2024

Covers 2000–2024. Earlier elections are not yet loaded.

Select a state — on the map or in the table — to see its record.

The record ends here. What follows is this site’s argument.

One person, one vote — except for president

Everything above is the documentary record: official certificates, verifiable, taking no side. This section is the author’s position. It is kept separate so you can use the record without adopting the argument.

Three things the record shows

1

Your vote weighs more or less depending on where you live

Electoral votes are not handed out in proportion to people. Every state gets two regardless of size — the same bonus for Wyoming as for California. Divide each state’s population by its electoral votes and the gap is plain:

Residents per electoral vote, 2020 Census. Fewer residents per electoral vote means each resident carries more weight.
StateResidents per electoral voteCompared with the national average

National average: 616,077 residents per electoral vote. Every state and D.C. is in the full table — .

2

Twice this century, the candidate with fewer votes won

In 2000, Al Gore received 50,999,897 votes to George W. Bush’s 50,456,002 — and Bush became president. In 2016, Hillary Clinton received 65,853,514 to Donald Trump’s 62,984,828 — and Trump became president. Both results followed the rules exactly. That is the point: the rules permit it. Slide the timeline above to either year and the record shows it.

3

538 people cast the actual votes, and the rules binding them are not the same everywhere

The president is not chosen by the national count. It is chosen by 538 individuals, named above, selected by state parties rather than elected by name. In Chiafalo v. Washington (2020) the Supreme Court held unanimously that a state may enforce an elector’s pledge. It did not hold that a state must.

So the protection varies. Some states cancel a faithless vote and seat a replacement — Colorado did exactly that to Micheal Baca in 2016, and his vote never counted. Some impose a fine and let the vote stand, which is what Washington did to four of its electors the same year. Some have no penalty at all. Which of those applies to you depends on your state, and most voters have never had reason to find out.

No faithless elector has ever changed a presidential outcome. The seven in 2016 changed nothing. But seven is not far from the margin of a close election, and the rules governing them were written state by state, for a system nobody expected to be tested.

A fix that needs no amendment

Amending the Constitution requires two-thirds of Congress and three-quarters of the states. The National Popular Vote Interstate Compact takes a different route. Article II leaves it to each state to decide how it appoints its electors. Member states agree to award theirs to whoever wins the most votes nationwide — and the agreement only takes effect once members hold at least 270 electoral votes, enough to decide the outcome. Until then, nothing changes.

Electoral votes committed 222 of 270 needed

48 short. Enacted by 18 states and the District of Columbia.

The case against — stated fairly

If you are going to repeat the argument above, you should know what people say back. These are real objections, not strawmen:

Supporters answer each of these. The point of listing them is that you should be able to.

What you can actually do

  1. Find out where your state stands. Select your state on the map above. The panel names your electors and shows what a vote there is worth. The meter above shows whether your state has joined the compact.
  2. Write to your state legislator, not Congress. This is decided in state legislatures. Your state representative and state senator vote on it; your member of Congress does not.
  3. Ask one specific question. Most offices will have to look it up — that is the point.
    Does our state legally bind its presidential electors to the statewide vote result, and what happens if an elector breaks that pledge?

  4. Send the record, not the opinion. Share the link. Let people read the certificates and reach their own conclusion — it travels further than an argument does.

Back to the record

Methodology, sources and limits

Everything on this site can be checked against a primary document. This section explains where each figure came from, how it was verified, which judgment calls were made, and what is known to be incomplete. If you intend to cite this work, read this part first.

Where the data comes from

Elector names
Each jurisdiction’s Certificate of Ascertainment — the document a governor signs naming the electors their state appoints — filed with the National Archives. Every state panel links to the certificate it was read from. National Archives, 2024 Electoral College
How electors voted
Certificates of Vote, signed by the electors after they met on December 17, 2024. Linked from each state panel alongside the ascertainment.
State outcomes and electoral vote counts
Certified state results, cross-checked against the National Archives’ published totals for each election year.
Population and vote weight
2020 Census resident population — the same count that produced the current apportionment. U.S. Census Bureau, 2020 Apportionment, Table 2
Faithless electors
Contemporaneous reporting and court records, including Chiafalo v. Washington (2020) and Baca v. Colorado Department of State (2020).
Compact status
State enactments of the National Popular Vote Interstate Compact as reported through April 2026.

