30 October, 2002


First, a disclaimer: I don't do this sort of thing professionally. Everything I list below might be entirely wrong. I take no responsibility if you use this information as gospel, without researching it on your own. Got that? Good.

This is what I've learned after researching the whole election/electoral college thing, using data from a few government sites.

Under the electoral college:

  • There are 538 electoral votes up for grabs.
  • Of these, 270 are needed to win.
  • The top 11 states will bring in the needed 270 votes. These include CA, NY, TX, FL, PA, IL, OH, MI, NJ and NC, plus either VA or GA.
  • There were 116,536,000 (57,777,007) voters in these 11 states.
  • Under the "winner claims all" policy, 58,384,536 (28,946,281) votes would give a candidate all 270 votes.
  • This is a mere 28.4% (14.1%) of the voter population.

Under the popular vote:

  • There were 205,815,000 (105,586,274) votes up for grabs.
  • A simple majority (50.1%) of the votes needed: 103,113,315 (52,898,723).
  • A mere 25.7% of the eligible voters could elect the president.

Just to make some sense, the big numbers in bold represent the eligible voters, while the number next to them, (in parantheses), represent the actual turnout in the 2000 election.

Basically, under the electoral college, 28.9 million people could elect the president. Which is kinda frightening. Voter apathy, of course, could cause this to happen in a popular election. I realize this, and hence some of my concerns in my previous entry.

Download my data as an Excel spreadsheet, and figure out whether or not I'm smoking crack while looking at this stuff. (Yes, I know, Excel is evil. But easy to use.)

Voting figures are based on the 2000 election. Population based on 2000 census. Electoral votes based on 1997 data. Voter turnout for 2000 was approximately 51.3%


  1. This is assuming a two party election, what happens to the numbers if a third party canidate makes the ballot in a majority of the 11 'important' states?

  2. The numbers start to get screwy when you add a third, fourth, etc candidate to the ballot. In a two-party system, you only need 50% + 1 of the ballots cast to win. However, with three parties, it's not 33% + 1.If you assume a fairly even distribution of votes between three candidates, then you're looking at approximately 34% of the vote to win. But how often does that happen? In all likelihood, you'll end up with a third party taking 5-15% of the vote, at least until Americans are over the two-party paradigm. If they take 10% of the vote, then you, as one of the two "major" parties, will need 45% + 1 to win the whole thing. But at least the third party does add an element of chance. You can't really figure out exactly how many are needed, until you know what percentage the third party will get.

  3. The algorithm is actually slightly flawed. You don't need 50% of the voters in (CA + NY + TX ...) +1 votes to take all 270 votes. You need (50% of CA + 1) + (50% of NY + 1) etc.Including VA as the 12th needed state (it has less voters) that gives you 58,268,011 (287,888,515). In other words even less people than you had calculated. ;-)If you start from the other side (states with least population) MA - DC in the spread sheet and add New Jersey you also get to 270. Is a politician was to win this set of states he/she would only need 46,984,540 (24,206,682) votes to become the President of the U.S.A.So it only takes roughly 22.9% of the actual voters to elect a president (assuming a two canidate race and winner takes all for each state).