.S. Elections Project explains here why the data from  the CPS might show more limited results.

For 2020, Reuters’ calculations found there were an estimated of 206,557,583 registered voters (found by adding the available voter registration statistics for the District of Columbia and all other states, except North Dakota which does not require voter registration, see here ).


The claim uses an average of 62.5% for the different “reported turnouts” to calculate the figure of the “133,500,000 registered voters” that voted. This percentage is lower than the voter turnout reported by multiple media outlets, including Reuters (estimated at 66.9% by Edison Research, here ) , the Washington Post (66.3% , here ), the New York Times ( 66.7%, here ) and CNBC (66.8%, here ).

Again, this is calculated with the voting-eligible population (VEP) in mind, not the total number of registered voters in the country (as this claim does).

MIT elections lab explains here why the VEP is the preferred denominator to calculate the voter turnout.

Using the number of registered voters as a denominator for calculating turnout nationwide is misleading as registration requirements vary within states (here) . North Dakota, for example, does not require voter registration (here) and some states allow registration on Election Day (here) .

According to the U.S. Election Project, a nonpartisan website run by Michael McDonald of the University of Florida , there were around 239,247,182 people that were eligible to vote in 2020, based on age, citizenship, and not considering felons (according to state law).

Out of the over 239 million eligible voters, the U.S. Election project reported that a total of 159,693,981 ballots were counted in the Nov. 3 election. This represents a VEP turnout of 66.7%, per their calculations.

According to the data by Edison Research reported by Reuters here , Biden received 81,283,098 votes (51.3%), while Trump received 74,222,958 votes (46.8%).

These posts feed into a false narrative pushed by the Trump campaign of widespread voter fraud during the presidential election (here) . U.S. election security officials have said the election was “the most secure in American history” (here   ,  here) .


False. Over 159 million people voted in the Nov. 3 election. The claim misleadingly uses a registered voters projection to calculate the number of people that voted. This is a misleading: turnout percentage reported by the media is based on the voting-eligible population, not the number of registered voters.

This article was produced by the Reuters Fact Check team. Read more about our work to fact-check social media posts  here  .