How each roster was verified

Every elector list was read directly off the certificate, then checked against an independent number: the state’s electoral vote count. Pennsylvania’s certificate must yield exactly 19 names, California’s exactly 54. A roster that does not match its state’s count is not published.

All 51 jurisdictions match. The 538 names on this site correspond one-to-one with the 538 electoral votes cast in 2024.

Where a certificate names its electors more than once — many do, once in a canvass table and again in the governor’s declaration — both lists were compared.

Judgment calls, and why

  • Appointed, not nominated. Where a state’s final certification differs from the slate originally put forward, the appointed list is used. Indiana nominated Tyler D. Warman; the governor’s certification names Martin E. Obst. Obst is listed.
  • Amended certificates supersede originals. California filed an amended certificate replacing five electors; the amended list is used, because those are the people who served.
  • Losing slates are excluded. Certificates list every party’s slate. Only appointed electors appear here. New Hampshire’s certificate names twelve people; four were electors.
  • Home addresses are omitted. Several certificates print electors’ street addresses. These are private citizens; republishing their addresses is not necessary to document who cast a vote. Cities and towns are omitted for the same reason. The certificates remain linked for anyone who needs the original.
  • Names appear as written. Including traditional names recorded on Washington’s certificate, and capitalization normalized only where a certificate prints entirely in capitals.
  • Nothing is estimated. No name was inferred, reconstructed, or filled in from a secondary source.

Known limits

  • Two rosters were transcribed by eye. Massachusetts and New Hampshire file their certificates in engraved script that machine text extraction cannot read. Those 15 names were read from the scanned image by a human and checked against the vote counts. They carry a different kind of risk than the rest — transcription error rather than extraction error.
  • Unusual surnames carry some extraction risk. Names were pulled from each PDF’s text layer. Characters such as rn/m and I/l can be misread. Every panel links to the certificate so any name can be checked at the source.
  • Elector names exist for 2024 only. The 2000–2020 elections show state outcomes, electoral vote counts, popular vote and faithless electors, but not full rosters. Those certificates are public and linked; they have not been read in yet.
  • National popular vote totals should be re-verified before publication. They are widely published figures, not read off a primary canvass document like the rest of this site.
  • State-by-state popular vote counts are not included.
  • Elections before 2000 are not included. Before 1824 most states had no popular vote at all, so the popular-versus-electoral comparison this site is built around does not exist for the early republic.

How to cite this

Replace the access date with the day you read it.

APA
MLA
Chicago

Citing a specific elector? Cite the certificate, not this site. Each state panel links to the governor-signed original at the National Archives, which is the primary source. This site is a finding aid; the certificate is the evidence.

Take the data

The full 2024 record is downloadable. No attribution required, though a link back helps people check your work.

CSV is one row per elector — 538 rows, with state, electoral votes, winning ticket, elector name and certificate URL. JSON keeps the same data grouped by jurisdiction.

Check the work. This whole site is a single public file. Every name, every source note and every link is readable in the source, and the full revision history shows what changed and when. View the repository on GitHub. Found an error? It is reproducible, so it is fixable.

Accessibility

This site was built to be usable without a mouse, without sight, and without fine motor control. That was a design requirement from the first version, not an afterthought.

  • Every state on the map is reachable by keyboard and announces its name, electoral votes and result to a screen reader.
  • The table view is a complete equivalent of the map, not a summary — anything the map shows, the table shows.
  • A skip link jumps straight to that table.
  • Result is never conveyed by color alone; split states are patterned and labeled.
  • Interactive controls meet a 44-pixel minimum touch target.
  • Map and meter animation is suppressed when the operating system requests reduced motion.
  • The page needs no external files to render, so it works on slow or filtered connections.

If something here is hard to use with your assistive technology, that is a defect worth reporting.

Corrections

If a name here does not match the certificate it is drawn from, that is an error and it will be fixed. Point to the state, the name, and the certificate.

Contact: [ add your preferred contact here before publishing ]

Last reviewed: . Record covers the 2000 through 2024 presidential elections; elector names complete for 2024.